Round 1

Reviewer 1 Report

I had the pleasure to read and review the manuscript “Pregnancy- and birth-related experiences among postpartum women during the third wave of the COVID-19 pandemic – a multinational European study”. I would like to congratulate the authors for their meticulous work and rigorous description of the methods. The study describes how maternity care services were adapted during an advanced stage of the COVID-19 pandemic, using data reported by postpartum women in an online survey in 5 European countries. The study also looks at associations between the experienced changes in care provision and the mental health status of these women. It fills an important gap in the literature by presenting women’s voices. The findings show that maternity care services continued to be adapted in an advanced stage of the pandemic, highlighting the necessity of prioritizing pregnant and postpartum women’s needs in a pandemic setting, and beyond.

I consider this manuscript to be fit for publication after minor adjustments described below.

 

General concept comments

 

The authors write in-text that supplementary material 3 describes country-level restriction measures during the study period although no such information is presented in the table (only infection and vaccination rates). I recommend adding a summary of how restriction measures were implemented per country, as this information could be useful for the interpretation of the findings. These restriction measures could also explain some of the between-country differences documented in the results (see last comment on “comparison”).  For this purpose, it is also important to take into consideration the “time” component, as women answering the survey “postpartum” in June-August 2021 means that they were pregnant since as early as October/Nov 2020 (particularly important when looking at reported changes during the prenatal period – schedule of prenatal visits).

 

Specific comments

Throughout the manuscript, recommend to consistently use the term COVID-19 (sometimes the word COVID is used instead).

 

Abstract:

Suggestion to specify the five countries in the abstract, and to provide some detail on the sampling and recruitment strategy.

 

Introduction:

Line 59-60: “Other studies describe limited birthing options (1), uncertainty or restrictions regarding partners’ visiting policies (1-3) and changes in birth locations due to the pandemic (7).” Authors mention “other studies” report “changes in birth location” and cite same study (#7) as the previous sentence. Please revise and align.

 

Methods:

Line 95: parenthesis missing before “questions 8 and 9).

 

Results and discussion:

Suggest editing terminology used for interpreting the results - replace terms such as “risk” and “likelihood” with the term “odds” which is more convenient for the study design.

 

Discussion: Line 311: authors mention that a strength of the study is the multi-country analysis “enabling comparisons across countries”. Yet the discussion of these comparisons is not addressed in detail in the text, only in the supplementary material, and not in regards to the findings. It is important to consider difference between countries when possible, taking into consideration existing guidelines before the pandemic (different models of care provision) and their adaptations during the pandemic, and the variance in restriction measures, which could explain some of the between-country differences (e.g. virtual ANC in the UK during COVID-19).

Author Response

Response to Reviewers

We thank the Editors and the Reviewers for the opportunity to revise our manuscript further, and we appreciate the positive and valuable feedback received. We provide our replies to the comments below, point-by-point. All changes to the manuscript were done using the track change mode, according to the MDPI guidelines. In the following, all line numbers refer to the revised version using track changes, and the changes in the manuscript are highlighted in red.

 

We also would like to make the Editors and Reviewers aware that we have re-run our adjusted logistic regression models in the revised version of the manuscript. Thanks to the reviewer comment, we noticed and corrected the missing data category for the two baseline chronic disease variables (see explanation below). When doing so, the variable “chronic somatic illness” was not associated with the number of birth-related changes. The adjusted logistic regression model did consequently not include “chronic somatic illness” as a confounder. We have updated Table 4 and re-written the following parts affected by this (updates in red):

 

Abstract: Lines 36-38 “The number of pregnancy- and birth-related changes were associated with the woman’s mental health status, as well as the type of change.”

Results: Lines 252-257: “Postpartum women who experienced >4 pregnancy and birth-related changes were more likely to experience major depressive symptoms (EDS≥13) than women with 0-4 changes in crude analyses. This association remained significant after adjusting for confounding factors (see Table 4, aOR 1.75; CI 1.2 – 2.55).

