Objective: The purpose of this study was to investigate the correlation between plasma homocysteine (Hcy) levels and cardiovascular disease (CVD) in United States adults based on the National Health and Examination Survey (NHANES) database of the United States.
Methods: Data from two survey periods (2003-2006) in the NHANES database were used as the research data set. Plasma Hcy levels are considered an independent variable, while CVD is a dependent variable. Weighted logistic regression, linear trend analysis, subgroup analysis and limiting cubic spline plots were used for analysis. A total of 4,418 samples were included.
Results: In the weighted logistic regression model, a significant positive correlation between Hcy level and CVD risk was observed (P for trend = 0.007).The subgroup analysis revealed that various characteristics such as age, race, education level, obesity, alcohol use, diabetes, and hypertension did not affect this positive correlation (P for interaction ≥0.05). The nonlinear association between Hcy level and CVD risk was explored by limiting cubic spline plots, revealing the overall significant trend (P for overall <0.0001) and the significant nonlinear trend (P for nonlinear <0.01).
Conclusion: In this large cross-sectional study, an increase in Hcy levels leads to an increased risk of CVD. There is a nonlinearly positive correlation between Hcy levels and the risk of CVD.
Keywords: National Health and Examination Survey; cardiovascular disease; homocysteine; linear trend analysis; logistic regression.
© 2025 Huang, Zhang, Zhang, Zhang, Huang and Xu.