Name: Acute Injury Inflammation, Ice and NSAIDs
Creator: Realised
Published: 2026-08-01

Summary

After an acute injury, inflammation is the first phase of repair, not the enemy — brief ice and a short course of painkillers are legitimate tools when they get you moving again, but reflexive around-the-clock suppression has no evidence of helping healing and a plausible cost; the goal is pain control in service of early gentle movement, using the least-suppressive tool that achieves it.

Why Moderate

This entry sits at Moderate overall. Why not the tier above (Strong Evidence): the entry's central conditional thesis — that reflexive prolonged suppression trades unproven healing cost for unneeded comfort — rests on mechanism, animal data, and expert consensus, not on human outcome trials; no human RCT demonstrates that ice or short NSAID courses harm healing, and the successor frameworks (POLICE, PEACE & LOVE) are editorials that have never been trial-tested against what they replaced. A thesis whose harm-arm is mechanistic cannot claim Strong. Why not the tier below (Emerging): the load-bearing practical claims are better than Emerging — early mobilisation has a blinded-assessor human RCT plus supporting meta-analyses; short NSAID courses for function carry national-guideline backing; and the ice-doesn't-improve-healing conclusion comes from systematic reviews of 22 human trials. The entry's advice never leans its full weight on the animal data.

Per-sub-area split: repair-phase biology and the muscle-pump physiology — Foundational (textbook mechanism, but mechanism only). RICE retirement — historical fact, no tier. POLICE / PEACE & LOVE frameworks — Moderate (peer-reviewed expert consensus, no head-to-head trial). Ice and healing outcomes — Moderate (systematic reviews; verdict "insufficient evidence," honestly an absence). Ice-delays-healing harm signal — Emerging (rodent models only, and internally split by injury severity). NSAIDs and fracture non-union — Emerging (confounded human observational plus solid animal mechanism; adjusted association non-significant). Short NSAID courses for acute pain and function — Strong Evidence (guideline-grade). Early mobilisation for graded soft-tissue injury — Strong Evidence (human RCT plus meta-analyses, boundary-limited to soft tissue).

Practical takeaway

The organising principle: pain control in service of movement. The question to ask about any tool — ice, tablets, compression, rest — is not "does this fight inflammation?" but "does this get me moving gently, sleeping properly, and functioning again?" Choose the least-suppressive tool that achieves that.

First, rule out the situations that end self-management. If you cannot bear weight, see visible deformity, develop rapid severe swelling, feel numbness or pins-and-needles, or the pain is wildly out of proportion to what happened — stop and get assessed. Those are fracture, rupture, and nerve-compromise patterns, and everything else in this section assumes they are absent.

In the first hours: protect the area from re-injury and from loads that sharply provoke it. Elevate when convenient. If pain is significant, ice is a reasonable analgesic used the way its own former champion now recommends: around 10 minutes on, 20 off, once or twice — for comfort. There is no evidence-based reason to run icing around the clock or beyond the first several hours, and doing so trades unknown healing cost for comfort you could get more cheaply.

Painkillers, honestly: if pain is blocking sleep or blocking gentle movement, a short course at the lowest effective dose is legitimate, evidence-backed medicine — for acute low back pain it is literally the first-line guideline recommendation. Comfort is a real treatment goal, not a moral failing. The reasonable cautions are about pattern, not use: prefer a few days over a few weeks; prefer as-needed over a fixed preventive schedule once the worst has passed; be more conservative around a known or suspected fracture, where many surgeons limit NSAIDs on mechanistic grounds; and mind your personal stomach, kidney, and cardiovascular risk. If a clinician has prescribed you an anti-inflammatory or any other medicine, take it as prescribed — this entry is about over-the-counter reflexes, and any change to prescribed treatment is a conversation with the prescriber, not a decision to make from an article.

From day one or two: gentle movement is the treatment. Pain-free or pain-tolerable range of motion, short frequent walks if the leg allows weight, light activation of the muscles around the injury. This is the intervention with human trial support. Let symptoms gate the dose: movement that produces mild, settling discomfort is generally productive; movement that sharply provokes or leaves the injury clearly worse the next day was too much.

From there: progressive loading. Gradually restore range, then strength, then speed and load, over days to weeks depending on severity. If function is not clearly improving week over week, or you keep re-flaring, get assessed — persistent failure to progress is itself a flag.

