اختلالات روانی کلامی در بیماری پارکینسون: تفاوت های فردی در مکانیسم های شناختی زمینه ای

Although it is now well established that patients with idiopathic Parkinson's disease (PD) frequently have verbal fluency deficits [1] and [2], the issue of individual differences has yet to be addressed experimentally. Evidence presented here demonstrates that the underlying information processing causes of PD verbal fluency deficits may differ from person to person. These studies have been carried out on non-demented PD patients. The theoretical issues of case-series research and population homogeneity in cognitive neuropsychology are also discussed.

A considerable literature on verbal fluency (v.f.) [3] impairments in Parkinson's disease patients has recently emerged [2], [4], [5], [6], [7], [8] and [9]. Its roots lie in: (1) cognitive psychology investigations of the structure of natural semantic categories [10], [11], [12], [13] and [14]; (2) cognitive neuropsychology research on word-finding under non-cued conditions [15]; and (3) behavioral neurology studies of continuous, speeded, internally generated performance associated with basal ganglia-thalamo-cortical loop functions [16], [17] and [18]. Controversies such as the clinical interpretations of v.f. deficits as a partial diagnostic for dementia [19], are theoretically distinct from these experimental issues that concern the functional architecture of v.f. task performance addressed here. Evidence from previous studies in our unit [1], [2], [30] and [21] has derived primarily from group studies. We have reported that many non-demented patients with Parkinson's disease (PD) have v.f. deficits [2]. Furthermore, their v.f. performance correlated with confrontation picture-naming accuracy on the Boston Naming Test [2] and with slowed word-search rate on a Neisser-paradigm [20]. The latter experiment focused on the role of hierarchical semantic structure (superordinate, basic and subordinate levels) in natural categories, and their effect as semantic contexts on lexical-semantic retrieval [22], [23] and [24]. A group of PDs were significantly impaired compared to matched controls, on a ‘pure’ measure of word search in which motor factors such as visual scanning and response times had been individually subtracted out. As the semantic proximity of the target to the foils (non-target word items in the list) increased, response times increased. For example, recognizing ‘any dog’ (basic level items), amongst a list of ‘animals’ (superordinate items), takes longer than the same target recognition from among a list of semantically unrelated items [21]. The group effect was only significant when such motor factors had been partialled out on an individual-by-individual basis. It is important to note that in any cognitive experiment on PD patients, motor factors will have a heterogeneous effect: motor deficits will vary in pattern, type, severity and fluctuation over time, as may cognitive factors [25]. The case-series or individual case study approach, though highly time consuming, will provide more accurate results when subjects with deficits that vary over time are drawn from such a heterogeneous population [26]. The current study is therefore motivated by the above factors. This further examination of the nature of PD word-finding deficits studied the individual patterns of performance of a series of PD patients, all of whom performed the same three different measures of word-finding: the tasks varied along the axes of ‘word-search’, versus articulatory load. Word search tasks were employed that maximized lexical-semantic retrieval load while minimizing articulatory load. Conversely, automatic speech (a.s.) tasks minimized semantic search while maximizing rapid, ongoing articulation. Given that verbal v.f. requires both word search and rapid articulatory production, these search and a.s. tasks were deployed to assess the information processing locus of breakdown in PD v.f. performance. In explaining why some non-demented PDs have v.f. deficits, it is crucial to distinguish experimentally between articulatory versus semantic deficits. There may be subcategories of PDs according to these criteria, and this may affect the interpretation of correlations between word search and v.f. [20] and [21]. The current study investigates such individual differences and their information-processing significance in a group of 22 non-demented PDs.

The individual subject score profiles are shown in Table 1 and Table 2. In order to make these scores maximally comparable, the values for ‘pure’ search mean, log v.f., and a.s. are represented as Z* scores. (These are Z-scores based on the control mean and standard deviation values. A value of 1.80 standard deviations below the mean (P<0.05) was taken as the cut-off for significant impairment here.) Table 1. Parkinson's disease (N=22) subjects' Z* scores on mean ‘pure’ search, logfluency and automatic speech a PD Search Fluency Automatic speech Pattern 1 +0.17 +0.35 +0.46 normal 2 −4.67 −6.83 −2.25 all impaired 3 −1.33 −1.37 −0.48 normal 4 +0.92 −4.49 −0.71 fluency 5 −0.25 −0.61 +0.09 normal 6 −1.83 −4.33 +1.07 search & fluency 7 −0.50 −5.42 −3.65 fluency & a.s. 8 −0.25 −1.89 −1.35 fluency & a.s. 9 +0.92 +0.61 +0.38 normal 10 −0.25 −0.28 +0.11 normal 11 +1.17 +0.83 +1.15 normal 12 −2.33 −2.88 −2.71 all impaired 13 −0.42 −3.22 −3.74 fluency & a.s. 14 +0.58 −0.66 −0.34 normal 15 −7.25 −5.84 −2.25 all impaired 16 +0.67 −1.43 −5.18 a.s. 17 +0.00 −2.92 −1.72 fluency 18 −0.92 −5.71 −0.38 fluency 19 +0.17 −4.80 −0.05 fluency 20 +0.67 −0.26 +0.24 normal 21 −0.08 −0.82 −1.37 normal 22 +0.67 −0.26 +0.24 normal Impaired (N): 4 11 6 (%) 18% 50% 27% a Abbreviations: a.s.=automatic speech, search=‘pure’ search, fluency=logfluency. Table options Table 2. Control subjects’ (N=21) Z*scores on mean ‘‘pure’ search, logfluency and automatic speech Control Search Fluency Automatic speech Pattern 23 −0.17 −0.50 +0.14 normal 24 −1.00 −0.33 −0.86 normal 25 −0.33 −0.33 +0.57 normal 26 +0.33 +1.00 +0.00 normal 27 +0.50 +1.33 −0.14 normal 28 +0.00 +1.00 −0.57 normal 29 +0.17 +1.33 −1.00 normal 30 +0.67 +0.50 −2.29 a.s. 31 +0.83 +1.83 +0.71 normal 32 −0.17 −1.17 +1.14 normal 33 +0.17 +0.17 −0.14 normal 34 +0.00 +1.00 +1.43 normal 35 +0.33 +0.17 +0.29 normal 36 −2.00 −2.50 −1.14 search & fluency 37 +0.00 +0.50 +0.86 normal 38 +3.00 −0.67 −0.29 normal 39 −0.33 −0.17 −1.29 normal 40 +0.00 −1.17 +2.14 normal 41 +0.83 +1.00 +0.29 normal 42 −0.17 −1.00 +0.29 normal 43 −0.50 +0.33 −1.29 normal Table options In the PD group, 18% were impaired on word-search, 50% were impaired on v.f., and 27% were impaired on a.s.. The percentage of PDs with values falling outside the control range was 9% on search, 50% on v.f. and 32% on a.s. Several different patterns of performance across the three different word finding tasks were found within this series of PD cases. The patterns of performance are classified as: normal (unimpaired)=46% PD, 85% controls; impaired on all tasks=14% PD, 0% controls; v.f. and search only impaired=5% PDs, 5% controls; v.f. only impaired=18% PDs, 0% controls; v.f. and a.s. only impaired=14% PDs, 0% controls; a.s. only impaired=5% PDs, 5% controls. Correlations between the tasks for PDs showed there to be a significant correlation between v.f. and search only (r=0.63, df=20, P<0.01). (The PD search by a.s. correlation was r=0.28; and that between v.f. and a.s. was r=0.41.) None of the between-task correlations reached significance in the control subjects (v.f. by search: r=0.39; search by a.s.: r=0.11; v.f. by a.s.: r=−0.04).