Does Cardio Kill Gains? What Concurrent Training Does to Muscle

Does cardio kill gains, and what would it cost you if it did

You lift three or four times a week and you also ride, row, or run. Somebody in the gym has told you that is a mistake, and the phrase they used was probably “interference effect”. The question underneath it is a real one: if the cardio is quietly taxing the muscle you are trying to build, you would want to know, because you are paying for it in hours either way.

So: does cardio kill gains? A systematic review with meta-analysis in Sports Medicine is the closest thing to a direct test. It took every training study it could find that ran the same lifting program twice — once on its own, once with aerobic training bolted on — and looked at what happened inside the muscle itself. The answer is small, specific, and not the one either camp in that gym argument wants.

Forest plot of 15 studies comparing muscle fiber growth after concurrent aerobic and strength training against strength training alone, with pooled effects of minus 0.34 for type I fibers, minus 0.13 for type II fibers and minus 0.23 overall

Figure 2 from Lundberg and colleagues (2022), reproduced under CC BY 4.0. Each row is one study. Dots to the left of the dashed line favour lifting alone; dots to the right favour lifting plus cardio. The diamonds at the bottom of each block are the pooled answers.

Does cardio kill gains, and what would it cost you if it did
What the study did: the same lifting program, with and without cardio
Does cardio kill gains at the muscle fiber level?
Running looked worse than cycling, on thin evidence
What did not matter: order, timing, frequency, training status
Where this sits in the wider evidence
What this means if you already lift and do cardio
The limits of this evidence
The practical answer
Common Questions About Cardio and Muscle Growth
How Pick It Up Helps
Further Reading & References

What the study did: the same lifting program, with and without cardio

Tommy Lundberg and colleagues searched six databases for supervised training studies lasting at least four weeks in which one group did aerobic and strength training together and another group did an identical strength-training prescription with no aerobic work at all. That control condition is what makes the comparison worth reading: the lifting is held constant, so anything that shows up is attributable to the cardio rather than to two different programs.

Fifteen studies met the bar, covering 300 healthy adults — 153 in the concurrent groups and 147 in the lifting-only groups. Twelve of the studies used cycling as the aerobic training; three used running.

What they measured is the unusual part. Most studies of this question use a tape measure, an ultrasound, or an MRI scan of the whole muscle. These studies took muscle biopsies and measured the cross-sectional area of individual fibers, split into the two families every muscle contains:

Type I fibers are the slow, fatigue-resistant ones. They do the work in endurance efforts and in the low-effort part of a set.
Type II fibers are the fast, forceful ones. Recruitment of them climbs as effort climbs, and in the thigh they make up a larger share of total muscle area than type I fibers do.

That distinction is the reason this paper exists. An earlier review by the same group had already reported no interference at the whole-muscle level, and the authors noticed that almost every individual study claiming an interference effect had measured fibers rather than whole muscles. This paper went and pooled the fiber studies to see whether the disagreement was real.

Does cardio kill gains at the muscle fiber level?

A little, and less than the phrase “kills gains” implies.

Pooled across all 15 studies and both fiber types, the concurrent groups grew their fibers slightly less than the lifting-only groups. The gap was 0.23 on a standardised effect-size scale, where researchers conventionally treat 0.2 as the boundary of “small”. Its range ran from 0.46 down to exactly zero — which is to say the analysis could not rule out no difference at all, and landed on a p-value of exactly 0.050.

Split by fiber type, neither half reached statistical significance on its own:

Type I fibers came out at 0.34 in favour of lifting alone, with a range running from 0.72 worse to 0.04 better.
Type II fibers came out at 0.13, with a range from 0.39 worse to 0.12 better. That is a dead heat.

The shape of that result matters more than the headline. The fibers that make up the bigger share of a thigh, and that heavy lifting leans on hardest, are the ones where adding cardio changed nothing measurable. What little interference showed up landed on the slow fibers, which is where you would expect endurance work to compete for the same adaptive machinery, and it is also the explanation the authors reach for.

Running looked worse than cycling, on thin evidence

The one subgroup that produced a clean-looking signal was the type of cardio.

Forest plot splitting the concurrent training studies by aerobic training type, showing cycling with pooled type I and type II fiber effects near zero and running with a type I effect of minus 0.81

Figure 4 from Lundberg and colleagues (2022), reproduced under CC BY 4.0. The left block is type I fibers, the right block type II. Cycling sits on the line; the three running studies sit to the left of it.

In the cycling studies, type I fibers came out at 0.22 and type II at 0.03 — nothing you would act on. In the running studies, type I fibers came out at 0.81 against, with a range from 1.26 to 0.36 that stays clear of zero. That is the single largest effect anywhere in the paper, and the authors’ own suggested explanation is mechanical: running loads the legs eccentrically, over and over, with a stretch-shortening cycle on every stride, while cycling is almost purely concentric. More inflammatory and metabolic stress per hour, in other words, competing with the lifting for the same recovery.

