Cross-sectional versus longitudinal associations of lean and fat mass with pQCT bone outcomes in children.

Abstract

Cross-sectional associations for lean mass (LM) and fat mass (FM) with bone may not reflect longitudinal associations. Cross-sectional and longitudinal associations of LM and FM with radial bone measurements in children were compared. We conducted a longitudinal study on 370 (232 females) children, 8-18 yr of age. LM and FM were measured by dual-energy absorptiometry. Peripheral quantitative computed tomography at the 4% radius (4R) and 20% radius (20R) measured bone mineral content (BMC), volumetric bone mineral density (vBMD), area, and strength [polar stress strain index (pSSI)]. Males at 20R had negative FM cross-sectional and longitudinal associations with cortical area and BMC and pSSI (P < 0.02); negative cross-sectional association with total area (P < 0.001); and negative longitudinal association with cortical thickness (P < 0.001). Females at 20R had FM cross-sectional association with total area, cortical BMC, and pSSI and longitudinal associations with cortical BMC and area, vBMD, and pSSI that went from positive to negative with age and, in some cases, varied with menarche. Both sexes at 4R had a negative FM cross-sectional association with BMC and area (P < 0.001) but negative longitudinal association with vBMD (P < 0.05). LM associations with bone outcomes were generally positive, except for negative longitudinal associations with cortical BMC and vBMD in young females (P < 0.01). LM associations were greater magnitude than FM associations and often depended on age. For males and older females, cross-sectional associations indicated a reduced bone size with higher FM, whereas longitudinal associations showed a decrease in cortical area without changes in bone size. LM was positively associated with BMC and area.