Association of serum water-soluble vitamin exposures with the risk of metabolic syndrome: results from NHANES 2003-2006

Abstract

Introduction: Existing evidence suggests an association between certain vitamins and metabolic syndrome (MetS), but few epidemiological studies have focused on the effects of multivitamin co-exposure on MetS. This study aims to investigate the associations of the individual or multiple water-soluble vitamins (i.e., vitamin C (VC), vitamin B9 (VB9), and vitamin B12 (VB12)) with co-exposure to MetS, as well as the dose-response relationships among them.

Methods: A cross-sectional study was conducted by employing the National Health and Examination Surveys (NHANESs) 2003-2006. Multivariate-adjusted logistic regression models were used to explore the association between individual serum water-soluble vitamins and the risk of MetS and its components, including waist circumference, triglyceride, high-density lipoprotein, blood pressure, and fasting plasma glucose. Restricted cubic splines were performed to explore the dose-response relationships among them. The quantile g-computation method was adopted to explore the associations of multiple water-soluble vitamins co-exposure with MetS risk and MetS components.

Results: A total of 8983 subjects were involved in the study, of whom 1443 were diagnosed with MetS. The MetS groups had a higher proportion of participants with age ≥60 years, BMI ≥30 kg/m², and insufficient physical activity. Compared with the lowest quartile, the third (OR=0.67, 95% CI: 0.48, 0.94) and highest quartiles (OR=0.52, 95%CI: 0.35, 0.76) of VC were associated with lower MetS risk. Restricted cubic splines showed negative dose-response relationships among VC, VB9 and VB12, and MetS. Regarding MetS components, higher VC quartiles were associated with lower waist circumference, triglyceride, blood pressure, and fasting plasma glucose, while higher VC and VB9 quartiles were associated with higher high-density lipoprotein (HDL). Co-exposure to VC, VB9, and VB12 was significantly inversely associated with MetS, with ORs (95% CI) of 0.81 (0.74, 0.89) and 0.84 (0.78, 0.90) in the conditional and marginal structural models, respectively. Furthermore, we found that VC, VB9, and VB12 co-exposure were negatively associated with waist circumference and blood pressure, while VC, VB9, and VB12 co-exposure were positively associated with HDL.

Conclusion: This study revealed negative associations of VC, VB9, and VB12 with MetS, while the high water-soluble vitamin co-exposure was associated with a lower MetS risk.

Keywords: NHANES; co-exposure; metabolic syndrome; vitamin; weighted quantile sum regression.