Associations and mediators of estimated sodium intake with cardiovascular mortality: data based on a national population cohort.

Abstract

Background: The current research on the association of sodium excretion with health outcomes is inconclusive, despite its linear relationship with blood pressure. Moreover, there is a lack of evidence regarding sodium intake. This study aims to estimate sodium intake, examine its associations with mortality, and explore potential factors that may mediate these associations.

Methods: Based on a nationwide community-based population cohort in China that covered 215 counties from 31 provinces, we collected fasting morning urine from 270,991 participants aged 35-75 years during Dec 2015 and Dec 2019. We calculated 24-h urinary sodium excretion using the Kawasaki formula, and estimated daily sodium intake by adjusting for sweat excretion based on a model with ambient temperature. We fitted Cox regression models to examine its independent hazard ratios (HR) and 95% confidence intervals (CI) on mortality, and assessed the causal mediation effects of metabolic factors.

Results: Among the included participants with an average age of 56 years, the mean of estimated daily sodium intake was 222.9 ± 71.0 mmol. Sodium intake was lower in elderly and women, but higher in participants living in the north or rural areas (P < 0.001 for both). In participants without antihypertensive treatment, blood pressure was positively related to sodium intake (4.14/1.58 mmHg per 100 mmol/day, P < 0.001). Compared with participants in the quintile 3 of sodium intake (i.e., 200.8-235.1 mmol/day), those in the quintile 1 (i.e., < 163.5 mmol/day) had an adjusted HR of 1.17 (95% CI: 1.05-1.30) for cardiovascular mortality, after adjusting for demographic, socioeconomic, behavioural, and clinical characteristics, as well as urinary potassium, with heart rate (29.4%) and blood glucose (18.5%) as major significant mediators. In the meantime, participants in the quintile 5 (i.e., > 278.8 mmol/day) had an adjusted HR of 1.18 (95% CI: 1.05-1.32), with systolic blood pressure (24.6%) and body mass index (2.4%) playing substantial mediating effects.

Conclusions: Both high and low sodium intake are associated with increased cardiovascular mortality. The lowest risk is observed at an estimated sodium intake of 200.8-235.1 mmol/day (equivalent to 4.6-5.4 g/day). Mediation analysis suggests that blood pressure, heart rate and glycaemic disorders could be plausible explanations for this U-shaped association.