Association between erectile dysfunction and lower urinary tract symptoms: a nationwide, population-based cross-sectional study

Lower urinary tract symptoms (LUTS) are common in men and include storage, voiding, and post-micturition symptoms that can impair quality of life. A recent Japanese nationwide survey reported an overall LUTS prevalence of approximately 80%, increasing with age.1 Erectile dysfunction (ED) is also prevalent in aging men, and LUTS and ED frequently coexist and share mechanisms, including vascular and autonomic dysfunction and psychosocial burden.2 However, nationwide data from Japan evaluating LUTS subtypes (including overactive bladder [OAB] and underactive bladder [UAB]) using a practical erection-hardness measure and multivariable adjustment remain limited. To the best of our knowledge, this is among the first population-based studies to comprehensively evaluate the association between ED and detailed LUTS subtypes, including OAB, UAB, and their coexistence, while also considering sexual activity as a potential mediator rather than a confounder. We conducted a cross-sectional analysis of the Japan Continence Survey 2023 (JaCS2023), an online nationwide survey administered in April 2023, using stratified sampling by age and region. Of 3122 male respondents, 2627 (aged 18-99 years) were included after excluding 495 who did not answer the erectile function item. No other missing responses were identified. Participation was anonymous, and ethical approval was obtained. Baseline demographic and clinical characteristics of the participants according to LUTS and ED status are summarized in Supplementary Tables S1 and S2, respectively. We also assessed frequency of sexual intercourse and self-reported awareness that LUTS affected sexual life. Erectile function was assessed with the question: “How would you evaluate the hardness of your penis when you have an erection?” Responses of “Does not get larger,” “Does not get hard,” or “Not hard enough for penetration” were classified as decreased erectile function (ED+). Responses of “Somewhat hard, but not completely” or “Completely hard” were classified as maintained erectile function (ED−). Because the survey did not include a multi-item validated instrument, ED was defined using a single erection-hardness question, which may not fully capture the standard definition of ED. Sexual intercourse frequency was categorized as none (0/month), low (1-3/month), moderate (4-8/month), or high (≥9/month). Lower urinary tract symptoms was defined as the presence of at least 1 symptom meeting prespecified frequency symptom thresholds: storage (daytime frequency ≥8/day, nocturia ≥1/night, urgency ≥1/week), voiding (interrupted stream, weak stream, straining to void; each ≥1/week), or post-micturition (feeling of incomplete emptying, post-void dribbling, post-void incontinence; each ≥1/week). Overactive bladder was defined using the Overactive Bladder Symptom Score.1 Underactive bladder was defined according to International Continence Society terminology as decreased urinary sensation, delayed initiation of voiding, and abdominal straining to void, each occurring at least once weekly.4 Coexisting OAB and UAB (COUB) was defined as the coexistence of OAB and UAB. Although we operationally defined nocturia as ≥1 void/night for consistency with the JaCS2023 criteria, we acknowledge that ≥2 voids/night is often used to indicate clinically significant or bothersome nocturia. Univariate comparisons were performed using chi-square tests for categorical variables and a Student’s t-test or Mann–Whitney U test for continuous variables depending on data distribution. Multivariable logistic regression estimated odds ratios (ORs) and confidence intervals (CIs) for ED with LUTS, adjusting for prespecified covariates. Because intercourse frequency markedly attenuated effect estimates when included and was plausibly part of the progression from LUTS to ED, it was treated as a mediating variable and excluded from the primary model. Intercourse frequency was added in a secondary model to evaluate its mediator role. Analyses were conducted using multivariable logistic regression with two-sided P < 0.05. Analyses were conducted using Stata SE (version 19.0). Among 2627 men, 2049 (78.0%) and 763 (29.0%) met the criteria for LUTS and ED, respectively. In univariate analysis, LUTS were strongly associated with ED (crude OR 3.01, 95% CI, 2.34-3.87; P < .001; Supplementary Table S3). After covariate adjustment, LUTS remained independently associated with ED (adjusted OR [aOR] 1.47, 95% CI, 