Adjustment of the GRACE score by 2-hour post-load glucose improves prediction of long-term major adverse cardiac events in acute coronary syndrome in patients without known diabetes

Adjustment of the GRACE score by 2-hour post-load glucose improves prediction of long-term major... Abstract Aims Global Registry of Acute Coronary Events (GRACE) risk score (GRS), a powerful predictor of prognosis after acute coronary event (ACE), does not include a glucometabolic measure. We investigate whether 2 h post-load plasma glucose (2h-PG) could improve GRS based prognostic models in ACE patients without known diabetes mellitus (DM). Methods and results A retrospective cohort study of 1056 ACE survivors without known DM who had fasting plasma glucose (FPG) and 2h-PG measured pre-discharge. Death and non-fatal myocardial infarction were recorded as major adverse cardiac events (MACE) during follow-up. GRS for discharge to 6 months was calculated. Cox proportional-hazards regression was used to identify predictors of event free survival. The predictive value of 2h-PG alone and combined with GRS was estimated using likelihood ratio test, Akaike’s information criteria, continuous net reclassification improvement (NRI>0), and integrated discrimination improvement (IDI). During 40.8 months follow-up 235 MACEs (22.3%) occurred, more frequently in the upper 2h-PG quartiles. Two-hour PG, but not FPG, adjusted for GRS independently predicted MACE (hazard ratio 1.091, 95% confidence interval 1.043–1.142; P = 0.0002). likelihood ratio test showed that 2h-PG significantly improved the prognostic models including GRS (χ2 = 20.56, 1 df; P = 0.000). Models containing GRS and 2h-PG yielded lowest corrected Akaike’s information criteria, compared to that with only GRS. 2h-PG, when added to GRS, improved net reclassification significantly (NRIe>0 6.4%, NRIne>0 24%, NRI>0 0.176; P = 0.017 at final follow-up). Two-hour PG, improved integrated discrimination of models containing GRS (IDI of 0.87%, P = 0.008 at final follow-up). Conclusion Two-hour PG, but not FPG, is an independent predictor of adverse outcome after ACE even after adjusting for the GRS. Two-hour PG, but not FPG, improves the predictability of prognostic models containing GRS. Acute coronary syndrome , Myocardial infarction , GRACE , Global Registry of Acute Coronary Events , Prognosis , Diabetes , Oral glucose tolerance test Introduction The Global Registry of Acute Coronary Events (GRACE) risk score (GRS) for mortality and re-infarction up to 6 months post-discharge is a powerful predictor of short and long-term prognosis after acute coronary syndrome (ACS).1–4 Although it is well established that post-ACS prognosis is worse in patients with known diabetes mellitus (DM) than in those without, DM is not included as a variable in the GRS model. Several studies show that hyperglycaemia, newly diagnosed after myocardial infarction (MI) on admission plasma glucose (APG), fasting plasma glucose (FPG), admission glycosylated haemoglobin (HbA1c) and oral glucose tolerance test (OGTT), in patients without known DM adversely affects long-term prognosis. None of these studies have included GRS or all of its components in their models to predict outcomes or demonstrated an independent effect of 2-hour post-load glucose (2h-PG) on prognosis. A few studies that have included GRS in addition to the glycaemic indices in their prognostic models have yielded variable results.5–16 Thus, it is still unclear as to which glycaemic index best predicts prognosis after MI in patients without known diabetes and whether 2h-PG, in addition to the GRS, independently predicts post-MI prognosis. In this study, we investigate the value of FPG and 2h-PG in addition to GRS in predicting major adverse cardiac events (MACEs) in patients with MI but without known DM and the potential incremental prognostic value of adding FPG and 2h-PG to models including GRS only. Methods We retrospectively analysed data, prospectively collected for a mandatory national audit, the Myocardial Infarction National Audit Project (MINAP), on all consecutive MI survivors without known DM, admitted between November 2005 and October 2008, who underwent pre-discharge OGTT as part of routine clinical care and were followed up.17 This observational study includes all patients for whom FPG, 2h-PG and the GRS were available. Data on age, gender, risk factors for coronary artery disease (CAD), past medical history, pre-hospital and discharge medications, troponin I levels, heart rate, systolic blood pressure, creatinine level, presence of congestive heart failure, previous history of MI, revascularisation status, and presence of ST-segment depression were recorded. Web-based GRS calculator was used to calculate the risk of death or MI from discharge to 6 months for each patient. Patients with known diabetes were excluded. Patients were classified as having pre-admission DM if the patient had been informed of the diagnosis by a physician or was on treatment. Glycosylated haemoglobin was not used for diagnosing pre-hospital diabetes as it was not recommended in contemporary guidance.18,19 Fasting plasma glucose (after an overnight fast of ≥8 h) and OGTT (venous plasma glucose measured 2 h after administration of 75 g glucose (2h-PG) in 200 mL water) were done on/after the third day of admission on consecutive patients without known DM. Patients who died before or did not tolerate the OGTT and were transferred to other centres before OGTT were excluded. Discharge was not delayed for the OGTT. Plasma glucose was enzymatically determined using the glucose oxidase method. Intravenous glucose solutions were not allowed, but anti-adrenergic agents were used if clinically indicated. Clinically unstable patients were tested later. The patients with impaired glucose tolerance (IGT) and new diabetes mellitus (NDM) were referred to the diabetologists for appropriate outpatients management. Participants were followed for up to 5 years (median 3.4 years) for outcomes. Completeness of follow-up was ensured by manual review of hospital and general practice records. The first occurrence of a MACE defined as death or non-fatal re-infarction, the events that the GRS predicts, was obtained from patient records. Mortality data was collected from the hospital care records for patients who died in hospital. For patients who died in the community, mortality data was obtained from the general practitioner medical records confirmed by the office of public health intelligence. Permission was sought from the East Yorkshire and North Lincolnshire Research Ethical Committee to analyse the data. As the study retrospectively analysed routinely collected anonymized data on standard clinical practice to contribute to a National Audit database, the Committee waived the need for formal ethical approval and patient consent.17 Statistical analysis Continuous variables are presented as median (inter-quartile range) and categorical variables as counts and proportions (%). Baseline characteristics are presented as quartiles of 2h-PG. The differences were compared between groups using the one-way analysis of variance and the Kruskal–Wallis test for parametric and non-parametric data, respectively for continuous variables and the χ2 test for categorical variables. Event free survival was estimated in the four quartiles of 2h-PG from the Kaplan–Meier curves that were compared using the log-rank test. Cox proportional-hazards regression modelling was used to analyse the effect of several variables on event free survival. All covariates known to affect prognosis after MI including gender, smoking status, hypercholesterolaemia, hypertension, history of previous acute MI, diagnosis at discharge, discharge prescription of aspirin, clopidogrel, beta-blockers, angiotensin-converting enzyme inhibitors and statins, inpatient revascularization status, GRS for 6 months from discharge for death and MI, FPG and 2h-PG were ‘entered’ into the model. The GRS variables (i.e. age, resting heart rate, systolic blood pressure on arrival, creatinine, congestive heart failure, history of MI, ST-segment depression, elevated troponin, and in-hospital revascularization) were not entered separately. Results are reported as hazard ratios with associated 95% confidence intervals (CIs). Multicollinearity was examined using variance inflation factor (VIF) (MedCalc Statistical Software version 17.0.4, Ostend, Belgium), and variables with VIF < 4 were included in the same model. Nested models were compared using the χ2 likelihood ratio tests to determine whether the logistic regression model that included GRS and FPG or 2h-PG provided a significantly better fit than those with GRS alone. Comparison of nested and non-nested models including GRS, or its combination with FPG or 2h-PG was performed by calculating corrected Akaike’s information criterion (AICc), delta-AICc (δAICc), and Akaike weights (wi), to estimate the probability that a given model is the “best” fitting model of those studied.20 Logistic regression models using the above covariates along with GRS, FPG, and 2h-PG individually and in combination were used to generate predicted probabilities of MACE. The incremental predictive value from adding FPG and 2h-PG to models with GRS was analysed from these predicted probabilities using several measures of improvement in discrimination: increase in the area under the receiver operating characteristic curve (AUC) (MedCalc Statistical Software version 17.0.4, Ostend, Belgium), category-free continuous net reclassification improvement (cNRI>0) and integrated discrimination improvement (IDI). In the absence of clearly pre-defined clinical risk thresholds for the models including GRS, categorical NRI was not used. The event NRI (NRIe) was defined as net percentage of persons with the event of interest correctly assigned a higher predicted risk and non-event NRI (NRIne) as net percentage of persons without the event of interest correctly assigned a lower predicted risk. The overall NRI defined as sum of the net proportion of persons with and without the event of interest correctly assigned a different predicted risk is reported