Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

review

In the case of continuing research on this problem, it is worth considering increasing the field of psychological variables, including anxiety, depression, identity, body image, self-esteem - in two measurements - before starting treatment and after 1-3 months - to recognize changes.

Comments for author File: Comments.pdf

Author Response

Response to Reviewer 1

We sincerely appreciate the reviewer’s comprehensive and constructive evaluation of our manuscript. We are pleased that the reviewer recognizes the importance of addressing the consequences of cancer treatment on women’s intimate lives and the experience of sexual coercion. We also acknowledge the reviewer’s emphasis on the sensitivity required when conducting research in this domain.

We are encouraged by the reviewer’s overall support for our main findings, particularly the observed higher levels of perceived coercion among women receiving LHRHa therapy and among younger patients (≤36 years old).

Below, we provide a detailed point-by-point response to each comment, along with the corresponding revisions made in the manuscript.

Comment 1:
“I question the validity of the use of the term 'experiment' in the text. Was the selection into the groups random? I suggest clarifying this issue at the beginning of the paper.”

Response:
We thank the reviewer for raising this point. The study was designed as a multicenter cross-sectional observational study, and participants were assigned to groups based solely on the endocrine therapy they were already receiving in routine clinical practice. No randomization or experimental allocation was performed.

Changes in Manuscript: To ensure complete clarity for readers, we have further emphasized the observational nature of the study design in the Methods section and we have changed the experimental statements to observational.

• Page 3, Lines 130–134 (Methods → Study Design):
  “This study was purely observational in nature, without any random assignment or experimental procedures. Treatment groups reflected the endocrine therapy regimen that patients were already receiving in routine clinical practice.’’

Comment 2:
“It is worth considering the possibility of psychosocial factors influencing women's functioning and perceived sexual coercion, including information obtained from the medical team about the consequences of treatment resulting in medical menopause. This raises the question of women's knowledge about the effects of treatment and the potential consequences for their functioning, including sexual functioning. Were such discussions held? Were women informed about the effects of treatment before it began?’’

Response 2:
We completely agree. These psychosocial factors may play a key role in mediating relational dynamics and sexual coercion. However, due to the cross-sectional nature of this study, these variables could not be included. We have addressed this clearly in the Limitations section and highlighted the need for future longitudinal studies.

Changes in Manuscript: In line with your valuable suggestions, the limitations paragraph has been revised below and marked in red in the manuscript.

 

• Page 11, Lines 404–408 (Limitations):
  Although counseling regarding ovarian suppression is routinely offered, the extent to which women understood the sexual and psychosocial consequences of treatment was not formally assessed; this unmeasured factor may influence functioning and perceived coercion.

Comment 3:
“Were women informed about the effects of treatment before it began?”

Response 3:
We thank the reviewer for this insightful question. In our centers, counseling about treatment-induced menopause is routinely provided; however, the patients’ understanding or perception of such information was not systematically measured. We acknowledge this as a limitation.

Changes in Manuscript: In line with your valuable suggestions, the limitations paragraph has been revised below and marked in red in the manuscript.

• Page 11, Lines 404–409 (limitations): Added:
  ’’ Although counseling regarding ovarian suppression is routinely offered, the extent to which women understood the sexual and psychosocial consequences of treatment was not formally assessed; this unmeasured factor may influence functioning and perceived coercion. Additionally, socioeconomic status, motherhood status, and relationship duration or partner history were not collected, and these unmeasured factors may further contribute to relational vulnerability.’’

Comment 4:
“In traditional societies… partner attitudes may change because of infertility or loss of reproductive capacity.”

Response 4:
We agree that sociocultural norms around femininity, fertility, and sexuality could intensify relational strain. We expanded the Discussion to emphasize how cultural expectations may contribute to coercive partner behaviors, particularly in younger women.

Changes in Manuscript: Additions have been made to the discussion section below, and the limitations section has been revised in line with your valuable suggestions.

• Page 10, Lines 360–365 (Discussion): Added:
  ‘’Moreover, in sociocultural settings where fertility and reproductive capacity are clo-sely tied to femininity, identity, and marital expectations, abrupt treatment-induced menopause may heighten emotional distress and alter partner perceptions [31]. The awareness or fear of reduced fertility can influence both women’s sense of self and their partners’ relational behaviors, potentially intensifying vulnerability to coercive dynamics—especially among younger patients [32]’’

 

Comment 5:
“Younger women reported higher coercion scores—this mechanism deserves consideration.”

Response 5:
We agree that age-related mechanisms warrant clarification. We have expanded the Discussion to offer psychological and relational explanations for higher coercion levels in younger women.

Changes in Manuscript:

• Page 9, Lines 333–344 (Discussion): In line with your valuable suggestions, we have endeavored to explain the mechanisms underlying the higher prevalence of coercion at younger ages.
  “Age emerged as another key factor, with younger women (≤36 years) reporting significantly higher coercion across all SCIRS domains. This observation aligns with prior studies showing that younger breast cancer survivors report more severe sexual distress, higher levels of marital conflict, and an increased risk of separation or divorce compared with older women [27,28]. Younger women are more likely to be in sexually active relationships, raising young families, or navigating earlier stages of marital de-velopment, which may intensify vulnerability to relational strain during cancer treat-ment. Additionally, abrupt treatment-induced menopause may challenge expectations related to femininity, sexuality, and reproductive potential, leading to heightened concerns about fertility loss and disruptions in identity. These mechanisms may am-plify emotional distress and relational tension, thereby contributing to the higher coer-cion scores observed in this group.’’

Comment 6:
“ANCOVA is generally used when compared samples come from a normally distributed population.”

Response 6:
Thank you for raising this methodological point. We reiterate that all ANCOVA assumptions—including normality of residuals, homogeneity of regression slopes, and variance equality—were checked and met. Nonetheless, we acknowledge that SCIRS data are ordinal and that using ANCOVA presents inherent limitations.

Changes in Manuscript: In line with your valuable suggestions, the ANCOVA limitations have been specified in the limitations paragraph and marked in red in the manuscript.

• Page 10-11, Lines 393–400 (Discussion—Limitations):
  “Third, because SCIRS scores originate from ordinal Likert-type items, the use of parametric ANCOVA models should be interpreted with caution. The modest sample size increases the risk of overfitting, particularly when multiple covariates are included. Although all ANCOVA assumptions—including normality of residuals, variance equality, and homogeneity of regression slopes—were examined and satisfied, variance heterogeneity was detected in some models and addressed using Welch-adjusted F statistics and robust standard errors. Therefore, the adjusted estimates should be interpreted cautiously.’’

Comment 7:
“The study results provide a basis for extending psychological care to partners.”

Response 7:
We appreciate this important recommendation. Given our findings, integrating partner-focused or dyadic psychosocial interventions may be beneficial. We have added this implication to the Discussion.

