ناتوانی در ادراک بیماری، غفلت، انقراض و سایت آسیب اختلال در زندگی روزمره به دنبال سکته مغزی نیمکره راست را پیش بینی می کند

Abstract Introduction Right-hemispheric stroke can give rise to manifold neuropsychological deficits, in particular, impairments of spatial perception which are often accompanied by reduced self-awareness of these deficits (anosognosia). To date, the specific contribution of these deficits to a patient's difficulties in daily life activities remains to be elucidated. Methods In 55 patients with right-hemispheric stroke we investigated the predictive value of different neglect-related symptoms, visual extinction and anosognosia for the performance of standardized activities of daily living (ADL). The additional impact of lesion location was examined using voxel-based lesion-symptom mapping. Results Step-wise linear regression revealed that anosognosia for visuospatial deficits was the most important predictor for performance in standardized ADL. In addition, motor-intentional and perceptual-attentional neglect, extinction and cancellation task performance significantly predicted ADL performance. Lesions comprising the right frontal and cingulate cortex and adjacent white matter explained additional variance in the performance of standardized ADL, in that damage to these areas was related to lower performance than predicted by the regression model only.

1. Introduction Right-hemispheric stroke frequently gives rise to severe impairments of visuospatial processing resulting in spatial neglect or extinction. While neglect patients generally show reduced awareness for contralesional sensory events, extinction patients fail to respond to contralesional events when simultaneous ipsilesional stimulation is present (Vossel et al., 2011). Furthermore, neglect is a multifaceted disorder and can, for example, differentially affect perceptual-attentional as well as motor-intentional functions (e.g., Bisiach et al., 1998; Vallar, 2001; Vossel et al., 2010). Right-hemispheric lesions and neglect are moreover closely associated with anosognosia for both motor impairment (Pia et al., 2004; Orfei et al., 2007) and visuospatial deficits (Vossel et al., 2012), so that the patients show reduced self-awareness for their stroke-induced functional impairments. While the cognitive processes and the underlying lesion anatomy of each of the above-mentioned symptoms and syndromes after stroke have been investigated by several studies, their functional significance for activities of daily living (ADL) to date remains less clear. For example, while paper-and-pencil or computer-based tests of neglect or extinction might be very sensitive in detecting visuospatial impairment, whether and how such deficits translate into difficulties in day-to-day activities remains to be established. In a previous study, we could show that performance in paper-and-pencil neglect tests and performance in standardized ADL correlate significantly, suggesting con-siderable ecological validity of the conventional paper-and-pencil neglect tests (Eschenbeck et al., 2010; see also Azouvi et al., 2002; Wilson et al., 1987a). However, despite the close relation of task performance in both test batteries, a considerable amount of patients showed neglect in neuropsychological tests, but not in standardized ADL (while the reverse pattern was observed in one of the examined patients only; Eschenbeck et al., 2010; see Wilson et al., 1987a for similar results). Hence, not all deficits that are detected by neuropsychological paper-and-pencil tests translate into impaired ADL. The present study used a step-wise linear regression analysis to investigate how different neglect-related symptoms (as assessed with paper-and-pencil tests), visual extinction (as assessed by a computer test), as well as the self-awareness for visuospatial deficits (i.e., anosognosia) impact on the performance in standardized ADL. Though it is well established that the presence of neglect significantly impacts upon rehabilitation (Jehkonen et al., 2006; Stone et al., 1992), the influence of extinction or anosognosia for neglect upon ADL or rehabilitation to date has hardly been investigated. In the tactile modality, one study has already provided preliminary evidence for a significant impact of extinction for predicting a patient's functional status (Rose et al., 1994). The present study moreover used statistical voxel-based lesion mapping to test whether lesion site can explain additional variability in ADL performance over and above what is explained by the neuropsychological predictors.

