Not sure when to use ice or heat? Get a quick overview in The Great Ice vs. Heat Confusion Debacle, or a detailed look at why you should (almost) never ice low back pain. In general, heating is better for almost anything that isn’t a fresh injury.
Icing — “cryotherapy” for therapy geeks — is an essential injury management skill. Everyone should understand icing the same way everyone knows how to put on a Band-Aid. It is a cheap, effective, drugless method for relieving the pain of injuries. Safe application of ice to your skin can relieve symptoms from sprains, strains, bruises, and tendinitis — virtually any situation in which superficial tissues are inflamed by trauma. Icing could also assist recovery from repetitive strain injuries like tendinitis or iliotibial band syndrome, although that is less clear.
But there also hazards and complications that anyone with chronic pain absolutely needs to know. This article is full of tips and insights about icing that will take you way beyond the basics!
My strained thigh muscle felt like a broken bone at first, but I'm using your icing method and it’s responding brilliantly.
Paul Farley, West Sussex, England
In late 2013, a personal trainer wrote an article that got a lot of attention: Why Ice and Anti-inflammatory Medication is NOT the Answer. More importantly, the guy who first proposed the RICE rehab protocol back in 1978 has actually recanted the I-for-ice part, declaring that “icing delays recovery.” Many readers have asked me about these articles, and many similar ones.
There is one important kernel of truth in these attempts to debunk icing: inflammation is indeed biology business as usual, a normal part of healing. But not everything “normal” in biology is wise and good and never needs to be controlled or treated.1 Diarrhea is natural, too. Inflammation can get out of hand, and we know that at least some of it is actually “bad” biology that does not benefit us in any way.2
I think the death of RICE (rest, ice, compression, elevation) has been declared prematurely by some overeager mythbusters, even including Dr. Mirkin. I’ve read their arguments and don’t find them very persuasive.3 Icing is mostly intended to conservatively, temporarily treat the worst symptoms of inflammation, not to try to put it out like a fire. And so the whole “inflammation is healthy” argument is mostly irrelevant, and conventional icing wisdom is perfectly sound and safe with the modest goal of pain control.
If it works, it is useful mainly where tissue are damaged and/or “inflamed.” (Although some things that we think of as “inflamed” aren’t actually inflamed. And vice versa! More about this below.)
Icing is primarily an analgesic — a pain-reliever — and not an actual treatment. That is, it doesn’t “fix” anything. Use it like you use ibuprofen. It may help to resolve chronic problems (much more about this below), but it’s mostly intended to simply numb painfully inflamed or other hurting tissues.
The most commonly iced acute injuries are fresh injuries — ligament sprains, muscle strains, and severe bruises. (When the skin is broken, things get a little trickier.) And what’s a “fresh” injury? Any time tissue has been physically damaged, it will be inflamed for a few days, give or take, depending on the seriousness of the injury. If superficial tissue is sensitive to touch, if the skin is hot and red, if there is swelling, these are all signs that your injury is still fresh, and should definitely not be heated. Heat will increase the circulation and significantly facilitate the immune system activity.
Ice is also often helpful with chronic overuse or tissue fatigue injuries like carpal tunnel syndrome, tennis elbow, supraspinatus tendinitis, iliotibial band syndrome, patellofemoral pain syndrome, shin splints, and plantar fasciitis. There are others, of course, but these are the most common.
Ice may also be useful for garden-variety “wear and tear” arthritis, and sometimes the nasty inflammatory arthritides (rheumatoid, ankylosing spondylitis).
Ice can mildly aggravate the pain of muscle spasm and trigger points (muscle knots). Trigger point pain is extremely common, and is routinely mistaken for an “iceable” injury, especially in the low back. Here’s the short explanation:
Back pain is rarely an injury — that is, the pain is rarely caused by inflammation which might be helped by ice. Even in cases where inflammation is present, it is usually deep in the back under a thick layer of insulating muscle and the ice cannot “reach” it. However, back pain almost always involves muscular trigger points (muscle knots), which are more likely to be aggravated by ice and helped by heat! For this reason, the majority of people with back pain prefer heat, and a few have negative reactions to ice. For similar reasons, neck pain usually should also not be iced. Although experiments have shown that both ice and heat are modestly helpful for low back and neck pain, there are good reasons to err on the side of heat. Ice should only be used on the back by patients who clearly prefer it (for whatever reason), or when there is definitely a fresh injury.
The stakes are not high. Studies have shown that people usually get mild benefit from ice — about the same as heat in fact.6 But there are good reasons to err on the side of heat nevertheless. If you have low back or neck pain, this topic is covered in much more detail in (Almost) Never Use Ice on Low Back Pain!