What not to do: do not immobilise a routine sprain or strain for a week "to be safe"; do not run ice or tablets on a schedule long after they stopped serving movement and sleep; and equally — do not white-knuckle through pain that is stopping you sleeping and moving because you have been told suppression is sacrilege. Both reflexes cost you the same thing: the early movement that actually helps.

Evidence detail

Why This Entry Exists

When you roll an ankle or strain a muscle, two instincts compete. The first is the one most of us were taught: ice it immediately, keep icing it, take ibuprofen on a schedule, rest completely. The second is the newer internet counter-doctrine: never ice, never touch an NSAID, inflammation is sacred and any interference sabotages healing. Both are reflexes, and neither survives contact with the actual evidence.

The strange history here is that the man who invented the old advice retracted it. Gabe Mirkin coined RICE — Rest, Ice, Compression, Elevation — in his 1978 Sportsmedicine Book, and decades later publicly recanted the R and the I, writing that both ice and complete rest "may delay healing, instead of helping." The physiotherapy field has since moved through POLICE (which replaced Rest with Optimal Loading) to PEACE & LOVE (which explicitly advises avoiding anti-inflammatory medications and ice in early soft-tissue management). But a field changing its consensus is not the same as a field proving its new position — the successor frameworks are expert editorials, not trial-tested protocols, and the loudest "ice is harmful" evidence comes from rats, not people.

Meanwhile the counter-doctrine has its own casualty: the person in genuine pain who has been told medication is a betrayal of their body, and who consequently cannot sleep, cannot walk, and cannot do the early gentle movement that is the single best-evidenced intervention in this whole topic. Comfort is a legitimate treatment goal. A short course of painkillers that restores your sleep and gets you walking is evidence-backed medicine, not weakness.

What bad advice this protects against, in all directions:
• "Ice it around the clock for the first few days — you have to get the inflammation down." Wrong because there is no good human evidence that icing improves healing outcomes at all; the systematic reviews found the trials weak and the evidence insufficient. Ice is a decent short-term painkiller and nothing more. Prolonged icing has no evidential support and a plausible mechanistic cost seen in animal models.
• "Never ice, never take an NSAID — inflammation is your body healing and you must not touch it." Wrong because it overclaims thin evidence: the ice-harms-healing signal is a rodent crush-injury study (and the same rodent lineage later found icing helped regeneration after milder injury), and the NSAID-harms-healing signal in humans largely dissolves when studies adjust for confounding. Worse, it abandons people to pain that blocks sleep and movement — and early movement is the intervention with actual human trial support.
• "Rest it completely until it stops hurting." Wrong for most soft-tissue injuries: a randomised trial in acute ankle sprains found early therapeutic exercise beat standard rest-based care for function, and exercise-based rehab reduces re-injury. Total rest is right for fractures and suspected ruptures pending assessment — not for a routine sprain or strain.
• "Just push through it — pain is information you should ignore." Wrong and dangerous at the edges. Some presentations demand assessment, not self-management: inability to bear weight, visible deformity, rapid severe swelling, numbness or pins-and-needles, pain wildly out of proportion to the event. Toughing those out is how a fracture or rupture gets missed.
• "Take ibuprofen preventively on a fixed schedule for a couple of weeks so it never flares." Wrong direction of use: the open questions about NSAIDs and tissue healing concern sustained, high-dose, weeks-long suppression — not a few days for function. Running a chemical blanket over the entire repair window buys you nothing the shorter course did not, and it is where the plausible costs live.

This entry OWNS: the acute-injury question — what to do about inflammation, ice, and short-term painkillers in the hours and days after a sprain, strain, or similar soft-tissue injury; the RICE-to-PEACE-and-LOVE history; the fracture-and-NSAID question; the early-movement evidence; when to seek assessment instead of self-managing.

This entry DEFERS: whether NSAIDs and ice blunt training adaptations and how to handle ordinary post-exercise soreness — that is exercise-induced muscle damage, not injury, and it is owned entirely by muscle_soreness_recovery_doms. The general philosophy of when a defensive symptom is worth blunting is owned by defensive_symptoms_and_when_to_blunt_them. Chronic joint pain management is owned by joint_pain_conservative_management; back pain specifics by low_back_pain_evidence_and_management.

