Sexual function concerns and attitudes toward discussing them with psychiatrists among psychiatric outpatients in Japan: a cross-sectional study

Abstract

Background

Sexual dysfunction, defined as persistent difficulties in sexual desire, arousal, or orgasm, affects treatment adherence to psychotropic medication and quality of life in people with mental disorders, yet patients rarely disclose such concerns to clinicians. Understanding patient-perceived barriers to disclosure is critical for improving care, but empirical data from psychiatric outpatient settings remain limited, particularly in Japan where cultural norms around sexual health may further inhibit communication.

Methods

We conducted an anonymous cross-sectional online survey of psychiatric outpatients (N = 264) at 9 Japanese institutions between July and September 2024. Participants reported sexual function concerns, self-attributed main contributors, and attitudes toward discussing these concerns with their psychiatrist, classified as Active, Passive, or No-intention/other. Multivariable multinomial logistic regression examined factors associated with these attitudes (N = 223), adjusting for sex, age, depressive symptoms, and socioeconomic factors.

Results

Current sexual function concerns were reported by 22.3% of respondents, with lifetime concerns reaching 33.3%. Among those with current concerns, one-third attributed their concerns primarily to medication. Only 30% held an Active attitude toward consulting their psychiatrist about sexual function. In the combined analytic sample spanning current and hypothetical response contexts, women had 2.4-fold higher adjusted odds of a No-intention/other attitude versus an Active attitude (aOR = 2.42, 95% CI 1.07–5.46, p = 0.033) after adjustment for age, depressive symptoms, and socioeconomic factors.

Conclusions

Most psychiatric outpatients with sexual function concerns held passive or non-disclosing attitudes toward discussing them with their psychiatrist. In the combined analysis of current and hypothetical response contexts, women had higher adjusted odds of a No-intention/other rather than an Active attitude. Given that many patients attribute their concerns to medication, routine clinician-initiated screening and patient-centered communication may help identify modifiable contributors and thereby support adherence to psychiatric treatment.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12888-026-08393-3.

Background

Sexual dysfunction, defined as persistent or recurrent difficulties in sexual desire, arousal, or orgasm that cause distress [1], is commonly reported among people with mental disorders and those receiving psychotropic medication, adversely affecting quality of life and treatment engagement [2–6]. Psychiatric symptoms may influence sexual functioning through motivational, cognitive, interpersonal, and neurobiological pathways [5]. Treatment-emergent sexual dysfunction (for example, selective serotonin reuptake inhibitor [SSRI]-induced anorgasmia or antipsychotic-related hyperprolactinaemia) often goes under-recognized and may reduce adherence to psychotropic medication if not proactively addressed [7, 8].

Sexual dysfunction remains under-addressed in psychiatric clinical practice. Patient-level barriers to disclosure include embarrassment, self-stigma and anticipated public stigma, and perceptions that sexual health is not an appropriate medical topic; clinician-level barriers include time constraints, discomfort, and limited training in sexual history-taking and in the management of treatment-emergent sexual dysfunction [9, 10]. Many patients prefer clinicians to initiate these conversations [11, 12]. Disclosure enables assessment and interventions such as psychoeducation or medication adjustments [4, 6], yet help-seeking remains limited even when sexual function concerns are distressing [12].

In Japan, sociocultural norms around privacy may further constrain disclosure about sexual health. Prior studies have examined sexual dysfunction among Japanese adults in online and primary care settings [13, 14]. Although population-based estimates for Japan remain scarce, the available data indicate a substantial burden: in a community-based cohort of Japanese women, symptoms of sexual desire disorder were reported by 27.7–57.9% and arousal-related symptoms by 29.7–57.9% of women in their 30s to 60s [15]; a health-screening study of Japanese men reported the prevalence of erectile dysfunction (all degrees) at 63.7%, 78.2%, and 90.3% in the 40s, 50s, and 60 years or older age bands, respectively [16]; and in the primary care study above, 92% of men aged 40–69 years met erectile dysfunction criteria on a validated scale whereas only 39% self-identified sexual dysfunction [14]. Nevertheless, studies on attitudes toward discussing sexual dysfunction among people with mental disorders in psychiatric outpatient care remain limited. Much prior research has assessed sexual dysfunction using standardized instruments, whereas fewer studies have focused on patients’ subjective experiences, that is, what they perceive as troubling and what prevents disclosure. Understanding subjective concerns is relevant because disclosure is driven by perceived burden and barriers rather than diagnostic thresholds alone. Accordingly, we operationalized sexual function concerns as patient-perceived being “troubled” by sexual function rather than clinician-diagnosed dysfunction.

