FIGURE 1.

Body weight, liver measures and metabolic ratios. Changes in (A) total body weight^a, (B) NAFLD liver fat score^a, (C) liver fat content^a, (D) paraxanthine/caffeine ratio^a,b, (E) 5‐OH‐omeprazole/omeprazole ratio^a,b,c and (F) losartan/LCA ratio^a,b in the RYGB and diet groups. Linear mixed‐effects models, with log‐transformed dependent variables, were used to estimate mean change over time compared with baseline. Data are presented as mean (95% CI). Boxplots of the (G) paraxanthine/caffeine ratio, (H) 5‐OH‐omeprazole/omeprazole ratio and (I) losartan/LCA ratio^d in patients with severe obesity (n = 81) and controls (n = 18) at baseline. Student's t‐test on log‐transformed data was used to compare patients with obesity and controls at week 0. ^aStatistically significant differences (P < .05) within group (RYGB or diet) compared with baseline are symbolized by *. ^bPredicted values from the mixed‐effect model. ^cPatients with genotype CYP2C19 *2/*2 or *2/*4 were excluded from the plot. ^dDue to visualization purposes, one losartan/LCA ratio in the control group (49.05) was excluded from plot i. CYP, cytochrome P450; LCA, losartan carboxylic acid; NAFLD, nonalcoholic fatty liver disease; RYGB, Roux‐en‐Y gastric bypass