Retinal Vessel Calibers in Predicting Long-Term Cardiovascular Outcomes: The Atherosclerosis Risk in Communities Study

Abstract

Background

Narrower retinal arterioles and wider retinal venules have been associated with negative cardiovascular outcomes. We investigated whether retinal vessel calibers are associated with cardiovascular outcomes in long-term follow-up and provide incremental value over the 2013 American College of Cardiology/American Heart Association pooled cohort equations (PCE) in predicting Atherosclerotic Cardiovascular Disease Events (ASCVE).

Methods

10,470 men and women without prior ASCVE or heart failure (HF) in the Atherosclerosis Risk in Communities (ARIC) study underwent retinal photography at visit 3 (1993–1995).

Results

During a mean follow up of 16 years, there were 1779 incident CHD events, 548 ischemic strokes, 1395 HF events and 2793 deaths. Rates of all outcomes were higher in those with wider retinal venules and narrower retinal arterioles. Subjects with wider retinal venules [hazard ratio (HR) 1.13 (95% CI: 1.08–1.18), HR 1.18 (1.07–1.31) and HR 1.10 (1.00–1.20) per standard deviation (SD) increase] and narrower retinal arterioles [HR 1.06 (1.01–1.11), HR 1.14 (1.03–1.26) and HR 1.13 (1.03–1.24) per SD decrease] had a higher risk of death and stroke in both sexes as well as incident CHD in women but not men (interaction p=0.02) after adjustment for the PCE risk-score variables. The association between retinal vessel caliber and HF was non-significant after adjustment for systolic blood pressure. Among women with PCE-predicted 10-year ASCVE risk <5% (overall risk 3.9%), women in the narrowest arteriolar quartile had a 10-year event-rate of 5.6% compared to 2.8% for the widest quartile (5.0% vs. 3.4% for wider vs. narrower venules). Retinal vessel caliber reclassified 21% of low-risk women (11% of all women) as intermediate-risk (>5%).

Conclusions

Narrower retinal arterioles and wider retinal venules conferred long-term risk of mortality and ischemic stroke in both genders and CHD in women. These measures serve as an inexpensive, reproducible biomarker that added incremental value to current practice guidelines in ASCVE risk prediction in low-risk women.

Coronary and cerebral microvascular disease is associated with significant cardiovascular morbidity with respect to incident coronary heart disease (CHD), ischemic stroke, congestive heart failure (HF) and mortality from these diseases¹. The retinal vasculature provides a readily accessible window to assess the microvasculature. With advances in technology, imaging of the fundus is relatively inexpensive, accurate, reproducible, and radiation-free. The retinal arterioles [measured as the central retinal arteriolar equivalent (CRAE)], retinal venules [measured as the central retinal venular equivalent (CRVE)] and retinal microvascular abnormalities have been associated with atherosclerotic cardiovascular disease events (ASCVE, defined as incident CHD and ischemic stroke) ^2–4: incident CHD^5–8 particularly in women, incident stroke^{6, 9–13} and cardiovascular mortality^{7, 14–17}. While fundus photography has been recognized as a promising imaging modality to evaluate the pathogenesis of cardiovascular diseases, its ability to be applied as a test to predict ASCVE in the clinical setting has yet to be evaluated using current practice guidelines.

In the 2013 Guideline on the Assessment of Cardiovascular Risk, the American College of Cardiology/American Heart Association (ACC/AHA) released Pooled Cohort Equations (PCE), multivariable risk equations developed to predict 10-year risk of ASCVE, derived using data from multiple cohort studies including ARIC, CARDIA (Coronary Artery Risk Development in Young Adults) and CHS (Cardiovascular Health Study), in part to include the outcome of ischemic stroke into treatment guidelines as well as to address the limitations of the Framingham Risk score (FRS) and its application to the Adult Treatment Panel Guidelines (ATP III)¹⁸. The ACC/AHA simultaneously published the Guideline on the Treatment of Blood Cholesterol creating new standards to guide statin therapy. While previous ATP III guidelines generally recommended statin therapy for patients whose estimated 10-year FRS was >20% for a CHD event, the new guidelines recalibrate the threshold for treatment, recommending statin therapy for adults whose 10-year ASCVE risk is above 7.5%.

Our study aimed to elucidate the association of retinal vessel calibers with long-term cardiovascular outcomes and death in an average of 16-year follow-up in a closely followed, well adjudicated large population cohort of asymptomatic adults in the Atherosclerosis Risk in Communities (ARIC) study and to assess the incremental value of retinal vessel calibers to the 2013 AHA/ACC PCE risk score.

