Abstract Background Low socioeconomic position (SEP) in childhood is robustly associated with adolescent mental health difficulties, yet this relationship is changing over time due to contextual and conceptual differences in these constructs. Consistent SEP and mental health data from different cohorts are required to understand changes over time, which is uncommon. This study used retrospectively harmonised data from three UK birth cohorts to investigate the changing relationship between SEP and mental health between the 1970s and 2000s. Methods Mental health and SEP data from three UK birth cohorts (British Cohort Study BCS70, Avon Longitudinal Study of Parents and Children ALSPAC, Millennium Cohort Study MCS) were harmonised and tested for measurement invariance. Latent factors reflecting internalising, behavioural, and ADHD symptoms, were examined as distal outcomes of SEP latent classes. Results Partial scalar invariance was supported with multi‐group confirmatory factor analysis. Although the structure of the SEP latent classes differed considerably across cohorts, the most disadvantaged class in each cohort experienced greater levels of mental health difficulties. In the ALSPAC, only behavioural symptoms were significantly predicted by disadvantaged class membership; in both the BCS70 and MCS, all symptom domains were significant. The strongest association between the most disadvantaged class and mental health symptoms was observed in the MCS. Conclusion This study assessed the relationship between SEP and mental health using a novel methodological approach to harmonise and compare cohort data. Findings indicate that mental health may be more strongly related to socioeconomic disadvantage in recent cohorts, despite differences in the structure of SEP. Future research should utilise harmonisation to understand time trends in prevalent societal issues.
Rawers et al. (Sat,) studied this question.