Is Muscle Hyperplasia Possible? What the Science Says

Muscle hyperplasia, the formation of entirely new muscle fibers in an adult, has long been considered impossible or at least irrelevant in humans. The textbook position for decades was that you are born with a fixed number of muscle fibers and that all growth after birth comes from making those existing fibers bigger. That consensus is now less certain than it used to be. Animal studies have repeatedly demonstrated hyperplasia under extreme loading conditions, and a 2024 human study found that long-term resistance-trained individuals had roughly a third more muscle fibers than untrained controls, reopening a debate the field thought it had settled.

Why the Textbook Answer Was “No”

For most of the twentieth century, exercise scientists worked under the assumption that postnatal skeletal muscle growth in humans comes from enlarging existing fibers, a process called hypertrophy, rather than from making new ones.¹ This made intuitive sense: muscle biopsies taken before and after training programs consistently showed that individual fibers got bigger without any obvious increase in fiber count. Early human studies reinforced the idea. One well-known 1996 trial had college men perform 12 weeks of heavy resistance training and found increases in overall muscle cross-sectional area and individual fiber size but no change in estimated fiber number.² Results like these, replicated in other short-term human training studies, cemented the idea that hyperplasia was either nonexistent or negligibly small in people.

But “not detectable in a 12-week study of college students” is not the same thing as “impossible.” As researchers started running longer, more extreme experiments in animals, the picture got more complicated.

Strong Animal Evidence

The case for hyperplasia is much stronger in animals than in humans, and the gap in evidence quality is worth understanding. A meta-analysis pooling data from multiple animal studies found that mechanical overload produced a statistically significant increase in muscle fiber number, averaging about 15%, alongside a roughly 32% increase in individual fiber area.³ Those results came from several species under several types of loading, suggesting that the phenomenon is a general feature of mammalian and avian muscle biology, not an artifact of one quirky experimental setup.

Some of the most dramatic findings come from bird models. When researchers attached progressively heavier weights to the wings of adult quail for 28 days, the stretched muscle grew enormously in mass and showed an 82% increase in fiber number measured histologically.⁴ Interestingly, the increase in fiber number appeared only after prolonged overload. Quail stretched for 16 days showed massive hypertrophy but no increase in fiber count; the new fibers showed up between days 16 and 28. A related study found that fiber splitting, where existing fibers branch and eventually separate into two, rose to about 5% of fibers in the longest-stretched muscles, suggesting this may be one mechanism by which new fibers appear.⁵

Mammalian models tell a similar story. In rats subjected to weight-lifting exercise, researchers found biochemical and electron microscopy evidence of hyperplasia appearing during the second week of training, apparently triggered by damage-and-repair cycles in muscle fibers.⁶ Genetic models are even more compelling. Mice bred without the myostatin gene, which normally puts the brakes on muscle growth, develop both larger and more numerous muscle fibers from early development.⁷ The same gene is responsible for the exaggerated musculature seen in Belgian Blue and Piedmontese cattle, breeds that carry natural myostatin mutations and develop what is called “double muscling.”⁸⁹

None of this directly proves hyperplasia happens in human gym-goers, of course. Animal overload models tend to be far more extreme than any realistic human training program. Hanging a weight equal to a third of your body mass from one limb for a month straight is not something people do. But the consistency of hyperplasia across species raised the reasonable question: is the human body truly incapable of this, or has the evidence just been too hard to collect?

A Provocative Human Finding

That question got new fuel in 2024 when a study compared muscle tissue from long-term resistance-trained individuals against untrained controls. The trained group had about 70% greater maximum muscle cross-sectional area, which was expected. What was less expected was that they also had an estimated 34% more muscle fibers, along with 29% larger individual fiber area.¹⁰ Each fiber also packed roughly 49% more myofibrils, the contractile strands inside a fiber. In total, the trained muscles had about double the myofibrils of untrained muscles.