Health behavior theories offer useful frameworks for understanding disclosure of sexual function concerns. We drew on the Health Belief Model because disclosure can be understood as a health behavior shaped by the patient’s weighing of perceived benefits against perceived barriers, together with cues to action such as a clinician’s invitation to discuss the topic [17]. We additionally drew on the Common-Sense Model of self-regulation, which explains how a patient’s own interpretation of a symptom, including its perceived cause and controllability, shapes help-seeking [18]; this is directly relevant because our survey captured patients’ attributions of their sexual function concerns to mental illness, medication, or other causes. These frameworks predict that symptom interpretation and anticipated barriers jointly influence disclosure.

We analyzed data from the psychiatric outpatients (N = 264) in a multi-department outpatient survey in Japan. We hypothesized that (a) a substantial proportion of psychiatric outpatients with current sexual function concerns would hold non-Active attitudes toward discussing them with their psychiatrist, and (b) sex would be associated with these attitudes after adjustment for age, depressive symptoms, and socioeconomic factors. Our aims were to: (1) examine the proportion reporting current sexual function concerns and their perceived main contributors; (2) describe attitudes toward discussing sexual function concerns with their psychiatrist; and (3) examine demographic and clinical factors associated with these attitudes.

Methods

Study design and setting

This cross-sectional study used data from an anonymous, self-administered online questionnaire (Google Forms) conducted between July 1 and September 30, 2024, at University of Tsukuba Hospital and 27 affiliated institutions. The questionnaire was administered in Japanese and was developed by the multidisciplinary research team (psychiatry, urology, and endocrinology) to assess age-related symptoms and their associations; it comprised approximately 80 items organized into 10 sections covering demographic and socioeconomic attributes, medical history, and validated symptom scales. The present analysis concerns the section on sexual function concerns and the attitude toward discussing them, which was presented to respondents who reported a history of a mental disorder. This study was conducted in accordance with the Declaration of Helsinki, and the study protocol was reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [19]. This study was approved by the Clinical Research Ethics Committee of University of Tsukuba Hospital (R06-027, approved May 30, 2024). At the start of the survey, all respondents read a study information page and provided electronic informed consent by selecting a statement confirming that they were aged 18 years or older and agreed to participate; those who declined or were under 18 were routed out of the survey. Reception staff distributed flyers containing a QR code linking to the survey. The questionnaire did not request direct identifiers (e.g., name, medical record number). Survey data were stored in password-protected files accessible only to the research team. Because the number of flyers distributed and QR-code page views were not recorded, response rates could not be calculated. To minimize duplicate responses, the survey included a screening question confirming first-time completion; however, duplicates cannot be entirely ruled out given the anonymous design.

Participants

Among 593 valid respondents across departments, this analysis focused on 264 participants who reported a history of a mental disorder and attended psychiatric, psychosomatic, or mental-health outpatient services. These 264 participants were recruited from 9 of the 28 institutions, all of which operated psychiatric outpatient services; the remaining 19 institutions did not provide psychiatric outpatient services. Respondents without a history of a mental disorder were not asked the sexual function and consultation-attitude items and were therefore not eligible for the present analysis. Eligible participants were adults aged ≥ 18 years who provided electronic informed consent.

Measures

Sexual function concerns and self-attributed main contributor

Participants were asked a single item, “Are you currently troubled by your sexual function?”, with an explanatory note stating that sexual function refers to sexual desire and the capacity to feel pleasure from intimacy or self-stimulation in both sexes, and additionally erectile and ejaculatory function in men. Response options included: currently troubled with mental illness, medication, or other/unknown as the perceived main contributor; troubled in the past only; never troubled; unsure; or prefer not to answer. The item did not specify a time frame for “currently”; “in the past only” denoted concerns experienced at some earlier point but not at the time of the survey, without a fixed cut-off. Current sexual function concerns were defined as endorsement of any “currently troubled” category, with the comparison group comprising those reporting past-only or no concerns. This single-item measure captures patient-perceived distress rather than clinically diagnosed sexual dysfunction, and perceived contributors reflect subjective attributions. The full wording of the item and its response options is provided in Additional file 1.

