evidence·published august 19, 2026·last verified august 24, 2026

how long were these drugs actually tested?

shorter than you are taking them. we counted the trial length of every study in the Resolv evidence library as it stood on 18 august 2026 — 235 verified records, 72 of them individual randomised trials. fifty-seven of those 72 state how long they ran. forty-seven of the fifty-seven measured their headline result at three months or sooner. six ran past six months. two studied an everyday prescription psychiatric drug for a year or more. meanwhile the same library contains a meta-analysis pooling people who had been on benzodiazepines for about ten years, and a cohort following antipsychotic patients for up to seven and a half. that gap is the whole page.

0
of 235 records carry a duration field on the study itself
72
individual randomised trials counted
57
of the 72 state how long they ran
47
of those 57 ran 3 months or less
8
of 93 pooled reviews state trial length

the field we could not fill

The Resolv evidence library records eighteen things about every study: what kind of study it is, how many people were in it, the journal, the year, the DOI, the PubMed ID, who funded it, whether it was preregistered, whether conflicts were disclosed, whether the authors were independent of the product, a trust score, and more.

Trial duration is not a field on the study record. Zero of the 235 records carry a structured duration for the study they describe.

One precise caveat, because the version of this sentence we published on 19 August was too strong and a reader could have falsified it in one query. Seven records do carry a structured trial_duration string, inside the approved_indications block: five quetiapine records (“8-week randomized trials”) and two esketamine records (“4-week pivotal trial, TRANSFORM-2, day-28 endpoint”). That field describes the trials behind the medication’s FDA approval, not the study the record is about, and it is null on the other 97 records that carry the block at all. The claim we should have made, and now make, is the narrower one: no record carries a structured duration for its own study.

We could have quietly added one and backfilled it from what we know about each drug. We did not, because the absence is itself the finding — and because a duration inferred from "SSRI trials are usually about 8 weeks" is a fabrication wearing a data field’s clothes.

What we did instead: read the plain-English summary of all 235 records, pull out every horizon stated verbatim, and hand-classify one record at a time. Records that state no horizon are counted in a "not stated" row, not dropped from the denominator and not estimated. Every number below is reproducible by reading the same records.

57 of 72 randomised trials say how long they ran. 47 of those ran three months or less.

Of the 235 records, 72 are individual randomised trials — the design that establishes whether a drug works. Fifty-seven of those state a treatment period or a primary-outcome horizon. Fifteen do not.

Of the 57 that do: 47 measured their headline result at three months or sooner. That is 82% of the trials that tell you. Four of them measured it the same day.

Six of the 72 ran past six months. Three ran past twelve. Those three are: a lithium relapse-prevention trial (up to two years), a 1986 methylene-blue crossover in 31 people with bipolar disorder (two years, 17 completers), and a fish-oil prevention trial with a median of 5.3 years that found no benefit.

every randomised trial in the library, by how far out it measured

72 records. rows sum to 72, and the not-stated row stays in the table rather than disappearing into a footnote.

primary-outcome horizonrecordswhich ones
same day — record states the outcome was measured within hours4 of 72CBD before a simulated public-speaking test (one of the two CBD records states the horizon; the other does not); L-theanine before a stress task (twice); a single kratom decoction
2 days to 1 month13 of 72esketamine at day 28; ketamine versus ECT over 3 weeks; ayahuasca at day 7; MM120 at week 4; COMP360 psilocybin at week 3; hydroxyzine over 4 weeks; a 3-week methylene-blue trial from 1987
more than 1 up to 2 months18 of 72the 1988 alprazolam panic trial (8 weeks); PANDA sertraline (6 weeks); BOLDER I quetiapine (8 weeks); both flagship ashwagandha trials (60 days); both creatine trials (8 weeks); psilocybin versus escitalopram (week 6)
more than 2 up to 3 months12 of 72TADS fluoxetine in adolescents (12 weeks); escitalopram for older adults with GAD (12 weeks); the MIR mirtazapine augmentation trial (12 weeks); OPTIMUM aripiprazole (10 weeks); all three NAC trials (12 weeks); the 1993 clonazepam social-phobia trial (10 weeks)
more than 3 up to 6 months4 of 72the 2017 methylene-blue bipolar crossover (6 months); the NAC bipolar trial (16 weeks); LSD-assisted therapy for anxiety (16 weeks); psilocybin in life-threatening cancer (6-month follow-up)
more than 6 up to 12 months3 of 72ANTLER, the antidepressant discontinuation trial, counted twice because it appears under both fluoxetine and sertraline (1 year); psilocybin for alcohol use disorder (32 weeks)
more than 12 months3 of 72BALANCE lithium relapse prevention (up to 2 years); a 1986 methylene-blue crossover in bipolar disorder (2 years, 31 people, 17 completers); VITAL-DEP omega-3 (median 5.3 years)
no horizon stated anywhere in the record15 of 72EAGLES; VAST-D; both STAR*D switch and augmentation arms; Restoring Study 329; the 1997 clonazepam dose-response trial; both MDMA phase 3 trials; SUSTAIN-1 esketamine maintenance; the 2015 aripiprazole augmentation trial; a microdosing trial; an internet sex-therapy pilot; both intravenous-lorazepam status-epilepticus trials and one of the two CBD public-speaking trials, which were previously counted as same-day and should not have been — see the correction note below

One accounting note, published rather than smoothed: the 6-to-12-month row contains three records but only two distinct papers. ANTLER appears twice because it is filed under both fluoxetine and sertraline — the trial randomised people taking either. We count records, and we tell you when a record count and a paper count diverge.

the syntheses almost never tell you either

This is the part we did not expect. Meta-analyses and systematic reviews are supposed to be the layer that contextualises a scattered literature. In this corpus, they overwhelmingly report an effect size with no time attached to it.

