Review Article - Neuropsychiatry (2018) Volume 8, Issue 2

Broad Impairment of Executive Functions in Patients with Parkinson’s disease: A Meta-Analysis

Corresponding Author:
Jin H Yan
Laboratory of Neuromotor Control and Learning, Shenzhen University, 3688 Nan Hai Avenue, Shenzhen, 518060, China
Tel: 86-0755-8617-2031
Fax: 86-0755-2671-6888

Abstract

Abstract

Objective: Executive functions (EF) impairments have been observed in patients with Parkinson’s disease (PD). However, the pattern of EF deficits in this population remains unclear. This article aimed to examine the influence of PD on different EF domains through meta-analysis of published data.

Methods: This article aimed to compare the EFs of PD patients with those of healthy controls (CON) in different EF domains. We searched electronic databases for articles reporting comparisons of EF performance between non-demented/non-depressed PD patients and CON. Accordingly, we identified 140 studies investigating 6 EF domains (attention, inhibition, planning, reasoning, shifting and working memory) in 4683 PD patients and 4247 CON.

 Results Results showed that PD patients exhibited impaired attention (Hedges’ g= -0.48), inhibition (Hedges’ g=-0.48), planning (Hedges’ g = -0.49), reasoning (Hedges’ g = -0.31), shifting (Hedges’ g = -0.55) and working memory (Hedges’ g = -0.53). They exhibited a moderately impaired overall EF (Hedges’ g = -0.49). EF deficits were not moderated by age, years of education, disease severity, motor deficits, disease duration, medication dose or global cognition.

Conclusions: The findings suggest that among PD patients, EFs in which reasoning is least affected are broadly impaired

Keywords

Executive functions, Meta-analysis, Movement disorder, Parkinson’s disease

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by dopamine depletion subsequent to the loss of dopaminergic neurons in the substantia nigra pars compacta [1]. The most salient symptoms exhibited by PD patients are motor impairments, including bradykinesia, rigidity, resting tremor, gait problems and postural instability [2-4]. In industrialised countries, PD affects 1% of older adults (age > 60 years) [5], and estimates suggest that the population of PD patients will reach 8.7–9.3 million worldwide by 2030 [6]. This increase is expected to place significant burdens on caregivers and healthcare systems in the future.

Recent research has demonstrated that in addition to motor impairments, PD patients exhibit decreases in cognitive capabilities, particularly executive functions (EF) [7]. EFs comprise a set of inter-related, effortful cognitive processes directed toward goal-directed behaviors [8]. According to the unity/diversity model, 3 core EFs exist: inhibition, working memory and shifting [9]. Inhibition refers to the suppression and control of attention, thoughts and responses required to reach a goal. Working memory describes the ability to retain information temporarily for processing and manipulation [10]. Shifting is defined as changing between task sets or response rules. The core EFs serve as foundation for higher-order EFs, such as planning and reasoning [11]. Planning involves the identification and organization of a sequence of steps needed to achieve a goal [12], whereas reasoning describes the application of knowledge to draw logical inferences [13]. As EFs enable us to address a variety of everyday tasks in a flexible manner, EF impairments can reduce the quality of life and functional outcomes of PD patients.

In addition to declines in behavioral performances, EF dysfunctions exhibited by PD patients may be related to abnormal activity in the basal ganglia and dorsolateral prefrontal cortex [14-17]. As a result, compensatory brain activity is often observed in regions related to EFs. For instance, in the Tower of London task, PD patients can normally activate the prefrontal cortex, despite harbouring subcortical lesions, and can additionally activate the hippocampus as a compensatory mechanism [14]. Regarding set shifting, the behavioral performances of PD patients and healthy controls were comparable, although PD patients exhibited increased activation in the inferior parietal cortex and superior frontal gyrus that likely indicated certain neuro-compensatory mechanisms [18]. During the n-back task, PD patients exhibited reduced connectivity between the dorsolateral prefrontal cortex and other task-related regions, indicating deficits in the executive network. These findings suggested that hyperactivity in the dorsolateral prefrontal cortex, caudate nucleus and inferior parietal cortex plays a crucial role in counteracting decreases in the EF performance in PD patients [19]. Hence, EF dysfunctions in PD patients can be associated with dysfunctional frontostriatal loops caused by dopamine pathway abnormalities.

According to a previous meta-analysis, PD patients exhibited dysfunctional EFs with effect sizes (Hedges’ g) ranging from -0.43 to -0.94 [7]. However, that meta-analysis failed to adequately address EF deficits in PD patients. First, the numbers of included studies related to different EF tasks were rather small (k = 2–14). Second, the authors did not report results for separate EFs, and thus it was difficult to compare the extents of deficits among different EFs. To overcome these limitations in our understanding, we conducted a meta-analysis to quantitatively summarize the existing results and compare different EFs between PD patients and healthy controls. Moreover, as a handful of studies have suggested the moderating effects of age, education, overall cognitive ability and PD-associated motor deficits on EF [20-24], we also examined the potential moderators of cognitive declines in PD patients. The results might provide insights for practitioners and clinicians that would allow them to focus on the most impaired cognitive abilities and devise suitable strategies to improve the functional abilities and quality of life of PD patients.

Methods

▪ Inclusion and exclusion criteria

In this review, we included studies that measured EFs in PD patients and healthy controls. Inclusion was limited to English-language research articles published in peer-reviewed journals that provided sufficient data for effect size computation. Additionally, patients in the included studies should not have been clinically depressed or demented or exhibited other neurological diseases. Commentaries, review articles and studies including PD patients experiencing deep brain stimulation were excluded.

▪ Literature search

The studies were identified by searching electronic databases and scanning the bibliographies of articles published from the first available date to May 10, 2016. The electronic databases Web of Knowledge, PubMed, Medline and PsycINFO were searched for literature using the term ‘Parkinson’s disease’ in combination with the terms ‘executive functions’, ‘working memory’, ‘cognitive control’, ‘inhibition’, ‘set shifting’, ‘flexibility’, ‘planning’, ‘reasoning’ or ‘task switching’.

▪ Data extraction

We developed an electronic spreadsheet for data extraction. The screening and eligibility assessment was performed by 3 reviewers in a non-blinded, standardized manner. During training, the reviewers assessed 300 articles and achieved a rate of agreement above 90%. Disagreements between the reviewers were resolved by consensus. Subsequently, the remaining articles were divided into 3 groups and each was assigned to a reviewer who performed data extraction and coding. The spreadsheet completed by each reviewer was double-checked by another reviewer.

The following information was extracted from each included article: (1) participants’ characteristics, including the mean age, years of education, global cognition (Mini-Mental State Examination score, MMSE) [25], disease severity (Hoehn and Yahr stage, HY) [26], motor deficits (motor score of the Unified Parkinson’s Disease Rating Scale motor, UPDRS) [27], disease duration and dose of medication (levodopa equivalent daily dose, LEDD); (2) EF domain and tasks used and (3) behavioral outcomes (EF performance).

Statistical Analyses

EF was the primary outcome measure. Mean data were converted to Hedges’ g for the metaanalysis. Multiple tests that assessed the same EF in one study were combined. The data analysis was conducted in Comprehensive Meta-Analysis 2.0. A random-effects model was used to account for variability among the samples and assessment tools across the included studies. A positive Hedges’ g indicated a better EF performance among PD patients relative to healthy controls.

▪ Heterogeneity

Cochran’s Q test was used to assess heterogeneity. An I2 statistic was included to quantify the proportion of heterogeneity across studies that could not be explained by chance (I2 values of 25%, 50%, and 75% corresponded to low, moderate, and high heterogeneity, respectively) [28].

▪ Publication bias

Publication bias was assessed through a visual inspection of funnel plots and Egger’s asymmetry test [29]. In a funnel plot, asymmetry can result from the non-publication of null or negative results. In Egger’s asymmetry test, the standardized effect estimate (effect size/standard error) is regressed on precision (1/standard error). A significant deviation of the y-intercept from zero might indicate the presence of publication bias. In addition, a fail-safe N test was also used to determine the number of hypothetical missing studies required to nullify the overall effects. A fail-safe N that exceeds the threshold (N ≥ 5k+10) has been well accepted as an indicator of a meta-analytic result robust to publication bias [30]. The trim-and-fill method was used to calculate an adjusted effect size corrected for the effects of missing studies in asymmetrical funnel plots [31].

