Caffeine’s Knee-Strength Boost May Change With Cycle Phase

Researchers wanted to know if caffeine gives the same strength boost during every part of the menstrual cycle. Caffeine is often used as a neuromuscular performance aid. The focus was maximal strength in resistance-trained women, and the question was whether the follicular or luteal phase changed the response.

Sixteen women took part. Their average age was 26.4 years, with a standard deviation of 3.9 years. The design was randomized, double-blind, and placebo-controlled, with a crossover setup. During both cycle phases, each woman completed trials with 6 mg of caffeine per kg of body weight and trials with a placebo, all under standardized conditions.

The tests measured acute strength changes. Peak torque was checked during isokinetic knee movement at 60 degrees per second and 180 degrees per second. The study also tracked average rate of force development, time to peak torque, and maximal voluntary isometric contraction. Linear mixed-effects models were used to analyze the data.

The clearest gains appeared in the knee extensors. Caffeine raised absolute and relative peak torque at 60 degrees per second, with p = .009 for absolute torque and p = .005 for relative torque. At 180 degrees per second, both measures improved too, and both p values were below .001. The partial eta-squared values were .146, .166, .293, and .310 in that order. The pattern suggested that caffeine may help more during the follicular phase, especially for knee extensor peak torque.

Knee flexors told a more mixed story. The phase-by-condition interaction was significant for knee extensor absolute and relative peak torque at both speeds. At 60 degrees per second, p = .016, with partial eta-squared values of .125 and .126. At 180 degrees per second, p = .025 for absolute peak torque and p = .022 for relative peak torque, with values of .109 and .114. For knee flexors, the interaction was not significant for absolute peak torque at 180 degrees per second, p = .069, but it was significant for relative peak torque, p = .050. Other neuromuscular outcomes did not show significant interactions, and all p values stayed above .05.