Search tips
Search criteria 


Logo of diacareAmerican Diabetes AssociationSubscribeSearchDiabetes Care Journal
Diabetes Care. 2009 September; 32(9): 1721–1726.
Published online 2009 June 5. doi:  10.2337/dc08-2337
PMCID: PMC2732157

Postprandial Glucose Improves the Risk Prediction of Cardiovascular Death Beyond the Metabolic Syndrome in the Nondiabetic Population

Hung-Ju Lin, MD,1,2 Bai-Chin Lee, MD,1,2 Yi-Lwun Ho, MD, PHD,1 Yen-Hung Lin, MD,1 Ching-Yi Chen, PHD,3 Hsiu-Ching Hsu, PHD,1 Mao-Shin Lin, MD,1 Kuo-Liong Chien, MD, PHD,1,4 and Ming-Fong Chen, MD, PHD1



With increasing evidence about the cardiovascular risk associated with postprandial nonfasting glucose and lipid dysmetabolism, it remains uncertain whether the postprandial glucose concentration increases the ability of metabolic syndrome to predict cardiovascular events.


This was an observational study of 15,145 individuals aged 35–75 years without diabetes or cardiovascular diseases. Postprandial glucose was obtained 2 h after a lunch meal. Metabolic syndrome was diagnosed using the criteria of the U.S. National Cholesterol Education Program Adult Treatment Panel III. Cardiovascular and all-cause deaths were primary outcomes.


During a median follow-up of 6.7 years, 410 individuals died, including 82 deaths from cardiovascular causes. In a Cox model adjusting for metabolic syndrome status as well as age, sex, smoking, systolic blood pressure, LDL, and HDL cholesterol levels, elevated 2-h postprandial glucose increased the risk of cardiovascular and all-cause death (per millimole per liter increase, hazard ratio 1.26 [95% CI 1.11–1.42] and 1.10 [1.04–1.16], respectively), with significant trends across the postprandial glucose quintiles. Including 2-h postprandial glucose into a metabolic syndrome–included multivariate risk prediction model conferred a discernible improvement of the model in discriminating between those who died of cardiovascular causes and who did not (integrated discrimination improvement 0.4, P = 0.005; net reclassification improvement 13.4%, P = 0.03); however, the improvement was only marginal for all-cause death.


Given the risk prediction based on metabolic syndrome and established cardiovascular risk factors, 2-h postprandial glucose improves the predictive ability to identity nondiabetic individuals at increased risk of cardiovascular death.

Proposed in 1988 by Reaven (1), the construct that impaired glucose regulation and insulin resistance to glucose stimulation frequently coexist with elevated blood pressure and dyslipidemia has been evolving into several diagnostic criteria of metabolic syndrome (24), a well-established risk for development of cardiovascular events and type 2 diabetes. Of those, the definitions recommended by the U.S. National Cholesterol Education Program Adult Treatment Panel III (ATP III) (2) and the International Diabetes Federation (IDF) (3) are the most widely used, partly because only fasting glucose and lipid concentrations are required. It has been increasingly recognized that postprandial glucose might be a better risk predictor for cardiovascular events than fasting glucose in nondiabetic and diabetic populations alike (5). A study using confirmatory factor analysis indicated that postprandial glucose appeared to be more correlated with metabolic syndrome than fasting glucose (6). However, concerns have not been fully addressed for whether elevated postprandial glucose would increase the risk of adverse outcomes of metabolic syndrome diagnosed according to the ATP III or IDF criteria.

Although postprandial glucose is traditionally obtained using an oral glucose tolerance test (OGTT), accumulating evidence shows that postprandial glucose after a mixed meal might be more likely to reflect physiological responses to a glucose challenge in daily life than an OGTT (7,8). Therefore, the aim of this study was to examine the relationship of postmeal postprandial glucose with cardiovascular and all-cause deaths in nondiabetic individuals.


