How the mind–body split hides solutions, from autism to aging
An ESF perspective. This essay describes a conceptual framework, not medical guidance. Nothing here is a substitute for clinical evaluation, and none of it claims to reverse or treat any neurological condition.
I. Two rooms, one problem
Picture two rooms.
In the first, a nine-year-old autistic boy is under his desk at school. Twenty minutes ago he was doing math. Now he is not doing anything anyone asked him to do, and the incident report that will be written this afternoon will use the word behavior four times. No one in the room knows that he has needed the bathroom since before math started, that he could not find the words or the window to say so, that the fluorescent light above his desk has been flickering at a frequency he can feel in his teeth, and that keeping his body upright in a chair that does not fit him has been, for the last hour, a job.
In the second room, in a memory-care unit forty years up the lifespan, an eighty-one-year-old woman who was pleasant at lunch is now pacing at six in the evening, pulling at the door handle, unable to answer questions she answered easily this morning. The note in her chart will say increased confusion and agitation. No one on the evening shift knows that she has not urinated since early afternoon, that her hip hurts in a way she can no longer name or localize, and that walking — which used to be free — now requires her to consciously supervise her own feet.
We have built entire professional apparatuses around these two rooms, and the apparatuses barely speak to each other. One belongs to developmental psychology and special education; the other to geriatrics and neurology. One produces behavior plans; the other produces medication reviews. What neither apparatus is built to see is the thing the two rooms have in common: in both, a nervous system is doing an enormous amount of unrecognized work to manage a body, and the cost of that work is being read, from the outside, as a problem of the mind.
This essay is about that misreading — where it comes from, what it costs, and why some of the most useful things we could do for people across the entire lifespan are hiding in plain sight, misfiled under comfort, custodial care, or behavior management, because our explanatory frameworks inherited a split between mind and body that the nervous system itself never agreed to.
II. The seam
That medicine’s specialties each inherited a piece of the whole, and that no discipline inherited the integrative level itself, is an argument I have made at length elsewhere and will take as given here. This essay is about one particular fault line in that partition — the one between mind and body — and about what falls through it.
The mind–body split that shapes modern care is less a philosophical position than an institutional fact; it lives in org charts, not arguments. Medicine divided the body among organ specialties and gave the mind to psychiatry and psychology; education took cognition and handed the body to PE and OT. So when a whole person arrives, they are administratively disassembled — bladder to urology, gait to physical therapy, mood to psychiatry, cognition to neuropsychology, “behavior” to whoever is left in the room when it happens. The person is the seam. Anything whose mechanism runs from the body through regulation into thought and action has no department, and gets described from whichever side happens to be watching, in whichever vocabulary that side owns.
That is why the boy under the desk gets a mental-vocabulary description — noncompliance, dysregulation — for what may be substantially a body event, and why the woman at the door gets one too: confusion, agitation, sundowning. Because the description determines the response, the response aims at the mind — redirection, reinforcement, reassurance, sedation — while the body’s contribution goes unaddressed, unmeasured, and unimagined.
The claim here is not that these are “really” body problems rather than mind problems. That is the same split run in reverse, and replacing one reductionism with another still leaves the mechanism missing. The claim is that the split itself is the error: mind and body are not two systems that interact but one regulatory system described at two scales — a claim this essay states as testable rather than settled, and whose payoff is that conditions from autism to dementia become tractable in ways the split renders invisible.
One clarification belongs here, because the seam runs through how we argue about disability itself. The medical model locates the problem inside the person and sets out to fix it. The social model locates it in a disabling environment and sets out to change that. The disagreement is real, and the social model’s corrective was necessary. But both positions accept the partition — they differ over which side of the line the problem sits on, not over whether the line exists. What neither describes well is what happens across it: a particular nervous system meeting a particular room, where the demand that room makes depends on the architecture doing the meeting. So “support the body” here is not a return to the medical model. The body in question may be injured, and in dementia something is. It is also working — and part of what it is working on is the room.
III. What the brain is actually doing all day
To see why, it helps to start with a picture of the brain’s actual workload — not the workload we test, but the workload it carries.
