Place · Level 3
Mobility & flexibility
末端 ROM 力量才是真拉伸 · 训练前动态、训练后静态 · 泡沫轴是神经的事 · 久坐先练臀
Last updated
Story path
Chapter 1
Flexibility vs mobility
Flexibility vs mobility
Flexibility is the maximum angle a joint reaches passively — the position someone else can pull you into. Mobility is the maximum angle you reach actively, under your own control — the position you can move into yourself. The two often diverge:
A yoga teacher has excellent flexibility but not necessarily strong end-range ROM (chronic 'over-flexible hypermobility' can hurt the knees)Strength athletes have large active ROM (deep squat with controlled end-range), with flexibility not necessarily extremeSedentary people are poor at both, but stretching won't fix mobility — that takes strength training
The most effective way to genuinely improve mobility isn't passive stretching but end-range loaded training (Behm 2018, Frontiers Physiology). A few comparisons: a deep squat beats passive hip-flexor and ankle-dorsiflexion stretches; a Romanian deadlift emphasizing hamstring stretch beats a standing toe-touch; full-ROM bench press (bar to chest) beats a passive pec doorway stretch; an overhead squat beats passive shoulder-mobility stretches.
Why more effective: end-range work improves neural control, the brain learns to actively suppress protective reflexes at full stretch, and active ROM grows; connective tissue thickens and remodels under tension (passive stretches mainly modulate neural perception); and you train strength at the same time, reusing your training time. The practical conclusion is simple: use full ROM in strength training and about 80% of mobility issues resolve — for most people, 'stretching needs its own session' is a misconception.
A yoga teacher has excellent flexibility but not necessarily strong end-range ROM (chronic 'over-flexible hypermobility' can hurt the knees)Strength athletes have large active ROM (deep squat with controlled end-range), with flexibility not necessarily extremeSedentary people are poor at both, but stretching won't fix mobility — that takes strength training
The most effective way to genuinely improve mobility isn't passive stretching but end-range loaded training (Behm 2018, Frontiers Physiology). A few comparisons: a deep squat beats passive hip-flexor and ankle-dorsiflexion stretches; a Romanian deadlift emphasizing hamstring stretch beats a standing toe-touch; full-ROM bench press (bar to chest) beats a passive pec doorway stretch; an overhead squat beats passive shoulder-mobility stretches.
Why more effective: end-range work improves neural control, the brain learns to actively suppress protective reflexes at full stretch, and active ROM grows; connective tissue thickens and remodels under tension (passive stretches mainly modulate neural perception); and you train strength at the same time, reusing your training time. The practical conclusion is simple: use full ROM in strength training and about 80% of mobility issues resolve — for most people, 'stretching needs its own session' is a misconception.
实操 · 用哪个动作替掉哪个拉伸
把拉伸换成在末端有负荷地动, 具体是这样对应的:深蹲到底 胜过被动的髋屈肌和踝背屈拉伸 —— 蹲到底时髋和踝同时被推到范围末端, 而且是在承重、由你自己控制着下去的状态下罗马尼亚硬拉 (强调腘绳肌被拉长的那一段) 胜过站立体前屈 —— 同样是把腘绳肌拉长, 但一个是松着挂在那里, 一个是绷着控制住全程卧推 (杠铃触胸) 胜过门框被动拉胸 —— 胸大肌在它最长的位置上还要发力把杠铃推起来过头深蹲 胜过被动的肩活动度拉伸 —— 肩、胸椎、髋、踝必须同时到位, 缺一个就站不稳
为什么末端负荷比被动拉伸更有效
神经这一层: 关节被带到陌生角度时, 神经系统会保守地收着劲 —— 它没把握在那里稳得住, 于是提前踩刹车。反复在末端有控制地发力, 等于反复告诉它这里是安全的; 保护性抑制被调低, 你主动能到的角度就变大。这也解释了为什么活动度的进步常常来得比想象中快: 先松口的是神经的许可, 不是组织的长度。组织这一层: 结缔组织在张力下会增厚、按受力方向重排。这是慢的那一半, 但也是留得住的那一半。被动拉伸主要调的是前一层 (神经感觉), 所以它的效果消退得也快 —— 今天拉完能多下去两指, 一周不练又回去了。时间这一层: 力量和活动度用同一组动作一起练完, 不用额外挤出一个时段。
所以对大多数人来说, 拉伸需要单独 session 是个误区 —— 不是拉伸没用, 是它的位置被放错了。
Who actually needs dedicated flexibility work
Few people genuinely need dedicated flexibility training: specific tight areas from prolonged sitting (hip flexors, pecs, neck), with stretching paired alongside strength rather than replacing it; professions requiring extreme ROM such as gymnastics, martial arts, ballet, and yoga teaching; the ROM-rebuilding phase after injury; and aging joint-capsule and connective-tissue adhesions.Those who don't need dedicated flexibility: typical recreational runners or strength trainers, for whom full-ROM training is enough; and people who 'want to be flexible because they heard it's healthy' — normal mobility is sufficient, no need to chase extremes.
