Comparison with existing literature

Service use has previously been examined in this cohort when the participants were aged 9–17 years. ^{Reference Potter, Mars, Eyre, Legge, Ford and Sellers8} At that time, although only a third of those with psychiatric disorder were in contact with services, only formal service use was examined, including educational, social, youth justice and health services. The current study builds on this work by including informal services and, with the focus on early adult life, a developmental transition to independence associated with the emergence of mental health difficulties, ^{Reference Kessler, Angermeyer, Anthony, De Graaf, Demyttenaere and Gasquet9,Reference Solmi, Radua, Olivola, Croce, Soardo and Salazar de Pablo10} changes in support services and personal and social changes and challenges (e.g. education, employment, relationships). ^{Reference Hovish, Weaver, Islam, Paul and Singh34} The current work also helps to address the lack of long-term studies in this population, and suggests that access to formal support among those with a disorder increases from childhood/adolescence to young adulthood (from a third to just over half the sample).

The levels of support accessed in this sample are higher than those reported in some earlier studies involving young adults with mental health difficulties in the UK. Salaheddin et al showed that 65% of 16- to 25-year-olds with mental health difficulties had accessed formal or informal help (including peer support). ^{Reference Salaheddin and Mason15} This compares with 82% of those with a disorder who accessed any support in our study, and might suggest that individuals with a parent (known to services) with mental health difficulties may be more likely to seek support, or could be explained by differences in methodology (e.g. participant characteristics, definition of difficulties and support).

Factors associated with access to support in the current study are consistent with those found in previous studies of young adults in the UK, such as mood disorders, severity of difficulties, comorbidity, suicide risk and female gender. ^{Reference Hodgson, Shelton and van den Bree16,Reference Duncan, Rayment, Kenrick and Cooper18,Reference Twomey, Baldwin, Hopfe and Cieza20} The current study extends this work by looking at a wide range of sociodemographic factors (including age, social support, living alone, NEET, education status and personal income) and suggests that, for some of these factors, associations may be different for formal and informal support.

Strengths and limitations

Data were drawn from a large study of young adults at elevated risk for psychopathology, recruited mainly from general practice, and who were followed prospectively over 13 years (from childhood/adolescence) and across key developmental phases. Assessments were rigorous, involving multiple informants and diagnostic interviews, and access to a broad range of support was considered.

The findings must, however, be interpreted considering the following limitations. The interview used to capture mental health support relied on the individual’s recall and interpretation of the question on whether they were currently receiving/using support. Therefore, some forms of support may have gone unreported. Broad definitions for support were used, although sensitivity analyses were conducted excluding social networks from informal support. Data on satisfaction with help received were based on lifetime reporting and were not specific to the type of services accessed. Social desirability bias may have led to over-reporting of satisfaction, although participants were not reporting back to service providers. A large number of analyses were conducted, and it is therefore possible that some results may have arisen by chance. Replication of these findings is needed in further studies.

As with all longitudinal studies, there was some attrition. Of the original 337 families in the sample, 197 (58.5%) took part in wave 4, with 144 participants having data on both disorder and support, leading to small subsamples for certain analyses (e.g. those with a disorder). However, IPW was used to account for attrition. Most of the parents in the cohort at baseline were depressed mothers (>90%), which may have affected the generalisability of the findings. This might reflect how research participation and help-seeking, in general, are lower in males. ^{Reference Ennis, McLafferty, Murray, Lapsley, Bjourson and Armour17–Reference Klineberg, Biddle, Donovan and Gunnell21} Further research is needed to understand risk of psychopathology and access to support in children of depressed fathers. The general strengths and limitations related to the EPAD cohort are also discussed in other publications. ^{Reference Powell, Lennon, Bevan Jones, Stephens, Weavers and Osborn4,Reference Mars, Collishaw, Smith, Thapar, Potter and Sellers24–Reference Powell, Agha, Jones, Eyre, Stephens and Weavers27}

Implications for practice

Young adults in this study accessed a range of support, and distinct levels and types of support are likely to be required based on individual needs and the severity of difficulties/impairment. UK guidelines for depression recommend tailored approaches including guided self-help, psychotherapy and medication, depending on the presentation from subthreshold/mild to severe depression. ³⁵ A UK study of older adolescents ^{Reference Neufeld, Dunn, Jones, Croudace and Goodyer36} found beneficial treatment effects on depressive symptoms only in those who met the criteria for psychiatric disorder or who had high subthreshold symptoms and impairment, suggesting that this is a suitable threshold for formal services.