A useful body of contemporary neuroscience describes the brain as a prediction machine: rather than passively receiving the world and the body, it continuously anticipates them, and it spends its effort on the places where anticipation fails. A prediction error — a signal that arrived when it wasn’t expected, or didn’t arrive when it was — is not just information. It is work. It has to be noticed, interpreted, prioritized, and resolved, and until it is resolved it keeps recruiting attention, autonomic adjustment, and behavioral control.
Now consider how much of the brain’s predictive traffic concerns the body itself. Interoception — the sensing of the body’s internal condition — streams a continuous report: bladder and bowel state, thirst, hunger, temperature, pain, breath, heart, the position and load of every joint. Movement adds its own channels: proprioception, balance, touch, the visual flow of a moving world, the constant checking of intended action against actual result.
When these channels are quiet, predictable, and well-resolved, they cost almost nothing. The body runs in the background and the foreground is available for what we conventionally call thinking. But every unresolved bodily signal changes that arithmetic. A full bladder is not “just a body problem” waiting politely for attention. The nervous system has to notice it, infer what it means, judge its urgency, inhibit the immediate response, plan a resolution, and keep the whole person safe and socially coherent while doing so. That is a cognitive workload — and it is a workload that never appears on any cognitive test, because our tests, like our institutions, assume the body has been handled offstage.
There is converging evidence for this picture at the component level. Experimental work has found that sustained bladder retention degrades attention even in healthy adults. Studies of older adults have associated bladder dysfunction with slower performance on executive tasks — a correlation that plausibly runs in both directions, since executive difficulty also makes continence harder to manage. Neuroimaging consistently places the representation of bodily state in regions deeply entangled with salience, attention, and emotion, not in some sealed somatic annex. None of these findings alone proves the integrated claim. Together they make the integrated claim hard to avoid: the systems that monitor the body and the systems that allocate thought are not neighbors.
IV. Movement is not free
The second half of the hidden workload is movement — and here the mind–body split has produced one of its most consequential blind spots.
We tend to treat movement as either automatic (and therefore cognitively free) or impaired (and therefore a motor problem, filed away from cognition). What the research on aging and neurological conditions shows instead is that automaticity is an achievement, and when it degrades, movement starts consuming the very control processes we call cognitive. Walking while performing another task is measurably costly for older adults, and the size of that cost tracks processing speed and executive function. A person who seems “more confused” while walking may not have lost knowledge in the last thirty seconds; they may be spending their attention on not falling. Conversely, peppering someone with questions during a difficult transfer can destabilize both the walking and the answering, because both are drawing on the same control architecture at the same moment.
The same blind spot runs through the neurodevelopmental literature, where it has taken decades for motor and sensorimotor differences in autism to be treated as central rather than incidental. A growing research tradition — the sensorimotor work of investigators like Elizabeth Torres, the developmental dynamics of Thelen and Smith — has argued that movement is not downstream of cognition but developmentally and computationally entangled with it: that atypical sensory feedback and motor variability change what every action costs, and that a child whose body returns noisy or unpredictable feedback is solving a harder control problem with every reach, every step, every attempt to sit still. From the outside, the visible surplus of that problem gets labeled fidgeting, clumsiness, avoidance, or noncompliance. From the inside, it is engineering under load.
Autism and aging are usually studied as if they occupied different universes. Notice what they share once the split is removed: in both, the assumed-free foundation — a predictable body, automatic movement — is not free. In both, the cost of the foundation is invisibly deducted from everything built on top of it. And in both, the deduction is systematically misattributed to the top floor, because the top floor is the only part our instruments were pointed at.
V. The energy story is real — and incomplete
At this point it is tempting to summarize everything so far in energetic terms, and the temptation should be partly indulged, because the energetics are real. A significant and welcome development in recent medicine is the return of energy to the center of health science: the recognition that the body–brain system operates under genuine energetic constraint, that stress responses are metabolically expensive, and that chronic demand forces trade-offs against long-term maintenance and repair. Recent work in this tradition — including the energy-constraint framework published this year by Behnke and colleagues in Trends in Endocrinology & Metabolism — is convergent with everything this essay has described. Prediction errors cost energy. Compensatory movement control costs energy. Unresolved interoceptive alarm costs energy. “Brain energy” is not a metaphor.
