Generalized abdominal visceral fat prediction models for black and white adults aged 17–65 y: the HERITAGE Family Study

Abstract

OBJECTIVE: To determine if the relationship between abdominal visceral fat (AVF) and measures of adiposity are different between Black and White subjects and to develop valid field prediction models that accurately identify those individuals with AVF levels associated with high risk for chronic disease.

DESIGN: Cross-sectional measurements obtained from 91 Black men, 137 Black women, 227 White men, and 237 White women subjects, ages 17–65 y, who were participants in the HERITAGE Family Study, both at baseline and following 20 weeks of endurance training.

MEASURMENTS: AVF, abdominal subcutaneous fat (ASF), abdominal total fat (ATF), and sagittal diameter (SagD) were measured by computed tomography (CT). Body density was determined by hydrostatic weighing and was used to estimate relative body fat. Arm, waist (WC), and hip circumferences and skinfold thickness measures were taken, and BMI was calculated from weight (kg) and height (m2). Since CT abdominal fat variables were skewed, a natural log transformation (Ln) was used to produce a normal distribution. The General Linear Model (GLM) procedure was used to test the relationship between AVF and two different groups of variables—CT and anthropometric.

RESULTS: The AVF of White men and women was significantly higher than that of Black men and women, independent of BMI, WHR, WC, and age, and was greater for men than for women. The CT model showed that the combination of SagD, Ln (ASF), age, and race accounted for 84 and 75% of the variance in AVF in men and women, respectively. The anthropometric model provided two valid generalized field AVF prediction equations. The Field-I equation, which included BMI, WHR, age and race, had an r2 of 0.78 and 0.73 for men and women, respectively. The Field-II equation, which included BMI (women only), WC, age, and race, had an r2 of 0.78 and 0.72 for men and women, respectively. The field model equations became less accurate as the estimated AVF increased.

CONCLUSIONS: (1) At the same age and level of adiposity, Black men and women have less AVF than White men and women. These differences are greater in men than in women. (2) The field regression equations can be generalized to the diverse group of adults studied, both in an untrained and trained state. However, their accuracy decreases with increasing levels of AVF.

Introduction

Abdominal visceral obesity is associated with an increased risk for cardiovascular disease and metabolic and endocrine disorders.1 Computed tomography (CT)2, 3, 4, 5, 6 and magnetic resonance imaging7, 8, 9, 10 are used to measure abdominal visceral fat (AVF), but these measures are not routinely obtained due to the high cost and unavailability of the associated equipment. This has led investigators7, 8, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 to examine potential predictors of AVF from more accessible demographic, anthropometric, and body composition variables. Most of these studies have included only White subjects, but there is evidence that when controlled in some manner for body fatness, the relationship of anthropometric and body composition measures to AVF may be different for Black and White subjects. Després et al25 observed that Black men and women had less AVF than did White men and women for a given amount of total body fat mass. This is in agreement with most11, 13, 17, 21 but not all22 studies that have compared AVF levels in Black and White subjects after adjusting for some measure of adiposity. These results suggest that regression equations established with White subjects to estimate AVF might be biased when applied to Black subjects.

Further, most prediction equations developed to estimate AVF have had relatively high standard errors of estimate (s.e.e.). However, it is possible that these equations could identify those individuals who are at highest risk due to excessive AVF. Previous studies26, 27 have identified an AVF level of 130 cm2 as a critical threshold above which glucose, insulin, and lipid–lipoprotein metabolic abnormalities become more prevalent. Rankinen et al,28 in a study of 789 French Canadians, determined that waist circumference (WC) had the best combined sensitivity (probability of correctly detecting those with an AVF >130 cm2) and specificity (probability of correctly detecting those with an AVF <130 cm2). Using WC as a predictor of AVF, their sensitivity values ranged from 81 to 91% and their specificity values ranged from 75 to 90% for men and women of varying ages. If an AVF level of 130 cm2 is a critical threshold for Black as well as White subjects, then the anthropometric values used to identify individuals with an AVF level of ≥130 cm2 could be different for Black and White subjects.

Using the HERITAGE Family Study data, we examined the multivariate effect of race, age, and subcutaneous fat on AVF variation. The goals of this study were (a) to determine if the relationship between AVF and measures of adiposity are different between Black and White subjects and (b) to develop valid field prediction models that accurately estimate AVF from easily obtained measures.