Attitudes toward discussing sexual function concerns with a psychiatrist

Participants reported their attitude toward discussing sexual function concerns with their psychiatrist (or, for those treated in psychosomatic medicine or a mental-health clinic, that physician). For those with current concerns, a single item assessed whether they had already consulted and, if not, their willingness and any hesitation. For those without current concerns, a parallel hypothetical single item assessed the attitude they anticipated if such concerns were to arise.

Responses to this single item were classified into a three-level attitude variable: Active (already consulted, or would consult without hesitation); Passive (would like to consult but would hesitate); and No-intention/other (would not consult, no need, unsure, or prefer not to answer). Because the classification was based on one item, no summative score or cut-off was used. The exact response options mapped to each category are listed in Additional file 1.

Participants reporting past-only concerns were asked about coping strategies instead; these responses were not comparable and were excluded from the primary regression analysis.

Depressive symptoms

Depressive symptoms were measured using the Japanese version of the Patient Health Questionnaire-9 (PHQ-9), with higher scores indicating greater severity [20]. The Japanese PHQ-9 has been validated against the Mini-International Neuropsychiatric Interview, with good internal consistency and criterion validity for detecting major depression, including in primary care settings [20–22]. Total scores were analyzed as a continuous measure; severity bands were not applied.

Insomnia symptoms

Insomnia symptoms were measured using the Japanese version of the Athens Insomnia Scale (AIS-J) [23], an eight-item self-report scale (e.g., difficulty falling asleep, night-time or early-morning awakening, daytime functioning) with higher scores indicating more severe insomnia. The AIS-J has demonstrated satisfactory internal consistency and validity in Japanese respondents [23]. Total scores were analyzed continuously without severity banding.

Demographic and socioeconomic variables

We included age, sex (women/men), highest educational attainment (lower-secondary, upper-secondary, vocational/junior college, or university/graduate), employment status, and household income (grouped into six bands, with an “unknown/prefer not to answer” option).

Use of AI-assisted technologies

During the preparation and revision of this work, one author (HM) used ChatGPT (OpenAI) to refine the English-language expression and clarity of the manuscript, assist with organizing author-written content, and assist with drafting and checking Python code for post hoc quality-control checks of selected reported values. The GPT-5.2 model was used during initial manuscript preparation and the GPT-5.5 model during revision. The authors executed the code locally, verified the variable definitions and outputs against the source data, and made all analytical decisions. No participant-level data or raw datasets were entered into the tool. The tool did not generate new scientific content, perform the primary statistical analysis, interpret the results, or formulate the conclusions. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article. A completed disclosure of generative-AI use following the GAMER checklist [24] is provided as Additional file 2.

Statistical analysis

Participant characteristics and response distributions were summarized using descriptive statistics. Comparisons between participants with and without current concerns used Welch’s t-tests for continuous variables and chi-square or Fisher’s exact tests for categorical variables, based on non-missing data for each variable. Effect sizes were reported as Cohen’s d for continuous variables and Cramér’s V for categorical variables.

For multivariable modeling, age, PHQ-9 score, education, and household income were standardized (mean = 0, SD = 1). Education and household income were treated as continuous ordinal variables to preserve power given the sample size (N = 223). Missing PHQ-9 and income values were handled using the missing-indicator method: missing values were imputed with the median before standardization, and a binary missingness indicator was included as a covariate. To examine the composition of the No-intention/other category, Fisher’s exact tests compared the proportion endorsing each underlying response option between sexes among participants without current concerns.

The primary analysis used multinomial logistic regression with the three-level attitude outcome, using the Active category as reference. Covariates included sex, age, education, employment, household income, depressive symptoms, and current concern status. Current concern status (current concern vs. no current concern) was included as a covariate because the outcome combined participants who reported current concerns and participants who responded to the parallel hypothetical item; adjusting for it allowed the sex association to be estimated across both contexts while accounting for whether the attitude was reported about an actual or a hypothetical concern. Results are reported as adjusted odds ratios (aOR) with 95% confidence intervals and p-values. Three sensitivity analyses examined robustness: (S1) excluding participants whose sexual-concern response was “unsure” or “prefer not to answer”; (S2) excluding “unsure” and “prefer not to answer” responses to the attitude item; and (S3) additionally adjusting for insomnia (AIS-J). We also tested a sex × current-concern interaction with a likelihood-ratio test.

