Carbohydrate periodization: why 'train low' improved running economy

The idea of "train low" (training with low carbohydrate stores) sounds paradoxical: why deliberately cut back on your main fuel if endurance running relies precisely on carbohydrates? The logic here isn't about starvation, it's about adaptation. When the body regularly trains against a backdrop of carbohydrate deficit, it is forced to burn fat more actively and remodels its mitochondria. The result is so-called metabolic flexibility: the ability to switch flexibly between fat and carbohydrates depending on intensity. For years this remained an elegant hypothesis. A recent randomized study from Austria tested it on runners — and got a surprisingly clear answer: what matters isn't the low-carb phase itself, but its periodization.

What was studied

Scientists from the University of Vienna (Kripp, Feichter, König) ran an 8-week randomized controlled trial. The final analysis included 24 recreationally active men (training 2–3 times a week, average VO₂peak around 51 ml·kg⁻¹·min⁻¹), split into three groups:

PER (periodization): the first 4 weeks ≤50 g of carbohydrates a day (more than 60% of energy from fat), then 4 weeks of returning to 50–60% of energy from carbohydrates.
LCHF (constant low-carb): ≤50 g of carbohydrates a day for all 8 weeks.
CHO (control): 50–60% of energy from carbohydrates for all 8 weeks.

The key point: the training was identical across all three groups. The first 4 weeks were base endurance (5 heart-rate-based sessions a week), the next 4 weeks were interval work. In other words, only the diet differed. Diet adherence was tracked through food diaries and ketone body measurements (in the first 4 weeks, PER and LCHF averaged 1,9 mmol/l — confirmation of a genuine keto state). Running economy was measured at the lactate threshold in ml·kg⁻¹·km⁻¹ (how much oxygen it takes to cover a kilometer — the lower, the more economical the running).

Results

Running economy at the lactate threshold improved only in the periodization group. A significant "time × group" interaction (p = 0,007); in PER, economy improved by −27 ± 21 ml·kg⁻¹·km⁻¹ (p = 0,026), while there were no changes in LCHF and CHO. When recalculated at a fixed speed of 7,5 km/h, the effect came out borderline significant (p = 0,053) — meaning the finding is real, but not rock-solid.
Fat burning increased in both PER and LCHF. Maximal fat oxidation (MFO) jumped from 0,4 to 0,8 g/min by week 4 (p < 0,001). But from there the paths diverged: in LCHF, fat burning stayed high, whereas in PER it dropped after carbs were reintroduced — but the ability to rapidly oxidize carbohydrates came back (p = 0,013), and that is precisely metabolic flexibility.
Periodization delivered no net gain in "performance." VO₂peak, peak speed and time to exhaustion rose in all groups by roughly the same amount — thanks to the training itself, with no advantage from any diet ("time × group" interaction not significant).
Body mass fell in the low-carb groups. In PER, −2,1 kg over the first 4 weeks (mostly fat); in LCHF the decline continued for all 8 weeks; in the control group weight didn't change.

How to apply this

The practical takeaway: a low-carb phase is a tool for a few weeks, not a way of life. The PER model looks like this: 3–4 weeks of "train low" in a base high-volume block, where intensity is low and fat handles the fueling, followed by a return to carbs by the time quality speed work and the competitive period begin.

Who might find this interesting: recreational runners and triathletes who build their season in blocks and want to "level up" fat metabolism without losing carbohydrate power. To estimate how much glycogen you actually store and how many carbs you need for a long effort, the carbohydrate storage calculator will help. And it's worth assessing your glycolytic power (VLamax) — how aggressively the body burns carbohydrates — before and after the block, to see which way your metabolism has shifted.

Risks and nuances: a low-carb phase almost inevitably lowers the quality of high-intensity training, so it makes more sense to schedule it during the base period rather than the peak. In the study, well-being, GI, and load tolerance didn't differ between groups — but the subjects were healthy men under supervision, and the authors honestly note that returning to the recommended carbohydrate intake after keto turned out to be no easy task.

Limitations

Small sample (just 24 people, 7–9 per group) and limited statistical power for the "performance" metrics — some of the null results may be a Type II error.
Diet was assessed by self-report (food diaries), with possible systematic underreporting of calories.
Only recreationally active men — the results can't be transferred directly to women and elite athletes.
The effect was assessed in a step test, not in a real race/time trial, so "improved economy" ≠ a guaranteed result on race day.
Economy at the lactate threshold depends on the speed of the threshold itself, which changed over the course of the study.

Key takeaways

Carbohydrate periodization (low-carb phase → return to carbs) improved running economy at the lactate threshold; constant keto and constant carbs did not.
Periodization delivered the best of both worlds: fat burning increased and carbohydrate power was preserved (metabolic flexibility).
The diet added no direct gain in VO₂peak or work capacity — those metrics grew from the training itself.
In practice: "train low" is a short block in the base period, not a permanent regimen.
The data are preliminary: small sample, men only, assessed by a test rather than a race.

Source: Kripp AM, Feichter A, König D. Periodized carbohydrate intake influences metabolic flexibility and indices of running economy during endurance training in recreationally active males. Frontiers in Nutrition, 2026. https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2025.1750042/full (DOI: 10.3389/fnut.2025.1750042)