Over-flexibility is itself a problem: joint stability is impaired, raising chronic dislocation or subluxation risk (the Ehlers-Danlos hypermobility type is the textbook case); end-range weakness raises fall and injury risk. Some people have congenital hypermobility and need stability work, not more flexibility. The Beighton score self-assesses over-flexibility with 9 items (thumb to forearm, pinky past 90°, knee hyperextension past 10°, and so on); over 5/9 may suggest hypermobility syndrome and calls for a completely different approach.
Chapter 2
Static vs dynamic stretching
Static vs dynamic stretching
'Stretch before exercise' is an 80s gym-class leftover that treats two distinct things as one.
Static stretching holds a muscle at end range for 15-60 seconds, and suits post-training or a standalone flexibility session for long-term ROM — but done pre-training it acutely depresses strength. Dynamic stretching / dynamic warm-up is movement-based, simulating the training pattern with progressively larger range (leg swings, hip openers, shoulder circles, light jogging); it's mandatory before training, raising core temperature, activating neural drive, and lubricating joints, without impairing acute performance. The two names are similar but the activities are completely different.
Simic 2013 (Scand J Med Sci Sports), a meta-analysis of 104 studies, gives the numbers: after pre-training static stretching, maximal strength falls about 5.4% and explosive performance about 2.0% (the power sub-band is −1.9% and not statistically significant). Behm 2016 (Appl Physiol Nutr Metab), a systematic review, supplies the dose relationship: holding ≥ 60 s per muscle costs about 4.6% of performance, while under 60 s costs only about 1.1% (the mechanism: muscle-tendon compliance rises and force transmission efficiency drops, plus transient reflex inhibition of neural drive). Pre-training dynamic warm-up, by contrast, raises strength and power 1-2% versus no warm-up and lowers injury risk by ~50% (Olsen 2005, a multi-component warm-up in youth handball), with ROM improvement matching static stretching. Simic 2013 and Kay 2012 follow-up metas replicate the same conclusion — this is modern sports-science consensus.
So where does static stretching belong? In the post-training cool-down, 1-2 stretches per major muscle, 15-30 seconds each, for relaxation and short-term ROM; or in a standalone flexibility session on off-days or a morning/evening routine, with 30+ second holds, where long-term ROM genuinely improves (visible within a few weeks of consistent practice, Konrad 2024).
Static stretching holds a muscle at end range for 15-60 seconds, and suits post-training or a standalone flexibility session for long-term ROM — but done pre-training it acutely depresses strength. Dynamic stretching / dynamic warm-up is movement-based, simulating the training pattern with progressively larger range (leg swings, hip openers, shoulder circles, light jogging); it's mandatory before training, raising core temperature, activating neural drive, and lubricating joints, without impairing acute performance. The two names are similar but the activities are completely different.
Simic 2013 (Scand J Med Sci Sports), a meta-analysis of 104 studies, gives the numbers: after pre-training static stretching, maximal strength falls about 5.4% and explosive performance about 2.0% (the power sub-band is −1.9% and not statistically significant). Behm 2016 (Appl Physiol Nutr Metab), a systematic review, supplies the dose relationship: holding ≥ 60 s per muscle costs about 4.6% of performance, while under 60 s costs only about 1.1% (the mechanism: muscle-tendon compliance rises and force transmission efficiency drops, plus transient reflex inhibition of neural drive). Pre-training dynamic warm-up, by contrast, raises strength and power 1-2% versus no warm-up and lowers injury risk by ~50% (Olsen 2005, a multi-component warm-up in youth handball), with ROM improvement matching static stretching. Simic 2013 and Kay 2012 follow-up metas replicate the same conclusion — this is modern sports-science consensus.
