Erectile dysfunction (ED) is defined as the consistent inability to obtain and/or maintain a penile erection which is sufficient to permit satisfactory sexual intercourse.1,2 The prevalence of ED is estimated at 48% among Portuguese men aged 40–69 years old.3

Obstructive sleep apnea syndrome (OSAS) is characterized by repetitive collapse of the upper airway due to the laxity of pharyngeal dilator muscles4 during sleep.

OSAS affects 4% of men between 30 and 60 years,5 but it is believed that the proportion of clinically diagnosed OSAS is underestimated.6–8 This is one of the most important medical conditions to have been identified in the last 50 years.9 It is related to increased morbidity and mortality due to clinical complications such as hypertension,10 congestive heart failure,11 acute myocardial infarction,12 stroke,13 diabetes,14 cognitive dysfunction15 and depression.16

In every REM sleep stage most men experience a sleep related erection (SRE),17 an event that ensures functional and morphological integrity to erectile tissue.18 In OSAS, intermittent hypoxic events and sleep fragmentation limit SRE, with serious consequences for erectile physiology.19–22 The literature review also defines hormonal,23–26 neural,27–30 endothelial31–35 and psychogenic16 mechanisms to explain ED complaints in OSAS.

OSAS and ED may also be connected through co-morbidities such as hypertension and diabetes.21

As there are some inconsistencies in regard to the prevalence data,18 the main purpose of this study was to estimate the prevalence of ED in a population of OSAS patients sent to Hospital S. João for diagnosis and follow-up. Additionally, clinical and demographic information was collected to obtain the ED determinants in our population.

The general characteristics of the population are described in Table 1. The mean age was 52 years. Of 62 patients studied, 28 (45.1%) were obese, 23 (37.1%) had dyslipidemia, 11 (17.7%) metabolic syndrome (MS), 36 (58.1%) arterial hypertension, 10 (16.1%) diabetes and 4 (6.5%) heart failure. Previous acute myocardial infarction, stroke and pelvic surgery were identified in 2 (3.2%), 6 (9.7%) and 2 (3.2%) participants, respectively. Mild OSAS was diagnosed in 30 (48.4%), moderate OSAS in 14 (22.6%) and severe OSAS in 17 (27.4%) individuals. The ED prevalence was 64.4%, it was mild in 24 (38.7%), mild to moderate in 12 (19.3%), moderate in 1 (1.6%) and severe in 3 (4.8%) participants.

Univariate analysis results are shown in Table 2. Aging was found to be significantly related to ED: OR (95% CI) of 3.90 (1.00–15.28); 5.50 (1.16–26.14); and 6.00 (1.00–35.91) in categories 46–55, 56–65 and >65 years, respectively. Past smoking habits were associated with ED (OR 4.32 (1.09–17.11). Hypertension and ACEI (Angiotensin Converting Enzyme Inhibitors) or ARB (Angiotensin II Receptor Blockers) therapy revealed a statistically significant association with ED (OR 3.25 (1.09–9.65) and 7.39 (1.52–35.99)).

No association was found between dyslipidemia (p=0.239), metabolic syndrome (p=0.215), chronic therapy with beta blockers (p=0.217), calcium antagonists (p=0.827) and serotonin-selective reuptake inhibitors (p=0.250) and ED.

All patients with diabetes, stroke, cardiac failure and previous pelvic surgery referred to ED on questionnaire, so it was not possible to calculate odds ratio (OR) for these variables.

PCSS parameter analysis is described in Table 3. No association was verified between ED and OSAS severity measured by ODI, minimum and medium oxygen saturation and AHI (p=0.494, p=0.657, p=0.498 and p=0.403, respectively).

Multivariate analysis results are shown in Table 4. Age and diabetes were independent risk factors for more severe degrees of ED: OR (95% CI) of 1.226 (1.062–1.415) and 31.205 (1.222–796.557), respectively. There was no evidence that BMI (p=932), smoking habits (p=853), alcohol consumption (p=0.683), hypertension (p=0.077) and OSAS severity (p=0.661) were independent risk factors for ED.

