quetiapine (Seroquel)

the boxed warning, part one: deaths among older people with dementia-related psychosis

the first part of the boxed warning says that elderly patients with dementia-related psychosis treated with antipsychotic drugs are at an increased risk of death, and that quetiapine is not approved for them [label, boxed warning and 5.1]. the number behind it is a pooled analysis of 17 placebo-controlled trials, mostly of atypical antipsychotics and mostly about 10 weeks long: the death rate was about 4.5% on drug against about 2.6% on placebo, 1.6 to 1.7 times higher. most of the deaths were cardiovascular, such as heart failure or sudden death, or infectious, such as pneumonia [label 5.1].

two things about it are easy to miss. it is a class warning, written from trials of several antipsychotics rather than of quetiapine alone, and the label adds that observational studies suggest the older, conventional antipsychotics may carry the same risk [label 5.1]. and it is specific to one population: older people with psychosis caused by dementia, a group the drug was never approved for. a related warning notes that in trials of risperidone, aripiprazole and olanzapine in elderly people with dementia there were more strokes and transient ischaemic attacks, including fatal ones, than on placebo [label 5.3].

the boxed warning, part two: suicidal thoughts and behaviors, by age

the second part is the antidepressant class warning, which quetiapine carries because it is approved for bipolar depression. it says antidepressants increased the risk of suicidal thoughts and behavior in children, adolescents and young adults in short-term studies, and that everyone started on one should be watched closely for worsening, especially in the first months and after a dose change [label, boxed warning and 5.2]. the label states the difference against placebo per 1,000 people treated: under 18, 14 additional people with suicidal thoughts or behaviors; 18 to 24, 5 additional; 25 to 64, 1 fewer; 65 and older, 6 fewer [label 5.2, table 2]. the pooled trials were of antidepressants generally, 24 pediatric trials of 9 drugs in more than 4,400 patients and 295 adult trials of 11 drugs in more than 77,000, not of quetiapine specifically. no suicides occurred in the pediatric trials; the adult trials had too few to conclude anything about suicide itself [label 5.2].

the label adds two instructions. because a depressive episode can be the first sign of bipolar disorder, anyone with depressive symptoms should be screened for bipolar risk before starting, with a family history of suicide, bipolar disorder and depression [label 5.2]. and prescriptions should be written for the smallest quantity of tablets consistent with good management, to reduce the risk of overdose [label 5.2]. quetiapine is not approved for anyone under ten [label, boxed warning].

the metabolic warning, with the label’s tables

warning 5.5 is the one most people on this drug will live with. the label says atypical antipsychotics as a class cause hyperglycemia and diabetes, undesirable lipid changes and weight gain, that each drug has its own profile, and that in some patients more than one of the three worsened at once. it asks for fasting glucose testing at the start and periodically in anyone with diabetes risk factors, fasting lipid testing at the start and periodically, and regular weight monitoring [label 5.5]. the tables are below, as the label prints them.

• weight, adults [5.5, table 1]: the share gaining 7% or more of body weight was 23% against 6% on placebo in the schizophrenia trials (up to 6 weeks; 391 against 206 people), 21% against 7% in the 12-week bipolar mania monotherapy trials, 13% against 4% in the 3-week adjunct trials, and 8% against 2% in the 8-week bipolar depression trials (554 against 295).
• weight, children and adolescents [5.5, table 2]: 21% against 7% gained 7% or more in the 6-week schizophrenia trial (mean 2.0 kg against a loss of 0.4 kg) and 12% against 0% in the 3-week mania trial (1.7 kg against 0.4 kg). in the open-label extension that followed, 241 of 380 completed 26 weeks; mean weight gain was 4.4 kg, 45% had gained 7% or more, and after adjusting for normal growth 18.3% met the label’s threshold for a clinically significant rise in body-mass index.
• blood sugar, adults [5.5, table 3]: in placebo-controlled trials up to 12 weeks, 2.4% of people who started with a normal fasting glucose crossed the diabetes threshold of 126 mg/dl, against 1.4% on placebo (2,907 against 1,346). in the two long-term bipolar maintenance trials, mean fasting glucose rose 5.0 mg/dl on quetiapine over an average 213 days and fell 0.05 mg/dl on placebo; the exposure-adjusted rate of a reading at or above 126 mg/dl was 18.0 per 100 patient-years against 9.5. the label is candid that the relationship between these drugs and diabetes "is not completely understood" and that precise risk estimates are not available.
• lipids, adults [5.5, table 4]: total cholesterol crossed 240 mg/dl in 18% against 7% in the 6-week schizophrenia trials and 9% against 6% in the 8-week bipolar depression trials; triglycerides crossed 200 mg/dl in 22% against 16% and 14% against 9%; ldl crossed 160 mg/dl in 6% against 5% (bipolar depression only; the schizophrenia registration trials did not measure it); hdl fell to 40 mg/dl or below in 14% on each arm.
• lipids, children and adolescents [5.5, table 5]: total cholesterol crossed 200 mg/dl in 12% against 2% (schizophrenia, 6 weeks) and 10% against 3% (mania, 3 weeks); triglycerides crossed 150 mg/dl in 17% against 8% and 22% against 13%.
• what the tables cannot show: they stop at 3 to 12 weeks for most indications, and the maintenance trials ran a median of about seven months. the one-year and ten-year numbers are in the research section below.

