Background:
Previous studies have suggested that sodium reduction could be a viable strategy for reducing heart failure-related disease burden. Relatively few studies had been conducted on dietary sodium and the incidence of heart failure (HF) and its major subtypes; HF with preserved ejection function (HFpEF) and HF with reduced ejection function (HFrEF).
Hypothesis:
We hypothesized that dietary sodium was positively associated with the risk of HF and its major subtypes with a linear dose-response relationship, whereas dietary potassium was inversely associated with the risks of these outcomes.
Methods:
Our observational cohort study included 118,057 racial/ethnically diverse postmenopausal women recruited during 1993-1998 and followed up until 2015 in the Women’s Health Initiative. Women who reported a history of HF, were underweight, or had implausible/missing food frequency questionnaire (FFQ) data were excluded at baseline. The exposures of our study were FFQ-measured dietary sodium and potassium calibrated by recovery biomarkers estimated from 24-hour urine excretion collections. The main outcomes were hospitalized heart failure, including HFpEF and HFrEF subtyping, as adjudicated by trained physicians.
Results:
During up to 22 years of follow-up, 2,533, 1,048 and 673 participants developed HF, HFpEF and HFrEF, respectively. The mean age of the study population was 63.4 years, in which 84.3% (99,297 of 118,057) were white, 7.8% (9,150 of 118,057) were African American, 3.8% (4,469 of 118,057) were Hispanic and 4.1% (4,832 of 118,057) were other race/ethnicity groups. The median of calibrated dietary sodium intake was 2,712.4 mg/day (interquartile range: 2,503.3 mg/day-2,948.4 mg/day) and the median of dietary potassium intake was 2,494.5 mg/day (interquartile range: 2,249.8 mg/day-2,718.2 mg/day). After adjusting for potential confounding variables and risk factors of HF, calibrated sodium intake was positively associated with the incidence of HF (HR Q5 vs. Q1 =2.59, 95% CI: 2.26-2.98, P-trend <0.001). Comparing extreme quintiles of sodium intake, the HR was 2.87 for HFpEF (95% CI: 2.29-3.60) and 1.71 for HFrEF (95% CI: 1.28-2.30, both P-trend<0.001). The dose-response relationships between calibrated sodium intake and the risk of HF and HFrEF were non-linear with accelerated increasing risks at higher intake level, while the dose-response relationship for HFpEF was linear. Similar positive associations were observed for the analyses on calibrated sodium/potassium ratio, whereas calibrated potassium intake was not associated with the risk of HF and its major subtypes.
Conclusions:
Higher sodium intake is associated with increased risk of HF and its major subtypes. The observed positive association appears to be stronger for HFpEF than for HFrEF. These findings help to inform dietary recommendations for primary heart failure prevention.
Author Correction: Meta-analytic research of the dose-response relationship between salt intake and risk of heart failure
