Introduction
Depression is very common, chronic, and increases markedly in adolescence (Costello et al. Reference Costello, Mustillo, Erkanli, Keeler and Angold2003; Ford et al. Reference Ford, Goodman and Meltzer2003; Hankin et al. Reference Hankin, Abramson, Moffitt, Silva, McGee and Angell1998). Adolescent depression reliably predicts long-term mental health difficulties (Harrington et al. Reference Harrington, Fudge, Rutter, Pickles and Hill1990; Dunn & Goodyer, Reference Dunn and Goodyer2006; Gregory et al. Reference Gregory, Caspi, Moffitt, Koenen, Eley and Poulton2007) and carries burden of social and educational impairment (Puig-Antich et al. Reference Puig-Antich, Kaufman, Ryan, Williamson, Dahl, Lukens, Todak, Ambrosini, Rabinovich and Nelson1993; Katon et al. Reference Katon, Richardson, Russo, McCarty, Rockhill, McCauley, Richards and Grossman2010; Riglin et al. Reference Riglin, Petrides, Frederickson and Rice2014). Maladaptive cognitions, such as biases in how individuals attend to, interpret and remember emotional information, have been implicated in the development and maintenance of depression (Jacobs et al. Reference Jacobs, Reinecke, Gollan and Kane2008), and are targeted by recommended first-line treatments such as cognitive behavioural therapy (CBT) (AACAP, 2007). Therefore, it is of high importance to understand the aetiology of depression-related cognitions operating across adolescence and emerging adulthood.
Hopelessness theory of depression
Negative attributional style and hopelessness are two maladaptive cognitions associated with adolescent depression. Negative attributional style refers to the attribution of negative events to internal (directed to the self), stable (likely to persist over time) and global (likely to affect many aspects of life) causes, and positive events to external, unstable and specific causes (Abramson et al. Reference Abramson, Seligman and Teasdale1978). Hopelessness is an expectation that the future will be negative and that this outcome cannot be altered (Beck et al. Reference Beck, Weissman, Lester and Trexler1974). A number of theories specified negative attributional style and hopelessness as causal cognitive vulnerabilities in depression, including the hopelessness theory of depression (Abramson et al. Reference Abramson, Metalsky and Alloy1989). Specifically, the theory posits that in the context of negative life events, attributional style contributes to the formation of hopelessness, which in turn contributes to development of depression symptoms. In other words, hopelessness is thought to mediate the prospective relationship between attributional style and depression. Although to date many aspects of the theory have gained empirical support (Metalsky & Joiner, Reference Metalsky and Joiner1992; Joiner, Reference Joiner2001; Hankin, Reference Hankin2008), evidence for the mediation is mixed in both child and adult samples (Alloy & Clements, Reference Alloy and Clements1998; Abela, Reference Abela2001; Hankin et al. Reference Hankin, Abramson and Siler2001).
Cognitive specificity
Given high co-morbidity of depression with anxiety disorders (Angold et al. Reference Angold, Costello and Erkanli1999; Costello et al. Reference Costello, Mustillo, Erkanli, Keeler and Angold2003; Kessler et al. Reference Kessler, Chiu, Demler, Merikangas and Walters2005; Cummings et al. Reference Cummings, Caporino and Kendall2014), it is of theoretical and clinical interest to differentiate them based on maladaptive cognitions. The cognitive content hypothesis posits that although anxious and depressed individuals both have distorted cognitions, the content differs across these disorders (Beck & Perkins, Reference Beck and Perkins2001). Specifically, it is hypothesized that depressed individuals tend to think negatively about the self and focus on experiences of loss whereas anxious individuals focus on perceived threat or danger. In line with this model, cognitive concerns targeted in CBT tend to vary across anxiety and depressive disorders (Brewin, Reference Brewin1996).
