The Relationship Between Climate Change Worry and Symptoms of Stress, Anxiety, and Depression in Turkish Pregnant Women: A Cross-Sectional Study

Reviewer 1 Comment

Response

Abstract & Methods: "Data were collected using Data were collected using..." is a duplication error.

The duplication error in the Abstract has been corrected.

Cross-Sectional Design -The Fundamental Limitation: The authors correctly note this, but its implications warrant greater emphasis. The design cannot establish causality or directionality.

The Discussion section has been revised to more explicitly emphasize that the cross-sectional design does not allow causal or directional interpretations. The findings are now consistently framed as associations rather than causal effects, and the study is positioned as identifying a vulnerable subgroup experiencing concurrent distress.

It is equally plausible that: Climate change worry causes increased anxiety/depression; Preexisting anxiety or depressive tendencies lead to higher levels of worry about global threats like climate change; A third, unmeasured variable drives both climate worry and psychological distress.

Alternative explanations suggested by the reviewer have been explicitly acknowledged and discussed in the Discussion as potential explanations for the observed associations.

Sampling and Generalizability – Convenience Sampling: The use of a convenience sample from a single hospital in one region (Van) severely limits the representativeness of the findings.        

The Limitations section has been expanded to clearly state that the findings are preliminary, context-specific, and limited by convenience sampling from a single geographic region.

Sample Characteristics: Nearly half (48.2%) reported no concern about climate change, which is a striking finding in itself. This suggests the sample may include a large proportion of individuals for whom climate change is not a salient issue, potentially skewing the mean scores and correlations. The generalizability to populations in urban centers or regions more directly impacted by climate events (e.g., coastal areas, agricultural zones) is questionable.

This finding has been addressed in the Discussion section in line with the reviewer’s suggestions, and additional references have been added.

References:

40.       Negiz, N.; Hayta, Y. Environmental behavior and environment education: A study on university students. Çağdaş Yerel Yöne-timler Dergisi 2021, 30(1), 149-179.

41.       Oğuz, D.; Çakcı, I.; Kavas, S. Environmental awareness of students in higher education. SDU Faculty of Forestry Journal 2011, 12(1), 34-39.

42.       Toptaş Acar, B.; Gerçek Öter, E. Climate change awareness in pregnant women: A qualitative study. Ordu Üniversitesi Hemşirelik Çalışmaları Dergisi 2024, 7(1), 38-45. https://doi.org/10.38108/ouhcd.1263410

Conceptual and Measurement Issues – “Worry” vs. “Anxiety”: The manuscript uses "Climate Change Worry Scale" but frequently discusses "anxiety."

Terminology has been revised throughout the Results and Discussion sections to consistently refer to “climate change worry” when describing findings derived from the CCWS, with distinctions clarified in the Introduction.

Controlling for Confounds: The analysis lacks control for critical confounding variables.

A detailed statement has been added to the Limitations section acknowledging the absence of controls for baseline mental health, direct climate event exposure, media exposure, and social support, and emphasizing this as a direction for future research.

Recommendation: Future research must employ multivariate regression models or structural equation modeling.      

This recommendation has been addressed in the Discussion and Limitations sections.

Minor Typos and Formatting:

Table 1: "Seconder school" is a typo for "Secondary school".

Methods, Section 2.3: "Completion of the Personal Information Form..." -The line break in the PDF makes it seem like the sentence is cut off. Ensure the text flows correctly in the final format.

References: Ensure consistent formatting of journal names, use of italics, and DOI presentation per the journal's style guide.

These issues have been corrected.

It is hoped that the revisions made have resolved this problem. Sincerely.

The reference list has been reviewed and formatted in accordance with the journal’s guidelines. New references have been added to the newly included paragraphs in line with the reviewers’ suggestions, and these references have been highlighted in yellow.

Introduction: The explanation of terms (solastalgia, eco-anxiety) is useful. However, the claim that eco-anxiety is a "broader concept" with climate anxiety as a subdimension is debatable and should be referenced more precisely (reference 14 is appropriate but could be engaged with more).

An additional reference has been added.

Reference: 15.Coffey, Y.; Bhullar, N.; Durkin, J.; Islam, M.S.; Usher, K. Understanding eco-anxiety: A systematic scoping review of current literature and identified knowledge gaps. The Journal of Climate Change and Health 2021, 3, 100047.https://doi.org/10.1016/j.joclim.2021.100047

Ethics: Approval is noted from "XXX". The name of the Institutional Review Board (IRB) or Ethics Committee must be disclosed for transparency.

