Clinical Question: Is there an association between homocysteine levels and the risk of fracture?

Setting: Population-based

Study Design: Cohort (prospective)

Synopsis: The authors used data from three Dutch cohort studies of 2,406 patients 55 years or older who were followed for an average of 2.7 years in the largest study and an average of 5.7 and 8.1 years in the other two studies. Any osteoporotic fracture was recorded by the patient’s primary care physician or ascertained at a follow-up visit. Bone mineral density and homocysteine levels were measured using standard methods, apparently at the beginning of each cohort study. It is unclear whether patients with a history of osteoporosis or osteoporotic fracture were excluded, which usually is done in a cohort study. The percentage of patients lost to follow-up also was not provided, although given the stability of these populations it is likely that few were lost.

The primary outcome was the relative risk of osteoporotic fracture adjusted for age, sex, body mass index, smoking, a history of falls, and serum creatinine level. The authors adjusted the data to prevent undue influence of overly high homocysteine levels (log transformation), which is reasonable with this type of data. They found a direct relationship between the relative risk of fracture and increased homocysteine levels. They also found an adjusted relative risk of fracture of 2.0 when comparing the highest quartile of homocysteine levels with the rest of the patients. This association is similar to that of other known risk factors, such as age, bone mineral density, smoking, recent fall, and cognitive impairment.

Bottom Line: There appears to be a moderate, independent association between homocysteine levels and the risk of osteoporotic fracture in older adults. It is unclear whether modifying homocysteine levels reduces the risk of fracture. Patients who are particularly concerned about osteoporosis might add a folate supplement after considering the balance among cost, benefit, and inconvenience. (Level of Evidence: 2b)