Why Do Two People With the Same Injury Recover Differently?

The same injury does not produce the same recovery

Two drivers struck from behind at the same intersection at close to the same speed can live two different years afterward. One is fine in three weeks. The other is still not right, and the collision explains neither outcome.

One thing has to be said before any of it. A person who has not recovered did not fail to try hard enough. Every mechanism in the recovery gap is the state of a regulating system, never the quality of a character. The difference between those two drivers is biological, historical and partly measurable, and none of it is a verdict on either of them.

Medicine has known about that gap for as long as it has kept records, and has mostly filed it under individual variation. The phrase names the observation and explains nothing. Naming what actually varies is the whole task.

The gap shows up wherever the event can be held steady. Henrik Kehlet, Troels Jensen and Clifford Woolf, a surgeon, a neurologist and a pain neuroscientist, reviewed what happens after ordinary operations. Hernia repair, breast surgery, chest surgery, amputation, coronary bypass. Between 10 and 50 percent of people are still in pain long afterward. For 2 to 10 percent of them that pain is severe.

The design is what makes those numbers sharp. Same operation. Frequently the same surgeon, the same operating room, the same week. The authors add the detail that makes it sharp. Only a proportion of patients with the same nerve damage during the operation go on to develop chronic pain.

So the event does not carry the outcome. Something else does, and part of that something can be counted before the event ever happens.

Age, genes and willpower do not close the recovery gap

Three explanations arrive before any evidence does, and each contains something true. Age moves the clock of healing without moving the destination. One pain gene covers 96 percent of people and still cannot call an outcome, and willpower names an output while missing the loop behind it.

Age moves the clock, not the destination

Two wound-healing researchers, Ankush Gosain and Luisa DiPietro, reviewed what actually changes about tissue repair as a body gets older. The old belief was that healing in the elderly is defective. The consensus they report is more precise. Repair in older people is delayed rather than defective, and the finished wound is qualitatively similar to a young person's. In tissue, age moves the clock and not the destination.

Age predicts more than tissue repair. Older age is one of four things that predicted a disabled course six months after whiplash. Being 13 or over earns a place in a risk score for slow recovery after a childhood head injury. So age is a real and partial predictor, and it still does not close the gap, because two people the same age diverge.

A gene tilts the recovery, and none runs it

The strongest version of the genetic story is real and worth teaching properly. An enzyme called COMT clears adrenaline-family chemicals out of the nervous system. The pain geneticist Luda Diatchenko tested whether a common difference in that enzyme could explain why two people report different pain from the same stimulus.

Working with Gary Slade, William Maixner and colleagues, she found three common versions of the gene that builds it. They named them low, average and high pain sensitivity, and showed the three together cover 96 percent of people. Carrying even one copy of the low-sensitivity version lowered the risk of a common jaw pain condition by a factor of up to 2.3.

Then the field checked it. The Finnish pharmacologists Anne Tammimäki and Pekka Männistö pooled the whole literature across three chronic pain conditions. The association survived for fibromyalgia and chronic widespread pain only. For migraine and for chronic musculoskeletal pain, low activity of the enzyme carried no measurable risk at all. A gene can tilt a system. It does not run one.

Willpower names the output and misses the loop

Willpower is the answer people reach for most, and the one that does the most harm. The behavioral scientists Johan Vlaeyen and Steven Linton reviewed why some people with a painful back return to their lives while others narrow around the injury. What they describe is a loop, and every step of it can be measured. Fear of pain drives avoidance of movement. Avoidance costs physical capacity and raises watchfulness. The lost capacity makes the next movement more threatening than the last.

Confrontation reduces the fear over time and avoidance amplifies it, which sounds like a moral about courage until you look at what is varying. The loop is driven by how much threat the nervous system assigns to movement, and that assignment is set by the state of the system rather than chosen by the person inside it. Willpower names the output and misses the regulation behind it.

Individual response is a claim that has to be proven

Stephen Senn, a medical statistician, showed that trials sorting patients into responders and non-responders almost never test the sort. An exercise experiment then abolished non-response to training entirely by raising the dose, from 69 percent of trainees to zero.

Senn spent a career on what trials can and cannot show, and he put the hardest available question to the idea of individual response. People can vary randomly from one occasion to the next without dividing into fixed types at all.

The distinction is sharp. If a patient improves on Monday and not on Thursday, the variation lives inside the person rather than between people, and there is no responder type to find. Establishing that true individual response exists requires exposing the same person repeatedly and showing the pattern holds. Most of the recovery literature has never done that.

Grant the first half of that without argument, because the model asks for it. The state a body is in is not a fixed trait, so the same person on two days is not quite the same system, and Monday differing from Thursday is what this model expects. The distinction Senn's objection really creates is about evidence. A stable difference between two people needs a measure recorded before the outcome was known, and that standard is what the strongest recovery evidence meets.

Raise the dose and the non-responders disappear

Two exercise physiologists then produced the result that should worry anyone who claims people differ. David Montero and Carsten Lundby went after the people who fail to get fitter, to find out whether they are a biological type or an artifact of the training program.

They put 78 healthy young men into five groups training one, two, three, four or five hourly sessions a week for six weeks. Non-response fell from 69 percent in the one-session group to zero in the four and five-session groups. Everyone still classed as a non-responder trained a second six weeks with two extra sessions per week, and non-response disappeared entirely.

In that experiment the individual difference was dose in disguise. What looked like a biological type was a person who had never received enough of the input to register it.

Both results stand, and every recovery claim below has to survive them. They do not show that everyone responds identically to everything, since the experiment used healthy young men, one measure of fitness and six-week blocks. They do show that a difference between two people is a claim requiring evidence rather than an assumption worth building on.

The model reads the training result as a statement about dose. The curve has four parts. Below a threshold an input is not registered at all, and a system cannot reveal a response to something it never received. A matched input is integrated and the system reorganizes around it. Past the system's tolerance the same input provokes defense. Higher still it destabilizes or injures. More is not inherently better, and the correct dose is the amount a system can actually use.

That is also how far the training result travels. Raising the dose worked in rested young men with capacity to spare. A depleted system meeting the same increase is a different experiment, and it has not been run.

A scan reports structure, and structure does not decide pain

Disc degeneration appears in 37 percent of pain-free 20-year-olds and 96 percent of pain-free 80-year-olds. The MRI is accurate about anatomy and mostly silent about who hurts, in the spine and in the knee alike.

The vocabulary of a spine report

One piece of anatomy first, because the vocabulary carries the argument. Between each pair of spinal bones sits a pressurized cushion called a disc. A bulge is that cushion spreading outward all round. A protrusion is part of it pushing further out in one place. An extrusion is material that has broken past the outer ring. An annular fissure is a crack in that ring. Modic changes are alterations in the bone next to a disc, sorted by radiologists into numbered types.

