End-Range Strength: Why Stretching Alone Does Not Hold

By Ada L. Wrenford

What Happens When You Only Stretch

Static stretching works, in the narrow sense that matters first: hold a muscle near its end range for a sustained period, repeated over weeks, and range of motion goes up. A 2025 systematic review and meta-analysis by Konrad and colleagues, published in Sports Medicine – Open, pooled 13 studies and 556 participants across stretching programs lasting five to fifteen weeks and reported a large average improvement in range of motion (effect size around 0.93) from baseline to the end of training. That part of the story is well established and consistent across decades of stretching research, including systematic reviews of acute stretching techniques showing that most common stretching methods produce measurable, if temporary, range increases within a single session.

The part that gets less attention is what happens after people stop. The same Konrad review followed participants through two to six weeks of detraining after their stretching programs ended. Range of motion declined during that window, a moderate drop (effect size around -0.41) relative to where people had peaked. The encouraging half of that finding is that range stayed above where people started (effect size around 0.55 versus baseline), so stretching gains are not wiped out entirely the moment sessions stop. The less encouraging half is that a meaningful share of the gain erodes quickly once the stimulus is removed, and the people who kept doing some activity during the detraining period retained more range than those who went fully sedentary. In practical terms, static stretching behaves like a maintenance habit: skip it for a month and you likely give some of it back.

What Loaded End-Range Training Studies Show

A separate line of research asks what happens when, instead of passively stretching, a person loads a muscle at or near its longest position, through movements such as deep split squats, deficit lunges, stiff-leg deadlifts, or dumbbell pullovers taken through a full stretch. A 2025 systematic review, meta-analysis, and meta-regression by Favro and colleagues in the Journal of Strength and Conditioning Research pooled 36 studies and 1,469 participants and found a medium-to-large pooled effect of resistance training on flexibility (g = 0.63) compared with non-training controls. Intensity mattered: high-intensity resistance protocols produced roughly double the flexibility gain of low-intensity ones in that analysis (0.75 versus 0.28), which points toward mechanical tension, not simply moving a joint through a range, as a driver of the effect.

Two randomized controlled trials give a more direct comparison. Rosenfeldt and colleagues (2024), in BMC Sports Science, Medicine and Rehabilitation, randomized 18 healthy young adults to eight weeks of full-range straight-leg deadlift training, static hamstring and low-back stretching, or a control condition. Sit-and-reach flexibility improved by a similarly large margin in both the training and stretching groups, with no significant difference between them, but the resistance-trained group produced far larger gains in isometric strength at deep hip flexion (g = 2.69 to 3.36 relative to stretching), meaning they got comparable flexibility plus usable strength at that range, something static stretching alone did not deliver. Murakami and colleagues (2025), in Frontiers in Physiology, ran a six-week trial in 36 untrained adults comparing static calf stretching against eccentric-emphasis bodyweight calf raises. Both groups gained similar ankle dorsiflexion range, but only the stretching group reduced passive muscle stiffness, and only the resistance-trained group increased calf muscle thickness and strength. Range of motion gains in both groups correlated with how much stretch sensation participants could tolerate, not with stiffness changes, an important detail returned to below.

Two Roads to the Same Range, Different Engines

If stretching and loaded end-range training can produce similar short-term range of motion gains, the interesting question is what is actually changing inside the tissue, and whether that changed state is the kind that persists. Two mechanisms come up repeatedly in this literature.

The first is stretch tolerance, a change in how much stretch sensation a person can accept before pulling back, driven by the nervous system rather than the muscle fiber itself. Murakami and colleagues found that in both their stretching group and their resistance-training group, the amount of range gained tracked with how much passive torque participants could tolerate, not with any change in tissue stiffness measured directly. This is a real, useful adaptation, but it is also plausibly the kind of adaptation that regresses once a person stops routinely working near end range, since there is nothing to keep “reminding” the nervous system that the position is safe.

