Medication Adherence in Women with Early-Stage Breast Cancer and Active Parenting Responsibilities: The Mediating Role of Parenting Stress and Spiritual Well-Being

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Medication adherence among women with early-stage breast cancer who have active parenting responsibilities is a significant yet understudied aspect of breast cancer care, addressed in this manuscript. Notable advantages include the use of validated psychometric tools, a multicenter design, and a comparatively large sample size. The mediation analysis goes beyond simple associations, and the conceptual framework linking parenting stress, spiritual well-being, and adherence is cohesive and clinically relevant.
However, before the manuscript can be considered for publication, several issues must be resolved.

-"Active Parenting" definition and operationalization
Although a definition is given, the idea is still ambiguous (e.g., children under the age of 18 vs. adult children in need of care). Participants' levels of caregiving burden probably differ significantly from one another. It is advised to make clear whether subgroup or sensitivity analyses based on the child's age, the number of dependents, or the level of caregiving were taken into account. If not, note this as a limitation and talk about how the results might be impacted by parenting role heterogeneity.

-Higher scores on the MMAS-6 indicate worse adherence, which may seem counterintuitive to readers. Furthermore, MMAS is susceptible to social desirability bias because it is a self-reported scale. In the Methods, Results, and Tables, reiterate the directionality of MMAS-6 scoring consistently. The limitations of self-reported adherence measures and the absence of objective adherence data should be more clearly addressed in the limitations section.

-A modest 11–12% of the variance in adherence is explained by the final model. Extend the Discussion to include the missing variables (e.g., depression, anxiety, social support, health literacy, physician–patient communication) and place the modest R2 within the larger adherence literature.

Author Response

1-"Active Parenting" definition and operationalization
Although a definition is given, the idea is still ambiguous (e.g., children under the age of 18 vs. adult children in need of care). Participants' levels of caregiving burden probably differ significantly from one another. It is advised to make clear whether subgroup or sensitivity analyses based on the child's age, the number of dependents, or the level of caregiving were taken into account. If not, note this as a limitation and talk about how the results might be impacted by parenting role heterogeneity.

Response:
We thank the reviewer for this thoughtful comment regarding the conceptualization of active parenting. In response, we have clarified the operational definition of active parenting in the Methods section by explicitly stating that it includes women with at least one child younger than 18 years of age, as well as those with adult children who continue to live in the same household and require ongoing caregiving responsibilities.

Although subgroup or sensitivity analyses based on child age, number of dependents, or caregiving intensity were not performed, we now explicitly acknowledge this heterogeneity in parenting roles as a study limitation. In the Discussion section, we have added a corresponding statement noting that variability in caregiving burden across participants may have influenced the observed associations and should be considered when interpreting the findings. We also emphasize that future studies incorporating more granular measures of caregiving intensity may help to further refine and contextualize these relationships.

2-Higher scores on the MMAS-6 indicate worse adherence, which may seem counterintuitive to readers. Furthermore, MMAS is susceptible to social desirability bias because it is a self-reported scale. In the Methods, Results, and Tables, reiterate the directionality of MMAS-6 scoring consistently. The limitations of self-reported adherence measures and the absence of objective adherence data should be more clearly addressed in the limitations section.

Response:  We have revised the manuscript to ensure that the directionality of the MMAS-6 is stated clearly and consistently throughout the Methods, Results, tables, and figure legends, explicitly noting that higher MMAS-6 scores indicate poorer medication adherence.

In addition, we have expanded the Discussion section to more clearly address the limitations associated with the use of a self-reported adherence measure. Specifically, we now acknowledge the potential for recall bias and social desirability bias inherent to questionnaire-based adherence assessments, as well as the absence of objective adherence measures. We emphasize that these factors should be taken into account when interpreting the findings.

3-A modest 11–12% of the variance in adherence is explained by the final model. Extend the Discussion to include the missing variables (e.g., depression, anxiety, social support, health literacy, physician–patient communication) and place the modest R2 within the larger adherence literature.