Of 93 pooled-evidence records — 72 meta-analyses of randomised trials and 21 systematic reviews of randomised trials — only 8 state how long the underlying trials ran. Split: 5 of the 72 meta-analyses (/r/51597 amphetamine, /r/51607 bupropion, /r/51660 mirtazapine, /r/51766 NAC, /r/51786 St John’s wort) and 3 of the 21 systematic reviews (/r/51682 trazodone in dementia, /r/51804 sildenafil, /r/51825 valerian). Proportionally the systematic reviews report duration more often than the meta-analyses, 3 of 21 against 5 of 72 — the opposite of what an earlier version of this page said.

So when a clinician or an article tells you "a meta-analysis of 133 trials shows this drug reduces symptoms", the time horizon behind that sentence is usually not recoverable from the thing being quoted. Two records in this library do it properly and deserve the credit: the Cochrane review of amphetamines for adult ADHD, which reports a mean trial length of 5.3 weeks, and the St John’s wort meta-analysis, which states that every one of its 27 included trials ran only 4 to 12 weeks and that it is therefore unclear whether the finding holds beyond that.

the longest randomised horizon, by drug class

Read this table as "the longest time horizon stated anywhere in this corpus for that class" — across individual trials and pooled evidence together, because several classes have no individual trial record at all. Three rows say nothing stated. Those are not errors; they are the honest state of what this library can tell you.

drug classlongest horizon statedwhere that comes from
SSRI antidepressants (4 drugs, 19 records)12 months — but it is a stopping trialANTLER (Lewis 2021, NEJM, n=478) randomised people who felt well enough to come off long-term antidepressants; 56% of those who stopped relapsed within a year versus 39% of those who stayed on. the longest horizon in any SSRI efficacy trial here is 12 weeks (TADS 2004; Lenze 2009).
SNRI antidepressants (2 drugs, 10 records)nothing statednot one of the 10 duloxetine or venlafaxine records in this corpus states how long its trial or its pooled trials ran. nine are meta-analyses and reviews that report effect sizes without a time horizon; the single individual trial (the STAR*D level-2 switch) does not state one either.
other antidepressants (3 drugs, 15 records)6 months or longer — for quitting smoking, not for depressionthe Cochrane review of antidepressants for smoking cessation (Hajizadeh 2023, 50 trials, 18,577 people) reports six-month-or-longer quit rates for bupropion. for depression the longest stated horizon is 12 weeks (the MIR mirtazapine trial, 2018).
benzodiazepines (3 drugs, 15 records)10 weeksa 1993 double-blind trial of clonazepam for social phobia in 75 outpatients (Davidson 1993). the alprazolam panic trial ran 8 weeks; the 2011 clonazepam-versus-paroxetine trial ran 8 weeks. against that, the lorazepam FDA label states that effectiveness beyond 4 months has not been systematically established, and this corpus separately contains a meta-analysis pooling people who had taken benzodiazepines for about 10 years.
z-drug hypnotics (zolpidem, 5 records)nothing statednone of the five zolpidem records states a trial length. two are meta-analyses that report the effect in minutes of sleep latency without a study duration, one is a falls-and-fractures review, and two are FDA safety communications. the AMBIEN label indication is short-term treatment of insomnia.
non-benzodiazepine anxiolytics (2 drugs, 8 records)12 weeksLlorca 2002, a 12-week hydroxyzine trial in 334 adults with generalised anxiety — the paper is literally subtitled "a 3-month double-blind study". the hydroxyzine FDA label states effectiveness beyond 4 months has not been established; the buspirone label indication is short-term relief of anxiety symptoms.
antipsychotics (2 drugs, 10 records)10 weeksOPTIMUM (Lenze 2023, NEJM, n=619) measured well-being gain over 10 weeks. the corpus separately contains a Swedish national cohort following 29,823 people with schizophrenia on antipsychotics for up to 7.5 years — observational, not randomised.
mood stabilisers (2 drugs, 10 records)2 yearsBALANCE (Geddes 2010, Lancet, n=330) followed people with bipolar I for up to two years, and a pooled analysis of two 18-month maintenance trials (Goodwin 2004, n=638) found lamotrigine roughly doubled median time to the next mood episode, 197 versus 86 days. this is the only class in the corpus where the long-term question was actually asked with randomisation.
ADHD stimulants (2 drugs, 10 records)12 weeksthere is no individual randomised-trial record for either stimulant in this corpus — the evidence is meta-analytic. Cortese 2018 pooled 133 double-blind trials at a 12-week efficacy horizon, and its own contested note reads that head-to-head evidence at 12 weeks says little about long-term outcomes. the Cochrane amphetamines review reports a mean trial length of 5.3 weeks. the long-term data are cohorts: the MTA follow-up to about age 25, and a Swedish mortality cohort over 2 years.
ketamine and esketamine (5 records)3 weeksELEKT-D (Anand 2023, NEJM, n=403) found IV ketamine noninferior to ECT over 3 weeks. TRANSFORM-2 measured esketamine at day 28. SUSTAIN-1, the relapse-prevention trial, does not state a horizon in its record.
psychedelics (4 substances, 13 records)32 weeksBogenschutz 2022 followed heavy-drinking days for 32 weeks after psilocybin-assisted psychotherapy for alcohol use disorder. most psychedelic records here measure at 3 to 6 weeks; the COMP360 phase 2b record notes that benefit at week 3 had faded for many by week 12.
MDMA (3 records)nothing statedneither MAPP1 (2021) nor MAPP2 (2023) states a follow-up horizon in its record. both report CAPS-5 change after three MDMA-assisted therapy sessions inside a roughly 40-hour therapy package.
supplements (11 substances, 52 records)5.3 yearsVITAL-DEP (Okereke 2021, JAMA, n=18,353, median 5.3 years) — the longest randomised horizon anywhere in this corpus, and it is a prevention trial of a fish-oil capsule that found no benefit. the St John’s wort meta-analysis states plainly that every one of its 27 included trials ran only 4 to 12 weeks.
other non-prescription substances (6, 28 records)2 yearsa 1986 double-blind crossover of methylene blue in 31 people with bipolar disorder, one year on each dose, 17 completers. the CBD evidence in this corpus is almost entirely single-dose laboratory challenges.

