OUP user menu

Retinal vessel diameters and cerebral small vessel disease: the Rotterdam Scan Study

M. Kamran Ikram, Frank Jan De Jong, Ewoud J. Van Dijk, Niels D. Prins, Albert Hofman, Monique M. B. Breteler, Paulus T. V. M. De Jong
DOI: http://dx.doi.org/10.1093/brain/awh688 182-188 First published online: 29 November 2005

Summary

The direct visualization of retinal vessels provides a unique opportunity to study cerebral small vessel disease, because these vessels share many features. It was reported that persons with smaller retinal arteriolar-to-venular ratio tended to have more white matter lesions on MRI. It is unclear whether this is due to arteriolar narrowing or venular dilatation. We investigated whether smaller arteriolar or larger venular diameters or both were related to severity and progression of cerebral small vessel disease. We studied 490 persons (60–90 years) without dementia from a population-based cohort study. At baseline (1990–1993), retinal arteriolar and venular diameters were measured on digitized images of one eye of each participant. In 1995–1996, participants underwent cerebral MRI scanning. We rated the severity of periventricular white matter lesions on a 9-point scale, approximated a total subcortical white matter lesion volume (range: 0–29.5 ml) and rated the presence of lacunar infarcts. On average 3.3 years later, 279 persons had a second MRI. Changes in periventricular and subcortical white matter lesions were rated with a semi-quantitative scale, and progression was classified as no, minor and marked. An incident infarct was a new infarct on the follow-up MRI. Neither venular nor arteriolar diameters were related to the severity of cerebral small vessel disease. Larger venular diameters were, however, associated with a marked progression of cerebral small vessel disease. Age and gender adjusted odds ratios (ORs) per standard deviation increase were 1.71 [95% confidence interval (CI): 1.11–2.61] for periventricular, 1.72 (95% CI: 1.09–2.71) for subcortical white matter lesion progression and 1.59 (95% CI: 1.06–2.39) for incident lacunar infarcts. These associations were independent of other cardiovascular risk factors. Only the OR for incident lacunar infarcts was attenuated (1.24; 95% CI: 0.72–2.12). No association was observed between arteriolar diameters and progression of cerebral small vessel disease. In conclusion, retinal venular dilatation was related to progression of cerebral small vessel disease. The mechanisms underlying venular dilatation deserve more attention, as they may provide new clues into the pathophysiology of cerebral small vessel disease.

  • population-based cohort study
  • retinal arteriolar and venular diameters
  • cerebral small vessel disease
  • white matter lesions
  • lacunar infarcts
  • AVR = arteriolar-to-venular ratio

Introduction

Cerebral MRI in elderly people frequently reveals white matter lesions and lacunar infarcts. Prevalence of white matter lesions ranges from 5 to 90% (van Dijk et al., 2002), whereas 20% of elderly have at least one brain infarct on MRI (Vermeer et al., 2003c). These lesions are related to incident stroke and may contribute to the development of dementia (Kobayashi et al., 1997; Barber et al., 1999; Vermeer et al., 2003b; Prins et al., 2004a). They reflect ischaemic small vessel disease, of which the exact pathophysiological mechanism is unknown (Pantoni and Garcia, 1997). Several studies point towards increasing age, hypertension and markers of atherosclerosis as main risk factors (Bots et al., 1993; Longstreth et al., 1996; Liao et al., 1997; Schmidt et al., 1997). The pathological status of cerebral small vessels is difficult to assess in vivo. Most non-invasive markers of vascular pathology are related to major blood vessels outside the brain and may not represent local cerebral abnormalities. The retinal vessels provide unique opportunities to study cerebral small vessel disease, because they share similar anatomy, physiology and embryology (Wong et al., 2001).