Methods

Study Design and Study Population

The ARIC Study is an ongoing, prospective observational study of atherosclerosis risk factors in four U.S. communities (Forsyth County, NC, Jackson, MS, suburbs of Minneapolis, MN, and Washington County, MD) originally comprised of 15,792 men and women, aged between 45 to 64 years recruited between 1987 and 1989 (Visit 1)¹⁹. The participants were examined, initially, every three years with the second examination in 1990–92, the third in 1993–95, the fourth in 1996–98, and the fifth in 2011–13. Institutional Review Board at each participating site approved the study protocol and informed consent was obtained in writing at each examination.

Identification of Cardiovascular Outcomes and Death

Outcomes evaluated in this study are incident ischemic stroke, CHD, HF, and death subsequent to the third visit until the end of 2012 (average of 16-year follow up). Methodology for quality control, detection and adjudication of CHD events has been presented previously²⁰. Trained abstractors retrieved information on out of hospital deaths via death certificates, hospitalized patients, physician questionnaires and next-of-kin interviews. We defined ASCVE as new ischemic stroke (fatal or non-fatal) or incident CHD (as previously described²¹: acute (definite or probable) myocardial infarction (MI), fatal coronary heart disease, silent MI and myocardial revascularization including coronary angioplasty or coronary artery bypass graft surgery from third visit (1993–95) until the end of 2012 among subjects who were free of these outcomes and HF at the beginning of visit three when retinal photography was performed. Incident HF was defined by HF hospitalization or HF death, according to the International Classification of Diseases-Ninth Revision (ICD-9), code 410 in any position, obtained by ARIC Study retrospective surveillance of hospital discharges^{22, 23}.

Measurement of Retinal Vessel Calibers

Retinal fundus photographs were taken at the third examination (1993–1995) in which 12,887 participants attended and retinal vessel calibers were measured as: Central retinal arteriolar equivalent (CRAE), Central retinal venular equivalent (CRVE), arteriolar-to-venular diameter ratio (A/V-ratio) = CRAE/CRVE. A/V-ratio eliminates the denominator for arteriolar and venular measurements and, thereby, magnification differences between photographs. However, as these two measures may represent distinct processes, the A/V-ratio analysis is presented in the Supplementary Tables. After a period of 5 minutes of adaptation to the dark, one random eye was selected and a 45° retinal photograph was taken. Photographs were digitized, and trained and masked graders measured the caliber of individual arterioles and venules running through an area one half to one disk diameter from the optic disk margin. CRAE and CRVE were calculated, representing the average of estimated calibers for the central retinal vessels. The reproducibility for CRAE and CRVE based on repeat readings of the same fundus photograph, was excellent, as previously reported ²⁴.

Components of the AHA/ACC pooled cohort equation risk score and multivariate models

Age, gender, race (self reported), total cholesterol (TChol, measured by blood plasma assay), high density lipoprotein (HDL, measured by blood plasma assay), systolic and diastolic blood pressure (SBP, DBP measured 3 times using a random-zero sphygmomanometer and the average of the 2nd and 3rd measurements used for analysis), hypertension status (defined as antihypertensive medication use within the past two weeks of examination, either self-reported or taken from prescription bottles), diabetes mellitus status (defined on the basis of use of anti-diabetic medications, self-report of a physician diagnosis, fasting glucose value ≥126 mg/dl or a non-fasting glucose of ≥200), and current smoking status (self reported) were analyzed. Unadjusted and multivariate models for CRAE and CRVE also included correction for retinal photography magnification.

Statistical Analysis

Of 12,887 participants who attended visit 3, 1,854 were excluded for non-gradable or missing retinal photographs, another 500 were excluded for previous stroke, MI or CHF, and approximately 63 had insufficient data to calculate AHA/ACC pooled cohort equation risk score (i.e., race other than black or white), leaving 10,470 for analysis. As previously reported, those individuals with ungradable or missing retinal photographs were more likely to be older, black, hypertensive, diabetic, and of higher CV risk²⁵. A/V-ratio and magnification-adjusted CRAE and CRVE were analyzed as a continuous and as categorical variables (population quartiles, with the first quartile representing the most narrow arterioles and/or venules and the fourth quartile representing the least narrow arterioles and/or widest venules). Magnification-adjusted CRAE was calculated as the residual value after fitting a linear regression model with CRAE as the outcome and CRVE as the predictor. Magnification-adjusted CRVE was calculated similarly. Baseline characteristics were summarized using descriptive statistics for continuous variables and number counts and percentages for categorical variables. Linear regression and chi-squared tests for trend were used to test for a relationship between continuous or categorical baseline characteristics with quartiles of CRAE and CRVE. Incident risk was calculated for each outcome. Follow-up time was calculated as the time from the third exam to the time of the first event. Hazard ratios were calculated using Cox Proportional Hazards Regression. Two multivariate models investigating the association between A/V-ratio quartiles and outcomes were created, one adjusting for age, race and gender; the second adjusting for the components of the PCE—age, race, gender, smoking, diabetes, SBP, hypertension treatment, HDL, and TChol. Ten- and fifteen-year risks of events were calculated for each outcome, based on Kaplan Meier failure curves.