as a number. The IDI was equal to the increase in discrimination slope defined as the mean difference in predicted risks between those with and without events. Results The 1056 patients, included in the study, were divided into quartiles of 2h-PG (Q1 ≤6.5 mmol/L, Q2 6.6–8.1 mmol/L, Q3 8.2–10.4 mmol/L, and Q4 >10.4 mmol/L) (Table 1) The patients in the upper quartiles were older, had more risk factors, were less frequently on clopidogrel, had higher heart rate and creatinine, more frequent heart failure, ST segment depression and high-risk GRS, higher mean GRS and FPG. Table 1 Baseline characteristics of the study population categorized by quartiles of 2 h post-load glucose Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Table 1 Baseline characteristics of the study population categorized by quartiles of 2 h post-load glucose Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Outcomes During the median follow-up of 40.8 months (range 6–60 months), there were 235 MACEs (22.3%), 112 deaths (10.6%), and 123 non-fatal re-infarctions (11.6%). Major adverse cardiac events were more frequent in the upper glucose quartiles (Table 1). Death and non-fatal re-infarction increased with increasing quartiles of 2h-PG even in those where the level of 2h-PG did not cross the conventional threshold for the diagnosis of DM (Figure 1). On Cox proportional hazard regression analysis 2h-PG and GRS, but not FPG, were consistently independent predictors of MACE at the final follow-up when included in the same model as GRS (Table 2). The risk of MACE increased by 9% for each mmol/L rise in 2h-PG. Table 2 Candidate predictors of event-free survival Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Table 2 Candidate predictors of event-free survival Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Figure 1 View largeDownload slide The event free survival in the quartiles of 2 h post-load glucose. Figure 1 View largeDownload slide The event free survival in the quartiles of 2 h post-load glucose. Nested models were compared using the likelihood ratio tests to determine whether logistic regression models that included GRS and FPG or 2h-PG provided a significantly better fit than that limited to the GRS. This showed that addition of the 2h-PG as a continuous variable significantly improved the ability of a model including GRS score to predict MACE (Table 3). Addition of FPG did not improve the model fit. Table 3 Akaike’s information criteria and likelihood ratio test to determine the best fitting model for predicting MACE Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 AICc, corrected Akaike’s information criteria; δAICc, delta AICc is a measure of each model relative to the best model; wi, Akaike weights, the ratio of δAICc values for each model relative to the whole set; wj/wi, evidence ratios compare the wi of the ‘best’ model and competing models to test the extent to which it is better than another. Table 3 Akaike’s information criteria and likelihood ratio test to determine the best fitting model for predicting MACE Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 AICc, corrected Akaike’s information criteria; δAICc, delta AICc is a measure of each model relative to the best model; wi, Akaike weights, the ratio of δAICc values for each model relative to the whole set; wj/wi, evidence ratios compare the wi of the ‘best’ model and competing models to test the extent to which it is better than another. Comparing models containing GRS alone, GRS with FPG and GRS with 2h-PG, the later yielded the lowest corrected AIC, highest Akaike’s weight and evidence ratio compared to that with only GRACE score (Table 3). This suggests that the model with GRACE score and 2h-PG is more likely to be the ‘best’ fitting model compared with the other models tested. Entering 2h-PG, but not FPG, into a logistic regression model containing GRACE score alone significantly improved the net reclassification of later model in predicting events during follow-up (Table 4). Using continuous NRI (NRI>0) 2h-PG improved reclassification by 6.4% for those with events and by 24% for those without, resulting in a significant overall improvement in net reclassification (NRI 0.176, P = 0.017 at final follow-up). The model including the GRS and 2h-PG seems to predict a lower risk of MACE than that with GRS only both in the event and non-event groups. This reduction in the predicted risk, results in 24% improvement in net reclassification in the non-event group. Addition of FPG did not improve reclassification. The addition of 2h-PG, but not FPG, to a model including GRS improved integrated discrimination at all time points during follow-up (Table 4). It yielded an IDI of 0.87%, P = 0.008 at final follow-up. Table 4 Net reclassification improvement for model improvement with the addition of 2 h PG or FPG to GRACE score alone GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 Table 4 Net reclassification improvement for model improvement with the addition of 2 h PG or FPG to GRACE score alone GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 The c-statistic was 0.746 (95% CI 0.719–0.772, P <0.0001) for the prognostic model containing the GRS only, 0.719 (95% CI 0.691–0.746, P < 0.0001) for the model containing 2h-PG only and 0.754 (95% CI 0.726–0.779, P <0.0001) for the model including GRS and 2h-PG. The AUC for the GRS-only was better than the 2h-PG only model (δAUC 0.0274, P = 0.045). The c-statistic did not increase significantly when 2h-PG was added to the GRS only model (δAUC 0.00744, P = 0.165) but did so when GRS was added to the 2h-PG only model (δAUC 0.0348, P = 0.002). This suggests that GRS, as expected, is a more powerful predictor of events than 2h-PG. Discussion This study shows that (i) 2h-PG, but not FPG, independently predicts prognosis after ACS after adjusting for the GRS and (ii) 2h-PG, but not FPG, improves the ability of models containing GRS to predict long-term adverse events after an ACS in patients without known DM. The GRS is a powerful predictor of prognosis after MI at different time points up to 4 years.1–4 Even though it is well established that patients with ACS and DM have poorer outcomes than those without; the GRS does not include DM or any of the glycaemic indices as a variable in the model. In the GRACE, DM independently predicted in-hospital2 but not the 6-month post-discharge mortality.3 The initial logistic regression models developed from GRACE to predict prognosis incorporated several variables including DM as a dichotomous categorical variable. This model was reduced to include only the eight most predictive variables to make it clinically usable.2 Diabetes mellitus and other variables were removed as the c statistics of models with and without these variables were similar.2,3 In the GRACE, FPG increased the risk of in-hospital mortality both when FPG was used to group patients and when used as continuous variable irrespective of a history of DM.5 The 6 months post-discharge mortality was high only if FPG was in the diabetic range.5 Almost all studies suggesting that FPG, APG, HbA1c, or AGT, are independent predictors of adverse prognosis after ACS have not included the GRS (or all its individual components) within their regression models. The results, in a few studies that did, are variable. When adjusted for GRS, FPG, APG, and HbA1c have independently predicted outcomes in some6,8,9,12,15 but not other5,7,11,13,14,16 studies. APG, FPG, and HbA1c improved the predictive ability of models containing GRS in some9,12,21 but not all studies.10,13,14 This is the only study to show that 2h-PG independently predicts prognosis after ACS after adjusting for the GRS and improves the ability of models containing GRS to predict prognosis. In contrast to our study, Aronson et al.9 showed that in patients without known diabetes FPG, adjusted for the GRS, predicted mortality after MI and improved the prognostic models containing GRS. That study included mainly (73%) ST-elevation myocardial infarction (STEMI) patients, measured FPG within 24 h of admission and 2h-PG were not measured. Only 44% of our patients had STEMI and FPG and 2h-PG were measured at 3–5 days. As the troponin in the GRS is not a continuous variable it does not reflect the prognostic effect of the extent of myonecrosis. Glucose levels are higher when measured within the first 24–48 h of MI than later and after STEMI compared to NSTEMI.22,23 The higher FPG in these STEMI patients when combined with GRS, a variable not influenced by the volume of myonecrosis, may have affected the model favourably improving its performance. In addition, it is unclear whether FPG would remain an independent predictor if 2h-PG was included in the models in this study. This could explain the difference in the two studies. The increased macrovascular morbidity associated with higher 2h-PG rather than FPG as seen in this study may be related to progression of atherosclerosis demonstrated with post-challenge rather than fasting hyperglycaemia.24–28 Without HbA1c, we do not have all the glycaemic indices to compare their effect on prognosis. Glycosylated haemoglobin has predicted post-MI prognosis in some 8,29–31 but not all studies.32–36 The relative ability of APG, FPG, 2h-PG, and HbA1c to predict post-MI prognosis in patients without previously known diabetes has rarely been studied.32,35,37 In the EUROASPIRE IV,35,38 neither FPG nor HbA1c predicted the primary outcome, whereas the 2h-PG did. In another study,32 HbA1c ≥6.5%, in the same model as OGTT, did not show any significant increase in mortality. However, there was significantly increased mortality in patients with HbA1c <6.5% categorized as newly diagnosed DM by OGTT. Kowalczyk et al.37 suggest that the HbA1c may be useful in further risk stratifying patients diagnosed with IGT and NDM but do not report the effect of HbA1c on prognosis of patients without. Sattar and Preiss39 suggest that HbA1c and FPG are better than OGTT for cardiovascular disease risk prediction citing two studies40,41 to argue in favour. The first,40 specifically excluded people with history of cardiovascular disease at baseline. The second,41 DETECT-2, looks