The hopelessness theory of depression has been proposed specifically to explain some of the causal factors in depression. In support of this claim, evidence generally suggests that hopelessness is uniquely associated with depression and not anxiety (Alloy & Clements, Reference Alloy and Clements1998; Beck et al. Reference Beck, Perkins, Holder, Robbins, Gray and Allison2001, Reference Beck, Wenzel, Riskind, Brown and Steer2006; Miranda & Mennin, Reference Miranda and Mennin2007; Alloy et al. Reference Alloy, Black, Young, Goldstein, Shapero, Stange, Boccia, Matt, Boland and Moore2012; Hendriks et al. Reference Hendriks, Licht, Spijker, Beekman, Hardeveld, de Graaf and Penninx2014). This indicates that negative inferences about the future might constitute unique content in depression that differentiates it from anxiety disorders. However, there is mixed evidence about the specificity of negative attributional style to depression (Ahrens & Haaga, Reference Ahrens and Haaga1993; Luten et al. Reference Luten, Ralph and Mineka1997; Waschbusch et al. Reference Waschbusch, Sellers, LeBlanc and Kelley2003; Hankin et al. Reference Hankin, Abramson, Miller and Haeffel2004; Brozina & Abela, Reference Brozina and Abela2006; Reardon & Williams, Reference Reardon and Williams2007), suggesting that maladaptive interpretations of events might instead be a transdiagnostic cognitive risk factor for internalizing problems. Notably, the specificity of attributional style and hopelessness has largely been studied in adults. As maladaptive cognitions are thought to emerge in development (Cole et al. Reference Cole, Ciesla, Dallaire, Jacquez, Pineda, LaGrange, Truss, Folmer, Tilghman-Osborne and Felton2008; Field & Lester, Reference Field and Lester2010), it is important to gain a better understanding of disorder-specific and transdiagnostic cognitions in depression and anxiety in young people.
Aetiology
There is growing evidence that individual differences in maladaptive cognitions stem from both genetic and environmental influences. Our team has previously shown that attributional style is moderately heritable in adolescence (Lau et al. Reference Lau, Rijsdijk and Eley2006; Lau & Eley, Reference Lau and Eley2008; Zavos et al. Reference Zavos, Rijsdijk, Gregory and Eley2010). These studies have also found genetic and environmental overlap between attributional style, depression and anxiety symptoms, suggesting that maladaptive cognitive processes in part represent a genetic vulnerability to internalizing problems, in addition to being reflections of the individual's environment (Beck, Reference Beck2008). This also indicates that both inherited predispositions and environmental circumstances might play a role in the hopelessness theory of depression, an idea originally hypothesized by Abramson et al. (Reference Abramson, Metalsky and Alloy1989). To date, a composite measure of hopelessness and guilt was found to be moderately heritable in adults, with substantial influence of the individual-specific environment (Jang et al. Reference Jang, Livesley, Taylor, Stein and Moon2004), however no study has yet investigated the aetiology of hopelessness in young people, or its genetic and environmental associations with depression or attributional style. Examining aetiological influences on the joint associations between attributional style, hopelessness and depression could help to clarify some of the mechanisms that underpin these relationships in the hopelessness theory of depression. It may also help to disentangle common and specific influences on these traits.
Aims
The current study investigated the hopelessness theory of depression from an aetiological perspective. Three waves of data from a large epidemiological sample of adolescent twins and siblings were employed. First, the potential cognitive specificity of negative attributional style and hopelessness to depression v. anxiety symptoms was investigated, both concurrently at mean age 17 years, and prospectively across adolescence and young adulthood. Based on the existing, largely adult literature we hypothesized that hopelessness would be uniquely associated with depression while attributional style would be associated with both depression and anxiety. Second, we aimed to determine whether hopelessness mediates the relationship between attributional style and depression, both concurrently and across time. In line with the hopelessness theory of depression, we expected to observe at least partial mediation. Third, we explored what proportion of variance in hopelessness was accounted for by genetic and environmental influences. We hypothesized that similarly to attributional style, hopelessness would be moderately heritable, in line with adult estimates of hopelessness and guilt. Fourth, we examined the shared aetiology between attributional style, hopelessness and depression. We expected that common genetic influences would explain most of the shared variance between these three traits, in line with the generalist genes hypothesis (Eley, Reference Eley1997), which proposes that traits co-vary due to shared genetic influences, while non-shared environmental influences are generally symptom-specific and contribute to the differentiation between the traits. Based on the generalist genes hypothesis, we hypothesized that there would be no unique genetic association between hopelessness and depression independent of attributional style.
Method
Sample
We used data from waves 2–4 (hereon referred to as times 1–3, respectively) of a longitudinal twin and sibling study, the Genesis 1219 (G1219; McAdams et al. Reference McAdams, Gregory, Rowe, Zavos, Barclay, Lau, Maughan and Eley2013). The study was given ethical approval by the Research Ethics Committee of the Institute of Psychiatry, King's College, London, South London and Maudsley NHS Trust and Goldsmiths, University of London. Informed consent was obtained from parents of adolescents aged <16 years and from participants aged ⩾16 years. The sample characteristics are presented in Table 1. Concurrent analyses were conducted at time 2, the only time at which hopelessness was measured. Longitudinal analyses spanned times 1–3, using attributional style measure at time 1, hopelessness at time 2 and depression and anxiety symptoms at time 3.