▪ Meta-regression

The influences of age, years of education, disease severity, motor deficits, disease duration and global cognition and dose of medication on the study outcome were assessed through a mixedeffects meta-regression analysis based on an unrestricted maximum likelihood model.

Results

One hundred and forty articles involving 4639 PD patients and 4219 healthy controls were retrieved and included in the meta-analysis. Figure 1 shows the number of studies retained at different stages of the literature search and screening process. Details of the individual studies are presented in Table 1. The mean ages of the PD patients and controls were 64.34 and 63.81 years, respectively. A total of 275 effect sizes were included in the meta-analysis: tapping on attention (k = 11, NPD = 556, NCON = 521), inhibition (k = 56, NPD = 1837, NCON = 1572), planning (k = 22, NPD = 771, NCON = 698), reasoning (k = 13, NPD = 429, NCON = 419), shifting (k = 82, NPD = 2827, NCON = 2648) and working memory (k = 91, NPD = 2901, NCON = 2650).

Study Sample size (PD/CON) Age (PD/CON) Education (yrs) MMSE HY UPDRS (motor) Disease duration (Mo) LEDD (mg) Assessment Outcome Hedges’ g
(95% CI)
Attention
Agosta et al. [39] 41/34 64/63 11 27.7 NA 27.1 85.2 NA ACE-R Test score -0.80 (-1.27, -0.34)
Colman et al. [40] 27/25 61.39/62.93 13.21 28.11 1.79 15.68 72.48 786.94 Sustained Attention Omissions -0.42 (-0.97, 0.12)
Crescentini et al. [41] 19/14 66.7/65.6 8.1 29 2.05 26.2 76.8 540.3 Divided Attention Omissions -0.24 (-0.93, 0.44)
Crescentini et al. [42] 16/14 63.6/61.6 9.7 28.8 1.94 22.9 72 386.4 TAP-A Omissions -0.03 (-0.73, 0.67)
Duncan et al. [43] 125/50 66/65.8 12 29 2 26.8 6.15 175 CDR Time of completion -0.67 (-1.01, -0.34)
Elwan et al. [44] 19/25 61.84/56.35 NA 26.57 2.67 47.90 28.60 NA PASAT Test score 0.41 (-0.59, 1.41)
Ibarretxe-Bilbao et al. [45] 24/24 56.13/57.58 10.96 29.63 1.73 14.67 36.72 299.58 CCPT Detectability -0.68 (-1.25, -0.10)
Lord et al. [46] 121/184 67/69.4 NA NA 1.94 25.5 NA 124.6 POA Time of completion -0.45 (-0.69, -0.22)
Murray and Rutledge, [47] 11/9 68.36/67.89 16 NA NA 17.91 NA NA TEA Composite score -0.69 (-1.56, 0.18)
Poletti et al. [48] 103/100 66/66.8 9.20 27.90 NA 16.90 10.70 NA Visual Search Number of correct response -0.32 (-0.60, -0.04)
Rodriguez-Ferreiro et al. [49] 50/42 72.92/74.08 7.04 26.94 1.96 NA 114.00 NA Visual Search Number of correct response -0.77 (-1.19, -0.35)
Inhibition
Abe et al. [50] 32/20 65.9/65.5 11.7 28.3 2.5 18.1 49.2 507 Stroop Task Accuracy (incongruent) -0.10 (-0.65, 0.46)
Anderson et al. [51] 27/16 63.8/65.4 NA NA 2.7 24.9 NA NA Hayling Test Accuracy -0.21 (-0.82, 0.40)
Baggio et al. [52] 39/23 63.5/61 11.4 28.7 1.8 16.5 67.2 560.3 Stroop Task Interference -0.69 (-1.21, -0.17)
Baggio et al. [53] 43/36 64/63.4 10.8 29.35 1.58 14.1 73.2 646.7 Stroop Task Interference -0.26 (-0.70, 0.18)
Barnes and Boubert, [54] 20/20 63.73/68.54 NA 27.2 2.9 NA 116.64 465 Stroop Task Interference -0.82 (-1.46, -0.19)
Beste et al. [55] 15/17 66.8/65.2 NA NA NA 15.90 NA NA Flanker Task Error (incongruent) -0.22 (-0.89, 0.45)
Beyer et al. [56] 114/99 65.8/65 11.3 28.3 1.8 21 NA NA Stroop Task Test score (incongruent) -0.16 (-0.43, 0.10)
Bezdicek et al. [57] 46/41 57.7/60.78 14.84 NA 1.95 10.3 124.8 1029 Stroop Task Interference 0.01 (-0.41, 0.42)
Bohlhalter et al. [58] 12/12 59.1/46.6 NA NA NA 16.5 92.4 766 Stroop Task Interference -0.29 (-1.07, 0.48)
Bohnen et al. [59] 13/14 70.8/69.7 NA 28.1 NA 26.4 70.8 NA Stroop Task Test score (incongruent) -0.96 (-1.74, -0.19)
Broeders et al. [60] 59/40 62.5/61.4 11.60 27.90 1.70 16.00 17.50 153.90 Stroop Task Time of completion (incongruent) -0.64 (-1.05, -0.23)
Cohen et al. [61] 13/16 65.3/66.6 4.8 NA 2.1 29.9 78 714 Stroop Task Interference -0.28 (-1.00, 0.44)
Colman et al. [41] 27/25 61.39/62.93 13.21 28.11 1.79 15.68 72.48 786.94 Stroop Task Time for color-word card divided by time for color card -0.17 (-0.71, 0.37)
Crescentini et al. [42] 19/14 66.7/65.6 8.1 29 2.05 26.2 76.8 540.3 Stroop Task Test score (incongruent) 0.46 (-0.23, 1.14)
Crescentini et al. [43] 16/14 63.6/61.6 9.7 28.8 1.94 22.9 72 386.4 Stroop Task Test score (incongruent) -0.69 (-1.41, 0.03)
Dujardin et al. [63] 18/18 60.17/59.5 11.56 NA NA 17.58 11.33 NA Stroop Task Interference -0.79 (-1.45, -0.12)
Edelstyn et al. [64] 17/17 65.4/64.5 NA 28 NA NA 92.4 448.59 Hayling Test Test score -0.83 (-1.52, -0.15)
Ell, 2013 [65] 36/35 71.1/65.8 16.7 NA 2 NA 42 NA DKEFS-CWI (inhibition) Time of completion -0.39 (-0.86, 0.08)
Fales et al. [66] 21/25 66.9/68.8 16.9 28.8 2 NA 69.6 NA Stroop Task Interference 0.26 (-0.31, 0.83)
Fling et al. [67] 15/12 65.5/66.7 NA NA 2 29.3 76.8 485 Stroop Task Time of completion (incongruent) -0.41 (-1.16, 0.33)
Galtier et al. [68] 43/20 59.19/60.85 7.88 27.58 2.28 28.46 99.60 NA Stroop Task Interference 0.28 (-0.24, 0.81)
Gawrys et al. [69] 30/18 56.03/57.11 13.51 28.93 2 NA NA 851.58 Stroop Task Time of completion (subtest 3 - 1) -0.74 (-1.34, -0.15)
Green et al. [70] 10/10 54.05/53.9 16.10 NA 1.30 9.20 NA NA Stroop Task Interference -0.06 (-0.90, 0.78)
