Sciatica: what is actually pressing on the nerve, and what each cause does over a year

The symptoms that mean stop reading and go to hospital

These override every other thing on this page. Go to an emergency department now, not tomorrow, if you have:

• Loss of bladder or bowel control, or new trouble starting or stopping urination
• Numbness in the area a bicycle seat would touch — groin, inner thighs, buttocks
• Weakness in both legs, or weakness in one leg that is getting worse by the day
• Pain in both legs at once with any of the above

That cluster is cauda equina syndrome — the bundle of nerve roots at the bottom of the spinal cord being squeezed. A 2026 review of it draws a line that matters enormously for your outcome: whether you can still empty your bladder when you present. Incomplete cases, where bladder function is impaired but working, do far better than cases that arrive already unable to urinate. The window is measured in hours.

Nobody reading a page on the internet should be managing that. Everything below assumes you have ruled it out.

What is actually pressing on the nerve, and why it changes everything

The word sciatica covers at least four different situations that behave differently:

A disc pushing out and touching a nerve root. The commonest cause by a wide margin. The soft centre of the spinal disc pushes through the tough outer ring and contacts the nerve. This one has a natural history that works in your favour, and the further out the fragment has travelled, the better it tends to clear.

A narrowed nerve channel. The bony canal the nerve passes through has closed down, usually with age. Pain comes on with walking and eases when you sit or lean forward. This does not resorb, because nothing has extruded that the body can clear away.

The nerve compressed outside the spine. Deep buttock muscles, most often the piriformis, squeezing the nerve after it leaves the spine. Local to the buttock, and it responds to entirely different things than a disc does.

Something else. Infection, a tumour, a fracture, hip joint disease that refers pain down the leg. Rare, and the reason unexplained sciatica that does not follow the expected course gets imaged rather than treated for longer.

Two people with identical leg pain and different causes on this list should not be doing the same things. Most of the confusion in what people are told about sciatica comes from advice generated for one of these being applied to another.

Your leg is weaker than you think, and your walk hides it

A 2026 study measured people with one-sided sciatica from a disc herniation while they walked, comparing the affected leg to the good one. The affected leg was measurably weaker. The walk itself looked normal — no meaningful difference in any of the force or timing measurements.

That combination is worth understanding, because it explains something people get wrong about their own recovery. Your body compensates well enough to hide the strength loss from you and from anyone watching you walk across a room. So "I'm walking fine now" is not evidence the nerve has recovered, and going back to loading that leg because your gait looks normal is how people re-injure themselves at week six. The weakness has to be measured, not eyeballed.

What the strongest trial in this field actually found

One randomised trial dominates this literature, and its result is more interesting than either side usually reports. It took 283 people with severe sciatica that had already lasted 6 to 12 weeks and assigned them either to early surgery or to continued conservative care, with surgery held in reserve.

Two findings, in order of importance to you.

First: over a third of the people assigned to wait ended up having surgery anyway. Of 142 assigned to conservative care, 55 crossed over and were operated on, after an average of 18.7 weeks.

Second: across the whole first year, the two groups' disability scores were not meaningfully different (P=0.13). Surgery got people out of pain faster. It did not get them to a better place at twelve months.

So the honest framing of surgery for disc-caused sciatica is not better versus worse. It is sooner versus later, with a real operation and its real risks as the price of sooner — and with a one-in-three chance that waiting ends in the operation regardless.

The large American cohort study that ran alongside this reached the same conclusion and then said the quiet part about its own design: people who chose surgery reported bigger improvements, but they chose it, and the authors warned plainly that comparisons like that are open to confounding and must be read cautiously.

That caution applies with more force to every before-and-after testimonial you will read about anything sold for sciatica. People who try a thing are different from people who do not.

The finding that should change how you take anti-inflammatories

When a disc fragment breaks free, the body clears it. Immune cells move in, release signalling molecules and enzymes that break tissue down, and the fragment shrinks. That process is inflammation doing something useful.

A 2026 review of how and when disc material resorbs states the awkward consequence directly: the immune response that removes the fragment is inflammatory, and standard anti-inflammatory treatment may work against it.

This is not a reason to be in unnecessary pain. It is a reason to know what you are buying. Weeks of scheduled anti-inflammatory medication to sit more comfortably may be trading against the mechanism that is clearing the thing causing the pain. That trade may be worth making. It should be a decision, not an accident.

