Accumulation of perceived discrimination over time and likelihood of probable mental health problems in UK adults: A longitudinal cohort study.

Background: Limited research has examined whether accumulation of discrimination over time is associated with worse mental health and whether such experiences are related to socioeconomic status (SES).

Methods: A sample of UK adults with self-reported discrimination experiences (n = 3863) was taken from three waves of the UK Household Longitudinal Study (2015-2020). Multinomial logistic regression assessed associations between SES (income, education, occupation) and cumulative discrimination (number of timepoints discrimination was reported). Logistic regression models assessed prospective associations between cumulative discrimination and probable mental health problems (GHQ-12; 4+ threshold).

Results: Those with lower income were more likely to report discrimination at one timepoint (vs. none). No SES measures were associated with experiencing discrimination at multiple timepoints. Participants who reported one timepoint of discrimination (vs. no experiences) were significantly more likely to report probable mental health problems (OR = 1.47, p < .001, 95% CI 1.20-1.80). Moreover, compared to those experiencing one timepoint, participants reporting multiple timepoints of discrimination were significantly more likely to report probable mental health problems (OR = 1.46, p = .002, 95% CI 1.15-1.86), indicating a cumulative association between discrimination and mental health. There was limited evidence that SES moderated this cumulative association.

Limitations: Mental health measures were based on self-report questionnaires and not a clinical diagnosis.

Conclusions: Amongst a sample of UK adults, perceiving discrimination at multiple timepoints increased the likelihood of experiencing probable mental health problems. There was limited evidence that this cumulative association differed by SES. National measures designed to reduce discrimination may benefit mental health.