After three years on sertraline, Sarah finally felt steady, so together with her GP she thought it was time to come off it. The plan is the one most people follow when they decide to stop, which is halving the dose for two weeks, then stopping completely. However, day four brought unusual dizziness and nausea that made Sarah wonder if she had stopped too soon and had been ready for it. She worried the symptoms meant she was relapsing, even though that was not necessarily the case.
What Withdrawal Can Feel Like
The truth is, a dose cut that looks ridiculously small on paper can be a lot for your brain to handle. The standard 2-week halving schedule is a very generalized way of tapering and might not be suitable for everyone. So many factors are not considered: your unique metabolism, generics, how long you have been under the treatment, your emotional state right now, and so on.
In this guide, we explain what really happens in your brain if and when you reduce your dose, what the current clinical guidelines recommend, and how to build a proper taper plan you and your prescriber can work through together.
Why The Standard Tapering Advice Often Falls Short
Most tapering advice is linear: halve the dose, wait two weeks, halve it again, stop. Feels simple, right? It is not complex to explain or remember, and surprisingly well matches the way pills are manufactured (e.g. 100 mg, 50 mg, 25 mg, and so on). Unfortunately, the brain does not respond in the same way.
You did not notice changes after the first pills while your brain adapted to pharmacological effects caused by prescribed antidepressants. Stopping the use of antidepressants requires brain adaptation too. The brain simply has to change how many receptors should be on the cell surface and how many neurotransmitters should be produced. The precise mechanisms driving withdrawal remain a subject of active scientific inquiry. A prominent theory suggests that when the medication is discontinued, the brain’s recycling proteins resume clearing neurotransmitters before the system has sufficiently adapted its own production levels. This resulting discrepancy is thought to trigger the symptoms of withdrawal. While this physiological explanation is widely supported in clinical literature, it continues to be refined through ongoing research.
In 2024, researchers pooled 79 trials involving more than 21,000 people to work out how often withdrawal actually happens. About 31% of those who came off an antidepressant had some symptoms, but so did 17% of people who stopped taking a placebo, a dummy pill with no active medicine in it. Once that placebo response is accounted for, roughly one person in six or seven (about 15%) has symptoms that can be put down to stopping the medicine itself, and severe symptoms affecting around one person in 35. A previous systematic review used different criteria and a different set of literature and indicated a far higher average incidence of 56%, varying between 27% and 86% in various studies. More importantly, those patients who do experience withdrawal often find it to be severe and prolonged, frequently mistaking it for a relapse.
But that is only part of the story. The other problem with a one-size-fits-all schedule is that no two people handle medicine in exactly the same way. Your liver breaks the drug down using enzymes, and the genes you inherited (called CYP2D6 and CYP2C19) decide how well those enzymes work. That is what sets how quickly the medicine leaves your body. Some of us are naturally poor metabolizers and can hold onto a drug for longer periods of time. Others can metabolize it unusually fast and will feel the dose drop within 24 hours.
It implies that if two individuals are on the same dosage of the same drug, under the same conditions and over the same period of time, their experiences will still be different in real life. The guidelines are there because they are endorsed by the medical community and for a good reason. They confirm the effectiveness and, most importantly, the safety of the medication in millions of patients. Yet sometimes, they need to be tailored a little more to the individual.
Why Those Last Milligrams Hit Hardest
SSRIs work by blocking a protein called the serotonin transporter, or SERT, which mops serotonin back up out of the gap between two nerve cells. Serotonin is the clearest part of the story, but not the only one. Some antidepressants also act on a second messenger chemical called noradrenaline, and for those drugs serotonin alone does not explain withdrawal. Normally SERT pulls serotonin back into the cell to be used again. Building serotonin from scratch is slow and takes a lot of energy so SERT allows the cell to reuse serotonin that has already been produced, speeding up the cell response.
When SSRIs block more of SERT, more serotonin stays available for stimulating receptors on the receiving cell. It prolongs the cell response but also allows it to have a response if the cell can not release enough serotonin.
Regarding the topic of the usage and tapering of the antidepressants, SERT occupancy is one of the main criteria that should be taken into account. It reflects the effect of a particular dose on SERT blocking.
The link between the dose you take and how much SERT gets blocked is not a straight line but a curve. It climbs steeply at low doses, then flattens out. Brain-scan (PET) studies show that an ordinary treatment dose already blocks around 80% of these transporters, which puts you well up on the flat part of the curve. From that point, taking more barely adds anything and, just as importantly, a small cut from the top of the curve barely takes anything away either. That is why early reductions often feel like nothing at all.
