Review papers with raw data transparency

Concise critical appraisal

The narrative review synthesizes epidemiology, mechanisms, and prevention strategies linking early life (preconception to early childhood) exposures to later obesity, emphasizing DOHaD mechanisms including epigenetics, breast feeding and infant feeding, maternal obesity and GDM, obesogens, microbiome, activity, sleep and screen time, and interventions — but is limited by the narrative design and heavy reliance on observational studies with potential confounding Paper.

Full critical review and synthesis

1 Purpose and scope

The article aims to synthesize evidence linking exposures from preconception through early childhood to later obesity risk and to evaluate mechanisms and prevention strategies within the Developmental Origins of Health and Disease framework. The authors included human observational studies, animal models, epigenetic and microbiome literature and summarized interventions where available Scope.

2 Key claims made by the paper (evidence anchored)

• Maternal pre pregnancy BMI and excessive gestational weight gain are independently associated with higher childhood overweight risk; population attributable fractions estimated as substantial (21.7% to 41.7% for maternal overweight/obesity) Maternal
• Breastfeeding duration associates with lower early-childhood obesity risk and more favorable BMI trajectories in many cohort studies, but randomized evidence is inconclusive due to short follow up Breastfeeding
• Formula feeding higher protein content likely contributes to more rapid infant growth and later adiposity through insulin IGF1 and amino acid mediated mechanisms; trials of lower protein formula reduce but do not eliminate growth differences compared to breastfeeding Protein
• Epigenetic programming and microbiome differences are plausible mechanistic intermediates linking early exposures to later obesity but causal chain in humans remains incompletely proved Epigenetics
• Evidence for perinatal exposures such as cesarean delivery and obesogen (EDC) exposure is mixed and inconsistent in humans despite mechanistic plausibility from animal models Cesarean

3 Strengths of the paper

• Comprehensive scope covering preconception pregnancy infancy and early childhood with mechanistic and translational links across epigenetics microbiome and behaviour supported by 333 references Comprehensive
• Balanced discussion of mechanisms and explicit statements highlighting limits of available randomized evidence and confounding in observational studies Balanced

4 Limitations and critical concerns

1. Design limitation: narrative review format without preregistration or formal systematic meta analytic methods increases risk of selection bias and subjective emphasis; authors acknowledge this Narrative
2. Heavy reliance on observational studies with confounding: many key human associations (breastfeeding, cesarean, GDM) are vulnerable to socioeconomic confounding maternal health confounders and measurement heterogeneity; the review notes attenuation after adjustment in several studies and inability to infer causality in many cases Confounding
3. Heterogeneity across populations and exposures reduces generalizability: studies represent varied socioeconomic contexts and definitions of exposures and outcomes, and translatability from animal obesogen models to human risk is uncertain Heterogeneity
4. Intervention evidence weak: while the review discusses prevention strategies (preconception counseling diet and exercise in pregnancy breastfeeding promotion responsive feeding and early activity) it accurately reports that RCTs showing durable reductions in childhood obesity are limited or show modest effects Weak

5 Missing details and blindspots

• Quantitative synthesis missing: pooled effect estimates, heterogeneity statistics and publication bias assessment are not provided because of narrative design; this reduces precision in assessing effect sizes and comparability across exposures No
• Limited discussion of measurement error and exposure misclassification in key exposures such as obesogen assays dietary intake and breastfeeding exclusivity; these biases can change effect directions Measurement
• Transgenerational claims are presented cautiously but require stronger germline human evidence; animal models suggest mechanisms yet human multigenerational causal proof is scarce Transgenerational

6 Practical takeaways for researchers and clinicians

• Prioritize preconception maternal health and weight optimization as plausible high impact target given strong epidemiologic signal for maternal BMI and GWG and plausible biologic mediators Preconception
• Breastfeeding promotion remains biologically plausible and supported by cohort data but RCT level causation remains unresolved so policy should combine promotion with rigorous evaluation of long term body composition outcomes Breastfeeding
• Design future trials to measure body composition not only BMI, include long follow up, pre-specify confounder control, and embed mechanistic sampling (epigenetics metabolomics microbiome) to move from association to causation Design

7 Suggested improvements to strengthen the work

1. Perform or integrate formal systematic reviews and meta analyses for major exposures to deliver quantitative pooled estimates and heterogeneity assessment.
2. Provide a structured risk of bias table per major topic (breastfeeding formula composition obesogens GDM cesarean) summarizing study designs, confounding adjustment, and exposure measurement quality.
3. Include explicit recommendations for minimum mechanistic sampling in future RCT protocols (placenta cord blood epigenetics targeted metabolomics serial microbiome sampling) with suggested timepoints.
4. Highlight population differences and equity considerations explicitly as biological effects may interact substantially with socioeconomic context.

8 Confidence and falsifiability

Conclusions on association strength vary by exposure: confidence is higher for maternal BMI and gestational weight gain links to child adiposity (higher quality epidemiology and consistent signal) and lower for obesogens and cesarean where human causal data are inconsistent; the authors explicitly state how evidence could be falsified (robust studies showing null associations after confounder control and mechanistic disconfirmation) Confidence

9 Final evaluation summary

The paper is a timely, broad, and well referenced narrative synthesis that usefully maps mechanistic and epidemiologic pathways from preconception to early childhood; it responsibly states limitations and research gaps. Its narrative nature, however, constrains quantitative inference and raises risk of selection bias. The review is most valuable as a roadmap for research priorities rather than as a definitive evidence grading document.