Skip main navigation

Increased Vascular Disease Mortality Risk in Prediabetic Korean Adults Is Mainly Attributable to Ischemic Stroke

Originally published 2017;48:840–845


Background and Purpose—

Prediabetes is a known risk factor for vascular diseases; however, its differential contribution to mortality risk from various vascular disease subtypes is not known.


The subjects of the National Health Insurance Service in Korea (2002–2013) nationwide cohort were stratified into normal glucose tolerance (fasting glucose <100 mg/dL), impaired fasting glucose (IFG) stage 1 (100–109 mg/dL), IFG stage 2 (110–125 mg/dL), and diabetes mellitus groups based on the fasting glucose level. A Cox regression analysis with counting process formulation was used to assess the mortality risk for vascular disease and its subtypes—ischemic heart disease, ischemic stroke, and hemorrhagic stroke.


When adjusted for age, sex, and body mass index, IFG stage 2, but not stage 1, was associated with significantly higher all-cause mortality (hazard ratio [HR], 1.26; 95% confidence interval [CI], 1.18–1.34) and vascular disease mortality (HR, 1.27; 95% CI, 1.08–1.49) compared with normal glucose tolerance. Among the vascular disease subtypes, mortality from ischemic stroke was significantly higher (HR, 1.60; 95% CI, 1.18–2.18) in subjects with IFG stage 2 but not from ischemic heart disease and hemorrhagic stroke. The ischemic stroke mortality associated with IFG stage 2 remained significantly high when adjusted other modifiable vascular disease risk factors (HR, 1.51; 95% CI: 1.10–2.09) and medical treatments (HR, 1.75; 95% CI, 1.19–2.57).


Higher IFG degree (fasting glucose, 110–125 mg/dL) was associated with increased all-cause and vascular disease mortality. The increased vascular disease mortality in IFG stage 2 was attributable to ischemic stroke, but not ischemic heart disease or hemorrhagic stroke in Korean adults.


Prediabetes, consisting of impaired fasting glucose (IFG) and impaired glucose tolerance, is defined as an intermediate hyperglycemic state that precedes overt diabetes mellitus (DM).1,2 It is regarded as a metabolic condition closely associated with insulin resistance or metabolic syndrome, as well as a precursor stage of type 2 DM.3 In addition, many previous studies have reported that prediabetes is a risk factor for the development of vascular diseases.48

Importantly, a few studies suggest that prediabetes contributes differently to each vascular disease subtype, including ischemic heart disease (IHD), ischemic stroke (IS), and hemorrhagic stroke (HS).911 These suggest either heterogeneity of the prediabetic state or existence of different mechanisms through which dysglycemia leads to each vascular event. Furthermore, only a few studies have investigated the association between prediabetes and hard outcomes, the mortality risk associated with each vascular disease subtype.

Prediabetes is a highly flexible state that may progress to DM or regress to normal glucose tolerance (NGT), and most previous studies were limited in not considering this changeable nature of prediabetes. To date, only one study has shown that the high risk of vascular disease mortality is confined to the prediabetic subjects who progress to DM.12 This dearth of relevant literature indicates the need for more sophisticated methodologies to understand the association between prediabetes and the risk of vascular disease events, or mortality.

We, therefore, aimed to investigate the association between IFG and the all-cause and cause-specific mortality in a prospective manner, using a population-based large-scale cohort (2002–2013).