A prospective observational cohort enrolled 17,299 self-referred individuals aged 30–75 years, who attended health examinations at the National Taiwan University Hospital between January 1995 and December 2002. At baseline, structured questionnaires were used to collect family history, current smoking status, and medical histories. We excluded individuals with diabetes or with a medical history of myocardial infarction, coronary heart disease, stroke, carotid artery stenosis, or peripheral artery disease at baseline. Diabetes was defined if one of following existed: current use of antidiabetic agents, diet modifications because of diabetes, fasting glucose ≥7.0 mmol/l (126 mg/dl), or 2-h postprandial glucose ≥11.1 mmol/l (200 mg/dl). Also excluded were those who had incomplete baseline data, missing fasting glucose or 2-h postprandial glucose data, severely impaired renal function (serum creatinine levels ≥530 μmol/l [6 mg/dl]), or comorbidity with cancer. Accordingly, we included a total of 6,024 women and 9,121 men for the analysis. The study protocol was approved by the hospital's institutional review board.

Laboratory measurements

Venous blood samples were collected from each participant after at least a 12-h overnight fast, and 2-h postprandial glucose was obtained 2 h after a standard 650-kcal lunch meal, which was composed of 110 g carbohydrates, 27–30 g protein, and 8–10 g fat. Total cholesterol levels were measured using the CHOD-PAP method (Boehringer Mannheim, Mannheim, Germany), and HDL cholesterol was measured after precipitation of apolipoprotein B–containing lipoproteins, with phosphotungstic acid and magnesium ions (Boehringer Mannheim). Triglyceride concentrations were measured by the GPO-DAOS method (Wako Pure Chemicals, Tokyo, Japan). LDL cholesterol concentrations were calculated using the Friedewald formula. Plasma fasting glucose and 2-h postprandial glucose concentrations were measured using a Hitachi 7450 automated analyzer (Hitachi, Tokyo, Japan). A1C was measured using a DCA 2000 analyzer (Bayer Diagnostics, Elkhart, IN).

Definition of metabolic syndrome and postprandial hyperglycemia

Of the criteria available for diagnosis of metabolic syndrome, the two recommended by the IDF and ATP III, respectively, do not require postprandial glucose as a clinical measurement of insulin resistance. Because a previous study has proved substantial agreement between the ATP III and IDF criteria (9), we did not assess the effects of the IDF criteria on the results of this study.

Metabolic syndrome was defined using the modified ATP III definition (9). Individuals with three or more of the following five criteria are regarded as having metabolic syndrome: 1) BMI of ≥27 kg/m2 for an Asian population (10); 2) systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg; 3) fasting plasma glucose ≥6.1 mmol/l (100 mg/dl); 4) triglyceride level ≥1.69 mmol/l (150 mg/dl); and 5) HDL cholesterol level <1.29 mmol/l (50 mg/dl) for women or <1.04 mmol/l (40 mg/dl) for men. Individuals with a 2-h postprandial glucose of ≥7.8 mmol/l (140 mg/dl) were defined as having postprandial hyperglycemia, because those with normal glucose regulation rarely have a plasma glucose level of ≥7.8 mmol/l (140 mg/dl) or higher (11).

Ascertainment of the causes of death

Mortality and causes of death were ascertained according to a national register database, which was provided by the National Health Administration, and was updated at the end of 2007, by linking identification numbers of individuals to the database. Cardiovascular death was defined as deaths from cardiovascular causes if the registered death codes ranged from 390 to 459, based on ICD-9.

Statistical analysis

Data are summarized as means ± SD for continuous variables and as percentages for categorical variables. Baseline characteristics are represented by quintiles of 2-h postprandial glucose, and differences among quintile groups were tested with ANOVA for mean values and by the χ2 test for categorical values. The Cox proportional hazards regression model was applied to assess the association of 2-h postprandial glucose with the incidence of cardiovascular death and all-cause death after adjustment for covariates as follows: model 1, age-groups (35–44, 45–54, 55–64, 65–74, and ≥75 years old), sex, and current smoker (yes/no); model 2, those in model 1 plus systolic blood pressure (quintile groups), HDL cholesterol (quintile groups), and LDL cholesterol (quintile groups); and model 3, metabolic syndrome (presence/absence) was added into model 2. P values for linear trend were calculated with the median from each quintile group.

