Factor Analysis of Metabolic Syndrome Using Direct Measurement of Insulin Resistance in Chinese With Different Degrees of Glucose Tolerance

By Wu, Chung-Ze Lin, Jiunn-Diann; Li, Jer-Chuan; Hsiao, Fone-Ching; Hsieh, Chang-Hsun; Kuo, Shi-Wen; Hung, Yi-Jen; Lu, Chieh-Hua; He, Chih-Tsueng; Pei, Dee

Background & objectives: With the increasing prevalence of type 2 diabetes and cardiovascular disease in Taiwan, the understanding of metabolic syndrome (MetS) becomes more important. The purpose of this study was to investigate the clustering patterns of the risk variables of the MetS with factor analysis (FacAn). Methods: A total of 564 Chinese individuals with normal glucose tolerance (N, n=345), impaired glucose tolerance (IGT, n=164) or diabetes mellitus (DM, n=55) were enrolled. Insulin resistance was measured by insulin suppression test (IST). The components of MetS such as waist hip ratio (WHR), fasting plasma glucose (FPG), systolic blood pressure (SBP), diastolic blood pressure (DBP), triglyceride (TG), high density lipoprotein cholesterol (HDLC) and steady state plasma glucose (SSPG) from IST were put into the model of exploratory FacAn.

Results: In spite of the minor different loading patterns, three dimensions were identified in the three subgroups; a “blood pressure” dimension, loading with mainly SBP and DBP, an “insulin resistance” dimension, loading mainly with SSPG, and an “adiposity/ glucose” dimension loading with TG, WHR or FPG.

Interpretation & conclusions: Our results were consistent with different ethnic groups in earlier reports that more than two dimensions were identified and that the MetS is not unified by a single underlying aetiology, i.e., insulin resistance. Longitudinal analysis in this and other populations will be required to validate our findings and to test their generalisability.

Key words Factor analysis – insulin resistance – metabolic syndrome

Criqui et al1 were the first who recognized that subjects with obesity, glucose tolerance, dyslipidaemia, and hypertension have higher risk for cardiovascular disease. In 1988, Reaven further proposed the term ‘syndrome X’ to describe the clustering of these factors and also suggested that the insulin resistance was the center of this syndrome2. Since the prevalence of type 2 diabetes and coronary heart disease are increasing in Taiwan3, they have already become an important public health issue. Recently, both World Health Organization (WHO) and National Cholesterol Education Program coined the term ‘Metabolic syndrome’ (MetS) in the hope to early detect and prevent the occurrences of diabetes and coronary heart disease4,5. Since then, whether the insulin resistance is the only single latent factor underling the metabolic syndrome, is of great interest. However, it is generally accepted that the nature of MetS is considerably complex. Each of the factors is highly intercorrelated with others. This presents a methodological challenge to researchers in this field.

Factor analysis (FacAn) is a suitable approach to address this challenge6. This multi-variate statistical tool could reduce a larger number of intercorrelated variables to a smaller set that accounts for most of the variances in the data. By this mean, a set of dimensions which are not easily observed in the original variables, could be identified.

A number of studies using FacAn to explore the clustering pattern of MetS in different populations provided convincing evidences that the dimensions do occur simultaneously more than could be expected by chance7. In general, two to four dimensions could be identified and the blood pressure was always a separate factor.

In spite of the possible important role of insulin resistance in the MetS, the majority of previous FacAn used only surrogate measurements of insulin resistance. Among them, only three used relatively precise methods such as intravenous glucose tolerance test8, 9 or clamp10. However, only small numbers of subjects were enrolled in two of these studies. The third one, done by Hanley et al10, although had significant number of subjects, did not include diabetic subjects.

In this study, we used insulin suppression test to directly measure the insulin resistance in 564 Chinese having normal glucose tolerance, impaired glucose tolerance or type 2 diabetes. Thus, we can have a better understanding of the relationships among the factors of MetS in different degree of glucose tolerance and compare our results with other different ethnic groups.