Hausdorff et al. [71] 30/25 71.3/70 14.1 28.1 NA 18.1 NA NA Go/Nogo Task Performance index -0.71 (-1.26, -0.16)
Hsieh et al. [72] 26/27 63.3/63.48 9.07 NA NA NA 40.80 NA Stroop Task Interference -0.52 (-1.05, 0.02)
Koerts et al. [73] 43/25 63.7/62.8 5.2 27.5 2.2 24.6 61.2 561.7 Stroop Task Interference -0.29 (-0.78, 0.20)
Koerts et al. [74] 88/65 62.5/61.9 NA 27.5 2 21.4 67.2 550.3 Stroop Task Interference -0.66 (-1.00, -0.32)
Koerts et al. [75] 43/25 63.6/62.8 5.30 27.50 2.20 24.60 63.60 583.70 Stroop Task Interference -1.07 (-1.59, -0.55)
Lewis et al. [76] 20/20 62.3/65.1 13.90 27.90 NA 23.30 70.80 806.30 Stroop Task Time of completion -0.37 (-0.98, 0.24)
Lord et al. [47] 121/184 67/69.4 NA NA 1.94 25.5 NA 124.6 Hayling Test Test score -0.34 (-0.57, -0.11)
Marzinzik et al. [77] 11/11 70.1/72.4 9.1 28.6 1.6 13.7 70.8 300.4 Go/Nogo Task Error (Nogo) -0.12 (-0.93, 0.69)
McNamara et al. [78] 20/10 71.8/69 12.00 NA NA NA 96.00 NA Stroop Task Interference -1.50 (-2.33, -0.67)
McNamara et al. [79] 22/22 73/70.5 12.8 27.5 NA NA NA NA Stroop Task Interference -0.83 (-1.45, -0.21)
McNamara et al. [80] 28/32 66.5/56.3 13.90 26.60 3.00 NA NA 628.90 Stroop Task Interference -0.15 (-0.65, 0.35)
Miller et al. [81] 42/28 64.8/63.9 17.3 NA 2 20.1 80.4 356.5 Stroop Task Number of correct response -0.96 (-1.44, -0.48)
Mitchell and Barbosa Bouças [82] 33/33 63.6/66.9 15.2 NA NA NA 96 NA Stroop Task Accuracy (incongruent) -0.51 (-0.99, -0.03)
Murray and Rutledge, [48] 11/9 68.36/67.89 16 NA NA 17.91 NA NA Flanker Task Accuracy (incongruent) -0.72 (-1.60, 0.15)
Obeso et al. [83] 17/16 69.41/65.69 13.53 29.35 2.12 23.48 114 915.94 SST Stop-signal reaction time -1.23 (-1.96, -0.50)
O’Callaghan et al. [84] 50/27 63.8/65.6 13.4 28 2.1 NA 69.6 775.5 Hayling Test Test score -0.49 (-0.96, -0.03)
Pellicano et al. [85] 13/13 58.8/60.3 11.80 28.40 1.90 18.50 51.60 NA Stroop Task Interference 0.11 (-0.63, 0.86)
Pereira et al. [86] 20/20 64/59.1 NA 28.50 2.40 24.90 81.60 627.00 Stroop Task Interference -0.71 (-1.34, -0.08)
Pettit et al. [87] 18/19 68.8/66.2 14.06 NA 1.78 NA 78.36 NA Hayling Test Test score -1.16 (-1.84, -0.47)
Pillon et al. [88] 20/14 62.4/64.3 11.00 28.90 2.50 18.10 97.20 630.30 Stroop Task Interference -3.71 (-4.81, -2.60)
Pillon et al. [89] 10/14 64.8/64.3 11.60 29.70 1.30 18.10 16.80 NA Stroop Task Interference -0.32 (-1.12, 0.47)
Poletti et al. [49] 103/100 66/66.8 9.20 27.90 NA 16.90 10.70 NA Stroop Task Interference -0.09 (-0.37, 0.18)
Ranchet et al. [90] 19/21 66.1/69.1 13 27.4 2.1 16.4 90 741.8 Stroop Task Interference -0.53 (-1.15, 0.09)
Raskin et al. [91] 54/34 61.9/61 14.7 NA 1.99 NA NA NA Stroop Task Interference -0.53 (-0.96, -0.10)
Segura et al. [92] 43/32 60.77/64.69 12.02 29.47 1.7 13.16 74.76 692.81 Stroop Task Interference 0.01 (-0.45, 0.46)
Stavitsky et al. [93] 35/18 66.2/64.4 16.8 NA 2 25.1 105.6 604.3 Stroop Task Interference -0.79 (-1.37, -0.21)
Theilmann et al. [94] 25/26 68/65.9 16.9 28.7 2.36 25.4 86.4 970.4 Stroop Task Interference 0.04 (-0.51, 0.59)
van Spaendonck et al. [95] 51/24 53.9/52.7 NA NA NA NA 39.60 NA Stroop Task Interference -0.14 (-0.62, 0.34)
Wild et al. [96] 18/18 69.33/69.44 6.22 26.39 1.97 16.22 100.68 NA Stroop Task Interference -1.33 (-2.04, -0.62)
Wylie et al. [97] 16/16 64.8/65.4 15.7 NA 1.8 NA 99.6 NA Flanker Task Reaction time (incongruent) -0.48 (-1.16, 0.21)
Wylie et al. [98] 28/17 65.5/62.3 15.8 28.6 NA 19 NA NA Flanker Task Time cost -0.31 (-0.91, 0.28)
Zgaljardic et al. [99] 32/29 66.9/66.7 15.4 NA 1.92 NA NA NA Stroop Task Interference -0.88 (-1.40, -0.36)
Zhang et al. [100] 42/36 62.2/62.7 9.8 27.59 2 20.02 50.4 298.2 Stroop Task Time of completion (incongruent) -0.51 (-0.96, -0.06)
Planning
Altgassen et al. [101] 16/16 61.1/62.6 NA NA 1.38 13.83 57.72 NA TOL Number of moves -0.81 (-1.52, -0.11)
Broeders et al. [61] 59/40 62.5/61.4 11.60 27.90 1.70 16.00 17.50 153.90 TOL Number of trials completed -0.81 (-1.22, -0.39)
Cipresso et al. [102] 15/15 69/61.7 7.93 27 NA NA NA NA TOL Test score -0.83 (-1.56, -0.10)
Dujardin et al. [103] 12/12 65.92/59.25 10.25 NA 2.13 30.58 103 405 SSG Number of sequence produced -0.90 (-1.72, -0.09)
Engels et al. [104] 48/57 72.9/67.9 NA 26.14 NA NA NA NA BADS-KS Test score -0.66 (-1.06, -0.27)
Fales et al. [66] 21/25 66.9/68.8 16.9 28.8 2 NA 69.6 NA TOL Number of moves -0.11 (-0.68, 0.47)
Foster et al. [105] 24/30 59/60 14.9 NA NA 18.95 54 NA TOL Number of correct response -0.02 (-0.55, 0.51)
Koerts et al. [73] 43/25 63.7/62.8 5.2 27.5 2.2 24.6 61.2 561.7 BADS-ZM Test score -0.21 (-0.70, 0.28)
Koerts et al. [74] 43/25 63.6/62.8 5.30 27.50 2.20 24.60 63.60 583.70 BADS-ZM Hit-error -0.26 (-0.75, 0.23)
Lord et al. [46] 121/184 67/69.4 NA NA 1.94 25.5 NA 124.6 OTS Problems solved -0.52 (-0.76, -0.29)
McKinlay et al. [105] 30/30 65.77/66.43 14.08 28.9 2.3 27.13 87.6 NA TOL Accuracy -0.65 (-1.18, -0.13)
McNamara et al. [77] 20/10 71.8/69 12.00 NA NA NA 96.00 NA TOL Time/move -0.20 (-0.94, 0.55)
Miah et al. [106] 23/21 62.6/60.3 NA NA NA 2 17.09 64.18 397.69 OTS Problems solved -0.18 (-0.73, 0.37)
Monetta et al. [107] 11/11 67.1/71.2 16.60 NA 2.50 NA 109.20 NA TOL Number of moves -1.30 (-2.21, -0.38)