The full name of the approving Ethics Committee has been added to the Ethics section.

Inconsistent Terminology: The manuscript fluctuates between multiple related terms.

Terminology has been standardized throughout the manuscript in line with the reviewer’s recommendation.

Results, Caption for Table 2: "Obtained min–max" and "Attainable min–max" are understandable but not standard. Preferred terms are "Observed range" and "Possible range".

This has been corrected in line with the reviewer’s recommendation.

The "No Opinion" Group: In Table 4, the "No opinion" group on regional impact had the highest depression scores. This is a fascinating result that is not adequately discussed. This group may represent individuals feeling overwhelmed, disengaged, or information-avoidant—a potentially high-risk psychological profile meriting exploration.

This finding has been addressed in the Discussion section in line with the reviewer’s suggestions, and additional references have been added.

References:

54.       Ådnøy, T.; Solem, S.; Hagen, R; Havnen A. An empirical investigation of the associations between metacognition, mindful-ness experiential avoidance, depression, and anxiety. BMC Psychol. 2023, 11, 281. https://doi.org/10.1186/s40359-023-01336-7

55.       Akbari, M.; Seydavi, M.; Hosseini, Z.S.; Krafft, J.; Levin, M. E. Experiential avoidance in depression, anxiety, obses-sive-compulsive related, and posttraumatic stress disorders: A comprehensive systematic review and meta-analysis. Journal of Contextual Behavioral Science 2022, 24(2022) 65–78. https://doi.org/10.1016/j.jcbs.2022.03.007

56.       Fledderus, M.; Bohlmeijer, E.T.; Pieterse, M.E. Does experiential avoidance mediate the effects of maladaptive coping styles on psychopathology and mental health?. Behavior modification 2010, 34(6), 503–519. https://doi.org/10.1177/0145445510378379

Controlling for Confounds: The analysis lacks control for critical confounding variables:

General Anxiety/Depression: Without controlling for baseline or pre-pregnancy mental health, it is impossible to attribute DASS-21 scores specifically to climate worry.

Direct Exposure: The study did not measure personal exposure to climate-related events (e.g., floods, droughts, extreme heat), which is a potent predictor of mental health outcomes. Perceived impact is measured, but not actual experience.

Media Exposure: The frequency and type of media consumption about climate change could be a major driver of both worry and distress.

Social Support: This is a key moderator of stress in pregnancy and was not included.

·                 We thank the reviewer for their valuable suggestion. The following sentence has been added.

“the study focused only on certain demographic characteristics and psychological variables; potential confounding factors such as environmental exposure levels, frequency of exposure to information regarding the impacts of climate change, or social support systems were not controlled. Therefore, the use of multivariate models is recommended in future research.”

Tables 3 and 4 Readability: The tables, while comprehensive, are dense and difficult to parse. Using letters (e.g., 4>1,2,3) to denote post-hoc group differences is non-standard and confusing. A clearer presentation (e.g., superscript letters) is needed for publication.

We express our gratitude to the referee for this recommendation.

Tables 3 and 4 have been completely revised. The previous tables used parametric tests. The new tables use non-parametric tests. The reporting has also been completely changed. Significant differences between groups are shown with letters instead of numbers.

Power Analysis: Post-hoc power analysis is reported. It is generally more informative to report a priori power analysis. However, the large effect size (d=0.64) and achieved power of 1.00 indicate the sample was adequate to detect the observed effects.

We express our gratitude to the referee for this recommendation.

The reviewer indicated that the power analysis should have been conducted a priori. As authors, we clearly explained how and when we conducted the power analysis. We evaluated the adequacy of our sample size through the analysis performed and concluded that it was sufficient.

Table Presentation (Critical for Clarity):

Tables 3 and 4: The notation for post-hoc comparisons (e.g., 4>1.2.3) is confusing and not standard academic practice. It must be changed. Recommendation: Use superscript letters (e.g., a, b, c) to denote statistically homogeneous subsets, explained in a table footnote. For example: "Means with different superscript letters within a column are significantly different (p < .05)."

We express our gratitude to the referee for this recommendation.

Tables 3 and 4 have been completely revised. The reporting has also been completely changed. Significant differences between groups are shown with letters instead of numbers.

