Big Gear, Low Cadence: Are Torque Intervals Strength Training on the Bike?

Grinding a big gear up a hill at 50 rpm feels like strength work. Your legs burn, each pedal stroke takes real effort, and your heart rate stays lower than it would at the same power with a faster cadence. That feeling has made "torque intervals", also called low-cadence, big-gear or "muscle tension" intervals, a fixture of winter training plans and indoor training apps. The pitch is attractive for a time-crunched triathlete: build strength on the bike, skip the gym.

This article looks at whether that pitch holds up. It covers what low-cadence pedalling actually asks of your muscles, the training trials that compared torque intervals with normal-cadence riding and with gym work, who those trials tested, and how a triathlete might use big-gear work without expecting it to do a job it can't. It's the bike-side partner to TriForward's "Heavy Lifting for Bike Power", which covers the gym evidence in detail.

What a torque interval is

There's no agreed definition. In training plans, a torque interval usually means riding at a cadence well below your normal one, commonly 40 to 60 rpm, at a moderate to hard intensity, often seated and often uphill. A typical session might be several efforts of 4 to 10 minutes around threshold power at 50 rpm.

The logic is simple physics. Power is the product of torque and cadence. If you hold the same power and halve your cadence, you have to roughly double the force on the pedals. More force per stroke, the argument goes, recruits more muscle fibres and trains strength in a cycling-specific way.

The question is whether "more force than usual" is the same as "enough force to count as strength training". For that you need to compare the force with what the rider can produce at maximum.

How hard do your legs actually push?

A 2024 study by David Barranco-Gil and colleagues, published in the Journal of Science and Medicine in Sport, measured exactly this. They tested 11 highly trained male cyclists (average age 29, VO2max around 65 ml/kg/min). First, each rider's maximal dynamic force was found on a friction-loaded ergometer: short all-out efforts against progressively heavier loads until they couldn't complete a full pedal revolution, a cycling version of a one-repetition maximum. Then the riders pedalled at four intensities, from ventilatory threshold up to maximal aerobic power, at a free cadence and at 40, 60 and 80 rpm, plus short all-out sprints.

The results, as a percentage of maximal force:

• Free cadence: 14% at ventilatory threshold, 17% at maximal lactate steady state, 20% at the respiratory compensation point and 25% at maximal aerobic power.
• 40 rpm: 30%, 37%, 44% and 54% at the same four intensities.
• All-out sprints from a standstill: 21 to 31%, depending on the resistance.

Lower cadence did raise relative force, roughly doubling it at each intensity. But even the hardest combination, 40 rpm at maximal aerobic power, stayed at 54% of maximum. The authors concluded that relative pedalling forces are low across intensities and cadences, including during so-called torque training.

For comparison, heavy strength training in the cycling trials covered in "Heavy Lifting for Bike Power" typically works at around 80% of a one-repetition maximum or more. Inference: a torque interval at threshold, which in this study meant somewhere between 37 and 44% of maximum force at 40 rpm, is closer to a moderate muscular endurance load than to heavy strength training. It may still be a useful stimulus, but it isn't the same stimulus.

The study had limits. It involved only 11 men, all highly trained. Force was measured as total force on the pedal, without separating the component that actually turns the crank. And the sprints were done from a standstill against relatively light resistances.

The trial that put torque work against squats

The most direct test of the "strength on the bike" idea comes from the same Spanish research group. Raúl de Pablos, David Barranco-Gil and colleagues presented a randomised controlled trial at the 2025 Science and Cycling congress, published as an abstract in the Journal of Science and Cycling.

• Participants: 27 well-trained male cyclists, randomly split into three groups of 9.
• Gym group: 5 sets of 7 squats at 70% of one-repetition maximum, twice a week, with 4 minutes' rest between sets.
• Torque group: 5 intervals of 4 minutes at 70% of the power at VO2max, at a target cadence of 50 rpm, with 2 minutes' recovery, twice a week.
• Control group: normal training only.
• Duration: 10 weeks.

The squat group improved maximal aerobic power (effect size 0.60) and power at both ventilatory thresholds (effect sizes 0.47 and 0.48). The torque group and the control group showed no significant changes. The authors concluded that off-bike resistance training was superior for improving pedalling performance and that torque protocols need refining.

This is a useful result, but treat it with care. It's a congress abstract, not yet a full peer-reviewed paper, so the methods and the full data aren't available to check. Each group had only 9 riders, all men. And 70% of VO2max power at 50 rpm is a moderate intensity; a harder torque protocol might behave differently.

Twelve weeks at 40 rpm in veteran cyclists

An older trial is more relevant to many age-group triathletes because of who it studied. Kristoffersen and colleagues, writing in Frontiers in Physiology in 2014, recruited 22 well-trained male veteran cyclists aged around 47.

