Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research

Jelle Brouwer; Floor van den Berg; Remco Knooihuizen; Hanneke Loerts; Merel Keijzer

doi:10.3390/bs10090132

Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research

Brouwer, Jelle; van den Berg, Floor;Knooihuizen, Remco;Loerts, Hanneke;Keijzer, Merel 2020-08-31 00:00:00 Review Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research 1, 1 1 2 Jelle Brouwer *, Floor van den Berg , Remco Knooihuizen , Hanneke Loerts and Merel Keijzer Department of Linguistics & English as a Second Language, University of Groningen, 9712 EK Groningen, The Netherlands; f.a.van.den.berg@rug.nl (F.v.d.B.); r.m.knooihuizen@rug.nl (R.K.); m.c.j.keijzer@rug.nl (M.K.) Department of Minorities and Multilingualism, University of Groningen, 9712 EK Groningen, The Netherlands; h.loerts@rug.nl * Correspondence: jelle.brouwer@rug.nl Received: 31 May 2020; Accepted: 26 August 2020; Published: 31 August 2020 Abstract: Late-life depression (LLD) affects about an eighth of community-dwelling seniors. LLD impacts well-being, with loneliness and small social networks being typical. It has also been linked to cognitive dysfunction and an increased risk of developing dementia. Safety and efficacy of pharmacological treatments for LLD have been debated, and cognitive dysfunction often persists even after remission. Various cognitive interventions have been proposed for LLD. Among these, one has received special attention: foreign language learning could serve as a social intervention that simultaneously targets brain structures affected in LLD. Lifelong bilingualism may significantly delay the onset of cognitive impairment symptoms by boosting cognitive reserve. Even late-life foreign language learning without lifelong bilingualism can train cognitive flexibility. It is then counterintuitive that the effects of language learning on LLD have never been examined. In order to create a theoretical basis for further interdisciplinary research, this paper presents a status quo of current work through two meta-analyses investigating cognitive functioning in LLD on the one hand and in senior bilinguals or seniors following a language course on the other hand. While LLD was consistently associated with cognitive dysfunction, inconsistent results were found for bilingualism and language learners. Possible reasons for this and suggestions for future research are subsequently discussed. Keywords: late-life depression; language learning; bilingualism; seniors; cognitive function; meta- analysis; systematic review 1. Introduction The world’s population is aging, both in relative and absolute numbers. Since the early 1900s, the mean life expectancy has increased by more than 30 years in some countries [1]. This increased longevity combined with declining birth-rates and a so-called ‘baby boom’ following World War II means that the proportion of those aged 65 and older will increase significantly in the coming years. This relative aging of the population is accompanied by many societal challenges. After the age of 50, for instance, the cost of healthcare grows exponentially every year [2]. Apart from the increased financial burden, there are also challenges on a personal level: a number of age-specific illnesses reduce cognitive functioning and quality of life. Most pertinently, neurodegenerative illnesses such as Alzheimer’s Disease (AD) are predicted to take over as a primary cause of death in developed Behav. Sci. 2020, 10, 132; doi:10.3390/bs10090132 www.mdpi.com/journal/behavsci Behav. Sci. 2020, 10, 132 2 of 36 countries [3]. The number of AD patients in the United States, for instance, is estimated to grow from 4.7 million in 2010 to 13.8 million in 2050, affecting half the population over 65 [4]. However, while research pertaining to AD is undeniably important, it overshadows a lesser-known threat to healthy aging: late-life depression (LLD). Falling in the realm of major depression disorder, LLD can be diagnosed on the basis of symptoms stipulated in the Diagnostic and Statistical Manual (DSM) [5]. LLD includes both early-onset depressions that persist or recurs after the age of 60 as well as the first diagnosis of depression after age 60 [6]. A diagnosis of depression can follow if an individual currently displays at least one of the two core symptoms: (1) depressed mood or (2) loss of interest/pleasure; and at least four of the following symptoms during at least a two-week period of time: (3) unintended large changes in weight (+/− 5% in a month), (4) insomnia or hypersomnia, (5) psychomotor agitation or retardation, (6) fatigue or loss of energy, (7) feelings of worthlessness or inappropriate guilt, (8) diminished ability to think or concentrate, or indecisiveness, and (9) recurrent thoughts of death or suicidal ideation [5]. Despite limited public awareness, LLD is extremely common: in high-income countries, major depressive disorder was found in 0.9–9.4% of community-dwelling seniors and in 14–42% of those living in care homes; minor depression (the presence of one core symptom and one to three of symptoms 3–9 mentioned above) was found in up to 49% of older adults [7]. Comparable numbers have been attested in low to middle-income countries [8]. Despite a high prevalence of LLD in large-scale epidemiological studies, awareness appears poor, and the illness is often difficult to diagnose. Part of the reason why LLD is so underdiagnosed can be traced back to its symptomatology. Many of the symptoms are somatic in nature, commonly leading older patients and/or their caretakers to mistake these signs for physical illness [9]. A number of studies also suggest that in some LLD patients, especially men, low mood is not necessarily present, further complicating diagnosis [6,10]. Patients suffering from LLD may also feel less inclined to see their general practitioner (GP) for experienced mental health issues, due to stigmas attached to psychiatric disorders [9,10]. Finally, the underrepresentation of LLD may also be associated with healthcare providers. A meta-analysis by Mitchell et al. suggested that GPs often do not take into account the high prevalence of depression in seniors [11]. The study found, for instance, that conversations regarding mental health were quite rare, and often not initiated by the GPs themselves. The impact of LLD on mental health and cognition is all-encompassing. Not only are mood disorders present in up to 87% of older adults who die by suicide [12], but nonsuicide mortality is also higher in these populations [6]. This can be partially explained through higher incidences of unhealthy behaviors such as smoking and excessive alcohol use among this group [13,14], but also potentially through the lack of social contacts [15]. Additionally, there is a strong and consistent finding that LLD patients suffer from cognitive dysfunction [16,17]. Related to this, LLD significantly increases a person’s risk of developing AD or other types of dementia [18,19]. This consistent association between LLD and cognitive dysfunction has sparked theories that depression is a prodromal form of dementia or that LLD and dementia are underpinned by a similar pathology [20]. An influential review paper by Butters et al. identified a number of possible associations between LLD and AD [21]. One theory they postulated was that depression-related disturbances in glucocorticoids (hormones associated with stress-response, among other processes) lead to hippocampal atrophy, which may increase the risk of developing AD. However, Butters and colleagues proposed that, out of a number of potential pathways, the most common scenario would be that “accumulate[d] AD neuropathology over many years along with co-occurring cerebrovascular disease [i.e., related to blood flow in the brain] damages the frontostriatal circuitry [i.e., the neural pathways connecting the frontal lobe to the striatum], leading to late-life depression” [21]. Another overview paper reached similar conclusions but emphasized that the cause-effect relationship needed to be investigated more thoroughly for a solid understanding [22]. While there is thus some support for LLD as a prodromal manifestation of AD, there are many other (preventable) biopsychosocial causes, associations, and risk factors for LLD that are “complementary and almost always transactional” (i.e., the risk factors interact with each other) [23]. Indeed, external psychological and social variables seem equally important in modulating LLD: Behav. Sci. 2020, 10, 132 3 of 36 seniors with smaller social networks, for example, seem to be more at risk of developing or having LLD [24,25]. Similarly, those who do not have a spouse or confidant develop LLD more often than those who do [15,26–28]. Another important factor is the presence of life experiences that build up cognitive reserve. Cognitive reserve can succinctly be described as an individual’s resilience to age- related brain pathology [29]. This reserve can, for example, be built up through higher educational attainment, having (had) a cognitively demanding job, and maintaining cognitively stimulating hobbies later in life [30]. To illustrate, many epidemiological studies investigating LLD reported a higher prevalence of depression in individuals with low educational attainment [24,31,32] and those with less complex occupations during their working life. Likewise, seniors who did not read for leisure were more often depressed, while playing cognitively and socially stimulating card games seemed to protect against LLD [31]. Considering the effects of LLD on personal well-being, and the negative prognosis, the need for effective treatment seems both urgent and evident. Current treatment methods can be problematic, however. Antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), are generally deemed safe in younger populations, but seniors have been found to be much more sensitive to negative side-effects [33]. This can largely be attributed to drug-drug interactions between some SSRIs and other medications often prescribed to seniors [34]. Moreover, while successful implementation of pharmacotherapy has been reported in individual papers [35,36], a 2011 meta- analysis found variable evidence for the efficacy of antidepressants [37]. Conversely, a wide range of psychological treatment modalities, such as cognitive behavioral therapy, life review therapy, and problem-solving therapy have been coined for their potential to alleviate late-life depressive symptoms in several meta-analyses [38,39]. However, while these methods are able to improve patients’ mood and well-being, cognitive dysfunction often persists even after remission [40–42]. Koenig et al. suggested that this persistent cognitive impairment could indicate a prodromal form of dementia, but acknowledged that patients may have had cognitive impairments before the onset of LLD [41]. Taking into account the evidence synthesized above, maintaining cognitive function or preventing cognitive decline during or after a depressive episode in older age seems difficult, if not impossible. Indeed, current treatment methods do not seem to restore cognitive functioning. In addition, since a pervasive public sentiment exists that aging is synonymous with cognitive decline [43,44], it may follow that not enough neural plasticity remains to improve again. However, lifelong learning initiatives such as the university of the third age [45], underscored by a growing body of research [46,47] have shown that improving cognition in healthy seniors through several training modalities is most definitely possible. In contrast, a number of researchers noted that current LLD treatment methods do little to nothing to combat cognitive dysfunction. In response to this, these researchers approached the treatment of LLD specifically through training cognitive functions. One study [48] did so by employing serious gaming software, targeting cognitive control through multitasking paradigms. Employing a cognitive training modality in depressed seniors was found to result in a significant improvement in mood. This change in mood in the training group was statistically indistinguishable from that in a control group following a recognized treatment modality (namely, problem-solving therapy). Interestingly, however, performance on previously untrained cognitive tests pertaining to working memory and sustained attention improved markedly in the cognitive training group, while no such improvement was found in the conventional therapy group. Building on these outcomes, Morimoto et al. developed a computerized cognitive training intervention that used tasks known to specifically train domains associated with treatment-resistant LLD, such as inhibitory control and cognitive flexibility [49]. This intervention was adaptive to the patients’ performance level, such that it was never impossible, but always challenging. In their pilot study, treatment-resistant participants enrolled in a cognitive intervention were compared to depressed controls (not treatment-resistant) receiving a conventional 12-week pharmacological treatment (escitalopram, 20 mg). The researchers reported a statistically significant reduction in depressive symptoms for both groups. Strikingly, however, in the cognitive intervention, this result was obtained after merely four weeks, instead of 12. Additionally, performance on the Stroop task Behav. Sci. 2020, 10, 132 4 of 36 and Trail Making Task Part B (a task in which numbers and letters distributed on a piece of paper need to be connected in ascending order: 1, A, 2, B, 3, C, etc.) increased in the cognitive intervention group, while staying constant in the pharmacological intervention group. While larger-scale studies are needed to consolidate these results, it seems that cognitive training may benefit LLD patients, both in terms of mood and cognition. While results are promising, criticism regarding these interventions must be acknowledged. One general issue with computerized cognitive training, both for treatment of depression and general improvement in cognition, is that it may “teach to the test” [50,51]: the outcome measures used to assess changes in cognition are too similar to the tasks used in the training. As a result, transfer effects are less meaningful. Another issue is the operationalization of these training programs. Computerized cognitive training often consists of somewhat repetitive tasks, done alone by participants in their own homes. A large-scale study in the Netherlands investigating several exclusively home-based cognitive interventions (e.g., language training, brain-training games) in healthy seniors [52] reported high participant attrition, with over 40% of participants dropping out before the two-month interventions were completed. In response, the researchers halved the duration of the interventions, but this, too, failed to reduce drop-out rates. Olfers et al., therefore, argued that adding a social component (such as chatrooms for participants) could increase adherence. Depressed seniors, in particular, could potentially benefit from a stronger emphasis on interaction and social networks, considering that loneliness is strongly associated with LLD [24,25]. One presently unexplored intervention for LLD that emphasizes this social component, while still training cognitive functions, is language learning. It has long been theorized that people who already know two or more languages (i.e., bilinguals) enjoy a certain cognitive advantage in comparison to monolinguals, as has been found in studies examining, for instance, word learning [53] and conflict processing [54]. Bilinguals essentially juggle two languages at all times: while one language is used, the other needs to be suppressed, with control transferring from the language domain to the general cognitive domain. It does need to be pointed out that the precise mechanisms underlying this advantage have been questioned recently [55] Additionally, depending on the individual bilingual experience, switches between the spoken languages may be frequent. It is theorized that these patterns of constant (de)activation and switching enhance cognitive control (that is, a bilingual experience causes transfer effects to general domains of cognition) [56]. It is also important to acknowledge that a fierce debate on the bilingual advantage has gone on since the inception of the concept. A large meta-analysis from 2018 compiled evidence from 152 published and unpublished studies [57]. While evidence in favor of the bilingual advantage seemed to be present, initially, these effects disappeared after correcting for publication bias. The authors concluded, therefore, that no systematic evidence existed for a bilingual advantage. They also emphasized the need for pre-registration of studies to ensure that negative or null-findings are published, too. Similarly, a large-scale study containing data from 11,000 participants found no conclusive evidence of bilinguals outperforming monolinguals [58]. It is important to point out, however, that individual differences in bilingual experiences were not taken into account in these larger reviews. Indeed, another critical review noted that research on bilingualism often contains methodological flaws that influence results [59], concluding that it is pertinent “to focus more research attention on the individual features of bilingual experience, as this is where the putative cognitive effects stem from”. As was mentioned before, bilinguals’ languages are always active, which is theorized to constantly train cognitive control. Indeed, a number of studies have found that a lifetime of bilingualism delays the average onset of Alzheimer’s disease symptoms by multiple years [60–63]. Additionally, a study comparing bilinguals and monolinguals with AD, who were comparable in terms of cognitive functioning, found that the degree of neuropathology in the bilinguals was more extensive [64]. This suggests that compensatory mechanisms pertaining to cognitive reserve in lifelong bilinguals are much stronger. This does not mean, however, that a positive result can only be found if bilingualism is lifelong. The constant interference that follows from juggling two or more languages makes language learning unique from other types of cognitive training, as any type of activity requiring language will automatically lead to these activation/deactivation patterns, Behav. Sci. 2020, 10, 132 5 of 36 interfering with earlier acquired languages in ways that learning other new skills will not. For instance, one study investigating children looked into the cognate facilitation effect, a well-known phenomenon where co-activation of two orthographically and phonologically similar words (e.g., “dokter” in Dutch and “doctor” in English) leads to faster response times [65]. Interestingly, these researchers not only found this facilitation effect in children who were lifelong bilinguals but also in children who were still learning their second language. Other studies have also suggested that such effects could also arise early in the language learning process of seniors. Indeed, in a 2013 paper, Antoniou et al. argued that learning a foreign language would likely activate those neural networks that decline with aging [66]. This includes networks affected in LLD. Therefore, they proposed that late-life language learning could have neuroprotective effects, effectively training the cognitive functions affected in aging (and LLD). Language learning in seniors is a rapidly emerging field. The focus thus far, though, has mainly been on boosting seniors’ cognitive reserve in order to stave off dementia [66]. However, as we have seen, AD and LLD potentially have similar underpinnings [21]. Like Alzheimer’s, LLD is characterized by pathological declines in cognitive function [16]. Similarly, a risk factor for both LLD and AD seems to be reduced cognitive reserve [24,28,31,32] and LLD has even been considered a prodromal expression of AD pathology [20,21]. Successfully ameliorating or reversing cognitive dysfunction in LLD, then, could potentially delay or prevent further cognitive decline. Additionally, learning and using a foreign language is an inherently social activity, especially when done in a classroom setting. A further advantage of classroom language learning is that it potentially tackles the common problems of loneliness and small social networks LLD patients experience [24,25]. Taking away these two risk factors for LLD, then, could potentially improve mood as well (although this is not the focus of the present paper). Published research directly investigating the efficacy of language learning to improve cognition and mood in LLD, is currently non-existent, although work on language learning in seniors is being carried out at the moment of writing, e.g., [67]. 2. Aims and Objectives As the field of late-life language learning as a potential therapy for LLD is in its infancy, it is essential to work towards a stronger theoretical and methodological basis for such interventions. This paper contributes to this aim by conducting two separate meta-analyses. First, an in-depth overview of performance on (standardized) neuropsychological tests (including measures of cognitive flexibility, processing speed, working memory, and inhibition) in seniors with LLD will reveal how cognitive domains are (differentially) affected. Subsequently, an overview of studies investigating the performance of bilingual seniors on these same tests will be compiled (both life-long bilinguals and late-life language learners). We acknowledge that it would be more ideal to only include studies on late-life language learners, but the field of third-age language acquisition is still emerging. It was therefore decided to include studies on lifelong bilinguals, too to substantiate claims relating to future research avenues. We expect the overall cognitive functioning of depressed seniors to be less efficient than in healthy controls, but based on evidence for the neuroprotective effect of bilingualism, we hypothesize that bilingual seniors will outperform their monolingual peers on these tasks. It must be stated that the results of these two meta-analyses will not be directly comparable. Indeed, if healthy older bilinguals outperform healthy older monolinguals, it does not necessarily point to the potential efficacy of language learning as an intervention in LLD. Rather, the two analyses should serve to bridge two currently separate fields (i.e., linguistics and geriatric psychiatry) by giving an overview of the status quo in both fields. Evidence from the present meta-analyses, then, will provide the first steps for a theoretical basis for interdisciplinary work targeting language learning as a potential therapy for LLD. Additionally, as a secondary objective, critical attention will be paid to methodological choices made in the literature on both late-life depression and late-life language learning, providing a framework for researchers from various disciplines interested in these topics. Based on the results of this review, concrete suggestions for future research will be presented. Behav. Sci. 2020, 10, 132 6 of 36 3. Methods 3.1. Inclusion and Exclusion Criteria A number of inclusion and exclusion criteria were established to meet the study’s aims. To be included in the literature review on cognitive functioning in LLD, studies had to compare a group of seniors with depressive symptoms or diagnosed depression to healthy controls (e.g., cross-sectional) on the basis of one or more neuropsychological tests from a pre-specified set. This set included a number of widely-used tests tapping into working memory (forward and backward digit span tasks [68]), processing speed (Trail Making Test A [69]), cognitive flexibility (Trail Making Test B [69], Wisconsin Card Sorting Test [70], color-shape switch task [71], letter-number sequencing task [68], digit-symbol substitution task [68]), inhibition (antisaccade task [72]), and episodic memory (Visual Association Task—Extended [73]). Most of the neuropsychological tasks included in the present analyses were selected because they are commonly used and clinically relevant (e.g., the TMT is often used to indicate potential dysfunction in clinical settings [74]). Two tasks were chosen because they were of particular interest to ongoing work (i.e., the color-shape switch task and the VAT-E). All neuropsychological tasks in the test battery. The neuropsychological tasks were also selected because they would be used in ongoing work of the authors. Longitudinal studies on LLD (e.g., randomized controlled trials) were included, but only baseline data were inspected. For the searches looking into bilingualism, all studies comparing older bilinguals to older monolinguals using this same test battery were included, as were intervention studies looking into language learning in seniors. Other inclusion criteria for eligibility were: 1. Participants were ≥55 years old (Studies on LLD used various cut-off points for a minimum age of inclusion ranging from 45 [75] to 65 [76]. 55 was chosen as a middle ground that primarily identified studies where participants had a mean age of 65 or higher and resulted in more included studies to meet the current investigation’s aims) 2. The study was published in English. Exclusion criteria were: 3. Participants having a comorbid psychiatric Axis-I disorder (except anxiety, which is highly comorbid with LLD [77]); 4. Participants having a comorbid neurological disorder (e.g, Parkinson’s disease or Alzheimer’s disease) [78,79]; 5. Participants reporting a dependency on alcohol, medication, or other substances [80]; 6. Participants using benzodiazepines [81] or beta-blockers [82]; 7. Qualitative studies; 8. Case studies; 9. Meta-analyses or (systematic) reviews. In preliminary searches, only studies that fully adhered to the criteria were considered. However, it soon became clear that a large number of potentially relevant studies, especially those in the realm of bilingualism, had not collected (complete) data on medication use and the presence of comorbidities. For this reason, studies that did not collect these data were also included. The cut-off of 55 years, however, was treated as a hard criterion. 3.2. Search Strategies All searches were done in PubMed and EBSCOhost search engines. Preliminary search strategies aimed to look into broader domains of cognition (e.g., (“late-life depress*” OR “geriatric depress*”) AND (“inhibitory control” OR …). However, due to a large number of results (>8000), it was determined that the searches would be limited to tests used in the test battery of the author’s lab group. The complete search strategy can be seen in Table 1. No limits were set on publication dates. Behav. Sci. 2020, 10, 132 7 of 36 Table 1. Complete search strategy for the present study. Date Database Search String (“late-life depress*” OR “geriatric depress*”) AND (“Trail making t*” OR “TMT” OR “color-shape switch t*” OR “colour-shape switch t*” OR “phonemic verbal fluency” OR “Visual association t*” OR “Vat-E” OR 23-01- PubMed, “wisconsin card sorting t*” OR “WCST” OR “modified wisconsin card 2020 EBSCOhost sorting t*” OR “mwcst” OR “digit-span” OR “digit span” OR “letter number sequencing” OR “letter-number sequencing” OR “symbol coding t*” OR “symbol-coding t*” OR “digit substitution t*” OR “DSST” OR “antisaccade” OR “anti-saccade”) (“senior*” OR “older adul*” OR “third-age” OR “third age” OR “65*” OR “aged” OR “elder*” OR “pension*”) AND (“lifelong bilingualism” OR “bilingualism” OR “language course” OR “language learning” OR “language training” OR “language acquisition” OR “multilingualism” OR “foreign language” OR “language teaching”) AND (“Trail making 17-03- PubMed, t*” OR “TMT” OR “color-shape switch t*” OR “colour-shape switch t*” 2020 EBSCOhost OR “phonemic verbal fluency” OR “Visual association t*” OR “Vat-E” OR “wisconsin card sorting t*” OR “WCST” OR “modified wisconsin card sorting t*” OR “mwcst” OR “digit-span” OR “digit span” OR “letter number sequencing” OR “letter-number sequencing” OR “symbol coding t*” OR “symbol-coding t*” OR “digit substitution t*” OR “DSST” OR “antisaccade” OR “anti-saccade”) 3.3. Study Selection After exporting the search results from PubMed and EBSCOhost, title duplicates were removed manually by one reviewer (JB). Then a list of the remaining titles was compiled, such that the reviewer was blind to the authors and year of publication. Subsequently, titles were carefully screened by the same reviewer, such that only those papers which clearly violated one or more exclusion criteria were removed (e.g., titles indicating that the main focus was bipolar depression). As a next step, the same procedure was repeated for the remaining abstracts. A separate list containing only the title, abstract, and notes from the previous screening round was checked for clear violations of exclusion criteria. Full-text papers were then retrieved from the remaining list of articles. This time, the reviewer looked at all inclusion/exclusion criteria, to determine if the paper could be included in the final review selection. If information pertaining to one or more criteria was missing, the corresponding author for the study was contacted. In the absence of a reply, a first reminder was sent after approximately three weeks. A second, and final, reminder was sent two weeks after the first reminder. The reference lists of included studies were also inspected for potentially relevant work. 3.4. Data Collection Process & Data Items The data necessary to calculate effect size (e.g., mean, standard deviation, and number of participants) were extracted where available. In cases where this was not possible because the necessary data were not reported, the corresponding authors were contacted using the same procedure mentioned in Section 3.3. When these data could not be retrieved, the study was excluded from further analysis. Additional data were extracted to compare the studies’ methods: country where the study was conducted, age of participants, gender distribution, years of education, scores on mini-mental state exams (MMSEs) [83] or the Montreal Cognitive Assessment (MoCA) [84], and the operationalization of depression or bilingualism. Behav. Sci. 2020, 10, 132 8 of 36 3.5. Risk of Bias in Individual Studies Risk of bias in individual studies was assessed independently by two authors (JB, FvdB) using the National Heart, Lung and Blood Institute’s quality assessment tool [85]. This tool consists of 14 questions regarding for instance selection bias of participants, validity of the tools used, and blinding of experimenters. This tool was deemed the most suitable, as it enabled the assessment of cross- sectional studies (others were more aimed at RCTs). A highly similar tool also from the NIH was used for the intervention studies [86]. The scoring guide for these tools explicitly states that overall quality should not be determined by summing up the scores per question. Rather, assessors are instructed to consider how much risk of bias each rating of the questions introduces in the individual studies and to determine the study quality based on the presence of any risks (e.g., selection bias, measurement bias, or confounding). Consensus on the ratings was reached after a discussion between the two reviewers. 3.6. Summary Measures & Synthesis of Results A pervasive issue with systematic reviews is that they employ so-called ‘vote-counting’ methods [87]: if a review, for instance, finds that five out of ten retrieved studies are in line with their hypothesis, they may conclude that evidence is variable. This method is not ideal because it is done purely on the basis of significance; this becomes especially problematic when effect sizes or sample sizes are small, as described in detail by Combs et al. [88]. It is, therefore, recommended that one calculates a pooled average by conducting a meta-analysis [87]. This method reduces the chance of type II errors because it does not simply count p values of generally low-powered studies [89]. Additionally, a meta-analysis assigns more weight to studies with more precision (generally those with larger sample sizes) [87]; this also increases the precision compared to vote-counting. Hedges’ g was calculated for each of the neuropsychological tests included in this review, in R (version 4.0.0) [90] with the ‘esc’ package (version 0.5.1.) [91]. A pooled effect size was subsequently calculated for each neuropsychological test with two or more available effect sizes using the ‘meta’ package (version 4.12-0) in R [92]. This same software package was used to create forest plots displaying the overall effect sizes, and funnel plots to show potential publication bias. Since we expected some methodological variation between studies, we opted for a random-effects model [87]. Neuropsychological tests with only one available effect size were reported in a separate table. 4. Results 4.1. Study Selection The search for papers on cognitive functioning in LLD led to 23 studies that met the inclusion criteria. A total of 14 studies regarding cognition in older bilinguals or older language learners were identified (two of which investigated language learning in seniors). A PRISMA flowchart [93] displaying this process can be seen in Figure 1. Behav. Sci. 2020, 10, 132 9 of 36 Figure 1. PRISMA flow diagram of the screening process for late-life depression (left) and bilingualism (right). 4.2. Summary of Study Characteristics Late-Life Depression An overview of all study characteristics can be found in Table 2, below (for a more detailed version please refer to Table S2 in the supplement). The 23 studies that investigated cognitive function in LLD were published between 1994 and 2018. Two-thirds of these studies (n = 14) were published after 2010. The majority of studies were conducted in North America (n = 13). Of the other 10 studies, five were conducted in Europe, three in Asia, and two in South America. Six studies contained complete information on all inclusion and exclusion criteria. The 17 studies with missing information did not collect or report information regarding the following: comorbid neurological illness (n = 3), comorbid psychiatric illnesses (n = 11), substance or medication abuse (n = 12), or the presence of medication that led to exclusion in the current design (n = 16). 4.2.1. Demographic Information The 23 retrieved studies contained a total of 4366 depressed participants and 8875 healthy controls. LLD studies had an average of 154.9 depressed participants (SD = 359.1) and 385.9 healthy controls (SD = 928.9). Groups were not balanced for gender: both LLD and healthy control groups on average comprised more female participants (Depressed: M = 67.5%, SD = 13.0; Healthy controls: M = 64.8%, SD = 11.9). Age was explicitly reported in all but one study (which expressed age distribution as the percentage of participants within an age band). Depressed participants were on average 73.9 years old (SD = 6.6), while healthy controls had a mean age of 72.3 (SD = 5.8). In the 18 studies that reported the total years of education, depressed participants had—on average—one year of education less than the healthy controls (Depressed: M = 13.1, SD = 4.1; Healthy controls: M = 14.1, SD = 3.7). Lastly, judging from the 17 studies that included MMSE scores, there was no indication of the presence of global cognitive impairment, and it seemed that depressed and healthy groups were well- matched (Depressed: M = 26.9, SD = 2.9; Healthy controls: M = 26.8, SD = 2.5). 4.2.2. Types of Tasks Used The following outcome measures were implemented in the included studies: time to complete TMT B (n = 13), percentage of perseveration errors on TMT B (n = 1), time to complete TMT A (n = 13) (One study [94] reported TMT A and B scores in percentiles; these are reported separately), phonemic verbal fluency (n = 9), digit span forward (n = 9), digit span backward (n = 5), (modified) Wisconsin Behav. Sci. 2020, 10, 132 10 of 36 Card Sorting Test categories (n = 6) (The modified WCST is shorter and less difficult [95]), (m)WCST perseveration errors (n = 4), digit symbol substitution task (n = 8), and score on the letter-number sequencing task (n = 2). The n indicates the number of times a healthy group could be compared to a depressed group. Studies containing multiple comparable groups (e.g., age 65–75 healthy vs depressed and aged >75 healthy vs depressed) were included (and therefore counted) twice. 4.2.3. Operationalization of Late-Life Depression Nearly two-thirds of the studies (n = 14) did not report whether participants in the depressed group had early-onset (before age 60) or late-onset (after age 60) depression. A further five studies reported either only late-onset LLD patients or they provided separate analyses for early and late- onset depression. The remaining four studies reported that they included both early and late-onset patients in their analyses. Blazer states that, while there is some debate on whether there is a difference in etiology between early and late-onset depression, the neuropsychological performance between these two groups does not differ [6]. Therefore, the authors did not differentiate between early and late-onset LLD in their analyses. There was substantial heterogeneity in terms of tools and interviews used to classify depression and depression severity. DSM criteria were used in 14 studies (DSM-III-R, DSM-IV(TR), and DSM-V). Three studies used the International Statistical Classification of Diseases and Related Health Problems (ICD-10), Chinese Classification of Mental Disorders (CCMD-3), and Research diagnostic criteria (RDC), respectively. Most of these studies used additional valid and reliable shorter scales containing between 10 and 30 questions: Hamilton Depression scale (ham-D), the Montgomery Asberg Depression Rating Scale (MADRS), the 30-item or 15-item Geriatric Depression Scale (GDS), the Center for Epidemiological Studies Depression Scale (CES-D), Beck’s Depression Inventory (BDI-II), or the Detection of Depression in the Elderly Scale (DDES). Six of the 23 studies only used one of these shorter scales. Behav. Sci. 2020, 10, 132 11 of 36 Table 2. Study characteristics of individual studies on cognitive function in late-life depression. Demographic information includes where the study took place, participants’ age and gender, educational background, and performance on the mini-mental state examination. Additionally, information is provided on how each study operationalized depression (continued on the next pages). First author N Age Education MMSE Country Group Operationalization of LLD Notes (year) (fem.) (SD) (SD) (SD) 64.9 28.9 DSM-IV criteria + at least LLD 11 (10) 14.4 (3.1) Alalade (4.5) (1.7) some depressive symptoms On average early onset 33.5 (2011) USA 71.2 29.2 as measured with MADRS (18.5) [96] HC 18 (11) 15.8 (2.3) (≥ 8) (6.6) (1.2) 69.0 29.1 Alexopoulos LLD 16 (n/a) 17.2 (2.3) (5.5) (0.9) DSM-IV criteria + MADRS ≥ (2013) USA 68.6 28.5 18 [97] HC 10 (n/a) 16.3 (3.8) (7.0) (1.0) 69.5 27.4 Only the high education group LLD 30 (22) 11.9 (3.3) DSM-IV criteria + (HAM-D + Avila (2009) (7.5) (2.4) was used here since low Brazil MADRS. No minimum for [98] 68.0 28.4 education group had only inclusion) HC 33 (26) 11.9 (3.3) (7.1) (1.7) approx. 