Professional Documents
Culture Documents
Meta-Analysis of Neuropsychological Measures of Executive
Meta-Analysis of Neuropsychological Measures of Executive
Existing literature on the profile of executive dysfunction in autism spectrum disorder showed inconsistent results.
Age, comorbid attention-deficit/hyperactivity disorder (ADHD) and cognitive abilities appeared to play a role in con-
founding the picture. Previous meta-analyses have focused on a few components of executive functions. This meta-
analysis attempted to delineate the profile of deficit in several components of executive functioning in children and
adolescents with high-functioning autism spectrum disorder (HFASD). Ninety-eight English published case-control
studies comparing children and adolescents with HFASD with typically developing controls using well-known neuro-
psychological measures to assess executive functions were included. Results showed that children and adolescents
with HFASD were moderately impaired in verbal working memory (g 5 0.67), spatial working memory (g 5 0.58), flexi-
bility (g 5 0.59), planning (g 5 0.62), and generativity (g 5 0.60) except for inhibition (g 5 0.41). Subgroup analysis
showed that impairments were still significant for flexibility (g 5 0.57–0.61), generativity (g 5 0.52–0.68), and working
memory (g 5 0.49–0.56) in a sample of autism spectrum disorder (ASD) subjects without comorbid ADHD or when
the cognitive abilities of the ASD group and the control group were comparable. This meta-analysis confirmed the
presence of executive dysfunction in children and adolescents with HFASD. These deficits are not solely accounted
for by the effect of comorbid ADHD and the general cognitive abilities. Our results support the executive dysfunction
hypothesis and contribute to the clinical understanding and possible development of interventions to alleviate these
deficits in children and adolescents with HFASD. Autism Res 2016, 0: 000–000. V C 2016 International Society for
Keywords: executive function; high-functioning autism spectrum disorder; Asperger’s syndrome; meta-analysis
From the Castle Peak Hospital, Tuen Mun, Hong Kong (C.L.E.L., S.S.Y.L., E.L., V.T., Q.C., K.M.C., S.M.L., E.F.C.C.)
Received March 11, 2016; accepted for publication October 15, 2016
Address for correspondence and reprints: Chun Lun Eric Lai, Castle Peak Hospital, No 15 Tsing Chung Koon Road, Tuen Mun, Hong Kong.
E-mail: lai_eric@yahoo.com
Published online 00 Month 2016 in Wiley Online Library (wileyonlinelibrary.com)
DOI: 10.1002/aur.1723
C 2016 International Society for Autism Research, Wiley Periodicals, Inc.
V
were of smaller magnitude in the inhibition and plan- Heterogeneity of Effect Sizes
ning constructs. The effect sizes were still of medium to
The results of assessment of heterogeneity are also pre-
large magnitude in the verbal working memory, spatial
sented in Tables 2–4. Overall the effect sizes were het-
working memory, flexibility, and generativity con-
erogeneous across all constructs. Studying a subgroup
structs. Conversely, through focusing exclusively on
of ASD samples without comorbid ADHD led to statisti-
studies matching the ASD and control subjects on IQ,
cally insignificant Q values, which indicates that the
the effect sizes were of similar magnitude across all
results were relatively more homogenous. Conversely,
constructs.
the results were still heterogeneous in all components
except inhibition in the IQ-matched subgroup.
Meta-Regression Publication Bias
The results of meta-regression are presented in Table 5. The funnel plots are displayed in Figures 3–8. The
Deficits in flexibility appeared to decrease with age results of the trim-and-fill analysis and the fail-safe N