Changes in Manuscript:

• Page 10, Lines 371–380 (Discussion → Clinical Implications): In line with your valuable suggestions, the discussion paragraph has been revised below and marked in red in the manuscript.
  “Clinically, these findings highlight the need for comprehensive psychosocial and sexual health assessment in premenopausal women receiving LHRHa therapy. Oncology providers should openly discuss potential sexual and relational consequences during treatment planning and ensure confidential environments that facilitate disclosure of sensitive concerns. Integrating psycho-oncology services, sexual counseling, and supportive interventions may reduce relational stress and improve survivorship outcomes. Partner education about treatment-related changes can foster empathy and prevent maladaptive reactions. Because coercive dynamics arise within the couple, incorporating partner-focused or dyadic counseling may further enhance communication and reduce relational tension. Future survivorship programs should therefore consider structured psychological education and partner-inclusive support.’’

Reviewer 2 Report

Comments and Suggestions for Authors

Manuscript Title: Impact of LHRHa Therapy on Relationship Dynamics and Sexual Coercion in Premenopausal Breast Cancer Patients: A Multicenter Cross-Sectional Study

Manuscript ID: healthcare-4019495

1- The study lacks a clearly stated primary hypothesis and does not justify why LHRHa should theoretically increase sexual coercion, making the rationale incomplete, kindly revise and correct it.

2- As, the sample size is equal to (n=81) is small for multivariable (ANCOVA) with many covariates, raising a high risk of overfitting and unstable effect estimates.

3- ANCOVA is inappropriate when groups differ systematically on covariates and assumptions, homogeneity of regression slopes, variance equality, are not fully tested.

4- Variance heterogeneity is performed however, the authors still rely on ANCOVA results.

5- The use of subjective self-report on a stigmatized issue in a conservative society introduces major social desirability bias, however, the manuscript fails to address how these biases may distort findings.

6- The underlying ordinal likert data violate parametric assumptions, as the authors interpret higher coercion scores as being caused by partners reactions to menopause which is speculative and unsupported.

7- No details are provided on missing data handling, interviewer training, or assessments of interviewer effect, which are important for sensitive psychosocial surveys.

8-Revise the conclusion section last paragraph for better clarity.

9- Kindly include some most updated studies in the literature review section and improve the literature review section properly.

Author Response

Response to Reviewer 2

We sincerely thank the reviewer for the thorough and constructive evaluation of our manuscript. We greatly appreciate the time and expertise invested in providing thoughtful feedback. Each comment has substantially improved the clarity, theoretical grounding, and methodological rigor of our study. Below, we provide a point-by-point response and indicate all revisions made to the manuscript.

Comment 1:

“The study lacks a clearly stated primary hypothesis and does not justify why LHRHa should theoretically increase sexual coercion, making the rationale incomplete, kindly revise and correct it.’’

Response :

Thank you for this insightful comment. We fully agree that the Introduction required a more explicit theoretical justification for the proposed association between LHRHa therapy and sexual coercion, as well as a clearly stated primary hypothesis.

In response, we expanded the Introduction to articulate the dyadic mechanism through which abrupt ovarian suppression may influence partner behaviors. Specifically, we now explain how LHRHa-induced hypoactive sexual desire and dyspareunia can create a sexual desire discrepancy, which may be misinterpreted by partners as emotional withdrawal. Such misinterpretations can destabilize the relational power balance and lead to maladaptive coercive responses, including guilt induction or threats of abandonment.

In addition, we explicitly stated the study’s primary hypothesis at the end of this theoretical paragraph.

These revisions provide a clearer conceptual foundation for the study and directly address your comment.

Changes in the manuscript:
A new paragraph outlining the psychosocial and relational mechanisms linking LHRHa-induced menopausal symptoms to coercive partner behaviors has been added to the Introduction (Page 3, Lines 98–113):

Added text:

“The theoretical rationale linking LHRHa therapy to sexual coercion lies in the profo-und physiological and psychological disruption caused by abrupt ovarian suppression [20]. Unlike the gradual transition of natural menopause, LHRHa induces a sudden and severe decline in estrogen, often resulting in marked hypoactive sexual desire and dyspareunia [21]. This creates a substantial ‘sexual desire discrepancy’ between the patient and her partner. From a psychosocial perspective, this discrepancy may desta-bilize dyadic dynamics. Partners may misinterpret the patient’s biologically driven sexual withdrawal as emotional rejection [22]. In response, particularly within relati-onships shaped by traditional gender norms, partners may adopt coercive behavi-ors—ranging from guilt induction (‘Commitment Manipulation’) to threats of aban-donment or infidelity (‘Defection Threat’)—as maladaptive attempts to restore inti-macy or reassert relational control [23].

Based on this framework, we hypothesized that premenopausal women receiving LHRHa therapy would exhibit higher levels of sexual coercion and relational strain compared with those treated with tamoxifen monotherapy, independent of demo-graphic and clinical variables.”

Comment 2:

“As the sample size is equal to (n = 81), it is small for multivariable ANCOVA with many covariates, raising a high risk of overfitting and unstable effect estimates.”

Response:

Thank you for this critical methodological observation. We fully agree that a sample size of 81 requires careful management of degrees of freedom in multivariable analyses such as ANCOVA.

In response to your concern, we revised our analytic strategy to minimize the risk of overfitting:

1. Strict Covariate Selection Approach

Rather than including all demographic and clinical variables, we applied a strict variable selection rule.
Only covariates that (a) showed statistically significant associations in preliminary bivariate analyses or (b) were identified in the literature as essential confounders (e.g., age) were retained in the final ANCOVA models.
This approach improved the ratio of sample size to predictor variables, thereby enhancing model stability.

2. Emphasis on Effect Sizes

Instead of relying solely on p-values, we emphasized partial eta squared (ηp²) to demonstrate that the observed effects were of medium-to-large magnitude, increasing confidence in the robustness of the findings despite the modest sample size.

3. Transparency in Limitations

We added an explicit statement acknowledging the potential risk of overfitting and clarifying that adjusted estimates should be interpreted with appropriate caution.

We believe these modifications address your concern and enhance the statistical rigor of the manuscript.

Changes in the Manuscript:

1) Methods Section Updated (Statistical Analysis)

Page 4, Statistical Analysis, Lines 189–202 — Revised text added:

“Adjusted univariate ANCOVA models were constructed for each SCIRS domain and the total score, with LHRHa use entered as the fixed factor. To minimize the risk of overfitting given the sample size (n = 81), the number of covariates in the adjusted models was strictly limited. Only variables that demonstrated statistically significant associations in preliminary bivariate analyses or were identified as essential confounders in prior literature (e.g., age group) were retained. All ANCOVA assumptions were examined prior to model interpretation. Homogeneity of regression slopes was assessed by testing the interaction between each covariate and the treatment group; no significant violations were detected. Variance equality was evaluated using both Levene’s and Brown–Forsythe tests. When variance heterogeneity was present, Welch-adjusted F statistics and robust standard errors were applied. Residual distributions were evaluated via Q–Q plots and Shapiro–Wilk tests, confirming acceptable normality. Effect sizes were expressed as r values for Mann–Whitney U tests and partial η² for ANCOVA, using thresholds of 0.01, 0.06, and 0.14 to indicate small, medium, and large effects, respectively.”

