|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 1999; 33:612-619 © 1999 by the American College of Cardiology Foundation |
* Preventive Cardiology and Therapeutics Research Program, Hamilton Civic Hospitals Research Centre, Hamilton, Ontario, Canada
Divisions of Endocrinology and Metabolism, McMaster University, Hamilton, Ontario, Canada
Division of Cardiology, McMaster University, Hamilton, Ontario, Canada
¶ Department of Medicine, St. John’s Medical College, Bangalore, India
Manuscript received May 5, 1998; revised manuscript received September 18, 1998, accepted October 30, 1998.
Reprint requests and correspondence: Dr. H. C. Gerstein, Department of Medicine, Room 3V38, 1200 Main Street West, Hamilton, Ontario, L8N 3Z5, Canada
gerstein{at}fhs.csu.mcmaster.ca
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
To assess the relationship between dysglycemia and myocardial infarction in nondiabetic individuals.
BACKGROUND
Nondiabetic hyperglycemia may be an important cardiac risk factor. The relationship between myocardial infarction and glucose, insulin, abdominal obesity, lipids and hypertension was therefore studied in South Asians—a group at high risk for coronary heart disease and diabetes.
METHODS
Demographics, waist/hip ratio, fasting blood glucose (FBG), insulin, lipids and glucose tolerance were measured in 300 consecutive patients with a first myocardial infarction and 300 matched controls.
RESULTS
Cases were more likely to have diabetes (OR 5.49; 95% CI 3.34, 9.01), impaired glucose tolerance (OR 4.08; 95% CI 2.31, 7.20) or impaired fasting glucose (OR 3.22; 95% CI 1.51, 6.85) than controls. Cases were 3.4 (95% CI 1.9, 5.8) and 6.0 (95% CI 3.3, 10.9) times more likely to have an FBG in the third and fourth quartile (5.2–6.3 and >6.3 mmol/l); after removing subjects with diabetes, impaired glucose tolerance and impaired fasting glucose, cases were 2.7 times (95% CI 1.5–4.8) more likely to have an FBG >5.2 mmol/l. A fasting glucose of 4.9 mmol/l best distinguished cases from controls (OR 3.42; 95% CI 2.42, 4.83). Glucose, abdominal obesity, lipids, hypertension and smoking were independent multivariate risk factors for myocardial infarction. In subjects without glucose intolerance, a 1.2 mmol/l (21 mg/dl) increase in postprandial glucose was independently associated with an increase in the odds of a myocardial infarction of 1.58 (95% CI 1.18, 2.12).
CONCLUSIONS
A moderately elevated glucose level is a continuous risk factor for MI in nondiabetic South Asians with either normal or impaired glucose tolerance.
| ||||||||||||||||||||||||
South Asians constitute one sixth of humanity and experience much higher rates of coronary heart disease than other ethnic groups in the world (1–6). The World Bank estimates that the death rate from coronary heart disease will increase dramatically in the Indian subcontinent and is expected to contribute to more quality-adjusted life years lost over the next 20 years than in any other part of the world (7).
South Asians also have up to a five-fold (8,9) higher rate of type 2 diabetes mellitus, as well as a higher rate of glucose intolerance, low HDL, high triglycerides and abdominal obesity than people of European ancestry. Despite the high prevalence of both CAD and metabolic risk factors for CAD in the South Asian Indian population, there are sparse data linking the two in these populations. This article reports the relationship between myocardial infarction and body fat distribution, hypertension, glucose tolerance status, glucose, insulin and lipid levels in a case-control study completed in India.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
7.0 mmol/l (126 mg/dl) or 11.1 mmol/l (200 mg/dl) respectively (11). A subject was classified as having impaired glucose tolerance (IGT) if the PPBG was 7.78 mmol/l (140 mg/dl) and <11.1 mmol/l (200 mg/dl) and the FBG was <7.0 mmol/l (126 mg/dl); impaired fasting glucose (IFG) was diagnosed if the FBG was 6.1 mmol/l (110 mg/dl) and <7.0 mmol/l (126 mg/dl). Subjects without known diabetes who did not have a glucose tolerance test were classified on the basis of their fasting glucose alone. Variables measured. Age, gender, religion, monthly income, educational level, dietary details, smoking habits, alcohol use and a history of diabetes mellitus and hypertension were recorded for all subjects. Weight, height, and waist and hip circumference were determined for each subject; waist circumference was measured at the narrowest diameter between the costal margin and the iliac crest, and hip circumference was measured at the greatest diameter over the glutei.
