(095) Comparative analysis of anticholinergic burden scales to explain iatrogenic cognitive impairment in schizophrenia: results from the multicenter FACE-SZ cohort

Background: The anticholinergic properties of medications are associated with poorer cognitive performance in schizophrenia. Numerous scales were developed to assess anticholinergic burden and yet, there is no consensus indicating which scale is more relevant to assess the risk of anticholinergic cognitive effects in patients with schizophrenia.

Objectives: This study aimed to identify scales significantly associated with cognitive impairment that could be valuable in assessing the risk of iatrogenic cognitive impairment in clinical settings. Additionally, we compared the scales to traditional measures of iatrogenic burden, such as the number of psychotropic drugs.

Methods: We conducted a comprehensive literature review in November 2022 to identify available anticholinergic burden scales. We included French adults with schizophrenia or schizoaffective disorder from a nation-wide cohort (the FACE-SZ cohort) from 2010 to 2021, excluding those with alternative sources of cognitive impairment. In a cross-sectional study design, we examined the association between objective global cognitive performance (assessed by neuropsychology tests) and identified scales, the number of psychotropic drugs, chlorpromazine and lorazepam equivalents, using bivariate regressions. Regression models were then adjusted with covariates like symptom severity and history of psychosis after multiple imputations.

Results: We identified 27 anticholinergic burden scales in the literature. The study included 839 adults (26% females, 31.6 y-o) with cognitive impairment compared to the general population (mean z-score = -0.83). Eight scales, the number of psychotropic drugs, and drug equivalents showed significant albeit weak associations with cognitive impairment in multiple regression models (-0.11 ≤ Standardized β ≤ -0.07; .005 ≤ p ≤ .043). According to the eight valid anticholinergic burden scales and the threshold proposed by the scales’ authors, 24-43% of our sample had a high anticholinergic burden. The number of psychotropic drugs, the easiest valid predictor to compute, was linked to worse executive function (Standardized β = -0.12, p = .004) and reasoning (Standardized β = -0.08, p = .037).

Conclusions: Eight scales were valid tools to assess the risk of cognitive impairment in schizophrenia and schizoaffective disorder. However, weak associations between anticholinergic burden, number of psychotropic drugs, drug equivalents, and cognition suggest that either those predictors inaccurately assess cognitive side effects or that medication contributes poorly to cognitive impairment in these disorders. The number of psychotropic drugs emerged as the easiest method to assess the risk of iatrogenic cognitive impairment.