| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research Communications |
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: We assessed REE, body composition, and respiratory quotient before and after weight loss in obese black and white women.
Design: We measured REE by indirect calorimetry and body composition by densitometry before and after 20–24 wk of treatment with a 3870–4289-kJ/d diet. Subjects were 109 obese females (24 black, 85 white) with a mean (±SD) body mass index (in kg/m2) of 36.3 ± 5.0, weight of 95.7 ± 12.6 kg, and age of 42.3 ± 8.1 y.
Results: Before treatment, REE, adjusted for body composition, was significantly lower in black than in white subjects (P = 0.001). Black subjects lost significantly less weight during treatment than did white subjects (13.4 ± 5.9 kg or 14.2 ± 5.7% compared with 16.4 ± 5.6 kg or 17.0 ± 5.7%, respectively; P = 0.04). Analyses that controlled for initial REE and changes in fat mass and fat-free mass showed that blacks had significantly greater decreases in REE after treatment than did whites (9.9 ± 7.3% compared with 6.3 ± 7.4%; P = 0.02).
Conclusion: This study suggests that weight loss results in greater reductions in REE in obese black women than in obese white women. These data underscore the need to consider both biological and behavioral factors when setting expectations and assessing outcomes for obesity treatment in African American women. Am J Clin Nutr 1999;69:13–7.
Key Words: Obesity • ethnicity • resting energy expenditure • metabolism • body composition • respiratory quotient • weight loss • women
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
27.3, nearly 50% of African American women are obese compared with only 33% of white women (1). In addition, obese black women lose less weight than do white women with a variety of weight-loss treatments, including balanced deficit diets (2), very-low-energy diets (3, 4), and bariatric surgery (5). Explanations for these differences have focused primarily on behavioral and cultural factors such as diet, activity, and ideals of beauty (6–9). Recent data suggest that metabolic factors may also contribute to the observed racial differences in obesity prevalence and treatment outcome. Four studies reported within the past 2 y showed that among obese women, blacks have a significantly lower resting energy expenditure (REE) than do whites, even after weight and body composition are controlled for (10–13). A fifth study found similar racial differences in REE that were not statistically significant in a small sample (14).
Independent of any differences in REE that may exist before weight-loss treatment, it is unknown whether blacks and whites differ in how REE changes with weight loss. The finding that black women experience larger reductions in REE in response to dieting than do white women would further explain their smaller weight losses. Thus, the present study assessed measures of weight, REE, respiratory quotient, and body composition before and after weight loss in obese black and white women.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Treatment conditions
The first study (n = 128) from which subjects were selected was described previously (16). It compared 4 treatments: diet alone, diet plus aerobic exercise, diet plus strength-resistance training, and diet plus aerobic exercise and strength-resistance training combined. Subjects in the exercise conditions participated in 3 weekly, supervised sessions. Subjects attended weekly group sessions for 24 wk.
Subjects for the current analysis were also selected from a second clinical trial that compared the effects of various degrees of energy restriction on weight loss and eating behavior. Subjects (n = 25) in one of the treatment conditions in that study consumed a diet (described below) similar to that consumed by subjects in study 1 (16). Subjects in study 2 were encouraged to increase their physical activity (typically by a modest walking program), but the type of exercise was neither controlled nor supervised. Subjects in study 2 attended weekly group sessions for 20 wk.
Diet
During the first week of treatment, all subjects were asked to consume their typical diets. Thereafter, subjects in study 1 consumed a 3870-kJ/d diet for 16 wk. The diet consisted of 4 servings daily of a liquid meal replacement (628 kJ each), a portion-controlled frozen dinner entrée (1172–1255 kJ each), and 2 cups of vegetable salad. After this time, subjects were asked to consume a self-selected diet providing 5021–6276 kJ/d. Subjects in study 2 followed the same dietary protocol as those in study 1 with several minor exceptions. The diet of study 2 subjects included a fruit serving (251 kJ) and the liquid meal replacement provided 669 rather than 628 kJ, bringing the total energy intake to 4289 rather than 3870 kJ/d. In addition, study 2 subjects consumed the portion-controlled diet for 12 wk rather than 16 wk. Subjects in study 2 were also instructed to increase their intake to 5021–6276 kJ/d with self-selected foods after this time.
Exclusions
On the basis of their similar dietary treatments, 153 subjects (128 from study 1 and 25 from study 2) were eligible for participation in this study. Of these, 8 (3 black, 5 white) were excluded because of medical conditions that required treatment during the study. Of the remaining 145 subjects, 24 were excluded for the following reasons: 9 subjects (3 black, 6 white) were smokers, 1 Hispanic woman was not identified with respect to black or white race, 9 subjects (3 black, 6 white) had baseline REE measurements that could not be equilibrated after 2 separate attempts, and 5 subjects (2 black, 3 white) did not have valid baseline body-composition assessments.