So where does static stretching belong? In the post-training cool-down, 1-2 stretches per major muscle, 15-30 seconds each, for relaxation and short-term ROM; or in a standalone flexibility session on off-days or a morning/evening routine, with 30+ second holds, where long-term ROM genuinely improves (visible within a few weeks of consistent practice, Konrad 2024).
数字 · 训练前静态拉伸到底掉多少
静态拉伸的标准做法, 是把肌肉拉到末端范围保持 15-60 秒。Simic 2013 (Scand J Med Sci Sports) 的 meta 汇总了 104 项研究, 给出的数字是: 训练前静态拉伸之后, 最大力量约降 5.4%, 爆发性动作约降 2.0% (其中爆发力那一档是 −1.9%, 未达统计学显著)。Behm 2016 (Appl Physiol Nutr Metab) 的系统综述补上了剂量关系: 每块肌肉拉 ≥ 60 秒时整体表现降约 4.6%, 而拉不到 60 秒只降约 1.1%。
机制有两条, 都不神秘
肌腱被拉软了。 肌肉产生的力必须经过肌腱才能传到骨头上。肌腱像一根绳子: 绷紧的绳子拉一头另一头立刻动, 松垮的绳子要先把松弛量拉掉才开始传力。长时间静态拉伸暂时提高了肌腱的顺应性 (变得更松、更易延展), 于是同样的一次收缩, 传到骨头上的力少了一点、也慢了一点。神经驱动被压低了。 在末端长时间保持, 会通过反射通路暂时降低肌肉的主动激活程度 —— 大脑指挥肌肉发力的效率短暂下调。
反过来, 训练前做动态暖身, 力量和爆发力比不暖身高 1-2%, 受伤风险降约 50% (Olsen 2005, 青少年手球的多组件热身研究), 而关节活动范围的改善也不输静态拉伸。Simic 2013 和 Kay 2012 后续 meta 复盘出相同结论 —— 这已经是当代运动科学的共识, 不是某一派的主张。
好消息是这个下降是短暂的: 通常几分钟到十几分钟就恢复, 幅度对普通健身者的影响也有限。真正性价比为负的场景是需要爆发力的项目 (短跑、跳跃、举重) 赛前来那么一长段 —— 既没怎么防到伤, 还让你慢那么一点点。
执行 · 静态拉伸该放在什么时候
训练后冷身: 各主要肌群做 1-2 个, 每个 15-30 秒。目的是放松心情、短期改善活动范围 (ROM), 别指望它减轻第二天的酸痛。训练空档日, 或早晚的独立柔韧时段: 保持 30 秒以上, 长期 ROM 有真实改善, 坚持数周即可见 (Konrad 2024)。
为什么同一个动作换个时间就从减分变成加分
静态拉伸带来的那两件事 —— 肌腱变软、神经驱动下调 —— 在你马上要发力的场景里是纯粹的损失; 但在你今天不需要发力的场景里, 它们恰好是让组织能被推到更远、并在那里停留足够久的前提。改善柔韧需要的正是在末端待着, 而在末端待着必然伴随这两件事。
所以这不是静态拉伸好不好的问题, 而是它和接下来那件事撞不撞车的问题。撞车就往后放, 不撞车就随便做。
Chapter 3
The truth about foam rolling
The truth about foam rolling
Wiewelhove 2019 (Frontiers Physiology) meta (21 RCTs) gives the real effects of foam rolling (self-myofascial release, SMR).
What's helpful: acute ROM rises 8-12% within 10 minutes (primarily neural); DOMS subjective pain drops 5-10% short-term; light pre-training rolling doesn't suppress strength (unlike static stretching), so it can supplement a warm-up; and the subjective 'looseness' feels good, with real psychological value.
