Background There is ample evidence that hypertension prevalence increases when socioeconomic status (SES) decreases. However, sex differences in this relationship has been less studied. Investigating potential sex differences could help understand the mechanisms of social health disparities. The aim of this work was to explore the pattern of associations between several indicators of SES and hypertension across sexes in a large sample of French adults.
Methods In this cross-sectional analysis, participants are adults aged between 18 and 69 years old recruited to the CONSTANCES cohort over the period 2012–2015 in 16 recruitment centers. SES was estimated using education (individual level), income (household level) and an indicator of residential socioeconomic deprivation, FDep (municipal level). Log-binomial and Poisson regressions with robust variance estimations were used to estimate the Risk Ratios (RR) comparing the extreme levels of SES and to test for interaction of sex in the associations between SES and hypertension prevalence.
Results A total of 62,247 individuals (53% women, mean age 48±13 years) were included. Age-standardized prevalence of hypertension was 30.1% [95%CI=29.7–30.6], higher in men (37.3% [95%CI=36.6–38.0]) than in women (23.2% [95%CI=22.7–23.8]).
Globally, we found steep socioeconomic gradients of hypertension in both genders. Education showed the strongest association with hypertension prevalence, especially among women (p for interaction between sex and education <0.001): age-adjusted RR comparing the lowest versus highest level of education were 1.57 [95%CI=1.47–1.68] in women and 1.26 [95%CI=1.21–1.31] in men. Income and FDep also displayed strong associations with hypertension, but we found no interaction with sex (p for interaction=0.420 and 0.236 respectively). Age-adjusted RR comparing the lowest versus highest level of household income was 1.30 [95%CI=1.21–1.39] and RR comparing the most versus least deprived quintile of residential areas was 1.26 [95%CI=1.22–1.31]. Adding all the three indicators in a full model resulted in a more pronounced attenuation of the income and FDep gradients compared to education, however all the associations remain significant in men and women.
Discussion In this sample, individual education displayed different patterns of association with hypertension across sexes, but not household income nor neighborhood deprivation. This result underlines that systematic stratification by sex may not be appropriate in all analyses focusing on hypertension as outcome of interest. Besides, these results suggest that education attainment has a greater impact on women in preventing hypertension and therefore prevention should start early. More work is needed to understand sex differences in the causal pathway linking SES and hypertension.