(B 5 20.062, P 5 0.031). Regression using other variables test are presented in Table 2. With trim-and-fill analy-
and in other constructs did not yield any statistically sis, asymmetry was detected in the spatial working
significant results. memory and generativity constructs, with a reduced
Alderson-Day, 2014 ASD 28/2 ADOS/ ASD 13.84 ASD 105.07 WASI A, G, IQ 1 NR NA CFT Number of correct responses 1.09
TD 10/5 ADI-R (2.44) (14.06) LFT Number of correct responses 0.64
TD 14.05 TD 110.40
(2.72) (10.54)
Altgassen, 2009 ASD 9/2 3DI/ADOS ASD 9.6 (2.6) NR WISC A, IQ NR NR NA DS Score 0.37
TD 6/5 TD 10.6 (2.9)
Ambrosino, 2014 ASD 19/0 ADI-R ASD 11.5 (1.2) ASD 112.2 WISC A, G, IQ NR 7 SS GNG % of correct responses on no-go 0.43
TD 19/0 TD 11.1 (1.6) (15.3) trial
TD 120.2
(15.8)
Andersen, 2013 ASD 31/7 K-SADS-PL ASD 12.0 (2.3) ASD 98.2 WASI A, IQ 8 4 SS HVLT Acquisition score 0.88
TD 32/18 TD 11.6 (2.0) (17.8) LNS Number of correct responses 1.20
TD 103.8
(13.0)
Andersen, 2015 ASD 28/6 K-SADS-PL ASD 11.6 (2.0) ASD 99.9 WASI A, G, IQ 8 4 SS LNS Score 1.33
TD 29/16 TD 11.4 (1.5) (17.4) Stroop Reaction time in condition 4 0.94
TD 104.5 (D-KEFS)
(13.1)
Barron- ASD 28/2 ADI-R ASD 9.1 (1.3) ASD 107.2 WISC A, G, IQ NR 3 NR DFT Score 0.012
Linnankoski, 2015 TD 56/4 TD 9.1 (1.4) (17.3) Stroop Switching score (NEPSY) 20.14
TD NR VFT Number of correct responses 0.59
Bishop, 2005a, ASD 14/0 SCQ/ADOS ASD 8.3 (0.99) ASD 107.21 Raven A, G, IQ NR NR NA OW Time difference 1.35
2005b TD 15/3 TD 8.56 (1.00) (15.62) UCO Number of correct responses 1.00
TD 110.83 WDW Number of correct responses 1.15
(10.38)
Brenner, 2015 ASD 23/4 ADI-R/ ASD 12.68 ASD 101.31 WISC A, G, IQ NR NR NA LNS Score 0.44
TD 22/3 ADOS (2.85) (11.24)
TD 13.41 TD 106.96
(2.32) (11.46)
INSAR
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
INSAR
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Chantiluke, 2015 ASD 17/0 ADI-R/ ASD 15.2 (1.8) ASD 114 (14) WASI-R A, G, IQ 0 0 NA NBT Percentage of correct responses 0.21
TD 17/0 ADOS TD 14.0 (2.6) TD 111 (14)
Chien, 2014 ASD 316/38 ADI-R ASD 10.5 (3.1) ASD 96.75 WISC A 251 NR NA CPT Commission errors 0.43
TD 180/46 TD 11.8 (2.2) (21.88)
TD 111.95
(10.23)
Corbett, 2009 ASD 17/1 ADOS ASD 9.44 ASD 94.17 WASI A NR 6 SS CCTT Interference score 1.29
TD 12/6 (1.96) (17.79) CFT Number of correct responses 1.33
TD 9.56 (1.81) TD 112.22 IED Extra-dimensional stage errors 20.15
(14.84) LFT Number of correct responses 1.37
SOC Number of perfect solutions 0.89
SSP Score 1.17
Stroop Reaction time in condition 4 1.46
SWM (D-KEFS) 1.31
Between-search errors
Cui, 2010 ASD 11/1 Clinical ASD 7.46 ASD 100.03 WISC A, G, IQ NR NR NA BD Score 0.028
TD 24/6 (0.84) (17.13) BT Spatial span 0.91
TD 7.37 (0.48) TD 108.31 FD Score 21.15
(14.08) NBT Percentage of correct responses 20.063
(digit) 20.37
Percentage of correct responses
(figure)
Czermainski, 2014 ASD 9/2 Clinical ASD 11.73 ASD 31.36 Raven A, G, IQ 0 4 NR BD Score 0.16
TD 17/2 (1.9) (4.75)c BT Spatial span 0.96
TD 11.42 (1.8) TD 30.18 CFT Number of correct responses 0.84
(3.81) LFT Number of correct responses 0.88
NWR Score 0.64
RCFT Recall score 1.91
Stroop SCW score (SCWT) 0.69
TMT Time taken in part b 1.11
7
8
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Gabig, 2008 ASD 13/2 ADOS/ ASD 6.5 (0.7) ASD 95 (10.6) PPVT/DAS A NR NR NA DS Score 2.16
TD 10 ADI-R TD 6.67 (0.9) TD 106 (9.5) NWR Score 1.34
Geurts, 2004 ASD 41/0 ADI-R ASD 9.4 (1.8) ASD 98.3 WISC A, G 11 NR NA BVRT Number of correct responses 0.60
TD 41/0 TD 9.1 (1.7) (18.4) BT Spatial span 0.70
TD 111.5 CFT Number of correct responses 0.92
(18.0) LFT Number of correct responses 1.01
OW Time difference 0.37
SOPT Number of errors 0.57
TOL Scores based on the number of trials 0.43
WCST required to solve a problem 0.86
