Cognitive Profile of Children with Autistic Traits

Introduction

The distinctive cognitive profile of Autism Spectrum Disorder (ASD) has led to a proliferation of neuropsychological theories to elucidate the brain-behaviour association. Three main theories – Theory Mind, Central Coherence and Executive Function Skills have dominated the research scene.¹ While the primacy of these theories has been debated, what is known today is that each has some influence on the other and contribute to the development of ASD. Belmonte refers to ASD as a prime test case for understanding neuropsychology in the general population.² This underlines the importance of establishing the cognitive profile in ASD.

The synergy in any social interaction is contingent on the individuals’ ability to predicate on the behaviours of social partners.³ This ability, referred to as Theory of Mind (ToM) is a complex process involving multiple facets such as perceptually recognizing the direction of eye gaze, interpreting emotions, and understanding it. Consequently, the assessment measures that have evolved from this have also been similarly varied. Explicit measures such as Reading the Mind in the Eye assess an individual’s understanding of emotions in the eyes, while false belief tasks are more cognitive in nature.^3,4 More recently, implicit measurements of ToM have been carried out through eye movements.⁵

Central coherence (CC) or the ability to integrate individual components into a coherent whole can be related to the perceptual area or language ability. This theory is said to account for the decreased ability in global processing and enhanced local processing in individuals with autism as they are said to have weak central coherence.⁶ As with ToM, varied measures have been used to assess CC depending on the conceptualization. It has been studied perceptually and in the form of language processing.⁷

The cognitive control of behaviour is attained through a gamut of skills - working memory, impulse control, planning, sequencing, organization, and execution. These skills referred to as ‘Executive Function’ continue to develop throughout childhood. Individuals with disorders such as Autism and Attention Deficit Hyperactive Disorder typically have impairments in Executive Function (EF) skills.⁸ A range of tasks have been used to study the different domains of EF.^7,9

These three theories have been studied extensively both in ASD and in the Broader Autism Phenotype (BAP). Individuals with subclinical autistic traits not meeting sufficient criteria for a diagnosis are referred to as having the BAP. Assessing these individuals is often easier than those within the spectrum. Apart from family members of autistic individuals, individuals with high levels of autistic traits in the general population provide a wider pool for research.^9,10

Specific measures to assess levels of autistic traits have been developed over the years.^11,12

These measures are used in studies attempting to define the neuropsychological profile of those with broader autism phenotype. However, much of the research has been inconclusive. While Delorme et al., found executive function difficulties among parents of individuals with ASD, Losh et al., found no such difficulty.^7,13 Similar contradictions can be seen in Theory of Mind research.^14,15 Happe´ and others found fathers of children with autism having weak central coherence, but this was not evident among mothers or siblings.¹⁶

Studies examining the cognitive theories among individuals with BAP in the general population are limited and these have generally been conducted on adults.

Kunihira et al.⁹ conducted a large scale study (1364 participants) by administering the ‘Autism Quotient’ among college students and comparing those scoring high with those scoring low on the cognitive theories. They found no relationship between the Autism Quotient and theory of mind, executive functions, and central coherence.

More recently, DiCriscio and Troiani confirmed the predisposition to local bias among a sample of children with high levels of autistic traits.¹⁷ Another study found both theory of mind and executive functioning difficulties to be associated with higher levels of autistic traits among adolescents and adults.¹⁸ Theory of Mind has been studied through various means and has different aspects including affective and cognitive aspects. One study attempted to investigate the association between the affective and cognitive components and its relation to autistic traits. They used the ‘Reading the Mind in the Eyes’ test (affective) and ‘Unexpected Outcomes Test’ (cognitive) along with the ‘Broader Autism Phenotype Questionnaire’. Though the study found both the components of Theory of Mind to be correlated, both the broader autism phenotype and non-broader autism phenotype groups performed similarly on the Reading the Mind in the Eyes test.¹⁹

The exact profile among BAP is still clearly under debate. An understanding of the neuropsychological profile is essential not just for diagnosis, but also for designing more focussed therapy strategies. To the best of our knowledge, determining the profile among a large sample of children with high levels of traits in the general population has not been previously attempted.

Materials and Methods

This cross-sectional, analytical study was the result of a doctoral thesis and was carried out in mainstream schools. It was a community-based study carried out in two main phases. In the first phase, parents were administered the Social Responsiveness Scale, with the help of the school administration.¹¹ The highest and the lowest 100 scores were taken as those with High Autistic Traits (SRS High) and Low Autistic Traits (SRS Low), respectively. In the second phase, the children who were categorized to High and Low groups were administered a battery of neuropsychological tests at the school premises.

Sample

Using the purposive sampling method, mainstream schools catering to different economic strata were chosen from different parts of Urban Bangalore. All were private schools affiliated to the major school boards - CBSE, ICSE and SSLC. This was done to ensure a sample representative of the general community. Five schools consented to be part of the study. Parents were approached through the school administration and at no point of time were they directly available to meet the researcher. All the children from the lower economic strata were from a school run by a non-governmental organization for children from the slums.

