How long until I recover from a hard mixed workout?
Small CrossFit studies: this review says 48-72 h both "needed" and "insufficient"; two primaries it cites report baseline by 48 h. None measured a HYROX simulation.
Martinho and colleagues searched three databases to 23 February 2024 for adult CrossFit participants with reported training experience, and included 68 papers. More than half (38 papers) characterised participants, and five focused on recovery. Mean heart rate was 179 ± 8 beats/min in Fran, 182 ± 7 in Cindy and 169 ± 6 in Murph.
Oxygen uptake and lactate differed widely by workout and group, so "the demands of CrossFit" depend on the workout.
In the meta-correlations, CrossFit-type strength tests went moderately with as-many-rounds workouts when scaled to body mass (r = 0.38, 0.03 to 0.65), not clearly with for-time workouts (r = -0.18, -0.36 to 0.01), and with Grace (r = 0.478, 0.177 to 0.697); in Donkey Kong, the "non-specific" tests were the significant ones.
The authors find no unique predictor and favour specific tests. On recovery, the abstract's line is "48 to 72 h are needed".
We traced it (our inference; the review does not say which studies it means): the review's text gives 48 h for physical performance and 72 h for physiological markers after two workouts in 12 people, while that study's own abstract says all values were back by 48 h, its last time point; and 72 h is the last time point after Karen in 8 men, at which the review says creatine kinase was still up, while that study's own abstract says it was back at baseline by 48 h.
The discussion then says 48-72 h "were insufficient". Our reading: usable as a map of what has been measured, and as evidence that recovery differs by marker and by workout. It is not a measured recovery time for a hard mixed session, and not a basis for a simulation schedule.
Protocol
Key Results
What the recovery line rests on, as the review reports it (primary papers not read unless stated):
| Workout | Study, group | Marker, last time point | As reported here |
|---|---|---|---|
| Fran | Rios [20], N = 14 males, N = 6 females | jump, force, plank at 24 h | below baseline after 24 h |
| Karen | Sousa-Neto [43], N = 8 males | jump, creatine kinase, perceived recovery at 72 h | jump back at 72 h; creatine kinase 151.1 ± 68.3 at baseline, 223.0 ± 86.3 at 72 h. Table 1 says up at 24 h, perceived recovery down at 24 h, then rising |
| As many rounds, rounds for time | Timón [68], N = 12 | performance and biochemical markers | text: performance at baseline by 48 h, physiological markers by 72 h. Table 1: measured at 24 and 48 h, "After 48 h, values returned to baseline" |
| Open 18.4, Cindy | Forte [41], N = 23 | heart rate, lactate at 30 min | above baseline at 30 min |
| Isabel | Rios [21], N = 14 males | heart rate at 5 min | 177 to 128 bpm |
Meta-correlations (performance vs tests; the printed signs follow the paper's "positive indicates positive performance" note only in part):
| Workouts | Tests | r | 95% CI | I² or p |
|---|---|---|---|---|
| As many rounds | specific | 0.33 | -0.09 to 0.65 | I² 77.1% |
| As many rounds | specific, per kg body mass | 0.38 | 0.03 to 0.65 | p = 0.03, I² 67% |
| For time | specific | -0.18 | -0.36 to 0.01 | p = 0.07 |
| Fran | specific | -0.44 | printed 0.22 to 0.54 | p < 0.05, I² 75% |
| Fran | non-specific | 0.24 | not printed | p = 0.19 |
| Grace | specific | 0.478 | 0.177 to 0.697 | p = 0.003, I² 62% |
| Cindy (one study) | combined | 0.14 | -0.15 to 0.41 | p = 0.33 |
| Donkey Kong (one study) | specific | -0.26 | printed -0.06 to 0.53 | p = 0.11; non-specific significant |
Mechanisms
Practical Application
Limitations
How this entry was checked and changed
correlational and descriptive evidence is B by design; held a step lower because no risk-of-bias assessment is reported, the headline recovery statement contradicts the paper's own discussion and reads two primary studies differently from those studies' abstracts (Timón, de Sousa Neto), the flow numbers do not add up, and several printed intervals do not match their estimates