System · Level 3
Joints — The Cartilage & Tendon Economy
软骨无血管 · 加载喂养 · 肌腱核心终身少更新 · OARSI/ACR 首选 = 运动 + 减重 + 外用 NSAID
Story path
- 1What a joint actually is — seven partsWhat a joint actually is — seven parts
- 2Cartilage economics — 'use it or lose it' is literalCartilage economics — 'use it or lose it' is literal
- 3Tendon economics — the tendon you grew up with is the one you keepTendon economics — the tendon you grew up with is the one you keep
- 4Why athletes' joints don't fall apartWhy athletes' joints don't fall apart
- 5Levers that actually work — what the guidelines sayLevers that actually work — what the guidelines say
- 6Hype-grade interventions — injections + supplementsHype-grade interventions — injections + supplements
- 7Recipes by joint — knee / hip / shoulder / ankle / low backRecipes by joint — knee / hip / shoulder / ankle / low back
Chapter 1
What a joint actually is — seven parts
What a joint actually is — seven parts
A joint is not one thing — it is a team of tissues, and pain shows up when any of them falls behind:
Articular cartilage: a 2-4 mm layer of hyaline cartilage capping the bone ends — type II collagen + GAGs (glycosaminoglycans) + ~80 % water — avascular, aneural, alymphatic (Sophia Fox 2009 *Sports Health*). Its compressive resilience and near-frictionless surface come entirely from the water content + GAG network locking that water in placeSynovium + synovial fluid: a thin synovial-cell layer secretes hyaluronic acid (HA) + lubricin (PRG4); the fluid is shear-thinning — viscous at rest, slippery under load (Jay & Waller 2014 *Matrix Biology*)Joint capsule + ligaments: dense type I collagen — binds bones together, limits extreme anglesTendons: transmit muscle force to bone; type I collagen + a little elastinMenisci / labra (knee / hip / shoulder): fibrocartilage — distributes load + enlarges the contact surfaceBursae: reduce friction between tendons / skin and boneProprioceptive nerve endings: live in the capsule + ligaments, telling the brain where the joint is in space
The first thing to hold onto: of those seven, only the synovium, capsule, and ligaments are vascular. Cartilage and the inner meniscus eat through load + diffusion. The rest of this island unfolds from that fact.
Articular cartilage: a 2-4 mm layer of hyaline cartilage capping the bone ends — type II collagen + GAGs (glycosaminoglycans) + ~80 % water — avascular, aneural, alymphatic (Sophia Fox 2009 *Sports Health*). Its compressive resilience and near-frictionless surface come entirely from the water content + GAG network locking that water in placeSynovium + synovial fluid: a thin synovial-cell layer secretes hyaluronic acid (HA) + lubricin (PRG4); the fluid is shear-thinning — viscous at rest, slippery under load (Jay & Waller 2014 *Matrix Biology*)Joint capsule + ligaments: dense type I collagen — binds bones together, limits extreme anglesTendons: transmit muscle force to bone; type I collagen + a little elastinMenisci / labra (knee / hip / shoulder): fibrocartilage — distributes load + enlarges the contact surfaceBursae: reduce friction between tendons / skin and boneProprioceptive nerve endings: live in the capsule + ligaments, telling the brain where the joint is in space
The first thing to hold onto: of those seven, only the synovium, capsule, and ligaments are vascular. Cartilage and the inner meniscus eat through load + diffusion. The rest of this island unfolds from that fact.