Number of perseverative responses
Geurts, 2009 ASD 16/2 CCC/CSBQ/ ASD 10.3 (1.6) ASD 108.0 NR A, G, IQ NR 1 SS GNG Commission errors 0.42
TD 19/3 clinical TD 10.3 (1.4) (19.0)
TD 103.2
(24.1)
Geurts, 2014 ASD 20/4 SRS/CSBQ/ ASD 10.5 (1.1) ASD 108.5 SON-R A, G, IQ NR 9 NS DS Score 0.41
TD 17/7 clinical TD 10.6 (0.7) (17.4)
TD 103.6
(13.9)
Goddard, 2014 ASD 51/12 SCQ ASD 12.55 ASD 103.6 WASI A, G, IQ NR NR NA CFT Number of correct responses 20.008
TD 51/12 (2.81) (13.08) CMS Score – visual 0.57
TD 12.10 TD 104.76 JHT Score – verbal 20.03
(2.26) (11.79) Stroop Total score of section B 0.41
TOL Number of correct responses 0.49
WCST Move score 0.58
Number of perseverative errors 0.49
Goldberg, 2005 ASD 13/4 ADI-R/ ASD 10.3 (1.8) ASD 96.5 NR A, G NR 2 SS IED Extra-dimensional stage errors 0.12
TD 21/11 ADOS TD 10.4 (1.5) (15.9) SOC Number of perfect solutions 0.55
TD 112.6 Stroop SCW score (SCWT) 0.11
(12.1) SWM Between-search errors 0.97
INSAR
Table 1. Continued
INSAR
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Kawakubo, 2009 ASD 12/2 CARS/clinical ASD 12.7 (3.4) ASD 95.2 (8.4) WISC A, G, IQ NR 10 NR LFT Number of correct responses 20.20
TD 12/2 TD 10.6 (2.8) TD 102.1 (6.5)
Kilincaslan, 2010 ASD 18/3 Clinical ASD 12.44 ASD 105.52 WISC A, G, IQ 9 6 SS CFT Number of correct responses 0.47
TD 15/3 (2.87) (14.74) CPT Commission errors 0.61
TD 11.96 TD 107.27 LFT Number of correct responses 0.76
(2.36) (13.39) Stroop Interference score (correct 0.11
WCST responses) 0.87
Number of perseverative errors
Kimhi, 2014 ASD 25/4 ADI-R/ ASD 4.95 ASD 103.52 WISC/ A, G, IQ NR NR NA TOL Scores based on the number of trials 0.53
TD 26/4 ADOS (0.92) (17.21) WPPSI/ required to solve a problem
TD 4.61 (0.91) TD 107.60 Mullen
(14.13)
Kretschmer, 2014 ASD 18/3 ADI-R/ ASD 10.22 ASD 10.86 WISC A, G, IQ NR NR NA DS Score 0.35
TD 14/7 ADOS (1.55) (2.99)c GNG % of correct responses on no-go 0.48
TD 9.83 (2.36) TD 10.71 trial
(3.75)
Landa, 2005 ASD 19 ADI-R/ ASD 11.01 ASD 109.7 WISC A, G, IQ NR NR NA IED Extra-dimensional stage errors 20.96
TD 19 ADOS (2.89) (15.8) SOC Number of perfect solutions 0.99
TD 11.00 TD 113.4 SWM Between-search errors 0.99
(2.85) (14.34)
Lee, 2009 ASD 9/3 ADI-R/ ASD 10.17 ASD 113.33 NR A, G, IQ NR NR NA GNG Commission errors 0.26
TD 8/4 ADOS (1.57) (17.33)
TD 11.91 TD 114.92
(1.78) (10.28)
Li, 2014 ASD 30/8 CARS/ABC/clinical ASD 9.59 ASD 109.76 Raven A, G, IQ NR NR NA WCST Number of perseverative errors 0.58
TD 31 (2.29) (18.85)
TD 10.20 TD 113.00
(1.53) (15.12)
Loucas, 2010 ASD 16/0 ADI-R/ ASD 14.69 NR NR A NR NR NA NWR Number of correct responses 1.17
9
Table 1. Continued
10
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Maister, 2013a ASD 14/0 ADI-R/ ASD 12.2 (0.6) ASD 43.9 Raven A, G, IQ 0 0 NA IED Extra-dimensional stage errors 0.28
TD 13/1 ADOS TD 12.1 (0.2) (6.4)c SSP Score 0.064
TD 46.4 (7.4) Stroop Interference score (correct 20.18
VFT responses) 0.45
Number of correct responses
Maister, 2013b ASD 13/1 ADI-R/ ASD 11.8 (1.4) ASD 41.6 Raven A, G, IQ 0 0 NA IED Extra-dimensional stage errors 0.48
TD 11/3 ADOS TD 11.8 (1.1) (8.3)c SSP Score 0.53
TD 44.3 (6.3)
Mashal, 2012 ASD 32/5 Clinical ASD 13.02 NR NR A, G 0 NR NA CFT Number of correct responses 2.86
TD 16/5 TD 12.09 LFT Number of correct responses 20.087
Matsuura, 2014 ASD 11/0 ADOS/ ASD 12.0 (2.2) ASD 105.6 WISC A, IQ 0 2 SS DMS Number of errors 20.32
TD 12/7 ADI-R TD 11.4 (1.6) (14.3) DS Score 20.22
TD 111.8 LNS Score 20.057
(13.4) SSP Score 0.39
SWM Between-search errors 0.72
May, 2015 ASD 20/20 SRS/clinical ASD 9.62 ASD 100.5 WISC A, G, IQ NR 7 NS FD Score 0.10
TD 20/20 (1.57) (15.1) (PIQ) SR Score 20.055
TD 9.62 (1.34) ASD 100.7
(13.4) (VIQ)
TD 104.8
(14.4) (PIQ)
TD 107.2
(10.5) (VIQ)
