This retrospective, time-period–based cohort study included patients with a confirmed diagnosis of endometriosis who received treatment at our institution between January 2021 and December 2024. Clinical data were extracted from electronic medical records and relevant follow-up documentation. To reflect changes in routine clinical practice over the study period, patients treated from January 2021 to December 2022 were managed with standard conventional therapy and were categorized as the control group, whereas patients treated from January 2023 to December 2024 received conventional therapy plus omega-3 PUFA intake as part of the treatment strategy and were categorized as the observation group. Group allocation was therefore determined by the calendar period of care rather than individual randomization. Informed consent was obtained from all participants. The study was approved by the hospital's ethics committee and conducted in accordance with relevant guidelines and the Declaration of Helsinki. Personal identifiers were removed to ensure data confidentiality and participant privacy.

Patients were included if they: 1) Had a confirmed diagnosis of endometriosis documented in medical records (surgical/pathological and/or imaging-supported diagnosis); 2) Were treated at our institution between January 2021 and December 2024; 3) Received conventional therapy during the study period, with time-period allocation: January 2021–December 2022 as the control group and January 2023–December 2024 as the omega-3 PUFA plus conventional therapy group; 4) Had complete baseline demographic and clinical information; 5) Had available baseline and follow-up assessments of pain and at least one evaluation of inflammatory cytokines and QoL using validated instruments.

Patients were excluded if they: 1) Had major comorbidities that could substantially confound pain, inflammatory cytokines, or quality-of-life outcomes (e.g., active pelvic inflammatory disease, autoimmune or systemic inflammatory disorders, or malignancy); 2) Were pregnant or lactating during the primary evaluation window; 3) Had severe organ dysfunction that could affect inflammatory status or treatment comparability; 4) Had unclear or non-standardized omega-3 PUFA exposure (e.g., long-term self-initiated supplementation before baseline or insufficient documentation for accurate classification).

Control group (Conventional therapy): patients treated between January 2021 and December 2022 received standard conventional management for endometriosis. The core regimen consisted of symptom control and disease-suppressive therapy based on individual clinical presentation. Analgesic treatment primarily included non-steroidal anti-inflammatory drugs (NSAIDs) for pelvic pain and dysmenorrhea. Hormonal therapy was administered when clinically indicated, including combined oral contraceptives and/or progestin-based regimens, with gonadotropin-releasing hormone (GnRH) analog-based therapy considered for patients with refractory symptoms. Supportive measures (e.g., counseling regarding lifestyle and pain self-management) were provided as part of routine care. Treatment selection and duration were determined by attending physicians according to standard institutional practice, and key medication exposure was documented in the electronic medical record.

Observation group (Conventional therapy + omega-3 PUFA intake): patients treated between January 2023 and December 2024 received the same conventional management protocol as the control group, with the addition of omega-3 PUFA supplementation as an adjunctive strategy. Omega-3 PUFA supplementation was prescribed in a standardized oral fish oil–derived preparation containing eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) that was routinely used at our institution during this phase. According to the institutional formulary/drug dictionary, the primary preparation provided 180 mg EPA and 120 mg DHA per capsule (total 300 mg EPA+DHA per capsule). The typical prescribed regimen was two to four capsules per day (corresponding to a total EPA+DHA dose of approximately 600–1,200 mg/day, based on prescription records), initiated at the start of the treatment course (baseline) and intended to be continued throughout the routine follow-up period. Exposure to omega-3 PUFA was defined a priori based on documented prescription and/or pharmacy dispensing records indicating planned intake during the observation period. When available, continued use during follow-up was assessed from repeat prescriptions, dispensing records, and follow-up documentation. Patients with unclear exposure documentation precluding reliable classification, or with long-term self-initiated omega-3 supplementation before baseline, were excluded according to the pre-defined criteria. Analgesic therapy with non-steroidal anti-inflammatory drugs (NSAIDs) was administered on an as-needed basis according to routine clinical practice and symptom severity. Specific NSAID agents and doses were individualized and therefore not standardized; however, overall NSAID use was comparable between groups and was adjusted for in multivariable analyses.

