In one trial of 26 adults, a drink carrying 200mg of theanine and three minor ingredients cut rated stress at one hour, not at three; cortisol fell at three hours, not at one.
Thirty-four people finished the trial, but the treatment effects were computed on 26 — the ones who had actually reacted to the task at baseline, picked out after the fact. An industry grant funded the study, and the paper says the funder had no hand in the analysis. The other ingredients were phosphatidylserine, alpha-glycerylphosphorylcholine and chamomile, each in a small dose. Why the two effects arrived two hours apart is unexplained. Performance on the task itself never moved.
