Do Steroid Injections for Osteoarthritis, Shoulder Pain, Tennis Elbow, Plantar Fasciitis, and Back Pain Do More Harm Than Good?

Why do doctors continue to give patients treatments that can make them worse in the long run?

Discuss
Languages
Republish

Please consider volunteering to help out on the site.

The 1950 Nobel Prize in medicine went to those demonstrating the anti-inflammatory effects of steroids on arthritic joints. A Medicare sample of half a million patients with osteoarthritis in their knees found that about one-third were treated with at least one corticosteroid injection. This can help with pain in the short term, but it makes the condition worse in the long term.

Those getting steroid injections can end up with a worsening of pain, stiffness, and disability with accelerated joint deterioration and progression to total knee replacement surgery. Every cortisone injection may increase the absolute risk of knee replacement surgery by 9 percent at nine years compared to untreated patients. This is in addition to complications that include osteonecrosis (bone rot) and rapid joint destruction. In a randomized controlled trial, two years of steroid injections for knee osteoarthritis sufferers led to a significantly greater loss of cartilage volume—and, ironically, no better long-term pain relief—than a placebo injection of saline, which is basically water. The study may have been the “final nail in the coffin” for the practice.

Doctors have been injecting steroids like cortisone into arthritic knees for more than 70 years. In light of the available evidence, any case made today for the procedure, concluded a commentary on the cartilage loss study, “seems mostly to be based on rock solid prejudice.” An invited commentary in the journal of the Osteoarthritis Research Society International, entitled Pay Attention to the Evidence: In the Longer Term, Intraarticular Corticosteroid Injections Offer Only Harm for Knee Osteoarthritis reads: “The best time to have stopped using IACIs[MP1]  was years ago, and the second-best time to stop using them (and stop recommending them) is today.”

What about steroid injections for shoulder pain? The most common cause of shoulder pain is rotator cuff tendinosis, the chronic degeneration of tendons considered part of human aging. This is an encompassing term that includes descriptors like “impingement,” “rotator cuff fraying,” “partial thickness tears,” and “tendinitis.” A corticosteroid injection into the shoulder to suppress inflammation can provide short-term relief, explaining why millions of joints are injected with steroids like cortisone every year in the United States alone.

Up to 96 percent of musculoskeletal clinicians consider such injections to be an effective treatment. In fact, about one in five of those seeking help from their primary care physician for shoulder pain may get an injection on their first visit. Imaging-guided injections (for example, using ultrasound) don’t appear to offer any benefit over traditional targeting based on anatomic landmarks, and multiple injections at a time don’t appear to be more effective than a single shot.

Shoulder corticosteroid injections are said to “provide, at best, minimal pain relief in a small number of patients,” yet as many as one in five has their symptoms reduced to no more than mild pain at assessments between four and eight weeks after the injection. It’s true the benefit is gone by 12 weeks, but isn’t even a 20 percent chance of significant pain relief for a month or two worth getting the shot? Unfortunately, the downsides to shoulder corticosteroid injections are more than just the transient discomfort of the needle. Any potential short-term symptomatic relief has to be balanced against the potential for long-term damage to the joint, as corticosteroids may worsen the underlying disease process, accelerating tendon degeneration.

Corticosteroid injections have become one of the most commonly used treatments for chronic tendon lesions, even though we’ve known for more than fifty years that such injections can cause tendon necrosis (tissue death) starting as early as two to three days after injection. Animal studies also demonstrate that corticosteroid injections lead to decreased bone volume density, anchor pullout strength, load to failure, and tendon stiffness. The thought is that the steroids interfere with the healing process that is normally triggered by the inflammatory process. What about in people?

In an innovative way to ethically test the effects of steroid injections on tendons, fifteen patients scheduled for lower limb amputations had their Achilles tendons injected with hydrocortisone at varying intervals prior to surgery. The researchers demonstrated significant collagen fiber degeneration within three hours of the shot at the injection site (versus completely normal tendon architecture at saline-injected and non-injected control sites), demonstrating a “definite damaging local effect on human tendons.”