Reviewer 2 Comment

Response

One of the main problems with the manuscript lies in the justification and application of the statistical analyses. The authors indicate that the variables did not follow a normal distribution according to the Kolmogorov-Smirnov test, but subsequently decide to apply parametric tests (ANOVA and Pearson correlation) based solely on the sample size and acceptable values of skewness and kurtosis. This reasoning is methodologically weak. A large sample size does not guarantee normality of the data, and the central limit theorem does not, in itself, justify the use of parametric tests on individual data points. Furthermore, no key checks associated with these tests are reported, such as the homogeneity of variances for ANOVA or the inspection of residuals and outliers. The absence of these checks limits the validity of the results. To strengthen the analysis, the use of parametric tests should be more rigorously justified, or alternatively, the results should be supplemented with nonparametric analyses (e.g., Spearman correlations or robust Welch-Welch ANOVA where appropriate). In its current form, the reader cannot assess whether the statistical assumptions are adequately met.

We express our gratitude to the referee for this recommendation. The following sentence has been added.

Following the referee's recommendation, the data analysis has been restructured as follows. Previously performed parametric data analyses have been deleted, and all analyses have been redone using non-parametric tests.

“The first step is to establish whether the distribution of the data is suitable for the effective use of statistical procedures. The Kolmogorov-Smirnov test for normality showed that the data were not normally distributed (p < .05). Based on this result, Kruskal-Wallis H was used to look at the differences in group means when comparing three or more groups. For variables with statistically significant differences, Mann-Whitney U test was used to identify which groups differ from each other. Spearman rho analysis was performed to examine the relationships between continuous variables.”

Although the stated objective is to examine the relationship between concern about climate change and symptoms of depression, anxiety, and stress, the analytical approach is limited almost exclusively to bivariate correlations and simple comparisons between groups. This approach is insufficient given the complexity of the phenomenon under study. The relationship between climate concern and psychological distress may be influenced by multiple relevant confounding factors, such as socioeconomic status, education, trimester of pregnancy, number of previous pregnancies, or unassessed psychological vulnerabilities. Because it does not incorporate multivariate analysis, the study does not allow us to discern whether the observed association is independent or mediated by other variables. Furthermore, the cross-sectional design precludes any directional inference, but some passages in the discussion implicitly suggest that climate concern contributes to psychological distress, without sufficiently acknowledging that the relationship could be inverse or bidirectional (for example, women with higher baseline anxiety might show greater concern about climate change).

We express our gratitude to the referee for this recommendation.

While agreeing with the reviewer's comment, advanced analyses (e.g., linear regression/GLM) were not performed because the necessary assumptions for such analyses were not met. Additions have been made to the study's limitations in line with the reviewer's comment.

“the study focused only on certain demographic characteristics and psychological variables; potential confounding factors such as environmental exposure levels, frequency of exposure to information regarding the impacts of climate change, or social support systems were not controlled. Therefore, the use of multivariate models is recommended in future research.”

Another problematic aspect is the inclusion of variables such as “level of concern about climate change” or “expectation that the region will be affected by climate change” as explanatory factors for scores on the Climate Concern Scale (CCWS). These variables appear to measure constructs that are conceptually very close, if not identical, to the main construct assessed by the scale. This conceptual overlap creates a risk of analytical circularity, since the observed differences could simply reflect different ways of measuring the same phenomenon. If the goal of these comparisons is to provide evidence of convergent validity of the scale, this should be clearly stated and discussed as such. Otherwise, the explanatory value of these analyses is limited and should be reconsidered.

We express our gratitude to the referee for this recommendation.

The purpose of using these variables is not to provide evidence of the convergent validity of the scale. These variables are intended to collect data on climate change based on participants' self-reports, taking into account their region and socio-cultural context, independently of the scale.

The manuscript repeatedly emphasizes the statistical significance of the results (p < .001), but does not provide effect sizes or an interpretation of their clinical relevance. Given the relatively large sample size, statistically significant associations are to be expected even when the effects are modest. Furthermore, the mean scores for depression, anxiety, and stress appear to be in the low ranges, which raises questions about the true magnitude of the clinical impact of the phenomenon studied. Without an analysis of effect sizes or the proportion of participants exceeding clinically relevant cut-off points, it is difficult to assess the practical importance of the findings.

We express our gratitude to the referee for this recommendation.

Following the referee's recommendation, the data analysis has been restructured as follows. Previously performed parametric data analyses have been deleted, and all analyses have been redone using non-parametric tests.