• Low-cadence group (11 riders): 5 intervals of 6 minutes at 40 rpm, at 73 to 82% of maximum heart rate, twice a week for 12 weeks, on top of their usual training.
• Free-cadence group (11 riders): 90 minutes a week at the same heart-rate range at their freely chosen cadence.
• Total training time over the 12 weeks was similar: about 91 hours versus 88 hours.

The results:

• The low-cadence group didn't improve. VO2max went from 57.2 to 58.5 ml/kg/min, not a significant change, and mean power in a 30-minute test went from 278 to 276 W.
• The free-cadence group improved within the group. VO2max rose from 58.9 to 62.2 ml/kg/min and 30-minute power from 284 to 297 W.
• Leg strength, measured by one-repetition maximum leg press and leg extension, didn't change in either group.

The authors concluded that 12 weeks of twice-weekly 40 rpm intervals didn't improve aerobic capacity, cycling performance or leg strength. One caution: the improvements in the free-cadence group were within-group changes, and the paper didn't show a statistically significant difference between the groups in those outcomes. The safest reading is that low-cadence work did no better than normal-cadence riding, not that it was clearly worse. The leg-strength result fits the force data above: 40 rpm at a moderate heart rate wasn't heavy enough to make the legs stronger.

Specific gains for specific terrain

A 2012 trial by Alfred Nimmerichter and colleagues in the European Journal of Applied Physiology asked a slightly different question: does cadence-specific training improve cadence-specific performance? They randomised 18 male cyclists (VO2max around 59 ml/kg/min) to three groups for 4 weeks:

• Low-cadence uphill intervals at 60 rpm.
• High-cadence flat intervals at 100 rpm.
• Control, with no interval training.

The interval groups did two sessions a week of six 5-minute efforts at the power of their respiratory compensation point. Laboratory test results improved across the participants as a whole, with no significant differences between groups. But in 20-minute field time trials, a pattern appeared:

• Uphill time trial: +4.4% after low-cadence intervals, −1.3% after high-cadence intervals and +4.0% in the control group.
• Flat time trial: +1.5% after low-cadence intervals, +2.6% after high-cadence intervals and −3.5% in the control group.

The authors suggested the higher forces during low-cadence work may benefit performance and that field time trials pick up specific adaptations that a lab test misses. But look at the control group's uphill result: +4.0%, almost the same as the low-cadence group's +4.4%. Inference: with groups of around 6 riders over 4 weeks, the study hints at some terrain specificity but can't show that low-cadence intervals were better than no intervals for climbing.

The trial that found a benefit

Not every study is negative. A 2024 trial by Rafał Hebisz and Paulina Hebisz in PLoS ONE found a clear advantage for low-cadence work, and it's important to understand why it might differ.

• Participants: 26 well-trained female cyclists aged 17 to 20 were recruited and 24 were analysed, 12 per group. All had at least 3 years of cycling experience and trained at least 10 hours a week.
• Programme: 8 weeks of polarised training repeating a 4-day cycle: a sprint interval session (8 to 12 all-out 30-second efforts), a high-intensity session (4 to 6 efforts of 4 minutes at 90 to 100% of maximal aerobic power), a long low-intensity ride and an active-rest day.
• The difference: one group rode the sprint and high-intensity efforts at a freely chosen cadence above 80 rpm. The other rode them uphill on 6 to 9% gradients in big gears, at 50 to 60 rpm for sprints and 60 to 70 rpm for the 4-minute efforts.

Results:

• Maximal aerobic power rose 8.1% in the low-cadence group and 3.0% in the free-cadence group, a significant group-by-time interaction.
• Relative VO2max rose 8.7% versus 4.6%, also a significant interaction.
• The second ventilatory threshold improved only in the low-cadence group.

The authors concluded that low-cadence high-intensity intervals were more effective and that coaches should consider cadence, not just intensity and duration. They acknowledged the small sample and called for replication.

Note how different this is from a typical torque session. These were all-out sprints and near-maximal 4-minute efforts, ridden uphill, in young women training 10 or more hours a week. The low-cadence element was layered on top of very hard intervals, not used as a moderate "strength" session. Inference: if low cadence helps at all, it may help most when the intensity is already high, which is the opposite of how torque intervals are often prescribed.

Populations and sex split

Across the five studies in this article:

• Four studied only men: 11 highly trained cyclists in the force study, 27 well-trained cyclists in the congress trial, 22 veterans in their 40s and 18 trained cyclists in the terrain trial.
• One studied only women: 24 young, well-trained cyclists aged 17 to 20.
• None studied triathletes, and none tested how low-cadence riding affects the run that follows.