Likewise, postpartum women with more than four pregnancy and birth-related changes were more likely to experience anxiety symptoms (EDS-3A≥5) compared to women with 0-4 changes. Effect estimates remained significantly elevated after adjustment for unbalanced background characteristics, i.e., country, COVID-19 status, and breastfeeding experience before the pandemic (aOR 1.55; CI 1.13 – 2.12).”

Discussion: Lines 323-327: “Importantly, our logistic regression models showed a statistically significant association between women having experienced >4 changes and symptoms of major depression (aOR 1.75; CI 1.20 – 2.55) and anxiety (aOR 1.55; CI 1.13 – 2.12). This clearly demonstrates the impact of the continuous substantial COVID-19 regulations on postpartum women’s mental health during the third wave of the pandemic.”

Lines 356-359: “The results presented in Table 5 suggest that, in addition to the number of changes, the type of change might be even more relevant for maternal mental health postpartum. Previous findings show similar results, where unexpected changes (e.g., emergency C-sections and change from home to hospital birth) correlated with poor birth experiences and depressive symptoms (1-2).”

Conclusion: Lines 421-422: “We found that having more than four pregnancy- and birth-related changes increased the likelihood of experiencing major depressive and anxiety symptoms postpartum”.

 

All authors have read and approved the revised manuscript.

Sincerely, Robin Aasegg Araya

on behalf of the authors

Reviewer 1

I had the pleasure to read and review the manuscript “Pregnancy- and birth-related experiences among postpartum women during the third wave of the COVID-19 pandemic – a multinational European study”. I would like to congratulate the authors for their meticulous work and rigorous description of the methods. The study describes how maternity care services were adapted during an advanced stage of the COVID-19 pandemic, using data reported by postpartum women in an online survey in 5 European countries. The study also looks at associations between the experienced changes in care provision and the mental health status of these women. It fills an important gap in the literature by presenting women’s voices. The findings show that maternity care services continued to be adapted in an advanced stage of the pandemic, highlighting the necessity of prioritizing pregnant and postpartum women’s needs in a pandemic setting, and beyond.

I consider this manuscript to be fit for publication after minor adjustments described below.

 

Comment 1: The authors write in-text that supplementary material 3 describes country-level restriction measures during the study period although no such information is presented in the table (only infection and vaccination rates). I recommend adding a summary of how restriction measures were implemented per country, as this information could be useful for the interpretation of the findings. These restriction measures could also explain some of the between-country differences documented in the results (see last comment on “comparison”).  For this purpose, it is also important to take into consideration the “time” component, as women answering the survey “postpartum” in June-August 2021 means that they were pregnant since as early as October/Nov 2020 (particularly important when looking at reported changes during the prenatal period – schedule of prenatal visits)

 

Response 1: Thank you for your encouraging comments and for raising the point about country specific restriction measures. Our initial reference to Supplementary Material 3 should have stated that it describes infection rates and vaccination status during the study period in the various countries. We have now corrected this. The main text now reads “Infection status and vaccination status in the respective countries for the period of interest are summarised in Supplementary Material 3”.

We purposely avoided showing COVID19 restriction measures in the various countries because these have been changing repeatedly, also within countries and would probably be incorrect. We hope the Reviewer understands our rationale.

 

Minor

Comment 2: Throughout the manuscript, recommend to consistently use the term COVID-19 (sometimes the word COVID is used instead).

Response 2: Thank you, we have revised this accordingly.

 

Comment 3: Abstract: Suggestion to specify the five countries in the abstract, and to provide some detail on the sampling and recruitment strategy.

Response 3: As suggested, we have specified the countries in the Abstract as follows:

“An online questionnaire was distributed in five European countries (Belgium, Netherlands, Norway, Switzerland, UK) between June and August 2021”

We have also added the following text: «Participants were recruited though social media platforms including pregnancy- and motherhood-related websites, pregnancy fora and apps.»

 

Comment 4: Introduction: Line 59-60: “Other studies describe limited birthing options (1), uncertainty or restrictions regarding partners’ visiting policies (1-3) and changes in birth locations due to the pandemic (7).” Authors mention “other studies” report “changes in birth location” and cite same study (#7) as the previous sentence. Please revise and align.