Material & Methods

Subjects: Five hundred sixty four Chinese subjects were enrolled in Tri-Service General Hospital, Taipei, Taiwan between 1999 and 2003. They were participants in the hospital routine health check and were given a choice to be enrolled in the present study. All subjects had no significant medical or surgical history in the past. Before the study, they were instructed by the doctors and dieticians not to receive any medications known to affect glucose or lipid metabolism and to stay on a stable diet for at least one week before the study. Each participant visited the clinical research center on two different days. On the day of the visit, each had a complete routine work-up to rule out the presence of cardiovascular, respiratory, renal or endocrine disorders. Waist and hip circumferences were measured at the umbilicus and the gluteal fold in the supine and standing position, respectively. Blood pressure was measured according to the JNC VI (Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure) suggestions11.

On the first study day, the participants took the standard 75 g glucose oral glucose tolerance test (OGTT). Plasma glucose and insulin concentrations were measured before and 30, 60 and 120 min after the glucose challenge. The results of the OGTT were classified according to the 1997 American Diabetes Association criteria12. Among them, 345 were grouped as normal (N group), 164 were as either impaired glucose tolerance or impaired fasting glucose (IGT group), and 55 type 2 diabetes (DM group).

On a separate day, insulin resistance was measured by the insulin suppression test. After an overnight fast at least for 10 h, intravenous catheter was placed in both the arms. One arm was used for the administration of a 180-min infusion of somatostatin (250 [mu]g/h), insulin (25 mU/m/min), and glucose (240 mg/m^sup 2^/min), respectively. The other arm was used for collecting blood samples. Blood was sampled every 30 min initially and then at 10 min intervals from 150 to 180 min of the infusion to determine the steady-state plasma insulin (SSPI) and glucose (SSPG) concentrations for each individual. Because SSPI concentrations were similar for all the subjects, the SSPG concentration provided a direct measure of the ability of insulin to mediate disposal of an infused glucose load; the higher the SSPG, the more insulin resistant the individual.

Other than the SSPG, homeostasis model assessment was also used to estimate insulin resistance (HOMA-IR). It is a mathematical model based on glucose and insulin interaction in different organs, including the pancreas, liver, and peripheral tissues13. HOMA-IR (fasting plasma insulinxfasting plasma glucose/22.5) determines insulin resistance13,14. Application of HOMA-IR has also been widely used in epidemiological studies13,15.

All plasma obtained during tests were separated from blood within 1 h and stored at -30[degrees]C until analyzed. Insulin was measured by a commercial solid phase radioimmunoassay kit (Coat-A-Count insulin kit. Diagnostic Products Corporation, Los Angles. California. USA). The intra- and inter-assay coefficients of variation for insulin were 3.3 and 2.5 per cent respectively. Plasma glucose was measured using a glucose oxidase method (YSI 203 glucose analyzer, Scientific Division, Yellow Spring Instrument Company, Inc., Yellow Spring, Ohio, USA). Both plasma total cholesterol (TC) and triglyceride (TG) and low-density lipoprotein cholesterol (LDLC) were measured using the dry, multilayer analytical slide method in the Fuji Dri-Chem 3000 analyzer (Fuji Photo Film Corporation Minato- Ku Tokyo, Japan)16. Serum level of high-density lipoprotein cholesterol (HDLC) was determined by an enzymatic cholesterol assay method after dextran sulfate precipitation17.

Statistical analysis: Because the age (43+-12.2 yr for normal, 51.0+- 11.3 for IGT and 49.9+- 11.2 for diabetes) and BMI (23.5+- 3.4 kg/m^sup 2^ for normal, 25.4+-3.2 for IGT and 26.0+-3.7 for diabetes) were significantly lower in the N group, all the statistical calculations were adjusted for age in the ANOVA. However, adjustment was not done in the Karl Pearson correlation and FacAn since these methods were done only within each group. Associations between baseline anthropometric and metabolic variables were determined using Karl Pearson correlation analysis.