To assess the incremental value of retinal vessel caliber to the 2013 AHA/ACC PCE risk score, two models of ASCVE risk were created: 1) 10-year risk of ASCVE for each subject at visit three per 2013 ACC/AHA PCE ²⁶; 2) Model 1 plus CRAE, CRVE or A/V-ratio quartiles. Individuals were classified into “low”, “intermediate”, and “high” risk using the risk thresholds of <5%, 5–7.5%, >7.5% for each of these models. To assess reclassification, we computed the continuous net reclassification improvement (NRI)²⁷. The population was further stratified by PCE risk group and CRAE or CRVE or A/V-ratio quartiles in order to assess the performance of these measures in subgroups of women based on risk score. In order to define improvement in model discrimination with the addition of CRAE, CRVE or A/V-ratio quartiles in women, the Harrell C statistic was calculated for ASCVE risk for the base model and the base model plus CRAE, CRVE or A/V-ratio within risk subgroups.

All data were analyzed using STATA version 14.0.

Results

Those subjects with a wider retinal venules or narrower retinal arterioles were more likely to be older, male, black, to have diabetes, to be on anti-HTN medications, and to have higher SBP, total cholesterol, and lower HDL (Table 1). The mean length of follow up was 16 years during which there were 1779 incident CHD events, 548 ischemic strokes, 1395 HF events and 2793 deaths. Rates of all outcomes considered were higher in participants with wider retinal venules (CRVE) and narrower retinal arteries (CRAE) (Table 2). There was no evidence of non-linearity using restricted cubic splines to evaluate outcomes. CRAE and CRVE were significantly associated with death (p=0.02 and p<0.0001, respectively) and ischemic stroke (p<0.0001 for both) after adjustment for the PCE variables (Figure 1). The relationship between CRAE and CRVE and HF outcomes was non-significant after multivariate adjustment, primarily by the addition of SBP to the model (Table 2, Figure 1). We observed a significant interaction of gender with incident CHD and ASCVE (p=0.02 for both) but not for death, HF or stroke (p>0.2 for all three). Accordingly, separate gender analysis was performed for CHD and ASCVE. There was also a nominally significant interaction between CRAE/CRVE and diabetic status. However, this interaction was not apparent when restricting to the female subgroup. After adjustment for the PCE variables, CRAE and CRVE were significantly associated with CHD and ASCVE in women but not men (Table 2, Suppl Table 1). This differential relationship persisted after adjustment for PCE variables (Table 2, Figure 1), remaining significant in women but not in men [CRAE p=0.005 and p=0.933; CRVE p<0.0001 and p=0.354, respectively]. Additionally, BMI was not significant when added to the multivariate model and CRAE and CRVE remained independently associated with CHD and ASCVE. The association of the composite measure of A/V-ratio (CRAE/CRVE) with cardiovascular outcomes was equivalent to those described for the individual measures (Suppl Table 2 and 3).

Table 1.

Baseline Characteristics of Subjects in the Atherosclerosis Risk in Communities Study cohort stratified by retinal vessel caliber quartiles (N=10,470).

CRVE quartiles Venular widening →						CRAE quartiles Arteriolar narrowing →
Characteristic	1(N=2618)	2 (N=2617)	3 (N=2618)	4 (N=2617)	P-value trend	4 (N=2617)	3 (N=2618)	2 (N=2617)	1 (N=2618)	P-value trend
Age, mean (SD), y	60±5.6	60±5.7	59±5.5	59±5.5	P<0.001	59±5.6	60±5.5	60±5.6	60±5.6	P=0.001
Men, %	920 (35%)	1075 (41%)	1218 (47%)	1305 (50%)	P<0.001	915 (35%)	1071 (41%)	1183 (45%)	1349 (52%)	P<0.001
Black, %	337 (13%)	437 (17%)	622 (24%)	836 (32%)	P<0.001	485 (19%)	556 (21%)	605 (23%)	586 (22%)	P<0.001
Smoking, current, %	258 (10%)	385 (15%)	492 (19%)	712 (27%)	P<0.001	539 (21%)	440 (17%)	438 (17%)	430 (16%)	P<0.001
Diabetes, %	285 (16%)	340 (14%)	369 (14%)	451 (14%)	P<0.001	344 (13%)	353 (14%)	359 (14%)	389 (15%)	P=0.07
Treated HTN, %	642 (25%)	696 (27%)	794 (30%)	904 (35%)	P<0.001	556 (21%)	671 (26%)	849 (32%)	960 (37%)	P<0.001
SBP, mean (SD), mmHg	121±19	123±19	125±18	127±19	P<0.001	116±17	122±17	126±18	131±18	P<0.001
TChol, mean (SD), mg/dl	207±36	207±37	208±37	210±39	P=0.002	207±36	208±38	208±37	209±38	P=0.1
HDL, mean (SD), mg/dl	55±19	53±18	52±18	51±18	P<0.001	54±18	53±18	53±18	52±19	P<0.001
BMI, mean (SD), kg/m2	29±6	29±6	28±6	28±6	P<0.001	28±5	28±5	29±5	29±6	P<0.001

