Prevalence of Metabolic Syndrome and Its Relation to Body ...

Prevalence of Metabolic Syndrome and Its Relation to Body Composition in a Chinese Rural Population Yan Feng,*,** Xiumei Hong,†‡储** Zhiping Li,† Wenbin Zhang,† Delai Jin,† Xue Liu,† Yan Zhang,* Frank B. Hu,§ Lee-Jen Wei,¶ Tonghua Zang,† Xiping Xu,储 and Xin Xu*

Abstract FENG, YAN, XIUMEI HONG, ZHIPING LI, WENBIN ZHANG, DELAI JIN, XUE LIU, YAN ZHANG, FRANK B. HU, LEE-JEN WEI, TONGHUA ZANG, XIPING XU, AND XIN XU. Prevalence of metabolic syndrome and its relation to body composition in a Chinese rural population. Obesity. 2006;14:2089 –2098. Objective: To investigate the prevalence of metabolic syndrome (MetS) with three different working definitions in a rural Chinese population and to examine its relation to body composition. Research Methods and Procedures: A total of 18,630 adults 25 to 64 years old (mean age, 45.8 years; 51.2% men) from 5686 families were enrolled from Anhui province of China during 2004 to 2005. Anthropometric measurement, body composition, blood pressure, plasma lipids, and fasting glucose and insulin and a questionnaire-based interview were obtained from each participant. Three different working definitions for MetS, including the U.S. National Cholesterol Education Program’s Adult Treatment Panel III, a modified Adult Treatment Panel III that adopts the World Health Organization’s criterion for central obesity in Asian populations, and one recently proposed by the International Diabetes Federation, were used in the study.

Received for review February 15, 2006. Accepted in final form August 22, 2006. The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. *Program for Population Genetics and Departments of §Nutrition and ¶Biostatistics, Harvard School of Public Health, Boston, Massachusetts; †Anhui Institute of Biomedicine, Anhui Medical University, Hefei, Anhui, China; ‡School of Life Science, University of Science and Technology of China, Hefei, Anhui, China; and 储Center for Population Genetics, School of Public Health, University of Illinois, Chicago, Illinois. **Y.F. and X.H. contributed equally. Address correspondence to Xin Xu, Program for Population Genetics, Harvard School of Public Health, 665 Huntington Avenue, FXB-101, Boston, MA 02115. E-mail: [email protected] Copyright © 2006 NAASO

Results: According to the three definitions, the age-adjusted prevalence of MetS for adults 25 to 64 years old was 3.2%, 4.9%, and 3.9% in men and 7.2%, 11.5%, and 10.9% in women, respectively. MetS prevalence increases significantly with age in women, but not in men. Body fat percentage and BMI and waist circumference were significantly associated with each component of MetS, especially with triglyceride level, insulin resistance index, and number of MetS components (r ⫽ 0.28 to 0.49). Discussion: The age-adjusted prevalence of MetS in our study population is lower than that reported in other urban Chinese populations. Significant gender differences in MetS prevalence were observed. The waist circumference is a good surrogate for abdominal fat percentage. Key words: metabolic syndrome, body composition, prevalence, Chinese, DXA

Introduction

Metabolic syndrome (MetS),1 also known as syndrome X (1) and insulin resistance syndrome (2), is the clustering of central obesity, hypertension, hyperglycemia, and dyslipidemia. MetS is an important risk factor for cardiovascular disease and all-cause mortality (3,4). However, there is no consensus on the clinical diagnosis criteria for MetS in a population, let alone across populations. Currently, several different diagnosis criteria for MetS exist (5). The World Health Organization (WHO)’s definition of MetS (6) includes insulin sensitivity measurement, which is not very feasible in a large epidemiological study. The Adult Treatment Panel (ATP)-III criteria proposed by the National

1

Nonstandard abbreviations: MetS, metabolic syndrome; WHO, World Health Organization; ATP, Adult Treatment Panel; WC, waist circumference; IDF, International Diabetes Federation; CT, computerized tomography; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-density lipoprotein-cholesterol; HOMAIR, insulin resistance index of homeostatic model assessment; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; WHR, waist-to-hip ratio.

OBESITY Vol. 14 No. 11 November 2006

2089

Prevalence of MetS in Chinese, Feng et al.

