Table 1 Included studies on working memory
Authors | NASD/NTD | ASD age | Diagnosis method Diagnosis criteria Diagnosis | Material Memory task | Neuroimaging method Memory phase | Memory accuracy | Main results (ASD relative to TD groups) Conclusion |
|---|---|---|---|---|---|---|---|
Audrain et al. [67] | 24/30 | 28.2 ± 6.59 | ADOS HFA | Visuospatial: abstract images 1-back | MEG Encoding, maintenance, recognition | Similar | Similar WM-related network in the theta frequency band between encoding and recognition and between ASD and TD, differing on maintenance Different WM-related network in the alpha frequency band between encoding, maintenance, and recognition and between ASD and TD Atypical functioning of WM maintenance, lack of inhibition of task-irrelevant processes |
Barendse et al. [68] | 13/13 | 15.3 ± 1.2 | ADOS DSM-IV Autistic disorder, Asperger, PDD-NOS | Visuospatial: faces and houses 1- and 2-back | fMRI Recognition | Similar | Similar activity and functional connectivity Analyses of sparse matrices: reduced global efficiency suggesting an increase in network costs WM difficulties are compensated |
Braden et al. [64] | 16/17 | 50.1 ± 1.7 | ADOS-2 DSM-IV DSM5 ASD | Verbal: letters 0-, 1-, 2-back | fMRI Recognition | Similar | Similar activation of the WM-related network, and similar deactivation of the DMN and salient network Decreased connectivity in the cortico-striatal-thalamic network when increasing memory load Functional brain difference persists with aging |
Chantiluke et al. [59] | 17/22 | 15.2 ± 1.8 | ADI-R and ADOS ICD-10 ASD | Verbal: letters 1-, 2-, 3-back | fMRI Recognition | Reduced | Under placebo: underactivation of the right dorsolateral PFC, and enhanced DMN deactivation Under fluoxetine: normalization of the dorsolateral PFC activation, and sub-normalization of behavioral results Fluoxetine modulates WM brain activation |
Hawco et al. [69] | 24/20 | 22.5 ± 4.7 | ADOS-2 ASD | Spatial: location of a dot 0-, 2-back | fMRI Recognition | Reduced, marginally significant | Greater variability in the pattern of dorsolateral PFC activation, in spite of similar between-group activation and connectivity Idiosyncratic prefrontal activation leads to diminished performance |
Herrington et al. [55] | 12/19 | 13.4 ± 4.2 | ADI-R and ADOS DSM-IV Autism disorder, Asperger, PDD-NOS | Visuospatial: superimposed faces and houses 1-back | fMRI Recognition | Similar | When attending to faces: increased activation within multiple PFC areas, including dorsolateral part Suggests selective attention in order to further discriminate faces |
Kleinhans et al. [57] | 19/21 | 23.5 ± 7.8 | ADI-R and ADOS DSM-IV Autistic disorder, Asperger, PDD-NOS | Visuospatial: neutral faces, or houses 1-back | fMRI Recognition | Similar | Similar activation of bilateral fusiform face areas, less functionally connected with both the subcortical face processing structures and the limbic structures Diminished connectivity within the limbic system supports social impairments |
Koshino et al. [51] | 14/13 | 25.7 | ADI and ADOS HFA | Verbal: letters 0-, 1-, 2-back | fMRI Recognition | Similar | Reduced activation in the left hemisphere, enhanced activation in the right hemisphere and posterior regions Reduced frontal connectivity; enhanced connectivity within posterior/occipital areas Enhanced processing of low-level visual features |
Koshino et al. [56] | 11/11 | 24.5 ± 10.2 | ADI-R and ADOS HFA | Visuospatial: neutral and emotional faces 0-, 1-, 2-back | fMRI Recognition | Similar | Reduced activations in both left inferior prefrontal areas and right posterior temporal areas Reduced fronto-fusiform functional connectivity Less verbally and socially oriented memory for faces |
Larrain-Valenzuela et al. [66] | 21/20 | 21.9 ± 4.1 | AMSE DSM5 ASD | Verbal: consonants Modified Sternberg task | EEG Recognition | Reduced | Modulation of alpha power (occipital cortex) and theta power (left premotor cortex and the right PFC) in TD. Absence of load-dependent modulation of alpha and theta power in ASD Atypical oscillatory activity contributes to WM difficulties |
Luna et al. [60] | 11/6 | 32.3 ± 9.3 | ADI-R and ADOS Autism | Visuospatial: location of a stimulus Oculomotor delayed response | fMRI Recall | Reduced | Decreased activation in both the dorsolateral PFC and the posterior cingulate cortex Fronto-parietal activation is reduced during WM |
Rahko et al. [63] | 28/22 | 14.6 [11.4–17.6] | ADI-R, ADOS, ASSQ ICD1-10, DSM-IV HFA, Asperger | Visuospatial: location of a stimulus 0-, 2-back | fMRI Recognition | Similar | Similar activation of WM-related areas, and similar DMN deactivation. Persistence of DMN when switching from 2- to 0-back in ASD only Inadequate memory load-dependent modulation of brain activity |
Silk et al. [62] | 7/9 | 14.7 ± 2.9 | ADI DSM-IV Autistic disorder, Asperger | Visuospatial: 3-dimensional block shapes Mental rotation task | fMRI Recognition | Similar | Reduced activation in frontal cortical and subcortical structures. Similar parietal activation Dysfunction of the fronto-striatal network |
Urbain et al. [54] | 20/20 | 11.25 ± 1.58 | ADOS HFA | Visuospatial: abstract images 1-, 2-back | MEG Recognition | Similar | Greater load-dependent activation in the left dorsal parietal cortex in the TD, while in the left ventral parietal cortex in ASD More controlled (i.e., top-down) processing of visual stimuli in the TD group, while being more automatic (i.e., bottom-up) in the ASD group |
Urbain et al. [58] | 17/20 | 11.17 ± 1.69 | ADOS HFA | Visuospatial: abstract images 2-back | MEG Recognition | Reduced, marginally significant | Reduced fronto-temporal connectivity in the alpha band Reduced electrophysiological connectivity contributes to WM difficulties |
Vogan et al. [52] | 19/17 | 11.05 ± 1.43 | ADOS HFA | Visuospatial: colored figures 1-back Color Matching Task, 6 levels of difficulty | fMRI Recognition | Reduced | Enhanced posterior activation. Reduced activation of frontal areas and the precuneus Less reliance on the fronto-parietal network for WM |
Vogan et al. [61] | 27/30 | 12.56 ± 1.46 | ADOS ASD | Verbal: letters Letter matching task, 4 levels of difficulties | fMRI Recognition | Reduced | Reduced activation of fronto-parietal areas when increasing the memory load; similar DMN deactivation Inadequate modulation of fronto-parietal activity when increasing WM load |
Vogan et al. [53] | 14/15 | 13.4 ± 1.82 | ADOS ASD | Visuospatial: colored figures 1-back Color Matching Task, 6 levels of difficulty | fMRI Recognition | Similar | 2-years follow-up of Vogan et al. (2014) Minimal age-related changes in load-dependent activation, but greater longitudinal load-dependent deactivation in DMN Inadequate modulation of fronto-parietal activity when increasing WM load |
Yuk et al. [65] | 40/39 | 27.17 ± 6.27 | ADOS ASD | Visuospatial: abstract images 1-, 2-back | MEG Maintenance, recognition | Similar | Maintenance: similar phase synchrony in theta, alpha, or beta bands; additional recruitment of fronto-parietal areas Recognition: diminished theta coherence between right frontal and left parietal areas Decreased theta connectivity; compensatory alpha recruitment |