Epidural Injections for Lumbar Spinal Stenosis: What They Can Change

An injection is often described as the step that sits between exercises and an operation. That description does too much work. For lumbar stenosis, an epidural steroid injection has a narrower job: it tries to settle swelling and pain around nerves that are already in a tight space.

It does not widen the canal. It is not a cure, and it is not a requirement before surgery can be discussed.

Lumbar spinal stenosis is narrowing of the space that holds the spinal cord and the nerves, which can squeeze them. The squeeze comes from bone and ligament. A medicine placed nearby can change how irritable the nerves are for a while. It cannot remodel the passage they sit in.

What the injection is trying to change

An epidural steroid injection places steroid medicine in the epidural space, the area just outside the covering of the spinal nerves. In this condition the aim is local. AAOS OrthoInfo says epidural cortisone may decrease swelling and pain, and may reduce numbness, but not weakness.

Cortisone is a steroid. Swelling and pain are the targets it may affect. A foot that slaps the ground, or a leg that gives way, sits outside what AAOS says the medicine changes.

The narrowing itself stays. AAOS is plain that nonsurgical care does not improve the narrowing, even though many people report symptom relief. An injection is one of the nonsurgical options on that list, alongside physical therapy and anti-inflammatory medicines.

Feeling easier on a walk afterward, if that happens, is a change in symptoms. A later scan that still looks narrow is what the anatomy would lead you to expect. The goal was never a wider canal.

Numbness and weakness should not be treated as one complaint when you judge a response. Less tingling, with strength unchanged, fits the split AAOS describes. Less pain, with the same heavy limit on walking, is still a symptom change, and it is a partial one.

No change at all is also a result. None of those results rewrites the bone.

People may or may not have back pain with stenosis. Pressure on the nerves is what produces burning pain, numbness, or tingling in the buttocks or legs, called sciatica. An injection discussed for back ache alone, with quiet legs, is being asked to treat a different problem from the nerve picture AAOS attaches to the medicine.

Say which symptom you hope will move. “My stenosis” is not specific enough to judge a response against.

What the LESS trial compared

The clearest randomized test of this injection for stenosis is the LESS trial, published in the New England Journal of Medicine. Adults with lumbar spinal stenosis and moderate to severe leg pain were assigned by chance to one of two epidural injections. One group received a glucocorticoid, a steroid medicine, plus lidocaine, a numbing medicine. The other group received lidocaine alone.

At six weeks, the groups did not differ significantly on the Roland-Morris Disability Questionnaire or on leg-pain intensity. The Roland-Morris Disability Questionnaire is a set of statements about how back problems limit everyday tasks, such as walking and self-care. “Did not differ significantly” means the gap between the groups was not large enough to count as a clear separation on those two measures.

It does not mean every person stayed the same. Both groups improved from the start of the trial.

That pair of findings is easy to collapse into a slogan, and the slogan is usually wrong. “Injections do nothing” ignores the improvement from baseline in both groups, and it ignores the numbing medicine the comparison group received. “Steroid injections work” ignores the missing difference between steroid-plus-lidocaine and lidocaine alone, on the measures the trial chose, at six weeks.

The honest sentence is longer. People tended to report less disability and less leg pain than when they enrolled. Adding the steroid did not produce a significant further difference on those outcomes at that time point.

The trial does not say the canal changed. An injection does not change the bony narrowing. LESS was a study of symptoms and of a disability questionnaire in adults who already had stenosis and moderate to severe leg pain.

It was not a study of people with a narrow scan and no leg pain. It was not a study of weakness as the main problem. It does not set a personal forecast for the next person offered the procedure.

Six weeks is the comparison that was reported here. It is not a claim about every later month.

What a patient-education page adds

AAOS and the trial are not answering the same question, so they should not be forced into one verdict. AAOS tells patients what cortisone may do: decrease swelling and pain, and possibly reduce numbness, without treating weakness. It also sets a ceiling of no more than three injections in a year.

That is a ceiling. It is not a course you are supposed to complete, and it is not evidence that a third injection adds something the first one did not.

The trial asks whether, at six weeks, the steroid added a measurable difference beyond lidocaine on two outcomes. It did not find one. A page that says the medicine may help, and a trial that did not find an added benefit on disability or leg-pain intensity at six weeks, can both be reported.

The page is a general description of intent and of a yearly limit. The trial is a comparison in a defined group, at a defined time, on defined measures. Using the page to erase the trial, or the trial to claim the medicine can never ease swelling, overreaches both.

Other nonsurgical options sit on the same AAOS list, and they have the same limit on the anatomy. Physical therapy does not open the canal. NSAIDs, nonsteroidal anti-inflammatory drugs, are named with cautions about the stomach and the kidneys.