Two more words, because they differ by three letters and mean different things. Spondylolysis is a stress crack in a small bridge of bone at the back of a vertebra. Spondylolisthesis is one vertebra slipping forward on the one below it.

Degeneration climbs with age in people who feel fine

Radiologists and back-pain researchers went and looked at what those features do in people who feel nothing wrong. They pooled 33 studies covering 3,110 people with no back pain and sorted the scans by decade of life. Disc degeneration appeared in 37 percent of pain-free 20-year-olds and in 96 percent of pain-free 80-year-olds.

Nearly every pain-free 80-year-old carried it.

Disc bulges rose from 30 percent to 84 percent across the same span. Protrusions rose from 29 percent to 43 percent, and annular fissures from 19 percent to 29 percent. Their own conclusion is that many of these features are likely part of ordinary aging and unassociated with pain.

The study that startled the field first is even more teachable, because of how it was run. A radiology group led by Maureen Jensen, with Michael Modic and Jeffrey Ross, asked whether a scan could pick out the person in pain when the reader did not already know the answer. They scanned 98 people with no back pain.

Those images were shuffled in with 27 from people who did have pain, so the two neuroradiologists reading them were blind to who was who. Only 36 percent of the pain-free group had entirely normal discs at every level. Fifty-two percent had a bulge somewhere. Twenty-seven percent had a protrusion, and 38 percent had an abnormality at more than one disc.

The knee tells the same story as the spine

The knee evidence comes from outside a clinic. Framingham is a town in Massachusetts whose residents have been examined by researchers since 1948, in one of the longest-running health cohorts anywhere. Radiologists and arthritis epidemiologists there, among them Ali Guermazi, Tuhina Neogi and David Felson, asked how much damage sits inside knees that an x-ray calls normal. They scanned 710 residents over the age of 50 with no osteoarthritis on x-ray.

Eighty-nine percent had at least one abnormality on MRI. Cartilage damage was present in 69 percent and bone marrow lesions in 52 percent. A bone marrow lesion is a bruise-like change inside the bone beneath the cartilage, visible on MRI and invisible on x-ray. Then the comparison that decides it. At least one abnormality appeared in 86 to 88 percent of painless knees and in 90 to 97 percent of painful ones.

Six findings move the odds and none decides a case

The comparison runs the other way as well. The same spine imaging group went back and asked the follow-up question, comparing 1,193 adults without symptoms against 1,904 with them, all aged 50 or under. Six findings were genuinely more common in the people with pain. Disc bulge, spondylolysis, disc extrusion, Modic type 1 changes, disc protrusion, and disc degeneration itself.

The odds ratios ran from 7.54 for a disc bulge down to 2.24 for degeneration. An odds ratio near one would mean a finding is equally common on both sides, so all six shift the odds and none of them settles a case. Five other findings showed no association at all, among them annular fissures, spondylolisthesis and any Modic change.

Findings on a scan are not meaningless. Six of them move the odds. What none of them does is decide the recovery, because a scan reports structure with great accuracy and cannot report what this particular nervous system is doing with that structure.

The model expects exactly this. A symptom is an output of regulation, produced by a system weighing tissue condition, threat, prior experience and expectation together. It is not a readout of damage. The same structural finding passing through a different tone is a different biological event.

When a slow recovery needs a doctor now

A scan predicts pain poorly and injury severity predicts recovery poorly, and both facts become dangerous the moment they are read as a reason to skip a workup. Some causes of a bad recovery are structural, infectious, genetic, toxic or malignant, and each needs treating as what it is.

Nothing in this model erases those causes. What the model adds is an account of how a body receives such a cause, responds to it, compensates for it, and either contains or amplifies what follows. The model's claim is bounded: every disease has a regulatory expression, and many are initiated, maintained or amplified by failures of regulation. It does not reduce all illness to poor regulation. The model calls that regulation tone.

Signs that mean an emergency department now

After a head injury, some signs mean an emergency department now rather than later. A headache that keeps getting worse. Repeated vomiting. A seizure. Weakness or numbness down one side. Slurred speech. One pupil larger than the other. Unusual drowsiness, or a person who cannot be woken properly. Any of those needs urgent assessment.

With back pain, a separate list applies. Numbness in the area a saddle would touch. Loss of bladder or bowel control. Progressive weakness in a leg. Fever alongside back pain. Unexplained weight loss, or a history of cancer. Each is a reason to be seen promptly rather than to wait and watch.

Chest pain or breathlessness is a separate emergency. So is a calf that becomes swollen and painful after surgery or a long period of immobility. So is a wound turning red, hot and steadily more painful. Each carries its own urgency and its own treatment. Recognizing them is not pessimism. It is what keeps everything else here safe.

Post-traumatic stress and depression after injury belong to this subject too. If you are thinking about harming yourself, contact emergency services or the 988 Suicide and Crisis Lifeline.

Reserve is what a system can spend before a deficit shows

Ten nursing home residents carried plaque counts 80 percent as high as visibly demented residents and had stayed in the top fifth of their home for cognition. Robert Katzman named what those ten brains had been spending: reserve.

The imaging evidence leaves a gap. If the finding does not carry the outcome, something standing between them does, and neurology found it by accident.

The neurologist Robert Katzman, working with Robert Terry and colleagues, compared the damage found in a brain after death against the thinking of the person while alive. They examined the brains of 137 nursing home residents whose cognition and daily function had been measured while they lived. Ten of them stood out. Those ten had been in the top fifth of the whole home for cognition and function, performing as well as residents whose brains turned out clean at autopsy.

Their brains were not clean. They carried plaque counts 80 percent as high as the residents who had been visibly demented. A plaque is a sticky deposit of protein that gathers between brain cells in Alzheimer disease, and pathologists count them under a microscope after death. The authors proposed that these people had more of something to spend before a deficit could show, and the word they used was reserve. That paper is where the reserve idea in dementia research is traced to.

The nun study read reserve six decades early

The idea needed a test that ran forward rather than backward, and the epidemiologist David Snowdon found one in a filing cabinet. A convent had kept the autobiographies its young sisters wrote on entering, at an average age of 22. Snowdon realized he could measure something about a brain six decades before its outcome was known.

He scored each essay for idea density, meaning how many distinct ideas were packed into every ten words. Fifty-eight years later, low idea density in youth predicted poor cognitive scores in old age. Among the 14 sisters who had died and been examined, every woman with low early idea density had confirmed Alzheimer disease. None of the women with high idea density did.

Fourteen autopsies. That cohort is also unusually uniform in diet, housing and occupation, which sharpens the comparison and narrows what it generalizes to. Idea density may index education or early brain development rather than reserve as such.

What survives all of that is the shape of the finding. A measurement taken long before the event carried information about how the event would land. That shape, the advance reading that forecasts the landing, is the signature that recurs across every recovery literature, from surgery to combat deployment.

