Introduction
Infertility constitutes a major life crisis that exerts significant psychological distress on individuals. Women, as the primary bearers of infertility-related stress and the physical burdens associated with reproductive treatments, are particularly vulnerable to such distress. Research indicates that the prevalence of psychological distress among women experiencing infertility ranges from 37% to 71%.1,2 This distress not only reflects a compromised psychological state but also disrupts treatment adherence, often leading to irrational treatment discontinuation.3 Furthermore, women who experience high levels of psychological distress tend to have lower rates of oocyte retrieval, produce fewer high-quality embryos, and experience reduced success in embryo transfer.4 Consequently, these psychological challenges substantially decrease the likelihood of achieving favorable treatment outcomes.4 Moreover, infertility psychological distress (IPD) can adversely affect intimate relationships, diminish marital satisfaction, and sometimes contribute to marital breakdown.5,6
Given its profound negative consequences, IPD has garnered increasing attention from both scholars and clinicians. High-quality reviews and clinical guidelines, such as those from the European Society of Human Reproduction and Embryology, strongly advocate for the integration of routine psychosocial care to support patients navigating the complexities of medical reproduction.7 To implement effective interventions, researchers have sought to identify the key determinants of IPD. While demographic and disease-related factors, such as older age, lower socioeconomic status, and multiple failed treatment attempts, are robustly associated with higher levels of IPD,8–10 their unchangeable nature presents significant clinical challenges.
Consequently, scholarly focus has shifted toward modifiable psychosocial factors that can be targeted in clinical practice. Among these, social support, insomnia, and health information literacy have emerged as critical domains warranting comprehensive investigation.
Social support has long been recognized as an important factor in promoting psychological well-being. Among individuals experiencing infertility, greater perceived social support has been associated with lower psychological distress.1 Support from family and friends has also been negatively associated with anxiety and depression among women experiencing infertility.11 A systematic review further showed that stronger social support was consistently associated with better mental health outcomes in women with infertility.12 Beyond psychological well-being, social support has also been examined in relation to treatment-related experiences and outcomes.13 For example, online social support has been associated with fertility-treatment outcomes, with greater engagement in online social networks reported among women with lower miscarriage rates following embryo transfer.14 Taken together, these findings indicate that perceived social support is an important psychosocial factor to consider when examining psychological distress among women undergoing fertility treatment.
Another factor relevant to IPD is insomnia, which is common among women experiencing infertility and undergoing fertility treatment. Poor sleep has been associated with greater perceived stress, anxiety, and depressive symptoms among women undergoing infertility treatment,15,16 and sleep characteristics have also been associated with ART outcomes.17,18 In women undergoing IVF, sleep quality, perceived social support, anxiety, and depressive symptoms have also been shown to be closely related.19 Evidence from another clinical population has further suggested that sleep may represent an indirect pathway linking social support with psychosomatic symptoms.20 Taken together, these findings provide a rationale for examining whether insomnia may constitute an indirect pathway in the association between perceived social support and IPD among women undergoing ART.
Health information literacy (HIL) may further shape these associations. ART treatment involves repeated exposure to complex medical information and requires patients to understand treatment procedures, evaluate available information, and make ongoing decisions about their care.21 During this process, patients receive health-related information not only from healthcare professionals but also from online sources and their surrounding support networks, making the ability to identify, understand, evaluate, and use such information especially important.22 Evidence from infertility populations has begun to support the psychological relevance of these abilities. Among women attending an initial infertility consultation, lower health literacy was associated with higher anxiety.23 In addition, research on eHealth literacy in people experiencing infertility has shown that online information is commonly used to support treatment decisions, while interaction within virtual communities contributes to the sharing, interpretation, and use of health information.24 Thus, in the context of fertility treatment, patients’ ability to manage health information is closely connected with how they access and use information and support from others.
This connection provides a basis for considering HIL in the relationship between perceived social support and psychological outcomes. Social support during ART includes emotional, informational, and practical resources, but the extent to which patients can benefit from these resources may partly depend on their ability to understand, evaluate, and use the information they receive. Studies in other patient populations provide indirect support for this possibility.