The only positive trial was in young women, and the negative trials were in men. It's tempting to read a sex difference into that, but the protocols were too different to say. The positive trial also used much harder intervals. Inference: the evidence can't tell you whether women respond differently to low-cadence training.

What this means for triathletes

Several practical points follow from the evidence, with the caveat that none of it comes from triathletes.

• Don't count torque intervals as strength training. Pedal forces stayed under 54% of maximum in the force study, and leg strength didn't change after 12 weeks of 40 rpm work. If you want the benefits of strength training, TriForward's "Heavy Lifting for Bike Power" and "The Minimum Effective Dose" cover how to get them in the gym.
• Don't expect torque work to beat normal-cadence intervals. Two trials in men found no benefit, and one found the gains were similar to doing nothing extra on a climbing test.
• If you use low cadence, consider attaching it to hard efforts. The one positive trial used near-maximal and all-out efforts. That's inference from a single study in young women, not a tested rule for age-groupers.
• Use big-gear work as a skill, not a cure. If your race includes long climbs, or you find yourself grinding at a low cadence late in a long ride, some practice at that cadence may help you feel comfortable there. The terrain trial hints at specificity but doesn't prove it.

TriForward's "Raising FTP on 8 Hours a Week" covers how to choose the intensity of your main interval sessions, which matters more than their cadence.

A triathlon scenario

The following example is illustrative, not drawn from a study. A 44-year-old age-grouper training for a hilly 70.3 has about 9 hours a week. Their winter plan from a training app includes two torque sessions a week, 5 × 8 minutes at 55 rpm around threshold, and they've dropped their gym session because "the torque work covers strength".

Based on the evidence above, they might:

• Restore one or two gym sessions a week, since the torque work is unlikely to provide a strength stimulus.
• Keep one of the two interval sessions at a normal cadence, using the intensity guidance in the FTP article.
• Keep a little low-cadence climbing, if they enjoy it, as specific practice for the course's hills, ideally as part of a harder session rather than a moderate one.
• Watch their knees. If big-gear work leaves them with pain at the front of the knee, drop it; nothing in the evidence makes it essential.

Common mistakes

• Replacing the gym with torque intervals. The force data and the congress trial both suggest they aren't equivalent.
• Grinding through knee pain. Low cadence means more force per stroke. If your knees complain, it isn't a sign the session is working.
• Treating any low-cadence riding as "strength". At a moderate intensity, the force is a fraction of your maximum.
• Assuming one positive study settles it. The trial in young female cyclists is interesting, but it used very different sessions from a typical torque plan and hasn't been replicated.
• Letting low cadence leak into race day. Practising big gears is fine, but the freely chosen cadence is what most riders return to when racing. TriForward's "High or Low Cadence Before the Run?" covers what bike cadence does to the run.

What we don't know

• Whether torque intervals help triathletes, or affect the run off the bike. No trial has tested this.
• Whether harder or lower-cadence protocols work better. Most trials used moderate intensities; the one positive trial used much harder efforts.
• Whether men and women respond differently. The only female trial used different sessions from the male trials.
• Long-term effects. The trials lasted 4 to 12 weeks.
• The full results of the squat versus torque trial, which so far is available only as a congress abstract.

References

• Barranco-Gil D, Hernández-Belmonte A, Iriberri J, Martínez-Cava A, Buendía-Romero Á, Alejo LB, et al. Relative pedaling forces are low during cycling. Journal of Science and Medicine in Sport, 2024;27(9):660–663. doi:10.1016/j.jsams.2024.05.009
• de Pablos R, Pallarés JG, Alejo LB, Martínez-Cava A, Valenzuela PL, Sánchez-Redondo IR, Revuelta C, Lucía A, Barranco-Gil D. Traditional resistance training versus torque training: a randomized controlled trial. Journal of Science and Cycling, 2025;14(2) (Science and Cycling congress abstract). University of Murcia research portal record
• Kristoffersen M, Gundersen H, Leirdal S, Iversen VV. Low cadence interval training at moderate intensity does not improve cycling performance in highly trained veteran cyclists. Frontiers in Physiology, 2014;5:34. doi:10.3389/fphys.2014.00034
• Nimmerichter A, Eston R, Bachl N, Williams C. Effects of low and high cadence interval training on power output in flat and uphill cycling time-trials. European Journal of Applied Physiology, 2012;112(1):69–78. doi:10.1007/s00421-011-1957-5
• Hebisz R, Hebisz P. Greater improvement in aerobic capacity after a polarized training program including cycling interval training at low cadence (50–70 RPM) than freely chosen cadence (above 80 RPM). PLoS ONE, 2024;19(11):e0311833. doi:10.1371/journal.pone.0311833