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

Learn More →

Investigating Friendship Difficulties in the Pathway from ADHD to Depressive Symptoms. Can Parent–Child Relationships Compensate?

Investigating Friendship Difficulties in the Pathway from ADHD to Depressive Symptoms. Can... Attention deficit/hyperactivity disorder (ADHD) is associated with friendship difficulties. This may partly account for the increasingly recognised association between ADHD and subsequent depression. Little is known about the types of friendship dic ffi ulties that could contribute to the association between ADHD and depressive symptoms and whether other relationships, such as parent–child relationships, can mitigate against potential adverse effects of friendship difficulties. In a representative UK school sample (n = 1712), three main features of friendship (presence of friends, friendship quality and characteristics of the individual’s classroom friendship group) were assessed in a longitudinal study with two assessment waves (W1, W2) during the first year of secondary school (children aged 11-12 years). These friendship features (W1) were investigated as potential mediators of the prospective association between teacher-rated ADHD symptoms (W1) and self-rated depressive symptoms (W2) seven months later. Parent–child relationship quality (W1) was tested as a moderator of any indirect effects of ADHD on depression via friendship. ADHD symptoms were inversely associated with friendship presence, friendship quality and positive characteristics of classroom friendship groups. Depressive symptoms were inversely associated with presence and quality of friendships. Friendship quality had indirect effects in the association between ADHD and subsequent depressive symptoms. There was some evidence of moderated mediation, whereby indirect effects via friendship quality attenuated slightly as children reported warmer parent–child relationships. This highlights the importance of considering the quality of friendships and parent–child relationships in children with ADHD symptoms. Fostering good quality relationships may help disrupt the link between ADHD symptomology and subsequent depression risk. Keywords ADHD · Depression · Friendship · Peer · Parent–child relationship · Transition Introduction ADHD precedes depression in a way that is consistent with a potentially causal relationship (Riglin et al., 2020), but Attention Deficit/Hyperactivity Disorder (ADHD) is a the mechanisms that might explain the relationship between DSM-5 (American Psychiatric Association, 2013) defined ADHD and later depression are unclear. ADHD impacts on neurodevelopmental disorder that is associated with a range many aspects of a young person’s life, including function- of long term adverse outcomes, including employment and ing in social, school and home life (Harpin, 2005), that education difficulties, interpersonal problems, substance might in turn increase the risk of subsequent depression. abuse, suicide and depression (Harpin, 2005; Ljung et al., For instance, ADHD is associated with increased difficulties 2014; Meinzer et al., 2014). There is strong evidence that in friendships (Mikami, 2010). Studies comparing children with an ADHD diagnosis to typically developing controls show that children with ADHD are likely to have fewer sta- * Victoria Powell ble friendships (Blachman & Hinshaw, 2002; Marton et al., powellv@cardiff.ac.uk 2015) and may be more likely to be friends with individuals Cardiff University, Cardiff, UK reported as having a learning or behaviour problem (Marton University of Surrey, Guildford, UK et  al., 2015). Friendship difficulties such as poor quality University College London, London, UK Vol.:(0123456789) 1 3 1032 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 friendships may be risk factors for subsequent depression In addition to friendships, it is important to consider (Goodyer et al., 1989). children’s other sources of social support including parent– One potential explanation of the prospective association child relationships, which can mitigate against poor men- of ADHD and depression is that social stressors that com- tal health outcomes in the presence of adversity (Collishaw monly accompany ADHD, such as friendship difficulties, et al., 2016). Some research suggests a compensatory model, lead to an increased risk of depression (Capaldi, 1992), whereby children with low levels of warmth from peers potentially by creating feelings of failure or lowered self- showed better adjustment outcomes in the presence of high esteem (Cole, 1990; Patterson & Stoolmiller, 1991). Inter- parental warmth compared to low parental warmth (Stocker, personal stress is an important precipitator of depression, 1994). Warmer relationships with parents are also associated particularly during adolescence (Flynn & Rudolph, 2011) – a with more satisfactory friendships in young people (Deković period when individuals spend increasing amounts of time & Meeus, 1997), and the influence of parental behaviour on with friends (Larson & Richards, 1991). Studies have shown the child’s friendships may be particularly important in chil- that bullying (Roy et al., 2015), social stress (Humphreys dren with ADHD (Mikami et al., 2010). Thus, it is possible et al., 2013) and peer-relationship difficulties (Powell et al., that any mediating effects of friendship in the association 2020) contribute to the prospective relationship of ADHD of ADHD and depression could be moderated by the qual- and depressive symptoms in young people. However, a ity of the parent–child relationship. However, parent–child detailed investigation of friendship features including pres- relationships are often not considered in studies of ADHD ence of friends, friendship quality and characteristics of the and friendship (Mikami, 2010) and this is especially the case friendship group (Bukowski et al., 1996) and how these are for the father-child relationship (Cabrera et al., 2018). Inves- associated with ADHD and impact on later depression is tigating the influence of relationships with both mother and lacking (Mikami, 2010; Mrug et al., 2012). Important fea- father is important, as they may exert differential influences tures of friendship include whether an individual has friends, on children’s friendships (Flynn et al., 2018; Updegraff et al., the quality of these friendships including the level of com- 2001). For example, a previous study found that mother sup- panionship and closeness, as well as the characteristics of portive behaviour and hostile behaviour, and father problem- the people that an individual is friends with (Bukowski et al., solving behaviour and hostile behaviour, influenced their 1996). It is possible that some of these features of friendship adolescent children’s interaction styles with peers (Flynn might be more important than others in the association of et al., 2018). ADHD and depression. For instance, a previous study con- The aim of this study was to investigate which aspects of ducted in secondary school children found that while retain- friendship (presence of friends, friendship quality and char- ing your best friend over time was not associated with emo- acteristics of friends) are important in increasing vulnerabil- tional outcomes, retaining poor quality friendships over time ity to depressive symptoms in those with elevated ADHD with your top three friends was associated with subsequent symptoms. A secondary aim of this study was to investigate emotional problems (Ng-Knight et al., 2019). The potential whether any indirect effects via elements of friendship were role of friendships in the prospective association between moderated by mother–child or father-child relationship qual- ADHD and depression may also differ by sex. For exam- ity. A follow-up exploratory research question was whether ple, adolescent females have been found to value aspects of indirect effects differed according to sex. The study included friendship such as companionship more highly than males, children in the first year of secondary school followed over while adolescent males have been found to value the status a seven-month timespan. of their peers more highly (Hall, 2010). Identification of fac- tors that might explain the relationship between ADHD and depression, as well as factors that moderate risk, could help Methods pinpoint new ways of supporting young people with ADHD. School is a key context where children make friends Sample (Ng-Knight et al., 2019). Classroom expectations includ- ing sustained concentration on tasks, following rules and The School Transition Adjustment Research Study (STARS; self-regulation of emotions and behaviour may often be a Ng-Knight et al., 2016) includes pupils recruited from nine poor ‘fit’ for those with ADHD, which can exacerbate poor secondary schools in Greater London selected to be repre- social outcomes for this group (Richardson et al.,  2015). sentative of schools in the local region in terms of socio- The transition from primary to secondary school is a period economic disadvantage, exam pass rates and proportions of of change (Chung et al., 1998) when good friendships can pupils from minority ethnic backgrounds. Ethical approval protect against poor mental health (Ng-Knight et al., 2019), was granted by the University College London Research but also when there is a natural disruption to established Ethics Committee. At both assessment waves, informed friendships and a need to establish new ones. assent was obtained at school for the participating children 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1033 and parents were given the opportunity to opt-out. Baseline Friendship Presence data were collected during the second half of the first term of year 7 (children aged 11-12 years; pupil and teacher reports) At baseline and follow-up, children were asked to name up and follow-up data seven months later in the summer term to three friends in order of preference as their top 3 friends. of year 7 (pupil reports) via questionnaires completed in This method was used to indicate the presence of close classroom settings. At baseline, 1712 children participated friends, as used previously (e.g., Fowler et al., 2007; Ng- (53.5% male; response rate = 87.0%). For 80.1% of these Knight et al., 2019). The following variables were derived: children, teachers returned questionnaires reporting on the number of friends at baseline (range = 0–3), stability of best child’s mental health (n = 1372). For 1020 of these 1372 friend (from the start to the end of the school year; 0 = no, children (74.3%), teachers completed the section rating the 1 = yes) and stability of ‘top three friends’ (stability of any children’s baseline ADHD symptoms. Of the 1020 children, of the three top friends; 0 = no, 1 = yes). This method of self-rated depressive symptom data at follow-up were avail- deriving friendship stability has also been used previously able for 901. Of these, there were 752 (50.3% male) for (Ng-Knight et al., 2019). whom data were additionally available on friendship and sociodemographic variables. To address bias that might Friendship Quality be caused by missingness, predictors of missingness were investigated (Supplement 1) and missing data were imputed For their best friend, children completed a 15-item version for covariates and outcome using Multiple Imputation by of the Friendships Qualities Scale, a reliable measure of five Chained Equations (MICE; White et al., 2011) with 100 different aspects of friendship quality, at baseline (FQS; imputations, which formed the analysis sample (n = 1712; Bukowski et al., 1994). The FQS consists of five subscales: Supplement 2). companionship, conflict, closeness, help and security. Items were rated on a 5-point Likert scale from “Not at all” (0) to Measures “Very much” (4). Positive scales (companionship, closeness, help and security) were summed and then negative scales ADHD Symptoms (conflict) subtracted to generate a maximum total quality score of 48. The Cronbach alpha suggested that this score ADHD symptoms were measured at baseline with the is reliable (0.85), as did the omega value (0.85), which does teacher-rated hyperactivity-inattention subscale of the not assume one-dimensionality or equal contribution of each Strengths and Difficulties Questionnaire (SDQ; Goodman, item to the total score (Deng & Chan, 2017). In addition, 1997). Five items were rated “Not True” (0), “Somewhat friendship quality across the top three friends was measured True” (1) or “Certainly True” (2) (total score range = 0–10; at baseline by a single rating for each of these friends on Cronbach alpha = 0.82). The teacher-rated hyperactivity- a 5-point smiley face rating scale from 0 (least satisfied) inattention subscale is a useful and valid screening tool for to 4 (most satisfied). Scores for the top three friends were ADHD symptoms (Goodman, 1997) with good sensitivity summed to give an overall score (maximum = 12). and specificity for detecting DSM-IV (American Psychi- atric Association, 1994) ADHD in school-aged children Classroom Friendship Group Characteristics (Goodman et al., 2000; Ullebø et al., 2011). In UK sec- ondary schools, children are typically taught by different Sociometric methods were used to identify classroom friend- teachers for different subjects. Children usually attend a ship groups at baseline. In classroom groups, children were daily registration group led by the same teacher throughout asked to draw on paper a diagram describing “which chil- the school year (the form tutor). Form tutors completed the dren in your class hang around together” by circling groups teacher questionnaire, including the SDQ. of pupils who are friends (example shown in Supplement 3). Social Cognitive Mapping software (SCM; Cairns & Cairns, Depressive Symptoms 1994; Hamm et al., 2011) was used to identify classroom friendship groups – a reliable, valid tool for this purpose Depressive symptoms were assessed at follow-up with the (Cairns et al., 1988, 1995). To describe the characteristics self-rated Short Moods and Feelings Questionnaire (SMFQ; of the classroom friendship group, this was linked