Reply to Letter by Yarlioglues et al Entitled “Ethnicity and the Evaluation of Anthropometric Measurements”

We thank the authors for their interest in our article ¹ and for the issues addressed in their letter. ²

Yarlioglues et al ² state that ethnic disparities in body fat and fat-free mass remain a persistent public health concern. Unfortunately, we did not include this information in our text. Our study is limited to a Chinese population. The most commonly used measurement to define obesity is body mass index (BMI). However, BMI does not distinguish between elevations in body weight by fat tissue versus muscle mass. Because of its strong association with insulin resistance, estimates of abdominal fat are often used as better determinants of risk for cardiovascular disease. ³ Abdominal fat is better reflected by measuring waist to hip ratio (WHR). ³ As stated by Yarlioglues et al, ² for many populations, widely accepted cutoffs of BMI and WHR are in many instances markedly different from those used for Caucasians. Specifically, a large body of epidemiologic evidences has shown that the Chinese had lower cutoff values than Caucasians. ^3,4

Another question by Yarlioglues et al ² is whether percutaneous coronary intervention (PCI) procedures were not successful enough and/or postprocedure medical treatment was insufficient. Despite the successful PCI and sufficient postprocedure medical treatment from our center, it is a challenge to determine the actual role of normal weight central obesity in cardiovascular events because of the complex interaction among the various risk factors. ³

Furthermore, Yarlioglues et al ² paid much attention to the baseline characteristics (Table 1 in our text ¹ ) for selection of the study population. Mean age was 52.1 ± 3.6 years in patients without normal weight central obesity, since this group actually had some participants younger than 30 years which may affect the standard deviation. As discussed later regarding the major limitation of our article, patients without normal weight central obesity might be categorized into 5 subgroups based on the combination of BMI and central obesity. ¹ Accordingly, some patients without normal weight central obesity actually have WHR higher than 0.9. However, we were unable to assess the contribution of 5 subgroups of patients without normal weight central obesity due to small sample sizes of each subgroup. ¹ Obesity is a key and causal component in metabolic syndrome, ⁵ but the incidence of metabolic syndrome was not investigated in our study. We believe that it is time to further incorporate normal weight central obesity into our important target population for weight management across the lifespan. Identifying and modifying risk factors associated with development of normal weight central obesity need to be proved by future studies.