3. Results 3.1. Variables predicting difficulties in ADL Table 2 depicts the descriptive statistics for the variables of interest. There were no differences in any of the listed variables between the three different patient subgroups [group 1 (<1 month after stroke, n = 31), group 2 (>1 month but <3 months after stroke, n = 13), group 3 (>3 months after stroke, n = 11)]. The step-wise linear regression revealed that a set of five variables was optimal for the prediction of ADL (AICC = 74.21). The most important predictor variable was anosognosia (self-awareness for visuospatial deficits), followed by motor-intentional and perceptual-attentional neglect (as reflected in the RB and PB scores of the Landmark-M test), visual extinction, and the LQ from the cancellation tests (see Fig. 1). The coefficient of determination (R2) for this regression model amounted to .88. There were no significant differences in the squared residuals when patients were divided into subgroups according to time post-stroke. Table 2. Descriptive statistics of the neuropsychological variables. Mean SD Min Max External rating in ADL tasks 36.85 5.23 17 40 Mean LQ −.12 .29 −1.00 .04 PB, Landmark-M test 54.80 7.40 43.52 81.48 RB, Landmark-M test 50.32 3.51 38.89 62.96 Visual extinction index of the computer task 4.98 19.93 −29.00 70.00 Anosognosia index −.16 .38 −1.7 .12 Table options Illustration of the role of anosognosia (self-awareness for visuospatial ... Fig. 1. Illustration of the role of anosognosia (self-awareness for visuospatial deficits), RB and PB in the Landmark-M test, visual extinction and cancellation task performance (laterality quotient) for predicting deficits in ADL tests. The connecting bars are scaled relative to the importance of the different independent variables (as indexed by the residual sum of squares with the predictor removed from the regression model; normalized to a total score of all predictor variables of 1). Figure options Age and time post-stroke did not considerably contribute to the prediction of ADL and were hence removed from the regression model. Table 3 summarizes the statistics of the regression coefficients of the variables that entered the regression model. Fig. 2 shows the relationship between predicted and observed performance in ADL. Table 3. Statistics of the step-wise regression analysis for variables entering the regression model. Regression coefficient t-score p Mean LQ 4.08 2.18 .034 PB, Landmark-M test −.14 −3.33 .002 RB, Landmark-M test −.38 −4.76 <.001 Visual extinction index of the computer task −.03 −2.45 .018 Anosognosia index 7.34 5.98 <.001 Table options Illustration of the relationship between observed and predicted score in the ADL ... Fig. 2. Illustration of the relationship between observed and predicted score in the ADL tests. Patient subgroups with different time post-stroke are depicted in different colours. Figure options 3.2. Additional effects of lesion site on ADL The five variables in the regression model (self-awareness, RB, PB, visual extinction and LQ) explained 88% of the variance in ADL. To test whether lesion location explains additional variance, the residuals from the regression analysis (i.e., observed minus predicted total ADL scores) were entered in a VLSM analysis. This analysis revealed that damage to the frontal cortex (in the middle frontal gyrus, adjacent to the inferior frontal sulcus) and the mid cingulate cortex (and neighbouring white matter) was associated with poorer ADL performance than predicted by the regression model only (false discovery rate (pFDR) < .05; see Fig. 3). Results of the voxel-based lesion-symptom mapping. For illustrative purposes, ... Fig. 3. Results of the voxel-based lesion-symptom mapping. For illustrative purposes, results are depicted at a significance level of p < .005 (uncorrected for multiple comparisons). Numbers denote Montreal Neurological Institute (MNI) z-coordinates. Figure options To rule out that the observed results of the lesion analysis (particularly in frontal cortex) were confounded by presence of anosognosia for hemiplegia, we re-calculated the analysis with excluding those 10 patients who showed a discrepancy between self- and external rating of ≥1 in the AQ. This lesion analysis in n = 45 patients revealed almost identical results (in the same frontal and cingulate brain regions) as the analysis based upon the whole patient sample. 3.3. Inter-rater-reliability of the ADL rating The correlation between the rating (total score) of the ADL by the three independent raters was very high (mean Pearson's correlation coefficient r = .96, all 3 correlation coefficients >.95, p < .001), indicating good inter-rater-agreement in the evaluation of the ADL task performance.