Response: We thank the reviewer for this important observation. In response, we have revised the Discussion section to explicitly contextualize the modest proportion of explained variance within the broader medication adherence literature. We now acknowledge that several well-established determinants of adherence—such as depression, anxiety, social support, health literacy, treatment perceptions, and clinician–patient communication—were not included in the present model and may partly account for the observed R² value. Furthermore, we emphasize that this magnitude of explained variance is consistent with prior adherence studies in oncology, where adherence behavior is recognized as a multifactorial outcome shaped by complex and interacting psychosocial, cognitive, and health system–related factors.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors present a study addressing medication adherence in women with early-stage breast cancer who have active parenting responsibilities, with particular attention to the mediating roles of parenting stress and spiritual well-being. The aim of the study is to advance understanding of adherence behaviors within the lived psychosocial context of early-stage breast cancer. The overall conclusion demonstrates that, among women with early-stage breast cancer and active parenting responsibilities, medication adherence is shaped not only by clinical factors but also by broader psychosocial influences.

General: This is an interesting and well-structured article. The authors present their arguments clearly and support them appropriately.

Major
1. Although the Introduction is generally well written and provides a clear and coherent background, the rationale and novelty of the study could be further strengthened. The authors may consider adding a dedicated paragraph explicitly outlining the existing gaps in the literature and clearly articulating the innovative aspects of the present work, particularly the focus on women with active parenting responsibilities and the mediating role of spiritual well-being in medication adherence.
2. While the overall study design is appropriate and clearly presented, the Methods section would benefit from greater methodological detail to enhance transparency and reproducibility. In particular, the inclusion and exclusion criteria appear relatively limited for a large, multicenter study focusing on psychosocial determinants of medication adherence. The authors are encouraged to clarify several aspects of participant selection and treatment characteristics, including: (1) the specific types of “active systemic therapy” included in the study (e.g., endocrine therapy, oral targeted therapy, chemotherapy, or combinations thereof); (2) whether participants were assessed during ongoing treatment or after treatment initiation, and at what stage of therapy; (3) whether any minimum or maximum duration of systemic treatment was required for eligibility; (4) whether patients with pre-existing or concurrent psychiatric disorders were excluded or otherwise accounted for; and (5) whether eligibility criteria related to language proficiency or reading ability were applied, given the reliance on self-report questionnaires. Providing this additional information would strengthen the methodological rigor of the study and allow for a more precise interpretation of the findings.
3. The statistical analyses applied are generally appropriate for the study aims. However, the Statistical Analysis section would benefit from additional clarification. In particular, further justification for the use of linear regression with a non-normally distributed, ordinal adherence scale (MMAS-6) would strengthen the methodological transparency. Moreover, given the cross-sectional design, it would be helpful to explicitly acknowledge the interpretative limitations of mediation analysis and clarify the selection of variables included in the regression models.
4. The manuscript would benefit from the inclusion of a dedicated table presenting the baseline characteristics of the study population. In addition to oncological and treatment-related variables, such a table should include key sociodemographic factors such as place of residence (urban vs. rural), educational level, and other relevant contextual variables. Providing a comprehensive descriptive overview of the study sample would improve clarity and facilitate interpretation of the subsequent analyses.
5. Although several sociodemographic variables (e.g., place of residence, educational level, and economic status) are included in the analytical comparisons presented in Tables 1 and 2, the manuscript lacks a comprehensive descriptive characterization of the study population. These variables are neither sufficiently described in the Methods section nor summarized in the opening paragraph of the Results section. The inclusion of a dedicated baseline characteristics table, along with a brief narrative description of the study sample, would substantially improve the clarity and contextualization of the subsequent analyses. At present, sociodemographic variables are introduced directly in inferential analyses without prior descriptive presentation, which limits the reader’s ability to fully interpret the reported group differences.
6. The Discussion is well written and thoughtfully integrates the findings with existing literature. While the authors do touch upon potential clinical relevance, the practical implications of the results remain largely implicit. The manuscript would benefit from a more explicit articulation of how the observed associations between parenting stress, spiritual well-being, and medication adherence could be translated into routine clinical practice, supportive care strategies, or early identification of patients at risk for non-adherence.
7. The final paragraph of the Discussion begins with “In conclusion” and largely overlaps with the content of the subsequent Conclusion section. To improve clarity and avoid redundancy, the authors may consider revising the structure by either removing the concluding summary from the Discussion or streamlining the Conclusion section to ensure a clearer distinction between interpretation of results and final take-home messages.
8. While some limitations are mentioned in the Discussion, they are embedded within a broader narrative and would benefit from being more clearly delineated. The manuscript would be strengthened by a more explicit and structured discussion of study limitations, particularly those related to the questionnaire-based design and the generalizability of the findings.