“nothing stated” means no record for that class — trial or review — states a time horizon. it does not mean the trials were short or that none exist; it means this library cannot tell you, and neither can the sources it drew from without opening the full papers.

and how long do people actually take them?

We are careful here, because our library is a library of studies, not of prescriptions — it contains no data on how long any individual stays on a medication, and we are not going to import a statistic from somewhere else to make the contrast look sharper. What it does contain are studies whose own design tells you something about real-world duration, and two FDA labels that state their own limits.

what the record describesover how longdesignsource
people studied for the cognitive effects of long-term benzodiazepine useabout 10 years on averagemeta-analysis of 13 observational studiesBarker 2004, CNS Drugs (PMID 14731058)
Swedish patients with schizophrenia followed on antipsychotic treatmentup to 7.5 yearsnationwide cohort, 29,823 peopleTiihonen 2017, JAMA Psychiatry (PMID 28593216)
people randomised in ANTLER, all of whom had been on antidepressants long enough to consider stoppingthe trial itself ran 1 yearrandomised discontinuation trial, 478 primary-care patientsLewis 2021, NEJM (PMID 34587384)
children with ADHD in the MTA study followed into adulthooda 14-month randomised phase, then observational follow-up to about age 25observational follow-up of a randomised trial, 515 peopleSwanson 2017, J Child Psychol Psychiatry (PMID 28295312)
what the lorazepam FDA label says about the horizon it was tested overeffectiveness beyond 4 months not systematically establishedFDA-approved label, verified against DailyMed on 2026-08-17DailyMed SPL 787240e9-67b5-4af3-b16d-7cf314e3f2e3
what the hydroxyzine FDA label sayseffectiveness beyond 4 months has not been establishedFDA-approved label, verified against DailyMed on 2026-08-17DailyMed SPL b4d274ce-d616-4835-80ce-76270524ed36

One more, worth flagging because it is a thing we deliberately did not use. A cross-sectional study finding that long-term benzodiazepine use reached 31.4% of users aged 65 to 80 (Olfson 2015, JAMA Psychiatry) was verified during the harvest and then cut by the five-studies-per-medication cap. It is a real, verified number and it is not in our corpus, so it is not in our count — mentioned here so you can go and read it, not counted as if it were ours.

Similarly: a three-year follow-up comparing long-term clonazepam with paroxetine exists, and it was left out of the clonazepam file because its PubMed ID could not be verified in that pass. It would have been the longest benzodiazepine horizon in the corpus. We would rather have a 10-week ceiling that is true of our data than a 3-year one we could not check.

what we will not claim

Four things this page deliberately does not say, because the data cannot carry them.

  • We do not claim these drugs are unsafe over the long term. Short randomised evidence plus long observational evidence is a description of an evidence base, not a safety signal. Where this library holds real long-term harm signals — the benzodiazepine cognitive meta-analysis, the antipsychotic rehospitalisation cohort, the MTA height findings — they are recorded with their own caveats and they are observational.
  • We do not claim a typical trial length for any drug class. With 5 records per medication, "the median SSRI trial is X weeks" would be a number computed over four trials. That is not a median, it is an anecdote with a decimal point.
  • We do not fill the 12 not-stated trials from published papers we did not read in this pass. Their durations are knowable and we could look them up. Doing so mid-count would make the number unreproducible from the records, which is the one property that makes it worth publishing.
  • We do not extrapolate from 235 records to psychiatry. This is a curated library, weighted toward the highest-quality evidence for each drug. It is enough to describe itself precisely. It is not a survey of the field.
  • 235 is the 18 August 2026 harvest, not the current library size. The library held 251 published resources on 2026-08-19 and keeps growing. Every count on this page is fixed to the harvest so the arithmetic stays reproducible; the live library is free to read in full.

