The SIAMS nominated an expert task force to provide an updated guideline on the diagnosis and management of ED. The project was actively supported by several other National Societies including the Italian Society of Endocrinology (Società Italiana di Endocrinologia, SIE), the Association of Clinical Endocrinologists (Associazione Medici Endocrinologi, AME), the Italian Society of Diabetology (Società Italiana di Diabetologia, SID), the Association of Clinical Diabetologists (Associazione Medici Diabetologi, AMD), the Italian Society of Internal Medicine (Società Italiana di Medicina Interna, SIMI), The Italian Society of Obesity (Società Italiana dell’Obesità, SIO), The Italian Society of Metabolism, Diabetes and Obesity (Società Italiana Metabolismo, Diabete, Obesità, SIMDO), the Italian National Association of Hospital Clinical Cardiologists (Associazione Nazionale Medici Cardiologi Ospedalieri, ANMCO), the Italian Society of Interventional Cardiology (SICI-GISE), and the Italian Society of Psychopathology (Società Italiana di Psicopatologia, SOPSI). A representative member of each Society has signed these Guidelines. According to the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) system [13], here in a series of consensus recommendations have been provided after having widely discussed the best evidence available in PubMed. The GRADE system allows the quality of evidence and the strength of recommendations to be rated, adding value to the clinical advice by using a consistent language and graphical descriptions for standardizing the grading [13]. Concerning the strength of recommendations, the number 1 indicates a strong recommendation and is associated with the terminology “we recommend”; the number 2 denotes a weak recommendation and is associated with the wording “we suggest”. The four-level grading of the quality of evidence employs the following graphical descriptions: ⊕ “◯◯◯very low-quality evidence”, ⊕  ⊕ “lo◯◯w quality”, ⊕  ⊕  ⊕ “m◯oderate quality”, and ⊕  ⊕  ⊕  ⊕ “high quality”. Finally, statements, recommendations, or suggestions not supported by direct evidence, but potentially useful for the clinical practice, are marked as “expert opinion” whereas those with high-quality evidence for every day clinical activity as “Good clinical practice”.

According to the SIAMS work-up and agreement, these Guidelines come from the work of a team of experts on the topic coordinated by the senior author and two members of the SIAMS Guideline Committee. After the revision by the SIAMS Executive Committee and by the Directors of all SIAMS Excellence Centers, these guidelines have then been announced by email and published for two weeks on the SIAMS’s website, www.siams.info, so that all SIAMS Members could provide further comments and suggest additional minor revisions. Following this last step, the present guidelines have been submitted to the Journal of Endocrinological Investigation for the normal process of international peer reviewing.

Epidemiology and risk factors

Evidence supporting that a specific ED case is due to a single and unique etiological factor is scarce and not supported by common clinical practice. For this reason, in defining the well-known pathophysiological mechanisms related to ED, we discourage the use of the term “etiologies”. The expression to be “Risk Factors” is preferred, also considering that there are frequently multiple factors in the same patient (see also Table 1).

Table 1 Main risk factors associated with erectile dysfunctionEpidemiology

Recommendation #1. We recommend considering erectile dysfunction as a common male disorder whose incidence and prevalence are strongly associated with age and health status (1 ØØØØ).

Evidence

ED is a common medical disorder in men whose incidence and prevalence increases with age and is associated with poorer general health and presence of comorbidities (see below) [14]. The crude incidence of ED varies widely among studies and, based on the age of the studied cohorts, ranges from 4 to 66 cases per 1000 men/year [14,15,16,17]. Similarly, the prevalence of ED varies based on the age of the enrolled subjects [14]. The Massachusetts Male Ageing Study (MMAS) showed a combined prevalence of mild to moderate ED of 52% in men aged 40–70 years, and ED was strongly related to age, health status and emotional factors [18]. Conversely, the European Male Ageing Study, the largest European multi-center population-based study of ageing men (40–79 years), reported a prevalence of ED ranging from 6 to 64%, depending on different age subgroups and increasing with age, with an average prevalence of 30% [19]. The prevalence of ED seems higher in the United States and Eastern and South-Eastern Asian countries, compared to Europe or South America [20]. The prevalence of ED varies also according to ethnicity. Hispanic men had increased odds of moderate-severe ED, whereas Black men were less likely to report moderate-to-severe ED. The prevalence of ED among different racial and ethnic groups is likely the result of complex phenomena and depends upon the interplay of socioeconomic, demographic, cultural, and lifestyle characteristics [14]. Despite few data on the epidemiology of ED in men younger than 40 years, a very recent study reported a prevalence of 11.3% of mild ED and 2.9% of moderate-to-severe ED in a sample of 2660 sexually active men in the age range 18–31 years [21].