The manuscript also presents reporting deficiencies that affect its transparency and reproducibility. The recruitment process and the number of participants excluded or who declined to participate are not described in sufficient detail. Furthermore, the handling of missing data is not specified, nor are the post-hoc procedures used after the analysis of variance detailed.

We express our gratitude to the referee for this recommendation. The following sentence has been added.

The study analyses were modified to be non-parametric, and all findings were rewritten.

“Considering the participants' voluntary participation, the exact number of women who were informed about the study but declined to participate is unknown, but it is estimated to be between 40 and 60. Furthermore, because the data collection tools used in the study were very brief, there was no missing data.”

Reviewer 3 Comment

Response

Crucially, the current analyses lack a multivariate approach. Correlations and ANOVA do not control for important covariates such as age, trimester of pregnancy, number of pregnancies, socioeconomic status, subjective assessment of the impact of climate change, or the self-reported level of concern declared in the demographic questionnaire. I recommend regression-based models (e.g., linear regression/GLM), with DASS-21 subscales as dependent variables and CCWS as the main predictor, while controlling for relevant covariates. This would substantially strengthen the analytical argumentation.

We express our gratitude to the referee for this recommendation.

Parametric analyses used in the study were removed in light of other referee suggestions. Since the study data did not show a normal distribution, non-parametric tests were used. The linear regression/GLM analysis suggested by the referee was not performed because its assumptions could not be met.

Statistical analysis: consistency of assumptions and reporting

The authors report that the Kolmogorov–Smirnov test indicated non-normality, yet subsequently assume normality based on skewness and kurtosis and proceed with ANOVA and Pearson correlations. The analytical approach should be unified, either by consistently justifying the use of parametric methods (including tests of homogeneity of variance and, if appropriate, Welch’s ANOVA), or by applying non-parametric alternatives (e.g., Kruskal–Wallis tests, Spearman correlations) or robust methods.

We express our gratitude to the referee for this recommendation.

Following the referee's recommendation, the data analysis has been restructured. Previously performed parametric data analyses have been deleted, and all analyses have been redone using non-parametric tests such as Kruskal-Wallis H and Spearman rho.

Please also specify which post-hoc tests were used (e.g., Tukey, Games–Howell) and report effect size measures (e.g., η²/ω² for ANOVA, correlation coefficients with 95% confidence intervals). Currently, the results are reported mainly in terms of F statistics, p-values, and means. Moreover, given the large number of comparisons (multiple grouping variables × multiple subscales), it would be advisable to consider controlling for Type I error (e.g., Benjamini–Hochberg correction) or to explicitly justify the absence of such corrections in light of the exploratory nature of the study.

We express our gratitude to the referee for this recommendation.

Following the referee's recommendation, the data analysis has been restructured. Previously performed parametric data analyses have been deleted, and all analyses have been redone using non-parametric tests such as Kruskal-Wallis H and Spearman rho. All findings have been reported again.

Tables: please standardize notation (X±SD vs. M±SD), clarify abbreviations, and consistently report degrees of freedom (df) in ANOVA tables.

We express our gratitude to the referee for this recommendation.

Following the referee's recommendation, the data analysis has been restructured. Previously performed parametric data analyses have been deleted, and all analyses have been redone using non-parametric tests such as Kruskal-Wallis H and Spearman rho. All findings have been reported again.

Methods: clarify the coding of ordinal variables (e.g., concern levels), handling of missing data (if any), and whether the interview procedure was standardized.

We express our gratitude to the referee for this recommendation. The following sentence has been added.

“Considering the participants' voluntary participation, the exact number of women who were informed about the study but declined to participate is unknown, but it is estimated to be between 40 and 60. Furthermore, because the data collection tools used in the study were very brief, there was no missing data.”

Reviewer 1 Comment

Response

Tables 3 and 4 – Readability and Formatting

We sincerely thank the reviewer for pointing out the readability and formatting issues in Tables.

Current Issue: The tables are dense and challenging to interpret. The use of non-standard notation (e.g., d>a,b,c) for post-hoc comparisons is confusing.

We appreciate the reviewer’s careful observation regarding the notation and density of the tables. We agree that the previous format may have been difficult to interpret.

Requested Change: Please reformat Tables 3 and 4 using superscript letters (e.g., a, b) to denote statistically significant group differences. Include a clear footnote explaining the notation (e.g., “Groups sharing the same superscript letter do not differ significantly”). Consider simplifying the layout of Table 3 to improve readability.

