Key Takeaways
- Hyperbolic tapering means each cut is a fraction of the dose at the time, so the cuts shrink in milligrams as the dose comes down.
- The curve that links dose to effect on the brain is steep at low doses, which is why the last milligrams of a benzodiazepine taper are the hardest.
- A fixed milligram cut becomes a growing share of a shrinking dose, so a linear schedule speeds up just where the nervous system can least absorb it.
- Professor Ashton’s own worked schedules already slow as they descend, and she wrote that there is no magic rate of withdrawal.
- The Maudsley Deprescribing Guidelines put a curve and a name to that pattern, and this site presents hyperbolic tapering as a development of the Manual, not a correction.
- A hold, meaning a decision to stay at the current dose until symptoms settle, is part of the method and not a setback.
- The crossover to diazepam, made one dose at a time, is the Manual’s way of making thin slices possible on a short-acting drug.
- Which percentage to cut and how long to hold are decisions for the person and their prescriber, and the calculators on this site are planning tools only.
Why must benzodiazepine dose cuts get smaller as the dose falls?
Hyperbolic tapering is a way of reducing a benzodiazepine in which each cut is a fraction of the dose at the time, so the cuts shrink in milligrams as the dose comes down. The reason is not caution for its own sake. It is the shape of the curve that links dose to effect on the brain, and that curve is not a straight line.
Physical dependence on a prescribed benzodiazepine is not addiction. It is an adaptation of the nervous system to a drug taken as directed, and the way out is a gradual reduction paced by the person taking it. The open question is what “gradual” should mean once the dose is already small.
The effect of a benzodiazepine rises with dose along a curve, not a straight line.
Benzodiazepines act on gamma-aminobutyric acid (GABA), the brain’s main calming transmitter, by binding to its receptor, the GABA-A receptor. The drug does not switch the receptor on by itself. It makes the receptor answer more strongly to the GABA that is already there.
The relationship between dose and receptor occupancy, meaning the share of receptors with drug attached, is hyperbolic. At low doses the curve is steep, and a small amount of drug occupies many receptors. At high doses it flattens, because most receptors are already taken and extra milligrams find fewer empty sites.
So, read backwards, from a high dose toward zero, the first cuts remove little receptor effect and the last cuts remove a great deal. This site does not pretend to know the percentages. The direction is settled, and it is why the last milligrams are the hardest.
Three pairs of sunglasses show why the last cut is the hardest.
Imagine someone wearing three pairs of sunglasses at once. The first pair blocks most of the glare, the second blocks most of what is left, and the third makes almost no difference. Taking off the third pair is barely noticed, and taking off the last lets in the whole sky at once.
Each pair was the same object, yet the effect of removing it depended on how many were still underneath. In fact, that is why a taper removes the bottom layer most slowly, in the thinnest slices.
A fixed milligram cut is a growing percentage of a shrinking dose.
The arithmetic makes the same point without any pharmacology. A cut of 0.5 mg from a daily dose of 4 mg is 12.5 percent of the dose. The same 0.5 mg cut from 2 mg is 25 percent, and from 1 mg it is 50 percent.
A linear schedule, meaning the same milligram cut at every step, speeds up in percentage terms just where the nervous system can least absorb it. Patients often describe such a taper as easy at first and then suddenly impossible. Yet, nothing about them changed. The schedule did.
The curve adds a second problem on top of the arithmetic. Near the bottom, where the curve is steep, the same half milligram takes away far more receptor effect than it did near the top, so the late cuts of a linear schedule are large twice over: as a share of the dose, and in what the brain feels.
What did Professor Ashton actually recommend?
Professor C. Heather Ashton of Newcastle University wrote the Ashton Manual, “Benzodiazepines: How They Work and How to Withdraw.” Chapter 2 gives her rule for the size of a step: “You could aim at reducing dosage by up to one tenth at each decrement.” Each new dose is held for a week or two before the next.
Her worked schedules already get smaller toward the end. Her illustration for a person taking 40 mg of diazepam a day reduces by 2 mg every week or two to 20 mg, then by 1 mg, and from 5 mg “some people prefer to reduce by 0.5mg every week or two.” That is a taper that slows as it descends, set down in the 2002 revision, and this site presents it as her illustration, not as a plan for any reader.
She was equally clear about pace. “The rate of withdrawal, as long as it is slow enough, is not critical,” she wrote, and there is “no magic rate of withdrawal.” The reader “must be in control and must proceed at the pace that is comfortable for you,” and her rule for setbacks is plain: “As far as possible, never go backwards,” which allows standing still at a dose for a few weeks but means trying “to avoid ever increasing the dosage again.” In this site’s view, a small increase agreed with your prescriber can also be reasonable when a step proves too hard, decided case by case.
Hyperbolic tapering is the later name for the shape her schedules already had.
The Maudsley Deprescribing Guidelines describe the dose-effect curve as hyperbolic and draw the practical conclusion from it. Reductions are made as a proportion of the current dose rather than of the starting dose, so that each step removes a similar amount of receptor effect.
Of course, nothing in this contradicts the Manual. Her fraction of the daily dose, her shrinking diazepam steps, and her insistence that the pace belongs to the patient are consistent with a proportional taper. The Guidelines put a curve and a name to the pattern she described from clinical experience, and that is how Ashton Manual Guide presents hyperbolic tapering: a development, not a correction.
