Introduction

Cardiovascular disease risk factors have been shown to be more strongly associated with regional body fat distribution than with obesity per se (1,2,3). Anthropometric studies have clarified the relationships among particular forms of obesity, such as androidal central adiposity and gynoidal lower-body adiposity (4). The typical android male upper body obesity pattern has been seen to present a higher risk of coronary heart disease (CHD)¹ than the female gynoid pattern of fat distribution (5,6). Large-scale investigations such as the second National Health and Nutrition Examination Survey in the U.S. have identified increasing risk of the development of hypertension, diabetes, and dyslipidemia in overweight subjects with BMI >28 kg/m² (7). Similar findings have been reported in the Gothenburg study, which investigated 1462 Swedish women and found that waist-to-hip ratio correlated positively with BMI, plasma cholesterol, and development of cardiovascular diseases including hypertension (8). Furthermore, this and other studies have identified upper body obesity, as measured by the waist-to-hip ratio, to be an important predictor of metabolic abnormalities, such as dyslipidemia and hyperglycemia, as well as being predictive of CHD in men and women (4,8,9). These studies have shown that waist-to-hip ratio, a marker of central adiposity, was more predictive of CHD than BMI, a marker of general obesity, and was independent of overall obesity. The waist-to-hip ratio is clearly affected by secondary sexual characteristics (10), and the measurement of two circumferences contributes to summative measurement error. Furthermore, waist and hip circumferences seem to have differing effects (11), which may also hinder the effectiveness of this parameter as a marker of cardiovascular risk. Waist circumference (WC) alone may be a more suitable measure of risk than waist-to-hip ratio (10,12). The association of regional sites with metabolic abnormalities has also been investigated using techniques capable of directly measuring fat depots, such as by computer tomography. In a small study of white subjects, Després et al. (13) have reported that visceral obesity, rather than obesity per se, is associated with an adverse lipid profile. However, direct measurements of fat sites are limited by the cost, availability, and the required time and, therefore, are realistic only in relatively small studies in specialized centers, whereas simple anthropometry is better suited to larger population studies.

In Chinese subjects, a higher proportion of body fat than in whites for a given BMI has been suggested to increase risk at lower levels of obesity (14). This has prompted the World Health Organization to issue preliminary Asian obesity guidelines (15), in which obesity is classified as a BMI 25.0 kg/m² or a WC 80 or 90 cm in women and men, respectively. However, there are only limited Chinese population-based data, and the validity of the use of such cut-off criteria requires confirmation.

In this population-based study, we report the relative associations between central and general adiposity and the presence of cardiovascular risk factors and assess the continuity of these relationships in 2893 Chinese subjects.

Research Methods and Procedures

In a cardiovascular risk factors prevalence study, 7730 Chinese, 25 to 74 years old, were randomly selected for telephone interviews in Hong Kong from 1994 to 1996, with a response rate of 78%. Subjects with serious diseases such as cancer or who were hospitalized were excluded. A standardized questionnaire modified from the questionnaire used in the 1992 Singapore National Health Survey was used, with the addition of the World Health Organization Rose-Angina Questionnaire (previously translated into Chinese and validated in a study of elderly Chinese). Information collected included demographic characteristics, lifestyle factors, and history of cardiovascular diseases and diabetes. The method of telephone interview was validated in a morbidity survey in Hong Kong (16), and the study complied with the Declaration of Helsinki. The study was approved by the University of Hong Kong Ethics Committee, and all subjects gave written, informed consent before participating in the study.

A total of 2893 subjects had physical examinations, including anthropometry and blood tests (fasting and 2-hour post-75 grams anhydrous glucose load), in the Clinical Biochemistry Unit of Queen Mary Hospital, a teaching hospital of the Faculty of Medicine, the University of Hong Kong. The laboratory used standard methods and met international quality control programs. The attendees and nonattendees were shown to match the general population, and nonattendance bias should be small (17). The detailed methods of measurement have been reported elsewhere (17,18). Two-hour post-75 grams anhydrous oral glucose tolerance test (OGTT) insulin levels were available in a subset of the subjects (n = 839); otherwise, the biochemical and anthropometric parameters were measured in all subjects.

