Psychometric Validation of the Mother–Infant Bonding Scale in Greek Mothers up to 1 Year Postpartum

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This research constitutes the first cultural adaptation and psychometric evaluation of the Mother-to-Infant Bonding Scale (MIBS), specifically adapted for Greek mothers within the initial year following childbirth. The introduction is clear and gives a clear picture of topic`s importance and contemporary practice of mother-infant bonding evaluation. Material and methods are adequately presented in order to make possible to reproduce the study. Results are clearly presented with sufficient data. Discussion is concise and well-written. Limitations of the study are clearly stated. Conclusions are clear and well extrapolated from the results of the study.

Author Response

Reviewer #1

This research constitutes the first cultural adaptation and psychometric evaluation of the Mother-to-Infant Bonding Scale (MIBS), specifically adapted for Greek mothers within the initial year following childbirth. The introduction is clear and gives a clear picture of the topic`s importance and contemporary practice of mother-infant bonding evaluation. Material and methods are adequately presented in order to make possible to reproduce the study. Results are clearly presented with sufficient data. Discussion is concise and well-written. Limitations of the study are clearly stated. Conclusions are clear and well extrapolated from the results of the study.

Authors’ response:  We would like to thank the reviewer for the positive appraisal of our study and for taking the time to review our manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this paper. Overall, I found the manuscript clearly written. The study benefits from a large sample size and provides detailed information on the analytic approach. However, I noted several areas that could be strengthened, which I outline below:

* Introduction  

• There are spots where in-text citation use seems unnatural, such as on LL 70-71, "Research indicates (Trombetta et al., 2021; Le Bas et al., 2022) that..."
• While the paper notes that the MIBQ was developed using a sample of mothers described as having maternal bonding disorders, I wonder whether this label of “maternal bonding disorder” is accurate and appropriate in this context. 
• Line 136, the acronym MPAS should be spelled out. Moreover, in this section where the psychometric properties of the MIBS are discussed, I wonder if the information can be organized differently (e.g., grouped by validity type or construct names) rather than listing the measure names that are correlated with it. 
• From Line 139, where the dimensionality of the MIBS is discussed, could the underlying factor structures be elaborated further? Specifically, what factors were identified in these previous studies?
• From the Introduction, the role and meaning of the time and “postpartum period” is somewhat unclear. For example, Line 102 mentions that the PBQ is administered in the postpartum period, but how is this period defined? Another example is Line 159 and Line 160 mentioning "younger infants" and "later stages." Perhaps the meaning of these periods could be clarified. In addition, how long is the MIBS typically used, and what is the rationale for the authors’ decision to focus on the first year?
• The main goal of the study is the cultural adaptation of the measure (based on the Abstract). Could the authors expand on the cultural relevance from this sense? Beyond the linguistic translation, what specific cultural aspects were considered in the adaptation process? 
• I think the Present Study section would benefit from further elaboration. Although the factor structure was explored, were any hypotheses specified in advance? For example, what kinds of validity evidence the authors anticipated and how these were to be evaluated?

Method

• I wonder if the full demographic information could be included (e.g., age range, proportion of each breakdown for education level, marital status, etc.). Additionally, procedures can include information of how surveys were administered (e.g., online survey, compensation, other measures included in the questionnaires). 
• Detailed information on how the random split of the sample was conducted would be helpful.
• The inclusion of measurement invariance testing was somewhat unexpected. Could this be foreshadowed in the Introduction, such as why testing invariance is relevant to the study’s aims?
• I apologize if I overlooked this information, but would it be possible to include the prompt of the scale?

Results 

• I don't think the acronym EGA was defined earlier.

• I apologize if I misunderstood anything, but the kurtosis value for Item 3 appears to be quite extreme. Could the authors comment on whether this distributional issue might affect the validity of the results?

• I initially understood the EPDS to be included as a validity measure (to check its correlations with MIBS), but I noticed that it was incorporated into the CFA model. Could this analytic choice be justified earlier in the paper?

 

Author Response

Reviewer #2

* Introduction 

• There are spots where in-text citation use seems unnatural, such as on LL 70-71, "Research indicates (Trombetta et al., 2021; Le Bas et al., 2022) that..."