Cholesterol Education Program’s ATP-III (7) has been widely used in studies in different populations (8 –15). Some studies in Asian populations used the modified ATPIII criteria, which adopts the cut-off value for waist circumference (WC) in Asians suggested by the WHO (9,12–17). Recently, the International Diabetes Federation (IDF) has proposed a new population-specific criteria for MetS (18). China and other developing countries with fast economic growth have seen significant diet and lifestyle changes in the recent decades, along with a significant increase in the prevalence of obesity, diabetes, and cardiovascular disease. The reported prevalence of MetS in Chinese populations ranged from 8.6% to 16.7% (9,11,12,15,19,20), which is about one-half of the prevalence in the U.S. (10) and other western countries (8). Most of these reports were based on urban populations (9,12,15,19), although rural areas constitute ⬎80% of the population in mainland China. The prevalence of MetS and its components differ significantly between rural and urban populations (11). Obesity, especially central obesity, has been proposed as an important mechanism underlying MetS. WC has widely been used as an index for central obesity. Alternatively, body composition (percentage of body fat in mass or in volume) can be directly measured using various imaging techniques [computerized tomography (CT), magnetic resonance imaging, and ultrasound], as well as DXA. The correlation coefficients between body composition and MetS components were reported in some reports and varied (21). We have been conducting an epidemiological study of MetS in a rural population in Anhui province of China, with the ultimate goal of identifying environmental and genetic factors that contribute to MetS. In the current study, we examined the prevalence of MetS and its components using three different working definitions and compared the prevalence among different gender and age groups. We also examined the correlation between total body and abdominal fat percentage, directly measured by DXA, and MetS components.

Research Methods and Procedures Study Population We conducted a community-based screening with the purpose of identifying siblings with extreme trait values in MetS and related phenotypes in two counties (Dongzi and Wangjiang) in Anhui province of China during 2004 to 2005. Dongzhi and Wangjiang, separated by the Yangzi River, are two agricultural counties in southern Anhui with a population of 540,000 and 600,000, respectively. Eligible families, which included a minimum of three siblings 25 to 64 years old, of which two were 40 to 64 years old, were invited to participate in the study. Parents and siblings whose ages were not within 25 to 64 years were not enrolled. Also excluded were patients with type 1 diabetes, 2090

OBESITY Vol. 14 No. 11 November 2006

renal failure, chronic diseases such as tuberculosis, malignancy, rickets or other metabolic bone diseases, and thyrotoxicosis, chronic glucocorticoid use, and pregnant women. A total of 18,630 subjects from 5686 nuclear families were enrolled in our study. This study was approved by both the Human Subject Committees at the Harvard School of Public Health and at Anhui Medical University. Data Collection Data collection was conducted at two local research sites: the Dongzhi County Hospital and the Anqing Biomedical Institute of Anhui Medical University. Each participant was asked to fast after 10 PM the night before the visit. The following data were collected from each participant. Questionnaires. Questionnaires were administered to collect information on sociodemographic status, occupation, diet, lifestyle, health behavior, medical history, and medication, as well as reproductive history for women. The smoking status was classed as never, occasional (⬍10 packs total), former (ⱖ10 packs total but none in the past year), and current. The alcohol consumption was classed as never, occasional (less than once a week), former (used to have more than once a week but none in the past year), and current. Physical activity level was self-reported as mild, moderate, and heavy. Anthropometry. Height was measured without shoes to the nearest 0.1 cm on a portable stadiometer. Weight was measured in light indoor clothing without shoes to the nearest 0.1 kg. WC was measured as the minimum circumference between the inferior margin of the ribcage and the crest of the ileum. Hip circumference was measured at the level of maximum extension of the buttocks. BMI was calculated as weight/height squared (kg/m2). Blood Pressure. Resting blood pressures were measured in a sitting position after ⱖ10 minutes of rest using a mercury sphygmomanometer independently by two trained technicians, twice per person, with at least 30 minutes between the reads from the two technicians. The median of the four reads on the same participant was finally used in all analyses. Whole-Body Densitometry. Lunar DXA (GE-Lunar Prodigy, Waukesha, WI) was used to measure regional and whole-body fat mass and lean mass (kilograms) through a whole-body scan covering all major body parts including the head, neck, spine, hip, abdomen (waist), forearm, pelvis, and femur. Abdominal lean and fat mass were calculated as previously described (22,23). In brief, the central abdominal region was defined by a rectangle that extended horizontally between the top of the second and the bottom of the fourth lumbar vertebrae and laterally to the inner aspects of the ribcage. The body fat percentage was calculated as total body fat mass divided by the sum of total fat and lean mass. The abdominal fat percentage was calculated as abdominal fat mass divided by the sum of abdominal fat and lean mass.

Prevalence of MetS in Chinese, Feng et al.

