Benzodiazepine withdrawal is usually discussed as what happens after the taper — the symptoms that emerge as the dose comes down. But a substantial number of long-term benzodiazepine patients are already in withdrawal while they are still taking their medication at the prescribed dose. This is tolerance withdrawal, and recognizing it matters clinically because it changes what the taper is actually doing.
The phenomenon is documented in the Ashton Manual and elsewhere in the benzodiazepine literature as the state in which the original therapeutic effect of the medication has been lost while the patient remains on it. The anxiety or insomnia that the benzodiazepine was originally prescribed for returns, often with new features, and the patient typically concludes — along with the prescriber — that the original condition has worsened and needs a higher dose. It has not necessarily worsened. The receptors have adapted.
How Tolerance Develops
GABA-A receptors are the inhibitory infrastructure of the central nervous system. Benzodiazepines act as positive allosteric modulators at these receptors, increasing the frequency of chloride channel opening in response to endogenous GABA. With prolonged exposure, the receptor complex adapts: subunit composition shifts, receptor density decreases, and the magnitude of response to a given benzodiazepine dose falls. The clinical result is that the dose that produced the original anxiolytic or hypnotic effect stops producing it. The patient is now receiving the drug for a nervous system that has reshaped itself around it.
Recognizable Features
Tolerance withdrawal looks different from a straightforward relapse of the original condition, though the distinction is frequently missed in a brief clinic visit. Several features are characteristic.
Symptoms that were not present before the benzodiazepine was started. Patients often report new physical phenomena — perioral numbness, paresthesias, muscle twitching, inner tremor, photophobia, hyperacusis, gastrointestinal symptoms, tinnitus — that were absent at baseline and do not fit the original diagnosis.
Interdose symptoms with short-acting agents. Patients on alprazolam or lorazepam frequently describe a time-locked pattern: symptoms build through the interdose interval and are briefly relieved by the next dose. This is indistinguishable in form from a short-acting withdrawal course and essentially confirms tolerance.
Failure of dose increases to hold. A dose increase produces transient improvement, often only for a few weeks, before symptoms return at the new dose. The ladder keeps climbing without producing sustained relief.
Morning worsening. Overnight metabolism of the drug can produce a morning low point with agitation, anxiety, or physical symptoms that resolve an hour or two after the morning dose. The pattern is again time-locked to pharmacokinetics, not to the patient’s day.
Sensory sensitivity and autonomic features. Heightened startle, light or sound sensitivity, temperature dysregulation, and episodes of tachycardia are common and typically unrelated to the patient’s original presenting complaint.
Why This Is Usually Missed
The clinical problem is structural. A patient who presents with worsening anxiety on a long-term benzodiazepine prescription typically receives a dose increase or a second psychotropic. Neither intervention addresses tolerance. The dose increase buys a few weeks; the second agent (often an SSRI or a neuroleptic prescribed for sleep) adds new pharmacologic variables without resolving the underlying adaptation.
Two features of the standard office visit contribute. First, most prescribers have limited exposure to the tolerance literature; the framework is not a default diagnostic consideration for new anxiety complaints. Second, the symptom pattern overlaps genuinely with primary anxiety disorders, and a time-limited visit cannot easily distinguish a patient in tolerance withdrawal from a patient whose original condition is worsening. The prescription environment therefore drifts toward dose escalation and polypharmacy.
Clinical Implications for the Taper
Tolerance withdrawal changes what a taper means. A patient who is already in withdrawal on their current dose is not going to experience the first several reductions as introducing new symptoms; they are going to experience them against a baseline that already includes withdrawal features. This has several practical consequences.
The first is that a very slow initial pace is often better tolerated than a standard one. Patients in tolerance withdrawal frequently stabilize partially during the first weeks of a taper — paradoxical as that sounds — because stable receptor conditions begin to reassert themselves once the dose stops being a moving target.
The second is that interdose stabilization sometimes matters more than dose reduction, especially for short-acting agents. Switching a patient from alprazolam to an equivalent dose of a longer-acting benzodiazepine such as diazepam (the Ashton substitution) can reduce interdose symptoms and make subsequent dose reductions more tolerable. This is not always the right move — the Maudsley Deprescribing Guidelines (Horowitz and Taylor, 2024) refine the Ashton substitution and treat it as one tool rather than a universal step — but it is worth considering early.
The third is that treating tolerance withdrawal as if it were an indication for dose increase is actively harmful. The reflex to add dose, or to add another medication, when a patient on a long-term benzodiazepine develops new symptoms is, in this population, the step that produces the most downstream harm. Recognizing tolerance as a possibility changes the conversation from “how much more do we add” to “how do we unwind this safely.”
What Patients Can Bring to a Visit
For patients who suspect they are in tolerance, several things help at the prescriber visit.
A symptom timeline: when each symptom first appeared, in relation to starting or increasing the benzodiazepine or to dose timing within the day. Handwritten is fine; dated is essential.
A dose history: drug, dose, frequency, and any escalations over time.
The specific concern framed as tolerance, not relapse. “These symptoms are new and time-locked to dosing” is a different clinical question from “my anxiety is worse.”
A willingness to discuss a taper. Many prescribers will not initiate the tolerance conversation, but most will engage with it if the patient raises the possibility directly.
Tolerance withdrawal is not a rare presentation. It is common, under-recognized, and the most reliable predictor of the patients who arrive later in protracted withdrawal with symptoms that have been accumulating for years. Identifying it early is the single intervention most likely to change the trajectory.