Vitamin D-Containing Nutritional Interventions for Adults with Depressive Disorders or Clinically Relevant Depressive Symptoms: A Bayesian Hierarchical Systematic Review with Network and Component Feasibility Assessment

This dataset supports a systematic review and Bayesian hierarchical evidence synthesis evaluating vitamin D-containing nutritional interventions in adults with depressive disorders or clinically relevant depressive symptoms.

The repository contains search and screening records, screening-audit datasets, extracted study-level and arm-level data, contrast-level analysis datasets, component-coding files, frequentist and Bayesian analysis inputs and outputs, network/component feasibility summaries, figure source data, and R scripts used to reproduce the analyses. Continuous depressive-symptom outcomes were harmonized as Hedges’ g standardized mean differences, with negative values favoring vitamin D-containing interventions.

The primary quantitative analysis was a Bayesian random-effects pairwise meta-analysis of endpoint depressive symptom severity in the expanded oral vitamin D evidence layer. Frequentist endpoint and change-score analyses were conducted as robustness and data-validation layers. Network meta-analysis and component network meta-analysis were assessed for feasibility but were not implemented as confirmatory models because the endpoint treatment-node network was fragmented and non-vitamin-D components lacked sufficient separable cross-study variation.

This repository is intended to provide the data and code needed to reproduce the main quantitative results, robustness analyses, network/component feasibility assessment, and figure/table source outputs. Copyrighted full-text articles are not redistributed. Formal supplementary materials are submitted separately with the manuscript rather than deposited as the primary content of this data repository.