Introduction

Psoriasis is an immune-mediated dermatosis that is typically characterized by widespread erythematous plaques covered with silvery white scales involving the entire body surface [1]. Psoriasis has been gradually considered as a major global health problem. Epidemiological investigators reported the prevalence of psoriasis in adults varying from 0.91% (United States) up to 8.5% (Norway) [2]. It is suggested that psoriasis is a kind of psychophysiological skin disorder which may affect patients far beyond the skin. Higher risk of developing comorbidities, i.e., cardiovascular disorders (CVD) [3], metabolic syndrome [4], non-alcoholic fatty liver disease [5], Crohn’s disease [6], and cancer [7] was observed among patients with psoriasis than the general population. Moreover, psoriasis is associated with several psychiatric comorbidities, these include, but are not limited to depression, anxiety, poor self-esteem, and even suicidal ideation [8,9,10,11]. All these physical and psychological disorders may have a negative effect on the sufferers’ sexual lives.

A recent epidemiological study described that psoriasis was associated with an increased risk of sexual dysfunction in both female and male subjects (odds ratio = 5.5, 95% confidence interval: 2.6–11.3) [12]. In female patients with psoriasis, Maaty et al.[13]. found that the satisfaction, arousal, and desire significantly decreased with increasing severity of disease, they also observed that patients with genital psoriasis have worse sexual dysfunction than those without genital lesions.

In men patients with psoriasis, the most common complaint among them is erectile difficulties. Based on a large sample case-control study, it has been reported that patients with erectile dysfunction (ED) had a 3.85-fold higher odds in psoriasis subjects than controls [14]. Over the past several years, an increasing number of studies had addressed the relationship between psoriasis and ED [15,16,17]. However, reliable results have been inconsistent with this association. In Goulding et al.’s study [18], though the higher prevalence of ED was observed in men with psoriasis than the controls (58 vs. 49%), the authors concluded that psoriasis per se was not an independent risk factor for ED after conducting a multivariable logistic regression model.

Though a trend toward a higher prevalence of ED in patients with psoriasis was found, the evidence for the potential association is still controversial and the well-established information was still limited. Thus, we conducted a meta-analysis in an attempt to explore whether psoriasis was a risk factor for ED.

Methods

This meta-analysis was in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The PRISMA checklist was shown in Supplementary Table 1. The protocol for this meta-analysis is available from PROSPERO (CRD42018093025; http://www.crd.york.ac.uk/PROSPERO).

Data sources and searches

MEDLINE (PubMed), EMBASE, and the Cochrane Library were systematically searched from the inception dates of the databases to 7 March 2018. The search was restricted to human participants and the English language. The subject headings and the text keywords were used for the search. We used the search strategy for MEDLINE including the MeSH and text words was (((((((“Erectile Dysfunction”[Mesh]) OR sexual dysfunction, physiological) OR sexual dysfunctions, psychological) OR sexual dysfunction) OR sexual function) OR impotence)) AND ((((((“Psoriasis”[Mesh]) OR psoriasis) OR (pustulosis of palms and soles)) OR pustulosis palmaris et plantaris) OR palmoplantaris pustulosis) OR (pustular psoriasis of palms and soles)). Literature search, study selection, quality assessment, and data extraction were performed by two authors (S.K Z. and J.M W.) independently. Any disagreements were resolved by consensus or consultation with a third author (ZGZ).

Measurement of psoriasis and ED

Definitions of psoriasis and ED were in line with the international classification of diseases codes. Patients with psoriasis were diagnosed based on clinical symptoms, dermatological characteristics, and the clinical diagnosing criteria of psoriasis. ED (also known as impotence) was defined as the recurrent or persistent inability to attain or maintain an erection sufficient for satisfactory sexual performance [19]. We included the studies in which ED was measured by using any of the validated methods, including questionnaires, physical examination, and clinical assessment of erectile function, i.e., the 5-item International Index of Erectile Dysfunction (IIEF-5) questionnaire, the International Index of Erectile Function questionnaire (IIEF), and the pertinent history and physical examination.

