The Relationship of Retinal Vessel Caliber With Erectile Dysfunction in Patients With Type 2 Diabetes

Abstract

Purpose.: To investigate the relationship of retinal vessel caliber with erectile dysfunction (ED) in males with type 2 diabetes.

Methods.: A hospital-based cross-sectional study. Male patients with type 2 diabetes were recruited from the Diabetic Management Project. All underwent a complete eye examination, a comprehensive interview, and blood and urine tests. Retinal vessel diameter was measured from retinal photographs by trained graders using semiautomated software. ED was defined as problems achieving or maintaining an erection and was assessed using a self-reported questionnaire.

Results.: A total of 289 male patients with a mean (±SD) age of 65.3 years (±11.2) were assessed. After adjusting for age, diastolic blood pressure, duration of diabetes, HbA_1c, total cholesterol, presence of diabetic retinopathy, and any diabetic complication, narrower retinal arteriolar diameter (odds ratio [OR] 1.66; 95% confidence interval [CI] 1.09–2.54; P = 0.019) and wider venular diameter (OR 1.58; 95% CI 1.03–2.44; P = 0.038) were associated with ED.

Conclusions.: Narrower retinal arteriolar and wider venular diameter are independently associated with an increased risk of self-reported ED. These results suggest a microvascular component in the pathogenesis of this condition.

Introduction

Erectile dysfunction (ED) affects 35% to 90% of diabetic men ¹ and is associated with increased morbidity and mortality. ^2,3 Although ED has organic as well as psychogenic components, it is typically associated with atherosclerosis and diabetes. ⁴ More recently, impaired release of nitric oxide from endothelial dysfunction in addition to oxidative stress have been proposed as a mechanism by which ED develops. ⁵ These suggested pathophysiological mechanisms indicate a microvascular component mediated by endothelial dysfunction that has yet to be fully elucidated. 

Examination of the retinal vasculature is noninvasive and provides a means to study the structural and pathological changes of the microcirculation. ⁶ Changes in retinal microvascular caliber reflect different pathophysiological processes and may provide insight into the understanding of microvascular disease and its contribution toward the pathogenesis of ED. 

To our knowledge, no study has investigated the association between ED and retinal vascular caliber, although they share common cardiovascular risk factors including age, hypertension, and obesity (associated with narrower arteriolar diameter) ^7–9 ; diabetes, smoking, dyslipidemia, inflammation, and endothelial dysfunction (associated with wider venular diameter) ^7,10–12 ; in addition to being independent risk factors for incident coronary heart disease. ^13,14  

In this paper, we therefore investigated the association between retinal arteriolar and venular diameters with the prevalence of ED in men with type 2 diabetes. 

Methods

Study Population

The Diabetes Management Project (DMP) is a clinical study of adults with diabetes residing in Victoria, Australia. The methodology and objectives of the study have been reported in detail elsewhere. ¹⁵ English-speaking adults aged 18 years or older with type 2 diabetes, free of significant hearing and cognitive impairment, and living independently in the community were invited to participate. Participants were recruited from the Royal Victorian Eye and Ear Hospital (RVEEH). All study procedures adhered to the tenets of the Declaration of Helsinki, and all privacy requirements were met. Ethical approval for the DMP was provided by the RVEEH Human Research and Ethics Committee (08/815H). Each individual signed a consent form that outlined the purposes and methodology of the DMP, including possible adverse outcomes, confidentiality policies, and data storage procedures. 

Objective Measurements

Each participant underwent a BP assessment with an automated BP machine (model 5200-103Z; Welch Allyn, Rydalmere, NSW, Australia). The average of two separate measurements was recorded for both systolic (SBP) and diastolic (DBP) BP. In cases in which there was a difference greater than 10 mm Hg for SBP and 5 mm Hg for DBP, a third measurement was taken. The average of the two closest BP measurements was then used for the analysis. All individuals had height and weight measured with a wall-mounted, adjustable measuring scale (Surgical and Medical Products, China) and a calibrated digital scientific weight scale (PRC; Oregon Scientific, Tualatin, OR), respectively. Individuals were instructed to remove any footwear and heavy clothing before testing. Body mass index (BMI) was calculated as weight (in kilograms) divided by height in meters squared (kg/m²). Fasting blood (>8 hours) was obtained to assess HbA_1c levels, fasting glucose and lipids (total cholesterol [C], triglyceride, low density lipoprotein (LDL) cholesterol, and high density lipoprotein cholesterol). All biochemical parameters were analyzed at Melbourne Pathology. Fasting plasma glucose and serum lipids were assessed (Hitachi Modular P analyzer; Roche Diagnostics, Mannheim, Germany). 

