Abstract
Background
In animal models and healthy volunteers, the use of GABA A receptor agonists (GABA-AGs) seem deleterious for functional recovery. The agents are widely used for subacute stroke, but their effect on functional recovery remains unclear.
Objectives
We aimed to evaluate the association between GABA-AG use and functional recovery after stroke.
Methods
We retrospectively recruited 434 survivors of subacute stroke admitted for inpatient rehabilitation between 2000 and 2013 in our institution (107 with and 327 without GABA-AG use). We used multivariate regression to assess the association of GABA-AG use and successful functional recovery, defined as reaching, between admission and discharge, the minimal clinically important difference (MCID) of 22 points on the global Functional Independence Measure (FIM). Secondary analyses were the associations of GABA-AG with cognitive and motor FIM MCID and constant GABA-AG exposure (24 h/24 GABA-AG) with global, cognitive and motor FIM MCID. A new estimation of the MCID was performed with the standard error of measurement.
Results
Reaching the global FIM MCID was associated with GABA-AG use (adjusted odds ratio [aOR] 0.54 [95% CI 0.31–0.91], P = 0.02) as well as 24 h/24 GABA-AG use (aOR 0.25 [0.08–0.83]; P = 0.02). Furthermore, GABA-AG and 24 h/24 GABA-AG use was inversely but not always significantly associated with reaching the cognitive FIM MCID (aOR 0.56, P = 0.07; aOR 0.26, P = 0.06, respectively) and motor FIM MCID (aOR 0.51, P = 0.07; aOR 0.13, P = 0.01, respectively). The estimated MCID was 19 for global FIM, 4 for cognitive FIM, and 16 for motor FIM.
Conclusions
GABA-AG use is associated with not reaching successful functional recovery during stroke rehabilitation. Randomised trials are needed to formally establish the potential deleterious effect of GABA-AG use on functional recovery.
1
Introduction
Stroke, a frequent disease with major impact on mortality and morbidity, represents the main cause of neurological disability among adults in Western countries. Rational and evidence-based management after stroke is needed, during acute care and during rehabilitation. Pharmacological treatment during this phase may have major effects on functional recovery . The GABA A receptor agonists (GABA-AGs) represent a widely used drug class in neurological rehabilitation, for anxiety and depression, sleep disorders, and dependence troubles . Frequently used GABA-AGs include benzodiazepines, benzodiazepine-like sleep inducers (i.e., zolpidem, zopiclone, and zaleplon), and clomethiazole. All these drugs have similar pharmacological action .
In animal models, reducing excessive GABA-mediated tonic inhibition promoted functional recovery after stroke . In healthy volunteers, benzodiazepines had deleterious effects on exercise-related neuronal plasticity . The clinical implications of GABA-AG use have been described only in observational studies with small sample sizes or heterogeneous populations . Lazar et al. showed that benzodiazepines could re-induce prior motor and phasic defects in stroke patients who already reached a functional recovery plateau . Two observational studies were designed to assess the effect of antineuroplastic drugs on functional recovery during post-stroke rehabilitation. Neither study was specifically designed for GABA-AGs. The study by Goldstein et al. , which included 96 patients, showed a deleterious effect on motor recovery for 37 patients who were taking drugs that could potentially affect neuroplasticity. Because benzodiazepines, neuroleptics, clonidine, prazosin, phenytoin, phenobarbital, and anti-dopaminergics were grouped for analysis, the observed effect on motor recovery could be also due to non GABA mediated effects. In another study of 408 patients, Nadeau et al. showed no significant effect on functional walking level for 33 patients taking benzodiazepines.
Different drugs other than GABA-AGs are probably involved in neuronal plasticity and functional recovery. Indeed, other central nervous system depressants such as neuroleptics (e.g., haloperidol) and some anti-epileptic agents (e.g., phenytoin and phenobarbital) can have a deleterious effect on neuronal plasticity . However, some randomized studies with small sample sizes supported the potential benefit of dopaminergics on functional recovery . More evidence exists about a potential benefit of selective serotoninergic reuptake inhibitors (SSRIs) . The FLAME study , a randomised study including 113 patients with subacute stroke, showed a benefit of SSRIs for motor recovery after 3 months of neurologic rehabilitation.