The modern expression is short: each cut is a fixed percentage of the current dose, the new dose is held until symptoms settle, and the next cut is smaller in milligrams than the last. Which percentage, and how long a hold, are decisions for the person and their prescriber.
Why does the diazepam crossover help?
Cutting in thin slices is hard on a short-acting drug. Table 1 in Chapter 1 of the Manual gives alprazolam (Xanax) a half-life of 6 to 12 hours, lorazepam (Ativan) 10 to 20 hours, and temazepam (Restoril) 8 to 22 hours, against 20 to 100 hours for diazepam (Valium) and up to 200 hours for its active metabolite. A drug that wears off between doses adds interdose withdrawal on top of the taper.
The Manual’s answer is the crossover taper, a switch to diazepam made one dose at a time. Chapter 2 describes replacing the evening or night-time dose first, then the others one by one at intervals of a few days or a week. The Manual notes that the 2 mg diazepam tablet is scored and easily halved and that liquid preparations of some benzodiazepines exist, which is what makes the small late cuts possible.
Table 1 is also the reference behind the diazepam conversions in the Ashton Manual Guide calculator. Diazepam 10 mg is approximately equivalent to alprazolam 0.5 mg, lorazepam 1 mg, clonazepam (Klonopin) 0.5 mg, and temazepam 20 mg. The table’s own note says these figures rest on clinical experience and may vary between people.
Holds are part of the method, not a failure of it.
A hold is a decision to stay at the current dose until symptoms settle. The Manual treats standing still as ordinary, and the proportional method depends on it. Unfortunately, many patients have learned from a clinic calendar that a pause is a setback. It is not.
The doctor’s part is the brake. If a cut proves too large, the doctor should advise a smaller one or a longer hold, and if a stretch goes easily, patient and doctor may agree to move a little faster.
The calculators on this site draw this curve.
The calculators on Ashton Manual Guide are built around this shape: a percentage of the current dose per step, a hold between steps, and cuts that shrink toward the end. The benzodiazepine taper calculator draws a proportional schedule from any starting dose, the drug-specific calculators do the same for one drug at a time, and the Ashton Manual Guide calculator lays out a crossover to diazepam using the Table 1 equivalences. They are educational planning tools, not a substitute for a physician who knows the person.
Fortunately, the stretch where the steps become fractions of a milligram is also where a liquid preparation or an accurately compounded dose earns its place.
Questions readers ask about hyperbolic tapering.
Why can’t I reduce by the same amount each time?
Because the same milligram cut is a larger share of the dose each time, and near the bottom of the curve it also removes more receptor effect than the same cut did higher up. A 0.5 mg cut that was one eighth of the dose at the start becomes half of it near the end. Linear schedules tend to fail in their final stages for this reason, and the failure is arithmetic, not weakness.
Can I switch to a different benzodiazepine to make tapering easier?
Often, yes. The Ashton Manual’s method is the crossover taper to diazepam, made one dose at a time using the Table 1 equivalences. Still, some people already taking a long-acting drug such as clonazepam taper it directly, and whether to switch is a decision to make with a prescriber.
What is the difference between feeling uncomfortable and being in danger?
Discomfort is expected during a taper: some anxiety, poorer sleep, muscle tension, and mood changes that ease as a hold goes on. Danger is different in kind. A seizure, severe confusion, hallucinations, or thoughts of suicide are medical emergencies, so call 911 for a seizure or any medical emergency, and in the United States call or text 988, the Suicide & Crisis Lifeline, for thoughts of suicide. The September 2020 update to the Food and Drug Administration (United States Food and Drug Administration (FDA)) boxed warning for benzodiazepines states that stopping abruptly or reducing the dose too quickly can cause withdrawal reactions, including seizures.
How long does a hyperbolic taper take?
As long as it needs to. Professor Ashton wrote that “whether it takes 6 months, 12 months or 18 months is of little significance if you have taken benzodiazepines for a matter of years,” and the last part of a taper, with its smaller steps and longer holds, often takes longer than it looks on paper.
What percentage should each cut be in a hyperbolic taper?
There is no single figure. The Manual’s rule of thumb is a reduction of up to one tenth of the dose at each step, with each new dose held for a week or two. The Maudsley Deprescribing Guidelines express the same idea as a fixed proportion of the current dose, so that each step removes a similar amount of receptor effect.
Which fraction suits a given person, and how long each hold runs, is for the person and their prescriber to decide. Symptoms, not the calendar, decide when the next step comes.
Does a hyperbolic taper need a liquid?
Not at the start. Higher up, tablets do the work, and the Manual notes that the 2 mg diazepam tablet is scored and easily halved. Lower down, where the steps become fractions of a milligram, a liquid preparation or an accurately compounded dose earns its place, because a tablet is never split beyond a quarter and a quarter of the smallest tablet is still a large step at the bottom of the curve.
Fortunately, diazepam is available as a manufactured oral solution of 1 mg per mL, and the Manual describes reducing a liquid by decreasing the volume of each dose with a graduated syringe. Doses of a liquid are measured with an oral syringe, never a dropper.
The shape of a good taper was on the page in 2002, in a schedule that stepped down by 2 mg, then 1 mg, then 0.5 mg. Hyperbolic tapering gives that shape a curve and a name. The person taking the tablets still sets the pace, and their doctor still holds the brake.
This article is educational. It is not medical advice, and reading it does not create a doctor-patient relationship. Decisions about starting, continuing, or tapering any medication should be made with your own physician.
