Male- and female-factor infertility and partners' quality of life during assisted reproductive technology: A dyadic actor-partner analysis.

Infertility is a major life crisis affecting millions of couples worldwide, bringing heavy psychological, emotional, and social burdens. 1 While medical care often emphasizes diagnosis and treatment, the impact on quality of life (QoL) is profound. Research has traditionally focused on female partners, who carry the greater physical burden of assisted reproductive technologies (ART) and face stronger societal pressure around childbearing. 2 However, infertility is fundamentally a dyadic stressor experienced by the couple as a unit. 3 Psychological research confirms that partners’ emotional responses are interdependent; distress or depression in one partner frequently “crosses over” to affect the other’s well-being. 4 Consequently, individual-based assessments fail to capture the complex relational dynamics of infertility. Recent studies emphasize that dyadic coping—how couples manage stress together—is a critical mechanism linking fertility stress to marital quality, underscoring the necessity of evaluating the couple as an interdependent system rather than separate individuals. 3 Despite this interdependence, the influence of the specific infertility etiology on dyadic QoL remains unsettled. While some evidence suggests that the psychological burden is similar across diagnoses—implying that childlessness itself is the primary stressor 5 —other studies indicate that the specific cause matters. For instance, male-factor infertility (MFI) is frequently linked to impaired male self-esteem and sexual satisfaction. 6 – 8 Conversely, female-factor diagnoses may uniquely distress male partners through a “bystander effect,” where they witness the physical toll of invasive treatments on their wives. It remains unclear whether these diagnostic categories exert asymmetric “actor” and “partner” effects within the dyad. 9 Our previous institutional work has demonstrated that female partners’ QoL is associated with the diagnosis of endometriosis and influences ART outcomes 10 – 12 ; however, these studies were limited to female partners alone. Therefore, the present study applies the actor–partner interdependence model (APIM) to a cohort of infertile couples in Taiwan to disentangle these dynamics. We aim to rigorously evaluate how objective infertility diagnoses (male, female, mixed, or unexplained) influence the QoL of both partners, testing the hypothesis that etiology exerts distinct effects on both the individual and their spouse.

This retrospective cross-sectional study was conducted at the Assisted Reproductive Technology Center of National Cheng Kung University Hospital, a tertiary medical center in southern Taiwan. Under Taiwan’s Assisted Reproduction Act, assisted reproduction is restricted to married heterosexual couples; accordingly, all dyads in this study met these criteria, and findings should be interpreted within this legal context. Eligible couples were those undergoing ART at our center during the study period, in whom both partners completed the Fertility Quality of Life (FertiQoL) questionnaire at a treatment visit. We consecutively enrolled 105 couples between January and November 2024, with Institutional Review Board approval (A-ER-114-183). Demographics, infertility duration, and treatment history were collected; clinical data (gravidity, parity, serum anti-Müllerian hormone [AMH]) were extracted from electronic medical records. During ART, both partners completed the FertiQoL questionnaire before transvaginal oocyte retrieval or embryo transfer (ET). The cause of infertility was determined through a comprehensive diagnostic workup completed prior to the initiation of ART treatment. Patients were fully aware of their diagnosis at the time of enrollment. Etiology was categorized as: Female Factor (eg, ovulatory dysfunction, diminished ovarian reserve), Male Factor (based on WHO 2021 criteria), Combined Factor, or Unexplained. For the APIM analysis, these diagnoses were coded into two separate dichotomous variables (present/absent) for each couple. Specifically, the “female factor present” variable was coded as “yes” for couples with either a female-only or a combined-factor diagnosis. Similarly, the “male factor present” variable was coded as “yes” for those with a male-only or combined-factor diagnosis. For couples with an unexplained cause of infertility, both variables were coded as “no.” The validated Mandarin version of the FertiQoL tool, which had been checked for reliability and validity in our previous study, 13 was used. It consists of a Core module (Emotional, Mind-Body, Relational, Social domains) and a Treatment (environment, tolerability) module. Higher scores indicate better QoL. Participants in this study underwent a standard in vitro fertilization (IVF) protocol. Ovarian stimulation was primarily achieved using gonadotropins (eg, recombinant follicle-stimulating hormone with or without luteinizing hormone), with the dosage and protocol (such as the gonadotropin-releasing hormone [GnRH] agonist or antagonist protocol) tailored to each patient’s individual characteristics, including age, AMH level, and previous response. Follicular growth was closely monitored through transvaginal ultrasound and serum hormone assessments. Once the lead follicles reached an appropriate size (typically ≥18 mm), a maturation trigger (eg, human chorionic gonadotropin or GnRH agonist) was administered. Oocyte retrieval was scheduled approximately 34 to 36 hours post-trigger and performed under ultrasound guidance. For comparisons of patient characteristics, categorical variables were analyzed using the χ 2 test or Fisher’s exact test, as appropriate. Continuous variables with a normal distribution were analyzed using analysis of variance, while non-normally distributed data were analyzed using the Kruskal–Wallis test. For comparisons of FertiQoL scores between male partners and female partners, the paired t test was used. All descriptive statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC). The APIM was then used to examine the associations between couple-infertility factors and both partners’ outcomes. All APIM analyses were conducted using the online software package “APIM_MM.” 14 A two-sided p value <0.05 was regarded as statistically significant. To test the robustness of our findings, a sensitivity analysis was conducted by reanalyzing the primary APIM while adjusting for potential confounding variables, including partner age, body mass index (BMI), duration of infertility, and the number of previous ET attempts without clinical pregnancy, which were entered into the model as covariates. To assess potential selection bias, we compared the demographic and clinical characteristics of the included female partners with those of the excluded individuals (missing partner data). Continuous variables were analyzed using the independent t test or the Mann–Whitney U test, depending on data distribution, while categorical variables were compared using the χ 2 test.

