Treatment of Neuroleptic-Induced Akathisia With the 5-HT2A
*Rafael Stryjer, *†Rael D. Strous, *†Faina Bar, ‡Michael Poyurovsky, †§Abraham Weizman, and
*Beer Yaakov Mental Health Center, †Sackler Faculty of Medicine, Tel Aviv University, ‡Tirat Carmel Mental Health Center,
§Geha Mental Health Center, Israel
Summary: Akathisia is a common and distressful extrapyramidal adverse side effect
usually resulting from the use of antipsychotic medications. Early management of aka-
thisia is important because it may be associated with poor treatment response and medi-
cation noncompliance. Unfortunately many patients fail to respond to standard manage-
ment of akathisia. In addition to dopaminergic mechanisms, it has been hypothesized
This open-label pilot study investigates the efficacy of trazodone in the management of
akathisia. Nine female patients with a score of at least “mild akathisia” on the Barnes
Akathisia Scale, and receiving a stable dose of antipsychotic medication, were admin-
istered trazodone, titrated up to a dosage of 100 mg/day over a period of 5 days. The
improvement was noted in symptomatology of anxiety, depression, and psychosis.
cation for the treatment of antipsychotic medication-induced akathisia. Further study in
the context of a double-blind, placebo-controlled trial is mandated to substantiate these
preliminary findings. Key Words: Akathisia—Antipsychotic medication—Trazodone
Akathisia is a common and distressful extrapyrami-
dal adverse effect usually resulting from the use of an-
tipsychotic medication and, less frequently, from the
use of antidepressant medications. Early recognition
and adequate management of neuroleptic-induced aka-
thisia (NIA) is of importance because it may be associ-
behavior, and noncompliance with antipsychotic drug
treatment (1). Although many agents used in the treat-
ment of NIA, including beta blockers, anticholinergic
medications, benzodiazepines, and dopamine agonists,
are successful in the management of akathisia, many
patients nevertheless fail to respond (2) and the prob-
lem remains one of considerable distress. Because
native methods of managing this important clinical
Because the beta blocker propranolol (considered to
be the most efficacious antiakathisia medication) is
thought to act by means of 5-hydroxytriptamine (5-
HT)1Apresynaptic antagonistic activity (3), it has been
hypothesized that serotonin may play a prominent role
in the pathophysiology of NIA. This is evidenced fur-
ther by the observation that a reduction in brain 5-HT
neurotransmission (5-HT2Aantagonists, 5-HT1Aago-
ramidal effects induced by D2receptor antagonists (4).
Moreover, treatment with atypical antipsychotic
agents, which demonstrate 5-HT2antagonistic activity
in addition to D2receptor blockade, have resulted in a
reduced rate of extrapyramidal effects, akathisia, and
tardive dyskinesia (5). Along these lines, it has been
suggested that 5-HT2receptor antagonism might pre-
vent the onset or mitigate the severity of NIA by coun-
Address correspondence and reprint requests to Dr. Rael Strous, Beer
Yaakov Mental Health Center, PO Box 1, Beer Yaakov 70350, Israel;
Vol. 26, No. 3, pp. 137–141
© 2003 Lippincott Williams & Wilkins, Inc., Philadelphia
teracting the dopamine D2 blockade (6). Further sup-
porting the role of serotonin in the pathophysiology of
akathisia is the knowledge that selective serotonin re-
mission, have a propensity to induce “akathisialike
syndrome” (7). The role of serotonin in inducing aka-
thisia supports the use of agents with 5-HT2antagonis-
tic effects in the management of NIA. To date, it ap-
of NIA (1,5,8–13), including two double-blind studies.
One study demonstrated the efficacy of mianserin in
the treatment of NIA (1), and the second demonstrated
similar efficacy between cyproheptadine and propran-
olol for the treatment of NIA (13).
Trazodone, an agent with weak inhibition of seroto-
nin reuptake but with potent antagonism at the 5-HT2A
and 5-HT2Creceptors (14), has been used as an antide-
indicated. Because trazodone demonstrates prominent
serotonergic antagonistic properties, we have hypoth-
esized that it would be beneficial in the management of
NIA. Thus the objective of this open-label pilot study
was to evaluate the efficacy of trazodone in the man-
agement of NIA.
MATERIALS AND METHODS
Subjects eligible for the study were required to be in
the age range of 18 to 60 years and meet the Diagnostic
diagnostic criteria for schizophrenia (first episode or
recurrent) (15). All patients were recruited from a fe-
male inpatient unit at the Beer Yaakov Mental Health
(BAS) (16) was required to be included in the study.
