Prevalence of Clinical and Pre-Clinical Obesity at Six Months Postpartum Following Gestational Diabetes Mellitus

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The present manuscript describes how a population that the authors described as GDM and Non GDM women during postpartum develop obesity and have organ dysfunction. However, this is not reflected in the title or introduction. These sections should be improved, and the information should be declared.

In this study, there is a good number of women in the follow-up period, which is not common to have.

Figure 2 should be in the supplementary material.

The model proposed in Table 2 should be considered for both experimental groups; the same should be considered for Table 3.

 

 

Author Response

RESPONSE TO REVIEWER 1 - Manuscript nutrients-4051197, entitled, “Prevalence of clinical and pre-clinical obesity at six months postpartum”

We thank the reviewers for their constructive feedback. We have carefully addressed each point and revised the manuscript accordingly. Below is a point-by-point response, with references to line numbers in the revised manuscript. Text that has been inserted here from the manuscript is presented in italics.

Response: In our responses to the specific comments, we believe that we have addressed the reviewer’s concerns regarding the introduction (Comment 1), research design (Comment 3), and figures and tables (Comment 2). However, we were unsure how the reviewer wished us to address deficiencies in the conclusion, as we are clear that this is a post-GDM cohort and controls. We would be happy to address further comments.

1. The present manuscript describes how a population that the authors described as GDM and Non GDM women during postpartum develop obesity and have organ dysfunction. However, this is not reflected in the title or introduction. These sections should be improved, and the information should be declared.

Response: We agree that our focus on women following pregnancy complicated by GDM was not reflected in the title, which we have revised by adding, “…following gestational diabetes mellitus” (line 3). However, in the introduction, we have the following text referring to GDM, which we believe provides the rationale for our focus. We were unsure how to improve this, but are of course open to specific suggestions.

“…As these women have a ten-fold higher risk of future pre-diabetes or type 2 diabetes [8], and more than a two-fold higher risk of cardiovascular events in the first decade postpartum [9], we have prospectively followed a cohort of women with GDM (and non-GDM controls), and assessed many criteria for adiposity-related organ dysfunction.

In this study, we estimate the prevalence of obesity-related organ dysfunction post-partum among women with GDM in their most recent pregnancy (vs. non-GDM controls), as they are at potentially higher risk of cardio-renal-metabolic forms of adiposity-related organ dysfunction” (lines 68-77).

 

2. In this study, there is a good number of women in the follow-up period, which is not common to have. Figure 2 should be in the supplementary material.

Response: Thank you for this suggestion. Figure 2 has now been moved to the Supplementary Material, where it is now designated as Supplementary Figure 1 (lines 278, 282-283).

 

3. The model proposed in Table 2 should be considered for both experimental groups; the same should be considered for Table 3.

Response: Thank you for this comment. We apologise for our lack of clarity, which we have addressed in the Analyses for each table, as follows (with new text highlighted in yellow: Table 1 (“For women with excess adiposity (as defined above), the following were presented descriptively: baseline characteristics…”) (lines 178-179), Table 2 (“Among women with excess adiposity, backward multivariable logistic regression…”) (line 185), and Table 3 (“…likely to be causal, by calculating the prevalence of organ dysfunction in clinical obesity minus the prevalence of the same organ dysfunction among GDM and non-GDM controls with BMI <30kg/m2 (who had not otherwise been included in the previous analyses)” (line 199-200).

We understand that by “experimental groups”, the reviewer is referring to the cohort presented as women with excess adiposity and either preclinical or and clinical obesity, or with or without prior GDM, as presented in Table 1. These were presented for descriptive purposes, but the objective of the analysis presented in Table 2 (for prediction of clinical obesity) was to focus on the application of the Lancet Commission framework in the whole cohort, and to evaluate whether GDM was predictive. The results indicate that GDM was associated with clinical obesity in the univariate, but not multivariate, analysis. The purpose of the analysis presented in Table 3 was to estimate how much of the excess adiposity-associated organ dysfunction may be causal, by estimating how much of such organ dysfunction is present in women without obesity, and then subtracting it from the prevalence of organ dysfunction present in women with clinical obesity.

