main-content

## Swipe om te navigeren naar een ander artikel

29-01-2018 | Original Paper | Uitgave 6/2018 Open Access

# Allocentric Versus Egocentric Spatial Memory in Adults with Autism Spectrum Disorder

Tijdschrift:
Journal of Autism and Developmental Disorders > Uitgave 6/2018
Auteurs:
Melanie Ring, Sebastian B. Gaigg, Mareike Altgassen, Peter Barr, Dermot M. Bowler
Belangrijke opmerkingen

## Electronic supplementary material

The online version of this article (https://​doi.​org/​10.​1007/​s10803-018-3465-5) contains supplementary material, which is available to authorized users.

## Introduction

Autism spectrum disorder (ASD) is a developmental disorder that is characterised by difficulties in the areas of social interaction and communication and restricted and repetitive behaviours (American Psychiatric Association 2013). Individuals with ASD show a heterogeneous cognitive profile with a specific pattern of intact and compromised processes in memory (Boucher and Bowler 2008; Boucher et al. 2012). Rote memory, which is the ability to learn material without understanding its meaning, was found to be a strength of individuals with ASD (e.g. Hermelin and OConnor 1975). Spared performance was also reported in tests measuring priming (Bowler et al. 1997), immediate cued recall (e.g. Mottron et al. 2001) and recognition memory (e.g. Farrant et al. 1998). Given that these procedures provide more support at test, Bowler et al. ( 1997) proposed the ‘task support hypothesis’ stating that ASD individuals show less difficulties when they can rely on external sources of support (e.g. recognition rather than free recall of items; Bowler et al. 2004, 2015; Gaigg et al. 2008; Minshew et al. 1992; Mottron et al. 2001; Toichi and Kamio 2003). Bowler et al. ( 2011) in their relational binding account have argued that the reason for this difficulty on unsupported test procedures is a reduced capacity for relational binding in ASD, i.e. difficulties linking elements of experience to one another or to their spatial or temporal context to form a coherent episodic representation in order to enable flexible retrieval of that information. These relational binding difficulties become apparent when ASD participants are asked to remember the context of an item presentation, for example temporal (e.g. Bennetto et al. 1996; Bigham et al. 2010; Gaigg et al. 2014; Minshew and Goldstein 1993; Ring et al. 2016), spatial (e.g. Bowler et al. 2014, 2004; Cooper et al. 2015; Ring et al. 2015, 2016; Semino et al. 2017) or other types of context information (e.g. Hala et al. 2005; Lopez and Leekam 2003; Maister et al. 2013; O’Shea et al. 2005).
Another capacity that is reliant on relational binding is spatial navigation. One way for spatial navigation to be successful is that one needs to create an abstract map representation of the environment which depicts the relation among goal location, object cues and travel direction in the environment. Neurologically, relational binding has been demonstrated as a capacity of the hippocampus (Eichenbaum 2004; Opitz 2010) and also spatial navigation has been shown to be at least in part dependent on the hippocampus (e.g. Bohbot et al. 2004). In particular, allocentric navigation or “shifted view-point representation” refers to navigation dependent on the processing of the relations among goal and landmarks independent of a single view-point and is regulated through the (right) hippocampus (Bohbot et al. 2004). Egocentric navigation or “same view-point representation” on the other hand describes navigation using the self as a reference for navigating a route and is regulated through the caudate nucleus (e.g. Bohbot et al. 2004; Hartley et al. 2004). Following the relational binding account, one would expect specific difficulties with allocentric navigation in ASD, yet few studies have examined this issue.
Similarly to Feigenbaum and Morris ( 2004), we expected that participants in both groups would show learning across trials for all conditions and that adults with ASD would show particular difficulties in allocentric navigation, leaving egocentric navigation intact. Further, we expected a similar pattern of results for the two added conditions with individuals with ASD showing difficulties in allocentric 2 but not egocentric 2 compared to the TD group. In addition to the Morris Water Maze task, three tasks to assess participants’ visual short-term memory and mental rotation were administered (Feigenbaum and Morris 2004). This was to explore whether people with ASD show difficulties with the temporary storage and manipulation of spatial information per se, which would point to additional difficulties related to functions based outside the hippocampus, for example involving parietal brain regions (Silk et al. 2006; Tadi et al. 2009; Zacks 2008). Following the relational binding account (Bowler et al. 2011), we did not expect differences between groups on tests of visual short-term memory and mental rotation.