3 years of schooling 69.4 27.4 LLD 22 (9) 7.1 (0.9) Balardin (1.1) (0.4) ≥5 on the 15 –item Geriatric (2009) Brazil 67.8 27.7 Depression Scale (GDS) [99] HC 22 (16) 9.6 (0.6) (1.4) (0.3) 71.8 28.1 Baudic LLD 21 (n/a) 9.6 (3.3) (8.8) (1.4) DSM-IV criteria + MADRS > (2004) France 73.6 29.7 20 [100] HC 19 (n/a) 13.5 (5.6) (7.7) (0.6) 74.7 25.5 Beheydt LLD 28 (24) n/a Participants matched on y/of Switzer- (7.6) (3.8) DSM-IV-TR criteria + 30- (2015) education; MMSE scores in land 72.0 28.3 item GDS score ≥11 [101] LLD group significantly lower HC 20 (15) n/a (5.1) (1.4) 64.4 DSM-III-R criteria measured Young-old and old-old USA LLD 23 (n/a) 15.6 (2.6) n/a (2.8) with SCID interview + presented in the same study Behav. Sci. 2020, 10, 132 12 of 36 Boone (1994) (Ham-D. No minimum for 64.4 young-old HC 54 (n/a) 14.5 (2.4) n/a inclusion) (2.7)) [102] 75.5 DSM-III-R criteria measured LLD 14 (n/a) 15.3 (2.8) n/a Boone (1994) (5.1) with SCID interview + Gender distribution overall is old-old USA 74.7 (Ham-D. No minimum for balanced (53% female) [102] HC 51 (n/a) 14.3 (2.8) n/a inclusion) (4.1) 67.5 29.9 LLD 64 (47) 8.9 (4.2) Chinese Classification of (6.0) (0.2) Chen (2013) China Mental Disorder (CCMD-3) [103] 68.2 29.9 HC 31 (18) 9.0 (4.0) + 30-item GDS score ≥ 11 (8.6) (0.2) 68.7 27.8 Dickinson LLD 112 (63) 14.2 (2.5) Not mentioned (Ham-D + (6.3) (2.4) (2011) USA MADRS used to measure 70.5 28.9 [104] depression severity) HC 101 (73) 15.4 (1.8) (5.7) (1.3) 71.4 27.2 LLD 14 (10) 9.4 (2.2) Egger (2008) (7.5) (1.0) DSM-IV criteria + 30-item Austria [105] 72.3 28.6 GDS score ≥ 15 HC 20 (13) 10.8 (2.7) (7.7) (0.8) Elderkin- 70.0 29.0 LLD 95 (63) 14.9 (2.7) DSM-IV criteria + 17-item (7.9) (1.3) Thompson Combined minor and major USA Ham-D ≥ 8 (minor LLD) + (2007) depression 71.5 29.4 HC 71 (42) 14.9 (2.7) 17-item Ham-D ≥ 15 [106] (7.6) (1.1) 69.0 29.0 Elderkin- LLD 112 (77) 15.6 (2.7) DSM-IV criteria measured Thompson (7.9) (1.1) USA with SCID interview + 17- (2011) 71.0 29.3 item Ham-D ≥ 15 HC 138 (86) 15.5 (2.7) [107] (7.7) (1.0) ≅65% Modif. attended or MMSE LLD 204 (204) ≅70 Goveas completed 96.0 ≥5 on the 15 –item Geriatric (2014) USA Women’s health cohort study college (3.5) Depression Scale (GDS) [108] 2017 ≅ 75% Modif. HC ≅70 (2017) attended or MMSE Behav. Sci. 2020, 10, 132 13 of 36 completed 97.0 college (2.7) 73.4 28.4 Kindermann LLD 25 (16) 13.3 (2.5) (6.0) (1.8) Research Diagnostic Criteria (2000) USA 71.2 29.2 + 21-item Ham-D ≥ 17 [109] HC 20 (12) 13.8 (2.5) (6.2) (1.0) 67.9 28.1 LLD 15 (9) 13.9 (2.2) ≥4 on 15-item GDS + ≥ 8 on Leal (2017) (6.9) (0.5) USA 21-item Beck’s Depression [110] 72.2 28.9 HC 27 (18) 16.7 (2.2) Inventory II (7.6) (0.2) 69.4 27.6 Miebach LLD 21 (14) 12.6 (2.8) ICD-10 criteria + (15-item (8.0) (2.1) (2018) Germany GDS. No minimum for 67.5 29.0 [111] HC 21 (12) 15.1 (3.1) inclusion) (7.2) (1.2) 71.2 23.6 LLD group includes major LLD 63 (46) 7.3 (5.5) DSM-IV criteria + (15-item Park (2018) South- (5.1) (4.4) depression, minor depression, Korean GDS. No minimum [112] Korea 70.3 26.3 dysthymia, and subsyndromal for inclusion) HC 59 (29) 10.2 (5.8) (4.7) (3.5) depression Rajtar- 68.1 13.6 (3.3) LLD 87 (57) n/a DSM-V criteria + ≥ 7 on 15- Zembaty (6.0) Poland item GDS + no depressive (2017) 66.8 HC 100 (61) 14.6 (3.0) n/a episode before age 60 [17] (4.8) 83.3 23.1 LLD 40 (25) 10.2 (2.1) (8.6) (4.7) Rapp (2005) DSM-III-R or DSM-IV + ≥ 11 Combined early and late-onset USA [113] 84.1 26.0 on 30-item GDS LLD HC 39 (21) 9.8 (2.0) (6.8) (3.5) 33.3% less 2545 74.8 LLD than high n/a ≥5 on 20-item CES-D (1633) (5.3) Rosano school (subclinical depression) + Combined subclinical and (2016) USA 23.7% less ≥11 on CES-D (clinical clinical depression [114] 2146 74.6 depression) HC than high n/a (1179) (5.4) school Behav. Sci. 2020, 10, 132 14 of 36 71.4 44.6% < 10 26.2 Shimada LLD 657 (245) Combined depressive (4.4) years (2.5) ≥ 6 on 15-item GDS or a (2014) Japan complaints group with 3695 71.5 33.3% < 10 26.5 diagnosis of depression [115] depressed group HC (1921) (5.2) years (2.4) 68.3 DSM-IV criteria measured Smoski LLD 30 (n/a) n/a >26 (6.3) with SCID interview + (CES- (2014) USA 70.8 D. No minimum for [94] HC 40 (n/a) n/a >26 (7.1) inclusion) 70.3 ≥24 in LLD 117 (80) 14.0 (3.6) Steffens DSM-IV criteria + (HAM-D + (7.2) 90% (2001) USA MADRS + DDES. No 70.2 ≥24 in [116] HC 142 (103) 15.8 (2.6) minimum for inclusion) (6.0) 90% Behav. Sci. 2020, 10, 132 15 of 36 4.3. Study Characteristics of Studies on Bilingualism and Aging An overview of all study characteristics can be found in Table 3, below (for a more detailed version please refer to Table S3 in the supplement). The 14 studies that were identified on bilingualism and aging were published between 1997 and 2020, with the majority (n = 11) after 2010. While studies were published in a range of countries (e.g., China, France, Canada), there was a strong bias towards Europe and North-America in participant recruitment. Of the 14 studies, only five studies contained information on all selected inclusion or exclusion criteria. The remaining nine provided no information on comorbid psychiatric illness (n = 5), comorbid neurological illnesses (n = 5), substance or medication abuse (n = 8), or medication status (n = 8). Only two studies used language learning interventions in a group of seniors [117,118]. A complete overview of the individual studies can be found below in Table 3. 4.3.1. Demographic Information A total of 2147 monolingual and 1225 bilingual participants took part in the 12 identified cross- sectional studies. The two intervention studies had 80 language learners and 117 controls. Studies included on average 102.1 bilingual participants (SD = 260.9) and 178.9 monolingual participants (SD = 537.2). These averages were heavily influenced by one large cohort study containing 2812 participants (928 bi/multilinguals). Generally speaking, in the cross-sectional studies the bilingual groups comprised the same number of female participants (M = 58.1%, SD = 24.8) as monolingual groups (M = 59.5%, SD = 10.2), although it must be mentioned that seven of the selected studies either did not report a gender distribution, or they reported it for all participants instead of for the separate groups. Those studies that did not separately report gender for bilingual and monolingual participants had 58.8% female participants (SD = 12.4) on average. In terms of age, monolingual participants in the cross-sectional studies were slightly younger (M = 67.8, SD = 6.8) than bilinguals (M = 70.4, SD = 5.5) in the 10 studies that reported age using mean and standard deviation. 4.3.2. Type of Tasks Used The 14 studies used the following neuropsychological tasks: Digit span forward (n = 6), digit span backward (n = 6), digit span forward + backward (n = 3) (Because a forward + backward version of the digit span task was present, it was decided to pool all implementations of this task for the meta- analyses.), TMT A (n = 4), TMT B (n = 4), TMT B–TMT A (n = 1) (The effect size for this version was taken together with TMT B measures, as the difference between TMT A and B indexes cognitive flexibility [119]), phonetic verbal fluency (n = 4), (m)WCST (n = 3), letter-number sequencing (n = 2), color-shape switch task (n = 1), and the antisaccade task (n = 1). None of the studies (including those on LLD) used the VAT-E. 4.3.3. Operationalization of Bilingualism Substantial variety was attested in terms of how bilingualism was defined. Of the 11 studies comparing monolinguals to bilinguals, one study prompted participants to merely provide the number of languages spoken on a daily basis, without taking into account factors like language proficiency. Only one study used the Common European Framework of Reference for Languages [120] to gauge proficiency. Four studies asked participants for self-reported proficiency measures regarding reading, writing, speaking, and listening. Six studies further supplemented self-rated proficiency with measures regarding, for instance, the frequency and context of use, age or order of acquisition, and language dominance. One of the two studies investigating language learning in seniors required that participants had no previous knowledge of learning another language. The other required that participants had no functional knowledge of the target language (English) but allowed the inclusion of bidialectal participants. Behav. Sci. 2020, 10, 132 16 of 36 Table 3. Study characteristics of individual studies on cognitive function in bilingualism and in language learning interventions. Demographic information includes where the study took place, participants’ age and gender, educational background, and performance on Montreal cognitive assessment or the mini-mental state examination (continued on next page). First MOCA/ Operationali- N Age Education Other Language Author Country Group MMSE Bil. Lang. zation of Notes (fem.) (SD) (SD) Tests (Year) (SD) Bilingualism 10 74.2 MOCA Ansaldo Bi 17.2 (3.1) Bilingual (n/a) (7.4) 27.2 (1.6) French- LEAP-Q [122]; (2015) Canada aphasia test part English ≥30% L2 usage 10 74.5 MOCA [121] Mono 16.1 (3.28) C (n/a) (7.1) 27.7 (1.1) 15 68.8 Speaking two Bi 15.3 (3.7) n/a Context of Craik (n/a) (6.1) languages acquisition; (2006) Canada Varied daily from 15 70.3 frequency of L2 [123] childhood (≤ Mono 16.1 (3.5) n/a (n/a) (4.3) usage 10 years) 60.5 MMSE Semi-structured Authors sent Bi 88 (46) 14 (4) (4.3) 28.7 (1.23) Speaking two interview; a subset of Estanga Basque- languages Bilingual dataset that (2017) Spain 43 60.8 MMSE Spanish regularly and Language Profile adhered to [124] Mono 12 (2) (25) (4.4) 28.4 (1.31) fluently questionnaire inclusion [125] criteria 26 69.7 Self-rating Bi 16.3 (0.5) n/a (n/a) (0.8) reading, Speaking two speaking, Fernand languages listening, writing es (2007) Canada Varied regularly from (1–10); AoA; 16 74.1 [126] Mono 15.5 (0.5) n/a adolescence (n/a) (7.5) frequency of L2 (≤12 years) usage; language preference Friesen 21 71.1 MMSE Speaking two Language and Canada Bi n/a Varied (2015) (n/a) (3.8) >26 languages Social Behav. Sci. 2020, 10, 132 17 of 36 [127] fluently on a Background daily basis Questionnaire [128]; self-rated 20 70.9 MMSE Mono n/a prof.; frequency (n/a) (2.6) >26 of language usage; context of language use 70.0 Data from Bi 20 (11) 13.7 (4) n/a Gubar- (6.6) two Self-rating of chuk USA + Russian- experiments n/a overall English (1997) Russia 68.2 English in the same Mono 20 (10) 12.0 (0) n/a proficiency (1–4) [129] (7.5) study combined 928 Speaking ≥ 2 Bi ≥ 65 n/a n/a Ihle (n/a) languages Dialects did Switzer- (2016) Varied regularly not count as land 1884 [130] (regardless of languages Mono ≥ 65 n/a n/a (n/a) proficiency) 70.6 MOCA Reaching high Self-rating Bi 28 (7) 16.1 (1.27) Johns (5.7) 27.7 (1.3) proficiency in reading, French- (2016) Canada L2 in early speaking, 74.1 MOCA English [131] adolescence listening, writing Mono 16 (10) 15.1 (3.3) (7.5) 27.1 (1.9) (≤13 years) (1–5) 70.7 MOCA Reaching high Self-rating Combined Bi 36 (17) 16.1 (2.9) French- Kousaie (5.9) 27.5 (1.6) proficiency in reading, two English & (2014) Canada L2 in early speaking, monolingual 72.4 MOCA English- [132] Mono 61 (23) 15.7 (2.7) adolescence (≤ listening, writing groups for (6.5) 27.5 (1.5) French 13 years) (1–5) our analysis 71.4 Highly Self-rating López Bi 18 (10) 16 (2.6) 27.6 (1.6) (4.0) proficient in reading, Zunini French- Canada both speaking, (2019) 71.7 English Mono 18 (11) 15.6 (2.7) 27.8 (1.3) languages; no listening, writing [133] (3.5) functional (1–5) Behav. Sci. 2020, 10, 132 18 of 36 knowledge of other languages 16 72.3 MMSE Sociolinguistic Bi 16.0 (2.7) Bilinguals Massa (n/a) (5.0) 29.5 (0.8) French- questionnaire were all (2020) France Italian/Itali regarding 16 71.1 MMSE dominant in [119] Mono 15.1 (2.4) an-French frequency/conte French (n/a) (5.9) 28.9 (1.6) xt of usage Most ≥ 19 MMSE ≥ B2 on CEFR Bi >55 Undergrad English- German Prehn (n/a) 29.7 (0.5) measured with uate German & vocabulary and (2017) Germany Goethe Most ≥ Russian- grammar test [134] 18 MMSE Institute Mono >55 Undergrad German using cloze test (n/a) 29.3 (0.9) Placement Test uate First N Age Global Language Country Group Education Intervention author (fem.) (SD) cogn. learnt? 92% ≥ 8-month MMSE Lang 26 (12) 67.4 secondary language Ramos 28.4 education course by an (2017) Spain Basque 71% ≥ education [117] MMSE Ctrl 17 (9) 69.2 secondary center for 28.5 adults education Most < high ADAS- 53 70.8 Lang. school Cog (43) (6.0) 5 hr per week graduate 8.3 (4.6) 2 to 3 days a Wong Most < high ADAS- 51 71.1 week for 6 (2019) China Game school Cog English (44) (6.5) months in [118] graduate 9.1 (5.4) senior center Most < high ADAS- 49 71.1 (130 hr max) Music school Cog (43) (6.1) graduate 9.1 (5.6) Behav. Sci. 2020, 10, 132 19 of 36 4.4. Risk of Bias in Individual Studies The risk of bias analysis revealed that—out of the 34 identified cross-sectional studies—all but one clearly stated their aims, research questions, and hypotheses. The study populations were in general also adequately described (n = 22). Studies that did not accomplish this mostly failed to mention the location from which participants were recruited. Furthermore, whether groups within a study were recruited from similar populations (e.g., from the same time period or location) was unclear in many cases (n = 16). Strikingly, only one study reported a justification for their sample size [113]. Many of the studies that did not do this, however, did mention small sample sizes as a limitation. Around half of the studies accounted for different levels of depression or bilingualism in their analyses (n = 18). A little over two-thirds of identified studies (n = 22) defined their populations consistently (e.g., using the same screening tools). Those that did not screened for LLD on the basis of two versions of the DSM [113], or they used different measures of language proficiency between groups [127]. Having a double-blind study was essentially impossible in most studies due to targeted recruitment of participants with a certain background; therefore, participants were generally aware of their group-status. However, out of all identified studies only one intervention study stated that the experimenters were blinded to group-status of participants [118], and only one cross-sectional study stated that both mood and neuropsychological function were measured by the same clinician [17]. The remaining 35 studies did not mention blinding of experimenters. It must be noted, however, that for many studies investigating bilinguals, blinding experimenters is essentially impossible, since researchers would often know what participants’ first and second languages were. Lastly, around two-thirds of all studies (n = 21) controlled for potential confounding variables (e.g., age, gender, years of education), either during the recruitment phase or using subsequent statistical methods. While this might not seem high, only one study had significant group differences at baseline that were not controlled for [110]. The remaining studies did not provide enough information to determine whether confounders were present. For a complete overview of the risk of bias analysis with results for individual studies please refer to table S3 in the Supplementary Materials. 4.5. Cognitive Functioning in LLD and Bilingualism The pooled effect sizes for cognitive functioning in LLD are presented in a forest plot in Figure 2. Tasks with insufficient data (i.e, TMT A and B performance measured in percentiles and TMT B perseveration errors) are reported separately in Table 4. Since the two intervention studies looking into late-life language learning used different tests, their results can also be found in Table 4. Behav. Sci. 2020, 10, 132 20 of 36 Figure 2. Forest plot for cognitive functioning in late-life depression showing individual and pooled effect sizes. A positive standardized mean difference indicates a higher average score for the depressed group. Every square in the forest plot represents a study. The bigger the square, the more it weighs in the pooled effect. The blue diamond represents the pooled effect size (expressed through the standardized mean difference, or SMD). The effect on the left of the null line in Figure 2 indicates a lower average score for the LLD group. A negative effect size generally suggests a disadvantage for the LLD group on that particular task (e.g., fewer categories completed in the (m)WCST implies less cognitive flexibility). However, the opposite is the case for TMT A and B (positive effect size means longer completion times), as well as perseverative errors on the (m)WCST (positive effect size means more errors for the LLD group). Behav. Sci. 2020, 10, 132 21 of 36 When looking at the pooled random effect sizes, significant (i.e., with confidence intervals not intersecting zero) small negative effects were found for the digit span task and phonetic verbal fluency. A medium significant positive effect was found for perseveration errors on (m)WCST. Large significant effect sizes were found for TMT A, TMT B, and (m)WCST (categories completed). No significant effects were found for DSST and letter-number sequencing. Overall, judging by the direction of results, the LLD groups seemed to perform worse on neuropsychological tests from our prespecified list. The (pooled) effect sizes for cognitive functioning in bilinguals can be interpreted the same way (Figure 3): any standardized mean difference (SMD) on the left of the null line indicates a lower average score for the bilingual participants. Again, only for TMT A and TMT B does a negative effect size represent an advantage (faster completion times). The pooled effect sizes for all tasks were small. None of the pooled effect sizes reached significance (i.e., the confidence interval for every test intersected the null line), except for TMT A. In TMT A, a small, significant negative effect was found, indicating that bilinguals completed the task faster than monolinguals. The overview with tasks that were only used once in Table 4 contains three separate groups: studies on late-life language learning, studies on lifelong bilingualism, and studies on LLD. Only performance on the digit-span (backward) was found to significantly improve after a foreign language intervention [118]. The other study found no significant differences following a language course intervention. The studies on lifelong bilingualism reported no significant differences between bilingual and monolingual groups, except for Ansaldo et al.’s study [121], which reported an advantage for bilinguals in terms of accuracy on the TMT A and B with a medium and small effect size, respectively. Figure 3. Forest plot for cognitive functioning in older bilinguals showing individual and pooled effect sizes. A positive standardized mean difference indicates a higher average score for the bilingual group. Behav. Sci. 2020, 10, 132 22 of 36 Table 4. Effect sizes of neuropsychological tasks that were used only once. Author Group Test Hedges’ g Conclusion 0.09 (very small) language vs. No significant differences Wong Language Digit-span gaming between language/music/gaming (2019) learning (forward) 0.04 (very small) groups after completing the [118] language vs. intervention music −0.26 Language vs. games Significant gains in digit-span Wong Language Digit-span (small) performance in language learning (2019) learning (backward) −0.58 language intervention, but not in the music [118] vs. music or gaming interventions (medium) Color-shape Decrease in RTs on switch trials Ramos Language- switch 0.1 was slightly larger in language (2017) learning (RTs to (small) learning group than in controls, [117] switch trials) but this was not significant Massa Bilinguals made slightly more Lifelong Antisaccade −0.50 (2020) mistakes than monolinguals, but biling. (% correct) (medium) [119] this was not significant Bilinguals had slightly shorter Massa Antisaccade Lifelong 0.18 response times than (2020) (congruent – biling. (small) monolinguals, but this was not [119] incongruent) significant Ansaldo Lifelong TMT A −0.73 Bilinguals made fewer errors (2015) biling. errors (medium) than monolinguals [121] Ansaldo Lifelong −0.33 Bilinguals made fewer errors (2015) TMT B errors biling. (small) than monolinguals [121] Smoski LLD took slightly longer to TMT A −0.09 (2014) LLD complete TMT A than HC, but (percentile) (small) [94] this was not significant Smoski LLD took longer to complete TMT B −0.47 (2014) LLD TMT B than HC, but this was not (percentile) (medium) [94] significant Steffens TMT B (% LLD group made significantly 0.27 (2001) LLD perseveratio more perseveration errors than (small) [116] n errors) HC 4.6. Risk of Bias Across Studies Separate funnel plots for studies on LLD and studies on bilingualism were made (Figure 4). The results for LLD, while containing a small number of outliers, were mostly symmetrical, although one study had an abnormally large SMD. Results regarding bilingualism were somewhat asymmetrical, suggesting a publication bias for small studies with larger effects [87]. Figure SEQ Figure \* ARABIC 4: Funnel plots for studies on Behav. Sci. 2020, 10, 132 23 of 36 Figure 4. Funnel plots for studies on late-life depression (left) and bilingualism (right). An asymmetrical distribution suggests publication bias. 5. Discussion 5.1. Cognitive Function in LLD Our findings indicate that, in line with previous work [16], LLD is consistently associated with cognitive dysfunction. We reported large impairments in processing speed (TMT A), a small to large disadvantage on multiple measures of cognitive flexibility (TMT B, verbal fluency, (m)WCST categories, (m) WCST perseveration errors), and a small deficit in working memory (digit span). For both measures of cognitive flexibility (the DSST and letter-number sequences), mean effect sizes were in line with our expectations, namely that healthy controls outperformed LLD groups. However, the confidence intervals intercepted the null-line, meaning that this effect was not significant. For the DSST, the absence of a significant effect was potentially caused by one study that actually reported a higher number of correct responses for the depressed group. The authors of that study, however, also mentioned that the LLD patients also took more than twice as long to match a digit to a symbol, suggesting an overall processing disadvantage for the depressed group. Only two studies examined the performance of depressed patients and controls on the letter-number sequencing task. One study found a very large negative effect [110], while the other found no significant difference [106]. This led to a large, yet insignificant, pooled effect size. The study by Leal et al. [110] reflects a well-known issue with smaller sample sizes: they potentially overestimate the actual effect size [135]. This overestimation, combined with the small number of studies that used this test, potentially skewed the results. Two studies were not analyzed in the meta-analysis because they reported the TMT differently from the other studies (i.e., completion time expressed as a percentile score [94]; number of perseveration errors on TMT B [116]). Instead, they were analyzed separately. Both studies found depressed participants performed more poorly, as was hypothesized, although only Steffens et al.’s [116] results reached significance. This also provides further support for our hypothesis that LLD patients experience cognitive dysfunction. In spite of the generally consistent direction of our findings, most of the results showed substantial heterogeneity (as measured with I ). Heterogeneity may have been caused by differences between studies in terms of participants’ educational level, present in the studies measuring DSST, for instance. The heterogeneity may also have been caused by underlying group differences not captured in our data. Since the present study dealt with a small number of studies per comparison, heterogeneity was to some extent expected [136]. Additionally, because it was known beforehand that groups would by no means be the same, a random-effects meta-analysis was chosen a priori [87]. This method produces more conservative estimates, which partially counteracts this heterogeneity [137]. Despite the small number of studies per comparison, this meta-analysis has provided evidence in favor of our hypothesis, suggesting that generally speaking (non-demented) seniors with Behav. Sci. 2020, 10, 132 24 of 36 depressive complaints perform less well on most measures of cognitive functioning in our selected list of neuropsychological tests. These results, therefore, are also in line with previous research, which reported that seniors with depressive complaints have impaired cognitive functioning, overall [16,40]. 5.2. Cognitive Function in Bilinguals While the direction of the effects was generally consistent in the analysis on LLD (despite the heterogeneity), the effects of studies investigating cognitive functioning in bilingualism were highly inconsistent. Overall, none of the pooled effects, except performance on TMT A, reached significance, meaning that statistically speaking our meta-analysis did not support our hypothesis. Rather, our results seem to be in line with Lehtonen et al.’s meta-analysis [57]: no statistical support was present for better preserved cognitive functioning in older bilinguals in the identified studies. Studies specifically looking into language learning in older adults were very sparsely present in our dataset. This can be partially explained by the fact that studies on this topic are scarce, but also because some of the studies on this subject employed neuropsychological tests, not from our predefined list (e.g., [138,139]). Only two studies on late-life language learning could be included [117,118]. These two studies employed no common neuropsychological tests between them, meaning that a pooled average could not be calculated. They will, therefore, be discussed separately. Wong et al. [118] compared a six-month language learning course to a gaming intervention (active control) and a music appreciation intervention (passive control) of the same duration. A small but significant increase was reported exclusively in the language group for working memory, as measured through the backward digit span task. No such increase was reported for the forward digit span. However, the authors did conclude that global cognition, as measured on the basis of the Alzheimer’s Disease Assessment Scale-Cognitive Subscale [140], had improved significantly both in the group learning a foreign language as in those enrolled in the gaming intervention. Ramos and colleagues on the other hand asked participants to complete an eight-month language course [117]. Changes in cognitive flexibility were measured using a color-shape switch task. Reaction times for both switch and non- switch trials decreased considerably between pre-test and post-test in the language training group (by 129 and 157 ms, respectively), while staying relatively constant in a passive control group (69 and 66 ms, respectively). However, the switch cost (the difference in RT between switch and non-switch trials) did not decrease in the language group or the control group. The authors, therefore, concluded that eight months of language training in seniors did not lead to better switching ability, but stated that further research was needed. However, it should be noted that participants in this study could self-enroll for the language course (contrary to Wong et al.’s study, which randomly allocated participants to a group). This may have influenced results. In fact, the language learning group outperformed the control group significantly before the course started, suggesting that the two groups differed at baseline. Lastly, only one type of test was used both at baseline and after completion of the course. The addition of a broader test battery could have provided more nuance. The findings in this meta-analysis do not support our hypothesis that lifelong bilingualism or late-life language learning enhances cognitive functioning. Only processing speed, as measured through the TMT A, was significantly faster in bilinguals and significantly slower in LLD groups. However, the fact that the results both for the meta-analysis and the individual studies were inconsistent does not mean that the potential for language learning to combat cognitive dysfunction in LLD should be discredited altogether. Our meta-analysis on bilingualism was based on 12 of the 14 identified studies in total, which generally had small sample sizes. Furthermore, the small number of studies (in both the meta-analysis on LLD and on bilingualism) may have decreased the accuracy of the summary statistic; something that will be discussed in more detail in the section on Strengths and limitations. Similarly, rejecting the notion that language learning in late-life may lead to cognitive benefits based on two longitudinal studies would also be an overgeneralization. In looking at the study characteristics in Table 3, we see many variations in terms of how the target demographic was defined. As a result, it becomes difficult to reach conclusions on the basis of the meta-analysis presented above. Instead, it is clear that more research is needed. To ensure that results of future Behav. Sci. 2020, 10, 132 25 of 36 research (both regarding lifelong bilingualism and late-life language learning) are more reliable, the variation present in previous research will be critically evaluated, upon which recommendations for further research will be made. 5.3. Issues in Defining Bilingualism To preface this section, while not strictly relevant to research on late-life language learning, the authors noticed much variation in how bilinguals were described. As much of the work investigating the potential merits of bilingualism in later life compares lifelong monolinguals to lifelong bilinguals, we feel it is important to discuss the term ‘bilingual’. Both the studies on bilingualism and those on LLD used a variety of (diagnostic) criteria and screening tools to differentiate between the two groups they compared. What set studies on LLD apart from the bilingualism studies, however, was their use of validated tools such as the Geriatric Depression Scale and the 17-item Hamilton Depression Rating Scale. These scales were designed to distinguish between healthy and depressed older adults, according to widely accepted DSM and ICD-10 criteria, with high accuracy [141,142]. The studies on language, on the other hand, were much less consistent in defining their monolingual and bilingual participant groups. The term ‘bilingual’ was for the most part operationalized by combinations of proficiency, age of acquisition, and usage patterns. Some studies, for instance, required that bilingual participants had reached a high level of proficiency in the second language before a certain age, which ranged anywhere from 10 to 13 years [123,126,131,132]. What constituted high language proficiency, however, also varied substantially. This was often determined by means of self-reported scales regarding production and reception [129], while only one study used more standardized and objective measures (i.e., based on CEFR guidelines [120]) of language proficiency [134]. Other studies further defined the bilingual groups as those who spoke two languages on a regular basis [123,126,130]. Overall, though, the many ways in which bilingualism was operationalized very likely played a role in the inconsistent results reported above. These differences between the bilinguals in one study versus those in another are likely caused by the lack of consensus on a ‘case definition’ of bilingualism. In this regard, our findings strongly mirror those reported in a systematic review by Surrain and Luk [143]. They, too, noted that the bilingual experience is a multidimensional construct, consisting of the frequency of use, proficiency, contexts in which languages are used, and sociolinguistic background variables such as the attitudes towards bilingualism in a certain region. In their review of work on bilingualism published between 2000 and 2015, they found that definitions varied considerably. Most studies emphasized proficiency and language use, with less than half using objective measures; this was especially common in studies investigating adult participants. An additional issue was that oftentimes the exposure to languages was presented purely categorically (e.g., whether or not someone spoke multiple languages in the home), instead of on a continuous scale (e.g., percentage of first language usage in the home). As a result, relevant data regarding the situations in which participants used their languages were not present. Surrain and Luk noted that coming to a consensus regarding what makes a bilingual is not realistic. Instead, they argued that our understanding of the cognitive effects of bilingualism would only become stronger when studies included a more complete and transparent description of bilingual experiences. This should include not only proficiency, but also language history, language usage patterns, and sociolinguistic information (e.g., the status of bilingualism in the country where the study took place or personal attitudes towards one’s own bilingualism). 5.4. Issues Pertaining to Older Populations Specifically The inconsistent definition of bilingualism is problematic, but to some extent unavoidable due to the many opinions on what constitutes a bilingual experience and the many (confounding) factors at play in individual bilingual experience. Similarly, in studying older adults, researchers need to take into account certain confounding variables particular to the demographic. One such issue would be the presence of (comorbid) disorders known to affect cognition, such as dementia or mild cognitive impairment (MCI). A majority of the studies on bilingualism (n = 10) and those regarding LLD (n = Behav. Sci. 2020, 10, 132 26 of 36 20) that were included in our review employed screening tools for cognitive impairment and dementia such as the MMSE and the MoCa [83,84]. At the same time, nearly a third of the studies on bilingualism (n = 4), did not report using such tools, or the included participants self-reported that they were healthy. While they can never be a replacement for a diagnosis by a medical professional, tools like the MMSE are able to indicate if a participant has a cognitive impairment that could act as a confounder. An area where more studies (including those on LLD) were lacking, though, was ensuring that participants were not taking medication that could affect cognitive functioning. Out of the 14 studies on bilingualism, 10 did not have sufficient information available to ensure that participants were not prescribed benzodiazepines or beta-blockers. Additionally, 14 out of 23 studies on LLD did not report on this, either. As mentioned before, this is especially problematic for benzodiazepines [81], which are commonly prescribed to and even abused by seniors [144]. 6. Directions for Future Research Results of the present study seem to suggest that, as was hypothesized, cognitive dysfunction in LLD is consistently present. Our meta-analysis also shows that results investigating cognitive functioning in older bilinguals are inconsistent and that no significant evidence was found in our sample that indicated preserved cognitive function in older bilinguals. The authors also noticed that very little research had been done on the effects of late-life language learning on cognition. While the authors of the present study are primarily interested in the role that language learning may have in later life, we realize that cross-sectional study designs are not as resource-intensive as longitudinal designs. For this reason, they are more commonly used to investigate the potential effects of bilingual experiences in seniors. The following section will therefore also give directions for future research comparing lifelong bilinguals to lifelong monolinguals. 6.1. Language Learning Interventions Only two studies investigating the effects of late-life language learning on cognition were identified for this review. Given the strong potential of language learning as a cognitive training tool [66], more work is clearly needed. In addition to general methodological improvements mentioned further on, forthcoming longitudinal, research investigating the potential effects of late-life language learning can improve in quality by assigning participants to interventions randomly, like Wong et al. [118]. While this does not completely eliminate the selection bias caused by recruiting participants from the community, it at least negates it somewhat. Additionally, from an applied linguistics perspective, we recommend future intervention studies to tailor their language learning interventions specifically to characteristics, needs, and preferences of seniors [67,145] to maximize the potential benefits of late-life language learning on a motivational, cognitive, and perhaps social (well-being) level. Another important variable that the present analysis did not look into due to a lack of studies on late-life language learners is the appropriate intervention duration. One can imagine that an ideal outcome would follow a multiple year intensive program; indeed, the activities that build up cognitive reserve generally seem to be garnered over a lifetime of repeated exposure [30]. In practice, however, such trials would be incredibly impractical due to participant attrition and budgetary constraints. A 2017 meta-analysis investigated the efficacy of several types of cognitive training in seniors [146]. The interventions these researchers identified consisted of one to 180 sessions and total exposure lasted between 1 and 270 h. Their analysis concluded that 20 or more training sessions administered one or two times per week led to stronger net gains. Their study did not find that intervention duration by itself influenced the effects. Another review investigating primarily the effect of short language learning interventions in mostly student-aged populations found structural changes in brain anatomy (e.g., increased gray matter) after as little as three months of intensive language learning [147]. Improvements in performance on neuropsychological tasks could arise even faster, though. One study reported a sustained improvement in cognitive flexibility (i.e., still present nine months post-intervention) after only one week of intensive language training [139]. In short, the key for optimal results seems to be intervention intensity instead of intervention Behav. Sci. 2020, 10, 132 27 of 36 duration. Further research should focus on an ideal balance between the two. Finally, while this was not the focus of the present study, we feel that researchers in geriatric psychiatry and linguists alike should also study the potential effects of language learning on well-being. While research on this is incredibly sparse at the moment, the few published studies suggest that late-life language learning is an experience that fosters social interactions and feelings of empowerment over one’s aging process [138,148]. 