Brain activation, sleep disturbances, and psychological distress in women undergoing assisted reproductive treatment: a functional near-infrared spectroscopy study.

A total of 80 participants were included after data quality assessment (38 infertility patients and 42 healthy controls). The demographic and clinical characteristics of both groups are summarized in Table  1 . Infertility patients reported significantly higher AIS-8 and GAD-7 scores compared to healthy controls ( P  < 0.001), while other demographic parameters did not differ significantly between groups. Table 1 Demographic and clinical characteristics of participants Variables Patient group ( n  = 38) Healthy group ( n  = 42) Confidence Interval (Lower, Upper) P value Age (years) 35.55 ± 4.87 37.10 ± 6.02 (-3.995, 0.910) 0.214 Education (years) 13.42 ± 3.25 13.71 ± 2.92 (-1.862, 1.290) 0.759 AIS-8 scores 6.11 ± 3.38 1.43 ± 1.53 (3.479, 5.874) < 0.001** GAD-7 scores 3.68 ± 3.52 1.26 ± 1.36 (1.198, 3.646) < 0.001** PHQ-9 scores 3.32 ± 3.09 2.69 ± 2.15 (-0.551, 1.802) 0.293 Duration of infertility (years) 4.67 ± 4.62 N/A N/A N/A Type of infertility, n (%)  Primary infertility 18 (47.4%) N/A N/A N/A  Secondary infertility 19 (50.0%) N/A N/A N/A  Other 1 (2.6%) N/A N/A N/A History of childbirth, n (%)  Yes 10 (26.3%) N/A N/A N/A  No 28 (73.7%) N/A N/A N/A Etiology of infertility, n (%)  Female factor 25 (65.8%) N/A N/A N/A  Male factor 4 (10.5%) N/A N/A N/A  Combined factor 4 (10.5%) N/A N/A N/A  Unexplained infertility 5 (13.2%) N/A N/A  N/A Ovarian stimulation protocol, n (%)   Antagonist protocol 23 (60.5%) N/A N/A N/A  Long protocol 3 (7.9%) N/A N/A N/A  Other protocols 12 (31.6%) N/A N/A N/A  Number of previous ART cycles 2.08 ± 1.22 N/A N/A N/A Current ART protocol, n (%)  IVF 26 (68.4%) N/A N/A N/A  ICSI 7 (18.4%) N/A N/A N/A  Other/Combined protocols 5 (13.2%) N/A N/A N/A Area of residence, n (%)  Urban 19 (50.0%) N/A N/A N/A  Rural 19 (50.0%) N/A N/A N/A ART assisted reproductive technology, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, N/A not applicable ** p  < 0.01 Demographic and clinical characteristics of participants ART assisted reproductive technology, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, N/A not applicable ** p  < 0.01 Figures 3 and 4 depict channel-level and regions of interest (ROI)-level cortical hemodynamic responses. Patients showed significantly lower oxygenated hemoglobin levels in Broca's area and the frontal pole area compared to controls. Fig. 3 Channel-wise hemodynamic responses during the verbal fluency task. The curves display the dynamic changes in HbO and HbR concentrations across individual measurement channels during the verbal fluency task for the infertility and healthy control groups. Infertility group: HbO (red), HbR (blue). Healthy control group: HbO (green), HbR (yellow). The x-axis represents task time, and the y-axis indicates the relative change in hemoglobin concentration Channel-wise hemodynamic responses during the verbal fluency task. The curves display the dynamic changes in HbO and HbR concentrations across individual measurement channels during the verbal fluency task for the infertility and healthy control groups. Infertility group: HbO (red), HbR (blue). Healthy control group: HbO (green), HbR (yellow). The x-axis represents task time, and the y-axis indicates the relative change in hemoglobin concentration Fig. 4 ROI-level hemodynamic responses during the verbal fluency task. A Right Broca's area (Broca-R). B Left Broca's area (Broca-L). C Left dorsolateral prefrontal cortex (DLPFC-L). D Right dorsolateral prefrontal cortex (DLPFC-R). E Left frontal pole area (FPA-L). F Right frontal pole area (FPA-R). The curves depict the averaged dynamic changes in HbO and HbR concentrations within these regions of interest (ROIs) during the verbal fluency task for the infertility and healthy control groups. Infertility group: HbO (red), HbR (blue). Healthy control group: HbO (green), HbR (yellow). The x-axis represents task time, and the y-axis indicates the relative change in hemoglobin concentration ROI-level hemodynamic responses during the verbal fluency task. A Right Broca's area (Broca-R). B Left Broca's area (Broca-L). C Left dorsolateral prefrontal cortex (DLPFC-L). D Right dorsolateral prefrontal cortex (DLPFC-R). E Left frontal pole area (FPA-L). F Right frontal pole area (FPA-R). The curves depict the averaged dynamic changes in HbO and HbR concentrations within these regions of interest (ROIs) during the verbal fluency task for the infertility and healthy control groups. Infertility group: HbO (red), HbR (blue). Healthy control group: HbO (green), HbR (yellow). The x-axis represents task time, and the y-axis indicates the relative change in hemoglobin concentration A general Linear Model analysis revealed significantly reduced brain activation (β values) during the verbal fluency task (VFT) in infertility patients compared to healthy controls (FDR-corrected; Fig.  5 ). The patient group exhibited smaller activation areas relative to controls. Fig. 5 Topographic cortical activation maps (FDR-corrected). This figure presents topographic maps of brain activation t-values based on fNIRS channel data, generated from one-sample t-tests (FDR-corrected) and mapped onto a 3D head model. The color bar indicates activation strength, with deeper red representing stronger positive activation. Panels A and B show the spatial patterns of brain activation in the infertility patient group and the healthy control group, respectively Topographic cortical activation maps (FDR-corrected). This figure presents topographic maps of brain activation t-values based on fNIRS channel data, generated from one-sample t-tests (FDR-corrected) and