Relation of 24-hour urinary caffeine and caffeine metabolite excretions with self-reported consumption of coffee and other caffeinated beverages in the general population

Dusan Petrovic, Sandrine Estoppey Younes, Menno Pruijm, Belén Ponte, Daniel Ackermann, Georg Ehret, Nicolas Ansermot, Markus Mohaupt, Fred Paccaud, Bruno Vogt, Antoinette Pechère-Bertschi, Pierre-Yves Martin, Michel Burnier, Chin B. Eap, Murielle Bochud, Idris Guessous

Caffeine intake is generally estimated by self-reported consumption, but it remains unclear how well self-report associates with metabolite urinary excretion. We investigated the associations of self-reported consumption of caffeinated drinks with urinary excretion of caffeine and its major metabolites in an adult population. We used data from the population-based Swiss Kidney Project on Genes in Hypertension (SKIPOGH) study. Consumption of caffeinated coffee, decaffeinated coffee and other caffeinated beverages was assessed by self-administered questionnaire. Quantification of caffeine, paraxanthine, theobromine and theophylline was performed by ultra-high performance liquid chromatography tandem mass spectrometry in 24-h urine. Association of reported consumption of caffeinated drinks with urinary caffeine derived metabolites was determined by quantile regression. We then explored the association between urinary metabolite excretion and dichotomized weekly consumption frequency of caffeinated coffee, with Receiver Operator Characteristic (ROC) analysis. In the present analysis, we included 598 individuals (52% women, mean age =46 ± 17 years). Self-reported caffeinated coffee intake was positively associated with 24-h urinary excretions of paraxanthine, theophylline and caffeine (p < 0.001), whereas reported intakes of decaffeinated coffee and other caffeinated beverages showed no association. In ROC analysis, optimal discrimination between individuals consuming less than one caffeinated coffee/week, vs. at least one coffee, was obtained for 24-h urinary paraxanthine (Area Under Curve (AUC) = 0.868, 95% Confidence Interval (CI) [0.830;0.906]), with slightly lower performance for theophylline and caffeine, whereas theobromine did not allow any discrimination. Our results suggest that reported consumption of caffeinated coffee is positively associated with 24-h urinary excretion of caffeine, paraxanthine, and theophylline, and may be used as a marker of caffeine intake for epidemiological studies.