Sociodemographic Factors and Childhood Growth: Associations with Environmental Sanitation Phases

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper addresses an important public health question about the relationship between environmental sanitation and child growth. While the longitudinal data collection is commendable and the statistical methods are generally appropriate, I have major concerns about the study design, interpretation of findings, and framing of conclusions that must be addressed before this manuscript can be considered for publication.

Cohort Effects Confounded with Intervention

My most serious concern is that this study cannot support causal claims about sanitation effects on growth. The fundamental problem is that you are comparing three different birth cohorts (born in ~1997, 2000, and 2003) rather than following the same children before and after sanitation improvements. This means any differences between your "phases" reflect not only sanitation changes but also economic growth, healthcare improvements and expansions, other social programs and interventions to support socioeconomic status, and shifts in feeding practices. Your study design cannot disentangle sanitation effects from these temporal trends. This is not simply a limitation—it fundamentally undermines the causal interpretation that pervades your manuscript.

Additionally:

• Only 4.5% of intervention-phase children (46/1,007) were actually born during the pre-intervention period (you mention this in limitations, but do not discuss how it affects your interpretation or conclusions)
• Sample sizes are highly imbalanced (n=299, 1,007, 123)
• Children start at different ages across phases (means: 0.9, 1.4, 2.7 years)
• No control areas without sanitation improvements were measured

Table 1 also reveals substantial baseline differences that could independently explain growth variations, such as birth weight changed, breastfeeding patterns shifted dramatically: never exclusively breastfed went from 9.7% (pre) → 55.7% (intervention) → 16.3% (post), c-sections increased. 

These temporal changes could easily explain the growth differences without invoking sanitation. The discussion must seriously consider and address these alternative explanations rather than attributing differences primarily to sanitation.

Intervention Definition Missing

Throughout the manuscript, the intervention is never clearly operationalized. What exactly changed during these "sanitation phases"? Was it coverage expansion? Household connections to sewerage systems? Improved toilet facilities? Wastewater treatment infrastructure? Movement up the sanitation ladder? Hygiene education?

The methods mention "sanitation actions" and the Bahia Azul program but never specify what this meant at the household or community level. Without a clear definition of the intervention being evaluated, the findings cannot be meaningfully interpreted. The introduction needs substantial detail about what infrastructure was built, what changed for households, the implementation timeline, and geographic coverage.

Contradictory RCT Evidence Not Discussed

Your discussion completely ignores that major randomized controlled trials have consistently found little to no impact of sanitation interventions on child anthropometric outcomes, including:

• WASH Benefits trials (Kenya and Bangladesh)
• SHINE trial (Zimbabwe)
• Other sanitation RCTs systematically reviewed in the literature

If well-designed experiments with clearly defined interventions don't demonstrate growth effects, how should we interpret positive findings from an observational study with substantial confounding? You must engage with this body of evidence and discuss:

• Why your results might differ (longer follow-up? community-level coverage? different context?)
• How your findings relate to or contradict experimental evidence
• What unique aspects of Bahia Azul might explain different outcomes
• Without this discussion, the manuscript reads as if you are unaware of contradictory evidence that directly challenges your conclusions.

Overstated Causal Language

This needs attention: throughout the discussion, you use causal language not supported by your design:

• "improvements in sanitation showed better linear growth"
• "environmental interventions may contribute to improved child growth"

This framing is inappropriate because it sounds like you are attributing all differences in outcomes to sanitation, but how you are detangling the effects of sanitation programs specifically from other overall societal improvements is unclear. At most, you can state: "Children born in later time periods showed better growth than those born earlier, which may reflect sanitation improvements along with numerous other societal changes occurring in Brazil during this period." Even then, you need a thorough discussion of other variables that changed, from which you cannot detach one sanitation program.

Missing Statistical Comparisons and Effect Sizes

I am not sure, but perhaps, there is a way to formally test whether trajectories differ between phases. Table 2 shows associations within each phase separately but provides no statistical test of between-phase differences or phase × covariate interactions. Figure 2 presents visual differences and claims "children in the post-intervention phase demonstrated better growth," but no statistical test quantifies this.