Study Population

The Korean National Health Insurance System (NHIS) cohort study (2002–2013) is a population-based longitudinal study consisting of nearly one million Koreans, a representative 2.2% sample of the national population data. The NHIS cohort was longitudinally constructed as an initial 2002 cohort of 1 025 340 Koreans, who were followed up to 2013. Briefly, the cohort data are composed of demographic information, medical and pharmacy records, biannual health examination data, and death records. The demographic information contains age, sex, socioeconomic status based on income, and residency. Annually, ≈10% to 15% of the cohort population received health examinations. The medical and pharmacy records contain individuals’ disease codes, prescriptions, and medical histories. The date and cause of death recorded in the Korean National Statistical Office database were merged into this cohort. More detailed cohort information and data have been published previously.13

From the NHIS cohort, we selected study subjects (cases) aged >30 years who had received health examinations during 2002 to 2013 and classified them on the basis of their glucose tolerance state assessed through the fasting plasma glucose levels. In our cohort data, there were multiple records of fasting glucose levels and other covariates for each individual as they may have received a health screening several times during the study period (2002–2013). By rearranging the original cohort data using the counting process formulation, we could perform survival analysis using the time-dependent variables and reduce the possible loss of data. Moreover, we identified prediabetes and DM at the every moment when an event such as a death is occurred, depending on their fasting glucose levels, to minimize the confounding effects by changes in subjects’ glucose tolerance states over time.

We excluded the subjects with cancer or vascular disease in 2002 to avoid a possible confounding effect on the mortality. The total number of participants was 499 239 and deaths were 16 482 during 2002 to 2013. Death records and the cause of death were identified until the last follow-up year, 2013.

This study was based on the data from the NHIS; therefore, informed consent was not obtained individually. The study was approved by the Institutional Review Board of Korea University Anam Hospital (IRB number: ED14188).

Determinants of Glucose Tolerance State and Mortality

Individuals’ glucose tolerance state was determined through fasting plasma glucose assessment during annual health examinations: NGT (fasting glucose <100 mg/dL), IFG stage 1 (100–109 mg/dL), IFG stage 2 (110–125 mg/dL), and DM (≥126 mg/dL). DM was also identified by disease codes for DM and prescribed glucose-lowering medications.

Causes of death were classified according to the Korean version of the International Classification of Disease, TenthRevision: vascular disease (I00–I99), IHD (I20–I25), IS (I63–I66), and HS (I60–I62). The date and cause of deaths of each individuals were recorded in their medical records by physicians, and all of death records were included in the NHIS database.

Details for other variables were obtained from health examination data and medical records; body weight (kg), height (m), total cholesterol level (mg/dL), systolic and diastolic blood pressure (mm Hg) were measured after overnight fasting at regular medical checkup programs. Hypertension and dyslipidemia were identified by health examination data (systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg for hypertension and total cholesterol ≥240 mg/dL for dyslipidemia) or disease codes or specific medications for those diseases.

Smoking status (current, former, or never), alcohol consumption (≥3 times/wk, ≤2 times/wk, or never), and physical activity (≥3 times/wk, ≤2 times/wk, or never) were obtained from questionnaires in the regular health examinations. The NHIS also provided individuals’ socioeconomic status categorized into 20 groups based on income levels.

Statistical Analyses

Each subject had multiple time-dependent health-screening records as he/she received a health screening several times during the studying period of 2002 to 2013. There were lots of delayed entries of individuals who joined the cohort after the baseline. In accordance with the time-dependent nature of the health records and to include the delayed entries in the analysis, we used the counting process formulation for the Cox regression analysis.

The counting process formulation is a data-rearranging method based on the time interval of health screening such that the individuals’ health records can be represented by multiple observations, each identifying the nonoverlapping time interval between consecutive health screenings; the values of covariates over that interval; and an indicator of whether death occurred during each interval. Once the data set has been reconstructed in this manner, the multiple observations from each individual become independent cases that then allow delayed entries to be included in the analysis. The time-dependent covariates are then treated as time-invariant covariates during each time interval. Moreover, the covariate values at the time of death are more likely to be missed than others, and the counting process formulation can reduce this loss of data, especially in the event time, by using the last observed covariate values as the values at the time of death.

