Associations between medication use and homocysteine levels in an older population, and potential mediation by vitamin B12 and folate: data from the B-PROOF Study.

ABSTRACT

BACKGROUND: Elevated homocysteine levels are a risk indicator for cardiovascular disease, fractures and cognitive decline. Previous studies indicated associations between homocysteine levels and medication use, including antihypertensive, lipid-lowering and antidiabetic medication. However, results were often contradictory and inconclusive. Our objective was to study the associations established previously in more detail by sub-classifying medication groups, and investigate the potential mediating role of vitamin B12 and folate status. MATERIALS AND METHODS: Baseline data from the B-PROOF (B-vitamins for the PRevention Of Osteoporotic Fractures) study were used. We included 2,912 participants aged ≥65 years, with homocysteine levels of 12-50 µmol/L and creatinine levels ≤150 µmol/L, for whom self-reported medication data were available. We used multivariable linear regression models and analysis of covariance to assess the association between medication use and plasma homocysteine levels, and the potential mediation by serum vitamin B12 and folate. RESULTS: The mean age was 74 years (standard deviation, 6.5), 50 % were women, and median homocysteine levels were 14 µmol/L [interquartile range, 13-17 µmol/L]. Higher mean homocysteine levels were observed in users vs. non-users for diuretics (15.2 vs. 14.9, p = 0.043), high-ceiling sulphonamide diuretics (16.0 vs. 14.9, p < 0.001), medication acting via the renin-angiotensin system (15.2 vs. 14.9, p = 0.029) and metformin (15.6 vs. 15.1, p = 0.006). Non-selective ß-blocker use was associated with lower mean homocysteine levels (14.4 vs. 15.0, p = 0.019). Only this association was mediated by an underlying association with vitamin B12 and folate levels. CONCLUSION: The associations between homocysteine levels and medication use appear to be fairly modest. Our results suggest that medication use is unlikely to contribute to clinically relevant changes in plasma homocysteine levels.