Inclusion criteria

All children of consenting parents from the selected schools between the age of 5 and 11 years.

Exclusion criteria

None

Phase 1: Parents of 771 children were administered the Social Responsiveness Scale (SRS). Since the children were from mainstream schools, it was assumed that they were without any diagnosed developmental disorders.

Phase 2: Two hundred children out of the above 771 were administered a battery of neuropsychological tests. Tests were administered to only those children who were willing to participate.

Post-hoc analysis

Exploratory factor analysis was carried out on the SRS scores.²⁰ This was done by Principal Axis Factoring and subjected to Direct Oblimin rotation to extract the factors. The factor analysis yielded a two-factor solution. The factor analysis was further validated by repeating it with the sample split in half and sample split by gender. There were 28 common items across the repeated factor analyses in the first factor. These 28 items mapped onto all the domains of the SRS. The SRS scores were then recalculated for only these 28 items (SRS28) and again the highest (n=17) and lowest (n=17) scores were chosen. The neuropsychological profile of these children was then compared. These children were from among the existing 200 who had already been assessed.

Measures

The tests administered included

1. Social Responsiveness Scale (SRS):¹¹ This is a 65- item parent questionnaire measuring autistic traits in children aged 4 to 18 years. The questionnaire was translated into Kannada to reach a wider sample. The questionnaire has been found to be reliable for use among Indian sample.²¹
2. Behaviour Rating Inventory of Executive Functions (BRIEF):²² The BRIEF is a parent questionnaire consisting of 86 items for assessing a wide range of executive functioning behaviours in children and adolescents aged 5-18 years. This had not been previously used in Indian setting, but has been tried in a pilot sample as part of this study.
3. Hooper Visual Organization Test (HVOT):²³ This brief screening test measures the individual's ability to organize visual stimuli, a task that is particularly sensitive to neurological impairment. The individual is shown pictures of common objects that have been cut up into parts. The examinee is required to mentally piece together the parts and identify the whole. It can be used for children aged five years and upwards and takes about 15 minutes to administer. This was used to measure central coherence. 
4. Reading the Mind in the Eyes Test (RMET):⁴ Children are shown a pair of eyes, and they must select the appropriate emotion from a choice of four words. This has been developed by the Autism Research Centre, Cambridge. There are no norms available for this test. The child version, which has simple words to describe emotions was used. There are a total of 28 pairs of eyes in this test. This was used to assess the Theory of Mind.

Statistical Analysis

The distribution of SRS data was checked for normalcy through the test of normality using both Kolmogorov-Smirnov and Shapiro-Wilk tests. Mann-Whitney U test (non-parametric test) was used to compare means in the three areas of neuropsychology.

Results

SRS scores were distributed continuously in the general population.²⁴ The demographic spread of the scores is tabulated in Table 1. The SRS scores for the entire sample ranged between 1 to 119, with a standard deviation of 19.717.

The distribution of scores was examined by age and gender. The distribution by age was not significantly different; however there was a difference when examined in terms of gender. Boys exhibited lower mean scores than girls.

The SRS scores were also analysed by socioeconomic status (SES) and the lower economic group had statistically significant lower scores.

In phase 2 of the study, neuropsychological tests were administered for the children. The children falling within the SRS High group were compared with those falling within the SRS Low group.

The children with high levels of autistic traits had significantly poorer executive function skills than those with low levels of autistic traits. However, the performance on Central Coherence and Theory of Mind tests was not significantly different between the two groups.

Post Hoc Analysis

Children with high levels of autistic traits continued to have poorer executive function skills even when the analysis was carried out with SRS28. While there was a change in the median scores and range of scores of ToM, this was not significant. However in the area of Central Coherence, children with high levels of autistic traits were significantly poorer than those with low levels of traits.

Since the SRS scores were significantly affected by the SES status of the child with the low economic group scoring lower scores, the neuropsychological profile was again examined after removing the lower SES group (high autistic traits group (n=98) and low traits (n=53)). The results were found to be statistically significant in executive function skills and central coherence. While the results remained insignificant in theory of mind, the P value did however move closer to significance.

Discussion

The distribution of SRS scores demonstrated that girls had higher scores than boys. This could be due to the use of raw scores instead of T scores. According to the SRS manual, girls have a higher range of raw scores for corresponding T scores among boys.¹¹

The extent of contribution and interaction of each theory towards development of autism has never been completely established. It is evident that Executive Function skills impact other areas of development including language development and adaptive behaviours.^25,26 It has also been found to impact the development of Theory of Mind among children with autism.²⁷ Central Coherence in turn appears to impact the development of Theory of Mind.²⁸ One critical aspect of Theory of Mind is the ability to put together information coming from multiple sources – tone of voice, facial expression and context, in order to predicate on a person’s behaviour during a social interaction. All these skills are dependent on the individuals’ Central Coherence ability. Theory of Mind deficits in ASD form the crux of the social deficits, but this relies on Executive

Function and Central Coherence making both these deficits more central to the development of disorder. The relationship between Executive Function and Central Coherence, however, appears to be bidirectional with some aspects of each impacting the other.¹

The support for Executive Function deficits being an integral part of autistic traits remained robust through both our initial analysis and the post hoc analysis. It is however, imperative to note the impact of lower SES on it. The SES of children affects development, particularly cognitive development with home environment having a strong role to play. This impact remains through childhood into adolescence and has undeniable effects on behaviour.²⁹ The post hoc results after removal of lower SES group clearly points to the strength of association between autistic traits and EF deficits. The lower SES group had been removed as they appeared to have lower levels of autistic traits and could have confounded the data.