But the low back is only the most obvious example, because it is both much tougher and immune to injury than people think, and also much more prone to painful trigger point activity than most people realize. In fact, the vast majority of garden variety aches and pains are dominated not by arthritis and injury as people tend to assume, but “just” by muscle pain — which tends to be irritated by ice, and generally would rather have a hot pack and a massage.
Icing is traditionally used wherever people suspect inflammation. However, that includes at least three common and surprisingly different biological situations … and different reasons for icing, or not icing.
The third one in particular is a bit of a revelation for a lot of people. It’s covered in detail below.
Classic inflammation is pretty well understood, and it’s mainly an immune system reaction. When tissue is damaged, the body responds with a complex array of chemical and neurological changes collectively known as inflammation. For instance, the capillaries widen in a big way to bring extra oxygen and nutrients to the area. They also “loosen,” becoming more permeable, to give immune system cells easy access to the injury.7 Most of the pain and discomfort of inflammation is due to this immune system activity.
If the skin is broken (septic injury), there is a risk of infection, and the immune system reaction is essential — a pure physiological goodness. Inflammation in that context is a machine finely-tuned by evolution to optimize recovery, just as a fever is an effective physiological process for fighting infection (indeed, they are closely related processes). Strictly speaking, if you to want to heal well, don’t interfere with inflammation! For broken-skin wounds, use ice only a little to “take the edge off.” And course it’s best to use it in a sterile way, so that you don’t add to the infection risk!
If the skin isn’t broken (sterile injury), things are really quite different! In this context, inflammation is an absurd overreaction that causes collateral damage and excessive pain for no good reason at all — a genuine glitch in biology In this context, inflammation is an absurd overreaction that causes collateral damage and excessive pain for no good reason at all — a glitch in biology., with a recently discovered and clear evolutionary explanation.8 There is simply no need for the immune cells to get all fired up for a sterile, internal injury where there is no possibility of infection. Nevertheless, they do get fired up.
The important point here is that inflammation of sterile injuries is truly worse than it needs to be — and it’s quite reasonable to try to suppress it with ice, and anti-inflammatory medications like ibuprofen or Voltaren® Gel. Bet you didn’t know that. Very few people do — this is based on surprisingly new science about that glitch! This treatment logic was unavailable until quite recently.
Of course, the immune system reaction is not the only reason injuries hurt. Damaged cells put out many kinds of distress signals. As with most biological processes, our comfort is not really a priority. In fact, quite the opposite — injury and inflammation have partly evolved to be painful. Cavemen didn’t have ibuprofen and ice, nor did they have the benefit of understanding inflammatory chemistry. In the big picture, super painful inflammation was good for our species: victims were encouraged to stay relatively still while inflammation ran its course like a fever!
But for modern humans, inflammation is … well, it’s overkill. We can afford to “turn it down.” We can ignore the warning of the inflammation, to a point, and especially if we’re confident that we don’t really need an immune system reaction (for a sterile injury). Ice can only turn it down so much anyway, so there’s no risk of missing the pain alarm entirely!
In theory, cold slows metabolic activity, numbs nerve endings, constricts capillaries. It limits and controls inflammation. It makes it hurt less. It helps us get through the day. And that’s an especially good thing for the common sterile injuries, where the inflammation is largely pointless.
There’s yet another kind of “inflammation”…
Do you have a chronic overuse injury? Shin splints for two years? Plantar fasciitis for five? Then you are probably thinking, “Inflammation is not @#!!$% a valuable warning signal!” It’s more like an annoying car alarm.
Right you are. Sort of.
What’s going on in a repetitive strain injury like runner’s knee or tennis elbow or Achilles tendinitis is not really inflammation, per se, except perhaps in the earliest stages,9 but erratically painful degeneration1011 — tissue rot, which has more in common with arthritis than inflammation. The chemistry of these situations is very different than classic inflammation, and in particular involves relatively little immune system activity. The most obvious implication of this is that medicines intended to suppress immune system activity — the anti-inflammatories — are obviously not going to work well. And they don’t.
So what of ice, then? The most basic of all anti-inflammatory treatments? Ice can relieve many kinds of pain temporarily by numbing nerve endings, of course. But does it do anything else?
If ice can help a repetitive strain injury in any way beyond brief numbing, no one has ever actually proven it, or shown how it might work. There’s little doubt that it’s relieving in the short term, and there is some reasonable speculation that it could stimulate miscellaneous minor tissue healing processes. Virtually any stimulatory input to the body, up to a point, can provoke a healthy, adaptive response — the use-it-or-lose-it principle. Stress a tissue, and it will probably get a bit tougher. Maybe.12 In broad strokes, that is probably the only plausible therapeutic mechanism of icing. Ice may simply be one of the easiest delivery systems for a bit of non-toxic stimulation — a way to stimulating tissue without overloading it, while simultaneously getting some temporary pain relief from numbing. However, it certainly isn’t “anti-inflammatory”!
