The Dark Side of Epidural Steroid Injections

Epidural steroid injections carry a catalog of risks that patients rarely hear about during the five-minute conversation before signing a consent form. In 2014, the FDA issued a formal warning that injecting corticosteroids into the epidural space can cause blindness, stroke, paralysis, and death, and pointedly noted that it has never approved corticosteroids for this use.¹ Despite roughly nine million of these procedures performed each year in the United States alone, the gap between how casually they are offered and how serious the complications can be is striking.² The story behind that gap involves biology, economics, and a surprising amount of scientific uncertainty.

What the FDA Warning Actually Says

On April 23, 2014, the FDA released a safety alert stating that epidural corticosteroid injections “may result in rare, but serious adverse events, including loss of vision, stroke, paralysis, and death.”³ The agency also made two points that surprised many patients and even some practitioners. First, the effectiveness and safety of corticosteroids for epidural use have not been established through the FDA’s standard approval process. Second, a review of the FDA’s own Adverse Event Reporting System had turned up cases of spinal cord infarction, paraplegia, quadriplegia, cortical blindness, seizures, nerve injury, and brain swelling, all tied to epidural steroid injections.⁴ The warning did not ban the procedure. It required a new label. But the fact that one of the most commonly performed pain procedures in the country lacks formal FDA approval for the way it is used is a detail most patients never learn.

The Rare but Devastating Neurological Injuries

The headline-grabbing risks are the rarest ones: stroke, paralysis, and permanent spinal cord damage. These events are almost always traced to the needle accidentally penetrating or irritating an artery that feeds the spinal cord. When a particulate steroid (one that comes as a suspension of tiny solid particles rather than a clear solution) enters such an artery, the consequences can be catastrophic. An animal study showed that particulate steroids injected into the bloodstream cause red blood cells to rapidly change shape, clump together, and block small arterioles, cutting off blood flow.⁵ When that blockage happens in vessels supplying the spinal cord or brain, the result is an infarction, essentially a stroke of the spine or brain.

A large database study of cervical epidural steroid injections found that neurological complications occurred at a rate of about four to five per thousand patients, regardless of the needle approach used.⁶ That is low in percentage terms, but when you multiply it by the millions of injections performed annually, the absolute number of people harmed is not trivial. The risk is highest in the cervical spine, where critical arteries sit close to the injection targets.

Particulate Versus Non-Particulate Steroids

This distinction matters more than most patients realize. Particulate steroids like methylprednisolone and triamcinolone have been the traditional workhorses of epidural injections for decades. They linger at the injection site longer, which theoretically provides more sustained anti-inflammatory action. But their solid particles are what cause the arterial blockages described above. Non-particulate steroids like dexamethasone dissolve fully in solution and are far less likely to clog an artery if accidentally injected into one.

The trade-off is effectiveness. A trial comparing dexamethasone to methylprednisolone for lumbar radiculopathy found that the non-particulate steroid came close to matching the particulate version’s pain relief but showed a trend toward less reduction in pain and a shorter duration of action.⁷ A larger crossover trial comparing two non-particulate steroids, dexamethasone and betamethasone, found that only about a third of patients in either group qualified as effective responders at two weeks, and betamethasone produced modestly better pain and disability scores.⁸ For cervical and some thoracic injections, many guidelines now recommend non-particulate steroids as the default because the risk of catastrophic vascular injury outweighs the marginal gain in potency. For lumbar injections, where the arterial anatomy is somewhat less precarious, practice remains divided.

Blood Sugar Spikes in Diabetic Patients

If you have diabetes, your doctor may or may not mention that an epidural steroid injection will temporarily wreck your glucose control. A study of 30 diabetic patients found that average blood glucose nearly doubled after injection, jumping from about 160 mg/dL to about 286 mg/dL, and took roughly two days to return to baseline.⁹ That is a substantial spike. A separate study confirmed that blood glucose was significantly elevated one day after injection, especially in patients whose diabetes was already poorly controlled (those with higher baseline HbA1c levels) and those using insulin.¹⁰

For most diabetic patients, the spike is manageable with aggressive monitoring and temporary insulin adjustments over two to three days. But the problem is that many patients report not being warned about it and are caught off guard by dangerously high readings at home. If you have diabetes and are offered an epidural steroid injection, the conversation should include a concrete plan for monitoring your blood sugar for at least 48 hours afterward.

Adrenal Suppression and Hormonal Disruption

Your adrenal glands produce cortisol, the body’s natural stress hormone. Injecting a large dose of synthetic corticosteroid into the epidural space can suppress your adrenal glands’ own cortisol production for weeks. A study of patients undergoing spine surgery found that those who had received multiple epidural steroid injections beforehand had significantly lower baseline cortisol levels than patients who had not, and their cortisol did not rise appropriately after the stress of surgery.¹¹ That blunted stress response is clinically meaningful: it means the body may not be able to mount a proper reaction to physical stress, infection, or trauma during the weeks following the injections.

