Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/jmir-publications/mitigating-feelings-of-loneliness-and-depression-by-means-of-web-based-wl31cjhskT
- Publisher
- JMIR Publications
- Copyright
- Copyright © The Author(s). Licensed under Creative Commons Attribution cc-by 4.0
- ISSN
- 2561-7605
- DOI
- 10.2196/36515
- Publisher site
- See Article on Publisher Site
Abstract
Background: Physical activity (PA) is associated with benefits, such as fewer depressive symptoms and loneliness. Web- and print-based PA interventions can help older individuals accordingly. Objective: We aimed to test the following research questions: Do PA interventions delivered in a web- or print-based mode improve self-reported PA stage of change, social-cognitive determinants of PA, loneliness, and symptoms of depression? Is subjective age a mediator and stage of change a moderator of this effect? Methods: Overall, 831 adults aged ≥60 years were recruited and either allocated to a print-based or web-based intervention group or assigned to a wait-list control group (WLCG) in 2 community-based PA intervention trials over 10 weeks. Missing value imputation using an expectation-maximization algorithm was applied. Frequency analyses, multivariate analyses of variance, and moderated mediation analyses were conducted. Results: The web-based intervention outperformed (47/59, 80% of initially inactive individuals being adopters, and 396/411, 96.4% of initially active individuals being maintainers of the recommended PA behavior) the print-based intervention (20/25, 80% of adopters, and 63/69, 91% of maintainers) and the WLCG (5/7, 71% of adopters; 141/150, 94% of maintainers). The pattern regarding adopters was statistically significant (web vs print Z=–1.94; P=.02; WLCG vs web Z=3.8367; P=.01). The pattern was replicated with stages (χ =79.1; P<.001; contingency coefficient 0.314; P<.001); in the WLCG, 40.1% (63/157) of the study participants moved to or remained in action stage. This number was higher in the groups receiving web-based (357/470, 76%) or print-based interventions (64/94, 68.1%). A significant difference was observed favoring the 2 intervention groups over and above the WLCG (F =4.778; P<.001; η =0.098) and a significant interaction of time and group (F =2.778; P<.001; 19, 701 19, 701 η =0.070) for predictors of behavior. The effects of the interventions on subjective age, loneliness, and depression revealed that both between-group effects (F =8.668; P<.001; η =0.018) and the interaction between group and time were significant (F 3, 717 3, =6.101; P<.001; η =0.025). In a moderated mediation model, both interventions had a significant direct effect on depression in comparison with the WLCG (web-based: c′ path −0.86, 95% CI −1.58 to −0.13, SE 0.38; print-based: c′ path −1.96, 95% CI −2.99 to −0.92, SE 0.53). Furthermore, subjective age was positively related to depression (b path 0.14, 95% CI 0.05-0.23; SE https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al 0.05). An indirect effect of the intervention on depression via subjective age was only present for participants who were in actor stage and received the web-based intervention (ab path −0.14, 95% CI −0.34 to −0.01; SE 0.09). Conclusions: Web-based interventions appear to be as effective as print-based interventions. Both modes might help older individuals remain or become active and experience fewer depression symptoms, especially if they feel younger. Trial Registration: German Registry of Clinical Trials DRKS00010052 (PROMOTE 1); https://tinyurl.com/nnzarpsu and DRKS00016073 (PROMOTE 2); https://tinyurl.com/4fhcvkwy International Registered Report Identifier (IRRID): RR2-10.2196/15168 (JMIR Aging 2022;5(3):e36515) doi: 10.2196/36515 KEYWORDS physical activity; older adults; intervention; loneliness; depression; eHealth; mobile health; mHealth behavior if they received the web-based or print-based Introduction intervention than those who received the control condition. Web-Based and Printed Intervention Material A previous study by Boekhout et al [8] revealed the benefits of a printed delivery method compared with a web-based version. Loneliness is a key element, along with lifestyle factors such Specifically, the authors found higher participation and lower as physical activity (PA), which is interrelated with health and attrition rates in this group [8]. Golsteijn et al [9] compared well-being [1-3]. Although the concept of loneliness has a long printed materials with web-based materials in terms of history, many concerns exist that modern times increase social cost-effectiveness and cost-utility to promote PA among adults isolation among older people [3,4]. Since the beginning of the aged ≥50 years. The results revealed that the print-based material COVID-19 (SARS-CoV-2) pandemic, social isolation and was most cost-effective in terms of increasing PA and could loneliness have received heightened attention [1-4]. Reasons also contribute to better overall health at the population level for elevated concerns related to loneliness during the COVID-19 [9]. However, little is known about the effectiveness in terms pandemic were seen because of the required social distancing of loneliness and social-cognitive predictors, what actually (ie, because of distancing rules, citizens were not allowed to be explains the effects of the intervention, and in whom and how in close physical contact with others or generally to come the intervention works [10]. Depression can be an important together). In addition, people limited their personal contacts component of mental health [1], whereas a central factor of and followed stay-at-home orders and face mask mandates in well-being and successful aging is feeling subjectively fit [7]. public [2-4]. Steps are needed to bridge the gap between the Thus, this study investigated whether a web-based or print-based necessary actions for public health, individual health, and PA intervention improved outcomes such as social-cognitive well-being. Such bridging can be done by means of web-based predictors of PA behavior change, subjective age, feelings of and print-based interventions in comparison with no support loneliness, and symptoms of depression in comparison with a (ie, a wait-list control group [WLCG]). control group. Furthermore, we examined whether mediating Dickens et al [5] performed a systematic review of interventions and moderating mechanisms exist. Conceivable mechanisms targeting social isolation among older adults. They found that will be outlined in the following sections to set the stage for 86% of the interventions aimed at supporting activities (social this study. activities and PAs) were effective [5]. Specifically, these Potential of PA Interventions activities were comprised of group and psychosocial accomplishments and included besides exercise and PA also PA is imperative for health and well-being at any age and is arts, discussion rounds, therapeutic writing, group therapy, increasing in importance with older age [11,12]. Approximately reading to children, lectures, assistance with organizing social half of the population will be aged >60 years by 2030 [13]. behavior, outings, mutual help networks as well as leisure and Consequently, it is important to improve the health of this cultural events with different durations [5]. Activities, especially population. Regular PA, particularly cardiovascular training social components and PA, are key to preventing or overcoming (also called endurance training), is considered to have enormous social isolation and loneliness [4,6]. However, only 25% of potential for maintaining and improving the health and internet training interventions have revealed a successful well-being of older adults [14,15]. Following the World Health reduction in social isolation among older adults [5]. Organization (WHO) recommendations, PA should be conducted for ≥150 minutes per week with moderate to vigorous PA In conclusion, the evidence demonstrates that interventions (MVPA) in bouts of at least 10 minutes to improve and maintain fostering physical exercise and PA can improve mood; increase health [13]. physical, social, and cognitive activities; and decrease social isolation [7]. However, little is known about the delivery mode Cross-sectionally, more PA is related to better health and vice of the intervention (eg, the comparison of internet-based training versa, and the improvement of a healthy lifestyle has been interventions and traditional print-based interventions), as well demonstrated to pay off in terms of increasing or recovering as the mechanisms that may explain possible differential effects. health [16,17]. Olson and McAuley [18] demonstrated the Therefore, this study addressed this open question. The research effectiveness of an intervention, including walking exercise question was whether more adults reported changing their PA (endurance training) and theory-based group workshops, aimed https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al at improving the PA level of older adults in the short and long disease”). Risk perception in itself is insufficient to enable a term. person to form an intention. Rather, it sets the stage for a contemplation process and further elaboration of thoughts on An important limitation of PA programs for older adults is that consequences and competencies. Similarly, positive outcome they are often unattractive to older adults [19]. This might be expectancies (eg, “If I exercise five times per week, I will reduce overcome by addressing the individual-level characteristics and my cardiovascular risk”) are chiefly seen as important in the (technology-based) preferences of participants [20]. Concerning motivation phase when a person balances the pros and cons of the overall uptake of PA and sports offerings, demographic and certain behavioral outcomes. Furthermore, one needs to believe socioeconomic individual-level differences have been shown in one’s capability to perform the desired action (perceived to be relevant in past studies [21-23]. Those who are already self-efficacy; eg, “I am capable of adhering to my exercise actively involved in PA and sports are more likely to participate schedule despite the temptation to watch TV”). Perceived [23-25]. However, some people may also experience that being self-efficacy operates in concert with positive outcome active operates as a barrier to adopting activities such as a new expectancies, both of which contribute substantially to the physical exercise program [26], showing that previous behavior formation of intention. Both beliefs are needed to form is yet another predictor that depends on individual intentions to adopt difficult behaviors, such as regular physical circumstances. A study of older adults’ specific requirements exercise. for PA class meetings also revealed sex-specific differences. For example, men, in contrast to women, were more critical of After forming an intention, the volitional phase is entered. Once group activities [27]. Further identification of how and why a person is inclined to adopt a particular health behavior, the interventions work can help the development and organization good intention must be transformed into detailed instructions of attractive future health interventions [19,20]. A theoretical on how to perform the desired action. As soon as an action is framework that might explain the differences based on baseline initiated, it must be maintained. This is not achieved through a characteristics such as previous experience is described in the single act of will but involves self-regulatory skills and following section. strategies. Thus, the postintentional phase should be further broken down into more proximal factors, such as planning, Theory-Based Interventions and Social-Cognitive action control, social support, and recovery self-efficacy. Predictors of PA Behavior Change Social support is a factor that reflects the barriers and resources Research comparing the effectiveness of theory-based and part of the HAPA model: support represents a resource, and the non–theory-based health behavior change interventions has lack of it can be a barrier to adopting or maintaining health demonstrated a higher potential for theory-based approaches to behaviors. Instrumental, emotional, and informational