beta

You're viewing our new website - find out more

Publication - Research Publication

The microsegmentation of the autism spectrum: research project

Published: 26 Mar 2018
Part of:
Health and social care, Research
ISBN:
9781788516167

Economic research on autism and implications for Scotland, including how the economic cost of autism can inform strategy and planning.

357 page PDF

4.7MB

357 page PDF

4.7MB

Contents
The microsegmentation of the autism spectrum: research project
7 The Scottish Autism Survey

357 page PDF

4.7MB

7 The Scottish Autism Survey

Method

7.1 The Scottish Autism Survey was developed jointly by the full research team at the University of Strathclyde and the London School of Economics. The final draft of the survey was piloted with a small group of professionals and service-users who had scrutinised earlier drafts and the final version reflected the modifications they proposed. A copy of the full survey will be found as an Annex.

7.2 The survey comprised 32 questions divided into three sections and was designed to be completed by individuals with ASD, by a carer of an individual with ASD or by a professional completing it on behalf of the individual. The first section asked the respondent to state the capacity in which they were responding and to provide basic demographic information about the individual with ASD, as well as details relating to their diagnosis. The second section of the survey focussed on the day-to-day lives of the individual with ASD, including their educational, employment and residential status and their use of support services. The survey concluded with a final section aimed exclusively at parents and carers, which sought to assess the impact of ASD on their lives and the lives of their families.

Sample recruitment

7.3 The survey was hosted on the online data collection platform ‘ Qualtrics’, and a link to this online version was promoted on the websites and social media pages of ASD support groups including Scottish Autism, Autism Network Scotland and the National Autistic Society ( NAS). These groups in turn circulated information about the survey and the link to appropriate groups and individuals. In addition, the researchers used their own additional networks to contact mailing lists of individuals with ASD and their carers by post.

7.4 A total of 1,580 individuals logged-on to the online version of the survey and an additional 24 individuals requested a paper version which was sent and returned by post. Full details of the recruitment process are shown in the flow chart in Figure 7.1.

7.5 Questions 1-3, 6 and 7 of the survey related to key independent and dependent variables for the final statistical analyses. Any respondents not completing these questions (n = 612, 38% of individuals who logged-on) were excluded from the final analysis. One individual was exempt from this exclusion criterion because although they had not provided sex data, they had provided responses to all other key questions, and detailed responses for many other questions in the survey.

7.6 A series of final checks for duplicate responses were carried out on the remaining responses (n = 992) as part of data screening to ensure that no two responses related to a single individual. Details of this procedure may be found in Appendix C.1. The checks identified 42 duplicate responses (3% of the remaining responses). In each case, the most complete response was retained, and the duplicate removed from further analysis. Final analysis was based upon responses relating to 950 individuals, and a full description of respondents and the individuals with ASD described in these responses is provided in Figure 7.1.

Figure 7.1 Overview of the sample selection process

 Figure 7.1 Overview of the sample selection process

Statistical analysis

7.7 Statistical analysis was carried out in two stages. The first focussed on collating and summarising the raw demographic, diagnostic and service-use data provided by respondents. This purpose of this stage of the analysis (reported below) was to characterise the sample included in the research, and to construct a detailed understanding of the lives of those with ASD living in Scotland.

7.8 The aim of the second stage of analysis was to identify and model the factors associated with education, employment, relationships, independent living, and mental-health outcomes using binary logistic regression with dependent variables which related to life outcomes, and independent variables which related to demographic, diagnostic, or service-use data. These analyses all met an event per variable ( EPV) [2] rate of 10.

7.9 The modelling approach used investigated the effects of different levels of predictors of child and adult outcomes including service use (cf. Morton & Frith, 1995). The following five levels were utilised, each entered as a block in hierarchical regression analyses: (i) demographics, (ii) diagnostics, (iii) co-occurring conditions, (iv) educational, health and social independent variables, and (v) support service variables. There were times when some of the regression analyses carried out could have included a large number of potentially relevant independent variables. However, on most occasions the inclusion of all potentially relevant predictors would have resulted in the EPV ratio falling below 10. To avoid this, a set of candidate variables was identified for each model using a method recommended by Bursac et al. (2008). The result of this systematic approach is that variables are only included as independent predictors of a dependent variable in a multi-factor model if they have previously been found to be significant at a level of p = .25 or less when included in a single variable model focusing on the same dependent variable. If at the end of this process the number of candidate variables still exceeded the intended EPV ratio, separate analyses of combinations of the effects of these candidate variables were conducted and the models which could account for the greatest amount of variance reported in the main body of this chapter.

