Trigger Stacking in Dogs: What Is Measured and What Is Assumed

1. What the Model Claims

1.1 The Core Proposition

Multiple stressors occurring within a short period accumulate. Each raises arousal without the dog returning to baseline, and a stimulus the dog would normally tolerate produces a disproportionate response because it lands on top of the others.

1.2 The Threshold Metaphor

The model is usually drawn as a bucket or a column filling toward a line. Below the line the dog copes; above it, behavior appears — barking, lunging, snapping, or shutting down.

1.3 What Counts as a Trigger

Practitioner accounts include both obvious and subtle contributors: noise, unfamiliar dogs or people, restraint, handling, changes in routine, poor sleep, hunger, and physical discomfort. Positive arousal is usually included as well, which is worth noting: it makes the model harder to falsify, since almost any preceding event can be recruited as a contributor after the fact.

1.4 What This Article Does Not Argue

Not that trigger stacking does not exist. The observation the term names — that a dog copes on one day and not the next, and that a demanding morning makes a demanding afternoon harder — is real and reported by enough households to take seriously.

What is unexamined is the mechanistic model built on top of that observation: the container, the additive units, the threshold and the drain rate. The observation stands; the explanation offered for it has not been tested directly in this species, and those are different statements (separating an observation from the explanation offered for it).

1.5 Why the Model Is Worth Taking Seriously

It is not a fringe idea. It appears in professional training material, in behavior consultations and in veterinary behavior discussion, and it directs real decisions about how dogs are managed week to week.

A claim with that reach deserves examination rather than either adoption or dismissal. Dismissing it would also discard the practical advice, most of which is sound for other reasons.

2. Where the Term Comes From

2.1 No Primary Source

Trigger stacking does not originate in a study. It has no defining paper, no operational definition in the research literature, and no canine experiment testing whether the accumulation it describes occurs as stated.

It is a practitioner model that entered training vocabulary and spread, in the way that constructs without gatekeepers do. That is not a fatal criticism — practitioner models can be useful before they are validated — but it changes what can honestly be claimed for it.

2.2 How It Acquired Scientific Appearance

Secondary accounts attach real studies to the model retrospectively. Beerda's work on canine stress responses, research on individual variation in stress tolerance, and general HPA-axis physiology are cited as though they had tested stacking. They did not; they established that dogs mount measurable stress responses, which the model then assumes accumulate in a particular way.

2.3 Why This Matters Here

An article recommending sub-threshold work on the basis of stacking is recommending something defensible for reasons that may not hold. The recommendation can be right while the stated mechanism is unverified, and separating the two is the point of this article.

2.4 How Ideas Acquire Citations They Never Had

A plausible idea gets repeated, an early repetition attaches a vague reference, later authors cite the repetition, and within a decade the claim looks sourced. Nothing dishonest has to happen for this to occur.

Tracing a familiar claim back to its origin is often the most informative thing a review can do, and it is rarely done because it is slow and produces a negative result.

3. The Number That Has No Source

3.1 The Claim

The figure quoted throughout training material is that canine cortisol takes 24 to 72 hours to return to baseline after a significant stressor. It is the quantitative backbone of the model, and it underpins the related cortisol vacation — a rest period of days to weeks prescribed for stressed dogs.

3.2 What Happened When Someone Checked

A behavior professional writing in the IAABC journal described having advocated cortisol vacations and quoted the 24-to-72-hour figure for years, then being unable to find any evidence that the information rests on scientific findings when she came to cite it (Depta, 2020). The same article warns against quoting unbacked claims of this kind, including the parallel assertion that chronically stressed dogs need four to six weeks to recover.

3.3 What the Canine Data Show Instead

Measured cortisol recovery is faster and more variable than the figure suggests. In a study of stress reactivity in 32 puppies and 16 young adults, salivary cortisol was sampled at 10 and 40 minutes after a behavioral test — timepoints chosen because they capture reactivity and recovery respectively (Lensen, Moons & Diederich, 2019).

In shelter research, dogs' cortisol-to-creatinine ratios rose during an intervention day and had returned to pre-intervention levels the following day. Across a temporary fostering study of 207 dogs with over a thousand cortisol values, ratios dropped during fostering and returned to baseline after return to the shelter (Gunter et al., 2019).

None of this establishes a universal recovery time, in either direction. The point is not that dogs always recover quickly — recovery speed differs substantially between individuals and with the nature of the stressor, and replacing one general rule with another would repeat the error this chapter is about. What it does establish is that the 24-to-72-hour figure is not derived from these data (and cortisol is not a straightforward stress readout in any case).

3.4 Cortisol Is the Wrong Currency Anyway

Cortisol cannot distinguish distress from positive arousal, and researchers have repeatedly cautioned that no single metric captures the presence or extent of stress. A model built on the assumption that accumulated cortisol is what fills the bucket is building on a marker that does not carry that meaning (a measurement problem that recurs throughout welfare research).

3.5 Why This Particular Number Spread So Well

It is specific, memorable, actionable and kind: it tells a household exactly how long to wait and absolves the dog of blame. Those are the properties of a claim that travels, and none of them is evidence — the same four describe most of what circulates in this field without support.

3.6 What Checking It Involved

Someone followed the citation trail and found it ended without a study (Depta, 2020). That is ordinary scholarly work and it is the reason this article can say what it says. Without that check, the figure would appear here as an established finding like any other.

The figure is still in circulation, which is what usually happens when a correction appears in a professional journal rather than in the training material where the claim actually lives.

4. Why Cortisol Cannot Carry the Argument

4.1 It Is Not a Distress Meter

Cortisol rises with play, with anticipation of a walk and with greeting a returning owner. A raised value establishes that something activating happened, not that the dog was suffering.

4.2 The Matrix Decides the Timescale

Blood and saliva capture the last minutes to hours, urine a few hours, feces roughly a day, and hair the weeks over which it grew. A model about accumulation over a weekend needs a measure matched to that window, and most available data are not (how long-term stress is measured).

4.3 Measurement Carries Its Own Confounds

Sampling itself can raise the value, individual baselines differ substantially between dogs, and in hair the pigment of the coat affects the reading. None of that makes cortisol useless; it makes single values a weak foundation for a quantitative claim about accumulation across days.

4.4 The Currency Problem

Even with perfect measurement, the model requires that stressors add in some common unit. Cortisol is one output of one system, and behavioral thresholds are not obviously a function of it.

A dog can be physiologically activated and behaviorally fine, or behaviorally near its limit with an unremarkable cortisol reading. Neither state is predictable from the other.

4.5 What the Canine Cortisol Work Does Support

That dogs mount measurable stress responses, that recovery speed differs between individuals, and that sustained low-control conditions change the system (Beerda et al., 1998). Those are real findings and they are not the same claim as the model makes. The distance between them is where the assumptions live.

5. What Is Actually Supported

5.1 Dogs Mount Measurable Stress Responses

This is established. Acoustic and visual stimuli produce increased heart rate, elevated cortisol and behavioral change (Beerda, Schilder, van Hooff, de Vries & Mol, 1998), and the physiological machinery — sympathetic activation within seconds, HPA activation over minutes — is well described.

5.2 Recovery Speed Varies Between Individuals and Predicts Behavior

The most directly relevant finding is about recovery rather than accumulation. High reactivity 10 minutes after a behavioral test was associated with desirable traits, while slow recovery at 40 minutes was associated with undesirable ones — greater dog rivalry and stranger-directed fear (Lensen et al., 2019).

The authors read the strong early response as adaptive and the persistent elevation as unsuccessful coping. That is a real empirical anchor for the intuition behind stacking: recovery time matters, and it differs between dogs (consistent with what coping-style research describes).

5.3 Chronic Stress Alters the System

Prolonged stress changes HPA function rather than simply raising it. Dogs subjected to five weeks of social and spatial restriction showed increased daily cortisol but an attenuated cortisol response to an acute noise stressor — reported in a sample of eight dogs (Beerda et al., 1999).

This is an important complication for the bucket metaphor: chronic load does not simply mean more cortisol per event.

5.4 Sensitization, With a Caveat

Repeated exposure to an aversive stimulus can lower the response threshold rather than raise it, and fear learning generalizes from the original trigger to related stimuli (as the fear conditioning literature sets out). That produces something that looks like stacking over weeks — through a different mechanism than same-day accumulation, and one with an identifiable learning history (as extinction and return illustrate).

Two caveats belong here rather than in the appraisal. Sensitization is offered as one possible account of cases that look like stacking over weeks, not as a replacement model that explains every case; where it applies, there is an identifiable learning history, and where there is none, it does not. And sensitization is well established as a general learning process, and the experimental work demonstrating it comes overwhelmingly from other species. Canine studies describe sound sensitivity and its physiological correlates; they have not tested the sensitization trajectory itself. This section rests on extrapolation, which is exactly the charge the article levels at the stacking model.

5.5 What the Supported Findings Add Up To

Dogs mount stress responses; those responses take time to resolve; individuals differ in how long; sustained load changes the system. Four findings, all of them canine, and none of which requires the accumulation model in order to be useful (Beerda et al., 1998; Lensen et al., 2019).

Most of the practical advice associated with trigger stacking follows from these four directly. Space demanding events apart, allow time to settle, and treat individuals as individuals.

6. What Is Assumed

6.1 That Stressors Add Up

The model treats stressors as additive units filling a container. Whether canine stress responses combine additively, sub-additively, or interact in some other way has not been tested. Physiological systems more often show habituation, adaptation and interaction than simple summation.

6.2 That There Is a Threshold

The line on the diagram is the model's central visual claim and its least supported one. No study has identified a threshold in dogs above which behavior changes categorically. Behavioral output more plausibly varies continuously with arousal (with the inverted-U relationship itself contested), and inhibition degrades gradually rather than switching off (as the self-control evidence indicates).

6.3 That the Bucket Empties on a Schedule

The recovery figure is unsourced, and the available canine studies show recovery of the measured cortisol variables within minutes to roughly a day in the settings studied, rather than supporting a universal 24-to-72-hour rule. Whether a genuine refractory period exists during which a second stressor produces a disproportionate response is a testable question that has not been tested.

6.4 That Positive Arousal Counts the Same

Play, greeting and excitement are routinely included as stack contributors. Cortisol does not distinguish valence, which is often cited in support — but the inference runs backwards: the marker's inability to tell distress from excitement is a limitation of the marker, not evidence that the dog experiences them equivalently.

6.5 Why the Assumptions Matter Individually

Each could be false independently. Stressors might not add linearly, the threshold might not be a single quantity, recovery might not follow a schedule, and positive arousal might behave differently from negative.

The model requires all four to hold, and none has been tested in dogs. A model resting on four untested assumptions is a hypothesis with unusual reach.

6.6 The Positive Arousal Question Is the Most Interesting

Whether a morning of enjoyable play loads the same system as a morning of frightening encounters is an empirical question with a plausible answer either way, and it changes the recommendations substantially.

A household told that excitement and fear fill the same bucket will manage a dog's week very differently from one told they do not.

7. What Measuring This Would Require

7.1 The Study Nobody Has Run

Testing the model means exposing dogs to a controlled sequence of stressors at defined intervals and measuring whether responses to a fixed later stimulus differ by sequence. That design is feasible and has not been reported for dogs, though comparable designs exist in other species, which is where the idea came from.

7.2 What Would Have to Be Defined First

What counts as a trigger, how intensity is scored, what interval separates one event from the next, and what outcome measure stands for the accumulated state. None of those has an agreed definition, which is part of why the study has not been run.

A model that cannot specify its own variables is difficult to test, and that difficulty is a property of the model rather than of the researchers. Sharpening the definitions would be the first useful piece of work.

7.3 The Welfare Obstacle Is Real

Deliberately stacking stressors on dogs to see what happens is the kind of design that would struggle for approval, particularly where the hypothesis predicts the animals will be worse off.

That is a legitimate reason the direct evidence is missing, and it does not license treating the model as established in the meantime. Untestable for good reasons and untested are different situations with the same consequence.

7.4 What Could Be Done Instead

Observational work would go a long way: recording the events preceding incidents in a large number of cases and testing whether density predicts severity. That requires no manipulation, no ethics approval beyond the ordinary, and no funding to speak of.

Nobody has done it, which is the more telling gap. Twenty years of teaching a model and nobody has counted. The data would exist already in the case notes of any busy behavior practice.

7.5 Why the Absence Matters

A model taught this widely, for this long, with no direct test and a central number that turned out to have no source, is worth examining on its own terms. That is what this article is for, and the examination is friendlier to the model than it might sound.

8. Why the Model Persists

8.1 It Explains Something Real

Owners genuinely observe that their dog copes on some days and not others, and that a bad morning predicts a bad afternoon. The observation is sound even if the proposed mechanism is unverified.

The model also fills a real gap in owner perception. Owners are unreliable at identifying stress in their own dogs (Mariti et al., 2012), and a framework that directs attention to accumulated load is more useful than no framework at all.

8.2 It Is Kind

Reframing a dog from unpredictable to overloaded changes how it is handled, and generally for the better. A model that produces humane management has social utility independent of its evidential status (much as other useful framings do).

8.3 It Is Unfalsifiable as Usually Stated

Because almost any preceding event can be counted as a contributor, and because no threshold is specified in measurable terms, the model accommodates every outcome. A dog that reacts was stacked; a dog that does not was below threshold. That flexibility is why it feels so consistently confirmed.

And the number gave it weight: 24 to 72 hours converted a metaphor into something that sounded measured. Specific figures travel further than qualified ones, and this one traveled without its source (the same asymmetry that keeps null results invisible).

8.4 The Cost of an Unfalsifiable Model

It cannot be wrong, so it cannot be improved. Where a plan fails, the model explains the failure as easily as it explained the success, and nothing is learned.

That is the practical case against keeping it, separate from whether its predictions happen to be right. A model that could be wrong is more useful than one that cannot.

9. What Survives for Practice

9.1 The Recommendations Hold

Spacing demanding events, allowing recovery, reducing exposure after a difficult episode and treating a dog's prior day as relevant context are all defensible. They follow from individual differences in recovery speed (Lensen et al., 2019) and from what chronic stress does to the system, without requiring the bucket.

9.2 What to Stop Saying

That cortisol takes 24 to 72 hours to normalize. That a stressed dog needs a fixed number of rest days. That arousal accumulates additively toward a threshold. None of these has canine evidence, and stating them as fact is the kind of claim that gets repeated by others.

9.3 What to Say Instead

That dogs differ in how quickly they recover, that recovery speed relates to how they behave generally, and that a dog's response to a later event can depend on how well it has recovered from earlier demands. This is weaker than the usual formulation and it is what the evidence supports.

9.4 Why Training Through It Fails

The one part of the practitioner account that deserves reinforcing rather than qualifying concerns what happens if the session continues anyway.

As arousal rises, retrieval and performance of previously trained behavior can become increasingly unreliable — which is a statement about performance, not about a brain that has shut down. Pressing on increasingly produces failed repetitions and poor learning conditions rather than useful practice (what this means for dogs that lose it on leash).

Two further consequences are documented rather than inferred. Aversive handling in that state carries measurable costs — elevated cortisol, stress behaviors, a more pessimistic outlook, no gain in effectiveness (as the fallout evidence sets out) — and it damages the handler relationship the training depends on (with the welfare consequences quantified).

And what a frightened dog associates in that moment is not confined to the trigger. Fear learning can generalize to contextual cues present at the time — including places, surfaces and people. A session continued past the point of usefulness does not merely fail; it can add associations that were not there before.

None of this specifies a cut-off, and the working rule is a practical one rather than a measured threshold: when a dog stops responding to what it knows, the session is over.

9.5 Where the Real Diagnostic Question Sits

A dog whose tolerance drops suddenly and persistently is more likely to have a physical problem than an unusually full bucket. Pain features heavily in behavior caseloads and is the differential that stacking narratives displace (as the chronic pain evidence shows; including internal sources that are easily missed).

9.6 What This Article Is Not Saying

Not that stress does not accumulate, which may well be true and is untested. Not that the recommendations are wrong, since most follow from evidence that does not depend on the model. And not that practitioners using the term are being careless.

The argument is narrower: a widely taught model with no direct canine test and a central figure with no source should be presented as a working idea rather than as a finding.

10. How to Talk About This With a Household

10.1 The Observation Is Valid

Households notice that a dog reacts on Saturday to something it handled without difficulty on Tuesday. That observation is real and worth taking seriously; the explanation offered for it is the part under discussion. Telling a household the model is untested does not mean telling them they imagined the pattern.

10.2 What to Say Without the Metaphor

That dogs do not reset instantly after a demanding event, that recovery takes longer than most people assume, and that spacing demanding events apart is worth doing. All three are supportable and none requires a bucket, a capacity or a number of days.

10.3 Why the Metaphor Is Worth Dropping

It implies a quantity that can be measured, a capacity that can be known and a drain rate that can be timed. None of those exists, and a household that believes otherwise will draw conclusions the evidence does not support.

It also makes the model unfalsifiable in practice: any good day is a low bucket and any bad day a full one.

10.4 What to Record Instead

How many demanding events occurred, how much time fell between them, and how long the dog took to settle afterwards. Three countable things that answer the practical question without the theory, and all three are available to any household with a notebook.

Over a fortnight they also show whether the pattern the household describes is actually there, which is worth knowing before a plan is built on it. Sometimes it is not.

10.5 The Differential Nobody Runs

A dog whose tolerance has narrowed over weeks is a candidate for a medical workup before any accumulation account is accepted. Pain lowers thresholds and produces exactly the pattern the model is invoked to explain, with orthopedic and dental findings the common ones and neither visible on a walk.

10.6 What a Careful Version of the Claim Looks Like

Dogs do not return to baseline instantly after demanding events, recovery takes longer than most households assume, and events occurring close together are plausibly harder to handle than the same events spread out. Every word of that is defensible.

What it does not include is a quantity, a capacity, a drain rate or a number of days. Those are the parts that were added along the way, and they are the parts with no support behind them.

10.7 Why This Matters Beyond One Model

The pattern here recurs: a plausible idea, a memorable number, a citation trail that ends nowhere, and two decades of confident teaching. Recognizing the pattern is more useful than correcting this one instance.

The test is simple enough to apply anywhere: where does this number come from, and has anyone followed it back? Most confident figures in dog training have not been through that.

10.8 What Would Settle It

An observational study relating the density of preceding events to the severity of incidents across a large number of cases. No manipulation, no welfare problem, and a result either way would be worth having.

Until someone runs it, the honest position is that the practical advice is sound and the model behind it is untested. Those two statements sit together more comfortably than the field usually allows.

10.9 The Gunter Study and What It Contributes

Work on shelter dogs and short-term fostering (Gunter et al., 2019) is cited in this area because it measured cortisol across a change in circumstances. It found what a recovery account would predict and was not designed to test accumulation.

That distinction matters: a study consistent with a model is not a test of it, and several of the citations attached to trigger stacking are of this kind, which is worth checking whenever the model is presented as supported.

10.10 What Owner Perception Adds

Work on how owners read canine stress signals (Mariti et al., 2012) is relevant here for a different reason. If households misread the signs, then reports of a dog being fine on Tuesday and not on Saturday may partly reflect what was noticed rather than what occurred (how owners can spot escalation early).

That does not dissolve the phenomenon and it does complicate the evidence, since most of the observations supporting the model are owner reports rather than coded behavior.

10.11 Individual Variation Runs Through All of It

Recovery speed differs substantially between dogs and relates to how they behave (Lensen et al., 2019). Any general figure for how long a dog needs is therefore a claim about an average that may describe few individuals.

That is the finding most worth carrying into practice, and it argues against exactly the kind of general number the model supplied.

10.12 What a Practitioner Loses by Dropping the Term

A shorthand that households understand immediately and that frames the dog sympathetically. Those are real losses and they explain why the term persists among people who know the evidence is thin.

The replacement has to do the same work: dogs take longer to recover than people expect, and stacking demands close together makes the next one harder. That is nearly as memorable and it commits to considerably less.

10.13 What a Practitioner Gains

The ability to be wrong, and therefore to learn. Where a plan built on spacing fails, a practitioner who has not invoked a bucket can look for another explanation instead of concluding the bucket was fuller than it looked.

They also gain the position of not having quoted a number that a client may later find has no source, which is a small thing until it happens.

10.14 The Honest Summary

The advice is good. The mechanism is plausible. The model is untested, its central figure has no origin, and its assumptions have never been examined individually in this species.

All four of those can be true at once, and holding all four together is the whole point of the article.

11. Summary at a Glance

• No canine study has tested it — The term has no defining paper, no operational definition in the research literature and no experimental test in dogs.

• The 24-to-72-hour figure has no traceable source — A behavior professional who had quoted it for years reported being unable to find evidence that it rests on scientific findings (Depta, 2020).

• Measured cortisol recovery can be much faster — Canine studies sample cortisol at 10 and 40 minutes post-stressor (Lensen et al., 2019), and shelter cortisol returned to baseline the following day (Gunter et al., 2019).

• Cortisol cannot distinguish distress from excitement — Which undermines a model that treats accumulated cortisol as what fills the bucket.

• Recovery speed varies and predicts behavior — Slow recovery at 40 minutes was associated with dog rivalry and stranger-directed fear; high early reactivity with desirable traits (Lensen et al., 2019).

• Chronic stress attenuates rather than amplifies — Five weeks of restriction raised daily cortisol but reduced the response to an acute noise stressor, in eight dogs (Beerda et al., 1999).

• Additivity and thresholds are assumptions — Neither has been tested in dogs.

• Continuing the session adds risk — Rising arousal makes retrieval of trained behavior less reliable, aversive handling in that state carries documented costs, and fear learning can generalize to contextual cues present at the time.

• The practical advice survives without the model — Spacing demands and allowing recovery follow from individual differences in recovery speed.

12. Research Gaps and Critical Appraisal

The central claim has never been tested. No study has exposed dogs to sequential stressors at controlled intervals and measured whether responses accumulate as the model predicts. This is a straightforward experiment that has not been run.

The samples behind the supporting evidence are small. Thirty-two puppies and sixteen young adults in the reactivity study (Lensen et al., 2019); eight dogs in the chronic restriction work (Beerda et al., 1999). The shelter fostering data are larger but address a different question.

Cortisol is a limited marker. It does not distinguish valence, varies with time of day and handling, and no single metric has been found sufficient to characterize stress in dogs or humans.

Practitioner sources cite studies that did not test the model. Beerda and others are routinely invoked as though they had validated stacking. Checking the cited work against the claim is a step almost nobody takes.

The model is not falsifiable as usually stated. Without an operational definition of the threshold or a specified set of contributors, no observation can count against it.

Positive stress counts too rests on an inference from a marker's limitation. That cortisol cannot distinguish excitement from distress is a fact about cortisol, not about the dog's experience.

The sensitization account is itself extrapolated. Sensitization and generalization are established learning processes, demonstrated largely outside the dog. Canine work documents sound sensitivity and its physiology without testing the sensitization trajectory — so the article's own alternative explanation carries a version of the weakness it identifies in the model.

Nobody has looked for the origin. Where the 24-to-72-hour figure entered training literature, and from what, is unknown — and tracing it would be a genuinely useful piece of work.

No canine study tests the model directly. The design is feasible, the observational version requires no manipulation, and neither has been reported as of the literature search underlying this article in September 2026. This is the claim in the article most likely to be overtaken by a new publication, and it is worth re-checking before citing.

The central number has no source. Following the citation trail ends without a study (Depta, 2020), and the figure remains in circulation.

The variables are undefined. What counts as a trigger, how intensity is scored and what interval separates events have no agreed operationalization, which is part of why the study has not been run.

Cortisol is a weak currency for the claim. It rises with positive arousal, the matrix determines the timescale, and behavioral thresholds are not obviously a function of it.

13. Conclusion

Trigger stacking is one of the most widely taught concepts in dog training despite never having been tested as a model in dogs. There is no defining paper, no operational definition, no measured threshold, and no experiment testing whether stressors accumulate the way the bucket diagram shows. Its best-known number — that cortisol takes 24 to 72 hours to return to baseline — appears to have no traceable source at all: a behavior professional who had taught it for years went looking for the evidence and could not find any, while the actual canine measurements sample recovery at 10 and 40 minutes and show shelter cortisol back at baseline the next day in the settings studied. What survives is narrower and worth keeping. Dogs mount measurable stress responses; recovery speed differs between individuals and relates to how they behave in daily life, with slow recovery associated with fear and rivalry; chronic load alters the system in ways more complicated than a fuller bucket, sometimes blunting the acute response rather than amplifying it. That is enough to justify everything the model is used to recommend — spacing demands, allowing recovery, reading a difficult day as context — without asserting a mechanism nobody has demonstrated. The distinction is not pedantic. A field that repeats an unsourced figure for two decades is a field that will repeat the next one, and the professionals most likely to be quoted are the ones who checked (which is the standard the rest of this literature is held to).

Key Insights (Takeaways)

• The concept has no canine research base. Trigger stacking has no defining study, no operational definition in the literature, and no experiment testing whether canine stress responses accumulate as described. It is a practitioner model that acquired scientific appearance through retrospective citation of studies that tested something else.

• The load-bearing number has no source. The claim that cortisol takes 24 to 72 hours to normalize underpins both stacking and the cortisol vacation. A behavior professional who had quoted it for years reported finding no evidence that it rests on scientific findings — and canine studies measure recovery at 10 and 40 minutes (Lensen et al., 2019), with shelter cortisol back at baseline the following day (Gunter et al., 2019).

• What is supported is recovery speed, not accumulation. High cortisol reactivity 10 minutes after a behavioral test was associated with desirable traits, while slow recovery at 40 minutes was associated with dog rivalry and stranger-directed fear (Lensen et al., 2019). Dogs differ in how fast they return to baseline, and that difference relates to behavior — which is a real finding, and a different one.

• Chronic stress complicates the bucket. Five weeks of social and spatial restriction raised daily cortisol but produced an attenuated response to an acute noise stressor (Beerda et al., 1999). Accumulated load does not simply mean a stronger reaction to the next event.

• The practical advice does not need the model. Spacing demanding events, allowing recovery and treating a difficult prior day as relevant context all follow from individual differences in recovery speed. Dropping the unsourced mechanism costs nothing and removes a claim that will otherwise be repeated by everyone who reads it.

References

Beerda, B., Schilder, M. B. H., van Hooff, J. A. R. A. M., de Vries, H. W., & Mol, J. A. (1998). Behavioural, saliva cortisol and heart rate responses to different types of stimuli in dogs. Applied Animal Behaviour Science, 58(3–4), 365–381. https://doi.org/10.1016/S0168-1591(97)00145-7

Beerda, B., Schilder, M. B. H., Bernadina, W., van Hooff, J. A. R. A. M., de Vries, H. W., & Mol, J. A. (1999). Chronic stress in dogs subjected to social and spatial restriction. II. Hormonal and immunological responses. Physiology & Behavior, 66(2), 243–254. https://doi.org/10.1016/S0031-9384(98)00290-X

Depta, L. (2020). Defining and refining the “cortisol vacation”. The IAABC Foundation Journal, 18. https://doi.org/10.55736/iaabcfj18.1

Gunter, L. M., Feuerbacher, E. N., Gilchrist, R. J., & Wynne, C. D. L. (2019). Evaluating the effects of a temporary fostering program on shelter dog welfare. PeerJ, 7, e6620. https://doi.org/10.7717/peerj.6620

Lensen, R. C. M. M., Moons, C. P. H., & Diederich, C. (2019). Physiological stress reactivity and recovery related to behavioral traits in dogs (Canis familiaris). PLoS ONE, 14(9), e0222581. https://doi.org/10.1371/journal.pone.0222581

Mariti, C., Gazzano, A., Moore, J. L., Baragli, P., Chelli, L., & Sighieri, C. (2012). Perception of dogs' stress by their owners. Journal of Veterinary Behavior, 7(4), 213–219. https://doi.org/10.1016/j.jveb.2011.09.004