Methods

The HERITAGE Family Study is a large, multicenter clinical trial primarily focused on investigating the probable genetic basis for the variability in individual responses to aerobic exercise training with respect to risk factors for cardiovascular disease and Type II diabetes, as well as for various physiological measures. The study included four Clinical Centers located at Indiana University (formerly at Arizona State University), the Pennington Biomedical Research Center (formerly Laval University), the University of Minnesota, and Texas A&M University (formerly at The University of Texas at Austin). The Data Coordinating Center is located at Washington University School of Medicine, St Louis, MO, USA. The aims, experimental design, detailed inclusion and exclusion criteria, and measurement protocols have been presented in detail in a previous publication.29 The changes in body composition and CT measured abdominal fat data have been reported;25, 30 however, regression equations to predict AVF were not developed.

Participants

The Institutional Review Board for each Clinical Center had previously approved the study protocol, and informed consent was obtained from each participant. Data from 692 men and women from the HERITAGE Family Study were used in the analyses. This sample represents only those subjects who had complete data for all variables used in the analyses for this study, and included 91 Black men, 137 Black women, 227 White men, and 237 White women. Subjects were 17–65 y of age, with resting blood pressures <160 mmHg systolic and <100 mmHg diastolic, not taking any antihypertensive or lipid lowering medication, and sedentary. Baseline (pretraining) data were used in the initial analyses and response to training (post-training) data were used to crossvalidate the models developed with baseline data. Subject characteristics for the baseline sample, contrasted by race and sex, are provided in Table 1.

Table 1 Baseline subject characteristics (mean±s.d.) contrasted by sex and race.

Measures

CT scans were used to measure abdominal total fat (ATF), abdominal subcutaneous fat (ASF), and AVF using the procedures of Sjöström et al.10 The scanning was performed at 125 kV and a slice thickness of 8 mm. The CT scans were obtained between the fourth (L4) and fifth (L5) lumbar vertebrae while subjects were supine with arms extended above the head. ATF and AVF areas were calculated by delineating the areas with a graph pen and then computing the surface areas by using an attenuation range of −190 to −30 Hounsfield units.10, 16 The AVF area was measured by drawing a line within the inner portion of the muscle wall surrounding the abdominal cavity. ASF area was calculated by subtracting AVF from ATF. Using the recommendations of Sjöström,31 sagittal diameter (SagD) was measured from the CT scan by measuring the distance from the umbilicus through the vertebrae to the back.

Skinfold thickness was measured at the subscapular, triceps, biceps, midaxillary, suprailiac, abdominal, thigh, and calf skinfold sites as described by Lohman et al32 using a Harpenden skinfold caliper (Quinton Instruments, Inc., #03496–001). Research has documented that skinfold measures are highly intercorrelated and measure a common body fat factor.33 For this reason, the skinfolds were summed to create a total skinfold fat measure (ΣSkinfolds). Waist (WC), hip (HC), and arm circumference (AC) measurements were taken as described by Lohman et al.32 The waist and hip circumferences were used to compute the waist-hip ratio (WHR). Body density was determined by hydrostatic weighing and was used to estimate relative body fat (% body fat). Refer to other HERITAGE publications for details on the anthropometric, skinfold, and hydrostatic weighing techniques.30, 34

Quality assurance and quality control

Extensive quality assurance and quality control procedures, as described by Gagnon et al,35 were followed at each Clinical Center. A random sample of 10% of the X-ray films collected at each Clinical Center was sent for review by the consortium CT scan reading center at the Quebec Clinical Center. A calibration unit, composed of lard and sealed in a special plastic cylinder, was sent to each Clinical Center on two different occasions to assess potential intercenter differences. As previously reported, the reproducibility of anthropometric and body composition measures at all Clinical Centers was very high.34

Statistical procedures

Analysis of variance (ANOVA) and covariance (ANCOVA) were used to determine gender and race differences in the independent variables.36 Product–moment correlations defined the relationship between the CT scan abdominal fat variables and the independent variables of the study. The General Linear Model (GLM)37 procedure was used to test the relationship between the dependent variable, AVF, and the independent variables that consisted of two different types, CT and ‘field’ measures. The first GLM (ie, the CT models) used the CT variables of ASF and SagD in combination with age and race. While not completely necessary, this CT model, which included a direct measure of abdominal subcutaneous fat, was developed as an additional method for evaluating the quality of the field model. The independent variables for the second GLM (ie, the field models) included the anthropometric variables in combination with age and race. The correlation results were used to select the independent variables for the second GLM. In both GLM procedures, race was dummy coded: Black=0, and White=1. A step-down analysis was used to determine if race and age, and their interaction, accounted for AVF variance while statistically controlling for the CT or anthropometric variables. Multiple regression36 was used as a post hoc test to determine if the regression weight of each independent variable differed significantly from zero.37 Multiple regression defined the generalized AVF prediction models.