During the study period, 7 of the 121 subjects (1 black, 6 white) discontinued treatment. Thus, weight-loss data were available for 114 subjects (25 black, 89 white). We were unable to obtain equilibrated REE measurements for 2 subjects (both white) and valid body-composition measurements for 3 subjects (1 black, 2 white) after treatment. Thus, 109 subjects (24 black, 85 white) were included in the present analysis, all of whom had pre- and posttreatment assessments of weight, REE, and body composition. Most of the subjects (17 black, 74 white) were from study 1 and the remainder (7 black, 11 white) were from study 2.
Dependent measures
Body weight, REE, and body composition were assessed before and after weight loss. Subjects in study 1 were assessed at week 24 and those in study 2 at week 20. Thus, all subjects were assessed 7–8 wk after discontinuing the structured 3870–4289-kJ/d diet and when prescribed a self-selected diet of 5021–6276 kJ/d. Body weight was measured by using a balance-beam scale and height was measured with use of an attached vertical rod.
REE was measured by indirect calorimetry (DeltaTrac Metabolic Monitor; SensorMedics, Yorba Linda, CA). Gas exchange (oxygen consumption and carbon dioxide production) was measured with a canopy collection technique in the morning after subjects had fasted overnight and after they had rested in a semirecumbent position for 30 min. Subjects were instructed to consume only water, avoid exercise, and refrain from taking any over-the-counter medication for the 12 h preceding the measurement. Subjects were interviewed before their REE measurement to assess adherence to the pretest conditions.
The metabolic cart was calibrated daily by using standard gases of known concentration (96% O2, 4% CO2). Continuous measurements of gas exchange were computed at 1-min intervals until 5 consecutive data points were obtained for both oxygen consumption and carbon dioxide production, which had a CV <5% (17). These equilibrated (ie, steady state) measurements were averaged and applied to the abbreviated Weir (18) formula to determine daily REE.
Body composition was assessed by densitometry. Subjects were submerged in a 1.5 x 1.5 x 1.7 m tank with a light metal chair suspended from a 15-kg scale (Chatillon, Kew Gardens, NY) as described by Goldman and Buskirk (19). Underwater measurements were made 6 times until 3 readings within 25 g were obtained. Residual lung volume was measured by the oxygen dilution technique (20) with a metabolic cart (MMC Horizon II; SensorMedics) while subjects rested in the same position that they had been in underwater. Body density was calculated from the formula of Brozek et al (21) and percentage body fat from the formula of Siri (22). Fat-free mass (FFM) was defined as body weight minus body fat.
Statistical analyses
Pretreatment
Analysis of variance was used to compare the dependent variables in blacks and whites before treatment. Traditionally, investigators have adjusted REE by expressing it as kcal/kg body wt or FFM. As Ravussin and Bogardus (23) noted, such adjustments are flawed because they are based on the faulty assumption of a 1:1 relation between REE and weight or FFM. Although linear, the relation is not 1:1 (ie, it does not have a 0,0 intercept) and it varies considerably at the extremes of body weight (24). As suggested, regression analyses were performed to identify the principal determinants of REE in this sample, and subsequent analyses of covariance (ANCOVAs) controlled for the relation between REE and its principal determinant or determinants.
Before treatment, univariate analyses revealed that weight was the strongest correlate of REE (r = 0.75, P = 0.0001). Because body weight consists of FFM and fat mass, we performed separate regression analyses on these variables to determine their relative contribution to REE. When FFM was entered into the regression first, it accounted for 29.7% of the variance and fat mass accounted for an additional 26.7%. When fat mass was entered into the regression first, it accounted for 36.6% of the variance and FFM accounted for an additional 19.8%. Thus, body weight explained 56% of the variance in REE, and fat mass was a stronger predictor of REE than was FFM.
Posttreatment
The posttreatment data were analyzed by ANCOVA. For ease of understanding, data are presented as absolute change or percentage change. However, because change and percentage change data were not normally distributed, analyses were performed on the absolute values after treatment with the initial (pretreatment) values as the covariates. For the REE analyses, changes in fat mass and FFM were also used as covariates. Pearson product moment correlations were used to assess the factors associated with weight loss and change in REE.