social effectively promote PA [28-30], although not consistently [31]. support can enable the adoption and continuation of behaviors However, overall, it should be noted that health behavior change [35]. The theoretical assumptions not only improve the interventions to improve PA are very heterogeneous with regard prediction of behavior but also allow for designing of to theoretical approaches, designs, and effectiveness. In addition, interventions more effectively by tailoring the intervention some interventions have only been found to produce small to components to the needs of the recipient and, finally, enhance moderate effects [32-34]. For example, an aggregated effect of participation. The relevant factors are described in the following Cohen d=0.27 was determined by Rhodes et al [34] in a sections. high-level overview of published reviews of the literature, which has been interpreted as small but meaningful. This shows that Tailored Web-Based Intervention theoretical frameworks that take further relevant parameters Tailoring is a key aspect of making interventions more effective, and pathways into account are needed for the design of PA not only by considering the users’ stage of change but also by interventions. matching the users’ needs. For instance, such needs can be that participants prefer self-monitoring and activity tracking as Social-cognitive variables are imperative for predicting active components of their intervention (eg, by digital formats, as behavior change [35]. Knowledge of such variables enables the found by Powell et al [36]). design of interventions. For example, key social-cognitive variables are described in the Health Action Process Approach Digital modes have much more potential than paper-based (HAPA) [35]. The HAPA is a theory that organizes different intervention modes as they provide more options for social-cognitive variables into a meaningful structure [35]. The personalization. At the same time, information can be delivered HAPA has two layers: a continuum layer with social-cognitive in different forms, including textual, visual, and audiovisual variables and a stage layer with the stages of change. The HAPA information to suit individual preferences and abilities [37]. assumes three different stages of change: the nonintenders stage However, when preferences are considered, older people in with its processes that lead to a behavioral intention, the particular like print formats better [38] and accordingly might intenders stage with postintentional volition processes that lead benefit more from it. However, this requires more systematic to the actual health behavior, and the action stage where the research. goal behavior is performed. The tailoring of interventions is a method that aims to meet the Within different stages, different patterns of social-cognitive needs of all individuals more appropriately. However, meeting predictors may emerge. In the nonintenders stage, a person all these different needs is typically challenging. Therefore, it develops the intention to act. In this phase, risk perception is is necessary to evaluate whether all individuals benefit equally. seen as a distal antecedent (eg, “I am at risk for cardiovascular https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al For instance, in a previous study [39], it was found that younger are better off and more optimistic [48]. Previous studies participants who were not sufficiently physically active before have demonstrated that interventions can improve subjective the study (nonintenders and intenders in comparison with actors) age and general health status or even reverse frailty [47]. The found the intervention useful. In another study [40], the printed question remains whether the effect of a PA intervention is method was more effective than the internet method in mediated by subjective age; thus, this study investigated this in participants with a high baseline intention for PA (intenders). further detail. Thus, the question remains of whether the intervention is Interrelations and Stage of Change moderated by the stage of change in endurance activities. Other Loneliness and mental health issues, such as depression, are needs may be interrelated with sociodemographic characteristics interrelated [1]. Fortuna et al [10] summarized that older adults such as age, which necessitates further elaboration, as will be can benefit from digital services to overcome their mental health described in the following sections. limitations (such as loneliness and depression). Moreover, PA Aging, Loneliness, and Subjective Age can help reduce depression and loneliness using mastery Aging is typically stereotyped as feeling lonely. However, experience and self-regulation with regard to physical loneliness is not related to older age but the opposite—younger perceptions and repairing interpersonal skills and relationships cohorts feel more lonely than older cohorts [1]. The aging [6]. Accordingly, the PA intervention group (IG) allocation population is at higher risk of other health-related challenges should reduce the likelihood of depressive symptoms and [1,3,4]. Aging processes and the health of older adults are highly loneliness. important. Many older adults experience more health limitations However, much is still unresearched, such as whether and an increased burden, such as falling upon their caregivers internet-based services are as good as, or better than, traditional [41]. In addition, older adults might have the highest risk of services for older individuals’ mental health. Although the inactive lifestyles because of their reduced functioning [42]. advantages are obvious, the effects on outcomes such as Aging theories posit that older adults prefer to exercise with loneliness and symptoms of depression still need more other individuals instead of exercising alone [43]. Accordingly, systematic attention, which will be addressed in this study. blended web-based and print interventions for older individuals With respect to intervention studies, it is assumed that the promoting PA proved to be effective [44-46] as web-based assignment of participants to specific study arms with different materials would typically be used more for unaccompanied forms of content would have an effect on symptoms of modes. However, whether print and web-based materials are depression and that this effect is mediated by subjective age beneficial for older adults’ health (eg, symptoms of depression), and moderated by the stage of change in endurance training well-being (or the opposite, eg, loneliness), health behavior, (Figure 1). and its predictors requires further investigation. Furthermore, the question remains as to whether the same Typically, calendrical and subjective ages are distinguished relationship with loneliness as a dependent variable would be [47,48]. Calendrical age is determined by the date of birth [38]. feasible (Figure 2). Accordingly, research is required regarding In contrast, if a person is asked how old they feel, then the the key question of whether the intervention effects on loneliness perceived or subjective age can be determined [47]. The latter and symptoms of depression depend on subjective age and is associated with health status and well-being, as well as with whether the stage of change for endurance activities affects this behavioral, cognitive, and biological processes, including frailty effect. Thus, the hypotheses described in the following sections [47]. were investigated. Although calendrical age cannot be changed, subjective age contains many options for interventions: individuals who feel https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Figure 1. Proposed moderated mediation model for depression. CES-D: Center for Epidemiologic Studies Depression; T0: time point 0; T1: time point Figure 2. Proposed moderated mediation model for loneliness. T0: time point 0; T1: time point 1. The intervention effect on symptoms of depression (Center Goals of This Study for Epidemiologic Studies Depression [CES-D] score) is The goal of this study was to test the following research mediated by subjective age, and this mediation is moderated questions: do interventions delivered in a web or print mode by the stage of change in endurance activities (moderated improve self-reported PA stage of change, social-cognitive mediation). determinants of PA, feelings of loneliness, and symptoms of depression, and in this effect, does subjective age act as a Methods mediator and stage of change act as a moderator? Overview The following hypotheses were tested: The PROMOTE study comprised 2 cohorts: PROMOTE 1 and The proportion of older adults who self-report a PA 2. In this study, web- and print-based programs to promote PA behavior change is higher in the web-based and print-based in community-dwelling older adults were developed, analyzed, PA interventions than in the respective control conditions. and evaluated according to multiple theoretical models and Compared with the control condition, the web-based and intervention effects using randomized intervention trials [49-54]. print-based PA interventions improve social-cognitive These were conducted as part of the interdisciplinary Physical predictors of PA behavior change, subjective age, feelings Activity and Health Equity: Primary Prevention for Healthy of loneliness, and symptoms of depression. Aging research network [55]. The intervention’s effect on feelings of loneliness is The first trial (2015-2018, PROMOTE 1) compared the effects mediated by subjective age, and this mediation is moderated of 2 web-based interventions with a wait-listed control condition, by the stage of change for endurance activities (moderated whereas the second trial (2018-2021, PROMOTE 2) compared mediation). adapted versions of the web-based interventions (the program https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al was adapted to initially inactive older adults) with a print-based The main inclusion criteria were being aged 65 to 75 years in PA intervention. For the analyses in this study, the groups PROMOTE 1 and ≥60 years in PROMOTE 2, as well as living receiving the web-based intervention in PROMOTE 1 and 2 independently, having basic knowledge of German, being able were combined. The plan for pooling the data of the 2 to walk without a walking aid, participation in study assessments intervention studies is described in the study protocol of and weekly group meetings without external assistance, and PROMOTE 2 [51]. Accordingly, measures with the intervention providing informed consent. design were taken to synchronize the different interventions The exclusion criteria were as follows: a medical condition or from the beginning of PROMOTE 2: interventions did not diagnosis prohibiting PA, severe visual or other impairments, significantly differ in their content and with the levels of implanted cardiac devices, or occasional syncopal episodes recommended activity levels [51]. Measures were matched for leading to exclusion or cognitive impairment (Mini-Mental PROMOTE 1 and 2 to pool data from both trial periods for the State Examination [MMSE] <25 in PROMOTE 1 and MMSE overarching analyses. second edition: MMSE-2: Brief Version [MMSE-2:BV] score <13 in PROMOTE 2). Individuals were excluded from the study Ethics Approval if they were planning a vacation during the intervention period, PROMOTE 1 was approved by the Ethics Committee of the had certain medical conditions or severe health impairments, Technical University of Chemnitz, Faculty of Behavioral and or did not have a mobile device or internet access. As the results Social Sciences, on July 14, 2015, with the ethics approval of the first study indicated that predominantly already active number: V-099-17-HS-CVR-PROMOTE-03072015. Ethics individuals participated in the study, the exclusion criteria for approval for PROMOTE 2 was obtained from the Medical PROMOTE 2 were modified, and individuals were excluded if Association in Bremen on July 3, 2018, with the ethics approval they reported being regularly physically active for at least 2.5 number 635. The trials were conducted in accordance with the hours per week for >1 year before the start of the study. Potential ethical principles of the American Psychological Association study participants