# Disease as a Pathological Attractor **Absurd Health Institute** · Medical Concept Papers, Volume Three · Third editorial pass, 17 October 2026 **Article type:** Mechanistic research perspective · **Status:** Conceptual argument, not clinical guideline ## Disease persistence as an attractor with a measurable basin A disease state can become more than the continuing effect of an initial insult. The bold proposition here is that reinforcing feedback loops may produce a **self-maintaining physiological arrangement** that persists after the original trigger diminishes. This is what the term *pathological attractor* should mean. It does not mean a mystical pattern, an entrenched belief, or a diagnosis that has merely lasted a long time. A true attractor has observable stability: when the system is nudged away from its present state, its own feedbacks tend to pull it back. Consider chronic pain after a tissue injury that has substantially healed. Altered sensory gain, disturbed sleep, guarded movement, heightened threat monitoring, and reduced physical confidence can reinforce each other. The injury may have been necessary to initiate the cycle without being sufficient to explain its present strength. But the same presentation can also result from ongoing nerve injury, inflammatory disease, or structural compression. Those alternatives must be assessed rather than renamed. The attractor proposal becomes useful only when it identifies feedback variables and predicts which combination of changes could loosen the loop. ## Hysteresis and the evidence for two different paths If an initial insult requires a large perturbation to establish the disease state but a small change in the original exposure fails to undo it, the system displays a form of path dependence. An informative experiment would track the same physiological variables during deterioration and recovery. Does a defined threshold exist beyond which a marker returns slowly or becomes unusually resistant to improvement? Does the return path differ from the onset path when the exposure is reduced? If so, a static dose-response description misses part of the biology. If no such differential path appears, attractor terminology adds little. A stronger version of this paper therefore proposes mapping repeated safe measurements of sleep continuity, pain interference, activity, sensory sensitivity, and autonomic response while a known contributor changes. Investigators could compare a one-variable treatment with a coordinated intervention aimed at two reinforcing loops, measuring function rather than solely subjective relief. The thesis would lose credibility if identical improvement occurs whenever the original driver is removed, or if the extra “network” intervention offers no reproducible gain. The controversial assertion is not that all chronic illness is an attractor, but that some diseases may have **causal maintenance machinery separate from their initial cause**. ## The prediction that makes an attractor controversial An attractor account should imply that modest improvements in one maintaining variable sometimes fail while a coordinated change pushes the system beyond a transition threshold. Take pain and insomnia: better sleep for one night may not alter fear of movement, and a short walk may not change nocturnal hypervigilance. The model predicts a nonlinear improvement when both variables shift sufficiently and consistently. But that apparent threshold may simply represent accumulated treatment intensity. A credible comparison must measure and control exposure, expectations and behavioral adherence. This makes relapse as informative as recovery. If a former symptom state returns after a small stressor despite prior recovery, the model must specify whether the system returned to its old basin or whether a fresh biological injury occurred. Repeated trajectories could distinguish those hypotheses. A researcher should be able to sketch the proposed feedback arrows and then remove one experimentally or through an ethically acceptable natural change. If no arrow can be tested, “attractor” remains literary language. The Institute can defend a bold theory most successfully by making it easy to prove wrong in a carefully chosen setting. ## The strongest alternative is incomplete removal of the original driver One could misidentify an attractor because the supposedly removed cause remains present at an undetected level. Persistent infection, ongoing nerve injury, repeated environmental exposure or an untreated autoimmune process could all explain prolonged symptoms without invoking a self-sustaining dynamical basin. Therefore the most compelling attractor demonstration would begin in a context where the driver can be measured with reasonable confidence and shown to have resolved. The observation of chronic pain or fatigue after a negative test is not automatically sufficient; negative tests have limits and the initial diagnosis may itself be uncertain. Researchers would then seek evidence of feedback among maintaining variables, not merely their association. For example, does changing sleep alter threat reactivity and subsequent movement in a predicted sequence? Does reversing movement avoidance alter sleep or fear? If one step reliably leads another, the direction of causation becomes clearer. An attractor account wins its contest when it predicts intervention sequencing or relapse more accurately than continuing injury alone. It loses if the presumed network adds no explanatory or therapeutic value after the driver is measured properly. ## Detailed considerations from the original investigation An attractor in dynamical systems is a pattern toward which a system evolves under specified rules. Applied to illness, this suggests that recurring symptoms may reflect feedback-maintained configurations rather than a fresh insult on every occasion. Depression, metabolic dysregulation and persistent inflammatory states may provide useful analogies, but an analogy becomes a medical mechanism only when its variables and dynamics are specified. Stable illness is not the same as a stable organism; apparent steadiness can coexist with loss of reserve and vulnerability to disturbance. A clinical attractor model might define state variables such as insulin sensitivity, autonomic tone, activity, sleep regularity and inflammatory mediators. Feedback can make one pattern self-reinforcing: poor sleep raises sympathetic activation; activation impairs sleep and glucose regulation; fatigue reduces movement; inactivity worsens metabolic regulation. The system can settle into recurrent daily patterns even though its initial trigger was a temporary stressor. The important scientific question is whether feedback is sufficiently strong to support more than one stable regime under similar external conditions, not merely whether variables correlate. This requires observations over time rather than a single laboratory panel. A patient may describe a cycle of worsening sleep, inactivity and mood without a single new event each week. That recurrent pattern is compatible with a maintenance loop, but it cannot prove an attractor. To do more than rename the problem, an investigator would need to identify the feedback variables and show that their interaction produces persistence and, perhaps, a nonlinear response to change. There are alternatives. Continuing tissue damage, recurrent exposures, progressive anatomical disease or a poorly treated primary cause can generate persistent symptoms without any exotic attractor. The language of basins and transitions must not hide known pathophysiology or create the impression that disease severity is voluntary. A model without defined variables, units and prospective predictions is metaphor rather than dynamical science. If an attractor exists, graded perturbations should sometimes show nonlinear responses, delayed return to baseline, or hysteresis: two patients with the same measured value but different histories may respond differently. Evidence favoring simple linear regression with no memory or threshold behavior would weaken the claim. A valid clinical prediction should specify which variables must be sampled, over what interval, and what response would count as a transition. The useful shift is from asking only “what abnormal value is present?” toward asking “what repeatedly recreates this value?” The concept should enrich evaluation without replacing proven treatments or implying that every chronic illness can be pushed into remission with lifestyle changes. An attractor is a specific dynamical proposition: feedback keeps a system near a persistent state despite modest perturbations. It is not a synonym for a stubborn symptom. Chronic pain may be reinforced by avoidance, poor sleep, sensitization and distress; a single variable may improve without dissolving the cycle. Ongoing nerve injury would demand a different model. Map repeated state transitions and perturbation responses; demonstrate two stable states or characteristic return dynamics rather than merely prolonged symptoms. ## Selected scientific reading 1. Barabási, Gulbahce & Loscalzo (2011), Network medicine: a network-based approach to human disease, Nature Reviews Genetics. https://doi.org/10.1038/nrg2918 2. Kitano (2002), Systems biology: a brief overview, Science. https://doi.org/10.1126/science.1069492 *Editorial source note: The inherited selected readings are retained from the source manuscript and have not all been independently verified. Newly checked evidence is identified separately. The mechanistic extensions are hypotheses except where explicitly described as established physiology.*