Should I still be on this? 10 questions for your antidepressant review

About one in four Americans taking an antidepressant this month has been taking one for ten years or more [2]. The trials those drugs were approved on ran a median of about eight weeks [3]. Between those two numbers sits a conversation — are we still doing this on purpose? — that the system is not built to schedule: prescriptions renew in seconds, and no appointment exists whose stated job is asking whether the medication should continue.

This page is the checklist for scheduling it yourself. One thing before the list, stated as plainly as we can: this question belongs to you and your prescriber, and the most dangerous possible response to this page is stopping on your own. The evidence below cuts in both directions, and we've kept both directions in.

This is information to bring to your prescriber, not medical advice and not a reason to change anything on your own.

The checklist

1. Why did I start, and does that reason still exist?
2. What has the drug done for me lately — and how would we know?
3. What am I trading for the benefit?
4. What would staying on another ten years mean?
5. What are my odds if I stop — and if I don't?
6. If we ever taper, what would the plan actually look like?
7. How would we tell withdrawal from relapse?
8. Has my life changed in ways that change the answer?
9. What's the smallest change we could try first?
10. When do we ask all this again?

1. Why did I start — and does that reason still exist?

The prescription may be older than your current job, relationship, or city. Depression after a divorce, anxiety in a workplace you left in 2019 — the indication was real; the question is whether it's current. This is the anchor question, and everything else on the list hangs off your prescriber's answer to it.

2. What has it done for me lately — and how would we know?

Feeling fine is genuinely ambiguous: it can mean the drug is working, or that you no longer need it. No test distinguishes the two. What a good review does instead is look at your relapse history (how many episodes, how severe, how recent), which is the strongest predictor prescribers actually have. Ask directly: "knowing my history, what do you think this is still doing?"

3. What am I trading?

Long-term use has costs that creep quietly enough to stop being noticed — sexual side effects, emotional blunting, weight change — each drug's tested rates and real-world reports are on its page in the medication library, caveats first. The point isn't that the trade is bad; it's that a trade renewed automatically for a decade deserves to be re-read once a year.

4. What would another ten years mean?

Here is the honest state of the evidence: the approval trials ran weeks [3], and what we know about year ten comes from observational studies — a patchier, more confounded record in both directions. We plotted every drug's longest approval trial next to how many people take it. Your prescriber can't give you certainty here, because it doesn't exist; what they can give you is what's known for your specific drug, and honesty about the rest.

5. What are my odds if I stop — and if I don't?

The best single answer comes from ANTLER, a randomized trial in exactly your situation — long-term users who felt well: 56% of those who stopped relapsed within a year, versus 39% of those who continued [1]. Read it whole. Stopping raised relapse risk substantially — that's the case for staying. And four in ten people who kept taking the drug relapsed anyway, while roughly half of those who stopped did not — that's the case that stopping is survivable and staying is not a guarantee. Any page that quotes only one of those halves is selling something.

6. If we taper, what would the plan look like?

Not a cliff, and probably not even stairs of equal height. Pharmacology research argues that tapers should get proportionally smaller as doses get lower — "hyperbolic tapering" — because the drug's receptor occupancy falls steeply in the low-dose range, so the last milligrams are the loudest [4]. Practically, that can mean months, liquid formulations, and a prescriber willing to go slower than the standard schedule. The question to ask is concrete: "if we ever did this, what would week one and month three actually look like?"

7. How would we tell withdrawal from relapse?

The distinction that decides everything afterward. Withdrawal tends to arrive within days of a change, often with physical signatures — dizziness, electric-shock sensations — that were never part of your original condition, and eases if the dose is restored; relapse builds over weeks and looks like what you started with. A withdrawal wave misread as relapse becomes proof-you-can-never-stop, which is how tapers fail. The full stopping guide covers this in depth; agree on the telltales with your prescriber before any change.

8. Has my life changed in ways that change the answer?

New medications that interact (each drug's label interactions are in the library, verbatim), new diagnoses, pregnancy plans, age — the risk calculus at 65 differs from 35. The prescription hasn't changed in a decade; you have.

9. What's the smallest change we could try first?

The review isn't stop-versus-stay. Options between: a dose reduction with a review date, switching to a drug with a better long-term side-effect profile for your particular trade-offs, or explicitly deciding to continue — on purpose, re-decided, with a calendar date to re-ask. An affirmative decision to stay on is a perfectly good outcome of this checklist. The failure mode isn't staying; it's never deciding.

10. When do we ask all this again?

Once a year, on a date, in the calendar. The entire reason a decade slips by unexamined is that no mechanism ever fires. You are the mechanism. Set the date before you leave the appointment.

---

The companion to this page — the 12 questions for starting an antidepressant — ends with the exit-plan question that makes this review automatic. If you're reading both, you're already the best-prepared patient in the building.