Recently, a semi-automated system was developed to measure retinal vessel diameters (Hubbard et al., 1999; Wong et al., 2002b). Because these studies did not have enough data to correct for magnification differences due to refractive errors of eyes, an arteriolar-to-venular ratio (AVR) was introduced to bypass this problem. A smaller AVR was suggested to reflect generalized arteriolar narrowing and was associated with cardiovascular diseases (Hubbard et al., 1999; Wong et al., 2002b). In the Atherosclerosis Risk in Communities Study, persons with a smaller AVR tended to have more white matter lesions (Wong et al., 2002a). One may wonder whether the AVR reflects solely retinal arteriolar narrowing, because venular width does not remain constant in various pathological conditions. We reported that smaller arteriolar diameters were related to higher blood pressures and larger venular diameters to markers of atherosclerosis and inflammation (Ikram et al., 2004).

We investigated whether arteriolar or venular diameters in the retina were associated with the severity and progression of cerebral small vessel disease.

Participants and methods

Study population

The present study was performed as part of the Rotterdam Study, a population-based cohort study on chronic diseases in the elderly (Hofman et al., 1991). All inhabitants of a district of the city of Rotterdam aged 55 years or over were invited in random order to the study, and 7983 actually participated (overall response 78%). The study was conducted according to the Declaration of Helsinki, and the appropriate Medical Ethics Committees approved the study protocols. A written informed consent was obtained from all participants. Because the ophthalmic part became only operational after the study had started, a total of 6780 participants underwent baseline ophthalmic examination (1990–1993) (Ikram et al., 2004). From 1995 to 1996, we randomly selected participants (aged 60–90 years) of the Rotterdam Study cohort in strata of gender and 5 year age categories for participation in the Rotterdam Scan Study, a study on age-related brain changes on MRI (Breteler, 2000). Complete information, including a cerebral MRI scan, was obtained in 490 individuals, who also had had the ophthalmic examination at baseline. Of them, 435 participants were still alive and without contraindications in 1999–2000 and 279 of them underwent a second MRI scan. The most important reasons for dropout at the second MRI were death (11%), claustrophobia developed at the baseline MRI (9%), too much trouble (9%) or no interest (7%) as mentioned by participants, MRI contraindications (2%), institutionalization (2%) and other reasons (3%).

Retinal vessel measurements

The ophthalmic examination at baseline included taking fundus transparencies centred on the optic disc of both eyes using a telecentric fundus camera (pharmacological mydriasis, 20° field, Topcon Optical Company, Tokyo, Japan) (Ikram et al., 2004). These transparencies were digitized with a high-resolution scanner (Nikon LS-4000, Nikon Corporation, Japan) and for each participant one eye with the best image quality was analysed with a semi-automated system (Retinal Analysis, Optimate, WI; Department of Ophthalmology and Visual Science, University of Wisconsin-Madison). Per eye one summary value was calculated for the diameters of the blood column in the retinal arterioles and one for the venules (in μm) (Knudtson et al., 2003). Because eyes may have a different magnification in case of refractive changes due to corneal curvature, lens and axial length differences, we additionally adjusted this summary vessel measure for possible magnification variations with Littmann's formula to approximate absolute measures (Littmann, 1988). The AVR was defined as the ratio of arteriolar-to-venular diameters.

In a random subsample of 100 participants we found no statistically significant differences between right and left eyes for the arteriolar and venular diameters. Four trained graders performed all measurements masked for participant characteristics. Both inter- and intragrader studies (n = 40) showed good to excellent agreement (intraclass correlation coefficient = 0.49–0.95) (Ikram et al., 2004).

MRI scanning

In 1995–1996, axial T1-, T2- and proton density (PD)-weighted cerebral MR scans were made on a 1.5-tesla MRI scanner (MR Vision, Siemens) (de Leeuw et al., 2001). In 1999–2000, participants underwent a second MRI on the same MR Vision scanner with the same sequences.