Phlebotomy. Venous blood was drawn from the forearm of each participant in the fasting status. Serum and plasma were separated from blood cells in the field within 30 minutes and kept frozen at ⫺20 °C. Serum and plasma were used for laboratory assays. Laboratory Assays. Serum glucose was measured within 2 hours of the blood draw by a modified hexokinase enzymatic method using a Hitachi 7020 Automatic Analyzer (Hitachi, Tokyo, Japan). Serum total cholesterol (TC), triglyceride (TG), and high-density lipoprotein-cholesterol (HDL-C) were also measured on the Hitachi 7020 Automatic Analyzer. Serum low-density lipoprotein-cholesterol (LDL-C) level in individuals with a ⱕ4.5 mM/L TG level was calculated using the Friedewald equation: LDL-C ⫽ TC ⫺ HDL-C ⫺ TG/2.2 (24). Plasma insulin was measured by using an enhanced chemiluminescence method on an Elecsys 2010 system (Roche, Basel, Switzerland). The insulin resistance index of homeostatic model assessment (HOMA-IR) was calculated according to fasting glucose and insulin level using the HOMA2 calculator (http://www. dtu.ox.ac.uk). Definitions of MetS Three different criteria were used for the definition of MetS in the study population. National Cholesterol Education Program ATP-III Criteria. Affected individuals meet three or more of the following conditions: WC ⬎ 102 cm in men or ⬎ 88 cm in women; TG ⱖ 1.7 mM/L; HDL-C ⬍ 1.04 mM/L in men or ⬍ 1.29 mM/L in women; systolic blood pressure (SBP) ⱖ 130 mm Hg or diastolic blood pressure (DBP) ⱖ 85 mm Hg; or fasting plasma glucose (FPG) ⱖ 6.1 mM/L. Modified ATP-III Criteria. Affected individuals meet the same criteria as ATP-III, with the exception of WC: WC ⱖ 90 cm in men or ⱖ 80 cm in women. IDF Criteria. Affected individuals meet the criterion of WC ⱖ 90 cm in men or ⱖ 80 cm in women plus two of the following conditions: TG ⬎ 1.7 mM/L or under special treatment for TG abnormality; HDL-C ⬍ 1.04 mM/L in men or ⬍ 1.29 mM/L in women or under special treatment for HDL-C abnormality; SBP ⱖ 130 mm Hg or DBP ⱖ 85 mm Hg or under treatment for previously diagnosed hypertension; or FPG ⱖ 5.6 mM/L or previously diagnosed type 2 diabetes. Statistical Methods A total of 18,595 subjects from 5685 families with complete data on age, gender, and MetS-related phenotypes were used in the analyses. Distributions of each variable in different genders or age groups were compared using a Student’s t test and/or ANOVA. The estimated prevalence of MetS for adults ages 20 to 64 years was age-standardized to the 2006 population of Anqing of Anhui, China, using age groups with 5-year intervals. The difference of correlation coefficients between body composition indices and TG,

HDL, SBP/DBP, glucose, fasting insulin level, and HOMA-IR were compared after Fisher’s z transformation. All p values were two-tailed. All of the statistical analyses were performed in SAS 8.2 (SAS Institute, Cary, NC).

Results The demographic and clinical characteristics of the study subjects are presented in Table 1. The median age in men and women were 47.1 and 43.8 years, respectively. The majority (90.1%) reported farming as their occupation. In this population, job-related physical work was common (90.6% as moderate or heavy), and regular voluntary exercise was scarce (⬍1%). Compared with Western populations, this population was rather lean, with a mean BMI of 21.8 kg/m2. The percentages of overweight (BMI ⱖ 25 kg/m2) were 10.3% and 15.4% in men and women, respectively. Most characteristics differed significantly between men and women. Compared with men, women, on average, had higher BMI, total body and abdominal fat mass, higher fasting blood glucose (FBG), TG, TC, and LDL-C, lower HDL-C and blood pressure. Cigarette smoking and alcohol drinking were common in men (70.4% and 40.2%, respectively) but very rare in women (3.1% and 1.7%, respectively). According to the ATP-III, the modified ATP-III, and the IDF criteria, ⬃5.8%, 9.5%, and 8.5% of the study population had MetS, respectively. The age-adjusted prevalence of MetS for adults ages 20 to 64 years was 4.9%, 7.9%, and 7.2% respectively. The age-adjusted MetS prevalence in women was ⬃2.5 times of that in men (7.2%, 11.5%, and 10.9% vs. 3.2%, 4.9%, and 3.9%) using the above three criteria. We also examined the MetS prevalence in three different age groups (25 to 39 years old, 40 to 49 years old, and 50 to 65 years old) in both men and women (Figure 1). Interestingly, there was a significant gender difference in the effect of age on MetS prevalence. Although the prevalence did not change with age in men (p ⬎ 0.05), in women, it was 2⫻ higher in the oldest group than that in the youngest group (p ⬍ 0.01). We further looked at the distribution and prevalence of each MetS component. The gender- and age-stratified distributions of the five MetS components, including WC, TG, HDL-C, blood pressure, and FBG, are presented in Table 2. ANOVA showed significant differences in distribution of these phenotypes among the seven age groups in both men and women. All of the five phenotypes in women and three phenotypes (HDL-C, SBP/DBP, and FBG) in men increased significantly with age. The mean WC in men peaked at 40 to 44 years, and the differences among age groups were relatively small. The mean TG levels decreased significantly with age in men. The same trends of gender difference in the effects of aging were also observed in the age-specific prevalence of each MetS component (Figure 2). OBESITY Vol. 14 No. 11 November 2006

2091

Prevalence of MetS in Chinese, Feng et al.