A total of 393 patient records from infertility consultation and treatment at our facility during the study period were initially identified. To assemble the final dyadic cohort, we first excluded 119 individuals for whom corresponding partner data were unavailable. An additional 62 individuals with multiple records from different treatment cycles or timings were excluded by retaining a single relevant entry for each partner. Specifically, for female patients, the latest record was chosen, while for male patients, the record with the date most proximal to their partner’s was used. After excluding one couple for incomplete data, the final study population comprised 105 couples (210 individuals). Of the 105 participating couples, 67 (63.8%) proceeded to ET during the study period, resulting in 25 clinical pregnancies (37.3% per transfer) and 22 live births (32.8% per transfer). Of the 38 women who did not undergo ET within the observation window, two attended consultation only and did not proceed to ovarian stimulation or oocyte retrieval, and four had no fertilized embryos after IVF. The remaining 32 had at least one embryo but had not yet undergone ET: two were awaiting preimplantation genetic testing for monogenic disorders (PGT-M) results, five were awaiting elective preimplantation genetic testing for aneuploidies (PGT-A) results, and 25 were awaiting a subsequent transfer cycle. The statistics of the demographics and relevant clinical characteristics of the study population by sex are summarized in Tables 1 and 2 . For the purposes of this descriptive analysis, one couple with an unknown cause of infertility was classified into the female-only infertility group. The mean age of the final cohort was 38.2 ± 4.3 years for female partners and 40.7 ± 6.4 years for male partners. The average duration of infertility was 3.7 ± 3.4 years. BMI of the male partners showed a significant difference among the three infertility factor groups, with the highest BMI observed in the “female-only” infertility group (26.2 kg/m 2 vs 24.3 kg/m 2 vs 23.5 kg/m 2 , p = 0.044). Comparison of characteristics by infertility factor in male partners Continuous variables with normal distribution are presented as mean ± SD, data without normal distribution are presented as median (interquartile range); categorical variables are presented as counts (percentage). p value <0.05 is shown in bold. BMI = body mass index. Comparison of characteristics by infertility factor in female partners Continuous variables with normal distribution are presented as mean ± SD, data without normal distribution are presented as median (interquartile range); categorical variables are presented as counts (percentage). AMH = anti-Müllerian hormone; BMI = body mass index; ET = embryo transfer. p values calculated with the Kruskal–Wallis test. p values calculated with the χ 2 test. Include only patients who received embryo transfer (n = 67). To assess potential selection bias, we compared the clinical characteristics of the included couples (n = 105) with the excluded female patients (n = 119) for whom male partner data were unavailable. Crucially, there were no significant differences in the distribution of infertility diagnoses between the groups ( p = 0.832). However, the groups differed significantly by treatment phase, whereas 96.2% of the included group underwent ovarian stimulation and oocyte retrieval, 92.4% of the excluded group did not receive these interventions, consisting largely of patients returning for frozen ETs (Supplementary Table 1, https://links.lww.com/JCMA/A385 ). Consequently, while Core QoL scores were similar between the groups ( p = 0.286), the excluded group reported significantly higher Treatment domain scores ( p < 0.001), as shown in the Supplementary Table 2, https://links.lww.com/JCMA/A385 . To verify that this discrepancy was driven by the treatment protocol rather than patient characteristics, we performed a pooled analysis of all 224 female patients (included and excluded combined), stratified solely by receipt of ovarian stimulation (Supplementary Table 3, https://links.lww.com/JCMA/A385 ). This analysis showed that undergoing stimulation significantly lowered Treatment domain scores (Tolerability: 19.3 vs 38.9, p < 0.001) without affecting Core QoL ( p = 0.152). Thus, the observed variations reflect the temporary physical burden of the stimulation phase rather than a systematic selection bias in the excluded population. Baseline QoL scores showed that male partners reported significantly higher QoL than their female partners across the Emotional, Mind-Body, Social, and Total Core FertiQoL domains (all p < 0.01) (Table 3 ). Comparisons of FertiQoL scores by sex Continuous FertiQoL scores are presented as mean ± SD. p values <0.05 are shown in bold. FertiQoL = Fertility Quality of Life. The APIM analysis revealed a significant and asymmetrical partner effect. The presence of a female-factor infertility diagnosis was significantly associated with lower Mind-Body QoL ( β = −17.54, 95% CI = −31.89 to −3.19, p = 0.017) and lower Core QoL ( β = −10.15, 95% CI = −20.08 to −0.22, p = 0.045) in their male partners (Table 4 ). Conversely, we did not detect a significant partner effect of a diagnosis of MFI on the QoL of their female partners. The analysis also showed no significant actor effects; neither a male partner’s nor a female partner’s own infertility diagnosis was significantly associated with