REF-1445 · Bi et al., 2026 · Scandinavian Journal of Medicine & Science in Sports · Paper · PubMed

Intervals or continuous work: which moves what?

115 trials: HIIT raised VO2max more (g 0.39); thresholds and performance showed no significant difference, with wide intervals. Sedentary to elite pooled; abstract only.

Main caveat: Abstract only; closed access. Heterogeneity, bias ratings and comparator intensities unknown.
fairly strong evidenceMeasured on non-athletes to world-class athletes

Bi and colleagues pooled randomised trials comparing the chronic effects of high-intensity interval training (HIIT) with moderate-to-vigorous-intensity continuous training (MVICT). Against MVICT, HIIT gave significantly larger gains in relative VO2max (g = 0.39), absolute VO2max (g = 0.29), maximal aerobic power or speed (g = 0.31) and mean anaerobic power (g = 0.47).

No significant difference appeared for peak anaerobic power, first and second intensity thresholds, exercise economy or indices of physical performance. Training status, age, sex, interval type and exercise mode moderated the effect.

The authors conclude HIIT is superior for VO2max and anaerobic capacity and that "both modalities show comparable outcomes" for thresholds, economy and performance. Our reading: useful as the wide-angle view behind "continuous work can carry threshold gains and does not replace intervals for VO2max", and consistent with REF-989 (A-) in trained athletes.

But "comparable" here means no significant difference with intervals that still allow a moderate effect either way (first threshold g 0.43, CI −0.38 to 1.25). A reader should not take it as proof that tempo work equals intervals for threshold or performance.

Protocol

Eligibility: RCTs comparing chronic HIIT and MVICT; subgroup analyses and meta-regression for moderators.
Sample: n = 3196 in 115 studies; mean age range 8 to 68 years; untrained sedentary to elite/world-class.
Analysis: three-level meta-analysis, Hedges' g.
Not in the abstract: how MVICT intensities were defined or split, durations, risk of bias, heterogeneity, publication bias, work matching.
Funding and interests: none stated in the PubMed or Crossref record; no conflict statement in the abstract record.

Key Results

Outcome, HIIT vs MVICTg95% CIReading
Relative VO2max0.390.27, 0.51favours HIIT; 1.30 mL/kg/min (printed as "SD = 1.30")
Absolute VO2max0.290.15, 0.43favours HIIT; 0.09 L/min (printed as "SD = 0.09")
Maximal aerobic power or speed0.310.17, 0.47favours HIIT
Mean anaerobic power0.470.08, 0.86favours HIIT
Peak anaerobic power0.31−0.06, 0.68no significant difference
First intensity threshold0.43−0.38, 1.25no significant difference
Second intensity threshold0.06−0.25, 0.36no significant difference
Exercise economy0.26−0.03, 0.54no significant difference
Physical performance0.04−0.46, 0.54no significant difference
VO2max subgroups named as favouring HIIT: Tier 0 (g = 0.34), Tier 1 (0.57), Tier 3 (0.83); men (0.43) and mixed samples (0.42); short (0.55) and long intervals (0.57); rowing (0.71), running (0.53), cycling (0.29).
Not named: Tier 2, Tier 4-5 and women alone; the abstract gives no values for them, so absence is not shown to be a null.

Mechanisms

Measured: larger VO2max and maximal aerobic power gains with HIIT than MVICT in pooled RCTs.
Authors' interpretation: the advantage depends on participant and protocol characteristics ("one size does not fit all").
Inference: as far as the abstract shows, continuous arms below the first threshold and arms between the thresholds are in one bin, so these pooled numbers cannot say whether heavy-domain continuous work behaves more like easy riding or more like intervals.

Practical Application

Intervals or continuous work: which moves what? 115 trials: HIIT raised VO2max more (g 0.39); thresholds and performance showed no significant difference, with wide intervals. Sedentary to elite pooled; abstract only.
Read it with [REF-989] (A-) and [REF-150] (B+): in trained athletes adding work above the steady-state ceiling raised VO2peak, and 4 × 16 min at about 88% of peak heart rate did not beat low-intensity riding in one trial. Not from this paper; the base's own practice: threshold work takes one of the week's quality sessions, two only above about 8 h (SYNTH-27, SYNTH-26); VO2max formats are in SYNTH-33.
What it does not show: that continuous work equals intervals for threshold power or race performance (a null, wide CI); anything specific to tempo or to a band set from FTP; a result for women alone or for Tier 2 amateurs.

Limitations

Abstract only; closed access. Heterogeneity, bias ratings and comparator intensities unknown.
Very broad population; study mean ages 8 to 68, and the share of trained athletes is not stated in the abstract.
Comparator bin mixes moderate and vigorous continuous work.
Nulls with wide intervals on the outcomes athletes care about most (threshold, performance).
Reporting: the abstract labels 1.30 mL/kg/min and 0.09 L/min "SD"; we read them as raw mean differences, which it does not state.
Funding and interests: not stated in the record.
How this entry was checked and changed
Why this level

(from the abstract) — an SR/MA of RCTs held below A: mostly small trials (about 28 people per trial, our arithmetic), a very wide population, and no heterogeneity or risk-of-bias figures in the abstract