Other Concerns:

• Why do sanitation effects appear mainly in males (Figure 2a)?
• Why did obesity increase in the post-intervention phase if this environment was optimal?
• How do you reconcile highest cesarean rates in the post-intervention phase with your discussion of cesarean delivery as problematic?
• Effect sizes are unclear—how many centimeters taller were post-intervention children on average?
• Clinical significance is not discussed alongside statistical significance
• No adjustment for multiple comparisons

 

To make this manuscript suitable for publication, you must:

• Remove or substantially qualify all causal language throughout the abstract, results, and discussion by reframe the entire paper as a descriptive study of temporal trends in child growth, not a causal evaluation of sanitation effects
  • Revise the conclusion to something like: "Children born in later time periods showed better growth trajectories, which may reflect the combined effects of sanitation improvements, economic development, and other societal changes in Brazil during this period. Further research using experimental or quasi-experimental designs is needed to isolate the specific contribution of sanitation to child growth."
• Clearly define the sanitation intervention with specific operational details
• Add a comprehensive section discussing the contradictory RCT evidence and how your findings relate to it
• Consider adding formal statistical tests comparing trajectories between phases with appropriate effect sizes if it makes sense to do (optional)
• Expand the discussion of alternative explanations, including temporal trends in birth weight, feeding practices, healthcare access, and economic conditions
• Address contradictions like obesity increase, cesarean delivery patterns, and breastfeeding effects
• Add a limitations section that explicitly acknowledges that cohort effects are completely confounded with sanitation phases and that causal attribution is not possible

Comments for author File: Comments.pdf

Author Response

We sincerely thank all reviewers for their thoughtful and constructive comments, which have significantly strengthened our manuscript. Each comment has been carefully analyzed and addressed.
Due to the extensive nature of the revisions and the detailed point-by-point responses required, we have prepared a comprehensive document which is attached as supplementary material to this submission. This document contains: each original reviewer comment in full;  our detailed response to each point raised; the specific location of each change in the revised manuscript (page number, line numbers).

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Please see my comments attached.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The writing quality is good, but there are some places where the phrasing confused me. I thought this might be due to mental translations between languages. In any case, they can be caught and corrected.

Author Response

We sincerely thank all reviewers for their thoughtful and constructive comments, which have significantly strengthened our manuscript. Each comment has been carefully analyzed and addressed.
Due to the extensive nature of the revisions and the detailed point-by-point responses required, we have prepared a comprehensive document which is attached as supplementary material to this submission. This document contains: each original reviewer comment in full;  our detailed response to each point raised; the specific location of each change in the revised manuscript (page number, line numbers).

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript uses a longitudinal cohort to study the association between changes in environmental sanitation and anthropometric trajectories from infancy through adolescence. Overall, the manuscript is in good shape. I have a few comments that should be addressed before it can be accepted for publication:

1. Please explain why BIC was used to select the best model. In addition, could you clarify what is meant by “data not presented” for the BIC values? Does this refer to confidential data, or is there another reason these results are not shown?

2. Are the covariates included in both the quadratic time model with random intercept only, and the quadratic time model with random intercept and random slope for the time variable in each phase? Please clarify how covariates are handled in each model.

3. Please elaborate on whether the assumptions of the statistical models were checked prior to their application (e.g., distributional assumptions, independence, etc.), and how these diagnostics were performed.

4. In Figure 2, there appears to be a different pattern between boys and girls in HAZ. Could you please explain this difference in more detail and provide a possible interpretation?

Author Response

We sincerely thank all reviewers for their thoughtful and constructive comments, which have significantly strengthened our manuscript. Each comment has been carefully analyzed and addressed.
Due to the extensive nature of the revisions and the detailed point-by-point responses required, we have prepared a comprehensive document which is attached as supplementary material to this submission. This document contains: each original reviewer comment in full;  our detailed response to each point raised; the specific location of each change in the revised manuscript (page number, line numbers).

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Please see attachement. 

Comments for author File: Comments.pdf

Author Response

We sincerely thank all reviewers for their thoughtful and constructive comments, which have significantly strengthened our manuscript. Each comment has been carefully analyzed and addressed.
Due to the extensive nature of the revisions and the detailed point-by-point responses required, we have prepared a comprehensive document which is attached as supplementary material to this submission. This document contains: each original reviewer comment in full;  our detailed response to each point raised; the specific location of each change in the revised manuscript (page number, line numbers).

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors clearly explained many of their original decisions and made appropriate changes where needed. My original concerns were all addressed, and the manuscript is significantly improved. This is an interesting analysis and excellent presentation of the results. 

The authors' cover letter was cut off and did not include responses to comments in the Results and Discussion. I was able to check my original review to assess revisions, and I am satisfied without seeing those comments. However, I wanted to point out the potential issue with uploading or sharing.

Reviewer 4 Report

Comments and Suggestions for Authors

The extensive and detailed revisions certainly improve the overall submission.  Most important are the authors' transparent admissions as to design weakness and inadequate power due to low sampled numbers in some cohorts, and other sub-optimal design considerations (some of which were due to the secondary nature of the datasets.

The findings are potentially useful to the public health and environmental health fields, and I believe this study provides a small-scale "pilot" model for a more comprehensive design in the future.