Intersectional analysis of sex and socioeconomic status on obesity and comorbidity risk in Finland

Abstract Obesity is a public health concern with long-term health consequences. Sex and socioeconomic status (SES) influence obesity, but they have been studied separately. Intersectional theory suggests that different health risks arise from combined effects of multiple social dimensions. The study aims to identify whether intersection of sex and SES affect obesity (and comorbidity) risk beyond individual factors. The study used data from longitudinal NFBC1966 to examine whether intersection of sex and SES at age 31 was associated with risk of obesity (and comorbidity) at age 46. Analysis included 1,441 individuals with complete data. Binary outcome was defined: normal BMI and no comorbidities; obesity (and comorbidities). MAIHDA was used to estimate fixed effects for sex, income, occupation, education; and random intercepts for 24 intersectional strata from combination of these variables. Predicted probabilities of obesity (and comorbidity) varied across intersectional strata, from 60% to 80%. Highest-risk groups were women with low income and no employment; lowest-risk groups were men with high income and post-secondary education. But interaction effects and residuals showed several groups had higher or lower risk than predicted by main effects. Fixed effects for sex and SES were not statistically significant, stratum-level variance was unchanged from the null model (SD 0.37). Variance Partition Coefficient (VPC) was low in both models (∼4%), Proportional Change in Variance (PCV) was minimal (0.83%). Discriminatory accuracy was low (AUC 0.50 null model, 0.55 full model). Key limitation is small sample size within some strata, which may reduce estimate precision. Findings indicate intersectional effects contribute to obesity (and comorbidity) risk, even when individual predictors are weak. This supports value of MAIHDA in identifying disparities across social group combinations that association models may overlook. This can inform targeted prevention for high-risk groups. Key messages • MAIHDA identified 20 percentage point differences in obesity (and comorbidity) risk across intersectional groups, despite non-significant fixed effects for sex and SES. • Findings support stratified prevention by showing that obesity (and comorbidity) risk is shaped by combined effects of sex and SES, not by each factor alone.