Why synovial fluid doesn't freeze up — the shear-thinning trick
"Why does my knee crack on the first step after sitting, then settle after a few paces?"Synovial fluid is non-Newtonian: at rest, hyaluronic-acid (HA) long chains tangle and viscosity is high (gel-like); under load and shear the chains align along the motion direction, viscosity collapses, and a near-frictionless film forms between the cartilage facesLubricin (PRG4) sits on the cartilage surface and provides boundary lubrication — even at near-contact pressures the surfaces still glide (Jay & Waller 2014)That is why the first step out of a chair feels stiff and clears within 30 s: the HA chains haven't aligned yet
Clinical implications:
Joints that pop or click ≠ joints breaking; most painless pops are HA rearrangement + cavitation bubbles (no treatment needed)"My joints need lubrication" is not solved by swallowing it: oral hyaluronate barely reaches the joint cavity; intra-articular HA injection trials show mixed effects (Cochrane 2015), not a first-line treatmentThe cheapest way to make synovial fluid slippery again is to move — 5-10 min of walking beats any supplement
Chapter 2
Cartilage economics — 'use it or lose it' is literal
Cartilage economics — 'use it or lose it' is literal
Cartilage has no blood supply. How does it stay alive? Through repeated load → unload → reload cycles that squeeze nutrients from synovial fluid into the matrix and squeeze waste back out (Sophia Fox 2009).
This is why immobilization is cartilage's real enemy:
21 days of bedrest significantly shifts serum cartilage biomarkers (COMP, CTX-II) — Liphardt 2020 *J Orthop Res*Spinal cord injury and prolonged casting both produce measurable cartilage thinning within 6-12 weeksAfter spaceflight or long bedrest, cartilage recovery on return to 1 g takes months
Does that mean more is always better? No — it is a U-shape:
Lin 2013 OAI (n = 205, 4-year 3 T MRI): moderate-activity people had the slowest cartilage T2 progression; both very-low and very-high activity were worse**Roos & Dahlberg 2005 *Arthritis & Rheumatism*** (4-month RCT, at-risk knees): the exercise arm's dGEMRIC GAG index sat significantly higher than the non-exercising controls — not so much regrowing cartilage as maintaining or improving while controls declinedAlentorn-Geli 2017 JOSPT meta (17 studies, n = 114,829): hip/knee OA prevalence was 3.5 % in recreational runners, 10.2 % in sedentary controls, 13.3 % in elite competitive runners — running is not the wear-and-tear story
Hold on to this: cartilage doesn't wear out, it starves. Loading is its lunch, not its enemy.
This is why immobilization is cartilage's real enemy:
21 days of bedrest significantly shifts serum cartilage biomarkers (COMP, CTX-II) — Liphardt 2020 *J Orthop Res*Spinal cord injury and prolonged casting both produce measurable cartilage thinning within 6-12 weeksAfter spaceflight or long bedrest, cartilage recovery on return to 1 g takes months
Does that mean more is always better? No — it is a U-shape:
Lin 2013 OAI (n = 205, 4-year 3 T MRI): moderate-activity people had the slowest cartilage T2 progression; both very-low and very-high activity were worse**Roos & Dahlberg 2005 *Arthritis & Rheumatism*** (4-month RCT, at-risk knees): the exercise arm's dGEMRIC GAG index sat significantly higher than the non-exercising controls — not so much regrowing cartilage as maintaining or improving while controls declinedAlentorn-Geli 2017 JOSPT meta (17 studies, n = 114,829): hip/knee OA prevalence was 3.5 % in recreational runners, 10.2 % in sedentary controls, 13.3 % in elite competitive runners — running is not the wear-and-tear story
Hold on to this: cartilage doesn't wear out, it starves. Loading is its lunch, not its enemy.
Why sitting still is cartilage's chronic illness
"I sit 12 h a day — what about my joints?"The issue is not seated pressure (joint pressures sitting are actually moderate). It is the absence of load-unload cyclingUnder static pressure, chondrocytes shift toward catabolic expression: MMP-13 and ADAMTS-5 are up-regulated, type II collagen and aggrecan synthesis go downThe radiographic OA burden in lifelong-sedentary adults is therefore not lower — sometimes it is higher — because the cycle they need never happens
Practical:
Stand and move for 1-2 min every 30-45 min: this is a free cartilage-feeding signal, nothing to do with calorie burnAim for 5,000-10,000 steps a day total: this is the middle of Lin 2013's U-curve where cartilage markers are most stableUse full ROM in your training day: full-depth squats, lunges, hip hinges all feed cartilage better than half-range workAvoid the two extremes: complete immobility (bedrest / pure desk work) and high-volume high-impact (80+ km/week heavy running) both sit on the falling ends of the U
Chapter 3
Tendon economics — the tendon you grew up with is the one you keep
Tendon economics — the tendon you grew up with is the one you keep
**Heinemeier 2013 *FASEB J*** used nuclear-bomb ¹⁴C dating on 28 adult Achilles tendons to measure the *age* of the core collagen. The verdict: core Achilles tendon collagen is laid down during the body's height growth (roughly birth to ~17 years), and afterward shows essentially no renewal.