What marketing exaggerates: 'breaking up fascial adhesions' is physically impossible — fascia is connective tissue, and foam-roller pressure (~50-100 kPa) is far below what would physically break collagen crosslinks (Chaudhry 2008); 'flushing lactate' has no mechanism, since lactate was cleared long ago; 'remodeling connective tissue' isn't seen long-term; 'fat burning' is an outright lie.
The real mechanism (Behm 2019) is neural: pressure activates mechanoreceptors (Pacinian, Ruffini), reduces neural drive and raises the pain threshold, so it subjectively feels 'loose' and ROM rises short-term. This is nervous-system business, not connective-tissue business.
Usage is simple: 3-5 minutes of light pre-training rolling (30-60 seconds per muscle, pain 4-5/10, not pushed to maximum), then follow with dynamic warm-up — foam rolling doesn't replace the warm-up; 5 minutes post-training on the trained muscles; 5-10 minutes of general loosening on off-days. Avoid bony prominences (lumbar spine, anterior tibia — risk of nerve or periosteum injury) and acute injuries (use RICE on sprains/strains under 7 days). As for massage guns and vibrating rollers, Behm 2019 / Konrad 2020 show their acute effects are close to a plain roller with no significant clinical endpoints — a cheap roller used consistently beats expensive gear used occasionally; habit matters far more than equipment.
What's helpful: acute ROM rises 8-12% within 10 minutes (primarily neural); DOMS subjective pain drops 5-10% short-term; light pre-training rolling doesn't suppress strength (unlike static stretching), so it can supplement a warm-up; and the subjective 'looseness' feels good, with real psychological value.
What marketing exaggerates: 'breaking up fascial adhesions' is physically impossible — fascia is connective tissue, and foam-roller pressure (~50-100 kPa) is far below what would physically break collagen crosslinks (Chaudhry 2008); 'flushing lactate' has no mechanism, since lactate was cleared long ago; 'remodeling connective tissue' isn't seen long-term; 'fat burning' is an outright lie.
The real mechanism (Behm 2019) is neural: pressure activates mechanoreceptors (Pacinian, Ruffini), reduces neural drive and raises the pain threshold, so it subjectively feels 'loose' and ROM rises short-term. This is nervous-system business, not connective-tissue business.
Usage is simple: 3-5 minutes of light pre-training rolling (30-60 seconds per muscle, pain 4-5/10, not pushed to maximum), then follow with dynamic warm-up — foam rolling doesn't replace the warm-up; 5 minutes post-training on the trained muscles; 5-10 minutes of general loosening on off-days. Avoid bony prominences (lumbar spine, anterior tibia — risk of nerve or periosteum injury) and acute injuries (use RICE on sprains/strains under 7 days). As for massage guns and vibrating rollers, Behm 2019 / Konrad 2020 show their acute effects are close to a plain roller with no significant clinical endpoints — a cheap roller used consistently beats expensive gear used occasionally; habit matters far more than equipment.
数字 · 它到底有多少效果
Wiewelhove 2019 (Frontiers Physiology) 汇总 21 个 RCT, 给出了泡沫轴 (自我筋膜放松, SMR) 的真实效应。有效的部分
急性活动范围 (ROM) 在 10 分钟内提升 8-12% (主要是神经性的)延迟性肌肉酸痛 (DOMS) 的主观痛感短期降 5-10%训练前轻轻滚不抑制力量 (不像静态拉伸), 可作暖身补充主观松的感觉好, 心理价值真实
营销夸大的部分
打散筋膜粘连: 物理上不允许。筋膜是结缔组织, 泡沫轴那点压力 (约 50-100 kPa) 远不足以物理打散胶原交联 (Chaudhry 2008)。真要在体外把胶原交联弄断, 需要的力高到会先把皮肤和血管压坏。排出乳酸: 没有机制。乳酸在运动结束后很快就被肝脏和心肌取走用掉了, 你隔天再滚的时候它早就不在那里 —— 何况它本来也不是酸痛的原因。重塑结缔组织: 长期不可见。燃脂: 直接的谎言。局部施压不会让下面的脂肪细胞把脂肪放出来。
真实机制 (Behm 2019) 是神经性的: 压力激活皮肤和肌肉里的机械感受器, 降低神经驱动、提高痛阈, 于是主观觉得松、ROM 短期上升。
这也解释了它为什么消退得快: 被改动的是神经的设定, 而设定可以被改回去; 组织的长度从头到尾没被改过。
执行 · 怎么滚、滚多久、哪里别滚
训练前: 3-5 分钟轻度滚, 每肌群 30-60 秒, 痛度 4-5/10, 不顶到最痛。之后接动态暖身 —— 泡沫轴不替代暖身, 它只是把暖身的第一步做得舒服一点。训练后: 5 分钟, 重点滚今天练过的肌群。非训练日: 5-10 分钟整体放松。要避开的地方
