A 2026 Frontiers Meta-Analysis Just Ended The Cyclist-Vs-Runner Cross-Training Debate.
Seven controlled trials, four weeks minimum intervention, and one clear answer. Here is what the Menges review found about swapping runs for rides.
Training
Strength Training For Cyclists: What The 2026 Meta-Analysis Actually Found
September 23, 2026
A January 2026 meta-analysis of 17 studies on 262 cyclists found heavy strength training improves cycling performance and efficiency but not VO2 max.
Every cyclist over 30 has been told the same thing.
Lift heavy, squat weight, build the leg. Your VO2 max will follow.
A systematic review published in the European Journal of Applied Physiology in January 2026 just told the strength-training-for-cyclists community that half of that claim is wrong.
Cycling performance, cycling efficiency, and anaerobic power did improve with heavy strength training. But VO2 max, power at VO2 max, maximal metabolic steady state, and anaerobic capacity did not.
That is a much narrower finding than the internet consensus. It is also more honest.
The study, led by Cristian Llanos-Lagos of the Physical Performance Sports Research Centre at Universidad Pablo de Olavide in Seville, pooled 17 controlled trials on 262 endurance cyclists, including 60 women.
The programs in those trials ran 5 to 25 weeks. They used one to three strength sessions per week and defined heavy strength as at least 80 percent of one-repetition maximum, or 1RM.
The results were graded as low certainty of evidence.
That grading matters. It is not the same as no effect, but it is far from the heavy squats will make you a Grand Tour climber claim that has spread through cycling media over the past decade.
What Heavy Strength Training Actually Improved
The three variables that showed statistically significant pooled improvements were the ones most tied to neuromuscular and morphological adaptation.
Cycling performance. This is the practical outcome of a time trial or a set effort, and it did improve across the pooled studies.
Cycling efficiency. This measures how much oxygen a rider needs to produce a given power output, and it also improved. A more efficient cyclist rides at a lower percentage of their maximum to hit the same watts.
Anaerobic power. This is the ability to produce short, hard, non-oxidative bursts, and it improved as well.
Those three findings sit consistently with a body of research that has long shown heavy lifting builds the neuromuscular pathways for producing force. The meta-analysis just made the pooled effect statistically visible for the first time.
The effect sizes were moderate. Not transformational.
That distinction is important. Reviews of this kind rarely produce huge effect sizes because they average across many programs, many riders, and many training states.
What Heavy Strength Training Did Not Improve
Now the harder findings.
Maximal oxygen uptake, or VO2 max, did not show a statistically significant pooled improvement. This is the number that gets cited most in cycling media as the reason to lift, and this meta-analysis did not find it in the pooled data.
Power at VO2 max, known as pVO2max, also did not improve. This is the sustainable output associated with maximal aerobic capacity, and it also did not respond to heavy strength training in the pooled analysis.
Maximal metabolic steady state, or MMSS, did not budge. This is close to what most riders think of as their functional threshold power, or FTP, and the pooled result was not significant.
Anaerobic capacity did not respond either. This is the total non-oxidative work capacity, and it was flat across the pooled trials.
Four of the five core aerobic-and-anaerobic metrics did not respond to a strength intervention.
That is the finding that reframes the entire cyclist-lifting conversation.
Strength work helps you use your engine more efficiently and puts a bigger sprint in your legs. It does not, based on the pooled evidence, make the engine itself bigger.
Why This Contradicts What Everyone Has Been Saying
The internet claim goes like this: heavy squats build fast-twitch fibres, fast-twitch fibres are linked to VO2 max, therefore heavy squats build VO2 max.
That chain has always been slightly wrong.
VO2 max is primarily a function of central cardiac output and peripheral oxygen delivery, not muscle fibre composition. Muscle fibre composition affects how efficiently you use that oxygen delivery, not the ceiling of it.
The 2026 meta-analysis is the pooled evidence supporting that distinction.