Table 1. Sample characteristics
MZ, Monozygotic; DZS, dizygotic same sex; DZO, dizygotic opposite sex.
The inclusion of siblings inevitably resulted in large age ranges; however 72% of the participants were twins with a tighter age range (e.g. at time 2, age s.d. = 1.11, range = 15–19 for twins, age s.d. = 1.97, range = 15–23 for siblings).
Attrition was predicted by socio-economic status (responses were more likely from individuals with parents reporting higher qualifications and home ownership), delinquency (individuals reporting lower levels of delinquent behaviour were more likely to stay in the study) and sex (females were more likely than males to remain in the study), but not by zygosity and internalizing symptoms.
Measures
Attributional style (times 1 and 2)
Participants completed the revised Children's Attributional Style Questionnaire (Thompson et al. Reference Thompson, Kaslow, Weiss and Nolen-Hoeksema1998); a 24-item forced-choice questionnaire that described a positive or negative event (e.g. ‘You get an A on a test’) and asked about its possible cause (e.g. ‘I am clever’). The measure assesses three dimensions of attributional style (internal-external, global-specific, stable-unstable), with an overall lower composite score indicating more negative attributional style. The measure demonstrates moderate internal consistency reliabilities ranging α = 0.40–0.60, moderate 6-month test–retest reliability of 0.53 and small to moderate criterion-related validity assessed through association with depression symptoms (r = −0.40) (Thompson et al. Reference Thompson, Kaslow, Weiss and Nolen-Hoeksema1998).
Hopelessness (time 2)
Participants completed the Hopelessness Scale (Beck et al. Reference Beck, Weissman, Lester and Trexler1974), consisting of 20 true-false items assessing feelings of hopelessness (e.g. ‘My future seems dark to me’). Theories of hopelessness suggest that it may be underpinned by three subfactors reflecting feelings about the future, loss of motivation and future expectations (Beck et al. Reference Beck, Weissman, Lester and Trexler1974), although unidimensional solutions have also been reported (Dozois & Covin, Reference Dozois, Covin and Hersen2004). In the current analyses items were summed to create a total score. The measure has sound psychometric properties in both clinical and healthy adults and adolescent samples, with high internal consistencies up to α = 0.90 and high 3-week test–retest reliability of 0.85, demonstrating robust validity with related constructs, such as depression symptoms, and suicidal ideation and intent independently of depression (Beck et al. Reference Beck, Weissman, Lester and Trexler1974; Young et al. Reference Young, Halper, Clark, Scheftner and Fawcett1992; Dozois & Covin, Reference Dozois, Covin and Hersen2004).
Depression symptoms (times 2 and 3)
Participants completed the Short Mood and Feelings Questionnaire (Angold et al. Reference Angold, Costello, Messer, Pickles, Winder and Silver1995), a 13-item unidimensional self-report measure assessing how often depression symptoms occurred in the past 2 weeks. Responses were summed to give a total depression score. The measure demonstrates good reliability and validity (Angold et al. Reference Angold, Costello, Messer, Pickles, Winder and Silver1995).
Anxiety symptoms (times 2 and 3)
At time 2 adolescents completed the Spence Children′s Anxiety Scale (Spence, Reference Spence1998); a 38-item self-report questionnaire tapping anxiety disorder related symptoms, such as generalized anxiety, panic, separation anxiety and social anxiety symptoms. At time 3 participants completed the Revised Symptoms of Anxiety Scale (Gregory et al. Reference Gregory, Buysse, Willis, Rijsdijk, Maughan, Rowe, Cartwright, Barclay and Eley2011), an age-appropriate version of the Revised Child Anxiety and Depression Scale (Chorpita et al. Reference Chorpita, Yim, Moffitt, Umemoto and Francis2000), consisting of 36 self-report items designed to assess DSM-IV anxiety disorder symptoms. Responses were summed to create total scores. The measures have good reliability and validity (Spence, Reference Spence1998; Birmaher et al. Reference Birmaher, Brent, Chiappetta, Bridge, Monga and Baugher1999; Chorpita et al. Reference Chorpita, Yim, Moffitt, Umemoto and Francis2000; Gregory et al. Reference Gregory, Buysse, Willis, Rijsdijk, Maughan, Rowe, Cartwright, Barclay and Eley2011). Internal consistencies for all measures in the current study and descriptive statistics are presented in Table 2.