骨突部位 (腰椎、胫骨前): 那里神经和骨膜就在皮下, 没有肌肉垫着, 容易压伤。急性受伤部位: 7 天内的扭拉伤用 RICE, 别滚。
关于按摩枪、振动泡沫轴这类器材
Behm 2019 / Konrad 2020 显示它们的急性效应和普通泡沫轴接近, 临床终点上没有显著差别。既然作用点在神经而不在组织, 器材能提供的更大压力并没有换来更多收益 —— 一个用得勤的便宜泡沫轴, 胜过偶尔用的贵器材。习惯比器材重要得多。
Chapter 4
Sitting & hip flexors
Sitting & hip flexors
Sit for 8 hours a day and the body changes in several ways.
The most obvious is hip-flexor shortening: the psoas major and iliacus are held chronically shortened while seated, and after 8 motionless hours the muscle treats that length as its new neutral. On standing, the psoas still wants to stay short, producing anterior pelvic tilt and lumbar hyperlordosis, and over time low back pain.
But the bigger, more overlooked problem is glute inhibition: the glutes are essentially off-duty while sitting, and the brain gradually 'forgets' to recruit them (Stuart McGill calls this glute amnesia). When standing and walking, the glute max — which should dominate hip extension — becomes functionally weak, the hamstrings compensate, and the result is tight hamstrings that strain easily. Add thoracic flexion and forward head (each 2.5 cm of forward head shift adds about 4-5 kg of perceived head weight), and chronic tension headaches and shoulder-neck stiffness come from this.
So 'sitting is terrible' isn't an exaggeration — but only stretching the hip flexors treats the symptom; the real fix is getting the glutes back to work. Ranked by benefit, high to low: stand up and move every 30-60 minutes (zero cost, about half the benefit is here); 2-3 strength sessions per week with compound lifts, focused on the glute-bridge-to-hip-thrust progression, rear-foot-elevated split squats, and Romanian deadlifts (~30%); 5 minutes a day of active-recovery movements (glute bridge, dead bug, cat-cow, doorway chest stretch, ~10%); and only then aerobic work and stretching/massage for the last sliver.
A few marketing debunks along the way: an ergonomic chair improves comfort but you're still sitting, so it doesn't fix the root (Cooley 2018); 'a standing desk is automatically healthy' is also wrong — 8 hours of standing is just as bad, and the real answer is a sit-stand desk with alternating positions; 'sitting supplements' only ease symptoms; and 'stretching solves everything' is the bigger misconception.
The most obvious is hip-flexor shortening: the psoas major and iliacus are held chronically shortened while seated, and after 8 motionless hours the muscle treats that length as its new neutral. On standing, the psoas still wants to stay short, producing anterior pelvic tilt and lumbar hyperlordosis, and over time low back pain.
But the bigger, more overlooked problem is glute inhibition: the glutes are essentially off-duty while sitting, and the brain gradually 'forgets' to recruit them (Stuart McGill calls this glute amnesia). When standing and walking, the glute max — which should dominate hip extension — becomes functionally weak, the hamstrings compensate, and the result is tight hamstrings that strain easily. Add thoracic flexion and forward head (each 2.5 cm of forward head shift adds about 4-5 kg of perceived head weight), and chronic tension headaches and shoulder-neck stiffness come from this.
So 'sitting is terrible' isn't an exaggeration — but only stretching the hip flexors treats the symptom; the real fix is getting the glutes back to work. Ranked by benefit, high to low: stand up and move every 30-60 minutes (zero cost, about half the benefit is here); 2-3 strength sessions per week with compound lifts, focused on the glute-bridge-to-hip-thrust progression, rear-foot-elevated split squats, and Romanian deadlifts (~30%); 5 minutes a day of active-recovery movements (glute bridge, dead bug, cat-cow, doorway chest stretch, ~10%); and only then aerobic work and stretching/massage for the last sliver.