The prediction models were crossvalidated by two ways. First, the PRESS procedure38, 39 was used to crossvalidate the regression models. Second, the field models developed with the baseline (pretraining) data were applied to the HERITAGE final (post-training) data. This second crossvalidation method used simple linear regression to determine if the slopes between estimated and measured AVF were within sampling variation of 1.0 and the intercepts did not differ from 0. Product–moment correlation was used to determine the relationship between estimated and measured AVF. The difference between estimated and measured AVF values was then used to calculate the crossvalidation s.e.e. of the field models.

Results

Table 1 gives the descriptive statistics for the men and women contrasted by race. As previously reported,25, 30 the men were taller, heavier, and leaner than the women. ANOVA was used to evaluate race differences for each gender. The only male race differences were that White men were nearly 5 y older and had a larger WHR. The mean fat weight for Black women was 7.5 kg higher than that of White women, which was associated with a higher total body weight, BMI, percent body fat, WC, HC, and Σskinfolds in the Black women. The Black and White women did not differ in fat-free weight and WHR.

Table 2 presents the values for the CT variables contrasted by sex and race. The s.d. of the CT abdominal fat variables were large in relation to the mean, suggesting that the CT fat variables were skewed in a positive direction. A natural log (Ln) transformation was therefore applied to each abdominal fat variable, which normalized the distributions. All subsequent analyses were then conducted on the log-transformed CT fat values.

Table 2 Baseline values for CT variables (mean±s.d.) contrasted by sex and race.a

ANOVA showed there was a significant sex by race interaction for ATF (F(1, 588)=11.37; P<0.01). The data in Table 2 show that the mean ATF for White men was higher than that for Black men, but the mean ATF for White women was lower than that for Black women. The sex by race interaction supported the strategy of analyzing the men's and women's data separately. The Black and White men did not differ in ASF, while the mean for the Black women was significantly higher than that for the White women. Both Black men and women had lower AVF values than White men and women. However, product–moment correlation coefficients showed that the log-transformed ASF and AVF variables were significantly correlated (r=0.72 and 0.80 for women and men, respectively). The correlations between ASF and AVF in their original metric were much lower (r=0.60 for both women and men). Owing to the significant correlation between ASF and AVF, ANCOVA was used to compare the racial AVF difference, using ASF as the covariate. ANCOVA showed that the mean AVF of White men (F(1, 315)=55.2; P<0.01) and White women (F(1, 371)=20.1; P<0.01) was significantly higher than that of their Black counterparts. The SagD means for Black and White men did not differ, but the mean SagD for Black women was significantly higher than that for White women.

Zero-order correlations were used to examine the relationship between Ln (AVF) and the independent variables contrasted by sex. All correlations were significantly different from zero, but the patterns differed by sex. The range in correlations between the anthropometric variables and Ln (AVF) for men ranged from a low of r=0.66 for BMI and skinfold fat to a high of r=0.79 for WC and r=0.80 for WHR. The female anthropometric correlations ranged from a low of r=0.64 for WHR to a high of r=0.76 for WC. The correlations between Ln (AVF) and age were r=0.63 for females and r=0.66 for males. The correlation between SagD and Ln (AVF) was r=0.76 for both males and females.

Table 3 gives the GLM, multiple regression and PRESS crossvalidation results for the CT variables. The GLM showed that ASF and SagD (Model CT-I) accounted for significant, independent sources of AVF variance. Adding age and race (Model CT-II) accounted for a significant 18 and 16% increase in AVF variance for men and women, respectively, beyond the CT variables of Ln (ASF) and SagD. The age by race interaction terms for women (F(1, 373)=0.58; P>0.05) and men (F(1, 317)=0.41; P>0.05) were not statistically significant. The post hoc analysis showed that all regression coefficients of CT-II models were statistically significant. The polarity of regression coefficients revealed that AVF increased independently with age and that the age effect was 0.02 Ln (AVF) units per year for both men and women. The regression coefficients in Table 3 show that with ASF, SagD, and age statistically controlled, the AVF of White men and women was higher than that of the Black men and women. The 95% confidence interval for the race coefficients showed that the male race effect was greater than the female race effect.

Table 3 CT models demonstrating the effect of Ln (ASF), SagD, age, and race on Ln (AVF) contrasted by sex.