Table 1. Demographic and clinical characteristics of the study population 关mean (standard deviation) or percentage兴 Characteristics

Men (N ⴝ 9518)

Women (N ⴝ 9077)

Age (years) Height (cm) Weight (kg) BMI (kg/m2) WC (cm) Hip circumference (cm) WHR Total body fat (kg) Abdominal fat (kg) Total body fat (%) Abdominal fat (%) FBG (mM) Fasting insulin (␮U/mL)‡ Total cholesterol (mM) TG (mM)‡ HDL-C (mM) LDL-C (mM) SBP (mm Hg) DBP (mm Hg) Self-reported hypertension (%) Self-reported type 2 diabetes (%) Physical activity (%) Slight Moderate Heavy Smoking (%) Never Occasionally Former Current Alcohol consumption (%) Never Occasional Former Current Occupation Farmer (%) Education level (%) Illiterate Primary school Middle school or higher

46.7 (7.7) 163.4 (5.9) 57.5 (8.2) 21.5 (2.6) 75.5 (8.2) 88.9 (5.4) 0.85 (0.06) 7.93 (5.35) 0.76 (0.62) 13.2 (7.2) 16.7 (10.4) 5.52 (0.99) 2.55 (2.24) 4.37 (0.83) 1.04 (1.64) 1.44 (0.42) 2.40 (0.66) 119.6 (17.7) 76.9 (11.3) 2.8 0.3

44.8 (7.4)† 153.5 (5.1)† 52.3 (7.5)† 22.2 (2.8)† 73.8 (8.2)† 89.9 (5.7)† 0.82 (0.06)† 15.10 (5.36)† 1.09 (0.53)† 28.6 (6.8)† 28.3 (9.4)† 5.55 (0.93)* 4.60 (1.81)† 4.42 (0.85)† 1.20 (1.56)† 1.41 (0.37)† 2.41 (0.68) 118.6 (18.8)† 75.0 (10.9)† 3.0 0.4 † 10.7 63.3 26.0 † 94.3 2.4 0.2 3.1 † 81.2 16.9 0.2 1.7 † 91.0 † 62.5 28.2 9.3

8.1 62.7 29.2 13.4 7.8 8.4 70.4 20.2 37.4 2.2 40.2 89.2 16.3 45.3 38.4

WC, waist circumference; WHR, waist-to-hip ratio; FBG, fasting blood glucose; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure. * p ⬍ 0.05. † p ⬍ 0.0001 (Student’s t test or ␹2 test). ‡ Log-transformed before the analysis and geometric mean and anti-log standard deviation are presented.

2092

OBESITY Vol. 14 No. 11 November 2006

Prevalence of MetS in Chinese, Feng et al.

Figure 1: Gender- and age-specific prevalence of MetS using the ATP-III criteria, the modified ATP-III criteria, and the IDF criteria.

Lastly, we examined the correlation between MetS components and body composition and the performance of BMI, waist-to-hip ratio (WHR), and WC as surrogates for total and abdominal body fat percentage (Table 3). BMI, WHR, WC, and fat percentages were positively correlated with all of the MetS-related phenotypes except for HDL-C, which was negatively associated. The correlation between body fat percentage and FBG was the weakest, with a coefficient ranging from 0.01 to 0.09. The correlations of body fat percentage with TG, fasting insulin, and HOMA-IR were fairly strong with coefficients of 0.46 to 0.49 in men and 0.31 to 0.44 in women. Generally, these correlations were slightly higher in men than in women. An extremely high correlation between total and abdominal fat percentage was observed in both men (r ⫽ 0.98) and women (r ⫽ 0.93). The correlations among BMI, WC, and fat percentage were also very strong (r ⫽ 0.76 to 0.85). WC tended to be a slightly

Table 2. Distributions of MetS-related phenotypes 关mean (standard deviation)兴 in different age and gender groups Age (years)

25 to 34

Number of subjects Men 601 Women 764 WC (cm) Men 74.9 (8.0) Women 70.8 (7.3) TG (mM)* Men 1.07 (1.70) Women 1.08 (1.51) HDL-C (mM) Men 1.30 (0.35) Women 1.36 (0.35) SBP (mm Hg) Men 113.0 (13.0) Women 109.1 (11.7) DBP (mm Hg) Men 72.8 (10.0) Women 70.0 (8.9) FBG (mM) Men 5.37 (1.20) Women 5.35 (0.61)

35 to 39

40 to 44

45 to 49

50 to 54

55 to 59

60 to 64

1412 1756

1979 2302

2044 1888

2022 1499

1090 686

370 182

75.7 (8.3) 72.5 (7.6)

76.5 (8.5) 74.1 (8.0)

75.3 (8.1) 74.4 (8.3)

75.4 (8.2) 75.1 (8.6)

74.5 (7.9) 74.2 (8.6)

74.5 (8.6) 75.3 (9.5)

p*

⬍0.001 ⬍0.001

1.11 (1.66) 1.13 (1.51)

1.10 (1.69) 1.17 (1.55)

1.03 (1.64) 1.22 (1.56)

1.00 (1.59) 1.34 (1.60)