mapped onto a 3D head model. The color bar indicates activation strength, with deeper red representing stronger positive activation. Panels A and B show the spatial patterns of brain activation in the infertility patient group and the healthy control group, respectively Independent sample t-tests indicated significant differences between patient and control groups for β , integral, mean, and peak HbO values (Fig.  6 , FDR-corrected). No significant differences were observed for the slope values. ROI-level analyses (Fig.  7 ) further confirmed reduced activation in infertility patients, specifically in the frontal pole area, Broca’s area, and DLPFC. Fig. 6 Statistical maps of cortical activation differences (FDR-corrected) between groups during the verbal fluency task. Numbers in the map represent channel locations. Color intensity indicates the t‑value from independent two‑sample t‑tests, with red denoting that the healthy control group exhibited higher values than the infertility patient group for a given metric at the corresponding channel. The subpanels display the spatial distribution of group differences for the following parameters: A β value; B mean value; C integral value; D peak value; E centroid value Statistical maps of cortical activation differences (FDR-corrected) between groups during the verbal fluency task. Numbers in the map represent channel locations. Color intensity indicates the t‑value from independent two‑sample t‑tests, with red denoting that the healthy control group exhibited higher values than the infertility patient group for a given metric at the corresponding channel. The subpanels display the spatial distribution of group differences for the following parameters: A β value; B mean value; C integral value; D peak value; E centroid value Fig. 7 Mean differences in fNIRS parameters at the ROI level between groups. The bar graph displays the means and standard deviations of fNIRS parameters across brain regions for both the infertility and healthy control groups. A β values. B Mean values. C Integral values. D Peak values.* indicates p < 0.05, ** indicates p < 0.01, ns indicates not significant Mean differences in fNIRS parameters at the ROI level between groups. The bar graph displays the means and standard deviations of fNIRS parameters across brain regions for both the infertility and healthy control groups. A β values. B Mean values. C Integral values. D Peak values.* indicates p < 0.05, ** indicates p < 0.01, ns indicates not significant Significant correlations between fNIRS-derived measures and psychological scores (AIS-8 and GAD-7) are shown in Fig. 8 . In infertility patients, higher anxiety scores (GAD-7) correlated with lower integral and mean HbO values and slower response times (slope values). Poorer sleep quality (higher AIS-8 scores) correlated with posterior shifts in centroid values. Similar patterns, though weaker, were observed in healthy controls. Fig. 8 Correlation between psychological scale scores and fNIRS indices in infertility patients and healthy controls. A Patient group: integral values vs GAD scores. B Patient group: slope values vs GAD scores. C Patient group: mean values vs GAD scores. D Patient group: centroid values vs AIS scores. E Healthy control group: slope values vs GAD scores. F Healthy control group: centroid values vs GAD scores. Numbers in the map represent channel locations. Color indicates the correlation coefficient (see color bar), with deeper red representing stronger positive correlations and deeper blue representing stronger negative correlations. Only channels with statistically significant correlations ( p 0.05) are set to zero and thus masked Correlation between psychological scale scores and fNIRS indices in infertility patients and healthy controls. A Patient group: integral values vs GAD scores. B Patient group: slope values vs GAD scores. C Patient group: mean values vs GAD scores. D Patient group: centroid values vs AIS scores. E Healthy control group: slope values vs GAD scores. F Healthy control group: centroid values vs GAD scores. Numbers in the map represent channel locations. Color indicates the correlation coefficient (see color bar), with deeper red representing stronger positive correlations and deeper blue representing stronger negative correlations. Only channels with statistically significant correlations ( p 0.05) are set to zero and thus masked Further analyses within the infertility group were conducted. First, we examined correlations with infertility duration, finding significant associations with posterior shifts in centroid values (channels 17, 20, 21) and slower responses (channels 36, 37, 43, 48; Fig.  9 ). Second, an exploratory subgroup analysis was performed to compare primary and secondary infertility patients, acknowledging the limited sample size in each subgroup. This analysis revealed significant differences in AIS-8 and GAD-7 scores and task-related slope and peak HbO values, with primary infertility patients exhibiting greater psychological distress and lower neural activation (Tables  2 and 3 ). Fig. 9 Correlation between infertility duration and cortical activation parameters in infertility patients. A Slope values vs infertility duration. B Centroid values vs infertility duration. Numbers in the map represent channel locations. Color indicates the correlation coefficient (see color bar), with deeper red representing stronger positive correlations and deeper blue representing stronger negative correlations. Only channels with statistically significant correlations ( p 0.05) are set to zero and thus masked Correlation between infertility duration and cortical activation parameters in infertility patients. A Slope values vs infertility