Important framing:
This applies to the load-bearing core fibrils, not the whole tendon. Intracellular proteins, the peripheral tendon, and the peritenon still turn over normallyBut the bulk of the load-bearing collagen — what you have as an adult — is what you keep for life
So how does an injured tendon recover? Not by "replacing collagen" — but by upregulating matrix synthesis around the bearing fibres + neural-loop remodelling:
**Heavy eccentric loading (Alfredson 1998, *AJSM*, n = 15)**: 3 × 15 reps of straight-knee + bent-knee heel drops, twice a day, for 12 weeks, returned all 15 chronic Achilles tendinopathy patients to sport. Later RCTs confirm eccentric > concentric, but with response rates closer to 60-80 % rather than "every patient cured"Heavy Slow Resistance (Kongsgaard 2009): patellar tendinopathy RCT showing HSR is equivalent to eccentric at 12 weeks, with higher patient satisfaction and better long-term adherence on just 3 sessions/weekGrowth hormone is a known stimulus for tendon collagen synthesis (Doessing 2010 *J Physiol*): 14 days of GH raised tendon collagen synthesis ~6-fold; GH releases mostly during NREM3 deep sleep (Van Cauter 2000). The full chain "sleep deprivation → impaired tendon repair" hasn't been tested end-to-end as a primary endpoint in humans — but every individual link is RCT-grade. Treat it as a mechanism-grounded inference, not marketing
One line: the rebar in your tendon you can't replace, but the cement (matrix synthesis around it) you can re-train. Eccentric or HSR + protein + deep sleep — that bundle has real RCT evidence, not vibes.
Important framing:
This applies to the load-bearing core fibrils, not the whole tendon. Intracellular proteins, the peripheral tendon, and the peritenon still turn over normallyBut the bulk of the load-bearing collagen — what you have as an adult — is what you keep for life
So how does an injured tendon recover? Not by "replacing collagen" — but by upregulating matrix synthesis around the bearing fibres + neural-loop remodelling:
**Heavy eccentric loading (Alfredson 1998, *AJSM*, n = 15)**: 3 × 15 reps of straight-knee + bent-knee heel drops, twice a day, for 12 weeks, returned all 15 chronic Achilles tendinopathy patients to sport. Later RCTs confirm eccentric > concentric, but with response rates closer to 60-80 % rather than "every patient cured"Heavy Slow Resistance (Kongsgaard 2009): patellar tendinopathy RCT showing HSR is equivalent to eccentric at 12 weeks, with higher patient satisfaction and better long-term adherence on just 3 sessions/weekGrowth hormone is a known stimulus for tendon collagen synthesis (Doessing 2010 *J Physiol*): 14 days of GH raised tendon collagen synthesis ~6-fold; GH releases mostly during NREM3 deep sleep (Van Cauter 2000). The full chain "sleep deprivation → impaired tendon repair" hasn't been tested end-to-end as a primary endpoint in humans — but every individual link is RCT-grade. Treat it as a mechanism-grounded inference, not marketing
One line: the rebar in your tendon you can't replace, but the cement (matrix synthesis around it) you can re-train. Eccentric or HSR + protein + deep sleep — that bundle has real RCT evidence, not vibes.