0.98 (1.55) 1.30 (1.55)

0.98 (1.55) ⬍0.001 1.39 (1.56) ⬍0.001

1.32 (0.34) 1.36 (0.36)

1.38 (0.37) 1.38 (0.36)

1.45 (0.41) 1.43 (0.38)

1.51 (0.46) 1.47 (0.37)

1.53 (0.46) 1.52 (0.38)

1.60 (0.47) ⬍0.001 1.47 (0.39) ⬍0.001

113.8 (13.2) 116.7 (14.7) 119.3 (17.3) 122.8 (19.0) 127.3 (20.7) 130.5 (22.6) 111.6 (13.8) 115.5 (15.5) 121.2 (19.6) 125.9 (21.5) 127.7 (20.1) 135.7 (21.7) 74.2 (10.0) 72.3 (9.6) 5.42 (0.87) 5.40 (0.63)

76.1 (10.9) 74.3 (9.9) 5.47 (0.88) 5.49 (0.82)

77.4 (11.5) 76.3 (11.1) 5.54 (1.09) 5.59 (1.00)

78.7 (11.9) 78.0 (12.0) 5.60 (1.08) 5.72 (1.14)

79.2 (11.7) 77.7 (10.8) 5.58 (0.81) 5.81 (1.20)

78.8 (11.8) 79.7 (11.2)

⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001

5.60 (0.75) ⬍0.001 5.94 (1.14) ⬍0.001

WC, waist circumference; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose. Log-transformed before the analysis and geometric mean and anti-log standard deviation are presented. * ANOVA test.

OBESITY Vol. 14 No. 11 November 2006

2093

Prevalence of MetS in Chinese, Feng et al.

Figure 2: The prevalence of MetS components in men (A) and women (B). The A to G groups in the x-axis: A, WC ⬎ 102 cm in men or ⬎ 88 cm in women; B, WC ⱖ 90 cm in men or ⱖ 80 cm in women; C, TG ⱖ 1.7 mM/L; D, HDL-C ⬍ 1.04 mM/L in men or ⬍ 1.29 mM/L in women; E, SBP ⱖ 130 mm Hg or DBP ⱖ 85 mm Hg; F, FBG ⱖ 6.1 mM/L; and G, FBG ⱖ 5.6 mM/L.

better surrogate for fat percentage in men, was more or less equivalent in women when compared with BMI, and was much better than WHR in both men and women. Interestingly, the relationship between obesity phenotypes (BMI, WC, total and abdominal fat percentage) and number of MetS components (excluding the central obesity component) were almost linear in women but were S-shaped in men (Figure 3). An accelerated increase in obesity phenotypes over the component number was observed in individuals with two or more MetS components. According to the IDF criteria, 60.6% of men and 56.9% of women had at least two other abnormal components in the high WC population, which was 2 to 3 times that in the normal WC population (22.2% in men and 29.5% in women).

Discussion The prevalence of MetS has been widely investigated in various countries and ethnicities (8). However, there is no 2094

OBESITY Vol. 14 No. 11 November 2006

consensus on the criteria for MetS in a population. In addition, a valid criterion for a MetS component in one population may not be appropriate in other populations. Hence, caution should be taken in comparing MetS prevalence using different criteria or among different populations. The WHO definition (6) includes hyperglycemia and insulin resistance as essential criteria, which may require an oral glucose tolerance test and/or euglycemic clamp technique, limiting its use in epidemiological studies. In comparison, the ATP-III definition is simpler to practice. However, its WC criterion for central obesity is inappropriately stringent in leaner populations in Asian countries where overweight individuals have yet to reach the overweight definition (BMI ⱖ 25 kg/m2) of the Western population but still carry a substantially higher risk of type 2 diabetes and cardiovascular disease than their leaner counterparts (25). As a result, the modified ATP-III criteria were proposed and used in some Asian population studies (9,12–17). Very recently, the IDF proposed a new definition for MetS that puts central obesity as the essential criterion and uses different WC cut-off values in different populations (18). Whether the IDF definition is better than others in predicting an individual’s risk for cardiovascular disease remains to be established. In our study population, using the modified ATP-III criteria with an Asian-specific WC cut-off value boosted the MetS prevalence by 65%. The IDF criteria gave rise to a slightly lower prevalence than the modified ATP-III criteria in both men and women. The prevalence of MetS in our study population is much lower than in the U.S. (23.4%) (10) and also slightly lower than that reported in other urban Chinese populations (9,11,12,15). Chuang et al. (9) investigated 24,329 subjects who received health check-ups at four MJ Health Screening Centers in Taiwan and found that the prevalence of MetS was 9.5% by the ATP-III definition and 12.9% by the modified ATP-III definition. The prevalence of MetS in Hong Kong was ⬃9.6% to 16.7% (12,15). Gu et al. (11) conducted a cross-sectional survey in a national representative sample of 15,540 adults 35 to 74 years old in Mainland China. The prevalence of MetS by the ATP-III definition was 9.8% in men and 17.8% in women. Interestingly, the prevalence differed considerably between the southern and northern areas (10.9% vs. 18.2%) and between the rural and urban areas (12.7% vs. 18.6%). Our study population came from a rural area in southern China and had a younger mean age (45.8 years) than Gu et al.’s study (50.2 years). This may explain the lower prevalence we observed in our study population. Selection of our study samples was based only on age and sibship size. Although we cannot completely rule out selection bias, the age-adjusted prevalence we reported here should be a good approximation. Gender differences in MetS prevalence have been observed in many studies (11,26 –29). Similar to previous reports from Mainland China, India, Iran, and Turkey

Prevalence of MetS in Chinese, Feng et al.

