Why an arm is harder than a leg
You arrived at this guide having followed a leg from the amputation through limb shaping, socket, and gait training, and having seen a residual limb learn to drive a prosthetic foot and knee. It is tempting to assume the arm is the same story, moved upward. It is not. A leg, for all the engineering it takes, has a gloriously narrow job: in walking it must mostly bear weight and swing through, doing roughly the same cycle over and over. An arm has almost no fixed job at all. In a single morning a hand turns a key, buttons a shirt, lifts a full mug without crushing it, types, and shakes another hand — each task a different grip, a different force, a different angle, guided continuously by what the fingers feel.
Two facts make this brutally hard to replace. First, dexterity: the human hand has more than twenty independently controlled joints, and the brain devotes an outsized slice of its motor and sensory cortex to running them — far more than to the whole leg. Second, feedback: you can carry a heavy bag without watching your foot, but try buttoning a shirt with a numb hand and you will fumble, because skilled hand use is a tight loop of touch sensing pressure and slip, the muscles correcting in milliseconds. A prosthetic leg can largely get away with no feeling, because the ground reliably pushes back and you can glance down. A hand without feeling is half-blind, and that, more than the motors, is the wall this field keeps running into.
The terminal device: the hook and the hand
The business end of an arm prosthesis is called the terminal device — the piece that actually grips, the prosthetic equivalent of a hand. It comes in two honest families that have coexisted for decades, and newcomers are often surprised that the older, plainer one has not gone away. The first is the split hook: two curved metal or plastic fingers, opened and closed against a spring or a band. It looks nothing like a hand. But it is light, nearly unbreakable, cheap, lets you see exactly what you are grasping, and pinches a precise point with no bulk in the way — which is why a mechanic, a farmer, or anyone doing fine, dirty, or heavy work often prefers it to any lifelike hand.
The second family is the prosthetic hand: five fingers in roughly human shape, sometimes covered in a lifelike glove. Its great gift is not function but presence — it restores a normal silhouette, fills an empty sleeve, lets a person pass unremarked in a meeting or a photograph. The cosmetic cost is real and worth taking seriously; for many, looking whole matters as much as grasping. But a basic mechanical hand usually grips less precisely than a hook and blocks your view of the object. So the deeper truth is that 'hook versus hand' is the wrong contest. Most experienced users do not pick one forever; they own several terminal devices and swap them, choosing a hook for the workshop and a hand for dinner, the way you choose shoes for the task.
Two ways to drive it: cable or muscle signal
However clever the terminal device, something must tell it when to open and close. The two main answers feel a generation apart, yet both are in daily use. The first is body power. A harness of straps loops across the shoulders, and a steel cable runs from it down to the terminal device. When the wearer shrugs or pushes the opposite shoulder forward, the cable pulls and the hook or hand opens; relax, and a spring snaps it shut. It is purely mechanical — no battery, no electronics — and that is its strength: rugged, repairable, light, and, crucially, it gives a faint sense of how hard you are gripping through the tension you feel in the harness. That sliver of feedback is precious, and we will meet why again shortly.
The second answer is the myoelectric prosthesis — the one most people mean by 'bionic.' When you intend to move, even a muscle that no longer connects to a hand still fires, and that firing leaks a tiny voltage to the skin. A myoelectric arm plants electrodes against the residual limb, reads those faint signals from, say, the surviving forearm muscles, and uses them to command little electric motors in the hand. Tense one muscle group and the hand closes; tense the opposing group and it opens — no cables, no harness pulling at your shoulders. The newest hands take this further, with a motor in each finger and a menu of pre-set grips you cycle through, so one hand can pinch a coin, wrap a bottle, or point. They can be genuinely remarkable.
