Familial history of depression increases risk for offspring psychiatric disorder. Therefore, we investigated whether polygenic risks for psychiatric disorders contribute to intergenerational transmission of depression. We found that multigenerational depression history was positively correlated with higher genome-wide PGSs for depression and bipolar disorder. Moreover, a greater PGS for depression significantly mediated the effects of familial history of depression on offspring’s mental health indexed by KSADS diagnoses. To our best knowledge, this is the first report showing polygenic risk for depression as a potential mechanism for the transgenerational transmission of depression.
Addressing conflicting results from previous research on the association between first-degree family history and depression PGS [21–23], the present study utilized more recent and larger GWAS results for constructing the PGS of depression, which could partially account for the discrepancies between the findings as well as testing findings in a larger cohort. For the first time, our results reveal that higher depression PGS is significantly associated with a family history of depression spanning multiple generations, encompassing risks from both first- and second-degree family members. The children with two previous generations affected showed the highest PGS (estimate, 0.129 [95% CI, 0.070–0.187]) in line with the increased risk for psychopathology [6]. These offspring with two previous generations affected might form a subgroup of individuals with an exceptionally high genetic burden.
We found a similar association for bipolar PGS (estimate, 0.109 [95% CI, 0.051–0.168]). But when we limited the sample to those with a family history of depression without mania, bipolar PGS lost its significance in the association with family history. This shows that the significant associations with bipolar PGS might be due to including children with a family history of bipolar disorder and that genetic risk for and familial aggregation of depression and bipolar disorder are separate. This is in line with recent literature showing an association between family history of bipolar disorder and higher bipolar PGS in offspring [54–56], reports of familial transmission being specific to type of mood disorder [57] and genetic differences between depression and bipolar disorder [58].
Our results also showed that depression PGS and family history of depression together accounted for significant variances of offspring’s psychopathology. These results extend previous findings from recent studies that showed family history and PGS are predictive of future cases of depression and including both factors in a model modestly increases the prognostic power in predicting depression [59–61].
The depression PGS was positively correlated with a wide range of KSADS diagnoses, including any psychiatric disorder, any anxiety disorder, suicidal ideation, ADHD, and conduct/oppositional defiant disorder (C/ODD). On the other hand, bipolar PGS showed no significant correlations with KSADS diagnoses. Despite not surviving correction for multiple comparisons, depression PGS was associated with any depressive disorder even at this young age of 9 to 10 years old (unadjusted P = .019). Depression often starts later in adolescence and early adulthood, but higher family and polygenic risk of depression have been shown to be associated with earlier onset [3, 62]. Suicidal ideation and self-harm at this age, also associated with depression PGS, are becoming more common in recent years [63] and are predictive of later suicide attempts [64]. However, identifying those at risk for suicide is challenging. Our results indicate that the combination of family assessments and PGS for depression might increase predictive power in evaluating risk for suicidal thoughts and behaviors in children, especially as PGS are thought to be able to explain more variance in the future as GWASs include ever larger samples.
PGS for depression significantly mediates the effects of family history on offspring’s depression. Moreover, the mediation effect was significant for a broad range of psychiatric disorders, including any psychiatric disorder, any anxiety disorder, suicidal thoughts and behaviors, C/ODD, and ADHD. This highlights a role for genetic liability in intergenerational transmission of childhood psychopathology in general and depression specifically in periadolescent children (9–10 years old). Future research may examine the brain correlates of family history of depression and depression PGS.
Family history of depression may affect the offspring’s risk for psychiatric disorders through genetic or environmental pathways. Here we show that the genetic component is significant and scales with the number of generations affected. On the other hand, children of parents with depression may experience and even generate more stressful life events [65, 66]. These events may exacerbate risks of psychiatric disorders. Further research could identify protective or vulnerability factors that may moderate the impact of family history of depression and depression PGS through gene-environment interactions. Furthermore, because we did not have access to parent and grandparent genotypes, we are unable to distinguish between direct genetic effects of the child’s genotype on their phenotype and indirect genetic effects through genetic nurture and other mechanisms. For example, the parent and the child may share a genotype associated with higher neuroticism, which could lead to less affectionate caregiving of the parent to the child [67, 68]. This, in turn, might bias our estimates of the direct effect of the neuroticism PGS on child psychiatric outcomes through altered parenting.
Strengths and Limitations:
The current study has the strengths of a large, diverse multi-ancestry sample, multigenerational family history assessment, and the latest Bayesian polygenic prediction method to test our hypotheses. Limitations of this study include that in most GWASs, the PGSs were derived from European-ancestry samples and, therefore, the generalizability of the summary statistics to other ancestry samples might be limited. Nevertheless, our findings were consistent across the analyses of multi-ancestry and European-only samples. In addition, PGSs still only account for relatively small variances compared to the heritability of depression estimated from twin studies or DNA, but PGS may become stronger predictors in future. Family history was assessed retrospectively by the caregiver, which may bias reports and underestimate effect sizes compared to samples with gold-standard clinician-based diagnoses for all family members. However, this limitation is mitigated by the fact that we previously showed that clinical results were similar between an interview-based design study [7, 8] and the ABCD study [6]. Lastly, the effect sizes that we found, while similar to those of prior polygenic score studies [69, 70] are small, suggesting that these PGS do not yet have high clinical significance. As GWAS discovery samples become larger and include multiple genetic ancestries, clinical significance may improve. On the other hand, the children in our sample are still young and before the median age of onset of depression. Some of the currently unaffected children might develop psychiatric disorders as the cohort ages, forming a stronger association between PGS and psychiatric disorders, and increasing the effect sizes.
In summary, we show that PGSs for depression and bipolar disorder are associated with family history of depression and that depression PGS mediates part of the association between (multigenerational) family history and offspring’s psychopathology. We demonstrate that having more previous generations affected with depression is linked to having higher polygenic risk for psychiatric disorders. The findings also implicate polygenic risk for depression as a potential mechanism for intergenerational transmission of depression, suggesting that integrating depression PGS with depression history may aid in identifying children at higher risk for psychopathology.