Retinal Microvascular Abnormalities and MRI-Defined Subclinical Cerebral Infarction
The Atherosclerosis Risk in Communities Study
Background and Purpose— Retinal microvascular abnormalities reflect cumulative small vessel damage from elevated blood pressure and may reflect subclinical cerebral microvascular changes. We examined their associations with MRI-defined cerebral infarcts.
Methods— Population-based, cross-sectional study of 1684 persons 55 to 74 years of age without a history of clinical stroke, sampled from 2 US southeastern communities. Retinal photographs were obtained and graded for presence of retinal microvascular abnormalities, including arteriovenous nicking, focal arteriolar narrowing, retinal hemorrhages, soft exudates and microaneurysms. Photographs were also digitized, and retinal vessel diameters were measured and summarized as the arteriole-to-venule ratio (AVR). Cerebral MRI scans were graded for presence of cerebral infarct, defined as a lesion ≥3 mm diameter in a vascular distribution with typical imaging characteristics.
Results— There were a total of 183 MRI cerebral infarcts. After adjustment for age, gender, race, 6-year mean arterial blood pressure, diabetes, and other stroke risk factors, cerebral infarcts were associated with retinal microvascular abnormalities, with odds ratios 1.90 (95% CI, 1.25 to 2.88) for arteriovenous nicking, 1.89 (95% CI, 1.22 to 2.92) for focal arteriolar narrowing, 2.95 (95% CI, 1.30 to 6.71) for blot hemorrhages, 2.08 (95% CI, 0.69, 6.31) for soft exudates, 3.17 (95% CI, 1.05 to 9.64) for microaneurysms, and 1.74 (95% CI, 0.95 to 3.21) for smallest compared with largest AVR. In stratified analyses, these associations were only present in persons with hypertension.
Conclusions— Retinal microvascular abnormalities are associated with MRI-defined subclinical cerebral infarcts independent of stroke risk factors. These data suggest that retinal photography may be useful for studying subclinical cerebrovascular disease in population-based studies.
Stroke is an important cause of mortality and morbidity in the United States. Because current stroke risk prediction has inherent limitations, there is a need to search for additional predictors. The retinal vasculature permits noninvasive visualization of long-term, cumulative exposure to elevated blood pressure and other stroke risk factors.1 It is embryologically and anatomically part of the cerebral vasculature.2,3 Recent studies have shown retinal microvascular abnormalities (eg, arteriovenous nicking, focal arteriolar narrowing, and microaneurysms; Figure) are associated with current and past blood pressure4,5 and predict the development of incident clinical stroke, independent of vascular risk factors.6
It is unclear whether retinal microvascular abnormalities are associated with subclinical cerebrovascular disease. Cerebral infarction diagnosed via MRI is correlated with pathologic evidence of infarction7 and is strongly related to stroke risk factors,8,9 motor and cognitive dysfunction,10 and prevalent11 and incident12 clinical stroke. In the current study, we examined the relationship of retinal microvascular abnormalities with MRI-defined subclinical cerebral infarcts.
The Atherosclerosis Risk in Communities (ARIC) Study has been described previously.13 From 1987 to 1989, 15 792 women and men were recruited, selected as probability samples of 45- to 64-year-old residents of 4 US communities in Forsyth County, NC, Jackson, Miss (blacks only), suburbs of Minneapolis, Minn, and Washington County, Md.
The study population was derived from participants at the Forsyth County and Jackson study sites who returned for the third ARIC study examination conducted from 1993 through 1995, when cerebral MRI and retinal photography were first performed.14 During this examination, a sample of the cohort at the 2 sites (n=2874) was screened for eligibility for cerebral MRI. For participant safety, the following criteria were used to exclude ineligible persons from the magnetic resonance study. Persons were excluded for: (1) having ever undergone surgery for an aneurysm in the brain; (2) having metal fragments in the eyes, brain, or spinal cord; (3) having a valvular prosthesis, a cardiac pacemaker, cochlear implant, spinal cord stimulator, or other internal electrical device; (4) being pregnant; and (5) having occupational exposure to metal fragments. The MRI examinees could be considered a random sample of the full cohort because examination dates were allocated at baseline through randomly selected induction cycles, and re-examination visits were scheduled according to the anniversary date to the degree possible. Of these, 106 were ineligible, 649 declined to undergo the cerebral magnetic resonance examinations, and 204 had ungradeable MRI data. Of the remaining 1915 who had cerebral MRI data, 50 had a previous clinical stroke event, and 181 had no retinal photographs or ungradeable photographs, leaving 1684 for the current analysis. Institutional review boards at each study site approved the study. Informed consent was obtained from all participants, and the study was conducted in accordance with the Declaration of Helsinki.