Exploratory FacAn was conducted to investigate the relationships among the variables which constitute MetS. Principle FacAn was used to transform the original variables into a new set of components which are independent of each other. The number of components to be retained was based on Scree plot analysis (factors above the break in the curve were retained) and eigenvalue criteria (1.0). The eigenvalues are used to identify the minimum number of components6,18. Since the smallest group had 55 subjects, the factors put into the model should be limited to 10 (19). Varimax rotation was used to obtain a set of independent interpretable factors. The resulting factor pattern was interpreted using factor loadings of > 0.4. The results of the KMO test for each group were 0.588, 0.558 and 0.46 and Bartlett’s test were O, O and 0.698 for groups N, IGT and DM, respectively. Due to the differences in sample size and severity of glucose intolerance, the clustering of the risk factors was not exactly the same in the subgroups. The fasting plasma insulin (FPI) levels, postchallenged glucose concentrations, insulin and glucose area under curve of the oral glucose tolerance test and homeostasis model assessment were all considered as an indirect surrogate method for measuring insulin resistance and, compared to insulin suppression test, they are less accurate2-. Therefore, we only put the SSPG but not these variables into the model. Other variables which are considered to be related with MetS such as waist/hip ratio, systolic blood pressure (SBP). diastolic blood pressure (DBP), TG, HDLC and fasting plasma glucose (FPG) were also analyzed.

Calculations were performed using the SPSS (10.0) statistical package (SPSS Inc., Chicago, IL, USA). P<.05 was considered to be significant.

The study protocol was approved by the hospital ethics committee, and the purpose and potential risks of the study were explained to the participants before obtaining their written consent to participate.

Results

The BMI. WHR, FPG, 2 h glucose levels post challenge, SSPG and TG were lowest and HDLC was the highest in N group. But the SBP and DBP were only lower in N group than in the IGT group. Finally, the IGT group had lower FPG, FPI, 2 h glucose levels post challenge, SSPG and HOMA than the DM group (Table I). The results of Karl Pearson correlation analyses of baseline variables are presented in Table II. The SSPG correlated well with all the parameters in N group. But in DM and IGT group, only the correlation between SSPG and HOMA could be noted in both groups. Because the sample size was not the same, the significance of correlations was not consistent across the three groups. However, the relationships between these risk factors still agreed with other previously reported results7-10.

Table III displays the results of FacAn of core metabolic factors among these three groups. The Fig. depicts the same results graphically and the percentage of variance explained by each factor is also shown. In spite of the minor difference of the loading patterns, two dimension in N group, and a three-dimension solution in IGT and DM groups was found, which was supported by the retention criteria described in the method. These dimensions were interpreted as (i) “blood pressure” dimension, loading with mainly SBP and DBP, (H) “insulin resistance” dimension, loading mainly with SSPG, and (iii) “adiposity/glucose ” dimension loading with TG, WHR or FPG.

Discussion

In this study, we used FacAn to test the hypothesis that there is one single underlying aetiology which accounts for the mutual occurrence of the MetS factors. The identification of only one dimension from multiple risk factors would suggest that insulin resistance is the core of the metabolic syndrome. However, in our study, three dimensions (insulin resistance, blood pressure and adiposity/glucose) were found and the results rejected this hypothesis. It does not imply that insulin resistance is unrelated to blood glucose and hypertension, but instead, it indicates that more than one physiological processes mediate the phenomenon of risk factor clustering.

Although the number of subjects in our study was smaller than an earlier study8, it was still large enough to perform the FacAn because we did not put too many risk variables into the model. The insulin resistance was also measured by the relatively accurate test – the insulin suppression test. Other than normal and impaired glucose tolerance groups, which were already examined in Hanley’s study8, diabetes group was also added in our study. More importantly, in this study we enrolled only Han Chinese which is an ethnic group experiencing high rates of MetS and type 2 diabetes23; also little is known about the epidemiology and pathogenesis of the disease in this ethnic group.

