BMC Public Health · Aug 6, 2026

Habitual sodium intake and metabolic abnormalities in Korean adults with obesity: a cross-sectional analysis of a prospective cohort

By Park SH, Lee M, Kang HJ et al.

BACKGROUND: Sodium intake is well recognized for its contribution to hypertension, but growing evidence suggests broader metabolic effects, particularly among individuals with obesity who are vulnerable to metabolic disturbances. This study investigated the cross-sectional associations between habitual sodium intake and metabolic diseases and related markers using baseline data from a prospective cohort of Korean adults with obesity. METHODS: Participants with obesity (BMI ≥ 25 kg/m²) and at least one metabolic condition were recruited from Severance Hospital, Seoul, Republic of Korea. Sodium intake was assessed using a food frequency questionnaire, and associations with diabetes, dyslipidemia, hypertension, and cardiovascular disease were evaluated using multivariable regression models. Mean differences in metabolic markers across sodium tertiles were examined using analysis of covariance with sex stratification. RESULTS: Of the 633 individuals enrolled, 487 with obesity were included in the final analysis after excluding non-obese participants and those with implausible energy intake. No statistically significant associations were observed between sodium intake and the prevalence of type 2 diabetes, dyslipidemia, hypertension, or cardiovascular disease. In sex-stratified biomarker analyses, LDL-C, non-HDL-C, CRI-I, and CRI-II differed across sodium tertiles among women, whereas no such differences were observed in men. CONCLUSIONS: In Korean adults with obesity and pre-existing metabolic conditions, increasing sodium intake was not significantly associated with higher prevalence of major metabolic diseases. In exploratory sex-stratified analyses, several atherogenic lipid markers differed across sodium intake tertiles among women only. This within-women pattern should be regarded as hypothesis-generating and requires confirmation in future longitudinal analyses of this cohort.

doi.org/10.1186/s12889-026-28903-y Open access PubMed 42977925 Populations