But we don’t tend to inject steroids directly into people’s tendons. It’s true that even oral exposure to corticosteroids like prednisone may increase the risk of tendon disorders, but more definitive evidence of harm wasn’t available until researchers took biopsies of rotator cuff tendons before and after patients with shoulder pain received steroid injections. The researchers saw the same kind of negative effects found in surgical specimens of tendons exposed to corticosteroids in a petri dish and concluded that steroid injections may, ironically, exacerbate the underlying disease responsible for the pain for which the glucocorticoids are prescribed to treat. What may make us feel better in the short-term may make us feel worse in the long-term.

Steroid injections in the shoulder may raise the risk of subsequent rotator cuff tears by as much as sevenfold. Not only may they increase the risk of surgical intervention, but steroid injections can also make future surgery riskier, increasing the likelihood of such complications as post-operative infections, which can have devastating consequences. This may be due to an immunosuppressive effect and/or the injections directly seeding the joint space with bacteria.

If steroid injections can be detrimental in the long run, why are they still endorsed in clinical practice guidelines? Continued use is chalked up to force of habit, financial incentives, and a desire for a “quick fix.”

What about steroid injections for tennis elbow? Even though they may do more harm than good in the long run, the potent anti-inflammatory effects of steroids can lead to short-term symptomatic improvements, which helps explain why steroid injections continue to be mainstays of treatment for tennis elbow.

Epicondylitis (inflammation of the epicondyle, the bony bump on the side of the elbow) may be a misnomer, since tennis elbow may be more of a degenerative rather than actively inflammatory condition. Lateral epicondylalgia (painful bony bump) or lateral elbow tendinopathy (tendon disease) may be more appropriate, or even the clunkier eECRB (enthesopathy of the extensor carpie radialis brevis tendon anchor point). anthesopathy means a diseased enthesis, which comes from an Ancient Greek word for “insertion.” The reason the name makes a difference is that calling it lateral epicondylitis reinforces the deeply entrenched dogma that it’s a largely inflammatory condition, which is why practitioners reach for the steroid syringe.

Corticosteroid injections (often mixed with a local anesthetic) can reduce pain in the short-term, but ultimately lead to significantly worse outcomes—worse pain and function, a lower rate of recovery, and quadruple the rate of recurrence. So your doctor is not only making it worse, but quadrupling the chances your pain comes back—now that’s a business model. As an editorial in a sports medicine journal was entitled, “Stop injecting corticosteroid into patients with tennis elbow, they are much more likely to get better by themselves!”

What about steroid injections for back pain? The first therapeutic epidural injection, in which a drug is injected directly into the spinal canal, was performed in 1885. Cocaine was injected into a man’s spine to treat “addiction to masturbation.” Today, epidural steroid injections may be the most common pain management procedure in the world.

Spinal injections are strongly recommended against by some but not all clinical practice guidelines. For example, the Spinal Intervention Society is, not surprisingly, all in favor of them, but a meta-analysis of twenty-five clinical trials of epidural steroid injections for sciatica found only a five-point drop on a zero-to-100 pain scale compared to placebo, which is considered far from clinically significant. For spinal stenosis cases, there doesn’t appear to be any effect at all over placebo, yet millions of epidural steroid injections continue to be given each year in the United States alone.

Even though there isn’t a clinically relevant benefit on average, about one in seven injected patients do experience some degree of short-term pain relief. So, why not give it a try? After all, you might be lucky number seven. Or you may be unlucky.

Minor complications are rare, occurring in two to 17 percent of injections, and most commonly manifesting as increased pain or persistent numbness. Of course, this is assuming the preparation your doctor uses isn’t contaminated with a crop rot mold. A devastating outbreak of fungal meningitis in the United States in 2012 was tied to epidural injections of mold-contaminated steroids, resulting in more than 650 cases of fungal infection and 39 deaths.