Study selection

Any available epidemiologic evidence reported the association between psoriasis and risk of ED were included in the presented study. Participants were limited to the broad-spectrum population diagnosed with psoriasis and ED. According to the patient, intervention, comparison, outcome and study design (PICOS), the question that guided this meta-analysis was: Does psoriasis increase the risk of ED? The PICOS evidence was composed of the following combinations: an adult patient with ED or impotence (P); a history of psoriasis (I); compared with the general or non-psoriasis population (C); the diagnosis of ED or impotence (O); all study designs were accepted (S). In addition, those studies on the pertinent subjects which provided relative risk, hazard ratio, and odds ratios (OR) with 95% confidence intervals (CI) or adequate data to enable calculation of these efficiency values were also included. The exclusion criteria included the following: (1) female subjects; (2) the control population data were not available; (3) review articles, comments, meeting abstracts, editorials, letters, and case reports, etc.; (4) duplicated data; (5) animal experiments; (6) studies with <10 participants; (7) previous publications of the same clinical trial.

Data extraction

A standardized data collection form was conducted to extract the following pertinent information: the first authors’ names, year of publication, country of origin, study design, mean disease duration, the demographic and age of case and the control sample, methods of psoriasis and ED ascertainment, the OR with CI, and variable adjustment for confounding factors.

Quality assessment

The methodological quality evaluation of the cross-sectional study was based on cross-sectional study quality methodology checklist (low quality = 0–3, moderate quality = 4–7, high quality = 8–11). The quality of case-control or cohort study was in line with the Newcastle–Ottawa Scale (NOS, low quality = 0–3, moderate quality = 4–6, high quality = 7–9). The grading of recommendations assessment, development, and evaluation (GRADE) approach was conducted to exert the absolute estimates of the risk of ED in men with psoriasis, assess and rank the overall quality of the evidence.

Statistical analyses

The strength of association between psoriasis and the risk of ED in the included studies was estimated using OR and its 95% CI. Results with a two-tail P value < 0.05 were regarded as statistically significant. The heterogeneity of included studies was evaluated by using the I2 statistic and the Cochrane Q statistic (I2 > 50% was considered of substantial heterogeneity; a P value of Q test < 0.10 was considered statistically significant). On account of the high likelihood of between-study variance for differences in study design and study population, a random effects model rather than a fixed effects model was employed. Subgroup analyses were utilized to further investigate the origin of heterogeneity. Sensitivity analyses were applied to detect the potential source of heterogeneity by omitting studies one by one and evaluating the resulting effect. Publication bias analysis was assessed by the Begg’s and Egger’s test. The current statistical analysis was performed using the Stata (version 13.0, Stata Corp LP, College Station, Texas, USA).

Results

Literature search

Through a broad selection of the databases, a total of 297 relevant articles were identified in the initial search (97 articles from MEDLINE, 115 articles from EMBASE, and 85 articles from the Cochrane Library). Of these articles, underwent the title and abstract review, 245 were excluded after removing duplicates, and irrelevant articles, leaving 52 pertinent articles for further full-text review. Among them, 8 studies were excluded due to the control group was not available; 15 for failing to meet the inclusion criteria; 10 for inappropriate grouping; and 11 for insufficient outcome data. Lists of full-text excluded studies with reasons for exclusion are presented in Supplementary Table 2. Finally, 8 observational studies [18, 20,21,22,23,24,25,26] (6 cross-sectional studies [18, 21,22,23,24, 26], 1 cohort study [20] and 1 case-control study [25]) met the pre-defined eligibility criteria were included in this study. The selection process was illustrated in Fig. 1.

Fig. 1
figure 1

Flow chart of study selection

Full size image

Study characteristic

The 8 studies which published between 2010 and 2017 years have been enrolled a total of 1,829,449 participants, 39,490 of whom were psoriasis patients while the remaining 1,789,959 individuals were healthy controls. The mean age of the individuals among studies ranged from 44 years to 56.3 years. Six studies [18, 21,22,23, 25, 26] were conducted in Europe and 2 studies in Asia [20, 24]. The mean disease duration of psoriasis ranged from 6 months to 19.9 years. And 6 studies [18, 20, 21, 24,25,26] reported the adjusted OR with 95%CI. The detail characteristics of the 8 studies were summarized in Table 1.

Table 1 Characteristics of the included studies
Full size table

Study quality and overall quality of the evidence

The results of the methodological quality assessment of the 6 cross-sectional studies were listed in Supplementary Table 3, 3 studies [23, 24, 26] were judged of high quality and 3 studies [18, 21, 22] were considered with moderate quality. Using the NOS, 1 cohort study [20] was assessed to be high quality and 1 case-control study [25] was evaluated to be medium quality (Supplementary Table 4). The GRADE-relevant outcomes revealed that the rate of events of ED on average in patients with psoriasis was 4037/39490 (10.2%), whereas the control was 151710/1789959 (8.5%); the absolute effect of psoriasis on ED was 46 more per 1000 (from 28 more to 66 more); and the overall quality of the evidence was judged as moderate (Table 2).