In extreme cases, repeated injections can push the body into a state resembling Cushing’s syndrome, with symptoms like facial swelling, weight gain, elevated blood pressure, and fragile skin. A case report documented this in a middle-aged woman after just three epidural steroid injections over four months.¹² Three injections in four months is not an unusual cadence in clinical practice, which is exactly the concern.

Bone Density Loss and Vertebral Fractures

Corticosteroids are well known to weaken bones when taken systemically, and the steroid from epidural injections does enter the systemic circulation. A systematic review found that epidural steroids were associated with significantly decreased bone mineral density in four of six studies examined and with an increased risk of vertebral fracture in one of two studies that assessed it.¹³ Significant reductions were linked to cumulative methylprednisolone-equivalent doses of 200 mg over one year and 400 mg over three years. Those are thresholds that a patient receiving multiple injections can reach quickly.

A more recent retrospective study reinforced the concern, finding that patients with lumbar degenerative disease who received epidural steroid injections had a roughly 3.5-fold higher rate of osteoporotic vertebral compression fractures compared with controls. After adjusting for other risk factors, the injections remained an independent risk factor, with the hazard ratio climbing further.¹⁴ The irony is hard to miss: a treatment aimed at back pain may be accelerating the very spinal fractures that cause back pain. Patients already at risk for osteoporosis, particularly postmenopausal women and older adults, should have a serious conversation about bone density monitoring before committing to a series of injections.

Infection, Abscess, and Arachnoiditis

Any time a needle is inserted near the spinal canal, infection is possible. Epidural abscesses and meningitis have been reported after epidural corticosteroid injections, though they remain rare. A literature review identified 14 reported cases of epidural abscess or meningitis tied to these injections; over half of the patients were immunocompromised, and the most common organism was Staphylococcus aureus.¹⁵ A case series of seven patients who developed meningitis after lumbar epidural steroid injections found that headache and fever typically appeared within hours of the procedure.¹⁶

A rarer and more insidious complication is arachnoiditis, an inflammatory condition of the membranes surrounding the spinal cord. A case report documented a patient who developed arachnoiditis after caudal epidural steroid injections, presenting with new sciatica on the opposite side, worsening back pain, urinary retention, and eventually dense foot drop from nerve root clumping.¹⁷ Arachnoiditis is chronic, painful, and notoriously difficult to treat. It is exceedingly uncommon, but the possibility of a permanent inflammatory condition caused by a procedure meant to reduce inflammation is a bitter irony that warrants disclosure.

Procedural Complications and Dural Puncture

The needle for an epidural injection is supposed to stop in the epidural space, the fat-filled area outside the membrane (dura) encasing the spinal cord and nerves. Sometimes it goes too far and punctures the dura. In a multi-institutional study of over 16,000 consecutive epidural steroid injections, dural punctures occurred in about 0.06% of cases overall, more commonly with the interlaminar approach than the transforaminal approach.¹⁸ In the cervical spine, where the anatomy is tighter, an analysis of nearly 1,000 CT-guided cervical interlaminar injections found inadvertent dural punctures in 1.4% of cases, with the highest rate at the C5-C6 level.¹⁹

A dural puncture leaks cerebrospinal fluid, which typically causes a severe positional headache that gets worse when you sit up and improves when you lie down. In most cases the puncture seals itself within days, though some patients need a blood patch procedure to close the leak. More concerning are the downstream neurological complications: intracranial hypotension, subdural hematomas, and cranial nerve palsies (particularly of the sixth cranial nerve, which controls a muscle that moves the eye).²⁰

The Mild Side Effects Nobody Warns You About

Beyond the serious complications, the same multi-institutional study of over 16,000 procedures catalogued the annoying-but-not-dangerous side effects that are far more common. Vasovagal reactions (feeling faint, lightheaded, or nauseated during or just after the injection) occurred in about 1.2% of procedures. Central steroid response, which includes insomnia, facial flushing, and non-positional headache, appeared in about 2.6% of patients. Around 2% of patients reported temporarily increased pain after the injection.²¹ None of these caused lasting problems, but they can be alarming if you are not expecting them.

Less appreciated are the mood effects. A case series documented patients who met formal diagnostic criteria for medication-induced mood disorders after epidural steroid injections, presenting with symptoms ranging from severe anxiety and agitation to depressive episodes.²² Corticosteroid-induced mood changes are well-documented with oral and intravenous steroids, and there is no reason they should be any different when the drug is absorbed systemically from the epidural space. Patients with a history of anxiety, depression, or bipolar disorder should be aware of this possibility.

