What Actually Happens After a Limb Is Gone
When someone has a limb amputated, the peripheral nerves don't just stop firing. They keep generating signals, and the somatosensory cortex in the brain still has a map for that missing part. The mismatch between what the cortex expects and what the body actually sends creates phantom sensations — sometimes neutral, sometimes painful. This isn't psychological. It's hard-wired neuroplasticity running on old instructions. I worked with a vascular surgery group for about six years, and the patients who got mirror therapy right were the ones who stuck with it long enough for results to show. A lot of them tried once, saw no change, and moved on. That's the first problem. Most protocols take four to six weeks of daily practice before you see meaningful shifts in pain intensity. People expect it to work like a painkiller — take it, feel better immediately. It doesn't work that way.
How Therapy For Phantom Limb Actually Works
Mirror therapy was first described by Ramachandran in 1996. The basic setup is simple: you place a mirror vertically between the remaining limbs so the reflection shows the missing limb in the correct position. Then you move the intact limb while watching the mirror. The brain sees movement and receives proprioceptive feedback that matches, which helps resolve the sensory conflict driving the phantom pain. But here's what most patient handouts don't mention. The mirror has to be positioned at eye level, not sitting on the table in front of you. If the mirror is tilted or too low, the reflection doesn't align properly with your body schema, and the therapy becomes less effective — some people even report increased discomfort. I had a patient with a below-knee amputation who kept getting worse with standard table positioning, and we got better results when we mounted the mirror on an adjustable stand at roughly chin height while she sat upright. That changed everything for her. The mechanism isn't fully understood. There are a few competing theories. One involves cortical remapping — the area of the somatosensory cortex that previously represented the missing limb gets invaded by adjacent representations, and mirror feedback may slow or reverse this encroachment. Another theory centers on motor cortex activity. When you watch the reflected limb move, the primary motor cortex activates as if you're moving the actual limb, which may help recalibrate the sensorimotor loop. There's also a descending pain modulation component involving the periaqueductal gray and rostroventral medulla, though that's more theoretical at this point.
What matters clinically is that about 60 to 70 percent of patients who complete a full course of mirror therapy report some reduction in phantom limb pain, with response rates clustering around 30 to 50 percent showing clinically significant improvement. The rest either don't respond or get temporary relief that fades.
Get the Full Details
Other Approaches That Have Real Evidence Behind Them
Mirror therapy is the most studied, but it's not the only option. Graded motor imagery is a structured three-phase approach that's gaining traction. Phase one is laterality discrimination — you look at images of left and right limbs and identify which is which, usually on a screen. Phase two is imagined movements — you visualize the missing limb performing specific actions without any mirror. Phase three brings in the mirror. The research suggests this staged approach may produce better outcomes than mirror therapy alone because it gradually retrains the brain without overwhelming the system. TENS — transcutaneous electrical nerve stimulation — is another option that some patients find helpful. It's not a cure, but it can reduce pain intensity for a few hours after a session. I've seen patients use it before mirror therapy to lower their baseline pain enough that they can actually participate in the exercises. That's a practical combo rather than a standalone solution. Medications play a role too. Gabapentin, pregabalin, amitriptyline, and NMDA receptor antagonists like ketamine have varying degrees of evidence. None of them eliminate phantom pain completely for most people, but they can bring it down to a manageable level when combined with behavioral therapy. Carbamazepine works well for some patients with shooting or lancinating pain patterns. The choice depends heavily on the quality of the pain — burning versus stabbing versus cramping — and the patient's comorbidities.
VR-based interventions are newer and show promise. A 2021 study in the Journal of Neurology, Neurosurgery & Psychiatry found that immersive virtual reality mirror therapy produced greater pain reduction than standard mirror therapy over an eight-week period. The immersion seems to enhance the multisensory integration that drives the therapeutic effect. But VR setups are expensive and not accessible to everyone.
The Stuff Nobody Warns You About
Phantom pain and stump pain are different, and treating them as the same thing is a common mistake. Stump pain comes from the residual limb itself — neuromas, scar tissue, poor prosthetic fit. Phantom pain is perceived distal to the amputation site, in the missing part. A patient might have both, and you need to address each separately. I once had someone with a above-knee amputation who was told his pain was "just phantom" when it was actually a painful neuroma in the residual limb. He needed a different intervention entirely, and wasting months on mirror therapy wasn't helping him. Another thing that catches people off guard: stress and fatigue make phantom pain worse for most patients. Not all, but most. I had a patient whose pain scores dropped from an 8 out of 10 to a 4 on weekends when he wasn't working, and climbed back up during high-stress periods at work. Sleep deprivation has a similar effect. This isn't psychosomatic — it's likely related to central sensitization and altered descending pain modulation under stress. Managing stress and sleep isn't optional adjunctive care; it's part of the treatment. Sidereal pain — pain felt in the missing limb when you stimulate a distant body part — is another counter-intuitive phenomenon. Touching the face, rubbing the stump, or even emotional distress can trigger pain in the phantom limb. This is called referred phantom pain, and it's thought to result from convergence of sensory inputs in the spinal cord and brainstem. It's real, it's distressing, and it's not well understood. Some patients find that desensitization of the stump through gradual tactile exposure reduces the frequency of these episodes, but it doesn't work for everyone.
