We set out to investigate the potential sex differences in the association between diabetes and depressive symptoms by conducting an interaction analysis, and to investigate whether sex mediates the effect of diabetes on depressive symptoms.
We conducted analyses on cross-sectional data of adults aged 20 years or older in the Mexican National Health and Nutrition Survey 2018–2019 (ENSANUT 2018–2019). Diabetes was defined by self-reported medical diagnosis, and depressive symptoms were measured using the seven-item Centre for Epidemiologic Studies Depression scale. First, an unadjusted interaction analysis was conducted. Second, the inverse probability of treatment weighting was applied to account for imbalances and biases. Third, the four-way decomposition method was used to estimate the potential mediating effect of sex.
In the study population (N=43 074), the prevalence of diabetes was 9.3% for men and 11.7% for women. Depressive symptoms were more prevalent in women (19.0%) than in men (9.5%). Women with diabetes had the greatest odds of having depressive symptoms, compared with men without diabetes (ORwomen-diabetes3.49 (95% CI: 3.16 to 3.86)). The interaction analysis indicated that diabetes and sex interact on both, multiplicative and additive scales (ratio of ORs (95% CI) 1.22 (1.02 to 1.45), and relative excess risk due to interaction (95% CI) 0.99 (0.63 to 1.36)). The four-way decomposition analysis showed that the interaction effect between diabetes and sex is larger than the mediation effect.
We found a positive interaction between diabetes and sex in the odds of having depressive symptoms. Mental health and diabetes care services planning would benefit from adopting a sex-informed approach.