A few marketing debunks along the way: an ergonomic chair improves comfort but you're still sitting, so it doesn't fix the root (Cooley 2018); 'a standing desk is automatically healthy' is also wrong — 8 hours of standing is just as bad, and the real answer is a sit-stand desk with alternating positions; 'sitting supplements' only ease symptoms; and 'stretching solves everything' is the bigger misconception.
机制 · 坐着的时候身体还发生了什么
髋屈肌为什么会把缩短的长度当成新的默认值腰大肌 (psoas major) 和髂肌 (iliacus) 在坐位时长期处于缩短状态, 8 小时不动后肌肉会把这个长度当成新的中性长度记下来。这不只是肌肉变短了那么简单: 肌肉的中性长度是神经设定的, 神经按你最常处在的那个长度去设静息张力。一天里有三分之一的时间髋是屈着的, 那个长度就成了默认值。
站立时腰大肌仍想保持缩短, 表现为骨盆前倾、腰椎过度前凸, 慢慢就是下背痛。
上半身也在同步变形
胸椎屈曲和颈前移一起来: 头每前移 2.5 cm, 主观头重感增加 4-5 kg。原因是杠杆 —— 头本身没变重, 但它离颈椎这个支点更远了, 后颈那一排肌肉必须用更大的力才拉得住。这些是姿势肌, 本来的工作方式是全天低强度待命, 现在被要求全天高强度出力, 于是慢性紧张性头痛和肩颈僵硬就这么来了。
为什么站起来动一下的收益这么高
久坐真正的问题不是某个姿势不好, 而是任何姿势维持太久都不好。肌肉要靠交替收缩放松才能把静脉血挤回心脏; 关节软骨没有血管, 只能靠反复的加载卸载像海绵一样把营养泵进去。人不动的时候, 这两条路都停着。所以频率比姿势重要 —— 换姿势本身就是治疗。
执行 · 按收益从高到低排
每 30-60 分钟站起活动一下 —— 零成本, 大约一半的收益在这里。每周 2-3 次复合动作的力量训练 —— 重点是臀桥到髋推的进阶、后腿抬高分腿蹲、直腿硬拉 (约 30% 收益)。这一档针对的是臀肌抑制, 也就是更难看见的那一半。一天 5 分钟主动恢复动作 —— 臀桥、死虫、猫牛、门框胸肌拉伸 (约 10%)。然后才是有氧和拉伸按摩, 各占最后那一点。注意这个排序里拉伸排在最后, 而它恰好是大多数人唯一在做的那一项。
顺带拆几个营销
人体工学椅: 改善舒适没问题, 但人还在坐着, 不解决根本 (Cooley 2018)。站立桌一定健康: 也不对, 站 8 小时一样糟 —— 静止站立还多一份下肢静脉负担。真正的解法是坐站交替的 sit-stand desk, 关键词是交替, 不是站。久坐补充剂: 只缓解症状, 不动那个根因。拉伸解决一切: 上面这个排序已经回答了它。
Chapter 5
The 'knees-over-toes' myth
The 'knees-over-toes' myth
'Don't let your knees go past your toes when squatting' was the standard PE-textbook slogan for 60 years; the origin is a short 1961 paper by Karl Klein, who observed college wrestlers and found a correlation between deep full-squat depth (knees past toes) and knee ligament laxity.
The problem lies in the 'correlation to causation' jump: the correlation was read as 'knees past toes = injures the knee'. Klein's sample was a wrestling team (high ROM plus frequent rapid direction changes — not extrapolable to ordinary strength training), a few dozen people, no controls, no follow-up. But coaches adopted the rule in the 1960s and it has carried through to today.
Modern evidence (Hartmann 2013, Sports Medicine systematic review, 24 studies): full-ROM squats (knees past toes) do not increase knee injury risk; conversely, partial-ROM shallow squats carry higher risk due to compensatory hip-knee shear; full-ROM squats activate more glute work while producing more thorough, healthier stimulation of the quadriceps and patellar tendon; patellofemoral peak pressure is indeed highest when the knee passes the toes, but at a physiological, normal magnitude, not a pathological one.
