Original Paper Received: October 19, 2009 Accepted: February 15, 2010 Published online: April 9, 2010
Eur Neurol 2010;63:279–284 DOI: 10.1159/000287583
Gender-Specific Predictors of Functional Outcome after Stroke Rehabilitation: Potential Role of the Autonomic Nervous System Andrea Bassi a Furio Colivicchi b Massimo Santini b Carlo Caltagirone a, c
a c
IRCCS Santa Lucia Foundation, b Cardiovascular Department, San Filippo Neri Hospital, and Department of Neurology, University of Rome ‘Tor Vergata’, Rome, Italy
Key Words Acute ischemic stroke ⴢ Cardiac autonomic balance ⴢ Heart rate variability
95% CI 2.34–71.4, p = 0.006). HRV does not predict functional outcome after rehabilitation in women. Instead, insular damage appears to have a role in determining the final results of rehabilitation in women but not in men. Copyright © 2010 S. Karger AG, Basel
Abstract The aim of this study was to assess the impact of cardiac autonomic derangement on gender-associated functional outcome of patients with subacute ischemic stroke undergoing a hospital-based rehabilitation program. The study population included 126 consecutive first-ever stroke survivors (mean age 59.7 8 11.6 years). Time-domain measures of heart rate variability (HRV) by 24-hour Holter monitoring (HM) were considered in all cases. By the end of the rehabilitation program an unfavorable functional outcome with dependency (Barthel Index score of !75) was found in 27 men (40.3%) and in 31 women (52.5%; p = 0.168). Multivariate analysis demonstrated that high age, low Barthel Index score and low Rankin Scale score on admission were independent predictors of an unfavorable functional outcome in both men and women. The standard deviation of normal-to-normal RR intervals on HM was an independent predictor of rehabilitation outcome only in men (OR 15.29, 95% CI 2.47– 46.58, p = 0.001). The presence of insular damage on neuroimaging studies was independently associated with an unfavorable functional outcome only in female (OR 18.89,
© 2010 S. Karger AG, Basel 0014–3022/10/0635–0279$26.00/0 Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/ene
Introduction
Hospital-based rehabilitation of patients with recent stroke appears to be effective, but represents a limited resource and its effectiveness could be possibly improved by appropriate patient selection [1]. In fact, the great variability in rehabilitation results has led to a particular interest in the identification of accurate predictors of rehabilitation outcome [1]. As a matter of fact, age, initial stroke severity and the presence of significant major comorbidities are known to largely predict the results of a rehabilitation program in most cases. Besides, the presence of an autonomic nervous system dysfunction, as expressed by decreased heart rate variability (HRV), may also contribute to a less-favorable rehabilitation outcome [2]. Women have been reported to experience a less-satisfactory level of functional recovery after stroke, even when an in-patient rehabilitation program is followed [3– 6]. A complex interplay of psychosocial, clinical and enAndrea Bassi, MD Via Gerolamo Belloni, 94 IT–00191 Rome (Italy) Tel. +39 06 5150 1409, Fax +39 06 5150 1408 E-Mail a.bassi @ hsantalucia.it
vironmental factors has been repeatedly advocated in order to explain such sex differences in rehabilitation results [7]. However, even if the autonomic nervous system appears to act differently in women [8–12], the specific impact of autonomic dysfunction on rehabilitation outcome in female stroke survivors is still unknown. This prospective study was designed and undertaken to assess the impact of cardiac autonomic derangement on gender-associated functional outcome and loss of autonomy of patients with subacute ischemic stroke undergoing a hospital-based rehabilitation program.