Hold the input constant and recovery still separates

A measured dose of cold virus produced colds in 27 to 47 percent of volunteers depending on their stress score, and an identical wound healed 9 days slower in caregivers. The input was fixed by design. The person receiving it was the only variable left.

Observational studies cannot hold the input steady. This one small line of work can, which is why it matters more than its size suggests.

The method came from a British unit where volunteers agreed to have a measured dose of a cold virus dripped into the nose, then be quarantined and watched. The psychologist Sheldon Cohen used it first on psychological stress.

In 394 volunteers the same dose of the same virus infected 74 to 90 percent by stress score, and produced actual colds in 27 to 47 percent. Smoking, drinking, exercise, diet and sleep quality did not account for it. The stress and physical symptoms page tells that experiment in full.

The same virus, sorted by social ties

What belongs here is what the same design produced when the question changed. Cohen ran it again asking about company rather than pressure. Two hundred and seventy-six volunteers reported how many of twelve kinds of social tie they had, and were then given rhinovirus, the commonest of the cold viruses. People with the fewest kinds of tie were 4.2 times more likely to develop a cold than people with six or more.

The strength of that result is that it did not rest on what anyone said afterward. The volunteers were weighed for mucus, tested for how efficiently their nasal passages cleared, and sampled for how much virus they shed. The sparsely connected group was worse on all three. One questionnaire and three objective measurements moved together.

Sleep measured by machine before the virus arrived

A third round measured the reserve objectively instead of asking about it. Aric Prather and Martica Hall study sleep, and they put the question to a machine rather than a questionnaire. They had 164 volunteers wear wrist monitors for a week before the virus arrived. Those sleeping under six hours a night were more than four times as likely to develop a cold as those sleeping over seven. How fragmented the sleep was did not matter. How much of it there was did.

An identical wound, healing at two speeds

The closest analogue to an injury came from a different laboratory. A group at Ohio State studying how psychological state changes immune function wanted a wound they could standardize. They gave 13 women caring for a relative with dementia, and 13 matched women who were not caregivers, an identical punch biopsy 3.5 millimeters across.

That is the small circular sample of skin a dermatologist takes with a cylindrical blade. The wounds were then photographed and chemically tested until they closed. The caregivers took 48.7 days on average. The controls took 39.3. The wound was the same. The healing was not.

Nobody randomizes a person into caregiving, and nobody can. Thirteen women against thirteen matched women is what is available, and the matching does the work randomization cannot. Volunteers willing to be deliberately infected are not a random sample either, and the stress scores were self-reported. What this group of studies establishes is narrow, and it is exactly the point. Hold the input constant and the outcomes still separate.

Three fields found the same fact, and the model unifies it

Pharmacogenomics, stress physiology and pain science each discovered that the receiving system decides what an input becomes. None of those observations is new to medicine, and the model does not claim to have discovered them. The unification under one variable, tone, is the claim.

The same drug meets different machinery

The pharmacologists Mary Relling and William Evans summarized where the field stands. Inherited variation in roughly 20 genes now has clinical consequences for about 80 medications, and prescribing guidance has been rewritten around it.

One case teaches it faster than any statistic. A man in Geneva was given small, ordinary doses of codeine for a cough during pneumonia and developed life-threatening opioid poisoning. Codeine does nothing by itself. A single enzyme converts it into morphine inside the body.

Doctors at the Geneva teaching hospital, among them the intensivist Yvan Gasche and the clinical pharmacologist Pierre Dayer, read his genes to find out what had happened. Genotyping means reading which versions of a gene a person carries. He carried three or more working copies of the gene for that enzyme, so he converted codeine to morphine far faster than the dose assumed.

Two other things had to line up with it. Other medications he was taking blocked the alternative route that would have disposed of some of the codeine harmlessly. His kidneys were briefly not clearing what the morphine became. Three ordinary conditions met in one body and a normal dose turned into a poisoning. The dose was normal. The body was not.

Accumulated demand changes what the next demand costs

A body does not hold its internal values constant. It moves them in advance to meet what is coming. The neuroscientist Peter Sterling named that anticipatory regulation allostasis, and set out in a 2012 review what the word is for. In 1993 the neuroendocrinologist Bruce McEwen, writing with a physiological psychologist, took allostasis and ran it across time. They called the running cost of holding a state allostatic load.

Identical wounds, different pain

Henry Beecher was an anesthesiologist who had treated wounded soldiers and could not explain what he had seen. Men with severe battlefield wounds often asked for no pain relief, while civilians with surgical wounds of comparable size asked for it urgently. He argued that pain is not read off the size of a wound, and that what the wound means to the person carrying it belongs inside the mechanism rather than beside it.

The pain researcher Roger Fillingim brings that up to date. Demographic factors, genetic variants and psychosocial processes each contribute, and none acts alone. Genetic associations with pain shift across sex and ethnic group. Genes interact with stress, and with catastrophizing, meaning the habit of reading the worst into a sensation. What results is a combination unique to each person, and no single factor sets it.

The Unified Model of Tone holds that these three are one fact governed by one variable. The unification is the claim.

Tone, defined

The everyday sense of the word is a muscle held tight, and that sense is true and incomplete. Defined as tension, tone stays a property of whichever tissue is under examination, and the tone of a muscle, a blood vessel and a nerve stay three unrelated things. Defined as organization, all three become one variable read in three places.

Tone, in this model, is the integrated organization of the body's interacting state transitions. It is the mechanical, electrical, chemical, fluid and neural processes of a living body taken as one bound state rather than a list of parts. Oscillation is the carrier. Organization is the thing being carried. Tone is the chord rather than the notes.

Stated as one law it reads like this. An input does not create an outcome. An input interacting with a tone creates an outcome. The same input passing through a different tone is a different biological event. The input does not carry the outcome, and the system that meets it does. In recovery that law comes apart into three foundations of tone: load, coupling and time course.

Load is the bill a body was already paying

Teresa Seeman's ten-marker index of accumulated physiological cost, recorded in older adults before anything went wrong, predicted steeper decline in memory and physical function and more new cardiovascular disease. The score preceded the recovery it forecast.

Every state a body holds is paid for. The chemicals released to meet a challenge, cortisol and adrenaline among them, protect in the short run and damage if they stay switched on. That running bill is allostatic load, and it has a page of its own that builds the four ways the stress response goes wrong and the ten-marker index that scores them.

Recovery needs one property out of all of it. The epidemiologist Teresa Seeman built that index inside a long-running community study of older adults, and the score predicted what came next. Higher scores meant steeper decline in memory and physical function over the following years, and more new cardiovascular disease, independently of the usual risk factors. The score was recorded before the decline and it carried information about who would decline. That is the advance-reading shape the recovery gap keeps producing.

