LAUSR

BACKGROUND Caffeine is a widely consumed stimulant, and caffeine-containing products are increasingly available on the market. Few tools are available to capture caffeine intake, particularly among young adults. To estimate caffeine consumption in the previous 24 hours, the 24-Hour Caffeine Intake Recall (CIR-24) was modeled after the Automated Self-Administered 24-Hour Dietary Assessment Tool, using a brand-specific database of caffeine-containing foods, beverages, and supplements. OBJECTIVE To evaluate the accuracy of the CIR-24 compared with caffeine concentration biomarkers in urine and a caffeinated beverage intake frequency screener (CBQ) designed to assess usual intake among a young adult population in Canada. DESIGN/PARTICIPANTS In all, 79 young adults, aged 18 to 29 years, provided 24-hour urine samples and completed the CIR-24 and CBQ. MAIN OUTCOME MEASURES Excretion for caffeine and eight caffeine metabolites were quantified from urine samples using high-performance liquid chromatography-polarity switching electrospray ionization-tandem quadrupole mass spectrometry with stable isotope-labeled internal standards. STATISTICAL ANALYSES PERFORMED Pearson correlations and weighted κ coefficients were calculated for the self-report tools and caffeine biomarkers. RESULTS The CIR-24 was significantly positively associated with all caffeine biomarkers (rp=0.28 to 0.52, κ=0.39 to 0.59), and the CBQ was significantly positively associated with all but one biomarker (rp=0.21 to 0.40, κ=0.32 to 0.45). The CIR-24 yielded a higher mean intake of caffeine than the CBQ. There was strong linear correlation between the CIR-24 and CBQ (rp=0.60, P<0.001), but poor agreement in absolute caffeine consumed (t=2.83, P=0.006); quartile ranking concordance was 0.44 (P<0.001). The CIR-24 performed better than the CBQ across all biomarkers in both linear correlation and quartile ranking. CONCLUSIONS Although both the CIR-24 and CBQ performed reasonably well in capturing caffeine intake compared with urinary biomarkers of caffeine consumption, the CIR-24 had stronger agreement than the CBQ. The results suggest that the CIR-24 is a promising tool for evaluating caffeine intake among this population.