Determination of CYP1A2 Activity in the US Population by Use of Caffeine Metabolite Ratios in Spot Urine Samples: NHANES 2009–2010

Abstract

The cytochrome P450 enzyme CYP1A2 is a majorphase I enzyme in the liver, accounting for nearly 15 % of P450 content. It isintegral in the metabolism of numerous drugs, the elimination of environmentaltoxins, and the activation of procarcinogens such as dietary aryl‐andheterocyclic amines, mycotoxins, and tobacco‐specific nitrosamines. Genetic, lifestyle, dietary and environmental factors can affect CYP1A2 activity. CYP1A2is also the most important enzyme in caffeine metabolism, catalyzing thedemethylation of caffeine and its subsequent metabolites. A metabolite ratio (MR) of these compounds in the urine, resulting from either a test dose or regular dietary intake of caffeine, can be used to determine CYP1A2 activity. In this study we described CYP1A2 activity in the US population ≥6 y by selected demographic, lifestyle and physiologic variables. We used spot urine concentrations of 5‐acetylamino‐6‐amino‐ 3‐methyluracil(AAMU), 1‐methyluric acid (1U), 1‐methylxanthine (1X), and 1,7‐ dimethyluricacid (17U) from the cross‐sectional NHANES 2009–2010 to calculate CYP1A2 activity by use of the MR [(AAMU+1U+1X)/17U]. We excluded individuals taking prescription medications known to interfere with caffeine metabolism and individuals with evidence of insufficient caffeine exposure to reliably calculate the CYP1A2 MR (n = 2197). We used Spearman correlation to examine the relation of CYP1A2 MR with continuous variables, and performed bivariate comparisons of unadjusted geometric mean (GM) CYP1A2 MRs for categorical variables. CYP1A2 activity was higher in males vs. females (Wald F P <0.0001) and highest in non‐Hispanic blacks, followed by non‐Hispanic whites, and lowest in Hispanics (Wald F P = 0.0002). An inverse relation was observed for CYP1A2 MR with age (Spearman ρ = −0.32, P <0.001). CYP1A2 activity was lower in individuals with evidence of inflammation (serum CRP ≥5 mg/L vs. <5 mg/L; Wald F P <0.0001) and in diabetics(yes vs. no based on a physician's diagnosis; Wald F P = 0.0221). GM CYP1A2 MR was lower in non‐smokers vs. smokers(i.e. serum cotinine ≤10 ng/mL vs. >10 ng/mL) (Wald F P <0.0001), and no association was observed with alcohol consumption (limited to individuals ≥20 y). Our associations of CYP1A2 activity with age, sex, race‐ethnicity, smoking status, and inflammation were generally consistent with observations in other studies. We believe our study is the first report of using urine caffeine MRs in spot urine samples to estimate CYP1A2 activity in a representative sample of the US population.