Then the instrument turned out to be several instruments. Fifteen years on, the nurse scientist Theresa Beckie read the 58 studies that had used the index between 1997 and 2012. The associations held across socioeconomic position, social relationships, work conditions, stress exposure, physical and mental health, and death from any cause.

Her complaint sits upstream of all of it. Groups measure allostatic load with different biomarkers, different numbers of them and different cut-off points, which leaves the studies hard to compare at all. Her first priority for future research is prospective longitudinal studies, which is an admission about what the existing base mostly is not.

A quantity every group measures differently is not yet a measurement.

The model predicts something more specific than a summed index. Readings of tone are dynamic. Variability, coupling between rhythms, and how long recovery takes after a demand. The prediction is that those will outperform any list of static values, and that they will move first.

History changes the system and sentences nobody

Adults reporting four or more categories of childhood adversity carried four to twelve times the risk of depression and attempted suicide across 9,508 people. For forecasting one person's recovery, the same count performs barely better than a coin toss.

Load accumulates across a life. The largest body of work on that is also the most frequently misused, and the misuse matters as much as the finding. The trauma and load pages carry the finding itself in full.

A survey of 9,508 adults, published by Vincent Felitti, Robert Anda and colleagues, asked about seven categories of childhood adversity and produced a staircase. Set against people reporting none, those reporting four or more carried four to twelve times the risk of alcoholism, drug use, depression and attempted suicide. Heart disease, cancer, lung disease and liver disease climbed the same staircase.

Pooling 37 studies and 253,719 people, Karen Hughes, Mark Bellis and colleagues found the staircase holds outside the United States, and that its steps are very uneven. Odds under two for physical inactivity, obesity and diabetes. Two to three for smoking, cancer, heart and respiratory disease. Above seven for problematic drug use and violence.

Anda, one of the original investigators, later wrote with Laura Porter and David Brown to say the score was being misused. The count was designed to describe risk across a population and not to grade an individual. Treating the number as a personal diagnosis or a screening result misreads what the measure can do.

Population predictor, individual coin toss

Two birth cohorts then settled it with data. Jessie Baldwin, Avshalom Caspi, Terrie Moffitt, Andrea Danese and colleagues asked whether the score could pick out a person heading for trouble. They followed a British cohort of 2,009 eighteen-year-olds and a second cohort of 918 people to the age of 45.

Each additional adversity raised risk slightly and reliably, exactly as the population studies said. The score's ability to say who would actually develop a problem was barely better than a coin toss. The area under the curve, a number that runs from 0.50 for pure chance to 1.00 for a perfect call, was 0.58 for mental health and 0.60 for physical health.

Hold both results at once, because together they are the answer to determinism. Across a population, accumulated history is among the strongest predictors medicine has. For one person sitting in a room, it forecasts very little. History changes the tone that meets the next input. It does not sentence anyone, and a high count describes what a system has carried rather than what it will become.

One sleepless night turns up the pain channel

After a single night without sleep, the same heat, cold, pressure and pinprick all hurt more, while the ability to detect ordinary touch stayed intact. Reserve is not only decades deep. Some of it is one night old, which makes it the easiest part of load to demonstrate.

The experiments before this one had disagreed with each other. Different deprivation protocols, different volunteers and different pain measures had produced results pointing in opposite directions, and nobody had tested the entire sensory range inside one protocol. Pain physiologists led by Sigrid Schuh-Hofer, with Walter Magerl and Rolf-Detlef Treede, set out to settle it. They tested every sensory channel in 14 healthy young people twice. Once after a normal night, once after a night with no sleep at all.

After the sleepless night the same heat, the same cold, the same blunt pressure and the same pinprick all hurt more. What did not change is the revealing part. Their ability to detect non-painful touch and temperature was unaltered. Losing a night of sleep did not blunt the senses. It turned up the pain channel specifically.

Fourteen people and one night of total deprivation is a small experiment, and the same finding runs through the fibromyalgia and sleep pages inside longer arguments. What it gives the recovery question is not size. It is a mechanism with a direction, plus a control showing the shift was specific to pain rather than a general dulling of sensation. Reserve is one measurable part of tone, and it changes what the next input will cost.

Sleep before the crash predicted what came after

The same variable has been recorded before a real event. Thomas Neylan, Ronald Kessler, Samuel McLean and colleagues enrolled 666 people in the hours after a collision and asked how they had been sleeping in the month before it. Eight weeks later, 42 percent met criteria for post-traumatic stress disorder and 27 percent for a major depressive episode. Nightmares and a tendency for stress to disturb sleep, both present before the crash, predicted who ended up in either group. Insomnia predicted the post-traumatic stress alone.

Pre-crash sleep was recalled in an emergency department after the crash, which invites recall bias. Even so, the direction is the one the model expects. The state the system was in before the collision carried information about what the collision would become.

Coupling decides how large the same event becomes

Nobody is sensitive in the abstract. A person is strongly wired to some kinds of input and barely wired to others, and that wiring, which coupling names, is why one collision becomes four different events in four survivors.

Picture it as how many of a system's other parts are wired to a given input, and how strongly. A small input entering a heavily connected node produces a large response. A large input arriving where the receptors, pathways and thresholds are unavailable produces almost none. The size of a response does not reveal the size of the input. It reveals the input's relationship to the tone receiving it.

The Geneva codeine case is coupling read at the smallest scale. The molecule was fixed. What varied was how much enzyme stood ready to receive it, so the same milligrams reached a body that converted them at several times the expected rate. The individual variation Fillingim describes is coupling read at the scale of a whole person, where sex, genotype, prior stress and expectation each change what a stimulus is arriving into.

This is why one person after a crash is undone by noise, another by movement, another by the sound of braking tires, and another walks away untouched. One event met four differently wired systems and became four different events.

Coupling also accounts for a recovery pattern that otherwise looks like unreliability. A person tolerates a food, a workout, a medication or a difficult conversation on one day and not on another. The input barely changed. The state that met it did. That is the case Senn's objection turned on, and the model claims it rather than conceding it, because tone is a state and not a trait.

The two inputs people are least likely to count as inputs at all are expectation and other people. Both have been measured, and both carry limits that belong beside the measurements.

What a sentence spoken to a patient is worth, in millimeters

Holding an opioid infusion perfectly constant while changing only the sentence spoken doubled the pain relief or abolished it. Expectation is a real regulatory input in recovery, and its measured average size is about 6.5 millimeters on a 100 millimeter pain scale.

Expectation reaches a person only as far as that person is wired to meaning and context, which is why the same sentence spoken to two people is two different inputs. That makes it a question about coupling. It is also the cleanest demonstration that a non-physical input is a real regulatory one, and the experiments that show it are unusually well built.