Hence, in applying the counting process method, we identified prediabetes and diabeteDM at the moment when a death was occurred, depending on their fasting glucose levels in each observed date. We used the SAS program’s PROC PHREG procedure for data analysis.14


Table 1 displays the baseline characteristics of the 4 study groups (NGT, IFG stage 1, IFG stage 2, and DM) according to the glucose tolerance status during the entire study period (2002–2013). The total number of cohort participants was 499 239 (71 678 were the entries in the first year cohort [2002], 54 127 in 2003, 47 239 in 2004, 39 905 in 2005, 44 044 in 2006, 35 244 in 2007, 42 253 in 2008, 40 866 in 2009, 34 604 in 2010, 34 524 in 2011, 29 465 in 2012, and 25 290 in 2013). The prevalence of DM and IFG was 13.8% and 16.4%, respectively. Subjects’ age, body mass index (BMI), and the presence of concomitant diseases such as hypertension and dyslipidemia exhibited an increasing trend from NGT to IFG to DM groups.

Table 1. Characteristics of the 2002 to 2013 NHIS Cohort Participants According to Their Glucose Tolerance State

CharacteristicsNGTIFG Stage 1 (100–110 mg/dL of FPG)IFG Stage 2 (110–125 mg/dL of FPG)DMP Value
Proportion, %69.810.65.813.8
Age, y47.3±11.849.7±11.951.3±12.156.5±12.2<0.0001
Male sex, %51.860.764.959.7<0.0001
BMI, kg/m223.5±3.024.3±3.124.7±3.224.5±3.2<0.0001
Smoking, %<0.0001
Alcohol, %<0.0001
 <2 times/wk36.737.636.429.2
 >3 times/wk10.
Physical activity, %<0.0001
 <2 times/wk26.826.926.223.0
 >3 times/wk36.037.736.243.0
Socioeconomic status, %NS
 Lower 30%20.420.221.722.6
 Middle 40%36.636.137.433.6
 Upper 30%43.143.740.943.8
Fasting glucose, mmol/L87.2±7.8103.8±2.8115.3±4.3139.0±44.9<0.0001
Systolic BP, mm Hg121.3±15.1125.3±15.3128.0±15.9127.6±16.0<0.0001
Total cholesterol, mmol/L194.3±35.4201.7±36.8204.7±38.6196.9±40.3<0.0001
Family history of DM, %<0.0001

Data are presented as mean±SD or %. BMI indicates body mass index; BP, blood pressure; DM, diabetes mellitus; FPG, fasting plasma glucose; IFG, impaired fasting glucose; NGT, normal glucose tolerance; NHIS, National Health Insurance System; and NS, nonsignificant.

During the entire study period, a total of 16 482 deaths were recorded in the cohort. Among them, 14.6% (2418 deaths) were because of vascular diseases. Table 2 shows the adjusted hazard ratio (HR) of all-cause and vascular disease mortality in the IFG and DM groups, compared with the NGT group. In a minimally adjusted model (model A), IFG stage 2 and DM groups had significantly higher all-cause mortality (HR, 1.26; 95% confidence interval, 1.16–1.32 and HR, 1.53; 95% confidence interval, 1.48–1.58, respectively) and vascular disease mortality (HR, 1.27; 95% confidence interval, 1.08–1.49 and HR, 1.54; 95% confidence interval, 1.41–1.69, respectively). This trend was maintained, although slightly attenuated, when adjusted for modifiable cardiovascular risk factors including hypertension and dyslipidemia (model B) or when adjusted not only for the risk factors included in model B but also medical treatments such as antihypertensive drugs, antithrombotics, and statins (model C). In contrast, the IFG stage 1 group did not show an increased mortality risk compared with the reference group.

