Gavaler Judith S, Rosenblum Elaine
Department of Pharmaceutical Sciences, University of Pittsburgh School of Pharmacy, 1014 Salk Hall, Pittsburgh, PA 15261, USA.
J Am Coll Nutr. 2003 Aug;22(4):269-76. doi: 10.1080/07315724.2003.10719303.
The goal of the current study has been to examine systematically the respective roles of nutrition, exercise, menopausal weight gain, moderate drinking and smoking as determinants of body mass index (BMI) and waist hip ratio (WHR) in a setting in which the role of race or ethnic group could be simultaneously or individually evaluated as predictors of BMI and WHR. Because the use of estrogen replacement has been reported to affect estimates of body fat mass in postmenopausal women, endocrine factors have also been evaluated.
The design is cross-sectional with historical prospective elements. The study has a biomedical focus and is not an epidemiologic study. Data are from 649 women recruited into The Postmenopausal Health Disparities Study in Oklahoma. The study population was composed of 649 postmenopausal women: American Indian: 226 (34.9%), Asian: 21 (3.2%), Black: 78 (12.0%), Hispanics: 54 (8.3%) and Whites: 270 (41.6%). Recruitment occurred between 1994 and late 1999 in Oklahoma.
In this multi-racial, multi-ethnic study population, there was statistical heterogeneity in all nutrition/dietary and exercise variables as well as in other potential determinants of BMI and WHR. In contrast to the literature available for postmenopausal women in which postmenopausal status, estrogen replacement and race have rarely been taken into account, the results of multi-linear regression revealed the following: Significant predictors for BMI, with or without WHR specified, included the neuroendocrine factors, menopausal weight gain, smoking, mean fitness (i.e., difficulty performing physical activities), fat as percent of total calories, moderate drinking and being Asian or Black. When WHR was not included, total calories and socioeconomic status also entered the model. The statistical predictors of WHR in the total study population with BMI in the equation included BMI and the neuroendocrine variables of FSH, E2, but not ERT, as well as the interaction of higher intensity exercise fitness with frequency, socioeconomic status and being American Indian or Asian. When BMI was not included in the model, in addition to the neuroendocrine factors, the interaction of lower intensity exercise fitness with frequency, fat as percent of total calories, age living alone and being American Indian or and Black were significant predictors of WHR. The predictors of both BMI and WHR were found to differ among individual racial and ethnic groups.
Given the role of increased body fat and obesity in disease risk and the substantial differences in life expectancy among the racial and ethnic groups, the findings of this study, particularly in contrast to literature reports, strongly suggest that a whole variety of factors including hormonal status and race need to be considered when examining the role of dietary factors and physical activity in relation to estimates of body fat mass and disease risk.
本研究的目的是系统地考察营养、运动、绝经后体重增加、适度饮酒和吸烟作为体重指数（BMI）和腰臀比（WHR）决定因素的各自作用，该研究环境中种族或族裔群体的作用可作为BMI和WHR的预测因素同时或单独进行评估。由于据报道使用雌激素替代会影响绝经后女性体脂量的估计，因此也对内分泌因素进行了评估。
本研究采用具有历史前瞻性因素的横断面设计。该研究以生物医学为重点，并非流行病学研究。数据来自招募到俄克拉荷马州绝经后健康差异研究中的649名女性。研究人群由649名绝经后女性组成：美洲印第安人：226名（34.9%），亚洲人：21名（3.2%），黑人：78名（12.0%），西班牙裔：54名（8.3%），白人：270名（41.6%）。招募工作于1994年至19年年底在俄克拉荷马州进行。
在这个多种族、多民族的研究人群中，所有营养/饮食和运动变量以及BMI和WHR的其他潜在决定因素均存在统计学异质性。与绝经后女性的现有文献中很少考虑绝经后状态、雌激素替代和种族不同，多元线性回归结果显示如下：无论是否指定WHR，BMI的显著预测因素包括神经内分泌因素、绝经后体重增加、吸烟、平均健康状况（即进行体育活动的困难程度）、脂肪占总热量的百分比、适度饮酒以及为亚洲人或黑人。当不包括WHR时，总热量和社会经济地位也进入模型。在方程中包含BMI的整个研究人群中，WHR的统计学预测因素包括BMI以及促卵泡激素（FSH）、雌二醇（E2）的神经内分泌变量，但不包括雌激素替代疗法（ERT），以及高强度运动健康状况与频率、社会经济地位以及为美洲印第安人或亚洲人的相互作用。当模型中不包括BMI时，除神经内分泌因素外，低强度运动健康状况与频率、脂肪占总热量的百分比、独居年龄以及为美洲印第安人或黑人的相互作用是WHR的显著预测因素。发现BMI和WHR的预测因素在各个种族和族裔群体中存在差异。
鉴于体脂增加和肥胖在疾病风险中的作用以及种族和族裔群体之间预期寿命的巨大差异，本研究的结果，特别是与文献报道相比，强烈表明在研究饮食因素和体育活动与体脂量估计和疾病风险的关系时，需要考虑包括激素状态和种族在内的各种各样的因素。