Subjects were considered hypertensive if their systolic and/or diastolic blood pressures were 140/90 mm Hg or if they were receiving blood pressure-lowering drugs (19). Subjects were classified as having a normal glycemic profile if their fasting plasma glucose was <6.1 mM and OGTT level was <7.8 mM. Diabetes was classified as a fasting glucose of 7.0 or postload glucose of 11.1 mM or if subjects were receiving hypoglycemic medication, whereas glucose intolerance in the nondiabetics was classified as 6.1 and <7.0 or 7.8 and <11.1 mM, respectively (20). For the indices of insulin resistance, we used the fasting insulin-glucose product, which if divided by 22.5 is numerically equivalent to the homeostasis model assessment (21), which has been shown to correlate well with the results of the euglycemic hyperinsulinemic clamp in population-based studies and the glucose and insulin results of the OGTT (22). Dyslipidemia was classified as either fasting plasma triglycerides 2.3 mM and/or total cholesterol 6.2 mM or between 5.2 and 6.2 mM with a total-to-high-density lipoprotein (HDL)-cholesterol ratio > 5.0 or if subjects were taking treatment to lower lipid levels (23,24). General obesity was classified as a BMI 25.0 kg/m² and central obesity as WC 80 or 90 cm in women and men, respectively (15).

Data from normally distributed parameters are presented as mean SD, whereas skewed data were logarithmically transformed and expressed as geometric mean with 95% confidence intervals. SPSS/PC statistical program (version 11.0.1 for Windows; SPSS, Inc., Chicago, IL) was used for the above analyses. ANOVA with the Bonferroni post hoc test was used to determine differences among the groups classified based on obesity criteria and quartiles of BMI or WC.

Logistic regression analyses were used to determine the age- and gender-adjusted odds ratios for the presence of hypertension, diabetes, or dyslipidemia at a given level of BMI or WC. Step-wise multiple regression analysis was used to investigate predictors of the independent variables [HDL- and low-density lipoprotein (LDL)-cholesterol, triglycerides, systolic and diastolic blood pressures, and insulin resistance]. Gender was coded 0 and 1 for men and women, respectively. The variables included in the analyses were linearly related to the independent variables. BMI, WC, age, and gender were included in the analyses. The appropriateness of the regression model was judged from the Durbin-Watson statistic (testing for serial correlation of adjacent error terms) and partial plots of the residuals. The tolerance and variance inflation factors were taken as measures of collinearity, with low tolerance and high variance inflation factor being signs of collinearity that indicate that a variable should not be included in the model.

Results

When the subjects were stratified by traditional obesity criteria as specified above, a total of 39.2% were found to be obese by either criterion, with 13.8% and 3.0% obese by either the BMI or WC criteria only, respectively, and a total of 22.4% of the population obese by both indices (Table 1). Subjects classified as obese by BMI or WC alone also presented with significantly higher levels of the other obesity parameter. Levels of both obesity indices were significantly higher in those with combined obesity compared with those with either category of obesity alone. Those who had central adiposity were generally older. Obesity was associated with significantly higher fibrinogen concentrations and worse lipid, glycemic, and blood pressure profiles, with those subjects who were both centrally and generally obese having worse cardiovascular risk factor levels. Central adiposity was associated with a higher proportion of hypertension and hyperglycemia (Table 1). Indeed, the prevalence of diabetes in subjects with central adiposity alone was 3 times as much as those who were obese by the BMI criterion alone (24.7% vs. 8.3%). The observed differences were essentially similar when gender-specific analyses were performed. However, given the relatively small numbers of men obese by WC criteria alone, only the anthropometric parameters, LDL-cholesterol, triglycerides, and the fasting insulin-glucose product were significantly greater than in the nonobese group (data not shown).

Table 1 - Demographic characteristics of the Chinese subjects stratified based on general and central obesity criteria.

