Screening for osteoporosis using easily obtainable biometrical data: diagnostic accuracy of measured, self-reported and recalled BMI, and related costs of bone mineral density measurements.

Abstract

The aims of the present study were: to determine the diagnostic accuracy of objectively measured, self-reported and recalled body mass index (BMI) for osteoporosis and osteopenia; to determine the diagnostic costs, in terms of bone mineral density (BMD) measurements, per osteoporotic or osteopenic patient detected, using different BMI tests; and to determine the extent to which the results can be used within the framework of the current screening program for breast cancer in The Netherlands. Within the framework of a cross-sectional study on the prevalence of osteoporosis in the south of The Netherlands, 1155 postmenopausal women aged 50-80 years were asked for their present height and their weight at age 20-30 years. Subsequently their actual weight, height and BMD of the lumbar spine (DXA) were measured. The BMD cutoff was 0.800 g/cm2 for osteoporosis and 0.970 g/cm2 for low BMD (osteoporosis + osteopenia). After receiver operating characteristic analysis, age was cut off at 60 years and BMI at 27 kg/m2. Diagnostic accuracies of objectively measured, self-reported and recalled BMI were evaluated using predictive values (PV) and odds ratios. The resulting 'true positive' and 'false positive' rates were used to calculate diagnostic costs (i.e., DXA) for each osteoporotic patient or low-BMD patient detected. The prevalence of osteoporosis in the study population was 25%, that of low BMD 65%. Only the age-BMI tests 'age > or = 60, BMI < or = 27' showed PVs for osteoporosis (31-41%) and for low BMD (71-81%) that were higher than the prior probabilities for these conditions. Related odds ratios were 2.14-3.18 (osteoporosis) and 1.87-3.04 (low BMD). The objective BMI test detected 50% of the osteoporotic patients. Using the self-reported BMI test and the recalled BMI test, detection rates increased to 55% and 69%, respectively. Concomitant costs per osteoporotic patient detected rose by 24%. Detection of patients with a low BMD increased from 38% for objective BMI and 42% for self-reported BMI to 60% for recalled BMI. Related costs increased by 11%. If all women over 50 years of age (irrespective of their BMI) were to be referred for BMD measurement, costs per osteoporotic patient or low-BMD patient detected would be 304 and 116 Euros, respectively. Only in women over 60 years does a BMI below 27 kg/m2 provide a better prediction of the presence of osteoporosis or low BMD than could be expected solely on the basis of the relevant prevalences in postmenopausal women aged 50-80 years. If the use of BMI for the detection of osteoporotic or low-BMD patients is still considered, measuring weight and just asking for a person's height will do. Although age and BMI are the strongest risk factors for osteoporosis, they are of less significance when used for screening the population for osteoporosis. More research is needed before age and BMI can be included in any screening program. As regards practical considerations alone, measurements of BMD could be implemented within the screening program for breast cancer.