The great advantages of ice as a treatment are not its impressive biological effects, but its thrift, ease, and safety: treatment options simply don’t get any more innocuous while still having some plausible mechanism of benefit.It’s worth doing. The great advantages of ice as a treatment are not its impressive biological effects — unknown and unproven! — but its thrift, ease, and safety: treatment options simply don’t get any more innocuous while still having some plausible mechanism of benefit. Therefore ice remains firmly on my “worth a shot” list for RSIs. Keep your expectations low, but there are virtually no risks, other than ice burn (which takes at least a couple minutes of raw ice application, probably twice that).
What if ice actually doesn’t do anything for inflammation? What if it actually made it a little worse? Could you even tell? There is actually research that shows this.
A small 2013 study for severe muscle soreness had “unexpected” results, according to the researchers: it seemed to do more harm than good.13 The icing victims had higher blood levels of molecules associated with muscle injury and they felt more fatigued. Icing had no effect on recovery of strength, or any biochemical sign of inflammation. A small study, to be sure, but how good can icing be if it can generate this kind of data?
That study was specifically focussed on post-exercise muscle soreness, a very specific biological situation that is essentially impossible to treat.14 Even if this data is right, it doesn’t necessarily mean that ice is bad news for other kinds of pain and injury. Icing probably has different effects depending on the situation (which should be very clear after the last couple sections), and it’s particularly clear that it has useful effects on severe inflammation following trauma.
But this is sobering news for ice enthusiasts. It is possible that the main effect of ice is temporary numbing, which actually just makes it difficult to tell if you’re actually a little better or a little worse an hour later. It’s hard to judge a subtle change after a big one!
An excellent method of therapeutic icing is to use bare or “raw” ice — that is, ice applied directly to the skin, with no layer of plastic or fabric between you and your ice.
Raw ice delivers more of an icy punch! This is due to the spreading of melt water into every crevice, which conducts heat more efficiently away from the skin both directly into the ice, and via evaporation.
In comparison, gel packs and bean bags are comparitively wimpy cryotherapies (although they have their place, as you’ll see). They tend to warm up too quickly (especially where the skin is hottest and needs the most icing), and they sometimes cannot shape themselves well (or gently) to the contours of the injured body part. There are times when they are handy or easier when the stakes are low, but for serious icing of acute injuries or a stubborn tendinitis, you really need an ice cup.
A styrofoam cup is an elegant delivery system for raw ice …
The humble styrofoam cup is the cheapest and most effective injury management tool in my office and in my own home. It’s not the cup itself that’s so useful, of course, but its contents — ice! Don’t wait until you’re hurt to do this — have them ready and waiting.
In a pinch, with no cups around, just use an ice cube held in a dishtowel — less convenient, especially for larger areas, but nearly as effective. Over the years I’ve found that, despite the best intentions, 9 times out of 10 that I need an ice cup, I haven’t actually got one ready in the freezer, and I end up using ice cubes at first!
One of the downsides of the styrofoam cup option is that it’s a bit wasteful. Wouldn’t it be better to have a re-usable plastic cup designed for the same purpose? There are several brands of ice cups, like the CRYOCUP™ and the Pro-Tec Ice-Up Portable Ice Massager. I’ve never used one personally, but they seem like a great idea, especially for anyone who lives on the edge and brings home new sports and adventure injuries on a regular basis.
In fact, there are even insulated icing tools designed to be taken to the field or the mountains or wherever you think you might need ice.
And although it’s not at nice as ice, an alternative worth mentioning is a simple chemical ice pack — the kind you crush to activate. Despite their inferiority (they aren’t as cold as ice and won’t last as long), I have succumbed to their convenience, and these are what I’ve actually been taking on expeditions for a while now.
Slide the ice over the inflamed area in a slow but steady pattern. It’s important to keep moving, as long as you don’t try to ice such a large area that tissue gets a chance to warm up before you return to the starting point.
Continue ice massaging for 1–3 minutes, or until it is numb, whichever comes first — no more. “When you’re numb, you’re done,” is the rule of thumb for safety (see next section). Areas with thick tissue, like the top of the thigh, will take longer to get numb. Thin areas, like the side of the knee, will usually go numb quickly.