duration. B Centroid values vs infertility duration. Numbers in the map represent channel locations. Color indicates the correlation coefficient (see color bar), with deeper red representing stronger positive correlations and deeper blue representing stronger negative correlations. Only channels with statistically significant correlations ( p 0.05) are set to zero and thus masked Table 2 Psychological scale scores by infertility characteristics Grouping Criteria Score Type Group n Mean ± SD 95% Confidence Interval P value Types of Infertility AIS Secondary 19 5.11 ± 3.31 (-4.387, -0.069) 0.043* Primary 18 7.33 ± 3.14 GAD Secondary 19 2.47 ± 2.65 (-4.662, -0.058) 0.045* Primary 18 4.83 ± 4.02 PHQ Secondary 19 2.63 ± 2.85 (-3.439, 0.702) 0.188 Primary 18 4.00 ± 3.34 Number of Assisted Reproductive Cycles AIS First 15 6.27 ± 3.61 (-2.201, 2.461) 0.910 Multiple 22 6.32 ± 4.04 GAD First 15 3.93 ± 4.23 (-2.141, 2.734) 0.806 Multiple 22 3.64 ± 3.08 PHQ First 15 3.33 ± 3.70 (-2.249, 2.006) 0.909 Multiple 22 3.45 ± 2.69 Area of residence AIS City 19 5.89 ± 2.64 (-2.668, 1.826) 0.706 Rural 19 6.32 ± 4.04 GAD City 19 3.68 ± 3.13 (-2.348, 2.348) 1.000 Rural 19 3.68 ± 3.96 PHQ City 19 3.05 ± 2.46 (-2.583, 1.53) 0.607 Rural 19 3.58 ± 3.67 * p  < 0.05 Psychological scale scores by infertility characteristics * p  < 0.05 Table 3 Significant differences in task-related fNIRS parameters between infertility subgroups Indicator ROI Secondary infertility ( n  = 19) Primary infertility ( n  = 18) 95% Confidence Interval P value Peak DLPFC-L 0.16 ± 0.07 0.08 ± 0.09 (0.0225, 0.1317) 0.007** Broca-R 0.39 ± 0.24 0.25 ± 0.14 (0.0150, 0.2779) 0.030* Slope DLPFC-L 0.0014 ± 0.0016 -0.0004 ± 0.0021 (0.0006, 0.0032) 0.005** FPA-R 0.0028 ± 0.0020 0.0014 ± 0.0015 (0.0002, 0.0026) 0.022* * p  < 0.05, ** p  < 0.01 Significant differences in task-related fNIRS parameters between infertility subgroups * p  < 0.05, ** p  < 0.01

This study employed a cross-sectional design conducted at the Reproductive Medicine Center of the Third Affiliated Hospital of Guangzhou Medical University between December 2024 and March 2025. The sample size was estimated based on anticipated group differences in cortical activation. Previous fNIRS studies have consistently reported medium to large effect sizes for prefrontal activation differences [ 24 ]. Assuming a medium effect size (Cohen’s d = 0.8), a power analysis with a two-tailed alpha of 0.05 and 80% power indicated that a minimum of 26 participants per group was required. To account for potential data attrition and exclusions, an initial target of 45 participants per group was set. Participants were recruited using convenience sampling. Based on this sample size calculation, a total of 90 female participants were enrolled, including 45 infertility patients undergoing ART treatment and 45 age-matched healthy controls. Patients were recruited directly from the outpatient population of the reproductive medicine center, while healthy controls were enrolled through local community advertisements. As the study focused on the acute neurobiological state of infertility patients during the critical “trigger day” of ART treatment, the menstrual cycle phase was not matched between the two groups. Participants were required to meet the following inclusion criteria. For the infertility group: (1) clinical diagnosis of infertility and currently undergoing ART treatment; (2) age ≥ 20 years; (3) ≥ 6 years of formal education; (4) right-handed; (5) no history of psychiatric disorders, cardiovascular diseases, or neurological conditions; and (6) no current use of psychotropic medications. Healthy controls met the same criteria except for having no history of infertility or ART treatment. All participants were excluded if they met any of the following conditions: (1) currently receiving systematic psychiatric or psychological treatment; (2) current use of anxiolytics, antidepressants, or sedative-hypnotics; (3) history of traumatic brain injury; or (4) any contraindications to fNIRS testing. Following data collection, 10 participants were excluded based on predefined quality control criteria. Specifically, 7 were from the patient group (3 withdrew due to personal reasons, 4 failed to complete all assessments) and 3 from the healthy control group (2 withdrew informed consent, 1 had incomplete data). Consequently, data from 80 participants (38 patients, 42 controls) were included in the final statistical analysis. The detailed recruitment and screening process is illustrated in Fig.  1 . Fig. 1 Study design and participant flow diagram Study design and participant flow diagram Figure  1 illustrates the study design and participant flow. The study was approved by the Ethics Committee of the Third Affiliated Hospital of Guangzhou Medical University (Approval No. 2024 − 163) and strictly adhered to the ethical principles of the Declaration of Helsinki. All participants provided written informed consent. Infertility patients underwent all assessments on the ART “trigger day”; healthy controls were assessed at corresponding time points. The evaluation followed a standardized procedure: all participants first completed general demographic questionnaires and psychological scale assessments, followed by fNIRS brain function monitoring while performing the Chinese Verbal Fluency Task(VFT). Three validated psychometric instruments were administered for assessment. The Athens Insomnia Scale-8 (AIS-8) was used to evaluate insomnia severity; this scale consists of 8 items scored from 0 to 3, with a total score range of 0–24. According to established clinical thresholds, a total score ≥ 6 indicates clinically significant insomnia [ 25 ]. The Generalized Anxiety Disorder-7 (GAD-7) scale was