Table 2.

Hazards Ratios of Ischemic Stroke, all-cause mortality (Death), Heart Failure (HF), incident Coronary Heart Disease (CHD), and Atherosclerotic Cardiovascular Disease Events (ASCVE), per 1 Standard Deviation (SD) decrease in Central Retinal Arteriolar Equivalent (CRAE) and 1 SD increase in Central Retinal Venular Equivalent (CRVE). (CHD and ASCVE tested separately in Males and Females due to a significant gender interaction (P=0.02 for both).

Combined cohort (N=10,470)
Event	Adjustment	No. Events	HR (95% CI) per SD Decrease in CRAE	P-value	HR (95% CI) per SD Increase in CRVE	P-value
Ischemic Stroke	Unadjusted	548	1.36 (1.23–1.50)	P<0.001	1.34 (1.21–1.47)	P<0.001
	Multivariate		1.14 (1.03–1.26)	P=0.013	1.18 (1.07–1.31)	P=0.001
Death	Unadjusted	2793	1.19 (1.14–1.25)	P<0.001	1.24 (1.18–1.29)	P<0.001
	Multivariate		1.06 (1.01–1.11)	P=0.020	1.13 (1.08–1.18)	P<0.001
HF	Unadjusted	1395	1.17 (1.10–1.24)	P<0.001	1.19 (1.11–1.26)	P<0.001
	Multivariate		1.00 (0.94–1.07)	P=0.908	1.05 (0.99–1.12)	P=0.10
Women (N=5,952)
CHD	Unadjusted	655	1.23 (1.13–1.35)	P<0.001	1.26 (1.15–1.38)	P<0.001
	Multivariate		1.13 (1.03–1.24)	P=0.012	1.10 (1.00–1.20)	P=0.04
ASCVE	Unadjusted	862	1.26 (1.17–1.37)	P<0.001	1.27 (1.17–1.38)	P<0.001
	Multivariate		1.14 (1.05–1.24)	P=0.002	1.12 (1.03–1.21)	P=0.005
Men (N=4,518)
CHD	Unadjusted	1,124	1.07 (1.00–1.15)	P=0.058	1.09 (1.02–1.17)	P=0.014
	Multivariate		0.98 (0.91–1.06)	P=0.63	1.02 (0.95–1.10)	P=0.60
ASCVE	Unadjusted	1294	1.11 (1.04–1.18)	P=0.002	1.12 (1.05–1.20)	P=0.001
	Multivariate		1.00 (0.93–1.07)	P=0.93	1.04 (0.97–1.11)	P=0.27

^*Multivariate model including age, race, SBP, cigarette use, TChol, HDL, DM, and HTN meds.

Figure 1.

Risk of Death and Atherosclerotic Cardiovascular Events per 1 Standard Deviation widening of the retinal vein or narrowing in the retinal artery in the Atherosclerosis Risk in Communities Study cohort (N=10,470) assessed by fundus photography following multivariate adjustment for the components of the Pooled Cohort Equations Risk Score (age, gender, race, systolic blood pressure, TChol, HDL, HTN, DM, cigarette use). CHD and ASCVE represents incident Coronary Heart Disease risk in women only (N=5952).