Remarks

We highly value considering ED as a marker of general male health status and as an expression and part of many comorbidities, rather than a mere manifestation of ageing. Studies involving male populations older than 70 years are limited, resulting, therefore, in an underestimation of the problem. Similarly, ED has scarcely been evaluated in men younger than 40 years, generating the false assumption that this symptom is almost unheard of in young people [14].

Risk factors

Several systemic conditions as well as organic, relational and intrapsychic factors can contribute to the development of ED (Table 1). Moreover, ED is frequently comorbid with other sexual dysfunctions, either in the patients and/or in the partner, which may amplify the erectile failure in subclinical forms (see below) into an overt ED.

Systemic risk factors

Recommendation #2. We recommend investigating sexual function and ruling out erectile dysfunction in all patients with systemic diseases especially in those with organ failure (Good clinical practice).

Evidence and remarks

Systemic diseases characterized by organ failure (such as end-stage renal diseases [22] and liver cirrhosis[23], as well as those conditions related to systemic inflammation such as HIV infection [24, 25] or to some rheumatic and/or autoimmune diseases [26,27,28,29]) are frequently associated with ED, couple impairment and reduced quality of life. Similarly, considerations can be drawn for some malignancies with long life expectancy such as lymphomas and testis cancers [30]. Finally, emerging evidence has documented that ED can be considered a long-term consequence of coronavirus disease 2019 (COVID-19) infection [31,32,33]. In all the aforementioned problems, ED is the result of the interaction between multiple risk factors which will be analyzed in detail in the following sections. However, we want to emphasize that frequently sexual function and ED, in particular, are often poorly investigated, if not ignored, in the vast majority of cases. Some evidence indicates that an improvement of sexual function can ameliorate couple intimacy quality, and quality of life, as well as diseases-related distress [22, 23, 28, 30].

Cardiovascular and respiratory risk factorsArterial hypertension

Recommendation #3. We recommend investigating erectile dysfunction in individuals with arterial hypertension since it is strongly associated with hypertension duration and severity (1ØØØØ), and it might be related to the use of some anti-hypertensive medications (1ØOOO).

Evidence

Around 30–45% of adults are affected by arterial hypertension, and the prevalence increases with age, affecting more than 60% in people aged > 60 years [34]. The association between arterial hypertension and ED is very frequent and up to seven times higher when compared to that observed in the general population [35]. Moreover, in hypertensive patients, ED is usually more severe than in normotensive people [35]. Considering this close association, information on ED must be regularly collected in all hypertensive patients. ED is effectively treated by phosphodiesterase-5 inhibitors (PDE5i), also in hypertensive subjects, but this class of medication is absolutely contraindicated if concomitantly administering nitrates, and only relatively contraindicated in treatment with alpha-blockers [36].

Remarks

It should be recognized that several classes of anti-hypertensive drugs have been associated with a higher risk of ED, although only few placebo randomized clinical trials (RCTs) have correctly investigated the role of these medications in erectile function (see below) [37].

Cardiovascular diseases

Recommendation #4. We recommend checking for symptoms of coronary artery disease in all patients with erectile dysfunction at each visit and evaluating the cardiovascular risk profile using cardiovascular algorithms such as SCORE2 or SCORE2-OP (Systematic Coronary Risk Estimation 2 and Systematic Coronary Risk Estimation 2-Older Persons) (1 ØØØØ).

Evidence

Much evidence has clearly demonstrated that ED is frequently comorbid in patients with cardiovascular diseases (CVD) since the two conditions often share the same risk factors [6]. In addition, data from longitudinal studies performed either in symptomatic patients or in the general population have shown that ED could represent an early sign of forthcoming Major Adverse Cardiovascular Events (MACE) [6]. Hence, the presence of sexual symptoms must be ruled out in all patients with documented CVD and subjects with ED should be adequately investigated to stratify their cardiovascular (CV) risk since the appearance of symptoms allows for a valuable time window [2–5 years] for earlier modification of associated risk factors and potentially an improvement in outcomes [6].

Several risk engines have been developed for the quantification of CV risk profile. The Framingham risk engine represents the most used risk prediction equation, originally developed in the United States. However, its application in European populations showed an overestimation of CV risk profile [38]. Conversely, Systematic Coronary Risk Estimation 2 and the Systematic Coronary Risk Estimation 2-Older Persons (SCORE) chart [39] (https://www.escardio.org/Education/Practice-Tools/CVD-prevention-toolbox/SCORE-Risk-Charts) have demonstrated better utility in European subjects, including Italians [40]. The SCORE2 and SCORE2-OP charts estimate an individual’s 10-year risk of fatal and non-fatal CVD events (myocardial infarction, stroke) in apparently healthy people aged 40–69 years and ≥ 70 years with risk factors that are untreated or have been stable for several years. Lifetime benefit estimation of risk factor treatment, e.g. with the LIFE-CVD (LIFEtime. Perspective CardioVascular Disease) lifetime model, can facilitate the communication of treatment benefits [40].

