Objective It is purported that exercise-induced dehydration (EID), especially if ≥ 2% bodyweight, impairs endurance performance (EP). Field research shows that athletes can achieve outstanding EP while dehydrated > 2% bodyweight. Using the meta-analytic procedure, this study compared the findings of laboratory-based studies that examined the impact of EID upon EP using either ecologically valid (EV) (time-trial exercise) or non-ecologically valid (NEV) (clamped-intensity exercise) exercise protocols.

Methods EP outcomes were put on the same scale and represent % changes in power output between euhydrated and dehydrated exercise tests. Random-effects model meta-regressions and weighted mean effect summaries, mixed-effects model analogue to the ANOVAs and magnitude-based effect statistics were used to delineate treatment effects.

Main results Fifteen research articles were included, producing 28 effect estimates, representing 122 subjects. Compared with euhydration, EID increased (0.09±2.60%, (p=0.9)) EP under time-trial exercise conditions, whereas it reduced it (1.91±1.53%, (p<0.05)) with NEV exercise protocols. Only with NEV exercise protocols did EID ≥ 2% body weight impair EP (p=0.03).

Conclusions Evidence indicates that (1) EID ≤ 4% bodyweight is very unlikely to impair EP under real-world exercise conditions (time-trial type exercise) and; (2) under situations of fixed-exercise intensity, which may have some relevance for military and occupational settings, EID ≥ 2% bodyweight is associated with a reduction in endurance capacity. The 2% bodyweight loss rule has been established from findings of studies using NEV exercise protocols and does not apply to out-of-doors exercise conditions. Athletes are therefore encouraged to drink according to thirst during exercise.