Dietary Intakes of Soy and Isoflavones and Risk of Mortality in a Population with High Soy Consumption.

Abstract

Background: Soy and isoflavones have been linked to various health benefits, but their associations with all-cause and cause-specific mortality remain unclear.

Objectives: We examined the associations of dietary intakes of isoflavones, soy protein, and soy foods with mortality risk in a population-based cohort study.

Methods: We included 39,286 adults aged ≥ 19 y from the Korea National Health and Nutrition Examination Survey (2007-2015). Dietary intakes were assessed using the 24-h dietary recall. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). Dose-response relationships were evaluated using restricted cubic splines.

Results: Over a median follow-up of 8.5 y, 1715 deaths were documented. Lower isoflavone intake was associated with a higher mortality risk, with a plateau in risk at higher intake levels. Compared with the lowest quintile of isoflavone intake (<1.3 mg/d), HRs (95% CIs) for all-cause mortality in the second to fifth quintiles were 0.72 (0.62, 0.84), 0.72 (0.62, 0.84), 0.76 (0.66, 0.88), and 0.74 (0.65, 0.85), respectively. For cardiovascular disease mortality, a significantly lower risk was observed in the third quintile compared with the lowest quintile (HR: 0.69; 95% CI: 0.50, 0.96). The highest quintile of isoflavone intake was associated with lower risks of cancer (HR: 0.77; 95% CI: 0.59, 0.99) and respiratory disease mortality (HR: 0.50; 95% CI: 0.33, 0.77) compared with the lowest quintile. Similar patterns were observed for soy protein and soy food intake. Dose-response relationships were nonlinear, with the lowest all-cause mortality risk observed around median intake levels (9 mg/d for isoflavones, 4 g/d for soy protein, and 33 g/d for soy foods).

Conclusions: Our findings suggest that lower intakes of soy and isoflavones are associated with a higher risk of mortality in the population where soy consumption is prevalent.