at the relation of FPG and HbA1c prevalence of retinopathy, a microangiopathy, in epidemiological setting. Thus both these studies included populations very dissimilar to our study. The EUROSPIRE IV and SWEETHEART registry42 support the use of OGTT for predicting prognosis in these high risk patients. The c-statistic did not change significantly when 2h-PG was added to a model containing GRS. It is unsurprising that adding GRS to the model containing 2h-PG did. The increment in the c-statistic, used to quantify the added value offered by the new biomarker, is overly conservative and the ΔAUC depends on the performance of the underlying clinical model i.e. good clinical models are harder to improve on.43 Improving models containing powerful variables as the GRS may be difficult. To deal with this anomaly, Pencina et al.44,45 devised the IDI and NRI>0, for evaluating reclassification with novel biomarkers. These matrices improved when 2h-PG is added to GRS. The larger improvement in net reclassification in the non-event group when adding 2h-PG to the GRS may suggest that the 2h-PG tempers down the risk predicted by the GRS alone in these patients. As we aimed to evaluate whether adding FPG and/or 2h-PG improved prediction of post-MI prognosis by models containing GRS, we restricted ourselves to end-points predicted by the GRS i.e. death and non-fatal re-infarction as the only study end-points. We also entered the GRS as a composite rather than its individual covariates separately in the logistic regression analysis, as we wanted to see whether FPG and/or 2h-PG could improve the models containing GRS. Limitations Being an observational longitudinal cohort study using retrospective analysis of prospectively collected data from a single centre, it has its limitations. Although national death register was not consulted directly, there is no reason to doubt the accuracy of a linked general practice database used. As the study includes only the re-infarctions admitted to the local hospital, a few admitted to other hospitals may have been missed. Although every effort was made to ensure completeness of the data, information not recorded could not be used in statistical models. Exclusion of small number of patients, albeit for valid reasons, and mainly Caucasian study population could affect the generalizability of the results. The effect of random glycaemic fluctuations or stress hyperglycaemia on the results can not be excluded. As OGTT was not repeated pre- or post-discharge, it is uncertain whether random fluctuation in glycaemia or stress hyperglycaemia affected results. However, OGTT done at or after 5 days seems to reliably predict long term glucometabolic state.46,47 As pre-discharge post-challenge hyperglycaemia, irrespective of its pathophysiological mechanism, predicted outcomes in post-MI patients, the reproducibility of these measurements and its relation to long term glucometabolic status, though important in establishing a diagnosis of MI, may be less relevant when assessing prognostic risk. Conclusion This study suggests that in patients without known diabetes 2h-PG, but not FPG, is an independent predictor of adverse outcome after ACS even after adjusting for the GRS. The 2h-PG, but not FPG, improves the ability of models including GRS to predict long-term post-ACS prognosis. As the choice of diagnostic tests for detection of glycaemic abnormalities in this population is hotly debated it may be reasonable to suggest that the most important test would be the one that determines long term prognosis after ACS i.e. 2h-PG rather than the one deemed sufficient for use in the low-risk general population for epidemiological purposes even if simpler and more feasible i.e. HbA1c. This is especially so when clear evidence in favour HbA1c against 2h-PG in this high risk population is lacking. The 2h-PG should at least be considered as a marker of post-MI prognosis in these patients. Conflict of interest: none declared. Footnotes See page 2746 for the editorial comment on this article (doi: 10.1093/eurheartj/ehy320) References 1 Tang EW , Wong CK , Herbison P. Global Registry of Acute Coronary Events (GRACE) hospital discharge risk score accurately predicts long-term mortality post acute coronary syndrome . Am Heart J 2007 ; 153 : 29 – 35 . Google Scholar CrossRef Search ADS PubMed 2 Granger CB , Goldberg RJ , Dabbous O , Pieper KS , Eagle KA , Cannon CP , Van de WF , Avezum A , Goodman SG , Flather MD , Fox KA. Predictors of hospital mortality in the global registry of acute coronary events . Arch Intern Med 2003 ; 163 : 2345 – 2353 . Google Scholar CrossRef Search ADS PubMed 3 Eagle KA , Lim MJ , Dabbous OH , Pieper KS , Goldberg RJ , Van de WF , Goodman SG , Granger CB , Steg PG , Gore JM , Budaj A , Avezum A , Flather MD , Fox KA. A validated prediction model for all forms of acute coronary syndrome: estimating the risk of 6-month postdischarge death in an international registry . JAMA 2004 ; 291 : 2727 – 2733 . Google Scholar CrossRef Search ADS PubMed 4 Bradshaw PJ , Ko DT , Newman AM , Donovan LR , Tu JV. Validity of the GRACE (Global Registry of Acute Coronary Events) acute coronary syndrome prediction model for six month post-discharge death in an independent data set . Heart 2006 ; 92 : 905 – 909 . Google Scholar CrossRef Search ADS PubMed 5 Sinnaeve PR , Steg PG , Fox KA , Van de WF , Montalescot G , Granger CB , Knobel E , Anderson FA , Dabbous OH , Avezum A. Association of elevated fasting glucose with increased short-term and 6-month mortality in ST-segment elevation and non-ST-segment elevation acute coronary syndromes: the Global Registry of Acute Coronary Events . Arch Intern Med 2009 ; 169 : 402 – 409 . Google Scholar CrossRef Search ADS PubMed 6 Kosiborod M , Rathore SS , Inzucchi SE , Masoudi FA , Wang Y , Havranek EP , Krumholz HM. Admission glucose and mortality in elderly patients hospitalized with acute myocardial infarction: implications for patients with and without recognized diabetes . Circulation 2005 ; 111 : 3078 – 3086 . Google Scholar CrossRef Search ADS PubMed 7 Foo K , Cooper J , Deaner A , Knight C , Suliman A , Ranjadayalan K , Timmis AD. A single serum glucose measurement predicts adverse outcomes across the whole range of acute coronary syndromes . Heart 2003 ; 89 : 512 – 516 . Google Scholar CrossRef Search ADS PubMed 8 Liu XJ , Wan ZF , Zhao N , Zhang YP , Mi L , Wang XH , Zhou D , Wu Y , Yuan ZY. Adjustment of the GRACE score by HemoglobinA1c enables a more accurate prediction of long-term major adverse cardiac events in acute coronary syndrome without diabetes undergoing percutaneous coronary intervention . Cardiovasc Diabetol 2015 ; 19 : 110 . Google Scholar CrossRef Search ADS 9 Aronson D , Hammerman H , Kapeliovich MR , Suleiman A , Agmon Y , Beyar R , Markiewicz W , Suleiman M. Fasting glucose in acute myocardial infarction: incremental value for long-term mortality and relationship with left ventricular systolic function . Diabetes Care 2007 ; 30 : 960 – 966 . Google Scholar CrossRef Search ADS PubMed 10 Baeza-Román A , de Miguel-Balsa E , Latour-Pérez J , Carrillo-López A. Predictive power of the GRACE score in population with diabetes . Int J Cardiol 2017 ; 248 : 73 – 76 . Google Scholar CrossRef Search ADS PubMed 11 Su G , Mi SH , Tao H , Li Z , Yang HX , Zheng H , Zhou Y , Tian L. Impact of admission glycemic variability, glucose, and glycosylated hemoglobin on major adverse cardiac events after acute myocardial infarction . Diabetes Care 2013 ; 36 : 1026 – 1032 . Google Scholar CrossRef Search ADS PubMed 12 Timoteo AT , Papoila AL , Rio P , Miranda F , Ferreira ML , Ferreira RC. Prognostic impact of admission blood glucose for all-cause mortality in patients with acute coronary syndromes: added value on top of GRACE risk score . Eur Heart J Acute Cardiovasc Care 2014 ; 3 : 257 – 263 . Google Scholar CrossRef Search ADS PubMed 13 de Mulder M , van der PT , de Waard GA , Boersma E , Umans VA. Admission glucose does not improve GRACE score at 6 months and 5 years after myocardial infarction . Cardiology 2011 ; 120 : 227 – 234 . Google Scholar CrossRef Search ADS PubMed 14 Correia LC , Rocha MS , Bittencourt AP , Freitas R , Souza AC , Almeida MC , Pericles EJ. Does acute hyperglycemia add prognostic value to the GRACE score in individuals with non-ST elevation acute coronary syndromes? Clin Chim Acta 2009 ; 410 : 74 – 78 . Google Scholar CrossRef Search ADS PubMed 15 David RB , Almeida ED , Cruz LV , Sebben JC , Feijo IP , Schmidt KE , Avena LM , Gottschall CA , Quadros AS. Diabetes mellitus and glucose as predictors of mortality in primary coronary percutaneous intervention . Arq Bras Cardiol 2014 ; 103 : 323 – 330 . Google Scholar PubMed 16 She J , Deng Y , Wu Y , Xia Y , Li H , Liang X , Shi R , Yuan Z. Hemoglobin A1c is associated with severity of coronary artery stenosis but not with long term clinical outcomes in diabetic and nondiabetic patients with acute myocardial infarction undergoing primary angioplasty . Cardiovasc Diabetol 2017 ; 16 : 97 – 0578 . Google Scholar CrossRef Search ADS PubMed 17 George A , Bhatia RT , Buchanan GL , Whiteside A , Moisey RS , Beer SF , Chattopadhyay S , Sathyapalan T , John J. Impaired glucose tolerance or newly diagnosed diabetes mellitus diagnosed during admission adversely affects prognosis after myocardial infarction: an observational study . PLoS One 2015 ; 10 : e0142045 . Google Scholar CrossRef Search ADS PubMed 18 Ryden L , Standl E , Bartnik M , Van den BG , Betteridge J , de Boer MJ , Cosentino F , Jonsson B , Laakso M , Malmberg K , Priori S , Ostergren J , Tuomilehto J , Thrainsdottir I , Vanhorebeek I , Stramba-Badiale M , Lindgren P , Qiao Q , Priori SG , Blanc JJ , Budaj A , Camm J , Dean V , Deckers J , Dickstein K , Lekakis J , McGregor K , Metra M , Morais J , Osterspey A , Tamargo J , Zamorano JL , Deckers JW , Bertrand M , Charbonnel B , Erdmann E , Ferrannini E , Flyvbjerg A , Gohlke H , Juanatey JR , Graham I , Monteiro PF , Parhofer K , Pyorala K , Raz I , Schernthaner G , Volpe M , Wood D. Guidelines on diabetes, pre-diabetes, and cardiovascular diseases: executive summary. The Task Force on Diabetes and