how we counted

where these numbers come from. the same corpus as our funding-transparency count: two study harvests completed 2026-08-17 (21 prescription medications plus psychedelic-assisted therapy, 123 records) and 2026-08-18 (17 non-prescription substances and 7 topics, 112 records). 235 records, 216 distinct papers, 46 source files. every citation was fetched live from PubMed (NCBI E-utilities) and Europe PMC and re-verified mechanically in an independent second pass.

there is no duration field on the study, and we did not invent one. the harvest schema records 18 fields per study and the length of that study is not among them — 0 of 235 carry it. (seven carry an FDA-approval trial_duration under approved_indications, which is a different thing; see the caveat above.) every horizon on this page was read verbatim out of the record’s own plain-English finding or its contested note, then hand-classified one record at a time into the eight buckets above. two people can check any single classification against the quoted phrase.

what we measured: the primary-outcome horizon. that is the point at which the trial measured its headline result — not the length of dosing, which is a different question and often shorter. we chose it because it is the one measure that is comparable across paradigms: a single-dose psilocybin trial measuring at week 3 and a daily-dosing SSRI trial measuring at week 12 can be placed on the same axis, where "how long were you on the drug" cannot. where a record states both a treatment period and a longer follow-up, the treatment period governs and the follow-up is noted (the LSD 2014 pilot is classified at its 2-month primary outcome, with its reported 12-month durability noted rather than counted).

what counts as "not stated". a record is counted as not stated if neither its finding nor its contested note gives a treatment period or an outcome horizon. fifteen randomised-trial records fall here. they are shown as a row in the table, not removed from the denominator — every percentage on this page is over 57 stated, and the 72 is always visible next to it.

what we excluded and why. nothing was excluded from the count; all 235 records were classified. exclusions happened upstream in the harvest and are documented there: four Russian-language phenibut trials dropped for unverifiable full text, roughly 30 studies verified and then deliberately left out for stated reasons (duplicate designs, post-hoc re-slices of one failed trial, one paper excluded on author-integrity grounds). two duration-relevant exclusions are named on this page because they would have changed a number: Olfson 2015 and a three-year clonazepam follow-up whose PMID could not be verified.

known limits. (1) our summaries were written to convey findings, so the 15 not-stated records reflect our record-keeping as much as the papers. (2) "longest horizon in this corpus" is not "longest trial that exists" — for several classes longer trials certainly exist and are simply not among the five records we selected per medication. we say "in this corpus" everywhere for that reason. (3) 5 records per medication is too few to compute a median or a distribution per drug, and we do not.

we are not clinicians. this page reports how long published studies ran. it is not medical advice, it is not a reason to stop or change a medication, and it cannot account for your diagnoses or history.

take the data. download all 72 rows as CSV — one row per trial, with the duration phrase quoted verbatim and a link to each record. Licensed CC BY 4.0: reuse the counts anywhere, including commercially, as long as you say where they came from and link back. If you recount and get a different answer, we want to know.

this is not medical advice, and it is emphatically not a reason to stop a medication. stopping psychiatric drugs abruptly can be dangerous, and several of the medications on this page have withdrawal or relapse risks documented in the same library. if the gap between trial length and your own time on a drug bothers you, that is a good conversation to have with your prescriber — not a decision to make alone. if you're in crisis, call or text 988 (u.s.), 24/7, free.

questions

how long is a typical psychiatric drug trial?

In this library, short. Of the 72 individual randomised trials, 57 state how long they ran, and 47 of those 57 measured their headline result at three months or sooner. The single largest bucket is 1 to 2 months (18 records). Only 6 of the 72 ran past six months, and only 3 past twelve. That matches what the trials themselves say: the flagship alprazolam paper is titled "efficacy in short-term treatment", and the St John’s wort meta-analysis notes that every one of its 27 pooled trials ran 4 to 12 weeks.

so is my medication unsafe because the trials were short?

No — short trials are not evidence of harm, and this page is not making that claim. Efficacy trials are short because that is how you detect whether a drug works at all, and because keeping people on placebo for years is often not ethical once a treatment is established. Long-term evidence exists; it is usually observational rather than randomised, which means it can tell you what happened to people who took the drug for years but not cleanly why. The honest statement is narrower and more useful: the randomised evidence that a drug works was mostly generated over weeks, and the question of what happens over years is answered by a weaker kind of study or, in several classes here, not answered at all.

which drug class has the weakest long-term randomised evidence in this library?

Two are tied for the starkest answer, but for different reasons, and the difference matters. Across 10 SNRI records (duloxetine, venlafaxine) not a single record states how long any trial ran — and duloxetine has no individual randomised trial in the library at all, only meta-analyses. Zolpidem is emptier still and it would be unfair to report it as a failure of disclosure: there is no individual zolpidem trial in this library to state a duration. Its 5 records are two meta-analyses, two FDA safety communications and one observational review. So the honest sentence is "the library has no zolpidem trial", not "zolpidem trials hide their length". That is not the same as saying those trials were short — it means the duration was never carried into the record, at any level, including the meta-analyses. Benzodiazepines are the sharpest contrast rather than the emptiest: the longest randomised horizon in this corpus is 10 weeks, while the corpus also contains a meta-analysis of people who had taken them for roughly a decade.

which class actually has long randomised evidence?