for PROMOTE 2 were excluded if they had and the 1964 Declaration of Helsinki and its later amendments already participated in PROMOTE 1 [52]. of comparable ethical standards. All participants were fully informed about the study and provided informed consent. Finally, participants were randomly assigned to the following study arms: Recruitment In PROMOTE 1 (N=589), to either a web-based Detailed information regarding data collection (recruitment and intervention with subjective PA self-monitoring (211/589, randomization strategies) can be obtained from the studies by 35.8%), web-based intervention with subjective and Muellmann et al [49] and Pischke et al [51]. Briefly, in 2016 objective PA self-monitoring (198/589, 33.6%), or WLCG (PROMOTE 1) and 2018 (PROMOTE 2), random samples of (180/589, 30.6%) [49] n=8299 older adults aged 65 to 75 years and n=3492 older adults In PROMOTE 2 (N=242), to a print-based intervention aged ≥60 years and living independently (without assisted with subjective PA self-monitoring via a printed PA living) were selected by the residents’ registration offices from pyramid (113/242, 46.7%) and web-based intervention with municipalities in the Bremen metropolitan region and invited subjective PA self-monitoring via a web-based PA pyramid to participate by mail. In addition, both study phases were (129/242, 53.3%); approximately 29.5% (38/129) of the promoted in the local press, as well as via prior discussions with latter were randomly selected and received a PA tracker the research staff with an offer to enroll voluntarily. Eligibility (objective PA self-monitoring) in addition [51] for study participation, which is described in detail in the published study protocols [49,51], was determined through In total, 831 individuals were randomized into one of these three computer-assisted telephone interviews with trained study groups: web-based PA intervention, print-based PA intervention, nurses. or WLCG. Further details are outlined in the flow chart in Figure https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Figure 3. Study flow chart. minutes per week of vigorous training or a combination of both Procedure and Interventions moderate and vigorous training intensities [59]. The PA interventions in this study were developed on the basis After randomization, a baseline assessment (time point 0 [T0]) of the self-regulation theory [56,57]; behavior change techniques was conducted. Following T0, the IGs in PROMOTE 1 received (ie, goal setting, planning, social support, and feedback) [58]; a print-based intervention in the form of a booklet, which was and (in PROMOTE 2) in a cocreative process with individuals tailored to the individual baseline PA levels. Feedback was of the target group and stakeholders, such as exercise instructors, tailored to the baseline motivational stage (nonintention, leaders of older adults’ facilities, and managers of older adults’ intention, or action) to engage in the recommended PA. In homes. The intervention aimed to improve overall addition, the material was tailored based on sex: pictures of men self-monitoring capabilities regarding PA and enable transfer for male participants and pictures of women for female for the time after the intervention. participants were used to model the recommended exercises. According to the PA recommendations of the WHO and the The web-based materials offered in the corresponding study American College of Sports Medicine [13], older adults arms in PROMOTE 1 and 2 included access to web-based randomized to the IGs were advised to engage in physical materials that contained and displayed the same information on exercises. The recommendations included suggestions to exercises for balance, flexibility, endurance, and strength, as improve balance (twice per week), flexibility (twice per week), for the print-based version. In addition, for PROMOTE 2, a and strength (twice per week for the 8 major muscle groups). print-based PA diary was developed in the form of an In addition, the participants were instructed to engage in at least expert-driven approach and contained all exercises; provided 150 minutes per week of moderate endurance training or 75 the option to enter data on performed exercises; and, thus, visualize personal progress. The web-based intervention from https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al PROMOTE 1 was adapted based on feedback obtained during Intention to engage in regular endurance and strength training the first study period. An additional web-based option (PA was assessed with 2 items (“I intend to engage in strenuous tracker) to capture the daily step count was included for the endurance training for at least 75 minutes per week and strength- randomly selected subgroups in both study arms (PROMOTE and balance training twice a week” and “I intend to engage in 1 and 2). moderate endurance training for at least 150 minutes per week and strength- and balance training twice a week”). All IGs were encouraged to use the material and engage in the recommended PA over 10 weeks. These were accompanied by These items were based on previous literature [60,61]. Both weekly group meetings conducted and moderated by trained items were measured at T0 and T1 on a 7-point Likert scale staff members. During these meetings, questions concerning ranging from 1=completely disagree to 7=completely agree. the program could be raised. In addition, theoretical inputs for The 2 items were kept separate as they had discriminant validity healthy aging were provided. Moreover, physical exercises were (Spearman ρ at T0=0.410; Spearman ρ at T1=0.467; P<.001). performed together with feedback from the participants The retest reliability was Spearman ρ T0 to T1 of 0.531 and regarding their exercise practice. At the same time, social 0.378, respectively (P<.001). interactions among the participants and their contacts with the Outcome expectancies, as suggested by Lippke et al [62] and study team were facilitated. After 10 weeks of group meetings, Schwarzer et al [63], were measured using 4 items in total at 2 a follow-up assessment for time point 1 (T1) was conducted 12 measurement time points. A total of 2 items measured positive weeks after the baseline assessment. Several collected variables outcome expectancies (“If I engage in 150 minutes of served as the basis for this study. More information on the moderately strenuous or 75 minutes of strenuous endurance interventions and procedures can be found in previous exercise of strength and balance training twice per week, it is publications [49-51,53]. good for my health.” and “If I engage in 150 minutes of moderately strenuous or 75 minutes of strenuous endurance Used Instruments exercise of strength and balance training twice per week, it Adherence was measured according to the WHO makes me feel better afterwards.”). recommendations of ≥150 minutes per week of MVPA in bouts of at least 10 minutes. The daily minutes for PA in terms of The remaining 2 items focused on negative outcome MVPA were assessed by asking the study participants what expectancies (“If I engage in 150 minutes of moderately activities they performed in bouts of at least 10 minutes. Minutes strenuous or 75 minutes of strenuous endurance exercise of per week for MVPA in the bouts were derived by multiplying strength and balance training twice per week, it takes too long.” the daily average minutes in the 10-minute bouts by 7. “If I engage in 150 minutes of moderately strenuous or 75 Subsequently, this measure was dichotomized as meeting or minutes of strenuous endurance exercise of strength and balance not meeting the WHO recommendations. This resulted in a training twice per week, it is too costly.”). All 4 items were dichotomous variable with 1=does not meet 150 minutes of measured on a 7-point Likert scale ranging from 1=completely MVPA recommendation and 2=meets 150 minutes of MVPA disagree to 7=completely agree. at T0 and T1. To determine adherence over time, a categorical The 2 items measuring positive outcome expectancies were variable regarding the change in meeting the recommended 150 kept separate as they had rather discriminant validity (Spearman minutes of MVPA was computed by subtracting the baseline ρ at T0=0.693; Spearman ρ at T1=0.703; P<.001). The retest value from the 12-week follow-up value. The resulting variable reliability was Spearman ρ T0 to T1 of 0.425 and 0.508, indicated whether study participants remained active or inactive respectively (P<.001). The 2 items measuring negative outcome (0), fell back into not meeting the recommendation anymore expectancies were also kept separate as they had rather (−1), or became active (1). discriminant validity (Spearman ρ at T0=0.474; Spearman ρ at To assess the stage of PA behavior, participants were asked T1=0.443; P<.001). The retest reliability was Spearman ρ T0 whether they had performed ≥150 minutes of endurance exercise to T1 of 0.339 and 0.441, respectively (P<.001). per week (eg, fast walking, walking, biking, and swimming) at Self-efficacy was measured with 5 items, in total, at both 2 measurement time points (T0 and T1). Participants were asked measurement time points T0 and T1 [61,64]. A single item was to respond on a 5-point Likert scale: 1=“No, and I do not intend used to assess task self-efficacy (“I am confident that I can to start” (precontemplation stage or nonintenders); 2=“No, but engage in 150 minutes of moderately strenuous or 75 minutes I am considering it” (contemplation stage or nonintenders); of strenuous endurance exercise and strength and balance 3=“No, but I seriously intend to start” (preparation stage or training twice a week, even if it gets difficult.”). intenders); 4=“Yes, but only for a brief period of time” (action stage or actors); and 5=“Yes, and for a long period of time” A total of 2 items assessed maintenance self-efficacy (“I am (maintenance stage or actors). confident that I can engage in 150 minutes of moderately strenuous or 75 minutes of strenuous endurance exercise and The stage item is based on items used by Lippke et al [60]. For strength and balance training twice a week, even if it takes long, this study, participants in the precontemplation and until it is a habit.” and “I am confident that I can engage in 150 contemplation stages were categorized as nonintenders, minutes of moderately strenuous or 75 minutes of strenuous participants in the preparation stage were categorized as endurance exercise and strength and balance training twice a intenders, and participants in the action and maintenance stage week, even if I am worried or face problems, e.g., scheduling were categorized as actors. difficulties.”). https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 8 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al In addition, recovery self-efficacy was measured by 2 items (“I To assess perceived loneliness at T0 and T1, the item I felt am confident that I can engage in 150 minutes of moderately lonely was used from the CES-D scale [67]. The item was strenuous or 75 minutes of strenuous endurance exercise and measured on a 4-point Likert scale from 0=rarely or never (<1 strength and balance training twice a week, even if I postponed day) to 1=some or little of the time (1-2 days), 2=often or a my plans several times.” and “I am confident that I can engage moderate amount of time (3-4 days), and 3=most of the time in 150 minutes of moderately strenuous or 75 minutes of (5-7 days). The retest reliability was Spearman ρ T0 to T1 of strenuous endurance exercise and strength and balance training 0.597 (P<.001). twice a week, even If I suspended several times.”). Sociodemographic data were collected using a questionnaire All 5 items were measured on a 7-point Likert scale ranging administered before the intervention (at the baseline level). The from 1=completely disagree to 7=completely agree. For this questionnaire was formulated according to the German Health study, the sum of all 5 items was calculated (Cronbach α at Interview and Examination Survey for Adults [68]. The collected T0=.884; Cronbach α at T1=.897). The retest reliability was variables included date of birth, sex of the participants (male Spearman ρ at T0 to T1 of 0.475 (P<.001). or female), height (in cm), and weight (in kg). To further assess social-cognitive predictors, planning was In addition to the date of birth, perceived age was measured measured using 6 items. The items were adapted for this study using an open-ended question. The participants were asked, from those used in previous studies on PA [65]. “How old do you feel?” Perceived age was assessed at T0 and T1. The retest reliability was Spearman ρ T0 to T1 of 0.826 Three items measured action planning: “For the next month, I (P<.001). Furthermore, country of birth, mother tongue, family have already planned where I will be physically active,” “For status, living alone, number of children, qualification, the next month, I have already planned how I will be physically educational level, and employment status were assessed. active,” and “For the next month, I have already planned when and how often I will be physically active.”