Patients and Methods Consecutive stroke inpatient survivors admitted to our rehabilitation institution were prospectively screened for inclusion in the study in a 12-month period. Our institution includes 318 beds and is a large, free-standing, university-affiliated rehabilitation institute. Admission to our institution is possible for all recent stroke survivors with functional disability (International Classification of Diseases, 9th revision, diagnosis codes 431 through 438) but without severe concurrent medical conditions, that may potentially contraindicate physical therapy. During the 12-month screening period 496 recent stroke survivors (243 men, 48.9%, and 253 women, 51.1%) were admitted to our institutions, with a mean rehabilitation length of stay of 59 8 28 days. Patients were included in the study only if they fulfilled all of the following criteria: (1) admission after a first-ever acute ischemic stroke; (2) absence of diabetes mellitus or any other concomitant nervous system, cardiac, or pulmonary disease possibly affecting the autonomic nervous system and HRV; (3) absence of any clinically relevant arrhythmia on admission, including atrial fibrillation; (4) absence of any pharmacological treatment, including betablockers, possibly affecting the autonomic nervous system and HRV; (5) absence of any major concurrent illness, including renal failure and malignancies; (6) absence of fever, hypoxia, severe hypertension, alterations in consciousness, or any relevant hemodynamic compromise on admission. As in previous studies [2, 13, 14], such strict inclusion criteria were chosen in order to select a homogeneous population of patients in whom any abnormality in HRV could be possibly determined only by the cerebral ischemic lesions. The study protocol was approved by the Ethics Committee of our institution. After prospective selection, 126 consecutive patients (67 men and 59 women; mean age 59.7 8 11.6 years) fulfilled all of the abovereported criteria, provided informed consent and were included in the study. No patient had received thrombolytic therapy during the index event. In fact, in the city of Rome only three major hospitals are currently authorized by the National Health Authority to perform thrombolytic therapy in acute stroke. Consequently, less than 1% of all acute stroke patients currently receive thrombolytic therapy in this geographic area.
280
Eur Neurol 2010;63:279–284
On admission, patients were submitted to clinical, neurological, and functional examinations. Stroke severity was assessed by the National Institutes of Health (NIH) stroke scale [15], autonomy in activities of daily living (ADL) by the Barthel Index (BI) [16], and global disability by the modified Rankin scale (RS) [17]. Independence in ADL and global disability were also evaluated at discharge after completion of the rehabilitation treatment. Neuroimaging studies (CT or MRI) were performed on admission in all patients to confirm brain infarct size and localization. As previously described [2, 13, 14], the presence of insular involvement was assessed on the basis of brain imaging by an experienced neuroradiologist blinded to clinical details. The presumed etiology of strokes was defined by the attending physician on the basis of clinical judgment and laboratory features, and subsequently classified according to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria [18]. In all cases, particular care was taken to exclude any major form of concurrent cardiac disease with a possible impact on HRV, as previously described [2, 13, 14]. In particular, congestive heart failure, moderate-to-severe valvular dysfunction, any form of cardiomyopathy, previous acute myocardial infarction, and left ventricular hypertrophy or dysfunction were ruled out by a comprehensive clinical evaluation, which included history, physical examination, 12-lead electrocardiography and echocardiography. All patients underwent 24-hour Holter monitoring (HM) on admission before the beginning of the rehabilitation program. HM was performed using a 3-channel bipolar recorder and was evaluated after digitization, as previously described [2, 13, 14]. The mean sinus heart rate was derived from the mean RR intervals (after exclusion of abnormal beats). HRV was analyzed from the Holter recordings using commercially available software (ELA Medical Synescope version 1.0 analysis system). The beat classification was verified and corrected appropriately by an experienced cardiologist, blinded to clinical details. The following time-domain measures of HRV were obtained by using the continuous data for 24 h [2, 13, 14]: (1) standard deviation of all normal-to-normal RR intervals (SDNN); (2) standard deviation of the averages of all normal-to-normal RR intervals intervals in all 5-minute segments of the entire recording (SDANN); (3) root-mean-square of differences of adjacent normal-to-normal RR intervals (rMSSD); (4) number of normal-to-normal RR intervals differing by more than 50 ms from adjacent interval divided by the total number of all normal-to-normal RR intervals (pNN50). Rehabilitation and Primary End Point In all patients rehabilitation was started within 30 days from the index event, with a mean time interval between the index event and rehabilitation beginning of 12.3 8 4.4 days. The rehabilitation treatment had a standard duration of 60 days and was designed by the same physiatrist for all patients, being essentially based on practical ADL skills. Individual physiotherapy was performed for 60 min twice a day for 6 days a week. Rehabilitation was started within 48 h of admission. If necessary, patients had access to daily training for unilateral spatial neglect or speech therapy or to individual training for swallowing, bowel, and bladder dysfunction. Physiotherapy and language treatment continued throughout the hospital stay. As in previous studies [19], the primary end point of the study was the occurrence of an unfavor-