Table 2 GRADEprofiler summary of evidence for the effects of psoriasis on erectile dysfunction
Full size table

Synthesis of results

As shown in Fig. 2, synthesis of results from 8 studies showed that the OR of psoriasis in patients with ED versus comparators was 1.62 (95%CI: 1.37–1.91, P < 0.001; heterogeneity: I2 = 62.6%, P = 0.009). The outcomes were consistent when restriction to multivariable adjustment (6 studies; combined OR = 1.5, 95%CI: 1.31–1.72, P < 0.001; heterogeneity: I2 = 53.5%, P = 0.056) (Fig. 3).

Fig. 2
figure 2

Forest plots of meta-analysis of the eight included studies on the association between psoriasis and erectile dysfunction

Full size image
Fig. 3
figure 3

Forest plots of meta-analysis of the six included studies with multivariable adjustment

Full size image

Subgroup analyses

To further elicit the relationship between psoriasis and the risk of ED and detect the source heterogeneity, subgroup analyses were performed based on mean disease duration of psoriasis, sample size, geographical area, publication year, and study design (Table 3). On account of the little variance in mean age among studies (ranged from 44 years to 56.3 years), thus we did not perform the subgroup analysis on age.

Table 3 Subgroup analysis of the association between psoriasis and erectile dysfunction
Full size table

Stratified analysis by the mean disease duration of psoriasis revealed that duration ≥ 10 years [18, 21, 22, 24] exhibited an increased odds of ED compared with non-psoriasis (OR = 1.76, 95% CI: 1.35–2.29, P < 0.001) and no substantial heterogeneity was detected; whereas when limited to < 10 years [20, 23], the results did not show this association (OR = 1.99, 95% CI: 0.74–5.34, P = 0.171).

Subgroup analysis by the sample size revealed that studies with the total participants < 300 [18, 22, 23, 25] (OR = 2.36, 95% CI: 1.45–3.85) had a higher risk of ED compared with >300 participants [20, 21, 24, 26] (OR = 1.49, 95% CI: 1.28–1.73). There was no significant heterogeneity in studies of <300 participants (I2 = 43.9%, P = 0.148), while substantial heterogeneity was observed in studies > 300 participants (I2 = 67.4%, P = 0.027).

Subgroup analysis based on the geographical area showed that the association between psoriasis and ED was more stronger in studies conducted in Europe region [18, 21,22,23, 25, 26] (RR = 1.94, 95% CI: 1.46–2.59, heterogeneity: I2 = 53.7%, P = 0.056) than those studies conducted in Asia region [20, 24] (OR = 1.31, 95% CI: 1.14–1.5, heterogeneity: I2 = 4.5%, P = 0.306). On the other hand, a stronger association was detected in the publication year during 2016–2017 [23,24,25,26] (OR = 1.84, 95%CI: 1.36–2.49 than those during 2010–2014 [18, 20,21,22] (OR = 1.61, 95%CI: 1.14–2.28). Stratified analysis by the study design revealed that a higher risk of ED among patients with psoriasis was observed (OR = 1.78, 95% CI, 1.42–2.24) in cross-sectional studies, while this association was not presented for cohort/case-control studies (OR = 1.57, 95%CI: 0.87–2.83).

Sensitivity analysis

To confirm the robustness of the synthetic effect estimate and assess the influence of individual study on the overall risk of ED, a sensitivity analysis was performed. After omitting any of the studies, there was no substantial change on the new overall pooled OR, which ranged from 1.53 (95%CI: 1.32–1.77) to 1.84 (95%CI: 1.36–2.47) (Table 4 and Supplementary Figure 1). On the other hand, similar heterogeneity was obtained after each study exclusion, the I2 ranged from 44.5 to 67.9%. These outcomes indicated that no single study dominated the overall pooled OR and heterogeneity.

Table 4 Sensitivity analysis after each study was excluded by turns
Full size table

Publication bias

Visualization of the funnel plot indicated that both Begg’s rank correlation test and Egger’s linear regression yielded non-significant publication bias among the included studies (Begg’s, P > |z| = 0.386; Egger, P > |t| = 1.04, 95%CI: −0.97–2.41) (Supplementary Figure 2).