Analyses used Python 3.11 with pandas, SciPy, and statsmodels. Statistical significance was set at p < 0.05 (two-sided).

Results

Participant characteristics

Our analysis included 264 survey respondents attending psychiatric outpatient clinics (164 women [62.1%], 98 men [37.1%], 2 not reported [0.8%]). Mean age was 42.9 ± 14.7 years (range 19–82). Baseline characteristics of all 264 participants are summarized in Table 1. Characteristics by current concern status are presented in Table 2.

Table 1.

Baseline characteristics of all participants (N = 264)

Characteristic	n	%
Total	264	100
Women	164	62.1
Men	98	37.1
Sex not reported	2	0.8
Age, mean (SD), years	42.9 (14.7)	—
Age range, years	19–82	—
Highest education: university/graduate	120	45.5
Currently employed	155	58.7
Household income ≥ JPY 4 million	68	25.8
PHQ-9, mean (SD)	11.1 (6.6)	—
AIS-J, mean (SD)	8.0 (4.5)	—
Sexual function concern: current	59	22.3
— attributed to mental illness	19	7.2
— attributed to medication	18	6.8
— attributed to other/unknown	22	8.3
Sexual function concern: past only	29	11.0
Sexual function concern: never	114	43.2
Sexual function concern: unsure	42	15.9
Sexual function concern: prefer not to answer	9	3.4
Sexual function concern: missing	11	4.2

Note: Values are n (%) unless otherwise indicated. Household income of JPY 4 million is approximately USD 27,000, GBP 21,000, or EUR 25,000 (2024 average exchange rates). PHQ-9 = Patient Health Questionnaire-9; AIS-J = Japanese version of the Athens Insomnia Scale [23]

Table 2.

Characteristics of participants by current sexual function concern status

Variable	Overall (N = 264)	No Current Concern (n = 143)	Current Concern (n = 59)	p
Age, years	42.9 (14.7)	43.0 (14.2)	41.0 (11.8)	0.292
Women	164 (62.1)	93 (66.0)	27 (45.8)	0.012
Education: University	120 (45.5)	65 (45.5)	29 (49.2)	0.746
Currently employed	155 (58.7)	85 (59.4)	36 (61.0)	0.830
Household income ≥ 4 M JPY	68 (25.8)	44 (30.8)	15 (25.4)	0.615
PHQ-9 total score	11.1 (6.6)	10.0 (6.6)	13.1 (6.3)	0.003
AIS-J total score	8.0 (4.5)	7.5 (4.4)	9.1 (4.5)	0.027

Note: Values are presented as mean (SD) or n (%). This comparison is restricted to the 202 participants who could be classified by current concern status (143 without and 59 with current concerns); the 62 participants who selected “unsure” (n = 42), “prefer not to answer” (n = 9), or had missing responses (n = 11) are not included. Percentages for sex are calculated among participants who reported their sex (n = 141 without and n = 59 with current concerns; two participants in the no-current-concern group did not report sex). “No current concern” includes participants reporting past-only concerns (n = 29) or never having concerns (n = 114). P-values were calculated using Welch’s t-test for continuous variables and chi-square or Fisher’s exact test for categorical variables, based on non-missing data for each variable. PHQ-9 = Patient Health Questionnaire-9 (range 0–27; higher scores indicate more severe depression); AIS-J = Japanese version of the Athens Insomnia Scale [23] (range 0–24; higher scores indicate more severe insomnia)

Frequency and perceived main contributors of sexual function concerns

Overall, 22.3% (59/264) reported current sexual function concerns. Among these, the perceived main contributor was mental illness in 32.2% (19/59), medication in 30.5% (18/59), and other/unknown factors in 37.3% (22/59), corresponding to 7.2%, 6.8%, and 8.3% of the total sample, respectively. Past-only concerns were reported by 11.0% (29/264), yielding a lifetime proportion of 33.3%. An additional 114 participants reported never having concerns, while others selected “unsure” (n = 42), “prefer not to answer” (n = 9), or had missing responses (n = 11).