The second is a structural one: the idea that a muscle fiber can add sarcomeres, its basic contractile units, in series when it is repeatedly loaded at a long length, a process sometimes called sarcomerogenesis. In theory, more sarcomeres in series means a longer resting fiber and a genuinely longer functional range, not just a nervous system more willing to tolerate the old range. This idea is biologically plausible and has support from animal models and some human eccentric-training studies, but the human evidence is inconsistent once you look closely. A 2021 systematic review and meta-analysis by Yagiz and colleagues in PLOS ONE, covering eight randomized trials of eccentric and Nordic hamstring training, found that estimated fascicle-length gains varied enormously depending on the measurement method used, from a small effect and roughly 0.02 cm using manual linear extrapolation, up to a very large effect and roughly 1.8 to 2 cm using trigonometric equation methods that the authors themselves flagged as likely to overestimate the true change. In plain terms, researchers agree loaded work at long muscle lengths does something to muscle architecture, but they do not yet agree on how much, and some of the more dramatic older numbers in this space were probably inflated by measurement error rather than reflecting real added tissue length.

Where the Case for Loaded End-Range Work Stands, Honestly

Put the pieces together and the case for loaded end-range training is reasoned rather than proven outright. What is well supported: stretching-only programs produce gains that partially regress once stopped, and loaded training at long muscle lengths produces flexibility gains in the same range as stretching while also building strength and, often, muscle size at that position. What is proposed but not yet nailed down: that the structural and strength-linked adaptations from loaded end-range work make the resulting range of motion more durable than range gained from stretching alone. That claim is plausible on mechanistic grounds, since a stronger muscle at a given length is less likely to feel unsafe there, and since people tend to keep lifting weights as an ongoing habit more consistently than they keep up a dedicated stretching routine. But it rests on indirect evidence, not a completed retention trial that tracked both groups through months of detraining and measured which one held on to more range. Treat the durability claim as the most likely reading of the current evidence, not as a settled conclusion.

A Realistic Example

Consider a composite scenario that reflects a pattern common in the training and rehabilitation literature discussed above, not a specific study or clinical record. A 34-year-old office worker with tight hips and hamstrings spends eight weeks doing daily static stretching before bed. Her hip flexion range improves noticeably, and she can finally touch her toes with straight knees. Then a busy work stretch and a short trip interrupt her routine for about three weeks. When she checks in again, some of that range has slipped back, consistent with the detraining pattern described by Konrad and colleagues, though she has not lost everything she gained.

For her next phase, she shifts to a program built around loaded end-range movements twice a week: deficit reverse lunges, Romanian deadlifts taken to a full, controlled stretch, and a light dumbbell pullover, all performed with slow lowering tempos and gradually increasing load over twelve weeks. She keeps a short general mobility routine during the first few weeks of the switch rather than dropping stretching all at once. Alongside the range of motion she is working to keep, she can now also load her hips and hamstrings near end range with real strength, something the stretching-only phase never gave her. When travel again interrupts her routine for a few weeks, her hip and hamstring range holds up better than it did after the stretching-only phase, which is exactly the pattern the mechanistic evidence above would predict, though this single scenario is illustrative rather than a controlled comparison.

What Researchers Say

Konrad and colleagues, whose 2025 meta-analysis anchors the detraining findings above, frame stretching’s fading effect as a reason to treat flexibility work the way trainers treat cardiovascular conditioning: something maintained through ongoing practice rather than banked permanently, noting that people who stayed active during a break from stretching kept more of their range than those who went fully sedentary.

Favro and colleagues, reviewing 36 resistance-training studies, point to exercise intensity as the strongest predictor of flexibility improvement, and suggest heavier loading recruits more motor units and produces more mechanical tension through the full range of a movement, which they propose is why some resistance protocols rival stretching for range of motion gains even though the exercises look nothing like a traditional stretch.

Murakami and colleagues, comparing calf stretching against loaded calf raises, are careful to separate the outcome, similar range of motion gains, from the mechanism, noting that only the stretching group changed passive stiffness while only the training group changed strength and muscle size, and conclude that the right tool depends on which underlying quality a person is actually trying to change.

Yagiz and colleagues, examining fascicle-length studies in eccentrically trained hamstrings, urge caution about overstating the structural story, since different measurement methods in the same body of research produce wildly different estimates of how much muscle architecture actually changes, a reminder that sarcomerogenesis, while a real phenomenon in principle, is measured with more uncertainty in humans than popular fitness explanations often suggest.

Frequently Asked Questions