Remarks

The SCORE algorithm should not be used for persons with established atherosclerotic CVD, diabetes mellitus (DM), chronic kidney diseases, and familial hypercholesterolemia, categories associated with moderate to very high CV disease risk [39]. It is important to recognize that not only traditional CV risk factors, but also unconventional components related to relational and intrapsychic determinants can contribute to the CV risk stratification of ED subjects [41,42,43]. The identification and the early recognition of these comorbid factors can allow a specific prevention program based on lifestyle modifications and optimization of the concomitant chronic diseases to be promoted, which might help not only in improving ED treatment outcomes but also in ameliorating couple fitness and decreasing the “residual CV risk” thus improving CV risk prevention (see below).

Chronic obstructive pulmonary disease and sleep apnea

Recommendation #5. We suggest investigating erectile dysfunction in all patients with chronic obstructive pulmonary disease and obstructive sleep apnea (2 ØØØΟ).

Evidence and remarks

Chronic obstructive pulmonary disease (COPD) is a frequent condition resulting from long-term exposure to harmful particles or gases (mainly smoking) which has been reported as the third leading cause of death globally [44]. ED is a frequent complication of COPD with a reported prevalence ranging from 72 to 87% of cases [44]. A recent meta-analysis including 31 studies, for a total of 1187 patients with COPD, and 224 age-matched, non-COPD controls, showed a pooled prevalence of ED in 74% (95% CI: 68–80%) of patients with COPD, compared to 56% (37–73%) observed in controls [45]. Similarly, another meta-analysis involving 58,307 subjects documented an up to threefold increased risk of moderate-to-severe ED in subjects with COPD when compared to controls [46]. Obstructive sleep apnea (OSA) is another common respiratory problem frequently associated with ED [47]. In the latter condition, ED prevalence increases as a function of age and body mass index (BMI) but it is also closely related to the apnea–hypopnea index which reflects OSA severity. A combination of vascular and hormonal (reduced testosterone, T, levels) factors represents the main pathogenetic determinants in both COPD and OSA, although the chronic reduced systemic oxygenation, as well as relational and intrapsychic factors, can contribute to the development and to the maintaining of ED in these subjects. Despite the aforementioned evidence, ED and sexual function is poorly investigated in COPD and OSA. Accordingly, although the Global Strategy for the Diagnosis, Management, and Prevention of Chronic Obstructive Lung Disease (GOLD) guidelines recommend the assessment of the state of sexual activity in COPD men, they do not contain any information regarding how to evaluate or manage the problem [44].

Metabolic risk factorsObesity

Recommendation #6. We recommend investigating sexual function in all male patients with obesity (1 ØØØØ).

Evidence

A large body of evidence has shown that obesity is a major independent risk factor for ED [6, 48,49,50] and that the incidence rate of ED increases proportionally to the degree of obesity [51, 52]. Among patients with ED, obesity has also been reported as a further risk factor for CVD [53]. It is conceivable that the higher prevalence of ED in obese subjects may be related to hypogonadism [54]; however, other pathophysiological mechanisms are also involved, including endothelial dysfunction, insulin resistance, psychological factors and physical inactivity [55]. Several RCTs on obese people have shown that lifestyle changes aimed at weight loss significantly improved erectile function [48, 49, 51], further supporting a cause-effect relationship between obesity and ED. This was also confirmed by prospective cohort studies of obese men undergoing bariatric surgery who significantly improved their T levels and sexual function upon the intervention [56].

Diabetes mellitus

Recommendation #7. We recommend investigating erectile dysfunction in all patients with diabetes mellitus since it is strongly associated with diabetes duration, metabolic control and coexistence of other diabetic complications (1 ØØØØ).

Evidence

ED prevalence is markedly increased in men with DM, but with a huge variability mainly due to the clinical features of the studied populations. According to a meta-analysis, overall prevalence is 52.5% with a clear difference between type 1 (37.5%) and type 2 (66.3%) DM; compared to healthy controls, people with diabetes had a 3.62 odds ratio of having ED [57]. Also incidence is 2–3 folds higher than in general age-matched populations [58]. Even when the same diagnostic tool is used, (i.e. International Index of Erectile Function; IIEF), prevalence can range from 35 to 90% [59]. Age, duration of the disease, metabolic control and concomitant chronic DM complications seem to be the major factors influencing prevalence and severity of ED in DM [59,60,61]. In the first large cross-sectional study performed in Italy on about 10,000 subjects, ED prevalence, assessed by means of an interview, was 35.8% and it was clearly influenced by age, diabetes duration, metabolic control, other chronic complications, and smoking [58]. More recently, in a population with newly diagnosed type 2 DM (TDM2), overall ED prevalence was 60% and its severity, assessed by means of IIEF-5, was associated with age, HbA1c levels, arterial hypertension, dyslipidemia and depression [60]. Due to the delay between TDM2 onset and its diagnosis, ED prevalence is high (43.3%) also in newly diagnosed subjects [62] and it is associated with hypogonadism, depressive symptoms and CV risk [60].

