Discipline of Genetics, Faculty of Medicine, Memorial University of Newfoundland, St John's, Newfoundland and Labrador, Canada.
Obesity (Silver Spring). 2011 Mar;19(3):624-30. doi: 10.1038/oby.2010.174. Epub 2010 Aug 12.
Nearly one-third of obese (OB) people are reported to be metabolically healthy based on BMI criteria. It is unknown whether this holds true when more accurate adiposity measurements are applied such as dual-energy X-ray absorptiometry (DXA). We compared differences in the prevalence of cardiometabolic abnormalities among adiposity groups classified using BMI vs. DXA criteria. A total of 1,907 adult volunteers from Newfoundland and Labrador participated. BMI and body fat percentage (%BF; measured using DXA) were measured following a 12-h fasting period. Subjects were categorized as normal weight (NW), overweight (OW), or OB based on BMI and %BF criteria. Cardiometabolic abnormalities considered included elevated triglyceride, glucose, and high-sensitivity C-reactive protein (hsCRP) levels, decreased high-density lipoprotein (HDL) cholesterol levels, insulin resistance, and hypertension. Subjects were classified as metabolically healthy (0 or 1 cardiometabolic abnormality) or abnormal (≥ 2 cardiometabolic abnormalities). We found low agreement in the prevalence of cardiometabolic abnormalities between BMI and %BF classifications (κ = 0.373, P < 0.001). Among NW and OW subjects, the prevalence of metabolically healthy individuals was similar between BMI and %BF (77.6 vs. 75.7% and 58.8 vs. 62.5%, respectively) however, there was a pronounced difference among OB subjects (34.0 vs. 47.7%, P < 0.05). Similar trends were evident using three additional definitions to characterize metabolically healthy individuals. Our findings indicate that approximately one-half of OB people are metabolically healthy when classified using %BF criteria which is significantly higher than previously reported using BMI. Caution should therefore be taken when making inferences about the metabolic health of an OB population depending on the method used to measure adiposity.
近三分之一的肥胖（OB）人群根据 BMI 标准被认为是代谢健康的。当应用更准确的肥胖测量方法（如双能 X 射线吸收法（DXA））时，这种情况是否仍然成立尚不清楚。我们比较了使用 BMI 与 DXA 标准分类的肥胖人群中代谢异常的患病率差异。共有 1907 名来自纽芬兰和拉布拉多的成年志愿者参与了这项研究。在禁食 12 小时后，测量了 BMI 和体脂肪百分比（%BF；使用 DXA 测量）。根据 BMI 和 %BF 标准，受试者被归类为正常体重（NW）、超重（OW）或肥胖（OB）。考虑到的代谢异常包括甘油三酯、葡萄糖和高敏 C 反应蛋白（hsCRP）水平升高、高密度脂蛋白（HDL）胆固醇水平降低、胰岛素抵抗和高血压。受试者被分类为代谢健康（0 或 1 项代谢异常）或异常（≥2 项代谢异常）。我们发现 BMI 和 %BF 分类之间代谢异常的患病率存在低一致性（κ=0.373，P<0.001）。在 NW 和 OW 受试者中，BMI 和 %BF 之间代谢健康个体的患病率相似（分别为 77.6%对 75.7%和 58.8%对 62.5%），但 OB 受试者之间存在明显差异（34.0%对 47.7%，P<0.05）。使用另外三种定义来描述代谢健康个体也存在类似的趋势。我们的研究结果表明，当使用 %BF 标准分类时，约有一半的 OB 人群代谢健康，这明显高于以前使用 BMI 标准报告的结果。因此，根据测量肥胖的方法，在推断 OB 人群的代谢健康时应谨慎。