Table 4 gives the GLM results for the field models (Field I and II). The independent variables used for Field-I were BMI and WHR. WC replaced WHR for Field-II. The accuracies of Field-I and -II for women were similar to CT-II, with an r2 of only 2–3% less than that obtained with CT-II. In contrast, the r2 of the two men's field models was 6% less than obtained with the male CT-II model. Replacing WHR with WC did not influence the accuracy of the field models, as the s.e.e. estimates were identical. There was a sex difference in Field-II, which replaced WHR with WC. The regression weights for both BMI and WC were statistically significant for the women's Field-II, while the BMI regression weight for the men's Field-II was not statistically significant. This gender difference can be traced to the high correlation (r=0.93) between BMI and WC for the sample of men with similar correlations between Ln (AVF) and WHR (r=0.80) and WC (r=0.79).

Table 4 Field models for estimating Ln (AVF) from anthropometric variables in combination with age and race.

The race regression weights in Table 4 show that the AVF values of White men and women were significantly higher than for Black men and women. In addition, the race regression weights for men were higher than for women. The 95% confidence intervals for the race effect of the CT-II models were 0.16 to 0.30 for women and 0.28 to 0.43 for men. The women's and men's race regression weights for the field measures were within these 95% confidence intervals suggesting that the race effect between the CT and Field models were within chance variation. The crossvalidation PRESS statistics for the field models were identical to the model values supporting the validity of the generalized field regression equations.

Table 5 gives the crossvalidation analysis of the field models for men and women. The equations developed from the HERITAGE baseline (pretraining) data were applied to all HERITAGE subjects who had post-training data. The crossvalidation correlations were nearly identical to the validation multiple correlations in Table 4. The 95% confidence interval showed that the crossvalidation correlations were within chance variation of the validation values. Simple linear regression analysis was used to examine the fit of the field equations applied to measured Ln (AVF). Table 5 gives the slopes and intercepts for these models and the 95% confidence intervals for both. This analysis showed that all slopes were within chance variation of 1.0 and the intercepts were within chance variation of 0. The crossvalidation s.e. of the field models were within 0.02 U of the validation s.e. (Table 4). These analyses showed that the equations developed with baseline data from untrained subjects could be generalized to trained subjects.

Table 5 Crossvalidation analysis of the field equations applied to trained HERITAGE subjects.

Discussion

The results of this study demonstrated that when age and subcutaneous fat, or other measures of adiposity, were controlled, Black men and women had significantly less AVF than White men and women. This difference was greater in men than in women. These results agree with the only other study comparing Black and White men. In a study of 196 young adult men, Hill et al17 reported that AVF, when adjusted for percent body fat, total body fat, BMI, WC WHR, or SagD, was significantly less in Blacks than in Whites. The present study results are also in agreement with those from most studies comparing Black and White women, although studies with larger sample sizes are equivocal. In studies with small sample sizes (8–25 subjects per race/sex cell), Albu et al11 determined that obese Black women had significantly less AVF than obese White women after controlling for total body mass and WHR. Conway et al13 found obese Black women to have less AVF than obese White women at L2-L3, but not at L4-L5. In a group of primarily obese women, Lovejoy et al21 found Black women to have less AVF than White women when the means were adjusted for BMI. In studies with larger sample sizes (55–105 subjects per race/sex cell), young Black women had significantly lower AVF than White women when adjusted for WC or SagD, but not when adjusted for percent body fat, total body fat, BMI, or WHR.17 In another study by Lovejoy et al22 there was a slight, nonsignificant trend for middle-aged Black women to have less AVF than middle-aged White women after adjusting for total body fat and age.

In the development of prediction models for AVF, this study demonstrated that variation in AVF was an independent function of race, age, and subcutaneous fat. The most accurate prediction models were obtained with age and race in combination with the CT scan variables of Ln (ASF) and SagD. While the field prediction models were less accurate than the CT-II models, the differences were small. Also, these differences were less between the women's models than between the men's models. The women's CT-II model accounted for 2% more Ln (AVF) variance and its s.e.e. was lower (0.31 vs 0.33) than the field models, while the men's CT-II model accounted for 6% more Ln (AVF) variance and its s.e.e. was lower (0.29 vs 0.33) than the field models.

The PRESS statistics demonstrated that the models were valid for inactive Black and White men and women who varied substantially in age. A limitation of the equations and PRESS crossvalidation results is that they were developed on the baseline data representative of untrained subjects. The crossvalidation results presented in Table 5 showed that the field equations were accurate when applied to the trained HERITAGE subjects, thus demonstrating that the derived field equations were also valid for physically active Black and White men and women.