to infor- Angold et  al., 1995). Thirteen items were rated “Not mation on the behavioural and emotional characteristics of true” (0), “Somewhat true” (1), or “True” (2) (total score children in that group: (i) the self-rated SDQ total difficulties range = 0–26; Cronbach alpha = 0.91). The SMFQ is a well- score (Goodman, 1997), and (ii) cooperative and disruptive validated, reliable measure of depression in adolescents with behaviour rated by peers who indicated from a list of the high sensitivity and specificity for detecting major depres - children in their class who they felt matched these descrip- sive disorder (MDD) (Thapar & McGuffin, 1998). tions (example shown in Supplement 4; the Guess Who 1 3 1034 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 peer-nomination method; Coie & Dodge, 1988; Parkhurst three friendship elements (presence/stability, quality, and & Asher, 1992). Classroom friendship group characteristics characteristics of friends), and to test associations of these were derived for each child by averaging the scores of the friendship indicators (standardised if continuous) with other pupils in their friendship group (the score of the indi- depressive symptoms. Regressions included the school vidual was not included to avoid group scores being biased class as a random effect to account for potential hierarchi- by the individual’s own score). The mean number of indi- cal data structure or clustering. viduals in a classroom friendship group was 5 (maximum: 15). A small number (5.3%) of children were social isolates (i.e., did not have a classroom friendship group) so their Indirect Effects via Friendship in the ADHD‑Depression scores could not be calculated. Association Parent–Child Relationship Quality To test whether the different features of friendship con- tributed to the association of ADHD and depressive At baseline, children who reported being in contact with symptoms, indirect effects were tested separately using their mother and/or father (or equivalents, e.g., stepparent, the [sureg] STATA command. Sureg conducts Seemingly carer) in the last month completed a measure assessing their Unrelated Regressions (Zellner, 1962), from which indi- perception of each parent’s warmth and hostility towards rect (mediated) and conditional indirect (moderated medi- them using the Iowa family interaction rating scale (Melby ation) effects can be derived in original and imputed data & Conger, 2001). The warmth scale consisted of six items (UCLA Statistical Consulting Group n.d.). Any friend- (example: “Acts loving and affectionate toward you”) rated ship variables found to have significant indirect effects on a 6-point Likert scale from “Never” (0) to “Always” were tested simultaneously in a multiple mediator model. (6) (score range = 0–36; Cronbach’s alpha for mother As indirect effects via friendship may differ according to score = 0.90; Cronbach’s alpha for father score = 0.92). The sex (Hall, 2010), a sensitivity analysis testing moderated hostility scale consisted of 4 items (example: “Criticises you mediation by the child’s sex was conducted using [sureg]. or your ideas”) rated in the same way (score range = 0–24; Cronbach’s alpha for mother score = 0.79; Cronbach’s alpha for father score = 0.82). Moderation of Indirect Effects by Parent–Child Relationships Confounders Moderated mediation was tested using [sureg] to inves- Analyses were adjusted for sex and socioeconomic and eth- tigate whether any indirect effects via friendship varied nic factors associated with depression (Gilman et al., 2002; according to the warmth or hostility of parent–child rela- Williams et al., 2015). These were socioeconomic disad- tionships. The model used tested combined moderation vantage (free school meals status), Black Minority Ethnic of the pathway from ADHD symptoms to friendship and status and English as a first language (data collected from the pathway from friendship to depression symptoms school records). (Preacher et al., 2007). Indirect effects at the mean level of the moderator and 1 standard deviation above and Data Analysis below this were calculated and plotted. For each modera- tor, significance tests (Z -tests) of the difference between Analysis was conducted in STATA (version 13). All analyses the observed indirect effect at the mean level of the mod- presented are adjusted for confounders by including these erator compared to the mean level of the moderator -1SD variables in the models as covariates. (for hostility) or +1SD (for warmth) were conducted. To investigate which path(s) in the indirect effects were being Associations of ADHD, Features of Friendship moderated, we additionally tested models with an inter- and Depression action effect on the path between ADHD symptoms and friendship or the path between friendship and depression ADHD symptoms were standardised meaning a point symptoms only. increase in SDQ hyperactivity score was equivalent to a standard deviation unit increase. Linear regression was used to test the association between ADHD symptoms and depressive symptoms 7 months later. Separate linear Results and logistic regressions were used as appropriate to test associations between ADHD symptoms and indicators of Descriptive statistics are shown in Table 1. 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1035 Table 1 Descriptive Statistics Variable Overall mean Female mean Male mean (SE) (SE) / proportion (SE) / proportion / proportion (%) (%) (%) ADHD 1.97 (0.07) 1.33 (0.08) 2.53 (0.11) Depression 3.36 (0.12) 3.85 (0.19) 2.94 (0.16) Presence of friends 2.90 (0.01) 2.93 (0.01) 2.87 (0.02) Stability: best friend 39.6% 42.1% 37.5% Stability: top three friends 62.3% 65.2% 59.8% Quality: best friend 34.81 (0.21) 37.14 (0.28) 32.78 (0.30) Quality: top three friends 10.84 (0.04) 10.81 (0.05) 10.86 (0.05) Classroom friendship group: total difficulties 7.99 (0.20) 7.18 (0.22) 8.70 (0.24) Classroom friendship group: cooperativeness 0.51 (0.01) 0.54 (0.01) 0.47 (0.01) Classroom friendship group: disruptiveness 0.12 (0.005) 0.07 (0.005) 0.17 (0.01) Number of people in classroom friendship group 4.98 (0.07) 4.59 (0.08) 5.33 (0.10) n = 1712 (53.5% male), ADHD attention deficit/hyperactivity disorder, SE standard error Associations of ADHD, Features of Friendship Indirect Effects via Friendship and Depression in the ADHD‑Depression Association ADHD symptoms at baseline were associated with nam- ADHD and depressive symptoms were associated directly ing fewer friends and lower friendship quality at baseline, and indirectly via friendship quality, both for best friend and both of which were associated with increased depressive top three friends, when tested separately (Table  2). They symptoms at follow-up. ADHD symptoms were also asso- showed independent effects when tested simultaneously in ciated with being part of a classroom friendship group that a multiple mediator model (Table 3), each accounting for had higher total difficulties and was rated as less coopera- approximately 9% of the total effect of ADHD symptoms on tive and more disruptive at baseline. Stability of the best subsequent depressive symptoms. Sensitivity analyses sug- friendship from baseline to follow-up was inversely associ- gested that the best friendship quality sub-scales of higher ated with depressive symptoms (Table 2). ADHD symp- friendship conflict and lower friendship security drove this toms were associated with depressive symptoms 7 months effect (Supplement 5). later (b = 0.48 (95% CI 0.17, 0.79) p = 0.002). The indirect effect of ADHD symptoms on depressive symptoms via friendship quality was larger for females than males (Supplement 6). Table 2 Associations of ADHD Symptoms, Friendship Variables and Depressive Symptoms Friendship variable ADHD symptoms association Variable association with Indirect effect via variable between with variable (b (95% CI) p) depressive symptoms ADHD and depressive symptoms (b (95% CI) p) (b (95% CI) p) Presence of friends -0.05 (-0.08, -0.02) 0.002 -0.65 (-1.24, -0.07) 0.029 0.03 (-0.004, 0.06) 0.095 Stability: best friend OR = 0.92 (0.80, 1.06) 0.267 -0.57 (-1.07, -0.06) 0.027 0.01 (-0.01, 0.03) 0.262 Stability: top three friends OR = 0.96 (0.82, 1.12) 0.568 -0.32 (-0.84, 0.21) 0.235 0.004 (-0.01, 0.02) 0.530 Quality: best friend -0.75 (-1.29, -0.20) 0.008 -0.72 (-0.97, -0.47) < 0.001 0.06 (0.01, 0.11) 0.011 Quality: top three friends -0.15 (-0.25, -0.05) 0.003 -0.69 (-0.94, -0.44) < 0.001 0.07 (0.02, 0.11) 0.008 Classroom friendship group: total difficulties 0.94 (0.61, 1.27) < 0.001 0.17 (-0.10, 0.43) 0.211 0.02 (-0.05, 0.08) 0.633 Classroom friendship group: cooperativeness -0.03 (-0.04, -0.02) < 0.001 -0.03 (-0.29, 0.23) 0.798 -0.01 (-0.06, 0.04) 0.663 Classroom friendship group: disruptiveness 0.03 (0.02, 0.04) < 0.001 0.15 (-0.12, 0.42) 0.262 -0.01 (-0.05, 0.08) 0.743 n = 1712, ADHD attention deficit/hyperactivity disorder, b unstandardized beta, CI confidence interval, OR odds ratio 1 3 1036 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 Table 3 Indirect Effects via Friendship Quality and Satisfaction in reach the conventional threshold for statistical significance the Association of ADHD and Depressive Symptoms in a Multiple (p = 0.084) (Fig. 1; Supplement 7). Mediator Model All results were very similar before and after imputation Effect b (95% CI) p Percentage of (Supplement 9). total effect medi- ated Indirect Effect Via 0.04 (0.01, 0.08) 0.024 9.05% Discussion Best Friendship Quality To our knowledge, this is the first detailed investigation of Indirect Effect Via 0.05 (0.006, 0.09) 0.024 9.48% how different aspects of friendships are involved in the pro- Top Three Friend- spective association of ADHD and depressive symptoms. ships Quality We also investigated whether parent–child relationship qual- Total Indirect Effect 0.09 (0.03, 0.15) 0.002 18.53% ity moderated any indirect effects via friendship in the pro- Direct Effect 0.39 (0.09, 0.70) 0.011 - spective association of ADHD and depressive symptoms. Total Effect 0.48 (0.18, 0.79) 0.002 - In a representative longitudinal study of children, ADHD n = 1712, ADHD attention deficit/hyperactivity disorder, b unstand- symptoms were associated with children having fewer ardized beta, CI confidence interval friends and lower quality friendships, which were both asso- ciated with depressive symptoms. ADHD symptoms were also associated with having a classroom friendship group Moderation of Indirect Eec ff ts by Parent–Child Relationships that had more total difficulties, was more disruptive and less cooperative. Retaining best friendships across the study Most (98.0%) children reported that they had been in period was inversely associated with depression symptoms. These findings build upon case–control studies that found an contact with their mother in the last month and 92.5% with their father. Indirect effects via top three friend- association between ADHD and having fewer friends, more friendship conflict and aggression (Blachman & Hinshaw, ships quality attenuated (p = 0.040) as mot her–child relationship warmth increased (Fig.  1; Supplements 7 2002) and an increased likelihood of being friends with a child with a learning or behaviour problem (Marton et al., & 8). Results were suggestive of this moderating effect by mother warmth acting on both paths in the indirect 2015), and an association between poor friendship quality and depression (Goodyer et al., 1989). Despite these var- effect via friendship quality between ADHD and depres- sive symptoms (Fig. 2; Supplement 8). Indirect effects via top ied associations, only friendship quality was identified as a pathway through which ADHD symptoms were associated three friendships quality also attenuated slightly as father- child hostility decreased, though this observation did not with subsequent depressive symptoms. The indirect effect a b 0.07 0.07 0.06 0.06 0.05 0.05 0.04 0.04 0.03 0.03 0.02 0.02 0.01 0.01 0 0 05 10 15 20 25 30 35 40 02468 10 12 14 Mother warmth Father hoslity Fig. 1 Plots of Moderated Mediation. Moderated mediation analy- conventional thresholds for statistical significance. The model tested sis suggested that when mother warmth increased, the indirect included interaction effects on the path between ADHD symptoms effect (beta) via quality of friendship with top three friends between and friendship quality and on the path between friendship quality and ADHD and depressive symptoms attenuated (a). When father hos- depressive symptoms. Indirect effect betas at the mean of the modera- tility decreased, the indirect effect via quality of friendship with top tor and the mean ± 1 standard deviation are plotted (n = 1712) three friends also attenuated slightly (b), though this did not meet 1 3 Indirect effect via top three friendships quality Indirect effect via top three friendships quality Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1037 Fig. 2 Moderated Mediation Model. Results supported a model whereby friendship quality accounted for part of the association between ADHD symptoms and depressive symptoms (black arrow) via mediated pathways (dashed dark grey arrows). There was some evidence to suggest that the indirect effect via top three friendships quality might be moderated by parent–child relationships (dotted light grey arrows), whereby warmer, less hostile relationships with mothers and fathers slightly decreased the size of the indi- rect effect via top three friendships quality varied slightly in magnitude predictive of depression in adolescent females than males according to the warmth and hostility of parent–child rela- (Shih et  al., 2006), particularly in females with ADHD, tionships, suggesting that positive parent–child relationships who may experience more peer-relationship difficulties than might mitigate some of the adverse effects in the indirect males with ADHD (Elkins et al., 2011). However, it seems pathway of ADHD to depressive symptoms via poor quality likely that the indirect effects observed are important to con - friendships. sider for both sexes. The findings on friendship quality align with theories of There was some evidence of moderated mediation sug- social difficulty explaining some of the