A lifter comes in with left shoulder pain. It shows up in snatches, heavy pulls, sometimes deadlifts. Everything about the complaint says “shoulder” — the location, the timing, the aggravating movements. But when we watch the lift, something else stands out: on the right side, the big toe barely touches the floor, and the foot rolls outward as the bar gets heavy. The shoulder is where it hurts. The foot is where it started.
This article is about why that’s not a stretch — it’s mechanics. We’re going to walk through levers and moment arms, what your foot is actually supposed to do under load, what happens when it can’t, and why the work reliably migrates up the body and often to the opposite side. We’ll also be clear about where this reasoning is well-supported and where it’s inference, because this is an area where confident overclaiming is common.
The mechanics here are solid: levers, moment arms, and force transfer are physics, and the anatomy is well-mapped. What’s less settled is the leap from “this foot moves poorly” to “therefore this specific shoulder got injured.” Reviews of the foot-to-proximal literature describe the evidence as controversial, and note that static structural measures don’t reliably predict dynamic function.[1] So the honest position isn’t “your toe caused your shoulder pain.” It’s: the chain is real, the compensation is measurable, and it’s worth assessing rather than assuming.
Part 1: Everything Your Body Does Is Levers
Start with the physics, because it makes everything downstream obvious.
A lever is a rigid bar rotating around a pivot. In your body, bones are the bars and joints are the pivots. A moment arm is the perpendicular distance from the pivot to the line of force — and it’s the term that matters most here, because moment arm determines how much rotational force (torque) a given muscle force actually produces.
Torque = force × moment arm. That simple equation has a big implication: you can lose leverage without losing an ounce of strength. Shorten the moment arm and the same muscular effort produces less useful torque. The body then has two options — produce more force somewhere, or find a different lever. It almost always does both.
The other principle worth naming is the trade-off between long and short levers. A long lever (a straight arm, a long stride) covers more distance and can generate more speed at the end point, but it’s harder to control and imposes bigger loads on the joints holding it. A short lever (a bent arm, a tucked position) is more stable and more efficient to stabilize, but produces less reach and speed. Good movement isn’t “long levers good, short levers bad.” It’s using the right lever length at the right moment — and being able to stiffen the segments in between so force actually transfers instead of leaking.
Part 2: Your Foot Is the First Lever in the Chain
In any standing lift, the foot is where force enters and exits your body. It has a genuinely difficult job — it has to be compliant when absorbing load and stiff when transmitting it, and it has to switch between those states quickly.
The tripod
A stable foot loads through three points, forming a tripod:
The Three Points of Contact
The medial anchor. Carries a large share of load and is the pivot for push-off.
The lateral anchor. Balances the medial side and resists rolling out.
The posterior anchor. Where ground reaction force first arrives.
Like any tripod, it only works with all three legs down. Lift one and the structure doesn’t just get slightly worse — it becomes a fundamentally different, less stable shape. When the big toe loses contact, the tripod collapses to a line running from heel to the outside of the foot, and the foot’s only remaining strategy for stability is to roll onto that outer edge. That’s supination — and it’s not a random quirk, it’s the geometrically predictable result of losing the medial point.
The windlass
Here’s the mechanism that makes the big toe so disproportionately important. Described by Hicks in 1954, the windlass mechanism works like a winch: the plantar fascia runs from the heel to the toes, wrapping around the metatarsal heads. When the big toe dorsiflexes (bends up), the fascia winds around that head like a cable around a drum, pulling the heel and forefoot toward each other and raising the medial arch. That arch rise stiffens the foot and converts it from a compliant shock absorber into a stable platform to push from.[2],[3]
The classic description says the windlass turns the foot into a “rigid lever.” Newer work using biplanar videoradiography has refined that: during running, the plantar fascia behaves quasi-isometrically while the arch rises, functioning as a stable base of support rather than a truly rigid lever — and one study found engaging the windlass may increase arch compliance rather than pure rigidity.[3],[4] The practical point survives intact: hallux dorsiflexion is what stiffens the medial arch and creates a stable platform. Lose it and you lose the platform, whether or not “rigid lever” is the perfect word.
Sources: Welte et al.; Proceedings of the Royal Society B 2021 (plantar fascia extensibility and the windlass during running).Notice what the windlass requires: the big toe has to be on the ground and able to extend. A hallux that never loads, or a first MTP joint that won’t dorsiflex, means the winch never engages. Research on restricted first-MTP dorsiflexion confirms the downstream consequence — when that motion is limited, the windlass fails to initiate during gait and lower limb biomechanics are measurably modified.[5]
Part 3: Lose the Arch, Lose the Lever
This is where it stops being anatomy and starts being measurable mechanics.
A gait study induced increased foot pronation using laterally wedged sandals and measured lower-limb kinematics and kinetics. The finding, stated plainly by the authors: increased pronation reduced foot lever arm function and may compromise the ankle plantarflexion moment during stance — which may overload the knee and hip.[6]
That’s the equation from Part 1, happening in a lab. A foot that can’t hold its shape is a foot with a compromised moment arm, and a compromised moment arm means less torque produced at the ankle for the same effort.