2) Limitations Section Updated

Page 9-10, Limitations, Lines 393–400 — New text added:

“Third, because SCIRS scores originate from ordinal Likert-type items, the use of parametric ANCOVA models should be interpreted with caution. The modest sample size increases the risk of overfitting, particularly when multiple covariates are included. Although all ANCOVA assumptions—including normality of residuals, variance equality, and homogeneity of regression slopes—were examined and satisfied, variance heterogeneity was detected in some models and addressed using Welch-adjusted F statistics and robust standard errors. Therefore, the adjusted estimates should be interpreted cautiously.”

3) Effect Size Justification Added

Page 7-8, Results (Adjusted Analyses), Lines 272–288 — Clarification added:

“To account for potential confounders, analysis of covariance (ANCOVA) models were applied with LHRHa use as the fixed factor. Consistent with the strict covariate selection approach described in the Methods section, only theoretically essential or bivariately associated variables (age group, treatment duration, surgery type, receipt of chemotherapy, receipt of radiotherapy, and education level) were retained in the adjusted models. All ANCOVA assumptions were examined prior to interpretation, including homogeneity of regression slopes, normality of residuals, and variance equality. Welch-adjusted F statistics and robust standard errors were applied when variance heterogeneity was detected. As shown in Table 5 and Figure 1, LHRHa therapy remained significantly associated with higher coercion scores across all domains (RM/V: p = 0.006; DT: p = 0.001; CM: p < 0.001; Total: p = 0.001). Partial η² values ranged from 0.098 to 0.177, indicating medium-to-large effect sizes and providing additional support for the robustness of these observed associations despite the modest sample size. These findings should be interpreted as associative rather than causal, given the cross-sectional design and the ordinal nature of the SCIRS scale.”

Comment 3:

“ANCOVA is inappropriate when groups differ systematically on covariates and assumptions, homogeneity of regression slopes, variance equality, are not fully tested.”

Response:

Thank you for this important methodological comment. We agree that ANCOVA is appropriate only when its core assumptions are satisfied, particularly when treatment groups may differ on relevant covariates. In response to your suggestion, we expanded the Statistical Analysis section to provide a detailed description of the diagnostic procedures performed before interpreting the adjusted models.

Specifically, we now report that:

• Homogeneity of regression slopes was tested by evaluating the interaction between each covariate and the treatment variable (LHRHa use). No significant violations were detected.
• Variance equality was assessed using Levene’s and Brown–Forsythe tests, and Welch-adjusted F statistics with robust standard errors were applied whenever heterogeneity was present.
• Normality of residuals was evaluated using Q–Q plots and Shapiro–Wilk tests, confirming acceptable distributional assumptions for ANCOVA.
• Additionally, strict covariate selection was applied to reduce the risk of overfitting, and this methodology is now clearly described.

These revisions clarify the appropriateness of ANCOVA for this dataset and enhance the transparency and statistical rigor of the manuscript.

Changes in the Manuscript:

Page 5, Statistical Analysis, Lines 188–205 — Updated paragraph:

“Adjusted univariate ANCOVA models were constructed for each SCIRS domain and the total score, with LHRHa use entered as the fixed factor. To minimize the risk of overfitting given the sample size (n = 81), the number of covariates in the adjusted models was strictly limited. Only variables that demonstrated statistically significant associations in preliminary bivariate analyses or were identified as essential confounders in prior literature (e.g., age group) were retained. All ANCOVA assumptions were examined prior to model interpretation. Homogeneity of regression slopes was assessed by testing the interaction between each covariate and the treatment group; no significant violations were detected. Variance equality was evaluated using both Levene’s and Brown–Forsythe tests. When variance heterogeneity was present, Welch-adjusted F statistics and robust standard errors were applied. Residual distributions were evaluated via Q–Q plots and Shapiro–Wilk tests, confirming acceptable normality. Effect sizes were expressed as r values for Mann–Whitney U tests and partial η² for ANCOVA, using thresholds of 0.01, 0.06, and 0.14 to indicate small, medium, and large effects, respectively. Statistical significance was set at p < 0.05 (two-sided). In addition, Mann–Whitney U tests were performed as non-parametric sensitivity analyses for each SCIRS domain to ensure consistency of findings given the ordinal nature of the scale.’’

Comment 4:

“Variance heterogeneity is performed however, the authors still rely on ANCOVA results.”

Response:

Thank you for drawing attention to this important issue. We agree that variance heterogeneity can affect the validity of ANCOVA estimates and therefore requires careful adjustment. In response to your concern, we have clarified in the Statistical Analysis section how variance heterogeneity was addressed.

Specifically, we now report that:

• Variance equality was formally tested using Levene’s and Brown–Forsythe tests for all SCIRS domains.
• When heterogeneity was detected, we applied Welch-adjusted F statistics and robust standard errors, which provide more reliable estimates under unequal variances.
• Sensitivity analyses using non-parametric tests (Mann–Whitney U) demonstrated that the direction and significance of the results were consistent with the ANCOVA findings, supporting the robustness of the conclusions.
• We have also acknowledged in the Limitations section that variance heterogeneity may still influence the precision of adjusted estimates.

By adding these clarifications, we aimed to improve transparency and demonstrate that the ANCOVA findings remain robust even when accounting for variance heterogeneity.

Changes in the Manuscript:

1) Methods — Statistical Analysis Updated

Page 5, Lines 188–205:

“Adjusted univariate ANCOVA models were constructed for each SCIRS domain and the total score, with LHRHa use entered as the fixed factor. To minimize the risk of overfitting given the sample size (n = 81), the number of covariates in the adjusted models was strictly limited. Only variables that demonstrated statistically significant associations in preliminary bivariate analyses or were identified as essential confounders in prior literature (e.g., age group) were retained. All ANCOVA assumptions were examined prior to model interpretation. Homogeneity of regression slopes was assessed by testing the interaction between each covariate and the treatment group; no significant violations were detected. Variance equality was evaluated using both Levene’s and Brown–Forsythe tests. When variance heterogeneity was present, Welch-adjusted F statistics and robust standard errors were applied. Residual distributions were evaluated via Q–Q plots and Shapiro–Wilk tests, confirming acceptable normality. Effect sizes were expressed as r values for Mann–Whitney U tests and partial η² for ANCOVA, using thresholds of 0.01, 0.06, and 0.14 to indicate small, medium, and large effects, respectively. Statistical significance was set at p < 0.05 (two-sided). In addition, Mann–Whitney U tests were performed as non-parametric sensitivity analyses for each SCIRS domain to ensure consistency of findings given the ordinal nature of the scale.’’

2) Results — Robustness Sentence Added

Page7-8, Lines 272–287:

‘’To account for potential confounders, analysis of covariance (ANCOVA) models were applied with LHRHa use as the fixed factor. Consistent with the strict covariate selection approach described in the Methods section, only theoretically essential or bivariately associated variables (age group, treatment duration, surgery type, receipt of chemotherapy, receipt of radiotherapy, and education level) were retained in the adjusted models. All ANCOVA assumptions were examined prior to interpretation, including homogeneity of regression slopes, normality of residuals, and variance equality. Welch-adjusted F statistics and robust standard errors were applied when variance heterogeneity was detected. As shown in Table 5 and Figure 1, LHRHa therapy remained significantly associated with higher coercion scores across all domains (RM/V: p = 0.006; DT: p = 0.001; CM: p < 0.001; Total: p = 0.001). Partial η² values ranged from 0.098 to 0.177, indicating medium-to-large effect sizes and providing additional support for the robustness of these observed associations despite the modest sample size. These findings should be interpreted as associative rather than causal, given the cross-sectional design and the ordinal nature of the SCIRS scale.’’

