Effect of Pre-Morbid Anxiolytic Medications on Ischemic Stroke Functional Outcome

Introduction

Preclinical and clinical data suggest anxiolytics such as benzodiazepines may have deleterious effects on stroke recovery.^1,2 This study investigates the association of pre-stroke use of anxiolytics and/or antidepressants on 90-day functional status following ischemic stroke.

Methods

This is a post-hoc analysis of the Albumin in Acute Ischemic Stroke Part 2 trial (ALIAS Part 2), ³ publicly available at https://www.ninds.nih.gov/Current-Research/Research-Funded-NINDS/Clinical-Research/Archived-Clinical-Research-Datasets. IRB approval was not required as data were deidentified. The primary outcome is 90-day post-stroke modified Rankin Scale (mRS) score 0–1 (good outcome), the secondary outcome is 90-day mRS 0–2. The primary exposure was patient or legally authorized representative reported use of anxiolytic or antidepressant medications in the month prior to stroke. We divided patients into: no anxiolytic or antidepressant, anxiolytic alone, antidepressant alone, or anxiolytic and antidepressant.

We fit logistic regression models to the outcomes with Model 1 unadjusted and Model 2 adjusted for variables selected with stepwise backwards elimination set at p<0.05. We conducted a sensitivity analysis to account for missing demographic data with multiple-imputation using a multivariate normal distribution.

Results

We included 804 patients: 664 not exposed to pre-stroke anxiolytic or antidepressant (39.8% with good outcome), 34 exposed to anxiolytics alone (20.6% with good outcome), 91 to antidepressants alone (30.8% with good outcome), and 15 to anxiolytics with antidepressants (33.3% with good outcome). Additional data is in Supplementary Table I. Compared to patients not taking anxiolytics or antidepressants, patients taking anxiolytics alone had lower odds of good outcome in the unadjusted model (OR 0.39, 95% CI 0.17–0.92, p=0.023); no association was seen in patients taking an antidepressant alone or anxiolytic with antidepressant (Table 1). These findings persisted in the adjusted model (Table 1), when defining a good outcome as mRS 0–2 (Table 1), and using multiple-imputation (results not shown).

Table 1.

Association between pre-stroke anxiolytic and antidepressant medication use and good outcome 90-days after ischemic stroke (n=754).

Outcome	Odds ratio*	95% CI	p-value
Outcome of 90-day modified Rankin Scale of 0–1
No pre-stroke anxiolytic or antidepressant (n=664)	ref	ref	ref
Pre-stroke anxiolytic alone (n=34)	0.30	0.11–0.85	0.023
Pre-stroke antidepressant alone (n=91)	0.76	0.44–1.28	0.300
Pre-stroke anxiolytic and antidepressant (n=15)	0.62	0.20–1.93	0.412
Outcome of 90-day modified Rankin Scale of 0–2
No pre-stroke anxiolytic or antidepressant (n=664)	ref	ref	ref
Pre-stroke anxiolytic alone (n=34)	0.40	0.17–0.95	0.037
Pre-stroke antidepressant alone (n=91)	0.83	0.50–1.38	0.425
Pre-stroke anxiolytic and antidepressant (n=15)	0.77	0.26–2.30	0.642

^*Adjusted for age, sex, race, ethnicity, baseline NIH Stroke Scale, tPA administration status, systolic blood pressure at baseline, diabetes, hypertension, statin use, pre-morbid mRS.

Discussion

We found pre-stroke use of anxiolytic medications was associated with worse functional outcome following acute ischemic stroke. There was no significant association for concurrent use of anxiolytics and antidepressants or antidepressants alone. Limitations include: the low number of patients on anxiolytics increases the possibility findings are due to chance, medication exposure obtained through self-report may decrease accuracy, and our inability to account for type of medication or severity of pre-stroke anxiety and/or depressive disorder.