In order to test whether or not CRVE and CRAE may add incremental value to the PCE risk score in the female population, we began by assessing reclassification. Despite no improvement in the integrated discrimination improvement index, category-free net reclassification yielded a net reclassification of 6.0% [(−0.4%, 11.1%), p=0.06 for CRVE], 7.6% [(2.3%, 12.2%), p=0.040 for CRAE], and 8.1% [(2.8%, 12.8%), p=0.007 for A/V-ratio]. 10-year risk for each PCE risk group, using categories of <5%, 5%–7.5%, >7.5%, was within expected range for each group: 3.9% for the low-risk group (<5% PCE risk score), 7.4% for the intermediate risk group (5–7.5% PCE risk score), and 16.9% for the high risk group (>7.5% PCE risk score) (Table 3). To further test how CRVE and CRAE may be clinically useful, women were classified into AHA/ACC PCE risk categories according to their estimated 10-year risks and CRVE or CRAE quartiles were then used to further stratify the women within these risk categories (Table 3). Further stratification by both measures provided additional discrimination in low-risk women but not in intermediate or high-risk groups (p≤0.001 for CRVE and CRAE; Table 3). For example, when examining CRAE, the smallest quartile had significantly higher risk of ASCVE compared to the highest quartile within the subset of women originally classified as low-risk (HR of 1.95, p<0.001; Table 3, Figure 2a). While the observed overall 10-year risk in the low-risk group was 3.9%, the low-risk women with the narrowest arterioles experienced a 10-year event rate of 5.6%, compared to 2.8% for those with the widest arterioles (Table 3, Figure 2a, b; 5.6% vs. 2.8% for narrower vs. wider CRAE; 5.0% vs. 3.4% for wider vs. narrower CRVE). Qualitatively, the same result, wherein retinal measures provide further risk stratification only in otherwise low-risk women and not in medium and high-risk women, can be seen when quartiles of CRAE/CRVE specific to low-risk women are used (Table 4) rather than quartiles derived from all women. Stratification of low-risk women using A/V-ratio yielded similar results (Suppl Table 4). When inverse probability weighting was used to account for differential rates of missing retinal photographs driven by previously identified risk factors, no major study findings differed.

Table 3.

Estimated ASCVE rates, cumulative 10- and 15-year hazard ratios in women (N=5952) stratified by Pooled Cohort Equations (PCE) risk score and Central Retinal Arteriolar Equivalent (CRAE) and Central Retinal Venular Equivalent (CRVE) quartiles.

PCE risk group	10-year risk (95% CI)	15-year risk (95% CI)	C-stat*	Quartile	N	No.	10-year risk (95% CI)	15 year risk (95% CI)	Hazard Ratio (p-value)
ARTERIOLAR NARROWING (CRAE)
(<5%)				1	682	73	5.6 (4.1, 7.6)	8.6 (6.7, 11)	1.95 (p<0.001)
	3.9 (3.2, 4.6)	6.4 (5.6, 7.3)	0.57	2	766	64	4.2 (2.9, 5.9)	7.3 (5.6, 9.4)	1.49 (p=0.03)
				3	874	65	3.4 (2.4, 4.9)	6.2 (4.8, 8.1)	1.32 (p=0.13)
(N=3277)				4	955	55	2.8 (1.9, 4.1)	4.4 (3.2, 5.9)	Reference
(5–7.5%)				1	260	39	9.6 (6.5, 14)	12.7 (9.1, 18)
	7.4 (5.9, 9.3)	12.6 (11, 15)		2	281	41	6.6 (4.2, 10)	13 (6.7, 14)	P trend=0.8
				3	232	36	7.1 (4.4, 11)	12.8 (9.8, 18)
(N=956)				4	183	27	5.8 (3.1, 11)	11 (7.0, 17)
(>7.5%)				1	546	146	18 (15, 22)	26 (22, 30)
	16.9 (15, 19)	24.8 (23, 27)		2	441	116	14 (11, 18)	25 (21, 30)	P trend=0.7
				3	382	109	18 (14, 22)	25 (21, 30)
(N=1719)				4	350	91	17 (13, 22)	23 (19, 28)
VENULAR WIDENING (CRVE)
(<5%)				4	683	73	5.0 (3.6, 6.9)	9.0 (7.1, 12)	1.78 (p=0.001)
	3.9 (3.2, 4.6)	6.4 (5.6, 7.3)	0.56	3	803	57	3.8 (2.7, 5.4)	5.7 (4.3, 7.6)	1.16 (p=0.42)
				2	864	70	3.5 (2.4, 4.9)	6.7 (5.2, 8.7)	1.35 (p=0.09)
(N=3277)				1	927	57	3.4 (2.4, 4.8)	4.8 (3.6, 6.4)	Reference
(5–7.5%)				4	262	45	9.5 (6.5, 14)	14 (10, 19)
	7.4 (5.9, 9.3)	12.6 (11, 15)		3	261	27	5.5 (3.3, 9.2)	11 (7.4, 15)	P trend=0.7
				2	220	33	7.0 (4.3, 11)	11 (7.4, 16)
(N=956)				1	213	38	7.4 (4.5, 12)	15 (10, 20)
(>7.5%)				4	543	156	21 (17, 25)	28 (24, 32)
	16.9 (15, 19)	24.8 (23, 27)		3	424	97	12 (9.5, 16)	21 (17, 25)	P trend=0.3
				2	404	122	18 (15, 22)	27 (22, 31)
(N=1719)				1	348	87	15 (11, 19)	22 (18, 27)

^*Calculated Harrell’s C-statistic of PCE risk groups when CRAE or CRVE is applied.