MRI Procedures and Definition of Cerebral Infarct
The MRI scanning and reading protocols and definition of cerebral infarcts are described in detail previously.15 A standardized protocol was used to scan participants at the field centers. These scans were read at the MRI reading center by trained readers, masked to subject identity. Cerebral infarct was defined as a lesion ≥3 mm in maximum diameter in a vascular distribution with typical imaging characteristics15 and was distinguished from smaller, more common cerebral white matter lesions. Lacunar infarct was defined as a cerebral infarct (by the above criteria) with maximum diameter <2 cm located in the brain stem, basal ganglia, internal capsule, thalamus, or deep cerebral white matter. All MRI scans were reviewed by a second reader, and disagreements were adjudicated by consensus. A total of 144 randomly selected cases were reread for the purpose of analyzing lesion detection reproducibility. Inter-reader and intrareader κ ranged from 0.52 and 0.78, respectively.15
Retinal Photographic Grading and Definitions
Retinal photography and the grading procedure have been described in detail previously.4,16 After 5 minutes of dark adaptation, a 45° retinal photograph was taken in 1 randomly selected eye, centered on the region of the optic disc and the macula, using an auto-focus camera. The photographic slides were evaluated by trained graders, masked to subject identity, using an established grading system.16 Focal microvascular abnormalities analyzed in this report include arteriovenous nicking, focal arteriolar narrowing, blot hemorrhages, soft exudates (cotton wool spots), and microaneurysms; other lesions were too infrequent for analysis.4 Quality control measures have been reported previously.17 In general, intragrader and intergrader κ statistics ranged from 0.61 to 1.00.
In addition to focal grading, a semiquantitative method estimated generalized arteriolar narrowing, expressed as arteriole-to-venule ratio (AVR).16 A smaller AVR represents narrower arterioles relative to venular caliber and is associated with higher blood pressure. This measurement has been shown to be reproducible17 and sensitive to modest blood pressure elevations and to reflect the effect of previous pressure elevations independent of the currently measured blood pressure.4
Definitions of Stroke Risk Factors
Participants underwent standardized evaluations of cardiovascular risk factors at all 3 examinations. Blood pressure was measured with a random-zero sphygmomanometer, and the average of the second and third readings of the measurements was used in this analysis. The mean arterial blood pressure was computed as two thirds of the diastolic plus one third of the systolic value, and the average of this over all 3 examinations (ie, 6-year mean arterial blood pressure) was used in all multivariate analyses to assess the effects of the retinal signs independent of measure blood pressure.
Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or use of antihypertensive medication during the previous 2 weeks. Diabetes mellitus was defined as a fasting glucose ≥7.0 mmol/L (≥126 mg/dL), a nonfasting glucose ≥11.1 mmol/L (≥200 mg/dL), or a self-reported history of treatment for diabetes. A person was considered to have hypertension or diabetes if these criteria were met at any of the 3 examinations.
Values from the third examination, if available, were used for other variables. Cigarette smoking status was characterized as never, former, or current cigarette smoker. Methods used for measurements of total cholesterol and triglycerides, high-density lipoprotein cholesterol, and serum fibrinogen are based on techniques described previously.13
All statistical procedures were based on SAS software. Focal retinal variables were dichotomized as present versus absent for analysis. Because the relationship between AVR and cerebral infarction was nonlinear, AVR was divided into quintiles for analysis, with the fifth or highest quintile (interpreted as no generalized arteriolar narrowing) serving as reference. χ2 and t test were used to compare categorical and continuous variables, respectively, whereas test of trend for infarct prevalence across AVR quintiles was based on the Mantel–Haenszel χ2 statistic. Multivariable odds ratios (ORs) for presence of MRI cerebral infarct as a function of retinal abnormalities were calculated using logistic regression analysis.
