Migraine and retinal microvascular abnormalities The Atherosclerosis Risk in Communities Study K.M. Rose, PhD T.Y. Wong, MD, PhD A.P. Carson, PhD D.J. Couper, PhD R. Klein, MD, MPH A.R. Sharrett, MD, PhD ABSTRACT Objective: This study examined the association between vascular headaches and retinal microvascular disease. Methods: We investigated the cross-sectional association between headaches (migraine/other headaches with aura, migraine without aura, other headaches without aura, no head- aches) and retinal microvascular signs (retinopathy, focal arteriolar narrowing, arteriovenous nicking; arteriolar and venular calibers) among middle-aged African American and white men and women from the third examination of the Atherosclerosis Risk in Communities Study (1993 through 1995). Results: After controlling for age, gender, race, study center, and cardiovascular risk factors, we de- termined that persons with headaches were more likely to have retinopathy than those without a history of headaches (odds ratio [OR]  1.38, 95% CI  0.96 to 1.99 for migraine/other headaches with aura; OR  1.49, 95% CI  1.05 to 2.12 for migraine without aura; and OR  1.28, 95% CI  0.99 to 1.65 for other headaches). Associations with migraine were stronger among the subset of participants without a history of diabetes or hypertension (OR  1.79, 95% CI  1.09 to 2.95 for migraine/other headaches with aura; and OR  1.74, 95% CI  1.11 to 2.71 for migraine without aura). Headaches were not associated with focal arteriolar narrowing or arteriovenous nicking. Per- sons with headaches tended to have smaller mean arteriolar and venular calibers; however, these associations did not tend to persist among those without hypertension or diabetes. Conclusion: Middle-aged persons with migraine and other headaches were more likely to have retinopathy signs, supporting the hypothesis that neurovascular dysfunction may underlie vascular headaches. NEUROLOGY 2007;68:1694–1700 Migraine headaches affect approximately 17% of women and 6% of men in the United States 1,2 and are more common among younger adults, 1,3-5 women, 1,3 and Caucasians. 1,3 Whereas the pathophysiologic mechanisms underlying migraine are unclear, there is evidence of both underlying vascular and neurologic components. 2,6,7 Migraine has been consistently associated with an increased risk of stroke, 8-10 but inconsistent associations have been re- ported for both hypertension 3,11-15 and coronary heart disease. 9,14,16-19 There have been further suggestions that the associations of migraine with cardiovascular disease (CVD) differ by migraine subtypes. For example, recent reports suggest that associations are stronger 19-21 or limited 22,23 to headaches with aura symptoms. Retinal microvascular abnormalities are associated with hypertension 24,25 and have re- cently been shown to predict clinical and subclinical stroke 26-28 and other cardiovascular outcomes. 25,27,29,30 Because the retinal and cerebral microcirculations share similar anatomy, embryology, and physiology, and because neurovascular mechanisms may underlie both mi- graine and retinal microvascular abnormalities, an association between the two is biologi- cally plausible. Case reports have linked retinal vein occlusions 31,32 and retinal infarctions 33,34 with migraine. A recent population-based study in a white Australian population (age 49 and older) reported smaller retinal arteriolar calibers among persons with a history of migraine without aura as compared with those without a history of migraine. 35 It is unclear if similar From the Departments of Epidemiology (K.M.R., A.P.C.) and Biostatistics (D.J.C.), University of North Carolina at Chapel Hill, Department of Ophthalmology and Visual Sciences (R.K.), University of Wisconsin Medical School, Madison, and Department of Epidemiology (A.R.S.), Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD; and Centre for Eye Research Australia (T.Y.W.), University of Melbourne, Melbourne, Australia. The Atherosclerosis Risk in Communities Study is a collaborative study supported by the National Heart Blood and Lung Institute contracts N01- HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022. Disclosure: In 2001 Dr. Rose was a paid consultant for and received a small grant from GlaxoSmithKline. These funds did not support the research presented in the current manuscript. Address correspondence and reprint requests to Dr. K.M. Rose, Department of Epidemiology, University of North Carolina at Chapel Hill, 137 E. Franklin St., Suite 306, Chapel Hill, NC 27514 kathyrn_rose@unc.edu 1694 Copyright © 2007 by AAN Enterprises, Inc.