Indra Steens^a,b; Yuekai Ji^c; Sanaz Sedaghat^c; Jim Pankow^c; Barbara Klein^d; Mary Frances Cotch^e; Robyn McClelland^f; Kathleen Hayden^g; Timothy Hughes^h,i; Michael Bancksⁱ; Alfons Houben^a,b; Abraham Kroon^a,b; Annemarie Koster^j,k; Hans Bosma^j,k; Tos Berendschot^l; Bastiaan de Galan^a,b,m; Thomas van Slotenⁿ.

^aCARIM School for Cardiovascular Diseases, UM, Maastricht, The Netherlands. ^bInternal Medicine, MUMC+, Maastricht, The Netherlands. ^cDivision of Epidemiology and Community Health, School of Public Health, University of Minnesota Medical School, Minneapolis. ^dOcular Epidemiology, University of Wisconsin-Madison. ^eDivision of Epidemiology and Clinical Applications, Intramural Research Program, National Eye Institute, National Institutes of Health, Bethesda, Maryland, USA. ^fDepartment of Biostatistics, University of Washington, Seattle, Washington, USA. ^gDepartment of Social Sciences and Health Policy, Wake Forest School of Medicine, Winston Salem, NC. ^hSection of Gerontology and Geriatric Medicine, Department of Internal Medicine, Wake Forest School of Medicine, Winston Salem, NC. ⁱDivision of Public Health Sciences, Department of Epidemiology & Prevention, Wake Forest School of Medicine, Winston-Salem, NC. ^jCAPHRI Care and Public Health Research Institute, UM, MD, The Netherlands. ^kDepartment of Social Medicine, UM, Maastricht, The Netherlands. ^lDepartment of Ophthalmology, MUMC+, Maastricht, the Netherlands. ^mDepartment of Internal Medicine, Radboud University Medical Center, Nijmegen, the Netherlands. ⁿDepartment of Vascular Medicine, University Medical Centre Utrecht, Utrecht, The Netherlands. Current e-mail address of the presenting (first) author: i.steens@maastrichtuniversity.nl

Background Type 2 diabetes increases the risk of stroke, but the underlying mechanism remains to be determined. Cerebral microvascular dysfunction is a potential mechanism underlying this increased risk. The retinal microvasculature, sharing similarities with cerebral microvasculature, offers a window for investigating cerebral microvasculature. We aimed to explore whether retinal microvascular dysfunction mediates the association between type 2 diabetes and incident stroke.

Methods We conducted a pooled analysis using data from three large cohort studies: The Maastricht Study, the Multi-Ethnic Study of Atherosclerosis (MESA), and the Atherosclerosis Risk in Communities study (ARIC). Causal mediation analysis was done to investigate whether central retinal venular equivalent (CRVE) and central retinal arteriolar equivalent (CRAE) mediated the association between type 2 diabetes and incident stroke.

Results A total of 16,806 individuals were included, comprising 5,469 from The Maastricht Study (mean age 59.5 ± 8.7 years, 50.2% women), 3,958 from MESA (mean age 62.30 ± 9.75 years, 53.8% women) and 7,379 from ARIC (mean age 60.0 ± 5.6 years, 54.4% women). Type 2 diabetes was associated with a higher incidence of stroke (hazard ratio [95% confidence interval]: 1.43 [1.13, 1.80]). Type 2 diabetes was also associated with a higher CRVE (0.11 [0.01, .22]) and a higher CRAE (0.13 [0.08, 0.17]) However, CRVE and CRAE did not mediate the association between type 2 diabetes and stroke, neither in the pooled analysis nor in analyses of the individual cohorts.

Conclusions Retinal microvascular dysfunction did not mediate the increased risk of stroke associated with type 2 diabetes. Future studies should explore whether microvascular dysfunction plays a role in distinct subtypes of diabetes-related stroke.