REF-996 · Montero & Lundby, 2017 · The Journal of Physiology, Vol. · Paper · PubMed

Refuting the myth of non-response to exercise training: 'non-responders' do respond to higher dose of training

B+moderate evidenceMeasured on non-athletes and recreational peopleWritten before October 2026, being re-checked

The study asked whether non-response of cardiorespiratory fitness (CRF) to endurance training depends on dose, using a between- and within-subject design. Seventy-eight healthy adults were divided into five groups doing one, two, three, four or five 60 min sessions per week of an otherwise identical 6-week endurance training programme.

Non-response was defined as a change in maximal incremental power output (Wmax) within the typical error of measurement (±3.96%). The share of non-responders fell as the weekly dose rose, and those classed as non-responders then completed a second 6-week period with two additional sessions per week, after which non-response was eliminated in all individuals.

Total haemoglobin mass was the strongest independent determinant of the VO2max change.

Participants

Seventy-eight healthy adults in five groups (1-5). Age, sex, baseline fitness and training history are not given in the abstract.

Protocol

Design: between-subject (five dose groups) and within-subject (non-responders repeat with more dose).
Period 1: identical 6-week endurance training programme; groups 1-5 did one, two, three, four and five 60 min sessions per week (the Key Points give this as 60, 120, 180, 240 or 300 min per week).
Definition of non-response: any change in Wmax within the typical error of measurement (±3.96%).
Period 2: participants classified as non-responders completed a successive 6-week period including two additional exercise sessions per week (Key Points: 120 min of additional exercise per week).
Measured before and after each period: VO2max, haematology and muscle biopsies.
Not in the abstract: training intensity, whether groups were randomised, how many people entered period 2.

Key Results

Wmax after period 1: increased (P < 0.05) in groups 2, 3, 4 and 5, but not 1.
Non-responders after period 1: 69%, 40%, 29%, 0% and 0% of individuals in groups 1, 2, 3, 4 and 5.
After period 2: non-response was eliminated in all individuals.
The change in VO2max independently determined the Wmax response (partial correlation coefficient ≥ 0.74, P < 0.001).
Total haemoglobin mass was the strongest independent determinant of VO2max (partial correlation 0.49, P < 0.001).
Authors' conclusion: individual CRF non-response is abolished by increasing the dose of exercise and is primarily a function of haematological adaptations in oxygen-carrying capacity.

Practical Application

"I trained for six weeks and my test did not move — am I a non-responder?" In 78 healthy adults (not described as trained) split into five groups doing one to five 60 min sessions a week for 6 weeks, 69%, 40%, 29%, 0% and 0% showed no gain in maximal power; when those non-responders then trained another 6 weeks with two additional sessions a week, every one of them responded (between-group dose comparison plus an uncontrolled within-person follow-up). In this sample a flat result was a property of the dose, not of the person.
What was measured: maximal incremental power (the responder definition), VO2max, haemoglobin mass, muscle biopsies. What was not: any performance test, any change of intensity, any trained athlete.
The within-person part has no control: everyone who entered it trained for two consecutive 6-week periods, so "more dose" and "more weeks" cannot be separated from this abstract.
Inference, not tested here: for an athlete whose completed block was flat, adding sessions is one exit; changing the stimulus is the other (REF-997, REF-995). Which works better was not compared.
Not from this paper: a single before/after test cannot by itself separate a real non-response from measurement noise (REF-430).

Limitations

The abstract does not state random allocation, group sizes, sex, age, baseline fitness or training intensity.
The second training period is uncontrolled and confounds added weekly dose with added weeks of training.
Non-response is defined on one variable (Wmax) with one threshold (typical error); a different definition would give different percentages.
Healthy adults, doses of 60 to 300 min per week, 6-week periods; nothing on trained athletes or on performance.
An erratum exists and was not read. Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

n=78 across five dose groups; the abstract does not say allocation was random, and the second period has no control arm

Content correction 2026-09-27
2026-10-02: body pointer repointed from the unverified composite REF-432 to REF-995 (Reuter 2023, higher intensity at equal energy expenditure).