Diabetic Retinopathy Grade as a Predictive Marker of Stroke and Mortality in Type 2 Diabetes Mellitus Patients with Amputations—Seven-Year Follow-Up

Abstract

We explored the association of grade of Diabetic Retinopathy (DR) with amputations in 2500, T2DM patients (males 1295, females 1205) who were screened for DR in year 2010, underwent Mini Mental State Examination (MMSE) for cognitive decline, and followed-up for 7 years, till 2017. Cohort was stratified into tertiles based on demographics, grade of DR and type of amputation. DR was observed in 56% patients [Non Proliferative DR (NPDR) n=962, 38%; Proliferative DR (PDR) n=438, 18%] demonstrating a strong association with all types of amputations (p<0.00001), with highest association for amputation of both limbs (n=640) (p<0.00001). Significant association was observed with comorbidities of hypertension, dyslipidemia and hyperhomocysteinemia (p<0.00001). The age tertile of 40-60 years with amputations reported highest number of NPDR and PDR, n=561 and n=263 respectively; highest tertile of HbA1c (>15%) and duration of diabetes (>15 years) corroborated with the high incidence of NPDR (n=421, 70%) and (n=449, 70%), respectively (p<0.00001). MMSE scores revealed a moderate to severe cognitive decline, NPDR (n=453) > PDR (n=231) (p<0.00001). Seven years follow-up revealed high incidence of new amputations (n=610; NPDR n= 372, PDR n=238) (p<0.00001). 637 patients with DR (NPDR, n= 515; PDR n=122) developed stroke (p<0.00001). Two or more fundus lesions increased the propensity of stroke and mortality. The mortality in patients with DR was 33.76% (n=844) of which 282 patients died due to stroke (p<0.00001). Overall, ischemic stroke was the leading cause of mortality in 83.82% patients, with lacunar stroke (n=340) as the commonest anatomical sub-type of stroke (p<0.00001). Patients with DR, especially with NPDR should be managed intensively to prevent fatal cerebrovascular events. Diagnosis of DR in patients who underwent amputations, can be used as a stratified model in association with comorbidities to predict stroke and the mortality. Disclosure K. Shah: None. N. Sundaram: None. A.A. Raut: None. K. Rao: None. N. Wadhwa: None.