Narzisi, 2013 ASD 22/0 Clinical ASD 9.77 ASD 99.09 WISC A, G NR NR NA CFT Number of correct responses 0.43
TD 44/0 (3.65) (14.23) CST Score (D-KEFS) 1.44
TD NR TD NR DFT Score 0.59
LFT Number of correct responses 0.51
NWR Score 0.78
SR Score 1.35
Stroop Switching score (NEPSY) 1.23
INSAR
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
INSAR
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Ozonoff, 1995b ASD 23/1 CARS/ ASD 12.0 (3.1) ASD 97.4 WISC A, G, IQ NR NR NA WCST Number of perseverative responses 0.89
TD 19/5 clinical TD 12.4 (1.5) (17.0)
TD 101.9
(14.8)
Ozonoff, 1999 ASD 40 ADI-R/ ASD 12.6 (3.4) ASD 95.2 WISC A NR NR NA Stroop NR 0.40
TD 29 ADOS TD 12.1 (3.0) (18.8) TOH NR 0.69
TD 107.8 WCST Number of perseverative responses 0.79
(10.8)
Pankert, 2014 ASD 16/1 SCQ/SRS/ ASD 11.6 (1.5) ASD 109.3 WISC A, G, IQ 8 2 SS GNG Commission errors 0.72
TD 12/5 ADOS TD 11.7 (1.2) (17.5)
TD 109.2 (9.9)
Pellicano, 2006 ASD 35/5 ADI-R ASD 5.60 ASD 113.58 Leiter-R A, G, IQ 0 NR NA LHG Number of correct responses in the 0.67
TD 31/9 (0.91) (14.11) Maze conflict condition 0.26
TD 5.48 (0.96) TD 112.52 TOL Score 0.98
(14.47) TBCS Number of problems solved in mini- 0.76
mal number of moves
Number of trials taken
Pellicano, 2010 ASD 40/5 ADI-R ASD 5.66 ASD 113.27 Leiter-R A, G, IQ 2 NR NA TBCS Number of perseverative errors 1.12
TD 37/8 (0.87) (13.93) TOL Number of problems solved in mini- 0.98
TD 5.43 (1.05) TD 115.61 mal number of moves
(16.42)
Perez, 2009 ASD 14/1 Clinical ASD 10.8 (3.4) ASD 111.2 WASI A, IQ NR NR NA CST Number of perseverative errors 0.60
TD 9/7 TD 11.1 (2.6) (15) Stroop Number of errors 0.94
TD 123 (8)
Planche, 2012 ASD 27/3 ADI-R ASD 8.4 ASD 101.8 WISC A, G, IQ NR NR NA Tower Score (NEPSY) 20.041
TD 12/3 TD 9.0 (21.96)
TD 106.02
(8.31)
Prior, 1990 ASD 9/3 Clinical ASD 13.75 ASD 88 Leiter-R A, G NR NR NA Maze Time taken 1.26
TD 9/3 TD 13.75 TD 100 RCFT Recall score 1.12
11
12
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Russell-smith, 2014 ASD 14/3 ADI-R ASD 11.93 PIQ WISC A, G, IQ NR NR NA CST Number of perseverative errors 0.11
TD 16/2 (1.9) ASD 101.6
TD 10.69 (2.3) (14.6)
TD 102.9
(11.7)
VIQ
ASD 101.4
(14.8)
TD 109.6
(14.7)
Salmanian, 2012 ASD 15 Clinical ASD 12.8 ASD 99 Raven A NR NR NA DMS Number of correct responses 1.21
TD 15 (3.23) (11.92) PAL First trial memory score 0.74
TD 10.53 TD 113.47 PRM Number of correct responses 0.76
(3.04) (8.29) SRM Number of correct responses 1.09
Samyn, 2015 ASD 31/0 SRS/clinical ASD 12.83 ASD 101.16 NR A 0 0 NA GNG Commission errors 0.46
TD 95/53 (1.41) (12.48) Stroop Interference score (correct 0.073
TD 12.73 TD 107.21 responses)
(1.48) (11.68)
Sawa, 2013 ASD 17/2 Clinical ASD 13.24 ASD 95.95 NR A, G, IQ 0 NR NA WCST Number of perseverative errors 0.91
TD 17/2 (1.79) (12.83)
TD 13.12 TD 97.32
(1.98) (9.48)
Schneider, 1987 ASD 14/1 K-SADS ASD 10.71 ASD 85.73 WISC A, G NR 1 NS WCST Number of perseverative responses 0.27
TD 22/6 (2.14) (14.56)
TD 11.00 TD NR
(1.72)
Schuh, 2012 ASD 16/2 ADOS/ ASD 12 (3) ASD 105 (10) SB-5 A, G, IQ 0 0 NA FW Score 0.98
INSAR
INSAR
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Sinzig, 2008 ASD 35/5 ADI-R/ ASD 12.6 (3.5) ASD 108 CFIT A, G 20 7 SA GNG Commission errors 0.15
TD 14/6 ADOS TD 13.1 (3.0) (16.0) IED Extra-dimensional stage errors 0.11
TD 113 (11.9) SOC Number of perfect solutions 20.22
SWM Between-search errors 0.57
Sinzig, 2014 ASD 26/0 ADI-R/ ASD 6.7 ASD 90.5 KAB-C A 17 9 NR GNG Commission errors 0.78
TD 11/19 ADOS 0(1.18) (13.5)
TD 5.19 (1.10) TD 107.4 (8.1)
Stieglitz, 2011 ASD 17/2 ADOS/SCQ ASD 12.1 (2.4) ASD 106.0 WASI A, G, IQ 0 NR NA DS Score 0.39
TD 21/2 TD 12.0 (2.1) (21.0)
TD 111.4
(16.5)
Terrett, 2013 ASD 23/7 SCQ/clinical ASD 9.50 ASD 115.63 NR A, G, IQ NR NR NA Stroop Switching score (NEPSY) 20.12
TD 21/9 (1.31) (16.8)