The Functional Independence Measure (FIM) is a validated scale that allows for assessing the functional capacities of patients with neurological impairment . The FIM contains 2 parts: cognitive FIM (maximum 35 points) and motor FIM (maximum 91 points). The global FIM (maximum 126 points) is obtained by adding the cognitive and motor FIM scores. Measuring FIM at admission and discharge allows for assessing functional recovery. An anchored study calculated the minimal clinically important difference (MCID) for the FIM during the rehabilitation of 113 survivors of subacute stroke. The established FIM MCID was 22 points for global FIM, 3 for cognitive FIM, and 17 for motor FIM.
In this retrospective study, we investigated the association of GABA-AG use and functional recovery of stroke survivors by using the FIM MCID.
2
Patients and methods
2.1
Patients
We recruited patients > 18 years old of both genders who were hospitalised for rehabilitation after subacute stroke (ischemic and/or hemorrhagic; i.e., 5–180 days after stroke) between January 2000 and December 2013 in a neurological rehabilitation centre (Institution de Lavigny, Switzerland). Exclusion criteria were rehabilitation duration < 14 days, occasional GABA-AG use, admission FIM > 104/126 (i.e., maximum global FIM minus global FIM MCID) and missing global FIM at admission or discharge. Moreover, we excluded patients with cognitive FIM > 32/35 on admission (i.e., maximum cognitive FIM minus cognitive FIM MCID) and motor FIM > 74/91 on admission (i.e., maximum motor FIM minus motor FIM MCID) from the cognitive and motor FIM subgroup analyses, respectively. These FIM thresholds were from global, cognitive, and motor FIM MCIDs previously reported . The study was approved by the local ethics committee (Switzerland, Vaud regional board). The patient sample is representative of classical neurological rehabilitation practice for stroke survivors usually provided in Switzerland.
2.2
Design
The primary aim was to assess an association between GABA-AG intake and reaching the global FIM MCID between admission and discharge. Secondary analyses were the association of GABA-AG intake and cognitive and motor FIM MCID and 24 h/24 GABA-AG use (i.e., drugs taken morning, noon, and evening or when the drug effect lasted > 12 h) and global, cognitive, and motor FIM. We chose the FIM because it is designed to assess the main aspects of functional recovery in patients after stroke. Moreover, to confirm the reliability of the selected MCIDs and for informative purposes, MCIDs were estimated with a distribution-based method with the standard error of measurement .
2.3
Data collection
All data were collected from digital patient records, which were systematically available between 2000 and 2013. GABA-AG intake (i.e., prescription for all days) and global FIM were systematically collected at admission and discharge. Cognitive and motor FIM were collected at admission and discharge. 24 h/24 GABA-AG intake was collected (e.g., for chlordiazepoxide, clorazepam, diazepam, flunitrazepam, flurazepam, halazepam, ketazolam, medazepam, nordazepam, prazepam, or quazepam once a day). We used the established FIM MCID of 22 points for global FIM, 3 for cognitive FIM, and 17 for motor FIM achieved between admission and discharge. We also collected potential confounding factors that could have an effect on our primary and secondary outcomes:
- •
age;
- •
sex;
- •
rehabilitation length;
- •
delay between stroke and rehabilitation;
- •
use of SSRIs or serotonin-norepinephrine reuptake inhibitors, tricyclic or tetracyclic antidepressants, antiepileptic or neuroleptic drugs;
- •
type of stroke (ischemic or hemorrhagic, left cortical with or without deep telencephalon or diencephalon, right cortical with or without deep telencephalon or diencephalon, deep telencephalon or diencephalon; infratentorial);
- •
aphasia;
- •
symptomatic stroke antecedent; anxiety or depression;
- •
neurologic or psychiatric debilitating condition (dementia, encephalopathy, multiple sclerosis, Parkinson disease, severe visual impairment, cerebral palsy, mental impairment, or psychosis);
- •
cardiac or pulmonary insufficiency; medical debilitating condition (acute care environment hospitalization, active malignancy, insulin-requiring diabetes mellitus);
- •
obstructive sleep apnoea syndrome.