Assessment of DR

To assess diabetic retinopathy, two 45° digital, nonstereo color fundus photographs of both eyes were taken from each individual with a nonmydriatic retinal camera (CR6-45NM; Canon, Inc., Tokyo, Japan), with all images being stored electronically (Digital Healthcare software; Cambridge, UK). Dilation was achieved with a single drop of tropicamide 1% (MINIMS; Chauvin Pharmaceuticals, Ltd., Kingston-on-Thames, UK) in each eye approximately 25 minutes before fundus photography. DR grading adhered to the two-standard-field color fundus photography procedure based on the modified Early Treatment Diabetic Retinopathy Scale used in the Multi-Ethnic Study of Atherosclerosis. DR was classified into two groups, nonproliferative DR (NPDR) and proliferative DR (PDR), according to the international clinical diabetic retinopathy and diabetic macular edema disease severity scales. ¹⁶  

Measurement of Retinal Vascular Caliber

All participants had fundus photography performed using a standardized protocol according to the Multi-Ethnic Study of Atherosclerosis. ⁷ An optic disc centered photograph was taken of each eye. Retinal vascular caliber was measured by a masked grader, using a computer-based program (IVAN, University of Wisconsin, Madison, Wisconsin), according to an established protocol. ^17,18 In all cases due to the disc centration of photographs, the macula status was not visible to the grader. For each photograph, all arterioles and venules coursing through an area one-half to one-disc diameter from the optic disc margin were measured and vessel calibers of the biggest six were summarized as the central retinal artery equivalent (CRAE) and central retinal vein equivalent (CRVE) using formulas developed by Parr and Hubbard and later modified by Knudtson. ^17,19 CRAE was derived from the mean CRAE for both eyes. If the mean could not be calculated (due to missing CRAE value in one eye) then the CRAE from the nonmissing eye was used. Pearson's correlations were 0.92 for mean CRAE and right eye CRAE, 0.95 for mean CRAE and left eye CRAE, 0.92 for mean CRVE and right eye CRVE and 0.94 for mean CRVE and left eye CRVE. 

Assessment of Self-Reported ED and Other Conditions

Participants were asked to complete a number of questionnaires administered by trained interviewers. The comprehensive list of questionnaires captured data from behavioral, lifestyle, and medical factors, which included the presence of erectile dysfunction in male patients. Male diabetic participants were asked about diagnosed impotence, which was explained to the patient as problems sustaining or maintaining an erection. Using categories “Yes,” “No,” or “Unsure,” participants indicated whether they had been diagnosed with ED, with those indicating “unsure” placed in the same group as those without ED. If the participant selected yes, they were asked to report their age of diagnosis. Patients were also asked about the use of phosphodiesterase inhibitors (e.g., sildenafil and its analogues) and were excluded from the study if they reported using it. Details about other self-reported microvascular complications included nephropathy (kidney disease), neuropathy (nerve damage), and retinopathy (diabetic eye disease). Self-reported macrovascular complications included cardiovascular disease (i.e., ischemic heart disease, myocardial infarction, and stroke); vascular ischemia (e.g., nontraumatic amputation); and peripheral vascular disease. 

Statistical Analysis

Patients' demographics and characteristics were summarized using mean and standard deviation or median and interquartile range (IQR) values for data measured on a continuous scale, and counts and percentages for categorical variables. Percentages have been rounded up to the nearest whole number and may not always add up to 100%. Student's unpaired t-test or Wilcoxon rank-sum test was used for the comparison of continuously distributed variables and the χ² test was used to assess for differences in frequency distributions. 

Potential prognostic factors associated with ED included age, blood pressure, BMI, duration of diabetes, smoking status, HbA_1c, total serum cholesterol, medications (for DM, BP, and dyslipidemia); any diabetic complications (nephropathy, peripheral vascular disease [PVD] or neuropathy); presence of DR; and retinal vessel caliber (CRAE, CRVE). Assessment of potential prognostic factors was made using χ² tests. 