6.2. Lifelong Bilinguals Versus Lifelong Monolinguals A number of potential reasons may explain the inconsistent findings regarding cognitive functioning in lifelong bilinguals. However, if we start at the root, the inconsistency may be (partly) caused by the operationalization of the term ‘bilingualism’, which ranged from an index of the total number of languages spoken daily (without taking into account dialects or language-proficiency) to more complex measures including frequency of use and context of acquisition. For future research, we recommend using a more transparent operationalization of bilingualism, in line with Surrain and Luk’s recommendations [143], which goes further than, for instance, summing the number of languages a participant reports to speak. This means being more explicit in describing the bilinguals: what languages do they speak, when did they learn them, how do they use their languages, what are their attitudes towards their languages? We believe that using background questionnaires such as the Language Experience and Proficiency Questionnaire (LEAP-Q) [122] and the Language and Social Background Questionnaire LSBQ [128] will help attain this. The information that comes from this, however, is often difficult to use in statistical analyses because resulting the data are quite extensive. An additional, very recently developed, measure that could be useful is language entropy [149]. Entropy essentially takes language background questionnaire data and converts it to an index expressing how balanced participants are in their language use. 6.3. General Methodological Improvements More generally, studies both on bilinguals and LLD were often lacking in the medical information collected and/or reported by the authors. It may be difficult to obtain a participant’s full medical history. However, if it is known whether participants take certain medications, or if they have cognitive impairments as measured through short screening tools such as the MMSE or MoCa, it is at least possible to control for this in analyses. We, therefore, recommend that forthcoming work targetting older bilinguals or seniors learning a language (whether healthy or depressed) records this information, too. On the one hand, this will improve the reliability of new research findings within the field of foreign language acquisition. For instance, it is well-known that long-term benzodiazepine use in seniors leads to considerable cognitive deficits [81]. Including these individuals without controlling for their medication use, then, introduces an unwanted confounder. At the same time, some relatively strong claims regarding bilingualism’s neuroprotective effects are currently being made. The inclusion of these measures will enhance the (clinical) relevance of research on bilingualism for those working in other fields, like gerontology. Controlling for these variables, then, will not only improve overall study quality, but it will also further stimulate understanding and cooperation between disciplines. Furthermore, based on what was seen in the risk of bias assessment, a number of recommendations for forthcoming cross-sectional studies both on LLD and bilingualism can be made. Firstly, researchers should reduce bias by including power analyses, and by blinding assessors for participants’ group-status as much as possible. If the latter is not possible due to, for instance, the language of a particular task giving away group status, researchers should be more transparent about this. Similarly, bias could be reduced by increasing transparency regarding where and how participant groups were recruited. Again, it may not always be possible to recruit subjects from one clinic. However, a considerable proportion of studies did not provide enough information to determine whether participants were recruited from similar populations. Lastly, while over half of the identified studies reported that the researchers controlled for potential confounding variables Behav. Sci. 2020, 10, 132 28 of 36 (such as age or education level), a large proportion of studies (n = 14) did not provide enough information to determine whether this was done. Again, we recommend that researchers explicitly report if participant groups differed on confounding variables and whether they controlled for this in their analyses. 7. Strengths and Limitations While it has been tentatively proposed that language learning may have a positive effect on cognitive functioning in seniors with depression [66], no study as of yet has bridged the two separate disciplines needed to research this (i.e., linguistics and geriatric psychiatry). A strength of the present research was the use of weighted averages across tests. Vote-counting would have resulted in an outcome that would have seemed more intuitive on the surface, but this method is flawed [87–89]. In interpreting the results of this study, a number of limitations should be considered, however. Firstly, the search strategies were developed to identify studies using specific validated neuropsychological tests that would likely also be used in ongoing work (including but not limited to our own lab). On the one hand, this was an advantage because it made the studies relatively easy to compare. A disadvantage, on the other hand, was that some relevant studies were excluded because they did not use tests from our predefined list. A study by Bialystok et al. [150], for instance, found that older bilinguals were better at certain types of conflict resolution than age-matched monolinguals. However, because the tests they employed (modified antisaccade tasks) did not match our list, their study had to be excluded. Another shortcoming of the present review was a result of it being conducted during the COVID-19 pandemic. A large majority of the corresponding authors who replied to our queries were more than willing to provide the information requested. However, due to travel restrictions, some were unable to access the data requested. Additionally, since the majority of identified studies did not report or collect data regarding all our inclusion criteria (e.g., medication use), there is a chance that some studies contained undesirable confounders. Regarding medication use, the exclusion of beta-blockers was chosen because reported side-effects include reduced neuropsychological functioning and increased risk for depression [151]. Recent reviews, however, have provided evidence contradicting this notion [152]. In practice, however, the majority of studies excluded for undesirable medication were due to benzodiazepines, which are known to affect cognition significantly [81]. Another limitation was that the selection of publications based on inclusion and exclusion criteria was done by one author (JB). Potential bias caused by this, however, was negated by showing the reviewer (JB) only the title and abstracts. Another major shortcoming is the small number of studies identified. Borenstein et al.’s comprehensive introduction to meta- analyses does, in fact, warn against a lowered accuracy when performing a random-effects meta- analysis on the basis of few studies [87]. However, immediately afterward they also state that “when faced with a series of studies people have an almost irresistible tendency to draw some summary conclusions from them” (i.e., vote-counting). Borenstein et al. therefore suggest that it is preferable to still compute a summary statistic with known shortcomings instead of inviting unknown properties, a procedure followed in the present investigation. The last shortcoming was using studies on lifelong bilinguals in lieu of papers focusing solely on language learning interventions. We argue that bilingualism is a continuum of experiences [153] and that part of the beneficial effect stems from juggling two competing lexicons [154], a process underlying both the language experience of lifelong bilinguals and new learners. This is why we opted to supplement data on late-life language learners with that of older lifelong bilinguals. However, to say that a lifelong bilingual and a late-life language learner are equivalent in terms of cognitive performance is impossible to say based on the limited existing research. 8. Conclusions A growing body of research is emerging regarding the potential positive effects that bilingualism and language learning may have on old-age disorders [66,155]. These studies have typically focused on building up cognitive reserve across the lifespan in order to stave off clinical symptoms of Behav. Sci. 2020, 10, 132 29 of 36 dementia [60–64]. However, language learning’s potential therapeutic role in late-life depression (LLD), which has been theorized to be a prodromal manifestation of Alzheimer’s [20,21], is underinvestigated. In order to lay out a theoretical basis for future (interdisciplinary) research, this review aimed to provide an overview of the cognitive domains affected in LLD, as well as an overview of cognitive functioning in older bilinguals and seniors learning a language. The results show that despite heterogeneity between studies, LLD was consistently associated with reduced performance on neuropsychological tests. Studies on bilingualism, however, like Lehtonen et al.’s meta-analysis [57] did not show consistent evidence in favor of enhanced cognitive functioning in bilinguals. We believe that this was likely caused by a varying operationalization of bilingualism and a lack of controlling for age-specific confounders. For further research, we suggest taking these potential pitfalls into account. Lastly, it was noticed that the body number of studies on longitudinal language interventions in seniors was very small Aside from a general call to investigate late-life language learning further, the authors recommend that linguists and researchers in gerontology alike be mindful of possibilities reduce their study’s risk of bias through, for instance, randomization of group allocation. Supplementary Materials: Supplementary materials can be found in the open science framework (https://osf.io/x3wqz/?view_only=627012de77b44e729cc3c04dcaab534f). They consist of an overview of the selection procedure for both the search on bilingualism (S1) and LLD (S2), as well as a complete overview of the risk of bias assessment (S3). Author Contributions: The conceptualization of the project and the development of methods was done by J.B., M.K., and R.K. Search strategies, extraction of results, screening, and analyses were done by J.B., with feedback from M.K. and R.K. Risk of bias assessment was done by F.v.d.B. and J.B. All authors helped draft the manuscript, and discussed content, results, and writing at length (J.B., H.L., F.v.d.B., R.K., M.K.). All authors have read and agreed to the published version of the manuscript. Funding: Funding for this project was provided by the Dutch Research Council NWO (grant no. 016.VIDI/185.190 awarded to Merel Keijzer). The funding agency had no active role in the design, execution, or analysis of the research. Acknowledgments: The authors would like to thank the three anonymous reviewers for their constructive feedback and useful comments that greatly improved the quality of our manuscript. Additionally, we would like to extend our gratitude to Marie-José van Tol for her helpful comments on an earlier version of the manuscript. Lastly, we are incredibly grateful for the corresponding authors of the identified papers for their willingness to answer our queries despite an ongoing pandemic. Conflicts of Interest: The authors declare no conflict of interest. References 1. Kangas, O. One hundred years of money, welfare and death: Mortality, economic growth and the development of the welfare state in 17 OECD countries 1900–2000. Int. J. Soc. Welf. 2010, 19, S42–S59. 2. Meerding, W.J.; Bonneux, L.; Polder, J.J.; Koopmanschap, M.A.; van der Maas, P.J. Demographic and epidemiological determinants of healthcare costs in Netherlands: Cost of illness study. BMJ 1998, 317, 111– 3. 2020 Alzheimer’s disease facts and figures. Alzheimer’s Dement. 2020, 16, 391–460. 4. Hebert, L.E.; Weuve, J.; Scherr, P.A.; Evans, D.A. Alzheimer disease in the United States (2010–2050) estimated using the 2010 census. Neurology 2013, 80, 1778–1783. 5. American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders 1; American Psychiatric Association: Arlington, VA, USA, 2013, doi:10.1176/appi.books.9780890425596.893619. 6. Blazer, D.G. Depression in Late Life: Review and Commentary. Focus (Madison) 2009, 7, 118–136. 7. Djernes, J.K. Prevalence and predictors of depression in populations of elderly: A review. Acta Psychiatr. Scand. 2006, 113, 372–387. 8. Guerra, M.; Prina, A.M.; Ferri, C.P.; Acosta, D.; Gallardo, S.; Huang, Y.; Jacob, K.S.; Jimenez-Velazquez, I.Z.; Llibre Rodriguez, J.J.; Liu, Z.; et al. A comparative cross-cultural study of the prevalence of late life depression in low and middle income countries. J. Affect. Disord. 2016, 190, 362–368. Behav. Sci. 2020, 10, 132 30 of 36 9. Wagenaar, D.B.; Mickus, M.A.; Gaumer, K.A.; Colenda, C.C. Late-Life Depression and Mental Health Services in Primary Care. J. Geriatr. Psychiatry Neurol. 2002, 15, 134–140. 10. Hinton, L.; Zweifach, M.; Tang, L.; Unützer, J.; Oishi, S. Gender Disparities in the Treatment of Late-Life Depression: Qualitative and Quantitative Findings From the IMPACT Trial. Am. J. Geriatr. Psychiatry 2006, 14, 884–892. 11. Mitchell, A.J.; Rao, S.; Vaze, A. do primary care physicians have particular difficulty identifying late-life depression? A meta-analysis stratified by age. Psychother. Psychosom. 2010, 79, 285–294. 12. Conwell, Y.; Van Orden, K.; Caine, E.D. Suicide in older adults. Psychiatr. Clin. North Am. 2011, 34, 451–468. 13. Covey, L.S.; Glassman, A.H.; Stetner, F. Cigarette smoking and major depression. J. Addict. Dis. 1998, 17, 35–46. 14. Brière, F.N.; Rohde, P.; Seeley, J.R.; Klein, D.; Lewinsohn, P.M. Comorbidity between major depression and alcohol use disorder from adolescence to adulthood. Compr. Psychiatry 2014, 55, 526–533. 15. Kim, J.-M.; Stewart, R.; Shin, I.-S.; Yoon, J.-S.; Lee, H.-Y. Lifetime urban/rural residence, social support and late-life depression in Korea. Int. J. Geriatr. Psychiatry 2004, 19, 843–851. 16. Koenig, A.M.; Bhalla, R.K.; Butters, M.A. Cognitive functioning and late-life depression. J. Int. Neuropsychol. Soc. 2014, 20, 461–467. 17. Rajtar, A.; Sałakowski, A.; Rajtar-Zembaty, J.; Starowicz-Filip, A. Executive dysfunction in late-life depression. Psychiatr. Pol. 2017, 51, 705–718. 18. Diniz, B.S.; Butters, M.A.; Albert, S.M.; Dew, M.A.; Reynolds, C.F. Late-life depression and risk of vascular dementia and Alzheimer’s disease: Systematic review and meta-analysis of community-based cohort studies. Br. J. Psychiatry 2013, 202, 329–335. 19. Zanetti, M.; Ballabio, C.; Abbate, C.; Cutaia, C.; Vergani, C.; Bergamaschini, L. Mild cognitive impairment subtypes and vascular dementia in community-dwelling elderly people: A 3-year follow-up study. J. Am. Geriatr. Soc. 2006, 54, 580–586. 20. Panza, F.; Frisardi, V.; Capurso, C.; D’Introno, A.; Colacicco, A.M.; Imbimbo, B.P.; Santamato, A.; Vendemiale, G.; Seripa, D.; Pilotto, A.; et al. Late-Life depression, mild cognitive impairment, and dementia: Possible continuum? Am. J. Geriatr. Psychiatry 2010, 18, 98–116. 21. Butters, M.A.; Young, J.B.; Lopez, O.; Aizenstein, H.J.; Mulsant, B.H.; Reynolds, C.F.; DeKosky, S.T.; Becker, J.T. Pathways linking late-life depression to persistent cognitive impairment and dementia. Dialogues Clin. Neurosci. 2008, 10, 345–357. 22. Crocco, E.A.; Castro, K.; Loewenstein, D.A. how late-life depression affects cognition: Neural mechanisms. Curr. Psychiatry Rep. 2010, 12, 34–38. 23. Blazer, D.G.; Hybels, C.F. Origins of depression in later life. Psychol. Med. 2005, 35, 1241–1252. 24. Chang, S.-C.; Pan, A.; Kawachi, I.; Okereke, O.I. Risk factors for late-life depression: A prospective cohort study among older women. Prev. Med. (Balt. ). 2016, 91, 144–151. 25. Hays, J.C.; Landerman, L.R.; George, L.K.; Flint, E.P.; Koenig, H.G.; Land, K.C.; Blazer, D.G. Social correlates of the dimensions of depression in the elderly. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 1998, 53B, P31–P39. 26. Lue, B.-H.; Chen, L.-J.; Wu, S.-C. Health, financial stresses, and life satisfaction affecting late-life depression among older adults: A nationwide, longitudinal survey in Taiwan. Arch. Gerontol. Geriatr. 2010, 50, S34– S38. 27. Kurlowicz, L.H. Social factors and depression in late life. Arch. Psychiatr. Nurs. 1993, 7, 30–36. 28. Kim, J.-M.; Shin, I.-S.; Yoon, J.-S.; Stewart, R. Prevalence and correlates of late-life depression compared between urban and rural populations in Korea. Int. J. Geriatr. Psychiatry 2002, 17, 409–415. 29. Stern, Y. Cognitive reserve. Neuropsychologia 2009, 47, 2015–2028. 30. Scarmeas, N.; Stern, Y. Cognitive reserve and lifestyle. J. Clin. Exp. Neuropsychol. 2003, 25, 625–633. 31. Leles da Costa Dias, F.; Teixeira, A.L.; Cerqueira Guimarães, H.; Borges Santos, A.P.; Rios Fonseca Ritter, S.; Barbosa Machado, J.C.; Tonidandel Barbosa, M.; Caramelli, P. Prevalence of late-life depression and its correlates in a community-dwelling low-educated population aged 75+ years: The Pietà study. J. Affect. Disord. 2019, 242, 173–179. 32. Gertner, A.K.; Domino, M.E.; Dow, W.H. Risk factors for late-life depression and correlates of antidepressant use in Costa Rica: Results from a nationally-representative longitudinal survey of older adults. J. Affect. Disord. 2017, 208, 338–344. Behav. Sci. 2020, 10, 132 31 of 36 33. Pollock, B.G. The pharmacokinetic imperative in late-life depression. J. Clin. Psychopharmacol. 2005; 25, S19– S23. 34. Solai, L.K.K.; Mulsant, B.H.; Pollock, B.G. Selective serotonin reuptake inhibitors for late-life depression: A comparative review. Drugs and Aging 2001, 18, 355–368. 35. Rapaport, M.H.; Schneider, L.S.; Dunner, D.L.; Davies, J.T.; Pitts, C.D. Efficacy of controlled-release paroxetine in the treatment of late-life depression. J. Clin. Psychiatry 2003, 64, 1065–1074. 36. Schneider, L.S.; Nelson, J.C.; Clary, C.M.; Newhouse, P.; Krishnan, K.R.R.; Shiovitz, T.; Weihs, K. An 8- Week Multicenter, Parallel-Group, Double-Blind, Placebo-controlled study of sertraline in elderly outpatients with major depression. Am. J. Psychiatry 2003, 160, 1277–1285. 37. Tedeschini, E.; Levkovitz, Y.; Iovieno, N.; Ameral, V.E.; Nelson, J.C.; Papakostas, G.I. Efficacy of antidepressants for late-life depression. J. Clin. Psychiatry 2011, 72, 1660–1668. 38. Bohlmeijer, E.; Smit, F.; Cuijpers, P. Effects of reminiscence and life review on late-life depression: A meta- analysis. Int. J. Geriatr. Psychiatry 2003, 18, 1088–1094. 39. Cuijpers, P.; van Straten, A.; Smit, F. Psychological treatment of late-life depression: A meta-analysis of randomized controlled trials. Int. J. Geriatr. Psychiatry 2006, 21, 1139–1149. 40. Bhalla, R.K.; Butters, M.A.; Mulsant, B.H.; Begley, A.E.; Zmuda, M.D.; Schoderbek, B.; Pollock, B.G.; Reynolds, C.F.; Becker, J.T. Persistence of neuropsychologic deficits in the remitted state of late-life depression. Am. J. Geriatr. Psychiatry 2006, 14, 419–427. 41. Koenig, A.M.; Delozier, I.J.; Zmuda, M.D.; Marron, M.M.; Begley, A.E.; Anderson, S.J.; Reynolds, C.F.; Arnold, S.E.; Becker, J.T.; Butters, M.A. Neuropsychological functioning in the acute and remitted states of late-life depression. J. Alzheimer’s Dis. 2015, 45, 175–185. 42. Liao, W.; Zhang, X.; Shu, H.; Wang, Z.; Liu, D.; Zhang, Z. The characteristic of cognitive dysfunction in remitted late life depression and amnestic mild cognitive impairment. Psychiatry Res. 2017, 251, 168–175. 43. Ramscar, M.; Hendrix, P.; Love, B.; Baayen, H. Learning is not decline. Ment. Lex. 2014, 8, 450–481. 44. Ramscar, M.; Hendrix, P.; Shaoul, C.; Milin, P.; Baayen, H. The Myth of cognitive decline: Non-linear dynamics of lifelong learning. Top. Cogn. Sci. 2014, 6, 5–42. 45. Formosa, M. Education and older adults at the university of the third age. Educ. Gerontol. 2012, 38, 114–126. 46. Kelly, M.E.; Loughrey, D.; Lawlor, B.A.; Robertson, I.H.; Walsh, C.; Brennan, S. The impact of cognitive training and mental stimulation on cognitive and everyday functioning of healthy older adults: A systematic review and meta-analysis. Ageing Res. Rev. 2014, 15, 28–43. 47. Kueider, A.M.; Parisi, J.M.; Gross, A.L.; Rebok, G.W. Computerized cognitive training with older adults: A systematic review. PLoS ONE 2012, 7, e40588. 48. Anguera, J.A.; Gunning, F.M.; Areán, P.A. Improving late life depression and cognitive control through the use of therapeutic video game technology: A proof-of-concept randomized trial. Depress. Anxiety 2017, 34, 508–517. 49. Morimoto, S.S.; Wexler, B.E.; Liu, J.; Hu, W.; Seirup, J.; Alexopoulos, G.S. Neuroplasticity-based computerized cognitive remediation for treatment-resistant geriatric depression. Nat. Commun. 2014, 5, 50. Motter, J.N.; Pimontel, M.A.; Rindskopf, D.; Devanand, D.P.; Doraiswamy, P.M.; Sneed, J.R. Computerized cognitive training and functional recovery in major depressive disorder: A meta-analysis. J. Affect. Disord. 2016, 189, 184–191. 51. Seitz, A.R. A New framework of design and continuous evaluation to improve brain training. J. Cogn. Enhanc. 2018, 2, 78–87. 52. Olfers, K.J.F.; Band, G.P.H.; Colzato, L.S.; Hommel, B. The Case of the Leiden “Lifestyle 2030” Study. In Theory-Driven Approaches to Cognitive Enhancement; Colzato, L.S., Ed.; Springer International Publishing:Cham, Germany, 2017; pp. 305–317. 53. Kaushanskaya, M.; Marian, V. The bilingual advantage in novel word learning. Psychon. Bull. Rev. 2009, 16, 705–710. 54. Costa, A.; Hernández, M.; Costa-Faidella, J.; Sebastián-Gallés, N. On the bilingual advantage in conflict processing: Now you see it, now you don’t. Cognition 2009, 113, 135–149. 55. Bialystok, E. Bilingualism Research: Visions For The Future. Available online: https://www.youtube.com/watch?v=Asyf1my3RP4&t=14m26(accessed on 27 August 2020). 56. Bialystok, E.; Craik, F.I.M. Cognitive and linguistic processing in the bilingual mind. Curr. Dir. Psychol. Sci. 2010, 19, 19–23. Behav. Sci. 2020, 10, 132 32 of 36 57. Lehtonen, M.; Soveri, A.; Laine, A.; Järvenpää, J.; de Bruin, A.; Antfolk, J. Is bilingualism associated with enhanced executive functioning in adults? A meta-analytic review. Psychol. Bull. 2018, 144, 394–425. 58. Nichols, E.S.; Wild, C.J.; Stojanoski, B.; Battista, M.E.; Owen, A.M. Bilingualism affords No general cognitive advantages: A population study of executive function in 11,000 people. Psychol. Sci. 2020, 31, 548– 59. Laine, M.; Lehtonen, M. Cognitive consequences of bilingualism: Where to go from here? Lang. Cogn. Neurosci. 2018, 33, 1205–1212. 60. Alladi, S.; Bak, T.H.; Duggirala, V.; Surampudi, B.; Shailaja, M.; Shukla, A.K.; Chaudhuri, J.R.; Kaul, S. Bilingualism delays age at onset of dementia, independent of education and immigration status. Neurology 2013, 81, 1938–1944. 61. Bialystok, E.; Craik, F. I. M.; Freedman, M. Delaying the Onset of Alzheimer Disease. Neurology 2010, 75, 1726–1729. 62. Woumans, E.; Santens, P.; Sieben, A.; Versijpt, J.; Stevens, M.; Duyck, W. Bilingualism Delays Clinical Manifestation of Alzheimer’s Disease. Bilingualism 2015, 18, 568–574. 63. Bialystok, E.; Craik, F. I. M.; Freedman, M. Bilingualism as a Protection against the Onset of Symptoms of Dementia. Neuropsychologia 2007, 45, 459–464. 64. Schweizer, T. A.; Ware, J.; Fischer, C. E.; Craik, F. I. M.; Bialystok, E. Bilingualism as a Contributor to Cognitive Reserve: Evidence from Brain Atrophy in Alzheimer’s Disease. Cortex 2012, 48, 991–996. 65. Poarch, G. J.; van Hell, J. G. Cross-Language Activation in Children’s Speech Production: Evidence from Second Language Learners, Bilinguals, and Trilinguals. J. Exp. Child Psychol. 2012, 11, 419–438. 66. Antoniou, M.; Gunasekera, G. M.; Wong, P. C. M. Foreign Language Training as Cognitive Therapy for Age-Related Cognitive Decline: A Hypothesis for Future Research. Neurosci. Biobehav. Rev. 2013, 37, 2689– 67. van der ploeg, M.; Keijzer, M.; Lowie, W. Methodological Concerns and Their Solutions in Third-Age Language Learning Studies. Dutch J. Appl. Psycholinguist. 2020, in press. 68. Wechsler, D. WAIS-IV : Wechsler Adult Intelligence Scale. San Antonio, TX Psychol. Corp. 2008, 36, 3245– 69. Partington, J. E.; Leiter, G. R. Partington’s Pathways Test. Psychol. Serv. Cebter J. 1949, 1, 11–20. 70. Grant, D. A.; Berg, E. A Behavioral Analysis of Degree of Reinforcement and Ease of Shifting to New Responses in a Weigl-Type Card-Sorting Problem. J. Exp. Psychol. 1948, 38, 404–411. 71. Prior, A.; Macwhinney, B. A Bilingual Advantage in Task Switching. Bilingualism 2010, 13, 253–262. 72. Hallett, P. E. Primary and Secondary Saccades to Goals Defined by Instructions. Vision Res. 1978, 18, 1279– 73. Meyer, S. R. A.; de Jonghe, J. F. M.; Schmand, B.; Ponds, R. W. H. M. The Visual Association Test-Extended: A Cross-Sectional Study of the Performance Validity Measures. Clin. Neuropsychol. 2017, 31, 798–813. 74. Bowie, C. R.; Harvey, P. D.; Administration and interpretation of the Trail Making Test. Nat. Protoc. 2006, 1, 2277-2281. 75. Naismith, S. L.; Rogers, N. L.; Lewis, S. J. G.; Terpening, Z.; Ip, T.; Diamond, K.; Norrie, L.; Hickie, I. B. Sleep Disturbance Relates to Neuropsychological Functioning in Late-Life Depression. J. Affect. Disord. 2011, 132, 139–145. 76. Morin, R. T.; Gonzales, M. M.; Bickford, D.; Catalinotto, D.; Nelson, C.; Mackin, R. S. Impaired Financial Capacity in Late-Life Depression: Revisiting Associations with Cognitive Functioning. J. Int. Neuropsychol. Soc. 2019, 25, 1088–1093. 77. Lamers, F.; Van Oppen, P.; Comijs, H. C.; Smit, J. H.; Spinhoven, P.; Van Balkom, A. J. L. M.; Nolen, W. A.; Zitman, F. G.; Beekman, A. T. F.; Penninx, B. W. J. H. Comorbidity Patterns of Anxiety and Depressive Disorders in a Large Cohort Study: The Netherlands Study of Depression and Anxiety (NESDA). J. Clin. Psychiatry 2011, 72, 342–348. 78. Hugo, J.; Ganguli, M. Dementia and Cognitive Impairment. Epidemiology, Diagnosis, and Treatment. Clin. Geriatr. Med. 2014, 30, 421–442. 79. Riedel, O.; Klotsche, J.; Spottke, A.; Deuschl, G.; Förstl, H.; Henn, F.; Heuser, I.; Oertel, W.; Reichmann, H.; Riederer, P.; et al. Cognitive Impairment in 873 Patients with Idiopathic Parkinson’s Disease: Results from the German Study on Epidemiology of Parkinson’s Disease with Dementia (GEPAD). J. Neurol. 2008, 255, 255–264. Behav. Sci. 2020, 10, 132 33 of 36 80. Kist, N.; Sandjojo, J.; Kok, R. M.; Van Den Berg, J. F. Cognitive Functioning in Older Adults with Early, Late, and Very Late Onset Alcohol Dependence. Int. Psychogeriatrics 2014, 26, 1863–1869. 81. Barker, M. J.; Greenwood, K. M.; Jackson, M.; Crowe, S. F. Cognitive Effects of Long-Term Benzodiazepine Use: A Meta-Analysis. CNS Drugs 2004, 18, 37–48. 82. Drayer, D. E. Lipophilicity, Hydrophilicity, and the Central Nervous System Side Effects of Beta Blockers. Pharmacother. J. Hum. Pharmacol. Drug Ther. 1987, 7, 87–91. 83. Folstein, M. F.; Folstein, S. E.; McHugh, P. R. “Mini-Mental State”. A Practical Method for Grading the Cognitive State of Patients for the Clinician. J. Psychiatr. Res. 1975, 12, 189–198. 84. Nasreddine, Z. S.; Phillips, N. A.; Bédirian, V.; Charbonneau, S.; Whitehead, V.; Collin, I.; Cummings, J. L.; Chertkow, H. The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool for Mild Cognitive Impairment. J. Am. Geriatr. Soc. 2005, 53, 695–699. 85. National Heart, Lung, and B. I. Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools (accessed 27 August 2020) 86. National Institutes of Health. Quality Assessment of Controlled Intervention Studies - NHLBI, NIH https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools (accessed 27 August 2020). 87. Borenstein, M.; Hedges, L. V.; Higgins, J. P. T.; Rothstein, H. R. Introduction to Meta-Analysis; John Wiley & Sons, Ltd: New York, USA; 2009. 88. Combs, J. G.; Ketchen, D. J.; Crook, T. R.; Roth, P. L. Assessing Cumulative Evidence within “Macro” Research: Why Meta-Analysis Should Be Preferred Over Vote Counting. J. Manag. Stud. 2011, 48, 178–197. 89. Friedman, L. Why Vote-Count Reviews Don’t Count (Multiple Letters). Biol. Psychiatry 2001, 49, 161–162. 90. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria. 2020. https://doi.org/R Foundation for Statistical Computing: Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. 91. Lüdecke, D. Esc: Effect size computation for meta analysis. Zenodo 2018, doi:10.5281/zenodo.1249218. 92. Schwarzer, G. Package “Meta.” Proc. - 2014 IEEE/ACM 7th Int. Conf. Util. Cloud Comput. UCC 2014, IEEE: London, UK; 2014. 93. Liberati, A.; Altman, D. G.; Tetzlaff, J.; Mulrow, C.; Gøtzsche, P. C.; Ioannidis, J. P. A.; Clarke, M.; Devereaux, P. J.; Kleijnen, J.; Moher, D. The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration. PLoS Med. 2009, 6, 1–34. 94. Smoski, M. J.; LaBar, K. S.; Steffens, D. C. Relative Effectiveness of Reappraisal and Distraction in Regulating Emotion in Late-Life Depression. Am. J. Geriatr. Psychiatry 2014, 22, 898–907. 95. Nelson, H. E. A Modified Card Sorting Test Sensitive to Frontal Lobe Defects. Cortex 1976, 12, 313–324. 96. Alalade, E.; Denny, K.; Potter, G.; Steffens, D.; Wang, L. Altered Cerebellar-Cerebral Functional Connectivity in Geriatric Depression. PLoS One 2011, 6, e20035. 97. Alexopoulos, G. S.; Hoptman, M. J.; Yuen, G.; Kanellopoulos, D.; Seirup, J. K.; Lim, K. O.; Gunning, F. M. Functional Connectivity in Apathy of Late-Life Depression: A Preliminary Study. J. Affect. Disord. 2013, 149, 398–405. 98. Avila, R.; Moscoso, M. A. A. A.; Ribeiz, S.; Arrais, J.; Jaluul, O.; Bottino, C. M. C. C. Influence of Education and Depressive Symptoms on Cognitive Function in the Elderly. Int. Psychogeriatrics 2009, 21, 560–567. 99. Balardin, J. B.; Vedana, G.; Ludwig, A.; Lima, D. B. de; Argimon, I.; Schneider, R.; Luz, C.; Schröder, N.; Bromberg, E. Contextual Memory and Encoding Strategies in Young and Older Adults with and without Depressive Symptoms. Aging Ment. Health 2009, 13, 313–318. 100. Baudic, S.; Tzortzis, C.; Dalla Barba, G.; Traykov, L. Executive Deficits in Elderly Patients With Major Unipolar Depression. J. Geriatr. Psychiatry Neurol. 2004, 17, 195–201. 101. Beheydt, L. L.; Schrijvers, D.; Docx, L.; Bouckaert, F.; Hulstijn, W.; Sabbe, B. Cognitive and Psychomotor Effects of Three Months of Escitalopram Treatment in Elderly Patients with Major Depressive Disorder. J. Affect. Disord. 2015, 188, 47–52. 102. Boone, K. B.; Lesser, I.; Miller, B.; Wohl, M.; Berman, N.; Lee, A.; Palmer, B. Cognitive Functioning in a Mildly to Moderately Depressed Geriatric Sample: Relationship to Chronological Age. J. Neuropsychiatry Clin. Neurosci. 1994, 6, 267–272. Behav. Sci. 2020, 10, 132 34 of 36 103. Chen, Z.; Lan, G.; Shen, X.; Pan, X.; Chen, X.; LI, J. Relationship of Changes in Cognitive and Depressive Symptoms during Antidepressant Treatment of Individuals with Geriatric Depression and Their Relationship to the APOE Epsilon 4 Allele. Shanghai Arch. Psychiatry 2013, 25, 99–106. 104. Dickinson, W. J.; Potter, G. G.; Hybels, C. F.; McQuoid, D. R.; Steffens, D. C. Change in Stress and Social Support as Predictors of Cognitive Decline in Older Adults with and without Depression. Int. J. Geriatr. Psychiatry 2011, 26, 1267–1274. 105. Egger, K.; Schocke, M.; Weiss, E.; Auffinger, S.; Esterhammer, R.; Goebel, G.; Walch, T.; Mechtcheriakov, S.; Marksteiner, J. Pattern of Brain Atrophy in Elderly Patients with Depression Revealed by Voxel-Based Morphometry. Psychiatry Res. Neuroimaging 2008, 164, 237–244. 106. Elderkin-Thompson, V.; Mintz, J.; Haroon, E.; Lavretsky, H.; Kumar, A. Executive Dysfunction and Memory in Older Patients with Major and Minor Depression. Arch. Clin. Neuropsychol. 2007, 22, 261–270. 107. Elderkin-Thompson, V.; Moody, T.; Knowlton, B.; Hellemann, G.; Kumar, A. Explicit and Implicit Memory in Late-Life Depression. Am. J. Geriatr. Psychiatry 2011, 19, 364–373. 108. Goveas, J. S.; Espeland, M. A.; Hogan, P. E.; Tindle, H. A.; Shih, R. A.; Kotchen, J. M.; Robinson, J. G.; Barnes, D. E.; Resnick, S. M. Depressive Symptoms and Longitudinal Changes in Cognition: Women’s Health Initiative Study of Cognitive Aging. J. Geriatr. Psychiatry Neurol. 2014, 27, 94–102. 109. Kindermann, S. S.; Kalayam, B.; Brown, G. G.; Burdick, K. E.; Alexopoulos, G. S. Executive Functions and P300 Latency in Elderly Depressed Patients and Control Subjects. Am. J. Geriatr. Psychiatry 2000, 8, 57–65. 110. Leal, S. L.; Noche, J. A.; Murray, E. A.; Yassa, M. A. Disruption of Amygdala–Entorhinal–Hippocampal Network in Late-life Depression. Hippocampus 2017, 27, 464–476. 111. Miebach, L.; Wolfsgruber, S.; Frommann, I.; Buckley, R.; Wagner, M. Different Cognitive Complaint Profiles in Memory Clinic and Depressive Patients. Am. J. Geriatr. Psychiatry 2018, 26, 463–475. 112. Park, J. H.; Lee, S. B.; Lee, J. J.; Yoon, J. C.; Han, J. W.; Kim, T. H.; Jeong, H.-G. G.; Newhouse, P. A.; Taylor, W. D.; Kim, J. H.; et al. Depression Plays a Moderating Role in the Cognitive Decline Associated With Changes of Brain White Matter Hyperintensities. J. Clin. Psychiatry 2018, 79, 1–7. 113. Rapp, M. A.; Dahlman, K.; Sano, M.; Grossman, H. T.; Haroutunian, V.; Gorman, J. M. Neuropsychological Differences Between Late-Onset and Recurrent Geriatric Major Depression. Am. J. Psychiatry 2005, 162, 691– 114. Rosano, C.; Perera, S.; Inzitari, M.; Newman, A. B.; Longstreth, W. T.; Studenski, S. Digit Symbol Substitution Test and Future Clinical and Subclinical Disorders of Cognition, Mobility and Mood in Older Adults. Age Ageing 2016, 45, 687–694. 115. Shimada, H.; Park, H.; Makizako, H.; Doi, T.; Lee, S.; Suzuki, T. Depressive Symptoms and Cognitive Performance in Older Adults. J. Psychiatr. Res. 2014, 57, 149–156. 116. Steffens, D. C.; Wagner, H. R.; Levy, R. M.; Horn, K. A.; Krishnan, K. R. Performance Feedback Deficit in Geriatric Depression. Biol. Psychiatry 2001, 50, 358–363. 117. Ramos, S.; Fernández García, Y.; Antón, E.; Casaponsa, A.; Duñabeitia, J. A. Does Learning a Language in the Elderly Enhance Switching Ability? J. Neurolinguistics 2017, 43, 39–48. 118. Wong, P. C. M.; Ou, J.; Pang, C. W. Y.; Zhang, L.; Tse, C. S.; Lam, L. C. W.; Antoniou, M. Language Training Leads to Global Cognitive Improvement in Older Adults: A Preliminary Study. J. Speech, Lang. Hear. Res. 2019, 62, 2411–2424. 119. Massa, E.; Köpke, B.; El Yagoubi, R. Age-Related Effect on Language Control and Executive Control in Bilingual and Monolingual Speakers: Behavioral and Electrophysiological Evidence. Neuropsychologia 2020, 138, 107336. 120. Little, D. Common European Framework of Reference for Languages. In The TESOL Encyclopedia of English Language Teaching; 2018; pp 1–6. 121. Ansaldo, A. I.; Ghazi-Saidi, L.; Adrover-Roig, D. Interference Control in Elderly Bilinguals: Appearances Can Be Misleading. J. Clin. Exp. Neuropsychol. 2015, 37, 455–470. 122. Marian, V.; Blumenfeld, H. K.; Kaushanskaya, M. The Language Experience and Proficiency Questionnaire (LEAP-Q): Assessing Language Profiles in Bilinguals and Multilinguals. J. Speech, Lang. Hear. Res. 2007, 50, 940–967. 123. Craik, F. I. M.; Bialystok, E. Planning and Task Management in Older Adults: Cooking Breakfast. Mem. Cogn., 2006, 34, 1236–1249. Behav. Sci. 2020, 10, 132 35 of 36 124. Estanga, A.; Ecay-Torres, M.; Ibañez, A.; Izagirre, A.; Villanua, J.; Garcia-Sebastian, M.; Iglesias Gaspar, M. T.; Otaegui-Arrazola, A.; Iriondo, A.; Clerigue, M.; et al. Beneficial Effect of Bilingualism on Alzheimer’s Disease CSF Biomarkers and Cognition. Neurobiol. Aging 2017, 50, 144–151. 125. COERLL University of Texas at Austin. Bilingual Language Profile: An Easy-to-Use Instrument to Assess Bilingualism (Spanish-English) sites.la.utexas.edu/bilingual (accessed 27 August 2020). 126. Fernandes, M. A.; Craik, F.; Bialystok, E.; Kreuger, S. Effects of Bilingualism, Aging, and Semantic Relatedness on Memory under Divided Attention. Can. J. Exp. Psychol. 2007, 61, 128–141. 127. Friesen, D. C.; Luo, L.; Luk, G.; Bialystok, E. Proficiency and Control in Verbal Fluency Performance across the Lifespan for Monolinguals and Bilinguals. Lang. Cogn. Neurosci. 2015, 30, 238–250. 128. Anderson, J. A. E.; Mak, L.; Keyvani Chahi, A.; Bialystok, E. The Language and Social Background Questionnaire: Assessing Degree of Bilingualism in a Diverse Population. Behav. Res. Methods 2018, 50, 250– 129. Gubarchuk, I.; Kemper, S. Effects of Aging on the Production of Russian. Discourse Process. 1997, 23, 63–82. 130. Ihle, A.; Oris, M.; Fagot, D.; Kliegel, M. The Relation of the Number of Languages Spoken to Performance in Different Cognitive Abilities in Old Age. J. Clin. Exp. Neuropsychol. 2016, 38, 1103–1114. 131. Johns, B. T.; Sheppard, C. L.; Jones, M. N.; Taler, V. The Role of Semantic Diversity in Word Recognition across Aging and Bilingualism. Front. Psychol. 2016, 7, 1–11. 132. Kousaie, S.; Sheppard, C.; Lemieux, M.; Monetta, L.; Taler, V. Executive Function and Bilingualism in Young and Older Adults. Front. Behav. Neurosci. 2014, 8, 1–12. 133. López Zunini, R. A.; Morrison, C.; Kousaie, S.; Taler, V. Task Switching and Bilingualism in Young and Older Adults: A Behavioral and Electrophysiological Investigation. Neuropsychologia 2019, 133, 107186. 134. Prehn, K.; Taud, B.; Reifergeste, J.; Clahsen, H.; Flöel, A. Neural Correlates of Grammatical Inflection in Older Native and Second-Language Speakers*. Bilingualism 2017, 21, 1–12. 135. Zhang, Z.; Xu, X.; Ni, H. Small Studies May Overestimate the Effect Sizes in Critical Care Meta-Analyses: A Meta-Epidemiological Study. Crit. Care 2013, 17, R2. 136. Ioannidis, J. P. A. Interpretation of Tests of Heterogeneity and Bias in Meta-Analysis. J. Eval. Clin. Pract. 2008, 14, 951–957. 137. Field, A. P. Meta-Analysis of Correlation Coefficients: A Monte Carlo Comparison of Fixed- and Random- Effects Methods. Psychol. Methods 2001, 6, 161–180. 138. Pfenninger, S. E.; Polz, S. Foreign Language Learning in the Third Age: A Pilot Feasibility Study on Cognitive, Socio-Affective and Linguistic Drivers and Benefits in Relation to Previous Bilingualism of the Learner. J. Eur. Second Lang. Assoc. 2018, 2, 1. 139. Bak, T. H.; Long, M. R.; Vega-Mendoza, M.; Sorace, A. Novelty, Challenge, and Practice: The Impact of Intensive Language Learning on Attentional Functions. PLoS One 2016, 11, 1–11. 140. Chu, L. W.; Chiu, K. C.; Hui, S. L.; Yu, G. K. K.; Tsui, W. J. C.; Lee, P. W. H. The Reliability and Validity of the Alzheimer’s Disease Assessment Scale Cognitive Subscale (ADAS-Cog) among the Elderly Chinese in Hong Kong. Ann. Acad. Med. Singapore, 2000, 29 (4), 474–485. 141. Almeida, O. P.; Almeida, S. A. Short Versions of the Geriatric Depression Scale: A Study of Their Validity for the Diagnosis of a Major Depressive Episode According to ICD-10 and DSM-IV. Int. J. Geriatr. Psychiatry 1999, 14, 858–865. 142. Reynolds, W. M.; Kobak, K. A. Reliability and Validity of the Hamilton Depression Inventory: A Paper- and-Pencil Version of the Hamilton Depression Rating Scale Clinical Interview. Psychol. Assess. 1995, 7, 472–483. 143. Surrain, S.; Luk, G. Describing Bilinguals: A Systematic Review of Labels and Descriptions Used in the Literature between 2005-2015. Bilingualism 2019, 1–15. 144. Landreat, M. G.; Vigneau, C. V.; Hardouin, J. B.; Bronnec, M. G.; Marais, M.; Venisse, J. L.; Jolliet, P. Can We Say That Seniors Are Addicted to Benzodiazepines. Subst. Use Misuse 2010, 45, 1988–1999. 145. Ramírez Gómez, D. Language Teaching and the Older Adult: The Significance of Experience; Channel View Books: Bristol, UK; 2016. 146. Mewborn, C. M.; Lindbergh, C. A.; Stephen Miller, L. Cognitive Interventions for Cognitively Healthy, Mildly Impaired, and Mixed Samples of Older Adults: A Systematic Review and Meta-Analysis of Randomized-Controlled Trials. Neuropsychol. Rev. 2017, 27, 403–439. 147. Li, P.; Legault, J.; Litcofsky, K. A. Neuroplasticity as a Function of Second Language Learning: Anatomical Changes in the Human Brain. Cortex 2014, 58, 301–324. Behav. Sci. 2020, 10, 132 36 of 36 148. Matsumoto, D. Exploring Third-Age Foreign Language Learning from the Well-Being Perspective: Work in Progress. Stud. Self-Access Learn. J. 2019, 10, 111–116. 149. Gullifer, J. W.; Titone, D. Characterizing the Social Diversity of Bilingualism Using Language Entropy. Bilingualism 2020, 23, 283–294. 150. Bialystok, E.; Craik, F. I. M.; Ryan, J. Executive Control in a Modified Antisaccade Task: Effects of Aging and Bilingualism. J. Exp. Psychol. Learn. Mem. Cogn. 2006, 32, 1341–1354. 151. Avorn, J.; Everitt, D. E.; Weiss, S. Increased Antidepressant Use in Patients Prescribed β-Blockers. JAMA J. Am. Med. Assoc. 1986, 255, 357–360. 152. Verbeek, D. E. P.; Van Riezen, J.; De Boer, R. A.; Van Melle, J. P.; De Jonge, P. A Review on the Putative Association Between Beta-Blockers and Depression. Heart Fail. Clin. 2011, 7, 89–99. 153. Pliatsikas, C. Understanding Structural Plasticity in the Bilingual Brain: The Dynamic Restructuring Model. Biling. Lang. Cogn. 2019, 23, 459–471. 154. Kroll, J. F.; Dussias, P. E.; Bice, K.; Perrotti, L. Bilingualism, Mind, and Brain. Annu. Rev. Linguist. 2015, 1, 377–394. 155. Bialystok, E.; Abutalebi, J.; Bak, T. H.; Burke, D. M.; Kroll, J. F. Aging in Two Languages: Implications for Public Health. Ageing Res. Rev. 2016, 27, 56–60. https://doi.org/10.1016/j.arr.2016.03.003. © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral Sciences Multidisciplinary Digital Publishing Institute http://www.deepdyve.com/lp/multidisciplinary-digital-publishing-institute/exploring-language-learning-as-a-potential-tool-against-cognitive-Bnz3LvnJZU