What does numb feel like? Just close your eyes and lightly touch the skin. If you can’t feel it at all, or if you can feel only pressure, that’s numb enough. Stop icing and let the tissue warm up.
Yes. You may have heard that bare ice is too cold to use directly on the skin in this way. That’s untrue for short periods. Although a cold-sensitive person may find raw ice too uncomfortable, tissue damage can only occur after sustained icing — well after you have gone completely numb, at least 3 minutes. Stopping roughly when you get numb pretty much guarantees that you won’t hurt yourself.
An ice treatment will feel like it is burning or stinging at first, and that’s okay. Icing this way can feel a bit nasty, especially at first in certain locations, but stick with it: the powerful anti-inflammatory effect is worthwhile. In many situations, this is a much better solution than an anti-inflammatory medication.When you’re numb, you’re done.
Once your tissues warm up again, you can repeat the treatment. In fact, you can apply the ice as often as you like, as long as your tissues have a chance to mostly warm up between treatments. In the case of tenditis, you can continue doing a lot of icing — many applications per day — as long as you still have symptoms, and even when you are feeling better.
In the case of injuries, icing is mostly just useful while the injury is still hot, red, swollen or painful — this phase may last for a few hours or several days. When these signs begin to fade, you may be certain that you would have been stuck with them for a lot longer if you had not been icing.
People often tell me that they have “tried icing” for an overuse injury like iliotibial band syndrome. A little quizzing usually reveals that this means that they have occasionally applied ice once or twice in a day, only when the need felt greatest. That is not really enough to know whether or not icing is going to help you.
Power icing is the name I’ve given to icing in megadoses — 20–50 applications of ice per day for 3–20 days at a time. As long each dose is no more than three minutes, and if you allow sufficient opportunity for tissue to warm up between applications, this is not risky — the worst case scenario is that you’ll waste your time.
I prescribe power icing only for serious chronic overuse injuries, especially the common tenditises (mostly tennis elbow), plantar fasciitis, some cases of carpal tunnel syndrome, iliotibial band syndrome, patellofemoral pain syndrome, and some kinds of shin splints.
Unfortunately, I have no scientific evidence that this works! It seems to make a certain amount of sense — more of a good thing? — and I have had some success with it with my own clients, a few slightly amazing recoveries with just a few days of power icing. Generally speaking, there is no excuse for prescribing self-treatments that lack even a clear rationale, let alone research evidence that they work. However, this one can be justified simply because it is low cost and low risk — so it is worth a shot!
I also prescribe power icing simply to make sure that people giving “normal” (lower dosage) icing a fair shot. Sometimes I will sneakily prescribe 50 applications of ice per day, knowing full well that the “type A” client isn’t actually going to fit in more than about one third that much — and that’s actually good enough for me.
The inevitable question with power icing is “how much?,” a conversation I find myself having with clients even after the numbers have been laid out. One way of clarifying is simply to say that it should be so much icing that you start to get sick of it. If you’re not thinking, “Wow, geez, this is really a lot of icing!” then you should probably do more.
If it doesn’t seem to be working within 3-5 days, it’s probably not going to do the trick for you. But I strongly recommend that you give it at least three days, so that you can be certain that you really have “tried icing!”
You can definitely enhance the effect of icing by using an ice gel afterwards. This is different than the gel-packs — it’s a goopy blue paste in a jar that feels very cold when you rub it into your skin. A variety of brands are available at any larger drugstore. There is one downside to this technique: if your goal is to do many sessions of icing, your skin tissue takes much longer to warm up between sessions when you’ve applied an ice gel.
If you want to try to chill deeper, thicker tissues, you need gentler, slower cooling. Raw ice is just too cold to leave on for longer periods. Ice packs to the rescue! Raw ice can probably chill the skinny joints of the hands and feet thoroughly and fairly quickly, and even knees and elbows, but hips or shoulders and spinal joints are deep.
You can also use fabric or towels to further ease the intensity of a gel pack, so that it can be used safely for even longer periods.
Thermoregulation of our tissues is extremely efficient: reflex responses in microvasculature can compensate quite a lot for changes in surface temperature. So what happens to the tissue temperature under ice? How cold does it get? At 1 centimetre? At 3? Can you chill a knee through and through?
Turns out you can! Most joints are less insulated and adapt to regulate their temperature, and can be cooled or heated much more easily. In experiments on arthritic knees, Oosterveld et al demonstrated that hot wax (ligno-paraffin) could raise deep knee temperatures by 1.7–3.5˚C … and the effect of cooling was even more dramatic, dropping the temperature as much as 9˚C. Brrr! That’s likely to have some effect on biology.
The effect on muscle is probably much less, but to date I’m unaware of any data that shows it.