How Well Do They Actually Work?

This is perhaps the most uncomfortable part of the story. The evidence for long-term benefit is weak. A meta-analysis of randomized controlled trials found that epidural steroid injections provided better pain relief than conservative treatment in the short and intermediate term but that the benefit did not hold at long-term follow-up.²³ There was no significant difference in functional improvement between injections and conservative care at any time point. In other words, patients felt somewhat less pain in the weeks after the shot, but they did not walk better, work better, or move better compared with people who received standard conservative care.

A retrospective study of patients with lumbar spinal stenosis painted a similar picture: about a third reported more than two months of pain relief, while nearly 30% reported no relief at all. A fifth of these patients eventually had surgery anyway.²⁴ On the question of whether injections prevent surgery, a systematic review found a borderline trend toward fewer surgeries in the first year but no reduction beyond one year.²⁵ The honest summary: the injections appear to speed up recovery that was probably going to happen on its own, buying time and comfort while nature does the actual healing.²⁶

A head-to-head trial comparing epidural steroid injections to gabapentin for lumbosacral radicular pain found no significant difference in the primary pain outcome at one or three months. The injection group had a slight edge in worst leg pain at one month, but that advantage disappeared by three months.²⁷ For cervical radicular pain, a trial found that combining epidural steroid injections with conservative treatment outperformed either approach alone, but the injection by itself did not beat conservative treatment.²⁸ The pattern across these studies is consistent: modest, short-lived pain relief, little functional improvement, and no clear advantage over cheaper, less invasive alternatives in the long run.

Why They Keep Getting Offered

Given all of the above, it is fair to ask why epidural steroid injections remain one of the most frequently performed pain procedures in the country. Part of the answer is that they genuinely do help some people in the short term, and a few weeks of meaningful pain relief can be life-changing for someone in crisis. The transforaminal approach, for instance, has been associated with roughly three times the odds of achieving a 50% or greater reduction in leg pain compared with the interlaminar approach.²⁹ For a patient who cannot function, that short-term bridge can matter.

But there are also structural forces at work. Reimbursement incentives, provider training norms, and patient expectations all perpetuate the use of procedures whose long-term value is questionable.³⁰ Epidural steroid injections are quick to perform, well-reimbursed, and easy to fit into a busy clinic schedule. Patients often arrive expecting an injection because they have heard about it from friends or the internet, and a provider who suggests physical therapy instead may seem like they are withholding treatment. The result is a self-reinforcing cycle in which the procedure’s popularity is sustained less by robust evidence and more by the economics and culture of pain management.

When Epidural Steroid Injections Still Make Sense

None of this means the procedure should be abandoned entirely. For a carefully selected patient experiencing acute radicular pain (pain shooting down a limb from a pinched or inflamed nerve root), a single epidural steroid injection under image guidance using a non-particulate steroid can provide weeks of relief that allow physical therapy and natural healing to take hold. The large multi-institutional study that tracked over 16,000 consecutive procedures found zero major adverse events, reinforcing that when performed by experienced practitioners with proper safety protocols, the immediate complication rate is very low.³¹

The trouble starts when injections become routine. Repeated injections pile up cumulative steroid exposure, increasing the risk of adrenal suppression, bone loss, and glucose instability. A patient who has had three injections without lasting relief is not a good candidate for a fourth. Yet many pain clinics offer three-injection series as a standard protocol, sometimes with little discussion of what happens if the first two do not work. The “dark side” of epidural steroid injections is not really about the freak catastrophic events. It is about the quiet accumulation of systemic harm in patients who keep returning for a procedure that was only ever designed to buy time.

Questions to Ask Before You Consent

If you are considering an epidural steroid injection, a few pointed questions can shift the conversation from routine procedure to informed decision:

Which steroid? Particulate or non-particulate? If particulate, why is non-particulate not being used, especially in the cervical spine?
Image guidance? Is fluoroscopy or CT being used to confirm needle placement? Blind injections without imaging carry higher complication rates.
Cumulative dose? How many injections have you had in the past year, and what is the total steroid dose? There are thresholds above which bone loss becomes a real concern.
Exit strategy? What happens if this injection does not provide lasting relief? Is there a plan beyond repeating the procedure?
Alternatives tried? Has physical therapy, oral medication, or activity modification been given a genuine trial? A randomized study found that gabapentin matched the injection’s pain relief at three months, without any of the injection-specific risks.³²

These are not questions designed to talk you out of the procedure. They are designed to ensure you are getting it for the right reason, at the right dose, with the right steroid, under the right guidance, and with a clear plan for what comes next.