Clin. Cardiol. 26, 25–30 (2003)
Prevalence and Risk Factors for Erectile Dysfunction in Men with Diabetes, Hypertension, or Both Diseases: A Community Survey among 1,412 Israeli Men ARIE ROTH, M.D., OFRA KALTER-LEIBOVICI, M.D.,* YEHUDA KERBIS, M.D.,† ELLA TENENBAUM-KOREN, M.D.,† JUZA CHEN, M.D.,‡ TAMAR SOBOL, M.D.,† ITAMAR RAZ, M.D.§ Departments of Cardiology and ‡Urology (Sexual Dysfunction Unit), Tel-Aviv Sourasky Medical Center, Tel-Aviv, and the *Unit of Cardiovascular Epidemiology, Gertner Institute for Epidemiology and Health Policy Research, Sheba Medical Center, Tel-Hashomer, both affiliated with the Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv; †Medical Division of Pfizer Pharmaceutical (Israel) Ltd.; §Department of Medicine, Hadassah Medical Center, Jerusalem, affiliated with the Faculty of Medicine, Hebrew University, Jerusalem, Israel
Summary
Background: Erectile dysfunction (ED) and cardiovascular disease share common risk factors and may be further aggravated by medical treatment for reducing them. Hypothesis: The study was undertaken to assess the prevalence of ED in patients with diabetes (DM), hypertension (HTN), or both diseases, and to evaluate the effect of patient age, medical treatment, and disease duration and control on the prevalence of ED in this population. Methods: A group of 150 primary practitioners who had patients with known DM and/or HTN conducted a survey, utilizing IIEF-15, a 15-item multidimensional, self-administered questionnaire used for the clinical assessment of ED. Results: In all, 1,412 patients were included: 37% had DM, 38% had HTN, and 25% had both diseases. Their mean age was 55, 58, and 60 years, and 62, 46, and 67% had some degree of ED, respectively. The prevalence of ED increased with age and disease duration in each age group and was higher in subjects with DM than in those with HTN, especially in those aged < 65 years. Poor glycemic control was associated with a higher prevalence rate of ED early in the course of the disease. There was no significant difference in the prevalence of ED according to type and number of antihypertensive drugs.
Address for reprints: Arie Roth, M.D. Department of Cardiology Tel-Aviv Sourasky Medical Center 6 Weizman Street Tel-Aviv, 64239 Israel e-mail:
[email protected] Received: October 22, 2001 Accepted with revision: February 27, 2002
Conclusions: Erectile dysfunction is common among patients at high risk for cardiovascular disease because of diabetes and/or HTN. Diabetic men are affected earlier than those with HTN. Given the high frequency of ED in young patients with these risk factors, physicians should encourage an open discussion on the subject during routine visits to promote early detection and treatment. Key words: hypertension, diabetes, erectile dysfunction
Introduction Erectile dysfunction (ED) is defined as the inability to achieve and/or maintain an erection sufficient to permit satisfactory sexual intercourse.1 It has been estimated to affect up to 30 million men in the United States alone.2, 3 Erectile dysfunction and cardiovascular diseases share common risk factors, for example, hypertension, diabetes mellitus, and hyperlipidemia,2, 3 and may be further aggravated by the medical treatment such as antihypertensive drugs often given to patients with these risk factors.4, 5 Normal penile erection is a hemodynamic process that is dependent upon corporal smooth muscle relaxation mediated by parasympathetic neurotransmission, nitric oxide (NO), electrophysiologic events, and possibly other regulatory factors. Erectile dysfunction can be due to vasculogenic, neurogenic, hormonal abnormalities, and cavernosal impairment with or without psychological factors, resulting in an imbalance in corporal smooth muscle contraction and relaxation.6 Nitric oxide activity is adversely affected in penile and vascular tissue by both hypertension and diabetes, and the association between ED and ischemic heart disease may be attributed to defective NO activity in both types of tissue.7 The aim of the current study was to evaluate the prevalence of ED among patients with hypertension, diabetes, or both diseases, who were being treated at primary care clinics. We further investigated the effect of age, disease duration, medications, and disease control on the frequency of ED.