To further evaluate the improvement in predictive ability of metabolic syndrome for cardiovascular and all-cause death after 2-h postprandial glucose was added to multivariate Cox models, we compared the measures of discrimination and calibration between the models with and without 2-h postprandial glucose. Three methods were used to assess the extent to which 2-h postprandial glucose increased the discriminative performance of risk prediction. The first was to compare the differences of the area under the receiver operating characteristic curves (AUC) between risk prediction models with and without 2-h postprandial glucose. Receiver operating characteristic curves were constructed by plotting sensitivity versus 1 − specificity obtained using multivariate logistic regression, and AUCs were compared using a nonparametric test (12).

Because AUC values may not be sensitive enough to express the improvement of discrimination performance, we also used two other methods: net reclassification improvement (NRI) and integrated discrimination improvement (IDI) as described by Pencina et al. (13). Based on the prior predictive risk, individuals were classified into four risk categories of 0–1%, 1–3%, 3–6%, and ≥6% derived from a probability distribution of 5-year risk for cardiovascular death (14) and then were reclassified according to the predictive risks in regression models including 2-h postprandial glucose. The NRI was the difference between the percentage of reclassifying those with events into higher risk categories and those without events into lower risk categories (indicating improvement in the discrimination performance) and the percentage of reclassifying those with events into lower risk categories and those without events into higher risk categories (indicating failure to improve the discrimination performance). The overall reclassification improvement was statistically examined by an asymptotic test (13). Similarly, individuals were grouped into four categories of 0–10%, 10 −25%, 25–35%, and ≥35%, which were defined on the assumption that cardiovascular death was accountable for ~15% of all-cause death according to the 2006 mortality statistics from the World Health Organization. Because the clinical consensus on risk cutoffs for cardiovascular and all-cause death has not been established, the IDI was applied to estimate the discrimination improvement in average sensitivity and any potential increase in average 1 − specificity and was examined using an asymptotic test (13).

Model calibration was assessed with the Hosmer-Lemeshow goodness-of-fit χ2 statistic, a measure of discrepancy describing how well the predicted probabilities fit observed outcomes. A model with a smaller Hosmer-Lemeshow statistic is better calibrated, whereas a model with a statistic of >20 is regarded as having a lack of calibration (15).

The metabolic syndrome by a postprandial hyperglycemia interaction term was used in a multivariate Cox model to explore the effect modification of metabolic syndrome with postprandial hyperglycemia on risk of cardiovascular and all-cause death, and multivariate-adjusted relative risks of primary outcomes were plotted by stratification.

All statistical tests were two-tailed with a type I error of 0.05, and P < 0.05 was regarded as statistically significant. All of the statistical analyses were performed with SAS (version 9.1; SAS Institute, Cary, NC) and Stata (version 10; StataCorp, College Station, TX).


Baseline characteristics

As summarized in Table 1, with increasing quintiles of 2-h postprandial glucose, individuals were more likely to have a higher prevalence of ATP III metabolic syndrome, higher total cholesterol, triglycerides, LDL cholesterol, fasting glucose, and A1C, but lower HDL cholesterol.

Table 1
Baseline characteristics among 6,024 women and 9,121 men according to the quintiles of 2-h postprandial glucose

Postprandial blood glucose as a predictor for cardiovascular and all-cause death independently of metabolic syndrome

Of the 15,145 nondiabetic participants, 410 died during a median follow-up of 6.7 years (interquartile range 3.3 years), including 82 deaths from cardiovascular causes. As outlined in Table 2, after adjustment for sex, age, smoking status, systolic blood pressure, and LDL and HDL cholesterol in a Cox model, elevated 2-h postprandial glucose was associated with increased risks of cardiovascular and all-cause death (hazard ratio 1.26 [95% CI 1.11–1.43] for cardiovascular death; 1.10 [1.03–1.16] for all-cause death). Furthermore, additional adjustment for the presence or absence of metabolic syndrome maintained the independent association between 2-h postprandial glucose and risks of cardiovascular and all-cause death. Moreover, the trends toward heightened risks of primary outcomes were also found to be significant across the increasing quintiles of 2-h postprandial glucose. Compared with the absence of postprandial hyperglycemia (Fig. 1), the presence of postprandial hyperglycemia raised the relative risk of 2.22 (95% CI 1.40–3.52) for cardiovascular death and of 1.56 (1.25–1.94) for all-cause death. There was no evidence of substantial heterogeneity in relative risks of primary outcomes associated with postprandial hyperglycemia between subgroups with or without metabolic syndrome.