TD 9.73 (1.02) TD 116.57
(17.26)
Tye, 2014 ASD 48/0 ADI-R/ ASD 10.99 ASD 112.08 WASI A, IQ 29 6 SS CPT Commission errors 0.32
TD 26/0 ADOS (1.77) (14.51)
TD 10.56 TD 120.04
(1.79) (13.42)
Unterrainer, 2016 ASD 41/0 ADI-R/ ASD 10.16 ASD 98.07 Raven A, G, IQ 23 14 SS TOL Number of problems solved in mini- 20.085
TD 42/0 ADOS (2.20) (14.19) mal number of moves
TD 9.76 (2.36) TD 97.59
(13.86)
Urbain, 2015 ASD 16/4 ADOS ASD 11.25 ASD 108.25 WASI A, G, IQ NR NR NA NBT Percentage of correct responses 0.46
TD 13/7 (1.58) (14.31)
TD 11.26 TD 115.95
(1.64) (10.97)
13
14
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Vara, 2014 ASD 12/3 ADI-R/ ASD 15.5 (1.2) ASD 103.8 WASI A, G NR 0 NA GNG Commission errors 0.62
TD 12/3 ADOS TD 15.6 (1.3) (13.6)
TD 112.4
(10.3)
Verte, 2006 ASD 99/13 ADI-R ASD 8.6 (1.9) ASD 100.5 WISC A, G, IQ 32 NR NA BT Spatial span 0.69
TD 40/7 TD 9.4 (1.6) (16.6) BVRT Number of correct responses 0.86
TD 112.1 (9.7) CFT Number of correct responses 0.93
LFT Number of correct responses 0.69
OW Time difference 0.35
SOPT Number of errors 0.65
TOL Scores based on the number of trials 1.34
WCST required to solve a problem 0.81
Number of perseverative responses
Voelbel, 2006 ASD 38/0 ADI-R ASD 10.16 ASD 99.37 WISC A, G 18 20 NS CFT Number of correct responses 0.50
TD 13/0 (1.92) (16.08) LFT Number of correct responses 0.068
TD 10.77 TD 115.15 Stroop SCW score (SCWT) 0.85
(1.48) (9.41) TMT Time taken in part b 0.96
WCST Number of perseverative errors 0.44
Vogan, 2014 ASD 16/3 ADOS ASD 11.05 ASD 109.42 WASI A, G, IQ NR 6 NS BD Score 0.74
TD 13/4 (1.43) (15.72) BT Spatial span 0.49
TD 11.12 TD 115.35 FD Score 0.46
(2.00) (9.27)
Weissman, 2010 ASD 52/3 ADI-R ASD 9.89 ASD 49.90 WISC A 19 43 NS CPT Commission errors 20.42
TD 19/8 (2.03) (9.73)c Stroop SCW score (SCWT) 0.41
TD 10.48 TD 58.36 (7.6)
(1.76)
Williams, 2005 ASD 22/2 ADOS/ ASD 11.75 ASD 109.67 WISC A, G, IQ NR NR NA FW Score 0.57
INSAR
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
INSAR
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Williams, 2013 ASD 21 3DI/SRS/clinical ASD 10.60 VIQ WASI A, IQ 0 NR NA WCST Number of perseverative errors 0.83
TD 21 (2.01) ASD 103.57
TD 10.59 (17.88)
(1.31) TD 106.48
(14.01)
PIQ
ASD 110.19
(16.35)
TD 107.48
(13.23)
Winsler, 2007 ASD 32/1 Clinical ASD 11.0 (2.3) ASD 107.07 NR A NR 19 SS WCST Number of perseverative errors 0.63
TD 19/9 TD 10.3 (3.2) (14.82)
TD NR
Xiao, 2012 ASD 19/0 ADI-R ASD 10.11 ASD 99.26 WISC A, G, IQ 0 NR NA GNG Commission errors 0.87
TD 16/0 (2.08) (9.03) Stroop Number of errors 0.34
TD 9.69 (1.74) TD 105.63
(13.12)
Yasumura, 2014 ASD 7/4 ADOS/ ASD 10.51 ASD 30.70 Raven A, G, IQ NR NR NA Stroop Interference score (correct 0.23
TD 6/9 ADI-R (2.30) (2.41)c Reversed stroop responses) 20.090
TD 9.56 (1.51) TD 29.47 Interference score (correct
(4.05) responses)
Yerys, 2009a ASD 38/11 ADI-R/ADOS ASD 9.68 ASD 114.75 WISC A, G, IQ 21 7 SS BD Score 0.58
TD 13/8 (1.90) (16.80) SWM Between-search errors 0.57
TD 10.30 TD 116.24 WDW Score 0.45
(1.76) (11.53)
Yerys, 2009b ASD 33/9 ADOS/ ASD 10.19 ASD 111.95 WASI/WISC A, G, IQ NR 2 SS IED Extra-dimensional stage errors 0.41
TD 65/19 ADI-R (2.00) (18.04)
TD 10.26 TD 113.18
(2.08) (11.94)
Yerys, 2011 ASD 21/7 ADOS/ ASD 10.89 ASD 113.86 WISC A, G, IQ NR 5 SS CTT Score 2.52
15
16
Table 1. Continued
Study Sample size Diagnostic Mean age Mean IQ EF
(1st author, year) (Male/Female) method M (SD) M (SD) IQ Ax Matching ADHDa Medb MS assessment Variable ES
Zinke, 2010 ASD 13/2 ADOS/ ASD 9.0 (1.5) ASD 96.4 WISC A, G NR 7 NS BT Spatial span 1.14
TD 14/3 ADI-R TD 9.8 (1.7) (14.5) FD Longest sequence 0.64
TD NR TOL Number of problems solved in mini- 0.95
mal number of moves
a
Number of subjects with comorbid ADHD in the ASD sample.