Table 2. Risk (Hazard Ratio) of All-Cause and Vascular Disease Mortality in Patients With Impaired Fasting Glucose and DM

NGTIFG Stage 1 (100–110 mg/dL of FPG)IFG Stage 2 (110–125 mg/dL of FPG)DM
All-cause mortality (n=16 482)
 No. of deaths6485168912067102
 Model A1.001.01 (0.95–1.06)1.26 (1.18–1.34)1.53 (1.48–1.58)
 Model B1.001.03 (0.97–1.09)1.24 (1.16–1.32)1.55 (1.49–1.61)
 Model C1.001.03 (0.96–1.11)1.26 (1.16–1.37)1.47 (1.40–1.54)
Vascular disease mortality (n=2418)
 No. of deaths9232291821084
 Model A1.000.94 (0.81–1.09)1.27 (1.08–1.49)1.54 (1.41–1.69)
 Model B1.000.93 (0.80–1.08)1.16 (0.98–1.38)1.54 (1.41–1.70)
 Model C1.000.97 (0.81–1.17)1.22 (0.99–1.51)1.41 (1.25–1.59)

Model A: adjusted for age, sex, and BMI; Model B: adjusted for age, sex, BMI, alcohol, smoking, physical activity, SES, systolic BP, and total cholesterol; Model C: adjusted for age, sex, BMI, alcohol, smoking, physical activity, SES, antihypertensive drugs, antithrombotics, and statins. BMI indicates body mass index; BP, blood pressure; DM, diabetes mellitus; IFG, impaired fasting glucose; NGT, normal glucose tolerance; and SES, socioeconomic status.

The same analyses were conducted for each vascular disease subtype (Table 3). Among the total vascular disease deaths, IHD was the most frequent cause of death (34.5%), followed by IS (23.9%) and HS (22.4%). The independence between mortality from vascular disease subtypes and fasting glucose levels was rejected with P value of 0.001, showing that mortality from IS was positively related to IFG stage 2, and mortality from HS was negatively related to DM. Subjects with DM had a significantly higher mortality risk because of all vascular disease subtypes. On the contrary, in all adjustment models, IFG stage 2 subjects exhibited an increased IS mortality, but not IHD or HS mortality. In particular, the subjects with IFG stage 2 had even higher IS mortality risk than those with DM in a full adjustment model including medical treatments (HR, 1.75 versus 1.41 in model C). Similar to the results of analysis for vascular disease mortality, the IFG stage 1 group did not demonstrate a different mortality risk associated with any of the vascular disease subtypes compared with the NGT group.

Table 3. Risk (Hazard Ratio) of Mortality From Each Vascular Disease Subtypes in Patients With Impaired Fasting Glucose and DM

NGTIFG Stage 1 (100–110 mg/dL of FPG)IFG Stage 2 (110–125 mg/dL of FPG)DM
IHD mortality (n=834)
 No. of deaths2938453404
 Model A1.001.06 (0.83–1.35)1.14 (0.85–1.52)1.76 (1.51–2.06)
 Model B1.001.10 (0.85–1.41)1.06 (0.77–1.45)1.82 (1.55–2.14)
 Model C1.001.09 (0.80–1.50)0.99 (0.66–1.48)1.74 (1.42–2.13)
Ischemic stroke mortality (n=577)
 No. of deaths1994553280
 Model A1.000.84 (0.61–1.17)1.60 (1.18–2.18)1.77 (1.47–2.13)
 Model B1.000.78 (0.55–1.11)1.51 (1.10–2.09)1.82 (1.50–2.21)
 Model C1.000.66 (0.42–1.06)1.75 (1.19–2.57)1.41 (1.09–1.82)
Hemorrhagic stroke mortality (n=541)
 No. of deaths2385339211
 Model A1.000.90 (0.66–1.23)1.10 (0.77–1.58)1.36 (1.11–1.66)
 Model B1.000.85 (0.63–1.16)1.00 (0.70–1.44)1.31 (1.08–1.60)
 Model C1.000.98 (0.67–1.41)1.09 (0.69–1.72)1.23 (0.95–1.59)

Model A: adjusted for age, sex, and BMI; Model B: adjusted for age, sex, BMI, alcohol, smoking, physical activity, SES, systolic BP, and total cholesterol; Model C: adjusted for age, sex, BMI, alcohol, smoking, physical activity, SES, antihypertensive drugs, antithrombotics, and statins. BMI indicates body mass index; BP, blood pressure; DM, diabetes mellitus; IFG, impaired fasting glucose; IHD, ischemic heart disease; NGT, normal glucose tolerance; and SES, socioeconomic status.

