How Is Treatment-Resistant Depression (TRD) Defined?

Treatment-resistant depression, or TRD, refers to major depressive disorder that does not improve after a person has tried at least two different antidepressant medications at adequate doses and for adequate durations. Both the U.S. Food and Drug Administration and the European Medicines Agency use this two-failure threshold as their working definition.¹ That sounds straightforward, but the reality is more complicated. What counts as “adequate,” whether the two medications need to be from different drug classes, and how to measure whether someone truly did not respond are all points of ongoing debate among researchers and clinicians.

What the Standard Definition Requires

The core idea behind TRD has been consistent since the late 1990s: a person with major depression whose symptoms persist despite at least two properly conducted antidepressant trials. An early operational definition proposed that those two trials should involve antidepressants from different pharmacological classes.² In practice, though, regulatory agencies have not always insisted on different classes. Some allow two trials of medications within the same class, provided each trial met minimum standards for dose and duration.

Those minimum standards matter. A diagnosis of TRD requires that each antidepressant trial reached at least the equivalent of 150 mg per day of imipramine (a benchmark dose used for comparison across drugs) and lasted six to eight weeks, and that both trials occurred within the same depressive episode.³ If someone was prescribed a low dose of an antidepressant for three weeks and then switched, that does not count as a failed trial. The bar is meant to ensure the medication actually had a fair chance to work before the label of “treatment-resistant” gets applied.

One persistent source of confusion is what “response” and “remission” mean in this context. Response typically means a 50 percent improvement in symptom severity from baseline, while remission means near-complete absence of depressive symptoms for at least three weeks.⁴ TRD is usually defined by failure to achieve response, but some researchers and staging models set the bar at failure to achieve remission, which raises the number of people who qualify. That distinction alone can shift prevalence estimates considerably.

Beyond a Simple Yes-or-No Label

Calling someone “treatment-resistant” suggests a binary: either the depression responds or it doesn’t. Researchers have pushed back against that framing for years. At least five formal staging models have been developed to capture the idea that treatment resistance exists on a spectrum. These models evolved from simple counts of how many medications a person had tried to more complex scoring systems that factor in the severity of the current episode, its duration, and whether treatments like electroconvulsive therapy had been attempted.⁵ The Maudsley Staging Model, for instance, assigns a numerical score based on multiple dimensions rather than simply marking someone as resistant or not.

The shift toward dimensional models reflects a clinical truth: a person who has failed two medications is in a very different situation from someone who has failed six medications plus electroconvulsive therapy. Lumping both under “TRD” is useful for regulatory purposes but limited for guiding next steps. Two research groups looking for biomarkers of TRD, for example, have used both the simpler Thase and Rush model and the more comprehensive Maudsley Staging Model side by side, precisely because each captures something the other misses.⁶

Pseudo-Resistance and Why It Matters

Before concluding that someone’s depression is truly treatment-resistant, clinicians need to rule out pseudo-resistance, a situation where symptoms persist not because the depression itself is resistant to medication, but because the medication never got a fair shot. The causes can sit on either side of the exam room. On the clinician’s side, the diagnosis might be wrong, the drug might have been prescribed at too low a dose, or the treatment course might have been too short. On the patient’s side, adherence might be spotty, the body might metabolize the drug unusually quickly, or an unrecognized medical or psychiatric condition might be interfering.⁷

The drug-metabolism piece is particularly tricky. Psychiatric medications need to cross the blood-brain barrier and bind to their targets in the brain. Factors like poor gut absorption, unusually fast liver metabolism, and low penetrance into the brain can all mean that a person is technically taking the right medication at the right dose but their brain never sees enough of it.⁸ Genetic variation in liver enzymes called cytochrome P450 enzymes contributes to this problem. People who carry gene variants that speed up enzyme activity end up with lower drug levels in their blood. One study found that CYP2C19 genotype significantly predicted blood levels of escitalopram, and CYP2D6 genotype predicted levels of nortriptyline, though interestingly neither genotype nor blood drug levels were significantly linked to whether the patient actually responded to treatment.⁹ That finding hints at how much about treatment resistance still can’t be explained by simple pharmacokinetics.

Misdiagnosis is another major contributor. Populations labeled as having TRD show high rates of unrecognized bipolar disorder.¹⁰ Treating bipolar depression with a standard antidepressant alone often does not work and can sometimes make things worse. For someone in that situation, the problem isn’t that their depression resists treatment; it’s that the wrong illness is being treated.

How Common TRD Is

Among U.S. adults already being treated with medication for major depression, roughly a third meet criteria for TRD. One study estimated that out of about 8.9 million Americans receiving medication for major depressive disorder, around 2.8 million, or about 31 percent, had treatment-resistant depression.¹¹ These numbers vary widely across settings and countries. A cross-sectional study at a specialized psychiatric hospital in Ethiopia found the prevalence of TRD among patients followed for major depression to be about 42 percent.¹² Differences in how TRD is defined, the populations studied, and the quality of initial treatment all push these numbers around.