Author Response

comment 1. Although the Introduction is generally well written and provides a clear and coherent background, the rationale and novelty of the study could be further strengthened. The authors may consider adding a dedicated paragraph explicitly outlining the existing gaps in the literature and clearly articulating the innovative aspects of the present work, particularly the focus on women with active parenting responsibilities and the mediating role of spiritual well-being in medication adherence.

Response1:Thank you for this valuable suggestion. In response, we added a dedicated paragraph to the Introduction explicitly outlining the gaps in the existing literature and clarifying the novelty of the present study. In this paragraph, we emphasize that prior adherence research has largely overlooked active parenting responsibilities and that the mediating role of spiritual well-being in medication adherence has not been systematically examined. We believe this addition strengthens the rationale and clearly highlights the innovative aspects of our work.

Comment2:While the overall study design is appropriate and clearly presented, the Methods section would benefit from greater methodological detail to enhance transparency and reproducibility. In particular, the inclusion and exclusion criteria appear relatively limited for a large, multicenter study focusing on psychosocial determinants of medication adherence. The authors are encouraged to clarify several aspects of participant selection and treatment characteristics, including: (1) the specific types of “active systemic therapy” included in the study (e.g., endocrine therapy, oral targeted therapy, chemotherapy, or combinations thereof); (2) whether participants were assessed during ongoing treatment or after treatment initiation, and at what stage of therapy; (3) whether any minimum or maximum duration of systemic treatment was required for eligibility; (4) whether patients with pre-existing or concurrent psychiatric disorders were excluded or otherwise accounted for; and (5) whether eligibility criteria related to language proficiency or reading ability were applied, given the reliance on self-report questionnaires. Providing this additional information would strengthen the methodological rigor of the study and allow for a more precise interpretation of the findings.

Response2:

In response, we have expanded the Methods section to improve methodological transparency and reproducibility by clarifying participant selection and treatment-related characteristics.

Specifically, we have now explicitly stated that active systemic therapy included adjuvant or neoadjuvant chemotherapy, endocrine therapy, and/or oral systemic treatments administered with curative intent. Participants were assessed during ongoing systemic treatment at routine outpatient follow-up visits, and no minimum or maximum duration of systemic therapy was required for eligibility.

Regarding psychiatric conditions, patients with severe psychiatric disorders or cognitive impairment that could interfere with the reliable completion of self-report questionnaires were excluded, and this criterion has been clearly specified. In addition, given the questionnaire-based design, we clarified that all participants were required to have sufficient Turkish language proficiency to complete the self-report measures. We believe that these additions strengthen the methodological rigor of the study and allow for a more precise interpretation of the findings, as suggested.

