Table 2.

Means and changes in insulin, glucose, and HOMA2 in exercisers versus controls

Geometric mean
Baseline4 mo12 mo% change (0–4 mo)% change (0–12 mo)βa (95% CI)
Insulin, μU/mL
 Control7.066.656.79−5.75−3.811.07 (0.96–1.19)
 Exercise6.747.056.764.660.39
Glucose, mmol/L
 Control5.205.175.17−0.39−0.581.01 (0.99–1.02)
 Exercise5.202.535.180.58−0.29
HOMA2
 Control0.800.750.77−5.84−3.971.07 (0.96–1.19)
 Exercise0.760.800.774.760.36

NOTE: Measures were available for all women at baseline (n = 189). Data were missing for 8 women at 4 months, and for 7 women at 12 months.

  • aThe β (95% CI) is the treatment effect ratio representing the overall intervention effect on metabolic variable (adjusted for baseline), including changes from baseline to 4 months and 4 to 12 months. Because metabolic variables were log-transformed for the analysis, the regression coefficient is the anti-logarithm of the original coefficient. Therefore, the anti-logarithm of the coefficient is a ratio that indicates whether the hormone level is, on average, higher in the intervention group compared with controls (>1) or lower (<1); e.g., 1.01 indicates that the metabolic variable is on average 1% higher in the intervention group compared with controls.