employed to assess anxiety symptoms; it comprises 7 items rated on a 4-point Likert scale (0="not at all” to 3="nearly every day”), with a total score ≥ 5 suggesting clinically significant anxiety [ 26 ]. The Patient Health Questionnaire-9 (PHQ-9) was utilized to measure depressive symptoms; this instrument contains 9 items similarly scored from 0 to 3, and based on established guidelines, a total score ≥ 5 indicates clinically significant depression [ 27 ]. VFT was adopted as the cognitive activation paradigm. The task protocol consisted of a 30-second pre-task baseline period (counting from 1 in response to auditory cues), a 60-second active task period (generating words beginning with the Chinese syllables “shang,” “shi,” “shuo,” and “jia,” with 15 s allocated per syllable), and a 30-second post-task baseline period. Participants were instructed to avoid word repetitions and proper nouns. All participants received syllable prompts in the same fixed order. The VFT was selected due to its established sensitivity to prefrontal executive function and its ability to elicit robust hemodynamic responses in fNIRS studies [ 28 ]. Although the task does not directly measure infertility-related stress, it effectively engages neural circuits involved in emotional regulation, making it suitable for investigating the impact of psychological distress on cognitive-neural function. Hemodynamic responses were measured using a 53-channel continuous-wave fNIRS system (BS-3000, Wuhan Znion Medical Technology Co., China) with wavelengths at 690 and 830 nm, a sampling rate of 20 Hz, and a source-detector distance of 3 cm. Probe placement followed the international 10–20 EEG system with source 9 positioned at Fpz. Channel positions and anatomical landmarks were digitized with a 3D digitizer (NirMap, Wuhan Znion), and data were projected to the Montreal Neurological Institute (MNI) space using NIRS-SPM [ 29 ]. Channels were categorized into five cortical regions: premotor and supplementary motor area (PreM & SMA), frontal eye fields (FEF), Broca’s area, dorsolateral prefrontal cortex (DLPFC), and frontal pole area (FPA). Figure 2 shows the spatial distribution of fNIRS channels and their corresponding cortical regions. Channels located in the PreM & SMA, FEF, and midline regions (channels 25, 28, 29) were excluded from subsequent analyses, as these areas are not primarily involved in the verbal fluency task employed in this study, consistent with previous fNIRS research protocols [ 30 ]. Fig. 2 Spatial distribution of the 53 fNIRS channels on the cerebral cortex. Channels (spheres) are numbered and grouped into five regions of interest (PreM & SMA, FEF, Broca’s area, DLPFC, FPA) in the left and right hemispheres Spatial distribution of the 53 fNIRS channels on the cerebral cortex. Channels (spheres) are numbered and grouped into five regions of interest (PreM & SMA, FEF, Broca’s area, DLPFC, FPA) in the left and right hemispheres Data preprocessing was performed using MATLAB 2020b and HOMER2 software [ 31 ]. Processing steps included converting raw intensity data into optical density changes, computing oxygenated and deoxygenated hemoglobin concentrations using the Modified Beer-Lambert Law, first-order detrending, and motion artifact removal with a 5-second moving average window [ 32 ]. Signal quality was assessed via the coefficient of variation (CV), calculated as: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:CV{\%}_{tw}\left(t\right)=\frac{{STD}_{tw}\left(t\right)}{{AVG}_{tw}\left(t\right)}\cdot\:100\%>Th$$\end{document} Where represents the time window to calculate CV, represents the standard deviation of the signal in the time window, represents the mean value of the signal in the time window, represents the time, and represents the exclusion threshold.In the analysis, was set to 25%. Channels with the CV value greater than 25% were marked as bad channels. For every participant, when the number of bad channels exceeds 20% of the total, the data was eliminated. A General Linear Model was applied to oxygenated hemoglobin concentration data using NIRS-KIT [ 33 ], generating β values to quantify task-related brain activation. Additional measures included peak (maximum HbO concentration), mean HbO concentration, and integral values (area under the positive HbO curve during the task period), calculated as: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Integral=\int_{\mathrm{0}}^{\mathrm{60}s}HbO(t)dt,\:\:\:HbO(t)\:>0$$\end{document} The centroid value, representing the temporal midpoint of activation, was also calculated. All statistical analyses were conducted using SPSS version 26.0. Continuous data are presented as mean ± standard deviation. Prior to the main analyses, the normality and homogeneity of variance of all continuous variables were assessed using the Shapiro-Wilk test and Levene’s test, respectively. For variables meeting parametric assumptions, independent-sample t-tests were used for between-group comparisons; for variables violating these assumptions, the non-parametric Mann-Whitney U test was applied. The specific analytical strategy included the following steps: First, independent-sample t -tests were used to compare demographic characteristics (age and years of education) and psychological scale scores between the infertility patient group and the healthy control group. Second, independent-sample t -tests were also employed to compare hemodynamic parameters (e.g., β values, mean, peak, and integral values) between the two groups at both the individual channel level and the aggregated region-of-interest (ROI) level. To address the multiple comparison issue arising from the numerous fNIRS channels, the Benjamini-Hochberg procedure was applied