Response 4: Thank you for this notification. We have now corrected this by omitted the last part of the sentence “and changes in birth locations due to the pandemic (7)”.

 

Comment 5: Methods: Line 95: parenthesis missing before “questions 8 and 9).

Response 5: This typo has now been corrected.

 

Comment 6: Results and discussion: Suggest editing terminology used for interpreting the results - replace terms such as “risk” and “likelihood” with the term “odds” which is more convenient for the study design.

Response 6: We have replaced the term “risk” with “odds” as suggested. We believe that “likelihood” remains an appropriate term, and have kept it the few places it occurs in the manuscript. To meet further the Reviewer’s concern, we have rephrased the word “impact” with “association” or “relationship”, to avoid use of causal terminology.

 

Comment 7: Discussion: Line 311: authors mention that a strength of the study is the multi-country analysis “enabling comparisons across countries”. Yet the discussion of these comparisons is not addressed in detail in the text, only in the supplementary material, and not in regards to the findings. It is important to consider difference between countries when possible, taking into consideration existing guidelines before the pandemic (different models of care provision) and their adaptations during the pandemic, and the variance in restriction measures, which could explain some of the between-country differences (e.g. virtual ANC in the UK during COVID-19).

Response 7: We thank the Reviewer for this important comment. We have now removed the sentence “enabling comparisons across countries”. As correctly pointed out, we did not discuss country-specific findings in details on the main text. The reasoning for this is that in some countries the sample size was small, which challenges country-specific result interpretation. We have already addressed this as limitation, which reads “Study participation was also not evenly distributed among participating countries, with an overrepresentation of postpartum women from Norway.»

 

Reviewer 2

Thank you for the opportunity to review this interesting manuscript on pregnancy- and birth-related experiences among postpartum women from five EU countries during the third wave of the COVID-19 pandemic. The study included 1730 postpartum women and found that experiencing changes related to delivery or prenatal examination was associated with much higher depressive and anxiety symptoms postpartum. The study is well conducted, well written and the focus is important and needs attention.

 

Comment 1: However, I have some major concerns about the design and methods used in this study. First, as this is a cross-sectional study, we have no way of knowing if the women had depressive symptoms prior to experiencing these changes. Second, measures of maternal depression and changes in pregnancy and birth experiences were self-reported and not confirmed by a professional. This could introduce the problem of common method variance and risk of dependent measurement error. Common method variance reflects the variance that is attributed to the measurement method of the variables of interest (in this case self-report) rather than the constructs of interest. Method variance can either inflate or deflate observed relationships between constructs and lead to both Type I and Type II errors. Factors that have been reported to produce dependent measurement error are negative affectivity and social desirability. In the current situation, it could be that mothers who are likely to over-report (exaggerate) depression are also likely to over-report having had changes to pregnancy or birth plans (negative affectivity). Similarly, mothers who are likely to under-report depression are also likely to under-report changes (social desirability). In other words, the mechanism is not through the changes in pregnancy and birth experiences itself but through the reporting being from the same source using the same method. This needs to be discussed.

Response 1: Thank you for raising this important concern. We agree with the Reviewer that by design, the study lacks information about the temporal component. We fully acknowledged this limitation in the original version of the manuscript. We do also agree with the Reviewer that bias attributable to common method variance cannot be ruled out. At the same time, if this latter bias was present, we would have obtained homogenously similar results for the associations between “number” and “type” of birth-related changes with maternal mental health outcomes. Women over-reporting (exaggerating) depressive symptoms would be more likely to over-report having had multiple changes related to the pregnancy and/or birth plans. We do not observe this association between number of changes and outcomes after adjusting for confounding. We do only observe association in the adjusted models between specific types of birth-related changes. The specificity of the observed associations, along with the objectivity of the birth-related changes women were asked to report about (e.g., change in schedule for C-section or induction of labour), point against a meaningful risk of bias attributable to common method variance in this study.

To accommodate the Reviewer’s concern, we have added the following limitation:

“Because we measured mental health outcomes and birth-related changes in the same questionnaire, we cannot exclude the risk of common method bias. However, we consider this risk to be minimal as women were asked to report objective changes that occurred during prenatal follow-up and in relation to childbirth. Our distinct results for number versus type of birth-related changes corroborate the above assumption; if common method bias was substantial, we would have obtained similar associations across analyses”.