 

3) Limitations Updated

    Page 11, Lines 401–407:

“Fourth, the study did not incorporate broader psychosocial constructs—such as anxiety, depression, body image, fertility concerns, self-esteem, marital satisfaction, or pre-existing relationship conflict—which may meaningfully influence both women’s vulnerability to coercive dynamics and their adjustment to treatment-induced menopause. Additionally, socioeconomic status, motherhood status, and relationship duration or partner history were not collected, and these unmeasured factors may further contribute to relational vulnerability.”

  Comment 5:

“The use of subjective self-report on a stigmatized issue in a conservative society introduces major social desirability bias, however, the manuscript fails to address how these biases may distort findings.”

Response:

Thank you for highlighting this important methodological concern. We fully agree that self-reported data regarding sexuality and coercive behaviors may be affected by social desirability bias, particularly in conservative sociocultural contexts such as Turkey. In response to your suggestion, we have expanded the Limitations section to explicitly acknowledge this issue and to discuss how it may influence our findings.

We added a detailed explanation stating that cultural stigma surrounding sexual topics may lead to underreporting of coercive experiences, potentially resulting in more conservative estimates of prevalence and effect sizes. To mitigate this risk, we also clarified that interviews were conducted by trained female clinicians in a private setting, anonymity was ensured, and participants were encouraged to answer freely without judgment. Nonetheless, we acknowledge that some degree of social desirability bias cannot be fully eliminated.

We believe that these revisions improve transparency and strengthen the interpretation of our findings.

Changes in the Manuscript:

Limitations Section Expanded

Page 10, Lines 388–392 — New text added:

“ Second, the reliance on self-reported data may introduce recall error and social desirability bias, especially in conservative sociocultural contexts where discussions about sexuality and relational difficulties may be stigmatized. Although all assessments were conducted privately by trained female clinicians, underreporting of coercive experiences remains possible and may have resulted in conservative effect estimates.’’

 

Comment 6:

“The underlying ordinal Likert data violate parametric assumptions, as the authors interpret higher coercion scores as being caused by partners’ reactions to menopause, which is speculative and unsupported.”

Response:

Thank you for this important clarification. We agree that SCIRS scores are derived from ordinal Likert-type items, and that parametric analyses should therefore be interpreted with caution. In response to your suggestion, we implemented several steps to improve the rigor and clarity of our analyses and interpretations.

First, we conducted parallel non-parametric comparisons (Mann–Whitney U tests) for each SCIRS domain, and the direction and significance of these results were consistent with the ANCOVA findings, supporting the robustness of the associations.

Second, we revised the manuscript to remove any language implying causal inference. We now clearly emphasize that our findings reflect associations rather than causal effects, and that coercion scores may be influenced by multiple psychosocial factors beyond treatment-induced menopausal symptoms.

Third, we added a statement to the Limitations section noting that SCIRS domain scores, while treated as continuous for ANCOVA, originate from ordinal scales, which may introduce analytic constraints.

We believe these revisions enhance the methodological transparency and accuracy of the interpretation.

Changes in the Manuscript:

1) Methods Section — Sensitivity Analysis Added

Page 5, Lines 203–205:

“In addition, Mann–Whitney U tests were performed as non-parametric sensitivity analyses for each SCIRS domain to ensure consistency of findings given the ordinal nature of the scale.’’

2) Results Section — Causal Language Removed

Page 7-8, Lines 271–287:

‘’ To account for potential confounders, analysis of covariance (ANCOVA) models were applied with LHRHa use as the fixed factor. Consistent with the strict covariate selection approach described in the Methods section, only theoretically essential or bivariately associated variables (age group, treatment duration, surgery type, receipt of chemotherapy, receipt of radiotherapy, and education level) were retained in the adjusted models. All ANCOVA assumptions were examined prior to interpretation, including homogeneity of regression slopes, normality of residuals, and variance equality. Welch-adjusted F statistics and robust standard errors were applied when variance heterogeneity was detected. As shown in Table 5 and Figure 1, LHRHa therapy remained significantly associated with higher coercion scores across all domains (RM/V: p = 0.006; DT: p = 0.001; CM: p < 0.001; Total: p = 0.001). Partial η² values ranged from 0.098 to 0.177, indicating medium-to-large effect sizes and providing additional support for the robustness of these observed associations despite the modest sample size. These findings should be interpreted as associative rather than causal, given the cross-sectional design and the ordinal nature of the SCIRS scale. ‘’

3) Limitations Section — Ordinal Data Note Added

Page 10-11, Lines 393–400:

“Third, because SCIRS scores originate from ordinal Likert-type items, the use of parametric ANCOVA models should be interpreted with caution. The modest sample size increases the risk of overfitting, particularly when multiple covariates are included. Although all ANCOVA assumptions—including normality of residuals, variance equality, and homogeneity of regression slopes—were examined and satisfied, variance heterogeneity was detected in some models and addressed using Welch-adjusted F statistics and robust standard errors. Therefore, the adjusted estimates should be interpreted cautiously.”

Comment 7:

“No details are provided on missing data handling, interviewer training, or assessments of interviewer effect.”

Response:

Thank you for this valuable methodological comment. We agree that, in studies addressing sensitive psychosocial constructs, it is essential to clarify procedures related to missing data, interviewer training, and interviewer-related variability. In response to your suggestion, we have revised both the Methods and Limitations sections to provide these details.

Missing data were minimal and handled using complete-case analysis. All interviews were conducted by trained female clinicians with at least two years of experience in breast cancer survivorship care, following a standardized administration protocol to reduce interviewer-related variation. Although these procedures were implemented to minimize interviewer effects, we have acknowledged in the Limitations section that some degree of interviewer influence cannot be completely excluded.

These revisions strengthen the methodological transparency and align with best practices for psychosocial research.

Changes in the Manuscript

1) Methods → Measures (NEW Sentence Added) Page 4, Lines 167–174:

 “All interviews were conducted individually in a confidential setting by trained female clinicians with at least two years of experience in breast cancer survivorship care. Interviewers received standardized training on SCIRS administration to minimize interviewer-related variability. When needed, participants were offered referrals for psychological counseling. Missing data were minimal and handled using complete-case analysis, and no systematic interviewer effects were identified during data quality checks.’’

2) Limitations → NEW Clause Added Page 11, Lines 408–411:

 “Fifth, interviewer influence cannot be completely excluded despite standardized SCIRS administration training. Moreover, partner-level factors—such as partners’ beliefs, expectations, coping styles, or psychological distress—were not assessed, though they may shape relational dynamics following endocrine therapy.’’

Comment 8: Revise the conclusion section last paragraph for better clarity.