Health literacy has been positively associated with perceived social support among patients with chronic conditions,25 and Li further showed that health literacy moderated the association between chronic disease and psychological distress.26 Although these findings were obtained outside infertility care and involved broader measures of health literacy, they indicate that health literacy can be closely linked to both social resources and psychological responses to health-related burdens. It is therefore plausible that HIL may influence the strength of the associations of perceived social support with both insomnia and IPD among women undergoing ART.
The potential role of HIL may also extend to the association between insomnia and psychological distress. Previous research has shown that poor sleep and lower health literacy are both associated with depressive symptoms.27 Further evidence from a large community sample showed that health literacy and eHealth literacy were associated with sleep hygiene and mental well-being, with sleep hygiene behaviors mediating the associations between health literacy and mental well-being.28 These findings suggest that health literacy is linked to both sleep-related behaviors and psychological well-being and raise the possibility that the psychological consequences associated with insomnia may also vary according to patients’ HIL. However, this possibility has not been directly examined among women undergoing ART.
Taken together, existing evidence suggests that HIL may be relevant to how women undergoing ART use social support and respond psychologically to sleep-related difficulties, but the conditional relationships among these factors remain poorly understood in infertility care. We therefore proposed the following hypotheses: H1, perceived social support is negatively associated with IPD; H2, perceived social support is indirectly associated with IPD through insomnia; and H3, HIL moderates the associations between perceived social support and insomnia, between perceived social support and IPD, and between insomnia and IPD. The proposed conceptual model is presented in Figure 1.
Accordingly, this study examined the direct, indirect, and conditional associations among perceived social support, insomnia, HIL, and IPD in women undergoing ART. Figure 1Hypothesized moderated mediation model linking perceived social support, insomnia severity, health information literacy, and infertility psychological distress.A model showing links between social support, insomnia, health literacy and infertility psychological distress.
Materials and Methods
Study Design
A cross-sectional survey design was employed for this study. Data collection was conducted between March and August 2024 at the reproductive medicine centers of two tertiary hospitals located in Wuhan, Hubei Province, central China.
Participants and Sampling
The study aimed to recruit women currently undergoing ART. Eligible participants were consecutively enrolled based on the following inclusion criteria: 1) married women of reproductive age (20–49 years); 2) clinical diagnosis of infertility and currently undergoing IVF-ET treatment; 3) sufficient cognitive and linguistic ability to comprehend and complete the survey; and 4) provision of voluntary informed consent. The exclusion criterion was a documented clinical history of psychiatric or severe cognitive disorders.
The minimum sample size was estimated using the conventional formula for a cross-sectional survey, \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$n = {{Z_{\alpha /2}^2P\left({1 - P} \right)} \over {{d^2}}}$\end{document}, where P represents the anticipated prevalence and d represents the allowable error. Previous studies reported that the prevalence of IPD ranged from 4% to 71%.1,2,29 Therefore, an anticipated prevalence of 37% was used. With a two-sided α of 0.05 and a relative allowable error of 10% of the anticipated prevalence (d=0.1P), the estimated sample size was 655.
This number was increased by 30% to account for potentially invalid responses, resulting in a minimum target sample size of 852. As participants were consecutively recruited throughout the prespecified data-collection period, 1007 valid participants were ultimately included in the analysis.
Measures
Infertility Psychological Distress
Infertility psychological distress (IPD) was assessed using the Infertility Psychological Distress Scale.30,31 This 25-item instrument measures negative emotional responses and affective states stemming from infertility and its corresponding clinical treatments. The scale encompasses two primary dimensions and six sub-dimensions: (1) distress related to the disruption of the former self (encompassing distress over self-worth questioning, strained family relationships, and female image deficit) and (2) distress linked to the struggling present self (encompassing distress over inadequate reproductive self-efficacy, maintaining family relationships, and social identity expansion). Items are rated on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).