Excluded were all patients who engaged in substance
abuse as well as individuals who exhibited neurologic
vous system function. Patients were also excluded if
there was evidence of a nonacute form of NIA ( i.e.,
withdrawal or chronic akathisia). Patients were ex-
cluded if they were receiving “conventional” antiaka-
thisia agents, beta blockers, anticholinergics, and ben-
zodiazepine medications that had been initiated before
the screening procedure. Also excluded were patients
who were uncooperative and patients who received
concomitant antihypertensive medication or who had
marked cardiac arrhythmia or prolonged Q-T, as well
as patients with symptomatic postural hypotension and
a history of syncope episodes. The study was approved
by the local institutional review board and all patients
gave informed consent after receiving a complete ex-
its potential adverse effects.
Before baseline rating, doses of all medications re-
mained unchanged for at least 4 days. The medications
remained stable during the entire study. A prospective
study of 5 days was performed in which trazodone was
ing manner: 50 mg/day the first day before sleep at
night, followed by an increase to 100 mg/day the fol-
lowing 4 days before nighttime sleep. A relatively low
dose of trazodone was selected because no antidepres-
sant effect was intended directly and to avoid any or-
thostatic hypotension, especially in combination with
antipsychotic medication, as well as to minimize its
tinuing the medication after the study for an indefinite
period of time.
The diagnosis and severity of NIA has been estab-
lished by clinical interview and the BAS (16). Because
NIA can be associated with dysphoric mood, the self-
stead (mLAS) (17), was completed. The modified
LAS12 (mLAS) contains five items—tension, panic,
irritability, palpitations, and impatience—scored on a
calculated as the sum of all subscales. Extrapyramidal
(SAS) (18). Assessments with the BAS, mLAS, and
Positive and Negative Symptom Scale (19) and the
Clinical Global Impressions–Severity of Illness and
–Improvement scales (20), and the Hamilton Rating
line and at completion of the 5-day study. All outcome
fore the first dose of trazodone. Vital signs were re-
peated twice a day until completion of the study.
The data were analyzed using analysis of variance
with repeated measures for the BAS, mLAS, and SAS.
The paired t-test was used for analysis of the HAM-D
R. Stryjer et al. 138
Clin. Neuropharmacol., Vol. 26, No. 3, 2003
at the completion of the study (day 5). All results are
expressed as mean ± SD.
Nine female patients (mean age, 36.4 ± 12 years;
range, 22–54 years) were recruited for study participa-
the study were risperidone 4 mg/day (n = 3) and 5
mg/day (n = 1), perphenazine 20 mg/day (n = 1), flu-
phenazine depot 25-mg administered once every 2
weeks (n = 1), olanzapine 15 mg/day (n = 1), haloper-
idol 10 mg/day (n = 1), and zuclopenthixol depot 200-
mg administered once every 2 weeks (n = 1).
Improvement was significant at p < 0.01 for the four
subitems of the BAS (Fig. 1). The mean scores of the
BAS, objective subitem on baseline, day 3, and day 5
were 2.66 ± 0.5 points, 0.55 ± 0.72 point, and 0.11 ±
0.33 point respectively. Mean scores of the subjective
subitem of the BAS on baseline, day 3, and day 5 were
BAS on baseline, day 3, and day 5 were 2.44 ± 0.52
points, 1.11 ± 0.6 point, and 0.33 ± 0.5 point respec-
tively. Mean scores of the BAS subitem global impres-
sion on baseline, day 3, and day 5 were 3.88 ± 0.78
points, 1.33 ± 0.7 point, and 0.44 ± 0.52 point respec-
Improvement in the SAS was not significant. Mean
scores of the SAS on baseline, day 3, and day 5 were
17.77 ± 5.44 points, 16.55 ± 4.47 points, and 16.22 ±
6.79 points respectively.
On the global score of the mLAS the p-value was
significant (p < 0.01) for baseline, day 3, and day 5.
3.04 ± 0.52 points, 1.82 ± 0.59 point, and 1.62 ± 0.49
point respectively. Improvement was significant for all
five subitems. The p-value was p < 0.01 for items one,
two, three, and five, and for subitem four the improve-
ment was less robust (p < 0.031) but still significant.
to the relief of inner tension, restlessness, and dyspho-
ria typically reported by patients with acute NIA.
