REF-997 · Bonafiglia et al., 2016 · PLOS ONE, Vol. · Paper · PubMed

Inter-Individual Variability in the Adaptive Responses to Endurance and Sprint Interval Training: A Randomized Crossover Study

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

Twenty-one recreationally active adults each completed three weeks of endurance training (END) and three weeks of sprint interval training (SIT) in a randomised crossover design with a three-month washout. Training produced a main effect on VO2peak, lactate threshold and submaximal heart rate, but individual responses varied considerably after both modes.

Individual changes after END were not positively related to individual changes after SIT for any variable. Non-responders, defined by two times the typical error of measurement, were seen for each variable after each mode, but the pattern differed between modes, and every individual responded in at least one variable when exposed to both.

Participants

Twenty-one recreationally active adults. Age, sex and baseline VO2peak are not given in the abstract.

Protocol

Design: randomised crossover, three-month washout period between training interventions.
END: three weeks, four days/week, 30 minutes at ~65% VO2peak work rate.
SIT: three weeks, four days/week, eight 20-second intervals at ~170% VO2peak work rate separated by 10 seconds of active rest.
Outcomes: VO2peak, lactate threshold, submaximal heart rate.
Non-response threshold: two times the typical error of measurement — VO2peak 0.107 L/min, lactate threshold 15.7 W, submaximal HR 10.7 bpm.

Key Results

Main effect of training on VO2peak, lactate threshold and submaximal HR.
Considerable variability in individual responses to both END and SIT.
No significant positive relationships between END and SIT for individual changes in any variable.
Non-responders in VO2peak, lactate threshold and submaximal HR were observed after both END and SIT, with different individual patterns after each.
All individuals responded in at least one variable when exposed to both END and SIT.
The abstract gives no group mean changes and no counts of non-responders.

Practical Application

"This block did nothing for me — will a different kind of training?" In 21 recreationally active adults who each did three weeks of endurance training and three weeks of sprint intervals (randomised crossover, within-person comparison), a person's response to one mode was not positively related to their response to the other, and everyone responded in at least one of VO2peak, lactate threshold or submaximal heart rate across the two. A flat result on one stimulus did not predict a flat result on the other.
What was measured: VO2peak, lactate threshold, submaximal HR. What was not: any performance test, any trained athlete, and — importantly — the same mode twice in the same person.
Because no mode was repeated, the study cannot tell whether a person's different responses to END and SIT are a stable trait or ordinary block-to-block variation (REF-430). The authors' reading is that changing the stimulus may reduce non-response.
Inference, not tested here: for an athlete whose completed block was flat, a change of stimulus is a reasonable next step; more dose is the other (REF-996).
This is the verifiable paper from the group named in REF-432, whose own sources could not be verified.

Limitations

Three-week blocks; no control period and no repeated exposure to the same mode, so true individual response and measurement variation are confounded.
"Responded in at least one variable" across three variables and two modes is a low bar: the more variables counted, the more likely a change beyond the threshold in one of them.
n=21 recreationally active adults; group mean changes, sex and age are not in the abstract.
No performance outcome.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

randomised crossover, n=21; three-week blocks and no repeat of the same mode, so a true difference in response cannot be separated from test-to-test variation