Zone 2 cardio 3×/week

What happens in the muscle

The stimulus is the repeated energy shortage in the muscle cell. In animals, a treadmill program led to a doubling of the capacity of the gastrocnemius and quadriceps to oxidize fatty acids; the activity of carnitine palmityltransferase, palmityl-CoA dehydrogenase and mitochondrial palmityl-CoA synthetase also doubled, and the protein content of the mitochondrial fraction rose by about 60 %.

In humans, moderate continuous training is followed by a higher mitochondrial volume and more mitofusin 2, while PGC-1α, TFAM and DRP1 remain unchanged at rest. The classic signaling cascade is thus an acute and not a chronic signal — which makes it invisible in the resting measurement.

Why the superiority of Zone 2 is not established

The core problem of all intensity comparisons is that two groups differ in many ways. One study solves this with a trick: single-leg cycling with a counterweight, so that the same person completes both forms of training — the same blood, the same hormones, the same diet. Ten young, active men trained one leg with intervals and the other with continuous training, over six sessions in two weeks, with matched work per session.

The interval leg won. Maximal citrate synthase activity was 10.2 ± 0.8 versus 8.4 ± 0.9 mmol·kg protein⁻¹·min⁻¹, oxidative phosphorylation capacity per mass 23.4 ± 3.2 versus 17.1 ± 2.8 for complex I and 58.2 ± 7.5 versus 42.2 ± 5.3 for complex I + II. Mitochondrial function per unit of citrate synthase remained unchanged: more mitochondria were formed, not better ones. COXIV at +24 %, NDUFA9 at +11 % and mitofusin 2 at +16 % rose equally in both legs.

The more honest wording

A more recent eight-week comparison with 14 men per condition does not arrive at a simple better or worse. Moderate continuous training increased mitochondrial content, complex I activity and the enrichment of proteins of the citric acid cycle and oxidative phosphorylation. Sprint interval training, by contrast, improved respiratory function and upregulated pathways of one-carbon metabolism and protein quality control; only after sprint intervals did COX7A2L accumulate in the III₂+IV₁ supercomplexes.

The honest summary is therefore not that intervals are better, but that the two intensities build different things. Easy training remains sensible, just not for the reason usually given.

What the studies show

MacInnis 2017: the cleanest comparison there is

Ten young, active men, single-leg cycling with a counterweight, each leg randomly assigned to one form of training. One leg completed six sessions of interval training with 4 × 5 minutes at 65 % of peak power and 2.5 minutes at 20 % in between, the other moderate continuous training with 30 minutes at 50 % of peak power, ten minutes apart on the same day, over two weeks. Work per session was matched: 143 ± 8.4 kJ versus 144 ± 8.5 kJ (p > 0.05). After the intervals, citrate synthase activity and respiratory chain capacity were higher. Limitations: ten young men, two weeks, single-leg model.

Botella 2025: the same question over eight weeks

Open-label, balanced randomized controlled trial with 14 men per condition. After eight weeks, moderate continuous training increased mitochondrial content, complex I activity and the enrichment of citric acid cycle and OXPHOS proteins, while sprint interval training improved respiratory function. Acutely, only sprint intervals triggered a mitochondrial stress response with an unfolded protein response and activation of mitochondrial quality control. The result contradicts the simple reading of the single-leg comparison, presumably because training there lasted only two weeks and the intensity was lower.

Reisman 2024: the methodological warning

Using a proteomics method at the single-fiber level, this paper showed that most training-induced changes in mitochondrial functional groups were no longer significant once normalized to markers of mitochondrial content. Much of what has been reported as qualitative adaptation may simply be more mitochondria and not better ones.

What the studies do not deliver

A study that explicitly confirms Zone 2 intensity as the optimum for mitochondrial formation does not exist. The available evidence is also small and male: ten young, active men over two weeks in the single-leg model, 14 men per condition in the eight-week comparison, 14 studies with a total of 184 participants in the meta-analysis, whose authors themselves rate the certainty of evidence as very low to low and name sex as a possible moderator.

How to do it

What was actually trained in the studies the mitochondria numbers come from: 30 minutes at 50 % of peak power, six sessions in two weeks, as single-leg cycling; in the comparison arm, 4 × 5 minutes at 65 % of peak power with 2.5 minutes at 20 % in between, with matched total work of 144 ± 8.5 kJ per session. The meta-analysis included moderate continuous training lasting two weeks or more and names training duration and modality as possible moderators. In the largest randomized long-term study, it was twice per week at around 70 % of peak heart rate over five years in 70- to 77-year-olds. There is no dose-finding study that tests weekly frequencies against each other for this goal — three sessions per week is a plausible practical figure, not a study result.

Safety

Easy endurance training at an intensity at which talking is still possible has no specific contraindications for healthy adults. In coronary heart disease, the largest available count recorded one fatal cardiac arrest per 129,456 hours of moderate training; the figure comes from supervised cardiac rehabilitation. With vertebral fractures and with knee and hip replacements, exercises with impact loading are contraindicated according to the DVO S3 guideline, so cycling and rowing rather than running. For bone, however, precisely these forms are of little use: the same guideline calls for weight-bearing loading and lists cycling and swimming as content that does not fall into this category. Under beta blockers, any percentage of maximum heart rate is useless because it is lowered by the medication. This is information, not a treatment recommendation.

Sources

• MacInnis MJ et al., The Journal of Physiology 2017
• Botella J et al., Nature Communications 2025
• Vabishchevich V et al., PLOS ONE 2026
• Reisman EG et al., Nature Communications 2024
• Molé PA, Oscai LB, Holloszy JO, Journal of Clinical Investigation 1971
• Holloszy JO, Journal of Biological Chemistry 1967 (rat study)
• Stensvold D et al., BMJ 2020 (Generation 100)
• Haugen T et al., Sports Medicine – Open 2022