Table 2. Descriptive statistics and phenotypic correlations: full correlations below diagonal and partial correlations above diagonal
s.d., Standard deviation.
95% Confidence intervals (CIs) are presented in parentheses. CIs not inclusive of zeros indicate significant correlations. Non-overlapping CIs mean significant difference between the values.
Partial correlations between two variables (e.g. attributional style and hopelessness) controlled for the associations with the other two variables (e.g. depression and anxiety symptoms).
Higher scores indicate more positive attributional style, greater helplessness and greater depression and anxiety symptoms.
Phenotypic analyses
Phenotypic analyses were conducted using Stata (StataCorp, 2007). Concurrent (time 2) and longitudinal (across times 1–3) associations between variables were first explored using full and partial correlations. Partial correlations allowed investigating unique associations between two variables over and above associations accounted for by other variables (e.g. unique association between attributional style and depression symptoms controlling for hopelessness and anxiety symptoms). Next, the Sobel-Goodman test was used to test whether hopelessness mediates the relationship between attributional style and depression, concurrently and across time (Preacher & Hayes, Reference Preacher and Hayes2004). Bootstrapping was used to obtain confidence intervals. All phenotypic analyses were conducted on untransformed and unregressed variables for comparison with other published samples. Analyses were conducted on a random selection of one twin from each twin pair to ensure the independence of observations.
Twin analyses
The twin design compares the similarity between monozygotic (sharing 100% of their genes) and dizygotic (sharing on average 50% of their segregating genes) twin pairs. Differences in within-pair correlations allows estimations of the influences of additive genetics (A), shared environment (C, factors that contribute to phenotypic similarity between siblings) and non-shared environment (E, factors that contribute to phenotypic differences between siblings). Quantitative genetic methods are described comprehensively elsewhere (Rijsdijk & Sham, Reference Rijsdijk and Sham2002; Plomin et al. Reference Plomin, DeFries, Knopik and Neiderhiser2013).
Twin models were fitted using OpenMx (Boker et al. Reference Boker, Neale, Maes, Wilde, Spiegel, Brick, Spies, Estabrook, Kenny and Bates2011) within R (http://www.R-project.org) (TeamRDC, 2010), a structural equation modelling package for genetically informative data. As is standard in model fitting analysis, all variables were regressed for age and sex (McGue & Bouchard, Reference McGue and Bouchard1984), and depression at both waves and hopelessness were log transformed. Transformation did not have an impact on the relationship between the variables in a way that would alter interpretation.
Models were fitted using raw data maximum likelihood. The core fit statistic was minus twice the log likelihood (-2LL) of the observations. This is not an overall measure of fit, but provides a relative measure of fit, since differences in -2LL between models are distributed as χ2. To examine the overall fit of the genetic model we compared the -2LL to that of a saturated model (which fully describes data using the maximum number of free parameters, estimating variances, covariances and means for the raw data to get a baseline index of fit). The fit of sub-models was assessed by χ2 difference tests, Akaike's Information Criterion (AIC) and Bayesian Information Criterion, with lower values suggesting a better fit. If the difference between the AIC of two models was <10, the more parsimonious model was selected (Wagenmakers & Farrell, Reference Wagenmakers and Farrell2004). Information about the precision of parameter estimates was obtained by likelihood-based confidence intervals.
Univariate twin analyses
Univariate genetic analyses were conducted for all variables. Males and females showed differences in variance on all variables and a scalar was fitted to account for this difference. Qualitative and quantitative sex differences were tested to see whether males and females differ in type and magnitude of genetic and environmental influences, but such differences were not found. The exception was attributional style at time 1, which showed quantitative sex differences, however differences in magnitude were small with overlapping 95% confidence intervals [AM = 0.38 (95% CI 0.25–0.49) v. AF = 0.49 (95% CI 0.39–0.58) and EM = 0.49 (95% CI 0.39–0.58) v. EF = 0.51 (95% CI 0.42–0.61)] and for simplicity homotypic models are presented. Finally, comparisons indicated that covariances, means and variances could be equated across dizygotic twins and singleton siblings for all variables, thus siblings were modelled alongside twins in the analyses.
Multivariate twin analyses
First, a one-factor independent pathway model was fitted to examine the genetic and environmental overlap on the three variables (Fig. 1a ). The model allows one set of common (AC, and EC) and variable-specific (AS, and ES) genetic and environmental influences on each variable. The model tests whether there is a single set of common aetiological factors that influence attributional style, hopelessness and depression symptoms, accounting for their associations, in addition to variable-specific factors.