Loading next page...

References (172)

P. Cuijpers, A. Straten, F. Smit (2006)
Psychological treatment of late‐life depression: a meta‐analysis of randomized controlled trials
International Journal of Geriatric Psychiatry, 21
J. Park, S. Lee, J. Lee, J. Yoon, Ji Han, T. Kim, Hyun-Ghang Jeong, P. Newhouse, W. Taylor, J. Kim, J. Woo, Ki Kim (2018)
Depression Plays a Moderating Role in the Cognitive Decline Associated With Changes of Brain White Matter Hyperintensities.
The Journal of clinical psychiatry, 79 5
E. Bialystok, J. Abutalebi, T. Bak, D. Burke, J. Kroll (2016)
Aging in two languages: Implications for public health
Ageing Research Reviews, 27
J. Partington, Russell Leiter (2020)
Partington's Pathways Test.
M. Smoski, K. LaBar, D. Steffens (2014)
Relative effectiveness of reappraisal and distraction in regulating emotion in late-life depression.
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 22 9
Albert Costa, M. Hernández, Jordi Costa-Faidella, N. Sebastián-Gallés (2009)
On the bilingual advantage in conflict processing: Now you see it, now you don’t
Cognition, 113
L. Covey, A. Glassman, F. Stetner (1998)
Cigarette smoking and major depression.
Journal of addictive diseases, 17 1
I. Gubarchuk, S. Kemper (1997)
Effects of aging on the production of Russian
Discourse Processes, 23
D. Blazer (2003)
Depression in late life: review and commentary.
The journals of gerontology. Series A, Biological sciences and medical sciences, 58 3
J. Avorn, D. Everitt, S. Weiss (1986)
Increased Antidepressant Use in Patients Prescribed β-Blockers
JAMA, 255
D. Lüdecke (2019)
Effect Size Computation for Meta Analysis [R package esc version 0.5.1]
A. Ihle, M. Oris, D. Fagot, M. Kliegel (2016)
The relation of the number of languages spoken to performance in different cognitive abilities in old age
Journal of Clinical and Experimental Neuropsychology, 38
高橋雅春, 久本博行 (1984)
WAIS〔Wechsler Adult Intelligence Scale〕類似問題の二次元構造と採点基準
, 15
A. Ansaldo, Ladan Ghazi-Saidi, Daniel Adrover-Roig (2015)
Interference Control In Elderly Bilinguals: Appearances Can Be Misleading
Journal of Clinical and Experimental Neuropsychology, 37
L. Kurlowicz (1993)
Social factors and depression in late life.
Archives of psychiatric nursing, 7 1
D. Drayer (1987)
Lipophilicity, Hydrophilicity, and the Central Nervous System Side Effects of Beta Blockers
Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy, 7
A. Mitchell, Sanjay Rao, A. Vaze (2010)
Do Primary Care Physicians Have Particular Difficulty Identifying Late-Life Depression? A Meta-Analysis Stratified by Age
Psychotherapy and Psychosomatics, 79
Wentao Bao, Xiao Xu, H. Ni (2013)
Small studies may overestimate the effect sizes in critical care meta-analyses: a meta-epidemiological study
Critical Care, 17
F. Craik, E. Bialystok, M. Freedman (2010)
Delaying the onset of Alzheimer disease
Neurology, 75
(2020)
National Heart, Lung, and Blood Institute
Definitions
W. Dickinson, G. Potter, C. Hybels, D. McQuoid, D. Steffens (2011)
Change in stress and social support as predictors of cognitive decline in older adults with and without depression
International Journal of Geriatric Psychiatry, 26
Y. Conwell, Kimberly Orden, E. Caine (2011)
Suicide in older adults.
The Psychiatric clinics of North America, 34 2
A. Cohen (1976)
The Redwood City Bilingual Education Project, 1971-1974: Spanish and English Proficiency, Mathematics and Language Use Over Time. Working Papers on Bilingualism, No. 8.
(2020)
2020 Alzheimer's disease facts and figures
Alzheimer's & Dementia, 16
(2013)
10.1176/appi.books.9780890425596.893619
Diagnostic and Statistical Manual of Mental Disorders 1
Minna Lehtonen, Anna Soveri, Aini Laine, Janica Järvenpää, Angela Bruin, J. Antfolk (2018)
Is Bilingualism Associated With Enhanced Executive Functioning in Adults? A Meta-Analytic Review
Psychological Bulletin, 144
Julie Hugo, M. Ganguli (2014)
Dementia and cognitive impairment: epidemiology, diagnosis, and treatment.
Clinics in geriatric medicine, 30 3
S. Baudic, C. Tzortzis, G. Barba, Latchezar Traykov (2004)
Executive Deficits in Elderly Patients With Major Unipolar Depression
Journal of Geriatric Psychiatry and Neurology, 17
Judith Hays, L. Landerman, Linda George, Elizabeth Flint, Harold Koenig, Kenneth Land, D. Blazer (1998)
Social correlates of the dimensions of depression in the elderly.
The journals of gerontology. Series B, Psychological sciences and social sciences, 53 1
E. Nichols, C. Wild, B. Stojanoski, Michael Battista, A. Owen (2020)
Bilingualism Affords No General Cognitive Advantages: A Population Study of Executive Function in 11,000 People
Psychological Science, 31
B. Lue, Liang-Ju Chen, Shwu-chong Wu (2010)
Health, financial stresses, and life satisfaction affecting late-life depression among older adults: a nationwide, longitudinal survey in Taiwan.
Archives of gerontology and geriatrics, 50 Suppl 1
Brendan Johns, Christine Sheppard, Michael Jones, V. Taler (2016)
The Role of Semantic Diversity in Word Recognition across Aging and Bilingualism
Frontiers in Psychology, 7
(2009)
The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration
PLoS Med., 6
Catherine Mewborn, C. Lindbergh, L. Miller (2017)
Cognitive Interventions for Cognitively Healthy, Mildly Impaired, and Mixed Samples of Older Adults: A Systematic Review and Meta-Analysis of Randomized-Controlled Trials
Neuropsychology Review, 27
Z. Nasreddine, N. Phillips, Valérie Bédirian, S. Charbonneau, V. Whitehead, Isabelle Collin, J. Cummings, H. Chertkow (2005)
The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive Impairment
Journal of the American Geriatrics Society, 53
John Anderson, Lorinda Mak, Aram Chahi, E. Bialystok (2018)
The language and social background questionnaire: Assessing degree of bilingualism in a diverse population
Behavior Research Methods, 50
(2009)
Cognitive reserve
Neuropsychologia, 47
Filipi Dias, A. Teixeira, Henrique Guimarães, Ana Santos, Simone Ritter, João Machado, Maira Barbosa, P. Caramelli (2019)
Prevalence of late-life depression and its correlates in a community-dwelling low-educated population aged 75+ years: The Pietà study.
Journal of affective disorders, 242
Anat Prior, B. MacWhinney (2010)
A bilingual advantage in task switching.
Bilingualism, 13 2
S. Naismith, N. Rogers, S. Lewis, Z. Terpening, T. Ip, Keri Diamond, L. Norrie, I. Hickie (2011)
Sleep disturbance relates to neuropsychological functioning in late-life depression.
Journal of affective disorders, 132 1-2
Viorica Marian, H. Blumenfeld, Margarita Kaushanskaya (2007)
The Language Experience and Proficiency Questionnaire (LEAP-Q): assessing language profiles in bilinguals and multilinguals.
Journal of speech, language, and hearing research : JSLHR, 50 4
Deborah Wagenaar, M. Mickus, Kristin Gaumer, C. Colenda (2002)
Late-Life Depression and Mental Health Services in Primary Care
Journal of Geriatric Psychiatry and Neurology, 15
(2017)
Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies
S. Pfenninger, S. Polz (2018)
Foreign language learning in the third age: A pilot feasibility study on cognitive, socio-affective and linguistic drivers and benefits in relation to previous bilingualism of the learner
Journal of the European Second Language Association
(2007)
Executive Dysfunction and Memory in Older Patients with Major and Minor Depression
Arch. Clin. Neuropsychol., 22
J. Gullifer, D. Titone (2019)
Characterizing the social diversity of bilingualism using language entropy
Bilingualism: Language and Cognition, 23
J. Ioannidis (2008)
Interpretation of tests of heterogeneity and bias in meta-analysis.
Journal of evaluation in clinical practice, 14 5
L. Solai, B. Mulsant, B. Pollock (2001)
Selective Serotonin Reuptake Inhibitors for Late-Life Depression
Drugs & Aging, 18
L. Chu, Patrick Chiu, W. Tsui, S. Hui, Peter Lee (2000)
The reliability and validity of the Alzheimer's diseases assessment scale cognitive subscale (ADAS-COG) among chinese elderly people in Hong Kong
Neurobiology of Aging, 21
S. Kousaie, Christine Sheppard, Maude Lemieux, L. Monetta, V. Taler (2014)
Executive function and bilingualism in young and older adults
Frontiers in Behavioral Neuroscience, 8
Bilingual Language Profile: An Easy-to-Use Instrument to Assess Bilingualism (Spanish-English)
R. Zunini, C. Morrison, S. Kousaie, V. Taler (2019)
Task switching and bilingualism in young and older adults: A behavioral and electrophysiological investigation
Neuropsychologia, 133
K. Egger, M. Schocke, E. Weiss, Simone Auffinger, R. Esterhammer, G. Goebel, T. Walch, S. Mechtcheriakov, J. Marksteiner (2008)
Pattern of brain atrophy in elderly patients with depression revealed by voxel-based morphometry
Psychiatry Research: Neuroimaging, 164
Evy Woumans, P. Santens, A. Sieben, J. Versijpt, Michaël Stevens, W. Duyck (2015)
Bilingualism delays clinical manifestation of Alzheimer's disease
Bilingualism: Language and Cognition, 18
O. Kangas (2010)
One hundred years of money, welfare and death: mortality, economic growth and the development of the welfare state in 17 OECD countries 1900–2000
International Journal of Social Welfare, 19
M. Laine, Minna Lehtonen (2018)
Cognitive consequences of bilingualism: where to go from here?
Language, Cognition and Neuroscience, 33
Michael Ramscar, Peter Hendrix, B. Love, R. Baayen (2014)
Learning is not decline: The mental lexicon as a window into cognition across the lifespan
The Mental Lexicon, 8
D. Steffens, H. Wagner, R. Levy, K. Horn, K. Krishnan (2001)
Performance feedback deficit in geriatric depression
Biological Psychiatry, 50
E. Bohlmeijer, F. Smit, P. Cuijpers (2003)
Effects of reminiscence and life review on late‐life depression: a meta‐analysis
International Journal of Geriatric Psychiatry, 18
M. Antoniou, Geshri Gunasekera, P. Wong (2013)
Foreign language training as cognitive therapy for age-related cognitive decline: A hypothesis for future research
Neuroscience & Biobehavioral Reviews, 37
Bilingual Language Profile: An Easy-to-Use Instrument to Assess Bilingualism
A. Field (2001)
Meta-analysis of correlation coefficients: a Monte Carlo comparison of fixed- and random-effects methods.
Psychological methods, 6 2
C. Rosano, S. Perera, M. Inzitari, A. Newman, W. Longstreth, S. Studenski (2016)
Digit Symbol Substitution test and future clinical and subclinical disorders of cognition, mobility and mood in older adults.
Age and ageing, 45 5
Shun-Chiao Chang, A. Pan, I. Kawachi, O. Okereke (2016)
Risk factors for late-life depression: A prospective cohort study among older women.
Preventive medicine, 91
M. Rapp, K. Dahlman, M. Sano, H. Grossman, V. Haroutunian, J. Gorman (2005)
Neuropsychological differences between late-onset and recurrent geriatric major depression.
The American journal of psychiatry, 162 4
Y. Stern, Daniel Barulli (2019)
Cognitive reserve.
Handbook of clinical neurology, 167
F. Lamers, P. Oppen, H. Comijs, J. Smit, P. Spinhoven, A. Balkom, W. Nolen, F. Zitman, A. Beekman, B. Penninx (2011)
Comorbidity patterns of anxiety and depressive disorders in a large cohort study: the Netherlands Study of Depression and Anxiety (NESDA).
The Journal of clinical psychiatry, 72 3
T. Bak, Madeleine Long, M. Vega-Mendoza, A. Sorace (2016)
Novelty, Challenge, and Practice: The Impact of Intensive Language Learning on Attentional Functions
PLoS ONE, 11
Quality Assessment of Controlled Intervention Studies-NHLBI, NIH
Sarah Surrain, G. Luk (2017)
Describing bilinguals: A systematic review of labels and descriptions used in the literature between 2005–2015
Bilingualism: Language and Cognition, 22
L. Hebert, J. Weuve, P. Scherr, Denis Evans (2013)
Alzheimer disease in the United States (2010–2050) estimated using the 2010 census
Neurology, 80
Deanna Friesen, L. Luo, G. Luk, E. Bialystok (2015)
Proficiency and control in verbal fluency performance across the lifespan for monolinguals and bilinguals
Language, Cognition and Neuroscience, 30
M. Formosa (2012)
Education and Older Adults at the University of the Third Age
Educational Gerontology, 38
E. Bialystok, F. Craik (2010)
Cognitive and Linguistic Processing in the Bilingual Mind
Current Directions in Psychological Science, 19
W. Appel (1986)
Antidepressant Use in Patients Prescribed β-Blockers
JAMA, 255
M. Pinto (2017)
Language Teaching and the Older Adult. The Significance of Experience
Frédéric Brière, P. Rohde, J. Seeley, D. Klein, P. Lewinsohn (2014)
Comorbidity between major depression and alcohol use disorder from adolescence to adulthood.
Comprehensive psychiatry, 55 3
A. Koenig, Rishi Bhalla, M. Butters (2014)
Cognitive Functioning and Late-Life Depression
Journal of the International Neuropsychological Society, 20
R. Team (2014)
R: A language and environment for statistical computing.
MSOR connections, 1
A. Estanga, M. Ecay-Torres, A. Ibañez, A. Izagirre, J. Villanúa, M. García-Sebastián, M. Gaspar, A. Otaegui-Arrazola, A. Iriondo, Monserrat Clerigue, P. Martínez-Lage (2017)
Beneficial effect of bilingualism on Alzheimer's disease CSF biomarkers and cognition
Neurobiology of Aging, 50
V. Elderkin‐Thompson, J. Mintz, E. Haroon, H. Lavretsky, Anand Kumar (2006)
Executive dysfunction and memory in older patients with major and minor depression.
Archives of clinical neuropsychology : the official journal of the National Academy of Neuropsychologists, 22 2
S. Morimoto, B. Wexler, Jiacheng Liu, Willie Hu, Joanna Seirup, G. Alexopoulos (2014)
Neuroplasticity-based computerized cognitive remediation for treatment resistant geriatric depression
Nature communications, 5
(2020)
of Texas at Austin
K. Olfers, G. Band, L. Colzato, B. Hommel (2017)
The Case of the Leiden “Lifestyle 2030” Study
S. Alladi, T. Bak, V. Duggirala, B. Surampudi, M. Shailaja, A. Shukla, J. Chaudhuri, S. Kaul (2013)
Bilingualism delays age at onset of dementia, independent of education and immigration status
Neurology, 81
A. Seitz (2017)
A New Framework of Design and Continuous Evaluation to Improve Brain Training
Journal of Cognitive Enhancement, 2
J. Djernes (2006)
Prevalence and predictors of depression in populations of elderly: a review
Acta Psychiatrica Scandinavica, 113
Margarita Kaushanskaya, Viorica Marian (2009)
The bilingual advantage in novel word learning
Psychonomic Bulletin & Review, 16
Myra Fernandes, F. Craik, E. Bialystok, Sharyn Kreuger (2007)
Effects of bilingualism, aging, and semantic relatedness on memory under divided attention.
Canadian journal of experimental psychology = Revue canadienne de psychologie experimentale, 61 2
Mara Ploeg, Merel Keijzer, W. Lowie (2020)
Methodological concerns and their solutions in third-age language learning studies
J. Anguera, F. Gunning, P. Areán (2017)
Improving late life depression and cognitive control through the use of therapeutic video game technology: A proof‐of‐concept randomized trial
Depression and Anxiety, 34
E. Tedeschini, Y. Levkovitz, N. Iovieno, Victoria Ameral, J. Nelson, G. Papakostas (2011)
Efficacy of antidepressants for late-life depression: a meta-analysis and meta-regression of placebo-controlled randomized trials.
The Journal of clinical psychiatry, 72 12
E. Crocco, K. Castro, D. Loewenstein (2010)
How Late-Life Depression Affects Cognition: Neural Mechanisms
Current Psychiatry Reports, 12
Rishi Bhalla, M. Butters, B. Mulsant, A. Begley, M. Zmuda, Beth Schoderbek, B. Pollock, C. Reynolds, J. Becker (2006)
Persistence of neuropsychologic deficits in the remitted state of late-life depression.
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 14 5
B. Diniz, M. Butters, S. Albert, M. Dew, C. Reynolds (2013)
Late-life depression and risk of vascular dementia and Alzheimer's disease: systematic review and meta-analysis of community-based cohort studies
British Journal of Psychiatry, 202
M. Folstein, M. Folstein, S. Folstein, S. Folstein, P. McHugh, P. McHugh (1975)
“Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician
Journal of Psychiatric Research, 12
Main Report Cognitive Reserve: Implications for Assessment and Intervention
Mahesh Patel (1989)
An introduction to meta-analysis.
Health policy, 11 1
C. Pliatsikas (2019)
Understanding structural plasticity in the bilingual brain: The Dynamic Restructuring Model
Bilingualism: Language and Cognition, 23
Joana Balardin, Gustavo Vedana, Andreine Ludwig, Daiane Lima, Irani Argimon, R. Schneider, C. Luz, N. Schröder, E. Bromberg (2009)
Contextual memory and encoding strategies in young and older adults with and without depressive symptoms
Aging & Mental Health, 13
L. Hinton, Mark Zweifach, S. Oishi, Lingqi Tang, J. Unützer (2006)
Gender disparities in the treatment of late-life depression: qualitative and quantitative findings from the IMPACT trial.
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 14 10
Gregory Poarch, Janet Hell (2012)
Cross-language activation in children's speech production: evidence from second language learners, bilinguals, and trilinguals.
Journal of experimental child psychology, 111 3
V. Elderkin‐Thompson, T. Moody, B. Knowlton, G. Hellemann, Anand Kumar (2011)
Explicit and implicit memory in late-life depression.
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 19 4
(2018)
Esc: Effect size computation for meta analysis
Zenodo
E. Bialystok, F. Craik, Jennifer Ryan (2006)
Executive control in a modified antisaccade task: Effects of aging and bilingualism.
Journal of experimental psychology. Learning, memory, and cognition, 32 6
O. Riedel, J. Klotsche, A. Spottke, G. Deuschl, Hans Förstl, F. Henn, I. Heuser, W. Oertel, H. Reichmann, P. Riederer, C. Trenkwalder, R. Dodel, H. Wittchen (2008)
Cognitive impairment in 873 patients with idiopathic Parkinson's disease
Journal of Neurology, 255
H. Shimada, Hyuntae Park, H. Makizako, T. Doi, Sangyoon Lee, Takao Suzuki (2014)
Depressive symptoms and cognitive performance in older adults.
Journal of psychiatric research, 57
Alex Gertner, M. Domino, W. Dow (2017)
Risk factors for late-life depression and correlates of antidepressant use in Costa Rica: Results from a nationally-representative longitudinal survey of older adults.
Journal of affective disorders, 208
T. Schweizer, J. Ware, Corinne Fischer, F. Craik, E. Bialystok (2012)
Bilingualism as a contributor to cognitive reserve: Evidence from brain atrophy in Alzheimer’s disease
Cortex, 48
M-G Landreat, C. Vigneau, J. Hardouin, M. Bronnec, M. Marais, J. Venisse, P. Jolliet (2010)
Can We Say that Seniors Are Addicted to Benzodiazepines?
Substance Use & Misuse, 45
E. Bialystok, F. Craik, M. Freedman (2007)
Bilingualism as a protection against the onset of symptoms of dementia
Neuropsychologia, 45
W. Meerding, L. Bonneux, J. Polder, M. Koopmanschap, P. Maas (1998)
Demographic and epidemiological determinants of healthcare costs in Netherlands: cost of illness study
BMJ, 317
Bilingualism Research: Visions For The Future
P. Hallett (1978)
Primary and secondary saccades to goals defined by instructions
Vision Research, 18
Sascha Meyer, J. Jonghe, B. Schmand, R. Ponds (2017)
The Visual Association Test-Extended: a cross-sectional study of the performance validity measures
The Clinical Neuropsychologist, 31
G. Alexopoulos, M. Hoptman, Genevieve Yuen, D. Kanellopoulos, Joanna Seirup, K. Lim, F. Gunning (2013)
Functional connectivity in apathy of late-life depression: a preliminary study.
Journal of affective disorders, 149 1-3
Lisa Miebach, S. Wolfsgruber, I. Frommann, R. Buckley, M. Wagner (2017)
Different Cognitive Complaint Profiles in Memory Clinic and Depressive Patients.
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 26 4
L. Friedman (2001)
Why vote-count reviews don’t count
Biological Psychiatry, 49
R. Morin, M. Gonzales, David Bickford, Daniel Catalinotto, C. Nelson, R. Mackin (2019)
Impaired Financial Capacity in Late-Life Depression: Revisiting Associations with Cognitive Functioning
Journal of the International Neuropsychological Society, 25
K. Boone, I. Lesser, B. Miller, M. Wohl, N. Berman, A. Lee, B. Palmer (1994)
Cognitive functioning in a mildly to moderately depressed geriatric sample: relationship to chronological age.
The Journal of neuropsychiatry and clinical neurosciences, 6 3
Jae-Min Kim, R. Stewart, I. Shin, Jin-Sang Yoon, Hyung-Yung Lee (2004)
Lifetime urban/rural residence, social support and late‐life depression in Korea
International Journal of Geriatric Psychiatry, 19
Anna Rajtar-Zembaty, A. Sałakowski, Jakub Rajtar-Zembaty, A. Starowicz-Filip (2017)
Executive dysfunction in late-life depression.
Psychiatria polska, 51 4
M. Rapaport, L. Schneider, D. Dunner, J. Davies, C. Pitts (2003)
Efficacy of controlled-release paroxetine in the treatment of late-life depression.
The Journal of clinical psychiatry, 64 9
M. Guerra, A. Prina, C. Ferri, D. Acosta, Sara Gallardo, Yu Huang, K. Jacob, I. Jimenez-Velazquez, J. Rodríguez, Zhaorui Liu, A. Salas, A. Sosa, Joseph Williams, R. Uwakwe, M. Prince (2016)
A comparative cross-cultural study of the prevalence of late life depression in low and middle income countries
Journal of Affective Disorders, 190
D. Grant, Esta Berg (1948)
A behavioral analysis of degree of reinforcement and ease of shifting to new responses in a Weigl-type card-sorting problem.
Journal of experimental psychology, 38 4
K. Prehn, Benedikt Taud, Jana Reifegerste, H. Clahsen, A. Flöel (2017)
Neural correlates of grammatical inflection in older native and second-language speakers*
Bilingualism: Language and Cognition, 21
Alexandra Kueider, Jeanine Parisi, A. Gross, G. Rebok (2012)
Computerized Cognitive Training with Older Adults: A Systematic Review
PLoS ONE, 7
M. Zanetti, Claudia Ballabio, C. Abbate, C. Cutaia, C. Vergani, L. Bergamaschini (2006)
Mild Cognitive Impairment Subtypes and Vascular Dementia in Community‐Dwelling Elderly People: A 3‐Year Follow‐Up Study
Journal of the American Geriatrics Society, 54
P. Wong, J. Ou, C. Pang, Ling Zhang, C. Tse, L. Lam, M. Antoniou (2019)
Language Training Leads to Global Cognitive Improvement in Older Adults: A Preliminary Study.
Journal of speech, language, and hearing research : JSLHR
Stephanie Leal, Jessica Noche, Elizabeth Murray, M. Yassa (2017)
Disruption of amygdala–entorhinal–hippocampal network in late‐life depression
Hippocampus, 27
L. Chu, Chiu Kc, Hui Sl, Yu Gk, Tsui Wj, Lee Pw (2000)
The reliability and validity of the Alzheimer's Disease Assessment Scale Cognitive Subscale (ADAS-Cog) among the elderly Chinese in Hong Kong.
Annals of the Academy of Medicine, Singapore, 29 4
W. Liao, Xiangrong Zhang, H. Shu, Zan Wang, Duan Liu, Zhijun Zhang (2017)
The characteristic of cognitive dysfunction in remitted late life depression and amnestic mild cognitive impairment
Psychiatry Research, 251
Emmanuel Alalade, K. Denny, G. Potter, D. Steffens, Lihong Wang (2011)
Altered Cerebellar-Cerebral Functional Connectivity in Geriatric Depression
PLoS ONE, 6
(2009)
Depression in Late Life: Review and Commentary
Focus (Madison), 7
(2014)
Package “Meta
C. Bowie, Philip Harvey (2006)
Administration and interpretation of the Trail Making Test
Nature Protocols, 1
H. Nelson (1976)
A Modified Card Sorting Test Sensitive to Frontal Lobe Defects
Cortex, 12
Emilie Massa, B. Köpke, R. Yagoubi (2020)
Age-related effect on language control and executive control in bilingual and monolingual speakers: Behavioral and electrophysiological evidence
Neuropsychologia, 138
J. Kroll, Paola Dussias, Kinsey Bice, L. Perrotti (2015)
Bilingualism, Mind, and Brain.
Annual review of linguistics, 1
F. Panza, V. Frisardi, C. Capurso, A. D'Introno, A. Colacicco, B. Imbimbo, A. Santamato, G. Vendemiale, D. Seripa, A. Pilotto, A. Capurso, V. Solfrizzi (2010)
Late-life depression, mild cognitive impairment, and dementia: possible continuum?
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 18 2
O. Almeida, S. Almeida (1999)
Short versions of the geriatric depression scale: a study of their validity for the diagnosis of a major depressive episode according to ICD‐10 and DSM‐IV
International Journal of Geriatric Psychiatry, 14
A. Liberati, D. Altman, J. Tetzlaff, C. Mulrow, P. Gøtzsche, J. Ioannidis, M. Clarke, P. Devereaux, J. Kleijnen, D. Moher (2009)
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration
The BMJ, 339
B. Pollock (2005)
The Pharmacokinetic Imperative in Late-Life Depression
Journal of Clinical Psychopharmacology, 25
D. Matsumoto (2019)
Exploring Third-Age Foreign Language Learning from the Well-being Perspective: Work in Progress
Studies in Self-Access Learning Journal
D. Blazer, C. Hybels (2005)
Origins of depression in later life
Psychological Medicine, 35
J. Goveas, M. Espeland, P. Hogan, H. Tindle, R. Shih, J. Kotchen, Jennifer Robinson, D. Barnes, S. Resnick (2014)
Depressive Symptoms and Longitudinal Changes in Cognition
Journal of Geriatric Psychiatry and Neurology, 27
M. Barker, K. Greenwood, M. Jackson, S. Crowe (2004)
Cognitive Effects of Long-Term Benzodiazepine Use
CNS Drugs, 18
L. Schneider, J. Nelson, C. Clary, P. Newhouse, K. Krishnan, T. Shiovitz, K. Weihs (2003)
An 8-week multicenter, parallel-group, double-blind, placebo-controlled study of sertraline in elderly outpatients with major depression.
The American journal of psychiatry, 160 7
W. Reynolds, K. Kobak (1995)
Reliability and Validity of the Hamilton Depression Inventory: A Paper-and-Pencil Version of the Hamilton Depression Rating Scale Clinical Interview
Psychological Assessment, 7
N. Kist, J. Sandjojo, R. Kok, J. Berg (2014)
Cognitive functioning in older adults with early, late, and very late onset alcohol dependence
International Psychogeriatrics, 26
J. Motter, J. Motter, Monique Pimontel, Monique Pimontel, D. Rindskopf, D. Devanand, P. Doraiswamy, J. Sneed, J. Sneed (2016)
Computerized cognitive training and functional recovery in major depressive disorder: A meta-analysis.
Journal of affective disorders, 189
(2020)
Diagnostic and Statistical Manual of Mental Disorders, 5th Edition
Definitions
Sara Ramos, Y. García, Eneko Antón, Aina Casaponsa, J. Duñabeitia (2017)
Does learning a language in the elderly enhance switching ability?
Journal of Neurolinguistics, 43
Jae-Min Kim, I. Shin, Jin-Sang Yoon, R. Stewart (2002)
Prevalence and correlates of late‐life depression compared between urban and rural populations in Korea
International Journal of Geriatric Psychiatry, 17
(2020)
Bilingualism Research: Visions For The Future. Available online: https://www.youtube.com/ watch?v=Asyf1my3RP4&t=14m26 (accessed on
猫田英伸, 猫田和明 (2002)
Common European Framework of Reference for Languagesの意義を考える : 日本の英語教育関係者の連携のために
, 13
M. Kelly, David Loughrey, B. Lawlor, I. Robertson, C. Walsh, S. Brennan (2014)
The impact of cognitive training and mental stimulation on cognitive and everyday functioning of healthy older adults: A systematic review and meta-analysis
Ageing Research Reviews, 15
S. Kindermann, B. Kalayam, Gregory Brown, Katherine Burdick, George Alexopoulos (2000)
Executive functions and P300 latency in elderly depressed patients and control subjects.
The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry, 8 1
D. Verbeek, Jerry Riezen, R. Boer, J. Melle, P. Jonge (2011)
A review on the putative association between beta-blockers and depression.
Heart failure clinics, 7 1
J. Gaugler, B. James, T. Johnson, K. Scholz, J. Weuve (2016)
2016 Alzheimer's disease facts and figures
Alzheimer's & Dementia, 12
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license
F. Craik, E. Bialystok (2006)
Planning and task management in older adults: Cooking breakfast
Memory & Cognition, 34
V. Buxton (1996)
Visions for the future.
Nursing times, 92 25
James Combs, David Jr, T. Crook, P. Roth (2010)
Assessing Cumulative Evidence within Macro Research: Why Meta-Analysis Should Be Preferred Over Vote Counting
SPGMI: Compustat Fundamentals (Topic)
A. Koenig, Isaac DeLozier, M. Zmuda, M. Marron, A. Begley, S. Anderson, C. Reynolds, S. Arnold, J. Becker, M. Butters (2015)
Neuropsychological functioning in the acute and remitted States of late-life depression.
Journal of Alzheimer's disease : JAD, 45 1
R. Ávila, M. Moscoso, S. Ribeiz, J. Arrais, O. Jaluul, C. Bottino (2009)
Influence of education and depressive symptoms on cognitive function in the elderly
International Psychogeriatrics, 21
N. Scarmeas, Y. Stern (2003)
Cognitive Reserve and Lifestyle
Journal of Clinical and Experimental Neuropsychology, 25
Zheli Chen, Guanghua Lan, Xin-hua Shen, X. Pan, Xiaoxun Chen, Jianhua Li (2013)
Relationship of changes in cognitive and depressive symptoms during antidepressant treatment of individuals with geriatric depression and their relationship to the APOE epsilon 4 allele
Shanghai Archives of Psychiatry, 25
L. Beheydt, D. Schrijvers, L. Docx, F. Bouckaert, W. Hulstijn, B. Sabbe (2015)
Cognitive and psychomotor effects of three months of escitalopram treatment in elderly patients with major depressive disorder.
Journal of affective disorders, 188
Ping Li, Jennifer Legault, Kaitlyn Litcofsky (2014)
Neuroplasticity as a function of second language learning: Anatomical changes in the human brain
Cortex, 58
Michael Ramscar, Peter Hendrix, Cyrus Shaoul, P. Milin, Harald Baayen (2014)
The Myth of Cognitive Decline: Non-Linear Dynamics of Lifelong Learning
Topics in cognitive science, 6 1
M. Butters, Jeffrey Young, O. Lopez, H. Aizenstein, B. Mulsant, C. III, S. DeKosky, J. Becker (2008)
Pathways linking late-life depression to persistent cognitive impairment and dementia
Dialogues in Clinical Neuroscience, 10