Other major complications are similarly extraordinarily rare, but can be equally serious. Should the needle puncture the spinal cord, it can result in paraplegia or quadriplegia. Accidentally injecting into a blood vessel can cause particulate steroid to embolize to the brain. Enough of these cases added up so the U.S. Food and Drug Administration added warning labels cautioning that epidural steroid injections can lead to blindness, stroke, paralysis, and death.

To prevent such catastrophic misplacement, the needle is typically inserted under fluoroscopy guidance, which is like a continuous real-time x-ray movie. This comes with the concomitant harms of radiation exposure. The dose of radiation for an epidural injection is approximately one-eighth that of a CT scan of the chest. To put that into perspective, the risk of dying from a cancer caused by a chest CT scan is the risk of death equivalent of dying in a car accident on a 2,000-mile drive. So, the radiation exposure from an epidural injection may be more like 250 miles’ worth of risk.

Among the most serious side effects of taking oral steroids are bone loss, skeletal fragility, and osteoporotic fracture risk. What might steroid injections do to spinal bone strength? A retrospective analysis of vertebral fractures following epidural steroid injections found that each successive injection appeared to increase the risk of bone fracture by 21 percent. So, even if the procedure goes well, there may be long-term, negative consequences even without considering the exposure to radiation.

Despite the likelihood that such injections offer risk with no clinical benefit, a study of Medicare recipients found that more than a million epidural steroid injections are still given for back pain every year. There was also an explosion in popularity for facet joint injections between the vertebrae, another procedure for which there is no evidence of efficacy over placebo.

Steroid injections are also commonly used to treat plantar fasciitis foot pain. They offer little or no pain relief compared to placebo, though—even in the short-term —and can cause rare but crippling complications. These include nerve injuries, bone infections, and, months later, plantar fat pad atrophy. They may even cause plantar fascial rupture in as many as one in ten individuals approximately ten weeks after injection. Why then do doctors continue to administer them? They trust experience over evidence. Health professionals often see with their own eyes that their injected patients experience relief, without understanding they would have likely experienced about the same relief without undergoing any real treatment at all.

In a section of my new book How Not to Hurt entitled “When to Say Neigh to Horse Shampoo,” I discuss how our natural inclination is to attribute whatever happens to whatever we did, a phenomenon known as post hoc bias or the post hoc ergo propter hoc fallacy, which translates to “after this, therefore because of this.” A study of post hoc bias on treatment decisions showed just how powerful this intuition is by randomizing hundreds of people to evaluate different scenarios:

“DK is a 75-year-old retired professor. He is going bald and starts applying horse shampoo once a day. However, researchers typically agree horse shampoo does not promote hair growth. A couple weeks later his hair seems either better or unchanged. Suppose he is undecided and asks you, would you recommend he continue or discontinue the horse shampoo?”

Half were told his hair was “better” and the other half “unchanged.” In the unchanged group, fewer than one in 10 (seven percent) said he should continue the horse shampoo, whereas most (65 percent) in the “better” group said he should stick with it.

Other scenarios described people with intermittent pain or insomnia happening to feel better versus unchanged the week after wearing a copper bracelet or taking a “sugar powder supplement.” There were the same large swings in opinion, with around 80 percent encouraging the “better” groups to stick with it versus only a small minority for those given the “unchanged” reading.

This can be as innocent as the “don’t mess with success” aphorism, but you can imagine how this could dupe the masses into questionable quackery or overinflate the apparent effectiveness of mainstream measures that may be just as dubious. The researchers conclude: “Our study highlights a core weakness around intuitions in health care.”

Below is an approximation of this video’s audio content. To see any graphs, charts, graphics, images, and quotes to which Dr. Greger may be referring, watch the above video.

Motion graphics by Avo Media

Please consider volunteering to help out on the site.