In other studies, factors which are not directly associated with insulin resistance were also put into the model, such as LDLC particle24, globulin, albumin, potassium25, alcohol consumption, albumin exertion rate or left ventricular mass26. Due to the complexity of the model, these factors were not considered in our study. This makes the comparison of the present study with other published studies hampered. However, in general, three dimensions identified in our study; blood pressure27, insulin resistance8,28- 30 and adiposity/ glucose30 were consistent with other studies in different ethnic groups or glucose tolerance.

In the present study, we used WHR (central obesity) to estimate the effect of obesity because we considered it could provide more information than body mass index alone31. Some of the previous studies showed that obesity is associated with glucose/insulin24 or with lipids factor27. In most of our subgroups, the WHR was loaded differently in different groups, but mainly in TG, FPG or SSPG. Another unique finding in our analysis was that the SSPG associated with TG and WHR in both N and IGT groups. In the study by Hanley et aP, the insulin sensitivity was linked with WHR, BMI, TG, HDLC, FPG, 2 hour glucose post loading and HOMA-IR. However, these parameters could be simplified to a smaller number, since WHR and BMI, low HDLC and high TG were correlated well with each other. Therefore, the only difference was the loading of the FPG. In our study, the FPG seemed to be more related to WHR or TG. This difference could be due to different ethnicities, measurements of insulin resistance and numbers of subjects.

It is also interesting to observe the pattern of clustering in different severity of glucose intolerance. In the Hanley’s report8, normal and impaired glucose tolerance subjects were studied and similar clustering in these 2 groups was noted. They drew the conclusions that differences in metabolic clustering do not become apparent until diabetes was fully developed. In other two studies also both normal and diabetes patients were investigated6,7. Although the dimensions identified were similar in normal and individuals with diabetes, the loading patterns were not the same. Obesity seemed to load more on the normal subjects and, after the subjects became diabetic, blood pressure played a more important role. In our study, instead of obesity, blood pressure universally loaded first in N, IGT and DM except for the DBP in DM group. This could be explained by the fact that BMI is smaller in Chinese than Caucasian. Thus the importance of obesity becomes less than the blood pressure.

In some of the other studies using FacAn to examine the loading pattern of the MetS, it could be noted that the blood pressure shares the correlation with insulin resistance through waist27,32. Although the cohort was non-diabetic in the study, by Meigs and coworkers21 in another study26, they were not only diabetic, but also Hong-Kong Chinese. The mutual association of the obesity (waist) could be explained by its correlation with insulin resistance and hypertension. In our study, the similar relationship could be found most evidently in N group. The WHR, BP and SSPG were all loaded in the same dimension. This result confirms the importance of the obesity in insulin resistance and BP. Meanwhile, in the IGT and DM groups, obesity (WHR) still has its important effect on insulin resistance (IGT group) or BP (DM group).

Some may argue that the inconsistent loading patterns in different groups in our study made the results hard to be interpreted. However, this is a universal phenomenon in other studies and the putative explanations might be the different severity of glucose intolerance. In the study by Gray et al., 4228 subjects were studied. The dimensions between normal (dimension 1: glucose BMI, insulin; factor 2: blood pressure) and diabetes (dimension 1: blood pressure; factor 2: lipids, glucose) were different. Similar results were also reported by Edwards et al., there were three dimensions being identified in normal subjects and four in the diabetic subjects. Not only this, many studies demonstrated that loading patterns were also different in different sex. Whether putting less risk variables into the model could improve the consistency is still remained to be investigated. However, we still think that the inconsistency in the loading pattern in different severity of glucose intolerance does not decrease the importance of our findings that there is not only one dimension in the MetS.

There were two limitations in our study. The first one was that the case numbers were not comparable in the three groups, especially in the ‘DM group’. Although, the intra-group FacAn was done, the data would have been more reliable if the numbers were more. Secondly, the study subjects were enrolled from the hospital cohort, not in a randomized way.