An examination of the AVF distribution showed that both the men and women's distributions were skewed in a positive direction. A natural log transformation produced a normal distribution. This is an important finding. The skewed distribution found in these data is consistent with the descriptive statistics published by Janssen et al.18 They reported that the mean (±s.d.) for L4-L5 visceral fat area (cm2) was 77 (±59) and 137 (±83) for men and women, respectively. The reported ranges were 3–290 cm2 for men and 3–482 cm2 for women. An examination of the plot of the residuals and estimated Ln (AVF) from the results of the present study (graph not shown) suggested that the skewed distribution affected prediction errors when the residuals were expressed in the original metric (cm2).

An important advantage of the derived field equations is the variables that comprise the model are easy to obtain. A limitation of the regression model is that the predicted value is log transformed which makes it more difficult to interpret. An examination of the relationship between estimated Ln (AVF) and the residuals (ie, measured−estimated Ln (AVF)] in the original unit of measurement (cm2) showed that predictions were less accurate with individuals at the higher levels of Ln (AVF). An important clinical objective of estimating AVF is to identify individuals at risk of a metabolic abnormality (AVF value ≥130 cm2). Logistic regression and receiver operating characteristic (ROC) analyses37, 40 were used to identify an estimated Ln (AVF) value to identify risk. These analyses showed that the sensitivity of the Ln (AVF) value of 4.5 was 88 and 92% for males and females. These sensitivity estimates were associated with false positive rates (ie, 1—specificity) of 13% for females and 15% for males. These findings suggest that an estimated Ln (AVF) value of 4.5 would be suitable to define the risk of a metabolic abnormality in both men and women for both Black and White subjects.

Using the Field-II equation, with a BMI of 27 kg/m2, WC values of 115, 101, and 88 cm for 20, 40, and 60-y-old White men, and 125, 112 and 99 cm for 20, 40, and 60-y-old Black men result in a Ln (AVF) of 4.5. WC values of 108, 88, and 68 cm for 20, 40, and 60-y-old White women and 118, 98, and 78 cm for 20, 40, and 60-y-old Black women, result in a Ln (AVF) of 4.5. This illustrates that if the same relationship exists between AVF and metabolic risk variables regardless of age and race, the WC value used to identify risk varies with age and race. These WC values are higher than the values of 90 and 81 cm determined by Rankinen et al28 and 90 and 91 cm determined by Lemieux et al41 for identifying French Canadian men <40 and ≥40 y of age, respectively, with an AVF ≥130 cm2. No similar data has been published on women. These are indications that further crossvalidation research is needed in this area.

The estimated Ln (AVF) values for men and women were correlated with hydrostatically determined fat weight in an effort to gain a better understanding of the validity issue. The correlations between estimated Ln (AVF) and measured fat weight were 0.75 and 0.77 for men and women, respectively. These correlations were lower than the correlations between the sum of skinfold fat and fat weight of 0.89 for men and 0.84 for women. Multiple regression showed that both the sum of skinfold fat and Ln (AVF) were independently related to total fat weight. The standardized regression coefficients were: males, skinfold fat=0.67, Ln (AVF)=0.31; and females, skinfold fat=0.57, Ln (AVF)=0.36. This demonstrated that the field regression models provide estimates of total body fat that differed from subcutaneous fat and supports the value of measuring both AVF and subcutaneous fat.

We found no difference in the accuracies of our Field-I equations, which in addition to age and race, included BMI and WHR, and our Field-II equations, which for men included WC instead of BMI and WHR, and for women included WC instead of WHR. Many other investigations11, 12, 14, 17, 19, 20, 28, 42, 43, 44 have found WC to be more strongly correlated with and a better predictor of AVF than WHR. The difference between our study and these other investigations is that the correlation between WHR and AVF was higher in our study. We cannot explain this difference. Our data also support using WC instead of WHR in estimating AVF since we found WC to be as good a predictor as WHR and it involves taking only one measurement instead of two.

In summary, these results show that at a similar age and level of adiposity, Black men and women have less AVF than White men and women. This race difference is greater in men than women. The validity of the field regression equations can be generalized to the diverse group of adults studied who were either untrained or trained. A limitation of the field prediction models is the accuracy decreases with increasing levels of estimated AVF. These data suggest that an estimated Ln (AVF) of 4.5 provides a reasonable estimate of an increased risk for a metabolic abnormality (AVF value ≥130 cm2), but needs additional crossvalidation research.

