In a bivariate linear regression model, change in total FFMQ score significantly predicted change in IDS (n = 82, r = 0.317, p =0.004; r = 0.286, p = 0.013 using transformed data; ). A stepwise multiple regression was then performed. The relationship between change in FFMQ and change in IDS remained significant after controlling for age (n =78, B = 0.054 with 95% CI of 0.018 to 0.090, p=0.004; B=0.009 with 95% CI of 0.002 to 0.016, p = 0.018 using transformed data), baseline BMI (n=70, B = 0.048 with 95% CI of 0.013 to 0.085, p = 0.009; B=0.009 with 95% CI of 0.002 to 0.016, p = 0.017 using normalized data), and baseline IDS score (n= 70, B= 0.037 with 95% CI of 0.007 to 0.068, p = 0.017; B=0.004 with 95% CI of 0.001 to 0.013, p =0.033 using transformed data), but significance was lost when change in PSS was added to the model (n=70, B= 0.026 with 95% CI of −0.009 to 0.062, p =0.147; B=0.005 with 95% CI of −0.002 to 0.012, p= 0.183). See .