In 1981, Schmidt and Wise were the first to describe a relationship between ED and sleep disorders.48 After that, several studies confirmed the increased prevalence of ED in OSAS patients51–54: Guilleminault et al. reported ejaculatory dysfunction and decreased libido in 48% men with OSAS49; Hirshkowitz et al. verified that 91.3% patients with ED symptoms also had OSAS50; Seftel et al. concluded that 40% OSAS patients had ED.51 However, the association was rejected by Schiavi et al.52 In the present study, a 64.4% ED prevalence was found, which is consistent with most published material on the subject.

In our sample, ED was not associated with severity of OSAS. However it can be seen that the group with severe OSAS did have a higher ED prevalence (Table 2) and patients who reported severe ED had the worst results in sleep study (Table 3), which indicates that studies with higher sample size are needed to support the association Margel et al. concluded that ED is associated exclusively with severe OSAS.53 Furthermore, trials evaluating OSAS treatment found improvements in ED,54–58 which shows that there may be an association between both disorders, which is understandable given that patients on CPAP treatment show decreased hypoxia, improved endothelial function, decreased blood pressure and sympathetic hyperactivity, all of which can improve erectile function.

In this study, age proved to be an independent risk factor for ED, as suggested by the literature.59 It was also confirmed that the risk of ED increases with aging. It is already established that age has more impact on erectile function than the severity of OSAS.60

In addition, the data confirmed that Diabetes is an independent risk factor for ED. It can be seen that diabetic patients have a 31 times higher risk for more severe ED, which is supported by the effect of hyperglycemia on penile neurovascular structure.61

Arterial hypertension was associated with ED in the univariate analysis. However, an association with more severe ED was not confirmed by the multivariate analysis. This difference is probably due to a confounding effect of diabetes, suggested by the fact that 9 out of 10 diabetic patients also suffered from hypertension (data not shown in tables).

An association between MS and ED was not found. Nevertheless, it is possible that the classification used underestimated the actual proportion of affected patients since it was based on patient report and did not include triglycerides or cholesterol plasmatic measurements.62 As a marker of systemic endothelial dysfunction, dyslipidemia may be more prevalent in ED patients.63 Fibrats and Statins, lipid-lowering drugs, can also induce ED as a side effect,64 which was not observed in our sample.

Similarly, this study did not prove an association between BMI and ED, with patients in BMI interval 25–29 reporting more symptoms than those included in higher BMI categories. This is probably related to the fact that the BMI interval 25–29 included older patients (43% patients between 56 and 65 years and 50% of patients older than 65 years), compared to the other categories (data not shown in tables). In our sample, aging is an important confounding factor when analyzing BMI data.

Concerning smoking habits, only ex-smokers showed a statistically significant association with ED, probably due to the fact that this group reported having had heavier smoking habits (34 pack-years), compared to current smokers (25 pack-years). In a recent cross-sectional study65 smoking habits of more than 23 years or 20 cigarettes per day were significantly associated with ED.

Regarding chronic medication, an association was found between ACEI/ARB and ED. In fact, anti-hypertensive therapy is associated with adverse sexual effects, most commonly involving diuretics and beta-blockers.66 ACEI/ARB were the most frequently used drugs and its association with ED may be due to hypertension itself.

This study has some strong points in its favor. We used a validated instrument to diagnose ED in clinical settings, which was designed by urologists, and so added a multidisciplinary approach to OSAS. Patients with co-morbidities were not excluded, which made it possible to measure the impact on erectile function. The effect of chronic medication was also evaluated.

However, this study also has limitations. Although it was a representative population for estimating prevalence, we analyzed multiple clinical variables, some of them shared by our small number of subjects, which may have prevented us from establishing more statistical significant associations. The size of our sample did not lend itself to such extended analysis. The study was small because there was a high refusal rate for patient participation (54%), possibly related to the embarrassment about describing such an intimate subject.

For future research larger scale studies are needed to be able to evaluate multiple clinical variables. Among the extended co-morbidities inherent to OSAS it is extremely important not to neglect complications such as ED. If we can use a bio-psychosocial approach with our patients, it will certainly have greater impact.

ED is very common among patients with OSAS and may be under recognized because patients do not spontaneously report the problem. Several factors are associated with ED among patients with OSAS and those include age, diabetes, past smoking habits, hypertension and ACEI/ARB therapy.

The authors have no conflicts of interest to declare.