## Methods

### Participants

Table 1
Descriptive statistics for individuals with autism spectrum disorder (ASD) and typically developing (TD) individuals
Measure
ASD (23m, 3f)
TD (18m, 8 f)
t(50)
p
Cohen’s d
M
SD
M
SD
Age (years)
38.81
11.82
42.12
12.14
1.00
.32
.28
VIQ a
109
16.6
111
16.3
.54
.59
.15
PIQ b
108
19.6
107
17.6
.22
.83
.06
FIQ c
110
18.7
110
17.9
.18
.86
.05
2.74 (0–5)
1.39

6.74 (3–13)
2.96

9.48 (3–17)
3.72

1.19 (0–2)
.60

1.41 (0–5)
1.30

aVerbal IQ (WAIS-III UK)
bPerformance IQ (WAIS-III UK)
cFull-scale IQ (WAIS-III UK)
fADOS total score − communication + reciprocal social interaction
hADOS—stereotyped behaviours and restricted interests. ADOS scores are presented with range in brackets

### Materials and Procedure

Participants were tested individually and testing took about 1.5 h. The order of tasks was counterbalanced across participants, with the ASD and TD members of each matched pair receiving the same order. Visual short-term memory and mental rotation tests were either given before or after the Water Maze task, which was counterbalanced across participants.

A computerized version of the Morris Water Maze written in Microsoft Visual Basic 6 was used to measure spatial navigation. This was an adaptation of Feigenbaum and Morris’ task ( 2004). The task was presented on a 19″ touch-sensitive screen ( http://​www.​elotouch.​com/​Products/​LCDs/​1939L/​), which was placed horizontally on a table located in a soundproof room. The table was surrounded by a small area, which was separated from the rest of the room by beige curtains placed on the ceiling and hanging from the ceiling forming a little cubicle to reduce the influence of external distracters or cues such as windows or features on the walls within the room to guide navigation. During the task, participants were asked to stand around the table looking down on the screen and to place their finger on it. There was sufficient space in the cubicle to allow participants to walk around the horizontal screen according to the task instructions. During task performance, room lights were turned off so that the only visible light came from the screen. This further reduced the influence of features in the room. On each trial, participants were presented with a virtual swimming pool environment. The display included a blue circle area representing the water in the pool, which was surrounded by an orange wall representing the wall around the pool. Outside the pool, a green area was presented representing grass growing around the pool area. On the grass, four objects were displayed in each corner of the screen, namely a chair, a life ring, a towel and a beach ball (see Fig.  1). The starting point, the location where participants were supposed to put their finger before moving it in the pool area, was indicated by a red dot on the orange wall around the swimming pool. On every trial, participants were asked to move their finger across the blue pool area until the platform appeared, which was represented by a brown box.
Participants were told that their task was to work out and learn the shortest way towards a hidden platform in the pool over several trials starting at the starting point and moving their finger across the blue space without lifting it from the screen and without crossing the orange perimeter line until the hidden platform appeared. The position of the participant’s finger was recorded during the entire experiment. After three practice trials during which participants learned how to use the touch-sensitive screen, participants were first presented with the place learning block and then with the 2 allocentric and the 2 egocentric blocks. Blocks were presented in counterbalanced order across participants with two individuals of a matched pair (one ASD and one TD person with similar IQ) performing the same order.
The starting point changed position on the orange wall in a random order. Each block consisted of 16 trials each lasting 60 s. If a participant could not find the platform within 60 s in one trial, a time out message together with the platform appeared on the screen and the participant was asked to move their finger to the platform. Each trial was followed by a distracter task, which consisted of a series of 12 blue circles (‘bubbles’) that appeared one after the other at random locations on a black screen and the participant was asked to ‘burst’ the bubbles by touching them. Measures taken were path length, time to find the target, path angle and percentage of time in the target quadrant. The target quadrant was defined as the quadrant where the participants’ path entered the platform. Control measures taken were: out of pool times (number of times participants left the pool area), finger lift times (number of times participants lifted the finger from the screen), time out times (number of times participants did not reach the target within 60 s), time before first movement (time from touching the starting point until the first movement in the pool area), and the duration of the distracter task (time for distracter task).
Place learning served as a control condition without any systematic manipulations taking place. The platform, the object cues and the participant stayed in the same location across the 16 trials of the task.
In the allocentric 1 condition (original), the platform and the object cues were presented in the same locations across the 16 trials but the participant had to move to another side of the table after every trial in a fixed randomized order. The direction of movement was indicated by an arrow on the screen and the phase ‘Please move to this side’. The movement could be 90 (in both directions) or 180° and it was used to disrupt egocentric processing.
The egocentric 1 condition (original) was designed to disturb allocentric processing in that the platform and the participant stayed in the same position across the 16 trials, but the object cues rotated in a fixed randomized order after every trial. Again, the movement could be 90° (in both directions) or 180°. The participant, however, did not see the objects rotate; the objects were only presented in their new rotated order to the participant.
In the allocentric 2 condition, participants stayed in the same position but the platform moved with the objects so that the relations among platform and objects stayed the same. This was to disrupt egocentric processing. The movement was 90° (in both directions) or 180°.
In egocentric 2, allocentric processing was disrupted in that the objects stayed in their same positions and did not relate to the platform position. The platform moved with the participants’ position, so that the relation between platform and participants position stayed the same.