Publisher: Multidisciplinary Digital Publishing Institute
Copyright: © 1996-2020 MDPI (Basel, Switzerland) unless otherwise stated Disclaimer The statements, opinions and data contained in the journals are solely those of the individual authors and contributors and not of the publisher and the editor(s). Terms and Conditions Privacy Policy
ISSN: 2076-328X
DOI: 10.3390/bs10090132
Publisher site: See Article on Publisher Site

Abstract

Review Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research 1, 1 1 2 Jelle Brouwer *, Floor van den Berg , Remco Knooihuizen , Hanneke Loerts and Merel Keijzer Department of Linguistics & English as a Second Language, University of Groningen, 9712 EK Groningen, The Netherlands; f.a.van.den.berg@rug.nl (F.v.d.B.); r.m.knooihuizen@rug.nl (R.K.); m.c.j.keijzer@rug.nl (M.K.) Department of Minorities and Multilingualism, University of Groningen, 9712 EK Groningen, The Netherlands; h.loerts@rug.nl * Correspondence: jelle.brouwer@rug.nl Received: 31 May 2020; Accepted: 26 August 2020; Published: 31 August 2020 Abstract: Late-life depression (LLD) affects about an eighth of community-dwelling seniors. LLD impacts well-being, with loneliness and small social networks being typical. It has also been linked to cognitive dysfunction and an increased risk of developing dementia. Safety and efficacy of pharmacological treatments for LLD have been debated, and cognitive dysfunction often persists even after remission. Various cognitive interventions have been proposed for LLD. Among these, one has received special attention: foreign language learning could serve as a social intervention that simultaneously targets brain structures affected in LLD. Lifelong bilingualism may significantly delay the onset of cognitive impairment symptoms by boosting cognitive reserve. Even late-life foreign language learning without lifelong bilingualism can train cognitive flexibility. It is then counterintuitive that the effects of language learning on LLD have never been examined. In order to create a theoretical basis for further interdisciplinary research, this paper presents a status quo of current work through two meta-analyses investigating cognitive functioning in LLD on the one hand and in senior bilinguals or seniors following a language course on the other hand. While LLD was consistently associated with cognitive dysfunction, inconsistent results were found for bilingualism and language learners. Possible reasons for this and suggestions for future research are subsequently discussed. Keywords: late-life depression; language learning; bilingualism; seniors; cognitive function; meta- analysis; systematic review 1. Introduction The world’s population is aging, both in relative and absolute numbers. Since the early 1900s, the mean life expectancy has increased by more than 30 years in some countries [1]. This increased longevity combined with declining birth-rates and a so-called ‘baby boom’ following World War II means that the proportion of those aged 65 and older will increase significantly in the coming years. This relative aging of the population is accompanied by many societal challenges. After the age of 50, for instance, the cost of healthcare grows exponentially every year [2]. Apart from the increased financial burden, there are also challenges on a personal level: a number of age-specific illnesses reduce cognitive functioning and quality of life. Most pertinently, neurodegenerative illnesses such as Alzheimer’s Disease (AD) are predicted to take over as a primary cause of death in developed Behav. Sci. 2020, 10, 132; doi:10.3390/bs10090132 www.mdpi.com/journal/behavsci Behav. Sci. 2020, 10, 132 2 of 36 countries [3]. The number of AD patients in the United States, for instance, is estimated to grow from 4.7 million in 2010 to 13.8 million in 2050, affecting half the population over 65 [4]. However, while research pertaining to AD is undeniably important, it overshadows a lesser-known threat to healthy aging: late-life depression (LLD). Falling in the realm of major depression disorder, LLD can be diagnosed on the basis of symptoms stipulated in the Diagnostic and Statistical Manual (DSM) [5]. LLD includes both early-onset depressions that persist or recurs after the age of 60 as well as the first diagnosis of depression after age 60 [6]. A diagnosis of depression can follow if an individual currently displays at least one of the two core symptoms: (1) depressed mood or (2) loss of interest/pleasure; and at least four of the following symptoms during at least a two-week period of time: (3) unintended large changes in weight (+/− 5% in a month), (4) insomnia or hypersomnia, (5) psychomotor agitation or retardation, (6) fatigue or loss of energy, (7) feelings of worthlessness or inappropriate guilt, (8) diminished ability to think or concentrate, or indecisiveness, and (9) recurrent thoughts of death or suicidal ideation [5]. Despite limited public awareness, LLD is extremely common: in high-income countries, major depressive disorder was found in 0.9–9.4% of community-dwelling seniors and in 14–42% of those living in care homes; minor depression (the presence of one core symptom and one to three of symptoms 3–9 mentioned above) was found in up to 49% of older adults [7]. Comparable numbers have been attested in low to middle-income countries [8]. Despite a high prevalence of LLD in large-scale epidemiological studies, awareness appears poor, and the illness is often difficult to diagnose. Part of the reason why LLD is so underdiagnosed can be traced back to its symptomatology. Many of the symptoms are somatic in nature, commonly leading older patients and/or their caretakers to mistake these signs for physical illness [9]. A number of studies also suggest that in some LLD patients, especially men, low mood is not necessarily present, further complicating diagnosis [6,10]. Patients suffering from LLD may also feel less inclined to see their general practitioner (GP) for experienced mental health issues, due to stigmas attached to psychiatric disorders [9,10]. Finally, the underrepresentation of LLD may also be associated with healthcare providers. A meta-analysis by Mitchell et al. suggested that GPs often do not take into account the high prevalence of depression in seniors [11]. The study found, for instance, that conversations regarding mental health were quite rare, and often not initiated by the GPs themselves. The impact of LLD on mental health and cognition is all-encompassing. Not only are mood disorders present in up to 87% of older adults who die by suicide [12], but nonsuicide mortality is also higher in these populations [6]. This can be partially explained through higher incidences of unhealthy behaviors such as smoking and excessive alcohol use among this group [13,14], but also potentially through the lack of social contacts [15]. Additionally, there is a strong and consistent finding that LLD patients suffer from cognitive dysfunction [16,17]. Related to this, LLD significantly increases a person’s risk of developing AD or other types of dementia [18,19]. This consistent association between LLD and cognitive dysfunction has sparked theories that depression is a prodromal form of dementia or that LLD and dementia are underpinned by a similar pathology [20]. An influential review paper by Butters et al. identified a number of possible associations between LLD and AD [21]. One theory they postulated was that depression-related disturbances in glucocorticoids (hormones associated with stress-response, among other processes) lead to hippocampal atrophy, which may increase the risk of developing AD. However, Butters and colleagues proposed that, out of a number of potential pathways, the most common scenario would be that “accumulate[d] AD neuropathology over many years along with co-occurring cerebrovascular disease [i.e., related to blood flow in the brain] damages the frontostriatal circuitry [i.e., the neural pathways connecting the frontal lobe to the striatum], leading to late-life depression” [21]. Another overview paper reached similar conclusions but emphasized that the cause-effect relationship needed to be investigated more thoroughly for a solid understanding [22]. While there is thus some support for LLD as a prodromal manifestation of AD, there are many other (preventable) biopsychosocial causes, associations, and risk factors for LLD that are “complementary and almost always transactional” (i.e., the risk factors interact with each other) [23]. Indeed, external psychological and social variables seem equally important in modulating LLD: Behav. Sci. 2020, 10, 132 3 of 36 seniors with smaller social networks, for example, seem to be more at risk of developing or having LLD [24,25]. Similarly, those who do not have a spouse or confidant develop LLD more often than those who do [15,26–28]. Another important factor is the presence of life experiences that build up cognitive reserve. Cognitive reserve can succinctly be described as an individual’s resilience to age- related brain pathology [29]. This reserve can, for example, be built up through higher educational attainment, having (had) a cognitively demanding job, and maintaining cognitively stimulating hobbies later in life [30]. To illustrate, many epidemiological studies investigating LLD reported a higher prevalence of depression in individuals with low educational attainment [24,31,32] and those with less complex occupations during their working life. Likewise, seniors who did not read for leisure were more often depressed, while playing cognitively and socially stimulating card games seemed to protect against LLD [31]. Considering the effects of LLD on personal well-being, and the negative prognosis, the need for effective treatment seems both urgent and evident. Current treatment methods can be problematic, however. Antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), are generally deemed safe in younger populations, but seniors have been found to be much more sensitive to negative side-effects [33]. This can largely be attributed to drug-drug interactions between some SSRIs and other medications often prescribed to seniors [34]. Moreover, while successful implementation of pharmacotherapy has been reported in individual papers [35,36], a 2011 meta- analysis found variable evidence for the efficacy of antidepressants [37]. Conversely, a wide range of psychological treatment modalities, such as cognitive behavioral therapy, life review therapy, and problem-solving therapy have been coined for their potential to alleviate late-life depressive symptoms in several meta-analyses [38,39]. However, while these methods are able to improve patients’ mood and well-being, cognitive dysfunction often persists even after remission [40–42]. Koenig et al. suggested that this persistent cognitive impairment could indicate a prodromal form of dementia, but acknowledged that patients may have had cognitive impairments before the onset of LLD [41]. Taking into account the evidence synthesized above, maintaining cognitive function or preventing cognitive decline during or after a depressive episode in older age seems difficult, if not impossible. Indeed, current treatment methods do not seem to restore cognitive functioning. In addition, since a pervasive public sentiment exists that aging is synonymous with cognitive decline [43,44], it may follow that not enough neural plasticity remains to improve again. However, lifelong learning initiatives such as the university of the third age [45], underscored by a growing body of research [46,47] have shown that improving cognition in healthy seniors through several training modalities is most definitely possible. In contrast, a number of researchers noted that current LLD treatment methods do little to nothing to combat cognitive dysfunction. In response to this, these researchers approached the treatment of LLD specifically through training cognitive functions. One study [48] did so by employing serious gaming software, targeting cognitive control through multitasking paradigms. Employing a cognitive training modality in depressed seniors was found to result in a significant improvement in mood. This change in mood in the training group was statistically indistinguishable from that in a control group following a recognized treatment modality (namely, problem-solving therapy). Interestingly, however, performance on previously untrained cognitive tests pertaining to working memory and sustained attention improved markedly in the cognitive training group, while no such improvement was found in the conventional therapy group. Building on these outcomes, Morimoto et al. developed a computerized cognitive training intervention that used tasks known to specifically train domains associated with treatment-resistant LLD, such as inhibitory control and cognitive flexibility [49]. This intervention was adaptive to the patients’ performance level, such that it was never impossible, but always challenging. In their pilot study, treatment-resistant participants enrolled in a cognitive intervention were compared to depressed controls (not treatment-resistant) receiving a conventional 12-week pharmacological treatment (escitalopram, 20 mg). The researchers reported a statistically significant reduction in depressive symptoms for both groups. Strikingly, however, in the cognitive intervention, this result was obtained after merely four weeks, instead of 12. Additionally, performance on the Stroop task Behav. Sci. 2020, 10, 132 4 of 36 and Trail Making Task Part B (a task in which numbers and letters distributed on a piece of paper need to be connected in ascending order: 1, A, 2, B, 3, C, etc.) increased in the cognitive intervention group, while staying constant in the pharmacological intervention group. While larger-scale studies are needed to consolidate these results, it seems that cognitive training may benefit LLD patients, both in terms of mood and cognition. While results are promising, criticism regarding these interventions must be acknowledged. One general issue with computerized cognitive training, both for treatment of depression and general improvement in cognition, is that it may “teach to the test” [50,51]: the outcome measures used to assess changes in cognition are too similar to the tasks used in the training. As a result, transfer effects are less meaningful. Another issue is the operationalization of these training programs. Computerized cognitive training often consists of somewhat repetitive tasks, done alone by participants in their own homes. A large-scale study in the Netherlands investigating several exclusively home-based cognitive interventions (e.g., language training, brain-training games) in healthy seniors [52] reported high participant attrition, with over 40% of participants dropping out before the two-month interventions were completed. In response, the researchers halved the duration of the interventions, but this, too, failed to reduce drop-out rates. Olfers et al., therefore, argued that adding a social component (such as chatrooms for participants) could increase adherence. Depressed seniors, in particular, could potentially benefit from a stronger emphasis on interaction and social networks, considering that loneliness is strongly associated with LLD [24,25]. One presently unexplored intervention for LLD that emphasizes this social component, while still training cognitive functions, is language learning. It has long been theorized that people who already know two or more languages (i.e., bilinguals) enjoy a certain cognitive advantage in comparison to monolinguals, as has been found in studies examining, for instance, word learning [53] and conflict processing [54]. Bilinguals essentially juggle two languages at all times: while one language is used, the other needs to be suppressed, with control transferring from the language domain to the general cognitive domain. It does need to be pointed out that the precise mechanisms underlying this advantage have been questioned recently [55] Additionally, depending on the individual bilingual experience, switches between the spoken languages may be frequent. It is theorized that these patterns of constant (de)activation and switching enhance cognitive control (that is, a bilingual experience causes transfer effects to general domains of cognition) [56]. It is also important to acknowledge that a fierce debate on the bilingual advantage has gone on since the inception of the concept. A large meta-analysis from 2018 compiled evidence from 152 published and unpublished studies [57]. While evidence in favor of the bilingual advantage seemed to be present, initially, these effects disappeared after correcting for publication bias. The authors concluded, therefore, that no systematic evidence existed for a bilingual advantage. They also emphasized the need for pre-registration of studies to ensure that negative or null-findings are published, too. Similarly, a large-scale study containing data from 11,000 participants found no conclusive evidence of bilinguals outperforming monolinguals [58]. It is important to point out, however, that individual differences in bilingual experiences were not taken into account in these larger reviews. Indeed, another critical review noted that research on bilingualism often contains methodological flaws that influence results [59], concluding that it is pertinent “to focus more research attention on the individual features of bilingual experience, as this is where the putative cognitive effects stem from”. As was mentioned before, bilinguals’ languages are always active, which is theorized to constantly train cognitive control. Indeed, a number of studies have found that a lifetime of bilingualism delays the average onset of Alzheimer’s disease symptoms by multiple years [60–63]. Additionally, a study comparing bilinguals and monolinguals with AD, who were comparable in terms of cognitive functioning, found that the degree of neuropathology in the bilinguals was more extensive [64]. This suggests that compensatory mechanisms pertaining to cognitive reserve in lifelong bilinguals are much stronger. This does not mean, however, that a positive result can only be found if bilingualism is lifelong. The constant interference that follows from juggling two or more languages makes language learning unique from other types of cognitive training, as any type of activity requiring language will automatically lead to these activation/deactivation patterns, Behav. Sci. 2020, 10, 132 5 of 36 interfering with earlier acquired languages in ways that learning other new skills will not. For instance, one study investigating children looked into the cognate facilitation effect, a well-known phenomenon where co-activation of two orthographically and phonologically similar words (e.g., “dokter” in Dutch and “doctor” in English) leads to faster response times [65]. Interestingly, these researchers not only found this facilitation effect in children who were lifelong bilinguals but also in children who were still learning their second language. Other studies have also suggested that such effects could also arise early in the language learning process of seniors. Indeed, in a 2013 paper, Antoniou et al. argued that learning a foreign language would likely activate those neural networks that decline with aging [66]. This includes networks affected in LLD. Therefore, they proposed that late-life language learning could have neuroprotective effects, effectively training the cognitive functions affected in aging (and LLD). Language learning in seniors is a rapidly emerging field. The focus thus far, though, has mainly been on boosting seniors’ cognitive reserve in order to stave off dementia [66]. However, as we have seen, AD and LLD potentially have similar underpinnings [21]. Like Alzheimer’s, LLD is characterized by pathological declines in cognitive function [16]. Similarly, a risk factor for both LLD and AD seems to be reduced cognitive reserve [24,28,31,32] and LLD has even been considered a prodromal expression of AD pathology [20,21]. Successfully ameliorating or reversing cognitive dysfunction in LLD, then, could potentially delay or prevent further cognitive decline. Additionally, learning and using a foreign language is an inherently social activity, especially when done in a classroom setting. A further advantage of classroom language learning is that it potentially tackles the common problems of loneliness and small social networks LLD patients experience [24,25]. Taking away these two risk factors for LLD, then, could potentially improve mood as well (although this is not the focus of the present paper). Published research directly investigating the efficacy of language learning to improve cognition and mood in LLD, is currently non-existent, although work on language learning in seniors is being carried out at the moment of writing, e.g., [67]. 2. Aims and Objectives As the field of late-life language learning as a potential therapy for LLD is in its infancy, it is essential to work towards a stronger theoretical and methodological basis for such interventions. This paper contributes to this aim by conducting two separate meta-analyses. First, an in-depth overview of performance on (standardized) neuropsychological tests (including measures of cognitive flexibility, processing speed, working memory, and inhibition) in seniors with LLD will reveal how cognitive domains are (differentially) affected. Subsequently, an overview of studies investigating the performance of bilingual seniors on these same tests will be compiled (both life-long bilinguals and late-life language learners). We acknowledge that it would be more ideal to only include studies on late-life language learners, but the field of third-age language acquisition is still emerging. It was therefore decided to include studies on lifelong bilinguals, too to substantiate claims relating to future research avenues. We expect the overall cognitive functioning of depressed seniors to be less efficient than in healthy controls, but based on evidence for the neuroprotective effect of bilingualism, we hypothesize that bilingual seniors will outperform their monolingual peers on these tasks. It must be stated that the results of these two meta-analyses will not be directly comparable. Indeed, if healthy older bilinguals outperform healthy older monolinguals, it does not necessarily point to the potential efficacy of language learning as an intervention in LLD. Rather, the two analyses should serve to bridge two currently separate fields (i.e., linguistics and geriatric psychiatry) by giving an overview of the status quo in both fields. Evidence from the present meta-analyses, then, will provide the first steps for a theoretical basis for interdisciplinary work targeting language learning as a potential therapy for LLD. Additionally, as a secondary objective, critical attention will be paid to methodological choices made in the literature on both late-life depression and late-life language learning, providing a framework for researchers from various disciplines interested in these topics. Based on the results of this review, concrete suggestions for future research will be presented. Behav. Sci. 2020, 10, 132 6 of 36 3. Methods 3.1. Inclusion and Exclusion Criteria A number of inclusion and exclusion criteria were established to meet the study’s aims. To be included in the literature review on cognitive functioning in LLD, studies had to compare a group of seniors with depressive symptoms or diagnosed depression to healthy controls (e.g., cross-sectional) on the basis of one or more neuropsychological tests from a pre-specified set. This set included a number of widely-used tests tapping into working memory (forward and backward digit span tasks [68]), processing speed (Trail Making Test A [69]), cognitive flexibility (Trail Making Test B [69], Wisconsin Card Sorting Test [70], color-shape switch task [71], letter-number sequencing task [68], digit-symbol substitution task [68]), inhibition (antisaccade task [72]), and episodic memory (Visual Association Task—Extended [73]). Most of the neuropsychological tasks included in the present analyses were selected because they are commonly used and clinically relevant (e.g., the TMT is often used to indicate potential dysfunction in clinical settings [74]). Two tasks were chosen because they were of particular interest to ongoing work (i.e., the color-shape switch task and the VAT-E). All neuropsychological tasks in the test battery. The neuropsychological tasks were also selected because they would be used in ongoing work of the authors. Longitudinal studies on LLD (e.g., randomized controlled trials) were included, but only baseline data were inspected. For the searches looking into bilingualism, all studies comparing older bilinguals to older monolinguals using this same test battery were included, as were intervention studies looking into language learning in seniors. Other inclusion criteria for eligibility were: 1. Participants were ≥55 years old (Studies on LLD used various cut-off points for a minimum age of inclusion ranging from 45 [75] to 65 [76]. 55 was chosen as a middle ground that primarily identified studies where participants had a mean age of 65 or higher and resulted in more included studies to meet the current investigation’s aims) 2. The study was published in English. Exclusion criteria were: 3. Participants having a comorbid psychiatric Axis-I disorder (except anxiety, which is highly comorbid with LLD [77]); 4. Participants having a comorbid neurological disorder (e.g, Parkinson’s disease or Alzheimer’s disease) [78,79]; 5. Participants reporting a dependency on alcohol, medication, or other substances [80]; 6. Participants using benzodiazepines [81] or beta-blockers [82]; 7. Qualitative studies; 8. Case studies; 9. Meta-analyses or (systematic) reviews. In preliminary searches, only studies that fully adhered to the criteria were considered. However, it soon became clear that a large number of potentially relevant studies, especially those in the realm of bilingualism, had not collected (complete) data on medication use and the presence of comorbidities. For this reason, studies that did not collect these data were also included. The cut-off of 55 years, however, was treated as a hard criterion. 3.2. Search Strategies All searches were done in PubMed and EBSCOhost search engines. Preliminary search strategies aimed to look into broader domains of cognition (e.g., (“late-life depress*” OR “geriatric depress*”) AND (“inhibitory control” OR …). However, due to a large number of results (>8000), it was determined that the searches would be limited to tests used in the test battery of the author’s lab group. The complete search strategy can be seen in Table 1. No limits were set on publication dates. Behav. Sci. 2020, 10, 132 7 of 36 Table 1. Complete search strategy for the present study. Date Database Search String (“late-life depress*” OR “geriatric depress*”) AND (“Trail making t*” OR “TMT” OR “color-shape switch t*” OR “colour-shape switch t*” OR “phonemic verbal fluency” OR “Visual association t*” OR “Vat-E” OR 23-01- PubMed, “wisconsin card sorting t*” OR “WCST” OR “modified wisconsin card 2020 EBSCOhost sorting t*” OR “mwcst” OR “digit-span” OR “digit span” OR “letter number sequencing” OR “letter-number sequencing” OR “symbol coding t*” OR “symbol-coding t*” OR “digit substitution t*” OR “DSST” OR “antisaccade” OR “anti-saccade”) (“senior*” OR “older adul*” OR “third-age” OR “third age” OR “65*” OR “aged” OR “elder*” OR “pension*”) AND (“lifelong bilingualism” OR “bilingualism” OR “language course” OR “language learning” OR “language training” OR “language acquisition” OR “multilingualism” OR “foreign language” OR “language teaching”) AND (“Trail making 17-03- PubMed, t*” OR “TMT” OR “color-shape switch t*” OR “colour-shape switch t*” 2020 EBSCOhost OR “phonemic verbal fluency” OR “Visual association t*” OR “Vat-E” OR “wisconsin card sorting t*” OR “WCST” OR “modified wisconsin card sorting t*” OR “mwcst” OR “digit-span” OR “digit span” OR “letter number sequencing” OR “letter-number sequencing” OR “symbol coding t*” OR “symbol-coding t*” OR “digit substitution t*” OR “DSST” OR “antisaccade” OR “anti-saccade”) 3.3. Study Selection After exporting the search results from PubMed and EBSCOhost, title duplicates were removed manually by one reviewer (JB). Then a list of the remaining titles was compiled, such that the reviewer was blind to the authors and year of publication. Subsequently, titles were carefully screened by the same reviewer, such that only those papers which clearly violated one or more exclusion criteria were removed (e.g., titles indicating that the main focus was bipolar depression). As a next step, the same procedure was repeated for the remaining abstracts. A separate list containing only the title, abstract, and notes from the previous screening round was checked for clear violations of exclusion criteria. Full-text papers were then retrieved from the remaining list of articles. This time, the reviewer looked at all inclusion/exclusion criteria, to determine if the paper could be included in the final review selection. If information pertaining to one or more criteria was missing, the corresponding author for the study was contacted. In the absence of a reply, a first reminder was sent after approximately three weeks. A second, and final, reminder was sent two weeks after the first reminder. The reference lists of included studies were also inspected for potentially relevant work. 3.4. Data Collection Process & Data Items The data necessary to calculate effect size (e.g., mean, standard deviation, and number of participants) were extracted where available. In cases where this was not possible because the necessary data were not reported, the corresponding authors were contacted using the same procedure mentioned in Section 3.3. When these data could not be retrieved, the study was excluded from further analysis. Additional data were extracted to compare the studies’ methods: country where the study was conducted, age of participants, gender distribution, years of education, scores on mini-mental state exams (MMSEs) [83] or the Montreal Cognitive Assessment (MoCA) [84], and the operationalization of depression or bilingualism. Behav. Sci. 2020, 10, 132 8 of 36 3.5. Risk of Bias in Individual Studies Risk of bias in individual studies was assessed independently by two authors (JB, FvdB) using the National Heart, Lung and Blood Institute’s quality assessment tool [85]. This tool consists of 14 questions regarding for instance selection bias of participants, validity of the tools used, and blinding of experimenters. This tool was deemed the most suitable, as it enabled the assessment of cross- sectional studies (others were more aimed at RCTs). A highly similar tool also from the NIH was used for the intervention studies [86]. The scoring guide for these tools explicitly states that overall quality should not be determined by summing up the scores per question. Rather, assessors are instructed to consider how much risk of bias each rating of the questions introduces in the individual studies and to determine the study quality based on the presence of any risks (e.g., selection bias, measurement bias, or confounding). Consensus on the ratings was reached after a discussion between the two reviewers. 