Cardiovascular Diseases of the European Society of Cardiology (ESC) and of the European Association for the Study of Diabetes (EASD) . Eur Heart J 2006 ; 28 : 88 – 136 . Google Scholar CrossRef Search ADS 19 Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus . Diabetes Care 1997 ; 20 : 1183 – 1197 . CrossRef Search ADS PubMed 20 Wagenmakers EJ , Farrell S. AIC model selection using Akaike weights . Psychon Bull Rev 2004 ; 11 : 192 – 196 . Google Scholar CrossRef Search ADS PubMed 21 Roffi M , Eberli FR. Diabetes and acute coronary syndromes . Best Pract Res Clin Endocrinol Metab 2009 ; 23 : 305 – 316 . Google Scholar CrossRef Search ADS PubMed 22 Hage C , Malmberg K , Ryden L , Wallander M. The impact of infarct type on the reliability of early oral glucose tolerance testing in patients with myocardial infarction . Int J Cardiol 2010 ; 145 : 259 – 260 . Google Scholar CrossRef Search ADS PubMed 23 Knudsen EC , Seljeflot I , Abdelnoor M , Eritsland J , Mangschau A , Arnesen H , Andersen GO. Abnormal glucose regulation in patients with acute ST- elevation myocardial infarction-a cohort study on 224 patients . Cardiovasc Diabetol 2009 ; 8 : 6 . Google Scholar CrossRef Search ADS PubMed 24 Ando T , Okada S , Niijima Y , Hashimoto K , Shimizu H , Tsuchiya T , Yamada M , Ohshima K , Mori M , Ono K. Impaired glucose tolerance, but not impaired fasting glucose, is a risk factor for early-stage atherosclerosis . Diabet Med 2010 ; 27 : 1430 – 1435 . Google Scholar CrossRef Search ADS PubMed 25 Choi ES , Rhee EJ , Choi JH , Bae JC , Yoo SH , Kim WJ , Park SE , Park CY , Lee WY , Cho YK , Oh KW , Park SW , Kim SW. The association of brachial-ankle pulse wave velocity with 30-minute post-challenge plasma glucose levels in korean adults with no history of type 2 diabetes . Korean Diabetes J 2010 ; 34 : 287 – 293 . Google Scholar CrossRef Search ADS PubMed 26 Mellen PB , Bittner V , Herrington DM. Post-challenge glucose predicts coronary atherosclerotic progression in non-diabetic, post-menopausal women . Diabet Med 2007 ; 24 : 1156 – 1159 . Google Scholar CrossRef Search ADS PubMed 27 Hanefeld M , Koehler C , Henkel E , Fuecker K , Schaper F , Temelkova-Kurktschiev T. Post-challenge hyperglycaemia relates more strongly than fasting hyperglycaemia with carotid intima-media thickness: the RIAD Study. Risk factors in impaired glucose tolerance for atherosclerosis and diabetes . Diabet Med 2000 ; 17 : 835 – 840 . Google Scholar CrossRef Search ADS PubMed 28 Temelkova-Kurktschiev TS , Koehler C , Henkel E , Leonhardt W , Fuecker K , Hanefeld M. Postchallenge plasma glucose and glycemic spikes are more strongly associated with atherosclerosis than fasting glucose or HbA1c level . Diabetes Care 2000 ; 23 : 1830 – 1834 . Google Scholar CrossRef Search ADS PubMed 29 Timmer JR , Hoekstra M , Nijsten MW , van der Horst IC , Ottervanger JP , Slingerland RJ , Dambrink JH , Bilo HJ , Zijlstra F , van 't Hof AW. Prognostic value of admission glycosylated hemoglobin and glucose in nondiabetic patients with ST-segment-elevation myocardial infarction treated with percutaneous coronary intervention . Circulation 2011 ; 124 : 704 – 711 . Google Scholar CrossRef Search ADS PubMed 30 Moura FA , Figueiredo VN , Teles BS , Barbosa MA , Pereira LR , Costa AP , Carvalho LS , Cintra RM , Almeida OL , Quinaglia E Silva JC , Nadruz JW , Sposito AC. Glycosylated hemoglobin is associated with decreased endothelial function, high inflammatory response, and adverse clinical outcome in non-diabetic STEMI patients . Atherosclerosis 2015 ; 243 : 124 – 130 . Google Scholar CrossRef Search ADS PubMed 31 Geng J , Zhang Y , Wang B , Xie J , Xu B , Li J. Glycosylated hemoglobin levels and clinical outcomes in nondiabetic patients with coronary artery disease: a meta-analysis . Medicine (Baltimore) 2017 ; 96 : e6784 . Google Scholar CrossRef Search ADS PubMed 32 Pararajasingam G , Hofsten DE , Logstrup BB , Egstrup M , Henriksen FL , Hangaard J , Egstrup K. Newly detected abnormal glucose regulation and long-term prognosis after acute myocardial infarction: comparison of an oral glucose tolerance test and glycosylated haemoglobin A1c . Int J Cardiol 2016 ; 214 : 310 – 315 . Google Scholar CrossRef Search ADS PubMed 33 Shin D , Ahn J , Cha KS , Park JS , Oh JH , Lee HW , Hong JY , Kim BW , Hong TJ. Impact of initial glycosylated hemoglobin level on cardiovascular outcomes in prediabetic patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention . Coron Artery Dis 2016 ; 27 : 40 – 46 . Google Scholar CrossRef Search ADS PubMed 34 Lazzeri C , Valente S , Chiostri M , Attana P , Mattesini A , Nesti M , Gensini GF. Glycated haemoglobin and long-term mortality in patients with ST Elevation Myocardial Infarction . J Cardiovasc Med (Hagerstown) 2015 ; 16 : 404 – 408 . Google Scholar CrossRef Search ADS PubMed 35 Shahim B , De Bacquer D , De Backer G , Gyberg V , Kotseva K , Mellbin L , Schnell O , Tuomilehto J , Wood D , Rydén L. The prognostic value of fasting plasma glucose, two-hour postload glucose, and HbA1c in patients with coronary artery disease: a report from EUROASPIRE IV: a survey from the European Society of Cardiology . Diabetes Care 2017 ; 40 : 1233 – 1240 . Google Scholar CrossRef Search ADS PubMed 36 Tailakh MA , Friger M , Zahger D , Sidi A , Mazor-Dray E , Novack V. Prospective study of the impact of diabetes mellitus newly diagnosed by glycated hemoglobin on outcomes in patients undergoing percutaneous coronary intervention . Eur J Intern Med 2017 ; 37 : 69 – 74 . Google Scholar CrossRef Search ADS PubMed 37 Kowalczyk J , Mazurek M , Zielinska T , Lenarczyk R , Sedkowska A , Swiatkowski A , Sredniawa B , Mencel G , Francuz P , Kalarus Z. Prognostic significance of HbA1c in patients with AMI treated invasively and newly detected glucose abnormalities . Eur J Prev Cardiol 2015 ; 22 : 798 – 806 . Google Scholar CrossRef Search ADS PubMed 38 Gyberg V , De Bacquer D , Kotseva K , De Backer G , Schnell O , Sundvall J , Tuomilehto J , Wood D , Rydén L. Screening for dysglycaemia in patients with coronary artery disease as reflected by fasting glucose, oral glucose tolerance test, and HbA1c: a report from EUROASPIRE IV—a survey from the European Society of Cardiology . Eur Heart J 2015 ; 36 : 1171 – 1177 . Google Scholar CrossRef Search ADS PubMed 39 Sattar N , Preiss D. Screening for diabetes in patients with cardiovascular disease: HbA1c trumps oral glucose tolerance testing . Lancet Diabetes Endocrinol 2016 ; 4 : 560 – 562 . Google Scholar CrossRef Search ADS PubMed 40 Di Angelantonio E , Gao P , Khan H , Butterworth AS , Wormser D , Kaptoge S , Kondapally Seshasai S Sr , Thompson A , Sarwar N , Willeit P , Ridker PM , Barr EL , Khaw KT , Psaty BM , Brenner H , Balkau B , Dekker JM , Lawlor DA , Daimon M , Willeit J , Njølstad I , Nissinen A , Brunner EJ , Kuller LH , Price JF , Sundström J , Knuiman MW , Feskens EJ , Verschuren WM , Wald N , Bakker SJ , Whincup PH , Ford I , Goldbourt U , Gómez-de-la-Cámara A , Gallacher J , Simons LA , Rosengren A , Sutherland SE , Björkelund C , Blazer DG , Wassertheil-Smoller S , Onat A , Marín Ibañez A , Casiglia E , Jukema JW , Simpson LM , Giampaoli S , Nordestgaard BG , Selmer R , Wennberg P , Kauhanen J , Salonen JT , Dankner R , Barrett-Connor E , Kavousi M , Gudnason V , Evans D , Wallace RB , Cushman M , D’agostino RB , Umans JG , Kiyohara Y , Nakagawa H , Sato S , Gillum RF , Folsom AR , van der Schouw YT , Moons KG , Griffin SJ , Sattar N , Wareham NJ , Selvin E , Thompson SG , Danesh J. Glycated hemoglobin measurement and prediction of cardiovascular disease . JAMA 2014 ; 311 : 1225 – 1233 . Google Scholar CrossRef Search ADS PubMed 41 Colagiuri S , Lee CM , Wong TY , Balkau B , Shaw JE , Borch-Johnsen K. Glycemic thresholds for diabetes-specific retinopathy: implications for diagnostic criteria for diabetes . Diabetes Care 2011 ; 34 : 145 – 150 . Google Scholar CrossRef Search ADS PubMed 42 Gitt A , Bramlage P , Towae F , Papp A , Deeg E , Senges J , Zeymer U , Zahn R , Fleischmann H , Boehler S , Schnell O , Standl E , Tschöpe D. Newly diagnosed diabetes and pre-diabetes in patients after acute myocardial infarction predicts adverse outcomes during a three-year follow-up: results of the Sweetheart Registry . J Am Coll Cardiol 2013 ; 61 : E176 . 10_S Google Scholar CrossRef Search ADS 43 Chen HC , Kodell RL , Cheng KF , Chen JJ. Assessment of performance of survival prediction models for cancer prognosis . BMC Med Res Methodol 2012 ; 12 : 102 – 112 . Google Scholar CrossRef Search ADS PubMed 44 Pencina MJ , D'Agostino RB Sr , Demler OV. Novel metrics for evaluating improvement in discrimination: net reclassification and integrated discrimination improvement for normal variables and nested models . Stat Med 2012 ; 31 : 101 – 113 . Google Scholar CrossRef Search ADS PubMed 45 Pencina MJ , D' Agostino RB , D' Agostino RB , 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 . Google Scholar CrossRef Search ADS PubMed 46 Wallander M , Malmberg K , Norhammar A , Ryden L , Tenerz A. Oral glucose tolerance test: a reliable tool for early detection of glucose abnormalities in patients with acute myocardial infarction in clinical practice: a report on repeated oral glucose tolerance tests from the GAMI study . Diabetes Care 2008 ; 31 : 36 – 38 . Google Scholar CrossRef Search ADS PubMed 47 Tenerz A , Norhammar A , Silveira A , Hamsten A , Nilsson G , Ryden L , Malmberg K. Diabetes, insulin resistance, and the metabolic syndrome in patients with acute myocardial infarction without previously known diabetes . Diabetes Care 2003 ; 26 : 2770 – 2776 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Heart Journal Oxford University Press

Adjustment of the GRACE score by 2-hour post-load glucose improves prediction of long-term major adverse cardiac events in acute coronary syndrome in patients without known diabetes

European Heart Journal , Volume 39 (29) – Aug 1, 2018

Loading next page...
 
/lp/ou_press/adjustment-of-the-grace-score-by-2-hour-post-load-glucose-improves-qDayul6mfO
Publisher
Oxford University Press
Copyright
Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com.
ISSN
0195-668X
eISSN
1522-9645
D.O.I.