Participants with current concerns had higher depressive symptom severity than those without (PHQ-9: 13.1 ± 6.3 vs. 10.0 ± 6.6, p = 0.003; Cohen’s d = 0.47) and higher insomnia severity (AIS-J: 9.1 ± 4.5 vs. 7.5 ± 4.4, p = 0.027; d = 0.36). The proportion of women was lower among those with current concerns (45.8% vs. 66.0%, p = 0.012; Cramér’s V = 0.18). Detailed results are presented in Table 2.

Attitudes toward discussing sexual function concerns among participants with current concerns

Among participants with current concerns (n = 59), 27.1% (16/59) had already consulted their psychiatrist about their sexual function concern, and an additional 3.4% (2/59) reported willingness to consult without hesitation. Thus, 30.5% (18/59) were classified as holding an Active attitude. In contrast, 32.2% (19/59) wanted to consult but would hesitate (Passive), 11.9% (7/59) did not want to consult, and 18.6% (11/59) reported “no need,” with the remainder selecting “unsure” or “prefer not to answer” (4/59).

Three-category attitude outcome

For the primary multinomial model, we excluded participants with past-only concerns (n = 29), missing sexual concern responses (n = 11), and one participant with unreported sex, yielding N = 223. This analytic sample comprised 59 participants with current concerns and 164 participants without current concerns (i.e., never troubled, unsure, or preferring not to answer) who responded to the parallel hypothetical item; participants with past-only concerns were a distinct group asked only about past coping and are not part of this no-current-concern group. This analytic grouping therefore differs from the descriptive comparison in Table 2, in which the no-current-concern column comprises past-only and never respondents and excludes those who selected “unsure” or “prefer not to answer.” In this analytic sample, 23.3% (52/223) were classified as Active, 34.5% (77/223) as Passive, and 42.2% (94/223) as No-intention/other. A full breakdown of the underlying response options by sex and by current-concern status is provided in Additional file 3. In exploratory item-level comparisons among participants without current concerns, women were less likely than men to report willingness to consult without hesitation (16.1% vs. 32.6%, p = 0.031) and more likely than men to indicate they “would not consult” (12.7% vs. 2.2%, p = 0.043), whereas proportions citing “no need” (15.3% vs. 13.0%, p = 0.81) or “unsure” (18.6% vs 17.4%, p > 0.99) were comparable between sexes. Among participants with current concerns, however, the proportion classified as Active was similar in women (8/27) and men (10/32); therefore, the combined-model association should not be interpreted as demonstrating lower actual consultation initiation among women with current concerns specifically.

Multivariable multinomial regression

In adjusted multinomial regression (reference=Active), women had higher adjusted odds of a No-intention/other attitude versus an Active attitude (aOR = 2.42, 95% CI 1.07–5.46, p = 0.033), indicating approximately 2.4-fold higher odds compared with men after adjusting for age, education, employment, income, depressive symptoms, and current concern status. Women were not significantly associated with the Passive category versus Active (aOR = 1.68, p = 0.198). Full results are shown in Table 3.

Table 3.

Multivariable multinomial logistic regression: factors associated with attitude categories (N = 223)

Variable	Passive aOR (95% CI)	p	No-intention/other aOR (95% CI)	p
Women (vs. men)	1.68 (0.76–3.72)	0.198	2.42 (1.07–5.46)	0.033
Age (per SD increase)	1.01 (0.69–1.48)	0.947	1.17 (0.84–1.63)	0.353
PHQ-9 (per SD increase)	1.20 (0.82–1.80)	0.340	1.15 (0.79–1.67)	0.474
PHQ-9 missing (yes vs. no)	0.12 (0.01–1.21)	0.074	0.97 (0.28–3.33)	0.967
Currently employed (yes vs. no)	1.50 (0.66–3.43)	0.338	0.98 (0.44–2.19)	0.959
Household income (per SD; ordinal)	1.01 (0.66–1.54)	0.980	1.17 (0.77–1.78)	0.465
Income unknown/prefer not (yes vs. no)	2.93 (0.50–16.99)	0.230	2.87 (0.50–14.47)	0.202
Current sexual function concerns (yes vs. no)	0.62 (0.26–1.46)	0.275	0.71 (0.31–1.63)	0.417
Education (per SD; ordinal)	1.34 (0.91–1.98)	0.144	1.08 (0.74–1.57)	0.681

