Drug tolerance
A Socratic walk-through of drug tolerance — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does a drug that worked well at first need a bigger dose later to do the same job?
The drug has not changed. The molecule that produced a strong effect in week one is chemically identical in week ten, and it still binds the same receptor. Yet the effect has faded, and restoring it takes more. Something in the person has moved. The obvious guess is that the body is "getting used to it" — but that phrase hides at least three different processes, running at different speeds, and one of them turns out not to be in the body's chemistry at all.
Reasoning it through
REASONING #Start with the simplest possibility. A drug works by occupying receptors and pushing a signalling pathway. What would a cell facing a persistently over-stimulated pathway do? The same thing any regulated system does when a signal is stuck high: turn down its own gain. Receptors are chemically modified so they couple less efficiently downstream; they are pulled off the surface into the cell's interior; over longer exposures the cell makes fewer of them. This is pharmacodynamic tolerance, and with opioids it goes further — the cell also upregulates the pathway the drug suppresses, so the opposing force grows while the drug's grip weakens.
Now ask what that implies for stopping. If the cell has built compensation against a drug and the drug is withdrawn, what is left? The compensation, unopposed. That is withdrawal: the same adaptation seen from the other side. Tolerance and withdrawal are not two phenomena but one, read forwards and backwards.
A second route lives in the liver. Many drugs induce the enzymes that metabolise them, so identical doses produce lower blood concentrations over time. Alcohol does this; several anticonvulsants do it dramatically. Note the difference in kind: nothing in the brain has adapted, the brain is simply seeing less drug.
Then the finding that makes tolerance genuinely strange. If tolerance were only receptors and enzymes it would travel with the person — the same dose should behave the same way anywhere. Shepard Siegel's experiments showed it does not. Rats made tolerant to morphine in one environment, then given a large dose in an unfamiliar one, died at substantially higher rates than tolerant rats dosed in the familiar setting. The receptors were identical. The room was not.
Why would a room matter? Because the compensation can be triggered by prediction rather than by the drug. Cues that reliably precede a dose — the place, the ritual, the paraphernalia — come to elicit the body's opposing response in advance, so by the time the drug lands the correction is already running. Take the drug where those cues are absent and it does not fire, and the routine dose becomes an overdose. This is Pavlovian conditioning, not pharmacology. It predicts that an experienced user is most vulnerable not when using more but when using the usual amount somewhere unfamiliar — though the animal work is the direct evidence, and the human inference, while supported, rests on observational data that rarely excludes other explanations.
One further wrinkle: tolerance is not uniform across a single drug's effects. With opioids, analgesia and euphoria fade considerably faster than the suppression of breathing or of gut motility. Escalating to recover pain relief therefore walks the dose toward a respiratory effect that has hardly adapted at all. The gap between the useful dose and the dangerous one narrows as tolerance grows — the opposite of the intuition that a tolerant person is a safer one.
The analogy
THE ANALOGY #Picture a cold room with a heater carried into it, and a caretaker whose job is to hold the temperature steady. He responds by opening vents. Keep the heater there for months and the open vents become the room's normal state — the temperature reads ordinary, so you bring in a second heater. Remove both and the room is freezing, which is withdrawal. And the caretaker learns: he hears your key in the lock and opens the vents before you enter. Come in through a door he cannot hear, and the same heater cooks the room.
The caretaker is one mechanism, whereas real tolerance is several independent ones — fewer receptors, faster liver clearance, learned anticipation — which build at different rates, fade at different rates, and crucially do not compensate for all of a drug's effects equally.
Clarifying the model
THE MODEL #Three words get used as synonyms and are not. Tolerance is a diminished effect from the same dose. Physical dependence is the adaptation that produces withdrawal on stopping. Addiction — clinically, substance use disorder — is compulsive use despite harm, involving craving and loss of control over the behaviour.
The distinctions are not academic. A patient on long-term opioids for cancer pain, or on a beta blocker, or on an antidepressant, is very likely tolerant and dependent and in no sense addicted; abruptly stopping any of them causes withdrawal in someone who never craved the drug. Conversely, addiction is not advanced tolerance — it involves motivational circuitry that tolerance alone does not touch. Collapsing the three is what leads to a person in pain being treated as a person with an addiction because their dose went up.
Two honest caveats. Tolerance is neither universal nor unidirectional: some drug effects show sensitisation with repetition, and to some no tolerance develops at all. And tolerance decays. After abstinence the adaptations reverse, often faster than people expect, so the dose that was routine before a hospital stay, a detox or a prison sentence is no longer routine on release. That reversal, together with the uneven tolerance across effects, is why the highest-risk moment is often a return to a previously ordinary dose.
A picture of it
THE PICTURE #How to readStart at the rounded terminal at the top and follow the three parallel arms — receptor adaptation, liver clearance, and learned cues — which are three independent sources of the same fading effect. The first diamond asks whether the dose still works; "yes" loops back through the green outcome to another round of dosing, which is the feedback that drives escalation. The two lower diamonds are where the danger lives: the red boxes are reached not by taking more than usual but by taking the usual amount without the cues, or after a gap long enough for the adaptations to unwind.
What became clearer
WHAT CLEARED #Tolerance is not the drug wearing out but the body building a counterweight — partly in the receptor, partly in the liver, and partly in learned anticipation that is triggered by the surroundings rather than by the chemistry. Because that counterweight can be cued, it can also be absent, which is why a familiar dose in an unfamiliar place is a different dose in practice. And because it develops unevenly across a drug's effects, growing tolerance narrows the margin of safety rather than widening it.
Where to go next
ONWARD #- Why some drug effects sensitise with repetition instead of tolerating.
- How opioid rotation exploits incomplete cross-tolerance between drugs at the same receptor.
Key terms
TERMS #| Term | What it means |
|---|---|
| Pharmacodynamic tolerance | reduced response at the target, through receptor desensitisation, internalisation or downstream counter-adaptation. |
| Pharmacokinetic tolerance | reduced drug exposure through faster metabolism, typically by induced liver enzymes. |
| Conditioned tolerance | a compensatory response elicited in advance by cues that reliably precede the drug. |
| Physical dependence | the adapted state whose unmasking on stopping produces a withdrawal syndrome. |
Every term the collection defines is gathered in the glossary.