to control the false discovery rate (FDR), with a significance threshold set at q  = 0.05. The correction procedure involved ranking all raw P -values from the channel-wise comparisons in ascending order and comparing them against the FDR-adjusted significance thresholds; only channels with a corrected P -value ( Q -value) less than 0.05 were considered statistically significant for group differences. Furthermore, within the infertility patient group, Pearson correlation analyses were performed to examine the relationships between fNIRS indices and psychological scale scores as well as infertility duration. Exploratory subgroup analyses based on infertility type (primary vs. secondary) were conducted using independent-sample t-tests to compare fNIRS parameters between the two subgroups. In all analyses, a two-tailed P  < 0.05 (or Q  < 0.05 for FDR-corrected channel comparisons) was considered statistically significant.

This study provides neurobiological evidence that infertility patients undergoing ART exhibit alterations in prefrontal cortical function, which are associated with increased psychological distress and impaired sleep quality. These neural markers indicate potential targets for personalized interventions. Future research could build upon these findings through systematic follow-up studies to further investigate the relationships between neural functional changes, treatment outcomes, and long-term psychological adaptation. Additionally, it would be valuable to examine whether interventions targeting prefrontal cortex function can improve both the mental health and reproductive prognosis of patients. Integrating neuroscience with reproductive medicine will offer new perspectives and approaches for the comprehensive health management of individuals with infertility.

This study presents the first comprehensive investigation of brain activation patterns, sleep quality, and psychological states in infertility patients undergoing ART treatment, utilizing fNIRS technology. The results show that on the trigger day, infertility patients exhibited significantly poorer sleep quality and higher anxiety levels compared to healthy controls. Notably, we observed marked reductions in cortical activation during cognitive tasks, especially in regions involved in language processing (Broca’s area) and executive function (DLPFC). These neural alterations were significantly associated with psychological distress and varied depending on specific infertility-related factors.This series of findings suggests that the trigger day, as a unique stressor during treatment, may pose specific challenges to the neuropsychological state of patients. The critical nature of this time point lies in the fact that it marks the end of prolonged hormonal stimulation and the approach of the oocyte retrieval procedure, constituting a psychological node where multiple stressors converge. In this study, the overall GAD-7 anxiety score of the infertility patient group on the trigger day was 3.68 ± 3.52. Although the mean did not reach the clinical threshold (GAD-7 ≥ 5), individual data analysis revealed that 28.95% of patients exceeded this threshold. This indicates that while the overall anxiety level of the patient group falls within the mild range, at this critical treatment juncture, there exists a clinically significant subgroup of non-negligible size experiencing anxiety of clinical relevance. This finding is consistent with previous studies reporting increased psychological vulnerability during key phases of ART treatment [ 22 ]. Concurrently, the sleep disturbance scores in the patient group were also significantly elevated, suggesting that anticipatory anxiety and treatment-related stress may manifest not only as daytime psychological symptoms but also as disruptions to nocturnal sleep architecture. These findings align with the existing literature and emphasize the need to better understand the interaction between the ART timeline and patients’ psychological functioning. Recent research has indicated that anxiety and depression among patients undergoing ART are influenced by multiple factors, with different treatment stages associated with varying degrees of psychological burden [ 34 ] In particular, the trigger day, due to the convergence of physical discomfort, procedural anxiety, and uncertainty regarding treatment outcomes, may represent an especially vulnerable period [ 35 ]. Notably, the co-occurrence of sleep disturbances and anxiety symptoms observed in this study suggests a close association between these two dimensions in the psychological stress experienced by this population. This observation is consistent with the perspective of the classical sleep emotion regulation model, which proposes that impaired sleep quality may compromise emotional regulation by impairing prefrontal cortex function, while emotional distress can further disrupt sleep, creating a bidirectional loop [ 36 , 37 ]. Although this study did not directly test the causal nature of this interaction, our findings that patients simultaneously showed significant sleep disturbances and anxiety on the highly stressful trigger day provide supportive contextual evidence for applying this model to the infertility treatment population. These identified associations have clear clinical implications. Given that sleep disturbance is a well-established risk factor for various severe psychiatric outcomes, including suicidal risk [ 38 ], and considering that our patient group exhibited a significant decline in sleep quality at this critical treatment juncture, it follows that during ART, particularly around the trigger day, proactive assessment and management of sleep problems could represent a feasible strategy