1. Introduction
The fact that obesity is a confirmed risk factor for chronic diseases including type 2 diabetes (T2DM) and cardiovascular diseases (CVD) has made it a major global health threat [
1,
2]. Body mass index (BMI) is the most commonly acknowledged diagnostic measurement for obesity. Nevertheless, individuals that are within a normal BMI range are not necessarily immune to metabolic disorders that are typically associated with obesity [
3,
4,
5,
6].
Lean individuals with abnormal metabolic profiles like hyperglycemia, hypertension, and dyslipidemia have been defined as “metabolically unhealthy normal-weight” (MUHNW), a concept first introduced by Ruderman et al. over 30 years ago [
3]. MUHNW individuals may also be predisposed to similar adverse health outcomes as those observed with obese patients. However, since they are not overweight or obese, they may not get adequate medical attention, thus increasing their risk for untreated complications, which has been supported by a limited number of studies. A Korean study showed MUHNW adults exhibited increased arterial stiffness and carotid atherosclerosis compared with metabolically healthy obese adults [
7]. MUHNW adults were three-fold more likely to develop diabetes than metabolically healthy obese individuals; and both groups had a similar risk in developing subclinical atherosclerosis as metabolically unhealthy obese adults [
8,
9]. Surprisingly, the highest risk of CVD and all-cause mortality of MUHNW adults was observed in a prospective cohort study with a median follow-up of ten years [
6].
There have been several studies that attempted to estimate the prevalence of the MUHNW phenotype in the general population. Because of different criteria used to define MUHNW, the prevalence can be quite heterogeneous, ranging from 2.2% to 53.6% [
10]. Even when defined by at least two metabolic abnormalities, the prevalence of MUHNW still differed in various population studies. For example, in a Korean study of the general population aged over 20 years old, MUHNW individuals accounted for 13% of the general population and 18.6% among normal-weight individuals, while in Americans of the same age range the proportions were 8.1% and 23.5%, respectively [
11,
12]. People with this phenotype have some common characteristics, including older age groups, alcohol drinker, smoker, and having a sedentary lifestyle [
10,
11,
12].
No study has investigated the abnormal metabolic phenotypes assessed by the indices of body composition in a population-based study with a large sample size. Of particular note are those individuals predisposed to adiposity but with relatively low BMI, which is typical of the Asian population [
13,
14]; such a study is well suited to be performed in the Chinese population. The results from this study will determine the prevalence of the abnormal metabolic phenotype in the normal-weight population, and provide the rationale to propose earlier prevention of metabolic diseases in this medically neglected population. Since body composition varies with gender and age [
15,
16], this study has added value in assessing the specific association of abnormal metabolic phenotypes and body composition stratified into different gender and age groups.
2. Materials and Methods
2.1. Study Population
An ongoing population-based survey of Chinese people encompassing health and basic physiological parameters was conducted from 2013 onwards, which covered five provinces, including Hainan, Shanxi, Qinghai, Gansu, and Jiangxi. A stratified multistage, random cluster process was employed to select subjects. This study was approved by the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences (approval number 029-2013). Written consent was obtained from all participants.
The data of the present study was from the site of the Shanxi province. The sampling method for this site was: first, four administrative regions were randomly sampled from the 104 districts in the Shanxi province, with two located in urban areas, one in a suburb area and one in a rural area, respectively. After that, six residential communities and six natural villages in those four regions were randomly chosen, based on a list provided by the Center for Disease Control of the Shanxi province. A total of 6410 residents aged 7–79 years old who lived in the selected communities were involved in this study. Among the 6410 participants, according to the results of a previous meta-analysis in the Chinese population [
17], 681 (298 males, 383 females) were underweight (BMI < 18.5), 3220 (1248 males, 1972 females) were normal-weight (BMI 18.5–23.9), 1894 (902 males, 992 females) were overweight (BMI 24.0–27.9), and 1894 (902 males, 992 females) were obese (BMI ≥ 28). Exclusion criteria included those aged below 18 years (
n = 588), BMI < 18.5 (
n = 681), BMI ≥ 24 (
n = 2509), and those with incomplete information to define the metabolic phenotype (n = 435). This left a total of 3015 normal-weight adults aged 18–79 years that were included in this study.
2.2. Data Collection
The information of demographic and socioeconomic status, smoking, and drinking status were obtained from participants using a questionnaire. Smoking status was divided into two categories: current smoker, or non-current smoker, according to the answers to “Are you currently a smoker”. Drinking status was categorized as current drinker and non-current drinker in the same way. Educational attainments were classified into three groups: ≤ middle school, high school, and ≥ college. Body compositions (fat mass, total body water, muscle mass), blood glucose and lipid profile, uric acid, alanine transaminase, and aspartate transaminase were collected. The latter three were chosen because they have been associated with insulin resistance in the liver [
18] and overall metabolism [
19,
20].