Discussion

Over the past few decades, a growing number of epidemiological studies have assessed the association between psoriasis and risk of ED but provided inconsistent results. Based on all of the available epidemiological evidence related to this topic, the current meta-analysis indicated a 1.62-fold increased risk of ED among subjects with psoriasis compared with those without psoriasis. Similarly, such an association between psoriasis and high risk of ED was consistent when conducted the multivariable adjustment analysis for confounding factors (OR = 1.5, 95%CI: 1.31–1.72). Ascertained by GRADEpro, the rate of events of ED on average in patients with psoriasis and without psoriasis was 10.2 and 8.5%, respectively; the absolute effect was 46 more per 1000. These results were in line with a recent meta-analysis reported by Wu et al. [27] and some other relevant studies which did not meet our inclusion criteria [8, 14,15,16,17, 28]. We also performed subgroup analyses to assess the degree to which potential confounders might have influenced these findings. Sensitivity analyses indicated that the quantification of the risk for the ED remained significantly higher in nearly all of the studies, suggesting that our findings in the present meta-analysis were robust. However, substantial heterogeneity has also appeared in the current study. The different mean disease duration of psoriasis, sample size, geographical area, study design, and varied characteristics of subjects could all be partly responsible for the heterogeneity. Of note, both Wu et al.’s study [27] and our study found a potential relationship between psoriasis and ED, but our study have a more restricted inclusion criteria (9 included studies in Wu et al.’ study and 8 included studies in the present study) and have been conducted more subgroup analyses (i.e., mean disease duration, sample size, geographical area, publication year, and study design) to further detect the heterogeneity, all of which might make the elaborate explanations on this topic.

ED, a multifactorial condition, can be affected by biological, psychological and interpersonal determinants [29]. The development of ED can be attributed to vascular, neuronal, hormonal, and metabolic factors, ultimately caused endothelial and smooth-muscle dysfunction [30]. A variety of independent risk factors have been confirmed for ED development, i.e., aging, cardiovascular disease, diabetes mellitus, dyslipidemia, and depression [30]. Nevertheless, some other potential risk factors have gradually been recognized to increase susceptibility to ED, such as some certain skin diseases [31, 32].

There has long been a recognition that skin disorder may have a negative effect on sexual aspects of the patient’s life [33]. Psoriasis is a common skin disorder with visible disfiguration, which is associated with a physical burden as well as a psychological burden, and ultimately has a deleterious effect on the sufferer’s sexual health. Psoriasis Area Severity Index (PASI) has been used to evaluate the severity of psoriasis, which was obtained by weighting the erythema, desquamation, induration, infiltration or thickness in the body surface area [34]. IIEF, a multidimensional scale for assessment of erectile dysfunction, is a widely used validation questionnaire for assessing the erectile function domains [35]. It was reported that IIEF score had a negative correlation with PASI scores in clinical practice [22]. In line with this finding, Egeberg et al.[26]. found that ED was significantly increased in men with psoriasis, the adjusted OR was slightly increased in severe psoriasis group than mild psoriasis group (1.13 versus 1.17). However, in a more recent study developed by Bardazzi et al.[23]., they found that a higher prevalence of ED was observed in patients with mild psoriasis compared to those patients with severe psoriasis (56.67 versus 46.68%).

Different comorbidities and treatments of the disease rather than psoriasis can modulate the ED risk [36]. In the present study, 6 studies provided the OR of multivariable adjustment for confounding factors, i.e., age, hypertension, systemic treatment, obesity (body mass index), diabetes mellitus, dyslipidemia, depression, anxiety, smoking, alcohol, genital lesions, and drug usage, etc. Moreover, these 6 studies also have the strict exclusion criteria in both the study group and control group. Therefore, the synthesis effect estimate in the current study seems reliable and robust. Though the overall OR adjusted for known potential causes for ED decreased slightly compared with the complete analysis (1.5 versus 1.62), our study indicated that patients with psoriasis were associated with a 50% increase in ED risk than non-psoriasis after adequate confounder adjustment.