26
Clin. Cardiol. Vol. 26, January 2003
Methods Primary practitioners working in primary care clinics of all four HMOs throughout Israel were asked to participate in the study. They invited their male patients with diabetes, hypertension, or both diseases to participate in the study; those who consented were entered consecutively. Data collection was completed within 12 months (Jan.–Dec. 2000). The diagnosis of ED was based on the patients’ answers to the 15-question International Index of Erectile Function (IIEF15),8 a validated, multidimensional, self-administered questionnaire used for the clinical assessment of erectile function (EF). Erectile dysfunction was defined both by the presence of any degree of ED according to a summarized score of 7 or below to questions 3 and 4, and a severity category according to the total score of the ED domain (questions 1–5, 15). The severity of ED was classified into five categories: no ED (EF score 26–30), mild (EF score 22–25), mild to moderate (EF score 17–21), moderate (EF score 11–16) and severe (EF score 6–10).9 The data on patients’ age, diagnosis, disease duration, the type of hypoglycemic and/or antihypertensive therapy as well as the current HbA1c value in diabetic patients was provided by the primary practitioners according to the information available in the patients’ clinical records. Statistical Analyses
Univariate analyses included the unpaired t-test or one-way analysis of variance for continuous variables with a normal distribution, and the Mann-Whitney or Kruskal-Wallis tests for other distributions. The chi-square test was employed to assess the significance of contingency tables. We calculated the common odds ratio (OR) using the Mantel-Haenszel statistic in order to look for significant associations with ED across various strata of a possible confounding factor. For multivariate analyses, we used the logistic regression model with forced entry of the dependent variables. Variables entered into the model were those found to be associated with the binary variable of any ED with a level of significance < 0.1 in the univariate analyses. The patient’s disease duration and HbA1c value were entered into the multivariate models either as below, equal to, or above the median value. We used the Hosmer and Lemeshow test to evaluate the model goodness
of fit, and the Wald test to look for variables that were significantly and independently associated with ED. All comparisons were two-tailed, and the critical value for statistical significance was set at 0.05.
Results In all, 150 primary practitioners recruited 1,412 subjects of whom 518 (37%) had diabetes, 541 (38%) had hypertension, and 353 (25%) had both diseases. The refusal rate had been about 15%. Most of the diabetic men (88%) had type 2 disease. Of the diabetic men, 59% were treated with oral hypoglycemic drugs, 12.2% were receiving insulin, and 8.4% were treated with insulin and oral hypoglycemic drugs, while 20.4% were on dietary treatment alone. The most frequent antihypertensive drugs were angiotensin-converting enzyme (ACE) inhibitors (58%), calcium-channel blockers (38%), beta-adrenergic blockers (36%), and diuretics (29%). Of the hypertensive men, 53% were on two or more antihypertensive drugs. The four most frequent antihypertensive drug combinations were ACE inhibitors and diuretics (7%), ACE inhibitors and calcium-channel blockers (7%), ACE inhibitors and beta-adrenergic blockers (6%), and calcium-channel blockers and beta-adrenergic blockers (4%). The men with combined hypertension and diabetes were significantly older and had longer disease duration than the men with either disease alone (Table I). General Associations
Erectile dysfunction was found among 57% of the subjects, 9% of them to a severe degree. The prevalence of ED was significantly lower among the men with hypertension than in the men with diabetes or with both diabetes and hypertension. The presence of hypertension did not significantly increase the prevalence of ED among diabetic men (p = 0.11). The prevalence of ED increased significantly with age. The mean age ± standard deviation (SD) of men with ED was 59 ± 10 years compared with 53 ± 10 years in men without ED (p < 0.001). Patients’ age increased with ED severity. The mean age ± SD was 62 ± 10 years in men with severe ED, 60 ± 10 years in men with moderate ED, 57 ± 10 years in men with
TABLE I Characteristics of the study subjects
Age, years (mean ± SD) Disease duration, years (median, range) Any ED (%) Severe ED (%)
Hypertension (n = 541)
Diabetes (n = 518)
Diabetes and hypertension (n = 353)
p Value
58 ± 11 6 (0–36) 46 6
55 ± 12 7 (0–59) 62 11
60 ± 9 10 (0–50) 67 12
< 0.001 a < 0.05 a < 0.001 < 0.001
a Each group is significantly different from the two other groups.