White matter lesions

White matter lesions were considered present when visible as hyper-intense on PD- and T2-weighted images, without prominent hypo-intensity on T1-weighted scans. We considered white matter lesions to be in the periventricular region if they were directly adjacent to the ventricle; otherwise, we called them subcortical. We scored periventricular white matter lesions semi-quantitatively in three regions (lesions adjacent to the frontal horns, the lateral walls and the occipital horns of the lateral ventricle) resulting in a total score ranging from 0 to 9. For subcortical white matter lesions we approximated a total volume based on number and size of lesions (volume range: 0–29.5 ml). Both inter- and intrarater studies (n = 100) showed a good to excellent agreement (κ = 0.79–0.90, intraclass correlation coefficient = 0.88–0.95) (de Leeuw et al., 2001).

For measuring change of white matter lesion severity over time, we used a specifically developed and validated white matter lesions change scale (Prins et al., 2004b). Two raters independently assessed the progression of white matter lesion severity on digital T2- and PD-weighted images by direct scan comparison. Raters were masked to all clinical information. To systematically evaluate differences in white matter lesion severity in all different brain regions they separately scored differences in the three periventricular regions of both hemispheres (periventricular score range −6 to +6) and in the subcortical white matter of the four lobes of both hemispheres (subcortical score range: −8 to +8). The rating showed good interobserver (intraclass correlation coefficient 0.72–0.73) and intraobserver agreement (intraclass correlation coefficient 0.70–0.93). If raters disagreed on one point or less on the scale, the mean of the ratings was used; if more, a consensus meeting was held. Progression was defined as an increase of one point or more between baseline and follow-up. Because this scale of white matter lesion progression was qualitative rather than quantitative, we categorized progression into no (score <1), minor (score 1–2.5) or marked progression (score 3 or higher).

Lacunar infarcts

We defined brain infarcts as areas of focal hyperintensity on T2-weighted images sized ≥3 mm. Areas of hyperintensity in the white matter also had to have corresponding prominent hypo-intensity on T1-weighted images in order to distinguish them from white matter lesions. Lacunar infarcts were defined as infarcts 3–20 mm in size and located in the subcortical white matter or basal ganglia. Non-lacunar infarcts were excluded from the analyses of lacunar infarcts. A new infarct on the follow-up MRI was classified as incident lacunar infarct (Vermeer et al., 2003a).

Cardiovascular risk factors

Baseline blood pressure was measured in sitting position at the right brachial artery with a random-zero sphygmomanometer. In the analyses we used the average of two measurements taken at one occasion. Body mass index (BMI) was computed as weight divided by height squared. Non-fasting serum total and HDL cholesterol levels were determined by an automated enzymatic procedure (van Gent et al., 1977). Diabetes mellitus was considered present if participants reported use of antidiabetic medication or when random or post-load serum glucose level was >11.1 mmol/l. The presence of atherosclerotic plaques was assessed by ultrasound at the bifurcation, common and internal carotid artery on both sides, resulting in a carotid artery plaque score ranging from 0 to 6 (Bots et al., 1996). Serum levels of high-sensitivity C-reactive protein (CRP) were determined by the Rate Near Infrared Particle Immunoassay method (Immage® high-sensitive CRP, Beckman Coulter, USA). Information on smoking (categorized as current, former or never) was obtained during the home interview.

Fig. 1

(A) The diameters of arterioles and venules that cross the area bounded by the dotted and outermost circles were measured. (B) Periventricular white matter lesions. (C) Subcortical white matter lesions. (D) Lacunar infarcts.

Statistical analyses

We used age and gender adjusted analysis of covariance to analyse whether baseline risk factors differed between people with and without a second MRI assessment. For the cross-sectional analyses, we used linear regression models to assess the relationship between baseline retinal vessel diameters and severity of periventricular and subcortical white matter lesions on the first MRI scan. Logistic regression models were used to study the association between retinal vessel diameters and prevalent lacunar infarcts. For the longitudinal analyses, we used multinomial logistic regression models to examine the relationship between retinal vessel diameters and white matter lesion progression, and logistic regression models were used to study the association between retinal vessel diameters and incident lacunar infarcts. All analyses were performed adjusting for age and gender, and additionally for other cardiovascular risk factors with SPSS Windows version 11.0 (SPSS Inc., Chicago, IL, USA).