1.11-1.95; P = .007; Supplementary Table S3). Other independent ED correlates included older age (60-99 years: OR 4.40, 95% CI, 3.25-5.97; P < .001), performance status≥1 (aOR 2.48, 95% CI, 1.90-3.22; P < .001), diabetes (aOR 1.84; P < .001), benign prostatic hyperplasia (aOR 1.60; P = .007), dysuria (aOR 1.63; P = .006), and awareness that LUTS affected sexual life (aOR 1.76, 95% CI, 1.15-2.72; P = .010). In the domain-specific multivariable models, voiding symptoms (aOR 1.70, 95% CI, 1.37-2.09; P < .001), storage symptoms (aOR 1.34, 95% CI, 1.05-1.71; P = .020), post-micturition symptoms (aOR 1.35, 95% CI, 1.10-1.66; P = .005), and OAB (aOR 1.57, 95% CI, 1.20-2.05; P = .001) were independently associated with ED (Supplementary Table S4, Supplementary Figure S1). In contrast, UAB did not reach statistical significance after adjustment (aOR 1.38; P = .065), and COUB was not significant (aOR 1.27; P = .331). In the secondary model, higher intercourse frequency was strongly associated with lower odds of ED and substantially attenuated the association between LUTS and ED (aOR for LUTS decreased from approximately 1.44 to approximately 1.10 and lost statistical significance; Supplementary Table S5), indicating that sexual activity may be involved in the observed association between LUTS and ED. Herein, LUTS were independently associated with ED after extensive adjustment, consistent with previous epidemiologic evidence that LUTS and ED frequently coexist in aging men.2 Domain-specific results suggest clinically relevant heterogeneity. Erectile dysfunction was more closely linked to dynamic dysfunction during the storage and voiding phases, particularly with voiding symptoms and OAB, than to UAB phenotypes (Supplementary Table S4). The independent associations of OAB and post-micturition symptoms with ED are consistent with previous reports linking LUTS domains to impaired sexual function.3 Contrastingly, UAB and COUB were not independently associated with ED in the multivariable analysis (Supplementary Table S4). Given that the International Continence Society definition of UAB emphasizes reduced sensation and contractility-related symptom patterns,4 the lack of an independent association may reflect confounding by age or comorbidity, phenotype heterogeneity in self-reported symptoms, or distinct mechanisms compared with OAB or voiding dysfunction. Sexual intercourse frequency was considered a potential mediator rather than a confounder in the association between LUTS and ED. Because adjustment for intercourse frequency attenuated the effect estimates, it was excluded from the primary multivariable model to avoid overadjustment and was examined separately in a secondary analysis. Given the cross-sectional design, these findings are hypothesis-generating and do not establish causality or directionality; reverse causation is also possible. Furthermore, reduced sexual activity itself should not be interpreted as evidence of ED. Additionally, awareness that LUTS affected sexual life remained independently associated with ED, indicating that subjective burden and anxiety related to urinary symptoms may worsen erectile function beyond objective symptom presence. Clinically, men presenting with LUTS, particularly those with voiding symptoms or OAB, should be assessed for ED and for the perceived impact of urinary symptoms on sexual life. Integrated management addressing urinary symptoms, comorbidities (eg, diabetes and benign prostatic hyperplasia), and psychosocial distress may improve overall quality of life. This study has some limitations. Its cross-sectional design precludes causal inference or temporal ordering. Lower urinary tract symptoms and ED were assessed by self-reporting, and nonresponse to the ED item may have introduced selection bias. Second, ED was assessed using a single self-reported erection-hardness item rather than a validated comprehensive questionnaire; thus, misclassification is possible and our measure may not fully reflect ED as defined by the ability to achieve satisfactory sexual activity. Residual confounding is possible despite extensive adjustment. Prospective studies and interventions are needed to clarify causality and to determine whether early LUTS management preserves sexual function. This study used data from the Japan Continence Survey 2023 (JaCS2023), supported by the Japanese Continence Society and Kochi Medical School Hospital President’s Discretionary Grant 2025 (N.S. and Y.K.). The authors declare that there are no conflicts of interest regarding the publication of this paper.