#### Visual Short-Term Memory and Mental Rotation

To measure participants’ visual short-term memory, a version of the Brooks matrix task (Brook 1967) was used. Participants were presented with sentences describing spatial relations of numbers in a grid. The grid was a 4 × 4 matrix with 16 squares. The sentences were formulated as prompts describing in which cells of the grid which numbers were to be put to encourage mental imagery of the numbers in the grid cells. Participants were asked to repeat the sets of sentences verbatim (e.g. ‘In the starting square put a 1. In the next square up put a 2.’). The task started with sets of two sentences describing the positions of two numbers in the grid. The last sets included eight sentences. Participants were given three trials at every level and the task stopped if they had got three trials wrong at one level or completed all eight levels. Dependent variables were the maximum level achieved (2–8), the number of correct trials (up to 24) and the number of trials attempted (up to 24).

#### Mental Rotation

In the Mental Rotations task, (version A from Peters et al. 1995, paper and pencil test) participants were presented with 3D objects made from ten blocks, presented from different angles and their task was to pick two out of four figures that matched a target figure. Dependent measures were the sum of credits (1 credit for 2 correct stimulus figures for an item, no credit was given if the participant chose one incorrect figure that did not match the target figure) and the number of trials attempted (out of 24).

## Results

The data were analysed using Chi-Squared test for nominal data, t-tests, repeated measures ANOVAs, point biserial and bivariate correlations. If the Sphericity assumption was violated, Greenhouse-Geisser correction (GG) was applied. In the case of significant differences, Bonferroni-corrected post hoc tests were conducted. The significance level was set at .05 for all tests.

All data are presented in Table  2. In addition, to show learning over trials for the different conditions, Figure S1 in the supplementary materials presents heat maps showing a comparison of the paths for each group between the first and the last trial of every condition. All analyses presented here focus on the percentage of time spent in the target quadrant as this is seen as the most suitable measure for this kind of analysis. The data on all other measures including control measures for allocentric and egocentric conditions are presented in the supplementary materials and Tables S1–S3.
Table 2
Percentage of time in target quadrant for place learning, allocentric 1, egocentric 1 (original conditions), allocentric 2 and egocentric 2 (added conditions) for individuals with autism spectrum disorder (ASD) and typical development (TD)
Measure
Condition
ASD
TD
Total
M (range)
SD
M (range)
SD
M (range)
SD
Percentage of time in target quadrant
Place learning
44.28 (0–100)
34.31
48.51 (0–100)
35.51
46.39 (0–100)
34.96
Percentage of time in target quadrant
28.87 (0–100)
29.74
34.46 (0–100)
32.17
31.67 (0–100)
31.09
37.57 (1–100)
32.79
36.93 (0–100)
31.32
37.25 (0–100)
32.05
Total
33.22 (0–100)
31.59
35.70 (0–100)
31.75
34.46 (0–100)
31.68
Percentage of time in target quadrant
Allocentric 2
35.75 (0–100)
28.88
36.09 (0–100)
29.29
35.92 (0–100)
29.07
Egocentric 2
27.96 (1–100)
26.58
29.86 (0–100)
28.12
28.91 (0–100)
27.36
Total
31.86 (0–100)
28.01
32.98 (0–100)
28.86
32.42 (0–100)
28.43