Table 3. Gender-specific age-adjusted partial Pearson correlation coefficients (r†) between body composition and metabolic phenotypes BMI Men

Women

WHR Men

WC

Women

Men

BMI 1.00 1.00 0.64 0.57 0.85 WHR 0.64 0.57 1.00 1.00 0.84 WC 0.85 0.83 0.84 0.83 1.00 Total fat (%) 0.78 0.78 0.67 0.55 0.83 Abdominal fat (%) 0.76 0.74 0.69 0.64 0.83 TG‡ 0.37* 0.29 0.36¶ 0.30 0.40 HDL-C ⫺0.17 ⫺0.16¶ ⫺0.16 ⫺0.22 ⫺0.18 SBP 0.24 0.25¶ 0.18¶ 0.15¶ 0.23 DBP 0.28 0.29 0.22¶ 0.19¶ 0.28 FBG 0.05 0.03¶ 0.09 0.07 0.07 Fasting insulin‡ 0.43 0.41 0.35¶ 0.32¶ 0.44 HOMA-IR‡ 0.43 0.40 0.35¶ 0.32¶ 0.43 Number of MetS components§ 0.34 0.28 0.31 0.28 0.36

Women 0.83 0.83 1.00 0.76 0.78 0.31 ⫺0.21 0.22 0.27 0.06 0.41 0.41 0.30

Total fat (%) Men Women

Abdominal fat (%) Men

Women

0.78 0.78 0.76 0.74 0.67 0.55 0.69 0.64 0.83 0.76 0.83 0.78 1.00 1.00 0.98 0.93 0.98 0.93 1.00 1.00 0.46¶ 0.31 0.48¶ 0.36¶ ⫺0.18 ⫺0.15¶ ⫺0.18 ⫺0.19 0.20* 0.17¶ 0.20* 0.17¶ 0.27 0.23¶ 0.28 0.24* 0.08 0.01¶ 0.07 0.05 0.49¶ 0.43 0.48¶ 0.44* 0.49¶ 0.43 0.48¶ 0.44* 0.37 0.25 0.38 0.30

WHR, waist-to-hip ratio; WC, waist circumference; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; HOMA-IR, homeostasis model assessment of insulin resistance; MetS, metabolic syndrome; ATP-III, Adult Treatment Panel III. * p ⬍ 0.05 compared with the r of WC. † p ⬍ 0.05 except FBG with total fat in women. ‡ Log transformed before the analysis. § According to the ATP-III criteria and excluding the WC criterion. ¶ p ⬍ 0.01 compared with the r of WC.

(11,26 –29), the prevalence in our study population was significantly higher in women than in men. MetS was slightly more prevalent in women than in men in Oman and Hong Kong and in African Americans and Hispanic Americans (10,12,30). In comparison, the MetS prevalence was slight higher in men than in women in Taiwanese and American whites (9,10). In our study, this gender difference can mostly be attributed to WC and HDL, two MetS components that have gender-specific cut-offs in the MetS definitions and population-specific distributions (Figure 2). Under the IDF and the modified ATP-III definitions, 5% of men and 20% of women met the criterion of central obesity (WC ⱖ 90 cm in men and ⱖ80 in women), and 15% of men and 40% of women met the criterion of decreased HDL-C level (HDL-C ⬍ 1.04 mM/L in men and ⬍ 1.29 mM/L in women). Interestingly, unlike in whites and urban Chinese populations (12,19,31), the mean serum HDL-C level in men was slightly higher than in women (1.44 vs. 1.41 mM/L) in our total study population (Table 1). Gu et al.’s nationwide study reported a higher mean HDL-C level in

women than in men (1.36 vs. 1.21 mM/L in urban Chinese but essentially no gender difference in rural area, 1.36 vs. 1.35 mM/L). It is not clear to us whether the aforementioned gender-specific cut-off criteria on WC and HDL-C serve well as an appropriate index in our study population for increased risk of cardiovascular disease. More study on this issue is warranted. There is a peculiar aging effect on lipid profiles in this study population that has not been observed in previous reports. The serum HDL-C level increased with age in both men and women in the study population. The TG level decreased with age in men but increased with age in women. The observed aging effect may, in part, reflect the differences in nutrition abundance and lifestyle between young and old generations in Mainland China, where the economy grew rapidly in the past 2 decades. The younger generation tends to have an excess of nutrition and tends to lead a more sedentary lifestyle. However, the generation difference could not explain the opposite directions of the aging effects on TG in men and women. OBESITY Vol. 14 No. 11 November 2006

2095

Prevalence of MetS in Chinese, Feng et al.

