This dataset contains the data files used in the Discovery, Validation/Replication, and sensitivity analyses of the manuscript “Aerobic fitness and health-related phenotypes: A two-stage phenome-wide Mendelian randomization study”. The study evaluated potentially causal associations between genetically predicted aerobic fitness and health-related phenotypes using a two-sample, two-stage phenome-wide Mendelian randomization framework. The main analyses used 160 independent genetic variants associated with aerobic fitness, as operationalized by Cai et al. (2023), as instrumental variables. Discovery and Validation/Replication analyses were conducted using publicly available European-ancestry GWAS summary statistics from OpenGWAS. A total of 712 Discovery–Validation GWAS pairs were evaluated. The deposited files include the databases used for the main Discovery and Validation/Replication analyses. These data support the analyses reported in the associated manuscript and allow users to identify the GWAS outcome datasets and genetic instruments retained for each Mendelian randomization analysis.

The dataset contains analysis databases derived from publicly available GWAS summary statistics. No individual-level data were generated or used. The GWAS summary statistics analyzed in the study are available from the MRC IEU OpenGWAS repository, and the specific OpenGWAS accessions used in the Discovery and Validation phases are listed in the supplementary materials of the associated manuscript.