Full table

When the 1759 nonobese subjects were subdivided into quartiles of BMI and WC, there was still a clear relationship between increasing levels of obesity and cardiovascular risk factors (Table 2). The BMI in BMI quartile 4 ranged from 23.5 to 25.0 kg/m², which resembles the overweight category in the Asian guidelines. Subjects in this group clearly had elevated levels of the lipid, blood pressure, and insulin resistance cardiovascular risk factors. For the 2893 subjects, Figure 1 shows a continuous relationship between increasing BMI and WC and the age- and gender-adjusted odds ratios for the presence of hypertension, diabetes, and dyslipidemia. The age- and gender-adjusted odds ratios for having hypertension, diabetes, or dyslipidemia at a BMI between 23 and 25 kg/m², the level currently defined in Asian populations as overweight (15), compared with those with a BMI below 19 kg/m² are 3.3, 2.1, and 3.1, respectively (Figure 1). The odds ratios of the above conditions in subjects with a WC between 75 and 80 cm, in men and women subjects combined compared with those with a WC below 65 cm are 3.2, 1.5, and 6.4, respectively.

Figure 1.

Age- and gender-adjusted odds ratios for the presence of hypertension, diabetes, and dyslipidemia in a Chinese population of 2893 subjects. The vertical lines represent the World Health Organization BMI overweight and obese cut-off points and women and men WC obesity cut-off points for Asian populations.

Full figure and legend (56K)

Table 2 - Characteristics of nonobese Chinese subjects (n = 1759) stratified by quartiles of BMI and WC.

Full table

Determination of the odds ratios for the presence of hypertension, dyslipidemia, and diabetes using logistic regression analyses indicated similar findings, with the risk being highest in the combined obesity group (Table 3). Those with only central obesity were at greater age- and gender-adjusted risk for dyslipidemia and diabetes than those with only general obesity, whereas the risk of hypertension was higher in the latter group. Similar observations were made when the risk was observed across quartiles of BMI or WC adjusted for age, gender, and the other anthropometric parameter, with odds ratios for the presence of dyslipidemia and diabetes being strongest for WC.

Table 3 - Odds ratios for the presence of hypertension, dyslipidemia, and diabetes across quartiles of BMI and WC and across types of obesity adjusted for age, gender, and WC or BMI.

Full table

WC and BMI levels were independent predictors of age- and gender-adjusted HDL-cholesterol, triglycerides, systolic and diastolic blood pressures, and the insulin resistance index (fasting insulin-glucose product), although only BMI was a predictor for LDL-cholesterol (Table 4). WC was more strongly associated with triglycerides, HDL-cholesterol, and insulin resistance than BMI (Table 4). Although central adiposity was associated with a greater proportion of hypertension (Table 1), in the multiple regression analyses, after adjusting for age and gender, the relationship was similar to that for BMI.

Table 4 - Standardized coefficients from multiple regression analyses investigating contribution of obesity to lipid, blood pressure, and insulin resistance parameters.

Full table

Discussion

Despite the categorization of the subjects based on dichotomous criteria of general and central obesity, the obese subjects by either WC or BMI alone also had significantly increased levels of the other obesity parameter compared with the nonobese group. Central adiposity was found to be strongly related to diabetes and dyslipidemia, particularly of the high triglyceride, low HDL-cholesterol type. Those with central obesity alone had 3 times as much diabetes as those obese by the BMI criterion alone. BMI is most often the only definition used when assessing levels of obesity; as such, patients with only central adiposity may not be recognized as being at risk for obesity-related increases in cardiovascular disease risk factors. However, there is a high prevalence of cardiovascular risk factors, with 26%, 25%, and 45% having hypertension, diabetes, or dyslipidemia, respectively, in the group of subjects who have central obesity but who may be classified as nonobese if only the BMI criterion is used. These data highlight the importance of evaluating body fat distribution when assessing the level of an individual's risk, which may otherwise be underestimated.

Although there was a 70% higher prevalence of hypertension in the centrally obese alone group, determination of odds ratios for the presence of hypertension suggests that BMI increases the risk of this condition as much as or more than central adiposity. However, being obese by both criteria generally presented the greatest level of risk for the presence of these conditions, in part because the subjects with combined obesity had significantly higher WC and BMI levels compared with those subjects categorized as obese by only a single parameter. The measures of general and central obesity contributed independently to disturbances in the levels of most of the metabolic parameters investigated in this study. However, from the multiple regression analyses, WC was most strongly associated with adverse HDL-cholesterol and triglyceride profiles and the insulin resistance index, whereas general adiposity was the only obesity determinant of LDL-cholesterol.

