Background: It is still unclear whether the strength of the association between elevated plasma homocysteine (HC) levels and peripheral arterial disease (PAD), coronary artery disease (CAD) and cerebrovascular disease (CVD) is similar.
Methods: Fasting homocysteine plasma levels were measured in 6880 unselected primary care patients aged 65 years or older. Presence of PAD was determined with the ankle brachial index, and both CAD and CVD were recorded according to patient history.
Results: Median homocysteine levels in the total sample (58.0% females, mean age 72.5 years, mean body mass index 27.3 kg m-2) differed between patients with and without PAD: 15.2 micro mol L-1 (95% confidence interval [CI] 14.8; 15.7, vs. 13.9 micro mol L-1 (CI: 13.8; 14.1; P < 0.001). Peripheral arterial disease prevalence moderately increased from 13.0% in the lowest HC quintile to 24.3% in the highest quintile (crude odds ratio [OR], 2.1 [CI: 1.7; 2.6]). The frequency of atherothrombotic manifestations in the patients' history increased nearly linearly across the homocysteine quintiles in the univariate analysis. However, the association diminished substantially after adjusting for known interfering variables: the ORs between the HC highest fifth vs. lowest fifth (adjusted for age, gender, smoking status, diabetes, hypertension lipid disorders, and estimated glomerular filtration rate levels) for PAD decreased to 1.4, for CAD to 1.0 (NS), and for CVD to 1.1. (NS).
Conclusions: Elevated HC is only slightly more related to PAD than to CAD and CVD. After adjustment for known risk factors, the effect size is small, and an association can no longer be observed between homocysteine and CAD and CVD.