Note: Reference category = Active attitude (already consulted or would consult without hesitation). Model adjusted for all variables shown in the table. Continuous variables (age, PHQ-9, household income, education) were standardized (mean = 0, SD = 1) before analysis. N = 223 after excluding participants with past-only concerns (n = 29), missing sexual concern responses (n = 11), and sex not reported (n = 1). Model fit: McFadden pseudo-R² = 0.052, AIC = 492.7. aOR = adjusted odds ratio; CI = confidence interval; SD = standard deviation; PHQ-9 = Patient Health Questionnaire-9

Predicted probabilities from the fitted model, with continuous covariates held at their sample means and categorical covariates at their reference levels (employment set to not currently employed; missing-data indicators set to non-missing), showed that women were more likely than men to hold a No-intention/other attitude in both the current-concern (0.48 vs. 0.32) and no-current-concern (0.50 vs. 0.36) contexts.

Sensitivity analyses

The point estimates for female sex in the No-intention/other versus Active contrast were similar in direction and magnitude across the three sensitivity analyses, although precision and statistical significance varied. In S1 (N = 172), excluding participants whose sexual-concern response was “unsure” or “prefer not to answer” yielded a similar point estimate, but statistical significance was not retained (aOR = 2.48, 95% CI 0.997–6.15, p = 0.051). In S2 (N = 187), excluding “unsure” and “prefer not to answer” responses to the attitude item yielded aOR = 2.88 (95% CI 1.13–7.33, p = 0.026). In S3 (N = 223), additionally adjusting for insomnia yielded aOR = 2.58 (95% CI 1.13–5.90, p = 0.025). The corresponding Passive-versus-Active estimates remained non-significant (aOR 1.6–1.8 across analyses). A sex × current-concern interaction was not significant (likelihood-ratio χ²=1.67, df = 2, p = 0.43).

Discussion

This multi-institution survey describes how psychiatric outpatients in Japan perceive discussing sexual function concerns with their psychiatrist and suggests a sex difference in disclosure attitudes that persisted after adjustment for age, depressive symptoms, socioeconomic factors, and current concern status. Rather than estimating the prevalence of sexual dysfunction alone, our findings highlight a clinically actionable issue: whether patients who experience sexual function concerns feel able to raise them in routine psychiatric care.

Because our outcome captured being currently troubled by sexual function, the observed 22.3% should not be interpreted as the prevalence of clinically defined sexual dysfunction. Prevalence estimates using multidimensional instruments are substantially higher and vary by diagnosis and measurement approach. A systematic review found that, even among medication-free patients with psychiatric disorders, sexual dysfunction was reported in 45–93% of those with depressive disorders and 33–75% of those with anxiety disorders, indicating that the disorders themselves, not only their treatment, contribute to sexual dysfunction [25, 26]. In Japan, sexual dysfunction assessed with the clinician-rated Udvalg for Kliniske Undersøgelser Side Effect Rating Scale (UKU-SERS-Clin) [27] was reported in approximately half of schizophrenia outpatients on antipsychotics [28, 29]. These data suggest that distress-based screening identifies a clinically meaningful subset but likely underestimates the broader burden.

Among respondents with current concerns, about one-third attributed them primarily to medication. The response option referred to medication only as a general category and did not record the specific drug or class, so the medications involved cannot be identified; in a psychiatric outpatient sample, however, they are most plausibly psychotropic agents—particularly antidepressants and antipsychotics, the classes most often implicated in treatment-emergent sexual dysfunction [2–4, 6]. Because these are subjective attributions rather than clinically adjudicated causes, causal interpretations warrant caution. Nevertheless, medication-related attributions are clinically relevant given that sexual adverse effects are often under-recognized and may undermine adherence to psychotropic medication if unaddressed [7, 8]. Because patients frequently do not raise these concerns spontaneously, proactively asking about patients’ perceived contributors and providing individualized information about management options may help clinicians identify potentially modifiable factors before unaddressed sexual concerns affect treatment adherence.

Our findings suggest links among sexual function concerns, depressive symptoms, and insomnia. Psychiatric outpatients with current sexual function concerns reported greater depressive and insomnia symptom severity than those without. These cross-sectional associations are plausibly bidirectional: depressive symptoms can impair sexual desire and arousal through motivational, interpersonal, and neurobiological pathways [5]; antidepressants and antipsychotics can themselves induce sexual dysfunction [2–4, 6]; and persistent sexual dysfunction and sleep disturbance may in turn worsen depressive symptoms and reduce quality of life [7]. Given the cross-sectional design, the direction of these relationships cannot be determined.

