Polarized > Threshold training over 6 weeks
Participants
n=12 male cyclists, described by the authors as already well-trained. Randomized crossover: two 6-week training periods separated by 4 weeks of detraining.
Protocol
Polarized model: 6.4 (± 1.4) h/week; 80% / 0% / 20% of training time in low- / moderate- / high-intensity zones.
Threshold model: 7.5 (± 2.0) h/week; 57% / 43% / 0%.
Before and after each period (after 2 days of diet and exercise control): fasted muscle biopsies (mitochondrial enzyme activity, MCT1/MCT4), urine metabolomics, 40-km time trial, incremental test, peak power output (PPO), and high-intensity exercise capacity (time to exhaustion at 95% of maximal work rate).
Key Findings
Endurance performance (40-km TT), PPO, lactate threshold (LT), MCT4 and high-intensity exercise capacity all improved over BOTH training periods.
Improvements were greater after polarized than after threshold for: PPO 8 (± 2)% vs 3 (± 1)%; LT 9 (± 3)% vs 2 (± 4)%; high-intensity exercise capacity 85 (± 14)% vs 37 (± 14)% (all P < 0.05).
The abstract does not report a between-model difference for the 40-km TT itself — it improved in both periods.
No change in mitochondrial enzyme activities or MCT1 after either period; metabolomics separated the threshold period but not the polarized one.
Consistent with the later multi-sport comparison by Stöggl & Sperlich 2014 REF-303.
Practical Application
In 12 well-trained male cyclists (randomized crossover, 6 weeks each), a polarized distribution (80/0/20, 6.4 h/wk) produced larger gains than a threshold distribution (57/43/0, 7.5 h/wk) in peak power output (+8% vs +3%), lactate threshold (+9% vs +2%) and time to exhaustion at 95% of max work rate (+85% vs +37%); 40-km TT time improved in both without a reported between-model difference — polarize the intensity distribution when a 6-week block has to move thresholds and peak power.
Inference, not tested: the polarized period did this on fewer weekly hours, so the gain is not a volume effect.
Not from this paper; the base's own practice: keep the middle zone empty on purpose during such a block — the moderate-intensity time in the threshold model is what the polarized model removes.
Limitations
n=12 men, crossover with a 4-week detraining washout; 6 weeks per period.
Muscle-level markers (mitochondrial enzymes, MCT1) did not change, so the mechanism of the systemic gain is not shown here.
How this entry was checked and changed
Why this level
regraded 2026-09-27 from A: randomized crossover, n=12
Content correction 2026-09-27
"Better TT and threshold test results" -> the abstract reports greater polarized gains for PPO, LT and high-intensity exercise capacity; the 40-km TT improved in both periods with no between-model difference stated.
Design: RCT -> randomized crossover, n=12 added; Level A -> B. Sport "cycling, running" -> cycling (the study was in cyclists only).
Added the training-distribution and hours figures from the abstract; "Confirms Stöggl 2014" -> "consistent with" (that paper came later).
Cross-references to duplicate stubs repointed to their canonical entries (REF-085->REF-303); the stubs were emptied on 2026-09-27.