It also aligns with what strength coaches like Ronnestad and Vikmoen have said for years in the underlying primary studies. Their individual trials often showed cycling efficiency and time-trial performance improvements but did not show consistent VO2 max shifts.
Aggregating those trials just made the pattern visible.
The finding that a systematic review contradicts online consensus is not the same as a finding that lifting is pointless. It is, however, the finding that lifting is misprescribed.
What This Means For Your Training
If you are a cyclist who lifts to build VO2 max, the meta-analysis says you should probably lift for a different reason.
Lift to become more efficient at your existing threshold. That change alone can be worth a real time-trial improvement.
Lift to build anaerobic power. That is what wins group-ride sprints and what lets you cover attacks on a climb.
Lift to build durability. This meta-analysis did not measure durability, but adjacent research shows heavy lifting probably reduces late-race power loss.
Do not lift with the expectation that your VO2 max, your FTP, or your ceiling of aerobic capacity will move much because of the lifting itself. Those numbers move with structured aerobic work, and this meta-analysis is direct evidence that strength training does not substitute for that.
For the training work that actually does move VO2 max, our guide on VO2 max intervals is here. And for the base-level Zone 2 work that supports all of it, our Zone 2 primer explains why riding slow is the fastest way to get faster.
The Volume And Frequency Question
One of the more useful sub-findings of the meta-analysis was that program duration and session frequency did not consistently moderate the results.
Programs from 5 weeks to 25 weeks all produced roughly the same effect sizes.
Programs at one, two, or three sessions per week all produced roughly the same effect sizes.
There is no minimum threshold in the pooled data that supports the you need two sessions per week or it does not count claim floating around cycling coaching. It also does not support the more is better claim.
The neuromuscular adaptations that heavy lifting produces appear to happen within a reasonable range of exposure, and pushing past that range does not produce further gains.
For a working amateur cyclist, one to two heavy sessions per week for 8 to 12 weeks is likely to produce most of the available benefit. Adding more sessions probably will not.
Where The Study Falls Short
This is where the low certainty of evidence grading becomes important.
The pooled sample was 262 cyclists. That is not small, but it is not huge either.
Only 60 of those cyclists were women. The pooled analysis for female-specific responses is genuinely underpowered.
The included trials used different lifting programs, different endurance protocols, different measurement techniques for VO2 max, and different populations. The heterogeneity was real.
Some of the null findings might be true nulls, and some of them might reflect the noise of averaging across very different studies. The authors themselves flagged this.
There is also a self-selection issue in the underlying studies. Cyclists who volunteer for a strength intervention are typically already high-performing, and the ceiling effects on VO2 max in those populations are real.
The certainty grade is low for a reason.
This is not a finding that closes the debate forever. It is the best pooled evidence available in 2026, and it should update our beliefs, but it should not fossilise them.
The Meta-Analysis Timing
The study was accepted for publication on June 19, 2025, and published online on July 9, 2025. It appeared in print in the January 2026 issue of the European Journal of Applied Physiology, volume 126, pages 193 to 222.
That is a serious peer-reviewed home. The journal is one of the leading outlets for exercise physiology research, and Ramirez-Campillo, one of the co-authors, is a well-cited name in the strength-and-conditioning literature.
If you are the kind of rider who reads the primary source, that is the primary source.
The Honest Takeaway
Heavy strength training helps cyclists become more efficient and more powerful in short bursts.
It does not measurably improve aerobic ceilings.
That is the two-line summary of a 30-page meta-analysis.
The distinction sounds small until you notice how many training plans, coaching claims, and gear-adjacent content pieces have been sold on the incorrect claim that lifting builds a bigger engine. It does not, based on the pooled evidence.
It builds a more efficient use of the engine you already have.
For a cyclist who understands the difference, this is still a useful reason to lift. For a cyclist who lifts hoping for a bigger aerobic ceiling, this is a course correction.
The Llanos-Lagos meta-analysis is not the final word.
It is the best word we have in 2026. That is worth paying attention to.
Become a Confident Cyclist in 5 Days
Perfect for the new riders!
No spam. Cancel anytime.