Whether one is appropriate is a decision for the treating clinician. This page does not suggest starting or stopping a medicine. Traction, acupuncture, and chiropractic care are mentioned by AAOS as options with less evidence, with extra caution if you have osteoporosis, bones that have thinned and break more easily, or a herniated disc, a disc that has pushed beyond its usual space.

Those are not injections. They are a reminder that “nonsurgical” is a group of different tools, none of which widens the passage.

How walking limits are recorded, if an injection is being judged by whether errands get longer, is covered in Walking Distance and Lumbar Spinal Stenosis: A Limit, Not a Grade. A longer walk, if it happens, is a symptom report. It is not a before-and-after measurement of the canal.

Not a requirement, and not a cure

Surgery, in the AAOS description, is generally reserved for poor quality of life from pain and weakness, including difficulty walking for extended periods. The operations named are laminectomy, removal of a portion of the bone at the back of the canal, done open or through smaller incisions, and fusion if there is instability. Instability means too much motion between vertebrae.

A slip seen on bending X-rays is a separate finding and does not, by itself, write the surgical plan. The comparison of those operations belongs in Lumbar Spinal Stenosis Surgery Options Compared.

Nothing on the AAOS page makes an epidural injection the required stop before that conversation. You can be a person for whom surgery is being discussed because walking and strength have already made daily life poor, without a stack of injections behind you. You can also be a person who has had an injection and is not a surgical candidate, because the symptoms and the goals do not match what AAOS reserves surgery for.

The injection does not graduate you, and skipping it does not disqualify you. Diagnosis does not dictate the procedure either. Stenosis on a scan, or even stenosis as a clinical diagnosis, is not an order for an operation or for a needle.

A JAMA review keeps the diagnosis itself in order. Clinical lumbar stenosis is a history of back and leg pain provoked by extension and relieved by flexion, then confirmed with imaging. Extension is the upright posture.

Flexion is bending forward. An injection offered for a label, without that history, is not aimed at the syndrome the review describes. In the Framingham imaging work, narrowing was common and often had no symptoms. A painless narrow canal is not an injection indication hiding in the report.

The wider question of symptom change without an operation, including what “better” has meant for groups of people, is the subject of Can Lumbar Spinal Stenosis Get Better Without Surgery?. An injection is one symptom tool inside that larger topic. It should not be asked to stand for all nonsurgical care, and nonsurgical care should not be asked to stand for a cure of the narrowing.

How to judge a response without inventing a rule

Decide in advance which symptom is supposed to move. Leg pain, walking tolerance, numbness, and weakness are not one outcome. AAOS already separates weakness from pain and numbness.

If you only ask “did it work?”, a quieter calf and a foot that still slaps will be averaged into a feeling. Write the pieces down separately, the way you would write a walking limit: what changed, on which days, and what did not change at all.

A response is not a forecast about surgery. Feeling better for a while does not prove an operation would help, and feeling no different does not prove an operation would fail. Those are different questions, asked of different treatments, in a person whose canal is still narrow either way.

The LESS trial does not answer the surgical question. It compared two injections.

A response is also not a schedule. The yearly ceiling from AAOS is a limit on how often the injection is repeated, not a reason to repeat it because the calendar allows another one. If function did not change, another injection is a new decision, not an automatic next dose.

The treating clinician makes that decision with you. Neighbors’ stories about a certain number of shots are not a plan.

One cluster should not be routed through an injection at all. AAOS describes cauda equina syndrome as rare and urgent: numbness in the private area, loss of bowel or bladder control, and legs that lose strength so walking fails. That needs prompt care. It is not a flare to cover with the next available procedure.

Questions that keep the injection in proportion

A few questions are enough before you treat the result as an answer.

Which symptom are we hoping to change, and which symptoms is the medicine not expected to change? If weakness is the reason you came, AAOS has already said cortisone does not treat it. The visit should say that out loud.

What was compared in the trial that fits me, and what was it compared with? For moderate to severe leg pain from lumbar stenosis, LESS compared steroid plus lidocaine with lidocaine alone, and looked at disability and leg-pain intensity at six weeks. If your situation is different from that group, say so. The result does not stretch by enthusiasm.

What else is on the list, and what will we use as the sign that this tool was or was not enough? Physical therapy, medicines your own clinician considers appropriate, and the injection are different tools with the same anatomical limit. The sign is usually something you can describe, such as the walk, not a newer adjective on a scan.

What would move the conversation toward an operation, or away from one, without using the injection as a gate? Poor quality of life from pain and weakness, including difficulty walking for extended periods, is the AAOS frame for reserving surgery. An injection can be irrelevant to that frame, or one experience inside it. It is not the frame.

The questions to prepare for a visit are a way to bring that list. They do not recommend the injection, and they do not recommend skipping it. The procedure remains a symptom tool with a defined target, a trial that did not show an added steroid benefit on two measures at six weeks, a yearly ceiling, and no effect on the width of the canal.