Bipolar Affective Disorder, Schizoaffective Disorder, Schizophrenia
Conditions
Brief summary
In this trial, researchers aim to investigate if prolonged-release melatonin can facilitate the withdrawal of chronic benzodiazepine administration in patients with schizophrenia. Furthermore, researchers will investigate the association of benzodiazepine dose reduction with the following clinically important variables: sleep, psychophysiology, cognition, social function, and quality of life.
Detailed description
Treatment of schizophrenia frequently includes prolonged administration of benzodiazepines despite lack of evidence of its use. It is often difficult to discontinue use of benzodiazepines because of development of dependence. After being randomized to prolonged-release melatonin (Circadin®) 2 mg daily versus matching placebo, participants are required to slowly taper off their benzodiazepine dose towards no intake. Data are collected at baseline and at 6 months follow-up regarding medical treatment, cognition, psychophysiology, sleep, laboratory tests, adverse events, psychopathology, social function, and quality of life. Data on medical treatment, cognition, adverse events, social function, and quality of life are also collected at 2 and 4 months follow-up. The results from this trial will assess if melatonin has a role in withdrawing long-term benzodiazepine administration in schizophrenia patients. This group of patients is difficult to treat and therefore often subject to polypharmacy which may play a role in the reduced life expectancy compared to the background population. In addition, the data of the trial are also analyzed as an observational cohort design to investigate the association of benzodiazepine dose reduction/discontinuation with psychophysiology, cognition, sleep, quality of life, and other selected variables (not further described below, see trial protocol). Knowledge of these important clinical aspects is lacking in this group of patients.
Interventions
DRUGPlacebo
Both Circadin and placebo are encapsulated in lactose containing gelatin capsules to optimize the blinding.
DRUGMelatonin
Prolonged-release melatonin (Circadin®) 2 mg, once daily, 1-2 hours before bedtime.
Sponsors
Glostrup University Hospital, Copenhagen
Copenhagen Trial Unit, Center for Clinical Intervention Research
Lone Baandrup
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
TREATMENT
Masking
QUADRUPLE (Subject, Caregiver, Investigator, Outcomes Assessor)
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Patients diagnosed with schizophrenia, schizoaffective disorder, or bipolar affective disorder (ICD-10 criteria for schizophrenia (F20), schizoaffective disorder (F25) or bipolar affective disorder (F31) must be fulfilled at inclusion or previously as documented by chart review; fulfillment of relevant DSM-IV-TR criteria will also be registered). * Treated with the same antipsychotic drug for at least 3 months before inclusion (change of dose, antipsychotic polypharmacy and prescription/discontinuation of add-on drugs allowed but the basic antipsychotic treatment should be the same). * Continuously treated with at least one benzodiazepine (chlordiazepoxide, diazepam, clobazam, clonazepam, flunitrazepam, nitrazepam, bromazepam, alprazolam, lorazepam, lormetazepam, oxazepam, triazolam) or benzodiazepine related drug (zolpidem, zopiclone, zaleplon) for at least 3 months before inclusion. * Age 18+. * Fertile women: negative pregnancy test at baseline and use of safe contraceptives (intrauterine devices or hormonal contraception) throughout the trial period and 1 day after withdrawal of trial medication. This does not apply to sterile or infertile participants, i.e. surgically sterilized or post menopausal (missing period for at least 12 months before inclusion) women. * Written informed consent.
Exclusion criteria
* Known aggressive or violent behavior. * Mental retardation, pervasive developmental disorder, or dementia. * Epilepsy, terminal illness, severe comorbidity or unable to understand Danish. * Allergic to compounds in the trial medication (melatonin, lactose, starch, gelatin, talc). * Hepatic impairment (known diagnosis). * Pregnancy and nursing. * Missing informed consent.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Benzodiazepine (including benzodiazepine related drugs) dose at 6 months follow-up. | 6 months follow-up. | The general linear model is used with the outcome measure (dose after 6 months) as the dependent variable and the indicator of intervention and the baseline value as the independent variables. If the assumptions of the model cannot be fulfilled either directly or after transformation a non-parametric method will be used. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| The fraction of participants who has completely discontinued benzodiazepines 6 months after initiating trial medication. | 6 months follow-up. | The analysis will be done using a logistic regression model where logit(p) is the dependent variable, p is the probability of completing the withdrawal, and a binary intervention indicator is the independent variable. |
| Pattern of P300 amplitude (psychophysiology) over time. | 2, 4, and 6 months. | The mixed model with repeated measures will be used to analyze the time course. The model is outcome measure = int + baseline + a\*t + b\*t\*t + c\*baseline\*t + d\*baseline\*t\*t + e\*I + f\*I\*t + g\*I\*t\*t where t is time, I the intervention indicator, int the intercept, and a through g the coefficients. Using Akaike's criterium the best co-variance matrix is first chosen among an unstructured, a compound symmetric, or a first order autoregressive. |
| Pattern of Brief Assessment of Cognition in Schizophrenia (BACS) composite score over time. | 2, 4, and 6 months. | The mixed model with repeated measures will be used to analyze the time course. The model is outcome measure = int + baseline + a\*t + b\*t\*t + c\*baseline\*t + d\*baseline\*t\*t + e\*I + f\*I\*t + g\*I\*t\*t where t is time, I the intervention indicator, int the intercept, and a through g the coefficients. Using Akaike's criterium the best co-variance matrix is first chosen among an unstructured, a compound symmetric, or a first order autoregressive. |
| Pattern of benzodiazepine dose over time. | 2, 4, and 6 months. | The mixed model with repeated measures will be used to analyze the time course. The model is outcome measure = int + baseline + a\*t + b\*t\*t + c\*baseline\*t + d\*baseline\*t\*t + e\*I + f\*I\*t + g\*I\*t\*t where t is time, I the intervention indicator, int the intercept, and a through g the coefficients. Using Akaike's criterium the best co-variance matrix is first chosen among an unstructured, a compound symmetric, or a first order autoregressive. |
| Pittsburgh Sleep Quality Index (PSQI) global score at 6 months follow-up. | 6 months. | The general linear model is used with the outcome measure (PSQI) as the dependent variable and the indicator of intervention and the baseline value as the independent variables. If the assumptions of the model cannot be fulfilled either directly or after transformation a non-parametric method will be used. |
| Pattern of Benzodiazepine Withdrawal Symptom Questionnaire (BWSQ-2) score over time. | 2, 4, and 6 months. | The mixed model with repeated measures will be used to analyze the time course. The model is outcome measure = int + baseline + a\*t + b\*t\*t + c\*baseline\*t + d\*baseline\*t\*t + e\*I + f\*I\*t + g\*I\*t\*t where t is time, I the intervention indicator, int the intercept, and a through g the coefficients. Using Akaike's criterium the best co-variance matrix is first chosen among an unstructured, a compound symmetric or a first order autoregressive. |
| Sleep efficiency (polysomnography) at 6 months follow-up. | 6 months. | The general linear model is used with the outcome measure (sleep efficiency) as the dependent variable and the indicator of intervention and the baseline value as the independent variables. If the assumptions of the model cannot be fulfilled either directly or after transformation a non-parametric method will be used. |
Countries
Denmark
Outcome results
None listed