Fasting total cholesterol, HDL cholesterol, low density lipoprotein (LDL) cholesterol, and total triglyceride were measured within 24 hours of the onset of chest pain for cases and prior to any surgery for controls. Fasting blood glucose in plasma was measured on the 9th or 10th day after admission for cases and at the time of lipid measurement for controls. In addition those subjects without a previous history of diabetes had blood drawn for fasting insulin levels at this time; they then had a PPBG measured two hours after drinking 75 grams of glucose. Patients were not on any intravenous therapy when glucose levels were measured.
Total cholesterol and triglycerides were estimated by enzymatic methods (cholesterol oxidase/peroxidase-aminophenazone for cholesterol, glycerol phosphate oxidase/peroxidase-aminophenazone for triglycerides) on an automated system using standard kits (Boehringer Mannheim Gmbh). High density lipoprotein cholesterol was estimated using a precipitation method and LDL cholesterol was calculated (total cholesterol-HDL cholesterol-triglycerides/5). Glucose was assayed by the glucose oxidase method and serum insulin was estimated by radioimmunoassay (kit manufactured by the Board of Radiation and Isotope Technology, Bombay). All biochemical analyses were conducted without knowledge of the clinical information.
Statistical methods. For the analyses in which all cases were compared to all controls, cases were matched with the next control subject recruited and estimates of odds ratios (ORs) and 95% confidence intervals were based on analyses of matched-pair data. Unmatched analyses were used to compare nondiabetic or diabetic cases to nondiabetic or diabetic controls because cases were not matched to controls on the basis of diabetes status. Because of the skewed distribution of continuous variables, nonparametric univariate analyses were used. Univariate comparisons of frequencies were done using chi square tests.
For multivariate model building, all nonnormal variables were transformed and stepwise multiple logistic regression analyses were performed to identify the best independent determinants of myocardial infarction (MI). Case-control matching was used for analyses of all cases and controls. For nondiabetic and non-IGT subgroups, an unmatched model was used with age and sex added as covariates. Based on these logistic regression models, cutpoints for the receiver operating curve (ROC) analyses were those that optimized the cross-product ratio.
The following variables were included in all of the multivariate analyses: fasting glucose, postprandial glucose (in nondiabetic subjects), waist/hip ratio, cholesterol, triglycerides, HDL cholesterol, LDL cholesterol, cholesterol/HDL and smoking status. In addition, diabetes status (yes/no) was entered in the regression analysis of all cases and controls; IGT status (yes/no), postprandial glucose and insulin level in the regression analysis of nondiabetic cases and controls; and impaired fasting glucose status (yes/no), postprandial glucose and insulin levels in the regression analysis of non-IGT cases and controls. BMDP LR (stepwise logistic regression) with a p-value for entry of a variable into the model of p = 0.10 and p = 0.15 to remove variables was used for this analysis.
Figures illustrating the unadjusted relationship between glucose values and odds ratios for all subjects and for nondiabetic subjects were constructed from univariate logistic regression coefficients. The SAS and BMDP statistical packages were used for all statistical analyses. p values are two-tailed.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
Continuous relationship between glucose and risk of myocardial infarction.