“Table 3 includes the 1137 women with BMI <30kg/m2, 695 with prior GDM and 442 without. The prevalence of organ dysfunction criteria…” (lines 287-288). Also, we have corrected the third column, which presents findings among women with excess adiposity (excluding the 9 women with BMI ≥30kg/m²but WHR ≤0.5).

Yours sincerely,

Kypros H. Nicolaides

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript titled, Prevalence of clinical and pre-clinical obesity at six months postpartum, assessed the prevalence of clinical and pre-clinical obesity at six months postpartum using the Lancet Commission framework, which expands beyond BMI to include organ dysfunction related to adiposity. In a UK cohort, the authors reveal that many women deemed non-obese by BMI already show signs of pre-clinical obesity, underscoring BMI's limitations in postpartum risk evaluation. The study also estimates organ dysfunction attributable to adiposity and investigates predictors of clinical obesity, finding that routine clinical variables have limited predictive ability. Although its observational nature and use of surrogate adiposity measures are limitations, the study offers valuable evidence for more refined obesity classification in postpartum health management. Although this study is well organized and informative, there are several limitations as follows. Addressing these limitations would improve the quality and scientific soundness of this study.

Major Concerns

1. Although the manuscript follows the Lancet Commission framework, it remains conceptually confusing to differentiate between excess adiposity, pre-clinical obesity, and clinical obesity. Its operationalization mainly depends on BMI and WHR, but these surrogate measures are recognized as imperfect. The lack of direct adiposity assessments such as DXA or BIA reduces the accuracy of classifying women as non-obese or pre-clinically obese.

 

2. The study is observational and cross-sectional at six months postpartum. However, some parts of the Discussion suggest a causal relationship between excess adiposity and organ dysfunction. Additionally, the “adiposity-attributable” prevalence estimate is obtained indirectly and should be approached with caution.

 

3. Subtracting the prevalence of organ dysfunction in women with BMI <30 kg/m² from that in women with excess adiposity assumes the groups are comparable and overlooks residual confounding factors such as genetics, prior disease, and socioeconomic status. This method oversimplifies intricate causal relationships.

 

4. The clinical obesity prediction model demonstrates modest discrimination with an AUC of 0.64, but the implications of this limited predictive ability are not thoroughly examined. Declaring that clinical obesity is “not predictable” may be an overstatement without considering alternative models or predictors such as lifestyle or biomarkers.

 

5. The cohort originates from a single UK tertiary center with a structured postpartum follow-up program. Attendance rates were notably high, which could restrict how well the findings apply to typical clinical settings or populations with fewer resources.

Minor Concerns

1. While justified pragmatically, the exclusive use of BMI ≥30 kg/m² (without ethnic-specific thresholds) may underestimate adiposity-related risk in some populations. This issue could be better integrated into the interpretation of results.

 

2. Table 1 is dense and hard to interpret. Consider simplifying them or moving detailed subgroup analyses to the supplementary material.

 

3. Sections comparing the findings with existing frameworks (EOSS, EASO, Commission) are informative but somewhat repetitive and could be streamlined.

 

Author Response

RESPONSE TO REVIEWER 2 - Manuscript nutrients-4051197, entitled, “Prevalence of clinical and pre-clinical obesity at six months postpartum”

We thank the reviewers for their constructive feedback. We have carefully addressed each point and revised the manuscript accordingly. Below is a point-by-point response, with references to line numbers in the revised manuscript. Text that has been inserted here from the manuscript is presented in italics.