Morris et al. [108] 12/18 64.58/63.72 9.5 NA 2.17 NA NA NA TOL Number of moves 0.16 (-0.55, 0.87)
Muslimovic et al. [109] 95/44 64.9/64.1 11.50 27.90 1.90 18.20 37.20 291.30 TOL Number of correct response -0.52 (-0.88, -0.16)
Parrao et al. [110] 44/17 63.5/63.1 11.2 NA 2.3 16.1 69.6 NA TOL Number of moves -0.51 (-1.07, 0.06)
Pell et al. [111] 15/16 70.9/70.4 15.40 NA NA 29.90 127.20 NA TOL Number of correct response -0.93 (-1.67, -0.19)
Perfetti et al. [112] 25/24 69.8/72.9 8.70 27.00 2.20 19.90 NA 1210 TOL Number of correct response -0.83 (-1.41, -0.25)
Raskin et al. [90] 53/34 61.9/61 14.7 NA 1.99 NA NA NA DKEFS-T Total achievement -0.47 (-0.90, -0.04)
Rosen et al. [113] 20/23 67.45/68.26 13.45 28.75 2.5 34.18 100.8 999.42 BADS-KS Test score -0.79 (-1.40, -0.18)
Schomaker et al. [114] 21/21 61.8/59.7 NA 28.81 NA 21.80 NA 851.10 TOH Test score 0.06 (-0.53, 0.66)
Reasoning
Basic et al. [115] 58/58 66.09/66.85 NA 26.10 NA NA 77.76 NA RPM Test score -0.12 (-0.48, 0.24)
Benke et al. [116] 22/18 58/60.9 9.9 NA 2.29 16.2 117.6 NA RPM Test score -0.14 (-0.75, 0.47)
Bodden et al. [117] 21/21 63.7/58.5 14.6 29 2.5 NA 61.2 432.1 LPS (subtest 4) Test score -0.43 (-1.03, 0.17)
Brand et al. [118] 20/20 66.85/64 9.1 28.2 3 NA 106.05 NA LPS (subtest 4) Test score -0.50 (-1.12, 0.12)
Costa et al. [119] 39/46 62.56/63.04 10.82 28.22 NA 9.09 81.48 NA RPM Test score -0.46 (-0.89, -0.03)
Costa et al. [120] 33/20 63.4/66 12.6 29.4 NA 19.7 82.8 607 RPM Test score -0.30 (-0.85, 0.25)
Crescentini et al. [41] 19/14 66.7/65.6 8.1 29 2.05 26.2 76.8 540.3 RPM Test score -0.49 (-1.17, 0.20)
Euteneuer et al. [121] 21/23 67.6/64.4 11.1 29 2.5 17.7 85.7 487.69 LPS (subtest 4) Test score -0.38 (-0.97, 0.20)
Mioni et al. [122] 21/25 68.95/71.12 7.76 27.28 NA 10.92 68.4 NA RPM Test score -0.63 (-1.22, -0.04)
Natsopoulosl et al. [123] 27/27 60.7/60.56 NA 29.41 NA NA 75.96 NA Modus Ponens Test score -0.25 (-0.77, 0.28)
Perfetti et al. [112] 25/24 69.8/72.9 8.70 27.00 2.20 19.90 NA 1210 Temporal Judgment Test score -0.65 (-1.22, -0.08)
Poletti et al. [48] 103/100 66/66.8 9.20 27.90 NA 16.90 10.70 NA RPM Test score -0.16 (-0.43, 0.12)
Rosen et al. [113] 20/23 67.45/68.26 13.45 28.75 2.5 34.18 100.8 999.42 LPS (subtest 4) Test score -0.32 (-0.91, 0.27)
Shifting
Abe et al. [50] 32/20 65.9/65.5 11.7 28.3 2.5 18.1 49.2 507 TMT Time (B-A) -0.61 (-1.18, -0.05)
Akamatsu et al. [124] 30/20 60.4/58.6 13.50 29.70 2.30 29.20 110.40 NA TMT Time (B) -0.90 (-1.49, -0.32)
Aksan et al. [125] 39/77 74.58/75.4 15.00 NA NA NA NA NA TMT Time (B) -1.04 (-1.45, -0.64)
Alonso-Recio et al. [126] 50/49 65.14/64.86 NA 29.00 NA NA 77.40 NA TMT Time (B-A) 1.58 (1.13, 2.03)
Baggio et al. [52] 39/23 63.5/61 11.4 28.7 1.8 16.5 67.2 560.3 TMT Time (B) -0.33 (-0.84, 0.18)
Baggio et al. [53] 43/36 64/63.4 10.8 29.35 1.58 14.1 73.2 646.7 TMT Time (B-A) 0.15 (-0.29, 0.59)
Bodden et al. [117] 21/21 63.7/58.5 14.6 29 2.5 NA 61.2 432.1 TMT Time (B-A) -0.15 (-0.74, 0.44)
Bogdanova and Cronin-Golomb, [127] 22/22 62.25/61.3 15.7 29.3 2 NA 98.4 486.35 TMT Time (B) -0.42 (-1.01, 0.17)
Bohnen et al. [59] 13/14 70.8/69.7 NA 28.1 NA 26.4 70.8 NA TMT Time (B-A) -0.43 (-1.18, 0.31)
Bokura et al. [128] 13/14 71/71 9.80 NA NA 29.23 86.77 NA WCST Number of category -1.40 (-2.22, -0.57)
Brand et al. [118] 20/20 66.85/64 9.1 28.2 3 NA 106.05 NA WCST Perseverative error 0.67 (0.04, 1.29)
Broeders et al. [60] 59/40 62.5/61.4 11.60 27.90 1.70 16.00 17.50 153.90 WCST Perseverative error -0.53 (-0.93, -0.12)
Broussolle et al. [129] 27/30 55.88/52.9 NA NA 2.61 14.72 83.91 471.30 WCST Perseverative error -0.28 (-1.16, 0.60)
Brown and Marsden [130] 16/16 59.2/56.1 11.30 NA NA NA 134.40 NA WCST Perseverative error -1.04 (-1.77, -0.32)
Camicioli et al. [131] 29/16 60.5/57.9 14.3 28.6 NA 15.3 NA NA TMT Time (B) -0.81 (-1.43, -0.19)
Cohen et al. [61] 13/16 65.3/66.6 4.8 NA 2.1 29.9 78 714 TMT Time (B-A) -0.17 (-0.89, 0.55)
Colman et al. [40] 27/25 61.39/62.93 13.21 28.11 1.79 15.68 72.48 786.94 TMT/OMO Time (B), error (OMO) -0.66 (-1.21, -0.11)
Cooper et al. [132] 60/37 59.8/59.6 9.93 NA NA NA 15.75 NA WCST Perseverative error -2.97 (-3.55, -2.39)
Costa et al. [119] 39/46 62.56/63.04 10.82 28.22 NA 9.09 81.48 NA WCST Perseverative error -0.62 (-1.06, -0.19)
Costa et al. [120] 33/20 63.4/66 12.6 29.4 NA 19.7 82.8 607 WCST Perseverative error -1.08 (-1.67, -0.49)
Crescentini et al. [41] 19/14 66.7/65.6 8.1 29 2.05 26.2 76.8 540.3 TMT Time (B) -0.63 (-1.32, 0.06)
Crescentini et al. [42] 16/14 63.6/61.6 9.7 28.8 1.94 22.9 72 386.4 TMT Time (B) -0.56 (-1.27, 0.16)
Dalrymple-Alford et al. [133] 8/8 65.6/62.4 10.1 NA 2.13 NA 52.8 NA WCST Percentage of perseverative error -0.13 (-1.06, 0.79)
Delazer et al. [134] 20/20 68.5/71.3 11.9 27.8 1.8 17.6 62.95 431 TMT Time (B) 0.30 (-0.31, 0.91)
Drag et al., 2009 24/24 69.04/68.67 16.58 NA 1.86 14.4 64.62 NA WCST Number of category -0.57 (-1.14, 0)
Dujardin et al. [135] 12/12 65.92/59.25 10.25 NA 2.13 30.58 103 405 WCST Perseverative error -0.61 (-1.41, 0.18)
Edelstyn et al. [63] 17/17 65.4/64.5 NA 28 NA NA 92.4 448.59 Brixton Test Test score -0.83 (-1.52, -0.15)
Ell, 2013 [64] 36/35 71.1/65.8 16.7 NA 2 NA 42 NA DKEFS-CWI (shifting) Time of completion -0.84 (-1.32, -0.36)