 

Comment 2:

Minor:

Besides this, I have a few minor issues that needs to be addressed.

 

• Line 82: maternal characteristics associated with changes? I think I know what you mean, but the sentence does not make much sense.

Response 2: We have rephrased this part of the Aims, which now reads «…and what factors characterised women having experienced these changes”. We hope the wording is clearer.

 

 

Comment 3: Could you elaborate why you have chosen to reduce changes in pregnancy and birth related experiences to “number of changes”? I guess that the experiences are qualitatively different and therefore reducing them to a number making them of equal importance could cause misclassification. At least this should be discussed. 

Response 3: Thank you for this important comment. We fully agree with the Reviewer, and that is why we explored pregnancy and birth-related changes in terms of number and type. We have addressed this comment in three sections. In the Methods, section 2.2.1, we have provided the rationale of looking at number of changes, which reads: “We examined number of changes in addition to type, as we assumed that multiple changes in pregnancy and birth-related experience may be of greater burden to the woman mental health.” In the Discussion, lines 338-339, we have added the following sentence: “Further, this classification may be too simplistic as it does not account for the different importance or seriousness of the individual changes”. In the Limitation section, lines 396-398, we have added: “Lastly, the pregnancy and birth-related experiences measured in the study are not equally important to women, and this may have affected our associations between postpartum mental health outcomes and number of pregnancy and birth-related changes.”

 

Comment 4: Please explain why numbers in Table 1 do not add up to 100% (e.g., Chronic mental illness and chronic somatic illness).

Response 4: Thank you for addressing this. We have checked, and the numbers should add up to 100% when including the missing figures as described in the footnote of the table. We have however corrected the variables “chronic mental illness” and “chronic somatic illness” as the online questionnaire was designed so that if a woman did not tick anything, it meant she did not have the illness, thus there were no missing values by default. Question 12 in the Appendix for the questionnaire demonstrates this by stating: “Please indicate for which chronic illnesses you have used a medication in the past 3 months? (i.e., chronic illnesses are conditions that already existed before your pregnancy/postpartum) (multiple answers possible)?».

We erroneously presented lack of presence of chronic illness as missing. When updating this, we also realized we needed to re-run analyses. See the explanation to the editor and reviewers above.

Reviewer 2 Report

Thank you for the opportunity to review this interesting manuscript on pregnancy- and birth-related experiences among postpartum women from five EU countries during the third wave of the COVID-19 pandemic. The study included 1730 postpartum women and found that experiencing changes related to delivery or prenatal examination was associated with much higher depressive and anxiety symptoms postpartum. The study is well conducted, well written and the focus is important and needs attention.

 

However, I have some major concerns about the design and methods used in this study. First, as this is a cross-sectional study, we have no way of knowing if the women had depressive symptoms prior to experiencing these changes. Second, measures of maternal depression and changes in pregnancy and birth experiences were self-reported and not confirmed by a professional. This could introduce the problem of common method variance and risk of dependent measurement error. Common method variance reflects the variance that is attributed to the measurement method of the variables of interest (in this case self-report) rather than the constructs of interest. Method variance can either inflate or deflate observed relationships between constructs and lead to both Type I and Type II errors. Factors that have been reported to produce dependent measurement error are negative affectivity and social desirability. In the current situation, it could be that mothers who are likely to over-report (exaggerate) depression are also likely to over-report having had changes to pregnancy or birth plans (negative affectivity). Similarly, mothers who are likely to under-report depression are also likely to under-report changes (social desirability). In other words, the mechanism is not through the changes in pregnancy and birth experiences itself but through the reporting being from the same source using the same method. This needs to be discussed.

 

Besides this, I have a few minor issues that needs to be addressed.

 

Line 82: maternal characteristics associated with changes? I think I know what you mean, but the sentence does not make much sense.

 

Could you elaborate why you have chosen to reduce changes in pregnancy and birth related experiences to “number of changes”? I guess that the experiences are qualitatively different and therefore reducing them to a number making them of equal importance could cause misclassification. At least this should be discussed. 