Compared with participants in the Forsyth County and Jackson study sites who returned for the third examination not included in the current study (n=4265), the study participants (n=1684) were younger, less likely blacks, current smokers, and to have previous myocardial infarction (Table 1). The study group also had lower diastolic blood pressure and body mass index and higher total and high-density lipoprotein cholesterol levels.
There were 183 MRI cerebral infarcts, 125 (68.3%) in hypertensives, and 58 (31.7%) in normotensives. The prevalence of MRI cerebral infarcts according to presence or absence of specific retinal microvascular abnormalities and to quintiles of AVR is shown in Table 2. The prevalence of MRI infarcts was higher in the presence of all retinal microvascular signs. This was stronger in persons with hypertension than in those without hypertension.
Table 3 shows the association of retinal microvascular signs and MRI infarcts in the total population and in persons with and without hypertension. After adjustment for age, gender and race/center, all retinal variables were significantly associated with MRI cerebral infarction, with OR ranging from 1.89 for arteriovenous nicking to 4.04 for blot hemorrhages. Additional adjustment for 6-year mean blood pressure, diabetes, and other risk factors attenuated some of these associations (OR ranging from 1.74 to 3.17). Among hypertensive persons, the presence of all retinal microvascular signs was associated with MRI infarcts (multivariable OR ranging from 2.30 to 4.24). Among normotensive persons, none of the retinal signs were associated with retinal microvascular signs.
Table 4 shows further analyses confined to persons with hypertension. The associations of the 3 most common retinal abnormalities (arteriovenous nicking, focal arteriolar narrowing, and AVR) were entered simultaneously in models adjusting for stroke risk factors. After multivariable adjustment, arteriovenous nicking (OR, 2.47), focal arteriolar narrowing (OR, 1.78), and smaller AVR (OR, 4.76 for first versus fifth quintile) were significantly associated with MRI cerebral infarction. Excluding persons with diabetes, with previous transient ischemic attacks, and diabetes and previous transient ischemic attacks had little impact on the results shown.
Results were overall similar when persons with ≥1 nonlacunar infarct were excluded, in persons older and younger than 65 years of age, and for past and current nonsmokers (data not shown).
These population-based data show that in middle-aged persons, retinal microvascular signs are associated with MRI-defined cerebral infarction, even after adjustment for blood pressure and other stroke risk factors. In stratified analyses, these associations were confined to persons with hypertension.
The retinal microvascular signs assessed here are thought to largely reflect arteriolar damage from hypertension,1,2 and possibly inflammation and endothelial dysfunction.18 The association of these retinal signs with MRI-defined cerebral infarcts, most of which is lacunar, may therefore reflect similar arteriolar processes in the cerebral microvasculature.
These findings are supported by data from epidemiological studies using less quantitative clinical assessment of retinal microvascular signs.19,20 In a Japanese study in the 1970s, retinal microvascular abnormalities diagnosed by ophthalmoscopy were associated with hemorrhagic and thrombotic stroke.19 A more recent study describes associations “retinal arterial sclerosis” and small infarcts detected by cerebral CT or MRI.20 The current findings are also consistent with our previous analyses of retinal microvascular signs, with clinical stroke6 cognitive impairment,21 and MRI white matter lesions,14 in the ARIC population.
Anatomical, histopathological, and functional studies provide further biological basis for our observations.3 In an autopsy study of stroke decedents, retinal arteriolar fibrous thickenings were seen in all 19 stroke cases but only 12 of 21 disease-free controls.22 Another study demonstrated functional alterations in the retinal vasculature (eg, a slow retinal arteriolar-venule passage time) in persons with lacunar stroke.23
Strengths of this study include applicability to contemporary populations with low prevalence of severe hypertension, standardized retinal and cerebral examinations, and the inclusion of blacks. Limitations include the following. First, because of the cross-sectional design, it is possible that some of the cerebral infarctions detected by MRI may have preceded the retinal lesions. However, prospective studies suggest that retinal findings precede clinical stroke.6,6 Second, possible selection biases should be considered because a significant proportion of participants did not have MRI or retinal data. Those excluded tended to be older and to have a higher prevalence of cardiovascular risk factors (Table 1). If these people were also more likely to have retinopathy signs and MRI infarcts, the associations could have been attenuated.