Results

Baseline characteristics of participants with and without repeated MRI assessments are presented in Table 1. Non-participants at the follow-up MRI examination were on average older than those who did have a repeated MRI assessment. With respect to all other cardiovascular risk factors, there were no significant differences between the two groups after adjusting for age and gender. The mean arteriolar diameter was 147.8 μm [range: 92.2–235.7; standard deviation (SD): 14.2], venular diameter 223.8 μm (range: 135.1–313.6; SD: 20.6) and AVR 0.66 (range: 0.48–1.02; SD: 0.06).

View this table:
Table 1

Baseline characteristics (1990–1993)

All participantsParticipants with repeated MRI assessmentNon-participants at follow-upAdjusted differencesa (95% CI)b
Number (n)490279211
Age (years)68.4 (7.8)67.0 (7.6)70.3 (7.8)3.3 (1.9; 4.7)c
Gender (% female)4949490.8 (−8.4; 9.9)
Diabetes mellitus (%)6.34.88.32.3 (−2.5; 7.0)
Smoking (% current)2223200.0 (−7.5; 7.6)
Systolic blood pressure (mmHg)136.7 (19.9)134.2 (19.3)139.9 (20.2)3.5 (−0.03; 6.9)
Diastolic blood pressure (mmHg)73.1 (10.8)72.6 (10.6)73.7 (11.2)1.5 (−0.4; 3.5)
BMI (kg/m2)26.3 (3.3)26.1 (3.2)26.5 (3.3)0.3 (−0.3; 0.9)
Carotid artery plaque score ≥4 (%)1715212.0 (−5.0; 8.9)
C-reactive protein (mg/l)1.72 (0.77; 3.34)1.64 (0.78; 2.93)1.75 (0.76; 3.62)0.57 (−0.30; 1.44)
Serum total cholesterol (mmol/l)6.65 (1.3)6.71 (1.3)6.58 (1.2)−0.10 (−0.32; 0.12)
Serum HDLd cholesterol (mmol/l)1.33 (0.4)1.33 (0.3)1.33 (0.5)0.00 (−0.07; 0.07)
Periventricular WMLe (range: 0–9)2.68 (2.19)2.46 (2.11)2.97 (2.25)0.08 (−0.28; 0.43)
Subcortical WMLe (ml)1.81 (3.40)1.59 (2.97)2.10 (3.88)−0.01 (−0.60; 0.57)
All infarcts (%)2623301.1 (−6.7; 8.8)
Lacunar infarctsf (%)2220250.0 (−7.0; 7.9)
Retinal arteriolar diameter (μm)147.8 (14.2)147.8 (13.9)147.9 (14.6)0.3 (−2.4; 2.9)
Retinal venular diameter (μm)223.8 (20.6)223.3 (20.1)224.5 (21.2)3.0 (−0.7; 6.9)
Retinal arteriolar-to-venular ratio0.66 (0.06)0.66 (0.06)0.66 (0.06)−0.008 (−0.018; 0.003)
  • Presented as unadjusted means (SD) or percentages, and for CRP as median (inter-quartile range).

  • a Age and gender adjusted if applicable.

  • b CI = confidence interval.

  • c Significant (P < 0.05) compared with those with repeated MRI assessment.

  • d HDL = high-density lipoprotein.

  • e WML = white matter lesions.

  • f Non-lacunar infarcts excluded.

The mean follow-up period between the first and second MRI was 3.3 years (SD: 0.2 years). During this period, 82 participants showed white matter progression in the periventricular region, of which 31 had marked progression. A total of 89 participants showed progression in the subcortical region, of which 25 had marked progression. There were 33 participants who had a new infarct on the follow-up MRI, of which 27 were lacunar (including 25 silent) and 6 non-lacunar (including five silent) infarcts.