The effects on lipid and blood pressure levels were due to the location of the fat deposits as has been described (4,5,6). Obesity, either independently or through the production of free fatty acids that may lead to the development of insulin resistance, probably underlies the clustering of hypertension, obesity, dyslipidemia, and type 2 diabetes (25,26,27). Central rather than general adiposity has been reported in a number of studies (28,29,30,31,32) to be most strongly associated with the development of insulin resistance. It has been proposed that centrally deposited fat is metabolically more active than that in the periphery and is more sensitive to catecholamine-induced lipolysis but less sensitive to the antilipolytic actions of insulin (26,27). Increased free fatty acid production with the subsequent rise in triglyceride levels, as very low density lipoprotein, is associated with a reduction in insulin clearance and increased gluconeogenesis and insulin resistance, the latter by reducing skeletal muscle glucose uptake (26,27). These abnormalities then promote the development of the constellation of disorders found with the metabolic syndrome.

However, as seen in the current study and in data from a number of populations, namely Chinese (33,34,35), Hispanics (36), and whites (1,37), general obesity also strongly contributes to insulin resistance. Lipolysis also occurs in subcutaneous fat, although the latter is less active relative to visceral fat. However, because over 80% of fat is subcutaneous (38), its relative contribution may be greater than that of the visceral fat depots. Subcutaneous abdominal fat area measured by magnetic resonance imaging correlates more strongly with plasma concentrations of free fatty acids than does the visceral fat area in both Chinese and white subjects (39). WC measures incorporate both visceral and subcutaneous abdominal fat, which together contribute to the close relationship between this measure and insulin resistance and HDL-cholesterol and triglyceride levels. Although obesity-mediated insulin resistance may be involved in the adverse HDL-cholesterol and triglyceride profiles in Chinese subjects, obesity has been reported to additionally contribute to dyslipidemia and hypertension in a manner independent of insulin resistance (34,35,40).

There was a clear increase in levels of cardiovascular risk factors in the subjects categorized as obese by either parameter, and the situation was worse in those with combined obesity. However, examination of the increasing prevalence of hypertension, diabetes, and dyslipidemia with increasing levels of BMI or WC as shown in Figure 1 shows a continuous escalation in the prevalence of these parameters. Similar findings were observed when men and women were examined individually (data not shown). The age- and gender-adjusted excess risk for having hypertension, diabetes, or dyslipidemia at a BMI between 23 and 25 kg/m², the level currently defined in Asian populations as overweight (15), compared with those with a BMI below 19 kg/m², was between 110% and 230%. These findings support the lower BMI levels used to define overweight and obesity in Asian populations. However, WCs of 80 and 90 cm for women and men, respectively (15), to classify obesity seem high, with excess risk of the above conditions at the lower level of between 75 and 80 cm in men and women subjects combined compared with those with a WC below 65 cm being 220%, 50%, and 540%, respectively. Indeed, using receiver operator characteristic curves analyses, we have previously determined the optimal cut-off values of BMI and WC in relation to one or more risk factors in men to be 23.4 kg/m² and 78.2 cm and in women to be 23.4 kg/m² and 74.7 cm, respectively (41). Given the stronger risk for the presence of the components of the metabolic syndrome and standardized coefficients in the regression equations with WC for a number of parameters, particularly the lipids and insulin resistance, compared with BMI, WC may be a better predictor of these conditions. Alternatively, the WC criteria used to define obesity in Asians may still be too high.

Although dichotomous criteria to define obesity provides useful reference points to physicians and policy makers, the risk associated with increasing levels of adiposity is a continuum and does not simply start at levels of adiposity classified as overweight or obese. The present analysis of the nonobese subjects stratified by quartiles of BMI or WC clearly showed that increasing adiposity was associated with a worsening cardiovascular risk factor profile with more adverse lipid, blood pressure, and insulin resistance levels. Increasing levels of adiposity were, therefore, related to disturbances in the metabolic parameters, even when the criteria used to define obesity suggested the subjects were normal, emphasizing the limitations of the use of dichotomous criterion for the determination of cardiovascular risk.