Comment 3:  The statistical analyses applied are generally appropriate for the study aims. However, the Statistical Analysis section would benefit from additional clarification. In particular, further justification for the use of linear regression with a non-normally distributed, ordinal adherence scale (MMAS-6) would strengthen the methodological transparency. Moreover, given the cross-sectional design, it would be helpful to explicitly acknowledge the interpretative limitations of mediation analysis and clarify the selection of variables included in the regression models.

response3: We thank the reviewer for this thoughtful methodological comment. In response, we have added a clarification to the Statistical Analysis section explaining that, although the MMAS-6 is an ordinal scale and demonstrated non-normal distribution, total scores were treated as approximately continuous variables, as commonly accepted in adherence research, thereby allowing the use of linear regression models. In addition, we explicitly addressed the interpretative limitations of mediation analysis in the Discussion, emphasizing that the mediation findings should be interpreted as statistical associations rather than evidence of temporal or causal mechanisms due to the cross-sectional study design. Finally, we clarified that the selection of parenting stress and spiritual well-being in the regression and mediation models was theory-driven and based on the study’s conceptual framework and prior literature, while acknowledging that other psychosocial and clinical determinants of medication adherence were not included in the models.

Comment4. The manuscript would benefit from the inclusion of a dedicated table presenting the baseline characteristics of the study population. In addition to oncological and treatment-related variables, such a table should include key sociodemographic factors such as place of residence (urban vs. rural), educational level, and other relevant contextual variables. Providing a comprehensive descriptive overview of the study sample would improve clarity and facilitate interpretation of the subsequent analyses.

Responce 4: In response, we have added a new table presenting the baseline sociodemographic and clinical characteristics of the study population (Table 1). This table provides a comprehensive descriptive overview of the cohort, including disease stage, treatment modalities, and key sociodemographic variables, to facilitate clearer interpretation of the subsequent analyses

comment5: 5. Although several sociodemographic variables (e.g., place of residence, educational level, and economic status) are included in the analytical comparisons presented in Tables 1 and 2, the manuscript lacks a comprehensive descriptive characterization of the study population. These variables are neither sufficiently described in the Methods section nor summarized in the opening paragraph of the Results section. The inclusion of a dedicated baseline characteristics table, along with a brief narrative description of the study sample, would substantially improve the clarity and contextualization of the subsequent analyses. At present, sociodemographic variables are introduced directly in inferential analyses without prior descriptive presentation, which limits the reader’s ability to fully interpret the reported group differences.

Responce5: we have added a dedicated baseline characteristics table (Table 1) summarizing the sociodemographic, caregiving-related, and clinical features of the study population. In addition, a brief descriptive sentence has been included at the beginning of the Results section to introduce and contextualize the cohort prior to inferential analyses.

comment 6:. The Discussion is well written and thoughtfully integrates the findings with existing literature. While the authors do touch upon potential clinical relevance, the practical implications of the results remain largely implicit. The manuscript would benefit from a more explicit articulation of how the observed associations between parenting stress, spiritual well-being, and medication adherence could be translated into routine clinical practice, supportive care strategies, or early identification of patients at risk for non-adherence.

Responce6:In response, we have revised the final paragraph of the Discussion to explicitly articulate the clinical implications of our findings. We now clarify how the observed associations between parenting stress, spiritual well-being, and medication adherence may inform early identification of vulnerable patients and support the development of integrated, psychosocially informed adherence strategies in routine clinical practice (Discussion, last paragraph).

Comment7: . The final paragraph of the Discussion begins with “In conclusion” and largely overlaps with the content of the subsequent Conclusion section. To improve clarity and avoid redundancy, the authors may consider revising the structure by either removing the concluding summary from the Discussion or streamlining the Conclusion section to ensure

Responce7:Thank you for this comment. To avoid redundancy between the Discussion and the Conclusion sections, we revised the opening of the final paragraph in the Discussion and reframed it to emphasize clinical implications rather than a general summary. The paragraph now begins with “From a clinical perspective,” and focuses on practical and translational aspects of the findings. In parallel, the Conclusion section was streamlined to retain a concise synthesis of the main results while preserving a clear clinical emphasis.

comment8. While some limitations are mentioned in the Discussion, they are embedded within a broader narrative and would benefit from being more clearly delineated. The manuscript would be strengthened by a more explicit and structured discussion of study limitations, particularly those related to the questionnaire-based design and the generalizability of the findings.

