Patterns of adversity, patterns of burden: a person-centred analysis of adult health and behavioural-risk profiles in Abu Dhabi

Abstract Background Adverse childhood experiences (ACEs) have been associated with multiple adult physical-health, mental-health and behavioural indicators. This study identified adult-reported health and behavioural-risk profiles in an Abu Dhabi convenience sample and examined their associations with alternative operationalisations of retrospectively reported ACEs. Methods Among 922 adults, 918 contributed to latent class analysis of 18 physical-condition, mental-health-condition and behavioural indicators. ACE exposure was represented by a standardised version of the archived 13-domain available-item count and by three re-estimated classes. In the count, unobserved domains were treated as unendorsed when at least one domain was observed; otherwise, the count was missing. Multinomial models used 200 pseudo-class draws to propagate adult-profile uncertainty and, for ACE-class models, uncertainty in both classifications. Unadjusted and age-, gender- and nationality-adjusted relative risk ratios (RRRs) were reported. Results The retained four-class solution comprised Low burden, Substance/prescription vulnerability, Cardiometabolic burden and Mental health and multi-burden. Their modal-assignment proportions were 57.1%, 15.9%, 11.5% and 15.5%, respectively. Each one-standard-deviation increase in ACE count was most strongly associated with Mental health and multi-burden versus Low burden (unadjusted RRR 2.94, 95% confidence interval 2.38–3.63; adjusted 2.86, 2.30–3.55). Adjusted count associations with Substance/prescription vulnerability (1.51, 1.15–2.00) and Cardiometabolic burden (1.39, 1.03–1.86) had unadjusted intervals including 1. Versus Low adversity, Household/interpersonal adversity was associated with Mental health and multi-burden (adjusted 7.21, 4.12–12.60); Polyvictimisation/high violence was associated with Mental health and multi-burden (21.31, 8.74–51.95) and Cardiometabolic burden (3.92, 1.21–12.72), but estimates were imprecise. Conclusions Retrospectively reported ACEs were associated most consistently with membership in the Mental health and multi-burden profile. Several smaller adjusted associations were specification-sensitive, and estimates involving Polyvictimisation/high violence were imprecise.