Hide the delivery and the drug works less well

Neuroscientists at the University of Turin, among them Luana Colloca, Leonardo Lopiano, Michele Lanotte and Fabrizio Benedetti, spent years measuring expectation without deceiving anyone, then gathered their own work into a single review. Instead of a dummy pill they gave the real drug through a preprogrammed pump.

Sometimes the patient watched a clinician deliver it. Sometimes the patient knew nothing had been given. Across pain, anxiety and Parkinson disease, the identical drug at the identical dose worked less well when it arrived unannounced. The gap between the two is the contribution of knowing.

Vary only the sentence and the brain follows it

Brain-imaging researchers at Oxford, including Ulrike Bingel, Katja Wiech, Markus Ploner and Irene Tracey, asked whether a sentence spoken to a patient could change what a fixed dose of drug does inside the brain. They gave healthy volunteers a steady concentration of a powerful opioid while a steady heat pain was applied, and scanned them throughout.

Told to expect nothing, they got the drug's baseline effect. Told to expect relief, the same infusion produced roughly double the pain reduction. Told to expect the pain to worsen, the same infusion produced no relief at all. Activity in pain-coding brain regions changed to match the reports.

Remove the deception and the effect survives

The field had assumed the effect required a patient to be fooled. Ted Kaptchuk has spent a career studying placebo, and with Irving Kirsch, Anthony Lembo and colleagues he tested whether an open-label one still works. They randomized 80 people with irritable bowel syndrome to one of two arms.

One was no treatment. The other was a bottle openly labeled as placebo pills made of an inert substance, with the rationale explained. Both groups had the same time with the same providers. The group knowingly taking sugar pills improved significantly more.

Ask before the operation and the answer carries forward

Clinical psychologists, with Charlotte Auer and Winfried Rief in Marburg and Elizabeth Broadbent in Auckland, pooled 21 prospective studies covering 2,611 surgical patients. They wanted to know whether what a person expected before an incision showed up in their life afterward. It did, at a correlation of 0.369 overall.

A correlation of 1.0 would mean expectation told you the outcome exactly, and 0 would mean it told you nothing, so 0.369 is a real signal a long way short of a forecast. It ran to 0.208 for mental quality of life and 0.126 for physical.

The counterweight: placebo against no treatment at all

Asbjørn Hróbjartsson and Peter Gøtzsche worked at the Nordic Cochrane Centre, where testing whether a celebrated result survives scrutiny is the job. They noticed that the power of placebo had almost never been tested against the right comparison, which is not treatment but nothing at all. They found 130 trials carrying a genuine no-treatment arm.

Placebo did nothing measurable for objective outcomes or for yes-or-no outcomes. For continuous subjective outcomes it produced a small effect that shrank as trials grew larger, which is what a literature shaped by its small trials looks like. For pain it produced a real but modest effect, about 6.5 millimeters on a 100 millimeter scale.

The open-label literature is thinner than its reputation as well. The clinical psychologists Melina von Wernsdorff and Stefan Schmidt, working in Freiburg, pooled what the field had produced a decade on and found a moderate effect, a standardized mean difference of 0.72. That means the average person given an honest placebo did about three quarters of one spread of scores better than the average person given nothing.

It rests on 13 studies, of which 11 could be pooled, all with subjective outcomes and moderate risk of bias throughout. Seventy-six percent of the variation between those studies came from genuine differences between them rather than chance, which is a great deal of disagreement. The authors call the field an infant.

The correct reading is the bounded one. Expectation is a real input with a real and limited size, larger than zero and much smaller than the folklore. Meaning is a high-order input that changes lower-order regulation through the body's own recursive wiring, and like every input its effect depends on the tone it meets.

That is also why its measured average is modest. An effect that depends on the receiving system will look small whenever it is averaged across systems nobody sorted first. The model states that as a prediction rather than a complaint.

Social ties are a measured input, with a measured limit

Across 148 studies following 308,849 people, stronger social relationships went with a 50 percent greater likelihood of being alive at follow-up. Delivered as a uniform intervention after heart attack, the same kind of input moved survival by a tenth of a point.

Other people are an input like any other, and how far one reaches a given person is what coupling names. The strongest association in the recovery literature is also the one most likely to be dismissed as soft, so it is worth taking at its measured size.

Julianne Holt-Lunstad and Timothy Smith are psychologists at Brigham Young University who set out to test whether relationships belong on the same risk list as smoking and blood pressure. Pooling 148 studies that followed 308,849 people, they found stronger social relationships associated with a 50 percent greater likelihood of still being alive at follow-up.

The effect held across age, sex, starting health and cause of death. It was largest for complex measures of social integration and smallest for the crude question of whether someone lives alone.

The rhinovirus quarantine study is the experimental version of the same thing. A measured dose of rhinovirus, and a 4.2-fold difference in who caught a cold, sorted by how many kinds of tie a person had.

A social network changed what pathology cost

The effect also shows up against pathology, and this is the cleanest coupling result available. David Bennett, Julie Schneider and Robert Wilson work at a memory and aging center in Chicago where older people agree to be tested every year and examined after death. They followed 89 people without dementia, measured the size of their social networks by interview, tested their thinking annually, and examined their brains.

More Alzheimer pathology meant worse thinking, as expected. The size of the network changed the slope of that relationship. At the same burden of pathology, people with larger networks were thinking better. The pathology arrived either way. What it cost differed, and that is coupling changing what an insult becomes rather than changing whether it arrived.

The trial that treated the correlate and changed nothing

If low social support and depression predict death after a heart attack, treating them should prevent some of those deaths. The social epidemiologist Lisa Berkman, with Robert Carney, Neil Schneiderman and the ENRICHD investigators, tested exactly that in 2,481 patients across eight centers.

Patients were randomized to usual care or to months of cognitive behavior therapy, with an antidepressant where indicated. The therapy worked on its targets. Depression scores improved and social support scores improved. Over an average of 29 months, survival free of another heart attack was 75.8 percent with the intervention and 75.9 percent without it.

That is as flat a result as trials produce. Usual-care patients also improved substantially on their own, which shrank the contrast the trial could detect.

Read through the model, the trial moved a correlate rather than the loop that sets it, and it moved it late. It also gave one predetermined input to everyone in a sample nobody had sorted by the tone that would decide who could use it. The model predicts that shape of result. Deliver the same thing to an unstratified sample, and the people it fits and the people it does not are collapsed into a mean describing neither.

What makes an input effective, in this model, is correspondence between the input and the organization it meets. How hard it is delivered is a separate axis, and magnitude is not the active ingredient. That turns the complaint into a prediction, and the convergence test below is built to isolate it.

Cold hyperalgesia, not crash severity, predicted chronic whiplash

A cold pain threshold at or above 13 degrees Celsius, measured within a month of a whiplash injury, raised the odds of the chronic severe recovery path 26.3 times. The strongest predictor was a reading of the person, not of the collision.