Data from southern China further emphasizes the development of cardiovascular risks with obesity even in lean populations (42,43). It has been reported that risk for hypertension, adjusted for a number of factors including age and gender, is 34% higher in subjects in the second BMI quintile compared with those in the lowest quintile, where mean BMIs are 19.4 and 18.0 kg/m², respectively (42). As in the present study, lipids and insulin resistance parameters were also adversely affected by increasing adiposity (42). Similar observations have been made in other lean Chinese populations, further supporting the relationship between cardiovascular risk factors at lower indices of obesity when compared with white populations (43). For a given BMI, Chinese subjects have been reported to have a higher proportion of body fat than white subjects, which may contribute to the increased risk observed in Chinese subjects (14,44).

In summary, Asian guidelines for obesity based on BMI seem reasonable, but those for WC still seem high, suggesting that the guidelines should be revised. Central adiposity appears to contribute to a greater extent than general adiposity to the development of cardiovascular risk. The relationship between obesity parameters and risk is a continuum, with risk factors significantly increasing even at levels usually considered healthy.

Notes

¹ Nonstandard abbreviations: CHD, coronary heart disease; WC, waist circumference; OGTT, oral glucose tolerance test; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

References

1. Abate, N., Garg, A., Peshock, R. M., Stray-Gundersen, J., Grundy, S. M. (1995) Relationships of generalized and regional adiposity to insulin sensitivity in men. J Clin Invest. 96: 88–98. | PubMed | ISI | ChemPort |
2. Yamashita, S., Nakamura, T., Shimomura, I., et al. (1996) Insulin resistance and body fat distribution. Diabetes Care. 19: 287–291. | PubMed | ISI | ChemPort |
3. Tuck, M. L. (1994) Obesity. Swales, JD eds.. Textbook of Hypertension. 576–592.Blackwell Scientific Publications Oxford, United Kingdom.
4. Kissebah, A. H. (1997) Central obesity: measurement and metabolic effects. Diabetes Rev. 5: 8–20.
5. Vague, J. (1956) The degree of masculine differentiation of obesities: a factor determining predisposition to diabetes, atherosclerosis, gout and uric calculous disease. Am J Clin Nutr. 4: 20–34. | PubMed | ISI | ChemPort |
6. Krotkiewski, M., Björntorp, P., Sjöström, L., Smith, U. (1983) Impact of obesity on metabolism in men and women: importance of regional adipose tissue distribution. J Clin Invest. 72: 1150–1162. | PubMed | ISI | ChemPort |
7. Najjar, M. F., Rowland, M. (1987) Anthropometric reference data and prevalence of overweight. National Center for Health Statistics, Series 11, No. 238. 87–1688.U.S. Government Printing Office Washington, DC.
8. Lapidus, L., Bengtsson, C., Larsson, B., Penner, K., Rybo, E., Sjöström, L. (1984) Distribution of adipose tissue and risk of cardiovascular disease and death: a 12-year follow-up of participants in the population study of women in Gothenburg, Sweden. Br Med J. 289: 1257–1261. | ChemPort |
9. Larsson, B., Svardsudd, K., Welin, L., Wilelmsen, L., Björntorp, P., Tibblin, G. (1984) Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and death: a 13-year follow-up of participants in the Study of Men Born in 1913. Br Med J. 288: 1401–1404. | ISI | ChemPort |
10. Thomas, G. N., Young, R. P., Tomlinson, B., et al. (1999) A sibling-pair analysis of fasting lipids and anthropometric measurements and their relationship to hypertension. Clin Exp Hypertens. 21: 1161–1176.
11. Seidell, J. C., Perusse, L., Despres, J. P., Bouchard, C. (2001) Waist and hip circumferences have independent and opposite effects on cardiovascular disease risk factors: the Quebec Family Study. Am J Clin Nutr. 74: 315–321. | PubMed | ISI | ChemPort |