To go deeper into this topic, see Icing, Heating & Tissue Temperature.
If you’ve read this far, you now know much more about icing than most people, and indeed most health care professionals. Congratulations!
I am a science writer, former massage therapist, and I was the assistant editor at ScienceBasedMedicine.org for several years. I have had my share of injuries and pain challenges as a runner and ultimate player. My wife and I live in downtown Vancouver, Canada. See my full bio and qualifications, or my blog, Writerly. You might run into me on Facebook or Twitter.
— Minor revision. Some more commentary about the popularity of debunking icing lately.
— Added evidence of efficacy of NSAIDs for common injuries.
— A series of substantial upgrades, details unlogged.
Bafflingly, therapeutic icing (cryotherapy) is one of really those basic things that you’d think modern medical science would have mastered by now, but no — not even close!
This is a 2008 review of the inadequate evidence: just six experiments, and only two of them any good, one with slightly positive results and the other showing nothing at all. So that’s two studies that showed little or no benefit, which is leaning towards bad news, but it’s not enough data to clinch it.
Four animal studies have showed that icing reduced swelling (and too much is harmful, duh!). That evidence is mildly encouraging, but of course we can’t take animal studies to the bank.
This really just isn’t enough data, and the bottom line is that we don’t know, which is what Collins concluded: “there is insufficient evidence.” A 2015 review (with a broader scope, see Malanga) had a similar non-conclusion, mostly confirming the absence of evidence.BACK TO TEXT
Although it’s 2015, “most recommendations for the use of heat and cold therapy are based on empirical experience,” not evidence, because the only evidence we have is still “limited.” Malinga, Yang, and Stark review the alleged benefits of heat and cold: pain relief for both, of course, plus ice reduces “blood flow, edema, inflammation, muscle spasm, and metabolic demand,” while heating increases “blood flow, metabolism, and elasticity of connective tissues.” Even these aren’t well-tested, and there are other possibilities that haven’t been tested at all.
Based on a handful of relevant trials, they concluded that “heat-wrap therapy provides short-term reductions in pain and disability in patients with acute low back pain and provides significantly greater pain relief of DOMS than does cold therapy.”
But the main take-home message from this paper is just “much more study needed.”BACK TO TEXT
What’s better for neck and back pain — ice or heat? This experiment, conducted at a university-based emergency department, compared the effectiveness of these two common treatments. Everyone studied received 400mg of ibuprofen orally and then thirty patients were given a half hour of either a heating pad or a cold pack.
The researchers concluded that adding heat or cold to ibuprofen therapy did not change the result. Both heat and cold resulted in “mild yet similar improvement in the pain severity.” They recommend that the “choice of heat or cold therapy should be based on patient and practitioner preferences and availability.”BACK TO TEXT
A well-written and important bird’s eye view of the subject of tendinopathy, presenting an updated way of thinking about the problem. Highly recommended, required reading for professionals.BACK TO TEXT
From the abstract: “If physicians acknowledge that overuse tendinopathies are due to tendinosis, as distinct from tendinitis, they must modify patient management …”BACK TO TEXT
BACK TO TEXT
It is generally thought that topical cooling can interfere with blood perfusion and may have positive effects on recovery from a traumatic challenge. This study examined the influence of topical cooling on muscle damage markers and hemodynamic changes during recovery from eccentric exercise. Eleven male subjects (age 20.2 ± 0.3 years) performed 6 sets of elbow extension at 85% maximum voluntary load and randomly assigned to topical cooling or sham groups during recovery in a randomized crossover fashion. Cold packs were applied to exercised muscle for 15 minutes at 0, 3, 24, 48, and 72 hours after exercise. The exercise significantly elevated circulating creatine kinase-MB isoform (CK-MB) and myoglobin levels. Unexpectedly, greater elevations in circulating CK-MB and myoglobin above the control level were noted in the cooling trial during 48-72 hours of the post-exercise recovery period. Subjective fatigue feeling was greater at 72 hours after topical cooling compared with controls. Removal of the cold pack also led to a protracted rebound in muscle hemoglobin concentration compared with controls. Measures of interleukin (IL)-8, IL-10, IL-1β, and muscle strength during recovery were not influenced by cooling. A peak shift in IL-12p70 was noted during recovery with topical cooling. These data suggest that topical cooling, a commonly used clinical intervention, seems to not improve but rather delay recovery from eccentric exercise-induced muscle damage.
“This drug is generally considered safe when used according to the directions on its labeling. But taking more than the recommended amount can cause liver damage, ranging from abnormalities in liver function blood tests, to acute liver failure, and even death.”BACK TO TEXT