Response:

Thank you for this helpful suggestion. We agree that the final par agraph of the Conclusion would benefit from improved clarity and concise wording. In response, we have rewritten this section to provide a more focused summary of the findings, to emphasize the associative (rather than causal) nature of the results, and to articulate the clinical and psychosocial implications more clearly. The revised paragraph also highlights areas for future research. We believe this modification substantially improves the readability and interpretability of the Conclusion.

REVISED CONCLUSION PARAGRAPH Page 11, Lines 424–436:

“In conclusion, premenopausal women receiving LHRHa therapy demonstrated significantly higher levels of intimate partner sexual coercion compared with those treated with tamoxifen alone; however, these findings are associative rather than causal. The results underscore the need for clinicians to recognize the relational and psychosocial challenges that may accompany abrupt ovarian suppression, particularly in sociocultural settings where open communication about sexuality is limited. Routine assessment of sexual well-being, relational strain, and partner dynamics should be incorporated into survivorship care, with special attention to younger women, who appear more vulnerable to coercive relational patterns. Future longitudinal and multicenter studies with larger and more diverse cohorts are needed to clarify temporal trajectories, identify high-risk subgroups, and better elucidate the mechanisms through which endocrine therapy influences intimate partner relationships.”

 

Comment 9:
“Kindly include some most updated studies in the literature review section and improve the literature review section properly.”

Response:
Thank you for this valuable suggestion. We agree that incorporating more recent and regionally diverse studies would strengthen the literature review and provide a broader conceptual foundation for our research question. In response, we have expanded the Introduction to include updated evidence published within the past five years from Europe, North America, and Latin America regarding the relational and psychosocial effects of endocrine therapy–induced menopausal symptoms in breast cancer survivors. These additions offer a more comprehensive international context and further justify the importance of examining intimate partner dynamics in women receiving LHRHa therapy. The revised Introduction now reflects a clearer and more current overview of the global literature.

Changes in the manuscript:
A new paragraph summarizing recent international findings has been added to the Introduction (Page 2, Lines 70–80),

‘’Recent studies from Europe, North America, and Latin America also highlight the complex relational impact of treatment-induced sexual dysfunction in breast cancer survivors. Large cohort studies from the United States and Canada have shown that endocrine therapy–related menopausal symptoms are associated with increased rela-tional tension, reduced intimacy, and partner-driven pressure to maintain sexual ac-tivity [9,10]. Similarly, research from Brazil and Mexico has documented how soci-ocultural expectations and limited communication around sexuality may predispose women to coercive or guilt-based partner behaviors during survivorship [11-13]. These findings underscore the need to better understand how ovarian suppression may in-fluence intimate partner dynamics across diverse cultural contexts.’’

 

Reviewer 3 Report

Comments and Suggestions for Authors

This study provides a valuable contribution to the literature on the psychosocial consequences of endocrine therapy among premenopausal breast cancer patients. The use of the SCIRS scale and the inclusion of adjusted analyses (ANCOVA) are methodological strengths, and the comparison between LHRHa therapy and tamoxifen is clinically relevant. The manuscript is well structured, follows STROBE recommendations, and presents the results clearly.However, the paper would benefit from additional conceptual clarification, a more rigorous methodological description, and a deeper explanation of the psychosocial mechanisms that may link ovarian suppression to coercive dynamics within intimate relationships.

Suggestions

1. Strengthen the theoretical rationale for the LHRHa–sexual coercion association

The introduction discusses sexual side effects of LHRHa, but the conceptual link between medically induced menopause and partner coercive behaviors needs expansion. It would be useful to:

• explicitly outline psychosocial mechanisms (e.g., loss of sexual desire → marital tension → emotional pressure),

• integrate international research on power dynamics and coercion in couples affected by chronic illness.

• 2. Discuss omitted psychosocial confounders

ANCOVA adjusts for relevant clinical variables, yet key psychosocial confounders are not considered, such as:

• marital satisfaction,

• pre-diagnosis relationship conflict,

• history of intimate partner violence,

• prior mental health status (anxiety, depression),

• social support.

These factors could influence both the likelihood of receiving LHRHa (through disease severity) and the risk of experiencing coercive behaviors.

3. Justify the age cutoff (≤36 years)

The selected threshold for age stratification is not explained. Please clarify whether:

• it is clinically grounded (e.g., SOFT/TEXT recommendations), or

• data-driven (e.g., median split).

This justification is necessary for interpretability and replicability.

4. Report SCIRS psychometrics for the current sample

Although the Turkish validation is cited, Cronbach’s α values for this sample (overall and for each subscale) should be reported, as internal reliability is essential when SCIRS is the primary outcome.

5. Address variance heterogeneity and ANCOVA robustness

Since heterogeneity of variances is mentioned, please specify:

• which tests were used (e.g., Levene, Brown–Forsythe),

• whether corrections were applied,

• and how variance heterogeneity may influence the validity of ANCOVA results.

6. Expand the ethical and clinical implications

The discussion could be strengthened by addressing:

• oncologists’ ethical responsibilities when identifying partner coercion,

• the value of interdisciplinary collaboration with psychologists and counselors,

• the risk of underreporting in conservative sociocultural context                                              The manuscript addresses a highly relevant and understudied topic in psycho-oncology and provides meaningful empirical evidence on relationship dynamics during endocrine therapy. The study is solid and well executed, but it would benefit from conceptual and methodological clarifications, expanded discussion of psychosocial confounders, and a more explicit treatment of ethical considerations. These refinements would enhance both the scientific rigor and the clinical applicability of the findings.

Comments on the Quality of English Language

The English language is generally clear, comprehensible, and suitable for academic publication. The manuscript is well written, with coherent structure and appropriate scientific terminology. However, several areas would benefit from minor linguistic refinement:

• Occasional phrasing is repetitive, particularly in the Introduction and Discussion, where similar ideas about LHRHa side effects appear multiple times.

• A few sentences are lengthy and could be made more concise for readability.

• Terminology should be standardized throughout (e.g., consistent use of “sexual coercion” rather than alternating with “coercion exposure”).

• Abbreviations in tables and figures should be clearly defined to avoid ambiguity.

• Minor grammatical adjustments (articles, word order, and transitions) would further improve fluency.

Overall, the manuscript requires minor English editing to enhance clarity and stylistic precision, but no major rewriting is needed.

Author Response

Response to Reviewer 3

We sincerely thank the reviewer for their thoughtful and encouraging evaluation of our manuscript. We greatly appreciate the recognition of the study’s methodological strengths, theoretical relevance, and contribution to understanding psychosocial outcomes associated with endocrine therapy in premenopausal breast cancer patients. We also appreciate the reviewer’s insightful suggestions regarding conceptual clarity, methodological refinement, and expansion of psychosocial mechanisms. These comments substantially improved the depth and rigor of the revised manuscript.

Below, we provide a detailed, point-by-point response to each comment along with the corresponding revisions made in the manuscript.

Comment 1:

“Strengthen the theoretical rationale for the LHRHa–sexual coercion association. The introduction discusses sexual side effects of LHRHa, but the conceptual link between medically induced menopause and partner coercive behaviors needs expansion. It would be useful to: explicitly outline psychosocial mechanisms (e.g., loss of sexual desire → marital tension → emotional pressure), integrate international research on power dynamics and coercion in couples affected by chronic illness.”