b
Number of medicated subjects in the ASD sample, irrespective of what the medications were and whether the medications were discontinued at the time of testing.
c
The reported IQ scores were on different scales.
General abbreviations: A 5 Age; G 5 Gender; NR 5 Not recorded; M 5 Mean; SD 5 Standard Deviation; IQ Ax 5 Method of IQ assessment; MS 5 Medication status (NS 5 Not stopped; SS 5 Stopped
stimulant; SA 5 Stopped all medications; NA 5 Not applicable); EF 5 Executive functioning; ADHD 5 Attention-deficit/hyperactivity Disorder; ASD 5 Autism Spectrum Disorder.
Abbreviations for diagnostic methods: 3DI 5 The Developmental, Dimensional and Diagnostic Interview; ABC 5 Aberrant Behavior Checklist; ADI-R 5 Autism Diagnostic Interview-Revised;
ADOS 5 Autism Diagnostic Observation Schedule; AQ 5 Autism Spectrum Quotient; CARS 5 Childhood Autism Rating Scale; CCC 5 Children’s Communication Checklist; CSBQ 5 Children’s Social Behavior
Questionnaire; KSADS-PL 5 Kiddie-Sads-Present and Lifetime Version; SCQ 5 Social Communication Questionnaire; SRS 5 Social Responsiveness Scale.
Abbreviations for inhibition assessments: CPT 5 Continuous Performance Test; GNG 5 Go-No-Go; JHT 5 Junior Hayling Test; LHG 5 Luria’s Hand Game; OW 5 Opposite World; Stroop 5 Stroop
Color Word Test; TEA-Ch 5 Test of Everyday Attention for Children; WDW 5 Walk-Don’t-Walk.
Abbreviations for verbal working memory assessments: BD 5 Backward Digit Span; CTT 5 Consonant Trigrams Test; DS 5 Digit Span; FD 5 Forward Digit Span; HVLT 5 Hopkin’s Verbal Learning
Test; LNS 5 Letter Number Sequencing; SR 5 Sentence repetition; WLI 5 Word list interference; WMTB 5 Working Memory Test Battery for Children.
Abbreviations for spatial working memory assessments: BT 5 Block Tapping; BVRT 5 Benton Visual Retention Test; DM 5 Designs Memory; DMS 5 Delayed Matching to Sample; FW 5 Fingers
Window; NBT 5 N-Back Task; PAL 5 Paired Associates Learning (CANTAB); PM 5 Picture Memory; PRM 5 Pattern Recognition Memory (CANTAB); RCFT 5 Rey Complex Figure Test; SRM 5 Spatial Recogni-
tion Memory (CANTAB); SSP 5 Spatial Span Test (CANTAB); SWM 5 Spatial Working Memory (CANTAB); SOPT 5 Self-ordered Pointing Test.
Abbreviations for flexibility assessments: CCTT 5 Children’s Color Trails Test; CST 5 Card Sorting Test; IED 5 Intra/Extra-Dimensional Shift; TBCS 5 Teddy Bear Card Sorting Test; TMT 5 Trails
Making Test; WCST 5 Wisconsin Card Sorting Test.
Abbreviations for planning assessments: SOC 5 Stockings of Cambridge; TOH 5 Tower of Hanoi; TOL 5 Tower of London.
Abbreviations for generativity assessments: DFT 5 Design Fluency Test; LFT 5 Letter Fluency Test; UCO 5 Use of Common Objects; VFT 5 Verbal Fluency Test.
Abbreviations for batteries: CMS 5 Children’s Memory Scale; CANTAB 5 Cambridge Neuropsychological Test Automated Battery; DKEFS 5 Delis-Kaplan Executive Function System; NEPSY 5 A Devel-
opmental Neuropsychological Assessment; WRAML 5 Wide Range Assessment of Memory and Learning.