How to actually run Alfredson / HSR protocols
Alfredson Achilles protocol (chronic tendinopathy — not acute rupture):Stand at a step edge on the injured foot's forefoot, use the uninjured leg to assist lifting up → slowly drop the heel below the step on the injured side → use the uninjured leg to return up (no concentric phase on the injured side)3 × 15 straight-knee + 3 × 15 bent-knee = 90 reps per sessionTwice daily = 180 reps/dayMaintain for 12 weeks; mild pain during the exercise is allowed ("not without pain" — pain that doesn't worsen is OK)After it stops hurting, progressively add load (weighted backpack)For acute rupture, complete tear, or red-hot-swollen presentation — stop and see a clinician
Heavy Slow Resistance (Kongsgaard patellar protocol):
Barbell half-squat / leg extension / leg press, 4 sets × 6 reps per exercise, 3 s eccentric + 3 s concentric (slow)Start at 15 RM (progress to 6 RM by week 6)3 sessions per week, 12 weeksFar less daily volume than Alfredson, with equivalent outcomes — easier to stick to
Neither is "just do it": get the first 4-6 weeks supervised by a physio or sports-med clinician. Acute injury and chronic tendinopathy are different conditions — acute = ice + temporary unloading; chronic is where these loading protocols belong.
Chapter 4
Why athletes' joints don't fall apart
Why athletes' joints don't fall apart
The standard misconception: "athletes wreck their joints — knees first." The data point the other way.
Alentorn-Geli 2017 JOSPT meta (17 studies, n = 114,829): hip/knee OA prevalence — recreational runners 3.5 %, sedentary controls 10.2 %, elite competitive runners 13.3 %. Recreational running is a *protective* signal, not a wear-and-tear oneQuadriceps weakness is a longitudinal risk factor for knee OA: Slemenda 1997 *Ann Intern Med* cohort + Roos & Arden 2016 review. Conversely, strong quads + surrounding stabilisers correlate with slower progression.Training's direct effect on cartilage morphology is modest, but unloading (immobilisation, bedrest, spinal-cord injury) reliably thins cartilage (Eckstein 2006 *J Anat*). The use-it-or-lose-it asymmetry is real — disuse damages faster than re-use rebuilds.Athlete ≠ broken joints. The real OA risks are prior joint injury, BMI, and inactivity — not weekly mileage
So what do athletes actually get right? (consensus across sports-medicine reviews):
1. Lifelong loading: cartilage and tendon are continuously told "you are needed", matrix metabolism stays on the synthesis side
2. Strong quads + hip abductors: the joint itself doesn't absorb the random-direction shocks — the muscles do
3. Sharp proprioception: the neural loop catches an ankle roll or fall *before* tissue tears
4. Body composition under control: every extra kilogram is ~3-4× force at the knee during landing
5. Sleep: GH release lives in NREM3 — and that is the tendon collagen synthesis signal
6. Full ROM use: they enter end-range; they don't live in the protective grey middle
Translation for non-athletes: you don't need to become an elite athlete. Picking just 2-3 of those six already moves you toward the bottom of the OA U-curve.
Alentorn-Geli 2017 JOSPT meta (17 studies, n = 114,829): hip/knee OA prevalence — recreational runners 3.5 %, sedentary controls 10.2 %, elite competitive runners 13.3 %. Recreational running is a *protective* signal, not a wear-and-tear oneQuadriceps weakness is a longitudinal risk factor for knee OA: Slemenda 1997 *Ann Intern Med* cohort + Roos & Arden 2016 review. Conversely, strong quads + surrounding stabilisers correlate with slower progression.Training's direct effect on cartilage morphology is modest, but unloading (immobilisation, bedrest, spinal-cord injury) reliably thins cartilage (Eckstein 2006 *J Anat*). The use-it-or-lose-it asymmetry is real — disuse damages faster than re-use rebuilds.Athlete ≠ broken joints. The real OA risks are prior joint injury, BMI, and inactivity — not weekly mileage
So what do athletes actually get right? (consensus across sports-medicine reviews):
1. Lifelong loading: cartilage and tendon are continuously told "you are needed", matrix metabolism stays on the synthesis side
2. Strong quads + hip abductors: the joint itself doesn't absorb the random-direction shocks — the muscles do
3. Sharp proprioception: the neural loop catches an ankle roll or fall *before* tissue tears
4. Body composition under control: every extra kilogram is ~3-4× force at the knee during landing
5. Sleep: GH release lives in NREM3 — and that is the tendon collagen synthesis signal
6. Full ROM use: they enter end-range; they don't live in the protective grey middle
Translation for non-athletes: you don't need to become an elite athlete. Picking just 2-3 of those six already moves you toward the bottom of the OA U-curve.
