The present study, based on a large Swedish cohort born in 1953 and their parents, shows that individuals whose parents suffered from mental health problems are more likely to also present with poor mental health in adulthood as measured by any hospital admission due to a number of psychiatric diagnoses, for example depressive disorders and bipolar affective disorders. These findings confirm existing research of intergenerational transmission of mental health [6,7,8,9,10,11,12]. However, this study contributes to the literature through its focus on adult mental health, in contrast to outcomes in childhood or adolescence. Another important difference compared to previous research is the low risk of response bias due to the use of register data instead of self-reported measures of mental health.
The main focus of this study was to explore whether intergenerational patterns of mental health differed depending on social/environmental circumstances in the child generation’s childhood years. It is well established that supportive social relationships are protective against mental health problems, whereas loneliness, weak peer networks, and peer problems (such as bullying) in childhood are risk factors of the same, both in childhood/adolescence [26, 41] and longitudinally into adulthood [36, 38]. The present study contributes with new knowledge through acknowledging the conjunction of both parental and child generation factors. As hypothesised, the identified intergenerational transmission of poor mental health was most evident among those in lower peer status positions at age 13. This suggests that childhood peer networks play an important role in the ‘spill-over effect’ of parental health .
A linear patterning of adult mental health problems depending on parental mental health and own childhood peer status was confirmed for men. In other words, the lower the peer status position of sons of parents with mental health problems, the greater the likelihood of presenting with own mental health problems in adulthood. Several possible mechanisms could explain why childhood peer status influences the strength of the intergenerational transmission of mental health problems. For example, children of mentally ill parents are not only at risk of poor attachment and neglect due to conflicts and inadequate parenting , they may also struggle with social stigma and shame of having a mentally ill parent which places those children in vulnerable positions in relation to peer interaction and making friends . It is also possible that boys to a higher degree than girls display less prosocial behaviour as a reaction to their parents’ mental health problems and thereby experience more trouble with friends, including bullying [18, 43].
Furthermore, previous research suggests that boys and men generally have fewer close friends and weaker social networks than girls and women [24, 44]. Given the strong mental health protective effects of social relationships, boys with low peer status from families with mental health problems may be particularly sensitive and at risk of developing own poor mental health. Typically, this pattern would also apply to girls. However, the findings suggest otherwise. Among women, the pattern was u-shaped; the combination of parental mental health problems and either very low (marginalised) or medium/high peer status increased the likelihood of adult mental health problems. These findings merit further attention. It is possible that gendered norms and expectations with regards to popularity and status play a role here. Studies suggest that aggressiveness is dominant in the perception of popularity in boys , while prosocial behaviour is valued in relation to popularity in girls [23, 46]. Hence, it is possible that for girls, being popular implies increased social pressure that might take its toll on their mental health through relational stress. Stress from social and relational responsibilities has been shown to be more prevalent as well as more detrimental for mental health in girls than in boys [46, 47].
To rule out the likely influence of childhood mental health status on both peer status and later mental health, analyses were adjusted for internalizing and externalizing problems in the child generation (data not shown). In accordance with Shepman , this did not substantially alter the results. In other words, childhood peer status matters for mental health in adulthood, especially for those with mentally ill parents, regardless of the child’s own mental health status.
Our findings did not show any clear differences in the strength of intergenerational transmission of mental health problems based on the gender of the parent. In the few existing studies exploring both maternal and paternal transmission of mental health, the associations seem stronger for same-gender-dyads than mother-son/father-daughter dyads [8, 16]. Andreas and colleagues  suggest that gendered parenting styles or gender cognition theory may explain such findings. Importantly, these hypotheses also suggest reciprocity, i.e., if the child identify with the parent of the same gender, this also applies to the parent. It is possible that this influences how parents report the mental health status of their child, a procedure applied in many studies . Our results may contradict previous findings because mental health data for both parents and children were drawn from registers instead of self-reports.
In line with previous studies, the current study has shown that positive peer relationships, as manifested through higher peer status positions, are overall beneficial for subsequent mental health outcomes. From a policy perspective, the results imply that school-based interventions aimed at improving children’s experiences with peers would provide an opening for also improving healthy development. This appears to apply particularly to boys. Important to note, however, is the possibility that girls how have mentally ill parents may come to respond differently to such efforts, at least in cases where more intensive peer interaction is accompanied by increases in social pressure and relational stress.
Strengths and limitations
The major strength of this study is that it builds on prospective data collected for a relatively large, community-based sample. Some limitations should nevertheless be addressed. One apparent shortcoming is that we have no access to genetic information which makes it impossible to determine if the intergenerational association in mental health problems is reliant upon the inheritance of genes related to mental illness or whether it is due to the assumingly troublesome circumstances that children living with mentally ill parents are exposed to. However, it should be emphasised that the primary focus of the current study was not to draw any causal inferences but, rather, whether the association varied according to experiences related to the children’s own social contexts, such as those reflected through peer status position. The empirical analysis did furthermore not set out to explain the associations between parental mental health problems, childhood peer status position, and own mental health problems in adulthood; therefore, we restricted the set of covariates to include only the most important confounders. An examination of potential pathways would require a careful consideration of temporality as well as appropriate statistical tools for assessing mediation, such as structural equation modelling. These are issues for future studies to look further into.
Another limitation concerns the measurement of mental health problems. For the child generation, we have relied on in-patient care, which captures only the most severe cases. Some other important, and potentially conflicting, sources of bias may also be present here. On the one hand, children whose parents’ mental health problems are registered by the authorities may be more closely monitored by the social and health care services and thus be more likely become diagnosed with a mental illness. On the other hand, help-seeking behaviours may be positively associated with peer status, which would lead individuals with lower status positions to become diagnosed to a lesser extent than expected. Moreover, due to the small proportion of individuals being hospitalised due to mental disorders, we were not able to examine specific diagnoses (e.g. depression and/or anxiety) which potentially could have provided deeper insights into possible mechanisms behind the patterning found here.
For the parental generation, only a very broad indicator of mental health problems was available in our data. A related issue has to do with the generalisability of the results when it comes to later cohorts. For example, having a mentally ill parent may have meant something different in our 1953 cohort compared to children of today: increased awareness in the general population around mental health problems could suggest that the stigma surrounding this topic has decreased since the 1950s and 1960s and, to the extent to which the intergenerational association in mental health problems is socially induced, this would lead to a weakening of the intergenerational transmission over time.
Finally, we only had information about peer status position at age 13. It would have been preferable to have assessments of peer status across multiple time points in order to account for stability and change.