Thank you for the constructive suggestion. We have reformatted Tables 3 and 4 using superscript letters to indicate statistically significant differences and added a clear footnote explaining this notation. Additionally, Table 3 has been simplified to enhance readability. We believe these changes address the reviewer’s concerns and improve the clarity of the presentation.

Conceptual Distinction Between “Worry” and “Anxiety”

Current Issue: While the CCWS measures “climate change worry,” its “anxiety” subscale and the broader DASS-21 “anxiety” scale are conceptually related but distinct. The text occasionally uses these terms interchangeably, which may cause confusion.

We thank the reviewer for this important observation. We acknowledge that “worry” and “anxiety” are conceptually related yet distinct constructs. We have carefully reviewed the manuscript and attempted to clarify the distinction where possible. We hope these revisions reduce potential confusion while maintaining the conceptual nuance of our measures.

Requested Change: Please add a brief clarifying sentence in the Methods section (when introducing the CCWS, around lines 120–130) to distinguish between the constructs. For example: “It should be noted that the ‘anxiety’ subscale of the CCWS assesses worry related to climate change, while the DASS-21 anxiety subscale measures broader symptoms of clinical anxiety.”

We appreciate the reviewer’s careful observation regarding the conceptual distinction between climate change–related worry and general anxiety. In response, we have added a clarifying sentence in the Methods section, immediately following the description of the CCWS subscales, to specify that:

“It should be noted that the ‘anxiety’ subscale of the CCWS assesses worry specifically related to climate change, whereas the DASS-21 anxiety subscale measures broader symptoms of clinical anxiety.”

We believe this addition enhances conceptual clarity and addresses the potential confusion between the constructs.

Minor Language and Consistency Check Please perform a final proofread to ensure consistency in terminology (e.g., favor “climate change worry” over “climate anxiety” where the CCWS is referenced) and correct any minor grammatical errors.

We sincerely thank the reviewer for this helpful suggestion. In particular, we have reviewed the manuscript to ensure consistent use of terminology by referring to the CCWS as "climate change worry," and we have also attempted to correct minor grammatical errors throughout the text.

Reviewer 2 Comment

Response

First, a significant improvement in the approach to the basic statistical analyses is evident. In the revised version, the authors have abandoned the use of parametric tests and consistently opted for non-parametric procedures (Kruskal-Wallis H, Mann-Whitney U, and Spearman correlations), which is consistent with the lack of normality in the data. However, the analysis remains incomplete, as effect sizes associated with the comparisons are not reported, nor is it specified whether corrections for multiple comparisons were applied in the post-hoc analyses. These omissions limit the quantitative interpretation of the results.

We thank our esteemed referee for this important warning. In our study, post-hoc analyses were conducted to determine the source of the difference between groups, and the Dunn-Bonferroni test with Bonferroni correction was used to control for Type I error. Thanks to this correction, p-values were normalized according to the number of groups, providing a more conservative and reliable analysis. In addition, the effect size (epsilon-square) values added to our analyses revealed not only the statistical significance of the differences but also their quantitative practical importance (small, medium, large), thus eliminating the limitations of interpretation and increasing the power of the findings.

Epsilon-squared calculations have been performed and are reported in the tables.

Second, one of the main limitations noted in the original review persists: the absence of multivariate analyses. The study continues to rely exclusively on bivariate comparisons and simple correlations, without adjusting for potentially confounding variables such as educational level, socioeconomic status, trimester of pregnancy, number of previous pregnancies, or other psychological vulnerabilities. Although the discussion and limitations sections explicitly acknowledge the cross-sectional nature of the design and the impossibility of establishing causal relationships, this clarification does not compensate for the lack of a more appropriate analytical approach for a phenomenon of this complexity.

Third, the previously mentioned problem of conceptual circularity has not been resolved. Variables such as the level of concern about climate change or the expectation that the region will be affected by climate change continue to be used as explanatory factors for scores on the Climate Concern Scale (CCWS). Since these indicators assess constructs that are conceptually very similar—if not equivalent—to the construct measured by the scale itself, the risk of analytical circularity persists. At no point is this approach justified as a convergent validity assessment, nor are its methodological implications discussed, which limits the interpretive value of these analyses.

We thank the reviewer for this important observation regarding the potential conceptual overlap between CCWS scores and variables such as ‘level of concern about climate change’ or ‘expectation that the region will be affected by climate change.’ We agree that these variables are conceptually related to the main construct assessed by the CCWS, and that using them as explanatory factors could lead to analytical circularity.