< 0.041) from baseline to day 5 (13.88 ± 6.52 points to
17 items on the HAM-D, there was a decrease in the
which may be related to improvement of inner tension
as a consequence of decrement of NIA. The improve-
ment of the depressed mood subitem in the HAM-D
was probably related to the decrease of dysphoria,
which is common in patients with NIA.
The improvement in the Positive and Negative
Symptom Scale was significant (p < 0.027) from base-
line to day 5 (105.88 ± 21.27 points to 95.22 ± 23.58
points). The improvement in the Positive and Negative
Symptom Scale was the result of specific item im-
provement: excitement, anxiety, tension, and depres-
sion, again probably related to the decrement of NIA.
Improvement on the Clinical Global Impressions–
Improvement score for akathisia was significant (p <
0.01): from Clinical Global Impressions–Severity of
Illness 4.55 ± 0.52 points at baseline to Clinical Global
Impressions–Severity of Illness 1.55 ± 0.72 point on
the global clinical scale subitem of the BAS. Response
0 point on the global clinical scale subitem (1). Nonre-
sponse was defined as no change or a decrease of 1
point in the global clinical scale subitem of the
Of the nine patients on day 5, five were full respond-
ers (55.5%) and four were responders (44.4%). If we
examine the results on the second evaluation day (day
3), six patients were responders (66.6%), one patient
had a full response (11.1%), and two patients did not
respond (22.2%). Of the two patients who did not re-
jective, and global) over the study period.
Barnes Akathisia Scale (BAS) scores (subjective, ob-
Treatment of Neuroleptic-Induced Akathisia 139
Clin. Neuropharmacol., Vol. 26, No. 3, 2003
spond on day 3, one patient responded and the second
patient had a full response on day 5.
Three patients were withdrawn from trazodone on
completion of the trial at their request: Two patients
said they wanted to know how they felt without the
drug and one patient complained of worsening of in-
somnia since beginning trazodone. All of them had re-
emergence of NIA according to clinical impression 1
day after the drug was withdrawn. The patient who
asked to stop trazodone because of worsening of in-
somnia had severe reemergence of NIA with suicidal
ideation. Because of her severe akathisia she requested
to reinitiate trazodone. One day after a single 100-mg
dose of trazodone a substantial relief was reported by
the patient and was observed by us.
Of the nine patients only one patient complained of
distressful side effects (worsening of insomnia). Al-
though expected, no patient complained of sedation or
women, is a dangerous side effect of trazodone and did
not occur in any of the participating patients. One pa-
tient, whose baseline blood pressure was 110/70
mmHg, had an asymptotic decrease of blood pressure
to 90/60 mmHg after one dose of 50 mg trazodone on
the first day of treatment. However the patient’s blood
pressure returned to normal despite increasing the drug
to 100 mg per day per the protocol.
Our findings indicate the probable efficacy of traz-
odone as an effective alternative agent for the manage-
ment of akathisia. Although the results are certainly in-
teresting, a recent review by Poyurovsky et al. (3) fur-
ther points to the antiakathisia effect of medications
that antagonize 5-HT2Apostsynaptic receptors, such as
mianserin, ritanserin, and cyproheptadine. Trazodone
addition to 5-HT2Cpostsynaptic antagonistic proper-
for the treatment effect in certain patients experiencing
NIA (1,5,8–13), we suggest trazodone’s property of
pal mechanism of activity responsible for the improve-
ment of NIA as observed in our patients. Our findings
thus provide further evidence for the notion that medi-
cations that antagonize postsynaptic 5-HT2Areceptors
exhibit antiakathisia effects.
Although it may be argued that the sedative effect of
trazodone may have played a role in the antiakathisia
effect, none of the participating subjects complained of
a source of the antiakathisia response.
Limitations of the study include the relatively small
sample size, the restriction of subject recruitment to
women, and the lack of a “blinded” nature of the study.
However, because the investigation was planned as a
pilot study, we consider the results to be of worth de-
spite the obvious shortfalls.
Because all of the patients were women, there ex-
isted less of a concern for the relatively dangerous side
effect of priapism (22), which is seen more commonly
in men (23,24). Thus, with an incidence reported as
much as 12%, priapism is an adverse effect that should
always be monitored. Despite this relatively uncom-
mon side effect it may come to limit the broad useful-
ness of trazodone in the management of NIA, espe-
cially in men.
It should be noted that despite the favorable safety
trazodone with atypical antipsychotic medications
leading to an increased risk of priapism has been re-
ported. This includes interactions with the medications
risperidone (25,26) and olanzapine (27). Another pos-
sible interaction of trazodone with antipsychotic
medication that may be of concern includes that of
symptomatic postural hypotension resulting from traz-
odone’s property of blocking peripheral ?1 recep-
tors—a feature observed in one patient in our study.