Elwan et al. [44] 5/12 61.84/56.35 NA 26.57 2.67 47.90 28.60 NA TMT Time (B) -0.04 (-0.62, 0.55)
Engels et al. [103] 48/57 72.9/67.9 NA 26.14 NA NA NA NA BADS-RS Number of correct response -0.49 (-0.88, -0.10)
Euteneuer et al. [121] 21/23 67.6/64.4 11.1 29 2.5 17.7 85.7 487.69 WCST Perseverative error -0.58 (-1.18, 0.01)
Fales et al. [65] 21/25 66.9/68.8 16.9 28.8 2 NA 69.6 NA WCST Perseverative error -0.27 (-0.84, 0.31)
Fama et al. [136] 20/38 63.1/65.3 16 27.4 NA NA 81.6 NA WCST Perseverative error -1.04 (-1.61, -0.47)
Galtier et al. [67] 43/20 59.19/60.85 7.88 27.58 2.28 28.46 99.60 NA WCST Number of category -0.87 (-1.42, -0.33)
Gawrys et al. [68] 30/18 56.03/57.11 13.51 28.93 2 NA NA 851.58 TMT Time (B-A) -1.06 (-1.67, -0.45)
Goebel et al. [137] 22/27 64.8/60.8 14 NA 1.27 NA 94.8 NA TMT Time (B) -1.32 (-1.93, -0.70)
Hartikainen et al. [138] 22/26 66.5/65.5 NA NA 1.9 NA NA NA TMT Time (B-A) -0.34 (-0.90, 0.22)
Katzen et al. [139] 222/108 66.41/67.63 14.06 25.18 2.30 NA 79.08 NA WCST Number of category -0.65 (-0.88, -0.41)
Kobayakawa et al. [140] 34/22 69.9/67.6 13.20 28.00 1.52 NA 76.80 391.00 WCST Number of category 0.60 (0.06, 1.14)
Koerts et al. [72] 43/25 63.7/62.8 5.2 27.5 2.2 24.6 61.2 561.7 TMT Time (B) -0.36 (-0.85, 0.13)
Koerts et al. [73] 88/65 62.5/61.9 NA 27.5 2 21.4 67.2 550.3 TMT Time (B) -0.28 (-0.62, 0.06)
Koerts et al. [74] 43/25 63.6/62.8 5.30 27.50 2.20 24.60 63.60 583.70 TMT Time (B) -0.36 (-0.85, 0.13)
Lewis et al. [75] 20/20 62.3/65.1 13.90 27.90   23.30 70.80 806.30 TMT Time (B) -0.61 (-1.23, 0.02)
Lin et al. [141] 31/50 65.3/67.7 9.60 26.60 1.84 14.99 NA NA TMT Time (B-A) -0.06 (-0.51, 0.38)
Lord et al. [46] 121/184 67/69.4 NA NA 1.94 25.5 NA 124.6 Brixton Test Test score -0.34 (-0.54, -0.14)
Miah et al. [106] 23/21 62.6/60.3 NA NA 2 17.09 64.18 397.69 IED Stages completed -0.42 (-0.96, 0.13)
Miller et al. [80] 42/28 64.8/63.9 17.3 NA 2 20.1 80.4 356.5 TMT Time (B) -0.67 (-1.15, -0.18)
Mioni et al. [122] 21/25 68.95/71.12 7.76 27.28 NA 10.92 68.4 NA TMT Time (B-A) 0.40 (-0.19, 0.99)
Miura et al. [142] 32/25 67.8/66.9 12 28.5 2.72 NA 67.2 318.9 WCST Perseverative error -0.52 (-1.04, 0)
Monetta et al. [107] 11/11 67.1/71.2 16.60 NA 2.50 NA 109.20 NA TMT Time (B) -0.92 (-1.77, -0.07)
Müller et al. [159] 20/20 55.3/55.7 10.7 NA 2.3 49.3 42.5 NA WCST Perseverative error -0.63 (-1.26, 0)
Muslimovic et al. [109] 95/44 64.9/64.1 11.50 27.90 1.90 18.20 37.20 291.30 WCST Perseverative error -0.77 (-1.14, -0.4)
Parrao et al. [110] 44/17 63.5/63.1 11.2 NA 2.3 16.1 69.6 NA WCST Number of category -1.15 (-1.74, -0.56)
Partiot et al. [144] 27/24 64.7/59.7 9.3 NA NA NA 93.6 NA WCST Number of category -1.41 (-2.02, -0.80)
Pellicano et al. [84] 13/13 58.8/60.3 11.80 28.40 1.90 18.50 51.60 NA WCST Perseverative error -1.08 (-1.89, -0.27)
Perfetti et al. [112] 25/24 69.8/72.9 8.70 27.00 2.20 19.90 NA 1210 TMT Time (B-A) -0.37 (-0.93, 0.19)
Pettit et al. [86] 18/19 68.8/66.2 14.06 NA 1.78 NA 78.36 NA Brixton Test Test score -1.47 (-2.19, -0.76)
Pillon et al. [87] 20/14 62.4/64.3 11.00 28.90 2.50 18.10 97.20 630.30 WCST Perseverative error -2.87 (-3.82, -1.91)
Pillon et al. [88] 10/14 64.8/64.3 11.60 29.70 1.30 18.10 16.80 NA WCST Perseverative error -3.41 (-4.66, -2.17)
Poletti et al. [48] 103/100 66/66.8 9.20 27.90 NA 16.90 10.70 NA TMT Time (B-A) 0.05 (-0.22, 0.33)
Ranchet et al. [89] 19/21 66.1/69.1 13 27.4 2.1 16.4 90 741.8 TMT Time (B-A) -0.61 (-1.23, 0.02)
Raskin et al. [90] 48/34 61.9/61 14.7 NA 1.99 NA NA NA TMT Test score -1.48 (-1.96, -1.00)
Rosen et al. [113] 20/23 67.45/68.26 13.45 28.75 2.5 34.18 100.8 999.42 WCST Perseverative error -0.77 (-1.38, -0.16)
Sagar et al. [145] 37/32 60.2/58.5 9.80 NA NA NA 12.00 NA WCST Perseverative error 0.02 (-0.45, 0.49)
Schomaker et al. [114] 21/21 61.8/59.7 NA 28.81 NA 21.80 NA 851.10 TMT Time (B/A) -0.48 (-1.09, 0.12)
Segura et al. [91] 43/32 60.77/64.69 12.02 29.47 1.7 13.16 74.76 692.81 TMT Time (B) 0.03 (-0.43, 0.48)
Stavitsky et al. [92] 35/18 66.2/64.4 16.8 NA 2 25.1 105.6 604.3 TMT Time (B) -0.96 (-1.55, -0.37)
Stolwyk et al. [146] 18/18 67.62/67.13 13.89 27.89 NA 11.67 80.04 470.84 TMT Time (B) -0.76 (-1.44, -0.08)
Tamura et al. [147] 24/24 60.9/61.7 12.7 28.5 2.21 NA 79.2 NA TMT Time (B-A) -0.32 (-0.88, 0.24)
Theilmann et al. [93] 25/26 68/65.9 16.9 28.7 2.36 25.4 86.4 970.4 TMT Time (B-A) -0.25 (-0.80, 0.30)
Uc et al. [148] 79/151 65.9/65.3 14.80 28.30 2.10 24.10 67.20 577 TMT Time (B-A) -0.69 (-0.97, -0.41)
Uc et al. [149] 76/152 65.9/65.3 14.80 28.40 2.20 23.70 68.40 588.00 TMT Time (B-A) -0.69 (-0.97, -0.41)
Vandenbossche et al. [150] 14/14 NA/NA 20.21 28.79 2.43 35.64 98.52 NA Brixton Test Error -0.77 (-1.52, -0.02)
Werheid et al. [151] 14/16 62.45/62.4 9.25 NA NA 44.25 NA NA TMT Time (B-A) -0.92 (-1.65, -0.18)
Wild et al. [95] 18/18 69.33/69.44 6.22 26.39 1.97 16.22 100.68 NA WCST Perseverative error -0.77 (-1.43, -0.11)
Yogev-Seligmann et al. [152] 20/20 70.3/70.9 15.9 NA 2.3 20.7 88.8 NA TMT Time (B-A) -0.34 (-0.96, 0.27)
Yogev-Seligmann et al. [33] 18/15 68.7/75.4 15.9 NA 2.5 23.3 NA NA TMT Time (B-A) -0.05 (-0.72, 0.62)
