Measuring the Effects of Caffeine and L-Theanine on Cognitive Performance: A Protocol for Self-Directed, Mobile N-of-1 Studies

Background: The growing consumer digital tools market has made using individual health data to inform lifestyle changes more accessible than ever. The n-of-1 trial -- a single participant, multiple crossover, comparative effectiveness trial -- offers methodological tools that link interventions directly with personalized outcomes to determine the best treatment for an individual. We have developed a complete digital platform to support self-directed n-of-1 trials, comprised of virtual study on-boarding, visual informed consent, device integrations, in-app assessments, and automated data analysis. Objective: To evaluate the n-of-1 platform, a pilot study was launched to investigate the effects of commonly consumed substances on cognition. The purpose of the study is to allow an individual to measure the effect of 2 treatments (caffeine alone vs. caffeine + L-theanine) on 3 measures of cognitive performance: creative thinking, processing speed, and visual attention. Upon completion of the study, individuals receive personalized results that compare the impact of the two treatments on each of the cognitive performance measures. Methods: After the onboarding process, participants are randomized to a study length (5, 15, or 27 days), starting treatment (caffeine or caffeine + L-theanine), and app notification frequency (light, moderate). Each trial begins with a baseline period, during which participants abstain from either treatment, followed by 2 randomized counterbalanced treatment sequences (either ABBA or BAAB). Throughout the trial, daily tests assess participant cognitive performance. These tests are digital versions of the Remote Associates Test, Stroop Test, and Trail Making Test, and are implemented directly in the n-of-1 mobile application (“N1”). Assessments are completed at a fixed time, defined by the individual during study setup. Treatments are taken daily within a fixed time window prior to the user-defined assessment time. Cognitive assessment results are analyzed using a linear model with factors for treatment and block, and each treatment is compared to baseline. Results: We launched our N1 app on the Apple App Store in mid-October 2019 and recruited over 40 participants within the first month. Conclusion: This platform provides individuals the opportunity to investigate their response to treatments through n-of-1 methods, empowering them to make data-driven, personalized lifestyle choices.