TY - JOUR
T1 - Obesity Increases Precision Errors in Dual-Energy X-Ray Absorptiometry Measurements
AU - Knapp, Karen M.
AU - Welsman, Joanne R.
AU - Hopkins, Susan J.
AU - Fogelman, Ignac
AU - Blake, Glen M.
PY - 2012/7
Y1 - 2012/7
N2 - The precision errors of dual-energy X-ray absorptiometry (DXA) measurements are important for monitoring osteoporosis. This study investigated the effect of body mass index (BMI) on precision errors for lumbar spine (LS), femoral neck (NOF), total hip (TH), and total body (TB) bone mineral density using the GE Lunar Prodigy. One hundred two women with BMIs ranging from 18.5 to 45.9 kg/m(2) were recruited. Participants had duplicate DXA scans of the LS, left hip, and TB with repositioning between scans. Participants were divided into 3 groups based on their BMI and the percentage coefficient of variation (%CV) calculated for each group. The %CVs for the normal (<25 kg/m(2)) (n = 48), overweight (25-30 kg/m(2)) (n = 26), and obese ( > 30 kg/m(2)) (n = 28) BMI groups, respectively, were LS BMD: 0.99%, 1.30%, and 1.68%; NOF BMD: 1.32%, 1.37%, and 2.00%; TH BMD: 0.85%, 0.88%, and 1.06%; TB BMD: 0.66%, 0.73%, and 0.91%. Statistically significant differences in precision error between the normal and obese groups were found for LS (p = 0.0006), NOF (p = 0.005), and TB BMD (p = 0.025). These results suggest that serial measurements in obese subjects should be treated with caution because the least significant change may be larger than anticipated.
AB - The precision errors of dual-energy X-ray absorptiometry (DXA) measurements are important for monitoring osteoporosis. This study investigated the effect of body mass index (BMI) on precision errors for lumbar spine (LS), femoral neck (NOF), total hip (TH), and total body (TB) bone mineral density using the GE Lunar Prodigy. One hundred two women with BMIs ranging from 18.5 to 45.9 kg/m(2) were recruited. Participants had duplicate DXA scans of the LS, left hip, and TB with repositioning between scans. Participants were divided into 3 groups based on their BMI and the percentage coefficient of variation (%CV) calculated for each group. The %CVs for the normal (<25 kg/m(2)) (n = 48), overweight (25-30 kg/m(2)) (n = 26), and obese ( > 30 kg/m(2)) (n = 28) BMI groups, respectively, were LS BMD: 0.99%, 1.30%, and 1.68%; NOF BMD: 1.32%, 1.37%, and 2.00%; TH BMD: 0.85%, 0.88%, and 1.06%; TB BMD: 0.66%, 0.73%, and 0.91%. Statistically significant differences in precision error between the normal and obese groups were found for LS (p = 0.0006), NOF (p = 0.005), and TB BMD (p = 0.025). These results suggest that serial measurements in obese subjects should be treated with caution because the least significant change may be larger than anticipated.
U2 - 10.1016/j.jocd.2012.01.002
DO - 10.1016/j.jocd.2012.01.002
M3 - Article
SN - 1094-6950
VL - 15
SP - 315
EP - 319
JO - JOURNAL OF CLINICAL DENSITOMETRY
JF - JOURNAL OF CLINICAL DENSITOMETRY
IS - 3
ER -