Figure 3: The relationship between BMI, WC, and total body and abdominal fat percentage and number of MetS components that are defined by the modified ATP-III criteria. The central obesity component was excluded from calculation of the number of MetS components.

The IDF definition includes central obesity as an essential criterion because of the strong evidence linking it with cardiovascular disease (32,33), insulin resistance, and type 2 diabetes (22,33,34) and other MetS components (35) and because of the likelihood that central obesity is an early step in the etiologic cascade leading to full MetS (18). WC is a practical anthropometric index preferred over WHR to estimate the amount of abdominal visceral fat (21). Its correlation with abdominal fat measured by DXA was fairly strong in our study. The DXA method we used to measure abdominal fat correlates strongly with CT measurement of intra-abdominal fat (r ⫽ 0.99) (36). The correlation coefficients between MetS components and WC are similar to that with abdominal fat percentage measured by DXA, showing the WC is a good index of abdominal fat. Although statistically significant, the body fat indices were only weakly correlated (r ⬍ 0.3) with fasting glucose, HDL-C, and blood pressure, and moderately correlated 2096

OBESITY Vol. 14 No. 11 November 2006

with TG and insulin resistance (r ⫽ 0.36 to 0.48). DiPietro et al. (37) reported a much higher correlation between CT-measured abdominal fat and HDL-C (r ⫽ ⫺0.75) in a small elderly female population, whereas Stolk et al. (38) found that the correlation coefficients between ultrasoundmeasured abdominal fat and plasma glucose, HDL-C, and TG were 0.13, ⫺0.13, and 0.25, respectively, in a white population. Although a much smaller proportion of men than women in our study population meet the criteria for central obesity and MetS, an accelerated increase of obesity phenotypes was observed in men with multiple MetS components (excluding the central obesity component) but not in women. This suggests that obesity may have a relatively bigger role in MetS in men than in women in this population. In summary, the age-adjusted prevalence of MetS for adults ages 25 to 64 years in a southern rural Chinese population ranges from 4.9% to 7.9%, depending on various

Prevalence of MetS in Chinese, Feng et al.

diagnostic criteria. Significant gender differences in MetS prevalence and in the change of the prevalence with age were observed in the study population. The WC is a good surrogate for abdominal fat percentage.

Acknowledgments We thank the local Bureaus of Health of Dongzhi and Wangjiang in Anhui province of China for their support. We thank Melissa Veno for English editing of the manuscript. This work was supported by NIH Grants R01 HL073882 and R01 AR45651. References 1. Reaven GM. Banting lecture 1988: role of insulin resistance in human disease. Diabetes. 1988;37:1595– 607. 2. DeFronzo RA, Ferrannini E. Insulin resistance: a multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care. 1991;14:173–94. 3. Isomaa B, Almgren P, Tuomi T, et al. Cardiovascular morbidity and mortality associated with the metabolic syndrome. Diabetes Care. 2001;24:683–9. 4. Sattar N, Gaw A, Scherbakova O, et al. Metabolic syndrome with and without C-reactive protein as a predictor of coronary heart disease and diabetes in the West of Scotland Coronary Prevention Study. Circulation. 2003;108:414 –9. 5. Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet. 2005;365:1415–28. 6. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications: I. Diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med. 1998;15:539 –53. 7. Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486 –97. 8. Cameron AJ, Shaw JE, Zimmet PZ. The metabolic syndrome: prevalence in worldwide populations. Endocrinol Metab Clin North Am. 2004;33:351–75. 9. Chuang SY, Chen CH, Chou P. Prevalence of metabolic syndrome in a large health check-up population in Taiwan. J Chin Med Assoc. 2004;67:611–20. 10. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA. 2002;287: 356 –9. 11. Gu D, Reynolds K, Wu X, et al. Prevalence of the metabolic syndrome and overweight among adults in China. Lancet. 2005;365:1398 – 405. 12. Thomas GN, Ho SY, Janus ED, Lam KS, Hedley AJ, Lam TH. The US National Cholesterol Education Programme Adult Treatment Panel III (NCEP ATP III) prevalence of the metabolic syndrome in a Chinese population. Diabetes Res Clin Pract. 2005;67:251–7. 13. Lee WY, Park JS, Noh SY, Rhee EJ, Kim SW, Zimmet PZ. Prevalence of the metabolic syndrome among 40,698 Korean metropolitan subjects. Diabetes Res Clin Pract. 2004;65: 143–9.

