Intersectional inequalities in youth mental health measurement and outcomes: evidence from age 23 sweep of the Millennium Cohort Study, a nationally representative UK birth cohort

Mental ill-health is among the top global contributors to disability, morbidity, and mortality in young people [1-4]. This is even after considering that figures on the impact of mental ill-health in the population may underrepresent the true burden as they often fail to appropriately include so-called ‘diseases and deaths of despair’, such as deaths by suicide (which are also among the top leading causes of death in young people) [5]. Similar to other health outcomes, youth mental health is socially patterned, unequally distributed across social groups within the population (e.g., by gender, ethnicity, sexuality, and socioeconomic background) due to the unequal and unfair distribution of power, privilege and resources. Evidence on youth population mental health shows, for instance, how young women and youth from sexual minorities experience substantially poorer mental health outcomes and, in the case of sexual (and gender) minorities, much higher self-harm and suicide rates than their heterosexual peers [6-9]. Evidence also shows inequities by socioeconomic position and background, with young people from more disadvantaged socioeconomic positions and backgrounds experiencing worse mental health outcomes [10-12], and complex patterns of inequalities by ethnicity [7, 13]. However, the systems of oppression that underlie these (and other) social inequalities –e.g., sexism, racism, heterosexism and homophobia, classism and the material effects of income inequality and socioeconomic deprivation– are not independent of each other but, rather, interrelated. This is one of the central tenets of intersectionality theory, which was developed by Black feminist scholars to highlight how the experiences of Black women are shaped by both racism and sexism in ways that are unique and go beyond the “sum of the parts” [14-16]. In line with this framework, the experiences of –for instance– a young Asian lesbian woman from a socioeconomically advantaged background is mutually affected by all these (and other) overlapping social identities and positions, leading to mental health outcomes that are socially patterned in complex yet systematic ways. When quantitatively analysing social inequalities at the population level, inequalities that may disproportionately impact people at specific combinations of social identities and positions will remain unnoticed unless this intersectional complexity is explicitly acknowledged [17]. Fortunately, there is a growing acknowledgement of this intersectional complexity in population (mental) health equity research. Methods like the Intersectional Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (I-MAIHDA) [18] are being increasingly adopted to provide sociodemographic mappings of inequalities from an intersectional framework. However, applications of this insightful method (and other methods) often ignore some key challenges in the context of population mental health, including measurement and missing data issues. First, due to their largely subjective nature, measuring mental health usually relies on the use of self-reported questionnaires with scores that are typically derived by summing the different items or indicators. One of the key assumptions when comparing levels in these scores across groups is that there are no systematic differences in the measurement characteristics of the questionnaire across those groups. This is, that the different groups will respond to the different items depending on their underlying levels in the mental health outcome being measured [19]. If this assumption does not hold, differences in those scores may be due to factors other than genuine differences in the outcome under study (e.g., certain indicators being representative of the outcome under study in some groups but not others). Although the assumption of invariance of the measures is not frequently assessed across individual social identities and positions, it is even more rarely checked in across intersectional positions, despite the availability of methods to empirically test it and the relevance of the results of such exploration to multiple parties interested in population mental health equity [20]. Recent methodological developments make measurement invariance testing from an intersectional perspective more streamlined than ever before [21]. Second, data in population-based studies are often missing in ways that are also socially patterned. The sole analysis of cases with complete data (complete-case analysis) can bias the results in systematic ways, limiting the inclusivity and generalisability of the research [22, 23]. Addressing these challenges when obtaining intersectionality-informed evidence on youth mental health inequalities can help more accurately and inclusively represent populations and inequalities that otherwise may remain invisible. Because mental health (e.g., well-being) and ill-health (e.g., anxiety and depression) are not two ends of the same continuum but, rather, can co-exist in different ways and show different social patterns, evidence including measures of both mental health and ill-health can help provide a more holistic picture, one that can further inform policies aimed at both fostering well-being and reducing (or, ideally, preventing) ill-health in equitable ways, which is critical considering the deteriorating levels of youth mental health at the population level [2, 4]. In this study, shaped by the involvement of a Young People Advisory Board (YPAG) comprising young people with lived experience of mental ill-health and/or social discrimination, we will provide a sociodemographic mapping or snapshot of social inequalities in youth mental health, including both measures of mental health and ill-health. This mapping will be done from an intersectional framework, acknowledging the intersectionality of multiple systems of oppression as well as the abovementioned challenges in measurement and missing data. We will provide updated, fine-grained evidence on these inequalities using the latest data release from the Millennium Cohort Study, a nationally representative UK birth cohort representing young people at a critical transitional age (age 23) marked by post-education transitions among those engaging in higher education, changing social networks from work and education, and increasing financial independence.