Table 2
Multivariate-adjusted associations of 2-h postprandial glucose with cardiovascular and all-cause death
Figure 1
Association of postprandial hyperglycemia with the risk of cardiovascular and all-cause death according to the presence or absence of metabolic syndrome. Postprandial hyperglycemia was defined as 2-h postprandial glucose ≥7.8 mmol/l (140 mg/dl). ...

Comparison in discrimination and calibration between metabolic syndrome–included multivariate prediction models before and after adding 2-h postprandial glucose

Table 3 and supplementary Tables A1 and A2 (available in an online appendix at present a summary of the statistics for the evaluation of the model discrimination performance and calibration. Although AUC values marginally increased after 2-h postprandial glucose was incorporated in the metabolic syndrome–included multivariate-adjusted models (Table 3), we observed significant improvements using the NRI and IDI methods. Using the predicted risk obtained from the metabolic syndrome–included prediction model incorporating 2-h postprandial glucose (supplementary Table A1), we found that, of 82 cardiovascular deaths, 17 individuals with events were reclassified into higher risk categories, whereas 6 were classified into lower risk categories; accordingly, the reclassification improvement was 13.4% ([17 − 6/82] × 100%). However, in 14,901 individuals who did not die of cardiovascular disease (supplementary Table A2), 558 were designated as lower risk and 567 as higher risk, thereby indicating 0.0% ([558 − 567/14,901] × 100%) in reclassification improvement. Therefore, the NRI was 13.4% (95% CI 2.2–24.5, P = 0.02), suggesting that the risk prediction for cardiovascular death improved after 2-h postprandial glucose was added into the model. Although the clinical implication of the absolute value of IDI awaits further clarification, the improvement in discrimination performance in the prediction model for cardiovascular death after addition of 2-h postprandial glucose was also demonstrated by the estimate of IDI of 0.4% (95% CI 0.1–0.7, P = 0.005), which might, however, be more pertinent than NRI when a consensus on clinical risk categories has not be established. Likewise, for all-cause death, NRI was 3.6% (95% CI 1.2–6.0, P = 0.003) (supplementary Table A2), and IDI was 0.2% (0.0–0.3, P = 0.03), indicating that the addition of 2-h postprandial glucose into prediction models would marginally improve discriminative ability of the models.

Table 3
Improvement in discrimination performance and calibration for risk prediction of cardiovascular death and all-cause death in the multivariate-adjusted model after including 2-h postprandial glucose

Assessed by the use of the Hosmer-Lemeshow test, the models with 2-h postprandial glucose had smaller a χ2 statistic than those without 2-h postprandial glucose, for predicting both cardiovascular death and all-cause death (Table 3). Moreover, all models with and without 2-h postprandial glucose represented adequate calibration.


In this study we investigated the independent and incremental utility of 2-h postprandial glucose in risk prediction for cardiovascular and all-cause death in the nondiabetic population. Our data demonstrated that increased 2-h postprandial glucose was an independent risk predictor of cardiovascular and all-cause death and that the addition of 2-h postprandial glucose into risk prediction, which includes metabolic syndrome status and established cardiovascular risk factors, could improve the predictive ability of the risk for cardiovascular death.

Postprandial hyperglycemia differs from fasting hyperglycemia in pathophysiology and the risk of adverse outcomes. Postprandial hyperglycemia results mainly from a moderate to severe decrease in muscle insulin sensitivity and late-phase insulin secretory response to oral glucose; whereas fasting hyperglycemia results mainly from a defect in hepatic insulin sensitivity and first-phase insulin response (16). Our findings concur with those of a meta-analysis of 38 prospective studies, which demonstrated that nondiabetic postprandial hyperglycemia significantly raises the likelihood of cardiovascular events or death and that the relationship remains after cardiovascular risk factors have been adjusted for (17). The underlying pathobiological changes linking elevated postprandial glucose and cardiovascular diseases consist mainly of elevated oxidative stress and endothelial dysfunction, leading to atherothombotic propensities of elevated postprandial glucose, such as oxidation of LDL cholesterol, vasoconstriction, and thrombogenicity (18). Moreover, interventions targeting elevated postprandial glucose are of clinical significance for optimizing comprehensive glycemic control in individuals with diabetes and with impaired glucose regulation to lower the risk of adverse outcomes (19).