. Employment status was measured with a single item taken from a questionnaire assessing demographic and sociostructural data Furthermore, three items assessed the construct of coping from German older adults and adapted for this study [69]. planning, respectively: “For the next month, I have already Qualification and educational level were measured with 2 items planned when I have to take care not to suspend,” “For the next and aggregated based on the 2016 version (volume 17) of the month, I have already planned what I can do in difficult International Standard of Education [70]. situations to stick to my intentions,” and “For the next month, I have already planned how I can remain physically active even BMI was calculated using height and weight and categorized if there are barriers.” into underweight, normal weight, overweight, and obese, according to the WHO BMI classification for adults aged ≥20 All 6 items were measured on a 7-point Likert scale ranging years [71]. from 1=completely disagree to 7=completely agree. For this study, the sum of all 6 items was calculated (Cronbach α at All the used instruments were validated before and are described T0=.932; Cronbach α at T1.899). The retest reliability was in the study protocols [49,51], as well as in previous publications Spearman ρ T0 to T1 of 0.492 (P<.001). [50,52-54]. Habits were measured using two items at two measurement Analysis Sample time points [66]: “Engaging in the recommended endurance, IG allocations from PROMOTE 1 and 2 were summarized strength and balance training is something that has become my within a pooled IG variable that included the following three habit.” and “Engaging in the recommended endurance, strength, categories: a=WLCG from PROMOTE 1 (reference), and balance training is something that I do without thinking b=web-based IG from PROMOTE 1 and 2, and c=print-based about it.” IG from PROMOTE 2. Both items were measured on a 7-point Likert scale from 1 Only the participants who completed the baseline assessment completely disagree to 7 completely agree (Spearman ρ at (T0) were included in the analysis. In PROMOTE 1, T0=0.474; Spearman ρ at T1=0.443; P<.001). The retest participants’ cognitive status was assessed using the MMSE reliability was Spearman ρ T0 to T1 of 0.339 and 0.441, [72] 1 week before the start of the intervention phase. In respectively (P<.001). PROMOTE 2, participants’ cognitive status was assessed using MMSE-2:BV [73] during the first weekly group meeting (ie, Symptoms of depression were measured using the CES-D [67] the start of the intervention phase). Participants who scored <25 scale both at T0 and T1. The scale comprises 20 items with a points on the MMSE or <13 points on the MMSE-2:BV were possible sum score range of 0 to 60. Each item was measured excluded from the analysis. Amendments to the cutoff values on a 5-point Likert scale ranging from 0=rarely or never (<1 for exclusion because of cognitive impairment have been day) to 1=some or little of the time (1-2 days), 2=often or a discussed in previous publications [50,52]. moderate amount of time (3-4 days), and 3=most of the time (5-7 days). After excluding individuals with cognitive impairment (37/831, 4.5%; Figure 3) and missing information on baseline For the purpose of analysis, the mean score of all 20 items was demographic characteristics (73/831, 8.8%), the analysis sample calculated for all participants (Cronbach α at T0=.605, Cronbach included 721 older adults (see Data Exclusion section and Figure α at T1=.587). The retest reliability was Spearman ρ T0 to T1 3). To determine adherence, a variable regarding meeting the of 0.759 (P<.001). https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 9 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al recommended 150 minutes of MVPA was computed by study completion. This was managed by missing value subtracting the baseline measure from the T1 measure. The imputation using an expectation-maximization algorithm. resulting variable thereby indicated whether study participants In addition, according to the study protocol, participants with remained active or inactive (0), fell back into not meeting cognitive impairments were excluded (37/831, 4.5%; Figure 3). recommendations anymore (−1), or became active (1). In addition, missing baseline demographic information was not imputed; thus, participants with missing information on sex, Statistical Analysis age, educational status, family status, or BMI were excluded Preparation from the analyses (73/831, 9%). All analyses were performed using SPSS (version 27; IBM Corp). The Little missing completely at random test (P>.05) Results suggested that data were not missing completely at random (ie, Hypothesis 1 it suggested that data were missing systematically). Assuming that existing data could be used to produce an estimate of the To test whether the 2 interventions outperformed the WLCG missing information (ie, assuming that data were missing at (hypothesis 1), the study participants who adopted or maintained random) [74], single data imputation was implemented by using an active lifestyle were analyzed, as recommended by the WHO. the expectation-maximization algorithm. First, those who did not meet the recommendation regarding PA at T0 based on self-reported adherence were investigated; Test of Hypotheses of those individuals, more individuals became adherent if they To assess whether the 2 IGs outperformed the WLCG were exposed to the web-based intervention (47/59, 80%; Table (hypothesis 1) on self-reported behavior and stage of change, 1) or received the print-based intervention (20/25, 80%) than frequency analyses and chi-square tests, Z tests, and contingency those who were not treated (WLCG; 5/7, 71%). At a descriptive coefficient tests were used to test the number of participants level, the numbers indicate the favoring of the IG over the who adopted or maintained an active lifestyle. control group. Changes in social-cognitive predictors (hypothesis 2) were Second, those who met the recommendation regarding PA analyzed with repeated-measures multivariate analysis of before the study were investigated; of these, 600 individuals variance (MANOVA) via mixed-effects generalized linear self-reported to be adherent at T1. More individuals remained models with group and time as factors. For the F values, the adherent if they were exposed to the web-based intervention Roy largest root was reported. (396/411, 96.4%) or not treated (WLCG; 141/150, 94%) than those receiving the print-based intervention (63/69, 91%). The primary aim of the moderated mediation analyses was to investigate whether the IG allocation (independent variable or The difference in the proportion of adopters between the predictor) had an effect on perceived symptoms of depression web-based and print-based interventions was statistically (CES-D score) at T1 (dependent variable or outcome), which significant (Z=-1.94; P=.02); as well as the differences between is mediated by subjective age at T1 and moderated by the stage the IGs and the WLCG were significant (WLCG vs print of change in endurance training at T0 (hypothesis 3; see Figure Z=2.3967 and WLCG vs web Z=3.8367; both P=.01). 1 for the proposed model). This finding was replicated by the stages of change in endurance The secondary aim of the moderated mediation analysis was to training. In Table 2, the number of study participants in the 3 investigate the same relationship with loneliness as the intervention conditions moving from the nonintenders, intenders, dependent variable (hypothesis 4; see Figure 2 for the proposed or actor stage to another stage or remaining in the former stage model). These associations were investigated using moderated is reported. mediation models within the PROCESS macro (version 3.0; In all stage groups, the percentage of individuals moving a stage Hayes, The International Association of Applied Psychology forward (from nonintentional stage to intentional or action, and mediation analysis). from intentional to action) or maintaining the stage when starting The models were adjusted for the following baseline variables: as an actor was higher in the web-based or print group than in loneliness and CES-D score, as well as subjective age, sex, age, the WLCG (Table 2). In contrast, in the WLCG, the percentage educational status (International Standard of Education), family of individuals remaining in the nonintentional or intentional status, and BMI (all at T0). A bootstrapping approach of 10,000 stage was higher than that in the IGs (Table 2). The pattern in samples and a specific seed (seed=1) was applied to ensure Table 2 was statistically significant (χ =79.1; P<.001; robust and replicable results. The effect sizes were represented contingency coefficient 0.314; P<.001). Z tests were performed by unstandardized regression coefficients. To calculate the to test whether group differences in the stage of change heteroscedasticity-robust SE, the HC3-Option in the process movements were statistically significant, which was the case function was used. Accordingly, the assumption of for initial intenders who moved to the actor stage (WLCG vs homoscedasticity could be avoided. web-based Z=−4.2325; P=.01 and WLCG vs print-based Data Exclusion Z=−5.349; P=.01) and initial actors who remained in the actor stage (WLCG vs print Z=−3.1853; P<.01). The analyses were conducted following the intention to treat principle; that is, participants were included in primary analyses Summarizing the findings regarding hypothesis 1 that the according to their original group allocation and disregarding web-based and print-based interventions outperformed the https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 10 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al control condition in terms of PA behavior change, we can of remaining in or a relapse into not meeting recommendations conclude that our results suggest this direction. The web-based (Table 1) and adopting or remaining in the intender or actor intervention seemed to work better in terms of the prevention stage (Table 2). Table 1. Numbers and percentages of study participants assigned to 1 of 3 experimental groups regarding who met or did not meet the recommended a b physical activity level at T0 and T1 . At T0 At T1, n (%) Total, N Not meeting the recommendation Meeting the recommendation Not meeting the recommendation 2 (28.6) 5 (71.4) 7 WLCG Web-based 12 (20.3) 47 (79.7) 59 Print-based 5 (20) 20 (80) 25 Total 19 (20.9) 72 (79.1) 91 Meeting the recommendation WLCG 9 (6) 141 (94) 150 Web-based 15 (3.6) 396 (96.4) 411 Print-based 6 (8.7) 63 (91.3) 69 Total 30 (4.8) 600 (95.2) 630 T0: time point 0. T1: time point 1. WLCG: wait-list control group. a b Table 2. Cross-tabulation of nonintenders, intenders, and actors at T0 moving to or remaining nonintenders, intenders, and actors at T1 depending on the experimental group they were in. T0 and T1 Nonintenders, n (%) Intenders, n (%) Actors, n (%) Total, N Nonintenders 36 (59) 16 (26.2) 9 (14.8) 61 WLCG Web-based 35 (16.7) 46 (22) 128 (61.2) 209 Print-based 9 (17.6) 13 (25.5) 29 (56.9) 51 Intenders WLCG 7 (23.3) 11 (36.7) 12 (40) 30 Web-based 5 (4.8) 11 (10.6) 88 (84.6) 104 Print-based 4 (16) 2 (8) 19 (76) 25 Actors WLCG 16 (24.2) 8 (12.1) 42 (63.6) 66 Web-based 9 (5.7) 7 (4.5) 141 (89.8) 157 Print-based 2 (11.1) 0 (0) 16 (88.9) 18 All stages together WLCG 59 (37.6) 35 (22.3) 63 (40.1) 157 Web-based 49 (10.4) 64 (13.6) 357 (76) 470 Print-based 15 (16) 15 (16) 64 (68.1) 94 Total 123 (17.1) 114 (15.8) 484 (67.1) 721 T0: time point 0. T1: time point 1. WLCG: wait-list control group. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 11 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Table S2). Both the between-group effect (F =8.668; P<.001; Hypothesis 2 3, 717 A total of 2 MANOVAs with 19 predictors (Multimedia η =0.018) and the interaction of group and time were significant Appendix 1, Table S1) and 3 outcomes (Multimedia Appendix (F =6.101; P<.001; η =0.025). The group effect was mainly 3, 717 1, Table S2) were calculated to test hypothesis 2. based on subjective age and symptoms of depression, and the interaction effect was based on symptoms of depression with The first MANOVA revealed a significant difference between regard to group and time on loneliness (see Multimedia the 3 groups, favoring the 2 IGs over and above the WLCG 2 Appendix 2, Table S1-S3 with means, SDs, and statistics). (F =4.778; P<.001; η =0.098), as well as a significant 19, 701 interaction between time and group (F =2.778; P<.001; Figure 5 highlights that the effect of loneliness comes from 19, 701 2 regression to the mean, with the WLCG increasing in its η =0.070). loneliness, the web-based IG maintaining its previous level, and This effect was mainly based on intention, negative outcome the print-based IG decreasing in its loneliness over time. With expectancies, planning, and habit (see Multimedia Appendix 1, subjective age, all groups showed an increase over time (Figure Table S1, for the means, SDs, and statistics). Figure 4 outlines 5). Differences from baseline values remained. Figure 5 also the development, indicating that the WLCG dropped slightly shows that all groups started off at almost the same level of with its intention over time, whereas the 2 IGs improved over depressive symptoms. However, over time, the WLCG increased time. Figure 4 shows that the WLCG retained its negative with regard to the reported symptoms of depression, whereas outcome expectancies, whereas the 2 IGs improved in terms of the 2 IGs decreased, with an even better effect of the print-based perceiving fewer negative outcomes. Figure 4 also demonstrates intervention than that of the web-based intervention. that the WLCG dropped with its self-efficacy, whereas Overall, the effect sizes were rather small, ranging from self-efficacy remained stable in the web-based IG and increased η =0.098 to 0.018. in the print-based IG. Finally, Figure 4 shows that the WLCG remained stable in terms of habit strength, whereas the 2 IGs Summarizing the findings regarding hypothesis 2, we found improved over time. overall support. The web-based and print-based interventions improved the social-cognitive predictors of PA behavior change, With the second MANOVA testing the outcomes, the effects subjective aging, loneliness, and depression compared with the of the interventions on subjective aging, loneliness, and control condition. The web-based and print-based interventions symptoms of depression were tested (Multimedia Appendix 1, were significantly different from the control condition. Figure 4. (A) Intention, (B) negative outcome expectancies, (C) self-efficacy, and (D) habit. T0: time point 0; T1: time point 1; WLCG: wait-list control group. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 12 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Figure 5. (A) Loneliness, (B) subjective age, and (C) symptoms of depression. CES-D: Center for Epidemiologic Studies Depression; T0: time point 0; T1: time point 1; WLCG: wait-list control group. age (subjective age), a significant positive relationship between Hypothesis 3 subjective age and loneliness at T1 was found (subjective age: To test the mechanisms and hypothesis 3, we conducted a b path 0.01, 95% CI 0.001-0.02; SE 0.01). However, the effect moderated mediation analysis. The results are shown in Figure size was very small, and no significant indirect effects of the 6 and Tables S1 and S2 in Multimedia Appendix 2. interventions on loneliness through subjective age were revealed. To summarize, the results suggest that subjective age could not None of the interventions had a significant direct effect on account for a significant proportion of the relationship between loneliness compared with the WLCG (web-based: c′ path −0.01, the IGs and loneliness. Furthermore, there was no significant 95% CI −0.11 to 0.08, SE 0.05; print-based: c′ path −0.09, 95% interaction, suggesting that there was no moderation of the stage CI −0.24 to 0.06, SE 0.08). Regarding the mediator subjective of change in endurance activities. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 13 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Figure 6. Moderated mediation model results for loneliness. The interaction between the intervention group and the stage of change in the a path is shown in parentheses. The first value represents the intention stage, and the second value represents the actor stage. The moderation of the ab path is shown on the upper right. The first value represents the nonintenders stage, the second value represents the intenders stage, and the third value represents the actor stage. The model was adjusted for the following baseline variables: loneliness, subjective age, sex, age, educational status (International Standard Classification of Education), family status, and BMI (all at time point 0). The wait-list control group was used as a reference. *Statistically significant value. T1: time point 1. b path 0.14, 95% CI 0.05-0.23; SE 0.05). An indirect Hypothesis 4 relationship between the intervention and the symptoms of To test hypothesis 4, another moderated mediation model was depression via subjective age was only present for participants tested. who were both in the actor stage of change for endurance activities and received the web-based intervention (web-based: The results are shown in Figure 7 and Tables S3 and S4 in ab path −0.14, 95% CI −0.34 to −0.01; SE 0.09). Multimedia Appendix 2. Summarizing the findings regarding hypothesis 4, only older Validating the previous analyses, both interventions had a adults in the actor stage of endurance training who received the significant direct effect on the symptoms of depression at T1 web-based intervention were associated with lower symptoms compared with the WLCG (web-based: c′ path −0.86, 95% CI of depression at T1, which was partially mediated by subjective −1.58 to −0.13, SE 0.38; print-based: c′ path −1.96, 95% CI age at T1. Thus, the existence of moderated mediation was −2.99 to −0.92, SE 0.53). Furthermore, subjective age at T1 confirmed, although all other mediation pathways were not was positively related to depressive symptoms (subjective age: significant, and the effect sizes were small. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 14 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Figure 7. Moderated mediation model results for depression. The interaction between the intervention group and the stage of change in the a path is shown in brackets. The first value represents the intention stage, and the second value represents the actor stage. The moderation of the ab path is shown in the upper right. The first value represents the nonintenders stage, the second value represents the intenders stage, and the third value represents the actor stage. The model was adjusted for the following baseline variable: depressive symptoms (CES-D), subjective age, sex, age, educational status (International Standard Classification of Education), family status, and BMI (all at time point 0). The wait-list control group was used as a reference. *Statistically significant values. CES-D: Center for Epidemiologic Studies Depression; T1: time point 1. of depressive symptoms and loneliness if they actually felt Discussion younger. Principal Findings However, compared with the wait-list control, the intervention did not help older adults feel less lonely, perhaps because of 2 This study aimed to compare the effects of web- and print-based aspects. We revealed a floor effect (ie, generally low loneliness PA interventions on self-reported PA, stage of change, levels). Moreover, one should also keep in mind that the determinants of PA, loneliness, and depression. Moreover, the intervention was not designed to reduce loneliness but to goal was to investigate whether subjective age is a mediator increase PA; accordingly, the relationship between the and whether the stage of change is a moderator of the intervention and feeling lonely was not strong enough to be of effectiveness in 831 older individuals participating in the statistical importance. Nevertheless, the effects in the IGs PROMOTE 1 or PROMOTE 2 study. underlined the importance of supporting active older adults to The main finding was that the print-based and web-based remain physically active to feel fit and subjectively young, as interventions both worked well and helped a higher proportion well as to lower symptoms of depression. In the following of individuals meet the recommendations for PA and move paragraphs, the hypotheses are reviewed in more detail, followed forward with their stage of change compared with the WLCG. by more discussion. Support for the effectiveness of the interventions was also found Hypothesis 1, assuming that web-based and print-based regarding the social-cognitive predictors of PA behavior. None interventions outperform the control condition in terms of of the interventions had a significant direct effect on loneliness self-reported PA behavior change, was confirmed by our data. compared with the control group. Thus, the main assumption The interventions seemed to work better in terms of preventing that a PA intervention always helps reduce loneliness does not a relapse into not meeting recommendations anymore and hold true. It seems more important to take mastery experience moving study participants into the actor stage. Without any into account: the results of our moderated mediation analyses intervention, 71% (5/7) of the previously inactive participants suggest that compared with the WLCG, receiving the web-based became active at the recommended level. This percentage was intervention was associated with lower symptoms of depression higher in the web-based (47/59, 80%) and print-based groups at T1 and that subjective age could explain a substantial (20/25, 80%). Without any intervention, 94% (141/150) of the proportion of variance. However, this holds true only for previously active study participants remained active at the participants in the actor stage of change for endurance activities. recommended level. This percentage was slightly higher in the The mechanisms are in accordance with the assumption that web-based group (396/411, 94%). mastery experience and self-regulation—operationalized with subjective age—help the study participants who are already When replicating this finding with the stages of behavior change, physically active at the baseline to reduce their symptoms the effects clearly demonstrated the benefits of the interventions; because of maintained or improved physical perceptions and in the untreated WLCG, individuals were more likely to remain repairing or maintaining interpersonal skills and relationships in or relapse to the nonintentional stage than the individuals in [6]. In this group, the PA IG allocation reduced the likelihood the 2 IGs. