BOSTON — The more women weigh, the greater their risk for incident pulmonary embolism, according to an analysis of prospective data from more than 85,000 women enrolled in the Nurses' Health Study.
The investigators found a relative risk for pulmonary embolism (PE) of 1.08 for every 1 kg/m
“We found that there is a strong, independent, positive, linear association between BMI and incident PE, and that this effect seems to impact not only obese subjects, but [also] subjects with relatively modest increases in their BMI,” reported Dr. Kabrhel, of Harvard Medical School and Massachusetts General Hospital in Boston, and his colleagues.
Cross-sectional and case-control studies have shown that patients who experience deep vein thrombosis and pulmonary embolism tend to have higher BMIs, and prospective studies have shown an association between severe overweight or obesity and pulmonary embolism, he noted.
The investigators examined the association between weight and thromboembolic events using data from 87,226 women enrolled in the prospective, longitudinal Nurses' Health Study, which has collected data on PE since its inception in 1976 and on diet, physical activity, and other risk factors for PE since 1984.
Participants enrolled in the Nurses' Health Study during 1984-2002, and were excluded from the current analysis if they had a PE diagnosis before 1984 or if their records were missing data necessary to calculate BMI.
The investigators divided participants into six BMI categories: less than 22.5 kg/m
The primary outcome was idiopathic PE, defined as cases of PE that were confirmed in the medical record and not associated with prior surgery, trauma, or malignancy. The authors also performed a secondary analysis of nonidiopathic PE. They used a multivariate Cox proportional hazards model to control for age, physical activity, caloric intake, smoking and pack-years, race, spouse's educational attainment, parity, menopausal status, NSAID use, warfarin use, multivitamin supplement use, hypertension, coronary heart disease, and rheumatologic disease.
During the period studied, there were 157 incident cases of idiopathic PE and 338 cases of nonidiopathic PE, and these correlated strongly with BMI. For both idiopathic and nonidiopathic PE, the relative risk for every 1 kg/m
Associations between BMI and nonidiopathic PE were similar, ranging from a relative risk of 1.48 for patients in the 22.5-24.9 range compared with those in the lowest BMI category, to a relative risk of 5.42 for patients in the highest vs. lowest BMI categories.
“There is a significant increase with the combined idiopathic PE and nonidiopathic PE. In other words, for our total PE, there is a significant increase in the risk of PE even with relatively modest increases in BMI—that is to say, subjects that would not be considered either overweight or obese, but within the normal range,” Dr. Kabrhel said.
In secondary analyses adjusted for all of the variables, neither waist-to-hip ratio nor weight change since age 18 were significantly associated with risk for PE.
A potential mechanism for the association between BMI and PE is the regulatory hormone leptin, which has been shown to induce tissue-factor activity in vitro and to be elevated in obese individuals, he said.
Alternatively, estrogen and progesterone, which have been linked to obesity and the risk of PE in women, may play a role, although there was no evidence of a hormone-PE interaction in their study, he said.
Dr. Kabrhel acknowledged that the study was limited by its inclusion of only women, and by the racial and ethnic imbalance of the Nurses' Health Study cohort, which represents a demographic sample of nurses in the United States. The study may also be subject to measurement bias because it relied on subject-reported weights.
The authors received grant and research support for the study from the National Institutes of Health. No other conflicts of interest were reported.