 

Please explain why numbers in Table 1 do not add up to 100% (e.g., Chronic mental illness and chronic somatic illness).

Author Response

Response to Reviewers

We thank the Editors and the Reviewers for the opportunity to revise our manuscript further, and we appreciate the positive and valuable feedback received. We provide our replies to the comments below, point-by-point. All changes to the manuscript were done using the track change mode, according to the MDPI guidelines. In the following, all line numbers refer to the revised version using track changes, and the changes in the manuscript are highlighted in red.

 

We also would like to make the Editors and Reviewers aware that we have re-run our adjusted logistic regression models in the revised version of the manuscript. Thanks to the reviewer comment, we noticed and corrected the missing data category for the two baseline chronic disease variables (see explanation below). When doing so, the variable “chronic somatic illness” was not associated with the number of birth-related changes. The adjusted logistic regression model did consequently not include “chronic somatic illness” as a confounder. We have updated Table 4 and re-written the following parts affected by this (updates in red):

 

Abstract: Lines 36-38 “The number of pregnancy- and birth-related changes were associated with the woman’s mental health status, as well as the type of change.”

Results: Lines 252-257: “Postpartum women who experienced >4 pregnancy and birth-related changes were more likely to experience major depressive symptoms (EDS≥13) than women with 0-4 changes in crude analyses. This association remained significant after adjusting for confounding factors (see Table 4, aOR 1.75; CI 1.2 – 2.55).

Likewise, postpartum women with more than four pregnancy and birth-related changes were more likely to experience anxiety symptoms (EDS-3A≥5) compared to women with 0-4 changes. Effect estimates remained significantly elevated after adjustment for unbalanced background characteristics, i.e., country, COVID-19 status, and breastfeeding experience before the pandemic (aOR 1.55; CI 1.13 – 2.12).”

Discussion: Lines 323-327: “Importantly, our logistic regression models showed a statistically significant association between women having experienced >4 changes and symptoms of major depression (aOR 1.75; CI 1.20 – 2.55) and anxiety (aOR 1.55; CI 1.13 – 2.12). This clearly demonstrates the impact of the continuous substantial COVID-19 regulations on postpartum women’s mental health during the third wave of the pandemic.”

Lines 356-359: “The results presented in Table 5 suggest that, in addition to the number of changes, the type of change might be even more relevant for maternal mental health postpartum. Previous findings show similar results, where unexpected changes (e.g., emergency C-sections and change from home to hospital birth) correlated with poor birth experiences and depressive symptoms (1-2).”

Conclusion: Lines 421-422: “We found that having more than four pregnancy- and birth-related changes increased the likelihood of experiencing major depressive and anxiety symptoms postpartum”.

 

All authors have read and approved the revised manuscript.

Sincerely, Robin Aasegg Araya

on behalf of the authors

Reviewer 1

I had the pleasure to read and review the manuscript “Pregnancy- and birth-related experiences among postpartum women during the third wave of the COVID-19 pandemic – a multinational European study”. I would like to congratulate the authors for their meticulous work and rigorous description of the methods. The study describes how maternity care services were adapted during an advanced stage of the COVID-19 pandemic, using data reported by postpartum women in an online survey in 5 European countries. The study also looks at associations between the experienced changes in care provision and the mental health status of these women. It fills an important gap in the literature by presenting women’s voices. The findings show that maternity care services continued to be adapted in an advanced stage of the pandemic, highlighting the necessity of prioritizing pregnant and postpartum women’s needs in a pandemic setting, and beyond.

I consider this manuscript to be fit for publication after minor adjustments described below.

 

Comment 1: The authors write in-text that supplementary material 3 describes country-level restriction measures during the study period although no such information is presented in the table (only infection and vaccination rates). I recommend adding a summary of how restriction measures were implemented per country, as this information could be useful for the interpretation of the findings. These restriction measures could also explain some of the between-country differences documented in the results (see last comment on “comparison”).  For this purpose, it is also important to take into consideration the “time” component, as women answering the survey “postpartum” in June-August 2021 means that they were pregnant since as early as October/Nov 2020 (particularly important when looking at reported changes during the prenatal period – schedule of prenatal visits)

 

Response 1: Thank you for your encouraging comments and for raising the point about country specific restriction measures. Our initial reference to Supplementary Material 3 should have stated that it describes infection rates and vaccination status during the study period in the various countries. We have now corrected this. The main text now reads “Infection status and vaccination status in the respective countries for the period of interest are summarised in Supplementary Material 3”.

