What to Do When Your Antidepressant Stops Working

The short answer

When an antidepressant stops working, do not stop it on your own. First rule out missed doses, a new interacting medication, more alcohol, a thyroid or sleep problem, a major stressor, or an unrecognized bipolar pattern. If depression has truly returned despite consistent dosing, the options are a dose increase, augmentation, a switch, or Deep TMS or Spravato.

A patient who has done well on sertraline for three years comes in and says the depression is back. Same medication, same dose, same pharmacy. Nothing changed, they say. And yet the mornings are heavy again, the interest is gone, and they are wondering whether the pill was ever doing anything at all.

This is one of the most common conversations in psychiatry, and the answer is rarely as simple as "the drug wore off." Here is how I think through it, and what the options are once we know what we are dealing with.

First, do not stop it

The instinct when a medication seems useless is to stop taking it. With antidepressants this is a mistake on two counts. Stopping abruptly, especially with paroxetine, venlafaxine or duloxetine, produces discontinuation symptoms: dizziness, electric-shock sensations, nausea, irritability and a rebound in anxiety that can be confused with the depression getting worse. And if the medication was still providing partial benefit, removing it makes the depression deeper before any new treatment has a chance to work. Call your prescriber. The change should be planned.

Is it really the medication?

Before I conclude that an antidepressant has lost effect, I look for the things that commonly mimic that. They account for the majority of cases.

Adherence

Not an accusation, just the most frequent finding. Missed doses accumulate. Mail-order refills get delayed. A pharmacy switches generic manufacturers, and for a small number of patients a different generic formulation is absorbed differently. I ask how often doses are missed in a typical week, and I ask whether the pills look different than they used to.

New medications and substances

Several common prescriptions reduce antidepressant levels or blunt their effect. Carbamazepine and some other anticonvulsants induce liver enzymes that clear antidepressants faster. Some new drugs added for other conditions, including certain acid reducers, antibiotics and antifungals, change metabolism in the other direction. Beta blockers, corticosteroids, hormonal contraceptives and isotretinoin can all lower mood on their own. And alcohol, which many people increase gradually without noticing, is a depressant that directly counteracts the medication. Cannabis use that has crept up is another frequent contributor.

Medical causes

Hypothyroidism is the classic one, and it develops slowly enough that it looks like a gradual return of depression. Anemia, vitamin B12 or D deficiency, untreated sleep apnea, perimenopause, low testosterone in men and new-onset diabetes can all present as fatigue, low mood and poor concentration. A basic lab panel is cheap and often decisive.

Life

Antidepressants raise the floor; they do not make you immune. A divorce, a job loss, a bereavement, a child in crisis or a chronic caregiving burden can overwhelm a medication that was adequate under ordinary conditions. In that case the medication has not failed. The load has changed, and the answer may be therapy, a temporary dose increase, or simply time and support.

A diagnosis that was incomplete

This is the one that matters most. A patient whose depression responds to an antidepressant, then returns, then responds, then returns, sometimes with periods of unusual energy or reduced sleep in between, may have a bipolar spectrum disorder rather than unipolar depression. Antidepressants alone are unreliable in bipolar depression and can destabilize mood over time. Before changing anything, I take a careful history of hypomanic symptoms and family history. Our article on bipolar versus unipolar depression covers this distinction in depth.

When the medication has genuinely lost effect

If adherence is solid, nothing new has been added, labs are normal and the history is clearly unipolar, then we are looking at what the literature calls antidepressant tachyphylaxis, sometimes nicknamed "poop-out." It is real. Estimates vary widely, but studies following patients on maintenance antidepressants suggest that somewhere between 10 and 30 percent experience a loss of response over several years despite continued treatment. The mechanism is not well understood; receptor adaptation and the natural course of recurrent depression are both proposed.

The STAR*D trial is the best guide to what happens next. It showed that among patients who did not reach remission on a first antidepressant, each subsequent step, whether a switch or an augmentation, brought additional patients into remission, with cumulative remission around two thirds after four steps. The practical lesson is that there is always a next step, and that the next step should be taken promptly rather than after months of waiting.