Two things stop this being the takeaway of the post.

It rests on three studies. The authors say so, and say caution is warranted. Three trials is a hypothesis, not a finding.

Running and cycling were not statistically distinguishable from each other. This is the part that is easy to miss and the paper is clear about it: the test for whether the two subgroups actually differ came out at p = 0.204 for type I fibers and p = 0.094 for type II. Neither clears the usual bar. What the paper established is that the running estimate was far enough from zero to be significant within its own subgroup — not that running and cycling differ from each other. The authors still put a recommendation on it — that someone chasing muscle should consider cycling over running — and it is a reasonable low-cost bet. It is not a proven one.

What did not matter: order, timing, frequency, training status

The list of things the paper looked for and did not find is the more useful half of it, because it is where most gym advice about cardio lives:

Lifting before cardio, or cardio before lifting, made no measurable difference when the two were done in one session.
Same session, same day, or different days made no measurable difference.
Training frequency made no measurable difference. The comparison was roughly four or five total weekly sessions against six or seven — about two lifting sessions a week in the lower group and three in the higher one.
Training status made no measurable difference. The untrained participants and the already-active ones showed the same pattern.

None of that means these variables are irrelevant to how you feel or how well a session goes. It means that across this body of evidence, none of them changed what happened to fiber size. Given how much confident advice exists about never running before you squat, that is worth knowing.

Where this sits in the wider evidence

This study does not overturn the field’s position on cardio and muscle. It refines it, and it comes from the same team that established the position it is refining.

The year before, Schumann and colleagues published the broader review this one is carved out of: 43 studies, using the same inclusion rules, looking at whole-muscle growth and at strength. Their pooled results were as close to nothing as a meta-analysis gets. Muscle growth came out at 0.01. Maximal strength — how much you can lift for one rep — came out at 0.06. Neither was statistically significant, and neither is a number anyone would train around.

One thing in that review did move: explosive strength, meaning jumping and other rapid force production, was attenuated by 0.28, and that was statistically significant. The attenuation was worse when the cardio and the lifting happened in the same session than when they were separated by at least three hours.

So the fair summary, stated as three findings rather than one:

Whole-muscle size and maximal strength are not compromised by adding aerobic training to a lifting program. That is the consensus, it rests on 43 studies, and this paper does not contradict it.
Fiber size may take a small hit, mostly in the slow fibers. That is this paper, it rests on 15 studies, and the authors state plainly that it “does not necessarily translate into differences in whole muscle hypertrophy” — that being the word researchers use for muscle growth.
Explosive power is the outcome with a real cost, and separating the sessions helps it.

These are not in conflict. A finding at the level of the individual fiber and a finding at the level of the whole muscle are two measurements of different things, taken with instruments of very different reliability, and the paper spends its discussion section on exactly why they might disagree — including the possibility that fiber-level changes show up first and whole-muscle measurements catch up later.

If you want the lever that actually moves muscle growth, it is not your cardio schedule. It is how many hard sets you do per week and how close to failure you take them. Both have better evidence behind them and both are bigger dials than anything in this paper.

What this means if you already lift and do cardio

Keep the cardio. Nothing here supports dropping it. The worst-case reading of this paper is a small effect on one fiber type, inside a body of evidence that finds no effect on the muscle you can actually measure, and set against cardiovascular benefits that are not in dispute.

If you are choosing a machine and muscle is the goal, take the bike. The evidence for this is thin and the authors label it preliminary, but the cost of acting on it is zero. If you have no particular attachment to running, ride instead. If you run because you like running or because you race, keep running — a possible 0.8 effect on type I fibers, from three studies, is not a reason to give up a sport.

Separate the sessions if power matters to you. In the broader review the explosive-strength cost was clearest when both were trained in one session, and did not reach significance when they were at least three hours apart. If you jump, sprint, or lift for a sport that rewards rate of force development, put the cardio in the evening and the lifting in the morning, or on a different day.

Stop worrying about the order. Cardio first, lifting first, same day, different days — none of it changed fiber growth. Do whichever one you will actually do. If lifting first means you lift better, lift first.

Do not let cardio quietly eat your lifting. This is the failure mode the research design deliberately excludes and real life does not. In every one of these studies the lifting program was identical in both groups. In your week it usually is not: the run happens, the session gets shortened, a set gets cut, and six weeks later the bar has not moved. That is not interference, it is lost volume, and it is by far the most common way cardio actually costs someone muscle.