7.10 All modelling testing logistic regression analyses were accompanied by residual and multicollinearity checks which have been reported alongside the main results. Such checks are carried out to confirm the validity of any models tested, and also to reduce the likelihood of type I and type II statistical errors through the removal of any data points which skew the overall results of an analysis. These checks included an analysis of Cook’s distances and studentised residuals and if a participant’s response was associated with Cook’s distances greater than 1 or studentised residuals greater than 2 then it was temporarily removed and the analysis was re-run. If this follow-up analysis showed an improvement of 2% in the accuracy of the classification table associated with a model then the original model was rejected, and the new model (minus the problematic cases as identified by the residual checks) was reported (in such cases original models have been included in the Appendices). If there was no such difference, then the results of the original analysis were reported.

7.11 For the analysis of categorical data a two-stage approach was employed. Firstly, Pearson’s Chi-Square Test was used to establish whether or not a relationship existed between variables, and secondly, if a relationship was found to exist, relative odds ratios were calculated (following the method proposed by Sharpe, 2015) to establish the magnitude and direction of these relationships.

Treatment of missing data

7.12 To address missing data in responses, multiple imputation by chained equation ( MICE) was carried out, an approach selected for its ability to account for the different types of variables included in our dataset (e.g. binary, ordinal and categorical). Binary variables were imputed using logistic regression analyses, while ordinal logistic regression and multinomial logistic regression were used when imputing ordinal and categorical variables.

7.13 This approach to multiple imputation creates new ‘blocks’ (i.e. alternative versions of the complete datasets), each of which represents a slightly different version of the original dataset in which the missing cases have been replaced with values informed by the data associated with other variables in the dataset (Sterne et al., 2009). Ultimately the results from each block are synthesised as part of a pooled analysis, and it is the results from these pooled analyses that are reported in the following chapter, in each case based on a dataset comprised of 20 imputation blocks.

Qualitative analysis

7.14 As part of the survey, participants were given the opportunity to provide more detail about any aspect of living with ASD, or caring for someone with ASD, which had not previously been addressed as part of a ‘free comments’ page at the end of the survey. Many of these comments included a good level of detail, and as a result an analysis of the comments has been included in this chapter.

7.15 Comments from individuals with ASD (N=9, 8% of the 114 of the individuals who responded to the survey as someone with an ASD) and the parent/carers of individuals with ASD (N=68, 10% of the 704 parents/carers who completed the survey as a whole) were analysed thematically using a ‘semantic’ approach (Braun & Clarke, 2006), whereby themes are identified using the explicit meanings of the text. Initial codes for the comments were generated by two of the authors and grouped into themes (second-level codes which show ‘patterns’ across the data codes) and constituent sub-themes (Miles & Huberman, 1994). Thematic networks were then constructed, again by the two authors, to illustrate the relationships between the themes and sub-themes (Miles & Huberman, 1994). We analysed the comments from the individuals with ASD and parents/carers separately to ensure that any distinctive comments from the two groups of respondents were captured. The comments and associated themes/sub-themes may be found in Tables 11.24 and 11.25.

Analysis of ASD Diagnostic Categories

7.16 Given the diversity of symptoms and behaviours that can affect individuals with ASD, one of the aims of this investigation was to focus on the differences in outcomes and service use across different types of ASD. However, a significant number of survey responses related to individuals whose diagnoses complicated this analysis.

7.17 The first group of these responses included individuals described (either by themselves or by their carers) as having non-specific ASD diagnoses (i.e. the severity of the autistic symptoms was not clear), while the second group related to individuals who had diagnoses of atypical autism or PDD-NOS (diagnoses which would have been given to those who met some but not all of the diagnostic criteria for autism or Asperger’s). Historically these diagnostic categories have been associated with a particularly broad and inconsistent range of behaviours and symptoms, which makes characterising these individuals, relative to some of the other individuals in the sample, a difficult task. As a result the team decided there was no benefit to analysing these individuals according to their diagnosis, but instead that those with these less precise diagnoses should be grouped as part of a composite category called ‘Other ASD’.

Results

Respondent Characteristics

7.18 Of the 950 responses included in the final analysis, 79% were provided by parents and family members who cared for an individual with ASD (n = 754), and an additional 4% of responses were provided by non-related carers (n = 33). A further 12% were provided by individuals with ASD themselves (n = 114), 4% were provided by professionals (n = 36), and the remaining 1% were provided by friends and volunteers who were close to an individual with ASD (n = 13). Table 7.1 summarises the respondent characteristics.

Table 7.1 Respondent characteristics

Respondent characteristics n (%)
Parents and family carers 754 (79)
Individuals with ASD 114 (12)
Non-related carers 33 (4)
Professionals 36 (4)
Others a 13 (1)
Total 950 (100)

a This category included close friends and volunteers who worked with people with ASD

Demographic Characteristics of the ASD sample

7.19 Sex data was available for all but one of the sample (n = 949) and an analysis of this data revealed that 77% of the sample (n = 735) were male and that the sex ratio for the sample was 3.4:1. This is compatible with the majority of findings in the ASD literature which show ASD to be more prevalent amongst males that females with sex ratios typically ranging from around 2.5 – 6.0: 1 (e.g. Baird et al., 2006; Chakrbarti & Fombonne 2005; Idring et al., 2012; Kocovska et al., 2012; Nygren, 2012).