Yu et al. [153] 55/30 62.47/64.2 12.33 28.29 1.46 16.65 43.44 NA WCST Perseverative error 0.16 (-0.28, 0.60)
Yu et al. [154] 39/40 62.7/61.9 11.4 27.9 1.6 18.9 51.6 562.9 WCST Perseverative error -0.06 (-0.49, 0.38)
Zamarian et al. [155]a 15/28 66.1/63.1 10.7 28.2 NA NA 63.6 NA TMT Time (B) -0.48 (-1.1, 0.15)
Zgaljardic et al. [98] 32/29 66.9/66.7 15.4 NA 1.92 NA NA NA OMO Test score 0.51 (0.01, 1.02)
Zhang et al. [99] 42/36 62.2/62.7 9.8 27.59 2 20.02 50.4 298.2 TMT Time (B) -0.64 (-1.09, -0.19)
Working Memory
Abe et al. [50] 32/20 65.9/65.5 11.7 28.3 2.5 18.1 49.2 507 Digit Span Span length -0.08 (-0.63, 0.47)
Agosta et al. [39] 41/34 64/63 11 27.7 NA 27.1 85.2 NA ACE-R Test score -0.41 (-0.86, 0.05)
Alonso-Recio et al. [126] 50/49 65.14/64.86 NA 29.00 NA NA 77.40 NA Digit Span Span length -0.53 (-0.93, -0.13)
Altgassen et al. [100] 16/16 61.1/62.6 NA NA 1.38 13.83 57.72 NA Digit Span Span length -0.56 (-1.26, 0.14)
Baggio et al. [52] 39/23 63.5/61 11.4 28.7 1.8 16.5 67.2 560.3 Digit Span Span length -0.22 (-0.73, 0.29)
Barnes and Boubert, [54] 20/20 63.73/68.54 NA 27.2 2.9 NA 116.64 465 2-back Task %hits-%false alarm -1.32 (-2.00, -0.64)
Basic et al. [115] 58/58 66.09/66.85 NA 26.10 NA NA 77.76 NA Digit Span Span length -0.78 (-1.15, -0.40)
Beato et al. [156] 18/21 52.7/ 13.8 NA 3 37.6 139.2 902 n-back Task Number of correct response -0.67 (-1.31, -0.03)
Benito-León et al. [157] 46/138 78.1/76.5 NA NA 2.4 NA 22.8 NA Immediate Free Recall Items recalled -0.17 (-0.50, 0.16)
Benke et al. [116] 22/18 58/60.9 9.9 NA 2.29 16.2 117.6 NA MGT Test score 0.07 (-0.54, 0.68)
Bezdicek et al. [57] 46/41 57.7/60.78 14.84 NA 1.95 10.3 124.8 1029 Digit Span Span length -0.26 (-0.68, 0.15)
Bodden et al. [117] 21/21 63.7/58.5 14.6 29 2.5 NA 61.2 432.1 Memo Test Test score -0.42 (-1.02, 0.18)
Bogdanova and Cronin-Golomb [127] 22/22 62.25/61.3 15.7 29.3 2 NA 98.4 486.35 Digit Span Span length -0.50 (-1.09, 0.09)
Bohlhalter et al. [58] 12/12 59.1/46.6 NA NA NA 16.5 92.4 766 Working Memory Span Span length -0.92 (-1.73, -0.10)
Bohnen et al. [59] 13/14 70.8/69.7 NA 28.1 NA 26.4 70.8 NA CVLT Test score -0.21 (-0.95, 0.52)
Bott et al. [20] 21/21 63.7/66.4 16.5 NA NA 27.1 NA NA EXAMINER Test score -0.65 (-1.26, -0.04)
Brand et al. [118] 20/20 66.85/64 9.1 28.2 3 NA 106.05 NA Immediate Recall Items recalled -2.02 (-2.77, -1.27)
Breitenstein et al. [158] 6/16 70.45/68.6 NA 29.2 NA 22.5 37.65 281 Digit Span Span length -0.34 (-1.17, 0.49)
Broeders et al. [60] 59/40 62.5/61.4 11.60 27.90 1.70 16.00 17.50 153.90 RAVLT Items recalled -0.57 (-0.98, -0.16)
Bublak et al. [159] 14/14 55.1/55.2 NA 28.9 2.3 NA 47.3 NA Working Memory Span Span length -0.58 (-1.33, 0.16)
Camicioli et al. [131] 29/16 60.5/57.9 14.3 28.6 NA 15.3 NA NA DOT Span length -1.15 (-1.79, -0.50)
Cipresso et al. [101] 15/15 69/61.7 7.93 27 NA NA NA NA Cors Block Test Span length -0.69 (-1.41, 0.03)
Cohen et al. [61] 13/16 65.3/66.6 4.8 NA 2.1 29.9 78 714 Digit Span Span length 0.07 (-0.64, 0.78)
Colman et al. [40] 27/25 61.39/62.93 13.21 28.11 1.79 15.68 72.48 786.94 Digit Span Span length -0.36 (-0.90, 0.18)
Cooper et al. [132] 60/37 59.8/59.6 9.93 NA NA NA 15.75 NA DOT Span length -2.43 (-2.97, -1.90)
Costa et al. [119] 39/46 62.56/63.04 10.82 28.22 NA 9.09 81.48 NA Immediate Visual Memory Items recalled -0.42 (-0.85, 0.01)
Costa et al. [120] 33/20 63.4/66 12.6 29.4 NA 19.7 82.8 607 Word List Recall Items recalled 0 (-0.55, 0.55)
Crawford et al. [160] 25/18 63/75 11.76 28.8 2.12 NA NA NA Digit Span Span length 0.17 (-0.43, 0.77)
Crescentini et al. [41] 19/14 66.7/65.6 8.1 29 2.05 26.2 76.8 540.3 Digit Span Span length -0.54 (-1.22, 0.15)
Crescentini et al. [42] 16/14 63.6/61.6 9.7 28.8 1.94 22.9 72 386.4 Digits Span Span length -0.93 (-1.69, -0.17)
Dalrymple-Alford et al. [133] 8/8 65.6/62.4 10.1 NA 2.13 NA 52.8 NA Digit Span Span length 0 (-0.93, 0.93)
Delazer et al. [134] 20/20 68.5/71.3 11.9 27.8 1.8 17.6 62.95 431 Digit Span Span length -0.66 (-1.28, -0.03)
Drag et al. [135] 24/24 69.04/68.67 16.58 NA 1.86 14.4 64.62 NA Digit Span Span length -0.42 (-0.99, 0.14)
Dujardin et al. [102] 12/12 65.92/59.25 10.25 NA 2.13 30.58 103 405 Digit Span Span length -0.95 (-1.77, -0.13)
Dujardin et al. [62] 18/18 60.17/59.5 11.56 NA NA 17.58 11.33 NA LNS Cost index -0.60 (-1.25, 0.05)
Ell, [64] 36/35 71.1/65.8 16.7 NA 2 NA 42 NA Digit Span Span length -0.93 (-1.41, -0.44)
Engels et al. [103] 48/57 72.9/67.9 NA 26.14 NA NA NA NA Digit Span Span length -0.53 (-0.91, -0.14)
Euteneuer et al. [121] 21/23 67.6/64.4 11.1 29 2.5 17.7 85.7 487.69 Word List Recall Items recalled -0.56 (-1.15, 0.03)
Fama et al. [136] 20/38 63.1/65.3 16 27.4 NA NA 81.6 NA Immediate Recall Items recalled -0.57 (-1.12, -0.03)
Foster et al. [104] 24/30 59/60 14.9 NA NA NA NA NA Letter Maintenance Accuracy -0.43 (-0.97, 0.10)
Fournet et al. [161] 15/15 66.1/66.6 9.7 NA 1.93 21.8 96 NA Verbal Span Span length -1.01 (-1.75, -0.26)
Gawrys et al. [68] 30/18 56.03/57.11 13.51 28.93 2 NA NA 851.58 Digit Span Span length -0.87 (-1.47, -0.27)
Gilbert et al. [162] 14/14 66.29/65.79 12.21 29.14 2.14 NA 87.48 NA Alphabetical Recall Items recalled -0.20 (-0.93, 0.52)
Goebel et al. [163] 14/22 66.17/62.06 NA 29.29 NA NA 106.8 NA Digit Span Span length -0.68 (-1.36, -0.01)
