Introduction:
Cerebrovascular small vessel disease, seen on brain imaging as lacunes and white matter hyperintensities (WMH), is a substrate for dementia in older adults. Diffusion tensor imaging (DTI) is thought to provide early signs of loss of white matter (WM) integrity due to microvascular disease and predicts WM hyperintensity volume. Retinal fundus photography provides surrogate measures of cerebral microvasculature. No studies have quantified the long-term association between retinal signs and DTI measures.
Hypothesis:
Microvascular retinal signs measured in midlife are associated with small vessel disease measured on brain magnetic resonance imaging (MRI) 18 years later, including reduced WM microstructural integrity (lower fractional anisotrophy [FA] and greater mean diffusivity [MD] by DTI), greater WM hyperintensity volume and greater lacune prevalence.
Methods:
In a biracial prospective cohort study, retinal signs were measured using fundus photography (1993-1995) with 3-T magnetic resonance imaging conducted in 2011-13. Multivariable-adjusted linear regression was used to quantify the relationships of retinal signs with WM measures. Prevalence of lacunar infarcts by retinal sign status was estimated using log binomial regression. Analyses were adjusted for age [linear and quadratic terms], education, sex, race, intracranial volume, body mass index, smoking, diabetes, hypertension, and ≥1
APOE
ε4 alleles.
Results:
In 1829 men and women (60% [N=1100] female, 27% [N=489] black race, aged 50-72 years when retinal signs were measured), a binary measure comprised of two retinal signs suggestive of arteriolar damage due to hypertension (focal arteriolar narrowing and/or arteriovenous nicking) was associated with worse (lower) FA (standardized β=-0.19, 95% confidence interval [CI]=-0.35, -0.02), worse (higher) MD (β=0.15, 95% CI=0.00, 0.30), greater WM hyperintensity volume (β=0.15, 95% CI=0.01, 0.30), and greater prevalence of lacunes (prevalence ratio=1.33, 95% CI: 0.99, 1.80). Generalized arteriolar narrowing, measured as the central retinal arteriolar equivalent (CRAE, narrowest quartile vs. widest three quartiles) was associated with worse FA (β=-0.13, 95% CI=-0.24, -0.01) and worse MD (β=0.12, 95% CI=0.01, 0.23). Results did not differ by sex, race, hypertension status or
APOE
ε4 genotype. No associations were found for retinopathy, but only 56 participants had retinopathy.
Conclusions:
Consistent with prior work, and as expected based on a common underlying pathology, retinal signs predicted WM disease and lacunar infarcts 18 years later. Novel to this study, we found that retinal signs related to arteriolar damage also predicted loss of white matter microvascular integrity measured using DTI.
Response by Loos et al to Letter Regarding Article, “Long-Term Morphological Changes of Symptomatic Lacunar Infarcts and Surrounding White Matter on Structural Magnetic Resonance Imaging”