Debunking 'my knees crack / I'm too old to lift'
**Fiatarone 1994 *NEJM*: 86-96 year olds, 8 weeks of progressive resistance training, +174 % quadriceps strength**, +48 % gait speed, more spontaneous activity — joints and tendons followed; zero dropouts.Hartmann 2013 squat meta: with correct technique, squats below parallel are neutral-to-protective for knee structure, not damaging (covered in detail in `knees-over-toes-myth`).
Three takeaways:
Joints that crack but don't hurt = generally nothing to fix (HA-network rearrangement + cavitation + tendon glide) — see `joints/anatomy` page 1"I'm too old to train" is a cultural narrative, not a biological one. Starting resistance training at 80+ still produces large strength + joint-function gainsThe real risk triad to watch for is prior joint injury + chronic inactivity + high BMI — that combination, not weekly mileage
Atlas cross-links:
Cartilage anatomy depth → this island"Do chondroitin / glucosamine really work?" → `glucosamine-chondroitin`"Eat collagen, grow collagen?" + Vitamin-C cofactor mechanism → `collagen-peptides`Resistance training in older adults → `elderly-resistance-training`Running injuries → `running-injuries`
Chapter 5
Levers that actually work — what the guidelines say
Levers that actually work — what the guidelines say
OARSI 2019 (Bannuru et al. *Osteoarthritis and Cartilage*) and ACR/AF 2019 (Kolasinski et al. *Arthritis & Rheumatology*) — two independent international guidelines — converge on the first-line management of knee OA:
Core / Strongly recommended:
Structured land-based exercise: resistance training + cardio + ROM — A-grade across every branch of the evidenceWeight loss (if overweight/obese): Felson 1992 *Ann Intern Med* Framingham cohort — women who sustained ~5 kg weight loss had ~50 % lower 10-year risk of symptomatic knee OA; Messier 2013 IDEA *JAMA* RCT (n = 454, 18 months) showed diet + exercise produced larger WOMAC pain reduction (−3.0) than diet alone (−1.8) or exercise alone (−1.7)Topical NSAID: diclofenac patch/gel — minimal systemic side effects, first choice in older adultsSelf-management education + tai chi: both in ACR's strong-recommendation tierShort courses of oral NSAID: ACR strong, OARSI conditionalShort-term intra-articular corticosteroid: real short-term pain relief — but don't repeat (see scene 6)
Conditional:
Intra-articular HA injection: mixed evidenceCane / walker / patellar taping: individualised
**For tendinopathy — which is *not* OA**:
Alfredson eccentric or Kongsgaard HSR — see scene 3**Rio 2015 *BJSM* isometric**: 5 × 45-second leg-extension holds produced ≥ 45 min of immediate analgesia (n = 6; replication is more modest, but it remains a useful acute pain-modulation tool)
Bottom line: the real levers for joint pain are not supplements. They are movement, weight loss, short-term topical NSAID, and targeted strength rehab — for most people, these four cover 80 % of the problem.
Core / Strongly recommended:
Structured land-based exercise: resistance training + cardio + ROM — A-grade across every branch of the evidenceWeight loss (if overweight/obese): Felson 1992 *Ann Intern Med* Framingham cohort — women who sustained ~5 kg weight loss had ~50 % lower 10-year risk of symptomatic knee OA; Messier 2013 IDEA *JAMA* RCT (n = 454, 18 months) showed diet + exercise produced larger WOMAC pain reduction (−3.0) than diet alone (−1.8) or exercise alone (−1.7)Topical NSAID: diclofenac patch/gel — minimal systemic side effects, first choice in older adultsSelf-management education + tai chi: both in ACR's strong-recommendation tierShort courses of oral NSAID: ACR strong, OARSI conditionalShort-term intra-articular corticosteroid: real short-term pain relief — but don't repeat (see scene 6)
Conditional:
Intra-articular HA injection: mixed evidenceCane / walker / patellar taping: individualised
**For tendinopathy — which is *not* OA**:
Alfredson eccentric or Kongsgaard HSR — see scene 3**Rio 2015 *BJSM* isometric**: 5 × 45-second leg-extension holds produced ≥ 45 min of immediate analgesia (n = 6; replication is more modest, but it remains a useful acute pain-modulation tool)
Bottom line: the real levers for joint pain are not supplements. They are movement, weight loss, short-term topical NSAID, and targeted strength rehab — for most people, these four cover 80 % of the problem.