Clinical data were retrospectively collected from the electronic medical record system and relevant outpatient follow-up documentation for patients with endometriosis treated at our institution between January 2021 and December 2024. Data abstraction was performed using a standardized case report form by trained investigators. Discrepancies were resolved through re-review of source records. Baseline variables included demographic characteristics (age, body mass index, smoking status, and alcohol use), disease-related characteristics (disease duration and endometriosis stage), and treatment-related information. Endometriosis stage was recorded based on available surgical or imaging documentation as documented in the medical records. Conventional treatment exposure was extracted, including the use of non-steroidal anti-inflammatory drugs and hormonal therapy, with details of regimen type and timing when available. Omega-3 PUFA exposure was defined according to medical records indicating planned or documented omega-3 PUFA intake as part of the institutional treatment strategy during January 2023 to December 2024. For each eligible patient, the form, initiation timing within the treatment course, and documentation of continued use during follow-up were recorded when available. Patients treated during January 2021 to December 2022 received conventional therapy alone and constituted the control group.

Outcome data were collected at two pre-specified time points based on routine clinical practice: baseline assessment at the initiation of the treatment course and post-treatment assessment at the end of the defined evaluation window documented in the records. Pain was evaluated using the visual analog scale (VAS, 0–10), including overall pain and, when available, domain-specific measures for dysmenorrhea, chronic pelvic pain, and dyspareunia. The “overall pain” outcome was operationally defined as pelvic pain attributable to endometriosis and was derived from clinician-documented VAS assessments reflecting endometriosis-associated pain domains, specifically dysmenorrhea, chronic pelvic pain, and dyspareunia. HRQoL was assessed using the SF-36, with extraction of Physical Component Summary (PCS), Mental Component Summary (MCS), and domain scores, including bodily pain, mental health, social functioning, physical functioning, role physical, and role emotional. Inflammatory status was evaluated using routinely obtained laboratory tests. Serum levels of IL-6, TNF-α, and C-reactive protein (CRP) were collected at baseline and post-treatment when paired measurements were available. If multiple results were recorded within a single assessment window, the value closest to the corresponding clinical evaluation date was used. To support adjusted analyses and reduce confounding, additional covariates were extracted, including baseline VAS, baseline PCS/MCS, and key comorbidities documented in the medical history. Data completeness was assessed prior to analysis; patients lacking core paired outcome measures for pain, inflammatory cytokines, or SF-36 were excluded according to pre-defined criteria.

Inflammatory biomarkers were assessed using peripheral venous blood samples collected as part of routine clinical care. Serum IL-6, TNF-α, and C-reactive protein (CRP) levels were measured by the hospital's central laboratory, which follows standardized procedures for each biomarker. IL-6 and TNF-α concentrations were quantified using immunoassay methods, specifically enzyme-linked immunosorbent assay (ELISA) or chemiluminescent immunoassay (CLIA), as per manufacturer protocols and internal quality control standards. CRP was measured using a high-sensitivity immunoturbidimetric assay or an equivalent method. All biomarker results were recorded in the laboratory information system, and data were retrospectively extracted for analysis. If multiple measurements were available within the same assessment window, the value closest to the corresponding clinical evaluation date was used.

All statistical analyses were performed using IBM SPSS Statistics, version 27.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation, and categorical variables as number (percentage). Baseline characteristics between the control and observation groups were compared using the independent-samples t-test for continuous variables and the χ² test for categorical variables. For outcome analyses, paired-samples t-tests were used to evaluate within-group changes in pain scores, inflammatory cytokines, and SF-36 indices from baseline to post-treatment. Between-group differences in post-treatment values and change scores (Δ) were assessed using independent-samples t-tests. Clinically meaningful improvement was defined a priori as overall pain reduction of ≥2 points on the VAS, and SF-36 improvement of ≥5 points for PCS and MCS.

To identify independent associations with clinically meaningful outcomes, multivariable logistic regression models were constructed to estimate odds ratios (ORs) and 95% confidence intervals (CIs). Covariates included demographic factors, disease-related variables, baseline severity measures, and conventional treatment exposure as recorded in the dataset. Subgroup analyses were performed stratified by disease stage (I–II vs. III–IV) and by hormonal therapy use (yes vs. no), and interaction terms were introduced to test for effect modification. Sensitivity analyses were conducted to assess robustness of the primary findings, including complete-case analyses, exclusion of patients with potential key confounders, additional adjustment for smoking and alcohol status, and restriction analyses excluding advanced-stage cases where applicable. Smoking status and alcohol consumption (current use: yes/no at baseline) were additionally adjusted for in sensitivity analyses. A propensity score–based inverse probability of treatment weighting (IPTW) approach was further applied as an alternative adjustment strategy. All tests were two-sided, and p < 0.05 was considered statistically significant.