Evidence

Read the tiers, not the thesis. The claims below range from textbook physiology to a single rat study, and the honest picture only emerges if you keep track of which is which. Every claim is labelled by evidence type.

Sub-area 1: What inflammation is doing after an injury

1. Inflammation is phase one of a repair sequence, not a malfunction. Injury triggers an inflammatory → proliferative → remodelling cascade: neutrophils arrive within roughly two hours (peaking at 6–24 hours) to clear debris; pro-inflammatory M1 macrophages dominate the first days, clearing dead tissue and activating the satellite cells that rebuild muscle; a switch to anti-inflammatory M2 macrophages around day four drives differentiation, resolution, and remodelling. Blocking or blunting the early M1 phase in animal models impairs the regeneration that follows. — Foundational as sequence; the phase biology is uncontested. Evidence type: mechanistic reviews — rodent models plus human histology; textbook physiology. Citation: Dort et al., Stem Cells International 2019. Cui bono: nobody monetises this directly, which is partly why it goes under-taught; wellness content borrows it to overclaim "all inflammation is sacred." Critical caveat: mechanism is not outcome. The leap from "macrophage depletion harms regeneration in rodents" to "any anti-inflammatory intervention harms human healing" is not proven by this literature.

2. Swelling is partly delivery, and clearance depends on movement. The exudate carries repair cells and signalling molecules to the injury; the lymphatic system clears the waste, and lymphatic return is driven substantially by skeletal-muscle contraction — which is the mechanistic case for gentle movement over total rest. — Foundational as physiology, but note honestly: this is textbook mechanism (the muscle pump), not a primary trial we verified, and the practical "movement beats rest" claim should hang on the human RCT below, not on the pump alone. Evidence type: textbook physiology / mechanistic reasoning. Cui bono: compression-garment vendors monetise the swelling-clearance story; the free version — walking, gentle contraction — sells nothing.

Sub-area 2: The rise and retirement of RICE

3. RICE was retired by its own author. Mirkin coined it in 1978 and publicly recanted, writing on his own site (verified verbatim): "Coaches have used my 'RICE' guideline for decades, but now it appears that both Ice and complete Rest may delay healing, instead of helping." His revised practical advice from the same article: ice for up to 10 minutes for pain, off 20, repeat once or twice, and no reason to ice beyond about six hours post-injury — ice as short-term pain relief, not treatment. — This is a historical fact, not efficacy evidence: one physician changing his mind on mechanistic grounds. Rhetorically powerful; evidentially it carries nothing on its own. Evidence type: primary-source author statement. Citation: Mirkin, "Why Ice Delays Recovery," drmirkin.com — first posted 2014; the current page is dated September 2015. Cui bono: the recantation is heavily amplified by "everything you know is wrong" content and ice-bath sceptics — note he recanted icing-for-healing, not icing-for-pain.

4. The successor frameworks are real, and they are editorials. POLICE (Protection, Optimal Loading, Ice, Compression, Elevation) replaced rest with loading in 2012; PEACE & LOVE (2019/2020) went further, explicitly advising avoidance of anti-inflammatory medications and ice in early soft-tissue management — while its authors acknowledge NSAIDs help pain and function. Both are peer-reviewed expert-consensus editorials in the British Journal of Sports Medicine, grounded in mixed human-trial and animal evidence. Neither has been trial-tested head-to-head against RICE. — Moderate: the current consensus direction of the field, honestly framed as "the field moved, and here is why," not "proven superior." Evidence type: expert-consensus editorials synthesising human RCT and animal data. Citations: Bleakley et al., BJSM 2012; Dubois & Esculier, BJSM 2020. Cui bono: physiotherapy as a profession benefits from active-rehab framing — more sessions than "go home and ice" — which does not make it wrong; the underlying loading trials are independent.