Multifactor analyses of retinal vessel caliber as potential prognostic factors for ED adjusting for other factors were carried out using multivariate regression. All probabilities quoted are two-sided and all statistical analyses were undertaken using statistical software (Intercooled Stata version 11.2; StataCorp, College Station, TX). 

Results

Out of the 290 type 2 diabetic male patients, one reported use of phosphodiesterase inhibitor and was excluded, leaving a total of 289 patients (mean age 65.3 years; SD, 11.2 years) for the analysis. Table 1 shows the baseline characteristics of the patients stratified by the presence or absence of ED, with 106 (36.7%) male patients reporting ED as being present. 

Table 1

 

Baseline Characteristics of Participants by ED Diagnosis

Characteristic*	ED		P Value
	Absent (n = 183)	Present (n = 106)
Continuous Variable
Mean, SD
Age, y	65.1 (12.3)	65.6 (9.0)	0.69
SBP, mmHg	140.7 (19.0)	139.6 (17.3)	0.63
DBP, mmHg	77.6 (8.4)	75.2 (8.2)	0.024
BMI, kg/m2	29.7 (5.7)	30.9 (5.7)	0.09
Duration of diabetes, y†	12.0 (15.0)	16.0 (11.0)	0.001
HbA1c (%)	7.60 (1.57)	8.0 (1.34)	0.026
Total serum cholesterol, mmol/L	4.62 (1.22)	4.32 (1.23)	0.056
Median (IQR)†
CRAE, μm	141.66 (22.45)	136.18 (24.95)	0.006
CRVE, μm	206.1 (37.4)	207.2 (31.3)	0.91
Smoker, y	20 (22.0)	20.0 (23.0)	0.388
Pack years	206.1 (37.4)	207.2 (31.3)	0.106
Categorical variable, n (%)
Current/Past smoker	115 (62.8)	70 (66.7)	0.51
Antihypertension medication	117 (71.8)	61 (71.8)	1.0
Lipid-lowering medication	104 (63.8)	58 (68.2)	0.49
Diabetes medication	143 (84.1)	85 (89.5)	0.23
Presence of DR	104 (57.1)	79 (76.0)	0.001
Diabetic complication (renal, PVD, or neuropathy)	48 (26.2)	51 (48.1)	<0.001

 Subjects who were unsure of the diagnosis of ED were grouped together with those with no ED.

*  Characteristics were expressed as the mean (SD) for continuous variables and the number (%) for dichotomous variables.

†  Median IQR and Wilcoxon rank-sum test was used for the comparison of nonparametric variables.

Risk factors significantly associated with ED in univariate analyses were presence of DR (odds ratio [OR] 2.37; 95% confidence interval [CI] 1.39–4.06); any diabetic complication (OR 3.76; 95% CI 2.07–6.86); low diastolic blood pressure (OR 0.97; 95% CI 0.94–0.995); increasing duration of diabetes (OR 1.03; 95% CI 1.01–1.05); increasing HbA_1c (OR 1.20; 95% CI 1.02–1.42); low total cholesterol (OR 0.82; 95% CI 0.66 to 1.01); CRAE per SD decrease (OR 1.66; 95% CI 1.16 to 2.39); CRVE per SD increase (OR 1.40; 95% CI 0.98 to 2.0; Table 2). A subgroup analysis of length and amount of cigarettes smoked demonstrated no association with ED (data not shown). In our sample of subjects, we also demonstrated a high level of correlation (r = 0.95) between total serum cholesterol and LDL, both of which were not associated with ED (data not shown). 

Table 2. 