References

  1. 1

    Bouchard C, Bray GA, Hubbard VS . Basic and clinical aspects of regional fat distribution. Am J Clin Nutr 1990; 52: 946–950.

    CAS  Article  Google Scholar 

  2. 2

    Grauer WO, Moss AA, Cann CE, Goldberg HI . Quantification of body fat distribution in the abdomen using computed tomography. Am J Clin Nutr 1984; 39: 631–637.

    CAS  Article  Google Scholar 

  3. 3

    Kvist H, Chowdhury B, Grangård U, Tylén U, Sjöström L . Total and visceral adipose-tissue volumes derived from measurements with computed tomography in adult men and women: predictive equations. Am J Clin Nutr 1988; 48: 1351–1361.

    CAS  Article  Google Scholar 

  4. 4

    Kvist H, Sjöström L, Tylén U . Adipose tissue volume determinations in women by computed tomography: technical consideration. Int J Obes 1986; 10: 53–67.

    CAS  Google Scholar 

  5. 5

    Rössner S, Bo WJ, Hiltbrandt E, Hinson W, Karstaedt N, Santago P, Sobol WT, Crouse JR . Adipose tissue determinations in cadavers—a comparison between cross-sectional planimetry and computed tomography. Int J Obes Relat Metab Disord 1990; 14: 893–902.

    Google Scholar 

  6. 6

    van der Kooy K, Seidell JC . Techniques for the measurement of visceral fat: a practical guide. Int J Obes Relat Metab Disord 1993; 17: 187–196.

    CAS  Google Scholar 

  7. 7

    Ross R, Léger L, Morris D, de Guise J, Guardo R . Quantification of adipose tissue by MRI: relationship with anthropometric variables. J Appl Physiol 1992; 72: 787–795.

    CAS  Article  Google Scholar 

  8. 8

    Ross R, Shaw KD, Martel Y, de Guise J, Avruch L . Adipose tissue distribution measured by magnetic resonance imaging in obese women. Am J Clin Nutr 1993; 57: 470–475.

    CAS  Article  Google Scholar 

  9. 9

    Seidell JC, Bakker CJG, van der Kooy K . Imaging techniques for measuring adipose-tissue distribution—a comparison between computed tomography and 1.5-T magnetic resonance. Am J Clin Nutr 1990; 51: 953–957.

    CAS  Article  Google Scholar 

  10. 10

    Sjöström L, Kvist H, Cederblad A, Tylén U . Determination of total adipose tissue and body fat in women by computed tomography, 40 K, and tritium. Am J Physiol 1986; 250: E736–E745.

    Google Scholar 

  11. 11

    Albu JB, Murphy L, Frager DH, Johnson JA, Pi-Sunyer FX . Visceral fat and race-dependent health risks in obese nondiabetic premenopausal women. Diabetes 1997; 46: 456–462.

    CAS  Article  Google Scholar 

  12. 12

    Bonora E, Micciolo R, Ghiatas AA, Lancaster JL, Alyassin A, Muggeo M, DeFronzo RA . Is it possible to derive a reliable estimate of human visceral and subcutaneous abdominal adipose tissue from simple anthropometric measurements? Metabolism 1995; 44: 1617–1625.

    CAS  Article  Google Scholar 

  13. 13

    Conway JM, Yanovski SZ, Avila NA, Hubbard VS . Visceral adipose tissue differences in black and white women. Am J Clin Nutr 1995; 61: 765–771.

    CAS  Article  Google Scholar 

  14. 14

    Conway JM, Chanetsa FF, Wang P . Intraabdominal adipose tissue and anthropometric surrogates in African American women with upper-and lower-body obesity. Am J Clin Nutr 1997; 66: 1345–1351.

    CAS  Article  Google Scholar 

  15. 15

    Després JP, Prud'homme D, Pouliot MC, Tremblay A, Bouchard C . Estimation of deep abdominal adipose-tissue accumulation from simple anthropometric measurements in men. Am J Clin Nutr 1991; 54: 471–477.

    Article  Google Scholar 

  16. 16

    Ferland M, Després JP, Tremblay A, Pinault S, Nadeau A, Moorjani S, Lupien PJ, Thériault G, Bouchard C . Assessment of adipose tissue distribution by computed axial tomography in obese women: association with body density and anthropometric measurements. Br J Nutr 1989; 61: 139–148.