Let us now look at the numbers on the other end of the curve. A reduction from 200 mg of sertraline to 100 mg, which is a 100 mg reduction, does not significantly affect the SERT occupancy, as both doses are in the region of the flat line in the graph. However, what if we make a 5 mg reduction from 5 mg to 0 mg? Yes, it is just a 5 mg reduction, but it results in SERT occupancy dropping from roughly 40% to zero. A small dose reduction leads to a significant biological effect.
This is exactly the reason why so many people say that tapering feels straightforward for several months and then suddenly much harder. No, people do not get any weaker and tapering does not go wrong in any way. They just reach the steep slope of the curve, where each remaining milligram carries far more weight than the ones before it. A linear schedule accelerates exactly when it should be slowing down.
Hyperbolic tapering is different. It determines how much biological effect each dose reduction will cause and maintains that effect at a roughly constant level. This implies that each cut in the dose becomes increasingly small. The initial dose reduction can be 20 mg. But one of the last ones might be just 0.4 mg.
This approach avoids the difficulty of the final stages. Each step of the protocol is sized to remove a similar slice of SERT occupancy rather than a similar number of milligrams, so the factual amount shrinks as you descend.
How Modern Clinical Guidelines Approach Tapering
The Maudsley Deprescribing Guidelines (2024) provide the most extensive information on the concept of hyperbolic tapering. These guidelines were authored by Dr. Mark Horowitz and Professor David Taylor. One can call these guidelines the protocol of hyperbolic tapering.
Their thinking first appeared in a 2019 Lancet Psychiatry paper, which set out the pharmacological reasoning behind the approach. What the 2024 edition adds is practicality: a drug-by-drug manual covering how far to cut, how long to hold before the next reduction, and which formulations make very small doses possible.
| Source | Year | What it says on tapering |
|---|---|---|
| Maudsley Deprescribing Guidelines | 2024 | Per-drug hyperbolic schedules; explicit guidance to slow further for people with severe or longstanding withdrawal reactions. |
| NICE NG222 (UK) | 2022 | Reduce in steps, prescribing a proportion of the previous dose at each step; taper more slowly for short half-life drugs; use liquid preparations when tablets can no longer go low enough. |
| Royal College of Psychiatrists | 2020 / 2024 | Acknowledges withdrawal can be severe and long-lasting; recommends slow, individualised reductions led by the patient’s response. |
| Dutch multidisciplinary document | 2018, rev. 2023 | A first national-level multidisciplinary document to endorse hyperbolic tapering and the use of standardised taper strips. |
What major clinical sources say about tapering antidepressants.
In a nutshell, NICE recommends reducing antidepressants by a percentage of the current dose, not by a fixed milligram amount, as well as waiting for withdrawal symptoms to settle before each reduction. This is a similar principle to hyperbolic tapering, just expressed in the language of a clinical guideline.
Unfortunately, the US has not yet adopted equivalent guidance and has no comparable FDA-endorsed national guideline. This gap has nothing to do with a lack of evidence, as the PET receptor occupancy studies underpinning hyperbolic tapering were conducted across the US, Canada, Japan, and Sweden.
In real life, many people who experience severe withdrawal have never heard the term “hyperbolic tapering” and only discover the concept after their taper has gone wrong.
Some Drugs Are Harder to Stop
The speed of removing drugs from the bloodstream is called half-life. It is quite self-explanatory and means the time needed for the blood levels of a drug to fall by half. A drug with a short half-life leaves your body quickly. This means the drug levels swing sharply between doses, drop steeply after a reduction, and leave your brain with very little time to adjust. A long half-life, however, feels softer as it effectively tapers itself a little between doses.
| Medication | Half-life | Reported withdrawal risk | Typical starting preset |
|---|---|---|---|
| Paroxetine (Paxil, Seroxat) | ~21 hours, no active metabolite | High | C — slower |
| Venlafaxine (Effexor) | ~5 h; metabolite ~11 h | High | C — slower |
| Duloxetine (Cymbalta) | ~12 hours | High | C — slower |
| Sertraline (Zoloft, Lustral) | ~26 h; metabolite longer | Moderate | B — moderate |
| Escitalopram (Lexapro, Cipralex) | ~27–32 hours | Moderate | B — moderate |
| Citalopram (Celexa, Cipramil) | ~35 hours | Moderate | B — moderate |
| Fluoxetine (Prozac) | 4–6 days; metabolite 4–16 days | Lower | A or B |
Generalised patterns only — not a prediction of your individual case. Presets refer to Claro’s starting speeds (A faster, B moderate, C slower).