One event arrives at two people. One system integrates it and comes back. The other keeps it, and reorganizes around what it kept. That fork is the third signature foundation, and the time course page owns the stages either side of it.

Whiplash is the best-measured example, because a research group suspected the crash was the wrong thing to be measuring and went and measured everything else instead.

Michele Sterling, Gwendolen Jull and Justin Kenardy worked in a whiplash research unit inside a physiotherapy department in Queensland. They assessed 76 people within weeks of a whiplash injury. Neck movement, muscle activity, sensitivity to pressure and to cold, blood vessel responses, psychological distress. Then they followed them for six months.

What predicted the people who ended up moderately or severely disabled was not the crash. It was cold hyperalgesia, meaning cold registering as painful, together with early post-traumatic stress, higher initial disability and older age.

Six years later the same group followed 155 people from within a month of injury out to three, six and twelve months. Then they let the statistics sort those people into recovery paths rather than assuming one average path. Distinct paths appeared.

The strongest baseline predictor of the chronic severe path was a cold pain threshold at or above 13 degrees Celsius. That sounds mild and is not. It means cold began to hurt while it was still relatively warm, so the higher the number, the more sensitive the person. Those people were 26.3 times more likely to land on the chronic severe path.

The range the data can support around that number, called the confidence interval, runs from 4.98 to 139.09. Certain in direction and close to useless in size. That specific estimate comes from one cohort and nobody has reproduced it.

The finding underneath the estimate has held up better. A review team in Cardiff searched the literature for prospective studies of cold hyperalgesia after whiplash and found six of them, run on four separate cohorts, one Danish and three Australian, covering 443 people. All four cohorts supported cold hyperalgesia as a predictor of long-term pain and disability. The reviewers graded that evidence moderate, called for the strength of the relationship to be pinned down, and stated plainly that the mechanism is unknown.

Concussion recovery reads the child, not the collision

Concussion tells the same story from the recovery side, and the concussion page builds it in full. What belongs here is one cohort. Roger Zemek, Martin Osmond, Keith Owen Yeates and colleagues enrolled 3,063 children across nine pediatric emergency departments in Canada within 48 hours of a head injury.

Thirty-one percent still had symptoms at four weeks. Their risk score is built from female sex, age 13 or over, a history of migraine, a previous concussion lasting over a week, and the child's own current headache, noise sensitivity, fatigue and slow answering.

The score's discrimination is modest, running at 0.71 in the group used to build it and 0.68 in the separate group used to check it, on the same scale where 0.50 is a coin toss. Its authors state plainly that it is not ready for routine clinical use. What the score describes, item by item, is a child rather than a collision.

In the model's terms this is the fork that time course turns on. A system either integrates the event and comes back with its range intact, or it retains a protective pattern that the rest of the body then reorganizes around. Which way it goes is not settled by the collision alone.

Impact scales with what an event demands, what it means and how new it is, and inversely with the capacity available to integrate it at the moment it lands. That capacity is a property of the tone the event arrives at. It is also why the predictors above mix the two terms, with a cold pain threshold reading the receiving system and an initial disability score reading the encounter.

Context around an injury changes its course, within limits

When Saskatchewan removed insurance payment for pain and suffering in 1995, median time from whiplash injury to claim closure fell from 433 days to 194. The two strongest demonstrations are both contested, and the contests are themselves informative.

In 1995 the Canadian province of Saskatchewan switched its traffic injury insurance from a system that paid for pain and suffering to one that did not. That handed researchers a natural experiment on 7,462 whiplash claimants, and the chance to ask what happens to recovery when the money around an injury changes. J. David Cassidy, Linda Carroll, Pierre Côté and colleagues found that claims fell, and the median time from injury to claim closure fell from 433 days to 194.

The physics of rear-end collisions did not change on the first of January. Something around the injury did.

Time to claim closure is an administrative endpoint rather than a clinical one, and the counter-reading carries equal weight. The same law that removed the payment also removed a reason to keep a file open, so the number may be measuring paperwork rather than necks. The result has been read both ways ever since, and it remains unsettled.

Imagined concussions match real symptom lists

The second finding is sharper. Psychologists asked whether people could describe concussion symptoms they had never experienced. They took volunteers with no personal experience or knowledge of head injury, had them imagine being concussed in a car accident, and asked which symptoms they would expect to have six months later. The imagined list matched the real post-concussion symptom pattern almost exactly.

That is a genuine problem for interpreting the whole post-concussion literature, and it should not be tidied away. It shows expectation can generate the pattern. It does not show expectation generated it in any particular patient, and the design cannot separate those two claims.

Neither result is embarrassing and neither is decisive. Meaning is a high-order input that reaches the body by changing tone. It changes threat interpretation, automatic nerve control, guarding, attention and sleep, and each of those changes what a tissue signal becomes on its way to being an experience.

A model of this kind has to predict that context alters outcome. It also has to insist that context altering an insurance file is a different claim from context altering a nervous system, and only one of those was measured here.

A pattern can outlive the injury that caused it

Systems with the right kind of positive feedback can hold two stable states, and the push required to leave a state is not the push that produced it. That mathematics is how a healed tissue and a continuing symptom coexist without contradiction.

The hardest version of the recovery question is why an injury goes on hurting after the tissue has healed.

Bistability licenses the concept

Start with the mathematics, because nothing else licenses the concept. David Angeli, James Ferrell and Eduardo Sontag work between control theory and cell biology, and they wanted a test that could be applied to a biological network of any size. What they showed is how to determine whether such a network can sit in two stable states rather than one. Systems built with positive feedback of the right shape are guaranteed to be bistable across some range of feedback strength.

Bistable systems show hysteresis. In plain terms, the push it takes to get out of a state is not the same size as the push that put you into it. The way back is not the way you came.

Be exact about what that result is. It is a mathematical proof demonstrated on cell-cycle and signaling cascades. It establishes that hysteresis is available to biology. It does not by itself demonstrate hysteresis in any human illness, and the time course page carries what human evidence there is.

Kindling: each episode needs less trigger than the last

The clinical version arrived from psychiatry. Robert Post, a psychiatrist studying recurrent mood disorder, noticed a pattern across decades of case records. A first episode usually follows a clear life stressor. Later episodes need less and less provocation, until eventually they arrive on their own.

He called it kindling, after the way a fire needs less each time to catch once it has burned before. His proposal was that stressors and episodes both leave lasting traces in gene expression, altering transmitters and receptors so the system becomes easier to tip.

Then it was tested forwards, which is where it acquired a boundary. Kenneth Kendler, Laura Thornton and Charles Gardner followed 2,395 women from a twin registry through four interviews over nine years, covering 1,380 onsets of major depression.

Through roughly nine episodes, the power of a stressful life event to trigger the next one declined steadily with each episode, exactly as kindling predicts. After nine it stopped declining. The pattern held after accounting for how severe the events were and for genetic risk. Sensitization is real and it appears to have a ceiling.