The 1950 Nobel Prize in medicine went to those demonstrating the anti-inflammatory effects of steroids on arthritic joints. A Medicare sample of half a million patients with osteoarthritis in their knees found that about one-third were treated with at least one corticosteroid injection. This can help with pain in the short term, but it makes the condition worse in the long term.

Those getting steroid injections can end up with a worsening of pain, stiffness, and disability with accelerated joint deterioration and progression to total knee replacement surgery. Every cortisone injection may increase the absolute risk of knee replacement surgery by 9 percent at nine years compared to untreated patients. This is in addition to complications that include osteonecrosis (bone rot) and rapid joint destruction. In a randomized controlled trial, two years of steroid injections for knee osteoarthritis sufferers led to a significantly greater loss of cartilage volume—and, ironically, no better long-term pain relief—than a placebo injection of saline, which is basically water. The study may have been the “final nail in the coffin” for the practice.

Doctors have been injecting steroids like cortisone into arthritic knees for more than 70 years. In light of the available evidence, any case made today for the procedure, concluded a commentary on the cartilage loss study, “seems mostly to be based on rock solid prejudice.” An invited commentary in the journal of the Osteoarthritis Research Society International, entitled Pay Attention to the Evidence: In the Longer Term, Intraarticular Corticosteroid Injections Offer Only Harm for Knee Osteoarthritis reads: “The best time to have stopped using IACIs[MP1]  was years ago, and the second-best time to stop using them (and stop recommending them) is today.”

What about steroid injections for shoulder pain? The most common cause of shoulder pain is rotator cuff tendinosis, the chronic degeneration of tendons considered part of human aging. This is an encompassing term that includes descriptors like “impingement,” “rotator cuff fraying,” “partial thickness tears,” and “tendinitis.” A corticosteroid injection into the shoulder to suppress inflammation can provide short-term relief, explaining why millions of joints are injected with steroids like cortisone every year in the United States alone.

Up to 96 percent of musculoskeletal clinicians consider such injections to be an effective treatment. In fact, about one in five of those seeking help from their primary care physician for shoulder pain may get an injection on their first visit. Imaging-guided injections (for example, using ultrasound) don’t appear to offer any benefit over traditional targeting based on anatomic landmarks, and multiple injections at a time don’t appear to be more effective than a single shot.

Shoulder corticosteroid injections are said to “provide, at best, minimal pain relief in a small number of patients,” yet as many as one in five has their symptoms reduced to no more than mild pain at assessments between four and eight weeks after the injection. It’s true the benefit is gone by 12 weeks, but isn’t even a 20 percent chance of significant pain relief for a month or two worth getting the shot? Unfortunately, the downsides to shoulder corticosteroid injections are more than just the transient discomfort of the needle. Any potential short-term symptomatic relief has to be balanced against the potential for long-term damage to the joint, as corticosteroids may worsen the underlying disease process, accelerating tendon degeneration.

Corticosteroid injections have become one of the most commonly used treatments for chronic tendon lesions, even though we’ve known for more than fifty years that such injections can cause tendon necrosis (tissue death) starting as early as two to three days after injection. Animal studies also demonstrate that corticosteroid injections lead to decreased bone volume density, anchor pullout strength, load to failure, and tendon stiffness. The thought is that the steroids interfere with the healing process that is normally triggered by the inflammatory process. What about in people?

In an innovative way to ethically test the effects of steroid injections on tendons, fifteen patients scheduled for lower limb amputations had their Achilles tendons injected with hydrocortisone at varying intervals prior to surgery. The researchers demonstrated significant collagen fiber degeneration within three hours of the shot at the injection site (versus completely normal tendon architecture at saline-injected and non-injected control sites), demonstrating a “definite damaging local effect on human tendons.”