Figure 2.

Kaplan Meier failure curves of Atherosclerotic Cardiovascular Event (ASCVE) Risk by Central Retinal Arteriolar Equivalent (CRAE) or Central Retinal Venular Equivalent (CRVE) quartiles in women with predicted 10-year ASCVE risk <5% according to the ACC/AHA pooled cohort equation (PCE) risk score (low-risk, N=3,277) in the Atherosclerosis Risk in Communities Study cohort. The first quartile represents the most narrow arterioles and venules and the fourth quartile represents the least narrow arterioles and widest venules. For CRAE quartiles the log rank p-value=0.0001; for CRVE quartiles the log-rank p=0.004.

Table 4.

Sensitivity Analysis based on quartiles defined separately for risk. Estimated ASCVE rates, cumulative 10- and 15-year hazard ratios in women (N=5952) stratified by Pooled Cohort Equations (PCE) risk score and Central Retinal Arteriolar Equivalent (CRAE) and Central Retinal Venular Equivalent (CRVE) quartiles defined separately for each PCE risk.

PCE risk group	10-year risk (95% CI)	15-year risk (95% CI)	C-stat*	Quartile	N	No.	10-year risk (95% CI)	15 year risk (95% CI)	Hazard Ratio (p-value)
ARTERIOLAR NARROWING (CRAE)
(<5%)				1	820	81	5.1 (3.8, 6.9)	7.7 (6.1, 9.8)	1.92 (p<0.001)
	3.9 (3.2, 4.6)	6.4 (5.6, 7.3)	0.56	2	819	70	4.3 (3.1, 5.9)	7.6 (5.9, 9.7)	1.65 (p=0.009)
				3	819	62	3.5 (2.4, 5.0)	6.1 (4.6, 8.0)	1.43 (p=0.1)
(N=3277)				4	819	44	2.5 1.6, 3.8)	4.2 (3.0, 5.9)	Reference
(5–7.5%)				1	239	37	9.7 (6.4, 14)	12.9 (9.1, 18)	1.09 (p=0.7)
	7.4 (5.9, 9.3)	12.6 (11, 15)		2	239	37	8.2 (5.3, 13)	14.2 (10, 20)	1.05 (p=0.8)
				3	239	33	6.1 (3.6, 10)	11.7 (8.1, 17)	0.93 (p=0.8)
(N=956)				4	239	36	5.7 (3.3, 11)	11.5 (7.9, 17)	Reference
(>7.5%)				1	430	120	20.0 (16, 24)	27.9 (24, 33)	1.13 (p=0.3)
	16.9 (15, 19)	24.8 (23, 27)		2	430	98	11.2 (8.5, 15)	21.5 (18, 26)	0.85 (p=0.2)
				3	430	130	19.4 (16, 24)	26.8 (23, 31)	1.20 (p=0.2)
(N=1719)				4	429	114	16.9 (14, 21)	23.1 (19, 28)	Reference
VENULAR WIDENING (CRVE)
(<5%)				4	819	87	5.0 (3.7, 6.8)	8.7 (6.9, 11)	1.63 (p=0.005)
	3.9 (3.2, 4.6)	6.4 (5.6, 7.3)	0.55	3	819	60	3.7 (2.6, 5.3)	6.2 (4.7, 8.2)	1.09 (p=0.6)
				2	819	55	2.9 (1.9, 4.3)	5.5 (4.1, 7.4)	1.02 (p=0.9)
(N=3277)				1	820	55	3.8 (2.6, 5.3)	5.2 (3.8, 7.0)	Reference
(5–7.5%)				4	239	45	10.4 (7.1, 15)	15.5 (11, 21)	1.11 (p=0.6)
	7.4 (5.9, 9.3)	12.6 (11, 15)		3	239	26	5.6 (3.3, 9.4)	11.3 (7.8, 16)	0.62 (p=0.06)
				2	239	30	6.1 (3.6, 10)	9.3 (6.2, 14)	0.69 (p=0.1)
(N=956)				1	239	42	7.4 (4.7, 12)	14.2 (10, 20)	Reference
(>7.5%)				4	429	123	21.5 (18, 26)	27.9 (28, 40)	1.15 (p=0.3)
	16.9 (15, 19)	24.8 (23, 27)		3	430	106	13.7 (15, 21)	23.5 (22, 32)	0.92 (p=0.5)
				2	430	115	15.9 (16, 23)	23.3 (24, 34)	0.99 (p=0.3)
(N=1719)				1	430	118	16.5 (16, 23)	24.5 (24, 35)	Reference

^*Calculated Harrell’s C-statistic of PCE risk groups when CRAE or CRVE is applied.