We further investigated whether specific subgroups in the IFG stage 2 group had a higher all-cause or vascular disease mortality risk than others. All subgroups had significantly risky effects on all-cause and vascular disease mortality except for dyslipidemia on vascular disease mortality. HRs were calculated by summing the main effects of each specific subgroup and IFG stage 2 and their interaction effect, although only 2 interaction effects between alcohol and IFG stage 2 and between BMI and IFG stage 2 on all-cause mortality were significant (Tables I and II in the online-only Data Supplement).


Our results indicate that prediabetes is capable of differentially contributing to the mortality from each vascular disease subtypes in the general population; compared with NGT, a higher degree of IFG (stage 2, 110–125 mg/dL of fasting plasma glucose) was associated with a significantly higher mortality risk from IS but not from IHD or HS. Given that the IFG stage 2 was also associated with a higher vascular disease mortality, our results further suggest that this increased vascular disease mortality was mainly attributable to the increased IS mortality. In contrast, a lower degree of IFG (stage 1, 100–109 mg/dL of fasting plasma glucose) did not increase mortality because of vascular disease or any of its subtypes.

The prevalence of prediabetes is increasing worldwide, with the current prevalence estimated at >30% in various regions and ethnic groups.3 Likewise, in Korea, an alarming increase in prediabetes prevalence has been reported, from 21.5% in 2006 to 25.0% in 2013.15 A growing body of epidemiological evidence has demonstrated that prediabetes or dysglycemia below the diabetic range is associated with increased vascular disease incidence and mortality.6,7,1618 However, some level of uncertainty exists, mainly in defining the fasting glucose cutoff value that is associated with an increased vascular disease risk and the type of vascular disease that demonstrates higher risk in the prediabetics. Although the newer IFG criteria (fasting glucose 100–125 mg/dL) defined by the American Diabetes Association in 2003 was based on the risk of developing DM in the future,2 most large-scale studies, including the DECODE (Diabetes Epidemiology: Collaborative analysis of Diagnostic criteria in Europe); Hoorn; and the Framingham heart study, generally indicated that the cutoff fasting glucose value associated with increased cardiovascular risk is 110 mg/dL (the 1997 American Diabetes Association criteria for IFG).7,12,18 The results of our study are concordant with these findings.

Furthermore, given that vascular disease reflects pathophysiologically and clinically heterogeneous vascular events, conflicting results have been reported for the association between prediabetes and the risk of various vascular disease subtypes. For instance, the Asia Pacific Cohort study reported that fasting plasma glucose in the prediabetic range was an important risk determinant of both IHD and stroke.5 Similarly, the Korean Heart study observed that a higher degree of IFG was associated with an increased risk of incident IHD and IS but not HS.9 In contrast, the Funagata study and the KNHS study (Korean National Health Insurance System) demonstrated that prediabetes increased the risk of incident IS, but not IHD and HS.10,11 The differences in the results of these studies could be explained by the differences in the study population, the study population size, and the study duration; however, these results strongly indicated different underlying mechanisms through which glucose dysregulation contributed to IHD, IS, and HS. Moreover, because few studies have addressed the prediabetes–vascular disease subtype association in terms of the related mortality, our findings may provide a new insight into their relationship. We demonstrated that prediabetes was strongly associated with IS mortality, even after adjusting for known cardiovascular risk factors but not with IHD or HS mortality. Interestingly, our finding is in concurrence with the results of the large-scale Korean cohort study—the KNHS study, which reported that prediabetes increased the incidence of IS alone and not the other vascular disease subtypes. The researchers explained that dysglycemia may have a more direct negative impact on IS as it comprises both micro- and macrovascular atherosclerotic lesions, which is different from the predominately macrovascular lesions of IHD.11,19