Retinal vessel diameters were associated neither with the severity of white matter lesions nor with prevalent lacunar infarcts on the first MRI scan (Table 2). Larger retinal venular diameters were associated with marked progression of both periventricular and subcortical white matter lesions and with incident lacunar infarcts (Table 3). Smaller arteriolar diameters were neither related to white matter lesion progression nor to incident lacunar infarcts.

View this table:
Table 2

Cross-sectional association between retinal vessel diameters and severity of cerebral small vessel disease (n = 490)

Periventricular WMLa (grade)Subcortical WMLa (ml)Lacunar infarct [OR (95% CI)]b
Arteriolar narrowingc0.002 (−0.17; 0.17)0.01 (−0.21; 0.36)1.14 (0.91–1.44)
Venular dilatationd−0.004 (−0.18; 0.17)0.08 (−0.21; 0.37)1.07 (0.85–1.35)
Arteriolar-to-venular ratioe0.04 (−0.14; 0.22)0.18 (−0.11; 0.47)1.24 (0.98–1.58)
  • All models adjusted for age and gender.

  • a WML = white matter lesions.

  • b OR = odds ratio; CI = confidence interval.

  • c Per SD decrease in arteriolar diameter.

  • d Per SD increase in venular diameter.

  • e Per SD decrease in arteriolar-to-venular ratio.

View this table:
Table 3

Odds ratiosa of white matter lesion progression and incident lacunar infarct per SD difference in baseline retinal vessel measurements (n = 279)

Periventricular WMLb progressionSubcortical WMLb progressionIncident lacunar infarct
Any (n = 82)Minor (n = 51)Marked (n = 31)Any (n = 89)Minor (n = 64)Marked (n = 25)All (n = 27)
Arteriolar narrowingc0.93 (0.70–1.22)1.10 (0.78–1.56)0.77 (0.52–1.14)0.85 (0.65–1.11)0.92 (0.67–1.25)0.76 (0.49–1.16)0.81 (0.54–1.22)
Venular dilatationd1.33 (0.99–1.78)1.12 (0.79–1.60)1.71 (1.11–2.61)1.22 (0.93–1.61)1.06 (0.78–1.45)1.72 (1.09–2.71)1.59 (1.06–2.39)
Arteriolar-to- venular ratioe1.22 (0.92–1.61)1.20 (0.85–1.67)1.25 (0.84–1.92)1.03 (0.79–1.33)0.95 (0.71–1.28)1.28 (0.83–2.00)1.29 (0.84–1.98)
  • a Age and gender adjusted odds ratios with corresponding 95% CIs.

  • b WML = white matter lesions.

  • c OR per SD decrease in retinal arteriolar diameters.

  • d OR per SD increase in retinal venular diameters.

  • e OR per SD decrease in arteriolar-to-venular ratio.

Finally, Table 4 presents the relation between retinal venular diameters and cerebral small vessel disease after additional adjustment for other cardiovascular risk factors. Each SD increase in venular diameters resulted in a 1.8 times increased risk of marked periventricular white matter lesion progression and in a 2.8 times increased risk of marked subcortical white matter progression. Persons with larger venular diameters had more incident lacunar infarcts, although this association attenuated after additional adjustments were made.

View this table:
Table 4

Odds ratiosa of white matter lesion progressions and incident lacunar infarcts per SD increase in retinal venular diameters additionally adjusted for other cardiovascular risk factors (n = 279)