A scan at week six predicted pain at month twelve

Pain has produced the cleanest prospective version. A group at Northwestern, among them Marwan Baliki and Thomas Schnitzer, scanned people whose back pain was only a few weeks old, then followed them for a year. The scan came before anyone knew who would recover. In those whose pain persisted, gray matter density fell across that year.

Gray matter is the layer where the nerve cell bodies sit. More striking is what was already there at the start. Stronger communication between a reward-related structure and the frontal cortex, present at the very first scan, predicted who would still be in pain twelve months later.

Small cohort, single center, and prediction from functional connectivity is statistically fragile. Functional connectivity means how closely two regions rise and fall together over time. The shape still holds. A measure of state taken early carried information about a fork that had not happened yet.

Watching a recovery approach the fork

There is also a proposal for watching a system approach that fork from outside. Marten Scheffer spent a career on why lakes flip suddenly from clear to green, and brought the same mathematics to bodies. A system near a tipping point recovers more slowly from small perturbations, so the pattern of tiny recoveries in a continuous signal reads out how much reserve remains. That is a perspective article rather than a result, and it should be labeled as one.

The proposal has one striking demonstration, and it came from an unlikely pairing. Ingrid van de Leemput models tipping points in lakes. Marieke Wichers studies the day-to-day course of mood disorder. Together they had healthy people and patients record their emotions repeatedly over time, then examined the shape of the fluctuations rather than the average mood.

Before a shift into or out of a depressive episode, the fluctuations became slower to return to baseline, more variable, and more tightly correlated with one another. Those three together are the recognized signature of a system approaching a tipping point.

Those are self-report time series rather than physiology, and the same statistical patterns can be produced by other processes. Prospective clinical use has not been established. What the work demonstrates is that the approach to a fork can become visible in the dynamics before the fork arrives.

The stable state a body settles into, in this model, is a tonal organization. The trigger can vanish while the organization persists, which is why a healed tissue and a continuing symptom are not a contradiction. The claim carries its own check.

A condition held in place this way should resist correction of any single variable while responding to several changed together, and should improve in steps rather than in proportion to the input. A chronic condition that resolves smoothly and proportionally under single-variable correction counts as evidence against the account, and the model accepts it as such.

Recovery is decided by the system the event meets

Put load, coupling and time course back together and the two drivers stop being a mystery. They carried different accumulated loads, different wiring to a rear-end collision, and different capacity to integrate the event at the moment it landed.

Return to the two drivers, and let the measurements above describe them.

Suppose the first arrives with a bill already running. Months of demand her body has not finished paying for, a short night behind her, a system strongly wired to the mechanical input a rear-end collision delivers. Every one of those is a quantity somebody has measured in somebody. The collision enters a heavily connected node in a system with nothing spare, and it stays long enough to become the organization the body then maintains.

The second meets the same physics with reserve to spend. The input reaches a body weakly wired to that kind of demand, registers as something the system can absorb, and resolves before anything has to be held in place around it.

Same collision. Two biological events. Neither driver chose which one they got.

Health is the width of the range

Health, in this model, is the width of the range a system can move through and return from. A healthy body can inflame, guard, brace, raise its pressure and mobilize fuel, and its health lies in being able to leave those states once they are no longer needed. Recovery is the return of the range. A failure to recover is a range that narrowed around the event and stayed narrow.

It also explains why one disturbance surfaces in such different places. A shared loss of regulation appears wherever a given body has the least capacity left to absorb it. One person after a crash presents with headaches, another with gut symptoms, another with panic, another with a shoulder that will not settle. The loudest tissue is frequently not the origin, and a presenting pattern often sits on a deeper one. Many of the unexplained conditions medicine files separately look like that from here.

Stated in the model's own words, every disease has a tonal expression, and many are initiated, maintained or amplified by failures of tonal regulation. That is a smaller claim than it first sounds, and a testable one.

Restoring tone and masking it

The same answer sets up the distinction that decides what counts as help.

Relief has real and honorable value. Unremitting pain that costs a person sleep and work is a regulatory burden in its own right, and quieting it sometimes returns the reserve a system needs before it can reorganize at all. Medication that does this is doing something worth doing. Surgery that removes a barrier the body could not resolve alone is the same principle at a larger magnitude.

Two results above say something further about how an input lands. A steady infusion of a powerful opioid and one sentence spoken to the person taking it moved the same measure in the same brain. The size of a person's social network and the plaque gathering inside their skull both showed up in a single cognitive score. Nothing about a molecule resembles a sentence, and nothing about company resembles a protein deposit. They meet at the measure because what they reach is one organization.

Widen that and the same statement covers every kind of care. A sentence, a molecule, a meal, a movement and an operation are not the same act, and none of them substitutes for another. Each enters the same organization at a different place, and which entrance suits a given person at a given moment is a question of judgment and match. The loop is continuous. A change delivered at any point of it travels through the rest.

The line the model draws runs across the professions rather than between them, and it separates interventions by aim. An input that restores tone resolves the organization, and the body reorganizes around the change. An input that masks tone improves the experience while that organization continues underneath. Every field delivers both, and both are usually entered in the record as the same success.

The convergence test

The difference between restoring and masking is measurable in principle, and here the model makes its sharpest claim. A drug is directional by construction. A drug that lowers a value lowers it in everyone who responds, including people in whom that value was already low. A restored regulator is not directional. It moves a value toward the middle of its healthy range from whichever side the person started on, bringing a high one down and a low one up.

Recovery is an unusually good arena for that test, because people who fail to recover are not uniform on the measurements. Some run fast resting heart rates and some run slow. Some drop their blood pressure on standing and some hold it high. Some are wired and hypervigilant, others flattened and unable to mount a response to anything. Same failure to recover, opposite deviations.

So the test is specifiable in advance, and one protocol would run it. Sort a mixed group by a measure of tone recorded before any input. Deliver a matched input to half of each side, against a sham matched for force, contact time and attention, so magnitude is held constant while correspondence varies. Then measure again. The model predicts convergence toward the middle in the treated arms that exceeds the sham arms, with the spread narrowing while the group average barely moves.

It predicts something further, and this is what separates the model from a list of independent set points. One input should do that for several separately regulated measures at once in the same person. No account assembled from independent set points expects it, so convergence on a single measure would leave the unification untested.

A uniform shift in one direction marks the input as one that pushes the output, helping whichever group it points at and carrying the other group further from the middle. Convergence no greater than the sham produces marks it the same way.

How well tone measured in advance predicts recovery

If a measure of tone recorded in advance carried no information about how differently the same event lands on different people, the principle behind the recovery gap would be empty. The score runs in both directions: real information, modest size, unstandardized instruments.