a decline in their own QoL (Fig. 1 ). APIM estimates of actor and partner effects of infertility factor presence on FertiQoL p values <0.05 are shown in bold. APIM = Actor–Partner Interdependence Model; FertiQoL = Fertility Quality of Life; M infertility = present with male factor infertility; M score = male partner’s score of specific FertiQoL domain; W infertility = present with female factor infertility; W score = female partner’s score of specific FertiQoL domain. APIM results for the FertiQoL Total Core score (A) and Mind-Body domain (B). Horizontal solid blue arrows denote actor effects (the impact of an individual’s diagnosis on their own QoL), while diagonal dashed orange arrows denote partner effects (the impact of one partner’s diagnosis on the other’s QoL). Unstandardized path coefficients are presented on the paths, with standardized estimates shown in parentheses. The curved arrow on the left depicts the covariance between partners’ infertility diagnoses. Values are presented as unstandardized covariance, with the standardized correlation coefficient ( r ) shown in parentheses. The curved arrow between E 1 and E 2 denotes the residual correlation between partners’ QoL scores. * p < 0.05. APIM = Actor–Partner Interdependence Model; QoL = quality of life. Residual errors, indicated as E 1 and E 2 , capture the variance in male and female partners’ QoL scores not explained by infertility diagnosis predictors. A significant positive correlation between these residuals signifies residual dependence, indicating that shared, unmeasured factors (eg, treatment stress, mutual coping) influence the couple’s QoL similarly, even after accounting for actor and partner effects of diagnosis. This reinforces the deeply interdependent nature of their experience (Supplementary Table 4, https://links.lww.com/JCMA/A385 ). Further sensitivity analyses were conducted for “Mind-body” and “Core” domains, with results summarized in Supplementary Table 4, https://links.lww.com/JCMA/A385 . Clinically relevant covariates were sequentially adjusted across models. For the “Mind-body” domain, the direction and statistical significance of the associations remained consistent across all models, indicating robust actor–partner effects under different covariate adjustment strategies. In contrast, for the “Core” domain, additional adjustment for age attenuated the association between female infertility status and the partner’s outcome. Although the effect direction remained consistent, the statistical significance was reduced to a borderline level in age-adjusted models (Supplementary Table 4, https://links.lww.com/JCMA/A385 ).

Supplementary data related to this article can be found at https://links.lww.com/JCMA/A385 .

Our study confirmed the well-established gender disparity in infertility: female partners consistently reported lower QoL than their male counterparts, reflecting the disproportionate physical burden of ART and societal pressures regarding childbearing. 8 , 9 However, the application of the APIM revealed a critical nuance beyond this baseline difference: a significant, asymmetrical partner effect where a female-factor diagnosis specifically impaired the male partner’s QoL. The decline in male QoL associated with a female-factor diagnosis is best interpreted through the lens of the “bystander effect.” 15 , 16 In this diagnostic context, the male partner is uniquely positioned as a passive observer. He witnesses his partner’s physical ordeal—injections, retrieval, and recovery—yet is unable to physically share the burden or medically resolve the underlying cause. Unable to fulfill the traditional role of protector, the male partner may internalize his partner’s suffering, experiencing a form of secondary traumatic stress characterized by profound helplessness. 17 This psychological state—manifesting as intrusive thoughts and exhaustion—aligns with the lower Mind-Body scores observed in our cohort. Furthermore, gendered coping styles likely exacerbate this dynamic, whereas women often seek social support to manage distress, men frequently rely on emotional distancing or avoidance. 18 Consequently, the male partner’s tendency to internalize this vicarious stress, combined with the inability to actively intervene, results in a significant decline in his own well-being. Conversely, a diagnosis of male-factor infertility did not significantly impact the female partner’s QoL. We propose that for female partners, the physical reality of treatment outweighs the etiological label. Regardless of the diagnosis, the female partner undergoes the same invasive procedures; her distress is likely driven by the burden of treatment and the unfulfilled desire for a child, rather than the specific medical cause. Additionally, our analysis showed that clinical history variables (duration of infertility, previous treatment failures) were not significant predictors of QoL. This null finding is likely attributable to the “tertiary center effect.” Our cohort, characterized by advanced maternal age and low ovarian reserve, represents a population with a universally high baseline of stress. This ceiling effect suggests that the objective severity of their clinical situation diminishes the predictive power of individual historical variables. 