Peripheral neuropathy: the longest nerves fail first, so it starts in both feet
It starts in the toes because they are furthest away. The nerves running to your feet are the longest cells in your body — a single one can run from the base of your spine to your big toe — and anything that starves or poisons a nerve hits the far end first, where the supply line is longest. That is why this begins in both feet at once rather than in one leg, why it creeps upward over months and years rather than jumping about, and why the hands only join in later, when the damage has climbed the legs to roughly the height of the knee.
That pattern is worth memorising, because it is what separates this from a pinched nerve. A pinched nerve in the back gives you one leg, in a stripe. This gives you both feet, evenly, in the shape of a sock.
The first job is not treating the pain. It is finding a cause you can reverse.
Most of what gets written about neuropathy is about pain medication. That skips the step that matters more, because some causes of this are correctable and the nerves recover if you correct them early enough.
The one most often missed is vitamin B12, and the reason it is missed is that a common diabetes drug causes it. Metformin interferes with B12 absorption over years. A 2026 case report describes exactly what that looks like when nobody checks: nerve and spinal cord damage in someone on long-term metformin, which improved substantially over the following weeks once treated.
If you have been on metformin for years and have numb feet, the assumption that your neuropathy is diabetic is a coin flip, not a diagnosis. A B12 level is a cheap blood test and the treatment is injections.
The other reversible or modifiable causes worth ruling out before settling for pain control: alcohol intake, thyroid function, kidney function, and chemotherapy exposure.
Even long-standing nerve damage is not necessarily one-way
The standard belief — patient and clinician alike — is that once the nerves are damaged, the direction is downward and treatment is about slowing it. A 2026 study followed people with type 1 diabetes over five years and found something more useful than that.
Yes, overall prevalence rose over the five years. But individual people moved in both directions: some regressed to less severe disease, and some no longer met the criteria at all.
The authors' conclusion is the sentence to carry: continued intervention remains worthwhile well after the diagnosis has been made. Someone told years ago that they have neuropathy, who concluded that glucose control no longer mattered for their feet, was told something the data does not support.
Screening exists, it is cheap, and it is often not done
There is an agreed way to detect this before it becomes a foot problem, and it requires two instruments that cost almost nothing.
A monofilament — a nylon thread pressed against the sole until it buckles — and a tuning fork. Annually. If you have diabetes and cannot remember the last time somebody touched your feet with either, that is a reasonable thing to ask for by name.
The reason to insist is in the numbers on what happens when nobody checks. In a 2026 study of people over 60 with diabetes, peripheral neuropathy raised the odds of being in the at-risk foot category by roughly ninefold.
Ninefold. Not the pain — the foot. Numbness means injuries do not announce themselves, and the sequence that ends in amputation usually starts with something the person never felt.
What the drugs actually do, and which one first
Be clear about what these are for: they reduce pain. None of them repairs a nerve. A 2026 review states the position plainly — treatment of the underlying nerve damage remains limited, and studies of causal therapy have produced conflicting results, so most treatment is glucose control plus symptom management.
Within symptom management, there is a real head-to-head answer. A trial comparing the first-line drugs found pregabalin better than amitriptyline and duloxetine, and about the same as gabapentin.
The same trial counted where the side effects landed: the duloxetine arm produced the most.
Between the two that came out on top, a 2024 pooled analysis separates them further: pregabalin worked better and faster than gabapentin, and the people taking it used fewer opioids and had fewer adverse events.
Note the contrast with nerve pain from a compressed root in the back, where these same drugs have repeatedly failed to beat placebo. That is not a contradiction. Pain from a mechanically squeezed nerve root and pain from diffusely damaged small nerve fibres are different problems that share a word, and evidence does not transfer between them. A drug can be a reasonable first choice here and close to useless for sciatica.
Ten people with numb feet
Ten people develop numbness in both feet. Most will be told it is diabetic neuropathy. In some of them — nobody can say how many without checking — the treatable driver is a B12 level lowered by years of metformin, and it will not be found unless someone looks. Over five years, the overall picture in a group like this gets worse, while individuals move in both directions and some improve. Of the ten, the ones who get a monofilament and a tuning fork on their feet once a year are the ones whose foot injuries get caught while they are still small, and the risk they are being screened for is roughly ninefold. Everyone offered a drug should understand they are buying pain relief and not nerve repair.
Why the compound argument belongs on a different page
No repair peptide has randomised human trial evidence in peripheral neuropathy. There is one compound with genuinely relevant human trial history in nerve conditions, but that argument belongs on a page whose title names both it and this condition, with its evidence tier stated — writing it here would turn a page about a condition into a case for a product, which is the failure this site is built to avoid. The operator of this site has a commercial interest in compounds of that kind, which is why this page states the absence rather than working around it.
What to ask for at the next appointment
- Check the pattern. Both feet, evenly, creeping upward. One leg in a stripe is a different problem and belongs in a different article.
- Get B12 measured — particularly if you take metformin. It is a blood test, the deficiency is common on long-term metformin, and treating it early is the difference between recovery and permanence.
- Rule out the other correctable drivers. Thyroid, kidney function, alcohol intake.
- Ask for the monofilament and the tuning fork by name, once a year. Neuropathy raises at-risk foot odds roughly ninefold, and the screening costs almost nothing.
- Do not accept that it is too late for glucose control to matter. Over five years, individuals regressed to less severe disease and some no longer met the criteria.
- If you need a drug, know what it is buying. Pain relief, not nerve repair — and in the head-to-head evidence pregabalin came out ahead of amitriptyline and duloxetine, with duloxetine producing the most side effects.
Ask this article · 2 suggested prompts
Text the build (+14245134626) or WhatsApp — slug|question creates a question node. Paste evidence with ingest slug|q:NODE_ID|your paste.