Abbreviations: ED = erectile dysfunction, SD = standard deviation.
A. Roth et al.: Erectile dysfunction and selected coronary risk factors 90
70 60 50 40 30 20 < 50
50–56 56–64 65+ Quartiles of subjects’ age (years)
FIG. 1 Frequency of erectile dysfunction by age and disease category. The dotted line denotes men with hypertension, the broken line men with diabetes, and the uninterrupted line men with both diabetes and hypertension.
mild to moderate ED, 54 ± 10 years in men with mild ED, and 51 ± 11 years in men with normal erectile function (p < 0.001). The prevalence of ED increased significantly with age in all three study groups, but less steeply among subjects with hypertension alone (Fig. 1). The prevalence of ED also increased significantly with disease duration both in men with diabetes and in those with hypertension: the prevalence of ED was higher among diabetic men with and without hypertension (p < 0.001). In diabetic men with concurrent hypertension there was no significant change in the risk of ED with disease duration (adjusted OR 1.04, 95% confidence interval [CI]: 0.75–1.45) (Fig. 2). Associations in Men with Diabetes
Frequency of erectile dysfunction (%)
The prevalence of ED among diabetic men with current HbA1c values equal to or below the median (HbA1c < 7.9%) was significantly lower than in those with HbA1c values equal to or above the median (58 vs. 69%, p = 0.02). This difference was more prominent in men with diabetes and hypertension
(57 vs. 78%, p = 0.001). The odds for having ED among diabetic men with current HbA1c values equal to or above the median were twice as high as for men with HbA1c values below the median (adjusted OR 2.0, 95% CI: 1.41–2.86). The prevalence of ED increased more steeply with diabetes duration among men with HbA1c values equal to or above the median (Fig. 3). Men with type 1 diabetes had a significantly lower prevalence of ED than did men with type 2 disease (53.5 vs. 65.5%, p = 0.03). Since the patients with type 1 diabetes were significantly younger than those with type 2 disease (48 ± 13 vs. 58 ± 10 years, p < 0.001), and given that age was significantly associated with ED, we stratified the patients according to the age quartiles and looked at the association between the type of diabetes and the prevalence of ED in each stratum. The association between the type of diabetes and ED was no longer statistically significant after correcting for the confounding effect of age (common OR = 1.04, 95% CI: 0.64–1.70, p = 0.86). There was no significant association between the type of hypoglycemic regimen and the prevalence of ED (data not shown). To look for variables that were significantly and independently associated with the presence of ED in diabetic men, we performed a multivariate logistic regression analysis. The variables entered into the model were disease duration, current HbA1c value, patient’s age, and an interaction term of disease duration and current HbA1c value. The following variables were found to be significantly and independently associated with ED among diabetic men in the multivariate model: current HbA1c value ≥ 7.9% (OR = 2.11, 95% CI: 1.43–3.11, p < 0.001), a 5-year increase in patient’s age (OR = 1.26, 95% CI: 1.15–1.39, p < 0.001), a disease duration ≥ 9 years (OR = 2.11, 95% CI: 1.42–3.10, p < 0.001), and the interaction term between diabetes duration and HbA1c (OR = 2.50, 95% CI: 1.16–5.38, p = 0.02). Thus, the odds of presence of any ED in patients with a current HbA1c value ≥ 7.9% and duration of diabetes ≥ 9 years were 2.5 times higher than those in subjects having both parameters below these values.
90
90
80
80
70 60 50 40 30