What actually helps, ranked by what is behind it

Physiotherapy — the best-supported non-surgical option, and it is not overwhelming. A 2025 review of what predicts success without surgery found a benefit on leg pain in 5 of the 7 studies at low to moderate risk of bias. The same review named the single most useful prognostic fact anyone can tell you: how bad you are at the start predicts how you do without surgery.

If your pain and disability are severe at presentation, the conservative route is less likely to work for you than it is for someone with milder symptoms — and knowing that early changes the conversation about timing rather than leaving you to discover it at month five.

Acupuncture, with a real comparison and a real number. A 2026 secondary analysis compared acupuncture with traction for leg pain during the settling phase after a disc herniation. Rated excellent in 67.57% of the acupuncture group against 47.22% of the traction group (P = .041). Back pain scores and disability scores did not differ. Side effects occurred in 8.1% of the acupuncture group and were mild and short-lived.

Read what that does and does not say. It beat traction on leg pain. It did not beat a placebo, because that was not the comparison, and it did nothing measurable for back pain or overall function.

Epidural steroid injection — buys time, with an unstudied gap where the important question is. The largest synthesis assembled 72 randomised trials. Injections relieve pain in the short term. The authors then name the gap that matters most to somebody deciding what to do: whether the injection tells you anything about whether you will need surgery has barely been studied.

What to be sceptical of, and what it costs you to try anyway

Gabapentin and pregabalin. These are prescribed constantly for nerve-related leg pain. A double-blind randomised trial in narrowed-canal leg pain found they did not beat placebo, and the people taking them had more side effects — dizziness and sedation especially.

The cost of trying them is not zero, and it is highest in exactly the group most likely to be offered them. A 2025 review and pooled analysis of anti-epileptic drugs used for nerve pain in older adults found a clinically significant fall risk that rose with dose, with gabapentin looking safer than pregabalin.

Oral steroid tablets. A 2026 review of the randomised evidence in pinched-nerve leg pain found two placebo-controlled trials showing no difference in pain or function, quality-of-life results that pointed in inconsistent directions, no difference in how many people ended up having surgery, and more short-term side effects in the steroid group.

Ten people with sciatica, counted

Take the trial evidence above and put it in units of people, because percentages hide what the year actually looks like.

Ten people present with severe sciatica that has already lasted six weeks and is caused by a disc. Left on the conservative route with surgery available if needed: roughly four will end up having the operation anyway, at around the four-and-a-half-month mark. The other six get there without it. At twelve months, the group that had early surgery and the group that waited are in about the same place on disability — but the early-surgery group got out of severe leg pain substantially sooner. Of the ten, the ones with the worst pain and function at the start are the most likely to be among the four.

Nothing in that paragraph tells you which of the ten you are. It tells you what the two roads look like, so that when someone offers you a third one you can ask what it beats.

Where a compound could and could not act on this

Substances get discussed for this — the tissue-repair peptides most often. This page is about sciatica, so the accurate statement is short: none of the compounds commonly promoted for nerve or disc recovery has randomised human trial evidence in sciatica. There is no trial to summarise here, and a page that implied otherwise by talking around it at length would be doing the thing this site exists not to do. Where a compound has laboratory or animal work behind a specific mechanism, that argument belongs on the page that names both the compound and the condition, and it belongs there with its evidence tier stated. The operator of this site has a commercial interest in compounds of that kind, which is exactly why the claim on this page is the absence of the trials rather than a case for the compounds.

Before you agree to any treatment

1. Rule out the emergency list at the top. Bladder, bowel, saddle numbness, both legs, worsening weakness. That is a same-day hospital visit, not a phone call.
2. Find out which of the four causes you have. The natural history, the useful treatments and the timeline are different for each one. An examination, and imaging if the picture does not fit, is what separates them.
3. Get your leg strength measured, not estimated. Your walk will look normal before the nerve has recovered.
4. Decide about anti-inflammatories deliberately. If the cause is an extruded fragment, sustained anti-inflammatory use may be working against the process clearing it.
5. Start physiotherapy, and be honest with yourself about your starting severity. It is the best-supported conservative option, and severe symptoms at the start are the clearest signal that it may not be enough.
6. Treat every claim, including every one above, against the baseline. At one year, the operated and the unoperated groups landed in the same place. Anything offered to you has to beat that, or be honest that it is selling you speed inside a recovery that was likely coming anyway.