Another possible explanation is the ethnic difference. The Asian population is more vulnerable to intracerebral vascular events than other populations; in a global comparative risk assessment study, hyperglycemia-related stroke burden was significantly higher in Asians, especially those from the East Asian region, than in the Western population.20 Moreover, the number of IS deaths is higher in the East Asians than in the number of IHD deaths, as has been shown in several Asian studies, including this study. Notably, the increased susceptibility to IS rather than IHD among people with prediabetes has mostly been reported in studies conducted on Asian study cohorts.10,11,21

We further noted that the subjects with prediabetes in our study cohort possessed a relatively lean phenotype compared with those in the Western population. Most studies in the Western countries have shown that the BMI of prediabetic subjects was usually in the 25 to 30 kg/m2 range4,7,16,22; however, the mean BMI of the IFG stage 2 group in our study cohort was only 24.2 kg/m2. Because prediabetes is a group of highly heterogeneous phenotypes and metabolic states, their cardiovascular risks are expected to vary depending on each individuals’ own risk factors. The findings of The Global Burden of Metabolic Risk Factors for Chronic Disease Collaboration study provided important suggestions to interpret our results. The study reported that obesity contributes differently to the risk of stroke and coronary heart disease; the effect of increasing BMI on the risk of vascular diseases was more pronounced on coronary heart disease than stroke, and conversely, the influence of metabolic mediators besides obesity was stronger on stroke than on coronary heart disease.23 Accordingly, the increased susceptibility to stroke rather than IHD that we observed may partly be explained by the lean and metabolically unhealthy phenotype of the Asian prediabetics.

As stated before, prediabetes is a flexible metabolic state. Large-scale observation studies have reported that ≈25% of the subjects with prediabetes regressed to NGT, whereas another 25% progressed to overt DM over a 3- to 5-year follow-up period.24,25 Consequently, most previous studies were fundamentally biased in not adjusting for the changing glucose tolerance state of the participants over time. In this context, the increased vascular disease risk observed only in the subjects who progressed to overt DM in the Hoorn study is suggestive. We eliminated this bias in our study by using the counting process analyses, which allowed the determination of glucose tolerance state as cases each year, which could then reflect regression to NGT and progression to overt DM. In fact, using a traditional Cox regression analysis to examine our data, we did not find an increased all-cause or cause-specific mortality risk in the IFG stage 2 group (data not shown).

This study still has several limitations. Because of the lack of relevant data, we could not use postprandial glucose or glycohemoglobin levels for analysis, even though they reportedly associate more significantly with increased vascular disease risk. Thus, our results may not be generalizable to all subjects with prediabetes, including isolated impaired glucose tolerance cases. Next, there exists a population bias in our study as we only used data from people who voluntarily participated in the health checkup programs and were thus probably more concerned about their health than those who did not participate in the checkups. Finally, we focused on mortality of each vascular diseases, not the incidence of them. So, the results of this study should be interpreted as the risk of fatal vascular events in subjects with prediabetes, rather than development of all types of vascular events.


To summarize, we observed that subjects with prediabetes, especially those with a higher fasting plasma glucose (110–125 mg/dL), had a significantly higher risk of ischemic stroke mortality. Therefore, an evaluation of the stroke-associated modifiable risks and formulation of stroke prevention strategies should be undertaken. Further investigations are needed to evaluate the different mechanisms underlying the development of IHD and stroke and to design beneficial strategies to decrease each vascular disease risks.


We thank all participants in the Korean Health Insurance Cohort study and the National Health Insurance Service who were a part of the NHIS-NSC (2002–2010) database (NHIS-2014-2-006).


The views expressed in this article are those of the authors and do not necessarily represent the official position of the department of Korean National Health Insurance Service.

*Drs Kim and Kwon contributed equally.