the other warnings, in plain words

the warnings section runs from 5.1 to 5.20; the newest, on prolactin, was revised in january 2025 [label, recent major changes]. beyond the boxed warnings and the metabolic one:

• neuroleptic malignant syndrome [5.4]: a rare, potentially fatal reaction to antipsychotics, with very high temperature, rigid muscles, confusion and unstable pulse or blood pressure. the label says stop the drug at once and treat in hospital. rare cases have been reported with quetiapine.
• tardive dyskinesia [5.6]: involuntary, potentially irreversible movements, most often after long treatment and high cumulative doses but sometimes after brief low-dose treatment or after stopping. the label says chronic antipsychotic treatment should be reserved for chronic illness with no equally effective, less harmful alternative, at the smallest dose and shortest duration that works, with the need reassessed periodically.
• low blood pressure and fainting [5.7] and falls [5.8]: orthostatic hypotension, especially during the first days of dose increase; fainting in 1% of patients (28 of 3,265) against 0.2% on placebo. the label asks for a fall-risk assessment in anyone with conditions or medicines that add to drowsiness or unsteadiness.
• raised blood pressure in children and adolescents [5.9]: a systolic rise of 20 mmhg or more in 15.2% against 5.5% on placebo and a diastolic rise of 10 mmhg or more in 40.6% against 24.5%. blood pressure is to be measured at the start and periodically. this did not occur in adults.
• low white cells [5.10]: leukopenia, neutropenia and agranulocytosis, including fatal cases and cases with no risk factor. a neutrophil count below 1.0 occurred in 0.3% against 0.1% on placebo [6.1]. anyone with a low white count or a history of drug-induced neutropenia needs frequent blood counts in the first months.
• cataracts [5.11]: lens changes seen in chronic dog studies and in people on long-term treatment, cause not established. the label recommends a lens examination at the start of treatment and every six months during chronic treatment.
• heart rhythm [5.12]: no persistent qt lengthening in the trials, but the label says the effect was never measured in a thorough qt study, that qt prolongation has been reported in overdose, and that quetiapine should be avoided with other qt-prolonging drugs (it names several antiarrhythmics, antipsychotics, antibiotics and methadone) and in people with low potassium or magnesium or a slow heart rate.
• seizures [5.13]: 0.5% of patients (20 of 3,490) against 0.2% on placebo.
• thyroid [5.14]: dose-related falls in thyroxine of about 20% at the high end of the dose range, mostly reversing when the drug stops; tsh rose above 5 in 3.2% against 2.7% on placebo overall and in 12% against 7% when added to lithium or divalproex. the label asks for tsh and free t4 at baseline and follow-up, and some patients needed thyroid replacement.
• prolactin [5.15, revised january 2025]: a clinically significant shift in 3.6% of adults against 2.6% on placebo; in children and adolescents, 13.4% against 4% of boys and 8.7% against 0% of girls. high prolactin can suppress reproductive hormones, cause milk production, missed periods, breast enlargement in men and impotence, and over years lower bone density.
• drowsiness and impaired thinking [5.16]: somnolence in 18% against 11% in schizophrenia trials, 16% against 4% in mania monotherapy, 34% against 9% as an adjunct, and 57% against 15% in bipolar depression. the label says not to drive or operate machinery until the effect is known, and that somnolence may lead to falls.
• heat [5.17], swallowing [5.18]: antipsychotics can impair the body’s cooling and have been linked to aspiration pneumonia, a common cause of death in older people with advanced dementia.
• stopping [5.19]: insomnia, nausea and vomiting after abrupt cessation; in the extended-release trials 12.1% had at least one discontinuation symptom against 6.7% on placebo, usually resolving within a week. "gradual withdrawal is advised."
• anticholinergic effects [5.20]: quetiapine’s active metabolite norquetiapine blocks muscarinic receptors, which adds to dry mouth, constipation and urinary retention, and the label reports intestinal obstruction, including fatal cases in people on several motility-slowing drugs. caution with other anticholinergic medicines.
• interactions and food [2.5, 2.6, 7.1]: quetiapine is cleared by the liver enzyme cyp3a4. with a strong inhibitor (ketoconazole, itraconazole, ritonavir, nefazodone) the label says the dose should be cut to one sixth; with a strong inducer taken for more than a week or two (phenytoin, carbamazepine, rifampin, st john’s wort) raised up to fivefold. grapefruit disables the same enzyme in the gut wall, and a review of grapefruit interactions lists quetiapine among the drugs affected [grapefruit]. the label also warns of false positives on urine screens for methadone and tricyclics [6.1].