Sub-area 3: Ice, on its own merits

5. Ice is a genuine short-term analgesic with no demonstrated healing benefit — and the verdict is "insufficient evidence," not "proven harmful." The anchor systematic review pooled 22 randomised trials and found them methodologically weak (mean PEDro quality score 3.4 out of 10), with insufficient evidence that cryotherapy improves clinical outcomes in soft-tissue injury; marginal signals favoured ice plus exercise after ankle sprain or surgery, and ice-plus-compression beat ice alone for pain. A second 2004 systematic review reached the same verdict. We could not confirm a dedicated Cochrane review on cryotherapy for acute soft-tissue injury; these two 2004 reviews are the field's cited anchors, and we say so rather than inventing one. — Moderate, with the absence of outcome benefit stated as an absence. Evidence type: systematic reviews of human RCTs, low-quality primaries; blinding impossible with ice. Citations: Bleakley et al., AJSM 2004; Hubbard & Denegar, J Athl Train 2004. Cui bono: cryotherapy chambers and recovery-tech vendors monetise ice-everything with "reduce inflammation" as a sales verb; "ice destroys healing" content monetises contrarianism; the verified middle — brief ice for pain, then move — sells no hardware.

6. The "ice delays healing" signal is real but comes from rats — and even in rats it depends on how bad the injury is. A single 20-minute ice-pack application after crush injury to rat muscle delayed the clearance of dead fibres and the migration of macrophages, produced smaller regenerating fibres at 28 days, and increased collagen (scar) deposition versus no icing. Note honestly what that dose means: 20 minutes of ice is within typical human use, so this result cannot be waved away as an overdose artefact — the genuine gaps are species and severity, because a crush injury with widespread muscle necrosis is far more severe than a typical human strain. And the same rodent lineage later cut the other way: a 2023 study from the same research group found that when the injury was mild (necrosis in only a small fraction of fibres), icing limited inflammatory macrophage invasion and actually facilitated regeneration. So even the animal evidence does not support a blanket "ice harms healing" — it supports "the effect of icing plausibly depends on injury severity, and none of it has been demonstrated in human clinical outcomes." Human translation: unproven, stated as an absence. — Emerging. Evidence type: rodent (crush-injury and mild-injury models); a human-like single 20-minute icing dose, but injury models and species that do not map cleanly onto human sprains and strains. Citations: Takagi et al., J Appl Physiol 2011; Nagata et al., Am J Physiol Regul Integr Comp Physiol 2023. Cui bono: the 2011 rat study carries most of the "ice is harmful" internet — quoting it without naming the species, or without the 2023 mild-injury result, is the tell for overclaim.

Sub-area 4: NSAIDs and healing

7. NSAIDs and fracture non-union: a real signal that shrinks under scrutiny. A 2010 meta-analysis of 11 human cohort and case-control studies found roughly threefold odds of non-union with NSAID exposure (pooled OR 3.0, 95% CI 1.6–5.6) — but lower-quality studies reported the higher odds, and restricting to higher-quality, confounder-adjusted studies gave OR 2.2 (0.8–6.3), not statistically significant. A 2024 updated systematic review and meta-analysis lands harder on the null: pooling adjusted results, it found no significant difference in non-union or delayed union between NSAID users and non-users (pooled adjusted OR 1.11, 95% CI 0.99–1.23). The key confound is confounding by indication — worse fractures hurt more, and people in more pain take more NSAIDs. The animal evidence that COX-2 inhibition impairs bone healing is consistent and mechanistically solid. No randomised human trial demonstrates harm. Honest verdict: prolonged or high-dose NSAID use after a fracture is a reasonable thing to be cautious about — many orthopaedic surgeons are — while a few days for pain has not been shown to cause non-union. — Emerging for the human harm claim. Evidence type: human observational (confounded by indication) plus robust rodent mechanism; no human RCT of harm. Citations: Dodwell et al., Calcif Tissue Int 2010; 2024 updated meta-analysis. Cui bono: the wellness market monetises "NSAIDs are poison" (it sells turmeric); the over-the-counter pharma market monetises reflexive scheduled dosing; neither side profits from "short course for function, don't run it for weeks." On tendon and muscle healing specifically: the concerns are largely animal-derived, we did not verify specific tendon studies for this entry, and the training-adaptation question is deferred entirely to muscle_soreness_recovery_doms.