This is one of the first human studies to report a statistically significant difference in fiber number between trained and untrained people. But it comes with a critical caveat: this was a cross-sectional comparison, not a longitudinal trial. The researchers compared people who had been training for years against people who had not. That design cannot prove the extra fibers grew in response to training. It is possible the people who happened to be born with more fibers were the ones who stuck with training for years because they responded well. Self-selection bias is real in studies of dedicated athletes.

Still, a 34% difference is large enough to be interesting, and the study used careful methodology including direct fiber counts. Previous short-term training studies that found no hyperplasia typically lasted only 8 to 16 weeks and used less experienced subjects, so they may simply not have looked long enough or at the right population.

The Fiber Splitting Debate

If new fibers do appear in adults, how would they form? One proposed mechanism is fiber splitting, where an existing fiber develops internal fissures and eventually divides into two or more daughter fibers. Researchers have observed split-looking fibers in biopsies from powerlifters, but whether these represent genuine splitting or something else entirely has been argued about for decades.

A biopsy study of powerlifters, including some who used anabolic steroids, found that fibers with fissures were present in both groups, with the steroid-using group showing significantly more of them in certain muscles. But serial cross-sections told a more complicated story: the fissured fibers changed their appearance dramatically over very short distances, and many showed signs of recent degeneration and regeneration rather than clean division. The researchers concluded that these “split fibers” were most likely the result of defective regeneration after damage, not true splitting into new independent fibers.¹¹

Work using electrical stimulation to induce controlled muscle damage has pointed in a similar direction. When researchers examined regenerating human muscle tissue 30 days after damage, they found small branches budding off parent fibers, but most of these branches fused back into the parent fiber further along its length. The branches typically had a more immature protein profile than the main fiber, suggesting they were new myotubes that formed during regeneration and then merged with the existing fiber rather than splitting off from it.¹² In other words, what looks like splitting in a single cross-section may actually be incomplete fusion during repair.

That said, some researchers have pushed back on the assumption that fiber splitting is always pathological. A 2019 hypothesis paper argued that splitting under extreme loading is a nonpathological physiological response in animals and that the mechanisms and consequences deserve further exploration rather than automatic dismissal.¹³ The truth probably sits somewhere in the middle: some fiber branching is regeneration-related, some may represent genuine splitting under extreme conditions, and distinguishing the two in a single biopsy snapshot is extremely difficult.

Satellite Cells and the Raw Machinery for New Fibers

The biological ingredients for making new muscle fibers clearly exist in adult human tissue. Satellite cells, the muscle-specific stem cells that sit just outside mature muscle fibers, can activate, multiply, and either fuse into existing fibers (contributing to hypertrophy) or fuse together to form entirely new fibers.¹⁴ Lab studies have shown that human satellite cells isolated from adult muscle can activate, differentiate, and form new fibers in culture within about eight days.¹⁵

So the cells capable of building new fibers are present and functional in adult humans. What is less clear is whether the signaling environment inside a living person ever favors creating a new independent fiber rather than simply patching an existing one. In a petri dish, satellite cells will happily form new myotubes. In a living muscle, the repair process overwhelmingly seems to funnel those cells toward reinforcing the fibers already there. Whether sufficiently extreme or prolonged mechanical stress can tip the balance toward new fiber formation is the core open question.

Why This Has Been So Hard to Settle

One reason the hyperplasia question remains unresolved after decades is that counting muscle fibers in a living person is remarkably difficult. The standard method involves taking a needle biopsy, a tiny cylinder of tissue, and counting fibers in the cross-section. The problem is that fiber composition varies greatly along the length and width of a muscle, so a single biopsy may not accurately represent the whole thing.¹⁶ Two biopsies taken from different spots in the same muscle can give meaningfully different fiber counts, making it easy to mistake sampling variation for real changes.

Estimating total fiber number from a biopsy requires additional assumptions and calculations: you measure the total cross-sectional area of the muscle (usually by MRI), divide by the average fiber area from the biopsy, and arrive at an estimate. Any error in either measurement gets amplified. Longitudinal studies, where you biopsy the same person before and after training, face the added complication of hitting the exact same spot twice.