It has been shown that the mixed-meal test is well correlated with the OGTT in measures of insulin sensitivity and is more likely to have a stronger pancreatic β-cell response to an equal amount of carbohydrate than the OGTT (7), perhaps suggesting that a mixed-meal test could be more suitable for evaluating the reserve of insulin secretion. Evidence has emerged about the association between postmeal glucose concentrations and the cardiovascular risk. In women with normal glucose regulation, those with elevated postmeal postprandial glucose tend to have increased carotid intima medial thickness (20). In a recent study assessing the relationship between postmeal glucose and incident cardiovascular complications, it was reported that elevated postmeal glucose, but not fasting glucose, predisposes patients with type 2 diabetes to increased cardiovascular risk (21). There is also growing interest in the increased risk of cardiovascular events in relation to postprandial lipid concentrations, which were obtained after normal food intake (22,23).

Some limitations and weakness in this study need careful concern. First, BMI, rather than waist circumference, was used as the obesity criterion of metabolic syndrome in this study. However, findings from previous studies have confirmed that, in this setting, using either BMI or waist circumference leads to no difference in predicting cardiovascular events (24). Second, the NRI method is affected by the risk categories used, and no consensus had existed on clinically significant risk categories of cardiovascular death when our analysis was conducted. Nonetheless, the improvement in reclassification for cardiovascular death was corroborated by the use of IDI, a method with no fixed cutoffs for risk categorization (13). Third, caution should be taken in generalizing our results to a community-based population because this study cohort was based on participants undergoing health check-ups at a university hospital. Fourth, compared with the survey in the general population including individuals with diabetes, the incidence of cardiovascular death seemed to be low in our study, partly because only individuals without diabetes were included. Thus, one should be cautious when the relative risks with wide CIs are considered.

Despite these concerns, our findings provide evidence that elevated postprandial glucose independently posed heightened risks of cardiovascular and all-cause death and that the inclusion of postprandial glucose into risk prediction for cardiovascular death increases the risk predictive ability. Future trials to examine the difference in the risk of adverse outcomes between definitions of metabolic syndrome using fasting or postprandial nonfasting glucose and lipid criteria may be of clinical interest.

Supplementary Material

Online-Only Appendix:


No potential conflicts of interest relevant to this article were reported.

We thank the staff of the Health Management Center, National Taiwan University Hospital, for their assistance in the study.


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.


