Research Proposal
Background: Antipsychotic drugs are the mainstay for the treatment of schizophrenia. Important known side effects are weight gain and disturbances in lipid and glucose metabolism (metabolic side effects). To date a comprehensive comparison of the propensity of the different antipsychotic drugs to cause metabolic side effects is still missing.
Objective: To estimate to which degree antipsychotic drugs differ in metabolic side effects during medium- to long-term treatment of individuals with schizophrenia.
Study Design: Systematic review and network meta-analysis
Participants: Individuals with a diagnosis of schizophrenia or related disorders who participated in randomized controlled trials which compared antipsychotic drugs head-to-head or versus placebo without restrictions in terms of age, gender, ethnicity or setting.
Main Outcome Measures: Parameters describing weight gain and changes in the glucose and lipid metabolism.
Statistical Analysis: Network meta-analysis of mean differences (MD) for continous outcomes and odds ratio (OR) for dichotomous outcomes.
Please find attached the PROSPERO protocol for further information.
Standardized mortality ratios (SMRs) show a significantly higher all-cause mortality of individuals with schizophrenia compared to the general population [1], and the life-span of people with schizophrenia has been estimated to be 14.5 years shorter [2]. Approximately 40% of the excess mortality is explained by suicides (and accidents). The remaining 60% are due to a variety of natural causes [3]. Particularly metabolic side effects of antipsychotics are thought to increase natural-cause-mortality because they are associated with increased cardiovascular mortality and morbidity [4,5,6].
Moreover, metabolic side effects, and weight gain in particular, reduce patients’ quality of life and lead to drug-non-adherence [7]. Drug-non-adherence often results in poor treatment outcome and psychotic relapse, thus leading to further hospitalisations and adding to the loss of productivity.
We will conduct a systematic review and network meta-analysis of randomized controlled trials to summarize the existing evidence of differences between antipsychotic drugs in metabolic side effects. These findings could help physicians and patients to choose the best drug for the individual.
Please find attached the PROSPERO protocol for further information.
The aim of this project is to examine the differences of antipsychotic drugs in propensity to cause metabolic side effects (weight gain and changes in the lipid and glucose metabolism) in the medium- to long-term treatment of individuals with schizophrenia.
Main hypothesis: Antipsychotic drugs differ in their propensities to cause metabolic side effects.
Subgroup analysis will explore effects of baseline metabolic parameters, age, sex, ethnicity, life-time exposure to antipsychotics (if not available duration of illness will be used as a proxy), antipsychotic dose, pharmaceutical sponsorship, study duration.
Please find attached the PROSPERO protocol for further information.
Search strategy: We ran a search in the Cochrane Schizophrenia Group’s Study-Based Register of Trials with no date/time, language, document type, and publication status limitations [8-11]. This register is compiled inter alia of regular searches in multiple electronic databases (including MEDLINE, Embase, PsycINFO, PubMed, ClinicalTrials.gov). Additionally, we will contact pharma companies for missing relevant data.
Included antipsychotic drugs are Amisulpride, Aripiprazole, Asenapine, Benperidol, Brexpiprazole, Cariprazine, Chlorpromazine, Clopenthixol, Clozapine, Flupentixol, Fluphenazine, Fluspirilene, Haloperidol, Iloperidone, Levomepromazine, Loxapine, Lumateperone, Lurasidone, Molindone, Olanzapine, Paliperidone, Penfluridol, Perazine, Perphenazine, Pimozide, Quetiapine, Risperidone, Sertindole, Sulpiride, Thioridazine, Tiotixene, Trifluoperazine, Ziprasidone, Zotepine, Zuclopenthixol.
Inclusion criteria:
Individuals with a diagnosis of schizophrenia or related disorders who participated in randomized controlled trials which compared antipsychotic drugs head-to-head or versus placebo.
Please find attached the PROSPERO protocol for further information.
The main outcome measure is the continuous measure of body weight.
The effect size measure will be mean differences (MD) and its 95% credibility intervals (CrIs). We will prefer change data over endpoint data. Estimates based on imputation methods to handle missing data (used by the original authors) will be preferred over completers’ data. Imputed data based on mixed-models of repeated measurement (MMRM) will be preferred over last-observation carried forward (LOCF), if available. We will estimate missing standard deviations (SDs) as described in the Cochrane Handbook [12].
Please find attached the PROSPERO protocol for further information.
We expect the different antipsychotic drugs to be the main predictor.
Other potential predictors (baseline metabolic parameters, age, gender, ethnicity, life-time exposure to antipsychotics, antipsychotic dose, pharmaceutical sponsorship and study duration) will be adressed in subgroup analyses.
Please find attached the PROSPERO protocol for further information.
We plan to analyse for each domain the parameter with the most data available:
1.Glucose metabolism: Continuous measures of fastening glucose, HbA1c, HOMA-IR and insulin. Dichotomous measures of number of patients with impaired fastening glucose and increased HbA1c.
2.Disturbances in total cholesterol metabolism: Continuous measures of total cholesterol. Dichotomous measures of number of patients with hypercholesterinemia.
3.Disturbances in LDL cholesterol metabolism: Continuous measures of LDL cholesterol. Dichotomous measures of number of patients with increased LDL cholesterol.
4.Disturbances in HDL cholesterol metabolism: Continuous measures of HDL cholesterol. Dichotomous measures of number of patients with reduced HDL cholesterol.
5.Disturbances in triglyceride metabolism: Continuous measures of triglycerides. Dichotomous measures of number of patients with hypertriglyceridemia.
Addtitionally we will extract BMI, waist circumference, and number of patients overweight, obese, and with clinically significant weight gain.
Please find attached the PROSPERO protocol for further information.