12. Pouliot, M. C., Despres, J. P., Lemieux, S., et al. (1994) Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol. 73: 460–468. | Article | PubMed | ISI | ChemPort |
13. Després, J. P., Moorjani, S., Lupien, P. J., Tremblay, A., Nadeau, A., Bouchard, C. (1990) Regional distribution of body fat, plasma lipoproteins and cardiovascular disease. Arteriosclerosis. 10: 497–511. | PubMed | ISI | ChemPort |
14. He, M., Tan, K. C., Li, E. T., Kung, A. W. (2001) Body fat determination by dual energy X-ray absorptiometry and its relation to body mass index and waist circumference in Hong Kong Chinese. Int J Obes. 25: 748–752.
15. World Health Organization (2000) The Asia-Pacific Perspective: Redefining Obesity and Its Treatment. 1–56.Health Communications Melbourne, Australia.
16. Lam, T. H., Kleevens, J. W. L., Wong, C. M. (1988) Doctor consultation in Hong Kong: a comparison between findings of a telephone interview with the general household survey. Community Med. 10: 175–179.
17. Lam, T. H., Liu, L. J., Janus, E. D., Bourke, C., Hedley, A. J. (1999) The relationship between fibrinogen and other coronary heart disease risk factors in a Chinese population. Atherosclerosis. 143: 405–413. | PubMed |
18. Janus, E. D., for the Hong Kong Cardiovascular Risk Factor Prevalence Study Group (1997) The Hong Kong Cardiovascular Risk Factor Prevalence Study 1995–1996. 1–145.The University of Hong Kong Pokfulam, Hong Kong.
19. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. (1997) Arch Intern Med. 157: 2413–2446. | PubMed | ISI |
20. Janus, E. D., Wat, N. M. S., Lam, K. S. L., et al. (2000) The prevalence of diabetes, association with cardiovascular risk factors and implications of diagnostic criteria (ADA1997 and WHO1998) in a 1996 community-based population study in Hong Kong Chinese. Diabetic Med. 17: 741–745. | PubMed |
21. Mathews, D. R., Hoskers, J. P., Rudenski, A. S., Naylor, B. A., Treacher, D. F., Turner, R. C. (1985) Homeostasis model assessment of insulin resistance and -cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 28: 412–419. | Article | PubMed | ISI | ChemPort |
22. Dowse, G. K., Qin, H., Collins, V. R., Zimmet, P. Z., Alberti, K. G., Gareeboo, H. (1990) Determinants of estimated insulin resistance and beta-cell function in Indian, Creole and Chinese Mauritians: the Mauritius NCD Study Group. Diab Res Clin Prac. 10: 265–279.
23. Wong, S. P., Cockram, C. S., Janus, E. D., et al. (1996) Guide to plasma lipids and lipoproteins for Hong Kong doctors. J Hong Kong Coll Cardiol. 4: 81–89.
24. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. (2001) 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). J Am Med Assoc. 285: 2486–2497. | ISI |
25. Modan, M., Halkin, H., Almog, S., et al. (1985) Hyperinsulinaemia: A link between hypertension, obesity and glucose intolerance. J Clin Invest. 75: 809–817. | PubMed | ISI | ChemPort |
26. Bonadonna, R. C., Bonora, E. (1997) Glucose and free fatty acid metabolism in human obesity. Diabetes Rev. 5: 21–51.
27. Ferrannini, E., Barrett, E. J., Bevilacqua, S., DeFronzo, R. (1983) Effects of fatty acids on glucose production and utilisation in man. J Clin Invest. 72: 1737–1747. | PubMed | ISI | ChemPort |
28. Fujioka, S., Matzuzawa, Y., Tokunaga, K., Tarui, S. (1987) Contribution of intra-abdominal fat accumulation to the impairment of glucose and lipid metabolism in human obesity. Metabolism. 36: 54–59. | Article | PubMed | ISI | ChemPort |
29. Banerji, M. A., Chaiken, R. L., Gordon, D., Kral, J. G., Lebovitz, H. E. (1995) Does intra-abdominal adipose tissue in black men determine whether NIDDM is insulin-resistant or insulin-sensitive? Diabetes. 44: 141–146.
30. McKeigue, P. M., Shah, B., Marmot, M. G. (1991) Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet. 337: 382–386. | Article | PubMed | ISI | ChemPort |
31. Matsuzawa, Y. (1997) Pathophysiology and molecular mechanisms of visceral fat syndrome: the Japanese experience. Diabetes Metab Rev. 13: 3–13. | Article | PubMed | ISI | ChemPort |