3.6. Summary Measures & Synthesis of Results A pervasive issue with systematic reviews is that they employ so-called ‘vote-counting’ methods [87]: if a review, for instance, finds that five out of ten retrieved studies are in line with their hypothesis, they may conclude that evidence is variable. This method is not ideal because it is done purely on the basis of significance; this becomes especially problematic when effect sizes or sample sizes are small, as described in detail by Combs et al. [88]. It is, therefore, recommended that one calculates a pooled average by conducting a meta-analysis [87]. This method reduces the chance of type II errors because it does not simply count p values of generally low-powered studies [89]. Additionally, a meta-analysis assigns more weight to studies with more precision (generally those with larger sample sizes) [87]; this also increases the precision compared to vote-counting. Hedges’ g was calculated for each of the neuropsychological tests included in this review, in R (version 4.0.0) [90] with the ‘esc’ package (version 0.5.1.) [91]. A pooled effect size was subsequently calculated for each neuropsychological test with two or more available effect sizes using the ‘meta’ package (version 4.12-0) in R [92]. This same software package was used to create forest plots displaying the overall effect sizes, and funnel plots to show potential publication bias. Since we expected some methodological variation between studies, we opted for a random-effects model [87]. Neuropsychological tests with only one available effect size were reported in a separate table. 4. Results 4.1. Study Selection The search for papers on cognitive functioning in LLD led to 23 studies that met the inclusion criteria. A total of 14 studies regarding cognition in older bilinguals or older language learners were identified (two of which investigated language learning in seniors). A PRISMA flowchart [93] displaying this process can be seen in Figure 1. Behav. Sci. 2020, 10, 132 9 of 36 Figure 1. PRISMA flow diagram of the screening process for late-life depression (left) and bilingualism (right). 4.2. Summary of Study Characteristics Late-Life Depression An overview of all study characteristics can be found in Table 2, below (for a more detailed version please refer to Table S2 in the supplement). The 23 studies that investigated cognitive function in LLD were published between 1994 and 2018. Two-thirds of these studies (n = 14) were published after 2010. The majority of studies were conducted in North America (n = 13). Of the other 10 studies, five were conducted in Europe, three in Asia, and two in South America. Six studies contained complete information on all inclusion and exclusion criteria. The 17 studies with missing information did not collect or report information regarding the following: comorbid neurological illness (n = 3), comorbid psychiatric illnesses (n = 11), substance or medication abuse (n = 12), or the presence of medication that led to exclusion in the current design (n = 16). 4.2.1. Demographic Information The 23 retrieved studies contained a total of 4366 depressed participants and 8875 healthy controls. LLD studies had an average of 154.9 depressed participants (SD = 359.1) and 385.9 healthy controls (SD = 928.9). Groups were not balanced for gender: both LLD and healthy control groups on average comprised more female participants (Depressed: M = 67.5%, SD = 13.0; Healthy controls: M = 64.8%, SD = 11.9). Age was explicitly reported in all but one study (which expressed age distribution as the percentage of participants within an age band). Depressed participants were on average 73.9 years old (SD = 6.6), while healthy controls had a mean age of 72.3 (SD = 5.8). In the 18 studies that reported the total years of education, depressed participants had—on average—one year of education less than the healthy controls (Depressed: M = 13.1, SD = 4.1; Healthy controls: M = 14.1, SD = 3.7). Lastly, judging from the 17 studies that included MMSE scores, there was no indication of the presence of global cognitive impairment, and it seemed that depressed and healthy groups were well- matched (Depressed: M = 26.9, SD = 2.9; Healthy controls: M = 26.8, SD = 2.5). 4.2.2. Types of Tasks Used The following outcome measures were implemented in the included studies: time to complete TMT B (n = 13), percentage of perseveration errors on TMT B (n = 1), time to complete TMT A (n = 13) (One study [94] reported TMT A and B scores in percentiles; these are reported separately), phonemic verbal fluency (n = 9), digit span forward (n = 9), digit span backward (n = 5), (modified) Wisconsin Behav. Sci. 2020, 10, 132 10 of 36 Card Sorting Test categories (n = 6) (The modified WCST is shorter and less difficult [95]), (m)WCST perseveration errors (n = 4), digit symbol substitution task (n = 8), and score on the letter-number sequencing task (n = 2). The n indicates the number of times a healthy group could be compared to a depressed group. Studies containing multiple comparable groups (e.g., age 65–75 healthy vs depressed and aged >75 healthy vs depressed) were included (and therefore counted) twice. 4.2.3. Operationalization of Late-Life Depression Nearly two-thirds of the studies (n = 14) did not report whether participants in the depressed group had early-onset (before age 60) or late-onset (after age 60) depression. A further five studies reported either only late-onset LLD patients or they provided separate analyses for early and late- onset depression. The remaining four studies reported that they included both early and late-onset patients in their analyses. Blazer states that, while there is some debate on whether there is a difference in etiology between early and late-onset depression, the neuropsychological performance between these two groups does not differ [6]. Therefore, the authors did not differentiate between early and late-onset LLD in their analyses. There was substantial heterogeneity in terms of tools and interviews used to classify depression and depression severity. DSM criteria were used in 14 studies (DSM-III-R, DSM-IV(TR), and DSM-V). Three studies used the International Statistical Classification of Diseases and Related Health Problems (ICD-10), Chinese Classification of Mental Disorders (CCMD-3), and Research diagnostic criteria (RDC), respectively. Most of these studies used additional valid and reliable shorter scales containing between 10 and 30 questions: Hamilton Depression scale (ham-D), the Montgomery Asberg Depression Rating Scale (MADRS), the 30-item or 15-item Geriatric Depression Scale (GDS), the Center for Epidemiological Studies Depression Scale (CES-D), Beck’s Depression Inventory (BDI-II), or the Detection of Depression in the Elderly Scale (DDES). Six of the 23 studies only used one of these shorter scales. Behav. Sci. 2020, 10, 132 11 of 36 Table 2. Study characteristics of individual studies on cognitive function in late-life depression. Demographic information includes where the study took place, participants’ age and gender, educational background, and performance on the mini-mental state examination. Additionally, information is provided on how each study operationalized depression (continued on the next pages). First author N Age Education MMSE Country Group Operationalization of LLD Notes (year) (fem.) (SD) (SD) (SD) 64.9 28.9 DSM-IV criteria + at least LLD 11 (10) 14.4 (3.1) Alalade (4.5) (1.7) some depressive symptoms On average early onset 33.5 (2011) USA 71.2 29.2 as measured with MADRS (18.5) [96] HC 18 (11) 15.8 (2.3) (≥ 8) (6.6) (1.2) 69.0 29.1 Alexopoulos LLD 16 (n/a) 17.2 (2.3) (5.5) (0.9) DSM-IV criteria + MADRS ≥ (2013) USA 68.6 28.5 18 [97] HC 10 (n/a) 16.3 (3.8) (7.0) (1.0) 69.5 27.4 Only the high education group LLD 30 (22) 11.9 (3.3) DSM-IV criteria + (HAM-D + Avila (2009) (7.5) (2.4) was used here since low Brazil MADRS. No minimum for [98] 68.0 28.4 education group had only inclusion) HC 33 (26) 11.9 (3.3) (7.1) (1.7) approx. 3 years of schooling 69.4 27.4 LLD 22 (9) 7.1 (0.9) Balardin (1.1) (0.4) ≥5 on the 15 –item Geriatric (2009) Brazil 67.8 27.7 Depression Scale (GDS) [99] HC 22 (16) 9.6 (0.6) (1.4) (0.3) 71.8 28.1 Baudic LLD 21 (n/a) 9.6 (3.3) (8.8) (1.4) DSM-IV criteria + MADRS > (2004) France 73.6 29.7 20 [100] HC 19 (n/a) 13.5 (5.6) (7.7) (0.6) 74.7 25.5 Beheydt LLD 28 (24) n/a Participants matched on y/of Switzer- (7.6) (3.8) DSM-IV-TR criteria + 30- (2015) education; MMSE scores in land 72.0 28.3 item GDS score ≥11 [101] LLD group significantly lower HC 20 (15) n/a (5.1) (1.4) 64.4 DSM-III-R criteria measured Young-old and old-old USA LLD 23 (n/a) 15.6 (2.6) n/a (2.8) with SCID interview + presented in the same study Behav. Sci. 2020, 10, 132 12 of 36 Boone (1994) (Ham-D. No minimum for 64.4 young-old HC 54 (n/a) 14.5 (2.4) n/a inclusion) (2.7)) [102] 75.5 DSM-III-R criteria measured LLD 14 (n/a) 15.3 (2.8) n/a Boone (1994) (5.1) with SCID interview + Gender distribution overall is old-old USA 74.7 (Ham-D. No minimum for balanced (53% female) [102] HC 51 (n/a) 14.3 (2.8) n/a inclusion) (4.1) 67.5 29.9 LLD 64 (47) 8.9 (4.2) Chinese Classification of (6.0) (0.2) Chen (2013) China Mental Disorder (CCMD-3) [103] 68.2 29.9 HC 31 (18) 9.0 (4.0) + 30-item GDS score ≥ 11 (8.6) (0.2) 68.7 27.8 Dickinson LLD 112 (63) 14.2 (2.5) Not mentioned (Ham-D + (6.3) (2.4) (2011) USA MADRS used to measure 70.5 28.9 [104] depression severity) HC 101 (73) 15.4 (1.8) (5.7) (1.3) 71.4 27.2 LLD 14 (10) 9.4 (2.2) Egger (2008) (7.5) (1.0) DSM-IV criteria + 30-item Austria [105] 72.3 28.6 GDS score ≥ 15 HC 20 (13) 10.8 (2.7) (7.7) (0.8) Elderkin- 70.0 29.0 LLD 95 (63) 14.9 (2.7) DSM-IV criteria + 17-item (7.9) (1.3) Thompson Combined minor and major USA Ham-D ≥ 8 (minor LLD) + (2007) depression 71.5 29.4 HC 71 (42) 14.9 (2.7) 17-item Ham-D ≥ 15 [106] (7.6) (1.1) 69.0 29.0 Elderkin- LLD 112 (77) 15.6 (2.7) DSM-IV criteria measured Thompson (7.9) (1.1) USA with SCID interview + 17- (2011) 71.0 29.3 item Ham-D ≥ 15 HC 138 (86) 15.5 (2.7) [107] (7.7) (1.0) ≅65% Modif. attended or MMSE LLD 204 (204) ≅70 Goveas completed 96.0 ≥5 on the 15 –item Geriatric (2014) USA Women’s health cohort study college (3.5) Depression Scale (GDS) [108] 2017 ≅ 75% Modif. HC ≅70 (2017) attended or MMSE Behav. Sci. 2020, 10, 132 13 of 36 completed 97.0 college (2.7) 73.4 28.4 Kindermann LLD 25 (16) 13.3 (2.5) (6.0) (1.8) Research Diagnostic Criteria (2000) USA 71.2 29.2 + 21-item Ham-D ≥ 17 [109] HC 20 (12) 13.8 (2.5) (6.2) (1.0) 67.9 28.1 LLD 15 (9) 13.9 (2.2) ≥4 on 15-item GDS + ≥ 8 on Leal (2017) (6.9) (0.5) USA 21-item Beck’s Depression [110] 72.2 28.9 HC 27 (18) 16.7 (2.2) Inventory II (7.6) (0.2) 69.4 27.6 Miebach LLD 21 (14) 12.6 (2.8) ICD-10 criteria + (15-item (8.0) (2.1) (2018) Germany GDS. No minimum for 67.5 29.0 [111] HC 21 (12) 15.1 (3.1) inclusion) (7.2) (1.2) 71.2 23.6 LLD group includes major LLD 63 (46) 7.3 (5.5) DSM-IV criteria + (15-item Park (2018) South- (5.1) (4.4) depression, minor depression, Korean GDS. No minimum [112] Korea 70.3 26.3 dysthymia, and subsyndromal for inclusion) HC 59 (29) 10.2 (5.8) (4.7) (3.5) depression Rajtar- 68.1 13.6 (3.3) LLD 87 (57) n/a DSM-V criteria + ≥ 7 on 15- Zembaty (6.0) Poland item GDS + no depressive (2017) 66.8 HC 100 (61) 14.6 (3.0) n/a episode before age 60 [17] (4.8) 83.3 23.1 LLD 40 (25) 10.2 (2.1) (8.6) (4.7) Rapp (2005) DSM-III-R or DSM-IV + ≥ 11 Combined early and late-onset USA [113] 84.1 26.0 on 30-item GDS LLD HC 39 (21) 9.8 (2.0) (6.8) (3.5) 33.3% less 2545 74.8 LLD than high n/a ≥5 on 20-item CES-D (1633) (5.3) Rosano school (subclinical depression) + Combined subclinical and (2016) USA 23.7% less ≥11 on CES-D (clinical clinical depression [114] 2146 74.6 depression) HC than high n/a (1179) (5.4) school Behav. Sci. 2020, 10, 132 14 of 36 71.4 44.6% < 10 26.2 Shimada LLD 657 (245) Combined depressive (4.4) years (2.5) ≥ 6 on 15-item GDS or a (2014) Japan complaints group with 3695 71.5 33.3% < 10 26.5 diagnosis of depression [115] depressed group HC (1921) (5.2) years (2.4) 68.3 DSM-IV criteria measured Smoski LLD 30 (n/a) n/a >26 (6.3) with SCID interview + (CES- (2014) USA 70.8 D. No minimum for [94] HC 40 (n/a) n/a >26 (7.1) inclusion) 70.3 ≥24 in LLD 117 (80) 14.0 (3.6) Steffens DSM-IV criteria + (HAM-D + (7.2) 90% (2001) USA MADRS + DDES. No 70.2 ≥24 in [116] HC 142 (103) 15.8 (2.6) minimum for inclusion) (6.0) 90% Behav. Sci. 2020, 10, 132 15 of 36 4.3. Study Characteristics of Studies on Bilingualism and Aging An overview of all study characteristics can be found in Table 3, below (for a more detailed version please refer to Table S3 in the supplement). The 14 studies that were identified on bilingualism and aging were published between 1997 and 2020, with the majority (n = 11) after 2010. While studies were published in a range of countries (e.g., China, France, Canada), there was a strong bias towards Europe and North-America in participant recruitment. Of the 14 studies, only five studies contained information on all selected inclusion or exclusion criteria. The remaining nine provided no information on comorbid psychiatric illness (n = 5), comorbid neurological illnesses (n = 5), substance or medication abuse (n = 8), or medication status (n = 8). Only two studies used language learning interventions in a group of seniors [117,118]. A complete overview of the individual studies can be found below in Table 3. 4.3.1. Demographic Information A total of 2147 monolingual and 1225 bilingual participants took part in the 12 identified cross- sectional studies. The two intervention studies had 80 language learners and 117 controls. Studies included on average 102.1 bilingual participants (SD = 260.9) and 178.9 monolingual participants (SD = 537.2). These averages were heavily influenced by one large cohort study containing 2812 participants (928 bi/multilinguals). Generally speaking, in the cross-sectional studies the bilingual groups comprised the same number of female participants (M = 58.1%, SD = 24.8) as monolingual groups (M = 59.5%, SD = 10.2), although it must be mentioned that seven of the selected studies either did not report a gender distribution, or they reported it for all participants instead of for the separate groups. Those studies that did not separately report gender for bilingual and monolingual participants had 58.8% female participants (SD = 12.4) on average. In terms of age, monolingual participants in the cross-sectional studies were slightly younger (M = 67.8, SD = 6.8) than bilinguals (M = 70.4, SD = 5.5) in the 10 studies that reported age using mean and standard deviation. 4.3.2. Type of Tasks Used The 14 studies used the following neuropsychological tasks: Digit span forward (n = 6), digit span backward (n = 6), digit span forward + backward (n = 3) (Because a forward + backward version of the digit span task was present, it was decided to pool all implementations of this task for the meta- analyses.), TMT A (n = 4), TMT B (n = 4), TMT B–TMT A (n = 1) (The effect size for this version was taken together with TMT B measures, as the difference between TMT A and B indexes cognitive flexibility [119]), phonetic verbal fluency (n = 4), (m)WCST (n = 3), letter-number sequencing (n = 2), color-shape switch task (n = 1), and the antisaccade task (n = 1). None of the studies (including those on LLD) used the VAT-E. 4.3.3. Operationalization of Bilingualism Substantial variety was attested in terms of how bilingualism was defined. Of the 11 studies comparing monolinguals to bilinguals, one study prompted participants to merely provide the number of languages spoken on a daily basis, without taking into account factors like language proficiency. Only one study used the Common European Framework of Reference for Languages [120] to gauge proficiency. Four studies asked participants for self-reported proficiency measures regarding reading, writing, speaking, and listening. Six studies further supplemented self-rated proficiency with measures regarding, for instance, the frequency and context of use, age or order of acquisition, and language dominance. One of the two studies investigating language learning in seniors required that participants had no previous knowledge of learning another language. The other required that participants had no functional knowledge of the target language (English) but allowed the inclusion of bidialectal participants. Behav. Sci. 2020, 10, 132 16 of 36 Table 3. Study characteristics of individual studies on cognitive function in bilingualism and in language learning interventions. Demographic information includes where the study took place, participants’ age and gender, educational background, and performance on Montreal cognitive assessment or the mini-mental state examination (continued on next page). First MOCA/ Operationali- N Age Education Other Language Author Country Group MMSE Bil. Lang. zation of Notes (fem.) (SD) (SD) Tests (Year) (SD) Bilingualism 10 74.2 MOCA Ansaldo Bi 17.2 (3.1) Bilingual (n/a) (7.4) 27.2 (1.6) French- LEAP-Q [122]; (2015) Canada aphasia test part English ≥30% L2 usage 10 74.5 MOCA [121] Mono 16.1 (3.28) C (n/a) (7.1) 27.7 (1.1) 15 68.8 Speaking two Bi 15.3 (3.7) n/a Context of Craik (n/a) (6.1) languages acquisition; (2006) Canada Varied daily from 15 70.3 frequency of L2 [123] childhood (≤ Mono 16.1 (3.5) n/a (n/a) (4.3) usage 10 years) 60.5 MMSE Semi-structured Authors sent Bi 88 (46) 14 (4) (4.3) 28.7 (1.23) Speaking two interview; a subset of Estanga Basque- languages Bilingual dataset that (2017) Spain 43 60.8 MMSE Spanish regularly and Language Profile adhered to [124] Mono 12 (2) (25) (4.4) 28.4 (1.31) fluently questionnaire inclusion [125] criteria 26 69.7 Self-rating Bi 16.3 (0.5) n/a (n/a) (0.8) reading, Speaking two speaking, Fernand languages listening, writing es (2007) Canada Varied regularly from (1–10); AoA; 16 74.1 [126] Mono 15.5 (0.5) n/a adolescence (n/a) (7.5) frequency of L2 (≤12 years) usage; language preference Friesen 21 71.1 MMSE Speaking two Language and Canada Bi n/a Varied (2015) (n/a) (3.8) >26 languages Social Behav. Sci. 2020, 10, 132 17 of 36 [127] fluently on a Background daily basis Questionnaire [128]; self-rated 20 70.9 MMSE Mono n/a prof.; frequency (n/a) (2.6) >26 of language usage; context of language use 70.0 Data from Bi 20 (11) 13.7 (4) n/a Gubar- (6.6) two Self-rating of chuk USA + Russian- experiments n/a overall English (1997) Russia 68.2 English in the same Mono 20 (10) 12.0 (0) n/a proficiency (1–4) [129] (7.5) study combined 928 Speaking ≥ 2 Bi ≥ 65 n/a n/a Ihle (n/a) languages Dialects did Switzer- (2016) Varied regularly not count as land 1884 [130] (regardless of languages Mono ≥ 65 n/a n/a (n/a) proficiency) 70.6 MOCA Reaching high Self-rating Bi 28 (7) 16.1 (1.27) Johns (5.7) 27.7 (1.3) proficiency in reading, French- (2016) Canada L2 in early speaking, 74.1 MOCA English [131] adolescence listening, writing Mono 16 (10) 15.1 (3.3) (7.5) 27.1 (1.9) (≤13 years) (1–5) 70.7 MOCA Reaching high Self-rating Combined Bi 36 (17) 16.1 (2.9) French- Kousaie (5.9) 27.5 (1.6) proficiency in reading, two English & (2014) Canada L2 in early speaking, monolingual 72.4 MOCA English- [132] Mono 61 (23) 15.7 (2.7) adolescence (≤ listening, writing groups for (6.5) 27.5 (1.5) French 13 years) (1–5) our analysis 71.4 Highly Self-rating López Bi 18 (10) 16 (2.6) 27.6 (1.6) (4.0) proficient in reading, Zunini French- Canada both speaking, (2019) 71.7 English Mono 18 (11) 15.6 (2.7) 27.8 (1.3) languages; no listening, writing [133] (3.5) functional (1–5) Behav. Sci. 2020, 10, 132 18 of 36 knowledge of other languages 16 72.3 MMSE Sociolinguistic Bi 16.0 (2.7) Bilinguals Massa (n/a) (5.0) 29.5 (0.8) French- questionnaire were all (2020) France Italian/Itali regarding 16 71.1 MMSE dominant in [119] Mono 15.1 (2.4) an-French frequency/conte French (n/a) (5.9) 28.9 (1.6) xt of usage Most ≥ 19 MMSE ≥ B2 on CEFR Bi >55 Undergrad English- German Prehn (n/a) 29.7 (0.5) measured with uate German & vocabulary and (2017) Germany Goethe Most ≥ Russian- grammar test [134] 18 MMSE Institute Mono >55 Undergrad German using cloze test (n/a) 29.3 (0.9) Placement Test uate First N Age Global Language Country Group Education Intervention author (fem.) (SD) cogn. learnt? 92% ≥ 8-month MMSE Lang 26 (12) 67.4 secondary language Ramos 28.4 education course by an (2017) Spain Basque 71% ≥ education [117] MMSE Ctrl 17 (9) 69.2 secondary center for 28.5 adults education Most < high ADAS- 53 70.8 Lang. school Cog (43) (6.0) 5 hr per week graduate 8.3 (4.6) 2 to 3 days a Wong Most < high ADAS- 51 71.1 week for 6 (2019) China Game school Cog English (44) (6.5) months in [118] graduate 9.1 (5.4) senior center Most < high ADAS- 49 71.1 (130 hr max) Music school Cog (43) (6.1) graduate 9.1 (5.6) Behav. Sci. 2020, 10, 132 19 of 36 4.4. Risk of Bias in Individual Studies The risk of bias analysis revealed that—out of the 34 identified cross-sectional studies—all but one clearly stated their aims, research questions, and hypotheses. The study populations were in general also adequately described (n = 22). Studies that did not accomplish this mostly failed to mention the location from which participants were recruited. Furthermore, whether groups within a study were recruited from similar populations (e.g., from the same time period or location) was unclear in many cases (n = 16). Strikingly, only one study reported a justification for their sample size [113]. Many of the studies that did not do this, however, did mention small sample sizes as a limitation. Around half of the studies accounted for different levels of depression or bilingualism in their analyses (n = 18). A little over two-thirds of identified studies (n = 22) defined their populations consistently (e.g., using the same screening tools). Those that did not screened for LLD on the basis of two versions of the DSM [113], or they used different measures of language proficiency between groups [127]. Having a double-blind study was essentially impossible in most studies due to targeted recruitment of participants with a certain background; therefore, participants were generally aware of their group-status. However, out of all identified studies only one intervention study stated that the experimenters were blinded to group-status of participants [118], and only one cross-sectional study stated that both mood and neuropsychological function were measured by the same clinician [17]. The remaining 35 studies did not mention blinding of experimenters. It must be noted, however, that for many studies investigating bilinguals, blinding experimenters is essentially impossible, since researchers would often know what participants’ first and second languages were. Lastly, around two-thirds of all studies (n = 21) controlled for potential confounding variables (e.g., age, gender, years of education), either during the recruitment phase or using subsequent statistical methods. While this might not seem high, only one study had significant group differences at baseline that were not controlled for [110]. The remaining studies did not provide enough information to determine whether confounders were present. For a complete overview of the risk of bias analysis with results for individual studies please refer to table S3 in the Supplementary Materials. 4.5. Cognitive Functioning in LLD and Bilingualism The pooled effect sizes for cognitive functioning in LLD are presented in a forest plot in Figure 2. Tasks with insufficient data (i.e, TMT A and B performance measured in percentiles and TMT B perseveration errors) are reported separately in Table 4. Since the two intervention studies looking into late-life language learning used different tests, their results can also be found in Table 4. Behav. Sci. 2020, 10, 132 20 of 36 Figure 2. Forest plot for cognitive functioning in late-life depression showing individual and pooled effect sizes. A positive standardized mean difference indicates a higher average score for the depressed group. Every square in the forest plot represents a study. The bigger the square, the more it weighs in the pooled effect. The blue diamond represents the pooled effect size (expressed through the standardized mean difference, or SMD). The effect on the left of the null line in Figure 2 indicates a lower average score for the LLD group. A negative effect size generally suggests a disadvantage for the LLD group on that particular task (e.g., fewer categories completed in the (m)WCST implies less cognitive flexibility). However, the opposite is the case for TMT A and B (positive effect size means longer completion times), as well as perseverative errors on the (m)WCST (positive effect size means more errors for the LLD group). Behav. Sci. 2020, 10, 132 21 of 36 When looking at the pooled random effect sizes, significant (i.e., with confidence intervals not intersecting zero) small negative effects were found for the digit span task and phonetic verbal fluency. A medium significant positive effect was found for perseveration errors on (m)WCST. Large significant effect sizes were found for TMT A, TMT B, and (m)WCST (categories completed). No significant effects were found for DSST and letter-number sequencing. Overall, judging by the direction of results, the LLD groups seemed to perform worse on neuropsychological tests from our prespecified list. The (pooled) effect sizes for cognitive functioning in bilinguals can be interpreted the same way (Figure 3): any standardized mean difference (SMD) on the left of the null line indicates a lower average score for the bilingual participants. Again, only for TMT A and TMT B does a negative effect size represent an advantage (faster completion times). The pooled effect sizes for all tasks were small. None of the pooled effect sizes reached significance (i.e., the confidence interval for every test intersected the null line), except for TMT A. In TMT A, a small, significant negative effect was found, indicating that bilinguals completed the task faster than monolinguals. The overview with tasks that were only used once in Table 4 contains three separate groups: studies on late-life language learning, studies on lifelong bilingualism, and studies on LLD. Only performance on the digit-span (backward) was found to significantly improve after a foreign language intervention [118]. The other study found no significant differences following a language course intervention. The studies on lifelong bilingualism reported no significant differences between bilingual and monolingual groups, except for Ansaldo et al.’s study [121], which reported an advantage for bilinguals in terms of accuracy on the TMT A and B with a medium and small effect size, respectively. Figure 3. Forest plot for cognitive functioning in older bilinguals showing individual and pooled effect sizes. A positive standardized mean difference indicates a higher average score for the bilingual group. Behav. Sci. 2020, 10, 132 22 of 36 Table 4. Effect sizes of neuropsychological tasks that were used only once. Author Group Test Hedges’ g Conclusion 0.09 (very small) language vs. No significant differences Wong Language Digit-span gaming between language/music/gaming (2019) learning (forward) 0.04 (very small) groups after completing the [118] language vs. intervention music −0.26 Language vs. games Significant gains in digit-span Wong Language Digit-span (small) performance in language learning (2019) learning (backward) −0.58 language intervention, but not in the music [118] vs. music or gaming interventions (medium) Color-shape Decrease in RTs on switch trials Ramos Language- switch 0.1 was slightly larger in language (2017) learning (RTs to (small) learning group than in controls, [117] switch trials) but this was not significant Massa Bilinguals made slightly more Lifelong Antisaccade −0.50 (2020) mistakes than monolinguals, but biling. (% correct) (medium) [119] this was not significant Bilinguals had slightly shorter Massa Antisaccade Lifelong 0.18 response times than (2020) (congruent – biling. (small) monolinguals, but this was not [119] incongruent) significant Ansaldo Lifelong TMT A −0.73 Bilinguals made fewer errors (2015) biling. errors (medium) than monolinguals [121] Ansaldo Lifelong −0.33 Bilinguals made fewer errors (2015) TMT B errors biling. (small) than monolinguals [121] Smoski LLD took slightly longer to TMT A −0.09 (2014) LLD complete TMT A than HC, but (percentile) (small) [94] this was not significant Smoski LLD took longer to complete TMT B −0.47 (2014) LLD TMT B than HC, but this was not (percentile) (medium) [94] significant Steffens TMT B (% LLD group made significantly 0.27 (2001) LLD perseveratio more perseveration errors than (small) [116] n errors) HC 4.6. Risk of Bias Across Studies Separate funnel plots for studies on LLD and studies on bilingualism were made (Figure 4). The results for LLD, while containing a small number of outliers, were mostly symmetrical, although one study had an abnormally large SMD. Results regarding bilingualism were somewhat asymmetrical, suggesting a publication bias for small studies with larger effects [87]. Figure SEQ Figure \* ARABIC 4: Funnel plots for studies on Behav. Sci. 2020, 10, 132 23 of 36 Figure 4. Funnel plots for studies on late-life depression (left) and bilingualism (right). An asymmetrical distribution suggests publication bias. 5. Discussion 5.1. Cognitive Function in LLD Our findings indicate that, in line with previous work [16], LLD is consistently associated with cognitive dysfunction. We reported large impairments in processing speed (TMT A), a small to large disadvantage on multiple measures of cognitive flexibility (TMT B, verbal fluency, (m)WCST categories, (m) WCST perseveration errors), and a small deficit in working memory (digit span). For both measures of cognitive flexibility (the DSST and letter-number sequences), mean effect sizes were in line with our expectations, namely that healthy controls outperformed LLD groups. However, the confidence intervals intercepted the null-line, meaning that this effect was not significant. For the DSST, the absence of a significant effect was potentially caused by one study that actually reported a higher number of correct responses for the depressed group. The authors of that study, however, also mentioned that the LLD patients also took more than twice as long to match a digit to a symbol, suggesting an overall processing disadvantage for the depressed group. Only two studies examined the performance of depressed patients and controls on the letter-number sequencing task. One study found a very large negative effect [110], while the other found no significant difference [106]. This led to a large, yet insignificant, pooled effect size. The study by Leal et al. [110] reflects a well-known issue with smaller sample sizes: they potentially overestimate the actual effect size [135]. This overestimation, combined with the small number of studies that used this test, potentially skewed the results. Two studies were not analyzed in the meta-analysis because they reported the TMT differently from the other studies (i.e., completion time expressed as a percentile score [94]; number of perseveration errors on TMT B [116]). Instead, they were analyzed separately. Both studies found depressed participants performed more poorly, as was hypothesized, although only Steffens et al.’s [116] results reached significance. This also provides further support for our hypothesis that LLD patients experience cognitive dysfunction. In spite of the generally consistent direction of our findings, most of the results showed substantial heterogeneity (as measured with I ). Heterogeneity may have been caused by differences between studies in terms of participants’ educational level, present in the studies measuring DSST, for instance. The heterogeneity may also have been caused by underlying group differences not captured in our data. Since the present study dealt with a small number of studies per comparison, heterogeneity was to some extent expected [136]. Additionally, because it was known beforehand that groups would by no means be the same, a random-effects meta-analysis was chosen a priori [87]. This method produces more conservative estimates, which partially counteracts this heterogeneity [137]. Despite the small number of studies per comparison, this meta-analysis has provided evidence in favor of our hypothesis, suggesting that generally speaking (non-demented) seniors with Behav. Sci. 2020, 10, 132 24 of 36 depressive complaints perform less well on most measures of cognitive functioning in our selected list of neuropsychological tests. These results, therefore, are also in line with previous research, which reported that seniors with depressive complaints have impaired cognitive functioning, overall [16,40]. 