Total scores range from 25 to 125, with higher scores reflecting greater levels of IPD. This instrument has demonstrated excellent psychometric properties in previous validation studies.30
Social Support
Perceived social support was evaluated using the Perceived Social Support (PSS) Scale, originally developed by Zimet et al32 and culturally adapted into Chinese by Jiang.33 The 12-item scale comprises three dimensions: family support (4 items), friend support (4 items), and support from significant others (4 items). Responses are recorded on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Total scores are calculated by summing the item scores, ranging from 12 to 84.
Higher scores denote higher levels of perceived social support. The scale demonstrates strong reliability, with a Cronbach’s alpha coefficient of 0.84.
Insomnia
Insomnia severity was assessed utilizing the Insomnia Severity Index (ISI).34 This 7-item instrument evaluates the nature, severity, and impact of insomnia symptoms over the recent period. Each item is rated on a 5-point Likert scale from 0 to 4, yielding a cumulative total score between 0 and 28. Higher scores indicate more severe insomnia symptoms.
Total scores are categorized into four clinical severity tiers: no clinically significant insomnia (0–7 points), subthreshold insomnia (8–14 points), clinical insomnia of moderate severity (15–21 points), and severe clinical insomnia (22–28 points). The Chinese version of the ISI has been robustly validated and exhibits excellent reliability.35
Health Information Literacy
Health information literacy was assessed using the everyday Health Information Literacy (HIL) questionnaire, originally developed by Finnish scholars.36 The core scale assesses individuals across three essential domains: willingness to seek health information, confidence in finding and understanding health information, and the ability to critically appraise health information. The current study utilized the Chinese culturally adapted and expanded 14-item version of the HIL, developed by Luo et al.37 Items are scored on a 5-point Likert scale, with higher total scores reflecting superior health information literacy. This adapted version has demonstrated robust construct validity and internal reliability in Chinese contexts.
Covariates
The following sociodemographic and clinical characteristics were prespecified as covariates: marital status, age, highest educational attainment, employment status, monthly family income, duration of infertility (in years), and cause of infertility. All seven prespecified covariates were retained in the mediation and moderated mediation models regardless of their statistical significance.
Data Collection
Data were collected utilizing an electronic questionnaire administered via Wenjuanxing, a standardized online survey platform in China. Participants were recruited consecutively on-site within the waiting areas of the ultrasound and injection departments at the respective reproductive medicine centers. Trained researchers rigorously explained the study’s purpose, procedures, and confidentiality protocols to potential participants.
Eligible women then scanned a designated Quick Response (QR) code using their smartphones to access the survey. The first page of the questionnaire provided information about the study purpose and procedures, the voluntary nature of participation, confidentiality, and participants’ right to withdraw. Participants could proceed to the questionnaire only after providing informed consent.
The electronic questionnaire was programmed with mandatory response settings to prevent missing data. Furthermore, to ensure data integrity, survey access was restricted to a single submission per IP address. Following data retrieval, researchers independently screened the submissions.
Questionnaires exhibiting logical inconsistencies, abnormally short completion times, or uniform response patterns (eg, selecting the identical option for all consecutive items) were systematically excluded prior to analysis.
Statistical Analysis
Data were analyzed using SPSS 24.0 and PROCESS 4.1. Descriptive statistics were used to summarize participant characteristics, with categorical variables presented as frequencies and percentages and continuous variables as means and standard deviations. Bivariate associations among PSS, ISI, HIL, and IPD were examined using Pearson correlation coefficients.
Hayes’s PROCESS macro was used to examine the hypothesized conditional process model.38 Model 4 was first used to estimate the statistical mediation of ISI in the association between PSS and IPD. Model 59 was subsequently used to examine whether HIL moderated the PSS–ISI, PSS–IPD, and ISI–IPD associations. Prior to constructing the interaction terms in Model 59, PSS, ISI, and HIL were mean-centered, and the variables were not standardized.
Indirect and conditional indirect effects were estimated using 5000 bootstrap samples with 95% bootstrap confidence intervals (CIs), with an effect considered statistically significant when the CI did not include zero. Statistical significance for regression coefficients and interaction terms was evaluated using two-sided tests with a significance level of P<0.05. Model R2 values were reported to indicate explained variance, and the incremental R2 (ΔR2) associated with each interaction term was used to quantify the additional variance explained by moderation.
