Trial
MASTERS
Research Observatory
Evidence Update
Current evidence
MASTERS trial intelligence: MASTERS found smaller whole-muscle/lean-mass gains with metformin during resistance training, but no significant between-arm difference in the powered type-II-fiber endpoint or reported strength/power measures. A later MASTERS reanalysis suggests a statin-associated modifier of muscle-fiber growth under metformin, but does not establish a causal combination-therapy benefit or independent replication.
What changed
The accepted result preserves endpoint hierarchy, the corrected relative-strength comparison and the later nonrandomized statin-subgroup context.
What did not change
The study does not establish generalized harm from metformin, causal benefit from adding a statin or an independent replication of its own subgroup analysis.
Why it matters
Exercise exposure and endpoint scale are essential to interpreting this trial: whole-muscle findings cannot be substituted for the powered fiber endpoint or clinical function.
Main limitation
Strength1RM,p=.055, isometric,p=.082, power,p=.064: trends, not significant harm or equivalence. Corrected relative-strength gains18.9% versus14.8%,p=.30. No no-exercise arm; no generalization to all metformin use or frail/diabetic populations. Statin users differed in sex and other medication use; small and varying biopsy denominators, multiple comparisons, no randomized statin assignment. Cannot reverse the whole-cohort result.
Next proof
Independent prespecified drug-by-exercise trials with clinically relevant function and adequately powered interactions. Prospectively randomized factorial or stratified trial testing the interaction.
Summary
MASTERS trial intelligence: MASTERS found smaller whole-muscle/lean-mass gains with metformin during resistance training, but no significant between-arm difference in the powered type-II-fiber endpoint or reported strength/power measures. A later MASTERS reanalysis suggests a statin-associated modifier of muscle-fiber growth under metformin, but does not establish a causal combination-therapy benefit or independent replication.
Intelligence Take
Exercise exposure and endpoint scale are essential to interpreting this trial: whole-muscle findings cannot be substituted for the powered fiber endpoint or clinical function.
Limitations
Strength1RM,p=.055, isometric,p=.082, power,p=.064: trends, not significant harm or equivalence. Corrected relative-strength gains18.9% versus14.8%,p=.30. No no-exercise arm; no generalization to all metformin use or frail/diabetic populations. Statin users differed in sex and other medication use; small and varying biopsy denominators, multiple comparisons, no randomized statin assignment. Cannot reverse the whole-cohort result.