2.4
Sample size determination
A previous (unpublished) pilot study of 100 eligible patients allowed for estimating sample size: 22/100 patients took GABA-AGs, among whom 6 reached the FIM MCID, and 78/100 patients did not take GABA-AGs, among whom 39 reached the FIM MCID. Confounding factors were not collected in this pilot study. Type 1 error probability associated with the null hypothesis and power were fixed at 0.05 and 80%, respectively. We used PS (v3.0.43) with estimated parameters from the pilot study and determined that we needed a sample size of at least 209 patients.
2.5
Statistical analysis
Statistical analysis involved use of R (v3.1.1). Non-normally distributed numeric variables were analysed as categorical variables. Variables with P < 0.2 on univariate analysis were entered into multivariate logistic regression. For cognitive and motor FIM MCID multivariate regressions, confounding factors were selected from univariate analysis. For 24 h/24 GABA-AG subgroup analysis, patients without GABA-AG use were used as a control. As recommended , collinearity between GABA-AG use and confounders significantly associated with FIM MCID assessment was evaluated by chi-square test or Student test, as appropriate. We did not correct for multiple comparisons because outcomes were preliminarily defined as primary and secondary . Finally, for evaluating the MCID, we calculated the standard error of measurement as follows: σ1 × √(1 − r ), where σ1 is the baseline SD and r is test–retest reliability, evaluated by the intraclass correlation coefficient between baseline and the end of hospital stay. One standard error of measurement should be a close approximation of the MCID . Statistical analysis requiring any of the global, cognitive, or motor FIM MCIDs involved use of the previously reported MCIDs , not the estimated ones. P < 0.05 was considered statistically significant.
2
Patients and methods
2.1
Patients
We recruited patients > 18 years old of both genders who were hospitalised for rehabilitation after subacute stroke (ischemic and/or hemorrhagic; i.e., 5–180 days after stroke) between January 2000 and December 2013 in a neurological rehabilitation centre (Institution de Lavigny, Switzerland). Exclusion criteria were rehabilitation duration < 14 days, occasional GABA-AG use, admission FIM > 104/126 (i.e., maximum global FIM minus global FIM MCID) and missing global FIM at admission or discharge. Moreover, we excluded patients with cognitive FIM > 32/35 on admission (i.e., maximum cognitive FIM minus cognitive FIM MCID) and motor FIM > 74/91 on admission (i.e., maximum motor FIM minus motor FIM MCID) from the cognitive and motor FIM subgroup analyses, respectively. These FIM thresholds were from global, cognitive, and motor FIM MCIDs previously reported . The study was approved by the local ethics committee (Switzerland, Vaud regional board). The patient sample is representative of classical neurological rehabilitation practice for stroke survivors usually provided in Switzerland.
2.2
Design
The primary aim was to assess an association between GABA-AG intake and reaching the global FIM MCID between admission and discharge. Secondary analyses were the association of GABA-AG intake and cognitive and motor FIM MCID and 24 h/24 GABA-AG use (i.e., drugs taken morning, noon, and evening or when the drug effect lasted > 12 h) and global, cognitive, and motor FIM. We chose the FIM because it is designed to assess the main aspects of functional recovery in patients after stroke. Moreover, to confirm the reliability of the selected MCIDs and for informative purposes, MCIDs were estimated with a distribution-based method with the standard error of measurement .