Variance inflation factors (VIFs) were examined to assess multicollinearity. All seven prespecified covariates—marital status, age, highest educational attainment, employment status, monthly family income, duration of infertility, and cause of infertility—were included and retained in both the mediation and moderated mediation models irrespective of statistical significance.
Results
Sample Characteristics
Table 1 summarizes the demographic and clinical characteristics of the study participants (N = 1007). More than half (53.53%) were women aged 30–35 years, 79% were in their first marriage, and 67.83% had been experiencing infertility for 1 to 3 years.Table 1Demographic and Clinical Characteristics of the Participants (N=1007)Variablesn (%)Variablesn (%)Marital statusEmployment status First marriage803 (79.74) Employed427 (42.40) Remarriage204 (20.26) Freelance work320 (31.78) Unemployed/ Separated260 (25.82)AgeHighest educational attainment <30203 (20.16) Middle school education or less207 (20.56) 30–35539 (53.53) High school education180 (17.87) 36–42238 (23.63) Bachelor or college education522 (51.84) >4327 (2.68) Master education or more98 (9.73)Duration of infertilityMonthly family income Less than 1 year50 (4.97) ≤5000 RMB374 (37.14) 1–3 years683 (67.83) 5001–8000 RMB336 (33.37) More than 3 years274 (27.21) >8000 RMB297 (29.49)Cause of infertility Male factor88 (8.74) Female factor474 (47.07) Mixed-factor175 (17.38) Unexplained270 (26.81)
Correlation Analysis of Study Variables
The bivariate correlations among the study variables are detailed in Table 2. PSS was negatively correlated with both IPD (r = −0.254, p < 0.001) and ISI (r = −0.254, p < 0.001), and positively correlated with HIL (r = 0.322, p < 0.001). ISI was positively correlated with IPD (r = 0.388, p < 0.001).
No significant correlations were observed between HIL and either ISI (r = −0.047, p > 0.05) or IPD (r = −0.014, p > 0.05).Table 2Correlations Between the Study Variables (N=1007)12341. Infertility psychological distress12. Perceived social support−0.254**13.
Insomnia severity Index0.388**−0.254**14. Health information literacy−0.0140.322**−0.0471Note: **P<0.01.
Mediation Analysis
As shown in Table 3, PSS was negatively associated with IPD (B = −0.364, p < 0.001) and ISI (B = −0.094, p < 0.001). When PSS and ISI were included simultaneously in the IPD model, PSS remained negatively associated with IPD (B = −0.231, p < 0.001), whereas ISI was positively associated with IPD (B = 1.401, p < 0.001). The models accounted for 8.6% of the variance in IPD for the total association model (R2 = 0.086), 7.0% of the variance in ISI (R2 = 0.070), and 19.4% of the variance in IPD when PSS and ISI were included simultaneously (R2 = 0.194).
Table 4 presents the total, direct, and indirect effects. The total effect of PSS on IPD was −0.364 (95% bootstrap CI: [−0.460, −0.267]), comprising a direct effect of −0.231 (95% bootstrap CI: [−0.325, −0.138]) and an indirect effect mediated through ISI of −0.132 (95% bootstrap CI: [−0.175, −0.094]). Because the bootstrap confidence interval for the indirect effect did not include zero, the indirect association through ISI was statistically significant.Table 3Regression Results for the Mediation Model (N=1007)PredictorsModel 1(IPD)Model 2 (ISI)Model 3 (IPD)B95% CItB95% CItB95% CItPSS−0.364[−0.460, −0.267]−7.368***−0.094[−0.118, −0.071]−7.792***−0.231[−0.325, −0.138]−4.844***ISI1.401[1.164,1.639]11.573***R20.0860.0700.194F11.671***9.431***26.638***Notes: All models were adjusted for marital status, age, highest educational attainment, employment status, monthly family income, duration of infertility, and cause of infertility. ***P<0.001.Abbreviations: B, unstandardized regression coefficient; CI, confidence interval; IPD, Infertility psychological distress; PSS, perceived social support; ISI, Insomnia Severity IndexTable 4Direct and Indirect Associations of Perceived Social Support with Infertility Psychological Distress Through Insomnia Severity (N=1007)EstimateBoot SEBootstrap 95% CIBoot LLCIBoot ULCITotal effect−0.3640.049−0.460−0.267Direct effect−0.2310.048−0.325−0.138Indirect effect−0.1320.020−0.175−0.094Notes: All models were adjusted for marital status, age, highest educational attainment, employment status, monthly family income, duration of infertility, and cause of infertility.Abbreviation: CI, confidence interval.