Abbreviations for IQ assessments: CFIT 5 Culture Fair Intelligence Test; DAS 5 Differential Ability Scales; KAB-C 5 Kaufman Assessment Battery for Children; Mullen 5 Mullen Scales of Early
INSAR
Table 2. Summary of Results
95% CI Heterogeneity test Publication bias
Measure N (ASD/TD) k g LL UL SE Z P Q I2 T&F gadj NFs
INHIBITION
Continuous Performance Test 488/316 5 0.20 20.16 0.55 0.18 1.10 0.28 13.07* 69.40 Nil NA 6
Go-No-Go 302/403 13 0.49 0.33 0.64 0.080 6.07 <.001 6.52 0.00 Nil NA 106
Opposite World 167/106 3 0.59 0.11 1.06 0.24 2.42 0.015 5.83# 65.71 Nil NA 11
Stroop 590/781 22 0.39 0.22 0.57 0.088 4.50 <.001 46.38** 54.72 Nil NA 234
Walk-Don’t-Walk 74/57 3 0.58 0.18 0.98 0.20 2.87 0.004 2.29 12.84 Nil NA 6
Inhibition Composite 1,534/1,517 42 0.41 0.31 0.51 0.050 8.22 <.001 61.34* 33.16 Nil NA 1,110
VERBAL WORKING MEMORY
Non-Word Repetition 113/147 7 0.96 0.70 1.21 0.13 7.25 <.001 4.17 0.00 R1 0.92 90
Digit Span 429/440 19 0.46 0.25 0.68 0.11 4.28 <.001 41.19** 56.30 Nil NA 187
Letter-Number Sequencing 188/239 7 0.54 0.13 0.96 0.21 2.56 0.011 25.71** 76.66 Nil NA 48
Sentence repetition 124/165 5 1.09 0.31 1.87 0.40 2.75 0.006 36.01*** 88.89 Nil NA 59
Verbal WM Composite 728/811 30 0.67 0.45 0.88 0.11 6.14 <.001 114.33*** 74.74 Nil NA 1,018
SPATIAL WORKING MEMORY
Block Tapping 321/267 7 0.65 0.46 0.84 0.097 6.70 <.001 6.06 1.05 R2 0.60 79
Delayed Match to Sample 51/59 4 0.48 20.16 1.12 0.33 1.46 0.14 8.46* 64.52 Nil NA 3
Rey Complex Figure Test 44/52 3 1.13 0.31 1.95 0.42 2.69 0.007 6.97* 71.30 Nil NA 17
SSP (CANTAB) 57/65 4 0.55 0.08 1.01 0.24 2.31 0.021 5.00 40.05 L1 0.71 6
SWM (CANTAB) 214/189 8 0.67 0.45 0.90 0.12 5.82 <.001 8.29 15.51 Nil NA 80
Finger Windows 126/144 4 0.73 0.49 0.98 0.13 5.82 <.001 0.93 0.00 R1 0.69 32
N-Back Task 146/152 5 0.099 20.17 0.36 0.14 0.73 0.46 4.82 17.03 Nil NA 0
Spatial WM Composite 816/738 28 0.58 0.45 0.71 0.070 8.56 <.001 41.04* 34.21 R6 0.48 809
FLEXIBILITY
Card Sorting Task 168/189 6 0.79 0.44 1.13 0.18 4.47 <.001 12.16* 58.87 L1 0.90 71
IED (CANTAB) 164/201 8 0.090 20.18 0.36 0.14 0.65 0.51 10.46 33.07 Nil NA 0
Trail Making Test 134/120 5 1.03 0.76 1.29 0.14 7.58 <.001 0.87 0.00 Nil NA 70
Wisconsin Card Sorting 739/641 21 0.62 0.51 0.73 0.056 11.16 <.001 16.49 0.00 R1 0.61 600
Flexibility Composite 1,112/1,080 37 0.59 0.48 0.71 0.058 10.17 <.001 58.77* 38.74 Nil NA 1,531
PLANNING
Stocking of Cambridge 104/99 5 0.43 20.061 0.93 0.25 1.72 0.086 11.64* 65.63 Nil NA 6
Tower of London and variants 577/518 14 0.67 0.42 0.92 0.02 5.17 <.001 52.42*** 75.20 Nil NA 373
Planning Composite 733/658 21 0.62 0.42 0.83 0.11 5.93 <.001 66.36*** 69.86 Nil NA 609
GENERATIVITY
Verbal Fluency 663/593 20 0.59 0.40 0.78 0.095 6.16 <.001 45.74** 58.46 R4 0.46 445
Design Fluency 132/184 4 0.35 0.054 0.65 0.15 2.31 0.021 5.04 40.50 R1 0.25 6
Use of Common Objects 103/107 3 1.03 0.50 1.56 0.27 3.79 <.001 6.17* 67.56 Nil NA 33
Generativity Composite 757/691 23 0.60 0.43 0.78 0.089 6.75 <.001 53.44*** 58.83 R4 0.49 624
Note. Only tasks utilised by more than three studies are listed in this table. SSP 5 Spatial span task of CANTAB; SWM 5 Spatial Working Memory
task of CANTAB; IED 5 Intra-extra Dimesntional Set-shifting task; CANTAB 5 Cambridge Neuropsychological Test Automated Battery; N 5 Number of
subjects; k 5 Number of studies; g 5 Hedges’s g; LL 5 Lower limit; UL 5 Upper limit; CI 5 Confident interval; SE 5 Standard error; Z 5 Z value; P 5 P-
value; Q 5 Q statistics; T&F 5 Trim and fill test; gadj 5 Adjusted Hedges’s g; NFs 5 Number of studies needed to nullify results in fail-safe N test;
Nil 5 No articles were trimmed by T&F; L 5 Studies were trimmed to the left of the mean with the number of articles trimmed denoted by the follow-
ing digit; R 5 Studies were trimmed to the right of the mean with the number of articles trimmed denoted by the following digit; NA 5 Not applica-
ble. Other abbreviations are listed under Table 1.