The manuscript titled, Prevalence of clinical and pre-clinical obesity at six months postpartum, assessed the prevalence of clinical and pre-clinical obesity at six months postpartum using the Lancet Commission framework, which expands beyond BMI to include organ dysfunction related to adiposity. In a UK cohort, the authors reveal that many women deemed non-obese by BMI already show signs of pre-clinical obesity, underscoring BMI's limitations in postpartum risk evaluation. The study also estimates organ dysfunction attributable to adiposity and investigates predictors of clinical obesity, finding that routine clinical variables have limited predictive ability. Although its observational nature and use of surrogate adiposity measures are limitations, the study offers valuable evidence for more refined obesity classification in postpartum health management. Although this study is well organized and informative, there are several limitations as follows. Addressing these limitations would improve the quality and scientific soundness of this study.

Response: We believe that we have made the requested improvements that address the reviewer’s concerns about our: introduction (Comment 3a); results, research design, methods and conclusions (Major comments 1, 2, 4 & 5, and Minor comment 1a & 3a); and figures/tables (Minor comments 2a).

Major Concerns

1. Although the manuscript follows the Lancet Commission framework, it remains conceptually confusing to differentiate between excess adiposity, pre-clinical obesity, and clinical obesity. Its operationalization mainly depends on BMI and WHR, but these surrogate measures are recognized as imperfect. The lack of direct adiposity assessments such as DXA or BIA reduces the accuracy of classifying women as non-obese or pre-clinically obese.

Response: We agree with this comment. Our hypothesis was formulated retrospectively, and the relevant information was not available. As such, we had already cited this as a limitation of our study, as follows: “Limitations of our study include the lack of information on additional criteria for assessment of excess adiposity (such as direct measurements of body fat) that would have allowed us to more accurately exclude obesity in women BMI < 30 kg/m²; as such, we may have under-estimated the prevalence of excess adiposity and clinical obesity postpartum” (lines 367-371, no changes made).

 

2. The study is observational and cross-sectional at six months postpartum. However, some parts of the Discussion suggest a causal relationship between excess adiposity and organ dysfunction. Additionally, the “adiposity-attributable” prevalence estimate is obtained indirectly and should be approached with caution.

Response: We thank the reviewer for this important comment. We agree that, given the observational and cross-sectional design of the study, caution should be exercised in making causal inferences. We have therefore revised the Discussion to remove or rephrase statements that could be interpreted as implying causality, and we now consistently describe the findings as associations rather than causal effects, as follows: “Our findings suggest that in a population of postpartum women following GDM and their non-GDM controls, excess adiposity may be common, over half of these women have end-organ dysfunction, and up to 1/3 of that is potentially associated with obesity (i.e., clinical obesity) and may be modifiable with weight loss” (lines 303-306).

 

3. Subtracting the prevalence of organ dysfunction in women with BMI <30 kg/m² from that in women with excess adiposity assumes the groups are comparable and overlooks residual confounding factors such as genetics, prior disease, and socioeconomic status. This method oversimplifies intricate causal relationships.

Response: We thank the reviewer for this comment and we agree with the limitations of our study, which was designed to estimate in our cohort, the frequency of end-organ dysfunction that the Lancet Commission has associated with obesity. However, as the reviewer correctly points out, the study was not intended to establish causal relationships. We have now tried to make this clearer. First, we have now stated this limitation in our Discussion: “We have no information about factors that affect adiposity and metabolic health (particularly diet, physical activity, genetic factors, biomarkers and other confounders) that would have allowed us to adjust for residual confounding of the relationship between adiposity and clinical obesity criteria” (lines 371-374). Second, we have been more cautious about potential causal relationships (as per our response to Comment 2, above).

 

4. The clinical obesity prediction model demonstrates modest discrimination with an AUC of 0.64, but the implications of this limited predictive ability are not thoroughly examined. Declaring that clinical obesity is “not predictable” may be an overstatement without considering alternative models or predictors such as lifestyle or biomarkers.

