HRV–3PAD: A Documentary, Numerical, and Methodological Audit of Reported Incremental Heart Rate Variability Prediction in the 3PAD Study — Part III: Final Investigation and Closure

This report presents Part III, the final stage of a three-part documentary, numerical, and methodological investigation of the 2026 study entitled "Exploring early and late pregnancy heart rate variability as incremental predictors of postpartum depression and anxiety symptoms", published in Psychoneuroendocrinology (Volume 188, Article 107813, 2026). The investigation examines the reported incremental predictive contribution of pregnancy heart rate variability (HRV) beyond clinical, psychological, and background predictors in the 3PAD study. Building on the numerical reconciliation work in Part I and the expanded methodological and documentary assessment in Part II, this final report consolidates the available evidence and establishes the overall scientific classification. The report addresses: • The published mean absolute error (MAE) differences for postpartum depression, state anxiety, and trait anxiety. • Numerical discrepancies between Supplementary Tables 9a and 9b, including seven of twelve comparisons that exceeded the approximate rounding tolerance. • The reporting and reproducibility of Random Forest modelling, feature selection, validation procedures, and multiple-imputation methods. • Differences between the doctoral dissertation and the final journal article in the description of HRV predictors and HRV difference-score direction. • Sample size, HRV measurement, model comparison, generalisability, and external validation. • The availability and verification status of executable code, repositories, imputation objects, and prediction-level evidence. The final classification is PARTIALLY SUPPORTED. The published analysis reports incremental MAE improvements for postpartum depressive symptoms and state anxiety, whereas the trait-anxiety result remains statistically inconclusive. The numerical reconciliation issue remains unresolved. Independent reproduction, exact-model external validation, and clinical utility have not been established. Importantly, this report does not claim to have independently reproduced the original models, does not establish a methodological defect or misconduct, and does not disprove the original study. It distinguishes published findings from unresolved documentary and methodological questions. This document is Part III of a continuous three-part investigation and provides its final evidence assessment and closure as of 9 November 2026. Original article DOI: https://doi.org/10.1016/j.psyneuen.2026.107813 PubMed PMID: 41780318 Author: Omar Magdy Mohamed Abdelkader Country: Egypt Document type: Documentary and methodological research audit Peer-review status: Not peer reviewed.