The limits of this evidence

Fiber measurements are noisy. The authors flag this themselves and it is the single biggest caveat on the paper. One reference study they cite found a 13% coefficient of variation for fiber-area measurements taken within the same person, and recommended averaging two or more biopsies from different sites for a reliable number. The review’s own recommendation for future work is that studies sample two sites, which tells you what the included ones mostly did not do.

The studies were small and the quality was middling. The review scored each one on the PEDro scale and got a mean of 4.3, which on their own thresholds is medium quality: above 6 is high, 4 to 5 is medium. Most studies were moderate quality, two were poor, and one was good.

“Trained” is doing a lot of work in the training-status subgroup. Studies were classified as untrained only when the participants were described as sedentary, previously untrained, or inactive; everything else went into a single “active” bucket that spans recreationally active through well-trained. If you have been lifting seriously for years, this paper cannot tell you whether you respond like the average participant in it.

Four weeks is the floor. Interventions had to last at least four weeks, which is short enough that some of the pooled evidence describes the first month of a training response rather than a year of one.

The overall result sits exactly on the significance line. A pooled p-value of 0.050 and a confidence interval whose upper bound is zero is the statistical equivalent of a coin landing on its edge. Small and uncertain is the whole of what it supports.

The practical answer

Cardio does not kill gains. Across 43 studies of whole-muscle growth and maximal strength, adding aerobic training to a lifting program changed neither.
There is a small cost at the fiber level, worth about 0.23 on an effect-size scale, concentrated in slow-twitch fibers, and it may not show up in muscle size you can measure.
Cycle rather than run if muscle is the priority and you have no preference. It is a free bet on thin evidence, not a rule.
Separate cardio and lifting by a few hours if explosive power matters. Same session was worse than sessions three or more hours apart for jumping-type strength.
Order and timing did not matter for muscle growth. Train in whatever order you will actually stick to.
Protect the lifting itself. The sets you skip because of a run will cost you more than the run does.

Common Questions About Cardio and Muscle Growth

Does cardio kill gains? Not at the level anyone can measure on a whole muscle. Pooling 43 training studies, adding aerobic training to an identical lifting program changed whole-muscle growth by 0.01 and maximal strength by 0.06 on a standardised scale, neither statistically significant. A separate analysis of 15 studies that measured individual muscle fibers found a small attenuation of 0.23, which its own authors say may not translate into whole-muscle differences.

Does running kill gains more than cycling? Possibly, and the evidence is thin. In this review the three running studies showed an effect of 0.81 against slow-twitch fiber growth while the cycling studies showed essentially nothing, but the test for whether running and cycling actually differ came out non-significant. The authors suggest choosing cycling if building muscle is the goal, and call the evidence preliminary.

Should I do cardio before or after lifting? For muscle growth, it made no measurable difference in this review, nor did doing them on the same day versus different days. For explosive strength, the broader review found the attenuation was worse when both were done in one session than when they were separated by at least three hours.

How much cardio is too much when lifting? This review compared about four or five total weekly sessions against six or seven and found no difference in fiber growth between them, so within that range frequency was not the deciding variable. What it cannot tell you is where the ceiling is, because nothing in the pooled evidence went far beyond that.

Does cycling build muscle on its own? Some, and much less than lifting does. A separate meta-analysis of 21 studies comparing aerobic training against resistance training found resistance training ahead by 0.66 for whole-muscle growth, 0.99 for type I fiber area and 1.44 for type II fiber area. Cardio is not a substitute for a lifting program.

Does cardio affect strength gains? Maximal strength, no — 43 studies put the difference at 0.06 and could not distinguish it from zero. Explosive strength, yes: the same review found a significant attenuation of 0.28, worse when the two were trained in the same session.

Do I need to change my program because of this? Almost certainly not. The adjustments this evidence supports are choosing a bike over a treadmill when you have no preference, and putting a few hours between cardio and lifting if you care about power. Everything else in your program matters more.

How Pick It Up Helps

The real risk to your lifting is not the interference effect. It is the Tuesday after a hard interval session, when the working weight that was fine last week feels heavy, and you either grind out a set that costs more than it gives or you quietly cut the session short. Do that often enough and you have lost volume, which is the thing the evidence says actually drives growth.

That is the problem autoregulation solves. Pick It Up gives every set a target weight, a target rep count, and a target difficulty on a simplified RPE scale, and then updates the targets for your next set based on the weight, reps and effort you actually reported for the last one. On a day when cardio has taken something out of you, the session adapts to what you have rather than to what the spreadsheet assumed. At the end of the block it re-evaluates your numbers across every session and builds the next block from what you actually did — so a training phase with heavy running in it shapes the next one instead of being ignored.

Start a free trial and have all the science taken care of for you, so you can get back to building a better you.