But we don’t tend to inject steroids directly into people’s tendons. It’s true that even oral exposure to corticosteroids like prednisone may increase the risk of tendon disorders, but more definitive evidence of harm wasn’t available until researchers took biopsies of rotator cuff tendons before and after patients with shoulder pain received steroid injections. The researchers saw the same kind of negative effects found in surgical specimens of tendons exposed to corticosteroids in a petri dish and concluded that steroid injections may, ironically, exacerbate the underlying disease responsible for the pain for which the glucocorticoids are prescribed to treat. What may make us feel better in the short-term may make us feel worse in the long-term.

Steroid injections in the shoulder may raise the risk of subsequent rotator cuff tears by as much as sevenfold. Not only may they increase the risk of surgical intervention, but steroid injections can also make future surgery riskier, increasing the likelihood of such complications as post-operative infections, which can have devastating consequences. This may be due to an immunosuppressive effect and/or the injections directly seeding the joint space with bacteria.

If steroid injections can be detrimental in the long run, why are they still endorsed in clinical practice guidelines? Continued use is chalked up to force of habit, financial incentives, and a desire for a “quick fix.”

What about steroid injections for tennis elbow? Even though they may do more harm than good in the long run, the potent anti-inflammatory effects of steroids can lead to short-term symptomatic improvements, which helps explain why steroid injections continue to be mainstays of treatment for tennis elbow.

Epicondylitis (inflammation of the epicondyle, the bony bump on the side of the elbow) may be a misnomer, since tennis elbow may be more of a degenerative rather than actively inflammatory condition. Lateral epicondylalgia (painful bony bump) or lateral elbow tendinopathy (tendon disease) may be more appropriate, or even the clunkier eECRB (enthesopathy of the extensor carpie radialis brevis tendon anchor point). anthesopathy means a diseased enthesis, which comes from an Ancient Greek word for “insertion.” The reason the name makes a difference is that calling it lateral epicondylitis reinforces the deeply entrenched dogma that it’s a largely inflammatory condition, which is why practitioners reach for the steroid syringe.

Corticosteroid injections (often mixed with a local anesthetic) can reduce pain in the short-term, but ultimately lead to significantly worse outcomes—worse pain and function, a lower rate of recovery, and quadruple the rate of recurrence. So your doctor is not only making it worse, but quadrupling the chances your pain comes back—now that’s a business model. As an editorial in a sports medicine journal was entitled, “Stop injecting corticosteroid into patients with tennis elbow, they are much more likely to get better by themselves!”

What about steroid injections for back pain? The first therapeutic epidural injection, in which a drug is injected directly into the spinal canal, was performed in 1885. Cocaine was injected into a man’s spine to treat “addiction to masturbation.” Today, epidural steroid injections may be the most common pain management procedure in the world.

Spinal injections are strongly recommended against by some but not all clinical practice guidelines. For example, the Spinal Intervention Society is, not surprisingly, all in favor of them, but a meta-analysis of twenty-five clinical trials of epidural steroid injections for sciatica found only a five-point drop on a zero-to-100 pain scale compared to placebo, which is considered far from clinically significant. For spinal stenosis cases, there doesn’t appear to be any effect at all over placebo, yet millions of epidural steroid injections continue to be given each year in the United States alone.

Even though there isn’t a clinically relevant benefit on average, about one in seven injected patients do experience some degree of short-term pain relief. So, why not give it a try? After all, you might be lucky number seven. Or you may be unlucky.

Minor complications are rare, occurring in two to 17 percent of injections, and most commonly manifesting as increased pain or persistent numbness. Of course, this is assuming the preparation your doctor uses isn’t contaminated with a crop rot mold. A devastating outbreak of fungal meningitis in the United States in 2012 was tied to epidural injections of mold-contaminated steroids, resulting in more than 650 cases of fungal infection and 39 deaths.

Other major complications are similarly extraordinarily rare, but can be equally serious. Should the needle puncture the spinal cord, it can result in paraplegia or quadriplegia. Accidentally injecting into a blood vessel can cause particulate steroid to embolize to the brain. Enough of these cases added up so the U.S. Food and Drug Administration added warning labels cautioning that epidural steroid injections can lead to blindness, stroke, paralysis, and death.