Weight was measured to the nearest 0.1 kg using a SECA 813 digital scale (Seca, Vogel & Halke GmbH & Co., Hamburg, Germany), with individuals wearing only light underwear and after emptying the bladder. Body height was measured to the nearest 0.1 cm using a flexible anthropometer [
21]. Waist circumference (WC) was measured at the midpoint between the lower border of the rib cage and the iliac crest. Blood pressure (BP) was measured in triplicate after a 10-min rest, using an Omron electronic sphygmomanometer (Omron Healthcare Co. Ltd., Kyoto, Japan). Handgrip was measured using a hand-held Takei dynamometer (Takei Scientific Instruments Co. Ltd., Niigata, Japan) in a standing position with the arm extended straight down to the side, and participants were asked to measure twice with their dominant hand. The larger reading of the two measurements was recorded as handgrip strength and expressed in kilograms.
Body fat mass, total body water, and impedance were measured by bioelectrical impedance (BI) analysis, which was performed with a TANITA body composition analyzer 420 (TANITA Co., Tokyo, Japan). BMI was calculated as weight (kg) divided by height squared (m
2). Skeletal muscle mass (kg) was estimated via the following BI analysis equation of Janssen et al. [
22]:
With
height measured in centimeters,
BI measured in ohms,
sex coded 1 for men and 0 for women, and
age measured in years. The percentage of body fat, body water, and skeletal muscle were calculated as the according mass divided by the body weight.
Venous blood samples were obtained after an overnight fast of at least 8 h. All blood samples were analyzed in a national central laboratory in Beijing using the Olympus auto-analyzer 2700 (Olympus Instruments Inc., Tokyo, Japan), with strict quality control. Fasting glucose was measured by the glucose oxidase method (GOD-PAP) method (Randox Laboratories Ltd., Crumlin, UK). Serum triglyceride (TG) was measured by the glycerol lipase oxidase (GPO-PAP) method (Kyowa Medex Co. Ltd., Tokyo, Japan). Low-density lipoprotein cholesterol (LDLC) and high-density lipoprotein cholesterol (HDL-C) concentrations were measured enzymatically (Kyowa Medex Co. Ltd., Tokyo, Japan). Alanine transaminase (ALT) and aspartate transaminase (AST) were measured enzymatically (Randox Laboratories Ltd., Crumlin, UK). Serum uric acid (UA) was measured by the enzymatic colorimetric method (Randox Laboratories Ltd., Crumlin, UK).
2.3. Definitions of Metabolic Phenotypes
We used the criteria established by the National Cholesterol Education Program Adult Treatment Panel III (ATP III) to identify the MUHNW phenotype [
23]. The presence of one or more of the following components: (1) high blood pressure (systolic blood pressure ≥ 130 or diastolic blood pressure ≥ 85 mmHg or known treatment for hypertension); (2) hypertriglyceridemia (fasting plasma triglycerides ≥ 1.69 mmol/L); (3) low HDL cholesterol (< 1.29 mmol/L); (4) hyperglycemia (fasting plasma glucose ≥ 5.6 mmol/L or known treatment for diabetes) was defined as metabolically unhealthy. The participants were classified into two groups according to the definition: metabolically healthy normal weight (MHNW) and MUHNW.
2.4. Statistical Analysis
All statistical analyses were conducted using the SPSS software version 18.0 for Windows (SPSS Inc., Chicago, IL, USA). The data was expressed as mean ± standard deviation (SD) for continuous variables and number (percentage) for categorical variables, respectively. Independent t-tests or ANOVA (analysis of variance) was used to compare the continuous variables between groups, and the chi-square test was used for the comparison of categorical variables. Factors associated with the unhealthy metabolic phenotype were evaluated by an unconditional logistic regression analysis, in which age, education, smoking, and drinking status were adjusted. Associations between body measurements and unhealthy metabolic phenotype within different age groups were also assessed in the logistic regression model, in which continuous variables were transformed to x’ = (x − mean)/SD to standardize OR (odds ratio). All tests were two-sided, and a value of p < 0.05 was considered as significant.