The broad takeaway is that TRD is not rare. It affects millions of people, and the one-in-three ratio among medicated depression patients has become a commonly cited benchmark. That proportion has major implications for healthcare planning: any system treating depression needs to anticipate that a substantial share of patients will not respond to first- and second-line medication strategies.

The Financial and Personal Burden

TRD costs significantly more than depression that responds to treatment. In the U.S., privately insured patients with TRD who also had physical health conditions averaged about $22,500 in annual healthcare costs, compared with roughly $17,500 for patients with non-resistant depression and about $10,000 for people without depression at all.¹³ Among those with available work-loss data, TRD patients missed roughly twice as many workdays as those with non-resistant depression, and the resulting work-loss costs were thousands of dollars higher per person.

Similar patterns show up internationally. A U.K. population-based study of over 110,000 patients found that people with TRD used primary care about 50 percent more often than those with standard depression and had on average 1.5 times the total healthcare costs.¹⁴ Another U.S. analysis showed that over a two-year period, direct costs for employees likely to have TRD were nearly double those of employees with responsive depression, and indirect costs were roughly double as well.¹⁵ These costs compound over time, because TRD by definition means the person is cycling through more treatments, more appointments, and more lost productivity.

What Happens in the Brain

Researchers have been looking for biological differences between people whose depression responds to medication and those whose does not. One of the most consistent findings involves inflammation. People with depression in general tend to have higher levels of inflammatory markers in their blood compared to healthy controls, but those with TRD appear to have even higher levels than those who respond to medication. One study found that C-reactive protein, a common marker of inflammation, was significantly higher in TRD patients than in treatment-responsive patients.¹⁶

A separate study, part of a larger project called BIODEP, examined gene expression patterns in blood and found that people with TRD had increased expression of multiple inflammation-related genes, including IL-1β, IL-6, TNF-α, and several others. They also showed signs of glucocorticoid resistance, meaning their body’s usual system for dampening inflammation via stress hormones was less effective.¹⁷ This kind of chronic, low-grade immune activation could potentially blunt the effectiveness of standard antidepressants, though the causal picture is still far from complete.

Brain imaging research adds another layer. An MRI study comparing TRD patients to those with treatment-sensitive depression found abnormal connectivity patterns within and between brain networks. Specifically, TRD patients showed stronger-than-normal connections between parts of the default mode network, which is involved in self-referential thinking, and different patterns of communication with visual processing regions.¹⁸ These findings are still early-stage and come from relatively small samples, but they suggest TRD may involve distinct brain-circuit patterns, not just “more severe” versions of ordinary depression.

Physical Health Conditions and TRD

People with TRD tend to carry a heavier burden of physical health problems than those whose depression responds to treatment. A large retrospective study using electronic health records found that TRD patients had higher rates of anxiety disorders and a range of physical conditions, including higher odds of heart disease, stroke, chronic kidney disease, arthritis, and cancer compared to depressed patients without TRD.¹⁹

A Scandinavian study took a longer view by looking at medical conditions both before and after a TRD diagnosis. Before TRD was identified, patients already had higher rates of immune, neurological, and musculoskeletal disorders. After the TRD diagnosis, the gap widened to include cardiovascular, endocrine, and additional neurological conditions.²⁰ The word “bidirectional” is key here: it is not simply that illness causes depression or that depression causes illness, but that the two likely feed each other. Chronic inflammation, for example, could plausibly contribute to both cardiovascular disease and treatment resistance in depression. Not all studies agree on which specific conditions are most tightly linked, however. An earlier study comparing TRD and non-TRD patients found no significant difference in the overall prevalence of physical conditions between the two groups, though it did note marginally higher rates of breast cancer and migraine among TRD patients.²¹ The discrepancy likely reflects differences in sample size and study design, but it underscores that the relationship between TRD and physical health is still being mapped.

Treatment Options When Standard Medications Fail

Once someone meets the threshold for TRD, the usual approach is augmentation: adding a second medication to the antidepressant already in place. A large network meta-analysis evaluated various augmentation strategies and found that several agents significantly improved response rates compared to placebo, including atypical antipsychotics like aripiprazole, brexpiprazole, quetiapine, and olanzapine combined with fluoxetine, as well as lithium, thyroid hormones, and a handful of other agents.²² For remission specifically, the list was shorter: thyroid hormone, aripiprazole, brexpiprazole, risperidone, quetiapine, and olanzapine-fluoxetine all outperformed placebo.