### Place Learning

Place learning was used to ensure that participants in both groups were able to use the equipment properly and that they show learning over trials. The data were analysed using a 2 (Group [ASD, TD]) × 16 (Trial [1–16]) repeated measures ANOVA, which showed a significant main effect of Trial, F(8.55,427.52) = 40.30, p < .0001, η p 2 = .45, GG. No Group main or Group × Trial interaction effects were significant, F max < 1.66, p min > .20, η p 2 max < .04, confirming similar learning across trials for both groups. Similar results were found when analysing the data for the three other measures (path length, time to target, path angle, see Table S1).

### Allocentric 1 vs. Egocentric 1

The data were analysed using a 2 (Group [ASD, TD]) × 16 (Trial [1–16]) × 2 (Condition [allocentric, egocentric]) repeated measures ANOVA. Next to a significant main effect of Trial, F(8.92,446.23) = 88.31, p < .0001, η p 2 = .64, GG, as well as a significant Trial × Condition interaction, F(8.05,402.54) = 74.70, p < .0001, η p 2 = .60, there was also a significant main effect of Condition, F(1,50) = 13.18, p < .001, η p 2 = .21, showing that percentage of time spent in the target quadrant increased and was higher for the egocentric compared to the allocentric condition for most trials. A significant Group × Condition interaction, F(1,50) = 4.10, p < .05, η p 2 = .08, showed that the ASD group spent a shorter percentage of time in the target quadrant compared to the TD group in the allocentric ( p < .05, Cohen’s d = .59) but not the egocentric condition ( p = .82, Cohen’s d = .06). There was no significant main effect of Group, F(1,50) = 1.20, p = .28, η p 2 = .02,

### Allocentric 2 vs. Egocentric 2

The data were analysed using a 2 (Group [ASD, TD]) × 16 (Trial [1–16]) × 2 (Condition [allocentric 2, egocentric 2]) repeated measures ANOVA. There was a significant main effect of Condition, F(1,50) = 28.54, p = .00, η p 2 = .36, with a higher percentage of time spent in the target quadrant for the allocentric 2 compared to the egocentric 2 condition. However, this was the case for both groups as there was no main effect of Group, F(1,50) = .28, p = .60, η p 2 = .01, or Group × Condition interaction, F(1,50) = .36, p = .55, η p 2 = .01.
Because of the slight difference in the number of men and women in each group who participated in the task and previous reports of TD women performing worse than men at spatial navigation (Astur et al. 2004), we investigated how gender might have affected the results of the current study. Point biserial correlation analyses showed that there were significant positive relations between gender and performance on the allocentric 1 (r = .32, p = .02) and the allocentric 2 (r = .37, p = .008) conditions indicating that women performed better than men in both allocentric conditions. There were no significant relations between gender and either of the egocentric conditions (r max < .14, p min > .34) making it unlikely that the slight difference in gender between groups may have hindered the detection of a between-group difference in performance on the egocentric conditions. In addition, including gender as a covariate in an ANCOVA repeating the analyses reported above showed that the direction of results stayed the same.

### Visual Short-Term Memory and Mental Rotation

The data are presented in Table  3, and they were analysed using independent samples t-tests. There were no significant differences in any of the measures for any of the tasks.
Table 3
Measures for visual short-term memory and mental rotation for participants with autism spectrum disorder (ASD) and typical development (TD)
ASD M ( SD)
TD M ( SD
t( df)
p
Cohen’s d
Total correct (out of 24)
12.92 (4.87)
12.04 (5.30)
.61 (48)
.54
.17
Maximum level (out of 8)
6.12 (1.62)
6.08 (1.96)
.08 (48)
.94
.02
Trials attempted (out of 24)
17.52 (3.94)
17.28 (5.00)
.19 (48)
.85
.05
Average accuracy
.91 (.14)
.83 (.16)
1.89 (46)
.07
.55
Average RT in ms
2964.46 (1215.11)
3195.15 (1144.24)
.68 (46)
.50
.20
Mental rotations test
Sum of credits (out of 24)
8.08 (6.09)
7.04 (4.51)
.70 (50)
.49
.19
Trials attempted (out of 24)
15.23 (5.70)
14.42 (5.11)
.54 (50)
.59
.15
aOnly 25 TD and 25 ASD individuals completed this task
bOne ASD individual did not complete this task. A further 2 ASD and 1 TD participants were excluded because they were at chance in discriminating between right and left in the control task. For both tasks, the remaining participants in both groups were still matched on VIQ, PIQ, FIQ, age ( t max = -.81, p max = .42, Cohen’s d max = .23) and gender ( X 2 max = 2.93, p max = .09).