Newer tools are starting to improve the situation. Deep learning-based image analysis software can now automatically segment and count fibers in stained muscle sections with high precision, reducing human error in the counting step.¹⁷ And three-dimensional imaging protocols using tissue clearing and confocal microscopy allow researchers to trace fibers along their length in rodent muscle, rather than relying on a single cross-sectional snapshot.¹⁸ These methods are still mostly limited to animal tissue and research settings, but they represent a genuine step forward. The next generation of hyperplasia studies in humans will have better tools than anything available to the researchers who shaped the current consensus.

Do Hormones or Steroids Drive Hyperplasia?

Given how common anabolic steroid use is among people who train at the extreme end, a natural question is whether hormones push muscle into hyperplasia territory. The answer, surprisingly, is that the evidence is thin. In the avian stretch-overload model, where hyperplasia reliably occurs, administering anabolic steroids did not enhance either the degree of fiber hypertrophy or the formation of new fibers beyond what stretch alone produced.¹⁹ Steroids clearly accelerate hypertrophy in humans, but their relationship to hyperplasia specifically seems to be a different question.

Growth hormone and IGF-1, another anabolic signaling molecule, tell a similar story. While muscle-specific IGF-1 delivery causes well-documented hypertrophy in animal models and cell culture, there is currently very little evidence that elevated systemic levels of growth hormone or IGF-1 drive muscle growth in healthy individuals beyond what training alone achieves.²⁰ The powerlifter biopsy study mentioned earlier did find more fissured fibers in steroid users, but the researchers interpreted those fissures as signs of impaired regeneration rather than productive fiber splitting. If anything, the continuous high-stress training combined with steroid use may have disrupted normal repair processes.

Aging Runs the Process in Reverse

While the debate about whether adults can gain fibers continues, there is no debate at all about losing them. Aging clearly reduces muscle fiber number. The progressive loss of motor neurons with age leads to denervation of muscle fibers, and when those fibers are not successfully reinnervated by remaining neurons, they atrophy and disappear.²¹ This process, broadly called sarcopenia, involves both a reduction in fiber number and a shrinkage of remaining fibers, with fast-twitch (type II) fibers being hit harder. Studies of aging muscle show a gradual decrease in both the number and size of type II fibers, reflecting a shift toward a slower, weaker fiber profile over time.²²

Fiber loss with aging has been documented in specific muscles too. In the pectoral muscle, type I fiber number declined significantly after age 60, while type II fibers shrank progressively starting around age 40.²³ If hyperplasia does occur with training, it may be particularly relevant for older adults: not as a way to build extraordinary muscle, but as a potential countermeasure to the natural attrition of fibers that accompanies aging. Whether resistance training can actually replace fibers lost to denervation, rather than merely enlarging the ones that remain, is an open question that carries real clinical significance for the aging population.

What Eccentric Training Does at the Molecular Level

If any type of exercise has the potential to nudge human muscle toward hyperplasia, eccentric contractions are a common candidate. Eccentric exercise, where the muscle lengthens under load (think the lowering phase of a bicep curl or running downhill), causes substantially more mechanical disruption to muscle fibers than concentric exercise. Gene expression studies show that eccentric contractions trigger a broader and more intense transcriptional response in muscle tissue, including activation of pathways involved in muscle remodeling and inflammation.²⁴ The damage-repair cycle set in motion by eccentric loading is, at least in theory, the kind of environment where satellite cell activation could lead to new fiber formation rather than just repair.

This does not mean eccentric training has been proven to cause hyperplasia in humans. It means the molecular conditions that would be necessary for hyperplasia, including satellite cell mobilization, inflammation-mediated remodeling, and activation of growth-related genes, are more strongly triggered by eccentric work than by other loading patterns. Whether the human body ever crosses the threshold from “repair existing fibers” to “build new ones” under real-world training remains the question researchers are still trying to answer. The animal data says the biological machinery is there. The human data says we have not yet caught it in the act convincingly enough to close the case.