Mood stabilisers. BALANCE followed people with bipolar I disorder for up to two years, and a pooled analysis of two 18-month maintenance trials found lamotrigine roughly doubled the median time to the next mood episode (197 versus 86 days). Bipolar disorder is a relapse-prevention problem, so the field designed relapse-prevention trials. The longest horizon anywhere in the corpus — a median of 5.3 years — belongs to VITAL-DEP, a fish-oil prevention trial that found no benefit.

why does your data not just have a "trial duration" field?

Because it does not, and we are telling you rather than filling it in. The harvest records 18 fields per study — design, sample size, journal, year, DOI, PMID, funding, venue, preregistration, conflict disclosure, independence, trust score and more — and trial duration is not one of them. Zero of the 235 records carry a structured duration. Every horizon on this page was read out of the record’s own plain-English summary by hand and classified one record at a time. Where the summary states no horizon, the record is counted in the "not stated" row. Nothing was estimated from the drug, the indication, or what is typical for that design.

why do 15 of the 72 trials not state a horizon?

Because our summaries were written to answer "what did this find", not "how long did it run" — the gap is ours as much as the literature’s. EAGLES, VAST-D, both STAR*D arms and both MDMA phase 3 trials all have well-documented durations in their published papers; our records simply do not carry them. We are counting what the record says rather than what we could go and look up, because the moment we start filling gaps from memory the count stops being reproducible. Those 15 are shown as their own row rather than dropped from the denominator, and every one of them is listed by name in the per-trial table below.

do meta-analyses tell you how long the trials they pooled ran?

Almost never, and that is the more troubling finding. Of 93 pooled-evidence records in this corpus — 72 meta-analyses of randomised trials and 21 systematic reviews of randomised trials — only 8 state how long the underlying trials ran. That is 5 of the 72 meta-analyses and 3 of the 21 systematic reviews, so proportionally the systematic reviews do it more often, not less. The honourable exceptions are worth naming: the Cochrane amphetamines review gives a mean trial length of about 5 weeks, and the St John’s wort meta-analysis states that every included trial ran 4 to 12 weeks. Those are the two records in this library that tell a reader, unprompted, the time horizon behind the effect size they are quoting.

what should I actually ask my prescriber?

"How long were the trials that showed this works, and how long do you expect me to be on it?" If those two numbers are far apart — and for most psychiatric medications they will be — the follow-up question is what the plan is for the gap: what gets monitored, how often, and what would trigger a review of whether to continue. That is a normal conversation, not a confrontational one.

every trial we counted, and what its record actually says

All 72 individual randomised trials, sorted by horizon. The third column is quoted from the record rather than summarised, so you can disagree with a specific classification instead of with the total. Every row links to its study summary, where the primary source and the trust score live; where the subject has a medication approval page, the first column links to that too.

Counts: 47 at three months or less · 4 more than 3 up to 6 months · 3 more than 6 up to 12 months · 3 more than 12 months · 15 stating no horizon. Sums to 72.

