s 491 sex, and body mass index (BMI). We previously reported total body values and trends with respect to age, BMI and time. However our objective was to further investigate the repeatability of both total body and subregional body composition measurements and their covariates in a large prospective self-selected sample. Methods: We recruited 651 participants of the National Health and Nutrition Examination Survey (NHANES) to receive two DXA scans. Participants varied in ethnicity and gender, with age ranging from 16 to 69 years (mean, 39.2 17.5), and BMI from 14.1 to 43.5 (mean, 26.5 5.0). Time between scans ranged from 1 to 30 days (mean, 15.1 8.8). Participants with avoidable scan artifacts were excluded. Precision estimates for whole body, and subregional measures, including android, gynoid, trunk, arm, leg and appendicular (arm + leg), were calculated as root mean square coefficients of variation (RMS-%CV) and standard deviations (RMS-SD). Pearson correlation coefficients and stepwise regression analysis was used to determine the influence of participant descriptors (Sex, BMI, ?BMI, time, age, BMD, BMC, fat, lean mass, total mass, %fat) on the precision of BMD, BMC, fat mass, lean mass, total mass, and %fat for each region. Results: The subregional precision values were worse than the total body values (CV and RMSE, 1.12 and 0.01g/cm2, 1.17 and 27.5g, 1.83 and 471g, 1.39 and 710g, 1.04 and 796g, and 1.62 and 0.53 for BMD, BMC, fat mass, lean mass, total mass, and %fat, respectively). The precision of the subregions vary with Android region having the worst overall (CV and RMSE, 3.44 and 73.7g, 3.02 and 108g, 2.97 and 1.06 for fat mass, lean mass and %fat, respectively). The precision of the android-gynoid percent fat ratio (CV and RMSE, 2.98 and 0.03) was also worse than the total body percent fat value, but similar to the trunk-leg percent fat ratio (CV and RMSE, 3.08 and 0.03). Fat mass was consistently measured with the worst precision in each region, with the arms having the highest value of 4.05 and 126g (CV and RMSE). Time between scans was not associated with any of the precision measures. Total body precision values were only associated with BMI, ?BMI, or sex, but subregional measures were associated with almost all other covariates. Conclusions: We conclude that long-term precision values for whole body DXA scan measures are better than subregional precision values and confirm that precision values are unique my type of measure and region of interest. 119 Other Density and Morphometric Measurements: Body Composition, Vertebral Fracture Analysis, Other RELATIONSHIP BETWEEN TOTAL BODY BONEMINERAL CONTENT AND TOTAL BODY, APPENDICULAR OR ABDOMINAL LEAN OR FAT MASS Jingmei Wang, Norlandda CooperSurgical Company Kathy Dudzek, Norlandda CooperSurgical Company; Chad Dudzek, Norlandda CooperSurgical Company; Tom Sanchez, Norlandda CooperSurgical Company A strong and positive relationship has been noted between total body lean mass and total body bone mineral in both male and female subjects by DXA. The same studies have demonstrated a significant relationship between total body fat mass and total body bone mineral in female subjects but not in male subjects. If the relationship between bone and lean relies on skeletal muscle a stronger relationship might be seen with appendicular lean mass. The present study examines if those relationships are strengthened or weakened when regional (appendicular or abdominal) lean or fat mass are examined. A population of 100 children between 7 and 19 years old (50 boys and 50 girls) underwent whole body studies to assess bone, lean and fat mass using a Norland XR-46 fitted with Illuminatus software. Regional results for lean and fat were obtained for total body, appendicular (left leg + right leg + left arm + right arm) and abdominal (midriff + pelvis) regions. A strong linear relationship was seen between appendicular lean mass and total body bone mass in the boys (y 5 0.1093x + 437.5, r 5 0.9774, RMSE 5 172.0) and girls (y 5 0.1133x + 500.2, r 5 0.9475, RMSE 5 202.1). Similar relationships were seen between abdominal lean mass and total body bone mass in the boys (y 5 0.2722x + 506.3, r 5 0.9656, RMSE 5 211.5) and girls (y 5 0.2672x + 579.6, r 5 0.9420, RMSE 5 212.1). As expected from earlier studies, a weaker relationship was seen between appendicular fat mass and total body bone mass in boys (y5 0.0278x + 2280.9, r5-0.1169, RMSE5 808.0) with a relatively better relationship being seen in girls (y 5 0.0735x + 1207.9, r 5 0.5026, RMSE 5 546.4). Similar findings were seen between abdominal fat mass and total body bone mass in boys (y5 -0.0243x + 2122.3, r5 -0.0517, RMSE5 812.5) and girls (y 5 0.1253x + 1429.6, r 5 0.4580, RMSE 5 561.8). Those results compare to a strong linear relationship between total body lean mass and total body bone in boys (y 5 0.0531x + 336.1, r 5 0.9795, RMSE 5 163.8) and girls (y 5 0.0560x + 371.7, r 5 0.9430, RMSE 5 210.3) and weaker relationship between total body fat and total body bone in boys (y 5 -0.0098x + 2211.8, r 5 Journal of Clinical Densitometry: Assessment of Skeletal Health -0.0869, RMSE 5 810.5) and girls (y 5 0.0322x + 1300.0, r 5 0.4649, RMSE 5 559.5). Results show very similar relationships between total body bone and total body, appendicular or abdominal region lean or fat mass suggesting that the relationship is not necessarily tied to regional lean or fat mass in boys or girls. 120 Other Density and Morphometric Measurements: Body Composition, Vertebral Fracture Analysis, Other CHANGE IN SERUM 25(OH)D FOLLOWING VITAMIN D SUPPLEMENTATION IS UNRELATED TO BODY