the low-dose use for sleep: not an approved use, and what the studies found

quetiapine has no approved indication for insomnia, anxiety or any sleep problem [label 1]. the only places sleepiness appears in the label are the side-effect tables and warning 5.16, where it is the most common adverse reaction, at 57% of people on 300 mg for bipolar depression [label 5.16; 6.1, table 11]. a prescription for sleep at 25 to 100 mg is off-label: legal for a prescriber to write, but written without a trial the fda has reviewed for that purpose.

what the trials say. a 2016 systematic review of atypical antipsychotics for insomnia found a single randomised trial of quetiapine for primary insomnia, in 13 people, with no significant improvement in sleep time, time to fall asleep or sleep satisfaction against placebo [thompson-2016]. for generalized anxiety the evidence is stronger: a cochrane review of 11 trials in more than 4,000 people found people about twice as likely to respond to quetiapine as to placebo, with more weight gain, more sedation and more dropouts from side effects, and the reviewers noted the use is not approved [depping-2010].

what low doses do to the body. the usual argument for low-dose use is that the metabolic effects belong to antipsychotic doses. a danish registry study tested that in 106,711 people who started quetiapine at 50 mg or less between 2008 and 2018 and had blood tests the year before and the year after. across everyone, triglycerides rose about 5% and hdl fell about 2%; among people whose results were normal at the start, hba1c, total cholesterol and ldl all rose, and the effect grew with dose, which argues against low doses being metabolically free [hojlund-2023]. the same group compared 60,566 low-dose quetiapine users with 454,567 users of z-drug sleeping tablets over 890,198 person-years: low-dose quetiapine went with a 13% higher rate of heart attack, stroke or cardiovascular death on intention-to-treat analysis and 52% higher among people who stayed on it, with cardiovascular death 26% and 90% higher respectively, and the excess was larger in women and in people over 65; against ssris the excess was 42%. the authors wrote that off-label low-dose use for sedation "should be discouraged" [hojlund-2022]. both are observational: people prescribed quetiapine for sleep may differ from people prescribed a z-drug in ways the records cannot fully adjust for.

what the approval trials measured, and for how long

schizophrenia rests on three 6-week trials of inpatients. one gave five fixed doses to 361 people: the four highest beat placebo on the brief psychiatric rating scale, the effect was maximal at 300 mg a day, and 150 to 750 mg were "generally indistinguishable". one titrated 286 people high or low: only the high-dose group, averaging 500 mg, beat placebo. the third, in 618 people, compared 450 mg with 50 mg and had no placebo arm. one trial carried a single fixed-dose haloperidol arm that the label itself calls inadequate for a valid comparison [label 14.1]. a 6-week trial in 222 adolescents aged 13 to 17 added 400 and 800 mg arms, each beating placebo by 8 to 9 points on the positive and negative syndrome scale [label 14.1, table 19]. the label states that "the effectiveness of quetiapine tablets for the maintenance treatment of schizophrenia has not been systematically evaluated in controlled clinical trials" [label 1.1].

