Intersection of ethnicity and physical multimorbidity on incidence of mental health conditions – retrospective cohort study from superdiverse south London

Problem

Over half of GP consultations in the UK involve patients with multimorbidity. Incidence of anxiety, depression and serious mental illness (SMI) vary substantially by ethnic group, which is relevant to create equitable health services that adequately address multimorbidity clusters. It is not known whether physical long-term conditions (LTCs) may increase ethnic disparities in mental LTCs.

Aim: Does the presence of a physical LTC exacerbate ethnic disparities in incidence of mental health diagnoses?

Approach

Retrospective cohort study of Lambeth DataNet, a pseudonymised database of adult patients registered to GP practices from the borough of Lambeth from 2005-2022. Patients without mental health LTCs were followed up with increasing age until diagnosis of a mental health LTC (anxiety, depression or SMI), death, deregistration or the end of follow-up. Two cohorts were derived: those with, and those without, comorbid physical LTCs. Upon diagnosis of a physical LTC, individuals were censored and entered the physical LTC cohort. For each cohort, cumulative incidence curves and mixed-effects multivariable Cox regression estimated the effect of ethnic group on mental health LTC diagnosis after adjustment for year of follow-up (modelled as a restricted cubic spline); sex; local-area deprivation; smoking status; substance use and practice-level variation in diagnosis.

Findings

658,500 individuals with a recorded ethnic group were followed up for a median 3.5 years. Cumulative incidence of anxiety, depression and SMI was 7.6%, 4.7% and 0.5%, respectively. Cumulative incidence of mental LTCs was higher for those with physical LTCs compared to those without physical LTCs for anxiety (12.9% vs 6.9%), depression (10% vs 4%) and SMI (0.9% vs 0.4%). Ethnic disparities in incident depression were greater amongst those with physical LTCs, particularly comparing the Black African, Black Caribbean and Black Other groups which showed lower incidence rates compared with the White British population (adjusted hazard ratio (aHR) for those with physical LTCs 0.5, 0.74, 0.63 respectively; without physical LTCs 0.62, 0.99, 0.84 respectively). The presence of a physical LTC diagnosis was not associated with greater ethnic disparities in anxiety diagnoses. Physical LTC diagnoses were associated with substantial attenuation of ethnic disparities in SMI incidence (aHRs compared to White British group for other ethnic groups without physical LTCs 0.78–3.12, with physical LTCs 0.88–1.82).

Consequences

Physical LTCs are associated with a higher incidence of anxiety, depression and SMI. Ethnic disparities in anxiety diagnosis remain constant regardless of physical LTC status, whereas ethnic disparities in depression diagnosis tend to increase for those with physical LTCs. This is most pronounced for the Black ethnic groups, which may represent underdiagnosis of mental LTCs in those with physical LTCs. Non-significant ethnic disparities in SMI diagnosis for those with physical LTCs may represent differential engagement with care for those with and without physical LTCs.