Urinary volatile organic compounds and stroke risk: A cross-sectional analysis of NHANES data

Volatile organic compounds (VOCs) are ubiquitous environmental pollutants. However, the effect of ambient VOCs, alone or in combination, on stroke incidence in the general population remains unclear and warrants further investigation. Urinary VOC metabolites were assessed using data from the National Health and Nutrition Examination Survey for the years 2011 and 2020. Three statistical techniques were used: weighted logistic regression, weighted quantile sum regression, and Bayesian kernel machine regression (BKMR). Mediation analysis was also carried out to identify potential mediating effects. In univariate weighted logistic regression, the metabolites CEMA, 3HPMA, and DHBMA showed significant associations with stroke. However, both weighted quantile sum and BKMR analyses did not identify a significant association between VOC mixtures and stroke risk in the overall population. In subgroup analyses, BKMR suggested a positive association between VOC mixtures and stroke risk only among individuals with diabetes, with DHBMA being the main contributor. No significant interactions between individual VOCs were observed in the total population. Several VOC metabolites, including N-acetyl-S-(2-carboxyethyl)-l-cysteine (CEMA), 3-hydroxypropyl mercapturic acid (3HPMA), and N-acetyl-S-(3,4-dihydroxybutyl)-l-cysteine (DHBMA), were associated with stroke risk, but no overall mixture effect was observed in the general population. The association was limited to the diabetic subgroup and warrants cautious interpretation and further longitudinal validation.