5.2. Cognitive Function in Bilinguals While the direction of the effects was generally consistent in the analysis on LLD (despite the heterogeneity), the effects of studies investigating cognitive functioning in bilingualism were highly inconsistent. Overall, none of the pooled effects, except performance on TMT A, reached significance, meaning that statistically speaking our meta-analysis did not support our hypothesis. Rather, our results seem to be in line with Lehtonen et al.’s meta-analysis [57]: no statistical support was present for better preserved cognitive functioning in older bilinguals in the identified studies. Studies specifically looking into language learning in older adults were very sparsely present in our dataset. This can be partially explained by the fact that studies on this topic are scarce, but also because some of the studies on this subject employed neuropsychological tests, not from our predefined list (e.g., [138,139]). Only two studies on late-life language learning could be included [117,118]. These two studies employed no common neuropsychological tests between them, meaning that a pooled average could not be calculated. They will, therefore, be discussed separately. Wong et al. [118] compared a six-month language learning course to a gaming intervention (active control) and a music appreciation intervention (passive control) of the same duration. A small but significant increase was reported exclusively in the language group for working memory, as measured through the backward digit span task. No such increase was reported for the forward digit span. However, the authors did conclude that global cognition, as measured on the basis of the Alzheimer’s Disease Assessment Scale-Cognitive Subscale [140], had improved significantly both in the group learning a foreign language as in those enrolled in the gaming intervention. Ramos and colleagues on the other hand asked participants to complete an eight-month language course [117]. Changes in cognitive flexibility were measured using a color-shape switch task. Reaction times for both switch and non- switch trials decreased considerably between pre-test and post-test in the language training group (by 129 and 157 ms, respectively), while staying relatively constant in a passive control group (69 and 66 ms, respectively). However, the switch cost (the difference in RT between switch and non-switch trials) did not decrease in the language group or the control group. The authors, therefore, concluded that eight months of language training in seniors did not lead to better switching ability, but stated that further research was needed. However, it should be noted that participants in this study could self-enroll for the language course (contrary to Wong et al.’s study, which randomly allocated participants to a group). This may have influenced results. In fact, the language learning group outperformed the control group significantly before the course started, suggesting that the two groups differed at baseline. Lastly, only one type of test was used both at baseline and after completion of the course. The addition of a broader test battery could have provided more nuance. The findings in this meta-analysis do not support our hypothesis that lifelong bilingualism or late-life language learning enhances cognitive functioning. Only processing speed, as measured through the TMT A, was significantly faster in bilinguals and significantly slower in LLD groups. However, the fact that the results both for the meta-analysis and the individual studies were inconsistent does not mean that the potential for language learning to combat cognitive dysfunction in LLD should be discredited altogether. Our meta-analysis on bilingualism was based on 12 of the 14 identified studies in total, which generally had small sample sizes. Furthermore, the small number of studies (in both the meta-analysis on LLD and on bilingualism) may have decreased the accuracy of the summary statistic; something that will be discussed in more detail in the section on Strengths and limitations. Similarly, rejecting the notion that language learning in late-life may lead to cognitive benefits based on two longitudinal studies would also be an overgeneralization. In looking at the study characteristics in Table 3, we see many variations in terms of how the target demographic was defined. As a result, it becomes difficult to reach conclusions on the basis of the meta-analysis presented above. Instead, it is clear that more research is needed. To ensure that results of future Behav. Sci. 2020, 10, 132 25 of 36 research (both regarding lifelong bilingualism and late-life language learning) are more reliable, the variation present in previous research will be critically evaluated, upon which recommendations for further research will be made. 5.3. Issues in Defining Bilingualism To preface this section, while not strictly relevant to research on late-life language learning, the authors noticed much variation in how bilinguals were described. As much of the work investigating the potential merits of bilingualism in later life compares lifelong monolinguals to lifelong bilinguals, we feel it is important to discuss the term ‘bilingual’. Both the studies on bilingualism and those on LLD used a variety of (diagnostic) criteria and screening tools to differentiate between the two groups they compared. What set studies on LLD apart from the bilingualism studies, however, was their use of validated tools such as the Geriatric Depression Scale and the 17-item Hamilton Depression Rating Scale. These scales were designed to distinguish between healthy and depressed older adults, according to widely accepted DSM and ICD-10 criteria, with high accuracy [141,142]. The studies on language, on the other hand, were much less consistent in defining their monolingual and bilingual participant groups. The term ‘bilingual’ was for the most part operationalized by combinations of proficiency, age of acquisition, and usage patterns. Some studies, for instance, required that bilingual participants had reached a high level of proficiency in the second language before a certain age, which ranged anywhere from 10 to 13 years [123,126,131,132]. What constituted high language proficiency, however, also varied substantially. This was often determined by means of self-reported scales regarding production and reception [129], while only one study used more standardized and objective measures (i.e., based on CEFR guidelines [120]) of language proficiency [134]. Other studies further defined the bilingual groups as those who spoke two languages on a regular basis [123,126,130]. Overall, though, the many ways in which bilingualism was operationalized very likely played a role in the inconsistent results reported above. These differences between the bilinguals in one study versus those in another are likely caused by the lack of consensus on a ‘case definition’ of bilingualism. In this regard, our findings strongly mirror those reported in a systematic review by Surrain and Luk [143]. They, too, noted that the bilingual experience is a multidimensional construct, consisting of the frequency of use, proficiency, contexts in which languages are used, and sociolinguistic background variables such as the attitudes towards bilingualism in a certain region. In their review of work on bilingualism published between 2000 and 2015, they found that definitions varied considerably. Most studies emphasized proficiency and language use, with less than half using objective measures; this was especially common in studies investigating adult participants. An additional issue was that oftentimes the exposure to languages was presented purely categorically (e.g., whether or not someone spoke multiple languages in the home), instead of on a continuous scale (e.g., percentage of first language usage in the home). As a result, relevant data regarding the situations in which participants used their languages were not present. Surrain and Luk noted that coming to a consensus regarding what makes a bilingual is not realistic. Instead, they argued that our understanding of the cognitive effects of bilingualism would only become stronger when studies included a more complete and transparent description of bilingual experiences. This should include not only proficiency, but also language history, language usage patterns, and sociolinguistic information (e.g., the status of bilingualism in the country where the study took place or personal attitudes towards one’s own bilingualism). 5.4. Issues Pertaining to Older Populations Specifically The inconsistent definition of bilingualism is problematic, but to some extent unavoidable due to the many opinions on what constitutes a bilingual experience and the many (confounding) factors at play in individual bilingual experience. Similarly, in studying older adults, researchers need to take into account certain confounding variables particular to the demographic. One such issue would be the presence of (comorbid) disorders known to affect cognition, such as dementia or mild cognitive impairment (MCI). A majority of the studies on bilingualism (n = 10) and those regarding LLD (n = Behav. Sci. 2020, 10, 132 26 of 36 20) that were included in our review employed screening tools for cognitive impairment and dementia such as the MMSE and the MoCa [83,84]. At the same time, nearly a third of the studies on bilingualism (n = 4), did not report using such tools, or the included participants self-reported that they were healthy. While they can never be a replacement for a diagnosis by a medical professional, tools like the MMSE are able to indicate if a participant has a cognitive impairment that could act as a confounder. An area where more studies (including those on LLD) were lacking, though, was ensuring that participants were not taking medication that could affect cognitive functioning. Out of the 14 studies on bilingualism, 10 did not have sufficient information available to ensure that participants were not prescribed benzodiazepines or beta-blockers. Additionally, 14 out of 23 studies on LLD did not report on this, either. As mentioned before, this is especially problematic for benzodiazepines [81], which are commonly prescribed to and even abused by seniors [144]. 6. Directions for Future Research Results of the present study seem to suggest that, as was hypothesized, cognitive dysfunction in LLD is consistently present. Our meta-analysis also shows that results investigating cognitive functioning in older bilinguals are inconsistent and that no significant evidence was found in our sample that indicated preserved cognitive function in older bilinguals. The authors also noticed that very little research had been done on the effects of late-life language learning on cognition. While the authors of the present study are primarily interested in the role that language learning may have in later life, we realize that cross-sectional study designs are not as resource-intensive as longitudinal designs. For this reason, they are more commonly used to investigate the potential effects of bilingual experiences in seniors. The following section will therefore also give directions for future research comparing lifelong bilinguals to lifelong monolinguals. 6.1. Language Learning Interventions Only two studies investigating the effects of late-life language learning on cognition were identified for this review. Given the strong potential of language learning as a cognitive training tool [66], more work is clearly needed. In addition to general methodological improvements mentioned further on, forthcoming longitudinal, research investigating the potential effects of late-life language learning can improve in quality by assigning participants to interventions randomly, like Wong et al. [118]. While this does not completely eliminate the selection bias caused by recruiting participants from the community, it at least negates it somewhat. Additionally, from an applied linguistics perspective, we recommend future intervention studies to tailor their language learning interventions specifically to characteristics, needs, and preferences of seniors [67,145] to maximize the potential benefits of late-life language learning on a motivational, cognitive, and perhaps social (well-being) level. Another important variable that the present analysis did not look into due to a lack of studies on late-life language learners is the appropriate intervention duration. One can imagine that an ideal outcome would follow a multiple year intensive program; indeed, the activities that build up cognitive reserve generally seem to be garnered over a lifetime of repeated exposure [30]. In practice, however, such trials would be incredibly impractical due to participant attrition and budgetary constraints. A 2017 meta-analysis investigated the efficacy of several types of cognitive training in seniors [146]. The interventions these researchers identified consisted of one to 180 sessions and total exposure lasted between 1 and 270 h. Their analysis concluded that 20 or more training sessions administered one or two times per week led to stronger net gains. Their study did not find that intervention duration by itself influenced the effects. Another review investigating primarily the effect of short language learning interventions in mostly student-aged populations found structural changes in brain anatomy (e.g., increased gray matter) after as little as three months of intensive language learning [147]. Improvements in performance on neuropsychological tasks could arise even faster, though. One study reported a sustained improvement in cognitive flexibility (i.e., still present nine months post-intervention) after only one week of intensive language training [139]. In short, the key for optimal results seems to be intervention intensity instead of intervention Behav. Sci. 2020, 10, 132 27 of 36 duration. Further research should focus on an ideal balance between the two. Finally, while this was not the focus of the present study, we feel that researchers in geriatric psychiatry and linguists alike should also study the potential effects of language learning on well-being. While research on this is incredibly sparse at the moment, the few published studies suggest that late-life language learning is an experience that fosters social interactions and feelings of empowerment over one’s aging process [138,148]. 6.2. Lifelong Bilinguals Versus Lifelong Monolinguals A number of potential reasons may explain the inconsistent findings regarding cognitive functioning in lifelong bilinguals. However, if we start at the root, the inconsistency may be (partly) caused by the operationalization of the term ‘bilingualism’, which ranged from an index of the total number of languages spoken daily (without taking into account dialects or language-proficiency) to more complex measures including frequency of use and context of acquisition. For future research, we recommend using a more transparent operationalization of bilingualism, in line with Surrain and Luk’s recommendations [143], which goes further than, for instance, summing the number of languages a participant reports to speak. This means being more explicit in describing the bilinguals: what languages do they speak, when did they learn them, how do they use their languages, what are their attitudes towards their languages? We believe that using background questionnaires such as the Language Experience and Proficiency Questionnaire (LEAP-Q) [122] and the Language and Social Background Questionnaire LSBQ [128] will help attain this. The information that comes from this, however, is often difficult to use in statistical analyses because resulting the data are quite extensive. An additional, very recently developed, measure that could be useful is language entropy [149]. Entropy essentially takes language background questionnaire data and converts it to an index expressing how balanced participants are in their language use. 6.3. General Methodological Improvements More generally, studies both on bilinguals and LLD were often lacking in the medical information collected and/or reported by the authors. It may be difficult to obtain a participant’s full medical history. However, if it is known whether participants take certain medications, or if they have cognitive impairments as measured through short screening tools such as the MMSE or MoCa, it is at least possible to control for this in analyses. We, therefore, recommend that forthcoming work targetting older bilinguals or seniors learning a language (whether healthy or depressed) records this information, too. On the one hand, this will improve the reliability of new research findings within the field of foreign language acquisition. For instance, it is well-known that long-term benzodiazepine use in seniors leads to considerable cognitive deficits [81]. Including these individuals without controlling for their medication use, then, introduces an unwanted confounder. At the same time, some relatively strong claims regarding bilingualism’s neuroprotective effects are currently being made. The inclusion of these measures will enhance the (clinical) relevance of research on bilingualism for those working in other fields, like gerontology. Controlling for these variables, then, will not only improve overall study quality, but it will also further stimulate understanding and cooperation between disciplines. Furthermore, based on what was seen in the risk of bias assessment, a number of recommendations for forthcoming cross-sectional studies both on LLD and bilingualism can be made. Firstly, researchers should reduce bias by including power analyses, and by blinding assessors for participants’ group-status as much as possible. If the latter is not possible due to, for instance, the language of a particular task giving away group status, researchers should be more transparent about this. Similarly, bias could be reduced by increasing transparency regarding where and how participant groups were recruited. Again, it may not always be possible to recruit subjects from one clinic. However, a considerable proportion of studies did not provide enough information to determine whether participants were recruited from similar populations. Lastly, while over half of the identified studies reported that the researchers controlled for potential confounding variables Behav. Sci. 2020, 10, 132 28 of 36 (such as age or education level), a large proportion of studies (n = 14) did not provide enough information to determine whether this was done. Again, we recommend that researchers explicitly report if participant groups differed on confounding variables and whether they controlled for this in their analyses. 7. Strengths and Limitations While it has been tentatively proposed that language learning may have a positive effect on cognitive functioning in seniors with depression [66], no study as of yet has bridged the two separate disciplines needed to research this (i.e., linguistics and geriatric psychiatry). A strength of the present research was the use of weighted averages across tests. Vote-counting would have resulted in an outcome that would have seemed more intuitive on the surface, but this method is flawed [87–89]. In interpreting the results of this study, a number of limitations should be considered, however. Firstly, the search strategies were developed to identify studies using specific validated neuropsychological tests that would likely also be used in ongoing work (including but not limited to our own lab). On the one hand, this was an advantage because it made the studies relatively easy to compare. A disadvantage, on the other hand, was that some relevant studies were excluded because they did not use tests from our predefined list. A study by Bialystok et al. [150], for instance, found that older bilinguals were better at certain types of conflict resolution than age-matched monolinguals. However, because the tests they employed (modified antisaccade tasks) did not match our list, their study had to be excluded. Another shortcoming of the present review was a result of it being conducted during the COVID-19 pandemic. A large majority of the corresponding authors who replied to our queries were more than willing to provide the information requested. However, due to travel restrictions, some were unable to access the data requested. Additionally, since the majority of identified studies did not report or collect data regarding all our inclusion criteria (e.g., medication use), there is a chance that some studies contained undesirable confounders. Regarding medication use, the exclusion of beta-blockers was chosen because reported side-effects include reduced neuropsychological functioning and increased risk for depression [151]. Recent reviews, however, have provided evidence contradicting this notion [152]. In practice, however, the majority of studies excluded for undesirable medication were due to benzodiazepines, which are known to affect cognition significantly [81]. Another limitation was that the selection of publications based on inclusion and exclusion criteria was done by one author (JB). Potential bias caused by this, however, was negated by showing the reviewer (JB) only the title and abstracts. Another major shortcoming is the small number of studies identified. Borenstein et al.’s comprehensive introduction to meta- analyses does, in fact, warn against a lowered accuracy when performing a random-effects meta- analysis on the basis of few studies [87]. However, immediately afterward they also state that “when faced with a series of studies people have an almost irresistible tendency to draw some summary conclusions from them” (i.e., vote-counting). Borenstein et al. therefore suggest that it is preferable to still compute a summary statistic with known shortcomings instead of inviting unknown properties, a procedure followed in the present investigation. The last shortcoming was using studies on lifelong bilinguals in lieu of papers focusing solely on language learning interventions. We argue that bilingualism is a continuum of experiences [153] and that part of the beneficial effect stems from juggling two competing lexicons [154], a process underlying both the language experience of lifelong bilinguals and new learners. This is why we opted to supplement data on late-life language learners with that of older lifelong bilinguals. However, to say that a lifelong bilingual and a late-life language learner are equivalent in terms of cognitive performance is impossible to say based on the limited existing research. 8. Conclusions A growing body of research is emerging regarding the potential positive effects that bilingualism and language learning may have on old-age disorders [66,155]. These studies have typically focused on building up cognitive reserve across the lifespan in order to stave off clinical symptoms of Behav. Sci. 2020, 10, 132 29 of 36 dementia [60–64]. However, language learning’s potential therapeutic role in late-life depression (LLD), which has been theorized to be a prodromal manifestation of Alzheimer’s [20,21], is underinvestigated. In order to lay out a theoretical basis for future (interdisciplinary) research, this review aimed to provide an overview of the cognitive domains affected in LLD, as well as an overview of cognitive functioning in older bilinguals and seniors learning a language. The results show that despite heterogeneity between studies, LLD was consistently associated with reduced performance on neuropsychological tests. Studies on bilingualism, however, like Lehtonen et al.’s meta-analysis [57] did not show consistent evidence in favor of enhanced cognitive functioning in bilinguals. We believe that this was likely caused by a varying operationalization of bilingualism and a lack of controlling for age-specific confounders. For further research, we suggest taking these potential pitfalls into account. Lastly, it was noticed that the body number of studies on longitudinal language interventions in seniors was very small Aside from a general call to investigate late-life language learning further, the authors recommend that linguists and researchers in gerontology alike be mindful of possibilities reduce their study’s risk of bias through, for instance, randomization of group allocation. Supplementary Materials: Supplementary materials can be found in the open science framework (https://osf.io/x3wqz/?view_only=627012de77b44e729cc3c04dcaab534f). They consist of an overview of the selection procedure for both the search on bilingualism (S1) and LLD (S2), as well as a complete overview of the risk of bias assessment (S3). Author Contributions: The conceptualization of the project and the development of methods was done by J.B., M.K., and R.K. Search strategies, extraction of results, screening, and analyses were done by J.B., with feedback from M.K. and R.K. Risk of bias assessment was done by F.v.d.B. and J.B. All authors helped draft the manuscript, and discussed content, results, and writing at length (J.B., H.L., F.v.d.B., R.K., M.K.). All authors have read and agreed to the published version of the manuscript. Funding: Funding for this project was provided by the Dutch Research Council NWO (grant no. 016.VIDI/185.190 awarded to Merel Keijzer). The funding agency had no active role in the design, execution, or analysis of the research. Acknowledgments: The authors would like to thank the three anonymous reviewers for their constructive feedback and useful comments that greatly improved the quality of our manuscript. Additionally, we would like to extend our gratitude to Marie-José van Tol for her helpful comments on an earlier version of the manuscript. Lastly, we are incredibly grateful for the corresponding authors of the identified papers for their willingness to answer our queries despite an ongoing pandemic. Conflicts of Interest: The authors declare no conflict of interest. References 1. Kangas, O. One hundred years of money, welfare and death: Mortality, economic growth and the development of the welfare state in 17 OECD countries 1900–2000. Int. J. Soc. Welf. 2010, 19, S42–S59. 