How Well Do Measures of Bone Mineral Density Predict Occurrence of Osteoporotic Fractures

A systematic review 1 including 11 prospective cohort studies (25 articles) was abstracted in DARE. Measurement at the spine had a better predictive ability for a decrease of 1 standard deviation in bone density for spine fractures, with a relative risk of 2.3 (95% CI 1.9 to 2.8). Measurement at the hip was better for predicting hip fractures, with a relative risk of 2.6 (95% CI 2.0 to 3.5). All other measurements had similar predictive ability. The sensitivity, specificity, positive predictive value, and population attributable risk for a cut poin in bone density of one standard deviation below the age adjusted mean were calculated for 3 different lifetime incidences of hip fracture: 3%, 15%, and 30% (the relative risk of hip fracture was assumed to be 2.6). The sensitivity and population attributable risk decreased as incidence increased, and the positive predictive value was much larger for higher incidence (9%, 36% and 58% for an incidence of 3%, 15% and 30% respectively).

The authors comment that the predictive ability of a one standard deviation decrease in bone mass was similar to that of a one standard deviation increase in blood pressure for stroke and better than a one standard deviation increase in serum cholesterol concentration for cardiovascular disease. The authors conclude that measurement of bone mineral density cannot identify individuals who will have a fracture, and they do not recommend a programme of screening menopausal women for osteoporosis by measuring bone density.

Another systematic review 2 including 12 prospective cohort studies (number or participants not stated) and 8 case-control studies (1 133 participants) was abstracted in DARE.

Analysis of prospective cohort studies showed that the relative risk for fracture at one standard deviation lower bone density was in the range of 1.4 to 2.6, depending on fracture type and the site of measurement. For hip fractures, measurement of hip bone density yielded a relative risk of 2.6 (95% CI 2.0 to 3.5).

Results from the case-control studies were largely similar. The odds ratios for hip fracture at 1 SD deviation lower bone density at measurement site were as follows: femoral neck, women 2.68, men 1.94: trochanter, women 2.79, men 2.70; Ward's triangle, women 2.10, men 2.20; lumbar spine, women 1.81, men 2.00.

Comment: The quality of evidence is upgraded by large magnitude of effect.

References

• Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996 May 18;312(7041):1254-9. [PubMed] [DARE]
• Ringertz H, Marshall D, Johansson C, Johnell O, Kullenberg RJ, Ljunhall S, Saaf M, Wedel H, Hallerby N, Jonsson E, Marke LA, Werko L. Bone density measurement: a systematic review. J Int Med 1997;241(suppl 739):1-60. [DARE]