Responce 8:We have revised the Discussion to more clearly delineate the study limitations by explicitly grouping them within a dedicated paragraph. This section now highlights the constraints related to the questionnaire-based, cross-sectional design, self-reported adherence, and the limited generalizability of the findings. These revisions aim to provide a more transparent and structured presentation of the study’s limitations.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The study evaluates medication adherence within the psychosocial context of active parenting responsibilities, a critical but underexplored aspect of breast cancer treatment. The multicenter design, adequate sample size, and use of validated tools (MMAS-6, PSS, FACIT-Sp-12) enhance the study's robustness. The mediation analysis provides a unique and clinically beneficial addition. The study is well-executed and recorded; nonetheless, several conceptual, methodological, and reporting issues must be resolved to enhance clarity, rigor, and interpretability.

Major Comments

While the authors use mediation analysis appropriately, the cross-sectional methodology constrains causal conclusions. The phrasing in several sections (Abstract, Results interpretation, Discussion) suggests explicit associations (for instance, "Parenting stress reduces adherence through spiritual well-being"). Modify the content to emphasize correlations instead of causation. The mediation findings should be characterized as statistical mediation rather than mechanistic or temporal mediation.

While Hayes' PROCESS Model 4 is commonly utilized, its utility with cross-sectional data is debated. Provide a strong methodological justification for performing mediation analysis in cross-sectional data.  The limitations section should clarify that temporal ordering of parental stress, spiritual well-being, and adherence cannot be confirmed.

The multiple linear regression model accounts for about 11.6% of the variance in MMAS-6 outcomes.  The correlation between this moderate explanatory power and other established determinants of adherence, such as cognitive traits, health literacy, treatment perceptions, and clinician-patient communication, warrants discussion.  The discussion should elucidate the therapeutic implications of statistically significant but small effect sizes.

Medication adherence was measured using self-reported MMAS-6, which is susceptible to recall and social desirability biases.  This constraint should be addressed. The authors should explain whether MMAS-6 scores were regarded as continuous variables and justify their methodology given their ordinal nature.

While numerous sociodemographic and clinical factors were examined descriptively, the rationale for the inclusion of only parental stress and spiritual well-being in the final regression models remains ambiguous.  Please clarify the variable selection technique for multivariate and hierarchical regression studies. Discuss also possible confounders, such as financial status, pharmaceutical use, and hormone treatment, and explain why they were excluded.

The study population comprised women receiving treatment at various cancer facilities in Turkey. Please elaborate on the discourse regarding external validity, emphasizing its significance to non-Turkish populations, diverse cultural contexts, and healthcare systems characterized by differing family support frameworks.

The conclusion recommends integrated care approaches, psychological screening, and meaning-centered therapies. While acceptable, these suggestions go beyond what can be concluded from cross-sectional data.  The authors should emphasize that these implications generate hypotheses and require confirmation through longitudinal or interventional investigations. 

The conclusion basically reiterates concerns raised in the later portion of the discussion, notably those concerning multidimensional adherence, parental stress, and spiritual well-being. Please reduce the length of the conclusion and concentrate on essential findings and future directions.

Minor Comments

The title is informative yet quite lengthy. Consider slight simplifying to improve readability while maintaining key constructs.

The background is strong but could be slightly condensed, particularly in the sections describing general psychosocial burden, to sharpen focus on the study’s unique contribution.

Ensure that effect sizes are consistently reported, accompanied by p-values when applicable.  Also, clarify if linear regression assumptions (e.g., residual normality and homoscedasticity) have been tested or visually evaluated.