Why glucosamine + chondroitin aren't on the first-line list
ACR 2019 gives a strong recommendation against glucosamine alone or combined with chondroitin for knee / hip OA. OARSI 2019 likewise does not recommend them as core management for knee OA.Why are the guidelines this firm?
**GAIT 2006 *NEJM* (n = 1,583): the primary endpoint (≥ 20 % improvement on WOMAC pain at 24 weeks) was overall negative; a prespecified moderate-to-severe pain subgroup (n = 354) showed combination 79 % vs placebo 54 % (P = 0.002), but the 2-year extension (Sawitzke 2008/2010) and the X-ray joint-space-width structural endpoint were both null** — guideline bodies therefore treat the subgroup as exploratory, not basis for recommendationWandel 2010 BMJ meta (10 RCTs, n > 3,800): glucosamine + chondroitin in any combination was clinically not effective (pain reduction below the minimum clinically important difference)
"Isn't European prescription-grade glucosamine sulfate different?"
The Rotta-formulation prescription sulfate showed some signal in older European RCTs, but more recent higher-quality meta-analyses (LEGS trial, Roman-Blas 2017) have pulled that signal back to near zeroIf you are already taking it, it won't harm you, but don't expect structural disease modification and don't let it crowd out exercise + weight management — see the `glucosamine-chondroitin` decision tree
On "collagen peptides fix joints": see `collagen-peptides` for the full debunk. Briefly: some RCTs (Clark 2008) show subjective exercise-related joint discomfort improvement, but clinical guidelines do not list collagen peptides as first-line for OA. The group where collagen peptides have the strongest evidence is high-load connective-tissue athletes (collagen + Vitamin C before jumping/landing training) — not general OA patients.
Chapter 6
Hype-grade interventions — injections + supplements
Hype-grade interventions — injections + supplements
"Premium" joint-pain treatments all sound high-tech, but the RCTs disagree. This scene body covers oral + repeated injection (①②③); page 2 covers PRP / HA / immobilisation / patches (④⑤⑥⑦).
① Glucosamine + chondroitin
GAIT primary negative + Wandel 2010 *BMJ* network meta null — covered in scene 5 + `glucosamine-chondroitin` in full
② Collagen peptides / "type II collagen"
`collagen-peptides` already debunks the "eat-it-to-grow-it" intuition. With Vitamin C in athletes there is some connective-tissue-synthesis signal (Shaw 2017 *Am J Clin Nutr*), but not a standard OA therapy — no major guideline lists it as first-line
③ Repeated intra-articular corticosteroid
**McAlindon 2017 *JAMA*: n = 140 knee OA, triamcinolone vs saline every 12 weeks for 2 years. The steroid arm lost more cartilage volume (−0.21 vs −0.10 mm) with no advantage on pain**A single short-term injection is still fine for an acute flare; repeated dosing doesn't just fail to help — it accelerates structural loss
Practical filter — when you hear about any "premium" option, ask two questions:
1. Where does it sit in OARSI 2019 / ACR 2019: strong / conditional / not-recommended / against?
2. Was the largest RCT (≥ 200 patients, ≥ 12-month follow-up) positive on its primary endpoint?
If neither is yes, it is not an "upgrade" — it is marketing.