Response:

Thank you for this insightful and constructive comment. We agree that the theoretical rationale required further strengthening. In the revised manuscript, we have expanded the Introduction to provide a clearer conceptual model describing how abrupt ovarian suppression may disrupt dyadic functioning, create sexual desire discrepancies, and trigger maladaptive partner responses.

Specifically, we now outline how hypoactive sexual desire, dyspareunia, and emotional distress associated with LHRHa therapy may be misinterpreted by partners as relational withdrawal. This misinterpretation may increase relational tension and lead to coercive behaviors—such as guilt induction, pressure for intimacy, or threats of abandonment—as documented in the broader literature on power asymmetry and coercion in couples coping with chronic illness. These additions integrate international research and provide a more comprehensive framework for the hypothesized association.

Changes in the manuscript: Page 3, Lines 98–112 (Introduction):

‘’The theoretical rationale linking LHRHa The theoretical rationale linking LHRHa therapy to sexual coercion lies in the profound physiological and psychological disruption caused by abrupt ovarian suppression [20]. Unlike the gradual transition of natural menopause, LHRHa induces a sudden and severe decline in estrogen, often resulting in marked hypoactive sexual desire and dyspareunia [21]. This creates a substantial ‘sexual desire discrepancy’ between the patient and her partner.

From a psychosocial perspective, this discrepancy may destabilize dyadic dynamics. Partners may misinterpret the patient’s biologically driven sexual withdrawal as emotional rejection [22]. In response, particularly within relationships shaped by traditional gender norms, partners may adopt coercive behaviors—ranging from guilt induction (‘Commitment Manipulation’) to threats of abandonment or infidelity (‘Defection Threat’)—as maladaptive attempts to restore intimacy or reassert relational control [23].

Based on this framework, we hypothesized that premenopausal women receiving LHRHa therapy would exhibit higher levels of sexual coercion and relational strain compared with those treated with tamoxifen monotherapy, independent of demographic and clinical variables. ‘’

Comment 2:

“ Discuss omitted psychosocial confounders ANCOVA adjusts for relevant clinical variables, yet key psychosocial confounders are not considered, such as: marital satisfaction, pre-diagnosis relationship conflict, history of intimate partner violence, prior mental health status (anxiety, depression),social support. These factors could influence both the likelihood of receiving LHRHa (through disease severity) and the risk of experiencing coercive behaviors.”

Response:

Thank you for this insightful comment. We fully agree that key psychosocial constructs—such as marital satisfaction, pre-diagnosis relationship conflict, intimate partner violence history, mental health status, and social support—may meaningfully influence both women’s vulnerability to coercive relational dynamics and their psychological responses to treatment-induced menopause.

These variables were not systematically collected across participating centers, and therefore could not be included in the adjusted analyses. As recommended, we have now explicitly acknowledged these omitted psychosocial confounders in the Limitations section and clarified that their absence may have contributed to residual confounding. We also emphasize that future studies should incorporate these constructs to more comprehensively model the relational and emotional pathways associated with ovarian suppression.

Changes in the manuscript:

• Page 11, Lines 402–408 (Limitations):
  ‘’Fourth, the study did not incorporate broader psychosocial constructs—such as anxiety, depression, body image, fertility concerns, self-esteem, marital satisfaction, or pre-existing relationship conflict—which may meaningfully influence both women’s vulnerability to coercive dynamics and their adjustment to treatment-induced menopause. Additionally, socioeconomic status, motherhood status, and relationship duration or partner history were not collected, and these unmeasured factors may further contribute to relational vulnerability.’’

Comment 3:
“ Justify the age cutoff (≤36 years) The selected threshold for age stratification is not explained. Please clarify whether:it is clinically grounded (e.g., SOFT/TEXT recommendations), or data-driven (e.g., median split). This justification is necessary for interpretability and replicability.”

Response

Thank you for this important comment. We agree that the rationale for selecting the ≤36-year cutoff needed clearer explanation. In the revised manuscript, we have clarified that the cutoff was based on both clinical and empirical considerations.

Clinically, subgroup analyses from the SOFT and TEXT trials identify women in their mid-30s as a distinct premenopausal subgroup with more pronounced ovarian suppression–related symptoms and greater vulnerability to treatment-induced menopausal distress. Empirically, ≤36 years closely matched the median age of our sample, producing balanced comparison groups and improving the stability of adjusted analyses.

This clarification has now been added to the Methods section.

Changes in the manuscript

• Page 4, Lines 155–159 (Methods → Treatment Groups):
  “The age cutoff (≤36 years) was chosen based on subgroup findings from the SOFT and TEXT trials, which identify mid-30s as a clinically distinct premenopausal group with more severe ovarian suppression–related symptoms. This threshold also corresponded to the median age of our sample, enabling balanced group sizes for adjusted analyses.’’

Comment 4:

“Report SCIRS psychometrics for the current sample. Internal reliability (Cronbach’s α) should be provided for the total scale and subscales.”

Response

Thank you for this valuable methodological suggestion. We fully agree that reporting internal reliability metrics is essential, particularly because SCIRS serves as the primary outcome measure in this study. In response, we recalculated Cronbach’s α values for the total scale and each of the three SCIRS subscales in our sample (n = 81). The internal consistency of the SCIRS was excellent overall and remained high across all subdomains, fully supporting the reliability of the instrument within this cohort.

These psychometric indices have now been incorporated into both the Methods and Results sections of the revised manuscript, as recommended.

Changes in the manuscript

Methods → Measures (Page 4, Lines 175–178):

“The SCIRS consisted of 39 items divided into three subscales consistent with its original structure: CM (20 items), DT (12 items), and RM/V (7 items). Internal consistency in the present sample was excellent, with Cronbach’s α values of 0.966 for the total scale, 0.952 for CM, 0.912 for DT, and 0.831 for RM/V.’’

Results → Internal Reliability (Page 6, Lines 224–230):

“3.2 Internal Reliability of the SCIRS

Internal reliability analyses demonstrated excellent internal consistency for the SCIRS total score (Cronbach’s α = 0.966). Subscale reliability was also high: Commitment Manipulation (α = 0.952), Defection Threat (α = 0.912), and Resource Manipulation/Violence (α = 0.831). These findings indicate that the scale performed reliably within this breast cancer cohort.’’

Comment 5:

“ Address variance heterogeneity and ANCOVA robustness Since heterogeneity of variances is mentioned, please specify: which tests were used (e.g., Levene, Brown–Forsythe), whether corrections were applied, and how variance heterogeneity may influence the validity of ANCOVA results.

Response

Thank you for highlighting this important analytical issue. We agree that greater clarity regarding variance heterogeneity and ANCOVA robustness strengthens the methodological transparency of the study. In the revised manuscript, we have expanded the Statistical Analysis section to explicitly describe all assumption checks performed prior to conducting ANCOVA.

More specifically, we now report that homogeneity of variances was evaluated using Levene’s test and the Brown–Forsythe test, and homogeneity of regression slopes was assessed by examining the interaction between covariates and group status. When variance heterogeneity was detected, Welch-adjusted F statistics and robust standard errors were applied. We also conducted non-parametric sensitivity analyses, which produced results consistent with the ANCOVA findings and support the robustness of observed group differences.