Whether the failure looks like collapse (arch drops, foot pronates) or rolling out (foot supinates to find stability the collapsed medial side can’t provide), the outcome is the same: the foot is no longer an efficient lever, and the ankle’s contribution to the lift drops.
Part 4: The Work Has to Go Somewhere
Here’s the finding that turns a plausible story into a mechanical one. Researchers used a device emulator to experimentally vary how much push-off power the ankle contributed during gait — deliberately turning the ankle’s contribution up and down. What happened upstream?
Ankle power and hip power have an
inverse relationship.
Reduce power generation at the ankle, and hip power generation increases to compensate.[6] This wasn’t observed after the fact — the ankle’s contribution was manipulated directly and the hip’s response was measured. It’s about as clean a demonstration as this field offers that work removed from a distal joint reappears at a proximal one.
The authors of the pronation study drew the logical conclusion: individuals with increased pronation in late stance may have to increase hip moment and power to make up for the reduced ankle contribution — and over time, those elevated hip demands may contribute to hip pathology.[6]
Apply that to a heavy snatch or pull. The load doesn’t negotiate. If the right foot can’t produce its share of the drive, the right hip and trunk have to make it up — every rep, under progressively heavier loads, at higher velocities than walking. The compensation doesn’t stop at the hip.
Part 5: Why the Opposite Shoulder?
This is the part that sounds like a stretch until you look at the anatomy. Force doesn’t travel straight up your body — it travels diagonally, and the structure that carries it is called the posterior oblique sling.
The sling links three things: the gluteus maximus on one side, the thoracolumbar fascia across the low back, and the latissimus dorsi on the opposite side. The fiber orientation of these structures runs roughly perpendicular to the sacroiliac joint, so when the glute and the contralateral lat contract together they compress and stabilize the SI joint — and transfer rotational force between the lower and upper body.[7]
That’s the anatomical answer to “why the left shoulder?” The right glute max is functionally paired with the left lat dorsi. And the left lat attaches to the humerus — meaning it’s not just a back muscle, it’s a shoulder muscle. Load the right side of that diagonal abnormally and the left shoulder is where the sling terminates.
Yes, though it’s early. A 2026 study of 73 healthy participants tested whether the posterior spiral chain (lat → thoracolumbar fascia → contralateral gluteal region) behaves as a functionally connected unit. Isometric activation of the chain via right hip abduction and external rotation produced a marked redistribution of trunk rotation — significantly increasing rotation one direction and reducing it the other, with only a small change in total range.[8]
In other words: activating one end of the diagonal measurably changed motion at the other end. That’s direct evidence of functional connectivity, not just anatomical adjacency.
Source: Colonna S, Maietti G, Cuoghi F. Cureus. 2026. In vivo evidence of myofascial force transmission along the posterior spiral chain.Putting the chain together
The shoulder is the victim in that sequence. It’s the last link, so it’s where the tissue finally fails — but treating it in isolation means treating the place where the problem surfaced rather than where it originated. Which is exactly why a shoulder that’s been rehabbed three times keeps coming back.
Part 6: This Has a Name in the Literature
None of this is a fringe idea. The formal clinical model is regional interdependence, defined by Wainner and colleagues in 2007 as “the concept that seemingly unrelated impairments in a remote anatomical region may contribute to, or be associated with, the patient’s primary complaint.”[9] It was expanded by Sueki, Cleland and Wainner in 2013 into a broader neuromusculoskeletal framework, and it’s now standard in orthopedic and sports physical therapy.[10]
The model’s practical implication is that interventions directed at one region often produce effects at remote areas — which is why, for example, thoracic spine manipulation is an evidence-supported approach for shoulder pain. The mechanisms are understood to be multiple: biomechanical load transfer, neuromuscular control, neurophysiological modulation, and psychosocial factors all contribute.[10]
Regional interdependence is genuinely useful and genuinely abusable. The discipline it requires, stated well in the literature: use it to prioritize likely contributors — not to justify random remote treatments. Local examination is necessary but not sufficient; interventions should target impairments that are common, measurable, and plausibly related to the patient’s actual task limitation.
Translated to the clinic: finding a foot problem in someone with shoulder pain doesn’t automatically make the foot the cause. It makes it a candidate — one you test, treat, and retest to see whether the shoulder actually changes. If it doesn’t, the foot wasn’t the driver, and honest practice means saying so.
Where the Evidence Actually Stands
That last row matters and it’s worth sitting with: how a foot looks standing still tells you relatively little about how it behaves under a heavy bar. Which is precisely why we assess movement under load rather than reading arches in standing.
What This Means for How You Get Assessed
-
The Painful Site Gets Examined — But Not Only It
Your shoulder still gets a thorough exam. Tissue that’s been overloaded for months needs direct treatment regardless of why. But stopping there is how a problem becomes recurrent.