2.3
Data collection
All data were collected from digital patient records, which were systematically available between 2000 and 2013. GABA-AG intake (i.e., prescription for all days) and global FIM were systematically collected at admission and discharge. Cognitive and motor FIM were collected at admission and discharge. 24 h/24 GABA-AG intake was collected (e.g., for chlordiazepoxide, clorazepam, diazepam, flunitrazepam, flurazepam, halazepam, ketazolam, medazepam, nordazepam, prazepam, or quazepam once a day). We used the established FIM MCID of 22 points for global FIM, 3 for cognitive FIM, and 17 for motor FIM achieved between admission and discharge. We also collected potential confounding factors that could have an effect on our primary and secondary outcomes:
- •
age;
- •
sex;
- •
rehabilitation length;
- •
delay between stroke and rehabilitation;
- •
use of SSRIs or serotonin-norepinephrine reuptake inhibitors, tricyclic or tetracyclic antidepressants, antiepileptic or neuroleptic drugs;
- •
type of stroke (ischemic or hemorrhagic, left cortical with or without deep telencephalon or diencephalon, right cortical with or without deep telencephalon or diencephalon, deep telencephalon or diencephalon; infratentorial);
- •
aphasia;
- •
symptomatic stroke antecedent; anxiety or depression;
- •
neurologic or psychiatric debilitating condition (dementia, encephalopathy, multiple sclerosis, Parkinson disease, severe visual impairment, cerebral palsy, mental impairment, or psychosis);
- •
cardiac or pulmonary insufficiency; medical debilitating condition (acute care environment hospitalization, active malignancy, insulin-requiring diabetes mellitus);
- •
obstructive sleep apnoea syndrome.
2.4
Sample size determination
A previous (unpublished) pilot study of 100 eligible patients allowed for estimating sample size: 22/100 patients took GABA-AGs, among whom 6 reached the FIM MCID, and 78/100 patients did not take GABA-AGs, among whom 39 reached the FIM MCID. Confounding factors were not collected in this pilot study. Type 1 error probability associated with the null hypothesis and power were fixed at 0.05 and 80%, respectively. We used PS (v3.0.43) with estimated parameters from the pilot study and determined that we needed a sample size of at least 209 patients.
2.5
Statistical analysis
Statistical analysis involved use of R (v3.1.1). Non-normally distributed numeric variables were analysed as categorical variables. Variables with P < 0.2 on univariate analysis were entered into multivariate logistic regression. For cognitive and motor FIM MCID multivariate regressions, confounding factors were selected from univariate analysis. For 24 h/24 GABA-AG subgroup analysis, patients without GABA-AG use were used as a control. As recommended , collinearity between GABA-AG use and confounders significantly associated with FIM MCID assessment was evaluated by chi-square test or Student test, as appropriate. We did not correct for multiple comparisons because outcomes were preliminarily defined as primary and secondary . Finally, for evaluating the MCID, we calculated the standard error of measurement as follows: σ1 × √(1 − r ), where σ1 is the baseline SD and r is test–retest reliability, evaluated by the intraclass correlation coefficient between baseline and the end of hospital stay. One standard error of measurement should be a close approximation of the MCID . Statistical analysis requiring any of the global, cognitive, or motor FIM MCIDs involved use of the previously reported MCIDs , not the estimated ones. P < 0.05 was considered statistically significant.
3
Results
3.1
Recruitment and study population
Between January 1, 2000 and December 31, 2013, 572 patients were hospitalised from 5 to 180 days after one stroke episode; 434 patients were considered eligible. In total, 138 were excluded for rehabilitation length < 7 days ( n = 7), occasional GABA-AG intake ( n = 21), and starting FIM > 104/126 ( n = 110). Among the 107/434 patients receiving GABA-AGs, 39 received zolpidem, 27 lorazeam, 17 oxazepam, and 14 clomethiazole. Among the 18 patients receiving 24 h/24 GABA-AGs, 5 received zolpidem, 5 oxazepam, 4 lorazeam, and 3 clomethiazole. Cognitive and motor FIM were not documented separately in 121/434 patients. All other data were collected without missing values. For cognitive and motor FIM, 306 and 272 patients, respectively, were eligible. Clinical characteristics were globally comparable between patients with and without GABA-AGs ( Table 1 ), except anxiety and depression, which were significantly more frequent in patients with than without GABA-AG ( P < 0.001).
Related posts:
Complex regional pain syndrome associated with hyperattention rather than neglect for the healthy side: A comprehensive case study
Validation of a novel activity monitor in impaired, slow-walking, crutch-supported patients
Care-related pain and discomfort in children with motor disabilities in rehabilitation centres
An update on the pathophysiology of osteoarthritis
Does the physiatrist replicate the functions of the neurologist or orthopedist?
Takotsubo cardiomyopathy as a reversible complication of intrathecal baclofen withdrawal
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