We further clarify in the Limitations section that, despite the application of these corrections, variance heterogeneity may still affect the precision of adjusted estimates and that results should be interpreted cautiously.

Changes in the manuscript

Methods → Statistical Analysis (Page 5, Lines 194–202):

“All ANCOVA assumptions were examined prior to model interpretation. Homogeneity of regression slopes was assessed by testing the interaction between each covariate and the treatment group; no significant violations were detected. Variance equality was evaluated using both Levene’s and Brown–Forsythe tests. When variance heterogeneity was present, Welch-adjusted F statistics and robust standard errors were applied. Residual distributions were evaluated via Q–Q plots and Shapiro–Wilk tests, confirming acceptable normality. Effect sizes were expressed as r values for Mann–Whitney U tests and partial η² for ANCOVA, using thresholds of 0.01, 0.06, and 0.14 to indicate small, medium, and large effects, respectively.”

Limitations (Page 11, Lines 396–401):

“Although all ANCOVA assumptions—including normality of residuals, variance equality, and homogeneity of regression slopes—were examined and satisfied, variance heterogeneity was detected in some models and addressed using Welch-adjusted F statistics and robust standard errors. Therefore, the adjusted estimates should be interpreted cautiously.’’

Comment 6:
“Expand the ethical and clinical implications

The discussion could be strengthened by addressing: oncologists’ ethical responsibilities when identifying partner coercion, the value of interdisciplinary collaboration with psychologists and counselors, the risk of underreporting in conservative sociocultural context. The manuscript addresses a highly relevant and understudied topic in psycho-oncology and provides meaningful empirical evidence on relationship dynamics during endocrine therapy. The study is solid and well executed, but it would benefit from conceptual and methodological clarifications, expanded discussion of psychosocial confounders, and a more explicit treatment of ethical considerations. These refinements would enhance both the scientific rigor and the clinical applicability of the findings.’’

Response

Thank you for this thoughtful and important comment. We agree that expanding the ethical and clinical implications strengthens the practical relevance of our findings. In the revised Discussion, we have elaborated on the ethical responsibilities of oncology providers to proactively assess relational distress and sexual coercion, particularly in sociocultural contexts where stigma may limit open disclosure.

We now emphasize that oncologists and survivorship teams have an ethical obligation to create confidential, nonjudgmental environments that facilitate patient disclosure of sensitive concerns. We also highlight the importance of interdisciplinary collaboration—including psycho-oncology, sexual health counseling, and social work—in supporting women who may be experiencing relational strain or coercive partner behaviors. Additionally, we expanded the discussion of underreporting risk in conservative cultural settings and clarified how such dynamics may obscure the true prevalence of coercion.

These revisions address the reviewer’s concerns and align with international psycho-oncology guidelines advocating for holistic, ethically grounded survivorship care.

Changes in the manuscript

Discussion (Page 10, Lines 371-380):

 “Clinically, these findings highlight the need for comprehensive psychosocial and sexual health assessment in premenopausal women receiving LHRHa therapy. Oncology providers should openly discuss potential sexual and relational consequences during treatment planning and ensure confidential environments that facilitate disclosure of sensitive concerns. Integrating psycho-oncology services, sexual counseling, and supportive interventions may reduce relational stress and improve survivorship outcomes. Partner education about treatment-related changes can foster empathy and prevent maladaptive reactions. Because coercive dynamics arise within the couple, incorporating partner-focused or dyadic counseling may further enhance communication and reduce relational tension. Future survivorship programs should therefore consider structured psychological education and partner-inclusive support.”

Comment 7:

“The English language is generally clear, but several areas require minor refinement, removal of repetition, standardization of terminology, and correction of a few long sentences.

Response

Thank you for your helpful observations regarding the language quality. In response, we have carefully revised the entire manuscript to improve clarity, remove repetitive phrasing, standardize terminology (including consistent use of “sexual coercion”), and correct overly long or complex sentences. We also refined transitions and adjusted minor grammatical issues to enhance overall readability. These revisions improve the linguistic precision of the manuscript while preserving its scientific meaning.

Additionally, the full revision of the manuscript’s language was carried out in coordination with an academic English specialist from our university who is experienced in scientific writing, to ensure that the manuscript meets high standards of linguistic accuracy and academic style.

Changes in the manuscript:
• Terminology standardized (e.g., consistent use of “sexual coercion” across all sections).
• Repetitive sentences removed or merged in the Introduction and Discussion.
• Long or complex sentences shortened for clarity and readability.
• Grammar, transitions, and article usage refined throughout the manuscript.
• Table and figure abbreviations clarified according to journal style requirements.

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript presents a very interesting study; however, I have several observations and important points that need to be addressed.

Tables (Tables 1–4).
It is necessary to specify which comparisons were calculated using UDMAT-Winnie and which were calculated using Fisher’s test. Additionally, please clarify why certain values appear in bold font and what specific aspect the authors wish to highlight with this formatting.

Figure 1.
The figure states that bars are used to illustrate the scores, yet no error bars are shown. It is essential to include them. Furthermore, the figure appears quite dim and needs improved clarity and sharpness.

Text adjustments.
Certain sections should be removed, specifically lines 2.17 to 2.19. Although the authors considered several covariates, it is important to acknowledge the limitations of the study. Moreover, I suggest including additional variables such as the socioeconomic status of the women evaluated, whether they are mothers or not, and the number of current partners, as these could provide a broader set of relevant factors. It would also be valuable to account for any additional therapies the participants may be using—such as complementary or alternative treatments—that could influence outcomes. These aspects should be incorporated into the manuscript.

Discussion section.
The discussion is generally well developed; however, there are a few words in bold whose purpose is unclear and should be corrected. It is also important to compare the present findings with similar studies conducted in other regions, such as Latin America, Europe, and elsewhere, where difficulties experienced by women with cancer have also been assessed. These comparisons would strengthen the discussion and contextualize the study more effectively.

Sample size and justification.
The authors should explain why the sample consisted of only 81 participants. Were additional participants available? How was this sample size justified? Furthermore, considering the relatively small number of participants in each group, it is crucial to explain how the inclusion of such a large number of variables in the models was justified. These points need to be clearly addressed in the manuscript.

References.
Overall, the references are adequate, though some errors must be corrected. I also recommend increasing the number of references, particularly by incorporating studies from Mexico, the United States, Latin America, Canada, and Europe that may serve as useful comparisons for the findings presented.

In summary, this is a very interesting study, but it would benefit from further expansion and clarification in several key areas.

Author Response

Response to Reviewer 4

We sincerely thank the reviewer for their detailed, thoughtful, and constructive evaluation of our manuscript. We appreciate the recognition of the study’s significance and value the reviewer’s suggestions, which have substantially improved the clarity, methodological rigor, and contextual relevance of the revised manuscript. Below, we provide a point-by-point response to each comment and summarize the revisions made accordingly.

Comment 1   Tables

“It is necessary to specify which comparisons were calculated using UDMAT-Winnie and which were calculated using Fisher’s test. Additionally, please clarify why certain values appear in bold font and what specific aspect the authors wish to highlight with this formatting.”