mania rests on two 12-week monotherapy trials (300 and 299 people) and one 3-week add-on trial (170 people). the placebo-subtracted change on the young mania rating scale was 4.0 and 7.9 points in the monotherapy trials and 3.8 points as an add-on; in the same trials haloperidol and lithium, included as comparators, scored 7.4 and 8.5 [label 14.2, table 3]. a second add-on trial of 200 people did not beat placebo, "possibly due to a higher placebo effect" [label 14.2]. a 3-week trial in children and adolescents aged 10 to 17 beat placebo at 400 and 600 mg.

bipolar depression rests on two identical 8-week trials in 1,045 people with bipolar i or ii at fixed doses of 300 or 600 mg. quetiapine beat placebo by 4 to 6.5 points on the montgomery-asberg scale from the first week, and "no additional benefit was seen with the 600 mg dose" [label 14.2, table 4]. the first of these, in 542 people, reported about 58% responding against 36% on placebo [calabrese-2005]. maintenance as an add-on rests on two randomised-withdrawal trials in 1,326 people who were already stable on quetiapine plus lithium or divalproex for an average of 15 weeks before being randomised to continue or switch to placebo; half of the quetiapine group had dropped out by day 280 and half of the placebo group by day 117 [label 14.2]. monotherapy maintenance has not been evaluated in controlled trials [label 1.2].

side effects, with the label’s numbers

in the 3- to 12-week schizophrenia and mania monotherapy trials (719 against 404 on placebo): headache 21% against 14%, agitation 20% against 17%, somnolence 18% against 8%, dizziness 11% against 5%, dry mouth 9% against 3%, constipation 8% against 3%, weight gain 5% against 1%, raised alt (a liver enzyme) 5% against 1%, dyspepsia 5% against 1%, postural hypotension 4% against 1% [label 6.1, table 9]. alt rose above three times the normal limit in about 6% against 1% in the schizophrenia trials, usually in the first three weeks and usually reversing during treatment [label 6.1].

in the 8-week bipolar depression trials (698 against 347): somnolence 57% against 15%, dry mouth 44% against 13%, dizziness 18% against 7%, constipation 10% against 4%, fatigue 10% against 8%, increased appetite 5% against 3%, lethargy 5% against 2%, akathisia 4% against 1%, weight increase 4% against 1%, extrapyramidal disorder 3% against 1% [label 6.1, table 11]. dropouts for side effects were 12.3% at 300 mg and 19.0% at 600 mg against 5.2% on placebo [label 6.1].

in children and adolescents with mania (3 weeks, 400 or 600 mg): somnolence 50% to 57% against 14%, dizziness 17% to 19% against 2%, increased appetite 9% to 10% against 1%, weight increase 6% against 0% [label 6.1, table 14]. on movement side effects the label is favourable: in the five-dose schizophrenia trial, parkinsonism, akathisia and dystonia were no more common on any dose than on placebo [label 6.1, table 12], which is consistent with the network meta-analysis below. in overdose, survival has been reported up to 30 grams, and one death followed 13.6 grams taken alone [label 10.1].

what the research says, in both directions

in favour. in a network meta-analysis of 402 randomised trials covering 53,463 adults with schizophrenia, quetiapine ranked mid-field of 32 antipsychotics for symptom reduction, with among the lowest rates of movement side effects, more sedation, and a small qtc prolongation of about 3 milliseconds against placebo [huhn-2019]. in bipolar depression it is one of few drugs with placebo-controlled evidence, and 300 mg did as well as 600 mg [calabrese-2005; label 14.2]. for generalized anxiety, an unapproved use, response was about twice placebo in a cochrane review [depping-2010].