 

Determinants of ED in Males With Type 2 Diabetes

Exploratory Variable	Univariate Model*		Multivariate Model†
	OR (95% CI)	P Value	OR (95% CI)	P Value
CRAE (per SD decrease)	1.66 (1.16, 2.39)	0.006	1.66 (1.09, 2.54)	0.019
CRVE (per SD increase)	1.40 (0.98, 2.00)	0.067	1.58 (1.03, 2.44)	0.038
Age, y	1.00 (0.98, 1.03)	0.69	1.02 (0.99, 1.05)	0.31
SBP, mm Hg	1.00 (0.98, 1.01)	0.63
DBP, mm Hg	0.97 (0.94, 0.995)	0.025	0.97 (0.93, 1.01)	0.11
BMI, kg/m2	1.04 (0.99, 1.08)	0.09	—	—
Duration of diabetes, y	1.03 (1.01, 1.05)	0.013	1.01 (0.97, 1.04)	0.74
HbA1c, %	1.20 (1.02, 1.42)	0.028	1.18 (0.96, 1.47)	0.12
Total cholesterol, mmol/L	0.82 (0.66, 1.01)	0.058	0.86 (0.67, 1.09)	0.21
Current/past smoker vs. nonsmoker	1.18 (0.71, 1.96)	0.52	—	—
Lipid-lowering medication	1.22 (0.70, 2.13)	0.49	—	—
Hypertension medication	1.00 (0.56, 1.79)	1.00	—	—
Diabetic medication	1.60 (0.74, 3.48)	0.23	—	—
DR	2.37 (1.39, 4.06)	0.002	1.39 (0.71, 2.72)	0.34
Diabetic complication (renal, PVD or neuropathy)	3.76 (2.07, 6.86)	<0.001	2.22 (1.25, 3.95)	0.006

*  Logistic regression unadjusted model.

†  Model for CRAE and CRVE as a variable for ED after adjusting for age and all variables found to be significant in the univariate regression analysis.

CRAE and CRVE were adjusted simultaneously in the multivariate regression model. Potential confounders including age and all variables that were significant in the univariate regression analyses (DBP, duration of diabetes, HbA_1c, blood cholesterol levels, presence of DR and any diabetic complication) were adjusted for. CRAE per SD decrease (OR 1.66; 95% CI 1.09–2.54); CRVE per SD increase (OR 1.58; 95% CI 1.03–2.44); and any diabetic complication (OR 2.22; 95% CI 1.25–3.95) remained independently associated with ED. However, DR was not found to be independently significantly associated with ED indicating that CRAE/CRVE is associated with ED, independent of DR (Table 2). 

Discussion

This study provides novel clinic-based data on the relationship of retinal arteriolar and venular diameters with the prevalence of ED in a clinical sample of men with diabetes and diabetic retinopathy. We demonstrated that narrower arteriolar diameter and wider venular diameter are all associated with ED independent of age, diastolic blood pressure, duration of diabetes, HbA_1c, total cholesterol, presence of diabetic retinopathy, and any diabetic complication in this population. Past studies have shown relationships between retinal vessel diameter with other vascular diseases including coronary heart disease, stroke, peripheral vascular disease, and renal disease, ^20–23 suggesting a microvascular component in its pathogenesis. This study contributes to the hypothesis that the microcirculation—easily visualized in the eye—is intimately linked to various vascular processes systemically. 

Traditionally, ED has been classified according to its pathophysiological mechanisms. For example, vasculogenic ED (accounting for the majority of cases) and neurogenic ED, both of which have diabetes as a risk factor. ^24,25 Although it is convenient to classify ED by separate etiologies as mentioned above, in reality it is not as straightforward. Recent papers have concluded that the pathophysiological mechanisms leading to ED is rarely due to a single cause but consists of a complex interplay of multiple factors. ^26,27  

With this in mind, we note that the relationship between narrowing CRAE and widening CRVE with ED provide some insight into its pathogenesis. Because systemic risk factors for arteriolar and venular diameter differ markedly, it supports recent advances that the development of ED is almost always multifactorial. 

We propose two hypotheses for the relationship between wider retinal venular diameter and ED. First, both ED and wider retinal venular diameter share many similar risk factors, including smoking, dyslipidemia, and inflammation, which contribute to accelerated atherosclerosis. Wider retinal venular diameter (CRVE) has been associated with higher levels of inflammatory biomarkers and thought to reflect inflammation. ⁷ Retinal venular diameter could therefore possibly be a surrogate marker for those at increased risk of atherosclerosis. This is consistent with results from the Hoorn Study, which found a relationship between retinal venular dilatation and increased intima-media thickness (an early marker of atherosclerosis), ²⁸ and the Rotterdam Study, which noted the association between wider venular diameter and atherosclerosis. ¹¹  

Second, a key event in the pathogenesis of ED is endothelial dysfunction, ²⁹ which may be reflected by wider CRVE. This suggestion lends further support to previous studies that demonstrated an association between increased markers of endothelial dysfunction with wider CRVE. ⁷ Recent studies have demonstrated a blunted vasodilatory response of the retinal vessels to nitric oxide in diabetic persons, ^30,31 reinforcing the role of endothelial dysfunction in widened venular diameter. In addition, ED is also associated with an increased risk of cardiovascular disease incidence, ^22,32 the common denominator being endothelial dysfunction, ³³ suggesting that endothelial dysfunction partly explains the association between retinal venular diameter and ED. 