Regarding help-seeking, our finding that only about 30% of participants with current concerns held an Active attitude toward consulting their psychiatrist aligns with prior reports of substantial nondisclosure. In an open-access survey of antidepressant users with self-attributed sexual dysfunction, fewer than half had discussed sexual side effects with their doctor [30]. In Japan, an Internet survey found sexual dysfunction to be the least likely antidepressant side effect reported to physicians [31], and primary care research suggests only a small minority of patients with sexual dysfunction have ever discussed it with their physician [14]. Provider-level data indicate that psychiatrists may underrecognize side effects when relying on routine documentation [32] and that sexual health is addressed routinely by only a minority of psychiatrists, often citing time constraints, embarrassment, and perceived lack of competence [33]. A multi-country Asian study similarly found sexual dysfunction seldom recorded by psychiatrists treating schizophrenia [34]. Our study extends this literature by quantifying attitudes toward discussion and perceived contributors in psychiatric outpatient settings, providing a patient-centered view of where the disclosure gap persists.

Women in the combined adjusted model were more likely than men to hold a No-intention/other attitude rather than an Active attitude after adjustment for age, depressive symptoms, socioeconomic factors, and current concern status. The association between female sex and a No-intention/other attitude remained similar after additional adjustment for insomnia severity (sensitivity analysis S3), suggesting that the observed sex difference in disclosure attitudes was not fully explained by measured insomnia symptoms. The response breakdown suggested that this difference reflected both higher endorsement of “would not consult” and lower endorsement of unhesitating willingness among women, rather than differences in “no need” or “unsure” responses. Several factors may be relevant. Differences in symptom profiles and medication effects may influence whether patients perceive sexual function concerns as clinically relevant. In addition, the single-item measure used in this study may not have captured the full range of women’s sexual concerns, such as desire, arousal, and pain. Social norms around embarrassment and the appropriateness of discussing sexual health with clinicians may also reduce disclosure, particularly when clinicians do not initiate the conversation; previous survey and qualitative research suggests that patients often do not raise sexual concerns unless clinicians ask, and that clinician-initiated inquiry can help reduce this communication gap [10–12]. Although clinician gender was not assessed, patient–clinician gender concordance may also affect willingness to disclose sexual concerns. More broadly, women in general populations report a higher burden of depressive and insomnia symptoms, and gender-specific psychosocial vulnerability, as documented for example during the COVID-19 pandemic [35], may shape both the experience of sexual function concerns and subsequent help-seeking.

A further consideration is that difficulty in identifying and communicating internal states may impede disclosure of sexual concerns. Alexithymic features have been associated with greater symptom severity and with suicidal ideation in psychiatric outpatients [36], suggesting that patients who find it hardest to articulate distress may also be those least likely to raise sexual concerns spontaneously and may be at higher clinical risk. This is an argument for clinician-initiated inquiry rather than reliance on patient disclosure.

From a behavioral perspective, these findings are consistent with the Health Belief Model and the Common-Sense Model [17, 18]. In terms of the Health Belief Model, patients may be less likely to consult when the expected benefits of disclosure are outweighed by perceived barriers, such as embarrassment or the belief that sexual health is outside the psychiatrist’s role. In terms of the Common-Sense Model, patients who interpret sexual function concerns as an unavoidable consequence of mental illness or psychotropic medication may see little reason to raise them. Anticipated social cost and perceived non-modifiability may therefore be clinically relevant targets for brief clinician-initiated communication.

These findings have practical implications for routine psychiatric care. First, psychiatrists can incorporate brief, normalized screening questions and explicitly invite discussion, consistent with evidence that patients often prefer clinician initiation [11, 12]. An example of such a normalizing question is: “Some people notice changes in their sexual desire or function when they have mental health symptoms or take psychiatric medication; have you experienced anything like that?” This approach is readily implementable because it requires no additional service or referral pathway and can be embedded in routine consultation. Providing clear explanations that sexual function concerns can be related to psychiatric symptoms or psychotropic medication, and that management options may be available, may increase the perceived benefits of disclosure [4, 6]. Clinicians may also consider gender-sensitive communication strategies, including offering follow-up with a clinician of the patient’s preferred gender where feasible. Finally, referral pathways to urology, gynecology, or sexual health specialists may be useful for patients requiring specialized assessment, particularly when symptoms fall within the clinical spectrum of sexual dysfunction described in ICD-11 conditions related to sexual health [37].

