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.