to alleviate patients’ acute psychological distress and potentially mitigate long term mental health risks. No significant difference in PHQ-9 depression scores was found between infertility patients and healthy controls in this study. Although this differs from the general observation that infertility is often accompanied by depressive mood, this result is explicable within the specific context of our study design. First, the stress of the “trigger day” is highly immediate and procedure-specific, and its psychological impact is more likely to manifest directly as anxiety about the impending medical procedure (captured by the GAD-7) rather than the broader and more persistent depressive symptoms assessed by the PHQ-9. Second, the patients enrolled in this study were all in the active phase of ART treatment, and their immediate psychological state may differ from that of individuals who have experienced long-term infertility but are not undergoing active treatment. Therefore, at this critical juncture of the “trigger day,” patients’ acute psychological response may be dominated by anxiety, which may explain the lack of a significant between-group difference in depressive symptoms observed here. This finding suggests that the manifestation of psychological distress may vary dynamically across different stages of the ART process. In parallel with these psychological findings, our fNIRS findings reveal a distinct pattern of prefrontal hypoactivation in infertility patients during cognitively demanding tasks.Clinically, many patients report experiencing mental cloudiness and reduced cognitive efficiency during treatment. Our neural findings may provide an objective correlate to these subjective reports.Reduced activation in Broca’s area may reflect not only difficulties in language production but also broader impairments in inner speech, self-regulatory functions, and the access and manipulation of semantic memory. Semantic memory is a core component of the verbal fluency task, and these capacities are essential for emotional coping [ 39 , 40 ]. Concurrent hypoactivation in the frontal pole, a region implicated in metacognitive processes and future-oriented thinking, suggests that infertility-related stress may hinder patients’ ability to employ adaptive cognitive strategies [ 41 ]. Furthermore, the observed hypoactivation suggests possible functional reorganization in these patients. Future studies with broader cortical coverage could examine whether other brain networks show compensatory recruitment in response to these frontal deficits. Such compensation would represent a potential adaptive mechanism to chronic treatment-related stress. These neurofunctional alterations may be further explained by chronic stress mechanisms.Emerging evidence has elucidated the neurobiological mechanisms underlying chronic stress–induced prefrontal dysfunction, emphasizing the pivotal role of hypothalamic–pituitary–adrenal axis dysregulation and neuroinflammatory processes [ 42 ]. Prolonged activation of the Hypothalamic pituitary adrenal(HPA) axis leads to excessive cortisol secretion, which impairs dendritic branching and synaptic plasticity in prefrontal regions. Simultaneously, elevated levels of proinflammatory cytokines further disrupt neural function and connectivity, compounding stress-related cognitive and emotional impairments. The left-lateralized DLPFC hypoactivation observed in our study is particularly noteworthy, given this region’s critical role in cognitive control and emotion regulation via its connections with limbic structures [ 43 ]. The observed correlation between reduced DLPFC activation and higher anxiety scores suggests that emotional distress may directly disrupt executive function networks, potentially initiating a vicious cycle in which cognitive impairments further intensify psychological distress.It is critical to note that this distress operates within the context of a profound desire for motherhood, which may itself shape neural responsiveness and exacerbate the cognitive-emotional load experienced during treatment.It is noteworthy that while the verbal fluency task employed here is not a direct measure of infertility-related stress, it effectively taps into cognitive domains that are susceptible to disruption under conditions of emotional and physiological strain [ 44 ]. The observed prefrontal hypoactivation during VFT likely represents a confluence of general cognitive load and the specific impact of infertility-associated psychological distress. This interpretation aligns with the significant correlations we found between task-related neural indices and scores on anxiety and insomnia scales. Further supporting the chronic impact of reproductive stress, our finding that longer infertility duration is associated with delayed hemodynamic responses—reflected by a posterior shift in centroid values—provides neurobiological evidence for the cumulative effects of chronic reproductive stress. This observation aligns with the allostatic load model, which posits that prolonged exposure to stress induces progressive neural adaptation and, ultimately, functional impairment [ 45 ]. The posterior shift may indicate compensatory recruitment of alternative neural networks as frontal resources become gradually depleted over time.When comparing primary and secondary infertility patients, we found the situation to be more complex. The results of exploratory analyses based on the current limited sample size suggest that the differences in brain activation patterns between the two groups warrant attention. Specifically, patients with primary infertility exhibited more pronounced hypoactivation in the dorsolateral prefrontal cortex. This may reflect their