Next, the Cholesky decomposition (Fig. 1b ) was used to examine whether there are any genetic and environmental influences shared between hopelessness and depression when accounting for the genetic overlap with attributional style. The Cholesky decomposition assumes three distinct sets of genetic and environmental influences on a variable at each time point. A1 and E1 are influences on the first variable (paths a11 and e11) that can also influence the remaining two variables (paths a12−3 and e12−3). A2 and E2 influence the second variable (paths a22 and e22) and can also influence the third variable over and above the influences accounted for by A1 and E1 (paths a23 and e23). A3 and E3 are unique, residual influences specific to the third variable only (paths a33 and e33).
Both multivariate models were fitted concurrently at time 2, and longitudinally at times 1–3. Although any ordering of variables explained the variance-covariance matrix between the variables equally well, the order of variables was based on the hopelessness theory of depression. Longitudinal models allowed investigating how aetiological influences operated across development, reflecting the developmental predictions of the theory. However, in addition to variable-specific influences, influences at later time points can also reflect age-specific genetic and environmental innovation that characterizes adolescence (Hannigan et al. Reference Hannigan, Walaker, Waszczuk, McAdams and Eleyin press; Kendler et al. Reference Kendler, Gardner and Lichtenstein2008; Waszczuk et al. Reference Waszczuk, Zavos, Gregory and Eley2016). Concurrent models allowed investigating variable-specific influences without the confounding effect of genetic and environmental innovation across time.
Results
Phenotypic results
Attributional style and hopelessness were moderately associated with internalizing symptoms (Table 2), both concurrently (r = −0.47 and 0.50 with depression, r = −0.32 and 0.36 with anxiety, respectively) and longitudinally (r = 0.−0.23 and 0.33 with depression, r = −0.20 and 0.27 with anxiety, respectively). However, partial correlations indicated that when controlling for depression symptoms, neither attributional style nor hopelessness remained significantly associated with anxiety symptoms. Conversely, both maladaptive cognitive styles were uniquely associated with depression symptoms after controlling for anxiety, both concurrently and across time (r p = −23 and −0.09 for attributional style and r p = .27 and 0.17 for hopelessness, respectively). Given the lack of unique association between both maladaptive styles and anxiety symptoms, only depression was taken forward to further analyses.
Mediation analyses revealed that hopelessness partially mediated the relationship between attributional style and depression symptoms (Fig 2). Specifically, hopelessness mediated about 37% of this relationship concurrently (total effect = −0.71, indirect effect via hopelessness = −0.26) (Fig. 2a ), and about 38% longitudinally (total effect = −0.37, indirect effect via hopelessness = −0.14) (Fig. 2b ).
Fig. 1. (a) Independent pathways model, (b) Cholesky decomposition. AE models are presented for clarity. Model fit statistics corroborated AE models and in the full models C estimates were very small.
Twin modelling results
Model fit statistics for comparisons to saturated models, and testing whether parameters can be dropped, are presented in Supplementary Table S1. Model fit statistics corroborate AE models and in the full models C estimates are very small. However, for completeness full ACE models are presented in the Supplementary material (Tables S2, S3). Dropping C from the models did not have impact on the interpretation of the results. The associations between attributional style and depression symptoms, including univariate ACE results for these variables, have been reported before (Lau et al. Reference Lau, Rijsdijk and Eley2006; Lau & Eley, Reference Lau and Eley2008; Zavos et al. Reference Zavos, Rijsdijk, Gregory and Eley2010).
Univariate results revealed that hopelessness was moderately heritable (A = 0.37, 95% CI 0.28–0.47), with remaining variance accounted for by non-shared environmental influences (E = 0.63, 95% CI 0.53–0.73). Attributional style and depression symptoms were also moderately heritable and univariate results for these variables have been presented before (Lau et al. Reference Lau, Rijsdijk and Eley2006; Lau & Eley, Reference Lau and Eley2008; Zavos et al. Reference Zavos, Rijsdijk, Gregory and Eley2010).
Independent pathway models examined common (i.e. influencing all three variables) and variable-specific genetic and shared environmental influences on attributional style, hopelessness and depression symptoms. Concurrently, common genetic influences accounted for majority of genetic influences on attributional style and depression (Ac = 0.31 and 0.30, accounting for 70% (95% CI 49–96) and 64% (95% CI 44–90) of total genetic variance in each variable, respectively), and about half genetic influences on hopelessness (Ac = 0.19, accounting for 51% (95% CI 34–71) of total genetic variance in hopelessness) (Fig. 3a ). Common non-shared environmental influences were significant and accounted for about a third of all non-shared environmental influences on variables [Ec = 0.21, 0.26 and 0.11 on attributional style, hopelessness and depression, respectively, accounting for 37% (95% CI 21–57), 42% (95% CI 25–59) and 20% (95% CI 10–33) of total non-shared environmental variance]. All variable-specific influences were significant (As = 0.13–0.18, Es = 0.35–0.42).