Resist the assumption that newer simply means better. A myoelectric hand is heavier, far costlier, must be charged, dislikes water and dust, and — this is the honest catch — it usually gives the wearer no sense of touch and only clumsy control of force, so people watch the object with their eyes to avoid crushing it. The rugged body-powered hook, by contrast, hands back that thread of grip feedback for free. So the real choice weighs the wearer's goals, work, environment, budget, and what their residual muscles can reliably signal — not a ranking of old against new. Many people, tellingly, end up owning one of each.
BODY-POWERED vs MYOELECTRIC: an honest trade-off
body-powered myoelectric ("bionic")
----------- --------------------
drive cable + harness muscle signals + motors
power source your shoulder rechargeable battery
weight lighter heavier
cost / repair low / easy high / specialist
durability rugged, wet/dirty OK dislikes water & dust
grip feedback some, via cable little to none
appearance visible harness no harness; can look natural
neither is "best" -- many users own both and swap by taskNew frontiers: rewiring nerves and anchoring to bone
The myoelectric arm has an old bottleneck: a forearm or upper-arm stump simply does not carry enough distinct muscle signals to command a hand's many movements. A bold surgery widens that channel. In targeted muscle reinnervation, surgeons take the major arm nerves that used to run all the way to the lost hand — nerves still alive but ending in nothing — and stitch them into spare chest or upper-arm muscles. Over months the nerves grow in and adopt those muscles. Now, when the person merely *thinks* 'close my hand,' the borrowed muscle contracts, and an electrode over it reads a clean, intuitive signal. TMR thus turns a few extra patches of muscle into living relay stations, giving a high-level amputee several independent, natural-feeling commands instead of one or two effortful ones.
TMR carries a striking bonus that circles back to feeling. Once those motor nerves are rerouted, some of the original sensory nerves can reconnect to the chest skin — so when that patch of skin is touched, the person feels their *missing hand* being touched. Surgeons and engineers are learning to exploit this: press a sensor in a fingertip, buzz the matching skin patch, and a flicker of hand sensation returns. It is early, partial, and not yet routine, but it is the most promising road back to the feedback loop the leg never needed and the hand cannot do without. TMR also tames the disordered nerve endings that cause painful neuromas and feed phantom limb pain — by giving a cut nerve a real muscle to grow into rather than a scarred dead end.
The other frontier solves a different problem: the socket. Through the whole leg journey you saw how a socket must hug the residual limb to hang the prosthesis on — and how, on an arm, that socket grips a short stump, can slip mid-task, sweats, and may smother the very skin electrodes a myoelectric arm depends on. Osseointegration removes the socket altogether. A titanium rod is implanted into the marrow of the residual bone; the living bone fuses to it, and a small post passes out through the skin, onto which the prosthesis clicks directly. The arm now hangs from the skeleton, as a real arm does. Wearers report a fuller range of motion, no socket sores, and a vivid sense of the ground or object transmitted up the bone — a kind of touch through structure called *osseoperception*.
Learning to use it: the brain has to be re-taught
A fitted arm is not a finished arm. The hardest part comes after the prosthetist's work: a person must learn to drive a limb that does not feel like part of them, and the brain must change to make it so. This is the cortical reorganization you met in the motor rung, now working for you instead of against you. With a myoelectric arm, the wearer first practices firing the right muscle in isolation — watching a screen turn their faint signal into a moving bar — long before any motor moves. Then come endless graded repetitions: open the hand to this width, hold an egg without cracking it, then a cup, then turn a key, the brain slowly stitching this strange tool into its body map until reaching for a mug stops being a calculation and becomes a habit.
This is also where the occupational therapist becomes central, and where the whole field's honesty is tested. Their job is not to make the prosthesis impressive in a clinic but to make a life work at home: can he cut food, do up a seatbelt, carry a toddler, return to his trade? Often the wisest plan is not the most advanced arm but a sturdy, simple one used well, blended with one-handed techniques and a few everyday adaptations. A high abandonment rate is the field's quiet verdict on devices that dazzled but did not fit a life. Success here looks modest — a man back at his workbench, a woman buttoning her own coat — and that modesty is exactly the point.