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 15 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al Our results also support hypothesis 2: participation in web-based Nevertheless, although it is helpful to compare rather and print-based PA interventions was associated with conventional print alternatives with web-based interventions, improvements in the social-cognitive predictors of PA behavior it should also be considered that future senior populations are change, self-reported PA behavior change, subjective age, likely to have more digital competence. Web-based interventions loneliness, and depressive symptoms compared with the WLCG. offer inclusionary benefits that can be of particular interest when The web-based and print-based interventions were significantly dealing with an older adult population (eg, text-to-speech options different from the WLCG, which matched expectations. and variable font sizes). Therefore, the design, content, formatting, and acceptance of digital or internet interventions Hypothesis 3 was not supported by our results, as we found that should be further researched and tailored when implementing the results were not consistent with the assumption that interventions for older adults. individuals in the IG would benefit from fewer feelings of loneliness as they would experience a decrease in subjective In addition, a critical point is selective dropout (eg, higher age. Regarding hypothesis 4, only older adults who were in the attrition in either the web-based IG, in the groups that received actor stage regarding endurance training and who received the only print-based material, or the WLCG). This was particularly web-based intervention revealed an intervention effect on evident in the first phase of the study among the groups with depressive symptoms (CES-D score), partially mediated by high technology requirements (web material and activity trackers subjective age at T1. This subgroup reported a lower subjective that had to be synchronized with the website). Accordingly, age than that of the WLCG, which was further associated with future projects should support those at risk of dropping out of reporting lower symptoms of depression at T1. the intervention to remain in the study. A further shortcoming of this study is that the interventions Limitations and Suggestions for Future Work were designed to improve PA and did not explicitly reduce In this study, only subjective data were included as most of the loneliness or symptoms of depression directly. As the findings variables such as social-cognitive variables, stage of change, are promising, future studies should follow up on this as PA loneliness, subjective age, and symptoms of depression could can also be a very effective tool to address these emotions and only be measured in this form. However, the behavior should cognitions. Parts of the data have already been published, such be assessed using objective measures (such as an activity as the effects of the interventions on the stages of change in tracker). Accordingly, validation studies are required to better PROMOTE 1 [50,53]. Consequently, when using these data for understand behavior changes more thoroughly. a systematic review or meta-analysis, this needs to be taken into In addition, the included study participants were not account. representative samples, as many individuals had already met In general, whether older adults benefit from improved health the behavioral criteria, especially in the first study phase (other indicators in addition to symptoms of depression) or (PROMOTE 1). Despite the adaptation of the inclusion criteria well-being (other indicators in addition to loneliness), health for the second phase of the study, the target group of individuals behavior, and its predictors from print-based material or with low activity was still not sufficiently reached, and their web-based interventions still requires more attention, and future needs were not adequately addressed. This is a common problem studies should follow up on the mechanisms that explain how in public health intervention research, which requires further the interventions work. effort (eg, to improve the recruitment of nonactive and low-motivation individuals). Therefore, future studies should Process Analysis attempt to overcome this problem. Therefore, to address the The results of the moderated mediation analysis revealed that needs of these individuals more effectively, need-based there is not only a simple linear effect of the intervention on assessments (ie, tailored to current circumstances and hindering symptoms of depression but also that the effects of the factors) should be conducted in future studies. intervention on symptoms of depression are modeled in a more Another limitation is access to, and availability of, web-based complex way. Therefore, it can be hypothesized that technology, particularly among the studied age groups. Thus, interventions generally tend to affect target variables in a in this context, on the one hand, some of the respondents nonlinear fashion and that, in general, interventions and health interested in the study invitation had to be excluded from should be understood as complex systems. With the actor stage participation as they did not meet the inclusion criterion of as a moderator, the conceivable conclusion is that only people owning a PC or did not have access to the internet. who were physically active before the intervention benefited from it with regard to the symptoms of depression. Therefore, On the other hand, personal affinity for and acceptance of the HAPA theory is a helpful tool in designing interventions as modern technology services, which in part becomes apparent it can assist in identifying a person’s current stage of change over the course of the study, might affect the success of and aid them in moving to the actor stage. information technology interventions and, therefore, should not be overlooked. These eligibility criteria partly exclude Li et al [47] found that subjective age is related to various health disadvantaged groups with different psychological preconditions benefits. In their study, subjective age acted as a mediator and developmental possibilities. In the future, such limitations between the intervention and symptoms of depression. With should be overcome by making study participation more this in mind, the following can be assumed: in interventions, a accessible to individuals with low computer or internet literacy set of variables that can act as potential mediators exists, as and technology affinity. these variables are linked to several health benefits. However, https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 16 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al it is easy to overlook these mediators when only focusing on to this group should be informed about the progress and linear and bivariate effects, when, in fact, a mediation might expected time of participation. As previous studies have only apply to a subgroup of individuals. Future research needs indicated that longer waiting times are associated with higher to identify this set of variables and implement it effectively in dropout rates, the study design should be adapted to decrease intervention frameworks. Subjective age might be one of these the waiting time for the WLCG [8,22,54]. variables, although there might be more underlying components. In addition, in the first phase of the study, the problem of increased dropout or greater dissatisfaction in the study arms Comparison With Prior Work with a need for greater technical skills and information Loneliness and mental health in the aging population are technology acceptance became relevant [50]. This has also been important topics, and our data match previous work showing reported by other authors [79]. As a reaction, the study design that physical exercise is a key factor in addressing these issues for PROMOTE 2 was adapted such that participants could [4,6]. However, the effects of the intervention were stronger change the type of materials provided at a defined time point regarding behavior adherence and stage (contingency coefficient on a preference basis, which could improve the loss of 0.314), as well as behavioral habits (η =0.025-0.026), than participants because of this aspect, as well as the satisfaction regarding loneliness and depression (η =0.005-0.024). The with the intervention in a recognizable way [52]. finding of stage-specific effects (only actors benefited) with Conclusions regard to depression matches the previous finding that interventions improving PA can also improve mood [7]. In times of physical distancing, for instance, during the COVID-19 pandemic, alternative forms of support, such as To date, little research has been conducted on whether print- and web-based PA exercise content for individual web-based or print-based interventions are more likely to result implementation at home, are essential and in high demand. As in successful behavior change regarding PA, which can users are very heterogeneous, tailoring PA interventions simultaneously affect feelings of loneliness and depression. By according to their specific needs (including differing motivations matching the findings of Golsteijn et al [9], this study revealed to engage in PA) and previous experiences (captured by stages), some benefits of the print-based intervention for this age group as well as according to their individual technology-based [9]. However, in general, the web-based and print-based preconditions, could be an effective approach to initiating interventions were more effective than the WLCG. behavior change with regard to PA. Particularly, for digital Aspects assumed to affect participation in, and effectiveness interventions, the varying availability and use preferences of of, interventions have been previously studied. With respect to digital devices should be considered. the uptake of PA offerings and intervention modalities, Tools for individual use, including activity monitoring (such as sociodemographic differences have emerged in the past. exercise diaries) and exercise instructions, are highly relevant Accordingly, women, those with higher levels of education, for location- and time-independent use. Finally, but importantly, and those who are already physically active are more likely to the relevance is determined by the reduced mobility in old age participate [23,75]. Physical inactivity, being overweight, and and, thus, the possibility or need for exercise at home. having a low educational status were indicators of discontinuation of the intervention before it was completed [22]. Successful aging in terms of helping older adults feel fit to Furthermore, it is known that younger individuals seem to prefer perform PA should be considered more explicitly, which can web-based services [8,76], whereas older adults or women help to ensure that a PA intervention actually translates into the appear more likely to favor print-based offerings [76]. reduction of symptoms of depression. At the same time, taking the stage of change-specific aspects into account can benefit This and former analyses of the PROMOTE 1 study [50] the knowledge that, similar to this study, the interventions identified higher attrition rates in the IGs than in WLCG, which worked well in intenders and actors for successful aging and is in line with other similar intervention studies [77]. This may symptoms of depression but not in nonintenders. Nonintenders be explained by the fact that the participants were more might need other support such as just-in-time adaptive motivated to stay in the study as they received the program interventions and more instant social support, not only through afterward. However, those who obtained the intervention later print-based and web-based modes. However, the delivered also predominantly maintained their behavior over the course interventions appeared to be supportive of intenders and actors of the intervention. Cunningham et al [78] found that wait-list and improve the predictors of behavior. Ingredients of the groups interrupt efforts to change and pointed to the importance intervention’s behavior change techniques (ie, goal setting, of activities toward changing different psychological planning, social support, and feedback [58]) paid off. preconditions and developmental possibilities. Therefore, to reduce the dropout rate in the WLCG, individuals randomized Acknowledgments This study was supported by research funding from the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung), with project numbers 01EL1822A, 01EL1822C, 01EL1822F, and 01EL1822I. The funder was not involved in the design of the study; in the collection, analysis, and interpretation of data; or in writing the manuscript. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 17 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al The authors would like to thank Grace Wire, Ronja Bellinghausen, and Veronika Hochdahl for manuscript editing and proofreading and Robin Rinn for his help in improving an earlier version of this manuscript. Authors' Contributions All authors made substantial contributions to the study conception (SL, CP, and CV-R), design of the study (SL, TR, and DJ), acquisition and analysis (SL, TR, FMK, DJ, and MP), data interpretation (SL, TR, FMK, DJ, MP, CP, and CV-R), manuscript drafting (SL, TR, and FMK), and substantial revision of the manuscript (SL, TR, FMK, DJ, MP, CP, and CV-R). All authors approved the submitted manuscript and agreed to be personally accountable for their own contributions and to ensure that questions related to the accuracy or integrity of any part of the work were appropriately investigated and resolved and the resolution documented in the literature. Conflicts of Interest None declared. Multimedia Appendix 1 Effect of the web and print interventions in comparison to the wait-list control group on social-cognitive variables (Table S1), and on loneliness, perceived age and symptoms of depression (Table S2). [DOCX File , 24 KB-Multimedia Appendix 1] Multimedia Appendix 2 Ordinal least squares regression model results for subjective age (Table S1), loneliness (Table S2), subjective age (Table S3) and depressive symptoms (Table S4). [DOCX File , 29 KB-Multimedia Appendix 2] References 1. Groarke JM, Berry E, Graham-Wisener L, McKenna-Plumley PE, McGlinchey E, Armour C. Loneliness in the UK during the COVID-19 pandemic: cross-sectional results from the COVID-19 psychological wellbeing study. PLoS One 2020;15(9):e0239698 [FREE Full text] [doi: 10.1371/journal.pone.0239698] [Medline: 32970764] 2. Noone C, McSharry J, Smalle M, Burns A, Dwan K, Devane D, et al. Video calls for reducing social isolation and loneliness in older people: a rapid review. Cochrane Database Syst Rev 2020 May 21;5:CD013632 [FREE Full text] [doi: 10.1002/14651858.CD013632] [Medline: 32441330] 3. Savage RD, Wu W, Li J, Lawson A, Bronskill SE, Chamberlain SA, et al. Loneliness among older adults in the community during COVID-19: a cross-sectional survey in Canada. BMJ Open 2021 Apr 02;11(4):e044517 [FREE Full text] [doi: 10.1136/bmjopen-2020-044517] [Medline: 33811054] 4. Macy JT, Owens C, Mullis K, Middlestadt SE. The role of self-efficacy and injunctive norms in helping older adults decide to stay home during the COVID-19 pandemic. Front Public Health 2021;9:660813 [FREE Full text] [doi: 10.3389/fpubh.2021.660813] [Medline: 34150704] 5. Dickens AP, Richards SH, Greaves CJ, Campbell JL. Interventions targeting social isolation in older people: a systematic review. BMC Public Health 2011 Aug 15;11:647 [FREE Full text] [doi: 10.1186/1471-2458-11-647] [Medline: 21843337] 6. Pels F, Kleinert J. Loneliness and physical activity: a systematic review. Int Rev Sport Exerc Psychol 2016 Jun 07;9(1):231-260. [doi: 10.1080/1750984x.2016.1177849] 7. Alves GS, Casali ME, Veras AB, Carrilho CG, Bruno Costa E, Rodrigues VM, et al. A systematic review of home-setting psychoeducation interventions for behavioral changes in dementia: some lessons for the COVID-19 pandemic and post-pandemic assistance. Front Psychiatry 2020;11:577871 [FREE Full text] [doi: 10.3389/fpsyt.2020.577871] [Medline: 33132937] 8. Boekhout JM, Peels DA, Berendsen BA, Bolman C, Lechner L. A web-based and print-delivered computer-tailored physical activity intervention for older adults: pretest-posttest intervention study comparing delivery mode preference and attrition. J Med Internet Res 2019 Aug 28;21(8):e13416 [FREE Full text] [doi: 10.2196/13416] [Medline: 31464186] 9. Golsteijn RH, Peels DA, Evers SM, Bolman C, Mudde AN, de Vries H, et al. Cost-effectiveness and cost-utility of a web-based or print-delivered tailored intervention to promote physical activity among adults aged over fifty: an economic evaluation of the Active Plus intervention. Int J Behav Nutr Phys Act 2014 Sep 28;11:122 [FREE Full text] [doi: 10.1186/s12966-014-0122-z] [Medline: 25262435] 10. Fortuna KL, Torous J, Depp CA, Jimenez DE, Areán PA, Walker R, et al. A future research agenda for digital geriatric mental healthcare. Am J Geriatr Psychiatry 2019 Nov;27(11):1277-1285 [FREE Full text] [doi: 10.1016/j.jagp.2019.05.013] [Medline: 31196619] 11. Paggi ME, Jopp D, Hertzog C. The importance of leisure activities in the relationship between physical health and well-being in a life span sample. Gerontology 2016;62(4):450-458 [FREE Full text] [doi: 10.1159/000444415] [Medline: 26974682] https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 18 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al 12. Sala G, Jopp D, Gobet F, Ogawa M, Ishioka Y, Masui Y, et al. The impact of leisure activities on older adults' cognitive function, physical function, and mental health. PLoS One 2019;14(11):e0225006 [FREE Full text] [doi: 10.1371/journal.pone.0225006] [Medline: 31703115] 13. Global recommendations on physical activity for health. World Health Organization. URL: https://apps.who.int/iris/bitstream/ handle/10665/44399/9789241599979_eng.pdf [accessed 2022-07-12] 14. Hupin D, Roche F, Gremeaux V, Chatard J, Oriol M, Gaspoz J, et al. Even a low-dose of moderate-to-vigorous physical activity reduces mortality by 22% in adults aged ≥60 years: a systematic review and meta-analysis. Br J Sports Med 2015 Oct;49(19):1262-1267. [doi: 10.1136/bjsports-2014-094306] [Medline: 26238869] 15. Warburton DE, Nicol CW, Bredin SS. Health benefits of physical activity: the evidence. Can Med Assoc J 2006 Mar 14;174(6):801-809 [FREE Full text] [doi: 10.1503/cmaj.051351] [Medline: 16534088] 16. Södergren M, McNaughton SA, Salmon J, Ball K, Crawford DA. Associations between fruit and vegetable intake, leisure-time physical activity, sitting time and self-rated health among older adults: cross-sectional data from the WELL study. BMC Public Health 2012 Jul 25;12:551 [FREE Full text] [doi: 10.1186/1471-2458-12-551] [Medline: 22830932] 17. Abdelbasset WK, Alsubaie SF, Tantawy SA, Elyazed TI, Elshehawy AA. A cross-sectional study on the correlation between physical activity levels and health-related quality of life in community-dwelling middle-aged and older adults. Medicine (Baltimore) 2019 Mar;98(11):e14895 [FREE Full text] [doi: 10.1097/MD.0000000000014895] [Medline: 30882705] 18. Olson EA, McAuley E. Impact of a brief intervention on self-regulation, self-efficacy and physical activity in older adults with type 2 diabetes. J Behav Med 2015 Dec;38(6):886-898 [FREE Full text] [doi: 10.1007/s10865-015-9660-3] [Medline: 26162648] 19. Seguin RA, Economos CD, Palombo R, Hyatt R, Kuder J, Nelson ME. Strength training and older women: a cross-sectional study examining factors related to exercise adherence. J Aging Phys Act 2010 Apr;18(2):201-218 [FREE Full text] [doi: 10.1123/japa.18.2.201] [Medline: 20440031] 20. Dolansky MA, Stepanczuk B, Charvat JM, Moore SM. Women's and men's exercise adherence after a cardiac event. Res Gerontol Nurs 2010 Jan;3(1):30-38 [FREE Full text] [doi: 10.3928/19404921-20090706-03] [Medline: 20128541] 21. Elzen H, Slaets JP, Snijders TA, Steverink N. Do older patients who refuse to participate in a self-management intervention in the Netherlands differ from older patients who agree to participate? Aging Clin Exp Res 2008 Jun;20(3):266-271. [doi: 10.1007/BF03324777] [Medline: 18594195] 22. Jancey J, Lee A, Howat P, Clarke A, Wang K, Shilton T. Reducing attrition in physical activity programs for older adults. J Aging Phys Act 2007 Apr;15(2):152-165. [doi: 10.1123/japa.15.2.152] [Medline: 17556782] 23. Chinn DJ, White M, Howel D, Harland JO, Drinkwater CK. Factors associated with non-participation in a physical activity promotion trial. Public Health 2006 Apr;120(4):309-319. [doi: 10.1016/j.puhe.2005.11.003] [Medline: 16473376] 24. Picorelli AM, Pereira LS, Pereira DS, Felício D, Sherrington C. Adherence to exercise programs for older people is influenced by program characteristics and personal factors: a systematic review. J Physiother 2014 Sep;60(3):151-156 [FREE Full text] [doi: 10.1016/j.jphys.2014.06.012] [Medline: 25092418] 25. Lakerveld J, Ijzelenberg W, van Tulder MW, Hellemans IM, Rauwerda JA, van Rossum AC, et al. Motives for (not) participating in a lifestyle intervention trial. BMC Med Res Methodol 2008 Apr 10;8:17 [FREE Full text] [doi: 10.1186/1471-2288-8-17] [Medline: 18402683] 26. Justine M, Azizan A, Hassan V, Salleh Z, Manaf H. Barriers to participation in physical activity and exercise among middle-aged and elderly individuals. Singapore Med J 2013 Oct;54(10):581-586 [FREE Full text] [doi: 10.11622/smedj.2013203] [Medline: 24154584] 27. Wichmann F, Pischke CR, Jürgens D, Darmann-Finck I, Koppelin F, Lippke S, et al. Requirements for (web-based) physical activity interventions targeting adults above the age of 65 years - qualitative results regarding acceptance and needs of participants and non-participants. BMC Public Health 2020 Jun 11;20(1):907 [FREE Full text] [doi: 10.1186/s12889-020-08927-8] [Medline: 32527251] 28. Gourlan M, Bernard P, Bortolon C, Romain AJ, Lareyre O, Carayol M, et al. Efficacy of theory-based interventions to promote physical activity. A meta-analysis of randomised controlled trials. Health Psychol Rev 2016;10(1):50-66. [doi: 10.1080/17437199.2014.981777] [Medline: 25402606] 29. van Genugten L, Dusseldorp E, Webb TL, van Empelen P. Which combinations of techniques and modes of delivery in internet-based interventions effectively change health behavior? A meta-analysis. J Med Internet Res 2016 Jun 07;18(6):e155 [FREE Full text] [doi: 10.2196/jmir.4218] [Medline: 27268104] 30. Wilson TE, Hennessy EA, Falzon L, Boyd R, Kronish IM, Birk JL. Effectiveness of interventions targeting self-regulation to improve adherence to chronic disease medications: a meta-review of meta-analyses. Health Psychol Rev 2020 Mar;14(1):66-85 [FREE Full text] [doi: 10.1080/17437199.2019.1706615] [Medline: 31856664] 31. McEwan D, Beauchamp MR, Kouvousis C, Ray CM, Wyrough A, Rhodes RE. Examining the active ingredients of physical activity interventions underpinned by theory versus no stated theory: a meta-analysis. Health Psychol Rev 2019 Mar;13(1):1-17. [doi: 10.1080/17437199.2018.1547120] [Medline: 30412685] 32. Johnson BT, Scott-Sheldon LA, Carey MP. Meta-synthesis of health behavior change meta-analyses. Am J Public Health 2010 Nov;100(11):2193-2198 [FREE Full text] [doi: 10.2105/AJPH.2008.155200] [Medline: 20167901] https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 19 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al 33. Murray JM, Brennan SF, French DP, Patterson CC, Kee F, Hunter RF. Effectiveness of physical activity interventions in achieving behaviour change maintenance in young and middle aged adults: a systematic review and meta-analysis. Soc Sci Med 2017 Nov;192:125-133. [doi: 10.1016/j.socscimed.2017.09.021] [Medline: 28965003] 34. Rhodes RE, Janssen I, Bredin SS, Warburton DE, Bauman A. Physical activity: health impact, prevalence, correlates and interventions. Psychol Health 2017 Aug;32(8):942-975. [doi: 10.1080/08870446.2017.1325486] [Medline: 28554222] 35. Schwarzer R, Lippke S, Luszczynska A. Mechanisms of health behavior change in persons with chronic illness or disability: the Health Action Process Approach (HAPA). Rehabil Psychol 2011 Aug;56(3):161-170. [doi: 10.1037/a0024509] [Medline: 21767036] 36. Powell SM, Larsen CA, Phillips SM, Pellegrini CA. Exploring beliefs and preferences for reducing sedentary behavior among adults with symptomatic knee osteoarthritis or knee replacement. ACR Open Rheumatol 2021 Jan;3(1):55-62 [FREE Full text] [doi: 10.1002/acr2.11216] [Medline: 33400397] 37. Nguyen MH, Smets EM, Bol N, Loos EF, Van Weert JC. How tailoring the mode of information presentation influences younger and older adults' satisfaction with health websites. J Health Commun 2018;23(2):170-180. [doi: 10.1080/10810730.2017.1421729] [Medline: 29345531] 38. Quittschalle J, Stein J, Luppa M, Pabst A, Löbner M, Koenig H, et al. Internet use in old age: results of a German population-representative survey. J Med Internet Res 2020 Nov 23;22(11):e15543 [FREE Full text] [doi: 10.2196/15543] [Medline: 33226351] 39. Neil-Sztramko SE, Smith-Turchyn J, Richardson J, Dobbins M. A mobility-focused knowledge translation randomized controlled trial to improve physical activity: process evaluation of the Move4Age study. J Med Internet Res 2019 Jun 20;21(6):e13965 [FREE Full text] [doi: 10.2196/13965] [Medline: 31223121] 40. Peels