Authors’ response: Thank you for pointing this out. We have corrected the in-text citations to improve flow. 

• While the paper notes that the MIBQ was developed using a sample of mothers described as having maternal bonding disorders, I wonder whether this label of “maternal bonding disorder” is accurate and appropriate in this context.

Authors’ response:  We thank the reviewer for raising this important conceptual point. We agree that the term “maternal bonding disorder” may be perceived as potentially ambiguous or overly pathologizing when used without clarification. In the original work by Kumar (1997), the term was employed descriptively to refer to clinically significant disturbances in mothers’ emotional responses toward their infants, rather than as a formally recognized diagnostic category in contemporary classification systems (e.g., DSM or ICD). To address this concern, we have revised the manuscript. We now refer more precisely to “clinically significant bonding disturbances” when describing the developmental basis of the MIBQ (line 251).

• Line 136, the acronym MPAS should be spelled out. Moreover, in this section where the psychometric properties of the MIBS are discussed, I wonder if the information can be organized differently (e.g., grouped by validity type or construct names) rather than listing the measure names that are correlated with it.

Authors’ response: We thank the reviewer for this helpful suggestion. In response, we have revised the relevant paragraph to improve conceptual clarity and psychometric coherence (lines 278 and below). Specifically, we reorganized the description of the MIBS by grouping evidence according to reliability and validity domains, rather than listing correlated instruments sequentially. The revised text now first presents evidence for reliability, followed by convergent validity with measures of maternal psychological distress (e.g., EPDS, Blues scale, HADS), and then concurrent validity with established measures of the mother–infant relationship (e.g., PBQ, MPAS). We believe this reorganization enhances the theoretical and methodological clarity of the section while retaining explicit references to the relevant scales, as suggested. We have also spelled out MPAS (line 413).

• From Line 139, where the dimensionality of the MIBS is discussed, could the underlying factor structures be elaborated further? Specifically, what factors were identified in these previous studies?

Authors’ response:  We thank the reviewer for this suggestion. The revised text now briefly summarizes the factor structures identified across studies, explicitly naming the factors reported (e.g., lack of affection, anger/rejection, positive vs. negative bonding) and clarifying how differences in item composition and cultural context may account for variability in findings (425-442).

• From the Introduction, the role and meaning of the time and “postpartum period” is somewhat unclear. For example, Line 102 mentions that the PBQ is administered in the postpartum period, but how is this period defined? Another example is Line 159 and Line 160 mentioning "younger infants" and "later stages." Perhaps the meaning of these periods could be clarified. In addition, how long is the MIBS typically used, and what is the rationale for the authors’ decision to focus on the first year?

Authors’ response:  Thank you for raising this issue. We have defined the postpartum period and we have corrected the text to improve precision regarding timing. We have also added a rationale for our decision to focus on the first year in section 1.2. “Present Study.”

 

• The main goal of the study is the cultural adaptation of the measure (based on the Abstract). Could the authors expand on the cultural relevance from this sense? Beyond the linguistic translation, what specific cultural aspects were considered in the adaptation process?

Authors’ response: We thank the reviewer for this valuable comment. In response, we have expanded Section 2.3 to clarify how cultural relevance was addressed beyond linguistic equivalence. Specifically, we now describe the role of the expert committee in evaluating the comprehensibility, acceptability, and conceptual equivalence of items within the Greek cultural context. We also explicitly discuss culturally salient factors, including strong familistic values and stigma related to perinatal psychological distress, and how these considerations informed item review and finalization.

• I think the Present Study section would benefit from further elaboration. Although the factor structure was explored, were any hypotheses specified in advance? For example, what kinds of validity evidence the authors anticipated and how these were to be evaluated?

Authors’ response: We thank the reviewer for this helpful suggestion. Although no specific factor structure was hypothesized a priori due to heterogeneity in previous findings, the study was guided by explicit expectations regarding the reliability and validity of the Greek MIBS. In particular, we anticipated acceptable internal consistency and evidence of convergent and concurrent validity through associations with maternal psychological distress and related bonding measures, while also examining discriminant validity to ensure that the construct was not redundant with depressive symptomatology. We have revised the Present Study section to clarify these a priori expectations and the corresponding analytic strategies.