Correction, 24 August 2026. The first version of this page counted 60 trials as stating a duration and 50 at three months or less. The correct figures are 57 and 47. Three records — both intravenous-lorazepam status-epilepticus trials and one of the two CBD public-speaking trials — were placed in the “same day” bucket because their design makes the answer obvious, which is exactly the inference this page’s own method forbids. They now sit in “no horizon stated”. Every other figure re-verified against the production corpus and is unchanged. The reviews figure also moved, from 10 of 93 to 8 of 93, and its split was backwards.
Every randomised trial in the Resolv evidence library, with the trial length the record itself states.
medication or topicstudywhat the record says about lengthbucketsource
5-htpComparative study of efficacy of l-5-hydroxytryptophan and fluoxetine in patients presenting with first depressive… (2013)randomised for eight weeks3 months or less/r/51716
5-htpThe Impact of 5-Hydroxytryptophan Supplementation on Cognitive Function and Mood in Singapore Older Adults: A… (2025)for 12 weeks3 months or less/r/51717
perinatalEffect of peer support on prevention of postnatal depression among high risk women: multisite randomised controlled… (2009)at 12 weeks3 months or less/r/51814
sexual healthA randomized trial comparing group mindfulness-based cognitive therapy with group supportive sex education and therapy… (2021)eight weekly sessions; still improved at 12 months3 months or less/r/51796
alprazolamAlprazolam in panic disorder and agoraphobia: results from a multicenter trial. I. Efficacy in short-term treatment (1988)8-week trial; panic-free at week 43 months or less/r/51593
aripiprazoleAntidepressant Augmentation versus Switch in Treatment-Resistant Geriatric Depression (2023)gain over 10 weeks3 months or less/r/51604
ashwagandhaAn investigation into the stress-relieving and pharmacological actions of an ashwagandha (Withania somnifera) extract:… (2019)daily for 60 days3 months or less/r/51720
ashwagandhaA prospective, randomized double-blind, placebo-controlled study of safety and efficacy of a high-concentration… (2012)for 60 days3 months or less/r/51721
ayahuascaRapid antidepressant effects of the psychedelic ayahuasca in treatment-resistant depression: a randomized… (2019)MADRS at day 73 months or less/r/51710
cannabidiolCannabidiol reduces the anxiety induced by simulated public speaking in treatment-naïve social phobia patients. (2011)single dose 90 minutes before challenge3 months or less/r/51724
clonazepamTreatment of social phobia with clonazepam and placebo. (1993)10-week double-blind trial3 months or less/r/51616
clonazepamPsychopharmacotherapy of panic disorder: 8-week randomized trial with clonazepam and paroxetine (2011)for 8 weeks3 months or less/r/51617
creatineA randomized, double-blind placebo-controlled trial of oral creatine monohydrate augmentation for enhanced response to… (2012)for eight weeks3 months or less/r/51733
creatineEfficacy and safety profile of oral creatine monohydrate in add-on to cognitive-behavioural therapy in depression: An… (2025)for eight weeks3 months or less/r/51734
dmtA short-acting psychedelic intervention for major depressive disorder: a phase IIa randomized placebo-controlled trial (2026)MADRS at 2 weeks; open-label to 3 months3 months or less/r/51711
escitalopramEscitalopram for older adults with generalized anxiety disorder: a randomized controlled trial (2009)responded within 12 weeks3 months or less/r/51629
esketamineEfficacy and Safety of Flexibly Dosed Esketamine Nasal Spray Combined With a Newly Initiated Oral Antidepressant in… (2019)4.0 MADRS points at day 283 months or less/r/51702
fluoxetineFluoxetine, cognitive-behavioral therapy, and their combination for adolescents with depression: Treatment for… (2004)after 12 weeks3 months or less/r/51631
hydroxyzineEfficacy and safety of hydroxyzine in the treatment of generalized anxiety disorder: a 3-month double-blind study (2002)12-week trial3 months or less/r/51636
hydroxyzineA multicentre double-blind comparison of hydroxyzine, buspirone and placebo in patients with generalized anxiety… (1998)4-week trial3 months or less/r/51637
ketamineKetamine versus ECT for Nonpsychotic Treatment-Resistant Major Depression (2023)noninferior to ECT over 3 weeks3 months or less/r/51704
ketamineAntidepressant effects of ketamine in depressed patients (2000)within 72 hours3 months or less/r/51706
kratomKratom and Pain Tolerance: A Randomized, Placebo-Controlled, Double-Blind Study. (2020)pain tolerance one hour after dose3 months or less/r/51740
l-theanineEffects of L-Theanine Administration on Stress-Related Symptoms and Cognitive Functions in Healthy Adults: A… (2019)four weeks each arm (crossover)3 months or less/r/51744
l-theanineAnti-Stress, Behavioural and Magnetoencephalography Effects of an L-Theanine-Based Nutrient Drink: A Randomised,… (2016)stress at 1 h, cortisol at 3 h3 months or less/r/51745
l-theanineA Randomized, Triple-Blind, Placebo-Controlled, Crossover Study to Investigate the Efficacy of a Single Dose of… (2021)alpha power at 3 h, cortisol at 1 h3 months or less/r/51746
lsdSingle Treatment With MM120 (Lysergide) in Generalized Anxiety Disorder: A Randomized Clinical Trial (2025)HAM-A at week 43 months or less/r/51707
lsdSafety and efficacy of lysergic acid diethylamide-assisted psychotherapy for anxiety associated with life-threatening… (2014)state anxiety at 2 months; sustained at 12 months3 months or less/r/51709
magnesiumRole of magnesium supplementation in the treatment of depression: A randomized clinical trial. (2017)six weeks on / six weeks off crossover3 months or less/r/51748
magnesiumEffects of Magnesium Supplementation on Unipolar Depression: A Placebo-Controlled Study and Review of the Importance… (2018)eight-week double-blind3 months or less/r/51750
methylene-blueEffects of post-session administration of methylene blue on fear extinction and contextual memory in adults with… (2014)fear at one-month follow-up3 months or less/r/51758
methylene-blueA controlled trial of methylene blue in severe depressive illness. (1987)three-week controlled trial3 months or less/r/51761
microdosingSelf-blinding citizen science to explore psychedelic microdosing. (2021)outcomes over four weeks3 months or less/r/51763
mirtazapineMirtazapine added to SSRIs or SNRIs for treatment resistant depression in primary care: phase III randomised placebo… (2018)at 12 weeks (>=6 wks prior SSRI)3 months or less/r/51661