Limitations

This study has several limitations. First, the cross-sectional design precludes causal inference; in particular, the associations between current concerns and depressive or insomnia symptoms are cross-sectional and may be bidirectional. Second, voluntary participation via self-administered online survey may limit generalizability due to self-selection and technology-related bias. Older patients or those less familiar with smartphones may have been underrepresented, potentially contributing to the relatively young mean age. Because flyer distribution and QR-code page views were not recorded, response rates could not be calculated. Although a screening question discouraged duplicate responses, multiple submissions could not be technically prevented. Third, key measures relied on self-report, and sexual function concerns and their attributed contributors were captured with a single item rather than a validated multidimensional instrument. Sexual pain was not explicitly assessed as a separate domain and may therefore have been undercaptured. The single-item design also precluded domain-specific analyses of desire, arousal, orgasm, and pain, which have distinct mechanisms and may differ by sex. Fourth, the attitude outcome combined current and hypothetical contexts; for participants without current concerns, responses reflect intentions rather than actual behavior, although the sex × context interaction was not significant. Fifth, the recruited sample was drawn from 9 of the 28 institutions and was predominantly from a single university hospital network, which may further limit generalizability. Sixth, because the survey was conducted in 2024, we did not directly assess the influence of the COVID-19 pandemic, which has been shown to heighten psychosocial vulnerability and to affect mental health differently by sex [35]; such contextual factors may shape help-seeking and warrant attention in future work. Finally, we did not adjust for clinical variables such as specific diagnoses, medication types/doses, or clinician characteristics, which may confound the observed associations.

Future research

Future work should incorporate validated multidimensional measures that explicitly assess sexual desire, arousal, orgasm, and sexual pain, alongside detailed medication and diagnostic data, and should examine specific domains of sexual dysfunction separately by sex. Qualitative studies, such as interviews with women receiving psychiatric care who report sexual concerns but hesitate to consult, could identify modifiable barriers including anticipated stigma, perceptions that sexual topics fall outside the psychiatrist’s role, or preferences regarding clinician gender. Intervention studies evaluating clinician-initiated screening and communication training could test strategies to reduce barriers in routine care.

Conclusions

In a Japanese psychiatry outpatient sample, about one in five respondents reported current sexual function concerns, yet only about 30% of those with concerns held an Active attitude toward discussing them with their psychiatrist. In the combined analysis of current and hypothetical response contexts, women had higher adjusted odds of a No-intention/other attitude than men after adjustment for age, depressive symptoms, socioeconomic factors, and current concern status. These findings suggest that an important obstacle in routine psychiatric care may be whether patients feel able to raise sexual function concerns. The disclosure gap is unlikely to be addressed by relying on patient-initiated reporting alone. Proactive clinician-initiated screening and patient-centered communication may help reduce barriers to discussing sexual function concerns in psychiatric care. Because the findings derive from a predominantly single-network Japanese sample using a single-item measure, they should be confirmed in broader multi-site studies using validated instruments; nevertheless, brief and normalized clinician-initiated inquiry may be a practical first step in comparable outpatient settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

PHQ-9

Patient Health Questionnaire-9

AIS-J

Japanese version of the Athens Insomnia Scale

aOR

Adjusted odds ratio

CI

Confidence interval

SD

Standard deviation

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

UKU-SERS-Clin

Udvalg for Kliniske Undersøgelser Side Effect Rating Scale, clinician-rated

ICD-11

International Classification of Diseases, 11th Revision

Author contributions

HM: Conceptualization, Investigation, Formal analysis, Methodology, Writing - original draft, Writing - review & editing. KK: Conceptualization, Investigation, Writing - review & editing. TM: Conceptualization, Investigation, Writing - review & editing. TT: Writing - review & editing. HN: Resources, Writing - review & editing. TA: Resources, Supervision, Writing - review & editing. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request, subject to ethical and institutional approvals.

Declarations

Ethics approval and consent to participate

This study was performed in accordance with the Declaration of Helsinki and was approved by the Clinical Research Ethics Committee of University of Tsukuba Hospital (R06-027, approved May 30, 2024). All participants provided electronic informed consent before participating in the survey.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.