Fig. 2. Mediation analyses results: (a) concurrently at time 2, (b) longitudinally across times 1–3. 95% Confidence Intervals (CIs) are presented in brackets. CIs not inclusive of zeros indicate significant coefficients. Non-overlapping CIs mean significant difference between the values.
Longitudinally, common genetic influences were the only source of common variance, and accounted for about half of genetic influences in attributional style time 1 [Ac = 0.22, accounting for 51% (95% CI 37–67) of total genetic variance in attributional style time 1] and the majority of genetic influences on hopelessness time 2 and depression symptoms time 3 (Ac = 0.36 and 0.30, accounting for 83% (95% CI 64–100) and 77% (95% CI 56–100) of total genetic variance in each variable respectively). Furthermore, there were only significant variable-specific genetic influences on attributional style (Fig. 4a ). Thus all genetic influence on both hopelessness and depression was common to all three variables. Common non-shared environmental influences did not emerge (for longitudinal independent pathway model with Ec see Supplementary Table S2, and for model fit comparisons see Supplementary Table S1 note), instead, all non-shared environmental influences were large and variable-specific (E s = 0.57–0.61).
Fig. 3. Concurrent multivariate results: (a) Independent pathways model, (b) Cholesky decomposition. All paths are squared. Square root of these values should be taken to obtain variance path.
Fig. 4. Longitudinal multivariate results: (a) Independent pathways model, (b) Cholesky decomposition. All paths are squared. Square root of these values should be taken to obtain variance path.
Cholesky decompositions examined whether there are any genetic and environmental influences shared between hopelessness and depression when accounting for the genetic overlap with attributional style. Both concurrent and longitudinal Cholesky decompositions indicate that after accounting for genetic influences shared with attributional style, there were no significant additional genetic associations between hopelessness and depression (standardized path a23 was non-significant in Figs 3b and 4b ). Although non-shared environmental influences were largely variable-specific, when accounting for non-shared environmental influences shared with attributional style, there was a small but significant non-shared environmental association between hopelessness and depression in concurrent analyses (standardized path e23 = 0.02).
Discussion
This study is the first to investigate the hopelessness theory of depression from an aetiological perspective, both within and across time-points during adolescence and into young adulthood. Phenotypic results supported the theory, as hopelessness partially mediated the relationship between attributional style and depression symptoms, and both maladaptive cognitive styles were unique to depression and not to anxiety (the latter phenotype was therefore dropped from twin analyses). Twin modelling results revealed that a set of common genetic influences largely accounted for the association between attributional style, hopelessness and depression symptoms, indicating shared genetic liability to maladaptive cognitions and depression underpinning the hopelessness theory of depression.
Aetiology of hopelessness theory of depression
Current phenotypic results indicate that hopelessness partially mediates the association between attributional style and depression symptoms, both concurrently and across development, as predicted by the hopelessness theory of depression. This is in line with some (Alloy & Clements, Reference Alloy and Clements1998), but not all (Abela, Reference Abela2001; Hankin et al. Reference Hankin, Abramson and Siler2001) previous studies. Furthermore, although prospective longitudinal results indicated that about a third of the association between early attributional style and later depression was mediated via hopelessness, both attributional style and hopelessness were also uniquely associated with depression symptoms, indicating that they may be independent cognitive risk factors for developing depression in adolescence.
Univariate twin modelling results highlighted the role of both genetic and environmental influences in the aetiology of adolescent hopelessness. The current heritability estimate was moderate, in line with the previous estimate of heritability of a composite hopelessness and guilt trait in adults (Jang et al. Reference Jang, Livesley, Taylor, Stein and Moon2004), but extends previous research by using a more comprehensive, validated measure of hopelessness and expands understanding of its aetiology to a novel developmental period.