During the study period, 302 patients with endometriosis were assessed for eligibility. Thirteen patients were excluded based on the pre-defined criteria, including incomplete key outcome data for pain, inflammatory cytokines, or QoL (n = 6), unclear documentation of omega-3 PUFA exposure that precluded reliable group classification (n = 3), major comorbid inflammatory conditions likely to confound cytokine profiles or pain assessment (n = 2), pregnancy or lactation during the evaluation window (n = 1), and insufficient follow-up records for standardized outcome assessment (n = 1). Ultimately, 289 patients were included in the final analysis, comprising 138 patients in the control group and 151 patients in the observation group.

In this retrospective cohort, baseline demographic and clinical variables were comparable between groups. The control group (n = 138) and the observation group (n = 151) showed no significant differences in age (32.8 ± 6.4 vs. 33.1 ± 6.1 years, t = −0.41, p = 0.684) or BMI (22.5 ± 2.8 vs. 22.7 ± 2.9 kg/m², t = −0.60, p = 0.551). The prevalence of current smoking (10.1% vs. 10.6%, χ² = 0.02, p = 0.900) and current alcohol use (15.9% vs. 17.2%, χ² = 0.08, p = 0.771) was similar across groups. Disease duration did not differ significantly (4.2 ± 2.1 vs. 4.1 ± 2.0 years, t = 0.41, p = 0.679). The distribution of endometriosis stages was well balanced (χ² = 0.04, p = 0.998). Key symptom profiles, including pelvic pain, dysmenorrhea, and dyspareunia, were also comparable (all p > 0.80). Regarding conventional treatment exposure, the proportions of NSAID use and hormonal therapy use were similar between groups (p = 0.750 and p = 0.809, respectively; Table 1).

Among patients in the observation group (n = 151), omega-3 PUFA supplementation was prescribed pre-dominantly as a standard fish oil capsule containing 180 mg EPA and 120 mg DHA per capsule (300 mg EPA+DHA); a minority received an alternative formulation with comparable EPA/DHA content according to the institutional formulary. The median prescribed daily dose of combined EPA and DHA was 900 mg/day (IQR: 600–1,200 mg/day; range: 300–1,800 mg/day), corresponding to a median of 3 capsules/day (IQR: 2–4; range: 1–6). The median duration of omega-3 PUFA exposure from baseline to the post-treatment assessment was 12 weeks (IQR: 10–12 weeks; range: 8–16 weeks). Detailed exposure characteristics, including formulation-specific EPA/DHA content, prescribed dose distributions, and exposure duration, are provided in Supplementary Table S1.

Across the study cohort, both groups demonstrated significant within-group reductions in overall pain after treatment. In the control group, the overall VAS score decreased from 6.7 ± 1.2 to 5.2 ± 1.5 (Δ = 1.5 ± 0.9; t = 20.37, p < 0.001). In the observation group receiving conventional therapy plus omega-3 PUFA intake, the overall VAS score declined from 6.6 ± 1.2 to 3.6 ± 1.6 (Δ = 3.0 ± 1.0; t = 36.42, p < 0.001). Between-group comparisons indicated lower post-treatment pain scores in the observation group than in the control group (t = 8.85, p < 0.001), accompanied by a larger magnitude of pain improvement (t = −13.91, p < 0.001). Analyses of specific pain domains yielded consistent patterns. Dysmenorrhea, chronic pelvic pain, and dyspareunia all improved significantly within each group; however, the observation group exhibited greater reductions across domains, reflected in both lower post-treatment scores and larger Δ values (all p < 0.001). These findings support a potential adjunctive benefit of omega-3 PUFA intake in alleviating multiple pain dimensions of endometriosis when added to conventional therapy (Table 2).

At baseline, inflammatory cytokine levels were comparable between groups. There were no significant differences in IL-6 (9.8 ± 3.0 vs. 9.7 ± 3.1 pg/ml, t = 0.28, p = 0.781), TNF-α (15.2 ± 4.2 vs. 15.0 ± 4.1 pg/ml, t = 0.41, p = 0.683), or CRP (4.6 ± 1.9 vs. 4.5 ± 1.8 mg/L, t = 0.46, p = 0.647). After treatment, the observation group receiving conventional therapy plus omega-3 PUFA intake showed lower post-treatment levels of IL-6, TNF-α, and CRP compared with the control group (all p < 0.001). Consistently, the magnitude of reduction (Δ) in each marker was greater in the observation group than in the control group (IL-6: 3.9 ± 1.8 vs. 1.8 ± 1.5; TNF-α: 4.2 ± 1.9 vs. 1.9 ± 1.6; CRP: 1.9 ± 1.1 vs. 0.8 ± 0.9; all p < 0.001; Table 3).

