Table 5: Regression Models predicting variability in log HOMA-IR and log ISI(0,120)

From: Enhanced insulin sensitivity in successful, long-term weight loss maintainers compared with matched controls with no weight loss history

PredictorsModel 1: predicting log HOMA-IRModel 2: predicting log ISI(0,120)
 Coeff.95% CIP-valueCoeff.95% CIP-value
  LowerUpper  LowerUpper 
% BW lost−0.038−0.058−0.018<0.0010.0250.0150.036<0.001
% BW regained0.0290.0020.0560.037−0.031−0.045−0.016<0.001
Light activity(min per day)−0.003−0.006−0.0010.048    
Vigorous activity (min per day)−0.033−0.061−0.0050.0230.0150.0010.0290.043
% BF0.019−0.0020.0400.073    
RQ ratio4.2231.3267.1200.005−1.823−3.372−0.2750.022
WHR    −1.273−2.8140.2680.103
Observations:50   50   
R2 (adjusted R2)0.614 (0.560)   0.597 (0.551)   
P-value<0.001   <0.001   
  1. Abbreviations: %BF, % body fat; %BW, % body weight; CI, confidence interval; Coeff., coefficient; HOMA-IR, Homeostatic Model Assessment; ISI, insulin sensitivity index; RQ, respiratory quotient; WHR, waist-to-hip ratio.