Subgroup analyses in resisted sprint training reviews: Methodological practices and credibility assessment in meta-analyses

Subgroup analyses are widely used in meta-analyses on resisted sprint training (RST) to explore potential moderators such as training variables and athlete characteristics. While potentially informative, these analyses are often susceptible to bias when conducted without methodological rigor. This meta-epidemiological review examined the reporting quality and credibility of subgroup analyses across 15 RST meta-analyses published up to September 2025, yielding a total of 90 subgroup comparisons. Of these, 40% reported statistically significant findings, with sprint distance and training load emerging as the most consistent moderators, particularly in the acceleration phase and with moderate-to-heavy resistance. However, only a minority of reviews provided adequate information for assessing credibility, and most subgroup hypotheses were not pre-specified. Interaction testing was rare, and reporting was inconsistent. Methodological quality varied considerably, with two-thirds of the included reviews rated as low or critically low quality. Overall, subgroup analyses in RST are prevalent but often exploratory, underpowered, and lacking in transparency. Greater pre-specification, methodological consistency, and detailed reporting are needed to enhance the credibility and utility of subgroup findings in this field.