The cleanest hit: the pain brake before surgery

A pain research group in Haifa, led by the neurologist David Yarnitsky with Elon Eisenberg and Michal Granot, went after the body's own pain brake. They wanted to know whether it could be measured before an operation, and whether the reading would identify who hurt afterward. The pain page builds that brake in full.

In outline, apply one painful stimulus, apply a second painful stimulus somewhere far away, and see how much the first one fades. They tested 62 people before chest surgery. Those whose brake worked well were far less likely to still be in pain months later. The relationship scaled. Every ten points the second stimulus knocked off the first, on a hundred-point pain scale, roughly halved the risk. Ordinary pain thresholds predicted nothing.

Sixty-two patients, one operation type, one center.

The spread: 4 to 54 percent of the variance

Mads Werner and Per Nielsen are anesthesiologists who asked how far that kind of testing has actually got. They reviewed 14 studies that tested people's pain systems before surgery and measured their pain after it. The preoperative tests explained between 4 and 54 percent of the variance in postoperative pain, depending on which test was used. That spread is itself the finding. Far stronger than demographics. Nowhere near a tool.

The mature verdict on preoperative testing

A decade later a pain research group led from Aalborg, with Kristian Kjaer Petersen, Lars Arendt-Nielsen and Audun Stubhaug among them, went back over 25 surgical and 11 pharmacological studies. Static measures, such as how much pressure it takes to hurt, performed poorly. Dynamic ones did better.

Temporal summation, meaning how much a repeated stimulus builds, was associated with outcome in half the studies that used it. The descending brake was associated with outcome in 44 percent of the studies that used it. Seventeen of the 25 surgical studies and all 11 drug studies found some association, so the literature is not empty.

What it is instead is unstandardized, and that is the sharper problem. The same review counts seven different ways of measuring the brake and four of measuring summation. Groups reporting on one variable are not always measuring the same thing.

The best-shaped test: heart rhythm before combat

First, what is being measured. The body runs an automatic control system nobody has to think about, setting heart rate, blood vessel width, sweating and gut activity. One branch of it works like an accelerator and the other like a brake, and the autonomic nervous system page builds it properly. Its balance shows up in the spacing of heartbeats. Every beat is followed by a slightly different gap, and how those gaps behave is a validated reading of how flexibly the heart is being regulated.

Researchers working with the Framingham cohort, among them Hisako Tsuji, Martin Larson and Daniel Levy, took that reading to ordinary people rather than to cardiac patients. They reanalyzed the first two hours of ambulatory heart recordings from 2,501 residents with no diagnosed heart disease. Over the next three and a half years, 58 of them had a first cardiac event.

Lower variability predicted those events after accounting for age, sex, smoking, diabetes and the rest. A drop of one standard deviation, meaning one ordinary step down through the spread of the measure, carried a hazard ratio of 1.47. Those people ran about half again the risk over the follow-up.

Two details in that paper matter as much as the headline. Fifty-eight events is a small number. And of the eight ways the group measured variability, seven predicted risk and one did not, the exception being the ratio between the slow and the fast components of the signal.

Then the test itself, and it ran on the exception. Arpi Minassian, Dewleen Baker and Victoria Risbrough study stress and the nervous system in military populations, and they recorded five minutes of the same signal in 2,160 Marines. Framingham read how widely the gaps between beats varied.

Minassian's group read the balance between the slow and the fast components of that variation. The reading was taken one to two months before deployment to combat, and the men were diagnosed again four to six months after they came home.

Accounting for how much combat each man actually saw, the pre-deployment reading still carried information. Read as a continuous measure, the odds ratio was 1.47, with a confidence interval from 1.10 to 1.98. Split at a cut-off instead, six of the 38 men with the least favorable balance developed post-traumatic stress disorder, which is 15.8 percent, against 3.7 percent of everyone else. Those are two readings of one dataset and the small one carries the dramatic number.

Every limit belongs next to it. The authors call the evidence initial and modest. Six cases among 38 men decide the headline contrast. All participants were male, and replication has not been established. The heart rate variability page carries the measure itself in full.

The flat verdict from the operating room

Mikkel Nicklas Frandsen and Jesper Mehlsen asked the same question of surgery and could not get an answer out of the literature. Screening 2,337 records and including 63 studies, they looked for whether a reading of automatic nerve control before an operation forecasts what happens during and after it.

Some signals emerged, for blood pressure collapse under spinal anesthesia and for an irregular heart rhythm after the operation. The studies were too heterogeneous to pool at all. A systematic review that could not perform a meta-analysis is itself the result.

Measures of tone recorded in advance do carry information about how differently the same event lands, in whiplash, in concussion, in surgery and in combat deployment. That information is real, modest and imprecisely measured, and the instruments carrying it have not agreed on how to measure it. The model does not get credit for the disagreement. It gets tested by what happens when the measurement settles.

That disagreement has a sharper consequence, and the model has to accept it. Four different readings appear in this section, taken from four different literatures. A descending pain brake, a sensory threshold, a cold pain threshold and the spacing of heartbeats. Treating them as four windows onto one organization is the unification, and the unification is the claim.

So they have to behave like windows onto one organization. Recorded together in the same people, they should load on a common underlying factor rather than vary independently of each other. Loading together on that factor confirms the unification, and makes one account of what would otherwise be four unrelated observations.

What would end the argument is equally specifiable. Record these measures in a properly powered cohort before the event, follow it to a fork nobody could call at baseline, and find the advance measures carry no information about who diverges. That result would empty the principle. It has not been produced, and the model would accept it if it were.

The last thing matters most to the person reading this. The gap between two people after the same injury was never a difference in effort. It is a difference in what each system was already carrying, what it happened to be wired to, and where on its own timetable the event arrived. Those are states, and states are the kind of thing that can be measured and moved.

How the recovery gap relates to the rest of the library

Recovery is where the library converges, because every foundation of tone shows up somewhere as a difference between two people meeting the same event.

Three foundations carry the signature and each has its own page. Load holds the full allostatic machinery behind the ten-marker index that predicted decline here. Coupling builds the wiring account of why a fixed input reaches one body and misses another. Time course owns the stages on either side of the fork where an event is integrated or kept.

The condition pages extend the evidence.

• Stress and physical symptoms tells the quarantine experiments in full, where measured doses of virus met measured lives.
• Concussion follows the head injury whose recovery reads the child rather than the collision.
• Pain builds the descending pain brake that sorted surgical patients months in advance.
• Trauma carries the childhood adversity evidence at full length, including its limits.
• Fibromyalgia and sleep take the one-night pain amplification finding into chronic territory.
• Idiopathic conditions is where the unexplained presentations that shared dysregulation produces get their own account.

Two instrument distinctions matter here as everywhere. Heart rate variability is the instrument that read 2,501 Framingham residents and 2,160 Marines before their outcomes arrived, and the autonomic nervous system is the anatomy that instrument samples.

Frequently asked