To address this concern, we have revised the manuscript to clarify in the Discussion section that the inclusion of these variables was intended to provide preliminary evidence for the convergent validity of the CCWS in our sample, and that the observed associations should be interpreted carefully and are not indicative of causality.

We believe these revisions adequately address the reviewer’s concerns while maintaining the relevance of these exploratory analyses.

Fourth, although there is a slight improvement in the reporting of correlation coefficients, the manuscript continues to emphasize statistical significance without sufficient discussion of the magnitude of the effects or their clinical relevance. Mean scores for depression, anxiety, and stress are low, but the proportion of participants exceeding clinically relevant cut-off points is not reported, nor is the practical impact of the associations found contextualized. Without this information, it is difficult to assess the clinical and public health significance of the findings.

We thank the reviewer for this valuable and constructive comment. In response, we have expanded the Discussion section to place greater emphasis on the magnitude and potential clinical and public health relevance of the observed associations. Specifically, we now explicitly discuss that the correlations between climate change worry and symptoms of depression, anxiety, and stress were of moderate magnitude, which is consistent with effect sizes commonly reported in psychosocial research, and may still be clinically meaningful at the population level.

The Depression Anxiety Stress Scale-21 (DASS-21) does not have predetermined cut-off scores for its subscales. Each subscale can be scored from 0 to 21, with higher scores indicating greater severity of symptoms in the respective domain. In this study, the classification of depression, anxiety, and stress levels was based on the ratings obtained in a previous study conducted on pregnant women using another Turkish-adapted version of the DASS-21 (Aker & Yılmaz Sezer, 2025).

This information was added under the section introducing the DASS-21 and was used as a reference in evaluating the psychological status of the participants. Accordingly, the following paragraph was added to the introduction of the Discussion section: “To contextualize participants’ psychological status, mean scores on the DASS-21 subscales were evaluated. Based on these scores, depression and stress were generally within the normal range, whereas anxiety may be considered moderate. These findings may indicate that clinically significant depressive or stress symptoms were uncommon, while a moderate level of anxiety could be present in the sample.”

Reference: Aker, MN., & Yılmaz Sezer, N.(2025). Frequency of nomophobia in pregnant women and its relationship with depression, anxiety, and stress. J Acad Res Nurs. 11(1):41-6. DOI: 10.55646/jaren.2025.12754

Fifth, improvements in methodological transparency are evident. The authors have added information on the approximate number of women who declined to participate, clarified the handling of missing data, and specified the post-hoc tests used. However, recruitment is still based on convenience sampling, and a clear outline of the participant flow is not presented, which continues to limit the study's reproducibility.

We thank the Reviewer for this valuable comment. We would like to clarify that the recruitment procedure and data collection process were described in detail in the Data Collection section. Specifically, the setting in which participants were approached, the inclusion criteria, the voluntary nature of participation, and the approximate number of women who declined participation were explicitly reported.

As the study was based on convenience sampling in a clinical setting, a formal participant flow diagram was not generated. However, all relevant steps of participant recruitment and data collection were narratively explained to ensure transparency. We agree that the use of convenience sampling represents a limitation, and this has been explicitly acknowledged in the Limitations section of the manuscript.

Finally, the manuscript shows a marked improvement in writing, style, and formal coherence. The typographical and consistency issues previously noted have been largely corrected, and the text now presents an editorially acceptable level.

We sincerely thank the reviewer for the positive and encouraging feedback regarding the improvements in writing quality, style, and overall formal coherence of the manuscript. We appreciate the reviewer’s careful evaluation and are pleased that the revisions have resulted in a text that is now considered editorially acceptable.

Overall, while the revised version represents an improvement over the original manuscript, fundamental limitations related to the analytical design, the lack of multivariate analysis, and the conceptual problems associated with the circularity of some key variables have not been fully addressed. Therefore, these issues continue to hinder a sound and robust interpretation of the results.

Regarding the remaining concerns, we acknowledge the reviewer’s comments on the analytical design, the absence of multivariate analyses, and the conceptual circularity related to some variables. As noted in the revised manuscript, these issues have been explicitly addressed in the Limitations section, where we emphasize the cross-sectional nature of the study, the reliance on bivariate analyses, and the potential conceptual overlap among certain measures. We are grateful to the reviewer for highlighting these important points, which have helped us to clarify the boundaries and interpretive scope of our findings.