① Glucosamine + chondroitin
GAIT primary negative + Wandel 2010 *BMJ* network meta null — covered in scene 5 + `glucosamine-chondroitin` in full
② Collagen peptides / "type II collagen"
`collagen-peptides` already debunks the "eat-it-to-grow-it" intuition. With Vitamin C in athletes there is some connective-tissue-synthesis signal (Shaw 2017 *Am J Clin Nutr*), but not a standard OA therapy — no major guideline lists it as first-line
③ Repeated intra-articular corticosteroid
**McAlindon 2017 *JAMA*: n = 140 knee OA, triamcinolone vs saline every 12 weeks for 2 years. The steroid arm lost more cartilage volume (−0.21 vs −0.10 mm) with no advantage on pain**A single short-term injection is still fine for an acute flare; repeated dosing doesn't just fail to help — it accelerates structural loss
Practical filter — when you hear about any "premium" option, ask two questions:
1. Where does it sit in OARSI 2019 / ACR 2019: strong / conditional / not-recommended / against?
2. Was the largest RCT (≥ 200 patients, ≥ 12-month follow-up) positive on its primary endpoint?
If neither is yes, it is not an "upgrade" — it is marketing.
Injections, rest, patches — the other four
④ Intra-articular PRP (platelet-rich plasma)**Bennell 2021 *JAMA* RESTORE (n = 288, 12 months, knee OA): PRP vs saline showed no difference in pain or cartilage volume**In high-quality blinded RCTs, PRP no longer beats placebo; early small-sample "positive" trials were mostly driven by lack of blinding + amplified placebo effect
⑤ Intra-articular hyaluronate (HA) injection
Cochrane meta: small short-term effect, uncertain long-term. Not first-line; sits in the conditional tier of every major guideline
⑥ Long-term joint rest
RICE applies for ~48-72 h after acute injury. After that, early movement is better for cartilage + tendon recovery (scenes 2 + 3). Prolonged immobilisation makes things worse, not better"My joint is broken so I should move less" inverts the causality — disuse is precisely what starves cartilage
⑦ Joint patches / infrared lamps / pellets / magnets
Most lack A- or B-grade RCT support. A few (e.g., kinesio tape) may produce short-term perceived analgesia, but that is not the same as "treating OA"
One more under-flagged "doesn't work" — arthroscopic surgery for degenerative meniscal tears:
**Sihvonen 2013 *NEJM* FIDELITY (n = 146): arthroscopic partial meniscectomy vs sham surgery** for degenerative tears — no difference in pain or function at 1 yearThe standard answer for non-traumatic degenerative meniscal tears is 12 weeks of exercise rehab, not surgery. If a PCP refers you straight to ortho for surgery, ask whether 12 weeks of conservative care could be tried first.
Chapter 7
Recipes by joint — knee / hip / shoulder / ankle / low back
Recipes by joint — knee / hip / shoulder / ankle / low back
Universal framework (applies to any joint):
1. Load: resistance training 2-3 ×/week + full ROM
2. Proprioception: single-leg stance / balance board / unstable surface
3. Body composition: BMI control beats any supplement (Felson Framingham)
4. Sleep: let GH release in NREM3 (Van Cauter 2000) — an all-nighter hurts more than just the brain
5. See a clinician if pain persists > 1 week
By joint (core movements, 2-3 ×/week, not every day):
Knee: goblet squat / Bulgarian split squat / lunges (full depth) + single-leg deadlift + terminal knee extension; add ATG knees-over-toes work (see `knees-over-toes-myth`)Hip: hip hinges (RDL, good morning) + hip abduction (clamshell, side-lying leg raise) + internal/external rotation (90/90 stretch). For desk workers, tight hip flexors + weak hip extensors is the #1 thing to fixShoulder: scapular retraction + depression + external-rotation drills (Y-T-W) + full push + pull (overhead press + pull-up). Don't only train the front delt — the posterior chain is the stability driverAnkle: single-leg calf raises + eccentric heel drops (do them even pain-free) + single-leg balance (30 s × 3); warm up before running or court sportsLow back: McKenzie extensions / hip hinge / dead bug + bird dog + side plank (Stuart McGill's three) + hip abduction. Avoid isolated spinal-extension work as the main strategy; what the low back actually needs is abdominal + hip + spinal co-contraction — the stiffness-sandwich pattern McGill describes