nacThe efficacy of adjunctive N-acetylcysteine in major depressive disorder: a double-blind, randomized,… (2014)for 12 weeks (follow-up week 16); primary week 123 months or less/r/51767
nacN-acetylcysteine, a glutamate modulator, in the treatment of trichotillomania: a double-blind, placebo-controlled study. (2009)for 12 weeks3 months or less/r/51769
nacN-Acetylcysteine in the treatment of pediatric trichotillomania: a randomized, double-blind, placebo-controlled add-on… (2013)for 12 weeks3 months or less/r/51770
psilocybinTrial of Psilocybin versus Escitalopram for Depression (2021)6 weeks escitalopram; primary at week 63 months or less/r/51693
psilocybinSingle-Dose Psilocybin for a Treatment-Resistant Episode of Major Depression (2022)MADRS at week 33 months or less/r/51694
psilocybinEffects of Psilocybin-Assisted Therapy on Major Depressive Disorder: A Randomized Clinical Trial (2021)8-week waitlist; response weeks 1 and 43 months or less/r/51695
psilocybinSingle-Dose Psilocybin Treatment for Major Depressive Disorder: A Randomized Clinical Trial (2023)through week 63 months or less/r/51697
quetiapineA randomized, double-blind, placebo-controlled trial of quetiapine in the treatment of bipolar I or II depression (2005)treated for 8 weeks3 months or less/r/51669
sameA double-blind, randomized, placebo-controlled clinical trial of S-adenosyl-L-methionine (SAMe) versus escitalopram in… (2014)for 12 weeks3 months or less/r/51781
sameS-adenosyl methionine (SAMe) augmentation of serotonin reuptake inhibitors for antidepressant nonresponders with major… (2010)for six weeks3 months or less/r/51782
sertralineThe clinical effectiveness of sertraline in primary care and the role of depression severity and duration (PANDA): a… (2019)at 6 weeks3 months or less/r/51674
sildenafilTreatment of antidepressant-associated sexual dysfunction with sildenafil: a randomized controlled trial. (2003)6 weeks flexible-dose3 months or less/r/51818
sildenafilSildenafil treatment of women with antidepressant-associated sexual dysfunction: a randomized controlled trial. (2008)8 weeks3 months or less/r/51819
lsdLysergic Acid Diethylamide-Assisted Therapy in Patients With Anxiety With and Without a Life-Threatening Illness: A… (2023)STAI 16 weeks latermore than 3, up to 6 months/r/51708
methylene-blueMethylene blue treatment for residual symptoms of bipolar disorder: randomised crossover study. (2017)six-month double-blind crossovermore than 3, up to 6 months/r/51759
nacA randomised controlled trial of a mitochondrial therapeutic target for bipolar depression: mitochondrial agents,… (2019)randomised for 16 weeksmore than 3, up to 6 months/r/51768
psilocybinPsilocybin produces substantial and sustained decreases in depression and anxiety in patients with life-threatening… (2016)80% improved at 6-month follow-upmore than 3, up to 6 months/r/51696
fluoxetineMaintenance or Discontinuation of Antidepressants in Primary Care (2021)relapsed within a yearmore than 6, up to 12 months/r/51634
psilocybinPercentage of Heavy Drinking Days Following Psilocybin-Assisted Psychotherapy vs Placebo in the Treatment of Adult… (2022)12 wks psychotherapy; drinking over 32 weeksmore than 6, up to 12 months/r/51698
sertralineMaintenance or Discontinuation of Antidepressants in Primary Care (2021)relapsed within a yearmore than 6, up to 12 months/r/51676
lithiumLithium plus valproate combination therapy versus monotherapy for relapse prevention in bipolar I disorder (BALANCE):… (2010)followed for up to two yearsmore than 12 months/r/51645
methylene-blueA two-year double-blind crossover trial of the prophylactic effect of methylene blue in manic-depressive psychosis. (1986)one year high dose + one year low dosemore than 12 months/r/51760
omega-3Effect of Long-term Supplementation With Marine Omega-3 Fatty Acids vs Placebo on Risk of Depression or Clinically… (2021)median 5.3 yearsmore than 12 months/r/51772
sexual healthInternet-based brief sex therapy for heterosexual men with sexual dysfunctions: a randomized controlled pilot trial. (2009)nothing stated in the recordno horizon stated/r/51805
aripiprazoleEfficacy, safety, and tolerability of augmentation pharmacotherapy with aripiprazole for treatment-resistant… (2015)nothing stated in the recordno horizon stated/r/51603
bupropionNeuropsychiatric safety and efficacy of varenicline, bupropion, and nicotine patch in smokers with and without… (2016)nothing stated in the recordno horizon stated/r/51608
bupropionEffect of Antidepressant Switching vs Augmentation on Remission Among Patients With Major Depressive Disorder… (2017)nothing stated in the recordno horizon stated/r/51609
buspironeMedication augmentation after the failure of SSRIs for depression (2006)nothing stated in the recordno horizon stated/r/51613
cannabidiolCannabidiol presents an inverted U-shaped dose-response curve in a simulated public speaking test. (2019)nothing stated in the recordno horizon stated/r/51725
clonazepamClonazepam in the treatment of panic disorder with or without agoraphobia: a dose-response study of efficacy, safety,… (1997)nothing stated in the recordno horizon stated/r/51615
esketamineEfficacy of Esketamine Nasal Spray Plus Oral Antidepressant Treatment for Relapse Prevention in Patients With… (2019)nothing stated in the recordno horizon stated/r/51703
lorazepamA comparison of four treatments for generalized convulsive status epilepticus (1998)nothing stated in the recordno horizon stated/r/51649
lorazepamA comparison of lorazepam, diazepam, and placebo for the treatment of out-of-hospital status epilepticus (2001)nothing stated in the recordno horizon stated/r/51650
microdosingMicrodosing with psilocybin mushrooms: a double-blind placebo-controlled study. (2022)nothing stated in the recordno horizon stated/r/51764
midomafetamineMDMA-assisted therapy for severe PTSD: a randomized, double-blind, placebo-controlled phase 3 study (2021)nothing stated in the recordno horizon stated/r/51699
midomafetamineMDMA-assisted therapy for moderate to severe PTSD: a randomized, placebo-controlled phase 3 trial (2023)nothing stated in the recordno horizon stated/r/51700
paroxetineRestoring Study 329: efficacy and harms of paroxetine and imipramine in treatment of major depression in adolescence (2015)nothing stated in the recordno horizon stated/r/51664
venlafaxineBupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression (2006)nothing stated in the recordno horizon stated/r/51685