The options, in rough order

Increase the dose

If there is room under the maximum and side effects are tolerable, this is the simplest move. Many patients were stabilized at a moderate dose years ago and never pushed higher. For SSRIs the dose-response curve is relatively flat above the middle of the range, so this helps some patients and not others, but it costs little to try for four weeks.

Augment

Adding a second medication to the existing antidepressant preserves whatever benefit remains and often restores response within two to four weeks. The options with the strongest evidence are the atypical antipsychotics aripiprazole, brexpiprazole and cariprazine, all FDA-approved as adjuncts for major depression; lithium, which has decades of data and a specific anti-suicidal effect; and bupropion, which is particularly useful when fatigue, low motivation or sexual side effects are prominent. Thyroid hormone augmentation, used in STAR*D, is an older option that still has a place.

Switch

If the current medication is doing nothing, switching makes more sense than adding to it. Switching within the same class (one SSRI to another) works for some patients; switching across classes (SSRI to SNRI, bupropion, mirtazapine or vortioxetine) is often more productive after a true loss of response. A cross-taper over two to four weeks avoids discontinuation symptoms and a gap in coverage.

Add psychotherapy

Cognitive behavioral therapy added to medication reduces relapse and is often the right move when a life stressor contributed to the return of symptoms. It is not a consolation prize. In several trials it matched the effect of a medication change.

Interventional treatment

Once two adequate antidepressant trials have failed in the current episode, the depression meets the usual definition of treatment-resistant, and the interventional options become appropriate rather than exotic. Deep TMS, FDA-cleared for major depressive disorder, is a course of 36 brief sessions over six to nine weeks with no systemic side effects and remission rates in the range of a third or more of patients who had already failed medication. We offer it at Palm Beach Gardens. Spravato, FDA-approved for treatment-resistant depression, is given twice weekly for four weeks and then tapered, in a certified clinic with two hours of monitoring, and often works within the first two weeks. We offer it at Boynton Beach, Palm Beach Gardens and Royal Palm Beach. Both are covered by most insurance and Medicare with prior authorization.

How we approach it at Stedfast Psychiatry

A patient whose antidepressant has stopped working gets a visit that looks a lot like a new evaluation: a full review of symptoms, adherence, medications, substances and life circumstances; screening for bipolar features; and labs when indicated. Then we choose the next step together, with a defined timeline for judging it, usually four to six weeks. We see patients at all 13 offices in Florida and South Carolina and by telehealth. If you are reading this because your own medication seems to have quit, the most useful thing you can do is keep taking it and make the appointment.

Frequently asked questions

Why would an antidepressant stop working after years?

True loss of effect, called tachyphylaxis, occurs in an estimated 10 to 30 percent of patients on long-term treatment. More often, though, something else explains it: missed doses, a new medication, increased alcohol, a thyroid or sleep problem, a major life stressor, or an unrecognized bipolar pattern. A psychiatrist checks those first.

Should I stop my antidepressant if it is not working?

Not on your own. Abrupt stopping causes discontinuation symptoms and removes any partial benefit before a replacement has a chance to work. Call your prescriber; the change should be a planned dose increase, augmentation or cross-taper to a new medication.

Is it better to switch antidepressants or add another medication?

If the current medication is still providing partial benefit, augmenting with aripiprazole, lithium or bupropion preserves that benefit and often works within two to four weeks. If it is doing nothing, switching to a different class is more productive. Both approaches produced additional remissions at each step in the STAR*D trial.

When should I consider TMS or Spravato?

After two adequate antidepressant trials have failed in the current episode, which is the standard definition of treatment-resistant depression. Deep TMS is FDA-cleared and Spravato is FDA-approved for this situation, and both are covered by most insurance and Medicare with prior authorization. Stedfast Psychiatry offers Deep TMS at Palm Beach Gardens and Spravato at three offices.

Could my antidepressant stopping working mean I have bipolar disorder?

It can. A pattern of response followed by relapse, especially with periods of elevated energy, reduced need for sleep or irritability between depressions, raises the possibility of a bipolar spectrum disorder. Antidepressants alone are unreliable in that case, so a careful history is taken before changing treatment.