 

Method

• I wonder if the full demographic information could be included (e.g., age range, proportion of each breakdown for education level, marital status, etc.). Additionally, procedures can include information of how surveys were administered (e.g., online survey, compensation, other measures included in the questionnaires).

Authors’ response: We thank the reviewer for this suggestion. We have revised the Procedure section to include all the requested information regarding survey administration, including the mode of administration, participation details, and the additional measures included in the study (lines 578-588). We have also expanded on our demographic information in the Participants section (lines 565-576).

• Detailed information on how the random split of the sample was conducted would be helpful.

Authors’ response: We thank the reviewer for this observation. The random split was created using a partial Fisher–Yates shuffle, creating a uniformly random subset of row indices of size ~50% of the sample. For clarity, we added this information to the paper.

 

• The inclusion of measurement invariance testing was somewhat unexpected. Could this be foreshadowed in the Introduction, such as why testing invariance is relevant to the study’s aims?

Authors’ response: We thank the reviewer for this helpful comment. We agree that the rationale for measurement invariance testing was not clearly introduced. Because we conducted mean comparisons across sociodemographic groups, it was necessary to first establish that the MIBS functions equivalently across these groups; otherwise, observed differences could reflect measurement bias rather than true differences in bonding. We have now clarified this rationale in the Aims and Data analysis sections.

 

• I apologize if I overlooked this information, but would it be possible to include the prompt of the scale?

 

Authors’ response:  We added the prompt in the description of the measure in the Methods section (line 631-636).

 

Results

• I don't think the acronym EGA was defined earlier.

            Authors’ response:  EGA is now spelled out (line 701).

• I apologize if I misunderstood anything, but the kurtosis value for Item 3 appears to be quite extreme. Could the authors comment on whether this distributional issue might affect the validity of the results?

Authors’ response:  We thank the reviewer for this careful observation. We agree that Item 3 exhibited elevated kurtosis and that multivariate normality was not supported (as indicated by Mardia’s test). However, this does not compromise the validity of our findings, as we did not rely on normal-theory maximum likelihood (ML) estimation. Instead, items were treated as ordered categorical variables and analyzed using estimators robust to non-normality, including weighted least squares with means and variances adjusted (WLSMV) for CFA, weighted least squares (WLS) for EFA, and Maximum Likelihood with Robust Standard Errors (MLR) for measurement invariance testing. These estimators apply corrections for non-normal distributions and non-continuous indicators, thereby minimizing bias in parameter estimates, standard errors, and fit indices. Accordingly, the distributional characteristics of Item 3 were appropriately accounted for in the analytic approach.

• I initially understood the EPDS to be included as a validity measure (to check its correlations with MIBS), but I noticed that it was incorporated into the CFA model. Could this analytic choice be justified earlier in the paper?

Authors’ response: We thank the reviewer for this important observation. The EPDS was included in the factor analytic models not to re-evaluate its construct validity, but to estimate associations with the MIBS at the latent level. Because correlations were derived from the factor models rather than from unit-weighted total scores, incorporating the EPDS into the CFA was necessary to obtain unbiased latent correlations that account for measurement error. Similarly, including the EPDS in the exploratory factor analysis allowed us to examine potential cross-loadings between depressive symptoms and bonding items. Modeling both constructs simultaneously reduces the risk that unmodeled cross-loadings inflate or distort the estimated association, thereby providing a more rigorous test of convergent and discriminant validity.

Reviewer 3 Report

Comments and Suggestions for Authors

This paper addresses the important problem of mother-infant bonding, with the purpose of designing a scale to identify Greek mothers who are at risk.  It carefully assesses and translates a scale that has been used in several other languages and establishes its reliability and validity in a Greek version.

The paper is very well written, with an excellent overview of maternal bonding and carefully distinguishing from attachment.  The results of this work will be useful in clinical settings.