link between early gesting that indirect effects of top three friendships qual- psychopathology and subsequent depression (Capaldi, ity on the link between ADHD and depressive symptoms 1992). Our findings suggest this is also a pathway that links decreased slightly as self-reports of mother–child relation- ADHD to depression symptoms. Friendship quality as a ship warmth increased and father-child relationship hostility form of perceived social support might reduce depression decreased, though the evidence for father hostility did not risk in those with elevated ADHD symptoms via increasing reach the conventional significance threshold. This draws a sense of connectedness and self-esteem, or by buffering attention to the need to consider the child’s social support against life stresses (Rueger et al., 2016). Those with ADHD across different contexts such as the parent–child relation- are more likely to experience adversity in various areas of ship. Findings were suggestive of mother warmth moderat- life (Harpin, 2005) and good social support can mitigate ing both the ADHD symptoms to friendship quality path against depressive outcomes in those who experience adver- and the friendship quality to depression symptoms path of sity (Collishaw et al., 2016; Lee et al., 2019). Sensitivity the indirect effect. This finding is consistent with that of analyses showed that conflict and security with friends a previous study that found an association between paren- (ability to disclose problems to friend and reconcile after tal behaviour and child peer-relationships in children with disagreement) appeared to be the specific elements of friend- ADHD (Mikami et al., 2010) in addition to studies that have ship quality that were important in the pathway from ADHD found parent–child relationships may be able to compensate to depressive symptoms. Elevated conflict with friends for a lack of friends (Stocker, 1994) and mitigate against has been reported for children with ADHD (Blachman poor mental health in the presence of adversity (Brennan & Hinshaw, 2002) and poor-quality friendships may also be et  al., 2003; Collishaw et  al., 2007, 2016; Lewandowski risk factors for depressive outcomes in school aged children et al., 2014). However, these moderated mediation findings (Goodyer et al., 1989). should be interpreted with caution given that results were Larger indirect effects via friendship quality were not consistent across mother and father or between warmth observed in females than males. This aligns with previous and hostility. Nevertheless, this may be explained in part evidence of females valuing aspects of friendship quality by previous findings that suggest mothers and fathers may such as companionship and intimacy more highly than males have differential effects on the friendships of their adoles- (Hall, 2010). Interpersonal stress may be more prevalent and cent children (Flynn et al., 2018; Updegraff et al., 2001). 1 3 1038 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 For instance, one study found that while mother supportive was minimal (Spratt et al., 2010). Those with ADHD may behaviour and hostile behaviour influenced their children’s under-report their depressive symptoms and over-report their interactional style with peers, for fathers, it was their prob- social competence, while those with depression might under- lem-solving behaviour and hostile behaviour that appeared rate their social ability, which could attenuate associations to be important (Flynn et al., 2018). Moderated mediation (Fraser et al., 2018; Ohan & Johnston, 2011; Whitton et al., results in the current study were also inconsistent across the 2008). In addition, we were unable to investigate whether two friendship quality measures used. While some evidence the friendships reported by the children were reciprocated in of moderation by mother warmth was observed for the indi- the current study, due to reported best friends not necessarily rect effect via friendship quality with the top three friends, attending the same school. However, children were asked strong evidence of moderation by parent–child relationship to report only their top three friendships – an approach that quality was not found for the indirect effect via friendship has been used previously (Fowler et al., 2007; Ng-Knight quality with the best friend. Measurement differences in the et al., 2019). Children were asked to report their best three two friendship variables used in the present study may have friends rather than to report a total number of friends to affected the findings, as a previous study found evidence of capture the children’s close friendships, thereby helping an interaction of friendship stability with a variable meas- to avoid some of the positive illusory bias that may affect uring quality of top three friendships, which was not found reporting on friendships in children with ADHD symptoms. for the quality of the best friendship only (Ng-Knight et al., While measures of how many friends a child has in total may 2019). capture how popular or liked that individual is on a group Limitations include that we relied on teacher reports of level (e.g., in their class or school), measures that capture child ADHD symptoms which provide a reliable measure children’s close friendships may be more predictive of later of ADHD symptoms in school, but most clinical research adjustment and depression (Narr et  al., 2019; Schneider and practitioners rely primarily on parental reports of symp- et al., 1994). In addition, while self-rated measures of friend- toms. In addition, while the SDQ hyperactivity-inattention ship presence and quality were used, peer-rated classroom subscale completed by teachers is a valid and useful screen- data was also used to measure the characteristics of the ing tool for ADHD (Goodman, 1997), it is not a diagnostic friendship group, which also might help to mitigate against measure and we were not able to investigate whether ADHD any effect of illusory biases. ADHD symptoms were associ- inattentive or hyperactive-impulsive subtypes were differen- ated with characteristics of the classroom friendship group tially associated with friendship or depression in this study. (a peer-rated variable), though it did not act as a mediator The inattentive subtype has been found to be the most com- of the association between ADHD and depressive symptoms mon subtype in population samples, particularly during ado- in the current study. lescence (Willcutt, 2012). Associations between ADHD and Strengths include use of a representative school-based emotional and peer problems may vary according to subtype sample during the first year of secondary school with (Graetz et al., 2001), and thus subtype differences are of detailed information from multiple informants on different interest for future research. Despite the longitudinal design features of friendship, in addition to parent–child relation- used, there is a possibility that reverse causation contributed ship quality. School life and transitions are important in ado- to observed associations. However, evidence suggests that lescent mental health and may be particularly challenging ADHD precedes depression in a potentially causal relation- for those with ADHD (Ford, 2020; Richardson et al., 2015). ship (Riglin et al., 2020) and that the prospective relationship Implications of this work include pinpointing quality of exists over long periods of time and when adjusting for prior friendships and parent–child relationships as important to emotional disorder symptomatology (Powell et al., 2020). consider clinically in those with ADHD for reducing depres- Moreover, a sensitivity check found that ADHD symptoms sion risk. Many peer relationship-focused interventions, were associated with subsequent depressive symptoms when which have mainly focused on peer acceptance and social adjusting for baseline self-reported SDQ emotional prob- skills thus far, have shown little success in children with lems (Goodman, 1997; Supplement 9). Although we tested neurodevelopmental disorders (Mikami, 2010). Promising indirect effects, this is not mediation analysis per se, due to directions for the development of enhanced programmes exposure and mediator being measured contemporaneously include those involving a parental component focused on (Selig & Preacher, 2009). This study had missing data, a dyadic friendship building (Gardner et al., 2019). Schools’ common problem in longitudinal data (Spratt et al., 2010). arrangements regarding awareness of friendship groups However, tests were adjusted for confounders that predicted (e.g., keeping together or separating friends) are also impor- missingness, helping to address potential bias arising from tant in ensuring children feel settled at the beginning of sec- missing data (Groenwold et al., 2012). Additionally, we con- ondary school (Keay et al., 2015), and may need additional ducted Multiple Imputation and results remained very simi- consideration in children with ADHD. Practical implications lar in imputed data, suggesting bias caused by missingness for children with ADHD in mitigating later risk of emotional 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1039 American Psychiatric Association. (1994). Diagnostic and Statistical difficulties may also involve focusing on interventions to Manual of Mental Disorders (DSM-IV) (4th ed.). Author. strengthen parent–child relationships (Abikoff et al., 2015; American Psychiatric Association. (2013). Diagnostic and Statistical Meinzer et al., 2020). Interventions aiming to improve par- Manual of Mental Disorders (DSM-V) (5th edn.) (5th ed.). ent–child interactions can have beneficial effects on the American Psychiatric Publishing. Angold, A., Costello, E., Messer, S., Pickles, A., Winder, F., & Silver, mental health of both child and parent (Sonuga-Barke et al., D. (1995). Development of a questionnaire for use in epidemiologi- 2001). cal studies of depression in children and adolescents. International Journal of Methods in Psychiatric Research, 5(4), 237–249. Blachman, D. R., & Hinshaw, S. P. (2002). Patterns of friendship among girls with and without attention-deficit/hyperactivity dis- Conclusion order. Journal of Abnormal Child Psychology, 30(6), 625–640. Brennan, P. A., Le Brocque, R., & Hammen, C. (2003). Maternal depression, parent-child relationships, and resilient outcomes We found in a large school-based sample that ADHD symp- in adolescence. Journal of the American Academy of Child and toms were associated with subsequent depressive symp- Adolescent Psychiatry, 42(12), 1469–1477. toms partly via decreased friendship quality, suggesting that  Bukowski, W. M., Hoza, B., & Boivin, M. (1994). Measuring friend- this aspect of friendship is one potential mechanism by which ship quality during pre- and early adolescence: the development and psychometric properties of the friendship qualities scale. ADHD symptoms increase risk for depressive symptoms. Journal of Social and Personal Relationships, 11(3), 471–484. Positive features of parent–child relationships seemed to Bukowski, W. M., Newcombe, A. F., & Hartup, W. W. (1996). slightly alleviate the indirect effect via friendship quality, Friendship and its significance in childhood and adolescence: highlighting the importance of considering different sources Introduction and comment. In W. M. Bukowski, A. F. Newcombe, & W. W. Hartup (Eds.), The company they keep: of social support in the child’s life. Friendship in childhood and adolescence (pp. 1–19). Cambridge University Press. Supplementary Information The online version contains supplemen- Cabrera, N. J., Volling, B. L., & Barr, R. (2018). Fathers are parents, tary material available at https: //doi.org/10.1007/s10802 -021-00798- w. too! Widening the lens on parenting for children’s development. Child Development Perspectives, 12(3), 152–157. Acknowledgements We thank the participating schools, pupils and par- Cairns, R. B., & Cairns, B. D. (1994). Lifelines and risks: pathways of ents in the School Transition and Adjustment Research Study (www.ucl. youth in our time. Cambridge University Press. ac.uk/stars). The S TARS project was funded by the Nuffield Foundation Cairns, R. B., Cairns, B. D., Neckerman, H. J., Gest, S. D., & Gariépy, (EDU/40065), but the views expressed are those of the authors and not J. L. (1988). Social networks and aggressive behavior: peer necessarily those of the Nuffield Foundation (www.nuffieldf oundat ion. support or peer rejection? Developmental Psychology, 24(6), org). This publication is the work of the authors and Victoria Powell 815–823. will serve as guarantor for the contents of this paper. Cairns, R. B., Leung, M.-C., Buchanan, L., & Cairns, B. D. (1995). Friendships and social networks in childhood and adolescence: fluidity, reliability, and interrelations. Child Development, 66(5), Compliance with Ethical Standards 1330–1345. Capaldi, D. M. (1992). Co-occurrence of conduct problems and depres- Conflict of Interest The authors declare that they have no conflict of sive symptoms in early adolescent boys: II. A 2-year follow-up interest. at Grade 8. Development and Psychopathology, 4(1), 125–144. Chung, H., Elias, M., & Schneider, K. (1998). Patterns of individual Open Access This article is licensed under a Creative Commons Attri- adjustment changes during middle school transition. Journal of bution 4.0 International License, which permits use, sharing, adapta- School Psychology, 36(1), 83–101. tion, distribution and reproduction in any medium or format, as long Coie, J. D., & Dodge, K. A. (1988). Multiple sources of data on social as you give appropriate credit to the original author(s) and the source, behavior and social status in the school: a cross-age comparison. provide a link to the Creative Commons licence, and indicate if changes Child Development, 59(3), 815–829. were made. The images or other third party material in this article are Cole, D. A. (1990). Relation of social and academic competence included in the article’s Creative Commons licence, unless indicated to depressive symptoms in childhood. Journal of Abnormal otherwise in a credit line to the material. If material is not included in Psychology, 99(4), 422–429. the article’s Creative Commons licence and your intended use is not Collishaw, S., Hammerton, G., Mahedy, L., Sellers, R., Owen, M. J., permitted by statutory regulation or exceeds the permitted use, you will Craddock, N., & Thapar, A. (2016). Mental health resilience in need to obtain permission directly from the copyright holder. To view a the adolescent offspring of parents with depression: a prospective copy of this licence, visit http://creativ ecommons .or g/licenses/b y/4.0/. longitudinal study. The Lancet Psychiatry, 3(1), 49–57. Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., & Maughan, B. (2007). Resilience to adult psychopathology follow- ing childhood maltreatment: evidence from a community sample. Child Abuse and Neglect, 31(3), 211–229. References Deković, M., & Meeus, W. (1997). Peer relations in adolescence: effects of parenting and adolescents’ self-concept. Journal of Abikoff, H. B., Thompson, M., Laver-Bradbury, C., Long, N., Adolescence, 20(2), 163–176. Forehand, R. L., Miller Brotman, L., & Sonuga-Barke, E. (2015). Deng, L., & Chan, W. (2017). Testing the difference between reliabil- Parent training for preschool ADHD: a randomized controlled trial ity coefficients alpha and omega. Educational and Psychological of specialized and generic programs. Journal of Child Psychology Measurement, 77(2), 185–203. and Psychiatry, 56(6), 618–631. 1 3 1040 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 Elkins, I. J., Malone, S., Keyes, M., Iacono, W. G., & McGue, M. Keay, A., Lang, J., & Frederickson, N. (2015). Comprehensive sup- (2011). The impact of attention-deficit/hyperactivity disorder port for peer relationships at secondary transition. Educational on preadolescent adjustment may be greater for girls than for Psychology in Practice, 31(3), 279–292. boys. Journal of Clinical Child & Adolescent Psychology, 40(4), Larson, R., & Richards, M. H. (1991). Daily companionship in late 532–545. childhood and early adolescence: changing developmental con- Flynn, H. K., Felmlee, D. H., Shu, X., & Conger, R. D. (2018). Moth- texts. Child Development, 62(2), 284–300. ers and fathers matter: the influence of parental support, hostility, Lee, H. Y., Oh, J., Kawachi, I., Heo, J., Kim, S., Lee, J. K., & Kang, and problem solving on adolescent friendships. Journal of Family D. (2019). Positive and negative social support and depressive Issues, 39(8), 2389–2412. symptoms according to economic status among adults in Korea: Flynn, M., & Rudolph, K. D. (2011). Stress generation and adolescent Cross-sectional results from the Health Examinees-Gem Study. depression: contribution of interpersonal stress responses. Journal British Medical Journal Open, 9(4), 1–10. of Abnormal Child Psychology, 39(8), 1187–1198. Lewandowski, R. E., Verdeli, H., Wickramaratne, P., Warner, V., Mancini, Ford, T. (2020). Transitional care for young adults with ADHD: A., & Weissman, M. (2014). Predictors of positive outcomes in transforming potential upheaval into smooth progression. offspring of depressed parents and non-depressed parents across 20 Epidemiology and Psychiatric Sciences, 29, e87. years. Journal of Child and Family Studies, 23(5), 800–811. Fowler, T., Shelton, K., Lifford, K., Rice, F., McBride, A., Nikolov, Ljung, T., Chen, Q., Lichtenstein, P., & Larsson, H. (2014). Common I., & Van Den Bree, M. B. M. (2007). Genetic and environmen- Etiological Factors of Attention-Deficit/Hyperactivity Disorder tal influences on the relationship between peer alcohol use and and Suicidal Behavior: A Population-Based Study in Sweden. own alcohol use in adolescents. Addiction, 102(6), 894–903. JAMA Psychiatry, 71(8), 958–964. Fraser, A., Cooper, M., Agha, S. S., Collishaw, S., Rice, F., Thapar, Marton, I., Wiener, J., Rogers, M., & Moore, C. (2015). Friendship A., & Eyre, O. (2018). The presentation of depression symp- characteristics of children with ADHD. Journal of Attention toms in attention-deficit/hyperactivity disorder: Comparing Disorders, 19(10), 872–881. child and parent reports. Child and Adolescent Mental Health, Meinzer, M. C., Felton, J. W., Oddo, L. E., Rubin, K. H., & Chronis- 23(3), 243–250. Tuscano, A. (2020). Do ADHD symptoms and relationship quality Gardner, D. M., Gerdes, A. C., & Weinberger, K. (2019). Examination with mothers and best friends across high school predict depres- of a parent-assisted, friendship-building program for adolescents sive symptoms for adolescents? Journal of Attention Disorders. with ADHD. Journal of Attention Disorders, 23(4), 363–373. Meinzer, M. C., Pettit, J. W., & Viswesvaran, C. (2014). The co- Gilman, S. E., Kawachi, I., Fitzmaurice, G. M., & Buka, S. L. (2002). occurrence of attention-deficit/hyperactivity disorder and unipolar Socioeconomic status in childhood and the lifetime risk of depression in children and adolescents: A meta-analytic review. major depression. International Journal of Epidemiology, 31(2), Clinical Psychology Review, 34(8), 595–607. 359–367. Melby, J. N., & Conger, R. D. (2001). The Iowa Family Interaction Goodman, R. (1997). The strengths and difficulties questionnaire: a Rating Scales: Instrument summary. Family Observational research note. Journal of Child Psychology and Psychiatry, 38, Coding Systems: Resources for Systemic Research., pp. 33–58. 581–586. Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers. Goodman, R., Ford, T., Simmons, H., Gatward, R., & Meltzer, H. Mikami, A. Y. (2010). The importance of friendship for youth with (2000). Using the strengths and difficulties questionnaire (SDQ) attention-deficit/hyperactivity disorder. Clinical Child and Family to screen for child psychiatric disorders in a community sample. Psychology Review, 13(2), 181–198. British Journal of Psychiatry, 177(6), 534–539. Mikami, A. Y., Jack, A., Emeh, C. C., & Stephens, H. F. (2010). Goodyer, I. M., Wright, C., & Altham, P. M. E. (1989). Recent friend- Parental influence on children with attention-deficit/hyperactivity ship in anxious and depressed school age children. Psychological disorder: I. Relationships between parent behaviors and child peer Medicine, 19, 165–174. status. Journal of Abnormal Child Psychology, 38(6), 721–736. Graetz, B. W., Sawyer, M. G., Hazell, P. L., Arney, F., & Baghurst, Mrug, S., Molina, B. S. G., Hoza, B., Gerdes, A. C., Hinshaw, S. P., P. (2001). Validity of DSM-IV ADHD subtypes in a nationally Hechtman, L., & Arnold, L. E. (2012). Peer rejection and friend- representative sample of Australian children and adolescents. ships in children with Attention-Deficit/Hyperactivity Disorder: Journal of the American Academy of Child & Adolescent contributions to long-term outcomes. Journal of Abnormal Child Psychiatry, 40(12), 1410–1417. Psychology, 40(6), 1013–1026. Groenwold, R. H. H., Donders, A. R. T., Roes, K. C. B., Harrell, F. E. Narr, R. K., Allen, J. P., Tan, J. S., & Loeb, E. L. (2019). Close friend- J., & Moons, K. G. M. (2012). Dealing with missing outcome data ship strength and broader peer group desirability as differential in randomized trials and observational studies. American Journal predictors of adult mental health. Child Development, 90(1), of Epidemiology, 175(3), 210–217. 298–313. Hall, J. A. (2010). Sex differences in friendship expectations: a meta- Ng-Knight, T., Shelton, K. H., Riglin, L., Frederickson, N., McManus, analysis. Journal of Social and Personal Relationships, 28(6), I. C., & Rice, F. (2019). ‘Best friends forever’? Friendship 723–747. stability across school transition and associations with mental Hamm, J. V., Farmer, T. W., Dadisman, K., Gravelle, M., & Murray, A. health and educational attainment. British Journal of Educational R. (2011). Teachers’ attunement to students’ peer group affiliations Psychology, 89(4), 585–599. as a source of improved student experiences of the school Ng-Knight, T., Shelton, K. H., Riglin, L., McManus, I. C., social–affective context following the middle school transition. Frederickson, N., & Rice, F. (2016). A longitudinal study of self- Journal of Applied Developmental Psychology, 32(5), 267–277. control at the transition to secondary school: Considering the Harpin, V. A. (2005). The effect of ADHD on the life of an individual, role of pubertal status and parenting. Journal of Adolescence, their family, and community from preschool to adult life. Archives 50, 44–55. of Disease in Childhood, 90, i2–i7. Ohan, J. L., & Johnston, C. (2011). Positive illusions of social com- Humphreys, K. L., Katz, S. J., Lee, S. S., Hammen, C., Brennan, P. A., petence in girls with and without ADHD. Journal of Abnormal & Najman, J. M. (2013). The association of ADHD and depres- Child Psychology, 39(4), 527–539.  sion: mediation by peer problems and parent-child difficulties in Parkhurst, J. T., & Asher, S. R. (1992). Peer rejection in middle school: subgroup differences in behavior, loneliness, and interpersonal two complementary samples. Journal of Abnormal Psychology, concerns. Developmental Psychology, 28, 231–241. 122(3), 854–867. 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1041 Patterson, G. R., & Stoolmiller, M. (1991). Replications of a dual fail- Stocker, C. M. (1994). Children’s perceptions of relationships with ure model for boys’ depressed mood. Journal of Consulting and siblings, friends, and mothers: compensatory processes and links Clinical Psychology, 59(4), 491–498. with adjustment. Journal of Child Psychology and Psychiatry, Powell, V., Riglin, L., Hammerton, G., Eyre, O., Martin, J., Anney, 35(8), 1447–1459. R., & Rice, F. (2020). What explains the link between childhood Thapar, A., & McGuffin, P. (1998). Validity of the shortened Mood and ADHD and adolescent depression? Investigating the role of Feelings Questionnaire in a community sample of children and peer relationships and academic attainment. European Child & adolescents: a preliminary research note. Psychiatry Research, Adolescent Psychiatry, 29(11), 1581–1591. 81(2), 259–268. Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing mod- UCLA Statistical Consulting Group. (n.d.). How can I compute indirect erated mediation hypotheses: Theory, methods, and prescriptions. effects with imputed data? (method 1) | STATA FAQ. Retrieved Multivariate Behavioral Research, 42(1), 185–227. January 31, 2020, from https ://stats .idre.ucla.edu/stata /faq/how- Richardson, M., Moore, D. A., Gwernan-Jones, R., Thompson-Coon, can-i-compu te-indir ect-effec ts-with-imput ed-data-metho d-1/ J., Ukoumunne, O., Rogers, M., & Ford, T. J. (2015). Non-phar- Ullebø, A. K., Posserud, M.-B., Heiervang, E., Gillberg, C., & Obel, C. macological interventions for attention-deficit/hyperactivity dis- (2011). Screening for the attention deficit hyperactivity disorder order (ADHD) delivered in school settings: systematic reviews of phenotype using the strength and difficulties questionnaire. quantitative and qualitative research. Health Technology Assess- European Child & Adolescent Psychiatry, 20(9), 451–458. ment, 19(45), 1–470. Updegraff, K. A., McHale, S. M., Crouter, A. C., & Kupanoff, K. Riglin, L., Leppert, B., Dardani, C., Thapar, A. K., Rice, F., (2001). Parents’ involvement in adolescents’ peer relationships: a O’Donovan, M. C., Thapar, A. (2020). ADHD and depression: comparison of mothers’ and fathers’ roles. Journal of Marriage investigating a causal explanation. Psychological Medicine, 1–8. and Family, 63(3), 655–668. Roy, A., Hartman, C. A., Veenstra, R., & Oldehinkel, A. J. (2015). White, I. R., Royston, P., & Wood, A. M. (2011). Multiple imputation Peer dislike and victimisation in pathways from ADHD symptoms using chained equations: issues and guidance for practice. to depression. European Child & Adolescent Psychiatry, 24(8), Statistics in Medicine, 30(4), 377–399. 887–895. Whitton, S. W., Larson, J. J., & Hauser, S. T. (2008). Depressive Rueger, S. Y., Malecki, C. K., Pyun, Y., Aycock, C., & Coyle, S. (2016). symptoms and bias in perceived social competence among young A meta-analytic review of the association between perceived adults. Journal of Clinical Psychology, 64(7), 791–805. social support and depression in childhood and adolescence. Willcutt, E. G. (2012). The prevalence of DSM-IV attention-deficit/ Psychological Bulletin, 142, 1017–1067. hyperactivity disorder: a meta-analytic review. Neurotherapeutics, Schneider, B. H., Wiener, J., & Murphy, K. (1994). Children’s friend- 9(3), 490–499. ships: the giant step beyond peer acceptance. Journal of Social Williams, E. D., Tillin, T., Richards, M., Tuson, C., Chaturvedi, N., and Personal Relationships, 11(3), 323–340. Hughes, A. D., & Stewart, R. (2015). Depressive symptoms are Selig, J. P., & Preacher, K. J. (2009). Mediation Models for doubled in older British South Asian and Black Caribbean people Longitudinal Data in Developmental Research. Research in compared with Europeans: associations with excess co-morbidity Human Development, 6(2–3), 144–164. and socioeconomic disadvantage. Psychological Medicine, 45(9), Shih, J. H., Eberhart, N. K., Hammen, C. L., & Brennan, P. A. (2006). 1861–1871. Differential exposure and reactivity to interpersonal stress predict Zellner, A. (1962). An efficient method of estimating seemingly sex die ff rences in adolescent depression. Journal of Clinical Child unrelated regressions and tests for aggregation bias. Journal of & Adolescent Psychology, 35(1), 103–115. the American Statistical Association, 57(298), 348–368. Sonuga-Barke, E. J., Daley, D., Thompson, M., Laver-Bradbury, C., & Weeks, A. (2001). Parent-based therapies for preschool attention- Publisher’s Note Springer Nature remains neutral with regard to deficit/hyperactivity disorder: a randomized, controlled trial with jurisdictional claims in published maps and institutional affiliations. a community sample. Journal of the American Academy of Child and Adolescent Psychiatry, 40(4), 402–408. Spratt, M., Carpenter, J., Sterne, J. A. C., Carlin, J. B., Heron, J., Henderson, J., & Tilling, K. (2010). Strategies for multiple imputation in longitudinal studies. American Journal of Epidemiology, 172(4), 478–487. 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Abnormal Child Psychology Springer Journals