8. The counterweight: short NSAID courses for acute pain are effective and guideline-supported. For acute low back pain, the UK national guideline (NICE NG59) makes oral NSAIDs the first-line pharmacological option — at the lowest effective dose for the shortest period, with stomach protection where indicated — and explicitly says do not offer paracetamol alone. — Strong Evidence. Evidence type: national clinical guideline synthesising human RCTs. Citation: NICE NG59, 2016 (updated 2020). Cui bono: pharma benefits from NSAID normalisation and wellness from demonising it; the guideline position — short, purposeful, lowest-dose — serves neither market. Caveat: this guideline is for low back pain specifically; NSAID net benefit varies by injury and by personal stomach, kidney, and cardiovascular risk. It is not a licence to generalise "NSAIDs first-line" to every injury — it is proof that a short course for function is legitimate medicine, not a failure of willpower.

Sub-area 5: The thing that actually has trial support — early movement

9. Early mobilisation beats immobilisation for most soft-tissue injuries. In a randomised controlled trial of 101 people with acute grade 1–2 ankle sprains recruited from emergency and sports-injury clinics, with blinded outcome assessment, early therapeutic exercise in week one improved short-term ankle function versus standard rest-based care. Later meta-analyses show exercise-based rehab reduces re-injury after lateral ankle sprain. This is the human-outcome evidence that gives POLICE and PEACE & LOVE their footing. — Strong Evidence, within its boundary: graded soft-tissue injury. Immobilisation remains correct for fracture, gross instability, and suspected complete ruptures pending assessment. Evidence type: human RCT (n=101) plus supporting meta-analyses; participants could not be blinded; generalisation beyond ankle sprains to hamstring or tendon is by analogy plus separate literatures. Citation: Bleakley et al., BMJ 2010. Cui bono: "move early" sells physiotherapy time but no products — it is the cheapest evidence-backed intervention in this entire topic.

Mechanism

The repair sequence runs in three overlapping phases. In the inflammatory phase (hours to days), damaged tissue releases signals that recruit neutrophils within about two hours to clear debris; pro-inflammatory macrophages follow, digesting dead fibres and — critically — activating the satellite cells that will rebuild the tissue. Around day four the macrophage population switches character to an anti-inflammatory profile that drives those repair cells to differentiate and mature, opening the proliferative phase (days to weeks) where new tissue is laid down, and then the remodelling phase (weeks to months) where it is reorganised along lines of mechanical stress.

Three practical consequences fall out of this biology. First, the early inflammation is doing recruitment and demolition work that the later rebuilding depends on — in animal models, deleting the early phase impairs everything downstream. That is the mechanistic core of "don't chemically flatten the first days without a reason." Second, swelling is a mixed bag: it is partly delivery of repair machinery and partly congestion, and its clearance runs through the lymphatic system, which is pumped substantially by muscle contraction — a mechanistic argument for gentle movement that the ankle-sprain trial then supports with human outcomes. Third, remodelling is load-directed: tissue organises along the stresses applied to it, which is why progressive loading, not protection, is the long middle of every modern rehab framework.

Ice works on this system by vasoconstriction and by slowing nerve conduction — the second is why it genuinely relieves pain. Whether the first meaningfully suppresses the repair cascade in human injuries is exactly the open question: shown in a rodent crush model, reversed in a milder rodent model, never demonstrated in human outcomes. NSAIDs work by inhibiting COX enzymes and thus prostaglandin production; prostaglandins are both pain signals and repair signals, which is why the same tablet that restores your sleep is theoretically capable — at sustained doses over weeks — of muting part of the rebuild. Mechanism cuts both ways here, and outcomes, not mechanisms, get the final word.

Risks And Contraindications

• Red flags — these end self-management and mean assessment now: inability to bear weight or use the limb; visible deformity; rapid, severe, or spreading swelling; numbness, tingling, or a cold/pale limb; pain grossly out of proportion to the injury (a pattern that can signal compartment syndrome, a genuine emergency); an audible pop with immediate giving-way; fever or spreading redness around a wound. None of the advice above applies until these are excluded.
• Fracture, instability, and complete rupture are the exceptions to early movement. Immobilisation is correct management for these pending professional assessment. "Move early" is for graded soft-tissue injuries, not broken bones.
• NSAIDs carry real, well-characterised risks unrelated to healing: stomach bleeding and ulcers (risk rises with age, dose, duration, alcohol, and combination with certain medicines), kidney strain (especially when dehydrated or on blood-pressure medicines), and cardiovascular considerations with some agents. Lowest effective dose, shortest useful period, food in the stomach, and a pharmacist's advice if you take other medicines.
• Never stop or alter a prescribed medicine on the strength of this entry. If a clinician has you on an anti-inflammatory, an anticoagulant, or anything else relevant to injury care, changes go through them.
• Ice on compromised tissue: avoid prolonged icing over areas with poor circulation or sensation (peripheral neuropathy, Raynaud's), and never ice to numbness against bare skin for extended periods — cold injury is a real, avoidable harm.
• Undertreatment is a risk too. Unmanaged severe pain wrecks sleep, blocks rehabilitation, and in some people contributes to pain becoming persistent. Refusing all analgesia on principle is not the safe option; it is a different risk.