Discussion

In this large, multiracial, prospective cohort study, arteriolar and venular retinal vessel calibers, as assessed on fundus photography, were associated with all-cause mortality and ischemic stroke in both genders and with CHD in women after a mean follow up of 16 years. The association between retinal vessel calibers with incident HF was non-significant after adjustment for systolic blood pressure. Narrower retinal arterioles and wider venules conferred higher risk for ASCVE in women and provided incremental value above and beyond the ACC/AHA PCE risk score in low-risk women, a large but understudied group. Low-risk women in the narrowest arteriolar quartile were twice as likely to suffer an ASCVE and approximately 4-fold as likely to die from a CHD event versus those with in the widest quartile. Utilizing data from retinal vessel imaging in addition to the PCE would result in 11% of all women (8.5% of total ASCVE in women) being identified at higher risk for ASCVE who would otherwise not be recognized using current practice guidelines. Our results were robust, replicated using quartiles of CRAE or CRVE specific to low-risk women or all women and also when using quartiles of A/V-ratio.

Both arteriolar narrowing and venular dilation are part of the pathological changes that occur within microvascular beds in both hypertensive and atherosclerotic disease ²⁸. Our data support that retinal arteriolar narrowing and venular dilatation are indeed associated with long-term cardiovascular outcomes. Prior reports have suggested that retinal microvascular disease was associated with cardiac mortality^{7, 14–17} but not necessarily all-cause death. Additionally, retinal vascular calibers have been associated with incident stroke ^{6, 9–13, 29, 30} and incident CHD ^5–8, particularly in women, and primarily after short-term follow up. Retinal vascular calibers has not previously been found to be associated with incident HF, but has been associated with left ventricular remodeling as assessed by cardiac MRI ³¹. In the ARIC study population, our findings show an interrelationship between microvascular and macrovascular dysfunction; the later, expressed by elevated systolic BP, seems to be important in the clinical manifestation of HF.

In differentiating between the retinal measures presented in this work, there are some findings worth noting. CRAE and A/V ratio seem to slightly outperform CRVE in prediction of ASCVE in all women as well as in low-risk women. However, CRVE may be a stronger predictor of death compared to CRAE. Different risk factors have been associated with wider CRVE (ie. inflammatory markers, BMI, smoking) versus narrower CRAE (ie. Age, SBP, DBP, mean arterial BP)^{32, 33}. These separate pathophysiological underpinnings likely reflect the different strengths of association observed between retinal vessel calibers and CV outcomes, herein. As CRAE and CRVE may reflect different disease processes, both measures are apparently useful to provide an overall assessment of the risk profile.

CHD poses a significant public health problem in women: it is the primary cause of death in this group, with rates of fatal CHD in women less than 55 years of age climbing ^{34, 35}. The expression of female-specific ischemic CHD is unlike its male counterpart—it is more often non-obstructive in nature with a larger burden of coronary microvascular disease ^{36, 37} and greater morbidity and mortality ^{34, 38}. In a prior analysis of ARIC data, reduced retinal A/V-ratio was predictive of incident CHD (average follow up time 3 years) in women but not men after adjusting for traditional risk factors, a finding that has also been observed in meta-analysis ^{5, 21, 39}. The current study provides longer term follow up and demonstrates incremental predictive value of retinal vessel calibers to our current practice guidelines, demonstrating that it offers significant additional information above and beyond the PCE risk score to low-risk women. Thirty percent of ASCVE occurred in low-risk women, a group that represents the majority (55%) of the female ARIC study population, as is the case for the general female population for which the PCE would be applied: the majority of U.S. women 40–79 years of age (per National Health and Nutrition Examination Survey 2007–2010 data) would be classified as low-risk by the PCE risk score (67% of women compared to 40% of men)²⁶. Compared to low-risk women in the widest arteriolar quartile, those in the narrowest quartile were twice as likely to suffer from an ASCVE.