But a budget, by itself, does not explain the two rooms we started in. Two people with comparable energetic resources can differ radically in what a given afternoon does to them, and the same person can absorb at ten in the morning a demand that undoes them at six in the evening. Allocation implies an allocator. The budget describes what is possible; it takes a regulator — an organized control architecture with its own structure, couplings, and rhythms — to determine what actually happens. This is not a criticism of the energy frameworks, which are asking exactly the right substrate questions; it is the completion of their own question calls for.
Seen from the regulator’s side, three features of the picture sharpen.
First, demand is channel-specific and coupled. Bladder urgency, postural instability, pain, an unfamiliar room, a hurried caregiver, and fear of falling are not one undifferentiated load; they arrive through different channels, and their interactions matter. None of them has to be extreme. It is the coupling — urgency requiring inhibition, while standing requires cardiovascular adjustment, while an unfamiliar corridor demands orientation, while someone is asking questions — that can carry a person toward a threshold.
Second, capacity is a profile, not a number, and it is rhythmically organized. The regulatory framework I work within treats present capacity as multidimensional and dynamically reconfigured by circadian and ultradian timing — which is why the evening is not simply “later in the day” but a genuinely different regulatory regime, and why the same demand can be tolerable at one phase of a person’s internal timing and destabilizing at another.
Third, what looks like sudden failure is often a state transition: a shift in the whole system’s operating mode when coupled demands cross what the current configuration can coordinate. The meltdown, the shutdown, the six-o’clock unraveling — these are not character events or arbitrary behaviors. They are what a regulatory system does at a threshold. That they arrive “for no reason” is itself diagnostic: the reasons live in channels nobody was watching.
VI. Same pattern, different labels
Line the phenomena up and the pattern is difficult to unsee.
An autistic child’s capacities fluctuate across the day in ways that baffle a trait-based model — fluent at breakfast, wordless by mid-afternoon — and the fluctuation gets narrated as inconsistency or willfulness. An older adult with dementia shows the same dissociation between what the brain still holds and what is currently reachable — oriented this morning, lost by evening — and it gets narrated as disease progression measured hour to hour, which it cannot be, since the neurodegeneration did not advance between lunch and dinner. A person in post-infectious fatigue can think clearly for an hour and then not at all, and gets narrated as deconditioned or depressed. A hospitalized elder becomes delirious, and the delirium resolves — revealing that the cognition was never gone, only inaccessible.
The vocabulary differs by department: behavior in the classroom, sundowning on the unit, brain fog in the clinic, agitation in the chart. The structure is the same everywhere: a distinction between what a nervous system has and what it can currently reach — between underlying architecture and state-dependent access. The mind–body split obscures this structure because access is precisely the thing that lives across the seam: it is set moment to moment by physiological state, bodily demand, timing, environment, and support, none of which the mind-side vocabularies were built to represent.
One well-known clinical finding makes the distinction concrete. A large cluster-randomized trial in nursing-home residents with dementia found that systematically assessing and treating pain substantially reduced agitation — yet did not improve scores on standard cognitive testing. Read through the old split, that is a null result for cognition. Read through the access model, it is the whole thesis in one study: treating the body did not restore lost neural function, and no one should claim it did; what it did was reduce a competing demand, and the person’s remaining capacities became more consistently expressible. Distress fell. Cooperation, communication, and presence improved. The brain did not get better. The person got more of their brain back.
That is why the precise claim matters so much, and why this essay has been careful with it. The defensible proposition is not treating the body improves dementia or sensory supports treat autism. It is: making the body easier to read, regulate, and move can reduce state-dependent interference, and thereby increase access to the cognition and capability the person already has. Nothing reversed; something released.
VII. The solutions the split has been hiding
Once the claim is stated that way, look at what has been sitting in the “merely comfort” file.
Scheduled toileting and bowel regularity. Hydration. Routine pain assessment in people who can no longer report pain in words. Glasses and hearing aids actually on the person. Stable footwear, supported seating, safe handholds. One clear instruction at a time, and no quizzing during transfers. Predictable routines and rooms. Clothing that fits the body’s sensory reality. Light and sound environments tuned to the nervous system that has to live in them. A calm companion — because co-regulation, the oldest technology humans have, extends what a person’s own system can currently absorb.