1. Reaven GM. Banting Lecture 1988: Role of insulin resistance in human disease. Diabetes 1988; 37: 1595– 1607 [PubMed]
2. 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) final report. Circulation 2002; 106: 3143– 3421 [PubMed]
3. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome–a new world-wide definition: a consensus statement from the International Diabetes Federation. Diabet Med 2006; 23: 469– 480 [PubMed]
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– 553 [PubMed]
5. Glucose tolerance and mortality: comparison of WHO and American Diabetes Association diagnostic criteria. The DECODE study group. European Diabetes Epidemiology Group. Diabetes Epidemiology: Collaborative analysis Of Diagnostic criteria in Europe. Lancet 1999, 354: 617– 621 [PubMed]
6. Pladevall M, Singal B, Williams LK, Brotons C, Guyer H, Sadurni J, Falces C, Serrano-Rios M, Gabriel R, Shaw JE, Zimmet PZ, Haffner S. A single factor underlies the metabolic syndrome: a confirmatory factor analysis. Diabetes Care 2006; 29: 113– 122 [PubMed]
7. Rijkelijkhuizen JM, Girman CJ, Mari A, Alssema M, Rhodes T, Nijpels G, Kostense PJ, Stein PP, Eekhoff EM, Heine RJ, Dekker JM. Classical and model-based estimates of β-cell function during a mixed meal vs. an OGTT in a population-based cohort. Diabetes Res Clin Pract 2009; 83: 280– 288 [PubMed]
8. Vollmer K, Holst JJ, Baller B, Ellrichmann M, Nauck MA, Schmidt WE, Meier JJ. Predictors of incretin concentrations in subjects with normal, impaired, and diabetic glucose tolerance. Diabetes 2008; 57: 678– 687 [PubMed]
9. Chien KL, Lee BC, Hsu HC, Lin HJ, Chen MF, Lee YT. Prevalence, agreement and classification of various metabolic syndrome criteria among ethnic Chinese: a report on the hospital-based health diagnosis of the adult population. Atherosclerosis 2008; 196: 764– 771 [PubMed]
10. Huang KC. Obesity and its related diseases in Taiwan. Obes Rev 2008; 9( Suppl. 1): 32– 34 [PubMed]
11. Polonsky KS, Given BD, Van Cauter E. Twenty-four-hour profiles and pulsatile patterns of insulin secretion in normal and obese subjects. J Clin Invest 1988; 81: 442– 448 [PMC free article] [PubMed]
12. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988; 44: 837– 845 [PubMed]
13. Pencina MJ, D'Agostino RB, Sr, D'Agostino RB, Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008; 27: 157– 172 [PubMed]
14. Pocock SJ, McCormack V, Gueyffier F, Boutitie F, Fagard RH, Boissel JP. A score for predicting risk of death from cardiovascular disease in adults with raised blood pressure, based on individual patient data from randomised controlled trials. BMJ 2001; 323: 75– 81 [PMC free article] [PubMed]
15. Ingelsson E, Schaefer EJ, Contois JH, McNamara JR, Sullivan L, Keyes MJ, Pencina MJ, Schoonmaker C, Wilson PW, D'Agostino RB, Vasan RS. Clinical utility of different lipid measures for prediction of coronary heart disease in men and women. JAMA 2007; 298: 776– 785 [PubMed]
16. Nathan DM, Davidson MB, DeFronzo RA, Heine RJ, Henry RR, Pratley R, Zinman B. Impaired fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care 2007; 30: 753– 759 [PubMed]
17. Levitan EB, Song Y, Ford ES, Liu S. Is nondiabetic hyperglycemia a risk factor for cardiovascular disease? A meta-analysis of prospective studies. Arch Intern Med 2004; 164: 2147– 2155 [PubMed]
18. Ceriello A. Impaired glucose tolerance and cardiovascular disease: the possible role of post-prandial hyperglycemia. Am Heart J 2004; 147: 803– 807 [PubMed]
19. Ceriello A, Colagiuri S. International Diabetes Federation guideline for management of postmeal glucose: a review of recommendations. Diabet Med 2008; 25: 1151– 1156 [PMC free article] [PubMed]
20. Alssema M, Schindhelm RK, Dekker JM, Diamant M, Kostense PJ, Teerlink T, Scheffer PG, Nijpels G, Heine RJ. Postprandial glucose and not triglyceride concentrations are associated with carotid intima media thickness in women with normal glucose metabolism: the Hoorn prandial study. Atherosclerosis 2008; 196: 712– 719 [PubMed]
21. Cavalot F, Petrelli A, Traversa M, Bonomo K, Fiora E, Conti M, Anfossi G, Costa G, Trovati M. Postprandial blood glucose is a stronger predictor of cardiovascular events than fasting blood glucose in type 2 diabetes mellitus, particularly in women: lessons from the San Luigi Gonzaga Diabetes Study. J Clin Endocrinol Metab 2006; 91: 813– 819 [PubMed]
22. Mora S, Rifai N, Buring JE, Ridker PM. Fasting compared with nonfasting lipids and apolipoproteins for predicting incident cardiovascular events. Circulation 2008; 118: 993– 1001 [PMC free article] [PubMed]
23. Nordestgaard BG, Benn M, Schnohr P, Tybjaerg-Hansen A. Nonfasting triglycerides and risk of myocardial infarction, ischemic heart disease, and death in men and women. JAMA 2007; 298: 299– 308 [PubMed]
24. Chien KL, Hsu HC, Lee YT, Chen MF. Renal function and metabolic syndrome components on cardiovascular and all-cause mortality. Atherosclerosis 2008; 197: 860– 867 [PubMed]

Articles from Diabetes Care are provided here courtesy of American Diabetes Association