In summary, we demonstrate a strong association between retinal microvascular signs and MRI-defined cerebral infarction. These associations were seen only in persons with hypertension. The data suggest that a careful examination of retinal vessels and their changes by retinal photography may be useful for investigating the contribution of small vessel disease to the pathogenesis of cerebrovascular24 and perhaps other cardiovascular diseases.25
This study was supported by contracts N01 HC-35125, N01 HC-35126, N01 HC-55015, N01-HC-55016, N01 HC-55018, N01 HC-55019, N01 HC-55020, and N01 HC-55022 from the National Heart, Lung, and Blood Institute, Bethesda, MD. The authors thank the staff and participants in the ARIC study for their important contributions. Additional support was provided by grant R21-HL077166 from NHLBI, NIH and the Sylvia and Charles Viertel Clinical Investigator Award (T.Y.W.).
- Received August 14, 2005.
- Revision received September 7, 2005.
- Accepted October 11, 2005.
Sharrett AR, Hubbard LD, Cooper LS, Sorlie PD, Brothers RJ, Nieto FJ, Pinsky JL, Klein R. Retinal arteriolar diameters and elevated blood pressure: the Atherosclerosis Risk in Communities Study. Am J Epidemiol. 1999; 150: 263–270.
Wong TY, Hubbard LD, Klein R, Marino EK, Kronmal R, Sharret AR, Siscovick DS, Burke G, Tielsch JM. Retinal microvascular abnormalities and blood pressure in older people: the Cardiovascular Health Study. Br J Ophthalmology. 2002; 86: 1007–1013.
Kobayashi S, Okada K, Koide H, Bokura H, Yamaguchi S. Subcortical silent brain infarction as a risk factor for clinical stroke. Stroke. 1997; 28: 1932–1939.
Howard G, Wagenknecht LE, Cai J, Cooper L, Kraut M, Toole JF. Cigarette smoking and other risk factors for silent cerebral infarction in the general population. Stroke. 1998; 29: 913–917.
Price TR, Manolio TA, Kronmal RA, Kittner SJ, Yue NC, Robbins J, Anton-Culver H, O’Leary DH. Silent brain infarction on magnetic resonance imaging and neurological abnormalities in community-dwelling older adults. The Cardiovascular Health Study. CHS Collaborative Research Group. Stroke. 1997; 28: 1158–1164.
Vermeer SE, Hollander M, van Dijk EJ, Hofman A, Koudstaal PJ, Breteler MM. Silent brain infarcts and white matter lesions increase stroke risk in the general population: the Rotterdam Scan Study. Stroke. 2003; 34: 1126–1129.
The ARIC Investigators. The Atherosclerosis Risk in Communities (ARIC) study: design and objectives. Am J Epidemiol. 1989; 129: 687–702.
Bryan RN, Cai J, Burke G, Hutchinson RG, Liao D, Toole JF, Dagher AP, Cooper L. Prevalence and anatomic characteristics of infarct-like lesions on MR images of middle-aged adults: the Atherosclerosis Risk in Communities study. AJNR Am J Neuroradiol. 1999; 20: 1273–1280.
Hubbard LD, Brothers RJ, King WN, Clegg LX, Klein R, Cooper LS, Sharrett AR, Davis MD, Cai J. Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the Atherosclerosis Risk in Communities Study. Ophthalmology. 1999; 106: 2269–2280.
Aoki N. Epidemiological evaluation of funduscopic findings in cerebrovascular diseases. II. A multivariate analysis of funduscopic findings. Jpn Circ J. 1975; 39–271–82.
Kwa VI, van der Sande JJ, Stam J, Tijmes N, Vrooland JL. Retinal arterial changes correlate with cerebral small-vessel disease. Neurology. 2002; 59: 1536–1540.
Wong TY, Klein R, Sharrett AR, Nieto FJ, Boland LL, Couper DJ, Mosley TH, Klein BE, Hubbard LD, Szklo M. Retinal microvascular abnormalities and cognitive impairment in middle-aged persons: the Atherosclerosis Risk in Communities Study. Stroke. 2002; 33: 1487–1492.
Goto I, Kimoto K, Katsuki S, Mimatsu T, Ikui H. Pathological studies on the intracerebral and retinal arteries in cerebrovascular and noncerebrovascular diseases. Stroke. 1975; 6: 263–269.
Schneider R, Rademacher M, Wolf S. Lacunar infarcts and white matter attenuation. Ophthamologic and microcirculatory aspects of the pathophysiology. Stroke. 1993; 24: 1874–1879.