Investigating Friendship Difficulties in the Pathway from ADHD to Depressive Symptoms. Can Parent–Child Relationships Compensate?

Download PDF
11 pages

Loading next page...
 
/lp/springer-journals/investigating-friendship-difficulties-in-the-pathway-from-adhd-to-Xl980sfBhD

References (97)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
ISSN
0091-0627
eISSN
2730-7174
DOI
10.1007/s10802-021-00798-w
Publisher site
See Article on Publisher Site

Abstract

Attention deficit/hyperactivity disorder (ADHD) is associated with friendship difficulties. This may partly account for the increasingly recognised association between ADHD and subsequent depression. Little is known about the types of friendship dic ffi ulties that could contribute to the association between ADHD and depressive symptoms and whether other relationships, such as parent–child relationships, can mitigate against potential adverse effects of friendship difficulties. In a representative UK school sample (n = 1712), three main features of friendship (presence of friends, friendship quality and characteristics of the individual’s classroom friendship group) were assessed in a longitudinal study with two assessment waves (W1, W2) during the first year of secondary school (children aged 11-12 years). These friendship features (W1) were investigated as potential mediators of the prospective association between teacher-rated ADHD symptoms (W1) and self-rated depressive symptoms (W2) seven months later. Parent–child relationship quality (W1) was tested as a moderator of any indirect effects of ADHD on depression via friendship. ADHD symptoms were inversely associated with friendship presence, friendship quality and positive characteristics of classroom friendship groups. Depressive symptoms were inversely associated with presence and quality of friendships. Friendship quality had indirect effects in the association between ADHD and subsequent depressive symptoms. There was some evidence of moderated mediation, whereby indirect effects via friendship quality attenuated slightly as children reported warmer parent–child relationships. This highlights the importance of considering the quality of friendships and parent–child relationships in children with ADHD symptoms. Fostering good quality relationships may help disrupt the link between ADHD symptomology and subsequent depression risk. Keywords ADHD · Depression · Friendship · Peer · Parent–child relationship · Transition Introduction ADHD precedes depression in a way that is consistent with a potentially causal relationship (Riglin et al., 2020), but Attention Deficit/Hyperactivity Disorder (ADHD) is a the mechanisms that might explain the relationship between DSM-5 (American Psychiatric Association, 2013) defined ADHD and later depression are unclear. ADHD impacts on neurodevelopmental disorder that is associated with a range many aspects of a young person’s life, including function- of long term adverse outcomes, including employment and ing in social, school and home life (Harpin, 2005), that education difficulties, interpersonal problems, substance might in turn increase the risk of subsequent depression. abuse, suicide and depression (Harpin, 2005; Ljung et al., For instance, ADHD is associated with increased difficulties 2014; Meinzer et al., 2014). There is strong evidence that in friendships (Mikami, 2010). Studies comparing children with an ADHD diagnosis to typically developing controls show that children with ADHD are likely to have fewer sta- * Victoria Powell ble friendships (Blachman & Hinshaw, 2002; Marton et al., powellv@cardiff.ac.uk 2015) and may be more likely to be friends with individuals Cardiff University, Cardiff, UK reported as having a learning or behaviour problem (Marton University of Surrey, Guildford, UK et  al., 2015). Friendship difficulties such as poor quality University College London, London, UK Vol.:(0123456789) 1 3 1032 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 friendships may be risk factors for subsequent depression In addition to friendships, it is important to consider (Goodyer et al., 1989). children’s other sources of social support including parent– One potential explanation of the prospective association child relationships, which can mitigate against poor men- of ADHD and depression is that social stressors that com- tal health outcomes in the presence of adversity (Collishaw monly accompany ADHD, such as friendship difficulties, et al., 2016). Some research suggests a compensatory model, lead to an increased risk of depression (Capaldi, 1992), whereby children with low levels of warmth from peers potentially by creating feelings of failure or lowered self- showed better adjustment outcomes in the presence of high esteem (Cole, 1990; Patterson & Stoolmiller, 1991). Inter- parental warmth compared to low parental warmth (Stocker, personal stress is an important precipitator of depression, 1994). Warmer relationships with parents are also associated particularly during adolescence (Flynn & Rudolph, 2011) – a with more satisfactory friendships in young people (Deković period when individuals spend increasing amounts of time & Meeus, 1997), and the influence of parental behaviour on with friends (Larson & Richards, 1991). Studies have shown the child’s friendships may be particularly important in chil- that bullying (Roy et al., 2015), social stress (Humphreys dren with ADHD (Mikami et al., 2010). Thus, it is possible et al., 2013) and peer-relationship difficulties (Powell et al., that any mediating effects of friendship in the association 2020) contribute to the prospective relationship of ADHD of ADHD and depression could be moderated by the qual- and depressive symptoms in young people. However, a ity of the parent–child relationship. However, parent–child detailed investigation of friendship features including pres- relationships are often not considered in studies of ADHD ence of friends, friendship quality and characteristics of the and friendship (Mikami, 2010) and this is especially the case friendship group (Bukowski et al., 1996) and how these are for the father-child relationship (Cabrera et al., 2018). Inves- associated with ADHD and impact on later depression is tigating the influence of relationships with both mother and lacking (Mikami, 2010; Mrug et al., 2012). Important fea- father is important, as they may exert differential influences tures of friendship include whether an individual has friends, on children’s friendships (Flynn et al., 2018; Updegraff et al., the quality of these friendships including the level of com- 2001). For example, a previous study found that mother sup- panionship and closeness, as well as the characteristics of portive behaviour and hostile behaviour, and father problem- the people that an individual is friends with (Bukowski et al., solving behaviour and hostile behaviour, influenced their 1996). It is possible that some of these features of friendship adolescent children’s interaction styles with peers (Flynn might be more important than others in the association of et al., 2018). ADHD and depression. For instance, a previous study con- The aim of this study was to investigate which aspects of ducted in secondary school children found that while retain- friendship (presence of friends, friendship quality and char- ing your best friend over time was not associated with emo- acteristics of friends) are important in increasing vulnerabil- tional outcomes, retaining poor quality friendships over time ity to depressive symptoms in those with elevated ADHD with your top three friends was associated with subsequent symptoms. A secondary aim of this study was to investigate emotional problems (Ng-Knight et al., 2019). The potential whether any indirect effects via elements of friendship were role of friendships in the prospective association between moderated by mother–child or father-child relationship qual- ADHD and depression may also differ by sex. For exam- ity. A follow-up exploratory research question was whether ple, adolescent females have been found to value aspects of indirect effects differed according to sex. The study included friendship such as companionship more highly than males, children in the first year of secondary school followed over while adolescent males have been found to value the status a seven-month timespan. of their peers more highly (Hall, 2010). Identification of fac- tors that might explain the relationship between ADHD and depression, as well as factors that moderate risk, could help Methods pinpoint new ways of supporting young people with ADHD. School is a key context where children make friends Sample (Ng-Knight et al., 2019). Classroom expectations includ- ing sustained concentration on tasks, following rules and The School Transition Adjustment Research Study (STARS; self-regulation of emotions and behaviour may often be a Ng-Knight et al., 2016) includes pupils recruited from nine poor ‘fit’ for those with ADHD, which can exacerbate poor secondary schools in Greater London selected to be repre- social outcomes for this group (Richardson et al.,  2015). sentative of schools in the local region in terms of socio- The transition from primary to secondary school is a period economic disadvantage, exam pass rates and proportions of of change (Chung et al., 1998) when good friendships can pupils from minority ethnic backgrounds. Ethical approval protect against poor mental health (Ng-Knight et al., 2019), was granted by the University College London Research but also when there is a natural disruption to established Ethics Committee. At both assessment waves, informed friendships and a need to establish new ones. assent was obtained at school for the participating children 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1033 and parents were given the opportunity to opt-out. Baseline Friendship Presence data were collected during the second half of the first term of year 7 (children aged 11-12 years; pupil and teacher reports) At baseline and follow-up, children were asked to name up and follow-up data seven months later in the summer term to three friends in order of preference as their top 3 friends. of year 7 (pupil reports) via questionnaires completed in This method was used to indicate the presence of close classroom settings. At baseline, 1712 children participated friends, as used previously (e.g., Fowler et al., 2007; Ng- (53.5% male; response rate = 87.0%). For 80.1% of these Knight et al., 2019). The following variables were derived: children, teachers returned questionnaires reporting on the number of friends at baseline (range = 0–3), stability of best child’s mental health (n = 1372). For 1020 of these 1372 friend (from the start to the end of the school year; 0 = no, children (74.3%), teachers completed the section rating the 1 = yes) and stability of ‘top three friends’ (stability of any children’s baseline ADHD symptoms. Of the 1020 children, of the three top friends; 0 = no, 1 = yes). This method of self-rated depressive symptom data at follow-up were avail- deriving friendship stability has also been used previously able for 901. Of these, there were 752 (50.3% male) for (Ng-Knight et al., 2019). whom data were additionally available on friendship and sociodemographic variables. To address bias that might Friendship Quality be caused by missingness, predictors of missingness were investigated (Supplement 1) and missing data were imputed For their best friend, children completed a 15-item version for covariates and outcome using Multiple Imputation by of the Friendships Qualities Scale, a reliable measure of five Chained Equations (MICE; White et al., 2011) with 100 different aspects of friendship quality, at baseline (FQS; imputations, which formed the analysis sample (n = 1712; Bukowski et al., 1994). The FQS consists of five subscales: Supplement 2). companionship, conflict, closeness, help and security. Items were rated on a 5-point Likert scale from “Not at all” (0) to Measures “Very much” (4). Positive scales (companionship, closeness, help and security) were summed and then negative scales ADHD Symptoms (conflict) subtracted to generate a maximum total quality score of 48. The Cronbach alpha suggested that this score ADHD symptoms were measured at baseline with the is reliable (0.85), as did the omega value (0.85), which does teacher-rated hyperactivity-inattention subscale of the not assume one-dimensionality or equal contribution of each Strengths and Difficulties Questionnaire (SDQ; Goodman, item to the total score (Deng & Chan, 2017). In addition, 1997). Five items were rated “Not True” (0), “Somewhat friendship quality across the top three friends was measured True” (1) or “Certainly True” (2) (total score range = 0–10; at baseline by a single rating for each of these friends on Cronbach alpha = 0.82). The teacher-rated hyperactivity- a 5-point smiley face rating scale from 0 (least satisfied) inattention subscale is a useful and valid screening tool for to 4 (most satisfied). Scores for the top three friends were ADHD symptoms (Goodman, 1997) with good sensitivity summed to give an overall score (maximum = 12). and specificity for detecting DSM-IV (American Psychi- atric Association, 1994) ADHD in school-aged children Classroom Friendship Group Characteristics (Goodman et al., 2000; Ullebø et al., 2011). In UK sec- ondary schools, children are typically taught by different Sociometric methods were used to identify classroom friend- teachers for different subjects. Children usually attend a ship groups at baseline. In classroom groups, children were daily registration group led by the same teacher throughout asked to draw on paper a diagram describing “which chil- the school year (the form tutor). Form tutors completed the dren in your class hang around together” by circling groups teacher questionnaire, including the SDQ. of pupils who are friends (example shown in Supplement 3). Social Cognitive Mapping software (SCM; Cairns & Cairns, Depressive Symptoms 1994; Hamm et al., 2011) was used to identify classroom friendship groups – a reliable, valid tool for this purpose Depressive symptoms were assessed at follow-up with the (Cairns et al., 1988, 1995). To describe the characteristics self-rated Short