Waist circumference as measure of abdominal fat compartments

Abstract

This study examines intercorrelations among waist circumference (WC), intraperitoneal fat (IPF), and subcutaneous abdominal fat (SAF) in ethnically diverse Dallas Heart Study consisting of 1538 women and 1212 men (50% Black). Correlations between fat depots and triglyceride or HOMA2-IR, biomarkers of metabolic syndrome, are also reported. Total abdominal fat (TAF), ASF, and IPF masses were measured by magnetic resonance imaging. The highest correlations with WC according to ethnicity and gender were noted for TAF (R (2) = 0.81 - 0.88) with progressively lower correlations with ASF (0.65-0.82) and IPF (0.29-0.85). The percentage of IPF relative to TAF was not significantly correlated with WC. For all WC categories, higher IPF/ASF ratios were associated with higher triglyceride levels. In contrast, differences in ratios had little or no association with HOMA2-IR. However, when all data were pooled, IPF was positively correlated with both triglyceride (r = 0.358 (men) and 0.363 (women)) and HOMA2-IR (r = 0.480 (men) and 0.517 (women)); after adjustment for ASF, IPF was still correlated with triglyceride (r = 0.353 (men) and 0.348 (women)) and HOMA2-IR (r = 0.290 (men) and 0.221 (women)). WC measures TAF reliably, but its association with IPF depends on IPF/ASF ratios that vary by gender and ethnicity.

Figure 1

Pearson's correlation coefficients (r ²) by ethnicity and gender for linear regression analyses of waist circumference versus total abdominal fat (TAF), subcutaneous abdominal fat (SAF), and intraperitoneal fat (IPF). Highest correlations were found for TAF, intermediate for SAF, and lowest for IPF. All correlations were significant at P < 0.001.

Figure 2

(a) Masses (kg) of intraperitoneal fat (IPF) for each waist circumference category for ethnicity and gender. Amounts of IPF increased for low, intermediate (MED), and high waist circumference categories. ^aSignificantly different from whites and Hispanics (P < 0.02); ^bSignificantly different from Hispanics (P < 0.02). (b) Masses (kg) of abdominal subcutaneous fat (ASF) for each waist circumference category for ethnicity and gender. Amounts of ASF increased for low, intermediate (MED), and high waist circumference categories. ^aSignificantly different from whites and Hispanics (P < 0.02); ^bsignificantly different from Hispanics (P < 0.02).

Figure 3

Percentage of intraperitoneal fat (IPF) or total abdominal fat (TAF) for each waist circumference category for ethnicity and gender. Blacks generally had a lower percentage IPF compared to whites and Hispanics. ^aSignificantly different from whites and Hispanics (P < 0.02); ^bsignificantly different from Hispanics (P < 0.02).

Figure 4

(a) Masses (kg) of intraperitoneal fat (IPF) plotted against masses of quintiles of total abdominal fat for all men. IPF masses increased progressively with TAF, and the distribution of IPF for each category widened. Boxes show mean and one standard deviation; whiskers show 2 standard deviations. The chart gives median values for each quintile, the mean waist girth, and ratio of IPF to abdominal subcutaneous fat. The latter ratio changed a little across ratios except in the highest quintile. (b) Masses (kg) of intraperitoneal fat (IPF) plotted against masses of quintiles of total abdominal fat for all women. IPF masses increased progressively with TAF, and the distribution of IPF for each category widened. Boxes show mean and one standard deviation; whiskers show 2 standard deviations. The chart gives median values for each quintile, the mean waist girth, and ratio of IPF to abdominal subcutaneous fat. The latter ratio changed a little across ratios except in the highest quintile.

Figure 5

(a) Plasma triglyceride (TG) and HOMA2-IR for upper and lower halves of the intraperitoneal fat/abdominal subcutaneous fat ratios for all men. For TG, those with the high ratios had significantly higher triglyceride in each waist circumference category (^a P < 0.05). For HOMA2-IR, there were no differences between higher and lower ratios. (b) Plasma triglyceride (TG) and HOMA2-IR for upper and lower halves of the intraperitoneal fat/abdominal subcutaneous fat ratios for all women. For TG, those with the high ratios had significantly higher triglyceride in each waist circumference category (^a P < 0.05). For HOMA2-IR, there were no differences between higher and lower ratios.

Figure 6

Black bars show Spearman correlation coefficients (r) for regression analyses of ln triglyceride and ln HOMA2-IR versus abdominal subcutaneous fat (ASF) and intraperitoneal fat (IPF) in women and men. White bars show partial correlation coefficients after adjustment for the opposite fat compartment. For this analysis, all men and all women were combined. For triglycerides, ASF was modestly correlated but became negatively correlated after adjustment for IPF. IPF was more strongly correlated with triglycerides and remained equally correlated after adjustment for ASF. Both ASF and IPF were strongly correlated with ln HOMA2-IR, but after adjustment for the opposing compartment, partial correlation coefficient was diminished. All correlation coefficients were significant at a level of P < 0.001.