against, or smaller than it sounds. in a swedish cohort of 29,823 people with schizophrenia followed up to 7.5 years and compared each with themselves on and off treatment, oral quetiapine had the highest rehospitalisation risk of the commonly used antipsychotics, barely better than no antipsychotic at all (hazard ratio 0.91), while clozapine and long-acting injections roughly halved the risk; several authors had industry ties and the sponsor’s product ranked first [tiihonen-2017; tiihonen-2017-plain]. in a finnish cohort of 18,018 people with bipolar disorder, quetiapine was the most prescribed drug and lowered the risk of rehospitalisation by less than a tenth, where lithium lowered it by about a third [lahteenvuo-2018]. in 6,671 uk patients on maintenance treatment, self-harm rates were roughly 40% to 50% higher on quetiapine, valproate or olanzapine than on lithium, with the usual caution that people prescribed different drugs differ [hayes-2016]. an english case-control study of 58,769 people with dementia found the highest cumulative exposure to strongly anticholinergic drugs, a group that includes some antipsychotics, associated with about 49% higher odds of dementia, 70% for the antipsychotic class; early dementia leading to prescriptions is the standard alternative explanation [coupland-2019].

the warnings, measured. the metabolic warning’s long-term size is in the danish studies above: measurable lipid and glucose shifts within a year at 50 mg or less, rising with dose [hojlund-2023], and a 13% to 52% higher rate of major cardiovascular events at low doses against sleeping tablets [hojlund-2022]. the label’s own tables show the first 3 to 12 weeks [label 5.5]. the boxed dementia warning’s size is the label’s: about two extra deaths per hundred people over ten weeks [label 5.1].

questions people ask

what does the seroquel (quetiapine) black box warning mean?

it has two parts. first, in 17 placebo-controlled trials of antipsychotics in elderly people with dementia-related psychosis, about 4.5% died over roughly ten weeks against about 2.6% on placebo, 1.6 to 1.7 times the rate, mostly from heart or infectious causes; quetiapine is not approved for that group. second, antidepressants increased suicidal thoughts and behavior in people under 25 in short-term trials (14 extra cases per 1,000 under 18, 5 extra at 18 to 24, 1 fewer at 25 to 64, 6 fewer at 65 and over), and quetiapine carries that warning because it is approved for bipolar depression. neither part is an instruction to stop the drug; both describe who should be watched, and for what.

is quetiapine approved for sleep?

no. quetiapine is approved for schizophrenia, bipolar mania, bipolar depression and bipolar maintenance as an add-on. a prescription for insomnia or anxiety is off-label, which is legal but means the fda has not reviewed a trial for that use. the only randomised trial of quetiapine for primary insomnia found by a 2016 systematic review had 13 people and found no significant improvement in sleep. two danish registry studies of people on 50 mg or less found measurable rises in blood sugar and cholesterol within a year, growing with dose, and a 13% to 52% higher rate of heart attack, stroke or cardiovascular death than in people on z-drug sleeping tablets. for generalized anxiety, also unapproved, a cochrane review found response about twice placebo, with more weight gain and sedation.

how much weight do people gain on quetiapine?

in the label’s adult trials, 23% gained 7% or more of their body weight in 6 weeks against 6% on placebo in schizophrenia, 21% against 7% over 12 weeks in mania, and 8% against 2% over 8 weeks in bipolar depression. in children and adolescents, mean gain was 2.0 kg over 6 weeks against a loss of 0.4 kg, and in the open-label extension the mean gain at 26 weeks was 4.4 kg, with 45% gaining 7% or more. the label asks for regular weight checks and fasting glucose and lipids at the start and periodically. the registry study of low doses found lipid and blood sugar shifts even at 50 mg or less.

what are the most common quetiapine side effects?

in the schizophrenia and mania trials: headache 21% vs 14% on placebo, somnolence 18% vs 8%, dizziness 11% vs 5%, dry mouth 9% vs 3%, constipation 8% vs 3%, weight gain 5% vs 1%, raised liver enzymes 5% vs 1%. in the bipolar depression trials at 300 to 600 mg: somnolence 57% vs 15%, dry mouth 44% vs 13%, dizziness 18% vs 7%, constipation 10% vs 4%. in children with mania, somnolence was 50% to 57% vs 14%. movement side effects were no more common than placebo in the fixed-dose schizophrenia trial, which is the drug’s main advantage in the network meta-analysis.

what happens if you stop quetiapine suddenly?

the label says insomnia, nausea and vomiting have been described after abrupt cessation of atypical antipsychotics including quetiapine; in the extended-release trials, 12.1% had at least one discontinuation symptom against 6.7% on placebo, usually resolving within a week. it says "gradual withdrawal is advised." separately, the illness the drug was treating can return: in the maintenance trials, half of the people switched to placebo had left the trial by day 117, against day 280 for those kept on quetiapine. a plan for reducing or stopping belongs with a prescriber.

is seroquel (quetiapine) a controlled substance or addictive?

it is not a controlled substance. the label says quetiapine "has not been systematically studied, in animals or humans, for its potential for abuse, tolerance, or physical dependence", that the trials showed no drug-seeking behaviour, and that this is not enough to predict misuse once marketed, so people with a history of drug misuse should be watched for tolerance, dose escalation and drug-seeking. the discontinuation symptoms above are a separate matter from addiction.

sources for this section

the library rows link to our one-page summary of each study, with the original paper, its funding and our evidence grade on that page.