Moods and Feelings Questionnaire (SMFQ; of the classroom friendship group, this was linked to infor- Angold et  al., 1995). Thirteen items were rated “Not mation on the behavioural and emotional characteristics of true” (0), “Somewhat true” (1), or “True” (2) (total score children in that group: (i) the self-rated SDQ total difficulties range = 0–26; Cronbach alpha = 0.91). The SMFQ is a well- score (Goodman, 1997), and (ii) cooperative and disruptive validated, reliable measure of depression in adolescents with behaviour rated by peers who indicated from a list of the high sensitivity and specificity for detecting major depres - children in their class who they felt matched these descrip- sive disorder (MDD) (Thapar & McGuffin, 1998). tions (example shown in Supplement 4; the Guess Who 1 3 1034 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 peer-nomination method; Coie & Dodge, 1988; Parkhurst three friendship elements (presence/stability, quality, and & Asher, 1992). Classroom friendship group characteristics characteristics of friends), and to test associations of these were derived for each child by averaging the scores of the friendship indicators (standardised if continuous) with other pupils in their friendship group (the score of the indi- depressive symptoms. Regressions included the school vidual was not included to avoid group scores being biased class as a random effect to account for potential hierarchi- by the individual’s own score). The mean number of indi- cal data structure or clustering. viduals in a classroom friendship group was 5 (maximum: 15). A small number (5.3%) of children were social isolates (i.e., did not have a classroom friendship group) so their Indirect Effects via Friendship in the ADHD‑Depression scores could not be calculated. Association Parent–Child Relationship Quality To test whether the different features of friendship con- tributed to the association of ADHD and depressive At baseline, children who reported being in contact with symptoms, indirect effects were tested separately using their mother and/or father (or equivalents, e.g., stepparent, the [sureg] STATA command. Sureg conducts Seemingly carer) in the last month completed a measure assessing their Unrelated Regressions (Zellner, 1962), from which indi- perception of each parent’s warmth and hostility towards rect (mediated) and conditional indirect (moderated medi- them using the Iowa family interaction rating scale (Melby ation) effects can be derived in original and imputed data & Conger, 2001). The warmth scale consisted of six items (UCLA Statistical Consulting Group n.d.). Any friend- (example: “Acts loving and affectionate toward you”) rated ship variables found to have significant indirect effects on a 6-point Likert scale from “Never” (0) to “Always” were tested simultaneously in a multiple mediator model. (6) (score range = 0–36; Cronbach’s alpha for mother As indirect effects via friendship may differ according to score = 0.90; Cronbach’s alpha for father score = 0.92). The sex (Hall, 2010), a sensitivity analysis testing moderated hostility scale consisted of 4 items (example: “Criticises you mediation by the child’s sex was conducted using [sureg]. or your ideas”) rated in the same way (score range = 0–24; Cronbach’s alpha for mother score = 0.79; Cronbach’s alpha for father score = 0.82). Moderation of Indirect Effects by Parent–Child Relationships Confounders Moderated mediation was tested using [sureg] to inves- Analyses were adjusted for sex and socioeconomic and eth- tigate whether any indirect effects via friendship varied nic factors associated with depression (Gilman et al., 2002; according to the warmth or hostility of parent–child rela- Williams et al., 2015). These were socioeconomic disad- tionships. The model used tested combined moderation vantage (free school meals status), Black Minority Ethnic of the pathway from ADHD symptoms to friendship and status and English as a first language (data collected from the pathway from friendship to depression symptoms school records). (Preacher et al., 2007). Indirect effects at the mean level of the moderator and 1 standard deviation above and Data Analysis below this were calculated and plotted. For each modera- tor, significance tests (Z -tests) of the difference between Analysis was conducted in STATA (version 13). All analyses the observed indirect effect at the mean level of the mod- presented are adjusted for confounders by including these erator compared to the mean level of the moderator -1SD variables in the models as covariates. (for hostility) or +1SD (for warmth) were conducted. To investigate which path(s) in the indirect effects were being Associations of ADHD, Features of Friendship moderated, we additionally tested models with an inter- and Depression action effect on the path between ADHD symptoms and friendship or the path between friendship and depression ADHD symptoms were standardised meaning a point symptoms only. increase in SDQ hyperactivity score was equivalent to a standard deviation unit increase. Linear regression was used to test the association between ADHD symptoms and depressive symptoms 7 months later. Separate linear Results and logistic regressions were used as appropriate to test associations between ADHD symptoms and indicators of Descriptive statistics are shown in Table 1. 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1035 Table 1 Descriptive Statistics Variable Overall mean Female mean Male mean (SE) (SE) / proportion (SE) / proportion / proportion (%) (%) (%) ADHD 1.97 (0.07) 1.33 (0.08) 2.53 (0.11) Depression 3.36 (0.12) 3.85 (0.19) 2.94 (0.16) Presence of friends 2.90 (0.01) 2.93 (0.01) 2.87 (0.02) Stability: best friend 39.6% 42.1% 37.5% Stability: top three friends 62.3% 65.2% 59.8% Quality: best friend 34.81 (0.21) 37.14 (0.28) 32.78 (0.30) Quality: top three friends 10.84 (0.04) 10.81 (0.05) 10.86 (0.05) Classroom friendship group: total difficulties 7.99 (0.20) 7.18 (0.22) 8.70 (0.24) Classroom friendship group: cooperativeness 0.51 (0.01) 0.54 (0.01) 0.47 (0.01) Classroom friendship group: disruptiveness 0.12 (0.005) 0.07 (0.005) 0.17 (0.01) Number of people in classroom friendship group 4.98 (0.07) 4.59 (0.08) 5.33 (0.10) n = 1712 (53.5% male), ADHD attention deficit/hyperactivity disorder, SE standard error Associations of ADHD, Features of Friendship Indirect Effects via Friendship and Depression in the ADHD‑Depression Association ADHD symptoms at baseline were associated with nam- ADHD and depressive symptoms were associated directly ing fewer friends and lower friendship quality at baseline, and indirectly via friendship quality, both for best friend and both of which were associated with increased depressive top three friends, when tested separately (Table  2). They symptoms at follow-up. ADHD symptoms were also asso- showed independent effects when tested simultaneously in ciated with being part of a classroom friendship group that a multiple mediator model (Table 3), each accounting for had higher total difficulties and was rated as less coopera- approximately 9% of the total effect of ADHD symptoms on tive and more disruptive at baseline. Stability of the best subsequent depressive symptoms. Sensitivity analyses sug- friendship from baseline to follow-up was inversely associ- gested that the best friendship quality sub-scales of higher ated with depressive symptoms (Table 2). ADHD symp- friendship conflict and lower friendship security drove this toms were associated with depressive symptoms 7 months effect (Supplement 5). later (b = 0.48 (95% CI 0.17, 0.79) p = 0.002). The indirect effect of ADHD symptoms on depressive symptoms via friendship quality was larger for females than males (Supplement 6). Table 2 Associations of ADHD Symptoms, Friendship Variables and Depressive Symptoms Friendship variable ADHD symptoms association Variable association with Indirect effect via variable between with variable (b (95% CI) p) depressive symptoms ADHD and depressive symptoms (b (95% CI) p) (b (95% CI) p) Presence of friends -0.05 (-0.08, -0.02) 0.002 -0.65 (-1.24, -0.07) 0.029 0.03 (-0.004, 0.06) 0.095 Stability: best friend OR = 0.92 (0.80, 1.06) 0.267 -0.57 (-1.07, -0.06) 0.027 0.01 (-0.01, 0.03) 0.262 Stability: top three friends OR = 0.96 (0.82, 1.12) 0.568 -0.32 (-0.84, 0.21) 0.235 0.004 (-0.01, 0.02) 0.530 Quality: best friend -0.75 (-1.29, -0.20) 0.008 -0.72 (-0.97, -0.47) < 0.001 0.06 (0.01, 0.11) 0.011 Quality: top three friends -0.15 (-0.25, -0.05) 0.003 -0.69 (-0.94, -0.44) < 0.001 0.07 (0.02, 0.11) 0.008 Classroom friendship group: total difficulties 0.94 (0.61, 1.27) < 0.001 0.17 (-0.10, 0.43) 0.211 0.02 (-0.05, 0.08) 0.633 Classroom friendship group: cooperativeness -0.03 (-0.04, -0.02) < 0.001 -0.03 (-0.29, 0.23) 0.798 -0.01 (-0.06, 0.04) 0.663 Classroom friendship group: disruptiveness 0.03 (0.02, 0.04) < 0.001 0.15 (-0.12, 0.42) 0.262 -0.01 (-0.05, 0.08) 0.743 n = 1712, ADHD attention deficit/hyperactivity disorder, b unstandardized beta, CI confidence interval, OR odds ratio 1 3 1036 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 Table 3 Indirect Effects via Friendship Quality and Satisfaction in reach the conventional threshold for statistical significance the Association of ADHD and Depressive Symptoms in a Multiple (p = 0.084) (Fig. 1; Supplement 7). Mediator Model All results were very similar before and after imputation Effect b (95% CI) p Percentage of (Supplement 9). total effect medi- ated Indirect Effect Via 0.04 (0.01, 0.08) 0.024 9.05% Discussion Best Friendship Quality To our knowledge, this is the first detailed investigation of Indirect Effect Via 0.05 (0.006, 0.09) 0.024 9.48% how different aspects of friendships are involved in the pro- Top Three Friend- spective association of ADHD and depressive symptoms. ships Quality We also investigated whether parent–child relationship qual- Total Indirect Effect 0.09 (0.03, 0.15) 0.002 18.53% ity moderated any indirect effects via friendship in the pro- Direct Effect 0.39 (0.09, 0.70) 0.011 - spective association of ADHD and depressive symptoms. Total Effect 0.48 (0.18, 0.79) 0.002 - In a representative longitudinal study of children, ADHD n = 1712, ADHD attention deficit/hyperactivity disorder, b unstand- symptoms were associated with children having fewer ardized beta, CI confidence interval friends and lower quality friendships, which were both asso- ciated with depressive symptoms. ADHD symptoms were also associated with having a classroom friendship group Moderation of Indirect Eec ff ts by Parent–Child Relationships that had more total difficulties, was more disruptive and less cooperative. Retaining best friendships across the study Most (98.0%) children reported that they had been in period was inversely associated with depression symptoms. These findings build upon case–control studies that found an contact with their mother in the last month and 92.5% with their father. Indirect effects via top three friend- association between ADHD and having fewer friends, more friendship conflict and aggression (Blachman & Hinshaw, ships quality attenuated (p = 0.040) as mot her–child relationship warmth increased (Fig.  1; Supplements 7 2002) and an increased likelihood of being friends with a child with a learning or behaviour problem (Marton et al., & 8). Results were suggestive of this moderating effect by mother warmth acting on both paths in the indirect 2015), and an association between poor friendship quality and depression (Goodyer et al., 1989). Despite these var- effect via friendship quality between ADHD and depres- sive symptoms (Fig. 2; Supplement 8). Indirect effects via top ied associations, only friendship quality was identified as a pathway through which ADHD symptoms were associated three friendships quality also attenuated slightly as father- child hostility decreased, though this observation did not with subsequent depressive symptoms. The indirect effect a b 0.07 0.07 0.06 0.06 0.05 0.05 0.04 0.04 0.03 0.03 0.02 0.02 0.01 0.01 0 0 05 10 15 20 25 30 35 40 02468 10 12 14 Mother warmth Father hoslity Fig. 1 Plots of Moderated Mediation. Moderated mediation analy- conventional thresholds for statistical significance. The model tested sis suggested that when mother warmth increased, the indirect included interaction effects on the path between ADHD symptoms effect (beta) via quality of friendship with top three friends between and friendship quality and on the path between friendship quality and ADHD and depressive symptoms attenuated (a). When father hos- depressive symptoms. Indirect effect betas at the mean of the modera- tility decreased, the indirect effect via quality of friendship with top tor and the mean ± 1 standard deviation are plotted (n = 1712) three friends also attenuated slightly (b), though this did not meet 1 3 Indirect effect via top three friendships quality Indirect effect via top three friendships quality Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1037 Fig. 2 Moderated Mediation Model. Results supported a model whereby friendship quality accounted for part of the association between ADHD symptoms and depressive symptoms (black arrow) via mediated pathways (dashed dark grey arrows). There was some evidence to suggest that the indirect effect via top three friendships quality might be moderated by parent–child relationships (dotted light grey arrows), whereby warmer, less hostile relationships with mothers and fathers slightly decreased the size of the indi- rect effect via top three friendships quality varied slightly in magnitude predictive of depression in adolescent females than males according to the warmth and hostility of parent–child rela- (Shih et  al., 2006), particularly in females with ADHD, tionships, suggesting that positive parent–child relationships who may experience more peer-relationship difficulties than might mitigate some of the adverse effects in the indirect males with ADHD (Elkins et al., 2011). However, it seems pathway of ADHD to depressive symptoms via poor quality likely that the indirect effects observed are important to con - friendships. sider for both sexes. The findings on friendship quality align with theories of There was some evidence of moderated mediation sug- social difficulty explaining