When Depression Treatment Isn't Working

There's a quiet, corrosive thought that shows up when a depression treatment doesn't work: maybe I'm the problem. Maybe nothing will work for me. The research says otherwise on both counts. Roughly two-thirds of people do not reach full remission on their first antidepressant — not a personal failure, just the expected statistics of a condition treated by informed trial and error.

Clinicians usually reserve the term treatment-resistant depression (TRD)for depression that hasn't responded adequately to twoantidepressant trials of adequate dose and duration. The label is imperfect — "resistant" can sound final, when it really just means the first-line playbook wasn't enough — and the deeper playbook is longer than most people realize.

The most influential study here is STAR*D (Sequenced Treatment Alternatives to Relieve Depression), a large NIMH-funded trial that followed thousands of real-world patients through up to four sequential treatment steps, switching or adding options at each step:

About one-third of participants reached remission at step one. Each additional step produced additional remissions — smaller percentages each time, but real ones — and cumulatively, roughly two-thirds of those who stayed through all four steps reached remission.

Two honest lessons sit side by side. First, the returns diminish — later steps help fewer people, and relapse was more common for those who needed more steps, so follow-up matters. Second, and more importantly: people who kept going kept getting better.If step one didn't work, you've barely opened the menu.

Before adding the TRD label, good clinicians re-examine the foundations — several fixable issues masquerade as resistance:

• Was the trial adequate?An antidepressant needs an adequate dose for roughly 6–8 weeks before it can be fairly judged; trials cut short by side effects or under-dosing don't count.
• Adherence and side effects. Missed doses are common and understandable — especially when side effects are rough. Naming that openly with a prescriber often leads to a workable adjustment.
• Is the diagnosis right? This matters most for bipolar disorder: depression is usually its presenting face, and antidepressants alone can be ineffective or destabilizing — screening for past hypomanic symptoms is standard before stacking more antidepressant trials. Unrecognized PTSD, ADHD, substance use, and psychotic features can also stall progress.
• Medical contributors. Thyroid problems, sleep apnea, anemia, some medications (including steroids), and regular alcohol or cannabis use can all cause or maintain depressive symptoms — basic labs and a sleep history are worth their weight.

When the foundations check out and depression persists, prescribers generally work through three well-studied moves — always as a conversation about trade-offs:

• Optimizing — raising the current medication to the top of its effective range and giving it full time, if side effects allow.
• Switching — to a different medication in the same class or a different one. Large network meta-analyses (Cipriani and colleagues, 2018) confirm antidepressants outperform placebo and differ somewhat in efficacy and tolerability.
• Augmentation — adding a second agent to boost a partial response. The best-studied options include lithium, atypical antipsychotics at low doses, and thyroid hormone (T3), each with its own monitoring and side-effect considerations.

One strategy deserves special emphasis because it's so often skipped: adding psychotherapy. Meta-analytic work by Cuijpers and colleagues consistently finds that combining psychotherapy with medication outperforms either alone, and protects better against relapse. If treatment so far has been medication-only, therapy is one of the strongest next moves on the board. Our medication reference and what to expect with psychiatric medication guides cover the practical details.

When medication steps aren't enough, a set of brain-stimulation and rapid-acting treatments comes into play — several with strong evidence.

• Electroconvulsive therapy (ECT) remains the single most effective treatment for severe depression, with response rates well above medication in the hardest cases. The modern procedure bears little resemblance to its movie image: brief general anesthesia, a muscle relaxant, a few short sessions a week. The main trade-off is memory — temporary confusion and, for some, gaps around the treatment period — weighed against how dangerous severe depression itself is.
• Repetitive transcranial magnetic stimulation (rTMS)is FDA-cleared for depression that hasn't responded to medication. A magnetic coil held against the scalp stimulates mood-related brain circuits — no anesthesia, no memory effects — in daily sessions over roughly four to six weeks. Gentler than ECT and less powerful, it occupies a useful middle ground.
• Esketamine (Spravato), FDA-approved in 2019 specifically for TRD, is a nasal spray taken alongside an oral antidepressant. Because of blood-pressure and dissociation effects, it's administered under supervision in a certified clinic through a REMS safety program — and it can work within days rather than weeks. IV ketamine shows similarly rapid effects; durability requires a maintenance plan, and long-term research is ongoing.
• Vagus nerve stimulation (VNS) is an implanted option for long-standing TRD — less common, but part of the map for chronic, severe cases.

None of these replace treatment — but as force multipliers, three behavioral moves have outsized evidence:

• Behavioral activation — deliberately, gradually re-engaging with rewarding activities before motivation returns. A full evidence-based therapy in its own right, its core insight (action precedes motivation, not the other way around) helps almost everyone with depression.
• Exercise— regular aerobic activity has meaningful antidepressant effects in trials. The dose that matters is the one you'll actually do; walking counts.
• Treating sleep directly — insomnia and sleep apnea both feed depression and blunt treatment response; CBT-I and apnea treatment can shift mood surprisingly far. See our guide to sleep and mental health.

Depression lies. When it persists through treatment, it tells you the treatments failed because you're beyond help — and presents that feeling as fact. The actual fact pattern: there are many sequential paths — optimizing, switching, augmenting, adding therapy, rTMS, esketamine, ECT — and most people who keep working through them with a good clinician find one that helps. "Nothing has worked yet" is a very different sentence from "nothing will work," and only the first one is true.

It's also okay for hope to be borrowed for a while — from a clinician, a partner, a friend who holds the long view when you can't.

Working through treatment steps goes best with two partners: a psychiatric prescriber (psychiatrist or psychiatric nurse practitioner) comfortable with augmentation strategies, and a therapist for the psychotherapy half of combined treatment. It's reasonable to ask directly whether a prescriber treats treatment-resistant depression and where they refer for rTMS, esketamine, or ECT.