Controversy

Nature of the controversy: a genuine scientific disagreement about whether standard anti-inflammatory care (ice, NSAIDs) trades healing quality for comfort — inflamed further by the unusual spectacle of a guideline's own author recanting it, and by two industries with opposite incentives amplifying opposite halves of the evidence.

Position A — "suppress by default": ice early and often, NSAIDs on schedule, rest until pain-free. Its best evidence: ice demonstrably relieves pain; NSAIDs demonstrably restore function and are first-line in a major national guideline for acute low back pain; no human trial has ever shown that brief icing or a short NSAID course worsens a soft-tissue outcome. Where it overreaches: it converts "good analgesics" into "healing treatments." There is no good human evidence that ice improves healing outcomes at all, the trials are weak, and around-the-clock protocols rest on habit, not data. Scheduled multi-week NSAID use extends into exactly the territory where the mechanistic and observational concerns live.

Position B — "inflammation is sacred": never ice, never NSAID, any suppression sabotages repair. Its best evidence: the phase biology of repair is real and Foundational; rodent models show icing after severe crush injury and macrophage-blunting impair regeneration; the field's own consensus frameworks now advise avoiding anti-inflammatories early; the fracture literature shows a crude association between NSAIDs and non-union. Where it overreaches: the rodent ice-harm data has never been replicated in human outcomes — and the rodent literature itself is split, with the same research lineage finding that icing helped regeneration after milder injury; the fracture association collapses to non-significance in adjusted, higher-quality studies, with confounding by indication the likeliest story; PEACE & LOVE is an editorial, never trial-tested against what it replaced; and the position's practical output — people afraid to take two days of ibuprofen, lying awake and not walking — directly blocks the one intervention with Strong Evidence behind it.

The funding/bias dimension — cui bono, both ways: the suppress-by-default world is monetised by over-the-counter analgesic marketing and by a recovery-tech industry (cryo chambers, cold tubs, compression gadgets) that sells "reduce inflammation" as a verb. The inflammation-is-sacred world is monetised by the wellness market that sells natural alternatives (turmeric, "anti-inflammatory protocols") and by contrarian content whose business model is "everything you were told is wrong." Note the asymmetry of what goes unsold: brief ice for comfort, a short purposeful course of tablets, and early gentle walking generate revenue for nobody, which is precisely why the honest middle position is the quietest one.

Realised Position: inflammation after injury is repair machinery and deserves respect, not reverence. We do not suppress it by reflex, and we do not protect it at the cost of the person. Ice is a painkiller — use it briefly if it helps, expect nothing more from it. A short course of NSAIDs to restore sleep and movement is legitimate, guideline-backed medicine; comfort is a real treatment goal, and nothing in this entry means "do not take medication" or licenses anyone to stop a prescribed medicine. The trade to actually avoid is weeks of scheduled suppression that nobody's evidence supports. The prize to actually chase is early, gentle, progressive movement — the cheapest, best-evidenced intervention in the topic. And a small set of presentations (can't bear weight, deformity, rapid severe swelling, numbness, pain out of proportion) are not management questions at all: they are assessment, today.