2. Meerding, W.J.; Bonneux, L.; Polder, J.J.; Koopmanschap, M.A.; van der Maas, P.J. Demographic and epidemiological determinants of healthcare costs in Netherlands: Cost of illness study. BMJ 1998, 317, 111– 3. 2020 Alzheimer’s disease facts and figures. Alzheimer’s Dement. 2020, 16, 391–460. 4. Hebert, L.E.; Weuve, J.; Scherr, P.A.; Evans, D.A. Alzheimer disease in the United States (2010–2050) estimated using the 2010 census. Neurology 2013, 80, 1778–1783. 5. American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders 1; American Psychiatric Association: Arlington, VA, USA, 2013, doi:10.1176/appi.books.9780890425596.893619. 6. Blazer, D.G. Depression in Late Life: Review and Commentary. Focus (Madison) 2009, 7, 118–136. 7. Djernes, J.K. Prevalence and predictors of depression in populations of elderly: A review. Acta Psychiatr. Scand. 2006, 113, 372–387. 8. Guerra, M.; Prina, A.M.; Ferri, C.P.; Acosta, D.; Gallardo, S.; Huang, Y.; Jacob, K.S.; Jimenez-Velazquez, I.Z.; Llibre Rodriguez, J.J.; Liu, Z.; et al. A comparative cross-cultural study of the prevalence of late life depression in low and middle income countries. J. Affect. Disord. 2016, 190, 362–368. Behav. Sci. 2020, 10, 132 30 of 36 9. Wagenaar, D.B.; Mickus, M.A.; Gaumer, K.A.; Colenda, C.C. Late-Life Depression and Mental Health Services in Primary Care. J. Geriatr. Psychiatry Neurol. 2002, 15, 134–140. 10. Hinton, L.; Zweifach, M.; Tang, L.; Unützer, J.; Oishi, S. Gender Disparities in the Treatment of Late-Life Depression: Qualitative and Quantitative Findings From the IMPACT Trial. Am. J. Geriatr. Psychiatry 2006, 14, 884–892. 11. Mitchell, A.J.; Rao, S.; Vaze, A. do primary care physicians have particular difficulty identifying late-life depression? A meta-analysis stratified by age. Psychother. Psychosom. 2010, 79, 285–294. 12. Conwell, Y.; Van Orden, K.; Caine, E.D. Suicide in older adults. Psychiatr. Clin. North Am. 2011, 34, 451–468. 13. Covey, L.S.; Glassman, A.H.; Stetner, F. Cigarette smoking and major depression. J. Addict. Dis. 1998, 17, 35–46. 14. Brière, F.N.; Rohde, P.; Seeley, J.R.; Klein, D.; Lewinsohn, P.M. Comorbidity between major depression and alcohol use disorder from adolescence to adulthood. Compr. Psychiatry 2014, 55, 526–533. 15. Kim, J.-M.; Stewart, R.; Shin, I.-S.; Yoon, J.-S.; Lee, H.-Y. Lifetime urban/rural residence, social support and late-life depression in Korea. Int. J. Geriatr. Psychiatry 2004, 19, 843–851. 16. Koenig, A.M.; Bhalla, R.K.; Butters, M.A. Cognitive functioning and late-life depression. J. Int. Neuropsychol. Soc. 2014, 20, 461–467. 17. Rajtar, A.; Sałakowski, A.; Rajtar-Zembaty, J.; Starowicz-Filip, A. Executive dysfunction in late-life depression. Psychiatr. Pol. 2017, 51, 705–718. 18. Diniz, B.S.; Butters, M.A.; Albert, S.M.; Dew, M.A.; Reynolds, C.F. Late-life depression and risk of vascular dementia and Alzheimer’s disease: Systematic review and meta-analysis of community-based cohort studies. Br. J. Psychiatry 2013, 202, 329–335. 19. Zanetti, M.; Ballabio, C.; Abbate, C.; Cutaia, C.; Vergani, C.; Bergamaschini, L. Mild cognitive impairment subtypes and vascular dementia in community-dwelling elderly people: A 3-year follow-up study. J. Am. Geriatr. Soc. 2006, 54, 580–586. 20. Panza, F.; Frisardi, V.; Capurso, C.; D’Introno, A.; Colacicco, A.M.; Imbimbo, B.P.; Santamato, A.; Vendemiale, G.; Seripa, D.; Pilotto, A.; et al. Late-Life depression, mild cognitive impairment, and dementia: Possible continuum? Am. J. Geriatr. Psychiatry 2010, 18, 98–116. 21. Butters, M.A.; Young, J.B.; Lopez, O.; Aizenstein, H.J.; Mulsant, B.H.; Reynolds, C.F.; DeKosky, S.T.; Becker, J.T. Pathways linking late-life depression to persistent cognitive impairment and dementia. Dialogues Clin. Neurosci. 2008, 10, 345–357. 22. Crocco, E.A.; Castro, K.; Loewenstein, D.A. how late-life depression affects cognition: Neural mechanisms. Curr. Psychiatry Rep. 2010, 12, 34–38. 23. Blazer, D.G.; Hybels, C.F. Origins of depression in later life. Psychol. Med. 2005, 35, 1241–1252. 24. Chang, S.-C.; Pan, A.; Kawachi, I.; Okereke, O.I. Risk factors for late-life depression: A prospective cohort study among older women. Prev. Med. (Balt. ). 2016, 91, 144–151. 25. Hays, J.C.; Landerman, L.R.; George, L.K.; Flint, E.P.; Koenig, H.G.; Land, K.C.; Blazer, D.G. Social correlates of the dimensions of depression in the elderly. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 1998, 53B, P31–P39. 26. Lue, B.-H.; Chen, L.-J.; Wu, S.-C. Health, financial stresses, and life satisfaction affecting late-life depression among older adults: A nationwide, longitudinal survey in Taiwan. Arch. Gerontol. Geriatr. 2010, 50, S34– S38. 27. Kurlowicz, L.H. Social factors and depression in late life. Arch. Psychiatr. Nurs. 1993, 7, 30–36. 28. Kim, J.-M.; Shin, I.-S.; Yoon, J.-S.; Stewart, R. Prevalence and correlates of late-life depression compared between urban and rural populations in Korea. Int. J. Geriatr. Psychiatry 2002, 17, 409–415. 29. Stern, Y. Cognitive reserve. Neuropsychologia 2009, 47, 2015–2028. 30. Scarmeas, N.; Stern, Y. Cognitive reserve and lifestyle. J. Clin. Exp. Neuropsychol. 2003, 25, 625–633. 31. Leles da Costa Dias, F.; Teixeira, A.L.; Cerqueira Guimarães, H.; Borges Santos, A.P.; Rios Fonseca Ritter, S.; Barbosa Machado, J.C.; Tonidandel Barbosa, M.; Caramelli, P. Prevalence of late-life depression and its correlates in a community-dwelling low-educated population aged 75+ years: The Pietà study. J. Affect. Disord. 2019, 242, 173–179. 32. Gertner, A.K.; Domino, M.E.; Dow, W.H. Risk factors for late-life depression and correlates of antidepressant use in Costa Rica: Results from a nationally-representative longitudinal survey of older adults. J. Affect. Disord. 2017, 208, 338–344. Behav. Sci. 2020, 10, 132 31 of 36 33. Pollock, B.G. The pharmacokinetic imperative in late-life depression. J. Clin. Psychopharmacol. 2005; 25, S19– S23. 34. Solai, L.K.K.; Mulsant, B.H.; Pollock, B.G. Selective serotonin reuptake inhibitors for late-life depression: A comparative review. Drugs and Aging 2001, 18, 355–368. 35. Rapaport, M.H.; Schneider, L.S.; Dunner, D.L.; Davies, J.T.; Pitts, C.D. Efficacy of controlled-release paroxetine in the treatment of late-life depression. J. Clin. Psychiatry 2003, 64, 1065–1074. 36. Schneider, L.S.; Nelson, J.C.; Clary, C.M.; Newhouse, P.; Krishnan, K.R.R.; Shiovitz, T.; Weihs, K. An 8- Week Multicenter, Parallel-Group, Double-Blind, Placebo-controlled study of sertraline in elderly outpatients with major depression. Am. J. Psychiatry 2003, 160, 1277–1285. 37. Tedeschini, E.; Levkovitz, Y.; Iovieno, N.; Ameral, V.E.; Nelson, J.C.; Papakostas, G.I. Efficacy of antidepressants for late-life depression. J. Clin. Psychiatry 2011, 72, 1660–1668. 38. Bohlmeijer, E.; Smit, F.; Cuijpers, P. Effects of reminiscence and life review on late-life depression: A meta- analysis. Int. J. Geriatr. Psychiatry 2003, 18, 1088–1094. 39. Cuijpers, P.; van Straten, A.; Smit, F. Psychological treatment of late-life depression: A meta-analysis of randomized controlled trials. Int. J. Geriatr. Psychiatry 2006, 21, 1139–1149. 40. Bhalla, R.K.; Butters, M.A.; Mulsant, B.H.; Begley, A.E.; Zmuda, M.D.; Schoderbek, B.; Pollock, B.G.; Reynolds, C.F.; Becker, J.T. Persistence of neuropsychologic deficits in the remitted state of late-life depression. Am. J. Geriatr. Psychiatry 2006, 14, 419–427. 41. Koenig, A.M.; Delozier, I.J.; Zmuda, M.D.; Marron, M.M.; Begley, A.E.; Anderson, S.J.; Reynolds, C.F.; Arnold, S.E.; Becker, J.T.; Butters, M.A. Neuropsychological functioning in the acute and remitted states of late-life depression. J. Alzheimer’s Dis. 2015, 45, 175–185. 42. Liao, W.; Zhang, X.; Shu, H.; Wang, Z.; Liu, D.; Zhang, Z. The characteristic of cognitive dysfunction in remitted late life depression and amnestic mild cognitive impairment. Psychiatry Res. 2017, 251, 168–175. 43. Ramscar, M.; Hendrix, P.; Love, B.; Baayen, H. Learning is not decline. Ment. Lex. 2014, 8, 450–481. 44. Ramscar, M.; Hendrix, P.; Shaoul, C.; Milin, P.; Baayen, H. The Myth of cognitive decline: Non-linear dynamics of lifelong learning. Top. Cogn. Sci. 2014, 6, 5–42. 45. Formosa, M. Education and older adults at the university of the third age. Educ. Gerontol. 2012, 38, 114–126. 46. Kelly, M.E.; Loughrey, D.; Lawlor, B.A.; Robertson, I.H.; Walsh, C.; Brennan, S. The impact of cognitive training and mental stimulation on cognitive and everyday functioning of healthy older adults: A systematic review and meta-analysis. Ageing Res. Rev. 2014, 15, 28–43. 47. Kueider, A.M.; Parisi, J.M.; Gross, A.L.; Rebok, G.W. Computerized cognitive training with older adults: A systematic review. PLoS ONE 2012, 7, e40588. 48. Anguera, J.A.; Gunning, F.M.; Areán, P.A. Improving late life depression and cognitive control through the use of therapeutic video game technology: A proof-of-concept randomized trial. Depress. Anxiety 2017, 34, 508–517. 49. Morimoto, S.S.; Wexler, B.E.; Liu, J.; Hu, W.; Seirup, J.; Alexopoulos, G.S. Neuroplasticity-based computerized cognitive remediation for treatment-resistant geriatric depression. Nat. Commun. 2014, 5, 50. Motter, J.N.; Pimontel, M.A.; Rindskopf, D.; Devanand, D.P.; Doraiswamy, P.M.; Sneed, J.R. Computerized cognitive training and functional recovery in major depressive disorder: A meta-analysis. J. Affect. Disord. 2016, 189, 184–191. 51. Seitz, A.R. A New framework of design and continuous evaluation to improve brain training. J. Cogn. Enhanc. 2018, 2, 78–87. 52. Olfers, K.J.F.; Band, G.P.H.; Colzato, L.S.; Hommel, B. The Case of the Leiden “Lifestyle 2030” Study. In Theory-Driven Approaches to Cognitive Enhancement; Colzato, L.S., Ed.; Springer International Publishing:Cham, Germany, 2017; pp. 305–317. 53. Kaushanskaya, M.; Marian, V. The bilingual advantage in novel word learning. Psychon. Bull. Rev. 2009, 16, 705–710. 54. Costa, A.; Hernández, M.; Costa-Faidella, J.; Sebastián-Gallés, N. On the bilingual advantage in conflict processing: Now you see it, now you don’t. Cognition 2009, 113, 135–149. 55. Bialystok, E. Bilingualism Research: Visions For The Future. Available online: https://www.youtube.com/watch?v=Asyf1my3RP4&t=14m26(accessed on 27 August 2020). 56. Bialystok, E.; Craik, F.I.M. Cognitive and linguistic processing in the bilingual mind. Curr. Dir. Psychol. Sci. 2010, 19, 19–23. Behav. Sci. 2020, 10, 132 32 of 36 57. Lehtonen, M.; Soveri, A.; Laine, A.; Järvenpää, J.; de Bruin, A.; Antfolk, J. Is bilingualism associated with enhanced executive functioning in adults? A meta-analytic review. Psychol. Bull. 2018, 144, 394–425. 58. Nichols, E.S.; Wild, C.J.; Stojanoski, B.; Battista, M.E.; Owen, A.M. Bilingualism affords No general cognitive advantages: A population study of executive function in 11,000 people. Psychol. Sci. 2020, 31, 548– 59. Laine, M.; Lehtonen, M. Cognitive consequences of bilingualism: Where to go from here? Lang. Cogn. Neurosci. 2018, 33, 1205–1212. 60. Alladi, S.; Bak, T.H.; Duggirala, V.; Surampudi, B.; Shailaja, M.; Shukla, A.K.; Chaudhuri, J.R.; Kaul, S. Bilingualism delays age at onset of dementia, independent of education and immigration status. Neurology 2013, 81, 1938–1944. 61. Bialystok, E.; Craik, F. I. M.; Freedman, M. Delaying the Onset of Alzheimer Disease. Neurology 2010, 75, 1726–1729. 62. Woumans, E.; Santens, P.; Sieben, A.; Versijpt, J.; Stevens, M.; Duyck, W. Bilingualism Delays Clinical Manifestation of Alzheimer’s Disease. Bilingualism 2015, 18, 568–574. 63. Bialystok, E.; Craik, F. I. M.; Freedman, M. Bilingualism as a Protection against the Onset of Symptoms of Dementia. Neuropsychologia 2007, 45, 459–464. 64. Schweizer, T. A.; Ware, J.; Fischer, C. E.; Craik, F. I. M.; Bialystok, E. Bilingualism as a Contributor to Cognitive Reserve: Evidence from Brain Atrophy in Alzheimer’s Disease. Cortex 2012, 48, 991–996. 65. Poarch, G. J.; van Hell, J. G. Cross-Language Activation in Children’s Speech Production: Evidence from Second Language Learners, Bilinguals, and Trilinguals. J. Exp. Child Psychol. 2012, 11, 419–438. 66. Antoniou, M.; Gunasekera, G. M.; Wong, P. C. M. Foreign Language Training as Cognitive Therapy for Age-Related Cognitive Decline: A Hypothesis for Future Research. Neurosci. Biobehav. Rev. 2013, 37, 2689– 67. van der ploeg, M.; Keijzer, M.; Lowie, W. Methodological Concerns and Their Solutions in Third-Age Language Learning Studies. Dutch J. Appl. Psycholinguist. 2020, in press. 68. Wechsler, D. WAIS-IV : Wechsler Adult Intelligence Scale. San Antonio, TX Psychol. Corp. 2008, 36, 3245– 69. Partington, J. E.; Leiter, G. R. Partington’s Pathways Test. Psychol. Serv. Cebter J. 1949, 1, 11–20. 70. Grant, D. A.; Berg, E. A Behavioral Analysis of Degree of Reinforcement and Ease of Shifting to New Responses in a Weigl-Type Card-Sorting Problem. J. Exp. Psychol. 1948, 38, 404–411. 71. Prior, A.; Macwhinney, B. A Bilingual Advantage in Task Switching. Bilingualism 2010, 13, 253–262. 72. Hallett, P. E. Primary and Secondary Saccades to Goals Defined by Instructions. Vision Res. 1978, 18, 1279– 73. Meyer, S. R. A.; de Jonghe, J. F. M.; Schmand, B.; Ponds, R. W. H. M. The Visual Association Test-Extended: A Cross-Sectional Study of the Performance Validity Measures. Clin. Neuropsychol. 2017, 31, 798–813. 74. Bowie, C. R.; Harvey, P. D.; Administration and interpretation of the Trail Making Test. Nat. Protoc. 2006, 1, 2277-2281. 75. Naismith, S. L.; Rogers, N. L.; Lewis, S. J. G.; Terpening, Z.; Ip, T.; Diamond, K.; Norrie, L.; Hickie, I. B. Sleep Disturbance Relates to Neuropsychological Functioning in Late-Life Depression. J. Affect. Disord. 2011, 132, 139–145. 76. Morin, R. T.; Gonzales, M. M.; Bickford, D.; Catalinotto, D.; Nelson, C.; Mackin, R. S. Impaired Financial Capacity in Late-Life Depression: Revisiting Associations with Cognitive Functioning. J. Int. Neuropsychol. Soc. 2019, 25, 1088–1093. 77. Lamers, F.; Van Oppen, P.; Comijs, H. C.; Smit, J. H.; Spinhoven, P.; Van Balkom, A. J. L. M.; Nolen, W. A.; Zitman, F. G.; Beekman, A. T. F.; Penninx, B. W. J. H. Comorbidity Patterns of Anxiety and Depressive Disorders in a Large Cohort Study: The Netherlands Study of Depression and Anxiety (NESDA). J. Clin. Psychiatry 2011, 72, 342–348. 78. Hugo, J.; Ganguli, M. Dementia and Cognitive Impairment. Epidemiology, Diagnosis, and Treatment. Clin. Geriatr. Med. 2014, 30, 421–442. 79. Riedel, O.; Klotsche, J.; Spottke, A.; Deuschl, G.; Förstl, H.; Henn, F.; Heuser, I.; Oertel, W.; Reichmann, H.; Riederer, P.; et al. Cognitive Impairment in 873 Patients with Idiopathic Parkinson’s Disease: Results from the German Study on Epidemiology of Parkinson’s Disease with Dementia (GEPAD). J. Neurol. 2008, 255, 255–264. Behav. Sci. 2020, 10, 132 33 of 36 80. Kist, N.; Sandjojo, J.; Kok, R. M.; Van Den Berg, J. F. Cognitive Functioning in Older Adults with Early, Late, and Very Late Onset Alcohol Dependence. Int. Psychogeriatrics 2014, 26, 1863–1869. 81. Barker, M. J.; Greenwood, K. M.; Jackson, M.; Crowe, S. F. Cognitive Effects of Long-Term Benzodiazepine Use: A Meta-Analysis. CNS Drugs 2004, 18, 37–48. 82. Drayer, D. E. Lipophilicity, Hydrophilicity, and the Central Nervous System Side Effects of Beta Blockers. Pharmacother. J. Hum. Pharmacol. Drug Ther. 1987, 7, 87–91. 83. Folstein, M. F.; Folstein, S. E.; McHugh, P. R. “Mini-Mental State”. A Practical Method for Grading the Cognitive State of Patients for the Clinician. J. Psychiatr. Res. 1975, 12, 189–198. 84. Nasreddine, Z. S.; Phillips, N. A.; Bédirian, V.; Charbonneau, S.; Whitehead, V.; Collin, I.; Cummings, J. L.; Chertkow, H. The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool for Mild Cognitive Impairment. J. Am. Geriatr. Soc. 2005, 53, 695–699. 85. National Heart, Lung, and B. I. Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools (accessed 27 August 2020) 86. National Institutes of Health. Quality Assessment of Controlled Intervention Studies - NHLBI, NIH https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools (accessed 27 August 2020). 87. Borenstein, M.; Hedges, L. V.; Higgins, J. P. T.; Rothstein, H. R. Introduction to Meta-Analysis; John Wiley & Sons, Ltd: New York, USA; 2009. 88. Combs, J. G.; Ketchen, D. J.; Crook, T. R.; Roth, P. L. Assessing Cumulative Evidence within “Macro” Research: Why Meta-Analysis Should Be Preferred Over Vote Counting. J. Manag. Stud. 2011, 48, 178–197. 89. Friedman, L. Why Vote-Count Reviews Don’t Count (Multiple Letters). Biol. Psychiatry 2001, 49, 161–162. 90. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria. 2020. https://doi.org/R Foundation for Statistical Computing: Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. 91. Lüdecke, D. Esc: Effect size computation for meta analysis. Zenodo 2018, doi:10.5281/zenodo.1249218. 92. Schwarzer, G. Package “Meta.” Proc. - 2014 IEEE/ACM 7th Int. Conf. Util. Cloud Comput. UCC 2014, IEEE: London, UK; 2014. 93. Liberati, A.; Altman, D. G.; Tetzlaff, J.; Mulrow, C.; Gøtzsche, P. C.; Ioannidis, J. P. A.; Clarke, M.; Devereaux, P. J.; Kleijnen, J.; Moher, D. The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration. PLoS Med. 2009, 6, 1–34. 94. Smoski, M. J.; LaBar, K. S.; Steffens, D. C. Relative Effectiveness of Reappraisal and Distraction in Regulating Emotion in Late-Life Depression. Am. J. Geriatr. Psychiatry 2014, 22, 898–907. 95. Nelson, H. E. A Modified Card Sorting Test Sensitive to Frontal Lobe Defects. Cortex 1976, 12, 313–324. 96. Alalade, E.; Denny, K.; Potter, G.; Steffens, D.; Wang, L. Altered Cerebellar-Cerebral Functional Connectivity in Geriatric Depression. PLoS One 2011, 6, e20035. 97. Alexopoulos, G. S.; Hoptman, M. J.; Yuen, G.; Kanellopoulos, D.; Seirup, J. K.; Lim, K. O.; Gunning, F. M. Functional Connectivity in Apathy of Late-Life Depression: A Preliminary Study. J. Affect. Disord. 2013, 149, 398–405. 98. Avila, R.; Moscoso, M. A. A. A.; Ribeiz, S.; Arrais, J.; Jaluul, O.; Bottino, C. M. C. C. Influence of Education and Depressive Symptoms on Cognitive Function in the Elderly. Int. Psychogeriatrics 2009, 21, 560–567. 99. Balardin, J. B.; Vedana, G.; Ludwig, A.; Lima, D. B. de; Argimon, I.; Schneider, R.; Luz, C.; Schröder, N.; Bromberg, E. Contextual Memory and Encoding Strategies in Young and Older Adults with and without Depressive Symptoms. Aging Ment. Health 2009, 13, 313–318. 100. Baudic, S.; Tzortzis, C.; Dalla Barba, G.; Traykov, L. Executive Deficits in Elderly Patients With Major Unipolar Depression. J. Geriatr. Psychiatry Neurol. 2004, 17, 195–201. 101. Beheydt, L. L.; Schrijvers, D.; Docx, L.; Bouckaert, F.; Hulstijn, W.; Sabbe, B. Cognitive and Psychomotor Effects of Three Months of Escitalopram Treatment in Elderly Patients with Major Depressive Disorder. J. Affect. Disord. 2015, 188, 47–52. 102. Boone, K. B.; Lesser, I.; Miller, B.; Wohl, M.; Berman, N.; Lee, A.; Palmer, B. Cognitive Functioning in a Mildly to Moderately Depressed Geriatric Sample: Relationship to Chronological Age. J. Neuropsychiatry Clin. Neurosci. 1994, 6, 267–272. Behav. Sci. 2020, 10, 132 34 of 36 103. Chen, Z.; Lan, G.; Shen, X.; Pan, X.; Chen, X.; LI, J. Relationship of Changes in Cognitive and Depressive Symptoms during Antidepressant Treatment of Individuals with Geriatric Depression and Their Relationship to the APOE Epsilon 4 Allele. Shanghai Arch. Psychiatry 2013, 25, 99–106. 104. Dickinson, W. J.; Potter, G. G.; Hybels, C. F.; McQuoid, D. R.; Steffens, D. C. Change in Stress and Social Support as Predictors of Cognitive Decline in Older Adults with and without Depression. Int. J. Geriatr. Psychiatry 2011, 26, 1267–1274. 105. Egger, K.; Schocke, M.; Weiss, E.; Auffinger, S.; Esterhammer, R.; Goebel, G.; Walch, T.; Mechtcheriakov, S.; Marksteiner, J. Pattern of Brain Atrophy in Elderly Patients with Depression Revealed by Voxel-Based Morphometry. Psychiatry Res. Neuroimaging 2008, 164, 237–244. 106. Elderkin-Thompson, V.; Mintz, J.; Haroon, E.; Lavretsky, H.; Kumar, A. Executive Dysfunction and Memory in Older Patients with Major and Minor Depression. Arch. Clin. Neuropsychol. 2007, 22, 261–270. 107. Elderkin-Thompson, V.; Moody, T.; Knowlton, B.; Hellemann, G.; Kumar, A. Explicit and Implicit Memory in Late-Life Depression. Am. J. Geriatr. Psychiatry 2011, 19, 364–373. 108. Goveas, J. S.; Espeland, M. A.; Hogan, P. E.; Tindle, H. A.; Shih, R. A.; Kotchen, J. M.; Robinson, J. G.; Barnes, D. E.; Resnick, S. M. Depressive Symptoms and Longitudinal Changes in Cognition: Women’s Health Initiative Study of Cognitive Aging. J. Geriatr. Psychiatry Neurol. 2014, 27, 94–102. 109. Kindermann, S. S.; Kalayam, B.; Brown, G. G.; Burdick, K. E.; Alexopoulos, G. S. Executive Functions and P300 Latency in Elderly Depressed Patients and Control Subjects. Am. J. Geriatr. Psychiatry 2000, 8, 57–65. 110. Leal, S. L.; Noche, J. A.; Murray, E. A.; Yassa, M. A. Disruption of Amygdala–Entorhinal–Hippocampal Network in Late-life Depression. Hippocampus 2017, 27, 464–476. 111. Miebach, L.; Wolfsgruber, S.; Frommann, I.; Buckley, R.; Wagner, M. Different Cognitive Complaint Profiles in Memory Clinic and Depressive Patients. Am. J. Geriatr. Psychiatry 2018, 26, 463–475. 112. Park, J. H.; Lee, S. B.; Lee, J. J.; Yoon, J. C.; Han, J. W.; Kim, T. H.; Jeong, H.-G. G.; Newhouse, P. A.; Taylor, W. D.; Kim, J. H.; et al. Depression Plays a Moderating Role in the Cognitive Decline Associated With Changes of Brain White Matter Hyperintensities. J. Clin. Psychiatry 2018, 79, 1–7. 113. Rapp, M. A.; Dahlman, K.; Sano, M.; Grossman, H. T.; Haroutunian, V.; Gorman, J. M. Neuropsychological Differences Between Late-Onset and Recurrent Geriatric Major Depression. Am. J. Psychiatry 2005, 162, 691– 114. Rosano, C.; Perera, S.; Inzitari, M.; Newman, A. B.; Longstreth, W. T.; Studenski, S. Digit Symbol Substitution Test and Future Clinical and Subclinical Disorders of Cognition, Mobility and Mood in Older Adults. Age Ageing 2016, 45, 687–694. 115. Shimada, H.; Park, H.; Makizako, H.; Doi, T.; Lee, S.; Suzuki, T. Depressive Symptoms and Cognitive Performance in Older Adults. J. Psychiatr. Res. 2014, 57, 149–156. 116. Steffens, D. C.; Wagner, H. R.; Levy, R. M.; Horn, K. A.; Krishnan, K. R. Performance Feedback Deficit in Geriatric Depression. Biol. Psychiatry 2001, 50, 358–363. 117. Ramos, S.; Fernández García, Y.; Antón, E.; Casaponsa, A.; Duñabeitia, J. A. Does Learning a Language in the Elderly Enhance Switching Ability? J. Neurolinguistics 2017, 43, 39–48. 118. Wong, P. C. M.; Ou, J.; Pang, C. W. Y.; Zhang, L.; Tse, C. S.; Lam, L. C. W.; Antoniou, M. Language Training Leads to Global Cognitive Improvement in Older Adults: A Preliminary Study. J. Speech, Lang. Hear. Res. 2019, 62, 2411–2424. 119. Massa, E.; Köpke, B.; El Yagoubi, R. Age-Related Effect on Language Control and Executive Control in Bilingual and Monolingual Speakers: Behavioral and Electrophysiological Evidence. Neuropsychologia 2020, 138, 107336. 120. Little, D. Common European Framework of Reference for Languages. In The TESOL Encyclopedia of English Language Teaching; 2018; pp 1–6. 121. Ansaldo, A. I.; Ghazi-Saidi, L.; Adrover-Roig, D. Interference Control in Elderly Bilinguals: Appearances Can Be Misleading. J. Clin. Exp. Neuropsychol. 2015, 37, 455–470. 122. Marian, V.; Blumenfeld, H. K.; Kaushanskaya, M. The Language Experience and Proficiency Questionnaire (LEAP-Q): Assessing Language Profiles in Bilinguals and Multilinguals. J. Speech, Lang. Hear. Res. 2007, 50, 940–967. 123. Craik, F. I. M.; Bialystok, E. Planning and Task Management in Older Adults: Cooking Breakfast. Mem. Cogn., 2006, 34, 1236–1249. Behav. Sci. 2020, 10, 132 35 of 36 124. Estanga, A.; Ecay-Torres, M.; Ibañez, A.; Izagirre, A.; Villanua, J.; Garcia-Sebastian, M.; Iglesias Gaspar, M. T.; Otaegui-Arrazola, A.; Iriondo, A.; Clerigue, M.; et al. Beneficial Effect of Bilingualism on Alzheimer’s Disease CSF Biomarkers and Cognition. Neurobiol. Aging 2017, 50, 144–151. 125. COERLL University of Texas at Austin. Bilingual Language Profile: An Easy-to-Use Instrument to Assess Bilingualism (Spanish-English) sites.la.utexas.edu/bilingual (accessed 27 August 2020). 126. Fernandes, M. A.; Craik, F.; Bialystok, E.; Kreuger, S. Effects of Bilingualism, Aging, and Semantic Relatedness on Memory under Divided Attention. Can. J. Exp. Psychol. 2007, 61, 128–141. 127. Friesen, D. C.; Luo, L.; Luk, G.; Bialystok, E. Proficiency and Control in Verbal Fluency Performance across the Lifespan for Monolinguals and Bilinguals. Lang. Cogn. Neurosci. 2015, 30, 238–250. 128. Anderson, J. A. E.; Mak, L.; Keyvani Chahi, A.; Bialystok, E. The Language and Social Background Questionnaire: Assessing Degree of Bilingualism in a Diverse Population. Behav. Res. Methods 2018, 50, 250– 129. Gubarchuk, I.; Kemper, S. Effects of Aging on the Production of Russian. Discourse Process. 1997, 23, 63–82. 130. Ihle, A.; Oris, M.; Fagot, D.; Kliegel, M. The Relation of the Number of Languages Spoken to Performance in Different Cognitive Abilities in Old Age. J. Clin. Exp. Neuropsychol. 2016, 38, 1103–1114. 131. Johns, B. T.; Sheppard, C. L.; Jones, M. N.; Taler, V. The Role of Semantic Diversity in Word Recognition across Aging and Bilingualism. Front. Psychol. 2016, 7, 1–11. 132. Kousaie, S.; Sheppard, C.; Lemieux, M.; Monetta, L.; Taler, V. Executive Function and Bilingualism in Young and Older Adults. Front. Behav. Neurosci. 2014, 8, 1–12. 133. López Zunini, R. A.; Morrison, C.; Kousaie, S.; Taler, V. Task Switching and Bilingualism in Young and Older Adults: A Behavioral and Electrophysiological Investigation. Neuropsychologia 2019, 133, 107186. 134. Prehn, K.; Taud, B.; Reifergeste, J.; Clahsen, H.; Flöel, A. Neural Correlates of Grammatical Inflection in Older Native and Second-Language Speakers*. Bilingualism 2017, 21, 1–12. 135. Zhang, Z.; Xu, X.; Ni, H. Small Studies May Overestimate the Effect Sizes in Critical Care Meta-Analyses: A Meta-Epidemiological Study. Crit. Care 2013, 17, R2. 136. Ioannidis, J. P. A. Interpretation of Tests of Heterogeneity and Bias in Meta-Analysis. J. Eval. Clin. Pract. 2008, 14, 951–957. 137. Field, A. P. Meta-Analysis of Correlation Coefficients: A Monte Carlo Comparison of Fixed- and Random- Effects Methods. Psychol. Methods 2001, 6, 161–180. 138. Pfenninger, S. E.; Polz, S. Foreign Language Learning in the Third Age: A Pilot Feasibility Study on Cognitive, Socio-Affective and Linguistic Drivers and Benefits in Relation to Previous Bilingualism of the Learner. J. Eur. Second Lang. Assoc. 2018, 2, 1. 139. Bak, T. H.; Long, M. R.; Vega-Mendoza, M.; Sorace, A. Novelty, Challenge, and Practice: The Impact of Intensive Language Learning on Attentional Functions. PLoS One 2016, 11, 1–11. 140. Chu, L. W.; Chiu, K. C.; Hui, S. L.; Yu, G. K. K.; Tsui, W. J. C.; Lee, P. W. H. The Reliability and Validity of the Alzheimer’s Disease Assessment Scale Cognitive Subscale (ADAS-Cog) among the Elderly Chinese in Hong Kong. Ann. Acad. Med. Singapore, 2000, 29 (4), 474–485. 141. Almeida, O. P.; Almeida, S. A. Short Versions of the Geriatric Depression Scale: A Study of Their Validity for the Diagnosis of a Major Depressive Episode According to ICD-10 and DSM-IV. Int. J. Geriatr. Psychiatry 1999, 14, 858–865. 142. Reynolds, W. M.; Kobak, K. A. Reliability and Validity of the Hamilton Depression Inventory: A Paper- and-Pencil Version of the Hamilton Depression Rating Scale Clinical Interview. Psychol. Assess. 1995, 7, 472–483. 143. Surrain, S.; Luk, G. Describing Bilinguals: A Systematic Review of Labels and Descriptions Used in the Literature between 2005-2015. Bilingualism 2019, 1–15. 144. Landreat, M. G.; Vigneau, C. V.; Hardouin, J. B.; Bronnec, M. G.; Marais, M.; Venisse, J. L.; Jolliet, P. Can We Say That Seniors Are Addicted to Benzodiazepines. Subst. Use Misuse 2010, 45, 1988–1999. 145. Ramírez Gómez, D. Language Teaching and the Older Adult: The Significance of Experience; Channel View Books: Bristol, UK; 2016. 146. Mewborn, C. M.; Lindbergh, C. A.; Stephen Miller, L. Cognitive Interventions for Cognitively Healthy, Mildly Impaired, and Mixed Samples of Older Adults: A Systematic Review and Meta-Analysis of Randomized-Controlled Trials. Neuropsychol. Rev. 2017, 27, 403–439. 147. Li, P.; Legault, J.; Litcofsky, K. A. Neuroplasticity as a Function of Second Language Learning: Anatomical Changes in the Human Brain. Cortex 2014, 58, 301–324. Behav. Sci. 2020, 10, 132 36 of 36 148. Matsumoto, D. Exploring Third-Age Foreign Language Learning from the Well-Being Perspective: Work in Progress. Stud. Self-Access Learn. J. 2019, 10, 111–116. 149. Gullifer, J. W.; Titone, D. Characterizing the Social Diversity of Bilingualism Using Language Entropy. Bilingualism 2020, 23, 283–294. 150. Bialystok, E.; Craik, F. I. M.; Ryan, J. Executive Control in a Modified Antisaccade Task: Effects of Aging and Bilingualism. J. Exp. Psychol. Learn. Mem. Cogn. 2006, 32, 1341–1354. 151. Avorn, J.; Everitt, D. E.; Weiss, S. Increased Antidepressant Use in Patients Prescribed β-Blockers. JAMA J. Am. Med. Assoc. 1986, 255, 357–360. 152. Verbeek, D. E. P.; Van Riezen, J.; De Boer, R. A.; Van Melle, J. P.; De Jonge, P. A Review on the Putative Association Between Beta-Blockers and Depression. Heart Fail. Clin. 2011, 7, 89–99. 153. Pliatsikas, C. Understanding Structural Plasticity in the Bilingual Brain: The Dynamic Restructuring Model. Biling. Lang. Cogn. 2019, 23, 459–471. 154. Kroll, J. F.; Dussias, P. E.; Bice, K.; Perrotti, L. Bilingualism, Mind, and Brain. Annu. Rev. Linguist. 2015, 1, 377–394. 155. Bialystok, E.; Abutalebi, J.; Bak, T. H.; Burke, D. M.; Kroll, J. F. Aging in Two Languages: Implications for Public Health. Ageing Res. Rev. 2016, 27, 56–60. https://doi.org/10.1016/j.arr.2016.03.003. © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Journal

Behavioral Sciences – Multidisciplinary Digital Publishing Institute

Published: Aug 31, 2020

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research

Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research

Exploring Language Learning as a Potential Tool against Cognitive Impairment in Late-Life Depression: Two Meta-Analyses and Suggestions for Future Research

References (172)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies