TAP Human Adaptation Framework

One framework, three questions

The TAP Human Adaptation Framework organizes research around three questions: what demands arise, how the system adjusts, and how function changes over time. It distinguishes demands, regulatory processes and recovery trajectories, placing reserve and cumulative costs in context.

TAP makes the connections between these levels of observation explicit, bringing theoretical sources, model assumptions and testable questions into one structure. The sections below show how to read the framework, interpret the illustration and formulate research questions.

Read the relationships

The map places prediction and adjustment inside a feedback loop. Current state and reserve may influence how a challenge is met; behavior may also change the environment. Recovery and accumulating costs can coexist over different timescales. Arrows organize this synthesis rather than certify causal effects.

1. Demands and context

Describe the type, timing, duration and predictability of demands, together with surrounding conditions. For example, a comparison of anticipated and unexpected tasks should record workload and advance information separately to specify the exposures being compared.

The focus on anticipated requirements comes from predictive-regulation theory; specifying the system and challenge is also central to the NIH resilience framework.^[1][4]

2. Prediction, physiology and behavior

Sterling’s account motivates including anticipation, coordinated adjustment and learning.^[1] TAP represents these together with actions that may change exposure. The environment is not merely a background variable: newer allostatic proposals explicitly include reciprocal organism–environment relationships.^[5]

Current state and domain-specific reserve are possible contributors to this process. Measured capacity should be interpreted for the particular task; performance under another challenge requires corresponding observations.^[6]

3. Function over time

A demand, a regulatory response and a functional outcome are separate observations. A larger physiological response can support performance; a smaller signal does not automatically mean better function. The direction of a useful change must be defined for the outcome and setting.

The NIH framework and physical resilience research emphasize following the maintenance, decline and recovery of function in a specified system.^[4][6] TAP therefore specifies reference conditions and observation windows, distinguishing maintained function, return toward a reference and other adaptive outcomes.

The next challenge meets a changed system

Allostatic load theory focuses on cumulative costs, while reactive scope considers changing response ranges.^[2][3] Connecting a short-term response with longer-term costs requires observations across challenges, systems and relevant time scales.

TAP’s synthesis keeps longer-term capacity and costs distinct from a single short response. Training, illness, resources and repeated demands may lead to different changes. Any claimed direction, timescale or causal relation needs evidence in the particular setting.

Explore a short task

Imagine a brief cycling task that starts at time 30 and stops at time 60. In this mathematical illustration, demand, regulatory activation and task function use separate, dimensionless scales. Accurate advance notice, an unexpected start, a cancelled task and persistent activation expose different assumptions.

What the illustration assumes

Demand is a fixed pulse of 0.65 during the task. Regulatory activation moves exponentially toward the required level. An advance cue starts preparation at time 20; in the cancelled-task scenario no task occurs and the cue is withdrawn at time 30. These times and values are arbitrary.

Task function is defined as 1 − 0.7 × max(0, demand − response) − 0.22 × max(0, response − demand). This invented relationship penalizes both shortfall and excess. It is not an estimated physiological law. Changing it could change the apparent benefit of anticipation.

No fatigue, tissue repair, learning, reserve depletion or allostatic-load index is estimated. Persistent activation changes only its settling time. The example explains distinctions; it cannot establish which response is healthiest or predict a person’s recovery.

What a study would need to observe

Specify the population, system, challenge and outcome before collecting data. Record a suitable reference period, the timing and size of exposure, repeated measures of the chosen response and function, and contextual factors such as sleep, medication and task familiarity. Different outcomes may require different schedules.^[4][6]

For a short-task study, externally recorded workload, a physiological signal and an independently measured performance outcome should remain distinct. Using the same signal as both predictor and definition of success risks a circular result. Reserve should be represented by a justified capacity measure, not inferred from the shape of the same curve.

Questions that could test the framework

Candidate question one: under comparable task demands, does accurate advance information change the relationship between regulatory response and function? False cues could also test the costs of preparing for demands that do not occur.

Candidate question 2: does a prior response trajectory add out-of-sample predictive information about performance at a subsequent challenge, beyond workload, baseline function and measured capacity? The proposal is weakened if it adds no reliable information beyond those simpler predictors.

These are TAP research proposals, not findings. A study would need an appropriate protocol, prespecified analyses, confounder handling, uncertainty estimates and ethical review before recruiting participants. Observational prediction alone would not identify a causal recovery mechanism.