Withdrawal risks described here only reflect generalized patterns that have been previously reported across the literature and in clinical practice. Do not take it as an actual prediction of your individualised case. Some people taper off paroxetine smoothly but will struggle with fluoxetine. Your own story is unique and is far more important to consider than the actual position of the drug in this table.
Several other factors feed into the decision. Fluoxetine, for example, stays in the body much longer than most antidepressants, which is why it is sometimes used as a “bridge”. A prescriber may move you from a fast-clearing antidepressant onto fluoxetine before the taper begins, so that drug levels fall away gradually instead of dipping sharply between doses. The trade-off is that fluoxetine interacts with a wide range of other medicines. A bridging switch is therefore a decision for your prescriber to make and monitor, never one to try on your own.
A possible limitation is the extended-release tablets or capsules. ER, XR, and CR drugs are engineered to have the medicine released gradually from the body, and when they are split, this mechanism is dramatically changed. In some cases, it may cause a sudden spike in the medication dose. For these types of drugs, any minor decrease of the dose would mean a whole new formulation, not just splitting the pill.
If you are interested in drug-specific schedules, formulations and real-life cases, see our dedicated guides on how to taper off Zoloft (sertraline), Lexapro (escitalopram), Effexor (venlafaxine), and Cymbalta (duloxetine).
Tips on How To Choose Your Taper Speed
A universal timeline does not exist. However, there is an obvious choice that you can modify in the future according to your reaction. You can try the Claro calculator, which offers three options. This calculator measures the reduction in occupancy that a single step would take in your case, as well as the time you will have to wait for the next one.
| Preset | Occupancy drop per step | Minimum hold | Typically suits |
|---|---|---|---|
| A — Faster | ~5 percentage points | 1 week | Short duration of use, no prior withdrawal difficulty, longer half-life drug |
| B — Moderate | ~2.5 percentage points | 2 weeks | The default starting point for most people and most drugs |
| C — Slower | ~1.5 percentage points | 4 weeks | Long-term use, previous failed attempts, or a short half-life drug |
Claro’s three starting speeds. Occupancy drop is the modelled SERT occupancy reduction per step.
Claro’s three presets all fall within the recommended per-drug ranges of the Maudsley Deprescribing Guidelines, NICE NG222, and the Royal College of Psychiatrists.
What May Influence Your Pace
- Duration of use. Years tend to require a slower tapering process compared to months.
- Your current dose. Starting from a high dose means going through more steps, and the early ones usually do feel the easiest.
- Whether you have attempted a taper before or if your taper has failed in the past. This may require a slower approach instead.
- The exact medication that you are on.
- How you metabolize the drug. If you have experienced unexpected sensitivity to dosage reduction/increase before, this is something to consider.
- Your current psychological state.
- What resources (prescriber and pharmacist support, liquid formulations) you have available that may shift the pace that will suit you.
The pre-taper risk quiz walks you through six of these factors and suggests a starting preset. Feel free to override it at any point. It is there to help you start the conversation.
How to use the output
The calculator produces an educational schedule, not a prescription. It is designed to give you and your prescriber a document to work with. Every dose change still goes through your prescriber.
How Can You Handle the Last Milligrams
Towards the bottom of the dose range, hyperbolic tapering calls for reductions that ordinary tablets simply cannot deliver. Your final few steps might be 3.2 mg, then 2.4 mg, then 1.8 mg. However, if the smallest tablet you can get hold of is 25 mg, none of those doses is reachable. There is only so far a tablet can be split, and this is the point at which many tapers come unstuck. The options below are worth knowing about well before you need them, as some take time to arrange.
Getting Those Small Doses
- Some antidepressants have licensed liquid formulations. For instance, Rosemont products in the UK, and Cipramil and Cipralex oral drops in some European countries. This is a great and straightforward option if it is available in your country.
- It is safe to split the tablet into two or four for the first few lower steps using a scored and immediate-release tablet. But its accuracy falls off drastically thereafter and cannot achieve the low doses that later steps call for.
- A compounding pharmacy can prepare a compounded liquid in a particular concentration. Availability and price range can vary quite a bit depending on the country and city.
- Taper strips are pre-packaged daily doses in a descending sequence that were pioneered by Regenboog Apotheek in the Netherlands and are referenced in the Dutch national guideline. They do solve the problem entirely, but unfortunately are not widely available outside the Netherlands.
- Extended-release products. ER, XR, and CR formulations must not be split. Reaching small doses on these normally means switching formulation first, but your prescriber should make this decision.