Tables are comprehensive but dense. Consider if any sociodemographic comparisons should be transferred to supplemental materials. 

Figure 1 should need a quick explanatory remark to emphasize that the pathways represent statistical data.

Ethical approval is reported; however, please clarify whether participants received any psychosocial support referrals if high stress or low spiritual well-being was identified.

 

 

 

Author Response

Response to the Reviewer

We would like to thank the Reviewer for the detailed and constructive evaluation of our manuscript. The positive comments regarding the originality of our study, its multicenter design, adequate sample size, and the use of validated measurement tools are highly encouraging for us. Below, we provide our point-by-point responses to all Major and Minor comments, together with the corresponding revisions made in the manuscript.

Major Comments

1. Use of causal language and interpretation of mediation

Reviewer’s comment: Although the study has a cross-sectional design, some sections contain language implying causality. The mediation findings should be interpreted in statistical terms.

Our response:

In line with this important feedback, all statements implying causality in the Abstract, interpretation of the Results, Discussion, and Conclusion sections were carefully reviewed and revised. The findings are now explicitly presented within the framework of “statistical associations” and “statistical mediation.” Throughout the manuscript, potentially causal expressions such as “effect,” “reduces,” or “causes” have been avoided and replaced with terms such as “is associated with,” “is related to,” and “can be statistically explained.”

2. Justification of mediation analysis in cross-sectional data

Reviewer’s comment: Although Hayes’ PROCESS Model 4 is commonly used, mediation analysis in cross-sectional data is debatable. A strong methodological justification should be provided, and the lack of temporal ordering must be clearly acknowledged.

Our response:

In response to this comment, explanatory statements were added to the Statistical Analysis section and the Limitations section. It is now clearly stated that the mediation analysis was not conducted to infer causality or temporal relationships, but rather to examine statistical pathways among variables within a pre-defined theoretical psychosocial framework. In addition, it is explicitly acknowledged that the temporal ordering of parenting stress, spiritual well-being, and medication adherence cannot be confirmed with the present study design.

3. Explanatory power of the regression model (R² ≈ 0.12)

Reviewer’s comment:The moderate explanatory power of the model should be discussed in relation to other known determinants of adherence. The clinical relevance of small effect sizes should also be clarified.

Our response:The relevant paragraph in the Discussion section was expanded to emphasize that medication adherence is inherently a multidimensional behavioral outcome. It was noted that well-established factors in the literature—such as cognitive characteristics, health literacy, treatment beliefs, and physician–patient communication—were not included in the model and may have limited its explanatory capacity. Furthermore, despite the small effect sizes, it was discussed that modifiable psychosocial factors such as parenting stress and spiritual well-being may still yield clinically meaningful outcomes when addressed through cumulative or holistic interventions.

4. Self-report and ordinal nature of the MMAS-6

Reviewer’s comment: The MMAS-6 is self-reported and may be subject to bias. Moreover, as an ordinal scale, its treatment as a continuous variable should be justified.

Our response:

This limitation has been explicitly acknowledged in the Limitations section. In addition, the Statistical Analysis section now justifies that, in line with common practice in the literature, MMAS-6 scores were treated as approximately continuous to allow regression modeling. The risks of recall bias and social desirability bias are also clearly stated.

5. Variable selection in the regression models

Reviewer’s comment: Why were only parenting stress and spiritual well-being included in the regression models? Why were potential confounders excluded?

Our response:

The variables included in the multivariable and hierarchical regression models were not selected through an automated statistical screening process but were chosen based on the study’s primary research question and a pre-defined theoretical framework. The main aim of the analyses was to examine the relationship between parenting stress and medication adherence and to evaluate the potential mediating role of spiritual well-being. Therefore, the regression and mediation analyses were specifically focused on testing the hypothesized psychosocial pathway (PSS → FACIT-Sp-12 → MMAS-6).

