Trial
Diabetes Prevention Program / DPPOS
Research Observatory
Evidence Update
Current evidence
The Diabetes Prevention Program (DPP) randomized 3,234 adults with impaired glucose tolerance to metformin 850 mg twice daily, intensive lifestyle intervention or placebo. DPPOS followed the same cohort, with later shared lifestyle programmes and treatment convergence. DPP/DPPOS shows sustained diabetes delay after lifestyle or metformin assignment, without an overall between-group aggregate microvascular benefit at 15 years. DPPOS did not establish lower all-cause mortality or major cardiovascular events from original metformin or lifestyle assignment over about 21 years.
What changed
DPP and DPPOS form one longitudinal programme with intervention convergence, not separate replications for each paper.
What did not change
Existing MILES/MASTERS results and TAME's design-only status are not superseded.
Why it matters
Adds a distinct patient-outcome question without converting diabetes delay into geroprotection.
Main limitation
Interventions converged; prevention/delay is not mortality benefit. These participants are distinct from the MILES, MASTERS and planned TAME populations. Dilution from crossover, outside cardioprotective therapy and intervention convergence; no equivalence or proof of universal ineffectiveness.
Next proof
Independent evidence for cardiovascular, microvascular and mortality outcomes; diabetes delay alone cannot establish those benefits.
Summary
The Diabetes Prevention Program (DPP) randomized 3,234 adults with impaired glucose tolerance to metformin 850 mg twice daily, intensive lifestyle intervention or placebo. DPPOS followed the same cohort, with later shared lifestyle programmes and treatment convergence. DPP/DPPOS shows sustained diabetes delay after lifestyle or metformin assignment, without an overall between-group aggregate microvascular benefit at 15 years. DPPOS did not establish lower all-cause mortality or major cardiovascular events from original metformin or lifestyle assignment over about 21 years.
Intelligence Take
Adds a distinct patient-outcome question without converting diabetes delay into geroprotection.
Limitations
Interventions converged; prevention/delay is not mortality benefit. These participants are distinct from the MILES, MASTERS and planned TAME populations. Dilution from crossover, outside cardioprotective therapy and intervention convergence; no equivalence or proof of universal ineffectiveness.