Metformin versus placebo treatment group effect on the risk (hazard) of diabetes onset before and after adjustment for each time-dependent covariate individually and in combination

Covariate (unit)Metformin versus placebo group effect*
Covariate effect
Percent group R2 explained
CoefficientSELikelihood ratio χ2PHR (risk reduction)R2 (%)HRR2 (%)Univariate (%)Jointly (%)Jointly (%)§
None−0.439220.0902123.7041<0.00010.645 (35.5%)1.150
Body weight (kg)−0.268960.092578.44140.00370.764 (23.6%)0.4091.1164.73464.429.67.9
Fasting insulin (pmol/l)−0.383590.0909017.8070<0.00010.681 (31.9%)0.8631.0052.24524.90.70.2
Proinsulin (pmol/l)−0.333440.0917513.20760.00030.716 (28.4%)0.6391.0413.67144.48.51.0
IGR (pmol/mmol)−0.431740.0903222.8478<0.00010.649 (35.1%)1.1080.9870.6163.72.30.2
Leisure activity (MET h/week)−0.440760.0902923.8320<0.00010.644 (35.6%)1.1560.9990.0070.00.00.0
All the above−0.198430.093674.48790.03410.820 (18.0%)0.21781.1
Fasting glucose (mmol/l)−0.206070.091725.04750.02470.814 (18.6%)0.2442.47812.99378.817.4
All the above−0.055080.094620.33890.56050.946 (5.4%)0.01698.6
  • *

    * Models adjusted for baseline age, sex, ethnicity, fasting glucose, 2-h glucose, A1C, recreational activity, percent calories from fat, and the baseline value of all time-dependent covariates. R2 values are based on likelihood ratio tests.

  • Covariate effect adjusted for baseline covariates in the combined groups, not adjusted for treatment group.

  • Not adjusted for fasting glucose over time.

  • §

    § Adjusted for fasting glucose over time. For example, if weight was removed from the model, the remaining covariates, including glucose, would explain 98.6–7.9 (90.7%) of the beneficial metformin effect.

  • HR for an increase of 10 units.