Mi. Goran et al., Influence of total vs. visceral fat on insulin action and secretion in African American and white children, OBES RES, 9(8), 2001, pp. 423-431
Objective: To examine whether total body fat (FAT) in general or visceral f
at (VFAT) in particular is associated with greater metabolic risk in white
and African American children.
Research Methods and Procedures: A total of 68 white and 51 African America
n children had measures of insulin sensitivity (Si) and acute insulin respo
nse (AIR) by a frequently sampled intravenous glucose tolerance test, total
body fat by DXA and abdominal fat distribution (visceral vs. subcutaneous)
by computed tomography. The influence of FAT and VFAT on insulin parameter
s were examined by comparing subgroups of children with high or low FAT vs.
high or low VFAT and by multiple regression analysis.
Results: In whites, fasting insulin, Si, and AIR were significantly influen
ced by FAT, but not VFAT (e.g., for Si, 9.8 +/-0.8 in low FAT vs. 4.6 +/-0.
7 x 10(-4)/min/[mu IU/mL] in high FAT, p < 0.05; 6.8<plus/minus>0.7 in low
VFAT vs. 7.5 +/-0.8 x 10(-4)/min/[mu IU/mL] in high VFAT, p>0.1). In Africa
n Americans, fasting insulin and Si were also primarily influenced by FAT (
e.g., for Si, 4.9 +/-0.4 in low FAT vs. 2.8 +/-0.5 x 10(-4)/min/[mu IU/mL]
in high FAT, p < 0.05) but not by VFAT, and there were no significant effec
ts of either fat compartment on AIR. In multiple regression analysis, Si wa
s significantly influenced by FAT (negative effect), ethnicity (lower in Af
rican Americans), and gender (lower in females), whereas fasting insulin wa
s significantly influenced by VFAT (positive effect), ethnicity (higher in
African Americans), and fat free mass (positive effect).
Discussion: Body fat in general is the predominant factor influencing Si, b
ut VFAT may have additional effects on fasting insulin. The lack of major e
ffects of VFAT on Si in children may be explained by lower levels of VFAT o
r because VFAT affects aspects of whole body insulin action that are not me
asured by the minimal model.