Estimating the size of the effects of co-medications on plasma clozapine concentrations using a model that controls for clozapine doses and confounding variables

F. J. Diaz, V. Santoro, E. Spina, M. Cogollo, T. E. Rivera, S. Botts, J. De Leon

Research output: Contribution to journalArticlepeer-review

86 Scopus citations


Aim: The purpose of this study was to estimate the effect sizes of drug interactions on plasma clozapine concentrations, adjusting for potentially confounding factors such as smoking. Methods: The estimation was performed by using a mixed model, and a combination of unpublished (N = 83) and published (N = 172) data that included patients taking phenobarbital, valproic acid, fluvoxamine, fluoxetine, paroxetine, sertraline, citalopram and reboxetine, and patients not taking co-medications. Results: The 255 patients provided a total of 415 steady-state trough plasma clozapine concentrations. Each patient provided 1 to 15 measures of plasma clozapine concentrations. Total plasma clozapine concentration, defined as the sum of plasma clozapine and norclozapine concentrations, was also investigated. A random intercept linear model of the natural log of plasma clozapine concentration with the natural log of dose and other variables as independent variables was built. The model confirmed that phenobarbital induces clozapine metabolism (effect size, E = -28%), and that fluoxetine (E = +42%), fluvoxamine (E = +263%) and paroxetine (E = +30%) inhibit it. Valproic acid appeared to inhibit clozapine metabolism in non-smokers (effect size, E = +16%), whereas it appeared to induce clozapine metabolism in smokers (E = -22%). The effect sizes of smoking on plasma clozapine concentration were -20% in patients not taking valproic acid, and -46% in patients taking valproic acid. Thus, smoking induces clozapine metabolism, and this induction may be stronger when the patient is taking valproic acid. The effect sizes allowed the computation of clozapine dose-correction factors for phenobarbital, 1.4 [95% confidence interval, CI, (1.1, 1.7)]; paroxetine, 0.77 (0.67, 0.89); fluoxetine, 0.70 (0.64, 0.78); fluvoxamine, 0.28 (0.22, 0.35); and valproic acid [0.86 (0.75, 1.0) in non-smokers, and 1.3 (0.96, 1.73) in smokers]. Sertraline, reboxetine and citalopram had no obvious effects. Discussion: The results for total plasma clozapine concentrations are similar to those for plasma clozapine concentrations. The main limitations of this study were that the computed effect sizes reflect only the doses and treatment-durations of the co-medications studied, and that the substantial "noise" of the clinical environment may make it difficult to detect the effects of some variables, particularly those with small effect sizes. Gender was not significant probably due to its relatively small effect size in the studied population, and age was not significant probably due to the limited age variability. Conclusion: This article contributes to the clozapine literature by describing a possible interaction between taking valproic acid and smoking, which modifies plasma clozapine concentrations, by estimating the effect sizes of other compounds on plasma clozapine concentrations after correcting for confounders, and by providing dose-correction factors for clinicians.

Original languageEnglish
Pages (from-to)81-91
Number of pages11
Issue number3
StatePublished - May 2008

ASJC Scopus subject areas

  • Psychiatry and Mental health
  • Pharmacology (medical)


Dive into the research topics of 'Estimating the size of the effects of co-medications on plasma clozapine concentrations using a model that controls for clozapine doses and confounding variables'. Together they form a unique fingerprint.

Cite this