Cross-Pillar Connections

• defensive_symptoms_and_when_to_blunt_them — the parent framework: inflammation after injury is a defensive programme like fever or sickness behaviour, and this entry is the acute-injury instance of the general question "when is blunting a defence costless comfort, and when is it a trade?"
• muscle_soreness_recovery_doms — the explicit boundary: ordinary post-training soreness is exercise-induced damage, not injury, and the whole NSAID/ice-blunts-adaptation question for training is owned there; this entry defers it entirely.
• joint_pain_conservative_management — where an acute episode shades into persistent joint pain, management shifts from the acute repair-phase logic here to the graded-exposure and load-management logic of that entry.
• low_back_pain_evidence_and_management — the NICE first-line NSAID evidence cited here comes from the acute low back pain guideline; that entry owns back pain's full picture, including why staying active beats bed rest there too.
• acl_and_knee_injury_prevention — the serious-knee-injury pathway: a pop-and-give-way knee is a red-flag presentation for this entry and an assessment-and-prevention story for that one.
• hormesis_and_adaptive_stress — the shared principle: inflammation is a stress response with an adaptive job, and both entries guard against the reflex that every uncomfortable biological response should be switched off.
• deload_and_recovery_management — the return journey: once the acute phase resolves, reintroducing training load after injury is a deload-and-rebuild problem, owned there.

What would change our mind

Falsifiability: explicit upgrade/downgrade criteria from source

• Toward stronger anti-suppression advice: a human randomised trial showing that early icing or a short NSAID course measurably worsens soft-tissue healing outcomes (function, re-injury, tissue quality) versus placebo. None exists today; one adequately powered trial would move the ice and NSAID sub-areas from Emerging-mechanistic to demonstrated harm, and this entry would harden accordingly.
• Toward relaxing the caution: high-quality prospective human studies of fracture patients with dose and duration captured, showing no adjusted association between short NSAID courses and non-union — the 2024 adjusted near-null already points this way, and a definitive study would let us soften the fracture caution to standard NSAID risk counselling only.
• On the frameworks: a head-to-head trial of PEACE & LOVE versus a standard-care protocol. If the avoid-anti-inflammatories arm won on healing outcomes, Position B gains real human footing; if it lost or tied, the "avoid" advice should be demoted to preference.
• On ice specifically: human studies with tissue-level or functional endpoints replicating the rodent regeneration findings at realistic icing doses would upgrade the harm signal from Emerging to something we act on; repeated failures to replicate — or human confirmation of the severity-dependence the rodent work now suggests — would retire or reshape it.
• What would NOT move us: more rodent crush-injury studies (the mechanism is already granted; species and injury severity are the gap, and the rodent literature already shows both directions); more expert editorials or consensus acronyms (the field's direction is already noted; consensus is not outcome data); observational fracture studies without confounder adjustment (the crude association is already known and already explained); Mirkin-recantation retellings (historical fact, zero incremental evidence); and testimonials in either direction — "I iced and healed fine" and "I quit NSAIDs and healed faster" are equally uninformative.

Industry bias note

Structural incentives the evidence base may reflect

The suppression-selling end. Over-the-counter analgesics are a mass consumer market whose advertising normalises reflexive, scheduled dosing for every ache — the profitable behaviour is exactly the multi-week pattern the healing question is about, not the short purposeful course guidelines describe. Alongside it, the recovery-technology industry (cryotherapy chambers, ice-bath hardware, cold-compression devices) sells "reduce inflammation" as an unexamined good; its marketing needs inflammation to be an enemy, because you cannot sell a machine that fights a repair process. Neither market has any incentive to tell you that a bag of frozen peas for ten minutes, twice, captures essentially all the demonstrated benefit of ice.

The suppression-demonising end. The wellness and supplement market runs the mirrored play: "NSAIDs are poison" is a conversion funnel to turmeric, proteolytic enzymes, and "natural anti-inflammatory protocols" whose own evidence is far weaker than the drugs they replace. Contrarian content platforms monetise the Mirkin recantation and the single rat icing study endlessly, because "the advice you grew up with is a lie" is reliable engagement — the species of the study rarely makes it into the caption, and the later rat study where icing helped a milder injury never does. Physiotherapy, more benignly, benefits from the active-rehab framing (a course of loading sessions bills more than "go home and ice"), though the loading evidence itself comes from independent trials and holds up.

The clean signal. Follow what no one profits from: brief ice for comfort with nothing expected of it, a short lowest-dose course of tablets when pain blocks sleep and movement, early gentle walking, and progressive loading. Every element of that package is free or nearly free, sells no hardware, no supplement, and no subscription — and it is where the human-outcome evidence actually points. When a topic has two well-funded opposite exaggerations, the unmarketable middle is usually the signal.

Acute Injury Inflammation, Ice and NSAIDs