Comments for author File: Comments.pdf

Author Response

Reviewer # 3

This paper addresses the important problem of mother-infant bonding, with the purpose of designing a scale to identify Greek mothers who are at risk.  It carefully assesses and translates a scale that has been used in several other languages and establishes its reliability and validity in a Greek version.

The paper is very well written, with an excellent overview of maternal bonding and carefully distinguishing from attachment.  The results of this work will be useful in clinical settings.

 

Authors’ response:  We would like to thank the reviewer for the positive appraisal of our study and for taking the time to review our manuscript.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

I congratulate the authors for their efforts in the study titled "Psychometric validation of the Mother-Infant Bonding Scale in Greek mothers up to 1 year postpartum.". Overall, the rationale for developing and validating the scale is clearly articulated, and the identified gap in the literature is well justified. The analytical strategy employed for the main scale appears appropriate and methodologically sound. Nevertheless, I identified several issues that warrant attention.

First, the ethical approval date raises concerns. The study relies on an ethics committee approval obtained in 2017, which is considerably earlier than the apparent time frame of publication. While this may be justified if data collection began at that time, the authors should explicitly clarify when the data were collected and whether the original ethical approval adequately covered all stages of the study. Without such clarification, the ethical validity of the study may be questioned.

Second, the rationale for conducting both exploratory and confirmatory factor analyses on the Edinburgh Postnatal Depression Scale (EPDS) is unclear. The EPDS is a well-established instrument with extensively validated factor structures, including in the Greek population. Given that the EPDS is used primarily to examine convergent validity, re-examining its factorial structure appears methodologically unnecessary and may be considered an overextension of the analyses.

Finally, I found the discussion section to be relatively limited. While the authors adequately summarize their findings, the discussion would benefit from a deeper theoretical interpretation and a more critical integration of the results within the broader literature. In particular, the implications of item removal and cross-cultural variability in factor structure deserve more thorough consideration. In addition, psychometric (statistical) references should be increased.

In sum, although the study is based on a solid methodological foundation, addressing these points would substantially strengthen the manuscript.

Best wishes.

Author Response

Reviewer #4

I congratulate the authors for their efforts in the study titled "Psychometric validation of the Mother-Infant Bonding Scale in Greek mothers up to 1 year postpartum.". Overall, the rationale for developing and validating the scale is clearly articulated, and the identified gap in the literature is well justified. The analytical strategy employed for the main scale appears appropriate and methodologically sound. Nevertheless, I identified several issues that warrant attention.

Authors’ response:  We would like to thank the reviewer for the positive appraisal of our study and for taking the time to review our manuscript.

First, the ethical approval date raises concerns. The study relies on an ethics committee approval obtained in 2017, which is considerably earlier than the apparent time frame of publication. While this may be justified if data collection began at that time, the authors should explicitly clarify when the data were collected and whether the original ethical approval adequately covered all stages of the study. Without such clarification, the ethical validity of the study may be questioned.

Authors’ response: We thank the reviewer for raising this important point. We have revised the manuscript to explicitly clarify the data collection period (11 August 2018 to 21 December 2019); see lines 541-542. 

Second, the rationale for conducting both exploratory and confirmatory factor analyses on the Edinburgh Postnatal Depression Scale (EPDS) is unclear. The EPDS is a well-established instrument with extensively validated factor structures, including in the Greek population. Given that the EPDS is used primarily to examine convergent validity, re-examining its factorial structure appears methodologically unnecessary and may be considered an overextension of the analyses.

Authors’ response: We thank the reviewer for this important observation. The EPDS was included in the factor analytic models not to re-evaluate its construct validity, but to estimate associations with the MIBS at the latent level. Because correlations were derived from the factor models rather than from unit-weighted total scores, incorporating the EPDS into the CFA was necessary to obtain unbiased latent correlations that account for measurement error. Similarly, including the EPDS in the exploratory factor analysis allowed us to examine potential cross-loadings between depressive symptoms and bonding items. Modeling both constructs simultaneously reduces the risk that unmodeled cross-loadings inflate or distort the estimated association, thereby providing a more rigorous test of convergent and discriminant validity.