Under the mind–body split, this list reads as kindness: worthy, low-status, unrelated to the “real” clinical problem. Under an integrated model, the same list reads as cognitive and regulatory support: each item removes an unresolved signal, an ambiguity, or a coupled demand that the brain would otherwise have to service out of the same finite control it uses for orientation, language, memory retrieval, and safe movement. The interventions did not change. Their meaning did — and with it, their priority, their timing, and who is expected to take them seriously.
The same reversal applies at the other end of the lifespan. Sensory accommodations, movement breaks, interoceptive support, communication that does not require the body to be silent and still first — under a behavior model these are indulgences to be earned; under an access model they are the substrate of the school day. A simple household and clinical heuristic falls out of the whole argument, and it works in both of our rooms: support the body before testing the brain. Before interpreting reduced performance as lost ability — before the incident report, before the cognitive screen, before the conclusion — ask what the body has been asking for, and for how long.
And this reframing carries one more gift, which has nothing to do with efficiency. When the six-o’clock unraveling is a character flaw or a behavior problem, the people involved — the person and their caregivers alike — inherit blame. When it is a regulatory system managing an extraordinary workload with the coordination it currently has, blame gives way to something closer to respect. The difficulty is real, but it attaches to the hour and the work, not to the person. They are living in their world, doing the best they can with the system they have. We can meet them there — and the meeting itself is regulation.
VIII. What this framework does not claim
Precision requires boundaries, so let me draw them explicitly.
This framework does not claim that autism, dementia, Parkinson’s disease, or any other condition is “really” an unmet bodily need. Architectural differences and neurodegeneration are real; no schedule of toileting reverses a tangle of tau, and no sensory diet rewires a developmental trajectory. It does not claim that meeting bodily needs improves cognition in the testable sense, and where trials have looked, they have often found exactly the dissociation described above — better state, unchanged scores — which is the finding this model predicts, not a finding it must explain away. It does not replace medical evaluation: a sudden change in cognition or function is a medical event until proven otherwise, full stop. And it must not curdle into a new checklist imposed on people, or a toxic-positivity mirror of the deficit frame in which struggle becomes unmentionable. Caregiver exhaustion is real; the work is real; naming the work honestly is part of respecting it.
What the framework claims is narrower and, I think, stronger: that state-dependent access is a genuine, mechanistically motivated layer of every one of these conditions; that this layer is systematically invisible to frameworks that split mind from body; that a meaningful fraction of preventable suffering lives in that invisible layer; and that the integrated hypothesis — interoceptive and sensorimotor predictability as a determinant of what remains accessible — is stated precisely enough to be tested, and should be.
IX. One system
The deepest version of the argument is not clinical but conceptual. We have spent several centuries perfecting the study of the mind as if it were housed in the body the way a driver is housed in a car — served by it, fueled by it, occasionally inconvenienced by it. Everything in this essay points the other way. The body is not the vehicle of cognition; it is part of the medium in which cognition happens. Thinking is something a whole regulated organism does, and the quality of the regulation — how readable the body is, how predictable its signals, how automatic its movement, how supported its needs, how well-timed its demands — is not the backdrop of mental life. It is a determinant of which mental life is reachable today, at this hour, in this room.
The boy under the desk and the woman at the door have never met. They are separated by seventy years, four medical specialties, and two entirely different literatures. But they are having the same kind of afternoon, for the same kind of reason, and they would be helped by the same kind of seeing. Supporting the body is not separate from supporting the mind, because there was never a separation to begin with — only a seam in our institutions, which the people we care for have been quietly falling through.
The solutions are not exotic. Many of them cost almost nothing. They have been here the whole time, filed under the wrong name.
The Evolutionary Stress Framework (ESF) is a conceptual research lens developed at the Center for Adaptive Stress. This essay synthesizes established component findings into an organizing hypothesis; the integrated model has not been tested as a causal whole and is presented as a framework for research and reflection, not as clinical guidance. Empirical claims reference published literatures on interoception, dual-task gait, pain management in dementia, and energetic constraint; full citations are being finalized and specific studies are described generically pending verification.
Leave a Reply