Hypertension is a possible explanation for the observed relationship between retinal arteriolar diameter and ED. Narrower arteriolar diameter is strongly associated with hypertension, ⁸ a risk factor for ED through accelerated atherosclerosis, damage to smooth muscle and use of antihypertensive medications. ^34,35 However, even after adjusting for hypertension in our study, narrower arteriolar diameter remained associated with ED. One possible explanation is that atherosclerosis, which causes flow-limiting stenosis in penile arterioles, is also a cause of increased peripheral resistance. ^36,37 This increased peripheral resistance may manifest as retinal arteriolar narrowing, which interestingly is also associated with future hypertension. ⁸ Another related mechanism is the failure of nitric oxide–dependent vasodilation accounting for ED and narrower retinal arteriolar diameter. In addition to its release from noncholinergic penile nerve endings, nitric oxide is also produced by the vascular endothelium. ^26,27 It is likely that damage to the endothelium from atherosclerosis and chronic inflammation caused by hypercholesterolemia, hypertension, and diabetes result in impaired nitric oxide release in both the retinal and penile vascular endothelium. However, the specific pathophysiological mechanisms resulting in impaired nitric oxide release are complex and beyond the scope of this article. ²⁷  

An additional point of discussion is the presence of DR as a risk factor for ED. In the univariate analysis, DR was associated with a 2-fold increased odds of having ED, consistent with the literature and a recent publication of ours. ^38,39 We note, however, that this association was no longer significant after adjusting for other factors, in particular CRAE and CRVE diameter. The reasons for this finding are unknown, but a possible inference is that retinal vessel diameter is intimately linked with DR which shares multiple risk factors that are known to influence retinal vessel diameter, namely endothelial dysfunction, hypertension, and body mass index. ^38,40,41  

Strengths of our study include a large clinical sample of people with diabetes, controlling for various confounders that increase risk of ED including medications and the use of standardized measurement methods for measuring retinal vessel diameter using dilated fundus photography. 

Limitations include the diagnosis of ED through the patients' history without objective measures (Doppler ultrasound of the penis). However, we believe that this would result in under- rather than overreporting of ED, as it is unlikely individuals would disclose a diagnosis of ED without suffering any symptoms. Additionally, previous studies have demonstrated a moderate-to-high correlation and agreement between patient self-assessments and validated questionnaires used in the diagnosis of ED. ^42,43 O'Donell et al. also concluded that a single question of self-reported erectile dysfunction correlated well with clinically diagnosed erectile dysfunction. ⁴⁴ Hence, we believe the diagnoses of significant ED made in this study to be reliable. Another limitation was the cross-sectional nature of the data which did not allow us to determine the temporal relation between retinal vessel caliber and ED. A further limitation was the diagnosis of diabetic complications through a self-reported patient questionnaire, in addition to a lack of data on testosterone levels. 

In conclusion, narrower retinal arteriolar and wider venular diameters are independently associated with the presence of ED, suggesting multiple microvascular processes involved in its pathogenesis. Cohort studies to establish which variable precedes the other will provide further insight into the pathogenesis of ED and retinal vessel diameter changes. 

Acknowledgments

Supported by an Australian Research Council Linkage grant [LP0884108]. CERA receives Operational Infrastructure Support from the Victorian Government. This study was partly supported by the National Health and Medical Research Council Centre for Clinical Research Excellence #529923 – Translational Clinical Research in Major Eye Diseases. 

Disclosure: S.K.H. Chew, None; Y. Taouk, None; J. Xie, None; T.E. Nicolaou, None; J.J. Wang, None; T.Y. Wong, None; E.L. Lamoureux, None