Bone Mineral Density and Stroke Risk
To the Editor:
In their recent article, Jorgensen et al1 examined the relationship between bone mineral density (BMD) and stroke. They reported that female, but not male, stroke patients had lower BMD than population controls. Like the few other previous reports, it was based on fewer than 85 total cases. We attempted to replicate their findings by analyzing data from a population-based study. We looked at stroke prevalence and BMD, using data from the cross-sectional Third National Health and Nutrition Examination Survey (NHANES III).2 NHANES III collected data from a nationally representative sample of the civilian noninstitutionalized US population from 1988 to 1994. Analyses were limited to 6298 non-Hispanic white, non-Hispanic black, and Mexican-American men and women whose BMD levels were measured and who were aged 45 years and older at the time of the NHANES III examination, because relatively few strokes occur at younger ages. BMD was measured by trained examiners in mobile examination centers. Total proximal femoral BMD was measured by dual-energy x-ray absorptiometry (Hologic QDR-1000; Hologic, Inc).3 Stroke prevalence was based on self-reported doctors’ diagnoses. A total of 323 stroke cases were identified. Baseline age, race-ethnicity, smoking status, alcohol consumption, and physical activity level were obtained by interview, and body mass index was calculated from measured height and weight. History of heart attack, congestive heart failure, and diabetes were based on self-reported doctors’ diagnoses. Hypertension was determined from blood pressure measurement at examination (systolic ≥140 or diastolic ≥90) or history of recent blood pressure medication.
No statistically significant differences in age-adjusted mean BMD for women with reported stroke or no stroke were found (0.809 g/cm2 vs 0.802 g/cm2, P=0.596). Results were similar for men (0.935 g/cm2 vs 0.947 g/cm2, P=0.358). Weighted multivariate logistic regression analyses with SUDAAN were performed. The number of prevalent cases of stroke by gender and BMD quartile are shown in the Table. In women, odds ratios for stroke were elevated in the first 3 quartiles after adjusting for age and race-ethnicity and multiple stroke risk factors, but confidence intervals were wide. Compared with women in the fourth BMD quartile (reference group), stroke risk was highest in the third BMD quartile, with a multivariate-adjusted OR of 2.59 (95% CI 1.00 to 6.73). The interaction of age and BMD was significant for women (P=0.031). Age-specific analyses suggested an association mainly in women aged 45 to 64 years, but CIs were wide. In men, no association of BMD with stroke could be demonstrated in the multivariate model. For example, compared with men in the highest quartile, men in the lowest quartile had a multivariate-adjusted OR of 0.91 (95% CI 0.34 to 2.44). In a test for linear trend, BMD was not significantly related to stroke in women or men. When treated as a continuous variable in the multivariate analyses, BMD was not associated with stroke in women (P=0.848) or men (P=0.819). In summary, in a large national study, no significant association of BMD and stroke prevalence was found in men. A trend toward elevated risk in women with BMD <0.936 g/cm2 with no dose-response relationship was found. Additional large-scale studies are needed to determine more precisely the nature (threshold vs dose response vs other) and magnitude of the BMD-stroke association, especially in women.
Jorgensen L, Engstad T, Jacobsen B. Bone mineral density in acute stroke patients: low bone mineral density may predict first stroke in women. Stroke. 2001; 32: 47–51.
Plan and operation of the Third National Health and Nutrition Examination Survey, 1988-94. Vital Health Stat 1. 1994; 32: 1–407.
Mussolino and coworkers present some interesting results from NHANES III as a comment to our recent article in Stroke.1 Similar to our findings, no association was found in men. In women, there were indications of a possible relationship between low bone mineral density and stroke risk, but no dose-response relationship was found in the analysis of NHANES data.
The 2 studies differ, however, in several ways. The results from NHANES III are based on a much higher number of stroke cases (323 prevalent cases) than those in our study (63 incident cases). Because our cases were incident cases, there has not been any poststroke reduction in bone mineral density, whereas it is likely that many of the stroke patients included in the analysis from NHANES have experienced a reduction in bone mineral density due to the stroke, particularly in the paretic leg.2 Only one of the stroke patients that met the inclusion criteria in our study declined to participate, whereas the stroke patients examined in NHANES may be a more self-selected group of stroke patients.
In our study, all stroke cases were admitted to a hospital for their first stroke, whereas in the NHANES study, the strokes were self-reported doctors’ diagnoses. The mean age of the stroke patients in the 2 studies may also differ, and the construction of the control (no-stroke) group is different. Finally, we measured the bone mineral density of the femoral neck at both sides, whereas Mussolino and coworkers measured the total proximal femural bone mineral density. We therefore fully agree with the conclusion of Mussolino and coworkers: There is a need for large-scale studies to determine the relationship between bone mineral density and stroke risk.