Moderated Mediation Analysis
Table 5 and Figure 2 present the results of the moderated mediation analysis. All seven prespecified covariates were retained in the models. VIFs ranged from 1.017 to 1.651, indicating no evidence of problematic multicollinearity.
The model for ISI explained 8.4% of the variance (R2 = 0.084), whereas the model for IPD explained 21.2% (R2 = 0.212). Significant interactions were observed between PSS and HIL in relation to ISI (B = −0.008, p < 0.001, ΔR2 = 0.012), between PSS and HIL in relation to IPD (B = −0.029, p < 0.001, Δ R2 = 0.010), and between ISI and HIL in relation to IPD (B = −0.071, p < 0.01, Δ R2 = 0.007). These findings indicate that HIL significantly moderated all three associations specified in the conditional process model.
Figure 2Conditional process model of perceived social support, insomnia severity, health information literacy, and infertility psychological distress.Path diagram for a conditional process model showing directional associations between perceived social support, insomnia severity index, health information literacy, and infertility psychological distress, with unstandardized regression coefficients and significance asterisks on each arrow.Note: Values shown are unstandardized regression coefficients. **P<0.01; ***P<0.001.Table 5Results of the Conditional Process Analysis (N=1007)PredictorsModel 1 (ISI)Model 2 (IPD)B95% CItB95% CItPSS−0.106[−0.131, −0.081]−8.263***−0.307[−0.407, −0.208]−6.064***HIL0.059[−0.012, 0.130]1.6260.343[0.072, 0.615]2.482*PSS×HIL−0.008[−0.012, −0.003]−3.590***−0.029[−0.045, −0.013]−3.481***ISI1.328[1.091, 1.566]10.979***ISI×HIL−0.071[−0.118, −0.024]−2.968**R20.0840.212F9.121***22.322***Note: PSS, ISI, and HIL were mean-centered before the interaction terms were created. All seven prespecified covariates were retained in both models. The interaction terms accounted for ΔR2=0.012 for PSS×HIL in Model 1, ΔR2=0.010 for PSS×HIL in Model 2, and ΔR2=0.007 for ISI×HIL in Model 2. *P<0.05; **P<0.01; ***P<0.001.Abbreviation: B, unstandardized regression coefficient; CI, confidence interval; IPD, Infertility psychological distress; PSS, perceived social support; ISI, Insomnia Severity Index; HIL, Health information literacy.
Simple slope analyses were conducted to further characterize these interactions. As shown in Figure 3, PSS was negatively associated with IPD at both low HIL (1 SD below the mean; B = −0.172, t = −2.942, p < 0.05) and high HIL (1 SD above the mean; B = −0.443, t = −6.428, p < 0.001), with a stronger negative association at higher HIL. For the ISI–IPD association (Figure 4), ISI was positively associated with IPD at both low HIL (B = 1.660, t = 10.390, p < 0.001) and high HIL (B = 0.997, t = 5.882, p < 0.001), with a weaker positive association at higher HIL.
Similarly, as shown in Figure 5, PSS was negatively associated with ISI at both low HIL (B = −0.071, t = −4.664, p < 0.001) and high HIL (B = −0.141, t = −8.251, p < 0.001), with a stronger negative association at higher HIL. Figure 3Interaction between health information literacy and perceived social support in relation to infertility psychological distress.Interaction line plot: X-axis: low and high Insomnia Severity Index (ISI); Y-axis: infertility psychological distress (IPD). Green dashed line = low health information literacy (HIL, 1 SD below mean); blue solid line = high HIL (1 SD above mean).