Remarks

Despite available evidence, ED in DM is still underestimated in routine clinical practice and consequently under-treated. This is even more important when considering that, in addition to sexual dysfunction, ED is associated with other conditions such as depression and CVD, which, all together, have a deeply negative impact on quality of life and life expectancy in male DM subjects [63].

Dyslipidemia and gout

Recommendation #8. We recommend investigating erectile dysfunction in all patients with dyslipidemia (1 ØØØØ) and gout (1 ØOOO).

Dyslipidemia

It is well known that dyslipidemia is clearly associated with MACE, [64]. Considering that ED is another well-known risk factor for MACE [6], associations between the two conditions have been thoroughly investigated.

Evidence

Thanks to the seminal study of Feldman et al. [18] in a cohort of 1290 men enrolled in the MMAS, an association between ED and dyslipidemia has been evident since the last decade of the past century. In fact, in the cross-sectional analysis of the MMAS cohort, the probability of ED ranged from 7 to 25% when high density lipoprotein cholesterol (HDL-C) decreased from 90 to 30 mg/dL [18]. In the same years, Wei et al. [65] published longitudinal results from a cohort of 3,250 Texan men without ED at study entry. After a mean of 22 months of follow-up, a 39 mg/dL increase in total cholesterol (TC) or in HDL-C was associated with a change of 1.32 [1.04–1.668] and 0.38 [0.18–0.8] times in the risk of ED, even after adjusting for other ED determinants. In particular, a TC > 240 mg/dL or a HDL-C < 30 mg/dL double the risk of ED [65]. Results were also confirmed in a recent study where high low density lipoprotein cholesterol (LDL-C)/HDL-C and low HDL-C were able to predict arteriogenic ED, as assessed by dynamic peak systolic velocity (PSV) at penile color Doppler ultrasonography (PCDU) in a small series of ED patients (n = 84) [66]. In a larger series of unselected ED patients (n = 2160), it was reported that high TC and triglyceride levels and low HDL-C were all associated with a reduced dynamic PSV [67]. However, when all these lipids were introduced in an age-adjusted multivariate model, only high TG levels retained significant association with impaired penile blood flow [67]. In a longitudinal study on a fraction of the above cohort, high triglyceride levels were also able to predict MACE [67]. Finally, in a study on patients with diabetes mellitus, HDL-C had a significant correlation with number of erectile episodes during the night, as documented by nocturnal penile tumescence studies such as Rigiscan [68]. A causal relationship between high lipids and ED is supported by intervention studies. Two meta-analyses investigating the effect of treating dyslipidemia on erectile function suggest that statin therapy is associated with a clinically meaningful increase by 3.3 points on the IIEF erectile function domain (IIEF-EFD) score [69] and that the addition of statins to sildenafil improves the response to sildenafil itself [70].

Remarks

ED is now considered a harbinger of MACE and it is now included in the algorithms of risk prediction of CV events [70]. Considering that lipid-lowering medications can substantially decrease CV risk and even ameliorate ED [69, 70], it is imperative for clinicians dealing with ED to obtain a full lipid profile in any ED patients and to treat any dyslipidemia accordingly.

Gout

The association between gout, the most common crystal arthropathy, and sexual dysfunction has often been investigated by studies in recent decades. Awareness of this association is frequently lacking and the pathogenetic mechanisms have only partially been identified.

Evidence

A recent meta-analysis [71] evaluating the risk of ED in subjects with gout indicates an overall statistically significant increased risk of ED in subjects with gout (RR = 1.2; 95%CI = 1.1–1.31). A subgroup analysis indicates that this increased risk was apparent only in longitudinal cohort studies (n = 5, RR = 1.22; 95%CI = 1.12–1.32) but not in the three cross-sectional studies investigated. Similar results were derived from a previous meta-analysis involving a lower number of studies [72]. In one of the aforementioned longitudinal studies conducted in Taiwan, it was also observed that in subjects with gout and without comorbidities treated for more than three months with urate-lowering agents, the risk of developing ED was similar to that of a control cohort without gout [73].

Remarks

Having gout there is a 20% increased risk of developing ED, even after adjusting for possible confounders, including comorbidities. Considering that hyperuricemia is the underlying condition favoring gout development and that hyperuricemia is associated with endothelial dysfunction and CVD [74], it is obvious that known or even unknown comorbidities might favor ED development [