The odds ratios for MI in all subjects increased with increasing FBG quartiles (Fig. 1) and was greater than 1 at FBG levels clearly below the nondiabetic and impaired glucose tolerance range. For example, cases were 3.4 times (95% CI 1.9–5.8) more likely to have an FBG between 5.2 mmol/l (94 mg/dl) and 6.3 mmol/l (114 mg/dl) than controls. This continuous relationship was maintained in the analysis of nondiabetic subjects and was also clearly maintained after removing all subjects with diabetes, IGT and IFG (Fig. 1D). Compared to subjects with an FBG 
4.5 mmol/l (81 mg/dl), the odds ratio for MI in subjects in this latter group with an FBG of 5.2 to 6.3 mmol/l (94 to 114 mg/dl) was 2.7 (95% CI 1.5–4.8).
|
4.9 mmol/l (89 mg/dl); these patients were 3.42 times more likely to be a case than a control (95% CI 2.42–4.83). Similar analyses for the postprandial glucose level yielded a cutpoint of 6.8 mmol/l (123 mg/dl; OR = 3.84; 95% CI 2.59–5.67). When the analysis was restricted to nondiabetic patients, similar fasting and postprandial glucose cutpoints of 5.1 mmol/l (91 mg/dl) and 7.1 mmol/l (128 mg/dl) were obtained (Table 3). A waist/hip ratio 0.89 best differentiated all cases from controls (OR = 3.24; 95% CI = 2.24–4.69; 81% of cases’ waist/hip ratios were 0.89). No fasting insulin level clearly separated cases from controls.
|
|
Because PPBG was not measured in the diabetic subjects, this variable was not included in this analysis. Similar findings were noted when the analysis was restricted to nondiabetic subjects and subjects without impaired glucose tolerance, in whom both FBG and PPBG as well as fasting insulin levels were measured and entered into the regression in addition to all of the other variables listed above. Insulin levels were not independent variables in any model tested.
Table 4 lists the odds ratios for an increase in 1 standard deviation of the variables included in the regression models for all subjects, nondiabetic subjects and non-IGT subjects. After adjustment for smoking status, waist/hip ratio and triglyceride level, an increase in PPBG of 1.2 mmol/l or 21 mg/dl (1 standard deviation) increased the odds of an MI 1.58 times (95% CI 1.18, 2.12) in non-IGT subjects.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Glucose levels and cardiovascular risk. These results are consistent with data from other populations suggesting a progressive relationship between glucose levels and an increased risk of cardiovascular disease in both nondiabetic (14–17) and diabetic patients (18–21). They are also consistent with a metaregression analysis of all cohort studies of nondiabetic patients in which baseline glucose levels were related to subsequent cardiovascular events and mortality (22). The observation that the glucose-associated risk persisted even after controlling for other risk factors suggests that glucose elevation is an independent marker for atherosclerosis in South Asians.
Lipid and insulin levels and cardiovascular risk. The fact that the cholesterol/HDL ratio was an independent determinant of MI in the multivariate analysis of all subjects and the fact that triglyceride levels were not an independent determinant of MI is consistent with other epidemiologic studies (23,24). The fact that this ratio was not an independent determinant of MI in the nondiabetic subjects may have been due to the lower power to detect such a relationship in this smaller subgroup. In the univariate analysis of the nondiabetic subgroup, triglyceride levels were nonsignificantly higher in cases than in controls (data not shown). After adjustment for other risk factors, however, there was an inverse correlation in this subgroup. This may have been due to the fact that some of the MI case subjects were treated with intravenous heparin—a therapy known to lower triglyceride levels (25).
The observation that insulin levels were not independent risk factors for MI in the logistic regression suggests that the moderately higher insulin levels observed in cases may have occurred in response to the elevated glucose levels (26,27). It is also consistent with epidemiologic data from other populations suggesting that hyperinsulinemia alone may not be a strong determinant of cardiovascular disease (28). The possibility that the preinfarct hyperinsulinemia may have been minimized in post-MI cases (at the time of insulin sampling), however, cannot be ruled out.
Waist/hip ratio and cardiovascular risk. The waist/hip ratio strongly discriminated cases from controls and was a strong independent risk factor for MI in this population. Conversely, body mass index had no discriminative value. Evidence that a high waist/hip ratio reflects visceral fat accumulation, and that visceral fat is associated with abnormalities in glucose and fatty acid metabolism suggest that it is a surrogate measure for an atherogenic metabolic state (29–31). Moreover, the independence of the high waist/hip ratio in the logistic regression suggests that this ratio reflects metabolic abnormalities that are distinct from those related to the higher glucose level. The nature of these abnormalities is currently unknown.
