A Randomised, Placebo-Controlled, Crossover Study Investigating the Optimal Timing of a Caffeine-Containing Supplement for Exercise Performance

Abstract

Background: Pre-exercise supplements containing low doses of caffeine improve endurance exercise performance, but the most efficacious time for consumption before intense endurance exercise remains unclear, as does the contribution of caffeine metabolism.

Methods: This study assessed the timing of a commercially available supplement containing 200 mg of caffeine, 1600 mg of β-alanine and 1000 mg of quercetin [Beachbody Performance Energize, Beachbody LLC, USA] on exercise performance, perception of effort and plasma caffeine metabolites. Thirteen cyclists (V̇O2max 64.5 ± 1.4 ml kg- 1 min- 1 (± SEM)) completed four experimental visits consisting of 30 min of steady-state exercise on a cycle ergometer at 83 ± 1% V̇O2max followed by a 15-min time trial, with perceived exertion measured regularly. On three of the visits, participants consumed caffeine either 35 min before steady-state exercise (PRE), at the onset of steady-state (ONS) or immediately before the time trial (DUR) phases, with a placebo consumed at the other two time points (i.e. three drinks per visit). The other visit (PLA) consisted of consuming the placebo supplement at all three time points. The placebo was taste-, colour- and calorie-matched.

Results: Total work performed during the time trial in PRE was 5% greater than PLA (3.53 ± 0.14 vs. 3.36 ± 0.13 kJ kg- 1 body mass; P = 0.0025), but not ONS (3.44 ± 0.13 kJ kg- 1; P = 0.3619) or DUR (3.39 ± 0.13 kJ kg- 1; P = 0.925), which were similar to PLA. Perceived exertion was lowest during steady-state exercise in the PRE condition (P < 0.05), which coincided with elevated plasma paraxanthine in PRE only (P < 0.05).

Conclusion: In summary, ingestion of a pre-exercise supplement containing 200 mg caffeine 35 min before exercise appeared optimal for improved performance in a subsequent fatiguing time trial, possibly by reducing the perception of effort. Whether this was due to increased circulating paraxanthine requires further investigation.

Trial registration: ClinicalTrials.Gov, NCT02985606 ; 10/26/2016.

Keywords: Ergogenic aid; Metabolism; Paraxanthine; Perceived exertion; Performance; Sports nutrition.

Conflict of interest statement

Francis B Stephens has received payments as a member of the Beachbody LLC scientific advisory board. Andrew D Davenport, Tom S O Jameson, Sean P Kilroe, Alistair J Monteyne, George F Pavis, Benjamin T Wall, Marlou L Dirks, Nima Alamdari and Catherine R Mikus have no competing interests to declare.

Figures

Fig. 1

Overview of the experimental protocol. CAFF, a commercially available beverage containing 200 mg of caffeine; PLA, placebo drink; HR, heart rate; RPE, rating of perceived exertion; VAS, visual analogue scales for alertness and paraesthesia. Treatments were randomly assigned in a crossover design

Fig. 2

a Mean work done per kilogramme of body mass during a 15-min cycling time trial in the PLA, DUR, ONS and PRE trials. Data are mean for all participants ± SEM. b Individuals participants’ work done per kilogramme of body mass during a 15-min cycling time trial in the PLA, DUR, ONS and PRE trials. *P < 0.01, significantly different from PLA. Mean work done per kilogramme of body mass during a 15-min cycling time trial in the PLA, DUR, ONS and PRE trials for c AA and d AC genotypes

Fig. 3

Calculated carbohydrate (CHO) and lipid (Lipid) oxidation during 30 min of SS cycling at 83 ± 1% VO2max in the PLA, DUR, ONS and PRE trials. Expired gases were measured for two periods of 5 min (SS1 and SS2)

Fig. 4

Plasma caffeine (a), paraxanthine (b), theobromine (c), theophylline (d), glucose (e) and lactate (f) during the PLA, DUR, ONS and PRE trials. Data are mean for all participants ± SEM. Main effect for time: *, significant difference from − 70; #, significant difference from − 35; †, significant difference from − 20. Main effect for condition: (a) PRE significantly different to ONS, PRE and PLA; (b) ONS significantly different to DUR and PLA; (c) DUR significantly different to PLA and (d) DUR significantly different to DUR and PLA

Fig. 5

a RPE and b heart rate during 30 min of SS cycling at 80% VO2max and a 15-min time trial in the PLA, DUR, ONS and PRE conditions. Data are mean for all participants ± SEM. Main effect for time: *, significant difference from − 35; #, significant difference from 0. Main effect for condition: (a) significant difference between PRE and ONS, (b) significant difference between PRE and DUR, (c) significant difference between PRE and PLA, (d) significant difference between ONS and DUR, (e) significant difference between ONS and PLA and (f) significant difference between DUR and PLA

Fig. 6

Alertness (a) and paraesthesia (b) during the PLA, DUR, ONS and PRE trials. Data are mean for all participants ± SEM. *, significant difference from − 25; #, significant difference from − 35