Lithium, one of the oldest medications in psychiatry, remains a strong option. A systematic review comparing lithium augmentation to atypical antipsychotic augmentation found no significant clinical difference between the two strategies, but lithium came out ahead in cost-effectiveness analyses.²³ Despite this, lithium augmentation is underused in many settings, partly because it requires blood-level monitoring and partly because newer, more heavily marketed options have drawn attention away from it.

Esketamine

Esketamine, a nasal spray derived from the anesthetic ketamine, was approved specifically for TRD and has changed the treatment landscape for many patients. A recent randomized trial tested esketamine as a standalone treatment (rather than as an add-on to another antidepressant) and found that both the 56 mg and 84 mg doses produced significant reductions in depression severity compared to placebo, with meaningful improvement appearing within about 24 hours of the first dose.²⁴ That speed is striking. Standard antidepressants typically take weeks to show an effect, so for someone who has already been struggling through months or years of failed treatments, rapid onset can be a significant relief. Esketamine does require in-clinic administration and monitoring, and its long-term profile is still being studied.

Brain Stimulation

When medications and augmentation strategies fall short, neuromodulation therapies are the next step. Electroconvulsive therapy, or ECT, is the oldest and most established of these. A comparison of ECT and repetitive transcranial magnetic stimulation (rTMS) found that ECT achieved a response rate of about 64 percent and a remission rate of about 53 percent, while high-frequency rTMS produced response and remission rates of roughly 49 percent and 32 percent, respectively.²⁵ A meta-analysis of randomized trials confirmed that while rTMS outperformed sham treatment, ECT held a statistically and clinically significant advantage over rTMS.²⁶

That said, one randomized trial from Nigeria produced a surprising durability result: at six weeks, ECT and rTMS had virtually identical response rates (95 percent and 90 percent), but at 12 months, the rTMS group maintained significantly higher response and remission rates than the ECT group.²⁷ This is a single small trial and should be interpreted cautiously, but it raises interesting questions about longer-term maintenance. For cases that remain resistant even to ECT and rTMS, more invasive options exist. Vagus nerve stimulation has FDA approval as an add-on treatment for major depression, and deep brain stimulation is being investigated experimentally.²⁸

Psychotherapy for TRD

Medication and brain stimulation dominate TRD conversations, but psychotherapy has a role too, particularly a specialized form called Cognitive Behavioral Analysis System of Psychotherapy, or CBASP. Developed specifically for chronic depression, CBASP focuses on helping people recognize how their behavior patterns affect their relationships and outcomes. A meta-analysis of randomized trials found that CBASP had small-to-moderate effects compared to treatment as usual and was comparable to antidepressant medication alone. Combining CBASP with medication produced benefits beyond medication on its own.²⁹ A group-adapted version of CBASP also showed significant reductions in depressive symptoms along with improvements in social functioning and coping strategies.³⁰ Psychotherapy will never replace medications for everyone with TRD, but it can be a meaningful complement, and for some patients it addresses interpersonal and behavioral patterns that medication simply cannot reach.

When the Patient Is a Child or Teenager

TRD definitions developed for adults do not translate neatly to younger populations. In children and adolescents, the threshold is usually lower. One working definition considers a young person treatment-resistant after one adequate medication trial and one adequate psychotherapy trial, either given back to back or together.³¹ That psychotherapy requirement is a meaningful distinction from adult definitions, reflecting the fact that evidence-based therapy like cognitive-behavioral therapy is considered a first-line treatment for adolescent depression, not just an add-on. Some definitions are even more specific, setting the bar at a two-month course of an antidepressant at a dose equivalent to 40 mg of fluoxetine per day, or 8 to 16 sessions of cognitive-behavioral or interpersonal therapy.³² The stakes of getting the definition right are high, because labeling a young person as treatment-resistant affects which interventions are considered next, and many of those interventions carry more risk in developing brains.

Psilocybin Research

Among the most closely watched developments in TRD research is psilocybin, the active compound in psychedelic mushrooms, used in combination with psychological support. An early open-label feasibility trial in patients with TRD found that depressive symptoms dropped substantially one week and three months after a high-dose session, with large effect sizes.³³ A larger randomized trial in the New England Journal of Medicine compared three dose levels of synthetic psilocybin. The 25 mg dose produced a significant reduction in depression scores at three weeks compared to the 1 mg control dose, while the 10 mg dose did not separate from the control.³⁴

More recent work has examined whether repeated psilocybin doses add further benefit. A randomized clinical trial found that the immediate-treatment group showed large reductions in depression severity compared to a waitlist group, and additional doses were associated with continued improvement.³⁵ The results are promising, but the field is still working through questions of optimal dosing, how long the effects last, who benefits most, and how to scale a therapy that requires hours of one-on-one psychological support per session. Psilocybin is not yet approved for clinical use in most countries, and the regulatory path forward remains uncertain.