unique psychological burden: these women have never experienced pregnancy, and their psychological distress may be linked to deeper issues of identity and existential anxiety [ 46 ]. Although these findings are preliminary and require validation in larger samples, they remain significant. On one hand, they suggest that clinical interventions should adopt a more individualized perspective, considering infertility type as a stratification factor. On the other hand, they also point the way for future research, indicating the necessity of incorporating infertility type as a key variable in more in-depth investigations. Taken together, our findings have several important clinical implications. The high prevalence of subclinical anxiety (28.95%) underscores the need for routine psychological screening during critical ART phases, particularly on the trigger day. Given the strong correlations among sleep quality, anxiety, and neural function, multimodal interventions targeting these interconnected domains appear warranted. The identified regions of hypoactivation—DLPFC and frontal pole—could serve as promising targets for non-invasive brain stimulation techniques. A meta-analysis [ 47 ] demonstrated that both repetitive transcranial magnetic stimulation and transcranial direct current stimulation targeting the DLPFC produced significant improvements in anxiety symptoms, suggesting potential therapeutic applications for ART patients experiencing treatment-related distress. Collectively, these insights highlight a critical therapeutic window during the trigger day for implementing early interventions that support sleep, emotional wellbeing, and cognitive resilience. This study has several limitations. First, its cross-sectional design limits causal inferences between brain activation and psychological symptoms and prevents the examination of trajectories of neurocognitive and psychological changes across different ART stages or in relation to treatment outcomes. Second, the statistical power of the study is constrained by the modest sample size, particularly within the infertility patient group ( n  = 38). This limitation is most pertinent to the exploratory comparisons between subgroups (e.g., primary vs. secondary infertility). Third, we focused solely on a verbal fluency task; including additional cognitive paradigms could provide a more comprehensive assessment of neural function. Fourth, the intensity of the desire for motherhood was not directly measured or quantified. This unmeasured variable may confound the observed associations between neural activation and treatment-related distress, underscoring the need for future research to incorporate its assessment to better disentangle these complex interactions. Nonetheless, this study offers several strengths. It is the first to integrate fNIRS neuroimaging with psychological and sleep assessments in ART patients, capturing a critical timepoint (trigger day) of heightened vulnerability. Moreover, our multimodal approach—integrating brain imaging, sleep assessment, and psychological evaluation—provides a holistic understanding of patient wellbeing during ART.

Infertility, defined as the inability to achieve clinical pregnancy after twelve months of regular unprotected intercourse, represents a significant health challenge affecting approximately 17.5% of the adult population worldwide [ 1 , 2 ]. This multifactorial reproductive disorder involves complex physiological, psychological, and social determinants, and is characterized by notable etiological heterogeneity. Common causes include polycystic ovary syndrome (PCOS), tubal factors, endometriosis, and male factors, each associated with distinct pathophysiological processes and psychosocial challenges [ 3 ]. For instance, PCOS is often accompanied by metabolic disturbances and body-image concerns [ 4 ], whereas endometriosis involves chronic pelvic pain; these concomitant symptoms and treatment burdens intersect with the core distress of childlessness, affecting patients’ quality of life across multiple dimensions [ 5 ]. Thus, infertility is not a monolithic condition but a heterogeneous health state involving physiological, psychological, and social dimensions, with stress stemming from reproductive failure, condition-specific symptoms, treatment procedures, and identity threats. This multi-dimensional health crisis generates an urgent need for effective interventions [ 6 ]. In this context, assisted reproductive technology (ART) has become the core treatment strategy for infertility, widely implemented globally [ 7 ]. However, this effective intervention introduces a significant secondary challenge: the ART process itself imposes a substantial and often cumulative psychological burden on patients, characterized by extended and frequently repetitive cycles, invasive procedures, and persistent outcome uncertainty [ 8 ]. As treatment outcomes are not guaranteed in a single attempt, patients often undergo multiple treatment cycles and face renewed pressure to make decisions and adjust protocols following failed attempts. This process, combined with the need to comprehend complex medical information and make sequential, high-stakes decisions, contributes to a chronic state of cognitive-emotional load. The psychological stress associated with ART stems from multiple interrelated factors, including treatment uncertainty, financial strain, social stigma, and the acute emotional toll of potential failure across multiple treatment attempts, which collectively overwhelm patients’ adaptive coping mechanisms and predispose them to anxiety and depressive disorders [ 9 ]. Systematic reviews and meta-analyses have consistently documented elevated rates of psychological distress