The online-only Data Supplement is available with this article at

Correspondence to Sin Gon Kim, MD, PhD, Division of Endocrinology and Metabolism, Department of Internal Medicine, Korea University Anam Hospital, Korea University College of Medicine, 126-1, Anam-dong 5-ga, Seongbuk-gu, Seoul 136–705, Korea, E-mail or Yousung Park, PhD, Department of Statistics, Korea University, Seoul, Korea, E-mail


  • 1. WHO, International Diabetes Foundation. Definition and Diagnosis of Diabetes Mellitus and Intermediate Hyperglycemia: Report of a WHO/IDF Consultation. Geneva: World Health Organization; 2006.Google Scholar
  • 2. Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, Kahn R, et al; Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Follow-up report on the diagnosis of diabetes mellitus.Diabetes Care. 2003; 26:3160–3167.CrossrefMedlineGoogle Scholar
  • 3. Tabák AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high-risk state for diabetes development.Lancet. 2012; 379:2279–2290. doi: 10.1016/S0140-6736(12)60283-9.CrossrefMedlineGoogle Scholar
  • 4. Glucose tolerance and cardiovascular mortality: comparison of fasting and 2-hour diagnostic criteria.Arch Intern Med. 2001; 161:397–405.CrossrefMedlineGoogle Scholar
  • 5. Lawes CM, Parag V, Bennett DA, Suh I, Lam TH, Whitlock G, et al; Asia Pacific Cohort Studies Collaboration. Blood glucose and risk of cardiovascular disease in the Asia Pacific region.Diabetes Care. 2004; 27:2836–2842.CrossrefMedlineGoogle Scholar
  • 6. Wen CP, Cheng TY, Tsai SP, Hsu HL, Wang SL. Increased mortality risks of pre-diabetes (impaired fasting glucose) in Taiwan.Diabetes Care. 2005; 28:2756–2761.CrossrefMedlineGoogle Scholar
  • 7. Levitzky YS, Pencina MJ, D’Agostino RB, Meigs JB, Murabito JM, Vasan RS, et al. Impact of impaired fasting glucose on cardiovascular disease: the Framingham Heart Study.J Am Coll Cardiol. 2008; 51:264–270. doi: 10.1016/j.jacc.2007.09.038.CrossrefMedlineGoogle Scholar
  • 8. Kim NH, Kim DJ, Park SW, Oh JY, Park JY, Shin C, et al; Committee on the Epidemiology of Diabetes Mellitus, Korean Diabetes Association. Plasma glucose regulation and mortality in Korea: a pooled analysis of three community-based cohort studies.Diabetes Metab J. 2014; 38:44–50. doi: 10.4093/dmj.2014.38.1.44.CrossrefMedlineGoogle Scholar
  • 9. Kim HK, Kim CH, Kim EH, Bae SJ, Choe J, Park JY, et al. Impaired fasting glucose and risk of cardiovascular disease in Korean men and women: the Korean Heart Study.Diabetes Care. 2013; 36:328–335. doi: 10.2337/dc12-0587.CrossrefMedlineGoogle Scholar
  • 10. Oizumi T, Daimon M, Jimbu Y, Wada K, Kameda W, Susa S, et al. Impaired glucose tolerance is a risk factor for stroke in a Japanese sample–the Funagata study.Metabolism. 2008; 57:333–338. doi: 10.1016/j.metabol.2007.10.007.CrossrefMedlineGoogle Scholar
  • 11. Sung J, Song YM, Ebrahim S, Lawlor DA. Fasting blood glucose and the risk of stroke and myocardial infarction.Circulation. 2009; 119:812–819. doi: 10.1161/CIRCULATIONAHA.108.776989.LinkGoogle Scholar
  • 12. Rijkelijkhuizen JM, Nijpels G, Heine RJ, Bouter LM, Stehouwer CD, Dekker JM. High risk of cardiovascular mortality in individuals with impaired fasting glucose is explained by conversion to diabetes: the Hoorn study.Diabetes Care. 2007; 30:332–336. doi: 10.2337/dc06-1238.CrossrefMedlineGoogle Scholar
