Table 2.
Attention
| Study | Age at different assessments (in months) | Total number of participants | Number of EL siblings | Number of siblings with ASD outcome | Number of siblings with TD outcome | Number of siblings with another outcome a | Topic | Assessment method | Diagnostic instruments | Difference EL-ASD vs EL-TD And/or predicts ASD diagnosis | Main findings |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Bedford et al. (2012) UK [39] | 7, 13, 24, 36 | 73 | 35 | 12 | 14 | 9 | Social attention without social interaction | Tobii 1750 (50Hz) Gaze following task | Autism Diagnostic Observation Schedule-Generic (ADOS-G) Autism Diagnostic Interview-Revised (ADI-R), expert clinical judgment according to ICD-10 | Yes |
At 7 m there was no difference between EL and TL or between EL-ASD, EL-Other, EL-TD and TL in the proportion of first look to congruent versus incongruent stimuli and in looking time towards congruent stimuli At 13 m EL-ASD showed reduced looking time to congruent stimuli compared to EL-TD and to TL and in EL-Other compared to EL-TD and to TL; no difference in the first look to congruent versus incongruent stimuli between any group was found |
| Bedford et al. (2014) UK/Canada [31] | 13, 36 | 88 | 40 | 16 | 24 | – | Disengagement of attention | Ocular-motor behaviour videotaped during the gap-overlap task and coded by raters | ADOS-G, ADI-R, expert clinical judgment according to ICD-10 | Yes | Increased saccadic reaction time at 13 m predicted ASD outcome at 36 m |
| Bryson et al. (2018) Canada [32] | 6, 12, 36 | 136 | 83 | 16 | 67 | – | Disengagement of attention | Ocular-motor behaviour videotaped during the gap-overlap task and coded by raters | ADOS-G, ADI-R, expert clinical judgment according to DSM-IV-TR | Yes |
Disengagement of attention: at 6 m no between group difference was found at 12 m EL-ASD had longer disengage latencies (particularly for latency towards stimuli appearing on the left) than EL-TD and TL. Disengage latencies did not differ between EL-TD and TL only EL-ASD showed an increase in left-sided saccadic latency from 6 to 12 m Disengagement (right and left latencies) at 12 m predicted ADOS severity score at 36 m |
| Attention shift |
Attention shift: no group differences in the shift trials were found |
||||||||||
| Cheung et al. (2018) UK [36] | 9, 15, 27, 36 | 140 | 113 | 17 | 64 | 32 | Visual search | Tobii T120 visual search task | Autism Diagnostic Observation Schedule-2 (ADOS-2), ADI-R, expert clinical judgment according to DSM-5 | Yes |
At 9 m EL-ASD showed longer looking time and higher number of valid trials than EL-TD and TL, while no difference was found between EL-ASD and EL-Other. EL-ASD showed higher proportion of first look to the target than EL-Other and TL No between group differences at 15 m and 2 years in looking time and number of valid trials At 15 m but not at 2 years EL-ASD showed superior visual search performance than EL-Other and TL |
| Elsabbagh et al. (2011) UK/Canada [58] | 10, 36–48 | 27 | 27 | Not reported | Not reported | – | Saccadic inhibitory control | Freeze-frame task (videotaped and coded) | ADOS-G | Not reported | At 10 m EL who looked more at the boring central targets than at the distractors had a higher level of impairment in social skills at 36-48 m, compared to TL |
|
Elsabbagh et al. (2013) UK [59] |
6–10, 12–15, 24, 36 | 104 | 54 | Not reported | Not reported | – | Disengagement of attention | Ocular-motor behaviour videotaped during the gap-overlap task and coded by raters | ADOS-G, ADI-R, expert clinical judgment according to ICD-10 | Yes |
During the first year no between-group difference in reaction time was found At 14 m EL-ASD had prolonged overlap RT compared to EL-TD, EL-Other and TL Different developmental courses were described: TL and EL-TD but not EL-ASD showed decreased RT in the overlap task over time |
| Elsabbagh et al. (2013) UK/Canada [60] | 7, 14, 24, 36 | 103 | 53 | 17 | 24 | 12 | Social attention without social interaction | Tobii (50 Hz) Face pop-out task | ADOS-G, ADI-R expert clinical judgment according to ICD-10 | No |
At 7 m and 14 m the proportion of trials with first look at faces was above chance in all groups At 14 m EL explored fewer AOIs than TL but this effect was not explained by the diagnosis EL spent more time than TL looking at the face area than other AOIs, and this effect was more clear at 14 m than 7 m |
| Falck-Ytter et al. (2018) Sweden/UK [40] | 10, 36 | 47 | 33 | 13 | 12 | − | Social attention without social interaction: multisensory integration during biological motion |
Tobii 1750 & Tobii TX300 Point light animations |