14. Oh JY, Hong YS, Sung YA, Barrett-Connor E. Prevalence and factor analysis of metabolic syndrome in an urban Korean population. Diabetes Care. 2004;27:2027–32. 15. Ko GT, Cockram CS, Chow CC, et al. High prevalence of metabolic syndrome in Hong Kong Chinese-comparison of three diagnostic criteria. Diabetes Res Clin Pract. 2005;69: 160 – 8. 16. Choi SH, Ahn CW, Cha BS, et al. The prevalence of the metabolic syndrome in Korean adults: comparison of WHO and NCEP criteria. Yonsei Med J. 2005;46:198 –205. 17. Song J, Kim E, Shin C, et al. Prevalence of the metabolic syndrome among South Korean adults: the Ansan study. Diabet Med. 2004;21:1154 –5. 18. Alberti KG, Zimmet P, Shaw J. The metabolic syndrome–a new worldwide definition. Lancet. 2005;366:1059 – 62. 19. Jia WP, Xiang KS, Chen L, Lu JX, Wu YM. Epidemiological study on obesity and its comorbidities in urban Chinese older than 20 years of age in Shanghai, China. Obes Rev. 2002;3:157– 65. 20. Further Study of Risk Factors for Stroke and Coronary Heart Disease Group. The prevalence of metabolic syndrome in a 11 provinces cohort in China. Zhonghua Yu Fang Yi Xue Za Zhi. 2002;36:298 –300. 21. Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev. 2000; 21:697–738. 22. Carey DG, Jenkins AB, Campbell LV, Freund J, Chisholm DJ. Abdominal fat and insulin resistance in normal and overweight women: direct measurements reveal a strong relationship in subjects at both low and high risk of NIDDM. Diabetes. 1996;45:633– 8. 23. Samaras K, Kelly PJ, Chiano MN, Spector TD, Campbell LV. Genetic and environmental influences on total-body and central abdominal fat: the effect of physical activity in female twins. Ann Intern Med. 1999;130:873– 82. 24. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499 –502. 25. World Health Organization Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004; 363:157– 63. 26. Azizi F, Salehi P, Etemadi A, Zahedi-Asl S. Prevalence of metabolic syndrome in an urban population: Tehran Lipid and Glucose Study. Diabetes Res Clin Pract. 2003;61:29 –37. 27. Gupta A, Gupta R, Sarna M, Rastogi S, Gupta VP, Kothari K. Prevalence of diabetes, impaired fasting glucose and insulin resistance syndrome in an urban Indian population. Diabetes Res Clin Pract. 2003;61:69 –76. 28. Onat A, Ceyhan K, Basar O, Erer B, Toprak S, Sansoy V. Metabolic syndrome: major impact on coronary risk in a population with low cholesterol levels: a prospective and cross-sectional evaluation. Atherosclerosis. 2002;165:285–92. 29. Ramachandran A, Snehalatha C, Satyavani K, Sivasankari S, Vijay V. Metabolic syndrome in urban Asian Indian adults: a population study using modified ATP III criteria. Diabetes Res Clin Pract. 2003;60:199 –204. OBESITY Vol. 14 No. 11 November 2006

2097

Prevalence of MetS in Chinese, Feng et al.

30. Al-Lawati JA, Mohammed AJ, Al-Hinai HQ, Jousilahti P. Prevalence of the metabolic syndrome among Omani adults. Diabetes Care. 2003;26:1781–5. 31. Li ZY, Xu GB, Xia TA. Prevalence rate of metabolic syndrome and dyslipidemia in a large professional population in Beijing. Atherosclerosis. 2006;184:188 –92. 32. Kannel WB, Cupples LA, Ramaswami R, Stokes J 3rd, Kreger BE, Higgins M. Regional obesity and risk of cardiovascular disease: the Framingham Study. J Clin Epidemiol. 1991;44:183–90. 33. McKeigue PM, Shah B, Marmot MG. Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet. 1991;337:382– 6. 34. Pedersen SB, Borglum JD, Schmitz O, Bak JF, Sorensen NS, Richelsen B. Abdominal obesity is associated with insulin resistance and reduced glycogen synthetase activity in skeletal muscle. Metabolism. 1993;42:998 –1005.

2098

OBESITY Vol. 14 No. 11 November 2006

35. Nieves DJ, Cnop M, Retzlaff B, et al. The atherogenic lipoprotein profile associated with obesity and insulin resistance is largely attributable to intra-abdominal fat. Diabetes. 2003;52:172–9. 36. Jensen MD, Kanaley JA, Reed JE, Sheedy PF. Measurement of abdominal and visceral fat with computed tomography and dual-energy x-ray absorptiometry. Am J Clin Nutr. 1995;61:274 – 8. 37. DiPietro L, Katz LD, Nadel ER. Excess abdominal adiposity remains correlated with altered lipid concentrations in healthy older women. Int J Obes Relat Metab Disord. 1999; 23:432– 6. 38. Stolk RP, Meijer R, Mali WP, Grobbee DE, van der Graaf Y. Ultrasound measurements of intraabdominal fat estimate the metabolic syndrome better than do measurements of waist circumference. Am J Clin Nutr. 2003;77: 857– 60.