TABLE 3.

Between‐group differences in within‐group changes in metabolic ratios

Metabolic ratio	RYGB versus diet
	W0‐Y2	W3‐W9	W3‐Y2
Paraxanthine/caffeine (CYP1A2)	−0.02 (−0.08, 0.03) n = 61	0.04 (−0.02, 0.09) n = 75	−0.03 (−0.09, 0.03) n = 61
5‐OH‐omeprazole/omeprazole (CYP2C19) a	−0.49 (−0.81, −0.17) n = 58	−0.30 (−0.63, 0.03) n = 71	−0.43 (−0.74, −0.12) n = 57
Losartan/LCA (CYP2C9) b	0.47 (0.11, 0.82) n = 62	0.10 (−0.20, 0.41) n = 75	0.13 (−0.18, 0.44) n = 62

Note: Data are presented as model estimated mean difference in change (95% CI). Bold values show statistically significant differences (P < .05). Linear mixed‐effects models, with a log‐transformed dependent variable, were used to estimate between‐group differences in within‐group changes, with diet as reference group. Not all included patients were able to supply metabolic ratios at all four study visits (RYGB = 4, diet = 1) due to technical difficulties.

Abbreviations: CYP, cytochrome P450; LCA, losartan carboxylic acid; W, week; Y, year.

^a Patients with genotype CYP2C19 *2/*2 or *2/*4 were excluded from the analysis.

^b One losartan/LCA ratio was excluded from week 0 due to analytical technicalities (diet group).