To prevent such catastrophic misplacement, the needle is typically inserted under fluoroscopy guidance, which is like a continuous real-time x-ray movie. This comes with the concomitant harms of radiation exposure. The dose of radiation for an epidural injection is approximately one-eighth that of a CT scan of the chest. To put that into perspective, the risk of dying from a cancer caused by a chest CT scan is the risk of death equivalent of dying in a car accident on a 2,000-mile drive. So, the radiation exposure from an epidural injection may be more like 250 miles’ worth of risk.

Among the most serious side effects of taking oral steroids are bone loss, skeletal fragility, and osteoporotic fracture risk. What might steroid injections do to spinal bone strength? A retrospective analysis of vertebral fractures following epidural steroid injections found that each successive injection appeared to increase the risk of bone fracture by 21 percent. So, even if the procedure goes well, there may be long-term, negative consequences even without considering the exposure to radiation.

Despite the likelihood that such injections offer risk with no clinical benefit, a study of Medicare recipients found that more than a million epidural steroid injections are still given for back pain every year. There was also an explosion in popularity for facet joint injections between the vertebrae, another procedure for which there is no evidence of efficacy over placebo.

Steroid injections are also commonly used to treat plantar fasciitis foot pain. They offer little or no pain relief compared to placebo, though—even in the short-term —and can cause rare but crippling complications. These include nerve injuries, bone infections, and, months later, plantar fat pad atrophy. They may even cause plantar fascial rupture in as many as one in ten individuals approximately ten weeks after injection. Why then do doctors continue to administer them? They trust experience over evidence. Health professionals often see with their own eyes that their injected patients experience relief, without understanding they would have likely experienced about the same relief without undergoing any real treatment at all.

In a section of my new book How Not to Hurt entitled “When to Say Neigh to Horse Shampoo,” I discuss how our natural inclination is to attribute whatever happens to whatever we did, a phenomenon known as post hoc bias or the post hoc ergo propter hoc fallacy, which translates to “after this, therefore because of this.” A study of post hoc bias on treatment decisions showed just how powerful this intuition is by randomizing hundreds of people to evaluate different scenarios:

“DK is a 75-year-old retired professor. He is going bald and starts applying horse shampoo once a day. However, researchers typically agree horse shampoo does not promote hair growth. A couple weeks later his hair seems either better or unchanged. Suppose he is undecided and asks you, would you recommend he continue or discontinue the horse shampoo?”

Half were told his hair was “better” and the other half “unchanged.” In the unchanged group, fewer than one in 10 (seven percent) said he should continue the horse shampoo, whereas most (65 percent) in the “better” group said he should stick with it.

Other scenarios described people with intermittent pain or insomnia happening to feel better versus unchanged the week after wearing a copper bracelet or taking a “sugar powder supplement.” There were the same large swings in opinion, with around 80 percent encouraging the “better” groups to stick with it versus only a small minority for those given the “unchanged” reading.

This can be as innocent as the “don’t mess with success” aphorism, but you can imagine how this could dupe the masses into questionable quackery or overinflate the apparent effectiveness of mainstream measures that may be just as dubious. The researchers conclude: “Our study highlights a core weakness around intuitions in health care.”

Below is an approximation of this video’s audio content. To see any graphs, charts, graphics, images, and quotes to which Dr. Greger may be referring, watch the above video.

Motion graphics by Avo Media

Doctor's Note

For more on osteoarthritis, see:

For more on shoulder pain, see:

For more on plantar fasciitis, see:

Check out my forthcoming book How Not to Hurt, available now for preorders. (All proceeds received from the book are donated directly to charity.)

If you haven't yet, you can subscribe to our free newsletter. With your subscription, you'll also get notifications for just-released blogs and videos. Check out our information page about our translated resources.

Subscribe to our free newsletter and receive the Preface and Introduction of Dr. Greger’s upcoming book How Not to Hurt.

Pin It on Pinterest

Share This