DA, Bolman C, Golsteijn RH, de Vries H, Mudde AN, van Stralen MM, et al. Long-term efficacy of a printed or a web-based tailored physical activity intervention among older adults. Int J Behav Nutr Phys Act 2013 Sep 02;10:104 [FREE Full text] [doi: 10.1186/1479-5868-10-104] [Medline: 24228862] 41. Tkatch R, Bazarko D, Musich S, Wu L, MacLeod S, Keown K, et al. A pilot online mindfulness intervention to decrease caregiver burden and improve psychological well-being. J Evid Based Complementary Altern Med 2017 Oct;22(4):736-743 [FREE Full text] [doi: 10.1177/2156587217737204] [Medline: 29228806] 42. Annear M, Keeling S, Wilkinson T, Cushman G, Gidlow B, Hopkins H. Environmental influences on healthy and active ageing: a systematic review. Ageing Soc 2012 Nov 06;34(4):590-622. [doi: 10.1017/s0144686x1200116x] 43. Beauchamp MR, Carron AV, McCutcheon S, Harper O. Older adults' preferences for exercising alone versus in groups: considering contextual congruence. Ann Behav Med 2007 Apr;33(2):200-206. [doi: 10.1007/BF02879901] [Medline: 17447872] 44. Mehra S, van den Helder J, Visser B, Engelbert RH, Weijs PJ, Kröse BJ. Evaluation of a blended physical activity intervention for older adults: mixed methods study. J Med Internet Res 2020 Jul 23;22(7):e16380 [FREE Full text] [doi: 10.2196/16380] [Medline: 32459652] 45. Volders E, de Groot RH, Coumans JM, Bolman CA, Lechner L. A randomized controlled trial into the cognitive effects of a computer-tailored physical activity intervention in older adults with chronic disease(s). Eur Rev Aging Phys Act 2021 Feb 25;18(1):3 [FREE Full text] [doi: 10.1186/s11556-021-00259-9] [Medline: 33632130] 46. Wentzel J, van der Vaart R, Bohlmeijer ET, van Gemert-Pijnen JE. Mixing online and face-to-face therapy: how to benefit from blended care in mental health care. JMIR Ment Health 2016 Feb 09;3(1):e9 [FREE Full text] [doi: 10.2196/mental.4534] [Medline: 26860537] 47. Li Y, Liu M, Miyawaki CE, Sun X, Hou T, Tang S, et al. Bidirectional relationship between subjective age and frailty: a prospective cohort study. BMC Geriatr 2021 Jun 29;21(1):395 [FREE Full text] [doi: 10.1186/s12877-021-02344-1] [Medline: 34187378] 48. Wiener J, Kuhlmann T, Fink S, Hambrecht R, Lippke S. Subjective age and planning strategies for physical activity: brief report about a domain-specific example from a cross-sectional observation study. Int J Sport Psychol 2017;48(4):448-458. 49. Muellmann S, Bragina I, Voelcker-Rehage C, Rost E, Lippke S, Meyer J, et al. Development and evaluation of two web-based interventions for the promotion of physical activity in older adults: study protocol for a community-based controlled intervention trial. BMC Public Health 2017 May 25;17(1):512 [FREE Full text] [doi: 10.1186/s12889-017-4446-x] [Medline: 28545506] 50. Muellmann S, Buck C, Voelcker-Rehage C, Bragina I, Lippke S, Meyer J, et al. Effects of two web-based interventions promoting physical activity among older adults compared to a delayed intervention control group in Northwestern Germany: results of the PROMOTE community-based intervention trial. Prev Med Rep 2019 Sep;15:100958 [FREE Full text] [doi: 10.1016/j.pmedr.2019.100958] [Medline: 31410347] 51. Pischke CR, Voelcker-Rehage C, Peters M, Ratz T, Pohlabeln H, Meyer J, et al. Implementation and effects of information technology-based and print-based interventions to promote physical activity among community-dwelling older adults: protocol for a randomized crossover trial. JMIR Res Protoc 2020 Apr 27;9(4):e15168 [FREE Full text] [doi: 10.2196/15168] [Medline: 32338622] https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 20 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al 52. Pischke C, Voelcker-Rehage C, Ratz T, Peters M, Buck C, Meyer J, et al. Web-based versus print-based physical activity intervention for community-dwelling older adults: crossover randomized trial. JMIR Mhealth Uhealth 2022 Mar 23;10(3):e32212 [FREE Full text] [doi: 10.2196/32212] [Medline: 35319484] 53. Ratz T, Lippke S, Muellmann S, Peters M, Pischke CR, Meyer J, et al. Effects of two web-based interventions and mediating mechanisms on stage of change regarding physical activity in older adults. Appl Psychol Health Well Being 2020 Mar;12(1):77-100. [doi: 10.1111/aphw.12174] [Medline: 31332957] 54. Ratz T, Voelcker-Rehage C, Pischke CR, Muellmann S, Peters M, Lippke S. Health-related lifestyle and dropout from a web-based physical activity intervention trial in older adults: a latent profile analysis. Health Psychol 2021 Aug;40(8):481-490. [doi: 10.1037/hea0001091] [Medline: 34472906] 55. Forberger S, Bammann K, Bauer J, Boll S, Bolte G, Brand T, et al. How to tackle key challenges in the promotion of physical activity among older adults (65+): the Aequipa network approach. Int J Environ Res Public Health 2017 Apr 04;14(4):379 [FREE Full text] [doi: 10.3390/ijerph14040379] [Medline: 28375177] 56. Fleig L, Lippke S, Pomp S, Schwarzer R. Intervention effects of exercise self-regulation on physical exercise and eating fruits and vegetables: a longitudinal study in orthopedic and cardiac rehabilitation. Prev Med 2011 Sep;53(3):182-187. [doi: 10.1016/j.ypmed.2011.06.019] [Medline: 21784096] 57. Pomp S, Fleig L, Schwarzer R, Lippke S. Effects of a self-regulation intervention on exercise are moderated by depressive symptoms: a quasi-experimental study. Int J Clin Health Psychol 2013 Jan;13(1):1-8. [doi: 10.1016/S1697-2600(13)70001-2] 58. Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med 2013 Aug;46(1):81-95. [doi: 10.1007/s12160-013-9486-6] [Medline: 23512568] 59. Nelson ME, Rejeski WJ, Blair SN, Duncan PW, Judge JO, King AC, et al. Physical activity and public health in older adults: recommendation from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc 2007 Aug;39(8):1435-1445. [doi: 10.1249/mss.0b013e3180616aa2] [Medline: 17762378] 60. Lippke S, Ziegelmann JP, Schwarzer R, Velicer WF. Validity of stage assessment in the adoption and maintenance of physical activity and fruit and vegetable consumption. Health Psychol 2009 Mar;28(2):183-193 [FREE Full text] [doi: 10.1037/a0012983] [Medline: 19290710] 61. Nigg CR. There is more to stages of exercise than just exercise. Exerc Sport Sci Rev 2005 Jan;33(1):32-35. [Medline: 15640718] 62. Lippke S, Ziegelmann J, Schwarzer R. Behavioral intentions and action plans promote physical exercise: a longitudinal study with orthopedic rehabilitation patients. J Sport Exerc Psychol 2004;26(3):470-483. [doi: 10.1123/jsep.26.3.470] 63. Schwarzer R, Schuz B, Ziegelmann JP, Lippke S, Luszczynska A, Scholz U. Adoption and maintenance of four health behaviors: theory-guided longitudinal studies on dental flossing, seat belt use, dietary behavior, and physical activity. Ann Behav Med 2007 Apr;33(2):156-166. [doi: 10.1007/BF02879897] [Medline: 17447868] 64. Luszczynska A. An implementation intentions intervention, the use of a planning strategy, and physical activity after myocardial infarction. Soc Sci Med 2006 Feb;62(4):900-908. [doi: 10.1016/j.socscimed.2005.06.043] [Medline: 16095786] 65. Lippke S, Fleig L, Pomp S, Schwarzer R. Validity of a stage algorithm for physical activity in participants recruited from orthopedic and cardiac rehabilitation clinics. Rehabil Psychol 2010 Nov;55(4):398-408. [doi: 10.1037/a0021563] [Medline: 21171799] 66. Verplanken B, Orbell S. Reflections on past behavior: a self-report index of habit strength. J Appl Social Pyschol 2003 Jun;33(6):1313-1330. [doi: 10.1111/j.1559-1816.2003.tb01951.x] 67. Radloff LS. The CES-D Scale: a self-report depression scale for research in the general population. Applied Psychol Measure 1977;1(3):385-401. [doi: 10.1177/014662167700100306] 68. Goeßwald A, Lange M, Kamtsiuris P, Kurth BM. DEGS: studie zur Gesundheit Erwachsener in Deutschland. Bundesweite Quer- und Laengsschnittstudie im Rahmen des Gesundheitsmonitorings des Robert Koch-Instituts. Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 2012;55:775-780. [doi: 10.5414/dbx00211] 69. Fragebögen nach Themenfeldern. Sozialplanung für Senioren. URL: http://www.sozialplanung-senioren.de/ frageboegen-bausteine/frageboegen-nach-themenfeldern/index.html [accessed 2022-01-12] 70. UNESCO United Nations Educational, Scientific and Cultural Organization. International Standard Classification of Education, ISCED 1997. In: Hoffmeyer-Zlotnik JHP, Wolf C, editors. Advances in Cross-National Comparison. Boston, MA: Springer; 2003. 71. Body mass index - BMI. World Health Organization. URL: https://www.euro.who.int/en/health-topics/disease-prevention/ nutrition/a-healthy-lifestyle/body-mass-index-bmi [accessed 2022-01-12] 72. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975 Nov;12(3):189-198. [doi: 10.1016/0022-3956(75)90026-6] [Medline: 1202204] 73. Baek MJ, Kim K, Park YH, Kim S. The validity and reliability of the mini-mental state examination-2 for detecting mild cognitive impairment and Alzheimer's disease in a Korean population. PLoS One 2016;11(9):e0163792 [FREE Full text] [doi: 10.1371/journal.pone.0163792] [Medline: 27668883] 74. Rubin DB. Inference and Missing Data. Biometrika 1976 Dec;63(3):581-592. [doi: 10.2307/2335739] https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 21 (page number not for citation purposes) XSL FO RenderX JMIR AGING Lippke et al 75. Noar SM, Benac CN, Harris MS. Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions. Psychol Bull 2007 Jul;133(4):673-693. [doi: 10.1037/0033-2909.133.4.673] [Medline: 17592961] 76. Vandelanotte C, Duncan MJ. Individual characteristics associated with physical activity intervention delivery mode preferences among adults. Int J Behav Nutr Phys Act 2014 Feb 25;11(1):25 [FREE Full text] [doi: 10.1186/1479-5868-11-25] [Medline: 24568611] 77. Spittaels H, De Bourdeaudhuij I, Vandelanotte C. Evaluation of a website-delivered computer-tailored intervention for increasing physical activity in the general population. Prev Med 2007 Mar;44(3):209-217. [doi: 10.1016/j.ypmed.2006.11.010] [Medline: 17197015] 78. Cunningham JA, Kypri K, McCambridge J. Exploratory randomized controlled trial evaluating the impact of a waiting list control design. BMC Med Res Methodol 2013 Dec 06;13:150 [FREE Full text] [doi: 10.1186/1471-2288-13-150] [Medline: 24314204] 79. Oenema A, Brug J, Dijkstra A, de Weerdt I, de Vries H. Efficacy and use of an internet-delivered computer-tailored lifestyle intervention, targeting saturated fat intake, physical activity and smoking cessation: a randomized controlled trial. Ann Behav Med 2008 Apr;35(2):125-135. [doi: 10.1007/s12160-008-9023-1] [Medline: 18363076] Abbreviations CES-D: Center for Epidemiologic Studies Depression HAPA: Health Action Process Approach IG: intervention group MANOVA: multivariate analysis of variance MMSE: Mini-Mental State Examination MMSE-2: BV: Mini-Mental State Examination second edition: Brief Version MVPA: moderate to vigorous physical activity PA: physical activity T0: time point 0 T1: time point 1 WHO: World Health Organization WLCG: wait-list control group Edited by T Leung, J Wang; submitted 18.01.22; peer-reviewed by A Joseph, M Leslie, K Wall; comments to author 24.02.22; revised version received 20.04.22; accepted 25.05.22; published 09.08.22 Please cite as: Lippke S, Ratz T, Keller FM, Juljugin D, Peters M, Pischke C, Voelcker-Rehage C Mitigating Feelings of Loneliness and Depression by Means of Web-Based or Print-Based Physical Activity Interventions: Pooled Analysis of 2 Community-Based Intervention Trials JMIR Aging 2022;5(3):e36515 URL: https://aging.jmir.org/2022/3/e36515 doi: 10.2196/36515 PMID: ©Sonia Lippke, Tiara Ratz, Franziska Maria Keller, Dennis Juljugin, Manuela Peters, Claudia Pischke, Claudia Voelcker-Rehage. Originally published in JMIR Aging (https://aging.jmir.org), 09.08.2022. This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Aging, is properly cited. The complete bibliographic information, a link to the original publication on https://aging.jmir.org, as well as this copyright and license information must be included. https://aging.jmir.org/2022/3/e36515 JMIR Aging 2022 | vol. 5 | iss. 3 | e36515 | p. 22 (page number not for citation purposes) XSL FO RenderX
Journal
JMIR Aging – JMIR Publications
Published: Aug 9, 2022
Keywords: physical activity; older adults; intervention; loneliness; depression; eHealth; mobile health; mHealth