  • 13. Kim NH, Lee J, Kim TJ, Kim NH, Choi KM, Baik SH, et al. Body mass index and mortality in the general population and in subjects with chronic disease in Korea: a nationwide cohort study (2002-2010).PLoS One. 2015; 10:e0139924. doi: 10.1371/journal.pone.0139924.CrossrefMedlineGoogle Scholar
  • 14. Allison PD. Survival Analysis Using SAS: A Practical Guide. 2nd ed. Cary, NC: SAS Institue; 2010.Google Scholar
  • 15. Noh J. The diabetes epidemic in Korea.Endocrinol Metab (Seoul). 2016; 31:349–353. doi: 10.3803/EnM.2016.31.3.349.CrossrefMedlineGoogle Scholar
  • 16. Barr EL, Zimmet PZ, Welborn TA, Jolley D, Magliano DJ, Dunstan DW, et al. Risk of cardiovascular and all-cause mortality in individuals with diabetes mellitus, impaired fasting glucose, and impaired glucose tolerance: the Australian Diabetes, Obesity, and Lifestyle Study (AusDiab).Circulation. 2007; 116:151–157. doi: 10.1161/CIRCULATIONAHA.106.685628.LinkGoogle Scholar
  • 17. Henry P, Thomas F, Benetos A, Guize L. Impaired fasting glucose, blood pressure and cardiovascular disease mortality.Hypertension. 2002; 40:458–463.LinkGoogle Scholar
  • 18. 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.CrossrefMedlineGoogle Scholar
  • 19. Fisher M, Folland E. Acute ischemic coronary artery disease and ischemic stroke: similarities and differences.Am J Ther. 2008; 15:137–149. doi: 10.1097/MJT.0b013e31816a61bb.CrossrefMedlineGoogle Scholar
  • 20. Danaei G, Lawes CM, Vander Hoorn S, Murray CJ, Ezzati M. Global and regional mortality from ischaemic heart disease and stroke attributable to higher-than-optimum blood glucose concentration: comparative risk assessment.Lancet. 2006; 368:1651–1659. doi: 10.1016/S0140-6736(06)69700-6.CrossrefMedlineGoogle Scholar
  • 21. Vermeer SE, Sandee W, Algra A, Koudstaal PJ, Kappelle LJ, Dippel DW; Dutch TIA Trial Study Group. Impaired glucose tolerance increases stroke risk in nondiabetic patients with transient ischemic attack or minor ischemic stroke.Stroke. 2006; 37:1413–1417. doi: 10.1161/01.STR.0000221766.73692.0b.LinkGoogle Scholar
  • 22. de Vegt F, Dekker JM, Ruhé HG, Stehouwer CD, Nijpels G, Bouter LM, et al. Hyperglycaemia is associated with all-cause and cardiovascular mortality in the Hoorn population: the Hoorn Study.Diabetologia. 1999; 42:926–931. doi: 10.1007/s001250051249.CrossrefMedlineGoogle Scholar
  • 23. Lu Y, Hajifathalian K, Ezzati M, Woodward M, Rimm EB, Danaei G. Metabolic mediators of the effects of body-mass index, overweight, and obesity on coronary heart disease and stroke: a pooled analysis of 97 prospective cohorts with 1.8 million participants.Lancet. 2014; 383:970–983.CrossrefMedlineGoogle Scholar
  • 24. Shaw JE, Zimmet PZ, de Courten M, Dowse GK, Chitson P, Gareeboo H, et al. Impaired fasting glucose or impaired glucose tolerance. What best predicts future diabetes in Mauritius?Diabetes Care. 1999; 22:399–402.CrossrefMedlineGoogle Scholar
  • 25. Gabir MM, Hanson RL, Dabelea D, Imperatore G, Roumain J, Bennett PH, et al. The 1997 American Diabetes Association and 1999 World Health Organization criteria for hyperglycemia in the diagnosis and prediction of diabetes.Diabetes Care. 2000; 23:1108–1112.CrossrefMedlineGoogle Scholar


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.