1. quetiapine tablets prescribing information, fda label on dailymed (set id 711f42c5-0719-46de-81fb-3d03801b9ed9, version 3, effective august 25, 2026; lupin, repackaged by preferred pharmaceuticals): boxed warning; sections 1, 2.5, 2.6, 5.1 to 5.20 with tables 1 to 5 and 8, 6.1 tables 9 to 14, 7.1, 9.1, 9.2, 10.1, 14.1 and 14.2 with tables 19, 3 and 4, recent major changes
2. drugs@fda, nda 020639 (seroquel), approved september 26, 1997
3. ahrq medical expenditure panel survey, prescribed medicines public-use files 2014 (hc-168a) and 2024 (hc-254a): people with at least one quetiapine purchase, weighted; computed by veisund, method on the usage section of this page
4. huhn m, et al. comparative efficacy and tolerability of 32 oral antipsychotics for the acute treatment of adults with multi-episode schizophrenia: a systematic review and network meta-analysis. lancet 2019 (library row)
5. calabrese jr, et al. a randomized, double-blind, placebo-controlled trial of quetiapine in the treatment of bipolar i or ii depression. american journal of psychiatry 2005 (bolder i, the registration trial; funding statement paywalled) (library row)
6. depping am, et al. second-generation antipsychotics for anxiety disorders. cochrane database of systematic reviews 2010 (library row)
7. thompson w, et al. atypical antipsychotics for insomnia: a systematic review. sleep medicine 2016 (library row)
8. højlund m, støvring h, andersen k, correll cu, hallas j. impact of low-dose quetiapine-use on glycosylated hemoglobin, triglyceride and cholesterol levels. acta psychiatrica scandinavica 2023 (public funding; authors’ unrelated industry ties recorded on the row) (library row)
9. højlund m, andersen k, ernst mt, correll cu, hallas j. use of low-dose quetiapine increases the risk of major adverse cardiovascular events: results from a nationwide active comparator-controlled cohort study. world psychiatry 2022;21:444-451 (pubmed 36073694; abstract read; not yet a library row)
10. tiihonen j, et al. real-world effectiveness of antipsychotic treatments in a nationwide cohort of 29,823 patients with schizophrenia. jama psychiatry 2017 (funded by janssen-cilag, which markets competing antipsychotics) (library row)
11. the same swedish cohort, written up for readers: about half the rehospitalisation risk on clozapine or a long-acting injection, and the cautions about authorship and stopping (library row)
12. lähteenvuo m, et al. real-world effectiveness of pharmacologic treatments for the prevention of rehospitalization in a finnish nationwide cohort of patients with bipolar disorder. jama psychiatry 2018 (library row)
13. hayes jf, et al. self-harm, unintentional injury, and suicide in bipolar disorder during maintenance mood stabilizer treatment: a uk population-based electronic health records study. jama psychiatry 2016 (library row)
14. coupland cac, et al. anticholinergic drug exposure and the risk of dementia: a nested case-control study. jama internal medicine 2019 (library row)
15. bailey dg, dresser g, arnold jm. grapefruit-medication interactions: forbidden fruit or avoidable consequences? cmaj 2013 (narrative review) (library row)
16. the quetiapine legal and safety record, a separate page on this site: court and regulator records, kept apart from the label and trial facts here

Weight gain on the quetiapine label

The quetiapine FDA label names weight gain in its Adverse Reactions section, in the list below.

Most-reported reactions

• Drug ineffective12,857
• Off label use12,337
• Toxicity to various agents10,154
• Insomnia9,282
• Fatigue9,208
• Weight increased8,383
• Diabetes mellitus8,355
• Drug interaction8,347
• Somnolence8,322
• Nausea8,284