Preliminary analyses showed that there was no significant effect of the 5 treatment conditions (4 from study 1 and 1 from study 2) on changes in weight, fat, or FFM. There was, however, a treatment effect for change in REE. Subjects in the diet plus modest, unsupervised walking program had significantly greater reductions in REE than did subjects in the diet plus aerobic exercise or the diet plus aerobic exercise and strength training combined program. As described previously, the last 2 conditions were supervised on-site 3 times per week. These results support the findings of the earlier report (16) in which subjects who received aerobic training tended to have smaller reductions in REE at week 24. For the purposes of this study, analyses of posttreatment REE controlled for the observed treatment condition effects before assessing any race effects. All data are reported as means ± SDs.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). ANCOVA with adjustment for fat mass and FFM did show, however, that blacks had a significantly lower REE than whites.
|
An ANCOVA that controlled for initial REE, change in FFM, change in fat mass, and treatment condition showed that blacks had significantly greater reductions in REE than did whites after weight loss (P = 0.02). In blacks, REE was reduced by 713.4 ± 516.7 kJ/d, or 170.5 ± 123.5 kcal/d (9.9 ± 7.3%), compared with 454.1 ± 516.7 kJ/d, or 108.7 ± 123.5 kcal/d (6.3 ± 7.4%), in whites (Figure 1). There were no differences between the groups in changes in respiratory quotient.
|
For whites, there was a strong relation between weight loss and initial weight (r = 0.43, P < 0.0001); additionally, initial REE (r = 0.34, P < 0.002) and change in REE (r = 0.38, P = 0.0001) showed a significant relation to weight loss. Thus, higher initial weights and REEs and greater reductions in REE were associated with greater weight losses. For blacks, weight loss was most strongly related to change in REE (r = 0.49, P = 0.02), followed by initial REE (r = 0.34, P = 0.1). There was no relation between initial weight and weight loss in blacks (r = 0.1, P = 0.66).
For whites, change in REE was most strongly related to initial REE (r = 0.41, P < 0.0001), followed by weight loss (r = 0.38, P < 0.001) and initial weight (r = 0.35, P < 0.001). Thus, higher initial REEs, larger weight losses, and higher initial weights were associated with greater reductions in REE. For blacks, change in REE was most strongly related to initial REE (r = 0.55, P = 0.005) and weight loss (r = 0.49, P = 0.02). Initial weight, however, showed no relation to change in REE (r = 0.05, P = 0.82) in black subjects.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The second and more novel finding of this study suggested an additional metabolic difference that may lead to smaller weight losses in obese black women. Specifically, black subjects in our study had significantly greater reductions in REE after energy restriction and weight loss than did white subjects, even after the smaller weight losses in the black subjects were controlled for. Thus, in addition to blacks having lower REEs than whites before starting weight-loss treatment (10–12), these new data suggest that there are also racial differences in how REE responds to energy restriction and weight loss. To our knowledge, these are the first data to compare changes in REE after weight loss between black and white subjects. Unfortunately, both pre- and posttreatment differences in REE suggest that blacks have a more difficult time losing weight than do whites. The reasons for these differences are unknown. Although blacks and whites showed differences in changes in weight and FFM, our analyses controlled for these differences and still found a significant effect of race. In addition, menopausal status of the women did not significantly affect REE (6959.2 ± 948.5, 6975.2 ± 872.3, and 6680.0 ± 759.4 kJ/d for premenopausal, postmenopausal, and undetermined menopausal status, respectively; F = 1.0, P = 0.37), and menstrual status categories were not significantly different between black and white women (chi-square = 4.2, P = 0.12).
The third principal finding of this study was that there were both similarities and differences in the determinants of weight loss and change in REE in black and white women. For example, initial REE and change in REE were similarly and significantly related to weight loss in black and white subjects. By contrast, initial weight was the strongest determinant of weight loss in whites but showed no relation to weight loss in blacks. Similarly, initial weight was significantly related to change in REE in whites but not in blacks. These data are difficult to interpret but they underscore the need for caution when extrapolating findings from studies of white populations to black and other nonwhite populations.
This study had several limitations. First, the use of an outpatient sample precluded the precise quantification of the behavioral and biological factors related to decreased weight loss in black subjects. Although our data suggest that racial differences in initial REE and changes in REE are likely explanations for the observed differences in weight loss, our study did not adequately assess the role of behavioral factors such as dietary (including alcohol) intake and physical activity. These factors would not only affect weight loss, but also may have independent effects on changes in REE. The questionable reliability and validity of self-reports of food intake make assessments of dietary adherence in an outpatient setting difficult. Future studies can better assess adherence by using the doubly labeled water method (26) to estimate total daily energy expenditure or by conducting studies in a controlled environment.
A second limitation was our inability to separate the effects of weight loss and those of energy restriction on REE because our subjects were prescribed a 5021–6276-kJ/d diet at the time of their second REE measurement. Assessing patients when they are in energy balance but at a reduced weight would help to clarify whether the greater reduction in REE in blacks persists when energy restriction is fully terminated.