Most of the genetic influences on each of the two maladaptive cognitions and depression symptoms were shared concurrently and across development. Furthermore, there were no shared genetic influences between hopelessness and depression over and above genetic influences shared with attributional style. Thus, as expected, the association between maladaptive cognitions and depression was explained largely by underlying genetic liability, in line with the generalist genes hypothesis (Eley, Reference Eley1997). The results are also in agreement with previous findings that associations between many different cognitive biases and internalizing problems are largely due to shared genetic influences (Lau & Eley, Reference Lau and Eley2008; Zavos et al. Reference Zavos, Rijsdijk, Gregory and Eley2010; Chen & Li, Reference Chen and Li2013; Moore et al. Reference Moore, Salk, Van Hulle, Abramson, Hyde, Lemery-Chalfant and Goldsmith2013; Waszczuk et al. Reference Waszczuk, Zavos and Eley2013; Brown et al. Reference Brown, Waszczuk, Zavos, Trzaskowski, Gregory and Eley2014; Lau et al. Reference Lau, Belli, Gregory and Eley2014). The results are suggestive of developmentally stable biological pathways underpinning the hopelessness theory of depression. Evidence for shared genetic effects has implications for molecular genetic studies, supporting the argument that including cases with a range of depression-related phenotypes would lead to increasing power to detect shared susceptibility loci (Hettema et al. Reference Hettema, Chen, Sun and Brown2015). It also suggests that the role of specific genes in the aetiology of internalizing symptoms and maladaptive cognition should be investigated as it might be possible to combine the genetic markers to create polygenic risk scores to predict an individual's vulnerability to depression (Demirkan et al. Reference Demirkan, Penninx, Hek, Wray, Amin, Aulchenko, Van Dyck, de Geus, Hofman and Uitterlinden2011).
Conversely, non-shared environmental influences were largely trait-specific, especially in longitudinal analyses. This indicates that environmental influences contribute to differences between cognitive vulnerabilities and depression, possibly explaining some of the phenotypic specificity observed. However, a significant common non-shared environmental factor suggests that it might be possible to identify environmental influences that contribute to the hopelessness theory of depression. In line with the theory, these could be negative life events that interact with attributional style in a diathesis-stress manner. Future studies should aim to identify these specific environmental influences to inform targeted clinical and resilience interventions in adolescence. Nonetheless, this common non-shared environmental factor might also to some extent reflect time-specific measurement error, as it does not replicate in the longitudinal analyses.
Cognitive specificity
Evidence for phenotypic specificity in associations between attributional style, hopelessness and depression, and unique environmental influences acting on these symptoms, has implications for therapeutic interventions. Furthermore, identifying disorder-specific maladaptive cognitions, and also those shared between co-morbid disorders such as depression and anxiety, could continue to inform the tailoring of CBT programmes to a given diagnosis. Current findings indicate that both attributional style and hopelessness are independently and uniquely associated with depression symptoms, but not with anxiety symptoms in adolescence. This is in line with previous research finding that hopelessness is a maladaptive cognition specific to depression (Beck et al. Reference Beck, Perkins, Holder, Robbins, Gray and Allison2001, Reference Beck, Wenzel, Riskind, Brown and Steer2006; Alloy et al. Reference Alloy, Black, Young, Goldstein, Shapero, Stange, Boccia, Matt, Boland and Moore2012; Hendriks et al. Reference Hendriks, Licht, Spijker, Beekman, Hardeveld, de Graaf and Penninx2014). Conversely, the finding that attributional style is unique to depression supports some previous studies in adults and children (Rodriguez & Pehi, Reference Rodriguez and Pehi1998; Hankin et al. Reference Hankin, Abramson, Miller and Haeffel2004; Brozina & Abela, Reference Brozina and Abela2006), but does not support the view, based largely on adult literature, that attributional style is a transdiagnostic cognitive risk factor for both depression and anxiety (Luten et al. Reference Luten, Ralph and Mineka1997; Alloy et al. Reference Alloy, Black, Young, Goldstein, Shapero, Stange, Boccia, Matt, Boland and Moore2012). Taken together, these results indicate that negative thoughts and interpretations about the present and future events constitute central and unique cognitive content in adolescent depression, in line with cognitive specificity hypothesis (Beck & Perkins, Reference Beck and Perkins2001). This supports the clinical evidence that modifying attributional style and hopelessness in CBT prevents and reduces adolescent depression, as well as other important depression-related symptoms such as suicidality (Brent et al. Reference Brent, Kolko, Birmaher, Baugher, Bridge, Roth and Holder1998; Voelz et al. Reference Voelz, Haeffel, Joiner and Wagner2003; Stanley et al. Reference Stanley, Brown, Brent, Wells, Poling, Curry, Kennard, Wagner, Cwik and Klomek2009). However, our findings also imply that targeting attributional style and hopelessness may not be as effective in reducing adolescent anxiety symptoms. Future research should aim to explore the unique and transdiagnostic content of depressive cognitions in more detail by combining multiple cognitive distortions within a single study.
