REF-1674 · Torres-Pérez et al., 2026 · The Journal of Sports Medicine and Physical Fitness, · Paper · PubMed

Live high-train low in cyclists raised blood values a little: pooling haemoglobin concentration and haemoglobin mass from six cycling studies gave a significant positive effect of 0.82% (confidence limits 0.62% to 1.03%); the six studies also reported gains in red cell mass, reticulocytes, VO2max and time-trial performance, but those with lower PEDro scores found only modest or non-significant changes, no pooled performance estimate is given, and the authors urge caution because influence analyses showed a large impact of individual studies (meta-analysis read from its abstract)

Main caveat: Six studies; the authors' influence and bias analyses showed that individual studies weighed on the pooled result, and they urge caution.
limited evidenceMeasured on trained athletesRead from the abstract only

Torres-Pérez and colleagues reviewed cycling studies that used the live high-train low model, in which athletes live at altitude (or in hypoxia) and train low, with the aim of raising red cell and haemoglobin mass. Two databases gave 34 candidate studies and six met the inclusion criteria.

Across those six, the authors describe improvements in red cell mass, haemoglobin concentration, haemoglobin mass, reticulocyte count, VO2max and time-trial performance, while noting that the studies with lower PEDro quality scores reported non-significant or modest gains.

The meta-analysis of haemoglobin concentration and haemoglobin mass found a significant positive effect of 0.82% (confidence limits 0.62% to 1.03%). Influence and bias analyses showed that individual studies weighed heavily on the result, and the authors conclude that LHTL had a positive haematological effect in cycling while calling for more studies.

Participants

Cyclists in six studies of LHTL. The number of participants, sex split, age, training level and competition level are not in the abstract.

Protocol

Search: PubMed/MEDLINE and SPORTDiscus; search dates not in the abstract; 34 studies identified, 6 included.
Inclusion: cycling studies applying LHTL that reported haematological or performance variables; whether a sea-level control group or randomisation was required is not in the abstract.
Quality: PEDro scale; the scores are not given, only that lower-scoring studies reported smaller or non-significant gains.
Analysis: meta-analysis of haemoglobin concentration and haemoglobin mass; influence and bias analyses. The abstract does not say whether the two variables were pooled into one estimate or analysed separately, or how the percentage was computed.
Not in the abstract: altitude or simulated altitude, hours per day in hypoxia, block length, the control condition, the performance tests, and any pooled performance estimate.

Key Results

OutcomeResult
Haemoglobin concentration and haemoglobin mass (meta-analysis)+0.82% (confidence limits 0.62% to 1.03%), significant
Red cell mass, reticulocyte countimprovements reported across studies; no pooled value in the abstract
VO2max, time-trial performanceimprovements reported across studies; no pooled value in the abstract
Studies with lower PEDro scoresnon-significant or modest gains
The influence and bias analyses "underscored individual study impacts"; their numbers are not in the abstract.

Practical Application

An LHTL block in cyclists moved the blood by under 1% on the pooled estimate (0.82%, limits 0.62-1.03%), and this review gives no pooled number for how much faster it makes a rider. It sits with SYNTH-06's "expect a few per cent at best": the blood gain here is small, and the performance gains are described study by study, with the weaker studies finding little.
Do not read the haemoglobin figure as a performance figure. A blood value is a step removed from race speed; the pooled sea-level performance estimates in the base come from other meta-analyses (REF-210, natural LHTL in subelite and elite athletes, with wide limits and placebo effects; REF-1677, VO2max in athletes).
Inference, not tested here: because the result leaned on single studies, one cyclist's response may sit anywhere from nothing to more than the average; SYNTH-06's iron check before a camp and its altitude dose are not tested by this review, which gives no dose in its abstract.

Limitations

Six studies; the authors' influence and bias analyses showed that individual studies weighed on the pooled result, and they urge caution.
The only pooled number is haematological; performance results are described, not pooled, in the abstract.
Lower-quality studies found smaller effects, which raises the question of how much of the reported benefit depends on study quality.
Written from the abstract: the PEDro scores, sample sizes, altitude dose, control conditions, whether the studies were randomised, heterogeneity and the exact basis of the 0.82% figure are unknown.
How this entry was checked and changed
Why this level

risk of bias not read (abstract only): the abstract names the PEDro scale but gives no scores, so the cap is B+, and the only pooled result is haematological (haemoglobin concentration and mass), a blood marker, which takes one step (rule 3); no pooled performance result is reported