10.1093/eurheartj/ehy233
Publisher site
See Article on Publisher Site

Abstract

Abstract Aims Global Registry of Acute Coronary Events (GRACE) risk score (GRS), a powerful predictor of prognosis after acute coronary event (ACE), does not include a glucometabolic measure. We investigate whether 2 h post-load plasma glucose (2h-PG) could improve GRS based prognostic models in ACE patients without known diabetes mellitus (DM). Methods and results A retrospective cohort study of 1056 ACE survivors without known DM who had fasting plasma glucose (FPG) and 2h-PG measured pre-discharge. Death and non-fatal myocardial infarction were recorded as major adverse cardiac events (MACE) during follow-up. GRS for discharge to 6 months was calculated. Cox proportional-hazards regression was used to identify predictors of event free survival. The predictive value of 2h-PG alone and combined with GRS was estimated using likelihood ratio test, Akaike’s information criteria, continuous net reclassification improvement (NRI>0), and integrated discrimination improvement (IDI). During 40.8 months follow-up 235 MACEs (22.3%) occurred, more frequently in the upper 2h-PG quartiles. Two-hour PG, but not FPG, adjusted for GRS independently predicted MACE (hazard ratio 1.091, 95% confidence interval 1.043–1.142; P = 0.0002). likelihood ratio test showed that 2h-PG significantly improved the prognostic models including GRS (χ2 = 20.56, 1 df; P = 0.000). Models containing GRS and 2h-PG yielded lowest corrected Akaike’s information criteria, compared to that with only GRS. 2h-PG, when added to GRS, improved net reclassification significantly (NRIe>0 6.4%, NRIne>0 24%, NRI>0 0.176; P = 0.017 at final follow-up). Two-hour PG, improved integrated discrimination of models containing GRS (IDI of 0.87%, P = 0.008 at final follow-up). Conclusion Two-hour PG, but not FPG, is an independent predictor of adverse outcome after ACE even after adjusting for the GRS. Two-hour PG, but not FPG, improves the predictability of prognostic models containing GRS. Acute coronary syndrome , Myocardial infarction , GRACE , Global Registry of Acute Coronary Events , Prognosis , Diabetes , Oral glucose tolerance test Introduction The Global Registry of Acute Coronary Events (GRACE) risk score (GRS) for mortality and re-infarction up to 6 months post-discharge is a powerful predictor of short and long-term prognosis after acute coronary syndrome (ACS).1–4 Although it is well established that post-ACS prognosis is worse in patients with known diabetes mellitus (DM) than in those without, DM is not included as a variable in the GRS model. Several studies show that hyperglycaemia, newly diagnosed after myocardial infarction (MI) on admission plasma glucose (APG), fasting plasma glucose (FPG), admission glycosylated haemoglobin (HbA1c) and oral glucose tolerance test (OGTT), in patients without known DM adversely affects long-term prognosis. None of these studies have included GRS or all of its components in their models to predict outcomes or demonstrated an independent effect of 2-hour post-load glucose (2h-PG) on prognosis. A few studies that have included GRS in addition to the glycaemic indices in their prognostic models have yielded variable results.5–16 Thus, it is still unclear as to which glycaemic index best predicts prognosis after MI in patients without known diabetes and whether 2h-PG, in addition to the GRS, independently predicts post-MI prognosis. In this study, we investigate the value of FPG and 2h-PG in addition to GRS in predicting major adverse cardiac events (MACEs) in patients with MI but without known DM and the potential incremental prognostic value of adding FPG and 2h-PG to models including GRS only. Methods We retrospectively analysed data, prospectively collected for a mandatory national audit, the Myocardial Infarction National Audit Project (MINAP), on all consecutive MI survivors without known DM, admitted between November 2005 and October 2008, who underwent pre-discharge OGTT as part of routine clinical care and were followed up.17 This observational study includes all patients for whom FPG, 2h-PG and the GRS were available. Data on age, gender, risk factors for coronary artery disease (CAD), past medical history, pre-hospital and discharge medications, troponin I levels, heart rate, systolic blood pressure, creatinine level, presence of congestive heart failure, previous history of MI, revascularisation status, and presence of ST-segment depression were recorded. Web-based GRS calculator was used to calculate the risk of death or MI from discharge to 6 months for each patient. Patients with known diabetes were excluded. Patients were classified as having pre-admission DM if the patient had been informed of the diagnosis by a physician or was on treatment. Glycosylated haemoglobin was not used for diagnosing pre-hospital diabetes as it was not recommended in contemporary guidance.18,19 Fasting plasma glucose (after an overnight fast of ≥8 h) and OGTT (venous plasma glucose measured 2 h after administration of 75 g glucose (2h-PG) in 200 mL water) were done on/after the third day of admission on consecutive patients without known DM. Patients who died before or did not tolerate the OGTT and were transferred to other centres before OGTT were excluded. Discharge was not delayed for the OGTT. Plasma glucose was enzymatically determined using the glucose oxidase method. Intravenous glucose solutions were not allowed, but anti-adrenergic agents were used if clinically indicated. Clinically unstable patients were tested later. The patients with impaired glucose tolerance (IGT) and new diabetes mellitus (NDM) were referred to the diabetologists for appropriate outpatients management. Participants were followed for up to 5 years (median 3.4 years) for outcomes. Completeness of follow-up was ensured by manual review of hospital and general practice records. The first occurrence of a MACE defined as death or non-fatal re-infarction, the events that the GRS predicts, was obtained from patient records. Mortality data was collected from the hospital care records for patients who died in hospital. For patients who died in the community, mortality data was obtained from the general practitioner medical records confirmed by the office of public health intelligence. Permission was sought from the East Yorkshire and North Lincolnshire Research Ethical Committee to analyse the data. As the study retrospectively analysed routinely collected anonymized data on standard clinical practice to contribute to a National Audit database, the Committee waived the need for formal ethical approval and patient consent.17 Statistical analysis Continuous variables are presented as median (inter-quartile range) and categorical variables as counts and proportions (%). Baseline characteristics are presented as quartiles of 2h-PG. The differences were compared between groups using the one-way analysis of variance and the Kruskal–Wallis test for parametric and non-parametric data, respectively for continuous variables and the χ2 test for categorical variables. Event free survival was estimated in the four quartiles of 2h-PG from the Kaplan–Meier curves that were compared using the log-rank test. Cox proportional-hazards regression modelling was used to analyse the effect of several variables on event free survival. All covariates known to affect prognosis after MI including gender, smoking status, hypercholesterolaemia, hypertension, history of previous acute MI, diagnosis at discharge, discharge prescription of aspirin, clopidogrel, beta-blockers, angiotensin-converting enzyme inhibitors and statins, inpatient revascularization status, GRS for 6 months from discharge for death and MI, FPG and 2h-PG were ‘entered’ into the model. The GRS variables (i.e. age, resting heart rate, systolic blood pressure on arrival, creatinine, congestive heart failure, history of MI, ST-segment depression, elevated troponin, and in-hospital revascularization) were not entered separately. Results are reported as hazard ratios with associated 95% confidence intervals (CIs). Multicollinearity was examined using variance inflation factor (VIF) (MedCalc Statistical Software version 17.0.4, Ostend, Belgium), and variables with VIF < 4 were included in the same model. Nested models were compared using the χ2 likelihood ratio tests to determine whether the logistic regression model that included GRS and FPG or 2h-PG provided a significantly better fit than those with GRS alone. Comparison of nested and non-nested models including GRS, or its combination with FPG or 2h-PG was performed by calculating corrected Akaike’s information criterion (AICc), delta-AICc (δAICc), and Akaike weights (wi), to estimate the probability that a given model is the “best” fitting model of those studied.20 Logistic regression models using the above covariates along with GRS, FPG, and 2h-PG individually and in combination were used to generate predicted probabilities of MACE. The incremental predictive value from adding FPG and 2h-PG to models with GRS was analysed from these predicted probabilities using several measures of improvement in discrimination: increase in the area under the receiver operating characteristic curve (AUC) (MedCalc Statistical Software version 17.0.4, Ostend, Belgium), category-free continuous net reclassification improvement (cNRI>0) and integrated discrimination improvement (IDI). In the absence of clearly pre-defined clinical risk thresholds for the models including GRS, categorical NRI was not used. The event NRI (NRIe) was defined as net percentage of persons with the event of interest correctly assigned a higher predicted risk and non-event NRI (NRIne) as net percentage of persons without the event of interest correctly assigned a lower predicted risk. The overall NRI defined as sum of the net proportion of persons with and without the event of interest correctly assigned a different predicted risk is reported as a number. The IDI was equal to the increase in discrimination slope defined as the mean difference in predicted risks between those with and without events. Results The 1056 patients, included in the study, were divided into quartiles of 2h-PG (Q1 ≤6.5 mmol/L, Q2 6.6–8.1 mmol/L, Q3 8.2–10.4 mmol/L, and Q4 >10.4 mmol/L) (Table 1) The patients in the upper quartiles were older, had more risk factors, were less frequently on clopidogrel, had higher heart rate and creatinine, more frequent heart failure, ST segment depression and high-risk GRS, higher mean GRS and FPG. Table 1 Baseline characteristics of the study population categorized by quartiles of 2 h post-load glucose Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Table 1 Baseline characteristics of the study population categorized by quartiles of 2 h post-load glucose Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Q1 ≤6.5 (n = 274) Q2 6.6–8.1 (n = 261) Q3 8.2–10.4 (n = 259) Q4 >10.4 (n = 262) P-value Male, n (%) 186 (67.9) 186 (71.3) 193 (74.5) 192 (73.3) 0.344 Current smoker, n (%) 114 (41.6) 83 (31.8) 84 (32.4) 82 (31.3) 0.034 Hypertension, n (%) 80 (29.2) 100 (38.3) 110 (42.5) 105 (40.1) 0.009 Hypercholesterolaemia, n (%) 45 (16.2) 69 (26.4) 57 (22.0) 62 (23.7) 0.039 Previous AMI, n (%) 35 (12.8) 45 (17.2) 47 (18.2) 54 (20.6) 0.107 Known IHD, n (%) 39 (14.2) 51 (19.5) 52 (20.1) 58 (22.1) 0.113 CVA, n (%) 7 (2.6) 8 (3.1) 15 (5.8) 22 (8.5) 0.006 Normal LVEF, n (%) 139 (50.7) 108 (41.4) 104 (40.2) 108 (41.2) 0.045 Diagnosis NSTEMI, n (%) 163 (59.5) 144 (55.2) 131 (50.6) 152 (58.0) 0.176 Discharge medications, n (%)  Aspirin 267 (97.5) 250 (95.8) 241 (93.1) 245 (93.5) 0.070  Clopidogrel 251 (91.6) 243 (93.1) 220 (84.9) 235 (89.7) 0.013  Beta-blocker 196 (71.5) 186 (71.3) 190 (73.4) 204 (77.9) 0.287  ACEI/ARB 210 (76.6) 204 (78.2) 208 (80.3) 214 (81.7) 0.489  Statin 237 (86.5) 223 (85.4) 219 (84.6) 228 (87.0) 0.853 GRACE variables  Age (years), median (IQR) 59.5 (18.8) 63.5 (17.4) 66.3 (17.5) 68.3 (18.2) <0.001  HR (b.p.m.), median (IQR) 73 (24) 76 (27) 74 (25) 81 (28) 0.005  SBP, median (IQR) 137 (31) 140 (38) 139 (38) 140 (36.5) 0.196  Creatinine (µmol/L), median (IQR) 94 (21) 96 (23) 100 (24) 102 (24.5) <0.001  HF 8 (2.92) 10 (3.8) 11 (4.25) 23 (8.8) 0.009  ST-segment depression 174 (63.5) 197 (75.5) 195 (75.3) 199 (75.9) <0.001  Troponin rise 273 (99.6) 260 (99.6) 256 (98.8) 258 (98.5) 0.369  Cardiac arrest 6 (2.2) 11 (4.2) 12 (4.6) 13 (5.0) 0.349 GRACE score, median, (IQR)  Admission 6m death 103 (39) 114 (40) 115 (37) 119 (43.3) <0.001  Admission 6m death/MI 154 (48) 166 (46) 167 (48) 167 (54.3) <0.001  Discharge 6m death 104 (42) 115 (39) 119 (38) 123 (42) <0.001  Discharge 6m death/MI 113 (37) 113 (37) 113 (37) 131 (45) <0.001 GRACE risk  High 97 (35.4) 117 (44.8) 131 (50.6) 149 (56.9) <0.001  Intermediate 89 (32.5) 95 (36.4) 88 (34.0) 88 (33.6) 0.810  Low 88 (32.1) 49 (18.8) 40 (15.4) 25 (9.5) <0.001 Glucometabolic category  NGT 267 (97.5) 198 (75.9) 0 (0) 0 (0) <0.001  IGT 0 (0) 56 (21.5) 253 (97.7) 61 (23.3) <0.001  NDM 7 (2.6) 7 (2.7) 6 (2.3) 209 (79.8) <0.001 FPG (mmol/L), median (IQR) 4.9 (0.6) 5 (0.6) 5.1 (0.8) 5.5 (1.13) <0.001 2HBG (mmol/L) median (IQR) 5.6 (1.3) 7.4 (0.7) 9.2 (1.4) 12.3 (3) <0.001 MACE 30 (10.9) 64 (24.5) 67 (25.9) 74 (28.2) —  Deaths 14 (5.1) 24 (9.2) 39 (15.1) 35 (13.4) —  Re-infarctions 16 (5.8) 40 (15.3) 28 (10.8) 39 (14.9) — Outcomes During the median follow-up of 40.8 months (range 6–60 months), there were 235 MACEs (22.3%), 112 deaths (10.6%), and 123 non-fatal re-infarctions (11.6%). Major adverse cardiac events were more frequent in the upper glucose quartiles (Table 1). Death and non-fatal re-infarction increased with increasing quartiles of 2h-PG even in those where the level of 2h-PG did not cross the conventional threshold for the diagnosis of DM (Figure 1). On Cox proportional hazard regression analysis 2h-PG and GRS, but not FPG, were consistently independent predictors of MACE at the final follow-up when included in the same model as GRS (Table 2). The risk of MACE increased by 9% for each mmol/L rise in 2h-PG. Table 2 Candidate predictors of event-free survival Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Table 2 Candidate predictors of event-free survival Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Covariates HR 95% CI P-value GRACE score 1.01 1.01–1.02 <0.0001 2 h PG 1.09 1.04–1.14 0.000 Hypercholesterolaemia 0.66 0.47–0.92 0.014 Previous MI 1.50 1.05–2.14 0.024 Discharged without BB 1.39 1.02–1.88 0.035 FPG 0.85 0.71–1.01 0.063 Discharged without clopidogrel 1.35 0.95–1.93 0.098 Hypertension 1.25 0.95–1.64 0.115 Previous revascularization 1.35 0.90–2.01 0.149 Discharged without ACEI 1.29 0.91–1.83 0.154 Discharged without aspirin 1.37 0.86–2.19 0.184 Female gender 1.14 0.85–1.51 0.379 Discharge diagnosis of STEMI 1.13 0.86–1.49 0.381 Discharged without statin 0.90 0.58–1.38 0.619 Current smoker 0.94 0.70–1.25 0.655 Figure 1 View largeDownload slide The event free survival in the quartiles of 2 h post-load glucose. Figure 1 View largeDownload slide The event free survival in the quartiles of 2 h post-load glucose. Nested models were compared using the likelihood ratio tests to determine whether logistic regression models that included GRS and FPG or 2h-PG provided a significantly better fit than that limited to the GRS. This showed that addition of the 2h-PG as a continuous variable significantly improved the ability of a model including GRS score to predict MACE (Table 3). Addition of FPG did not improve the model fit. Table 3 Akaike’s information criteria and likelihood ratio test to determine the best fitting model for predicting MACE Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 AICc, corrected Akaike’s information criteria; δAICc, delta AICc is a measure of each model relative to the best model; wi, Akaike weights, the ratio of δAICc values for each model relative to the whole set; wj/wi, evidence ratios compare the wi of the ‘best’ model and competing models to test the extent to which it is better than another. Table 3 Akaike’s information criteria and likelihood ratio test to determine the best fitting model for predicting MACE Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 Akaike’s information criteria Likelihood ratio test Model AICc δAICc Relative likelihood wi wj/wi Model χ2 df P-value GRS 1006.46 8.22 0.02 0.02 2.65 GRS vs. GRS + 2HBS 998.24 0.00 1.00 0.98 162.07 GRS + 2HBS 20.56 1 0.000 GRS + FBS 1008.41 10.18 0.01 0.01 1.00 GRS + FBS 0.21 1 0.645 AICc, corrected Akaike’s information criteria; δAICc, delta AICc is a measure of each model relative to the best model; wi, Akaike weights, the ratio of δAICc values for each model relative to the whole set; wj/wi, evidence ratios compare the wi of the ‘best’ model and competing models to test the extent to which it is better than another. Comparing models containing GRS alone, GRS with FPG and GRS with 2h-PG, the later yielded the lowest corrected AIC, highest Akaike’s weight and evidence ratio compared to that with only GRACE score (Table 3). This suggests that the model with GRACE score and 2h-PG is more likely to be the ‘best’ fitting model compared with the other models tested. Entering 2h-PG, but not FPG, into a logistic regression model containing GRACE score alone significantly improved the net reclassification of later model in predicting events during follow-up (Table 4). Using continuous NRI (NRI>0) 2h-PG improved reclassification by 6.4% for those with events and by 24% for those without, resulting in a significant overall improvement in net reclassification (NRI 0.176, P = 0.017 at final follow-up). The model including the GRS and 2h-PG seems to predict a lower risk of MACE than that with GRS only both in the event and non-event groups. This reduction in the predicted risk, results in 24% improvement in net reclassification in the non-event group. Addition of FPG did not improve reclassification. The addition of 2h-PG, but not FPG, to a model including GRS improved integrated discrimination at all time points during follow-up (Table 4). It yielded an IDI of 0.87%, P = 0.008 at final follow-up. Table 4 Net reclassification improvement for model improvement with the addition of 2 h PG or FPG to GRACE score alone GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 Table 4 Net reclassification improvement for model improvement with the addition of 2 h PG or FPG to GRACE score alone GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 GRACE score vs. GRACE score and 2 h PG GRACE score vs. GRACE score and FPG NRIe NRIne Total P-value NRIe NRIne Total P-value UP 110 312 422 143 481 624 DWN 125 509 634 92 340 432 Total 235 821 1056 235 821 1056 NRI>0 −0.064 0.240 0.176 0.017 0.217 −0.172 0.045 0.541 IDIe IDIne Total P-value IDIe IDIne Total P-value Final 0.0067 −0.0019 0.0087 0.008 0.0000 0.0000 0.0000 0.449 The c-statistic was 0.746 (95% CI 0.719–0.772, P <0.0001) for the prognostic model containing the GRS only, 0.719 (95% CI 0.691–0.746, P < 0.0001) for the model containing 2h-PG only and 0.754 (95% CI 0.726–0.779, P <0.0001) for the model including GRS and 2h-PG. The AUC for the GRS-only was better than the 2h-PG only model (δAUC 0.0274, P = 0.045). The c-statistic did not increase significantly when 2h-PG was added to the GRS only model (δAUC 0.00744, P = 0.165) but did so when GRS was added to the 2h-PG only model (δAUC 0.0348, P = 0.002). This suggests that GRS, as expected, is a more powerful predictor of events than 2h-PG. Discussion This study shows that (i) 2h-PG, but not FPG, independently predicts prognosis after ACS after adjusting for the GRS and (ii) 2h-PG, but not FPG, improves the ability of models containing GRS to predict long-term adverse events after an ACS in patients without known DM. The GRS is a powerful predictor of prognosis after MI at different time points up to 4 years.1–4 Even though it is well established that patients with ACS and DM have poorer outcomes than those without; the GRS does not include DM or any of the glycaemic indices as a variable in the model. In the GRACE, DM independently predicted in-hospital2 but not the 6-month post-discharge mortality.3 The initial logistic regression models developed from GRACE to predict prognosis incorporated several variables including DM as a dichotomous categorical variable. This model was reduced to include only the eight most predictive variables to make it clinically usable.2 Diabetes mellitus and other variables were removed as the c statistics of models with and without these variables were similar.2,3 In the GRACE, FPG increased the risk of in-hospital mortality both when FPG was used to group patients and when used as continuous variable irrespective of a history of DM.5 The 6 months post-discharge mortality was high only if FPG was in the diabetic range.5 Almost all studies suggesting that FPG, APG, HbA1c, or AGT, are independent predictors of adverse prognosis after ACS have not included the GRS (or all its individual components) within their regression models. The results, in a few studies that did, are variable. When adjusted for GRS, FPG, APG, and HbA1c have independently predicted outcomes in some6,8,9,12,15 but not other5,7,11,13,14,16 studies. APG, FPG, and HbA1c improved the predictive ability of models containing GRS in some9,12,21 but not all studies.10,13,14 This is the only study to show that 2h-PG independently predicts prognosis after ACS after adjusting for the GRS and improves the ability of models containing GRS to predict prognosis. In contrast to our study, Aronson et al.9 showed that in patients without known diabetes FPG, adjusted for the GRS, predicted mortality after MI and improved the prognostic models containing GRS. That study included mainly (73%) ST-elevation myocardial infarction (STEMI) patients, measured FPG within 24 h of admission and 2h-PG were not measured. Only 44% of our patients had STEMI and FPG and 2h-PG were measured at 3–5 