-
We Watch You Do the Thing That Hurts
Static assessment doesn’t reliably predict dynamic behavior. If the pain shows up in a snatch, the snatch is the test — because a foot that behaves fine standing can still collapse under 200 pounds moving fast.
-
Test, Treat, Retest
This is the discipline that separates a real kinetic-chain approach from storytelling. If we treat the foot and hip and your shoulder measurably improves on retest, the hypothesis held. If it doesn’t, we were wrong and we change direction.
-
Then You Have to Rebuild the Pattern
Restoring first-MTP motion and arch control creates the capacity for a working tripod. Loading it deliberately — and relearning to drive through the big toe under a bar — is what makes it stick.
Full Body ART, SFMA movement screening, and a whole-chain approach exist for exactly this scenario. When a shoulder keeps failing despite good local rehab, the most common reason is that the demand placed on it never changed. Treating the tissue without addressing what’s overloading it is treating the smoke.
And to be clear about the honest version: we don’t assume the foot is the culprit. We look, we test, we treat, and we retest — and the retest tells us whether we were right.
The Bottom Line
Your body moves through levers, and leverage is only as good as the platform it’s built on. Lose the big toe’s contact and you lose the windlass. Lose the windlass and the medial arch loses stiffness. Lose arch stiffness and the foot’s lever arm function drops, taking the ankle’s contribution with it. That work reappears at the hip — experimentally demonstrated, not just theorized — and travels the posterior oblique sling diagonally to the opposite lat and shoulder.
The shoulder is where it hurts. It may not be where the problem lives. And the only way to find out is to look at the whole chain — then test whether changing it actually changes your symptoms.
Shoulder Pain That Keeps Coming Back?
If you’ve rehabbed the same shoulder more than once and it keeps returning under heavy lifts, the demand on it may never have changed. Let’s assess the whole chain — from the floor up — and find what’s actually driving it.
Book an Assessment ?References
- Reviews of foot and ankle kinematics and proximal joint influence, noting the evidence is controversial and that structural variability lacks consistent support as a measure of dynamic function. Sports Health / related reviews. PMC
- Hicks JH. The mechanics of the foot: II. The plantar aponeurosis and the arch. Journal of Anatomy. 1954. (Original description of the windlass mechanism: hallux dorsiflexion winds the plantar fascia around the metatarsal heads, drawing calcaneus and metatarsals together and raising the medial longitudinal arch.)
- The extensibility of the plantar fascia influences the windlass mechanism during human running. Proceedings of the Royal Society B. 2021;288(1943):20202095. (Biplanar videoradiography; plantar fascia quasi-isometric during propulsion while arch rises; describes the foot as a stable base of support rather than a rigid lever.) Royal Society
- Welte L, et al. Influence of the windlass mechanism on arch-spring mechanics during dynamic foot arch deformation. (Found windlass engagement may increase arch compliance rather than producing pure rigidity.) PMC
- Research on first metatarsophalangeal joint dorsiflexion and windlass initiation, noting that when first-MTP dorsiflexion is restricted, windlass initiation during gait is prevented and lower limb biomechanics are modified. PMC
- Effects of foot pronation on lower limb sagittal plane biomechanics during gait. Gait & Posture. 2018. (Laterally wedged sandals induced pronation; increased pronation reduced foot lever arm function and may compromise ankle plantarflexion moment, potentially overloading knee and hip. Cites Caputo & Collins on the inverse relationship between ankle push-off power and hip power generation in late stance.) ScienceDirect
- Vleeming A, Willard F, Barker P, et al. Descriptions of the posterior oblique sling / posterior oblique subsystem: gluteus maximus, thoracolumbar fascia, and contralateral latissimus dorsi; fiber orientation perpendicular to the sacroiliac joint providing force closure and contralateral force transfer during gait and rotation.
- Colonna S, Maietti G, Cuoghi F. In vivo evidence of myofascial force transmission along the posterior spiral chain: functional connectivity linking the contralateral latissimus dorsi, thoracolumbar fascia, and gluteal region. Cureus. 2026. doi:10.7759/cureus.100760. (73 participants; isometric activation of the chain produced marked redistribution of trunk rotation.) PMC
- Wainner RS, Whitman JM, Cleland JA, Flynn TW. Regional interdependence: a musculoskeletal examination model whose time has come. Journal of Orthopaedic & Sports Physical Therapy. 2007;37(11):658-660. JOSPT
- Sueki DG, Cleland JA, Wainner RS. A regional interdependence model of musculoskeletal dysfunction: research, mechanisms, and clinical implications. Journal of Manual & Manipulative Therapy. 2013;21(2):90-102. (Expanded framework incorporating biomechanical, neurophysiological, and biopsychosocial mechanisms; emphasis on test-treat-retest.) PMC
This article is for general educational purposes and does not constitute medical advice. The clinical scenario described is a composite illustration, not a specific patient. Persistent shoulder or foot pain should be evaluated by a qualified provider.