some of the link between early gesting that indirect effects of top three friendships qual- psychopathology and subsequent depression (Capaldi, ity on the link between ADHD and depressive symptoms 1992). Our findings suggest this is also a pathway that links decreased slightly as self-reports of mother–child relation- ADHD to depression symptoms. Friendship quality as a ship warmth increased and father-child relationship hostility form of perceived social support might reduce depression decreased, though the evidence for father hostility did not risk in those with elevated ADHD symptoms via increasing reach the conventional significance threshold. This draws a sense of connectedness and self-esteem, or by buffering attention to the need to consider the child’s social support against life stresses (Rueger et al., 2016). Those with ADHD across different contexts such as the parent–child relation- are more likely to experience adversity in various areas of ship. Findings were suggestive of mother warmth moderat- life (Harpin, 2005) and good social support can mitigate ing both the ADHD symptoms to friendship quality path against depressive outcomes in those who experience adver- and the friendship quality to depression symptoms path of sity (Collishaw et al., 2016; Lee et al., 2019). Sensitivity the indirect effect. This finding is consistent with that of analyses showed that conflict and security with friends a previous study that found an association between paren- (ability to disclose problems to friend and reconcile after tal behaviour and child peer-relationships in children with disagreement) appeared to be the specific elements of friend- ADHD (Mikami et al., 2010) in addition to studies that have ship quality that were important in the pathway from ADHD found parent–child relationships may be able to compensate to depressive symptoms. Elevated conflict with friends for a lack of friends (Stocker, 1994) and mitigate against has been reported for children with ADHD (Blachman poor mental health in the presence of adversity (Brennan & Hinshaw, 2002) and poor-quality friendships may also be et  al., 2003; Collishaw et  al., 2007, 2016; Lewandowski risk factors for depressive outcomes in school aged children et al., 2014). However, these moderated mediation findings (Goodyer et al., 1989). should be interpreted with caution given that results were Larger indirect effects via friendship quality were not consistent across mother and father or between warmth observed in females than males. This aligns with previous and hostility. Nevertheless, this may be explained in part evidence of females valuing aspects of friendship quality by previous findings that suggest mothers and fathers may such as companionship and intimacy more highly than males have differential effects on the friendships of their adoles- (Hall, 2010). Interpersonal stress may be more prevalent and cent children (Flynn et al., 2018; Updegraff et al., 2001). 1 3 1038 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 For instance, one study found that while mother supportive was minimal (Spratt et al., 2010). Those with ADHD may behaviour and hostile behaviour influenced their children’s under-report their depressive symptoms and over-report their interactional style with peers, for fathers, it was their prob- social competence, while those with depression might under- lem-solving behaviour and hostile behaviour that appeared rate their social ability, which could attenuate associations to be important (Flynn et al., 2018). Moderated mediation (Fraser et al., 2018; Ohan & Johnston, 2011; Whitton et al., results in the current study were also inconsistent across the 2008). In addition, we were unable to investigate whether two friendship quality measures used. While some evidence the friendships reported by the children were reciprocated in of moderation by mother warmth was observed for the indi- the current study, due to reported best friends not necessarily rect effect via friendship quality with the top three friends, attending the same school. However, children were asked strong evidence of moderation by parent–child relationship to report only their top three friendships – an approach that quality was not found for the indirect effect via friendship has been used previously (Fowler et al., 2007; Ng-Knight quality with the best friend. Measurement differences in the et al., 2019). Children were asked to report their best three two friendship variables used in the present study may have friends rather than to report a total number of friends to affected the findings, as a previous study found evidence of capture the children’s close friendships, thereby helping an interaction of friendship stability with a variable meas- to avoid some of the positive illusory bias that may affect uring quality of top three friendships, which was not found reporting on friendships in children with ADHD symptoms. for the quality of the best friendship only (Ng-Knight et al., While measures of how many friends a child has in total may 2019). capture how popular or liked that individual is on a group Limitations include that we relied on teacher reports of level (e.g., in their class or school), measures that capture child ADHD symptoms which provide a reliable measure children’s close friendships may be more predictive of later of ADHD symptoms in school, but most clinical research adjustment and depression (Narr et  al., 2019; Schneider and practitioners rely primarily on parental reports of symp- et al., 1994). In addition, while self-rated measures of friend- toms. In addition, while the SDQ hyperactivity-inattention ship presence and quality were used, peer-rated classroom subscale completed by teachers is a valid and useful screen- data was also used to measure the characteristics of the ing tool for ADHD (Goodman, 1997), it is not a diagnostic friendship group, which also might help to mitigate against measure and we were not able to investigate whether ADHD any effect of illusory biases. ADHD symptoms were associ- inattentive or hyperactive-impulsive subtypes were differen- ated with characteristics of the classroom friendship group tially associated with friendship or depression in this study. (a peer-rated variable), though it did not act as a mediator The inattentive subtype has been found to be the most com- of the association between ADHD and depressive symptoms mon subtype in population samples, particularly during ado- in the current study. lescence (Willcutt, 2012). Associations between ADHD and Strengths include use of a representative school-based emotional and peer problems may vary according to subtype sample during the first year of secondary school with (Graetz et al., 2001), and thus subtype differences are of detailed information from multiple informants on different interest for future research. Despite the longitudinal design features of friendship, in addition to parent–child relation- used, there is a possibility that reverse causation contributed ship quality. School life and transitions are important in ado- to observed associations. However, evidence suggests that lescent mental health and may be particularly challenging ADHD precedes depression in a potentially causal relation- for those with ADHD (Ford, 2020; Richardson et al., 2015). ship (Riglin et al., 2020) and that the prospective relationship Implications of this work include pinpointing quality of exists over long periods of time and when adjusting for prior friendships and parent–child relationships as important to emotional disorder symptomatology (Powell et al., 2020). consider clinically in those with ADHD for reducing depres- Moreover, a sensitivity check found that ADHD symptoms sion risk. Many peer relationship-focused interventions, were associated with subsequent depressive symptoms when which have mainly focused on peer acceptance and social adjusting for baseline self-reported SDQ emotional prob- skills thus far, have shown little success in children with lems (Goodman, 1997; Supplement 9). Although we tested neurodevelopmental disorders (Mikami, 2010). Promising indirect effects, this is not mediation analysis per se, due to directions for the development of enhanced programmes exposure and mediator being measured contemporaneously include those involving a parental component focused on (Selig & Preacher, 2009). This study had missing data, a dyadic friendship building (Gardner et al., 2019). Schools’ common problem in longitudinal data (Spratt et al., 2010). arrangements regarding awareness of friendship groups However, tests were adjusted for confounders that predicted (e.g., keeping together or separating friends) are also impor- missingness, helping to address potential bias arising from tant in ensuring children feel settled at the beginning of sec- missing data (Groenwold et al., 2012). Additionally, we con- ondary school (Keay et al., 2015), and may need additional ducted Multiple Imputation and results remained very simi- consideration in children with ADHD. Practical implications lar in imputed data, suggesting bias caused by missingness for children with ADHD in mitigating later risk of emotional 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1039 American Psychiatric Association. (1994). Diagnostic and Statistical difficulties may also involve focusing on interventions to Manual of Mental Disorders (DSM-IV) (4th ed.). Author. strengthen parent–child relationships (Abikoff et al., 2015; American Psychiatric Association. (2013). Diagnostic and Statistical Meinzer et al., 2020). Interventions aiming to improve par- Manual of Mental Disorders (DSM-V) (5th edn.) (5th ed.). ent–child interactions can have beneficial effects on the American Psychiatric Publishing. Angold, A., Costello, E., Messer, S., Pickles, A., Winder, F., & Silver, mental health of both child and parent (Sonuga-Barke et al., D. (1995). Development of a questionnaire for use in epidemiologi- 2001). cal studies of depression in children and adolescents. International Journal of Methods in Psychiatric Research, 5(4), 237–249. Blachman, D. R., & Hinshaw, S. P. (2002). Patterns of friendship among girls with and without attention-deficit/hyperactivity dis- Conclusion order. Journal of Abnormal Child Psychology, 30(6), 625–640. Brennan, P. A., Le Brocque, R., & Hammen, C. (2003). Maternal depression, parent-child relationships, and resilient outcomes We found in a large school-based sample that ADHD symp- in adolescence. Journal of the American Academy of Child and toms were associated with subsequent depressive symp- Adolescent Psychiatry, 42(12), 1469–1477. toms partly via decreased friendship quality, suggesting that  Bukowski, W. M., Hoza, B., & Boivin, M. (1994). Measuring friend- this aspect of friendship is one potential mechanism by which ship quality during pre- and early adolescence: the development and psychometric properties of the friendship qualities scale. ADHD symptoms increase risk for depressive symptoms. Journal of Social and Personal Relationships, 11(3), 471–484. Positive features of parent–child relationships seemed to Bukowski, W. M., Newcombe, A. F., & Hartup, W. W. (1996). slightly alleviate the indirect effect via friendship quality, Friendship and its significance in childhood and adolescence: highlighting the importance of considering different sources Introduction and comment. In W. M. Bukowski, A. F. Newcombe, & W. W. Hartup (Eds.), The company they keep: of social support in the child’s life. Friendship in childhood and adolescence (pp. 1–19). Cambridge University Press. Supplementary Information The online version contains supplemen- Cabrera, N. J., Volling, B. L., & Barr, R. (2018). Fathers are parents, tary material available at https: //doi.org/10.1007/s10802 -021-00798- w. too! Widening the lens on parenting for children’s development. Child Development Perspectives, 12(3), 152–157. Acknowledgements We thank the participating schools, pupils and par- Cairns, R. B., & Cairns, B. D. (1994). Lifelines and risks: pathways of ents in the School Transition and Adjustment Research Study (www.ucl. youth in our time. Cambridge University Press. ac.uk/stars). The S TARS project was funded by the Nuffield Foundation Cairns, R. B., Cairns, B. D., Neckerman, H. J., Gest, S. D., & Gariépy, (EDU/40065), but the views expressed are those of the authors and not J. L. (1988). Social networks and aggressive behavior: peer necessarily those of the Nuffield Foundation (www.nuffieldf oundat ion. support or peer rejection? Developmental Psychology, 24(6), org). This publication is the work of the authors and Victoria Powell 815–823. will serve as guarantor for the contents of this paper. Cairns, R. B., Leung, M.-C., Buchanan, L., & Cairns, B. D. (1995). Friendships and social networks in childhood and adolescence: fluidity, reliability, and interrelations. Child Development, 66(5), Compliance with Ethical Standards 1330–1345. Capaldi, D. M. (1992). Co-occurrence of conduct problems and depres- Conflict of Interest The authors declare that they have no conflict of sive symptoms in early adolescent boys: II. A 2-year follow-up interest. at Grade 8. Development and Psychopathology, 4(1), 125–144. Chung, H., Elias, M., & Schneider, K. (1998). Patterns of individual Open Access This article is licensed under a Creative Commons Attri- adjustment changes during middle school transition. Journal of bution 4.0 International License, which permits use, sharing, adapta- School Psychology, 36(1), 83–101. tion, distribution and reproduction in any medium or format, as long Coie, J. D., & Dodge, K. A. (1988). Multiple sources of data on social as you give appropriate credit to the original author(s) and the source, behavior and social status in the school: a cross-age comparison. provide a link to the Creative Commons licence, and indicate if changes Child Development, 59(3), 815–829. were made. The images or other third party material in this article are Cole, D. A. (1990). Relation of social and academic competence included in the article’s Creative Commons licence, unless indicated to depressive symptoms in childhood. Journal of Abnormal otherwise in a credit line to the material. If material is not included in Psychology, 99(4), 422–429. the article’s Creative Commons licence and your intended use is not Collishaw, S., Hammerton, G., Mahedy, L., Sellers, R., Owen, M. J., permitted by statutory regulation or exceeds the permitted use, you will Craddock, N., & Thapar, A. (2016). Mental health resilience in need to obtain permission directly from the copyright holder. To view a the adolescent offspring of parents with depression: a prospective copy of this licence, visit http://creativ ecommons .or g/licenses/b y/4.0/. longitudinal study. The Lancet Psychiatry, 3(1), 49–57. Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., & Maughan, B. (2007). Resilience to adult psychopathology follow- ing childhood maltreatment: evidence from a community sample. Child Abuse and Neglect, 31(3), 211–229. References Deković, M., & Meeus, W. (1997). Peer relations in adolescence: effects of parenting and adolescents’ self-concept. Journal of Abikoff, H. B., Thompson, M., Laver-Bradbury, C., Long, N., Adolescence, 20(2), 163–176. Forehand, R. L., Miller Brotman, L., & Sonuga-Barke, E. (2015). Deng, L., & Chan, W. (2017). Testing the difference between reliabil- Parent training for preschool ADHD: a randomized controlled trial ity coefficients alpha and omega. Educational and Psychological of specialized and generic programs. Journal of Child Psychology Measurement, 77(2), 185–203. and Psychiatry, 56(6), 618–631. 1 3 1040 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 Elkins, I. J., Malone, S., Keyes, M., Iacono, W. G., & McGue, M. Keay, A., Lang, J., & Frederickson, N. (2015). Comprehensive sup- (2011). The impact of attention-deficit/hyperactivity disorder port for peer relationships at secondary transition. Educational on preadolescent adjustment may be greater for girls than for Psychology in Practice, 31(3), 279–292. boys. Journal of Clinical Child & Adolescent Psychology, 40(4), Larson, R., & Richards, M. H. (1991). Daily companionship in late 532–545. childhood and early adolescence: changing developmental con- Flynn, H. K., Felmlee, D. H., Shu, X., & Conger, R. D. (2018). Moth- texts. Child Development, 62(2), 284–300. ers and fathers matter: the influence of parental support, hostility, Lee, H. Y., Oh, J., Kawachi, I., Heo, J., Kim, S., Lee, J. K., & Kang, and problem solving on adolescent friendships. Journal of Family D. (2019). Positive and negative social support and depressive Issues, 39(8), 2389–2412. symptoms according to economic status among adults in Korea: Flynn, M., & Rudolph, K. D. (2011). Stress generation and adolescent Cross-sectional results from the Health Examinees-Gem Study. depression: contribution of interpersonal stress responses. Journal British Medical Journal Open, 9(4), 1–10. of Abnormal Child Psychology, 39(8), 1187–1198. Lewandowski, R. E., Verdeli, H., Wickramaratne, P., Warner, V., Mancini, Ford, T. (2020). Transitional care for young adults with ADHD: A., & Weissman, M. (2014). Predictors of positive outcomes in transforming potential upheaval into smooth progression. offspring of depressed parents and non-depressed parents across 20 Epidemiology and Psychiatric Sciences, 29, e87. years. Journal of Child and Family Studies, 23(5), 800–811. Fowler, T., Shelton, K., Lifford, K., Rice, F., McBride, A., Nikolov, Ljung, T., Chen, Q., Lichtenstein, P., & Larsson, H. (2014). Common I., & Van Den Bree, M. B. M. (2007). Genetic and environmen- Etiological Factors of Attention-Deficit/Hyperactivity Disorder tal influences on the relationship between peer alcohol use and and Suicidal Behavior: A Population-Based Study in Sweden. own alcohol use in adolescents. Addiction, 102(6), 894–903. JAMA Psychiatry, 71(8), 958–964. Fraser, A., Cooper, M., Agha, S. S., Collishaw, S., Rice, F., Thapar, Marton, I., Wiener, J., Rogers, M., & Moore, C. (2015). Friendship A., & Eyre, O. (2018). The presentation of depression symp- characteristics of children with ADHD. Journal of Attention toms in attention-deficit/hyperactivity disorder: Comparing Disorders, 19(10), 872–881. child and parent reports. Child and Adolescent Mental Health, Meinzer, M. C., Felton, J. W., Oddo, L. E., Rubin, K. H., & Chronis- 23(3), 243–250. Tuscano, A. (2020). Do ADHD symptoms and relationship quality Gardner, D. M., Gerdes, A. C., & Weinberger, K. (2019). Examination with mothers and best friends across high school predict depres- of a parent-assisted, friendship-building program for adolescents sive symptoms for adolescents? Journal of Attention Disorders. with ADHD. Journal of Attention Disorders, 23(4), 363–373. Meinzer, M. C., Pettit, J. W., & Viswesvaran, C. (2014). The co- Gilman, S. E., Kawachi, I., Fitzmaurice, G. M., & Buka, S. L. (2002). occurrence of attention-deficit/hyperactivity disorder and unipolar Socioeconomic status in childhood and the lifetime risk of depression in children and adolescents: A meta-analytic review. major depression. International Journal of Epidemiology, 31(2), Clinical Psychology Review, 34(8), 595–607. 359–367. Melby, J. N., & Conger, R. D. (2001). The Iowa Family Interaction Goodman, R. (1997). The strengths and difficulties questionnaire: a Rating Scales: Instrument summary. Family Observational research note. Journal of Child Psychology and Psychiatry, 38, Coding Systems: Resources for Systemic Research., pp. 33–58. 581–586. Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers. Goodman, R., Ford, T., Simmons, H., Gatward, R., & Meltzer, H. Mikami, A. Y. (2010). The importance of friendship for youth with (2000). Using the strengths and difficulties questionnaire (SDQ) attention-deficit/hyperactivity disorder. Clinical Child and Family to screen for child psychiatric disorders in a community sample. Psychology Review, 13(2), 181–198. British Journal of Psychiatry, 177(6), 534–539. Mikami, A. Y., Jack, A., Emeh, C. C., & Stephens, H. F. (2010). Goodyer, I. M., Wright, C., & Altham, P. M. E. (1989). Recent friend- Parental influence on children with attention-deficit/hyperactivity ship in anxious and depressed school age children. Psychological disorder: I. Relationships between parent behaviors and child peer Medicine, 19, 165–174. status. Journal of Abnormal Child Psychology, 38(6), 721–736. Graetz, B. W., Sawyer, M. G., Hazell, P. L., Arney, F., & Baghurst, Mrug, S., Molina, B. S. G., Hoza, B., Gerdes, A. C., Hinshaw, S. P., P. (2001). Validity of DSM-IV ADHD subtypes in a nationally Hechtman, L., & Arnold, L. E. (2012). Peer rejection and friend- representative sample of Australian children and adolescents. ships in children with Attention-Deficit/Hyperactivity Disorder: Journal of the American Academy of Child & Adolescent contributions to long-term outcomes. Journal of Abnormal Child Psychiatry, 40(12), 1410–1417. Psychology, 40(6), 1013–1026. Groenwold, R. H. H., Donders, A. R. T., Roes, K. C. B., Harrell, F. E. Narr, R. K., Allen, J. P., Tan, J. S., & Loeb, E. L. (2019). Close friend- J., & Moons, K. G. M. (2012). Dealing with missing outcome data ship strength and broader peer group desirability as differential in randomized trials and observational studies. American Journal predictors of adult mental health. Child Development, 90(1), of Epidemiology, 175(3), 210–217. 298–313. Hall, J. A. (2010). Sex differences in friendship expectations: a meta- Ng-Knight, T., Shelton, K. H., Riglin, L., Frederickson, N., McManus, analysis. Journal of Social and Personal Relationships, 28(6), I. C., & Rice, F. (2019). ‘Best friends forever’? Friendship 723–747. stability across school transition and associations with mental Hamm, J. V., Farmer, T. W., Dadisman, K., Gravelle, M., & Murray, A. health and educational attainment. British Journal of Educational R. (2011). Teachers’ attunement to students’ peer group affiliations Psychology, 89(4), 585–599. as a source of improved student experiences of the school Ng-Knight, T., Shelton, K. H., Riglin, L., McManus, I. C., social–affective context following the middle school transition. Frederickson, N., & Rice, F. (2016). A longitudinal study of self- Journal of Applied Developmental Psychology, 32(5), 267–277. control at the transition to secondary school: Considering the Harpin, V. A. (2005). The effect of ADHD on the life of an individual, role of pubertal status and parenting. Journal of Adolescence, their family, and community from preschool to adult life. Archives 50, 44–55. of Disease in Childhood, 90, i2–i7. Ohan, J. L., & Johnston, C. (2011). Positive illusions of social com- Humphreys, K. L., Katz, S. J., Lee, S. S., Hammen, C., Brennan, P. A., petence in girls with and without ADHD. Journal of Abnormal & Najman, J. M. (2013). The association of ADHD and depres- Child Psychology, 39(4), 527–539.  sion: mediation by peer problems and parent-child difficulties in Parkhurst, J. T., & Asher, S. R. (1992). Peer rejection in middle school: subgroup differences in behavior, loneliness, and interpersonal two complementary samples. Journal of Abnormal Psychology, concerns. Developmental Psychology, 28, 231–241. 122(3), 854–867. 1 3 Research on Child and Adolescent Psychopathology (2021) 49:1031–1041 1041 Patterson, G. R., & Stoolmiller, M. (1991). Replications of a dual fail- Stocker, C. M. (1994). Children’s perceptions of relationships with ure model for boys’ depressed mood. Journal of Consulting and siblings, friends, and mothers: compensatory processes and links Clinical Psychology, 59(4), 491–498. with adjustment. Journal of Child Psychology and Psychiatry, Powell, V., Riglin, L., Hammerton, G., Eyre, O., Martin, J., Anney, 35(8), 1447–1459. R., & Rice, F. (2020). What explains the link between childhood Thapar, A., & McGuffin, P. (1998). Validity of the shortened Mood and ADHD and adolescent depression? Investigating the role of Feelings Questionnaire in a community sample of children and peer relationships and academic attainment. European Child & adolescents: a preliminary research note. Psychiatry Research, Adolescent Psychiatry, 29(11), 1581–1591. 81(2), 259–268. Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing mod- UCLA Statistical Consulting Group. (n.d.). How can I compute indirect erated mediation hypotheses: Theory, methods, and prescriptions. effects with imputed data? (method 1) | STATA FAQ. Retrieved Multivariate Behavioral Research, 42(1), 185–227. January 31, 2020, from https ://stats .idre.ucla.edu/stata /faq/how- Richardson, M., Moore, D. A., Gwernan-Jones, R., Thompson-Coon, can-i-compu te-indir ect-effec ts-with-imput ed-data-metho d-1/ J., Ukoumunne, O., Rogers, M., & Ford, T. J. (2015). Non-phar- Ullebø, A. K., Posserud, M.-B., Heiervang, E., Gillberg, C., & Obel, C. macological interventions for attention-deficit/hyperactivity dis- (2011). Screening for the attention deficit hyperactivity disorder order (ADHD) delivered in school settings: systematic reviews of phenotype using the strength and difficulties questionnaire. quantitative and qualitative research. Health Technology Assess- European Child & Adolescent Psychiatry, 20(9), 451–458. ment, 19(45), 1–470. Updegraff, K. A., McHale, S. M., Crouter, A. C., & Kupanoff, K. Riglin, L., Leppert, B., Dardani, C., Thapar, A. K., Rice, F., (2001). Parents’ involvement in adolescents’ peer relationships: a O’Donovan, M. C., Thapar, A. (2020). ADHD and depression: comparison of mothers’ and fathers’ roles. Journal of Marriage investigating a causal explanation. Psychological Medicine, 1–8. and Family, 63(3), 655–668. Roy, A., Hartman, C. A., Veenstra, R., & Oldehinkel, A. J. (2015). White, I. R., Royston, P., & Wood, A. M. (2011). Multiple imputation Peer dislike and victimisation in pathways from ADHD symptoms using chained equations: issues and guidance for practice. to depression. European Child & Adolescent Psychiatry, 24(8), Statistics in Medicine, 30(4), 377–399. 887–895. Whitton, S. W., Larson, J. J., & Hauser, S. T. (2008). Depressive Rueger, S. Y., Malecki, C. K., Pyun, Y., Aycock, C., & Coyle, S. (2016). symptoms and bias in perceived social competence among young A meta-analytic review of the association between perceived adults. Journal of Clinical Psychology, 64(7), 791–805. social support and depression in childhood and adolescence. Willcutt, E. G. (2012). The prevalence of DSM-IV attention-deficit/ Psychological Bulletin, 142, 1017–1067. hyperactivity disorder: a meta-analytic review. Neurotherapeutics, Schneider, B. H., Wiener, J., & Murphy, K. (1994). Children’s friend- 9(3), 490–499. ships: the giant step beyond peer acceptance. Journal of Social Williams, E. D., Tillin, T., Richards, M., Tuson, C., Chaturvedi, N., and Personal Relationships, 11(3), 323–340. Hughes, A. D., & Stewart, R. (2015). Depressive symptoms are Selig, J. P., & Preacher, K. J. (2009). Mediation Models for doubled in older British South Asian and Black Caribbean people Longitudinal Data in Developmental Research. Research in compared with Europeans: associations with excess co-morbidity Human Development, 6(2–3), 144–164. and socioeconomic disadvantage. Psychological Medicine, 45(9), Shih, J. H., Eberhart, N. K., Hammen, C. L., & Brennan, P. A. (2006). 1861–1871. Differential exposure and reactivity to interpersonal stress predict Zellner, A. (1962). An efficient method of estimating seemingly sex die ff rences in adolescent depression. Journal of Clinical Child unrelated regressions and tests for aggregation bias. Journal of & Adolescent Psychology, 35(1), 103–115. the American Statistical Association, 57(298), 348–368. Sonuga-Barke, E. J., Daley, D., Thompson, M., Laver-Bradbury, C., & Weeks, A. (2001). Parent-based therapies for preschool attention- Publisher’s Note Springer Nature remains neutral with regard to deficit/hyperactivity disorder: a randomized, controlled trial with jurisdictional claims in published maps and institutional affiliations. a community sample. Journal of the American Academy of Child and Adolescent Psychiatry, 40(4), 402–408. Spratt, M., Carpenter, J., Sterne, J. A. C., Carlin, J. B., Heron, J., Henderson, J., & Tilling, K. (2010). Strategies for multiple imputation in longitudinal studies. American Journal of Epidemiology, 172(4), 478–487. 1 3

Journal

Journal of Abnormal Child Psychology – Springer Journals

Published: Mar 2, 2021

Keywords: ADHD; Depression; Friendship; Peer; Parent–child relationship; Transition

There are no references for this article.