ADOS-G/ADOS-2, ADI-R, expert clinical judgment according to DSM-5 | No |
EL-ASD at 10 m were less able to differentiate between audiovisual synchrony, synchronous to the upright animation, versus synchronous to the inverted/reversed animation compared to EL-TD and TL No between-group difference was found in the preference for the upright compared to the inverted animation |
| Gammer et al. (2015) UK/Canada [29] | 7, 14, 24, 36 | 104 | 54 | 17 | 36 | Visual tracking | Autism Observation Scale for Infants (AOSI) | ADI-R, Social Communication Questionnaire (SCQ), ADOS-G, expert clinical judgment according to ICD-10 | No | At 7 m EL-TD showed poorer visual tracking scores than TL | |
| Attention engagement | At 14 m between-group differences in engagement of attention were not significant after post-hoc tests | ||||||||||
| Gliga et al. (2015) UK [61] | 9, 15, 24 | 109 | 82 | Visual search |
Tobii T120 visual search task |
AOSI (15 m) ADOS-2 (24 m) | Yes | Visual search at 9 m predicted ASD symptoms at 15 m and ADOS score at 24 m | |||
| de Klerk et al. (2014) UK [37] | 7, 36 | 84 | 44 | 14 | 19 | 11 | Social attention without social interaction |
Tobii (50 Hz) Face recognition task |
ADOS-G, ADI-R, expert clinical judgment according to ICD-10 | No | Face recognition task: at 7 m TL performed above chance level on easy and difficult items; EL only on easy items; there was no difference between each EL subgroup (EL-ASD, EL-TD, EL-Other) and TL, and no association between ASD-like characteristics and the performance on the task |
| Face pop-out task | Face pop-out task: longer looking time at faces at 7 m was associated with poorer face recognition at 36 m in EL, but not in TL | ||||||||||
| Rutherford et al. (2015) US [38] | 3, 6, 9, 12, 36 | 62 | 31 | 10 | 21 | Social attention without social interaction |
Tobii 60x (60 Hz) Face’s free viewing |
ADOS-2 | Yes |
At 3 m EL-TD looked longer at the eye region than TL, EL-ASD looked at the eye region less than the other two groups At 3 m EL-TD had a bigger preference towards eyes over mouth than EL-ASD, while difference between EL-TD and TL and between EL-ASD and TL was not significant after Bonferroni correction At 6 m, 9 m, 12 m the between-group differences were not significant Across the first year EL-TD showed a decreased preference for the eyes, the same change was not significant for EL-ASD and was marginally significant for TL |
|
| Sacrey et al. (2013) Canada [62] | 6, 9, 12, 15, 18, 24, 36 | 30 | 20 | 10 | 10 | Disengagement of attention | AOSI, ADOS-G (videorecorded and coded) | ADOS-G, ADI-R, expert clinical judgment according to DSM-IV-TR | Yes | Disengagement of attention: EL-ASD were less likely to look away from the target before the grasp was complete and during the grasp compared to EL-TD and TL. Group differences started at 12 m, but were no longer evident at 36 m | |
| Attention engagement | Attention engagement: EL-TD and TL were more likely to move their hand towards a target before visually engaging it, compared to EL-ASD | ||||||||||
| Sustained attention | Sustained attention: EL-ASD disengaged and re-engaged the target prior grasp more than TL | ||||||||||
| Wass et al. (2015) UK [34] | 8, 36 | 94 | 45 | 15 | 30 | Sustained attention |
Tobii 1750 (50 Hz) Free viewing of static scenes |
ADOS-G, ADI-R, expert clinical judgment according to ICD-10 | Yes |
At 8 m shorter fixation duration was found in EL than in TL and in EL-ASD than in TL; no significant difference was present between EL-TD and TL or between EL-ASD and EL-TD Shorter fixation duration at 8 m was associated with higher scores on the social communication ADOS scale at 36 m |
|
| Zwaigenbaum et al. (2005) Canada [33] | 6, 12, 24 | 88 | 65 | 7 | 46 | 12 | Disengagement of attention | Visual orienting task (videorecorded), AOSI | ADOS-G, ADI-R, expert clinical judgment according to DSM-IV | Yes | Disengagement of attention: between 6 and 12 m EL showed prolonged latencies in attention disengagement compared to TL while no difference was seen at 6 m. Disengagement score from the AOSI at 12 m predicted ASD at 24 m |
| Attention shift and engagement |
Attention engagement: EL-ASD at 12 m showed longer duration of orienting to objects than EL-TD and TL and less attention shifting at 24 m Attention shift: results showed no between-group difference in the latency to shift attention |
||||||||||
| Sustained attention | Sustained attention: at 12 m EL-ASD had longer fixations on specific objects than EL-TD and TL |
aAtypical outcome: deficit in general cognition, motor functioning, language delay, Broader Autism Phenotype (BAP, social communication delay)