# P < .10; *P < .05; **P < .01; ***P < .001.
weighted mean effect sizes 0.48 and 0.49, respectively, adolescents with high-functioning ASD across six major
on subsequent incorporation of hypothetical missing components of executive function. It is also the first
studies. The Fail-safe N test showed that 609 to 1,531 meta-analysis to examine the impact of comorbid
negative articles had to be identified to render the effect ADHD on the expression of executive function deficits
sizes non-significant. in ASD. Positive weighted mean effect sizes were found
across all components of executive function. In particu-
lar, the impairment of verbal working memory, spatial
Discussion working memory, flexibility, and generativity were evi-
dent irrespective of whether comorbid ADHD was
This is the first meta-analysis conducted to examine the excluded or whether an IQ-matched control group was
profile of executive function in children and used.
Working Memory, Flexibility, and Generativity as Core by the effect of comorbid ADHD and IQ. When consid-
Deficits ering the results of individual tests, there was consider-
able variability in the effect sizes. In the verbal working
The pooled effect size of verbal working memory (g 5.
memory component, more impairment was elicited by
67) was comparable to that of spatial working memory
the Sentence Repetition task and the Non-Word Repeti-
(g 5 0.58) as a whole. The results were similar when
tion task. Compared with other verbal working memory
considering the subgroups excluding co-morbid ADHD
(g 5 0.53 for verbal; g 5 0.51 for spatial) or participants tasks (such as the digit span task and the letter-number
with matched IQ (g 5 0.56 for verbal; g 5 0.50 for spa- sequencing task), these two tasks place high demand on
tial). This suggests the absence of differential impair- language ability [Conti-Ramsden, Botting, & Faragher,
ment of these constructs, in contrast to the postulation 2001]. Given that individuals with ASD often have
by some of the researchers [Cui, Gao, Chen, Zou, & communication difficulties, encompassing syntax, pho-
Wang, 2010; Williams, Goldstein, Carpenter, & Min- nological skills, and communications skills [Bartak, Rut-
shew, 2005]. Moreover, the fact that the impairments ter, & Cox, 1975], it is difficult to distinguish whether
of both components in either condition in the sub- poor performance on these tasks actually reflects execu-
group analyses were of at least medium magnitude sug- tive dysfunction or poor language skills [Joseph,
gests that these impairments were not solely explained McGrath, & Tager-Flusberg, 2005]. Conversely, in the
Inhibition 42 20.027 0.026 0.30 17 20.011 0.011 0.31 32 0.35 0.48 0.47 21 0.006 0.005 0.24
Verbal Working Memory 30 20.018 0.057 0.75 13 0.038 0.036 0.30 21 0.94 1.18 0.42 15 20.009 0.008 0.22
Spatial Working Memory 28 20.028 0.042 0.51 14 20.018 0.019 0.33 21 20.062 0.58 0.92 16 0.004 0.006 0.53
Flexibility 36 20.062 0.029 0.031 14 20.0064 0.017 0.71 30 0.60 0.65 0.35 13 0.003 0.004 0.40
Planning 21 20.059 0.040 0.13 9 2<0.001 0.064 0.99 18 20.39 1.02 0.70 5 0.023 0.014 0.098
Generativity 22 20.073 0.050 0.14 11 20.0083 0.020 0.67 19 0.88 0.76 0.25 9 20.004 0.008 0.59
Note. k 5 Number of studies; B 5 Regression coefficient; SE 5 Standard error of the slope; P 5 P-value; % Male 5 Percentage of male in the ASD
sample; % Med 5 Percentage of medicated subjects in the ASD sample.
Figure 4. Funnel plot of verbal working memory. Figure 7. Funnel plot of planning.
Figure 5. Funnel plot of spatial working memory. Figure 8. Funnel plot of generativity.
as in the Verbal Fluency and Use of Common Object ADHD was considered. Apparently, our finding suggests
tasks, but still existed when it was non-verbal, as in the that inhibition and planning are not the most promi-
Design Fluency task. This may suggest that the genera- nent deficits in ASD. Friedman and Miyake [2004]
tivity impairment is further exacerbated when demand argued that inhibition is not a unitary construct but a
is placed on the participants’ communicative abilities as constellation of three different functions, including pre-
suggested by Bishop and Norbury [2005a]. potent response inhibition, resistance to distractor
Inhibition and planning were both shown to be interference and resistance to proactive interference.
impaired in ADHD [Willcutt et al., 2005]. This meta- However, our meta-analysis did not separately analyse
analysis demonstrated that the effect sizes of inhibition these different types of inhibition, and also excluded
and planning were of small to medium magnitude several experimental paradigms which specifically mea-
when the subgroup of participants without co-morbid sure interference controls (such as the Flanker Task and