Finally, we investigated correlations among allocentric navigation performance and performance on memory and mental rotation tasks. Since there were no significant correlations among any of the measures for either group (see Table  4), it seems unlikely that performance on the visual short-term memory and mental rotations tasks may have influenced performance on the allocentric navigation task.
Table 4
Bivariate correlations among spatial navigation performance in the allocentric condition and performance on visual short-term memory (Brooks matrix task) and mental rotation tests (Manikin, Mental rotations task) for groups of individuals with autism spectrum disorder (ASD) and typical development (TD) and for both groups in total

Allo ASD
Allo TD
Allo Total
− .10
.33
.13
− .22
.21
.06
− .22
.17
.03
.36 +
.19
.19
− .25
− .25
− .20
Mental rotations test—sum of credits
.11
.20
.11
Mental rotations test—trials attempted
.26
− .21
− .04
+ p < .10

## Discussion

The primary aim of this study was to explore if ASD individuals show spatial navigation difficulties particularly in allocentric spatial navigation. Such a deficit would be consistent with the relational binding account of autistic memory (Bowler et al. 2011). To test this, we compared matched groups of adults with and without ASD on navigation conditions that either required egocentric or allocentric processing (Bohbot et al. 2004) using a human virtual reality adaptation of the Morris Water Maze task. We predicted to find particular difficulties in ASD with forming view-point independent, allocentric representations. As control tasks visual short-term memory and mental rotation tasks were used to measure participants’ ability to process and manipulate spatial information. We did not expect difficulties in ASD on these tasks.
Our prediction was confirmed for the two original test conditions. Only for the allocentric condition ASD individuals spent less time in the target quadrant compared to TD individuals. This finding is supported by a number of other observations. First, there were no differences between groups in place learning (the baseline condition). Second, there were no differences between groups in participants’ ability to follow instructions (out of pool, finger lift times) and speed of learning (time out times). There was also no difference between groups in how long they took to complete the distracter task between blocks (bursting the bubbles), suggesting that both groups experienced the same time interval between tasks. Finally, consistent with our expectations no between-group differences were found for the control tasks of visual short-term memory and mental rotation. The absence of any significant correlations among allocentric navigation and visual short-term memory and mental rotation performance makes it unlikely that these abilities had any influence on the significant between-group difference in allocentric spatial navigation performance.

## Acknowledgments

We would like to thank all the participants, who have taken part in this research. The first author was supported by an Erasmus scholarship as well as a travel grant from “Die Gesellschaft der Freunde und Förderer der Technischen Universität Dresden” to work on this study.

### Author Contributions

MR, SBG and DMB conceived of the study, implemented the design and contributed to the analysis and interpretation of the results. MA offered critical comments on the design of the study and contributed to the analysis and interpretation of the results. PB programmed the task and contributed to the analysis and creation of heat maps. MR and SBG were responsible for data collection. MR drafted the manuscript and MA, SBG and DMB provided critical comments for revisions. All authors read and approved the final manuscript.

## Compliance with Ethical Standards

### Conflict of interest

The authors declare that they have no conflict of interest.

### Ethical Approval

All experimental procedures outlined involving human participants were in accordance with the ethical guidelines set out by the British Psychological Society, they were approved by City, University of London’s Research Ethics Committee and they adhered to the 1964 Helsinki declaration and its later amendments.

### Informed Consent

Informed consent was obtained prior to participation from all individuals included in this study.

## Onze productaanbevelingen

### BSL Psychologie Totaal

Met BSL Psychologie Totaal blijft u als professional steeds op de hoogte van de nieuwste ontwikkelingen binnen uw vak. Met het online abonnement heeft u toegang tot een groot aantal boeken, protocollen, vaktijdschriften en e-learnings op het gebied van psychologie en psychiatrie. Zo kunt u op uw gemak en wanneer het u het beste uitkomt verdiepen in uw vakgebied.

Extra materiaal
Literatuur
Over dit artikel

Naar de uitgave