Goebel et al. [137] 22/27 64.8/60.8 14 NA 1.27 NA 94.8 NA Digit Span Span length -0.68 (-1.25, -0.11)
Hartikainen et al. [138] 22/26 66.5/65.5 NA NA 1.9 NA NA NA DSST Test score -0.34 (-0.90, 0.22)
Hsieh et al. [71] 26/27 63.3/63.48 9.07 NA NA NA 40.80 NA Digit Span Span length -0.47 (-1.01, 0.07)
Ibarretxe-Bilbao et al. [45] 24/24 56.13/57.58 10.96 29.63 1.73 14.67 36.72 299.58 RAVLT Items recalled -0.05 (-0.61, 0.51)
Koerts et al. [72] 43/25 63.7/62.8 5.2 27.5 2.2 24.6 61.2 561.7 Digit Span Span length -0.62 (-1.12, -0.12)
Koerts et al. [73] 88/65 62.5/61.9 NA 27.5 2 21.4 67.2 550.3 Immediate Recall Items recalled -0.38 (-0.72, -0.04)
Koerts et al. [74] 43/25 63.6/62.8 5.30 27.50 2.20 24.60 63.60 583.70 Digit Span Span length -0.63 (-1.13, -0.13)
Koivisto et al. [164] 8/12 72.9/69.3 7 22.9 2.5 NA 115.2 NA Object Recall Items recalled -1.56 (-2.56, -0.57)
Lee et al. [165] 19/12 64.11/67.61 9 NA NA NA 94.32 NA PLM Accuracy -0.88 (-1.61, -0.14)
Lee et al. [166] 21/28 66.71/68.57 14.29 28.50 1.98 NA 80.40 NA Working Memory Span Span length -0.80 (-1.38, -0.22)
Lord et al. [46] 121/184 67/69.4 NA NA 1.94 25.5 NA 124.6 Digit Span Span length -0.38 (-0.58, -0.18)
Martin et al. [167] 16/22 65.4/68 15.10 29.10 NA 7.90 NA NA SDMT Test score -2.22 (-3.03, -1.42)
Miah et al. [106] 23/21 62.6/60.3 NA NA 2 17.09 64.18 397.69 Spatial Span Span length -0.32 (-0.87, 0.23)
Miller et al. [80] 42/28 64.8/63.9 17.3 NA 2 20.1 80.4 356.5 Digit Span Span length -0.59 (-1.07, -0.11)
Mioni et al. [122] 21/25 68.95/71.12 7.76 27.28 NA 10.92 68.4 NA Word List Recall Items recalled -0.25 (-0.84, 0.34)
Mitchell and Barbosa Bouças, [81] 33/33 63.6/66.9 15.2 NA NA NA 96 NA n-back Task Accuracy -0.47 (-0.95, 0.01)
Miura et al. [142] 32/25 67.8/66.9 12 28.5 2.72 NA 67.2 318.9 Digit Span Span length -0.49 (-1.01, 0.03)
Monetta et al. [107] 11/11 67.1/71.2 16.60 NA 2.50 NA 109.20 NA Working Memory Span Span length 0.47 (-0.34, 1.29)
Müller et al. [159] 20/20 55.3/55.7 10.7 NA 2.3 49.3 42.5 NA Digit Span Span length -1.07 (-1.73, -0.42)
Murray and Rutledge, [47] 11/9 68.36/67.89 16 NA NA 17.91 NA NA Sentence Reading Span Span length -1.34 (-2.28, -0.40)
Owen et al. [168] 7/7 53.9/56.47 NA NA 2.56 NA NA NA Working Memory Span Span length 0.06 (-0.94, 1.06)
Parrao et al. [110] 44/17 63.5/63.1 11.2 NA 2.3 16.1 69.6 NA Digit Span Span length -0.78 (-1.35, -0.21)
Peavy et al. [169] 16/30 64.2/63 15.40 NA NA NA 140.40 NA Digit Span Span length -0.39 (-0.99, 0.21)
Peigneux et al. [170] 17/17 63.2/63 11.30 NA 2.53 30.42 105.88 NA Digit Span Span length 0.37 (-0.30, 1.03)
Pell et al. [111] 15/16 70.9/70.4 15.40 NA NA 29.90 127.20 NA Digit Span Span length -0.64 (-1.35, 0.06)
Pellicano et al. [84] 13/13 58.8/60.3 11.80 28.40 1.90 18.50 51.60 NA RAVLT Items recalled -0.98 (-1.77, -0.19)
Pereira et al. [85] 20/20 64/59.1 NA 28.50 2.40 24.90 81.60 627.00 Digit Span Span length -1.09 (-1.75, -0.44)
Perfetti et al. [112] 25/24 69.8/72.9 8.70 27.00 2.20 19.90 NA 1210 Jigsaw Puzzle Test Test score -1.07 (-1.67, -0.48)
Poletti et al. [48] 103/100 66/66.8 9.20 27.90 NA 16.90 10.70 NA Digit Span Span length -0.34 (-0.62, -0.07)
Poliakoff and Smith-Spark [171] 24/24 62.6/62.8 12.30 29.40 2.21 21.00 88.56 NA Digit Span Span length -0.17 (-0.73, 0.39)
Pollux [172] 18/18 58.27/61.05 NA 28.38 2.5 NA 103.32 NA Digit Span Span length -0.57 (-1.22, 0.09)
Ranchet et al. [89] 19/21 66.1/69.1 13 27.4 2.1 16.4 90 741.8 Benton Test Test score -0.13 (-0.74, 0.48)
Raskin et al. [90] 54/34 61.9/61 14.7 NA 1.99 NA NA NA Digit Span Span length 0.59 (0.15, 1.03)
Rodriguez-Ferreiro et al. [49] 50/42 72.92/74.08 7.04 26.94 1.96 NA 114.00 NA Immediate Recall Items recalled -0.45 (-0.86, -0.04)
Segura et al. [91] 43/32 60.77/64.69 12.02 29.47 1.7 13.16 74.76 692.81 RAVLT Items recalled -0.26 (-0.71, 0.19)
Siepel et al. [173] 339/158 61.1/59.4 15.5 NA NA 21 NA NA HVLT-R Items recalled -0.35 (-0.54, -0.16)
Stavitsky et al. [92] 35/18 66.2/64.4 16.8 NA 2 25.1 105.6 604.3 Digit Span Span length -0.28 (-0.84, 0.28)
Stebbins et al. [174] 16/16 68.56/69.19 14.8 28.81 2.75 NA NA NA Free Recall Items recalled -0.80 (-1.51, -0.10)
Stolwyk et al. [146] 18/18 67.62/67.13 13.89 27.89 NA 11.67 80.04 470.84 Digit Span Span length -0.12 (-0.77, 0.53)
Tamura et al. [147] 24/24 60.9/61.7 12.7 28.5 2.21 NA 79.2 NA Digit Span Span length -0.63 (-1.20, -0.06)
Theilmann et al. [93] 25/26 68/65.9 16.9 28.7 2.36 25.4 86.4 970.4 Digit Span Span length -0.52 (-1.07, 0.03)
van Spaendonck et al. [94] 51/24 53.9/52.7 NA NA NA NA 39.60 NA RAVLT Items recalled -0.37 (-0.86, 0.11)
Werheid et al. [151] 14/16 62.45/62.4 9.25 NA NA 44.25 NA NA Digit Span Span length -0.53 (-1.25, 0.18)
Xu et al. [175] 20/20 65.9/68.9 12.40 27.60 1.40 26.60 72.00 622.25 DSST Reaction time for correct response -0.48 (-1.09, 0.14)
Zamarian et al. [155] 15/28 66.1/63.1 10.7 28.2 NA NA 63.6 NA Digit Span Span length -0.72 (-1.36, -0.09)
Zgaljardic et al.[98] 32/29 66.9/66.7 15.4 NA 1.92 NA NA NA Digit Span Span length -1.10 (-1.64, -0.57)
Zhang et al. [99] 42/36 62.2/62.7 9.8 27.59 2 20.02 50.4 298.2 Digit Span Span length -0.25 (-0.70, 0.19)