The cognitive shift: 'don't let the knee pass the toes' actually pins the knee in place rigidly, transferring load onto the lumbar spine and hip — and ends up hurting the back; whereas 'knee loaded through full ROM' lets hip, knee, and ankle work together, distributing force properly through the kinetic chain. The reverse-ROM training popularized by Ben Patrick (nicknamed Knees Over Toes Guy) from 2018 (ATG split squats, Reverse Nordics, backward walking) has become a significant school in knee rehab and basketball jump training, widely adopted by NBA and NFL players in post-injury rehab.
The problem lies in the 'correlation to causation' jump: the correlation was read as 'knees past toes = injures the knee'. Klein's sample was a wrestling team (high ROM plus frequent rapid direction changes — not extrapolable to ordinary strength training), a few dozen people, no controls, no follow-up. But coaches adopted the rule in the 1960s and it has carried through to today.
Modern evidence (Hartmann 2013, Sports Medicine systematic review, 24 studies): full-ROM squats (knees past toes) do not increase knee injury risk; conversely, partial-ROM shallow squats carry higher risk due to compensatory hip-knee shear; full-ROM squats activate more glute work while producing more thorough, healthier stimulation of the quadriceps and patellar tendon; patellofemoral peak pressure is indeed highest when the knee passes the toes, but at a physiological, normal magnitude, not a pathological one.
The cognitive shift: 'don't let the knee pass the toes' actually pins the knee in place rigidly, transferring load onto the lumbar spine and hip — and ends up hurting the back; whereas 'knee loaded through full ROM' lets hip, knee, and ankle work together, distributing force properly through the kinetic chain. The reverse-ROM training popularized by Ben Patrick (nicknamed Knees Over Toes Guy) from 2018 (ATG split squats, Reverse Nordics, backward walking) has become a significant school in knee rehab and basketball jump training, widely adopted by NBA and NFL players in post-injury rehab.
数字 · 现代证据怎么说
Hartmann 2013 (Sports Medicine 系统综述, 24 个研究) 的结论有四条:全活动范围 (ROM) 深蹲 (膝过脚尖) 不增加膝盖损伤风险。反过来, 部分 ROM 的浅蹲因为代偿性的髋膝剪切, 风险反而更高。全 ROM 深蹲激活更多臀肌, 同时对股四头肌和髌腱产生更全程、更健康的刺激。髌股关节压力峰值出现在膝屈约 90° 的位置, 而不是在最深处 —— 继续往下蹲, 大腿后侧和小腿贴合 (wrapping effect) 把负荷摊开、接触面也随之变大, 髌股压力反而降下来。这一条恰好和越深越伤膝的直觉相反, 也正是这条迷思一直听起来有道理的地方: 压力确实会先上去, 只是它在你最怕的那个位置已经在往下走了。。
最后一条值得多看一眼, 因为它正是这条迷思听起来有道理的来源。压力确实上去了, 这一点 Klein 那一代人没看错; 看错的是压力大等于会坏这个默认。组织对压力的反应是双向的: 在生理范围内, 压力是让软骨和肌腱变强的信号, 不是磨损它的力。真正会坏的情形是压力突然超出这个范围, 或者组织长期没有被加载过。所以躲开这个角度反而让它更弱, 下次被迫承受时更容易出事。
观念上的转变: Ben Patrick (绰号 Knees Over Toes Guy) 从 2018 年起推广的反向 ROM 训练 (ATG 分腿蹲、Reverse Nordic、后倾走), 现在已经成为膝伤康复和篮球跳跃训练的一个重要流派, NBA、NFL 球员伤后康复大量采用 —— 路线从躲开这个角度换成了在这个角度上练出力量。
What you should do
For typical trainees: squat to parallel or deeper, don't intentionally restrict ROM; knees passing toes is normal, don't over-shift weight backward to keep them behind; slow eccentric and end-range control beat bounce velocity; if you have no knee pain, don't change your form — 'fixing what isn't broken' carries more risk than benefit.For those with knee pain: don't avoid ROM — full ROM with appropriate load is a rehab tool, not the cause; check compensation patterns, like weak hip abductors producing knee valgus (cave-in), paired with glute activation; start ATG split squats from bodyweight and add load gradually over 12 weeks; if pain persists, get a detailed differential.