among infertility patients. A comprehensive meta-analysis of psychological assessment tools in infertility populations revealed significantly higher rates of anxiety, depression, and general psychological distress compared to fertile populations [ 10 ]. Concurrently, sleep disturbances represent a particularly prevalent and bidirectional comorbidity during ART treatment. A recent systematic review specifically examining sleep disturbances and female infertility found that sleep disorders are significantly more prevalent among women with infertility, with sleep quality deteriorating progressively throughout treatment cycles and potentially impacting reproductive outcomes [ 11 ]. These psychological and sleep-related symptoms not only diminish quality of life but also potentially impact treatment adherence and therapeutic outcomes [ 12 ]. Despite extensive research documenting the psychological burden of infertility treatment [ 13 , 14 ], the underlying neurobiological mechanisms remain largely a “black box”. While the majority of existing studies rely on subjective assessment tools, including self-reported questionnaires and clinical interviews, which are valuable for capturing psychological states, they lack objective neurophysiological data and fail to elucidate the cortical activation patterns associated with emotional fluctuations and sleep disturbances during ART treatment [ 15 ]. Understanding this brain-mind connection is crucial, as emerging evidence suggests that brain structure and function may be intimately linked to reproductive outcomes. Recent prospective neuroimaging studies have demonstrated that preconceptional brain structure patterns, particularly in areas associated with reward processing and emotional regulation, can predict conception likelihood in women attempting to conceive [ 16 ]. Elucidating the neural correlates of psychological distress in infertility patients could therefore inform the development of targeted neurobiological interventions and personalized treatment approaches. Functional near-infrared spectroscopy (fNIRS) offers a promising non-invasive neuroimaging approach for investigating brain activation patterns in clinical populations. This technique monitors real-time hemodynamic changes, specifically alterations in oxyhemoglobin(HbO) and deoxyhemoglobin(HbR) concentrations, providing indirect measures of neural activity in cortical regions [ 17 , 18 ]. The principal advantages of fNIRS include superior temporal resolution compared to functional magnetic resonance imaging, reduced susceptibility to motion artifacts, and the capacity for ecological validity through assessment in naturalistic settings [ 19 ]. Recent developments in fNIRS applications for mental health research have demonstrated its particular utility for investigating psychiatric conditions, including mood and anxiety disorders, with growing evidence supporting its role in identifying objective biomarkers of psychological distress [ 20 ]. While fNIRS has demonstrated clinical utility across diverse neuropsychiatric applications, including mood disorders and cognitive dysfunction [ 21 ], its application in ART-related infertility research remains nascent. ART treatment is a continuous process involving several key stages, typically including: controlled ovarian stimulation, trigger, oocyte retrieval, embryo culture, embryo transfer, and final pregnancy confirmation. Within this precisely timed sequence, the “trigger day” represents a critical juncture in ART treatment, marking the administration of human chorionic gonadotropin to induce final oocyte maturation prior to retrieval. This phase constitutes the culmination of weeks of hormonal stimulation and monitoring, representing a period of peak psychological stress and physiological activation that may significantly influence neural processing, sleep architecture, and emotional regulation. The temporal specificity of this intervention provides a unique opportunity to examine acute neurobiological responses to treatment-related stress in a standardized clinical context, as studies have shown sharp elevations in cortisol and sympathetic activity during this stage [ 22 ]. However, current understanding of the functional brain responses at this specific clinical moment remains limited. For instance, while studies utilizing structural MRI have explored the long-term association between pre-conceptional brain structure and conception likelihood [ 23 ], thereby advancing our knowledge of stable neural traits, such approaches have yet to capture the acute, state-dependent functional brain changes triggered by events like the “trigger day” within the treatment cycle. Consequently, applying real-time hemodynamic techniques such as fNIRS to investigate the immediate neurocognitive impact of ART procedures represents a worthwhile direction for further exploration. Based on this rationale, the present study employs fNIRS technology, focusing on the “trigger day” as a key clinical time point, with the following specific aims: (1) characterize brain activation patterns in infertility patients on trigger day using fNIRS technology during cognitive challenge; (2) examine associations between cortical hemodynamic responses and sleep quality and emotional states; and (3) investigate the relationship between infertility characteristics (duration, type) and neural activation patterns. By elucidating the neurobiological underpinnings of cognitive-emotional dysfunction during critical phases of ART, this research provides empirical evidence for developing personalized neurobiological interventions and optimizing psychological support strategies for infertility patients.

Supplementary Material 1. Supplementary Material 1.