Table 1: Summary of the included studies.

neuropsychiatry-Flowchart

Figure 1: Flowchart of literature search.

▪ Pooled effect size

The pooled effect sizes, heterogeneity and publication bias results are summarized in Table 2. A summary forest plot of all the relevant EFs is presented in Figure 2. The pooled effect size suggested a fairly moderate deficit in overall EF among PD patients. Small to moderate effect sizes were observed for attention (Hedges’ g = -0.48, 95% CI: -0.62 to -0.35), inhibition (Hedges’ g = -0.48, 95% CI: -0.59 to -0.36), planning (Hedges’ g = -0.49, 95% CI: -0.62 to -0.36), reasoning (Hedges’ g = -0.31, 95% CI: -0.45 to -0.18), shifting (Hedges’ g = -0.55, 95% CI: -0.68 to -0.42) and working memory (Hedges’ g = -0.49, 95% CI: -0.53 to -0.44).

EF domain N (PD/CON) k Hedges’ g (95% CI) Heterogeneity test Publication bias test
Q I2 FSN Egger’s test t-value ESadj (95% CI)
Attention 556/521 11 -0.48 (-0.62, -0.35)*** 12.56 12.44 134 0.56 -0.48 (-0.62, -0.35)
Inhibition 1837/1572 56 -0.48 (-0.59, -0.36)*** 135.85*** 59.51 2287 2.80** -0.33 (-0.46, -0.20)
Planning 771/698 22 -0.49 (-0.62, -0.36)*** 27.97 24.92 411 0.097 -0.47 (-0.61, -0.34)
Reasoning 429/419 13 -0.31 (-0.45, -0.18)*** 6.44 0 66 2.99** -0.20 (-0.32, -0.08)
Shifting 2827/2648 82 -0.55 (-0.68, -0.42)*** 414.86*** 80.48 6662 2.02* -0.31 (-0.45, -0.17)
Working memory 2901/2650 91 -0.53 (-0.62, -0.44)*** 215.28*** 58.19 7527 2.64** -0.38 (-0.48, -0.28)
Overall 4639/4219 275 -0.49 (-0.53, -0.44)*** 820.94*** 66.50 67088 3.93*** -0.36 (-0.42, -0.30)

Table 2: Summary of the meta-analytic results.

neuropsychiatry-Meta-analysis

Figure 2: Meta-analysis results for different executive functions. Positive effect size indicates better behavioral performance of Parkinson’s patients (PD).

▪ Heterogeneity

Heterogeneity was significant among the studies of overall EF (Cochran’s Q = 820.94, p< 0.001). After grouping by domain, however, significant heterogeneity was not observed among studies of attention (Cochran’s Q = 12.56, p= 0.32), planning (Cochran’s Q = 27.97, p = 0.14) and reasoning (Cochran’s Q = 6.44, p= 0.89), but remained significant among studies of inhibition (Cochran’s Q = 135.85, p< 0.001), shifting (Cochran’s Q = 414.86, p< 0.001) and working memory (Cochran’s Q = 215.28, p< 0.001). Heterogeneity was moderate for inhibition (I2 = 59.51%), working memory (I2 = 58.19%) and overall EF (I2 = 66.5%), but high for shifting (I2 = 80.48%).

▪ Publication bias

The fail-safe N exceeded the thresholds for all EF domains except reasoning, indicating that the meta-analytic results were generally robust to publication bias. Funnel plot asymmetry appeared to be present for all EF domains (Figure 3). However, Egger’s asymmetry test yielded significant results only for inhibition (t = 2.80, p < 0.01), reasoning (t = 2.99, p < 0.01), shifting (t = 2.02, p < 0.05), working memory (t = 2.64, p < 0.01) and overall EF (t = 3.93, p < 0.001). After a trim-and-fill procedure, the adjusted effect sizes generally decreased, and no significant measure became non-significant. On average, the mean effect size (Hedges’ g) decreased (i.e., became less negative) in the trimmed analyses by a value of 0.13 (Table 2).

neuropsychiatry-Funnel-plots

Figure 3: Funnel plots of executive functions.

▪ Meta-regression

The results of meta-regression analyses showed that age (β = -0.0046, p = 0.44), years of education (β = -0.0036, p = 0.83), Hoehn and Yahr stage (β = -0.11, p = 0.32), disease duration (β = -0.0025, p = 0.15), UPDRS motor score (β = -0.0046, p = 0.44), MMSE (β = 0.065, p = 0.21) and LEDD (β = 0.000009, p = 0.54) did not significantly predict the study outcomes.

Discussion

▪ Summary of evidence

This meta-analysis quantitatively compared the EFs of PD patients and healthy controls. A fairly moderate impairment of overall EF was observed among PD patients, consistent with the results of a previous meta-analysis [7]. The smaller effect size observed in this meta-analysis might have resulted from the inclusion of a greater number of studies. Small to moderate effect sizes were observed in different EF domains (Hedges’ g = -0.31 to -0.55), among which reasoning skill was the least affected. This finding was compatible with the results of an earlier study in which abstract thinking was demonstrated to be less affected than attentional control in PD patients [32]. Other EF domains were similarly affected to a similar extent. Furthermore, the EF deficits experienced by PD patients were not found to be influenced by age, years of education, disease severity, motor deficits, disease duration, medication dose or global cognition.

▪ Heterogeneity

Moderate heterogeneity was observed for overall EF. After grouping by EF domain, however, high heterogeneity was observed for shifting, whereas moderate heterogeneity was observed for inhibition and working memory. Heterogeneity may have resulted from variability among the assessment tools and performance indexes used by the included studies. In addition, the constructs of inhibition, shifting and working memory could possibly be divided into heterogeneous sub-constructs. For instance, PD patients might exhibit impairment in only one type of working memory. Thus, heterogeneity may arise if the different types of inhibition, shifting and working memory are treated as single constructs.

▪ Publication bias

According to Egger’s asymmetry test, the overall EF was subject to publication bias. Even after grouping by EF domain, obvious publication bias remained for inhibition, reasoning, shifting and working memory, suggesting that null results or results indicating better EF among PD patients may not have been published. The failsafe N indicated that the meta-analytic results for attention, inhibition, planning, shifting and working memory were robust to publication bias. However, the fail-safe N for reasoning was smaller than the threshold, indicating that the meta-analytic results for this domain were susceptible to publication bias. The effect sizes generally decreased after the trim-out-fill procedure was applied, but remained significantly different from zero. Hence, broad EF impairments were observed in PD patients, and reasoning remained the least affected EF domain.

▪ Implications

The current meta-analysis provided evidence indicating that PD patients exhibit substantial dysfunctions across different EF domains. In this population, the impairments in the attention, inhibition, planning, and reasoning, shifting and working memory domains were not moderated by age, years of education, disease severity, motor deficits, disease duration, medication dose or global cognition. These results better enable us to understand the EF profiles of PD patients, and this information will assist clinicians and caregivers with devising suitable strategies to improve the functional outcomes and quality of life of PD patients.

EFs are essential for performing daily tasks; accordingly, impairments in these domains can greatly degrade a PD patient’s quality of life and functional capability [33]. The existing literature suggests that EF can be altered through deliberate training and intervention. For instance, physical exercise has been shown to improve inhibition, planning and working memory in PD patients [34-36]. Computerized cognitive training and video games were also found to improve EF in PD patients [37,38]. The selection of an appropriate remedial approach could halt a decline in or even improve the EF of PD patients, thus reducing the challenges faced during daily life and the burdens placed on caregivers.

▪ Limitations and future research direction

This meta-analysis was limited to studies published in peer-reviewed journals. Although this criterion ensured the quality of the included studies, we risked missing relevant studies that had been published elsewhere. Additionally, some identified studies were excluded from the meta-analysis because they contained insufficient data for an effect size computation. This reduced the number of included studies and potentially introduced a non-random bias.

In addition, the selection of EF domains and tasks may have been arbitrary. According to an existing consensus, inhibition, working memory and shifting are the core EFs [9]. Although attention, planning and reasoning are crucial for goal-directed behaviors, their inclusion in EF is debatable. For instance, reasoning may depend on one’s prior knowledge and experience, rather than the ability to follow rules of logic. Moreover, more than one performance index (such as reaction time, accuracy and error rate) might apply to an EF task, and the index selection may be subjective. To ensure that the included performance indices were representative of the assessed EFs, we attempted to select the most appropriate indicators based on suggestions in the literature.

Furthermore, as most included studies failed to report the medication statuses of PD patients during testing, it was difficult to determine whether performance was affected by medication. Therefore, future reports should provide more details about the medication statuses of PD patients, which would allow a decoupling of genuine EF deficits from medication-influenced performance.

Conclusions

PD patients exhibit impairments in the EF domains of attention, inhibition, planning, reasoning, shifting and working memory; of these, reasoning is the least affected. Furthermore, these deficits are not influenced by age, years of education, disease severity, motor deficits, disease duration, medication dose or global cognition. However, as the results for reasoning were more susceptible to publication bias, compared to other EF domains, additional studies of this domain should be conducted and included in future meta-analyses. Finally, although numerous studies have demonstrated the plasticity of EF in PD patients, the efficacies of remediate EF training and intervention strategies require further verifications.

Funding

This paper was supported by the Natural Science Foundation of SZU to JHY (JCYJ20170302143406192).

Conflict of Interest

No conflict of interest is to be disclosed.

References