In conclusion, although there were minor differences including the loading patterns, FacAn identified three independent factors underlying clustering of the basic risk variables constituting the MetS in most of the subgroups of N, IGT and DM; a “blood pressure” factor, loading with mainly SBP and DBP, a “insulin resistance” factor, loading mainly with SSPG., and finally, a “adiposity/ glucose” factor loading with TG, WHR or FPG. The SSPG only associated with one of the three factors underlying the syndrome. A precise definition of the MetS, independent of the presence of determinants of glucose intolerance or hypertension, may help in the search for the genetic causes of shared risk for both type 2 diabetes and cardiovascular diseases. Longitudinal analysis in this and other populations will be required to validate our findings and to test their generalisability. References

1. Criqui MH, Barrett-Connor E, Holdbrook MJ, Austin M, Turner JD. Clustering of cardiovascular disease risk factors. Prev Med 1980; 9: 525-33.

2. Reaven GM. Banting Lecture 1988. Role of insulin resistance in human disease. Diabetes 1988; 37 : 1595-607.

3. Chu NF, Wang DJ, Shieh SM. Obesity, leptin and blood pressure among children in Taiwan: the Taipei Children’s Heart Study. Am J Hypertens 2001; 14 : 135-40.

4. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998; 15 : 539-53.

5. 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.

6. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood- glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998; 352 : 837-53.

7. Liese AD, Mayer-Davis EJ, Haffner SM. Development of the multiple metabolic syndrome: an epidemiologic perspective. Epidemiol Rev 1998; 20 : 157-72.

8. Edwards KL, Burchfiel CM, Sharp DS, Curb JD, Rodriguez BL, Fujimoto WY, et al. Factors of the insulin resistance syndrome in nondiabetic and diabetic elderly Japanese-American men. Am J Epidemiol 1998; 147 : 441-7.

9. Gray RS, Fabsitz RR, Cowan LD, Lee ET, Howard BV, Savage PJ. Risk factor clustering in the insulin resistance syndrome. The Strong Heart Study. Am J Epidemiol 1998; 148 : 869-78.

10. Hanley AJ, Karter AJ, Festa A, D’Agostino R Jr, Wagenknecht LE, Savage P, et al. Factor analysis of metabolic syndrome using directly measured insulin sensitivity: The Insulin Resistance Atherosclerosis Study. Diabetes 2002; 51 : 2642-7.

11. The sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Arch Intern Med 1997; 157 : 2413-46.

12. American Diabetes Association: clinical practice recommendations 1997. Diabetes Care 1997; 20 (Suppl 1): S1-70.

13. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta- cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28 : 412-9.

14. Emoto M, Nishizawa Y, Maekawa K, Hiura Y, Kanda H, Kawagishi T, et al. Homeostasis model assessment as a clinical index of insulin resistance in type 2 diabetic patients treated with sulfonylureas. Diabetes Care 1999; 22 : 818-22.

15. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 1999; 22 : 1462-70.

16. Spayd RW, Bruschi B, Burdick BA, Dappen GM, Eikenberry JN, Esders TW, et al. Multilayer film elements for clinical analysis: applications to representative chemical determinations. Clin Chem 1978; 24 : 1343-50.

17. Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantitation of high-density- lipoprotein cholesterol. Clin Chem 1982; 28 ; 1379-88.

18. Isomaa B, Almgren P, Tuomi T, Forsen B, Lahti K, Nissen M, et al. Cardiovascular morbidity and mortality associated with the metabolic syndrome. Diabetes Care 2001; 24 : 683-9.

19. Kline P. An easy guide to factor analysis. Lodon: Routledge; 1994.

20. Baron AD. Pathogenesis and measurement of insulin resistance in hypertension. Curr Opin Nephrol Hypertens 1994; 3 : 631-5.