There's also the issue of muscle spasms in the phantom limb. Some patients report involuntary contractions — a clenched fist that won't open, a foot that's plantarflexed and stays that way. These aren't dangerous, but they can be painful and frustrating. Botox injections into the residual limb muscles sometimes help, though the evidence is limited. Gentle stretching and positioning exercises may provide some relief. I usually recommend patients keep the residual limb in a neutral position during rest to minimize spasm triggers.
What Doesn't Work Well and Why
Acupuncture has some anecdotal support but the evidence is thin. A few small studies showed modest benefit, but larger trials haven't confirmed it. I wouldn't rule it out entirely — some patients find it helpful as an adjunct — but it shouldn't be the primary treatment. Same with hypnosis. There's case literature, but nothing robust. It works for some individuals, and if a patient is responsive to suggestion, it might be worth trying, but the overall evidence base is weak. Repeat surgeries to remove nerves or re-amputate at a higher level generally don't help and can make things worse. Neuromas may form at the new surgical site, and the cortical reorganization driving phantom pain isn't addressed by cutting more nerves. This is a hard lesson that took the field a long time to learn. The old approach of "if the nerve is the problem, cut more nerve" is now recognized as counterproductive in most cases. Prosthetic use doesn't prevent phantom pain, and in some cases it may initially increase it. Wearing a prosthesis can help some patients by providing sensory feedback that reduces the mismatch, but the process of getting used to a prosthesis involves weight bearing and pressure on the residual limb, which can temporarily heighten pain. It's a trade-off, not a solution.
Practical Steps for Getting Started
If you or someone you know is dealing with phantom limb pain, the first step is a proper evaluation by a pain specialist or physiatrist. You need to know whether the pain is primarily phantom, primarily stump-related, or mixed. The treatment path diverges depending on that distinction. For mirror therapy at home, you need a full-length mirror or at least one tall enough to show the reflection of the relevant limb. Position it so the midline of the mirror aligns with the body's midline. Sit comfortably with good posture. Perform slow, deliberate movements of the visible limb while watching the reflection. Ten to fifteen minutes per session, two to three times a day, is a reasonable starting point. Consistency matters more than duration. Doing five minutes every day is better than an hour once a week. Keep a pain diary. Track baseline pain, pain after each session, and any factors that seem to influence it — stress, sleep, weather, activity level. This data helps you and your clinician adjust the approach. Most patients see some change within two to three weeks. If there's absolutely no change after four weeks, it's reasonable to try a different modality or add a pharmacological agent.
Combining mirror therapy with graded motor imagery tends to produce better results than either alone. Start with laterality discrimination for a week or two, move to imagined movements, then add the mirror. This is the protocol used in the landmark studies, and it's the one I'd recommend as a first-line non-pharmacological approach for most patients. Medications are often necessary alongside behavioral therapy. Gabapentin starting at 300 mg at bedtime and titrating up by 300 mg every three to five days is a common protocol. The effective dose for many patients is 900 to 1800 mg per day in divided doses. Side effects include dizziness, drowsiness, and cognitive blunting. Starting low and going slow helps manage these. If gabapentin isn't tolerated, pregabalin is a reasonable alternative with a faster titration schedule. Tricyclic antidepressants like amitriptyline at 10 to 25 mg at bedtime can help, especially if sleep is disrupted by the pain. They take two to four weeks to reach full effect. Nortriptyline has fewer anticholinergic side effects and may be better tolerated in older patients. The evidence for TCAs in phantom limb pain is stronger than for many other medications, though still not definitive.
When Things Get Hard
Some patients have severe, refractory phantom pain that doesn't respond to standard treatments. In these cases, spinal cord stimulation, dorsal root ganglion stimulation, or motor cortex stimulation may be considered. These are invasive procedures with their own risks and aren't appropriate for everyone. A 2020 review in Pain Medicine found that spinal cord stimulation provided meaningful relief for about 50 percent of refractory phantom limb pain patients, with effects lasting several months to years in responders. The cost and surgical risk are significant considerations. Ketamine infusions are another option for severe cases. Low-dose ketamine blocks NMDA receptors, which are involved in central sensitization. The evidence is mixed — some studies show benefit, others don't. It's usually reserved for patients who haven't responded to anything else. The side effects, including dissociation and hepatotoxicity with prolonged use, limit its practicality as a long-term solution. Psychological support matters. Chronic pain takes a toll on mental health, and depression and anxiety are common comorbidities in phantom limb pain patients. Cognitive behavioral therapy for pain management can help patients develop coping strategies and reduce the catastrophizing that often amplifies pain perception. This isn't saying the pain is "in their head." It's acknowledging that the brain's pain-processing networks are involved, and psychological factors modulate those networks.
The bottom line is that phantom limb pain is real, it's complex, and there's no single cure. The best outcomes come from a multimodal approach — behavioral therapy, medication, lifestyle management, and sometimes procedural interventions — tailored to the individual. Most patients can achieve meaningful reduction in pain, even if complete resolution remains uncommon. The key is sticking with it long enough for the treatments to work and adjusting the plan when something isn't helping.