    CAS  Article  Google Scholar 

  17. 17

    Hill JO, Sidney S, Lewis CE, Tolan K, Scherzinger AL, Stamm ER . Racial differences in amounts of visceral adipose tissue in young adults: the CARDIA (Coronary Artery Risk Development in Young Adults) Study. Am J Clin Nutr 1999; 69: 381–387.

    CAS  Article  Google Scholar 

  18. 18

    Janssen I, Heymsfield SB, Allison DB, Kotler DP, Ross R . Body mass index and waist circumference independently contribute to the prediction of nonabdominal, abdominal, subcutaneous, and visceral fat. Am J Clin Nutr 2002; 75: 683–688.

    CAS  Article  Google Scholar 

  19. 19

    Kekes-Szabo T, Hunter GR, Nyikos I, Williams M, Blaudeau T, Snyder S . Anthropometric equations for estimating abdominal adipose tissue distribution in women. Int J Obes Relat Metab Disord 1996; 20: 753–758.

    CAS  PubMed  Google Scholar 

  20. 20

    Koester RS, Hunter GR, Snyder S, Khaled MA, Berland LL . Estimation of computerized tomography derived abdominal fat distribution. Int J Obes Relat Metab Disord 1992; 16: 543–554.

    CAS  Google Scholar 

  21. 21

    Lovejoy JC, de la Bretonne JA, Klemperer M, Tulley R . Abdominal fat distribution and metabolic risk factors: effects of race. Metabolism 1996; 45: 1119–1124.

    CAS  Article  Google Scholar 

  22. 22

    Lovejoy JC, Smith SR, Rood JC . Comparison of regional fat distribution and health risk factors in middle-aged white and African American women: The Healthy Transitions Study. Obes Res 2001; 9: 10–16.

    CAS  Article  Google Scholar 

  23. 23

    Seidell JC, Oosterlee A, Thijssen MAO, Burema J, Deurenberg P, Hautvast JGAJ, Ruijs JHJ . Assessment of intra-abdominal and subcutaneous abdominal fat: relation between anthropometry and computed tomography. Am J Clin Nutr 1987; 45: 7–13.

    CAS  Article  Google Scholar 

  24. 24

    Weits T, van der Beek EJ, Wedel M, Ter Haar Romeny BM . Computed tomography measurement of abdominal fat deposition in relation to anthropometry. Int J Obes Relat Metab Disord 1988; 12: 217–225.

    CAS  Google Scholar 

  25. 25

    Després JP, Couillard C, Gagnon J, Bergeron J, Leon AS, Rao DC, Skinner JS, Wilmore JH, Bouchard C . Race, visceral adipose tissue, plasma lipids, and lipoprotein lipase activity in men and women : The Health, Risk Factors, Exercise Training, and Genetics (HERITAGE) Family Study. Arterioscler. Thromb. Vasc. Biol. 2000; 20: 1932–1938.

    Article  Google Scholar 

  26. 26

    Després JP, Lamarche B . Effects of diet and physical activity on adiposity and body fat distribution: implications for the prevention of cardiovascular disease. Nutr Res Rev 1993; 6: 137–159.

    Article  Google Scholar 

  27. 27

    Hunter GR, Snyder SW, Kekes-Szabo T, Nicholson C, Berland LL . Intra-abdominal adipose tissue values associated with risk of possessing elevated blood lipids and blood pressure. Obes Res 1994; 2: 563–568.

    CAS  Article  Google Scholar 

  28. 28

    Rankinen T, Kim SY, Perusse L, Després JP, Bouchard C . The prediction of abominal visceral fat level from body composition and anthropometry: ROC analysis. Int J Obes Relat Metab Disord 1999; 23: 801–809.

    CAS  Article  Google Scholar 

  29. 29

    Bouchard C, Leon AS, Rao DC, Skinner JS, Wilmore JH, Gagnon J . The HERITAGE Family Study: aims, design, and measurement protocol. Med Sci Sports Exerc 1995; 27: 721–729.

    CAS  Article  Google Scholar 

  30. 30

    Wilmore JH, Després JP, Stanforth PR, Mandel S, Rice T, Gagnon J, Leon AS, Rao DC, Skinner JS, Bouchard C . Alterations in body weight and composition consequent to 20 wk of endurance training: the HERITAGE Family Study. Am J Clin Nutr 1999; 70: 346–352.

    CAS  Article  Google Scholar 

  31. 31

    Sjöström L . Measurement of fat distribution. In: Bouchard C, Johnston FE (eds). Fat distribution during growth and later health outcomes. Alan R Liss: New York, 1988, pp 43–61.