Bassi /Colivicchi /Santini /Caltagirone
Table 1. General characteristics of the study population
Age, years Right-sided stroke, patients (%) Insular damage, patients (%) NIH stroke scale score on admission BI score on admission RS score on admission NIH stroke scale score at discharge BI score at discharge RS score at discharge Unfavorable functional outcome number (%) Urinary incontinence on admission, patients (%) Stroke subtypes Large vessel disease, patients (%) Cardioembolic, patients (%) Undetermined etiology, patients (%)
Total cohort (n = 126) Men (n = 67)
Women (n = 59)
p value*
59.7811.6 72 (57.1) 51 (40.4) 7.883.7 48.0819.9 3.880.4 4.782.8 74.0819.6 3.080.5 58 (46.0) 39 (30.9)
59.389.8 40 (59.7) 26 (38.8) 7.783.2 49.4818.2 3.780.3 4.582.2 77.2817.3 2.980.6 27 (40.3) 21 (31.3)
60.1813.4 32 (54.2) 25 (42.3) 8.084.2 46.5821.8 3.980.6 5.083.3 70.4821.5 3.080.7 31 (52.5) 18 (30.5)
0.710 0.536 0.683 0.553 0.413 0.378 0.388 0.058 0.389 0.168 0.919
59 (46.8) 14 (11.1) 53 (42.1)
32 (47.7) 8 (11.9) 27 (40.4)
27 (45.8) 6 (10.1) 26 (44.1)
0.822 0.752 0.668
74 (58.7) 27 (21.5) 25 (19.8)
41 (61.2) 15 (22.4) 11 (16.4)
33 (56.0) 12 (20.3) 14 (23.7)
0.549 0.779 0.304
Localization of ischemia Cortico-subcortical, patients (%) Cortical, patients (%) Subcortical, patients (%)
* Comparison between men and women. NIH = National Institute of Health; BI = Barthel Index; RS = Rankin scale.
able functional outcome with dependency at discharge after the standardized 60-day rehabilitation program, defined as a BI score of !75. Physicians performing the final assessment of global disability and independence in ADL were blinded to the results of HM. Statistical Analysis Means (8 SD) were calculated for continuous variables, while frequencies were measured for categorical variables. Distributions of continuous variables were determined by the Kolmogorov-Smirnov test. In case of normal distribution, group differences for continuous data were examined by unpaired Student’s t test, while the Mann-Whitney two-sample test was applied in case of non-normal distribution. Group differences for categorical variables were examined by 2 or Fisher exact test, as appropriate. In particular, the Fisher exact tests was applied in case of an expected frequency of !5. The association between baseline clinical and laboratory features on admission, including HRV findings, and the primary end-point of the study (i.e. unfavorable functional rehabilitation outcome with dependency at discharge after 60-day rehabilitation program) was tested by means of multivariate logistic regression analysis. In order to define gender-specific outcome predictors, multivariate analysis was separately performed in men and women. All variables, as determined from the baseline evaluation on admission, were considered potential predictors of the study end point. Subsequently, only those variables possibly associated with the outcome of interest (p !0.25 after univariate analysis) were included in the multivariate models. Variables for which the p value exceeded 0.05 were eliminated in a stepwise fashion so that only those that had a statistically sig-
Gender Differences, Outcome after Stroke
nificant association with the outcome of interest were included in the final regression models. Final models were evaluated for goodness of fit by the Hosmer–Lemeshow test. The results are presented as the odds ratio (OR) and the 95% confidence interval (95% CI). As in previous studies [2, 13, 14], for the purpose of multivariate analyses, patients were categorized in two groups with lower (SDNN !100 ms), or higher SDNN values (SDNN 1100 ms). The cut-off value of 100 ms for SDNN was chosen as it allowed the best discrimination for the primary end point, as assessed by the receiver operating characteristic (ROC) curve [21]. Data analysis was performed by using the SPSS statistical software package (SPSS 12.0). p ! 0.05 was considered statistically significant.
Results
In the whole study population, the 60-day rehabilitation program yielded a significant functional improvement with an increase in mean BI score from 48.0 8 19.9 to 74.0 8 19.6, while mean RS score decreased from 3.8 8 0.4 to 3.0 8 0.5. As to the primary end point, by the end of the 60-day rehabilitation program an unfavorable functional outcome with dependency (BI score of !75) was found in 58 patients (46.0%). Baseline clinical characteristics of the entire study cohort, as well as unadjusted comparisons between men and women are shown in table 1. Men and women did not differ in terms of age, Eur Neurol 2010;63:279–284
281
Table 2. Heart rate variability measures in the study population
Heart rate, bpm SDNN, ms SDNN