Further, our data suggests that ASCVE in low-risk women with microvascular dysfunction are more deadly than in women without microvascular dysfunction as assessed by retinal imaging. Because this portion of the population would not typically know that they are at increased risk for cardiovascular disease and often have atypical presenting symptoms, it may represent a group that would particularly benefit from disease prevention education, lifestyle modification and/or therapeutic intervention from their healthcare provider. Prior work has shown that retinal arteriolar narrowing is reversible by antihypertensive therapy ⁴⁰ and, thereby, serial monitoring of retinal vessel calibers by fundus photography could potentially provide a secondary endpoint and visual aid to assist in patient motivation and compliance to dietary, exercise and therapeutic interventions. In recent European Society of Cardiology Guidelines, several new, simple markers were suggested that could specifically improve ASCVE risk prediction in women including: history of pre-eclampsia or pregnancy induced hypertension, gestational diabetes and polycystic ovary syndrome. While these markers typically forewarn of later illness (ie. preeclampsia increases risk for HTN and DM; gestational DM increases risk for DM, etc) it has yet to be established whether they predict ASCVE independent of the conventional risk factors that they predispose to in the ARIC study population. A particular advantage of using retinal imaging over these options is the possibility to “re-quantify” risk and response to life-style interventions via approximately 5-year measures.

Retinal arteriolar narrowing serves as an early marker of systemic microvascular dysfunction and prior work has shown an association between smaller retinal arteriolar caliber and lower hyperemic myocardial blood flow (MBF) and perfusion reserve (PR)⁴¹. Coronary microvascular dysfunction is frequently present in the absence of epicardial atherosclerosis (particularly in women) and non-obstructive microvascular disease is an independent risk factor for adverse cardiovascular events^{42, 43}. Interestingly, the relationship between retinal vessel caliber and MBF and PR was only observed in those subjects without coronary calcification, suggesting that these measures were less reflective of the coronary microvascular processes in more advanced epicardial coronary disease⁴¹. Presumably, the contribution of the microcirculation to myocardial perfusion is significantly reduced in the presence of epicardial stenosis, either from restriction of downstream MBF or via endothelial dysfunction diminishing flow-dependent vasodilation. In our work, we found retinal vessel caliber to be a better predictor of ASCVE in women, most likely for two reasons: 1) female-specific ischemic CHD is frequently non-obstructive with a larger encumbrance of coronary microvascular disease compared to men, ^{36, 37} and 2) women comprise the vast majority of the low-risk PCE category, which is the group in which retinal vessel caliber is best able to discriminate ASCVE risk. Once epicardial coronary disease becomes apparent, as is more often the case in higher–risk men or diabetics, retinal vessel caliber is less likely to be useful in predicting ASCVE.

This study has several limitations. As many retinal photographs were non-gradable (however, equally affecting men and women), it is possible that there was a selection bias influencing the observed associations or resulting in an underestimation of risk. However, when inverse probability weighting was used to account for missing data, no major study findings differed. And notably, low-risk individuals, for whom these measures are most likely to serve, were less likely to have non-gradable photographs. Secondly, utilizing retinal caliber measurements obtained from one randomly selected eye versus the average of both could obscure findings. In clinical practice, retinal photographs could be obtained from both eyes, increasing precision. Thirdly, measurement of retinal vessel size can vary by many factors including inter- and intra- grader reliability and image quality, leading to an underestimation of risk. Finally, these findings need to be replicated in other large cohort studies with well-adjudicated cardiovascular outcomes before these findings are applied on a population level.

In conclusion, our study suggests that retinal vessel calibers are associated with long-term cardiovascular outcomes in a large community-based multiracial cohort, particularly in women. These findings suggest that narrower retinal arterioles and wider retinal venules confer a greater risk of death, stroke and heart failure as well as CHD in women. It remains to be replicated whether adding fundus photography to obtain information to further risk stratify women with a PCE risk score of <5%, will be of benefit in lowering the risk for death or morbidity from ASCVE in this group, which would not otherwise be recognized using current practice guidelines.

Clinical Perspective.

What is new?

• The cornea provides a transparent window into the retinal microvasculature that has been related to cardiovascular outcomes.

• Whether retinal vessel caliber can provide incremental value to current practice guidelines (2013 AHA/ACC Pooled Cohort Equations (PCE)) in predicting Atherosclerotic Cardiovascular Disease Events (ASCVE) hasn’t been established.

• In 10,470 subjects, narrower retinal arterioles and wider retinal venules were associated with long-term risk of mortality and ischemic stroke in both genders and CHD in women independent of PCE risk-score variables.

• Retinal vessel caliber reclassified 21% of low-risk women (11% of all women) as intermediate-risk for ASCVE.

What are the clinical implications?

• The identification of CHD is frequently delayed or undiagnosed in women and many at-risk for adverse outcomes are not offered preventative or therapeutic options.

• This under recognition may be partly due to more prevalent non-obstructive CHD in women, with microvascular dysfunction largely contributing to myocardial ischemia.

• Whether adding retinal imaging to further risk stratify low-risk women will result in the attenuation of risk of death or morbidity from ASCVE in this group, which would be unrecognized using current practice guidelines, remains to be determined.