    Google Scholar 

  32. 32

    Lohman TG, Roche AF, Martorell F (eds). Anthropometric standardization reference manual. Human Kinetics Books: Champaign, IL, 1988.

    Google Scholar 

  33. 33

    Jackson AS, Pollock ML . Factor analysis and multivariate scaling of anthropometric variables for the assessment of body composition. Med Sci Sports Exerc 1976; 8: 196–203.

    CAS  Article  Google Scholar 

  34. 34

    Wilmore JH, Stanforth PR, Domenick MA, Gagnon J, Daw EW, Leon AS, Rao DC, Skinner JS, Bouchard C . Reproducibility anthropometric and body composition measurements: The HERITAGE Family Study. Int J Obes Relat Metab Disord 1997; 21: 297–303.

    CAS  Article  Google Scholar 

  35. 35

    Gagnon J, Province MA, Bouchard C, Leon AS, Skinner JS, Wilmore JW, Rao DC . The HERITAGE Family Study: quality assurance and quality control. Ann Epidemiol 1996; 6: 520–529.

    CAS  Article  Google Scholar 

  36. 36

    SAS. StatView reference. SAS Institute Inc.: Cary, 1998.

  37. 37

    Pedhauzur EJ . Multiple regression in behavioral research: explanation and prediction. Harcourt Brace College Publishers: New York, 1997.

    Google Scholar 

  38. 38

    Guo SS, Chumlea WC . Statistical methods for the development and testing of predictive equations. In: Roche AF, Heymsfield SB, Lohman TG (eds). Human body composition. Human Kinetics: Champaign IL, 1996, pp 191–202.

    Google Scholar 

  39. 39

    Holiday DB, Ballard JE, McKeown BC . PRESS-related statistics: regression tools for cross-validation and case diagnostics. Med Sci Sports Exerc 1995; 27: 612–620.

    CAS  Article  Google Scholar 

  40. 40

    Hosmer DW, Lemeshow S . Applied logistic regression. John Wiley: New York, 1989.

    Google Scholar 

  41. 41

    Lemieux S, Prud'homme D, Bouchard C, Tremblay A, Després JP . A single threshold value of wasit girth identfies normal-weight and overweight subjects with excess visceral adipose tissue. Am J Clin Nutr 1996; 64: 685–693.

    CAS  Article  Google Scholar 

  42. 42

    Räikkönen K, Matthews KA, Kuller LH . Anthropometric and psychosocial determinants of visceral obesity in healthy postmenopausal women. Int J Obes Relat Metab Disord 1999; 23: 775–782.

    Article  Google Scholar 

  43. 43

    Clasey JL, Bouchard C, Teates CD, Riblett JE, Thorner MO, Hartman ML, Weltman A . The use of anthropometric and dual-energy X-ray absorptiometry (DXA) to estimate total abdominal and abdominal visceral fat in men and women. Obes Res 1999; 7: 256–264.

    CAS  Article  Google Scholar 

  44. 44

    Pouliot MC, Després JP, Lemieux S, Moorjani S, Bouchard C, Tremblay A, Nadeau A, Lupien PJ . Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol 1994; 73: 460–468.

    CAS  Article  Google Scholar 

Download references

Acknowledgements

The HERITAGE Family Study is supported by the National Heart, Lung and Blood Institute through the following grants: HL45670 (C Bouchard, PI); HL47323 (AS Leon, PI); HL47317 (DC Rao, PI); HL47327 (JS Skinner, PI); and HL47321 (JH Wilmore, PI). Claude Bouchard is partially supported by the George A Bray Chair in Nutrition. Further, Art Leon is partially supported by the Henry L Taylor Professorship in Exercise Science and Health Enhancement. Thanks are also expressed to all of the co-principal investigators, investigators, co-investigators, local project coordinators, research assistants, laboratory technicians, and secretaries who have contributed to this study (see Bouchard et al29). Finally, the HERITAGE consortium is very thankful to those hard-working families whose participation has made these data possible.

Author information

Affiliations

Authors

Corresponding author

Correspondence to P R Stanforth.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Stanforth, P., Jackson, A., Green, J. et al. Generalized abdominal visceral fat prediction models for black and white adults aged 17–65 y: the HERITAGE Family Study. Int J Obes 28, 925–932 (2004). https://doi.org/10.1038/sj.ijo.0802563

Download citation

Keywords

  • computed tomography
  • age
  • race
  • intraabdominal adipose tissue

Further reading

Search

Quick links