Baseline SF-36 indices were comparable between groups, including PCS (t = −0.28, p = 0.777) and MCS (t = −0.27, p = 0.786). After treatment, both groups exhibited significant within-group improvements in overall QoL. The control group showed increases in PCS and MCS (PCS Δ = 5.3 ± 4.8, paired t = 12.97, p < 0.001; MCS Δ = 4.8 ± 4.6, paired t = 12.26, p < 0.001). The observation group demonstrated larger gains (PCS Δ = 12.1 ± 5.2, paired t = 28.59, p < 0.001; MCS Δ = 10.3 ± 5.0, paired t = 25.31, p < 0.001). Between-group comparisons indicated higher post-treatment PCS and MCS scores in the observation group (PCS: t = −6.77, p < 0.001; MCS: t = −5.38, p < 0.001), accompanied by greater improvement magnitudes (PCS Δ: t = −11.56, p < 0.001; MCS Δ: t = −9.74, p < 0.001). Domain-specific analyses further supported this pattern. Pain-related QoL improved in both groups, with a larger increase in BP in the observation group (Δ = 23.0 ± 9.0 vs. 9.5 ± 8.0, p < 0.001). Emotional/psychological and social functioning domains also showed greater enhancement in the observation group, including MH and SF (both Δp < 0.001). Functional dimensions reflecting daily activities and role limitations (PF, RP, and RE) improved more substantially in the observation group than in the control group (all post-treatment and Δp < 0.001; Table 4).

In the adjusted logistic model, adjunctive omega-3 PUFA intake was independently associated with a higher likelihood of clinically meaningful overall pain improvement (ΔVAS ≥2) compared with conventional therapy alone (OR = 3.06, 95% CI 1.85–5.06, p < 0.001). Higher baseline pain severity was also a significant predictor of achieving this level of improvement (OR = 1.58 per 1-point increase in baseline VAS, 95% CI 1.29–1.94, p < 0.001). In contrast, longer disease duration was associated with a reduced probability of substantial pain relief (OR = 0.90 per year, 95% CI 0.84–0.97, p = 0.006). Age, BMI, disease stage, and concomitant conventional treatment exposure were not independently associated with the pain improvement endpoint in this model. For quality-of-life outcomes, omega-3 PUFA intake also remained an independent factor associated with clinically meaningful gains in SF-36 components. Patients in the observation group were more likely to achieve ΔPCS ≥5 (OR = 2.74, 95% CI 1.67–4.50, p < 0.001) and ΔMCS ≥5 (OR = 2.41, 95% CI 1.50–3.88, p < 0.001). Higher baseline PCS and MCS scores were inversely associated with achieving clinically meaningful improvement in the corresponding domains (both p < 0.001), whereas longer disease duration showed modest negative associations with both PCS and MCS improvement thresholds (Table 5).

Subgroup analyses demonstrated a consistent association between adjunctive omega-3 PUFA intake and clinically meaningful improvements in both pain and physical QoL. The odds of achieving significant pain reduction (ΔVAS ≥2) were higher in the omega-3 PUFA group across disease severity strata, including Stage I–II (OR = 3.22, 95% CI 1.75–5.94) and Stage III–IV (OR = 2.84, 95% CI 1.42–5.67), with no evidence of effect modification by stage (p for interaction = 0.741). Similar consistency was observed for QoL improvement (ΔPCS ≥5) in Stage I–II (OR = 2.81, 95% CI 1.54–5.14) and Stage III–IV (OR = 2.61, 95% CI 1.31–5.20). When stratified by hormonal therapy use, omega-3 PUFA intake remained associated with increased odds of clinically meaningful pain and PCS improvement both among patients receiving hormonal therapy (pain: OR = 2.97, 95% CI 1.60–5.49; PCS: OR = 2.66, 95% CI 1.46–4.85) and those not receiving hormonal therapy (pain: OR = 3.18, 95% CI 1.55–6.53; PCS: OR = 2.88, 95% CI 1.38–6.00), with no significant interaction (p for interaction = 0.689; Table 6).

Sensitivity analyses further supported the robustness of the primary findings. The association of omega-3 PUFA intake with clinically meaningful pain improvement and PCS improvement remained stable in the complete-case analysis (pain: OR = 2.98; PCS: OR = 2.69), after excluding patients with potential inflammatory comorbidities (pain: OR = 3.12; PCS: OR = 2.81), with additional adjustment for smoking and alcohol status (pain: OR = 2.95; PCS: OR = 2.63), after excluding Stage IV disease (pain: OR = 3.21; PCS: OR = 2.88), and in the IPTW model (pain: OR = 2.90; PCS: OR = 2.58). Results remained consistent after additional adjustment for smoking and alcohol use (Table 7).