For coaches and fitness professionals: stop saying 'knees don't pass toes' — it's an outdated slogan and modern evidence has reversed it; replace it with 'keep knee aligned with toes' plus 'no knee valgus (caving in)', the cues that are actually safety-relevant; teach full ROM, not partial.
References · 10
- Behm, D. G. (2018). The Science and Physiology of Flexibility and Stretching: Implications and Applications in Sport Performance and Health. Routledge. Comprehensive textbook differentiating passive flexibility from active mobility; end-range loaded strength training, not static stretch, drives durable range-of-motion change.
- Hartmann, H., Wirth, K., & Klusemann, M. (2013). Analysis of the load on the knee joint and vertebral column with changes in squatting depth and weight load. Sports Medicine, 43(10), 993-1008. Full / deep squats do not increase knee or spine injury risk vs. partial squats when loaded progressively; the 'knees over toes' admonition is biomechanically unsupported. 10.1007/s40279-013-0073-6
- Konrad, A., Alizadeh, S., Daneshjoo, A., Anvar, S. H., Graham, A., Zahiri, A., et al. (2024). Chronic effects of stretching on range of motion with consideration of potential moderating variables: A systematic review with meta-analysis. Journal of Sport and Health Science, 13(2), 186-194. Meta-analysis of 77 studies: regular stretch training produces prolonged, moderate increases in joint ROM; static and PNF outperform ballistic/dynamic for long-term ROM. 10.1016/j.jshs.2023.06.002
- Behm, D. G., Blazevich, A. J., Kay, A. D., & McHugh, M. (2016). Acute effects of muscle stretching on physical performance, range of motion, and injury incidence in healthy active individuals: A systematic review. Applied Physiology, Nutrition, and Metabolism, 41(1), 1-11. Pre-exercise static stretching > 60 s reduces strength ~5% and power ~3%; dynamic stretching is neutral or beneficial. 10.1139/apnm-2015-0235
- Simic, L., Sarabon, N., & Markovic, G. (2013). Does pre-exercise static stretching inhibit maximal muscular performance? A meta-analytical review. Scandinavian Journal of Medicine & Science in Sports, 23(2), 131-148. Pre-exercise static stretching reduces strength by 5.4% and power by 1.9% on average. 10.1111/j.1600-0838.2012.01444.x
- Herbert, R. D., de Noronha, M., & Kamper, S. J. (2011). Stretching to prevent or reduce muscle soreness after exercise. Cochrane Database of Systematic Reviews, 2011(7), CD004577. 12 RCTs: no clinically meaningful effect of stretching (before or after) on DOMS. 10.1002/14651858.CD004577.pub3
- Wiewelhove, T., Döweling, A., Schneider, C., Hottenrott, L., Meyer, T., Kellmann, M., Pfeiffer, M., & Ferrauti, A. (2019). A meta-analysis of the effects of foam rolling on performance and recovery. Frontiers in Physiology, 10, 376. Foam rolling produces small acute ROM and recovery-perception benefits; the 'fascial release' mechanism is implausible — effects are neural. 10.3389/fphys.2019.00376
- Davis, H. L., Alabed, S., & Chico, T. J. A. (2020). Effect of sports massage on performance and recovery: A systematic review and meta-analysis. BMJ Open Sport & Exercise Medicine, 6(1), e000614. Modest improvements in flexibility and DOMS perception; no consistent effect on strength or power performance. 10.1136/bmjsem-2019-000614
- Schoenfeld, B. J. (2010). The mechanisms of muscle hypertrophy and their application to resistance training. Journal of Strength and Conditioning Research, 24(10), 2857-2872. 10.1519/JSC.0b013e3181e840f3
- Suchomel, T. J., Nimphius, S., & Stone, M. H. (2016). The importance of muscular strength in athletic performance. Sports Medicine, 46(10), 1419-1449. Reviews 200+ studies; muscular strength underpins virtually every athletic performance attribute (power, speed, change of direction) and reduces injury risk. 10.1007/s40279-016-0486-0