Benzodiazepine long-term administration is associated with impaired attention/working memory in schizophrenia: results from the national multicentre FACE-SZ data set

Eur Arch Psychiatry Clin Neurosci. 2018 Feb;268(1):17-26. doi: 10.1007/s00406-017-0787-9. Epub 2017 Mar 27.


Objective: The effect of benzodiazepine long-term administration (BLTA) in cognitive functioning of subjects with schizophrenia (SZ) has been partially explored to date. The objective was to assess BLTA-associated cognitive impairment with a comprehensive cognitive battery in a non-selected multicentric/national community-dwelling sample of stabilized SZ subjects.

Method: 407 community-dwelling stabilized SZ subjects were consecutively included in the FondaMental Academic Centers of Expertise for Schizophrenia Cohort (FACE-SZ). Patients taking daily benzodiazepine were defined as BLTA+ as all patients examined by the Expert Center were clinically stabilized and under stable dose of treatment for at least 3 months. Each patient has been administered a 1-day long comprehensive cognitive battery (including The National Adult Reading Test, the Wechsler Adult Intelligence Scale, the Trail Making Test, the California Verbal Learning Test, the Doors test, and The Continuous Performance Test-Identical Pairs).

Results: In the multivariate analyses, results showed that BLTA was associated with impaired attention/working memory (OR 0.60, 95% confidence interval 0.42-0.86; p = 0.005) independently of socio-demographic variables and illness characteristics. Verbal and performance current IQ-[respectively, OR 0.98, 95% CI (0.96;0.99), p = 0.016 and 0.98, 95% CI(0.97;0.99), p = 0.034] but not premorbid IQ-(p > 0.05) have been associated with BLTA in a multivariate model including the same confounding variables.

Conclusion: BLTA is associated with impaired attention/working memory in schizophrenia. The BLTA benefit/risk ratio should be regularly reevaluated. Alternative pharmacological and non-pharmacological strategies for comorbid anxiety disorders and sleep disorders should be preferred when possible. It seems reasonable to withdraw BLTA before the start of cognitive remediation therapy, as soon as possible, to improve the effectiveness of this therapy. Limits: the delay between the last benzodiazepine intake and testing, as well as the specific class of benzodiazepines (long half-life vs. short half-life), and the number of benzodiazepine daily intakes have not been recorded in the present study. The precise motive for BLTA prescription and sleep disturbances have not been reported, which is a limit for the interpretation of the present results.

Keywords: Anxiety; Anxiolytics; Benzodiazepine; Cognition; Schizophrenia.

Publication types

  • Multicenter Study

MeSH terms

  • Adult
  • Antipsychotic Agents / adverse effects*
  • Attention Deficit Disorder with Hyperactivity / chemically induced*
  • Benzodiazepines / adverse effects*
  • Cohort Studies
  • Female
  • Humans
  • Male
  • Memory Disorders / chemically induced*
  • Memory, Short-Term / drug effects*
  • Neuropsychological Tests
  • Principal Component Analysis
  • Psychiatric Status Rating Scales
  • Schizophrenia / drug therapy


  • Antipsychotic Agents
  • Benzodiazepines