Limitations of the case-control design. These data are limited by the case-control design. First, only one glucose tolerance test was done in cases and controls who did not have a history of diabetes. Even under ideal circumstances the postprandial (2 h) glucose result in this test is not highly reproducible (32). As this variability may be even more pronounced in patients who are within two weeks of a myocardial infarction, the classification of cases into those with diabetes and IGT according to postprandial glucose values may not be robust. Second, the postprandial glucose level is increased in people who are inactive or who are not consuming a high carbohydrate diet for several days prior to the test (33). It may also have been affected by drugs prescribed in the post-MI period. This may have magnified the elevation in postprandial glucose levels seen in nondiabetic cases (in whom glucose tolerance tests was done 9 to 10 days after admission for myocardial infarction) compared with nondiabetic controls (in whom glucose tolerance tests were done within 24 h of admission) and may have overestimated the number of cases who had (previously undiagnosed) diabetes. Despite these possibilities, the detection of clear and consistent differences in the distribution of the FBG (which are more robust than the PPBG [32,33]) as well as the PPBG among cases and controls regardless of diagnosis (i.e., in all subjects and in those without diabetes or IGT) strongly supports the inference that dysglycemia is an important and independent cardiac risk factor in this population.
Implications and conclusions. The finding of a graded risk of MI with glucose elevations within the "normal" range in South Asians strongly supports the need to explore this relationship in other ethnic groups. If these observations are confirmed, the population attributable risk of dysglycemia (i.e., the excess risk of MI in the general population attributable to glucose elevations above some low dysglycemic threshold) may be several times greater than the population attributable risk of diabetes alone. This would focus attention on the high prevalence of elevated glucose levels in the nondiabetic population and may lead to innovative ways of preventing cardiovascular disease in this group.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. J. HOOGWERF Does intensive therapy of type 2 diabetes help or harm? Seeking accord on ACCORD. Cleveland Clinic Journal of Medicine, October 1, 2008; 75(10): 729 - 737. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Gupta, P Joshi, V Mohan, K S Reddy, and S Yusuf Epidemiology and causation of coronary heart disease and stroke in India Heart, January 1, 2008; 94(1): 16 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. G Noordzij, E. Boersma, F. Schreiner, M. D Kertai, H. H H Feringa, M. Dunkelgrun, J. J Bax, J. Klein, and D. Poldermans Increased preoperative glucose levels are associated with perioperative mortality in patients undergoing noncardiac, nonvascular surgery Eur. J. Endocrinol., January 1, 2007; 156(1): 137 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. R. Gross, A. K. Hsu, and G. J. Gross Diabetes Abolishes Morphine-Induced Cardioprotection via Multiple Pathways Upstream of Glycogen Synthase Kinase-3{beta} Diabetes, January 1, 2007; 56(1): 127 - 136. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. T. Bloomgarden, R. Dodis, C. M. Viscoli, E. S. Holmboe, and S. E. Inzucchi Lower Baseline Glycemia Reduces Apparent Oral Agent Glucose-Lowering Efficacy: A meta-regression analysis. Diabetes Care, September 1, 2006; 29(9): 2137 - 2139. [Full Text] [PDF]
|
|
|
|
|
|