Response:

Thank you for this important comment. We agree that the statistical procedures used in the tables required clearer reporting. In the revised manuscript, we now specify in both the Statistical Analysis section and the table footnotes which tests were used for each comparison. The Mann–Whitney U test (UDMAT-Winnie) was applied for non-parametric continuous variables, and Fisher’s exact test for categorical variables with small cell frequencies. Additionally, we clarified that bold formatting indicates statistically significant p-values (p < 0.05).

Changes in the manuscript:

• Tables 1 Footnote:

Values are presented as n (%).
p values were calculated using the Chi-square test or Fisher’s exact test when expected cell counts were < 5.
Bold p-values indicate statistical significance (p < 0.05).
LHRHa = Luteinizing Hormone-Releasing Hormone agonist;
BCS = Breast-conserving surgery.

• Tables 2 Footnote:

Values are mean ranks from the Mann–Whitney U test (UDMAT-Winnie).
U, Z, and p values are reported for each comparison.
Bold p-values indicate statistical significance (p < 0.05).
RM/V = Resource Manipulation/Violence; DT = Defection Threat;
CM = Commitment Manipulation; SCIRS = Sexual Coercion in Intimate Relationships Scale.

• Tables 3 Footnote:

Values represent mean ranks from the Kruskal–Wallis test.
χ² statistics and corresponding p-values are provided for each domain.
Bold p-values indicate statistical significance (p < 0.05).
RM/V = Resource Manipulation/Violence; DT = Defection Threat;
CM = Commitment Manipulation; SCIRS = Sexual Coercion in Intimate Relationships Scale.

• Tables 4 Footnote:

Non-parametric comparisons were performed using the Mann–Whitney U test (UDMAT-Winnie).
r effect sizes were calculated as r = Z / √N. Bold p-values indicate statistical significance (p < 0.05).
RM/V = Resource Manipulation/Violence; DT = Defection Threat; CM = Commitment Manipulation; SCIRS = Sexual Coercion in Intimate Relationships Scale.

 

 Comment 2   Figure 1 Improvements

“The figure states that bars are used to illustrate the scores, yet no error bars are shown. It is essential to include them. Furthermore, the figure appears quite dim and needs improved clarity and sharpness..”

Response:

We appreciate this suggestion. Figure 1 has been fully revised to include error bars representing ± standard error for each group. The figure has been regenerated at higher resolution (300 dpi), with improved brightness, contrast, and label visibility to enhance clarity and meet publication standards.

Changes in the manuscript:

• Updated Figure 1: Added error bars and improved graphical quality.

Comment 3  Methods Section: Remove Lines 2.17–2.19; Consider Additional Variables

“Certain sections should be removed, specifically lines 2.17 to 2.19. Although the authors considered several covariates, it is important to acknowledge the limitations of the study. Moreover, I suggest including additional variables such as the socioeconomic status of the women evaluated, whether they are mothers or not, and the number of current partners, as these could provide a broader set of relevant factors. It would also be valuable to account for any additional therapies the participants may be using—such as complementary or alternative treatments—that could influence outcomes. These aspects should be incorporated into the manuscript.”

Response:

Thank you for this comment. As recommended, the lines referenced in the reviewer’s comment (previously 2.17–2.19) have been removed to avoid redundancy and improve coherence. With regard to additional variables such as socioeconomic status, motherhood status, number of partners, and use of complementary therapies, these data were not systematically collected across centers and therefore could not be included in the analyses.

To address this, we now explicitly acknowledge in the Limitations section that these unmeasured factors may contribute to residual confounding and represent important targets for future research.

Changes in the manuscript:

• Limitations section: Page 11, Lines 402–408

‘’Fourth, the study did not incorporate broader psychosocial constructs—such as anxiety, depression, body image, fertility concerns, self-esteem, marital satisfaction, or pre-existing relationship conflict—which may meaningfully influence both women’s vulnerability to coercive dynamics and their adjustment to treatment-induced menopause. Additionally, socioeconomic status, motherhood status, and relationship duration or partner history were not collected, and these unmeasured factors may further contribute to relational vulnerability.’’

Comment 4   Discussion Section Improvements

“The discussion is generally well developed; however, there are a few words in bold whose purpose is unclear and should be corrected. It is also important to compare the present findings with similar studies conducted in other regions, such as Latin America, Europe, and elsewhere, where difficulties experienced by women with cancer have also been assessed. These comparisons would strengthen the discussion and contextualize the study more effectively.’’

Response:
We thank the reviewer for this helpful recommendation. All unintended bold formatting has been removed from the Discussion. Additionally, we expanded the Discussion to include comparisons with international studies from Europe, North America, and Latin America that have examined sexual dysfunction, relational strain, and coercive partner behaviors among breast cancer survivors. These additions contextualize our findings within broader global evidence and strengthen the interpretative depth of the Discussion.

Changes in the manuscript:

• Discussion section (newly added text): Page 10, Lines 365–370

“Our findings are consistent with international evidence showing that treatment-related sexual dysfunction can reshape intimate partner dynamics. Studies from Europe, North America, and Latin America have similarly reported increases in relational tension, emotional pressure, and coercive partner responses during survivorship, suggesting that the psychosocial impact of ovarian suppression may extend across diverse cultural settings [9-13].”

Comment 5   Sample Size and Justification

“The authors should explain why the sample consisted of only 81 participants. Were additional participants available? How was this sample size justified? Furthermore, considering the relatively small number of participants in each group, it is crucial to explain how the inclusion of such a large number of variables in the models was justified. These points need to be clearly addressed in the manuscript.’’

Response:
Thank you for this important point. The final sample size (n = 81) reflects the total number of eligible premenopausal breast cancer patients who met inclusion criteria and completed the survey across all three participating centers during the study period. This sample was therefore determined by the available population rather than by selective recruitment.

We also clarify that although the sample size was modest, covariates included in the ANCOVA models were carefully selected based on clinical relevance and prior literature to minimize overfitting. This limitation has now been emphasized in the Limitations section.

Changes in the manuscript:

• Methods → Participants: Page 4, Lines 144–146

“The final sample (n = 81) reflects all eligible premenopausal women who met the inclusion criteria, were approached during the study period across the three centers, and voluntarily agreed to participate.’’

• Limitations: Page 10, Lines 392–393

“The modest sample size may increase the risk of overfitting in adjusted models; therefore, results should be interpreted with caution.”

Comment 6   References

“Overall, the references are adequate, though some errors must be corrected. I also recommend increasing the number of references, particularly by incorporating studies from Mexico, the United States, Latin America, Canada, and Europe that may serve as useful comparisons for the findings presented.’’

Response:

Thank you for this suggestion. The Introduction and Discussion sections have been updated to incorporate additional recent studies from Europe, North America, Canada, Mexico, and Brazil addressing endocrine therapy, relational strain, sexual dysfunction, and coercive partner behavior. These additions broaden the international scope and strengthen the contextual relevance of our findings.

Changes in the manuscript:

• Introduction and Discussion: Added multiple new international references.
• Reference list: Updated accordingly.

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

accept