A third limitation was the use of densitometry to assess body composition. Densitometry does not take into account the greater bone mineral density of African Americans (27–29), which may lead to an overestimation of FFM (30). This flaw, however, seems to have little, if any, effect on body-composition estimates (29, 31), especially when using the Siri (22) formula, as we did in this study. Because densitometry was used both before and after treatment, it is less likely that the trend for blacks to lose more FFM than whites was due to the assumptions of densitometry. Nevertheless, any possible artifacts of densitometry can be remedied by using dual-energy X-ray absorptiometry measurements in future studies.
In summary, the results of this study suggest that a greater reduction in REE after weight loss or energy restriction may partially contribute to the observed differences in weight loss between obese black and white women. It would be imprudent to use these data to support a solely biological explanation for reduced weight loss by African American women. Clearly, behavioral (eg, adherence) and cultural (eg, desired body shape) factors affect weight-loss outcomes. The current data suggest that, in addition to any behavioral factors, a lower initial REE and a greater decrease in REE after energy restriction and weight loss will make weight control efforts in African American women more difficult. Thus, biological as well as behavioral differences should be considered when setting expectations and evaluating treatment outcomes for obese African American women (32).
|
FOOTNOTES |
|---|
2 Supported in part by grants RO1 MH49451-03, KO2 MH00702-07, and P30-MH45178-07.
3 Address reprint requests to GD Foster, University of Pennsylvania, 3600 Market Street, Suite 737, Philadelphia, PA 19104-2648. E-mail: fosterg{at}mail.med.upenn.edu.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. C Walsh, G. R Hunter, B. Sirikul, and B. A Gower Comparison of self-reported with objectively assessed energy expenditure in black and white women before and after weight loss Am. J. Clinical Nutrition, June 1, 2004; 79(6): 1013 - 1019. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Jones Jr, W. Shen, M.-P. St-Onge, D. Gallagher, S. Heshka, Z. Wang, and S. B Heymsfield Body-composition differences between African American and white women: relation to resting energy requirements Am. J. Clinical Nutrition, May 1, 2004; 79(5): 780 - 786. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. Samuel-Hodge, L. M. Fernandez, C. F. Henriquez-Roldan, L. F. Johnston, and T. C. Keyserling A Comparison of Self-Reported Energy Intake With Total Energy Expenditure Estimated by Accelerometer and Basal Metabolic Rate in African-American Women With Type 2 Diabetes Diabetes Care, March 1, 2004; 27(3): 663 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. M Byrne, R. L Weinsier, G. R Hunter, R. Desmond, M. A Patterson, B. E Darnell, and P. A Zuckerman Influence of distribution of lean body mass on resting metabolic rate after weight loss and weight regain: comparison of responses in white and black women Am. J. Clinical Nutrition, June 1, 2003; 77(6): 1368 - 1373. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L Weinsier, G. R Hunter, Y. Schutz, P. A Zuckerman, and B. E Darnell Physical activity in free-living, overweight white and black women: divergent responses by race to diet-induced weight loss Am. J. Clinical Nutrition, October 1, 2002; 76(4): 736 - 742. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Nguyen, H. Dumon, L. Martin, B. Siliart, L. Ferrier, B. Humbert, M. Diez, S. Breul, and V. Biourge Weight Loss Does Not Influence Energy Expenditure or Leucine Metabolism in Obese Cats J. Nutr., June 1, 2002; 132(6): 1649S - 1651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M Tershakovec, K. M Kuppler, B. Zemel, and V. A Stallings Age, sex, ethnicity, body composition, and resting energy expenditure of obese African American and white children and adolescents Am. J. Clinical Nutrition, May 1, 2002; 75(5): 867 - 871. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Djuric, S. Lababidi, L. K. Heilbrun, J. B. Depper, K. M. Poore, and V. E. Uhley Effect of Low-Fat and/or Low-Energy Diets on Anthropometric Measures in Participants of the Women's Diet Study J. Am. Coll. Nutr., February 1, 2002; 21(1): 38 - 46. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Yanovski, S. Z. Yanovski, A. J. Boyle, P. W. Gold, K. N. Sovik, N. G. Sebring, and B. Drinkard Hypothalamic-Pituitary-Adrenal Axis Activity during Exercise in African American and Caucasian Women J. Clin. Endocrinol. Metab., August 1, 2000; 85(8): 2660 - 2663. [Abstract] [Full Text]
|
|
|
|
|
|
R. L Weinsier, G. R Hunter, P. A Zuckerman, D. T Redden, B. E Darnell, D E. Larson, B. R Newcomer, and M. I Goran Energy expenditure and free-living physical activity in black and white women: comparison before and after weight loss Am. J. Clinical Nutrition, May 1, 2000; 71(5): 1138 - 1146. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