In this sample, 44% of young adults with a disorder were not accessing any formal support, suggesting that many in need of support are not receiving it. Access to support among this subgroup was predicted by indicators of severity and impairment, alongside demographic factors such as living alone. Specifically, those with lower education status and personal income were less likely to seek formal support, suggesting they represent a hard-to-reach group where targeted interventions could improve access. Education status was also negatively associated with formal support in the whole sample, while those who were NEET were more likely to access support. These findings may appear contradictory, but could be explained by higher levels of difficulties and impairment among those who are unemployed.

The proportion of participants with a disorder receiving support increased from 56.5% (formal support) to 82% when including informal support, highlighting the reliance on less formal sources. This can be explained in part by the lack of formal services and growth in self-help, online resources and digital devices, as well as young adults’ comfort with informal support, attributable to factors including accessibility, convenience, trust, confidentiality and stigma. ^{Reference Michelmore and Hindley14,Reference Singh, Zaki and Farid37} This also reflects the importance of social networks, and the proportion accessing any support reduced to 68.7% following exclusion of family/friends.

Direct involvement of young adults is important to capture their lived experiences and inform both research and practice. ³⁸ To enrich our findings and guide future research in this area, a focus group was arranged with young adults to explore their experiences and perspectives on accessing support (Supplement 6). Of relevance to this study, they reported that their help-seeking behaviour could be influenced by how those close to them (e.g. parents/carers) managed their health difficulties and their experiences of accessing support. This aligns with previous findings suggesting that stigma associated with parental mental illness can deter family members from seeking support. ^{Reference Reupert, Gladstone, Helena Hine, Yates, McGaw and Charles39} The group also confirmed that young adults access a range of online and informal support and identified potential facilitators for help-seeking, including role models, public discussions about mental health and practitioner training. Reported barriers were consistent with those arising from the literature, including stigma, ^{Reference Salaheddin and Mason15,Reference Ennis, McLafferty, Murray, Lapsley, Bjourson and Armour17,Reference Pickles, Rhind, Miller, Jackson, Allister and Philp22,Reference Cage, Stock, Sharpington, Pitman and Batchelor23} difficulty identifying or expressing concerns ^{Reference Salaheddin and Mason15,Reference Klineberg, Biddle, Donovan and Gunnell21,Reference Cage, Stock, Sharpington, Pitman and Batchelor23} and uncertainty about where to seek help. ^{Reference Ennis, McLafferty, Murray, Lapsley, Bjourson and Armour17}

The group’s recommendations for educational and anti-stigma initiatives reflect wider calls for help-seeking interventions to improve mental health knowledge and stigma. ^{Reference Singh, Zaki and Farid37} Participants stressed the need for improved coordination across health and student services and charities, and acceptable and effective resources and services that are co-developed with users. ^{Reference Bevan Jones, Stallard, Agha, Rice, Werner-Seidler and Stasiak40} Further participatory work and mixed-methods research are required, particularly with children of depressed parents, to explore access to support.

In conclusion, all participants in this study were young adults with parents with recurrent depression and were therefore at elevated risk of mental health difficulties. Participants accessed a variety of sources of mental health support, with just over half of those with a psychiatric disorder accessing formal help, while one in five were not receiving any support. Further work is needed to ensure early identification of difficulties and access to support, and a better understanding of the types of support that meet the needs and preferences of young adults, including those at risk.