Monitoring Training Load to Understand Fatigue in Athletes
Key Findings
No single, definitive marker of fatigue exists in the literature; very few markers have strong scientific evidence supporting their use
Appropriate load monitoring helps determine whether an athlete is adapting and helps minimise the risk of non-functional overreaching, illness and injury
External load tools: power-output devices, time-motion analysis. Internal load measures: perception of effort, heart rate, blood lactate, training impulse (TRIMP)
Dissociation between external and internal load may reveal the athlete's state of fatigue — the same power at a higher perceived effort or heart rate, or lower power at the same internal load
Other tools used in high-performance programmes: heart-rate recovery, neuromuscular function, biochemical / hormonal / immunological assessments, questionnaires and diaries, psychomotor speed, sleep quality and quantity
Individualisation of load monitoring "cannot be over emphasized"; the approach may differ between individual and team sports
Detecting meaningful changes requires scientific and statistical approaches; monitoring systems should be intuitive, analyse data efficiently and return simple, scientifically valid feedback
Full text, checked 2026-09-27: questionnaires and diaries are a relatively simple and inexpensive tool, though subjective and open to manipulation; salivary cortisol and testosterone relate to overreaching to some degree, but their routine use has not been examined and hormonal markers are not currently justified
Practical Application
Halson's review finds no single definitive marker of fatigue and little strong evidence for most markers; what carries information is the dissociation between external load (power, pace) and internal load (RPE, heart rate, lactate, TRIMP) in an individually monitored athlete — read fatigue from the gap between what the athlete produced and what it cost, tracked against their own baseline, not from any one number.
Monitor both sides: an external measure (power / pace) and an internal one (session RPE, heart rate) every session, so the dissociation is visible.
Keep the system simple enough to be used daily and reported clearly — the review's own criterion for a monitoring system.
Not from this paper — the base's own heuristic, not Halson's: a minimum stack of morning HRV + session RPE + a short weekly wellness questionnaire (sleep, soreness, energy, mood, motivation); an extended stack adding morning HR, body mass and sleep duration; a research stack adding periodic CK, CRP and ferritin; and a trigger to investigate when 2+ metrics sit outside 1 SD of their rolling mean for 3+ days.These are the base's operational defaults built on the review's principle of individual baselines; the paper prescribes no stack and no threshold.
Limitations
Narrative review, 2014 — no systematic method, and wearable / HRV practice has moved on since
Monitoring overload possible — too many metrics, no action
Subjective self-report prone to athlete social-desirability bias (under-reporting fatigue)
No single metric predictive; requires context and trend analysis
How this entry was checked and changed
Content correction 2026-09-27
Monitoring stacks (minimum / extended / research) and the "2+ metrics outside 1 SD for 3+ days" trigger -> labelled as the base's own heuristic under "Not from this paper"; Halson prescribes neither.
Key points rewritten to the abstract: no single definitive marker, external-internal load dissociation as the fatigue signal, individualisation; wellness-questionnaire and T:C-ratio details marked full text.
Lines labelled "full text; not checked" were read against the open-access full text (PMC4213373): 1 corrected to what the paper says.
Resync 2026-09-28: Labelled REF-368 'biomarker use' -> subjective vs objective measures.