Primarily, we will perform random effects network meta-analysis fitted in a Bayesian environment in JAGS. A common heterogeneity parameter will be used in the model.
We will estimate the probability for the ranking of each intervention using SUCRA (surface under the cumulative ranking curve) [13].
We will assess transitivity of the network first epidemiologically by comparing the distribution of potential effect modifiers across studies grouped by comparison. Potential effect modifiers are listed under “Main Predictor" above. Statistical evaluation of the transitivity (consistency) will be performed using the design-by-treatment test [14], and the SIDE (Separating Indirect from Direct Evidence) approach [15]. In case of significant inconsistency, we will investigate possible sources of it (mistakes in data entry, clear differences in study characteristics). Small or moderate amounts of inconsistency will be further explored by subgroup analyses of the potential effect modifiers.
We will explore the association between study size and effect size with a comparison-adjusted funnel plot and a contour enhanced funnel plot of all active drugs versus placebo (assuming that at least 10 studies are available per outcome). Any asymmetry observed can be attributed to systematic differences between small and large studies, true heterogeneity or publication bias.
We will evaluate the confidence in estimates of the primary outcome with the framework Confidence in Network Meta-Analysis (CINeMA) [16].
If the requirements of network meta-analysis are not met (low likelihood of transitivity and/or large unexplained inconsistency) we will use pairwise meta-analysis for data synthesis. We will perform frequentist pairwise meta-analysis in R using the package “meta”. Heterogeneity will be investigated by visual inspection of forest plots, by estimating the between studies variance tau² and with the I² statistic.
Please find attached the PROSPERO protocol for further information.
Antipsychotic drugs differ little in efficacy but in the occurrence of side effects. Beside extrapyramidal motor disorders and prolactin increase, metabolic disorders are the most important side effects: Weight gain and changes in lipid and glucose metabolism increase cardiovascular risk and can lead to severe somatic diseases (e.g. diabetes, stroke and myocardial infarction). To date a comprehensive comparison of the metabolic side effects of the antipsychotic drugs during medium- to long-term treatment of individuals with schizophrenia is still missing. We will conduct a systematic review and network meta-analysis to fill this gap.
Start of project: 10/2019
First contact of data holders: 11/2020
Actual state of the project: Identification of included RCTs from literature search and data extraction.
It is planned to finish data extraction and to start data analysis by 06/2021.
First data presentations and publications are planned for the following months.
According to the framework of our grant, there is a deadline for data presentation in 11/2021.
Findings of this systematic review will be published in peer-reviewed scientific journals (such as Lancet Psychiatry, JAMA Psychiatry, Schizophrenia Bulletin or Schizophrenia Research) and it is likely that they will contribute to treatment guidelines.
1. McGrath J, Saha S, Chant D, Welham J. Schizophrenia: A concise overview of incidence, prevalence, and mortali-ty. Epidemiol Rev 2008; 30:67–76.
2. Hjorthøj C, Stürup AE, McGrath JJ, Nordentoft M. Years of potential life lost and life expectancy in schizophrenia: A systematic review and meta-analysis. The Lancet Psychiatry 2017; 4(4):295–301.
3. Saha S, Chant D, McGrath J. A systematic review of mortality in schizophrenia: Is the differential mortality gap worsening over time? Arch Gen Psychiatry 2007; 64(10):1123–31.
4. Hert M de, Detraux J, van Winkel R, Yu W, Correll CU. Metabolic and cardiovascular adverse effects associated with antipsychotic drugs. Nat Rev Endocrinol 2011; 8(2):114–26.
5. Kaur J. A comprehensive review on metabolic syndrome. Cardiol Res Pract 2014; 2014:943162.
6. Arango C, Bobes J, Aranda P, Carmena R, Garcia-Garcia M, Rejas J. A comparison of schizophrenia outpatients treat-ed with antipsychotics with and without metabolic syndrome: Findings from the CLAMORS study. Schizophr Res 2008; 104(1-3):1–12.
7. Perkins DO. Predictors of noncompliance in patients with schizophrenia. J Clin Psychiatry 2002; 63(12):1121–8.
8. Shokraneh F, Adams CE. Study-based registers of randomized controlled trials: Starting a sys-tematic review with data extraction or meta-analysis. Bioimpacts 2017; 7(4):209–17.
9. Shokraneh F, Adams CE. Study-based registers reduce waste in systematic reviewing: Discus-sion and case report. Syst Rev 2019; 8(1):129.
10. Shokraneh F, Adams C. Classification of all pharmacological interventions tested in trials rele-vant to people with schizophrenia: A study-based analysis; 2020.
11. Shokraneh F, Adams C. Cochrane Schizophrenia Group’s Study-Based Register of Randomized Controlled Trials: Development and Content Analysis; 2020.
12. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.1 (updated September 2020). Cochrane, 2020. Available from www.training.cochrane.org/handbook.
13. Salanti G, Ades AE, Ioannidis JPA. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: An overview and tutorial. J Clin Epidemiol 2011; 64(2):163–71.
14. Higgins JPT, Jackson D, Barrett JK, Lu G, Ades AE, White IR. Consistency and inconsistency in network meta-analysis: Concepts and models for multi-arm studies. Res Synth Methods 2012; 3(2):98–110.
15. Dias S, Welton NJ, Caldwell DM, Ades AE. Checking consistency in mixed treatment comparison meta-analysis. Stat Med 2010; 29(7-8):932–44.
16. Nikolakopoulou A, Higgins JPT, Papakonstantinou T, Chaimani A, Del Giovane C, Egger M et al. CINeMA: An approach for assessing confidence in the results of a network meta-analysis. PLoS Med 2020; 17(4):e1003082.
