Prevalence and outcomes of unrecognized diabetes mellitus and prediabetes among acute stroke patients with admission hyperglycemia at the Philippine General Hospital: DASH study

Abstract

Background: Hyperglycemia occurs in 60% of acute stroke patients, 12-52% of which maybe previously undiagnosed with diabetes. A retrospective study by Gacutan showed21% admission hyperglycemia at the Philippine General Hospital. Screening for glucose disorders among hyperglycemic stroke patients provides a venue for secondary prevention. There are no current guidelines on screening for post-stroke hyperglycemia by the Stroke Society of the Philippines. Objective: To determine the prevalence and outcomes of unrecognized diabetes and prediabetes among acute stroke patients. Secondarily, we aim to determine the clinical and biochemical predictors of unrecognized/newly diagnosed diabetes mellitus among the acute stroke patients. Methodology: Cross-sectional prospective study. Acute stroke patients admitted from January to December, 2008 with admission hyperglycemia were included. Random blood sugar, glycosylated hemoglobin & the National Institute of Health Stroke Scale (NIHSS) were done on admission. All were followed up six weeks post-discharge for an FBS and 75-grams OGTT except those discharged with anti-diabetes medications who were considered newly-diagnosed diabetes. Outcomes: Glucose disorders were classified at six weeks post-discharge. Mortality was measured at six weeks and 12 weeks. Functional outcomes using the Modified Rankin Score & the Barthel’s Index were measured at the OPD 6-weeks postdischarge. Results: There were 504 stroke patients of which 178(35%) had admission hyperglycemia. Majority were newly diagnosed diabetes 32.7%, stress hyperglycemia 29.56%, impaired glucosetolerance, (IGT) 13.8%, impaired fasting blood sugar, (IFG) 8.8%, combined IGT and IFG 3.7%, previous diabetes 11.32%. Age, body mass index, alcohol abuse, smoking, hypertension, nosocomial infections, acute coronary syndromewere associated with diabetes. No association was noted with diabetes and atrial fibrillation. By Kruskal-Wallis test, admission RBS and HbA1c were higher among diabetics compared to prediabetics and stress hyperglycemics. Triglycerides was higher and HDL was lower among diabeticversus the stress hyperglycemia. No trends were noted for total cholesterol and LDL cholesterol across groups. On multiple logistic regression, age (OR: 1.06, p0.07), smoking (OR: 9.81, p0.05) & Hba1c (OR:11.39, p<0.01) were the predictors for newly diagnosed diabetes classification among the acute stroke patients. Higher Modified Rankin and lower Barthel’s index were seen among diabetics compared to the other groups measured on follow-up. Mortality at six weeks was higher among diabetics (OR: 2.34, p<0.01) compared to non-diabetics. A trend towards higher mortality at 12 weeks was also noted among diabetics (OR:1.81, p0.063) compared to other groups. Conclusion: Admission hyperglycemia was seen in 35% of admitted stroke patients, majority of which were unrecognized diabetes. By screening for post-stroke hyperglycemia, we identified 59.1% (32.7% unrecognized diabetes; 26.4% prediabetes) patients eligible for treatment and secondary  prevention. Short-term mortality was higher among acute stroke patients with diabetes. Age, smoking and admission HbA1c were the clinical and biochemical predictors of a newly diagnoseddiabetes mellitus. Keywords: Admission hyperglycemia; diabetes; prediabetes; cerebrovascular disease Philippine Journal of Internal Medicine, Vol. 49 No. 2: Apr-Jun 2011, pp79-87