Red lines (any joint — stop training, see a clinician):
Acute swelling, redness, heat: possible infection / gout flare / acute crystal arthritis — see atlas `gout`Joint locking / inability to bear weight after an injury: possible meniscus or ligament injuryBilateral, symmetric small-joint pain (finger MCPs, wrist) + morning stiffness > 1 h + fatigue / low-grade fever: this is not the OA pathway — refer to rheumatology to evaluate RA / autoimmune diseaseChronic dull ache + morning stiffness < 30 min + better with movement: most likely OA — work the OARSI / ACR first-line framework from scene 5
1. Load: resistance training 2-3 ×/week + full ROM
2. Proprioception: single-leg stance / balance board / unstable surface
3. Body composition: BMI control beats any supplement (Felson Framingham)
4. Sleep: let GH release in NREM3 (Van Cauter 2000) — an all-nighter hurts more than just the brain
5. See a clinician if pain persists > 1 week
By joint (core movements, 2-3 ×/week, not every day):
Knee: goblet squat / Bulgarian split squat / lunges (full depth) + single-leg deadlift + terminal knee extension; add ATG knees-over-toes work (see `knees-over-toes-myth`)Hip: hip hinges (RDL, good morning) + hip abduction (clamshell, side-lying leg raise) + internal/external rotation (90/90 stretch). For desk workers, tight hip flexors + weak hip extensors is the #1 thing to fixShoulder: scapular retraction + depression + external-rotation drills (Y-T-W) + full push + pull (overhead press + pull-up). Don't only train the front delt — the posterior chain is the stability driverAnkle: single-leg calf raises + eccentric heel drops (do them even pain-free) + single-leg balance (30 s × 3); warm up before running or court sportsLow back: McKenzie extensions / hip hinge / dead bug + bird dog + side plank (Stuart McGill's three) + hip abduction. Avoid isolated spinal-extension work as the main strategy; what the low back actually needs is abdominal + hip + spinal co-contraction — the stiffness-sandwich pattern McGill describes
Red lines (any joint — stop training, see a clinician):
Acute swelling, redness, heat: possible infection / gout flare / acute crystal arthritis — see atlas `gout`Joint locking / inability to bear weight after an injury: possible meniscus or ligament injuryBilateral, symmetric small-joint pain (finger MCPs, wrist) + morning stiffness > 1 h + fatigue / low-grade fever: this is not the OA pathway — refer to rheumatology to evaluate RA / autoimmune diseaseChronic dull ache + morning stiffness < 30 min + better with movement: most likely OA — work the OARSI / ACR first-line framework from scene 5
A weekly template — joint-friendly
Mon / Thu (resistance days, 45-60 min)Warm-up: 5-10 reps of joint ROM through full range (shoulder circles, hip circles, unloaded hip hinge, walking lunges, calf raises)4-6 main lifts: squat / deadlift / push / pull / single-leg work / core. 6-12 reps × 3-4 sets, progressive loadFinish: cooldown + 2 min balance / proprio work
Tue / Fri / Sat (aerobic + movement)
30-50 min moderate intensity: brisk walk / easy jog / bike / elliptical / swimYou don't need to run every session; cross-modal beats single-modeHR 70-80 % HRmax (can talk, cannot sing)
Every day (free, highest leverage)
8,000-10,000 steps (middle of Lin 2013's U-curve)Stand and move for 1-2 min every 30-45 min (cartilage-loading signal)Sleep 7-9 h with at least one full deep cycle (for GH release — see `all-nighter`)
Once a week (optional, useful)
Full dynamic stretching + ROM session (10-15 min): not static — loaded ROM (see `flexibility-vs-mobility`)
Progress signals:
After 6-8 weeks: single-leg balance, squat depth, subjective joint creakiness / stiffness should all improveIf they don't: review your form on video, see a physio 1-2 ×, check nutrition + sleep baselinesIf symptoms get worse, or you develop swelling / redness: stop and see a clinician