A challenge that one may have is that there are usually two specialists involved. The prescriber determines how the taper should be done, while the pharmacist is aware of the medication that can be obtained in the country and in which dosages. It is recommended to contact your pharmacist well ahead of tapering toward the last milligram, since it gives them time to prepare any special orders you might need and saves a lot of last-minute stress.
There is no problem at all in considering the final phase to take as much time as the entire tapering phase taken before it. To spend several months going down from 5mg to zero does not amount to a stuck taper. Concerning the shape of the occupancy curve, this is actually where it counts.
Withdrawal, Relapse, and How to Respond
So what happens if you feel awful three weeks into a dose reduction? Is this withdrawal? Is it a relapse of the depression? These are important questions to ask that directly influence the outcome of the taper and, unfortunately, they are also the ones most guides omit.
If you are dealing with a withdrawal, the answer is to hold or step back up and move again at a slower pace. If it is relapse, the answer may be to stay longer on your prescribed treatment. But doing it the other way around can cause real harm, making some restart medication they no longer need, or push through a withdrawal that is unlikely to go away on its own.
| Factor | More typical of withdrawal | More typical of relapse |
|---|---|---|
| Timing | Within days of a dose reduction — often 1–4 days | Usually weeks to months later, not tied to a specific step |
| First symptoms | Physical: dizziness, nausea, brain zaps, headache, vivid dreams | Mood and cognition: low mood, loss of interest, hopelessness |
| Character | Symptoms often new — things you never had before treatment | Recognisable as your original illness returning |
| Pattern | Wave-like: peaks and eases, then peaks again | Steadier, gradually deepening |
| Response to reinstating | Improves quickly, often within days | Little immediate change; takes weeks if it responds |
Typical differences between withdrawal and relapse.
The Cosci & Chouinard model, set out in the table above, identifies three distinct withdrawal phenomena. Importantly, none of them is relapse. These are new withdrawal symptoms, rebound (the original symptoms returning, often more intensely), and persistent post-withdrawal disorder.
Pause and reinstate
This is perhaps the most overlooked option; holding at a reduced dose for longer. If a reduction brings on symptoms you cannot tolerate, returning to the last dose you were settled on and then attempting a smaller step is a clinically recognised approach. Understandably, many people experience this as going backwards. It is worth remembering, though, that milligrams are not the measure of success. A taper that pauses at 5 mg for three months and then finishes has succeeded. One that is pushed through at speed and ends in a return to the full dose has not.
Windows and waves
Most people expect recovery from withdrawal to be a steady climb. In practice, it rarely follows that pattern. Instead, spells of feeling well, often called “windows”, alternate with spells when symptoms return, known as “waves”. Over time the windows tend to lengthen and the waves tend to soften. Knowing that this pattern is normal makes the waves considerably less frightening when they arrive.
A minority of people experience protracted withdrawal, in which symptoms persist well beyond a few weeks. Case series and survey reports describe symptoms lasting several months. This is uncommon, but it does happen. If your symptoms are severe, if they persist for a long time, or if you have any thoughts of harming yourself, contact your doctor straight away. Do not wait for your next scheduled appointment.
Our symptom checker can help you see whether what you are experiencing is a commonly reported experience during withdrawal. It does not provide a diagnosis but it will definitely help you better understand the typical from the unusual. Telling withdrawal apart from relapse in any individual case requires a clinical assessment, not a self-assessment.
Talking to Your Prescriber About Tapering
It is very common for people to put off having this discussion. The concern is that bringing up this topic would make one come across as potentially non-compliant or questioning their doctor’s advice.
In most cases, physicians are quite open-minded about patients wanting to lower the dosage and are open to such ideas. The problem is mostly due to the lack of time and information rather than anything else. When you say, “I have been doing okay lately; maybe I can stop taking my medication,” then your doctor has to make an important decision based on limited information within just 10 minutes.
What changes the conversation is evidence. Having a symptom timeline covering at least the past two or three months, validated scores such as the PHQ-9 and GAD-7, which say far more as a trend than as a single snapshot, a record of side effects, your history with this medication, and a specific tapering plan rather than a vague request. The Claro app is built around exactly this. It is to turn your daily check-ins into a report you can hand to your prescriber at the appointment.
It is also worth preparing for the possibility that the conversation does not go the way you hoped. If your prescriber is unfamiliar with hyperbolic tapering, that is not obstruction on their part but it is simply a relatively new approach. Referring to the Maudsley or NICE guidance at your appointment gives the discussion a concrete reference point, and you are entitled to seek a second opinion if you still feel unheard.