We purposely avoided showing COVID19 restriction measures in the various countries because these have been changing repeatedly, also within countries and would probably be incorrect. We hope the Reviewer understands our rationale.

 

Minor

Comment 2: Throughout the manuscript, recommend to consistently use the term COVID-19 (sometimes the word COVID is used instead).

Response 2: Thank you, we have revised this accordingly.

 

Comment 3: Abstract: Suggestion to specify the five countries in the abstract, and to provide some detail on the sampling and recruitment strategy.

Response 3: As suggested, we have specified the countries in the Abstract as follows:

“An online questionnaire was distributed in five European countries (Belgium, Netherlands, Norway, Switzerland, UK) between June and August 2021”

We have also added the following text: «Participants were recruited though social media platforms including pregnancy- and motherhood-related websites, pregnancy fora and apps.»

 

Comment 4: Introduction: Line 59-60: “Other studies describe limited birthing options (1), uncertainty or restrictions regarding partners’ visiting policies (1-3) and changes in birth locations due to the pandemic (7).” Authors mention “other studies” report “changes in birth location” and cite same study (#7) as the previous sentence. Please revise and align.

Response 4: Thank you for this notification. We have now corrected this by omitted the last part of the sentence “and changes in birth locations due to the pandemic (7)”.

 

Comment 5: Methods: Line 95: parenthesis missing before “questions 8 and 9).

Response 5: This typo has now been corrected.

 

Comment 6: Results and discussion: Suggest editing terminology used for interpreting the results - replace terms such as “risk” and “likelihood” with the term “odds” which is more convenient for the study design.

Response 6: We have replaced the term “risk” with “odds” as suggested. We believe that “likelihood” remains an appropriate term, and have kept it the few places it occurs in the manuscript. To meet further the Reviewer’s concern, we have rephrased the word “impact” with “association” or “relationship”, to avoid use of causal terminology.

 

Comment 7: Discussion: Line 311: authors mention that a strength of the study is the multi-country analysis “enabling comparisons across countries”. Yet the discussion of these comparisons is not addressed in detail in the text, only in the supplementary material, and not in regards to the findings. It is important to consider difference between countries when possible, taking into consideration existing guidelines before the pandemic (different models of care provision) and their adaptations during the pandemic, and the variance in restriction measures, which could explain some of the between-country differences (e.g. virtual ANC in the UK during COVID-19).

Response 7: We thank the Reviewer for this important comment. We have now removed the sentence “enabling comparisons across countries”. As correctly pointed out, we did not discuss country-specific findings in details on the main text. The reasoning for this is that in some countries the sample size was small, which challenges country-specific result interpretation. We have already addressed this as limitation, which reads “Study participation was also not evenly distributed among participating countries, with an overrepresentation of postpartum women from Norway.»

 

Reviewer 2

Thank you for the opportunity to review this interesting manuscript on pregnancy- and birth-related experiences among postpartum women from five EU countries during the third wave of the COVID-19 pandemic. The study included 1730 postpartum women and found that experiencing changes related to delivery or prenatal examination was associated with much higher depressive and anxiety symptoms postpartum. The study is well conducted, well written and the focus is important and needs attention.