sources

  1. Lewis G, et al. Maintenance or Discontinuation of Antidepressants in Primary Care (ANTLER). New England Journal of Medicine, 2021. PMID 34587384. Randomised discontinuation trial, 478 patients, 1-year horizon. Accessed 2026-08-19. https://doi.org/10.1056/NEJMoa2106356 read our summary of this study →
  2. Geddes JR, et al. (BALANCE investigators and collaborators). Lithium plus valproate combination therapy versus monotherapy for relapse prevention in bipolar I disorder (BALANCE): a randomised open-label trial. The Lancet, 2010. PMID 20092882. Followed for up to two years. Accessed 2026-08-19. https://doi.org/10.1016/S0140-6736(09)61828-6
  3. Goodwin GM, et al. A pooled analysis of 2 placebo-controlled 18-month trials of lamotrigine and lithium maintenance in bipolar I disorder. Journal of Clinical Psychiatry, 2004. PMID 15096085. Median time to next mood episode 197 versus 86 days. Accessed 2026-08-19. https://doi.org/10.4088/jcp.v65n0321 read our summary of this study →
  4. Okereke OI, et al. Effect of Long-term Supplementation With Marine Omega-3 Fatty Acids vs Placebo on Risk of Depression or Clinically Relevant Depressive Symptoms and on Change in Mood Scores (VITAL-DEP). JAMA, 2021. PMID 34932079. 18,353 adults, median 5.3 years — the longest randomised horizon in this corpus. Accessed 2026-08-19. https://doi.org/10.1001/jama.2021.21187 read our summary of this study →
  5. Barker MJ, et al. Cognitive effects of long-term benzodiazepine use: a meta-analysis. CNS Drugs, 2004. PMID 14731058. Pooled 13 observational studies of people who had taken benzodiazepines for about 10 years on average. Accessed 2026-08-19. https://doi.org/10.2165/00023210-200418010-00004 read our summary of this study →
  6. Davidson JR, et al. Treatment of social phobia with clonazepam and placebo. Journal of Clinical Psychopharmacology, 1993. PMID 8120156. 10-week double-blind trial, 75 outpatients — the longest randomised benzodiazepine horizon in this corpus. Accessed 2026-08-19. https://doi.org/10.1097/00004714-199312000-00008 read our summary of this study →
  7. Cortese S, et al. Comparative efficacy and tolerability of medications for attention-deficit hyperactivity disorder in children, adolescents, and adults: a systematic review and network meta-analysis. Lancet Psychiatry, 2018. PMID 30097390. 133 double-blind trials, 12-week efficacy horizon. Accessed 2026-08-19. https://doi.org/10.1016/S2215-0366(18)30269-4
  8. Castells X, et al. Amphetamines for attention deficit hyperactivity disorder (ADHD) in adults. Cochrane Database of Systematic Reviews, 2018. PMID 30091808. 19 randomised trials, 2,521 adults, mean trial length 5.3 weeks — one of only two records in this corpus that states the pooled trial length. Accessed 2026-08-19. https://doi.org/10.1002/14651858.CD007813.pub3 read our summary of this study →
  9. Ng QX, Venkatanarayanan N, Ho CY. Clinical use of Hypericum perforatum (St John's wort) in depression: A meta-analysis. Journal of Affective Disorders, 2017. PMID 28064110. 27 trials, 3,808 patients; every included trial ran only 4 to 12 weeks. Accessed 2026-08-19. https://doi.org/10.1016/j.jad.2016.12.048 read our summary of this study →
  10. Tiihonen J, et al. Real-World Effectiveness of Antipsychotic Treatments in a Nationwide Cohort of 29,823 Patients With Schizophrenia. JAMA Psychiatry, 2017. PMID 28593216. Followed for up to 7.5 years. Accessed 2026-08-19. https://doi.org/10.1001/jamapsychiatry.2017.1322 read our summary of this study →
  11. Swanson JM, et al. Young adult outcomes in the follow-up of the multimodal treatment study of attention-deficit/hyperactivity disorder (MTA). Journal of Child Psychology and Psychiatry, 2017. PMID 28295312. 14-month randomised phase, observational follow-up to about age 25. Accessed 2026-08-19. https://doi.org/10.1111/jcpp.12684 read our summary of this study →
  12. Lorazepam tablets, FDA-approved label. DailyMed structured product label 787240e9-67b5-4af3-b16d-7cf314e3f2e3 (ANI Pharmaceuticals), verified 2026-08-17: indicated for the management of anxiety disorders or the short-term relief of anxiety symptoms; effectiveness beyond 4 months not systematically established. https://dailymed.nlm.nih.gov/dailymed/
  13. Hydroxyzine hydrochloride tablets, FDA-approved label. DailyMed structured product label b4d274ce-d616-4835-80ce-76270524ed36 (Cardinal Health), verified 2026-08-17: effectiveness beyond 4 months has not been established. https://dailymed.nlm.nih.gov/dailymed/
  14. Resolv study harvest, 2026-08-17 and 2026-08-18. 235 study records, 216 distinct papers, 46 source files, 72 of them individual randomised trials. The schema records 18 fields per study; trial duration is not one of them, which is why every horizon on this page was hand-read from the record text. The underlying records are published through the Resolv resource library. https://www.resolv.social/resources