Sources, synthesis and limits

TAP connects different research traditions in an explicit analytical structure: distinguishing demands, responses and function, documenting the illustration’s assumptions and proposing testable questions. The source theories and TAP’s synthesis are identified separately so readers can trace the basis of each component.

The accompanying map is TAP-FIG-0001 v2.1. The publication record identifies this framework’s version and review history. Figure and model editions are managed separately, with explicit links to the associated versions.

Model specification

// TAP educational scenario v1.0. Dimensionless and deliberately not fitted to physiology.
// A fixed-demand task separates external demand, a regulatory response and task function.
// Changing these assumptions changes the result; no parameter estimates a person's health.
export const adaptationScenarioVersion = "1.0";
export const scenarioIds = ["unexpected", "anticipated", "false-alarm", "persistent"] as const;
export type ScenarioId = (typeof scenarioIds)[number];
export type ScenarioPoint = { time: number; demand: number; response: number; function: number };
export const scenarioDefaults: ScenarioId = "unexpected";
export function isScenarioId(value: string): value is ScenarioId {
  return scenarioIds.some((id) => id === value);
}
export function adaptationScenario(id: ScenarioId): ScenarioPoint[] {
  if (!isScenarioId(id)) throw new Error("Unknown educational scenario");
  let response = 0;
  return Array.from({ length: 101 }, (_, time) => {
    const demand = id !== "false-alarm" && time >= 30 && time < 60 ? 0.65 : 0;
    // Advance notice starts preparation at t=20. A cancelled task is recognized at t=30.
    const advance = (id === "anticipated" || id === "false-alarm") && time >= 20 && time < 30;
    const target = advance ? 0.65 : demand;
    const tau = target > response ? 4 : id === "persistent" ? 15 : 4;
    response += (target - response) * (1 - Math.exp(-1 / tau));
    // Explicit toy assumption: insufficient matching impairs task function; excess activation
    // also carries an immediate penalty. Neither term measures allostatic load or reserve.
    const shortfall = Math.max(0, demand - response);
    const excess = Math.max(0, response - demand);
    const taskFunction = 1 - 0.7 * shortfall - 0.22 * excess;
    return { time, demand, response, function: taskFunction };
  });
}
export function scenarioPath(points: ScenarioPoint[], key: "demand" | "response" | "function") {
  return points.map((p, i) => `${i ? "L" : "M"}${(48 + p.time * 7.84).toFixed(2)},${(124 - p[key] * 84).toFixed(2)}`).join(" ");
}

Download this edition’s model source ↧

References

1. Peter Sterling (2012). Allostasis: a model of predictive regulation. Physiology & Behavior. https://doi.org/10.1016/j.physbeh.2011.06.004
2. Bruce S. McEwen, Eliot Stellar (1993). Stress and the individual. Mechanisms leading to disease. Archives of Internal Medicine. https://doi.org/10.1001/archinte.1993.00410180039004
3. L. Michael Romero, Molly J. Dickens, Nicole E. Cyr (2009). The Reactive Scope Model — A new model integrating homeostasis, allostasis, and stress. Hormones and Behavior. https://doi.org/10.1016/j.yhbeh.2008.12.009
4. LaVerne Brown, Barbara Cohen, Rebecca Costello, Olga Brazhnik, Zorina Galis (2023). Conceptualizing a resilience research framework at The National Institutes of Health. Stress and Health. https://doi.org/10.1002/smi.3260
5. Laura A. Harrison, Antonio J. Gracias, Karl J. Friston, J. Galen Buckwalter (2025). Resilience phenotypes derived from an active inference account of allostasis. Frontiers in Behavioral Neuroscience. https://doi.org/10.3389/fnbeh.2025.1524722
6. Heather E. Whitson, Wei Duan-Porter, Kenneth E. Schmader, Miriam C. Morey, Harvey J. Cohen, Cathleen S. Colón-Emeric (2016). Physical Resilience in Older Adults: Systematic Review and Development of an Emerging Construct. The Journals of Gerontology: Series A. https://doi.org/10.1093/gerona/glv202
7. Ramsay DS, Woods SC (2014). Clarifying the roles of homeostasis and allostasis in physiological regulation. Psychological review. https://doi.org/10.1037/a0035942

Publication record

TAP-MDL-0001 · v0.4 · English

The Allostasis Project is founded and supported by Resilio.

© The Allostasis Project. All rights reserved. Reproduction or adaptation requires separate permission, subject to applicable legal exceptions.

Download complete edition record (JSON) ↧