21. Del Prato S. Measurement of insulin resistance in vivo. Drugs 1999; 58 (Suppl 1): 3-6.

22. Hosker JP, Matthews DR, Rudenski AS, Burnett MA, Darling P, Bown EG, et al. Continuous infusion of glucose with model assessment: measurement of insulin resistance and beta-cell function in man. Diabetologia 1985; 28 : 401-11.

23. Zimmet PZ, McCarty DJ, de Courten MP. The global epidemiology of non-insulin-dependent diabetes mellitus and the metabolic syndrome. J Diabetes Complications 1997; 11 : 60-8.

24. Edwards KL, Austin MA, Newman B, Mayer E, Krauss RM, Selby JV. Multivariate analysis of the insulin resistance syndrome in women. Arterioscler Thromb 1994; 14 : 1940-5.

25. Leyva F, Godsland IF, Worthington M, Wallon C, Stevenson JC. Factors of the metabolic syndrome: baseline interrelationships in the first follow-up cohort of the HDDRISC Study (HDDRISC-1). Heart Disease and Diabetes Risk Indicators in a Screened Cohort. Arterioscler Thromb Vasc Biol 1998; 18 : 208-14.

26. Lempiainen P, Mykkanen L, Pyorala K, Laakso M, Kuusisto J. Insulin resistance syndrome predicts coronary heart disease events in elderly nondiabetic men. Circulation 1999; 100 : 123-8.

27. Meigs JB, D’Agostino RB Sr, Wilson PW, Cupples LA, Nathan DM, Singer DE. Risk variable clustering in the insulin resistance syndrome. The Framingham Offspring Study. Diabetes 1997; 46 : 1594- 600.

28. Chen W, Srinivasan SR, Elkasabany A, Berenson GS. Cardiovascular risk factors clustering features of insulin resistance syndrome (Syndrome X) in a biracial (Black-White) population of children, adolescents, and young adults: the Bogalusa Heart Study. Am J Epidemiol 1999; 150 : 667-74.

29. Guthrie RA, Guthrie DW. Pathophysiology of diabetes mellitus. Crit Care Nurs Q 2004; 27 : 113-25.

30. Leyva F, Godsland IF, Ghatei M, Proudler AJ, Aldis S, Wallon C, et al. Hyperleptinemia as a component of a metabolic syndrome of cardiovascular risk. Arterioscler Thromb Vasc Biol 1998; 18 : 928- 33.

31. Folsom AR, Kushi LH, Anderson KE, Mink PJ, Olson JE, Hong CP, et al. Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women’s Health Study. Arch Intern Med 2000; 160 : 2117-28.

32. Anderson PJ, Critchley JA, Chan JC, Cockram CS, Lee ZS, Thomas GN, et al. Factor analysis of the metabolic syndrome: obesily vs insulin resistance as the central abnormalily. Int J Obes Relat Metab Disord 2001; 25 : 1782-8.

Chung-Ze Wu, Jiunn-Diann Lin, Jer-Chuan Li**, Fone-Ching Hsiao*, Chang-Hsun Hsieh*, Shi-Wen Kuo, Yi-Jen Hung*, Chieh-Hua Lu*, Chih- Tsueng He* & Dee Pei[dagger]

Department of Internal Medicine, Buddhist Tzu Chi General Hospital, Taipei Brack, * Department of Internal Medicine, Tri- Service General Hospital, Taipei, ** Department of Internal Medicine, Buddhist TZU Chi General Hospital, Hualien & [dagger] Department of Internal Medicine, Cardinal Tien Hospital, Medical School, Fu Jen Catholic University, Taiwan, ROC

Received November 21, 2006

Reprint requests: Dr Dee Pei, Department of Internal Medicine, Cardinal Tien Hospital, No 362, Chung Cheng Rd.,

Hsintien Taipei Hsien 23137, Taiwan, R.O.C.

e-mail: peidee@gmail.com

Copyright Indian Council of Medical Research Apr 2008

(c) 2008 Indian Journal of Medical Research. Provided by ProQuest LLC. All rights Reserved.