19 , 20 These findings offer a practical application for the clinical setting by establishing infertility etiology as a risk stratification tool for psychological distress. While the diagnosis itself is a fixed clinical condition, it serves as a critical marker to guide resource allocation. Current practice often focuses psychosocial support on the partner undergoing invasive procedures (typically the female). However, our data indicate that in cases of female-factor infertility, the male partner is at significantly heightened risk of “silent” distress due to the bystander effect. Consequently, clinics should implement a targeted protocol: when a female-factor diagnosis is confirmed, clinicians should actively screen the male partner for burden and coping, rather than assuming his well-being is intact. By using the diagnosis to trigger specific dyadic check-ins, clinicians can move from a “patient-centered” model to a “couple-centered” model, ensuring that the nonmedical partner is not left unsupported. While these findings highlight the necessity of a dyadic approach to care, several limitations of the current study must be acknowledged. First, its retrospective, cross-sectional design precludes causal inference, while we modeled the directionality from diagnosis to QoL based on the APIM framework; longitudinal data would be required to confirm these pathways over time. Additionally, reliance on self-administered questionnaires introduces potential self-report biases, such as social desirability, where participants may underreport distress to minimize stigma. Second, a primary statistical limitation is the power to detect effects related to male-factor infertility. Although our cohort included 27 couples with a male-factor component (seven male-only and 20 combined-factor), this sample size is insufficient to confidently rule out a true partner effect of a male-factor diagnosis on female QoL. Therefore, the null finding for this specific pathway should be interpreted cautiously as a failure to detect an effect, rather than definitive evidence of its absence. Third, our analysis of the excluded population identified a specific selection bias related to the timing of data collection. While there was no difference in infertility etiology between participants and nonparticipants ( p = 0.832), the excluded group was significantly composed of patients returning for frozen ETs who had likely undergone oocyte retrieval before the start of male partner data collection in 2024. Because partners are less likely to attend routine monitoring or frozen transfer visits compared with the initial consultation or oocyte retrieval, dyadic data were frequently unavailable for this subgroup. Therefore, our findings are most representative of couples currently undergoing the active diagnostic and ovarian stimulation phases of ART, rather than those in the maintenance or frozen transfer phases. Additionally, while patients were aware of their etiology before assessment, we acknowledge that the cross-sectional design cannot fully disentangle the psychological impact of the diagnostic label from potential unmeasured characteristics inherent to specific patient groups (eg, specific symptom burdens or health behaviors associated with male vs female factor diagnoses). Fourth, a demographic finding warrants careful interpretation. We observed a significantly higher BMI among male partners in the “female-only” infertility group (26.2 kg/m 2 , p = 0.044). This raises the clinical hypothesis of a “hidden male factor,” as standard semen analysis cannot detect obesity-related impairments such as elevated sperm DNA fragmentation or reactive oxygen species. 20 , 21 Although our study was not designed to perform these advanced diagnostics, it is crucial to note that our sensitivity analysis confirmed this demographic difference did not confound our primary psychological findings; the asymmetrical partner effect remained significant after adjusting for male BMI (Fig. 1 ). Finally, the findings originate from a single tertiary referral center in Southern Taiwan, characterized by patients with challenging clinical profiles and potentially higher baseline stress than primary care populations. Furthermore, this sample included heterosexual couples only; findings may not generalize to LGBTQIA+ couples or single patients. In conclusion, this study demonstrates that the psychological burden of infertility is not determined solely by the physical demands of treatment, but also by the specific diagnostic context. Using the APIM framework, we identified a distinct vulnerability in male partners of women with female-factor infertility, who experience significant QoL impairment likely due to the “bystander effect.” These findings challenge the clinical assumption that the nonmedical partner is unaffected. We therefore recommend that fertility centers use the diagnosis not merely as a medical classification, but as a risk stratification tool. Specifically, confirming a female-factor diagnosis should trigger active psychological screening for the male partner, ensuring that care protocols transition from treating the individual patient to supporting the resilience of the couple as a dyad.

This study was supported by grants from the National Cheng Kung University Hospital (grant NCKUH-11303052) and the National Science and Technology Council (grant NSTC 114-2314-B-006-021-MY3). We thank all members of the Assisted Reproduction Center at National Cheng Kung University Hospital—including embryologists, counselors, and administrative staff—for their dedicated support of patient care and this study.