Children with ADHD also have Executive Function difficulties predominantly in the domains of vigilance and response inhibition.³⁰ But amongst those with ASD, the deficits are generally found to be pervading through all domains.³¹ It would be important to explore the executive function profiles in those who demonstrate high autistic traits to determine the origin of these deficits to ensure that the deficits do not reflect children with ADHD or other disorders.

Parent questionnaires evaluating Executive Function have been found to be valid tools.³² The fact that the parents who evaluated their children’s social responsiveness critically could be evaluating the EF skills in a similar critical manner cannot be ruled out.

Children with high levels of autistic traits showed significant difference in their performance on the Central Coherence test during the post hoc analysis. This was when autistic traits were stronger after factor analysis and in the sample without the lower SES group.

The test utilized estimates the perceptual aspect of Central Coherence and is more intrinsic in nature. It is not affected by schooling or age of the child. This association between Central Coherence and autistic traits is further evinced by fact that the two other areas it has been researched in are both associated with ASD - Anorexia & Savant Syndrome, demonstrating the likelihood of Central Coherence being more closely linked to autistic traits.^33,34 Central Coherence becomes more evident in individuals as autistic traits aggregate.

One previously observed key characteristic of Central Coherence is that unlike Theory of Mind and EF which develop over time, it appears to be constant and does not change with time.³¹

This constancy of its nature could make Central Coherence more central to the development of autistic traits and autism. While the presence of some EF deficits in a school sample can be attributed to other difficulties like ADHD, it appears to be additive effects of Central Coherence difficulties which lead to the rise of autistic traits.

Theory of Mind deficits have become congruent with the concept of ASD over the years. They are certainly well established in children with autism. Results from studies among the broader autism phenotype have not been similarly consistent. While family studies have lent support to the theory, the same has not been found among broader autism phenotype in general population.^7,9 On the other hand, imaging studies have found evidence for differences in the social brain network of individuals with high levels of autistic traits.³⁵ This appears to suggest that while direct measures such as Reading the Mind in the Eyes have not yielded results, the more implicit measures of the social brain appear to indicate differences. The explanation for this could lie in the drive theories.

The Social Motivation Theory proposed by Chevallier et al., suggests that social cognitive deficits such as Theory of Mind are a consequence of disruption in social orientation or social motivation.³⁶ While ToM has a role to play in social cognition, it may not be a primary driver of the syndrome manifestation. It is possible that ToM could be an intermediate cognitive phenotype rather than a primary deficit.

Another challenge in the assessment of ToM appears to be in the wide range of measures available, with each measure focussing on a different aspect of the concept. Camodeca studied both the cognitive aspect and the affective aspect of ToM by using the Unexpected Outcomes Test (for the cognitive aspect) and RMET (for the affective aspect).¹⁹ While both were found to be correlated, they did not directly affect scores on a measure of autistic traits. This could indicate difficulties in ToM being more subtle among the BAP derived from the general population.

One of the few studies which found positive results for all three theories amongst individuals with autistic traits was the study by Best et al.³⁷ However, they derived their subjects from a pool of individuals with different learning challenges and their results could be an artefact of the subjects’ varied psychiatric challenges rather than autistic traits.

While the ToM did not receive support in this study, there is robust evidence for the presence of executive function difficulties in children with high levels of autistic traits. Executive Function has a strong role to play in several areas of development. It has been linked to language skills in both children with autism and typically developing children.²⁵ It has also been linked to adaptive behaviour and the development of the theory of mind.^26,27 It clearly plays a strong role in the development of various disorders including autism.

While differences in weak central coherence were not significant in the SRS full sample, it appears to have a stronger role to play as autistic traits grow stronger, indicating its importance in the development of the disorder. One study carried among typically developing preschoolers found a link between central coherence and executive function skills at the perceptual level, but this link appeared to disappear when higher order tasks such as language were concerned.²⁸ Better CC in the early years has also been associated with better ToM.²⁷ This could indicate the importance of CC in development and needs to be explored further to come to a clearer understanding of the role it plays.

Conclusion

The results of this study advance the understanding of conceptualization of neuropsychological profile of ASD. Executive Function skills remained strongly associated with autistic traits even when the lower SES group was removed from the sample. The influence of Central Coherence is apparent in the fact that it is significantly stronger when autistic traits get stronger. Theory of Mind, though showed some change, was not significant and could be a secondary, more downstream effect of Central Coherence and Executive Function skills.

Conflict of Interest

None

Financial Support

None