Considering the theoretic value of postsynaptic
5-HT2a receptor antagonists in the management of
NIA, we suggest trazodone as a beneficial and rela-
tively safe medication for the treatment of NIA. Its
merit relative to other antiakathisia medications must
be determined. To support the hypothesis arising from
this pilot investigation, further study is indicated in the
context of a large-scale, double-blind, placebo-
1. Poyurovsky M, Shardorodsky M, Weizman A. Treatment of neuro-
leptic-induced akathisia with 5-HT2 antagonist mianserin. Double-
blind, placebo-controlled study. Br J Psychiatry 1999;174:238–42.
2. Fleischhacker WW, Roth SD, Kane JM. The pharmacological treat-
ment of neuroleptic-induced akathisia . J Clin Psychopharmacol
3. Poyurovsky M, Weizman A. Serotonin-based pharmacotherapy for
acute neuroleptic-induced akathisia: a new approach to an old prob-
lem. Br J Psychiatry 2001;179:4–8.
4. Kapur S, Remington G. Serotonin–dopamine interaction and its rel-
evance to schizophrenia. Am J Psychiatry 1999;153:466–76.
5. Miller GH, Fleischhacker WW, Ehrman H, et al. Treatment of neu-
R. Stryjer et al. 140
Clin. Neuropharmacol., Vol. 26, No. 3, 2003
chopharmacol Bull 1990;26:373–6.
ropsychopharmacology 1999;21(suppl 2):S106–15.
7. Lane RM. SSRI-induced extrapyramidal side-effects and akathisia:
implication for treatment. J Psychopharmacol 1998;12:192–214.
8. Poyurovsky M, Weizman A. Mirtazapine for neuroleptic-induced
akathisia. Am J Psychiatry 2001;158:819.
9. Miller GH, Fleischhacker WW, Ehrman H, et al. Treatment of neu-
roleptic induced akathisia with the 5HT2 antagonist ritanserin. Psy-
chopharmacol Bull 1990;26:373–6.
10. Miller CH, Hummer M, Pycha R, et al. The effect of ritanserin on
Neuropsychopharmacol Biol Psychiatry 1992;16:247–51.
neuroleptic-induced akathisia. J Clin Psychopharmacol 1998;18:
in neuroleptic-induced akathisia. Br J Psychiatry 1995;167:483–6.
13. Fischel T, Hermesh H, Aizenberg D, et al. Cyproheptadine vs pro-
pranolol for the treatment of acute neuroleptic-induced akathisia: a
comparative double-blind study. J Clin Psychopharmacol 2001;21:
14. Bryant SG, Ereshefsky L. Antidepressant properties of trazodone.
Clin Pharmacol 1982;1:406–17.
15. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders. 4th ed. Washington, DC: American
Psychiatric Association, 1994.
16. Barnes TRE. A rating scale for drug-induced akathisia. Br J Psy-
17. Halstead SM, Barnes TRE, Speller JC. Akathisia: prevalence and
associated dysphoria in an in-patient population with chronic
schizophrenia. Br J Psychiatry 1994;164:177–83.
18. Simpson GM, Angus JWS. A rating scale for extrapyramidal side
effects. Acta Psych Scand 1970;212(suppl):11–9.
19. Kay SR, Fiszbein A, Opler LA. The Positive and Negative Syn-
drome Scale (PANSS) for schizophrenia. Schizophr Bull 1987;13:
20. Guy W. ECDEU Assessment Manual for Psychopharmacology. US
Department of Health, Education, and Welfare publication no.
of vulvar pain. Obstet Gynecol 2001;97:S26–7.
23. Thompson JW Jr, Ware MR, Blashfield RK. Psychotropic medica-
tion and priapism: a comprehensive review. J Clin Psychiatry 1990;
24. Werner MD, Dorn MR, Peabody CA. Survey on the usefulness of
trazodone in patients with PTSD with insomnia and nightmares.
25. Nicolson R, McCurley R. Risperidone-associated priapism. J Clin
26. Emes CE, Milson RC. Risperidone-induced priapism. Can J Psy-
27. Deirmenjian JM, Erharst SM, Wirshing DA, et al. Olanzapine-
induced reversible priapism: a case report. J Clin Psychopharmacol
Treatment of Neuroleptic-Induced Akathisia141
Clin. Neuropharmacol., Vol. 26, No. 3, 2003