Table 4.
Baseline Model (n = 42) | Controlling for age | Controlling for age, BMI | Controlling for age, BMI, HbA1c | Controlling for age, BMI, HbA1c, HOMA‐IR | Controlling for age, BMI, HbA1c, HOMA‐IR, HOMA‐β | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
P‐Value | Effect size (%)a |
Adjusted 3‐mo meanb (95% CI) |
P‐value | Effect size (%)a | P‐value | Effect size (%)a | P‐value | Effect size (%)a | P‐value | Effect size (%)a | P‐value | Effect size (%)a |
Adjusted 3‐mo meanb (95% CI) |
|||||
Met | Vogli | Placebo | ||||||||||||||||
Met | Vogli | Placebo | ||||||||||||||||
PINK1 | .002 | 28.5 | 2.12 (1.63, 2.60) | 1.06 (0.59, 1.53) | 0.89 (0.39, 1.40) | .002 | 28.8 | .002 | 28.9 | .002 | 29.1 | .006 | 28.1 | .001 | 38.9 | 2.32 (1.76, 2.88) | 1.22 (0.67, 1.77) | 0.67 (0.09, 1.25) |
PARKIN | .004 | 24.8 | 2.26 (1.62, 2.91) | 1.05 (0.42, 1.67) | 0.75 (0.08, 1.41) | .002 | 28.0 | .004 | 26.9 | .003 | 28.1 | .001 | 34.3 | .004 | 33.8 | 2.57 (1.83, 3.30) | 1.13 (0.41,1.84) | 0.71 (−0.06,1.48) |
NIX | .161 | 9.2 | 1.60 (1.11, 2.09) | 1.23 (0.75, 1.70) | 0.92 (0.41, 1.42) | .145 | 9.9 | .187 | 8.9 | .186 | 9.2 | .421 | 5.4 | .304 | 8.4 | 1.58 (1.00, 2.15) | 1.23 (0.68, 1.79) | 0.91 (0.32,1.50) |
LC3‐II | <.001 | 33.4 | 2.24 (1.71, 2.77) | 0.75 (0.23, 1.26) | 0.96 (0.41, 1.52) | .001 | 32.4 | .001 | 33.1 | .001 | 32.9 | <.001 | 42.7 | .001 | 38.8 | 1.95 (1.51, 2.39) | 0.78 (0.34, 1.21) | 0.95 (0.48, 1.41) |
LAMP2 | <.001 | 45.4 | 3.62 (2.87, 4.36) | 1.26 (0.55, 1.98) | 0.93 (0.16, 1.70) | <.001 | 44.7 | <.001 | 44.6 | <.001 | 44.5 | <.001 | 44.8 | <.001 | 45.5 | 3.72 (2.82, 4.62) | 1.37 (0.50, 2.25) | 0.87 (−0.07, 1.80) |
MFN2 | <.001 | 45.9 | 2.61 (2.09,3.12) | 0.99 (0.49,1.48) | 0.71 (0.18,1.25) | <.001 | 49.3 | <.001 | 48.6 | <.001 | 48.6 | <.001 | 52.7 | <.001 | 52.2 | 2.87 (2.26, 3.49) | 0.95 (0.35, 1.56) | 0.71 (0.07, 1.36) |
Variance explained by the use of metformin; calculated from partial eta‐squared. Squared values of 1%, 6% and 13.8% indicate small, medium and large effect sizes, respectively.
Adjusted 3‐mo mean for outcome variables calculated controlling for effects of confounder(s).