IPD rises with higher ISI for both HIL groups, and the increase is more pronounced in the low HIL group.Note: Low and high HIL represent 1 SD below and 1 SD above the mean, respectively.Abbreviations: HIL, health information literacy; PSS, perceived social support; IPD, infertility psychological distress.Figure 4Interaction between health information literacy and insomnia severity in relation to infertility psychological distress.Interaction line plot: X-axis is low and high perceived social support (PSS); Y-axis is Insomnia Severity Index (ISI). Green dashed line = low health information literacy (HIL, 1 SD below mean); blue solid line = high HIL (1 SD above mean). ISI decreases with higher PSS in both HIL groups, and the decline is steeper for high HIL.Note: Low and high HIL represent 1 SD below and 1 SD above the mean, respectively.Abbreviations: HIL, health information literacy; ISI, Insomnia Severity Index; IPD, infertility psychological distress.Figure 5Interaction between health information literacy and perceived social support in relation to insomnia severity.Interaction line plot: X-axis shows low and high perceived social support (PSS); Y-axis shows Insomnia Severity Index (ISI).
Green dashed line indicates low health information literacy (HIL, 1 SD below mean); blue solid line indicates high HIL (1 SD above mean). ISI declines with increasing PSS in both groups, and the decline is steeper for high HIL.Note: Low and high HIL represent 1 SD below and 1 SD above the mean, respectively.Abbreviations: HIL, health information literacy; PSS, perceived social support; ISI, Insomnia Severity Index.
As shown in Table 6, the conditional indirect effect of PSS on IPD through ISI was statistically significant at low HIL (effect = −0.117, 95% bootstrap CI [−0.179, −0.063]), mean HIL (effect = −0.141, 95% bootstrap CI [−0.184, −0.101]), and high HIL (effect = −0.141, 95% bootstrap CI [−0.204, −0.085]).Table 6Conditional Direct and Indirect Associations at Different Levels of Health Information Literacy (N=1007)Health Information LiteracyEstimateBoot SEBootstrap 95% CIBoot LLCIBoot ULCIDirect effect41.75 (M-1SD)−0.1720.058−0.287−0.05746.41 (M)−0.3070.051−0.407−0.20851.07 (M+1SD)−0.4430.069−0.578−0.308The mediating effect of Insomnia severity Index41.75 (M-1SD)−0.1170.030−0.179−0.06346.41 (M)−0.1410.021−0.184−0.10151.07 (M+1SD)−0.1410.030−0.204−0.085Notes: HIL values are shown on the original scale; HIL was mean-centered for the interaction analyses. Bootstrap confidence intervals for the conditional indirect associations were based on 5000 bootstrap samples.
Regarding the prespecified covariates, none was significantly associated with ISI. In the IPD model, monthly family income was significantly associated with IPD (B = −1.492, p = 0.027), whereas the remaining covariates were not statistically significant.
Discussion
This study examined the associations among PSS, ISI, IPD, and HIL in women undergoing ART. PSS was negatively associated with IPD, and a statistically significant indirect association through ISI was observed. HIL further modified all three component associations in the conditional process model, namely, the associations between PSS and ISI, between PSS and IPD, and between ISI and IPD.
Specifically, higher HIL was associated with stronger negative associations of PSS with both ISI and IPD, whereas the positive association between ISI and IPD was weaker at higher levels of HIL. These findings extend previous research by showing that the associations among social support, insomnia, and psychological distress may vary according to patients’ HIL.
The negative association between PSS and IPD is consistent with previous studies showing that greater perceived social support is associated with lower psychological distress among women experiencing infertility.1,11 Our findings further showed a significant indirect association between PSS and IPD through insomnia. Previous research among women undergoing IVF has demonstrated close relationships among sleep quality, perceived social support, anxiety, and depressive symptoms.19 Evidence from other clinical populations has also suggested that sleep may constitute an indirect pathway linking social support with psychosomatic symptoms.20 One possible explanation is that social support may help reduce treatment-related worry, rumination, and emotional arousal, which are closely related to sleep disturbance, while lower insomnia severity may in turn be associated with less psychological distress. However, because the present data were collected cross-sectionally, this interpretation should be regarded as a possible explanation of the observed associations rather than evidence of a temporal mechanism.