 

Comment 1: However, I have some major concerns about the design and methods used in this study. First, as this is a cross-sectional study, we have no way of knowing if the women had depressive symptoms prior to experiencing these changes. Second, measures of maternal depression and changes in pregnancy and birth experiences were self-reported and not confirmed by a professional. This could introduce the problem of common method variance and risk of dependent measurement error. Common method variance reflects the variance that is attributed to the measurement method of the variables of interest (in this case self-report) rather than the constructs of interest. Method variance can either inflate or deflate observed relationships between constructs and lead to both Type I and Type II errors. Factors that have been reported to produce dependent measurement error are negative affectivity and social desirability. In the current situation, it could be that mothers who are likely to over-report (exaggerate) depression are also likely to over-report having had changes to pregnancy or birth plans (negative affectivity). Similarly, mothers who are likely to under-report depression are also likely to under-report changes (social desirability). In other words, the mechanism is not through the changes in pregnancy and birth experiences itself but through the reporting being from the same source using the same method. This needs to be discussed.

Response 1: Thank you for raising this important concern. We agree with the Reviewer that by design, the study lacks information about the temporal component. We fully acknowledged this limitation in the original version of the manuscript. We do also agree with the Reviewer that bias attributable to common method variance cannot be ruled out. At the same time, if this latter bias was present, we would have obtained homogenously similar results for the associations between “number” and “type” of birth-related changes with maternal mental health outcomes. Women over-reporting (exaggerating) depressive symptoms would be more likely to over-report having had multiple changes related to the pregnancy and/or birth plans. We do not observe this association between number of changes and outcomes after adjusting for confounding. We do only observe association in the adjusted models between specific types of birth-related changes. The specificity of the observed associations, along with the objectivity of the birth-related changes women were asked to report about (e.g., change in schedule for C-section or induction of labour), point against a meaningful risk of bias attributable to common method variance in this study.

To accommodate the Reviewer’s concern, we have added the following limitation:

“Because we measured mental health outcomes and birth-related changes in the same questionnaire, we cannot exclude the risk of common method bias. However, we consider this risk to be minimal as women were asked to report objective changes that occurred during prenatal follow-up and in relation to childbirth. Our distinct results for number versus type of birth-related changes corroborate the above assumption; if common method bias was substantial, we would have obtained similar associations across analyses”.

 

Comment 2:

Minor:

Besides this, I have a few minor issues that needs to be addressed.

 

• Line 82: maternal characteristics associated with changes? I think I know what you mean, but the sentence does not make much sense.

Response 2: We have rephrased this part of the Aims, which now reads «…and what factors characterised women having experienced these changes”. We hope the wording is clearer.

 

 

Comment 3: Could you elaborate why you have chosen to reduce changes in pregnancy and birth related experiences to “number of changes”? I guess that the experiences are qualitatively different and therefore reducing them to a number making them of equal importance could cause misclassification. At least this should be discussed. 

Response 3: Thank you for this important comment. We fully agree with the Reviewer, and that is why we explored pregnancy and birth-related changes in terms of number and type. We have addressed this comment in three sections. In the Methods, section 2.2.1, we have provided the rationale of looking at number of changes, which reads: “We examined number of changes in addition to type, as we assumed that multiple changes in pregnancy and birth-related experience may be of greater burden to the woman mental health.” In the Discussion, lines 338-339, we have added the following sentence: “Further, this classification may be too simplistic as it does not account for the different importance or seriousness of the individual changes”. In the Limitation section, lines 396-398, we have added: “Lastly, the pregnancy and birth-related experiences measured in the study are not equally important to women, and this may have affected our associations between postpartum mental health outcomes and number of pregnancy and birth-related changes.”

 

Comment 4: Please explain why numbers in Table 1 do not add up to 100% (e.g., Chronic mental illness and chronic somatic illness).

Response 4: Thank you for addressing this. We have checked, and the numbers should add up to 100% when including the missing figures as described in the footnote of the table. We have however corrected the variables “chronic mental illness” and “chronic somatic illness” as the online questionnaire was designed so that if a woman did not tick anything, it meant she did not have the illness, thus there were no missing values by default. Question 12 in the Appendix for the questionnaire demonstrates this by stating: “Please indicate for which chronic illnesses you have used a medication in the past 3 months? (i.e., chronic illnesses are conditions that already existed before your pregnancy/postpartum) (multiple answers possible)?».

We erroneously presented lack of presence of chronic illness as missing. When updating this, we also realized we needed to re-run analyses. See the explanation to the editor and reviewers above.

Round 2

Reviewer 2 Report

The authors have responded adequately to the review comments