32. Després, J. P., Lemieux, S., Lamarche, B., et al. (1995) The insulin resistance-dyslipidemic syndrome: contribution of visceral obesity and therapeutic implications. Int J Obes Relat Metab Disord. 19: (Suppl 1) S76–S86. | PubMed |
33. Chan, J. C. N., Cheung, J. C. K., Lau, E. M. C., Woo, J., Chan, A. Y. W., Cockram, C. S. (1996) The metabolic syndrome in Hong Kong Chinese: the interrelationships among its components analyzed by structural equation modeling. Diabetes Care. 19: 953–959. | PubMed |
34. Anderson, P. J., Critchley, J. A., Chan, J. C., et al. (2001) Factor analysis of the metabolic syndrome: obesity vs insulin resistance as the central abnormality. Int J Obes. 25: 1782–1788.
35. Thomas, G. N., Critchley, JAJH, Tomlinson, B., Anderson, P. J., Lee, Z. S. K., Chan, J. C. N. (2000) Obesity, independent of insulin resistance, is a major determinant of blood pressure in normoglycaemic Hong Kong Chinese. Metabolism. 49: 1523–1528. | Article | PubMed | ISI | ChemPort |
36. Haffner, S. M., Miettinen, H., Stern, M. P. (1997) The homeostatic model in the San Antonio Heart Study. Diabetes Care. 20: 1087–1092. | PubMed | ISI | ChemPort |
37. Vanhala, M. J., Pitkäjärvi, T. K., Kumpusalo, E. A., Takala, J. K. (1998) Obesity type and clustering of insulin resistance-associated cardiovascular risk factors in middle-aged men and women. Int J Obes. 22: 369–374.
38. Ross, R., Leger, L., Morris, D., de Guisse, J., Guardo, R. (1992) Quantification of adipose tissue by MRI: relationship with anthropometric variables. J Appl Physiol. 72: 787–795. | PubMed | ISI | ChemPort |
39. Anderson, P. J. (1998) Insulin Sensitivity in Chinese: Inter-Relationships with Obesity and Other Components of the Metabolic Syndrome. 1–380.Division of Clinical Pharmacology, Department of Medicine and Therapeutics, Chinese University of Hong Kong Hong Kong.
40. Ko, G. T. C., Chan, J. C. N., Cockram, C. S. (1998) The association between dyslipidaemia and obesity in Chinese men after adjustment for insulin resistance. Atherosclerosis. 138: 153–161.
41. Ho, S. Y., Lam, T. H., Janus, E. D., Hong Kong Cardiovascular Risk Factor Prevalence Study Steering Committee (2003) Waist to stature ratio is more strongly associated with cardiovascular risk factors than other simple anthropometric indices. Ann Epidemiol. 13: 683–691. | Article | PubMed |
42. Hu, F. B., Wang, B., Chen, C., et al. (2000) Body mass index and cardiovascular risk factors in a rural Chinese population. Am J Epidemiol. 151: 88–97. | PubMed | ISI | ChemPort |
43. Folsom, A. R., Li, Y., Rao, X., et al. (1994) Body mass, fat distribution and cardiovascular risk factors in a lean population of south China. J Clin Epidemiol. 47: 173–181. | Article | PubMed | ISI | ChemPort |
44. Deurenberg-Yap, M., Schmidt, G., van Staveren, W. A., Deurenberg, P. (2000) The paradox of low body mass index and high body fat percentage among Chinese, Malays and Indians in Singapore. Int J Obesity. 24: 1011–1017.

Acknowledgments

The Hong Kong Cardiovascular Risk Factor Prevalence Study Steering Committee consisted of the following members: Edward D. Janus (Chairman), Clive S. Cockram, Richard Fielding, Anthony J. Hedley, P. Ho, C.P. Lau, M. Lo, S.L. Lo, P.L. Ma, John R.C. Maserei, Y.T. Tai, Brian Tomlinson, S.P. Wong, and Jean L.F. Woo. This work was supported by the Hong Kong Health Services Research Committee (Grant HSRC 411026), by the University of Hong Kong Committee on Research and Conference Grants, by the Hong Kong Research Grants Council (Grant 407/94m), and by the Hong Kong Society for the Aged. We thank the late M.R. Janus, survey center nurse coordinator, S.F. Chung for her assistance in recruitment and telephone interview coordination, T.J.T. Cheung, R.W.Y. Lam, R.Y.H. Leung, and S.C.H. Wong for special assistance in laboratory analysis, and S.T.S. Siu for assistance in data processing, and all of the interviewers.