A. Ramachandran, S. Chamukuttan, S. Immaneni, R. M. Shanmugam, N. Vishnu, V. Viswanathan, and T. Jaakko High Incidence of Glucose Intolerance in Asian-Indian Subjects With Acute Coronary Syndrome Diabetes Care, October 1, 2005; 28(10): 2492 - 2496. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. TAYLOR, C. YOUNG, R. ESOP, C. PATON, and R. WALWYN Testing for diabetes in hospitalised patients prescribed antipsychotic drugs The British Journal of Psychiatry, August 1, 2004; 185(2): 152 - 156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. Blake, A. D. Pradhan, J. E. Manson, G. R. Williams, J. Buring, P. M. Ridker, and R. J. Glynn Hemoglobin A1c Level and Future Cardiovascular Events Among Women Arch Intern Med, April 12, 2004; 164(7): 757 - 761. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Prabhakaran and S. S Anand The metabolic syndrome: an emerging risk state for cardiovascular disease Vascular Medicine, February 1, 2004; 9(1): 55 - 68. [Abstract] [PDF]
|
|
|
|
 |
|
 T. F. Luscher, M. A. Creager, J. A. Beckman, and F. Cosentino Diabetes and Vascular Disease: Pathophysiology, Clinical Consequences, and Medical Therapy: Part II Circulation, September 30, 2003; 108(13): 1655 - 1661. [Full Text] [PDF]
|
|
|
|
|
|
J. E. Gerich Clinical Significance, Pathogenesis, and Management of Postprandial Hyperglycemia Arch Intern Med, June 9, 2003; 163(11): 1306 - 1316. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
References Circulation, December 17, 2002; 106(25): 3373 - 3421. [Full Text]
|
|
|
|
 |
|
 N. Mikhail, J. A. Beckman, M. A. Creager, and P. Libby Postprandial Hyperglycemia and Risk of Atherosclerosis JAMA, August 28, 2002; 288(8): 955 - 955. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Tanaka, F. Kehl, W. Gu, J. G. Krolikowski, P. S. Pagel, D. C. Warltier, and J. R. Kersten Isoflurane-induced preconditioning is attenuated by diabetes Am J Physiol Heart Circ Physiol, June 1, 2002; 282(6): H2018 - H2023. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. A. Lell, V. G. Nielsen, D. C. McGiffin, F. E. Schmidt Jr, J. K. Kirklin, and A. W. Stanley Jr Glucose-insulin-potassium infusion for myocardial protection during off-pump coronary artery surgery Ann. Thorac. Surg., April 1, 2002; 73(4): 1246 - 1251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Sacks, D. E. Bruns, D. E. Goldstein, N. K. Maclaren, J. M. McDonald, and M. Parrott Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus Clin. Chem., March 1, 2002; 48(3): 436 - 472. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Yusuf, S. Reddy, S. Ounpuu, and S. Anand Global Burden of Cardiovascular Diseases: Part I: General Considerations, the Epidemiologic Transition, Risk Factors, and Impact of Urbanization Circulation, November 27, 2001; 104(22): 2746 - 2753. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Mohan, R. Deepa, S. Shanthi Rani, and G. Premalatha Prevalence of coronary artery disease and its relationship to lipids in a selected population in South India: The Chennai Urban Population Study (CUPS No. 5) J. Am. Coll. Cardiol., September 1, 2001; 38(3): 682 - 687. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Raskin, M. Rendell, M. C. Riddle, J. F. Dole, M. I. Freed, and J. Rosenstock A Randomized Trial of Rosiglitazone Therapy in Patients With Inadequately Controlled Insulin-Treated Type 2 Diabetes Diabetes Care, July 1, 2001; 24(7): 1226 - 1232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Kersten, W. G. Toller, E. R. Gross, P. S. Pagel, and D. C. Warltier Diabetes abolishes ischemic preconditioning: role of glucose, insulin, and osmolality Am J Physiol Heart Circ Physiol, April 1, 2000; 278(4): H1218 - H1224. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. SELL, N. R. KLEINMAN, and V. M. MONNIER Longitudinal determination of skin collagen glycation and glycoxidation rates predicts early death in C57BL/6NNIA mice FASEB J, January 1, 2000; 14(1): 145 - 156. [Abstract] [Full Text]
|
|
|
|
|
|
N. Hosszufalusi, P. Panczel, L. Janoskuti, M. Pyorala, H. Miettinen, M. Laakso, and K. Pyorala Hyperinsulinemia Predicts Coronary Heart Disease Risk in Healthy Middle-Aged Men • Response Circulation, December 14, 1999; 100 (24): e118 - e118. [Full Text] [PDF]
|
|
|
|
|
|
K. Tanaka, F. Kehl, W. Gu, J. G. Krolikowski, P. S. Pagel, D. C. Warltier, and J. R. Kersten Isoflurane-induced preconditioning is attenuated by diabetes Am J Physiol Heart Circ Physiol, June 1, 2002; 282(6): H2018 - H2023. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