Sociodemographic and clinical variables (e.g., income status, place of residence, number of medications, and type of hormonal therapy) were evaluated in descriptive and contextual analyses prior to the regression models. Their associations with the scale scores were examined using appropriate non-parametric methods and are reported in detail in Tables 3 and 4. At this stage, no univariable regression analyses were conducted; relationships were assessed using correlation and group comparison approaches.

Although some of these variables were associated with certain scale scores, they were not conceptualized as direct confounders within the core psychosocial model and were therefore not included in the final regression analyses. This decision was made to avoid overadjustment and overfitting in a cross-sectional design and to maintain the explanatory, theory-driven nature of the regression models rather than a purely predictive approach.

6. External validity

Reviewer’s comment:The study was conducted in Türkiye. External validity across different cultures and healthcare systems should be discussed.

Our response:

An additional paragraph was included in the Discussion section to emphasize that the findings may be sensitive to cultural context, family support structures, and healthcare system characteristics. The results are therefore presented not as universally generalizable, but as findings that should be tested in similar sociocultural settings.

7. Scope and tone of the Conclusion section

Reviewer’s comment: The clinical recommendations in the Conclusion are too strong for cross-sectional data. The section is also lengthy and repetitive.

Our response:

The Conclusion section was substantially shortened, repetitive statements were removed, and the tone was revised to reflect a hypothesis-generating framework. Clinical implications were softened, and it is now clearly stated that the findings require confirmation in longitudinal and interventional studies.

 

Minor Comments

8. Title length

Reviewer’s comment:The title is long.

Our response:

The title was carefully re-evaluated; however, it was not shortened because it clearly reflects the target population, the key psychosocial variables, and the analytical framework of the study. Considering that descriptive titles transparently conveying the scope of a study are increasingly adopted in the field of psychosocial oncology, we believe that the current title is consistent with the content and is not excessively long.

9. Density of the Introduction

Reviewer’s comment: The Introduction is dense.

Our response:

The Introduction was reviewed; however, its current structure was considered necessary to establish the psychosocial and theoretical background of the study, to justify the conceptual model, and to clearly define the research gap addressed. Given the complexity of the constructs examined, further shortening was considered likely to lead to oversimplification of the framework.

10. Effect sizes and regression assumptions

Reviewer’s comment: Effect sizes and regression assumptions should be clarified.

Our response:

Effect sizes (standardized β coefficients, R² and ΔR² values) are now reported consistently together with p-values in the Results section (Tables 6 and 7). In addition, a statement has been added to the Statistical Analysis section indicating that regression assumptions (multicollinearity, normality of residuals, and homoscedasticity) were tested and met.

11. Density of tables

Reviewer’s comment: There are too many tables.

Our response:

All tables were carefully reviewed and retained in the main manuscript because each presents a distinct and clinically meaningful dimension of the data. In addition, upon the request of another reviewer, selected tables have been prepared for possible inclusion as supplementary material, if required by the Editorial Office.

12. Psychosocial support referral

Reviewer’s comment: Statements regarding referral to psychosocial support should be clarified.

Our response:

The relevant sentences in the Discussion and Conclusion sections were revised to emphasize that referrals to psychosocial support services are presented as supportive and hypothesis-generating considerations rather than prescriptive clinical recommendations, in line with the cross-sectional design of the study.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have addressed and incorporated all my comments and suggestions, and I accept the manuscript in its current form.

Reviewer 3 Report

Comments and Suggestions for Authors

All major and minor reviewer comments have been fully and completely addressed by the authors. Methodological clarity, statistical interpretation, and consistency between research design and findings are all clearly improved in the updated text. The manuscript text has adequately acknowledged and addressed issues with causal language, cross-sectional mediation analysis, the explanatory power of the regression models, and the use of self-reported adherence measures.

The changes improve the findings' interpretability and scientific rigor. As a result, I suggest that the submission be accepted (or, at the editor's option, accepted with minor editorial revisions).