In the present study, subgroup analysis by mean disease duration revealed that duration ≥ 10 years had been confirmed the relationship between psoriasis and ED, while no positive association was found in duration < 10 years. However, Cabete et al.[21]. observed an increased prevalence of ED in patients with psoriasis especially in those had a severe symptom of psoriasis, but they found that psoriasis duration did not statistically contribute to ED in these subjects. Stratified analysis by sample size revealed that more significant association was found in participants < 300 groups compared with >300 participants groups. It is possible that studies with small sample sizes are prone to generate the detection bias, selection bias, and performance bias. Another explanation could be the quality of the included studies with a small sample were lower than those with a large sample. Therefore, additional well-designed studies are still warranted to validate this finding. In the current study, we also observed that studies conducted in Europe region have a stronger association between psoriasis and ED than Asia region. This might probably due to the high prevalence of psoriasis was detected in the Europe countries. It was reported that the prevalence of psoriasis in adults was estimated as 3.73, 8.5, 3.1, and 5.2% in Denmark [37], Norway [38], Italy [39], and France [40], respectively; while in China [41] and Chinese Taipei [42] was 0.2–1.5% and 0.23%, respectively. It was also reported that the prevalence of psoriasis among the Caucasian population and Mongoloid population was 1–2% and 0.3%, respectively [20].

Though our meta-analysis has linked psoriasis to ED, no clear-cut pathogenic and etiological mechanism has been identified to elucidate this potential association. It was suggested that the etiology of ED in psoriasis patients may be multifactorial, including physical, organic and/or psychogenic factors [20, 43]. The connecting link between psoriasis and CVD may be a conceivable contributor [18]. Accumulating evidence has emerged demonstrating that psoriasis is an independent risk factor for the development of CVD [36]. Compared to the general population, patients with psoriasis were reported that have a higher risk of atherosclerosis and hypertension, both of which were recognized as the risk factors affecting ED [18, 21]. Psoriasis has been speculated to be associated with systemic chronic vascular inflammation and endothelial dysfunction [44]. It is known that endothelial dysfunction is the shared pathogenic mechanism between ED and CVD [45]. Endothelial dysfunction can reduce the bioavailability of nitric oxide and ultimately lead to ED [46]. On the other hand, patients with psoriasis had increased susceptibility to metabolic syndrome compared with the general population [47]. Intriguingly, an association between metabolic syndrome and ED has been confirmed in some studies [48, 49]. The relationship network has been linked metabolic syndrome to endothelial dysfunction, nitric oxide dysfunction, as well as vascular insufficiency, and that all of these pathogenic conditions might be involved in the development of psoriasis and ED [50]. Yet in several studies reported the mechanism underlying linkage between ED and psoriasis might contribute to systemic inflammation. Patients with psoriasis likely have a high level of pro-inflammatory cytokines, including leukotrienes, reactive oxygen species, TNF-α, IFN-γ, and Th-1, which may subsequently contribute to endothelial dysfunction—central to the pathogenesis of ED [22, 51,52,53].

Extensive analyses were carried out that psoriasis may be not per se a risk factor for ED, but the concomitant psychological distress [20, 54]. Psoriasis, especially in the presence of genital lesions, can cause an alteration in body image thus cause feelings of depression, anxiety, low self-esteem, stigmatization, and decreased confidence; all these psychological impairments could lead to ED [25, 43, 55]. On the other hand, there are reports [56, 57] revealing that some treatments prescribed for psoriasis might also attribute to the development of ED, such as methotrexate, retinoids, etc. It has been speculated that methotrexate could disrupt the balance of estrogen–testosterone, leading to ED [58]. However, some systemic anti-psoriatic therapies can have beneficial effects on the erectile function of men with psoriasis, i.e., anti-TNF-α, interleukin-23 inhibitors, etc[59,60,61]. Since the pathogenesis of ED in psoriatic patients attributed to various factors, thus a better alternative management for those patients should employ a combined approach, such as medical, psychological, and behavioral interventions together.

This systematic review and meta-analysis have been summarized all available evidence to confirm the association between psoriasis and the risks of developing ED. Nevertheless, we acknowledged there was an inherent limitation in the current study. As with any meta-analysis, substantial heterogeneity across included studies was detected. Subsequently, subgroup analysis shown that no substantial heterogeneity was observed in the subgroup of ≥ 10 years mean disease duration, sample size < 300 participants, study those studies conducted in Asia, and cross-sectional study design. Furthermore, no significant heterogeneity was detected when eliminated the study of Chen et al. during sensitivity analyses. As a result, high-quality well-designed cohort studies with large sample size are still warranted to validate the evidence of psoriasis predisposing to the development of ED.

Conclusions

In summary, our meta-analysis indicates a potential hazardous effect of psoriasis for developing ED. After multivariable adjustment for the confounding factors, psoriasis and ED was still closely associated, which reminds both dermatologists and urologists should realize the relationship between psoriasis and ED and provide specific treatments for those patients when necessary.