A notable finding of this study was the conditional role of HIL. The simple slope analyses showed that the negative associations of PSS with both IPD and ISI were stronger at higher levels of HIL, whereas the positive association between ISI and IPD was weaker. This pattern is consistent with evidence from other health contexts showing that health literacy can modify the relationship between health-related conditions and psychological distress.26 In infertility care, HIL may be particularly relevant because patients receive health-related information from healthcare professionals, online sources, and their social networks, and therefore differ in their ability to evaluate and use the information embedded in these sources of support.
Research among people experiencing infertility has shown that online information and virtual communities are frequently used in treatment-related decision making and that the ability to critically evaluate such information forms an important part of eHealth literacy.24 Higher HIL may therefore facilitate the interpretation and use of informational and practical support during ART, which could partly explain why the associations of PSS with insomnia and IPD were stronger among women with higher HIL.
The weaker association between insomnia and IPD at higher levels of HIL may reflect a related process. Previous studies have linked both poor sleep and lower health literacy with poorer psychological outcomes,27 while health literacy and eHealth literacy have also been associated with sleep-related behaviors and mental well-being.28 Patients with greater HIL may be better able to understand sleep-related health information, distinguish reliable advice from misinformation, and use appropriate resources when sleep problems occur. Nevertheless, this explanation remains tentative because the present study did not directly assess how participants used health information in response to insomnia.
Importantly, HIL was not significantly correlated with either ISI or IPD at the bivariate level. This suggests that, in the present data, HIL may be more relevant as a factor modifying the associations among social support, insomnia, and psychological distress than as an independent correlate of insomnia or IPD.
These findings have implications for psychosocial assessment during ART, but they should not be interpreted as evidence that increasing HIL will directly reduce psychological distress. Rather, the results support considering perceived social support, sleep problems, and HIL together when identifying patients who may have different psychosocial needs during treatment. Brief assessment of HIL may help identify patients who have difficulty understanding, evaluating, or using health information, while psychoeducational approaches designed to strengthen these abilities could be evaluated alongside social support and sleep-related interventions in future studies.
Longitudinal and intervention research is needed to determine whether improving HIL changes the extent to which patients benefit from social support or cope with sleep-related difficulties.
Several limitations should be considered. First, the cross-sectional design does not allow the temporal ordering of PSS, insomnia, HIL, and IPD to be established, and reverse associations cannot be excluded. For example, greater psychological distress may contribute to poorer sleep or lower perceived social support, rather than these factors occurring only in the direction specified in the statistical model.
The indirect and conditional associations observed in this study should therefore not be interpreted as causal pathways. Longitudinal studies are needed to examine the temporal sequence of these relationships, and intervention studies could further test whether changes in social support, sleep, or HIL are followed by changes in psychological distress. Second, participants were recruited from two tertiary hospitals in Wuhan, which may limit the generalizability of the findings.
Women receiving care in lower-tier facilities, rural areas, other regions, or different sociocultural settings may differ in their social support networks, access to health information, sleep problems, and psychological experiences. Multicenter studies involving more diverse clinical and geographic settings are therefore warranted. Despite these limitations, this study provides evidence that the associations among perceived social support, insomnia, and IPD may differ according to HIL in women undergoing ART.
The findings provide a basis for further longitudinal and intervention studies examining whether HIL influences how patients use social support and respond to sleep-related difficulties during fertility treatment.
Conclusions
In summary, this study identified a significant indirect association between perceived social support and IPD through insomnia severity among women undergoing ART. HIL significantly moderated the associations between perceived social support and insomnia, between perceived social support and IPD, and between insomnia and IPD. These findings suggest that the relationships among social support, sleep problems, and psychological distress may vary according to patients’ HIL.
The results support considering social support, insomnia, and HIL together in psychosocial assessment during ART and provide a basis for future longitudinal and intervention studies to test whether HIL-focused support can improve psychological outcomes.