Periventricular WMLb progressionSubcortical WMLb progressionIncident lacunar Infarcts
MinorMarkedMinorMarked
Age and gender1.12 (0.79–1.60)1.71 (1.11–2.61)1.06 (0.78–1.45)1.72 (1.09–2.71)1.59 (1.06–2.39)
Age, gender and cholesterolc1.10 (0.77–1.57)1.80 (1.17–2.78)1.05 (0.76–1.44)1.80 (1.12–2.89)1.59 (1.06–2.39)
Age, gender and body mass index1.14 (0.80–1.63)1.63 (1.07–2.52)1.05 (0.77–1.44)1.68 (1.06–2.66)1.58 (1.05–2.39)
Age, gender and carotid artery plaques1.16 (0.80–1.69)1.66 (1.08–2.55)1.11 (0.80–1.54)1.95 (1.21–3.17)1.43 (0.92–2.22)
Age, gender and C-reactive protein1.07 (0.75–1.53)1.60 (1.03–2.47)1.03 (0.75–1.41)1.58 (1.00–2.52)1.52 (1.00–2.30)
Age, gender and smoking1.08 (0.74–1.57)1.64 (1.05–2.57)1.06 (0.77–1.46)1.69 (1.07–2.68)1.67 (1.10–2.54)
Age, gender and diabetes mellitus1.25 (0.85–1.84)1.79 (1.13–2.83)1.06 (0.75–1.51)1.77 (1.10–2.85)1.40 (0.90–2.17)
Age, gender and blood pressured1.12 (0.78–1.61)1.73 (1.12–2.68)1.07 (0.78–1.47)1.83 (1.13–2.97)1.62 (1.07–2.45)
Fully adjusted1.25 (0.79–1.96)1.74 (1.02–2.95)1.09 (0.74–1.60)2.50 (1.30–4.81)1.24 (0.72–2.12)
  • a Odds ratios with corresponding 95% confidence intervals.

  • b WML = white matter lesions.

  • c Total and high-density lipoprotein cholesterol.

  • d Systolic and diastolic blood pressure.

Discussion

Especially larger retinal venular diameters were associated in this study with marked progression of periventricular and subcortical white matter lesions independent of other cardiovascular risk factors. Also, persons with larger venular diameters tended to have more incident lacunar infarcts.

Some methodological issues warrant consideration. Non-participants at follow-up were older than those who underwent the second MRI. However, there were no significant differences in other cardiovascular risk factors between the two groups, suggesting that selection bias played a limited role in the longitudinal analyses. Nevertheless, it is likely that those who remained in the study and came for re-examination were somewhat healthier at baseline. Selective dropout, if any, would therefore most likely have led to an underestimation of the strength of the relationship. Another limitation is that the first MRI was performed on average 3 years after baseline fundus photography. Because in a subsample of 300 persons we found hardly any changes in retinal vessel diameters [mean change of −0.8 (SD: 14) for arteriolar diameter and −0.3 (SD: 19) for venular diameter] over a period of 6 years, we concluded that baseline measurements are a good reflection of retinal vessel diameters at the time of the first MRI. Moreover, baseline (vascular) risk factors were not associated with the change in vessel diameters over the follow-up period. Finally, due to the low number of incident cases we were not able to examine the association between retinal vessel diameters and non-lacunar infarcts.

We observed that venular dilatation was only related to marked white matter lesion progression, but not to the severity of white matter lesions on the first MRI scan. A possible explanation is that persons with more severe white matter lesions at baseline show a greater rate of progression of white matter lesions during follow-up (Schmidt et al., 1999; Schmidt et al., 2003). Hence, in the longitudinal analyses, the inclusion of these persons with marked progression led to a broader range of white matter lesions and allowed us to detect the association with venular dilatation. In contrast, due to the limited range of white matter lesions, the cross-sectional analyses were probably not sensitive enough to uncover this association. This is further supported by the fact that minor lesion progression was also not associated with venular dilatation. Another explanation might be that the severity of white matter lesions at baseline can be viewed as a compound indicator of past risk factor exposure and genetic susceptibility. As such, it is not surprising that this is an important risk factor for lesion progression. Our measure of venular dilatation reflects a vessel condition that is present at baseline. Our findings suggest that, inherently conditional on the amount of white matter lesions already present, this measure is more predictive of the risk of lesion progression than that it would reflect all genetic and non-genetic factors that led to the occurrence of white matter lesions. When we restricted the cross-sectional analyses to the participants who also had follow-up scans the results remained virtually identical, suggesting that selective dropout did not underlie this discrepancy.