days. As the troponin in the GRS is not a continuous variable it does not reflect the prognostic effect of the extent of myonecrosis. Glucose levels are higher when measured within the first 24–48 h of MI than later and after STEMI compared to NSTEMI.22,23 The higher FPG in these STEMI patients when combined with GRS, a variable not influenced by the volume of myonecrosis, may have affected the model favourably improving its performance. In addition, it is unclear whether FPG would remain an independent predictor if 2h-PG was included in the models in this study. This could explain the difference in the two studies. The increased macrovascular morbidity associated with higher 2h-PG rather than FPG as seen in this study may be related to progression of atherosclerosis demonstrated with post-challenge rather than fasting hyperglycaemia.24–28 Without HbA1c, we do not have all the glycaemic indices to compare their effect on prognosis. Glycosylated haemoglobin has predicted post-MI prognosis in some 8,29–31 but not all studies.32–36 The relative ability of APG, FPG, 2h-PG, and HbA1c to predict post-MI prognosis in patients without previously known diabetes has rarely been studied.32,35,37 In the EUROASPIRE IV,35,38 neither FPG nor HbA1c predicted the primary outcome, whereas the 2h-PG did. In another study,32 HbA1c ≥6.5%, in the same model as OGTT, did not show any significant increase in mortality. However, there was significantly increased mortality in patients with HbA1c <6.5% categorized as newly diagnosed DM by OGTT. Kowalczyk et al.37 suggest that the HbA1c may be useful in further risk stratifying patients diagnosed with IGT and NDM but do not report the effect of HbA1c on prognosis of patients without. Sattar and Preiss39 suggest that HbA1c and FPG are better than OGTT for cardiovascular disease risk prediction citing two studies40,41 to argue in favour. The first,40 specifically excluded people with history of cardiovascular disease at baseline. The second,41 DETECT-2, looks at the relation of FPG and HbA1c prevalence of retinopathy, a microangiopathy, in epidemiological setting. Thus both these studies included populations very dissimilar to our study. The EUROSPIRE IV and SWEETHEART registry42 support the use of OGTT for predicting prognosis in these high risk patients. The c-statistic did not change significantly when 2h-PG was added to a model containing GRS. It is unsurprising that adding GRS to the model containing 2h-PG did. The increment in the c-statistic, used to quantify the added value offered by the new biomarker, is overly conservative and the ΔAUC depends on the performance of the underlying clinical model i.e. good clinical models are harder to improve on.43 Improving models containing powerful variables as the GRS may be difficult. To deal with this anomaly, Pencina et al.44,45 devised the IDI and NRI>0, for evaluating reclassification with novel biomarkers. These matrices improved when 2h-PG is added to GRS. The larger improvement in net reclassification in the non-event group when adding 2h-PG to the GRS may suggest that the 2h-PG tempers down the risk predicted by the GRS alone in these patients. As we aimed to evaluate whether adding FPG and/or 2h-PG improved prediction of post-MI prognosis by models containing GRS, we restricted ourselves to end-points predicted by the GRS i.e. death and non-fatal re-infarction as the only study end-points. We also entered the GRS as a composite rather than its individual covariates separately in the logistic regression analysis, as we wanted to see whether FPG and/or 2h-PG could improve the models containing GRS. Limitations Being an observational longitudinal cohort study using retrospective analysis of prospectively collected data from a single centre, it has its limitations. Although national death register was not consulted directly, there is no reason to doubt the accuracy of a linked general practice database used. As the study includes only the re-infarctions admitted to the local hospital, a few admitted to other hospitals may have been missed. Although every effort was made to ensure completeness of the data, information not recorded could not be used in statistical models. Exclusion of small number of patients, albeit for valid reasons, and mainly Caucasian study population could affect the generalizability of the results. The effect of random glycaemic fluctuations or stress hyperglycaemia on the results can not be excluded. As OGTT was not repeated pre- or post-discharge, it is uncertain whether random fluctuation in glycaemia or stress hyperglycaemia affected results. However, OGTT done at or after 5 days seems to reliably predict long term glucometabolic state.46,47 As pre-discharge post-challenge hyperglycaemia, irrespective of its pathophysiological mechanism, predicted outcomes in post-MI patients, the reproducibility of these measurements and its relation to long term glucometabolic status, though important in establishing a diagnosis of MI, may be less relevant when assessing prognostic risk. Conclusion This study suggests that in patients without known diabetes 2h-PG, but not FPG, is an independent predictor of adverse outcome after ACS even after adjusting for the GRS. The 2h-PG, but not FPG, improves the ability of models including GRS to predict long-term post-ACS prognosis. As the choice of diagnostic tests for detection of glycaemic abnormalities in this population is hotly debated it may be reasonable to suggest that the most important test would be the one that determines long term prognosis after ACS i.e. 2h-PG rather than the one deemed sufficient for use in the low-risk general population for epidemiological purposes even if simpler and more feasible i.e. HbA1c. This is especially so when clear evidence in favour HbA1c against 2h-PG in this high risk population is lacking. The 2h-PG should at least be considered as a marker of post-MI prognosis in these patients. Conflict of interest: none declared. Footnotes See page 2746 for the editorial comment on this article (doi: 10.1093/eurheartj/ehy320) References 1 Tang EW , Wong CK , Herbison P. Global Registry of Acute Coronary Events (GRACE) hospital discharge risk score accurately predicts long-term mortality post acute coronary syndrome . Am Heart J 2007 ; 153 : 29 – 35 . Google Scholar CrossRef Search ADS PubMed 2 Granger CB , Goldberg RJ , Dabbous O , Pieper KS , Eagle KA , Cannon CP , Van de WF , Avezum A , Goodman SG , Flather MD , Fox KA. Predictors of hospital mortality in the global registry of acute coronary events . Arch Intern Med 2003 ; 163 : 2345 – 2353 . Google Scholar CrossRef Search ADS PubMed 3 Eagle KA , Lim MJ , Dabbous OH , Pieper KS , Goldberg RJ , Van de WF , Goodman SG , Granger CB , Steg PG , Gore JM , Budaj A , Avezum A , Flather MD , Fox KA. A validated prediction model for all forms of acute coronary syndrome: estimating the risk of 6-month postdischarge death in an international registry . JAMA 2004 ; 291 : 2727 – 2733 . Google Scholar CrossRef Search ADS PubMed 4 Bradshaw PJ , Ko DT , Newman AM , Donovan LR , Tu JV. Validity of the GRACE (Global Registry of Acute Coronary Events) acute coronary syndrome prediction model for six month post-discharge death in an independent data set . Heart 2006 ; 92 : 905 – 909 . Google Scholar CrossRef Search ADS PubMed 5 Sinnaeve PR , Steg PG , Fox KA , Van de WF , Montalescot G , Granger CB , Knobel E , Anderson FA , Dabbous OH , Avezum A. Association of elevated fasting glucose with increased short-term and 6-month mortality in ST-segment elevation and non-ST-segment elevation acute coronary syndromes: the Global Registry of Acute Coronary Events . Arch Intern Med 2009 ; 169 : 402 – 409 . Google Scholar CrossRef Search ADS PubMed 6 Kosiborod M , Rathore SS , Inzucchi SE , Masoudi FA , Wang Y , Havranek EP , Krumholz HM. Admission glucose and mortality in elderly patients hospitalized with acute myocardial infarction: implications for patients with and without recognized diabetes . Circulation 2005 ; 111 : 3078 – 3086 . Google Scholar CrossRef Search ADS PubMed 7 Foo K , Cooper J , Deaner A , Knight C , Suliman A , Ranjadayalan K , Timmis AD. A single serum glucose measurement predicts adverse outcomes across the whole range of acute coronary syndromes . Heart 2003 ; 89 : 512 – 516 . Google Scholar CrossRef Search ADS PubMed 8 Liu XJ , Wan ZF , Zhao N , Zhang YP , Mi L , Wang XH , Zhou D , Wu Y , Yuan ZY. Adjustment of the GRACE score by HemoglobinA1c enables a more accurate prediction of long-term major adverse cardiac events in acute coronary syndrome without diabetes undergoing percutaneous coronary intervention . Cardiovasc Diabetol 2015 ; 19 : 110 . Google Scholar CrossRef Search ADS 9 Aronson D , Hammerman H , Kapeliovich MR , Suleiman A , Agmon Y , Beyar R , Markiewicz W , Suleiman M. Fasting glucose in acute myocardial infarction: incremental value for long-term mortality and relationship with left ventricular systolic function . Diabetes Care 2007 ; 30 : 960 – 966 . Google Scholar CrossRef Search ADS PubMed 10 Baeza-Román A , de Miguel-Balsa E , Latour-Pérez J , Carrillo-López A. Predictive power of the GRACE score in population with diabetes . Int J Cardiol 2017 ; 248 : 73 – 76 . Google Scholar CrossRef Search ADS PubMed 11 Su G , Mi SH , Tao H , Li Z , Yang HX , Zheng H , Zhou Y , Tian L. Impact of admission glycemic variability, glucose, and glycosylated hemoglobin on major adverse cardiac events after acute myocardial infarction . Diabetes Care 2013 ; 36 : 1026 – 1032 . Google Scholar CrossRef Search ADS PubMed 12 Timoteo AT , Papoila AL , Rio P , Miranda F , Ferreira ML , Ferreira RC. Prognostic impact of admission blood glucose for all-cause mortality in patients with acute coronary syndromes: added value on top of GRACE risk score . Eur Heart J Acute Cardiovasc Care 2014 ; 3 : 257 – 263 . Google Scholar CrossRef Search ADS PubMed 13 de Mulder M , van der PT , de Waard GA , Boersma E , Umans VA. Admission glucose does not improve GRACE score at 6 months and 5 years after myocardial infarction . Cardiology 2011 ; 120 : 227 – 234 . Google Scholar CrossRef Search ADS PubMed 14 Correia LC , Rocha MS , Bittencourt AP , Freitas R , Souza AC , Almeida MC , Pericles EJ. Does acute hyperglycemia add prognostic value to the GRACE score in individuals with non-ST elevation acute coronary syndromes? Clin Chim Acta 2009 ; 410 : 74 – 78 . Google Scholar CrossRef Search ADS PubMed 15 David RB , Almeida ED , Cruz LV , Sebben JC , Feijo IP , Schmidt KE , Avena LM , Gottschall CA , Quadros AS. Diabetes mellitus and glucose as predictors of mortality in primary coronary percutaneous intervention . Arq Bras Cardiol 2014 ; 103 : 323 – 330 . Google Scholar PubMed 16 She J , Deng Y , Wu Y , Xia Y , Li H , Liang X , Shi R , Yuan Z. Hemoglobin A1c is associated with severity of coronary artery stenosis but not with long term clinical outcomes in diabetic and nondiabetic patients with acute myocardial