REF-1549 · Cochran et al., 2015 · International Journal of Sport Nutrition and Exercise Metabolism · Paper · PubMed

Does restricting carbohydrate between two HIIT sessions help?

18 active people, 2 wk: 250-kJ TT rose more on 17 g than 195 g (211→244 vs 203→219 W, p = .02); mitochondrial markers no clear difference. Both arms doubled.

Main caveat: Read from the abstract. Sex split, blinding, test fuelling and marker values are missing.
limited evidenceMeasured on recreational people

Eighteen active people were split at random into two groups, matched on age, sex and fitness. On 6 days over 2 weeks everyone did two sessions a day of 5 × 4 min cycling at 60% of peak power with 2 min recovery. In the 3 h between sessions one group took 195 g of carbohydrate (about 2.3 g/kg) and the other 17 g (about 0.3 g/kg).

The 250-kJ time trial improved more on the low intake. Citrate synthase activity and protein and COXIV protein rose in both with no clear difference. Our reading: a rare performance gain from restriction between sessions, from people who are not athletes, and not one the measured mitochondrial markers account for.

Protocol

What it isolates: the fuel between two sessions. Everyone doubled, at the same gap and in the same order, so the double itself is not tested, and no one-a-day arm exists.
Who: 18 active, not highly trained; VO2peak 44 ± 9 mL/kg/min.
Training: 6 days over 2 weeks, 2 sessions a day, each 5 × 4 min at 60% of peak power with 2 min recovery; 3 h between sessions (12 sessions in all, our arithmetic).
Fuel: 195 g (HI-HI, about 2.3 g/kg) or 17 g (HI-LO, about 0.3 g/kg) of carbohydrate between sessions.
Measures: 250-kJ time trial; citrate synthase maximal activity and protein; COXIV protein.
Funding and interests: PubMed: Research Support, Non-U.S. Gov't; sponsor and conflicts not readable in the abstract record.

Key Results

Measure195 g between sessions (HI-HI)17 g between sessions (HI-LO)Between groups
250-kJ time trial, mean power, before to after (W)203 ± 53 to 219 ± 60211 ± 66 to 244 ± 75p = .02
Citrate synthase activity and proteinroseroseinteraction p > .05
COXIV proteinroseroseinteraction p > .05
Our arithmetic: +16 W (+7.9%) on 195 g and +33 W (+15.6%) on 17 g, a 17-W difference in the gain. Starting means differ by 8 W.
A between-group result for the time trial, since the abstract reports a greater gain with p = .02.
No clear difference, not "the same", for the mitochondrial markers: n = 18 and no equivalence test. The abstract does not give their values.

Mechanisms

Measured: whole-body time-trial power, and mitochondrial content by three markers.
Authors' interpretation: the first evidence of a short "train low, compete high" protocol improving whole-body capacity; they ask for a test in highly trained people.
Inference: a better time trial without a measured mitochondrial difference means the route is not shown; it may lie in something unmeasured or in the test itself.
Not measured, or not in the abstract: glycogen, power in the second session, how the test was fuelled, blinding (an obvious difference between 195 g and 17 g is possible), any result after the 2 weeks.

Practical Application

Does restricting carbohydrate between two HIIT sessions help? 18 active people, 2 wk: 250-kJ TT rose more on 17 g than 195 g (211→244 vs 203→219 W, p = .02); mitochondrial markers no clear difference. Both arms doubled.
Good for: the positive case on restriction, with its setting: not highly trained, 2 weeks, both arms doubling.
Good for: another case of performance and mitochondrial markers parting company; for signal against adaptation see PHYS-18.
Not good for: trained athletes. See the elite null in REF-1541 and the pooled null in REF-218.
Not good for: whether to double at all. No one trained once a day.
Not from this paper: no carbohydrate advice follows from one 2-week trial; the base's practice on a close next session is in SYNTH-85.

Limitations

Read from the abstract. Sex split, blinding, test fuelling and marker values are missing.
Small, short: 18 people, 2 weeks, one performance test.
Not highly trained; the authors ask for a test in highly trained people.
No once-a-day arm. Nothing here isolates the double.
Wide spread in power (SD 53-75 W) in a mixed-fitness sample.
How this entry was checked and changed
Why this level

(from the abstract) — parallel RCT with 18 (rubric B: <20). Not lowered: the outcome is performance and the abstract prints means, SD and the between-group p. Subscription article; no copy in PubMed Central or Europe PMC