Previously, we showed that larger venular diameters were related to atherosclerosis (as measured by more plaques in the carotid arteries), higher levels of total cholesterol, lower levels of HDL, higher levels of inflammatory markers (such as erythrocyte sedimentation rate and leucocyte count) and smoking (Ikram et al., 2004). Because these risk factors are also related to the development of white matter lesions and lacunar infarcts, they may explain the observed associations (Bots et al., 1993; Breteler et al., 1994). After adjusting one-by-one for these factors, the associations between venular dilatation and white matter lesions still remained significant. The associations attenuated most after including CRP, pointing towards the involvement of inflammation in the pathogenesis of cerebral small vessel disease. When we adjusted for all these risk factors simultaneously the association between venular dilatation and cerebral small vessel disease hardly changed suggesting that other mechanisms are involved.

It has been hypothesized that retinal venular dilatation occurs in response to retinal hypoxia (Klein et al., 2004). Venular dilatation has been described as one of the earliest changes not only in diabetic retinopathy, but also in the less well-known venous stasis retinopathy, both of which are characterized by retinal hypoxia (Zatz and Brenner, 1986; Klijn et al., 2002). Furthermore, treatment with photocoagulation decreases the oxygen requirement of retinal tissue eventually leading to disappearance of venular dilatation (Neupert et al., 1976; Carter, 1985). Based on these observations, a possible explanation might be that venular dilatation is a general marker of diffuse retinal ischaemia and by proxy cerebral ischaemia leading to increased susceptibility of brain tissue to the development of white matter lesions and lacunar infarcts.

Another explanation might be that retinal venular dilatation in some way reflects cerebral venular abnormalities. In one study, brains with severe white matter lesions showed wall thickening and obstruction of the periventricular veins, a condition known as periventricular venous collagenosis (Moody et al., 1995, 1997). It was suggested that these alterations resulted in an increased venous pressure, venular dilatation and venular blood–brain barrier disruption. This venous insufficiency could result in ischaemic stress from impaired clearance of cellular metabolites in the deep white matter eventually leading to white matter lesions (Moody et al., 1995; Sahlas et al., 2002). It is unkown to what extent such cerebral venular changes might be reflected in the retinal venular diameters.

With regard to arterioles, we previously suggested that arteriolar narrowing reflects vasoconstriction, intimal thickening and medial hyperplasia (Ikram et al., 2004). The lack of an association between smaller arteriolar diameters and cerebral small vessel disease could be due to the fact that, despite increased arterial stiffness, arterioles still possess the ability to dilate even in retinal or cerebral hypoxia, though less so than the venules (Klijn et al., 2002; Ikram et al., 2004; Klein et al., 2004). These opposing effects in the same arterioles, narrowing and widening, might result in no apparent association.

In conclusion, we found retinal venular dilatation to be associated with progression of cerebral small vessel disease. Future research will focus more on the mechanisms underlying venular abnormalities including dilatation, this might provide new clues into the pathophysiology of cerebral small vessel disease.

Acknowledgments

This study was supported by the Netherlands Organization for Scientific Research (NWO) grant 904-61-155, The Hague. Foundations: Alzheimer Nederland (grant V-2001-015); Optimix, Amsterdam; Physico Therapeutic Institute, Rotterdam; Blindenpenning, Amsterdam; Sint Laurens Institute, Rotterdam; Bevordering van Volkskracht, Rotterdam; Blindenhulp, The Hague; Rotterdamse Blindenbelangen Association, Rotterdam; OOG, The Hague; kfHein, Utrecht; Ooglijders, Rotterdam; Prins Bernhard Cultuurfonds, Amsterdam; Van Leeuwen Van Lignac, Rotterdam; Verhagen, Rotterdam; Netherlands Society for the Prevention of Blindness, Doorn; LSBS, Utrecht; and Elise Mathilde, Maarn. Unrestricted grants were obtained from Topcon Europe BV, Capelle aan de IJssel; Lameris Ootech, Nieuwegein; Carl Zeiss BV Nederland, Sliedrecht (all in The Netherlands); and Heidelberg Engineering, Dossenheim, Germany.

References

View Abstract