Limitations
The genetically-informative, representative sample and multiple time points are strengths of the study. However, a number of limitations are noteworthy. First, it was beyond the scope of the current study to investigate all aspects of the hopelessness theory of depression, such as the generally supported diathesis-stress interaction between negative life events and attributional style (Abela, Reference Abela2001; Hankin et al. Reference Hankin, Abramson and Siler2001; Abela & Sarin, Reference Abela and Sarin2002). Furthermore, Abramson et al. (Reference Abramson, Metalsky and Alloy1989) posited that the theory is specific to ‘hopelessness depression’, however the distinction of this depression subtype from major depression is debated (Alloy & Clements, Reference Alloy and Clements1998), and the theory has generally been studied with broad measures of depression. Future research should explore phenotypic and aetiological associations between different dimensions of depression, attributional style and hopelessness within the context of the theory, which was beyond the scope of the current study. It should also investigate whether there are bidirectional associations between attributional style, hopelessness and depression (Zavos et al. Reference Zavos, Rijsdijk, Gregory and Eley2010). Second, our analyses used self-report internalizing symptoms. Results should be replicated in clinical samples and using lifetime diagnostic interviews. However, symptoms of internalizing disorders are important markers of psychopathology (Pickles et al. Reference Pickles, Rowe, Simonoff, Foley, Rutter and Silberg2001; Fergusson et al. Reference Fergusson, Horwood, Ridder and Beautrais2005; Balázs et al. Reference Balázs, Miklósi, Keresztény, Hoven, Carli, Wasserman, Apter, Bobes, Brunner and Cosman2013). Common psychiatric disorders are now considered to be the extremes of quantitative traits (Plomin et al. Reference Plomin, Haworth and Davis2009; Insel et al. Reference Insel, Cuthbert, Garvey, Heinssen, Pine, Quinn, Sanislow and Wang2010) and there is evidence that differently defined internalizing problems have the same aetiology (Kendler et al. Reference Kendler, Heath, Martin and Eaves1987, Reference Kendler, Neale, Kessler, Heath and Eaves1992a , Reference Kendler, Neale, Kessler, Heath and Eaves b ). Nonetheless, reliance on self-report data may be associated with shared method variance that could inflate the correlations. Third, there was attrition in the sample – although not for internalizing symptoms. Attrition bias might complicate estimation of trait prevalence; however it is unlikely to affect the estimation of between trait associations (Wolke et al. Reference Wolke, Waylen, Samara, Steer, Goodman, Ford and Lamberts2009). Last, there are limitations inherent to the twin design, discussed comprehensively elsewhere (Plomin et al. Reference Plomin, DeFries, Knopik and Neiderhiser2013). These have minimal and contrasting effects on parameter estimates which should be taken as indicative rather than absolute.
Conclusions
The current study is the first to study the aetiological underpinnings of the hopelessness theory of depression, demonstrating that associations between attributional style, hopelessness and depression symptoms are largely due to shared genetic liability, suggesting developmentally stable biological pathways underpinning this influential theory. Furthermore, both attributional style and hopelessness were not related to anxiety, thus constituted unique cognitive content in depression. The results inform molecular genetics research and treatment approaches, as identifying specific cognitions in depression can inform the design of more precise clinical interventions for this disorder across development.
Supplementary material
For supplementary material accompanying this paper visit http://dx.doi.org/10.1017/S0033291716000489.
Acknowledgements
G1219 study was supported by a Medical Research Council Training Fellowship (G81/343) and a Career Development Award (G120/635) given to Thalia C. Eley. Waves 1–3 of G1219 were also funded by the W T Grant Foundation, the University of London Central Research fund and wave 4 was supported by the Economic and Social Research Council (RES-000-22-2206), the Institute of Social Psychiatry (06/07–11) to Alice M. Gregory who was supported by a Leverhulme Research Fellowship (RF/2/RFG/2008/0145). Monika A. Waszczuk was supported by a Ph.D. studentship funded by the Alexander von Humboldt Foundation. This study presents independent research partly funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. The views expressed in this article are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health. Monika A. Waszczuk had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. We thank the families for their participation as well as numerous staff and students from the Social Genetic Developmental Psychiatry Centre, Institute of Psychiatry, London and Goldsmiths, University of London.
Declaration of Interest
None.
Open access
