Clinical Feasibility Of BMD Correction By Body Weight, Body Mass Index And Body Surface Area For Osteoporosis Diagnosis

ObjectiveAs we know, World Health Organization(WHO) used bone mineral density(BMD) of lumbar spine and hip measured by dual X-ray absorptiometry (DXA) as a "golden standard" of osteoporosis(OP) diagnosis , but the shortcomings is that it provided area BMD(units, g/cm2) rather than the volume BMD (units, g/cm3). Area BMD is proportional to the bone size, ignoring the bone thickness, So using area BMD instead of volume BMD will inevitably lead to OP misdiagnosis of thin person and missed diagnosis of fat person, but how it affects T value accurately has not been reported. Body weight is the third most important factor which influences BMD following age and sex. The study was aimed at assessing the statistics change of T value after BMD corrected by body weight and it's derived variables body mass index(BMI),body surface area ,then discussing the feasibility and follow up.MethodsSelect 3072 subjects who were received DXA examination in our department measure their height, weight, and BMD of L1-L4, L2-L4, left neck, right neck ,left hip, right hip and average total hip. BMD after weight correction = BMD / body weight, Body mass index, body surface area were the similar way. then calculate before and after correction T value of all parts according to T = (BMD-peak BMD) / SD. To see whether the T value change is statistically significant on all parts after three methods correction by Paired T test ,ANOVA and correlation analysis. Compare the OP diagnosis results in men more than 50 years old and postmenopausal women group using T≤-2.0 as standard analyze , then follow-up. Results1. Only BMI correction T value has significant differences in all parts of men and women samples after paired T test , what's more, T value declined a little in all parts. Maybe owing to that high BMI people has a more higher rate in our sample. weight and body surface area corrections were not statistically significant in all parts, moreover, men and women samples are not in the same situation.2. Correlation coefficients of BMC and BMD of men and women all increased after BMI correction. But when we corrected BMC and BMD using weight, correlation coefficient only increased on lumbar spine and bilateral femoral neck of men sample. What's more, correlation coefficients decreased in all parts of men and women after body surface area correction..3. From the single factor analysis results , we can see all partsΔT after BMI correction were statistically significant with the changes of BMI.ΔT change trend of men and women were almost the same in various parts. The curve rise at low BMI group, and go down at high BMI group.ΔT equals 0 of all men and women samples when BMI was 22.0.The slope of men' sΔT change curve is higher than women's.4. After BMI correction, there was 21.2 percents people maybe missed diagnosis and 15.6 percents people maybe misdiagnosed for those inconsistent diagnosis cases. We found that those newly diagnosed OP persons all have varying degrees of osteoporosis symptoms from the follow up results and most laboratory tests have positive indicators if they did. Besides ,we hadn't found who was osteoporotic fracture in those maybe misdiagnosed 267 low BMI persons.ConclusionBMD correction by BMI is feasible in clinical since it could improve OP diagnostic accuracy in a certain extent for it's increasing OP detection rate in fat groups and decreasing OP misdiagnosis rate in thin groups .