Response: We agree with the reviewer that an AUC of 0.64 reflects only modest discriminatory performance and that the implications of this limited predictive ability require careful interpretation. We have revised the Discussion to more explicitly acknowledge the modest performance of the model and to avoid overstatement regarding the lack of predictability. Specifically, we have rephrased statements suggesting that clinical obesity is “not predictable” to clarify that, within the constraints of the available clinical variables and the current model, predictive performance was limited, as follows: “Importantly, clinical obesity in our cohort has limited predictability from available…” (line 306-307). Also, in the limitations, we have added to our previous statement about important sources of residual confounding. “We have no information about factors that affect adiposity and metabolic health (particularly diet, physical activity, genetic factors, biomarkers and other confounders) that would have allowed us to adjust for residual confounding of the relationship between adiposity and clinical obesity criteria” (lines 371-374).

 

5. The cohort originates from a single UK tertiary center with a structured postpartum follow-up program. Attendance rates were notably high, which could restrict how well the findings apply to typical clinical settings or populations with fewer resources.

Response: We acknowledge that the study was conducted in a single UK tertiary center with a structured postpartum follow-up program and high attendance rates, the potential reasons for which we have outlined in the Discussion (lines 319-356). However, we have now explicitly addressed this limitation in the Discussion, as follows: “Finally, as the study was conducted at a single tertiary care center with a structured postpartum follow-up program and high attendance rates, the generalizability of the findings and, consequently, their external validity should be considered when extrapolating the results” (lines 381-384).

Minor Concerns

1. While justified pragmatically, the exclusive use of BMI ≥30 kg/m² (without ethnic-specific thresholds) may underestimate adiposity-related risk in some populations. This issue could be better integrated into the interpretation of results.

Response: We agree with the reviewer that using a BMI threshold of ≥30kg/m² without considering ethnic-specific cut-offs may underestimate adiposity-related risk in certain populations, and had included this in our limitations (lines 374-379). However, to make this point more clearly, we have added, as a second paragraph of the Discussion: “While our findings indicate associations between obesity and organ dysfunction, it is important to note that obesity was defined using a BMI threshold of ≥30 kg/m² without accounting for ethnic-specific cut-offs. Consequently, the burden of adiposity-related risk may be underestimated in populations for whom lower BMI thresholds are recommended” (lines 313-317).

 

2. Table 1 is dense and hard to Consider simplifying them or moving detailed subgroup analyses to the supplementary material.

Response: Thank you for this feedback. We apologise for the sub-optimal presentation of study findings. We have now moved the pre-clinical and clinical obesity columns and their comparison to the supplementary material, as Supplementary Table 1, with appropriate referencing in the text (line 268).

 

3. Sections comparing the findings with existing frameworks (EOSS, EASO, Commission) are informative but somewhat repetitive and could be streamlined.

Response: Thank you for this feedback. We have now relied on information about these other frameworks that we provided in the introduction, and combined the third paragraph of 4.1 with the second paragraph of 4.2, into a shortened paragraph, as follows: “The recent Lancet Commission defined clinical and pre-clinical obesity to reflect the nuanced impact of excess adiposity at the individual-level, and in doing so, facilitate a medically-meaningful approach to obesity care. Our findings support implementation of this diagnostic framework in the clinical assessment of postpartum women, distinguishing between those with ongoing disease (clinical obesity) and those with a variable level of health risk without illness (pre-clinical obesity), although it appears that this approach would require systematic assessment of the signs and symptoms of clinical obesity, which is not yet part of standard clinical practice. Identifying individuals with clinical obesity can help with the choice and prioritisation of available anti-obesity interventions [3]. This approach aligns with EOSS[1] and the EASO framework[2]. While lifestyle interventions may play a greater role in decreasing overall health risk for women with pre-clinical obesity, these approaches are alone, likely insufficient to improve organ function and related clinical manifestations in women with clinical obesity. Qualitative work has highlighted that women find relevant changes challenging, and lower rates of compliance reduce their impact [23].” (lines 331-345).

 

Yours sincerely,

Kypros H. Nicolaides

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

the authors responded to all the suggestions made

Reviewer 2 Report

Comments and Suggestions for Authors

The authors addressed all of my concerns with the previous version of the manuscript. So, I recommend acceptance of this revised version.