Association between Dietary Patterns and Incident Hypertension in Regards (Reasons for Geographic and Regional Differences in Stroke) Participants

BACKGROUND Diet contributes to the development of hypertension. We examined associations between five dietary patterns and incident hypertension in the REasons for Geographic and Racial Differences in Stroke (REGARDS) cohort. METHODS Participants without prevalent hypertension (blood pressure < 130/80 mmHg or antihypertensive medication use) were evaluated at baseline and follow-up. Dietary patterns were developed using food frequency questionnaires. Diet adherence was divided into quartiles, and Modified Poisson regression estimated risk ratios of incident hypertension by quartile. We adjusted for covariates sequentially. Sex and race interactions were tested with a priori stratified analyses. An identical sensitivity analysis was completed using the Joint National Comission-7 definition of hypertension, the prevailing hypertension definition at the time of enrollment (>140/90 mmHg or antihypertensive medication use). RESULTS Of 30,239 REGARDS participants, 3,423 met inclusion criteria and 1,439 (42.0%) developed hypertension. Participants highly adherent to the Alcohol/Salad pattern had an increased risk of incident hypertension [RR = 1.12, 95% CI (1.00, 1.27)]. This persisted in White females after controlling for alcohol [RR = 1.27, 95% CI (1.04, 1.54)]. The remaining dietary patterns had no associations in fully adjusted models. Sex/race interactions were not significant. The sensitivity analysis included 6,582 participants, of whom 2,295 (34.9%) developed hypertension. No associations were found between diet and incident hypertension in fully adjusted models. CONCLUSIONS The Alcohol/Salad dietary pattern portends the highest risk of incident hypertension. No further associations existed between other dietary patterns and incident hypertension after adjusting for covariates. Our findings suggest that diet is contributory, but not an exclusive cause, of incident hypertension.