Ischemic stroke (IS) is often comorbid with type 2 diabetes mellitus (T2DM) wherein the determinants of long-term outcomes, beyond the acute stroke phase, are not adequately known. This study identified the determinants of long-term outcomes for diabetic IS patients through a prospective nested case-control study in 624 patients treated with conservative measures (38.60% females, mean age: 63.85 years). After 36-month follow-up, 117 (18.8%) patients with poor outcome were enrolled in the case group. The poor outcome was defined with a modified Rankin Scale (mRS) score ≥3. Meanwhile, 374 (59.9%) patients with good outcome, defined as (mRS score <3), were included in the control group. Patients who died (n = 32) or lost to follow-up (n = 101) were excluded in analysis. Poor prognostic outcome was positively associated with (1) the pulse rate at admission, (2) diastolic blood pressure (DBP), and (3) fasting blood glucose (FBG) during follow-up, whereas physical activity and lipid-lowering treatment during follow-up were negatively associated. Importantly, a forecasting model with these indicators distinguished the patients with good versus poor outcomes with 70.1% sensitivity and 73.5% specificity. Health care professionals and laboratory medicine scholars may want to monitor an increase in DBP and FBG during follow-up, as well as physical activity and lipid-lowering treatment, in relationship to the prognosis of IS with comorbid T2DM after conservative therapies. The proposed predictive model for personalized/precision medicine requires field testing in independent studies, and might help risk stratification with theranostic tests for patients with acute IS who also have a diagnosis of T2DM.
Keywords: diagnostic medicine; ischemic stroke; personalized medicine; public health; theranostics; type 2 diabetes mellitus.