infarction undergoing primary angioplasty . Cardiovasc Diabetol 2017 ; 16 : 97 – 0578 . Google Scholar CrossRef Search ADS PubMed 17 George A , Bhatia RT , Buchanan GL , Whiteside A , Moisey RS , Beer SF , Chattopadhyay S , Sathyapalan T , John J. Impaired glucose tolerance or newly diagnosed diabetes mellitus diagnosed during admission adversely affects prognosis after myocardial infarction: an observational study . PLoS One 2015 ; 10 : e0142045 . Google Scholar CrossRef Search ADS PubMed 18 Ryden L , Standl E , Bartnik M , Van den BG , Betteridge J , de Boer MJ , Cosentino F , Jonsson B , Laakso M , Malmberg K , Priori S , Ostergren J , Tuomilehto J , Thrainsdottir I , Vanhorebeek I , Stramba-Badiale M , Lindgren P , Qiao Q , Priori SG , Blanc JJ , Budaj A , Camm J , Dean V , Deckers J , Dickstein K , Lekakis J , McGregor K , Metra M , Morais J , Osterspey A , Tamargo J , Zamorano JL , Deckers JW , Bertrand M , Charbonnel B , Erdmann E , Ferrannini E , Flyvbjerg A , Gohlke H , Juanatey JR , Graham I , Monteiro PF , Parhofer K , Pyorala K , Raz I , Schernthaner G , Volpe M , Wood D. Guidelines on diabetes, pre-diabetes, and cardiovascular diseases: executive summary. The Task Force on Diabetes and Cardiovascular Diseases of the European Society of Cardiology (ESC) and of the European Association for the Study of Diabetes (EASD) . Eur Heart J 2006 ; 28 : 88 – 136 . Google Scholar CrossRef Search ADS 19 Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus . Diabetes Care 1997 ; 20 : 1183 – 1197 . CrossRef Search ADS PubMed 20 Wagenmakers EJ , Farrell S. AIC model selection using Akaike weights . Psychon Bull Rev 2004 ; 11 : 192 – 196 . Google Scholar CrossRef Search ADS PubMed 21 Roffi M , Eberli FR. Diabetes and acute coronary syndromes . Best Pract Res Clin Endocrinol Metab 2009 ; 23 : 305 – 316 . Google Scholar CrossRef Search ADS PubMed 22 Hage C , Malmberg K , Ryden L , Wallander M. The impact of infarct type on the reliability of early oral glucose tolerance testing in patients with myocardial infarction . Int J Cardiol 2010 ; 145 : 259 – 260 . Google Scholar CrossRef Search ADS PubMed 23 Knudsen EC , Seljeflot I , Abdelnoor M , Eritsland J , Mangschau A , Arnesen H , Andersen GO. Abnormal glucose regulation in patients with acute ST- elevation myocardial infarction-a cohort study on 224 patients . Cardiovasc Diabetol 2009 ; 8 : 6 . Google Scholar CrossRef Search ADS PubMed 24 Ando T , Okada S , Niijima Y , Hashimoto K , Shimizu H , Tsuchiya T , Yamada M , Ohshima K , Mori M , Ono K. Impaired glucose tolerance, but not impaired fasting glucose, is a risk factor for early-stage atherosclerosis . Diabet Med 2010 ; 27 : 1430 – 1435 . Google Scholar CrossRef Search ADS PubMed 25 Choi ES , Rhee EJ , Choi JH , Bae JC , Yoo SH , Kim WJ , Park SE , Park CY , Lee WY , Cho YK , Oh KW , Park SW , Kim SW. The association of brachial-ankle pulse wave velocity with 30-minute post-challenge plasma glucose levels in korean adults with no history of type 2 diabetes . Korean Diabetes J 2010 ; 34 : 287 – 293 . Google Scholar CrossRef Search ADS PubMed 26 Mellen PB , Bittner V , Herrington DM. Post-challenge glucose predicts coronary atherosclerotic progression in non-diabetic, post-menopausal women . Diabet Med 2007 ; 24 : 1156 – 1159 . Google Scholar CrossRef Search ADS PubMed 27 Hanefeld M , Koehler C , Henkel E , Fuecker K , Schaper F , Temelkova-Kurktschiev T. Post-challenge hyperglycaemia relates more strongly than fasting hyperglycaemia with carotid intima-media thickness: the RIAD Study. Risk factors in impaired glucose tolerance for atherosclerosis and diabetes . Diabet Med 2000 ; 17 : 835 – 840 . Google Scholar CrossRef Search ADS PubMed 28 Temelkova-Kurktschiev TS , Koehler C , Henkel E , Leonhardt W , Fuecker K , Hanefeld M. Postchallenge plasma glucose and glycemic spikes are more strongly associated with atherosclerosis than fasting glucose or HbA1c level . Diabetes Care 2000 ; 23 : 1830 – 1834 . Google Scholar CrossRef Search ADS PubMed 29 Timmer JR , Hoekstra M , Nijsten MW , van der Horst IC , Ottervanger JP , Slingerland RJ , Dambrink JH , Bilo HJ , Zijlstra F , van 't Hof AW. Prognostic value of admission glycosylated hemoglobin and glucose in nondiabetic patients with ST-segment-elevation myocardial infarction treated with percutaneous coronary intervention . Circulation 2011 ; 124 : 704 – 711 . Google Scholar CrossRef Search ADS PubMed 30 Moura FA , Figueiredo VN , Teles BS , Barbosa MA , Pereira LR , Costa AP , Carvalho LS , Cintra RM , Almeida OL , Quinaglia E Silva JC , Nadruz JW , Sposito AC. Glycosylated hemoglobin is associated with decreased endothelial function, high inflammatory response, and adverse clinical outcome in non-diabetic STEMI patients . Atherosclerosis 2015 ; 243 : 124 – 130 . Google Scholar CrossRef Search ADS PubMed 31 Geng J , Zhang Y , Wang B , Xie J , Xu B , Li J. Glycosylated hemoglobin levels and clinical outcomes in nondiabetic patients with coronary artery disease: a meta-analysis . Medicine (Baltimore) 2017 ; 96 : e6784 . Google Scholar CrossRef Search ADS PubMed 32 Pararajasingam G , Hofsten DE , Logstrup BB , Egstrup M , Henriksen FL , Hangaard J , Egstrup K. Newly detected abnormal glucose regulation and long-term prognosis after acute myocardial infarction: comparison of an oral glucose tolerance test and glycosylated haemoglobin A1c . Int J Cardiol 2016 ; 214 : 310 – 315 . Google Scholar CrossRef Search ADS PubMed 33 Shin D , Ahn J , Cha KS , Park JS , Oh JH , Lee HW , Hong JY , Kim BW , Hong TJ. Impact of initial glycosylated hemoglobin level on cardiovascular outcomes in prediabetic patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention . Coron Artery Dis 2016 ; 27 : 40 – 46 . Google Scholar CrossRef Search ADS PubMed 34 Lazzeri C , Valente S , Chiostri M , Attana P , Mattesini A , Nesti M , Gensini GF. Glycated haemoglobin and long-term mortality in patients with ST Elevation Myocardial Infarction . J Cardiovasc Med (Hagerstown) 2015 ; 16 : 404 – 408 . Google Scholar CrossRef Search ADS PubMed 35 Shahim B , De Bacquer D , De Backer G , Gyberg V , Kotseva K , Mellbin L , Schnell O , Tuomilehto J , Wood D , Rydén L. The prognostic value of fasting plasma glucose, two-hour postload glucose, and HbA1c in patients with coronary artery disease: a report from EUROASPIRE IV: a survey from the European Society of Cardiology . Diabetes Care 2017 ; 40 : 1233 – 1240 . Google Scholar CrossRef Search ADS PubMed 36 Tailakh MA , Friger M , Zahger D , Sidi A , Mazor-Dray E , Novack V. Prospective study of the impact of diabetes mellitus newly diagnosed by glycated hemoglobin on outcomes in patients undergoing percutaneous coronary intervention . Eur J Intern Med 2017 ; 37 : 69 – 74 . Google Scholar CrossRef Search ADS PubMed 37 Kowalczyk J , Mazurek M , Zielinska T , Lenarczyk R , Sedkowska A , Swiatkowski A , Sredniawa B , Mencel G , Francuz P , Kalarus Z. Prognostic significance of HbA1c in patients with AMI treated invasively and newly detected glucose abnormalities . Eur J Prev Cardiol 2015 ; 22 : 798 – 806 . Google Scholar CrossRef Search ADS PubMed 38 Gyberg V , De Bacquer D , Kotseva K , De Backer G , Schnell O , Sundvall J , Tuomilehto J , Wood D , Rydén L. Screening for dysglycaemia in patients with coronary artery disease as reflected by fasting glucose, oral glucose tolerance test, and HbA1c: a report from EUROASPIRE IV—a survey from the European Society of Cardiology . Eur Heart J 2015 ; 36 : 1171 – 1177 . Google Scholar CrossRef Search ADS PubMed 39 Sattar N , Preiss D. Screening for diabetes in patients with cardiovascular disease: HbA1c trumps oral glucose tolerance testing . Lancet Diabetes Endocrinol 2016 ; 4 : 560 – 562 . Google Scholar CrossRef Search ADS PubMed 40 Di Angelantonio E , Gao P , Khan H , Butterworth AS , Wormser D , Kaptoge S , Kondapally Seshasai S Sr , Thompson A , Sarwar N , Willeit P , Ridker PM , Barr EL , Khaw KT , Psaty BM , Brenner H , Balkau B , Dekker JM , Lawlor DA , Daimon M , Willeit J , Njølstad I , Nissinen A , Brunner EJ , Kuller LH , Price JF , Sundström J , Knuiman MW , Feskens EJ , Verschuren WM , Wald N , Bakker SJ , Whincup PH , Ford I , Goldbourt U , Gómez-de-la-Cámara A , Gallacher J , Simons LA , Rosengren A , Sutherland SE , Björkelund C , Blazer DG , Wassertheil-Smoller S , Onat A , Marín Ibañez A , Casiglia E , Jukema JW , Simpson LM , Giampaoli S , Nordestgaard BG , Selmer R , Wennberg P , Kauhanen J , Salonen JT , Dankner R , Barrett-Connor E , Kavousi M , Gudnason V , Evans D , Wallace RB , Cushman M , D’agostino RB , Umans JG , Kiyohara Y , Nakagawa H , Sato S , Gillum RF , Folsom AR , van der Schouw YT , Moons KG , Griffin SJ , Sattar N , Wareham NJ , Selvin E , Thompson SG , Danesh J. Glycated hemoglobin measurement and prediction of cardiovascular disease . JAMA 2014 ; 311 : 1225 – 1233 . Google Scholar CrossRef Search ADS PubMed 41 Colagiuri S , Lee CM , Wong TY , Balkau B , Shaw JE , Borch-Johnsen K. Glycemic thresholds for diabetes-specific retinopathy: implications for diagnostic criteria for diabetes . Diabetes Care 2011 ; 34 : 145 – 150 . Google Scholar CrossRef Search ADS PubMed 42 Gitt A , Bramlage P , Towae F , Papp A , Deeg E , Senges J , Zeymer U , Zahn R , Fleischmann H , Boehler S , Schnell O , Standl E , Tschöpe D. Newly diagnosed diabetes and pre-diabetes in patients after acute myocardial infarction predicts adverse outcomes during a three-year follow-up: results of the Sweetheart Registry . J Am Coll Cardiol 2013 ; 61 : E176 . 10_S Google Scholar CrossRef Search ADS 43 Chen HC , Kodell RL , Cheng KF , Chen JJ. Assessment of performance of survival prediction models for cancer prognosis . BMC Med Res Methodol 2012 ; 12 : 102 – 112 . Google Scholar CrossRef Search ADS PubMed 44 Pencina MJ , D'Agostino RB Sr , Demler OV. Novel metrics for evaluating improvement in discrimination: net reclassification and integrated discrimination improvement for normal variables and nested models . Stat Med 2012 ; 31 : 101 – 113 . Google Scholar CrossRef Search ADS PubMed 45 Pencina MJ , D' Agostino RB , D' Agostino RB , 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 . Google Scholar CrossRef Search ADS PubMed 46 Wallander M , Malmberg K , Norhammar A , Ryden L , Tenerz A. Oral glucose tolerance test: a reliable tool for early detection of glucose abnormalities in patients with acute myocardial infarction in clinical practice: a report on repeated oral glucose tolerance tests from the GAMI study . Diabetes Care 2008 ; 31 : 36 – 38 . Google Scholar CrossRef Search ADS PubMed 47 Tenerz A , Norhammar A , Silveira A , Hamsten A , Nilsson G , Ryden L , Malmberg K. Diabetes, insulin resistance, and the metabolic syndrome in patients with acute myocardial infarction without previously known diabetes . Diabetes Care 2003 ; 26 : 2770 – 2776 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Heart JournalOxford University Press

Published: Aug 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off