4. Discussion
To the best of our knowledge, the present study is the first to examine the factors associated with the abnormal metabolic phenotype in normal-weight Chinese adults of different age groups. Our findings suggest that higher adiposity indices (BMI, waist circumference, body fat %), and lower skeletal muscle % and body water % are associated with abnormal metabolism in lean Chinese adults, and that the impact of factors related to the unhealthy metabolic phenotype shows a decreasing trend with increasing age in females. Of note, there are disparities in the factors associated with the MUHNW phenotype in males and females aged over 60 years. Those factors remained unchanged in males throughout the age stages, while the association of BMI, body fat %, skeletal muscle %, and body water % to MUHNW attenuated and grip strength emerged as a protective factor in the elderly female.
Our finding of the association between adiposity indices and the abnormal metabolic phenotype in normal-weight adults is well in line with previous reports. The current study and other studies of the Chinese population [
24,
25] have shown that BMI and WC are higher in MUHNW individuals regardless of gender, and this result is further supported by studies conducted in Korea [
26,
27]. Moreover, the research conducted by Dvorak et al. [
28] shows that the body fat % in young women with MUHNW is higher than in normal women, which also holds true in men and women below 60 years of age in our study. However, in the Dvorak et al. and Conus et al. [
5] studies, the BMI and WC in women with abnormal metabolism were not significantly different from normal women. This inconsistency might have resulted from the ethnic differences, since Asians are verified to have more visceral fat than Europeans at a given WC or BMI [
14,
29]. Visceral fat (VAT) accumulation is a plausible mechanism for the metabolically unhealthy phenotype in our study. VAT not only acts as a fat-deposit site, but also as a highly secretory organ with a differential production of adipokines capable of regulating energy expenditure, lipid metabolism, insulin sensitivity, and inflammation [
30,
31]. A wealth of clinical studies have demonstrated that free fat acid (FFA), interleukin (IL)-6, C-reactive protein (CRP), and tumor necrosis factor (TNF)-α circulate at higher concentrations in individuals with greater VAT, indicating a pro-inflammatory feature [
32,
33,
34,
35]. In addition, increased macrophage infiltration has been found in both the subcutaneous and visceral adiposity tissue of individuals with abnormal metabolism [
36,
37], creating a low-grade chronic inflammation which is a common etiology of obesity-related complications.
Skeletal muscle is the most abundant tissue in non-obese adults, accounting for approximately 40% of the body weight and playing a critical role in energy expenditure and glucose homeostasis [
38,
39]. Although the impact of skeletal muscle mass on metabolism status has been less evaluated in normal-weight adults, previous reports can provide some clues. Recently, in the Korean sarcopenic obesity study, researchers used thigh muscle cross-sectional area corrected by weight as an index of muscle mass, and found that it decreased in MUHNW [
16]. Furthermore, a clinical study demonstrated that in normal-weight young adults, malfunction of skeletal muscle diverted ingested glucose to the liver, leading to increased hepatic de novo lipogenesis and hyperlipidemia [
40]. Since skeletal muscle accounts for the majority of glucose disposal, we presumed that the decreased skeletal muscle % accompanied by the increased fat accumulation observed in our study may play a critical role in the pathological process of the abnormal metabolic phenotype.
Our study supports previous observations that WC serves as a better indicator of metabolic risk in the elderly than BMI. Aging is associated with substantial changes in body composition, with a gradual loss of lean mass and a shift to central fat accumulation [
15,
41]. In this case, invisible obesity has already occurred with a perfectly normal BMI undermining metabolic health, which can partly explain a significantly higher proportion of the abnormal metabolic phenotype in the aged individuals in our study. On the contrary, WC reflects central obesity and has been validated in many epidemiological studies to be associated with increased FFA and adipokines, higher activity of inflammation, increased oxidative stress, blunt insulin sensitivity, and increased risk of developing insulin resistance and diabetes [
42,
43]. The shift of lean mass to fat accumulation in the elderly combined with the innate susceptibility to visceral fat deposition may account for the observed better performance of WC in both genders, even when the BMI can no longer predict unhealthy metabolism in the elderly females of our study. However, unlike elderly females, factors associated with MUHNW remained unchanged in elderly males; and although the difference in body composition is obvious, the underling mechanism of this gender disparity needs to be further investigated in a larger sample.
The present study has several limitations. First, the cross-sectional design limited our ability to infer causality from the associations observed. Second, no standard criteria for the definition of abnormal metabolism have been established. We adopted a strict definition in which satisfying any component of ATP III was considered as metabolically unhealthy. Our results might vary with different criteria. Third, the sample size of the elderly was relatively small, thus the disparity in the elderly needs to be further verified with a larger sample size and multi-ethnicities. Despite these limitations, there are several strengths of this study. We adopted a stratified multistage, randomized sampling, ensuring the representativeness of our population, thus enhancing the credibility of our results. Furthermore, the strength of association between body composition and the metabolically unhealthy normal-weight phenotype in different age groups has scarcely been investigated.