Finally, I found the discussion section to be relatively limited. While the authors adequately summarize their findings, the discussion would benefit from a deeper theoretical interpretation and a more critical integration of the results within the broader literature. In particular, the implications of item removal and cross-cultural variability in factor structure deserve more thorough consideration. In addition, psychometric (statistical) references should be increased.

Authors’ response: We thank the reviewer for this comment. In response, we have substantially expanded the Discussion section. Specifically, we elaborated on the conceptual distinction between maternal depression and mother–infant bonding, framing them as related yet distinct constructs and supporting our findings within longitudinal and meta-analytic evidence (lines 837-876). We also expanded the discussion regarding the removal of the “protective” item, addressing potential cultural influences and discussing psychometric/methodological considerations (lines 874-889).

In sum, although the study is based on a solid methodological foundation, addressing these points would substantially strengthen the manuscript.

Best wishes.

 

 

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for your thorough and careful revision. I have only a couple remaining comments.

• Regarding the cultural adaptation described in Section 2.3, I wonder if the authors could provide some concrete examples to help readers better understand how the measures were adapted across cultural contexts.
• I also continue to have some concern about Item 3, as it shows extremely low variability (M = 0.14, SD = 0.46), indicating a substantial floor effect with extreme kurtosis. I wonder if this may signal an issue beyond general non-normality and suggests limited informational value of the item. While I appreciate and respect the authors' response that robust estimators (e.g., WLSMV) mitigate bias due to non-normality, I am not sure they can resolve issues arising from severely restricted variance. Just a suggestion, but perhaps the authors could consider justifying the retention of this item or providing sensitivity analyses excluding it? 

Author Response

Reviewer comment #1:

Regarding the cultural adaptation described in Section 2.3, I wonder if the authors could provide some concrete examples to help readers better understand how the measures were adapted across cultural contexts.

Authors’ response: We have revised the manuscript to clarify the cultural adaptation process in greater detail. Examples of item adaptations (e.g., “neutral or felt nothing” and “aggressive”) have been added to illustrate how potential ambiguity and stigma were addressed.

Reviewer comment: “I also continue to have some concern about Item 3, as it shows extremely low variability … suggesting limited informational value… perhaps justify retention or provide sensitivity analyses excluding it?”

Authors’ response: Thank you for this thoughtful observation. We agree that severely restricted variance and pronounced floor effects can reduce an item’s informational value, which is conceptually distinct from general non-normality. While robust ordinal estimators (e.g., WLSMV/DWLS) mitigate bias in standard errors and test statistics under non-normal categorical indicators, they do not “recover” information when an item exhibits extreme range restriction. To directly address this concern, we conducted a sensitivity analysis excluding Item 3, in addition to the previously excluded Item 6 (i.e., retaining Items 1, 2, 4, 5, 7, and 8).

Across the two random splits, the results supported the same conclusions regarding unidimensionality and reliability. In the first subset, exploratory graph analysis indicated a unidimensional structure, and factorability was supported (KMO = .82; Bartlett’s test of sphericity, , p < .001). A one-factor EFA (WLS using polychoric correlations) yielded strong loadings (.71–.84) and explained 57% of the variance; a two-factor solution was inadmissible (ultra-Heywood case for Item 1), indicating overfactoring rather than meaningful multidimensionality. Consistent with these findings, van Bork et al.’s bootstrap test of the one-factor condition provided no clear evidence against a unidimensional solution (no item-pair confidence interval was entirely above 1.00). In the second subset, a unidimensional ordinal CFA demonstrated good fit (, p = .002; CFI = .984; TLI = .974; RMSEA = .070 [90% CI .039, .103]; SRMR = .052) with substantial standardized loadings (.70–.84) and only small residual correlations (largest rresid = .12). Reliability remained high in this sensitivity model (Cronbach’s α = .79; composite reliability ≈ .89; H = .90), indicating strong construct replicability.

Taken together, these sensitivity analyses show that our conclusions about the measure’s unidimensional structure and reliability are robust to the exclusion of Item 3. We have added a brief sensitivity analysis section to the manuscript reporting these results and have reported the analysis results as supplementary material.

Author Response File: Author Response.docx