Table 7.2 Age of ASD individuals (n = 950)

Age (years) 0 – 10 11 – 18 19 – 49 ≥ 50
n a 335 299 280 36
% 35 32 29 3

7.20 As shown in Table 7.2, the majority of individuals with ASD were children or young adults with 35% of the sample under the age of 10 (n = 335), and 32% aged between 11 and 18 (n = 634). Of the remaining sample, 29% were 49 and under (n = 280), and 3% were over the age of 50 (n = 36).

7.21 In terms of ethnicity, 97% of the individuals with ASD were described as white (n = 917). Of the remaining 3%, 10 individuals were described as Asian, Asian Scottish or Asian British, nine were described as being mixed race or from multiple ethnic groups, four were described as African, one was described as Caribbean and nine were described as being of ‘other’ ethnicity including Arab, Jewish, Turkish, Taiwanese, and White Chinese. These data are compared with data from the 2011 Scottish Census (National Records of Scotland, 2011) in Table 7.3.

Table 7.3 Comparison of the ethnicity of respondents in the Scottish Autism Survey sample with data from the 2011 Scottish Census

Ethnicity Scottish Autism Sample 2011 Scottish Census Data
n (%) n (%)
White 917 (97) 5,084,000 (96)
Mixed/multiple ethnic groups 9 (1) 20,000 (0)
Asian (including Asian Scottish/British) 10 (1) 141,000 (3)
African 4 (0) 30,000 (1)
Caribbean 1 (< 1) 7000 (0)
Other a 9 (1) 14,000 (0)
Total 950 (100) 5,296,000 (100)

a ‘Other’ ethnicities represented in the sample included Taiwanese, Jewish, Arabian and Turkish. Two participants preferred not to specify their ethnicity.


7.22 Figure 7.2 shows the number and percentage of ASD individuals in the sample living in each of the Scottish local council areas. Amongst the areas best represented in the sample were Glasgow City (n = 103), the City of Edinburgh (n = 90), North Lanarkshire (n = 86), South Lanarkshire (n = 60), and Fife (n = 56). The least represented areas in the sample included East Ayrshire (n = 11), North Ayrshire (n =11), Midlothian (n = 9) and Clackmannanshire (n = 8). These data are compared with the 2013 Scottish Census (National Records of Scotland, 2011) in Table 7.4 which takes into account the population of each of the areas and reveals close mapping (+/- 1 standard deviation, equivalent +/- 2%) in 24 of the 32 local areas (the remaining eight areas are italicised in Table 7.4).

Figure 7.2 Geographic location (determined by post code) of responses included in the final sample

Figure 7.2 Geographic location (determined by post code) of responses included in the final sample
Key Local Authority Responses Returned (% of total sample) Key Local Authority Responses Returned (% of total sample)
1 Fife 56 (6) 10 Inverclyde 12 (1)
2 Clackmannanshire 8 (1) 11 Renfrewshire 12 (1)
3 West Dunbartonshire 22 (2) 12 Glasgow City 103 (11)
4 East Dunbartonshire 21 (2) 13 Midlothian 9 (1)
5 North Lanarkshire 86 (9) 14 North Ayrshire 11 (1)
6 Falkirk 22 (2) 15 East Renfrewshire 14 (2)
7 West Lothian 33 (4) 16 South Ayrshire 15 (2)
8 City of Edinburgh 90 (10) 17 East Ayrshire 11 (1)
9 East Lothian 20 (2) 18 South Lanarkshire 60 (6)

Table 7.4 Comparison of number of responses relating to ASD individuals in each council area to the total population of each council area

Local government region Scottish Autism sample responses Total population (from 2011 census data)
n (%) n (%)
Aberdeen City 3 (< 1) 227,130 (4)
Aberdeenshire 43 (5) 257,740 (5)
Angus 7 (1) 116,240 (2)
Argyll & Bute 31 (3) 88,050 (2)
Clackmannanshire 8 (1) 51,280 (1)
Dumfries & Galloway 13 (1) 150,270 (3)
Dundee City 21 (2) 148,170 (3)
East Ayrshire 16 (2) 122,440 (2)
East Dunbartonshire 21 (2) 105,860 (2)
East Lothian 20 (2) 101,360 (2)
East Renfrewshire 14 (2) 91,500 (2)
City of Edinburgh 90 (10) 487,500 (9)
Falkirk 22 (2) 27,400 (1)
Fife 56 (6) 157,140 (3)
Glasgow City 103 (11) 366,910 (7)
Highland 77 (8) 596,550 (11)
Inverclyde 12 (1) 232,950 (4)
Midlothian 9 (1) 80,310 (2)
Moray 9 (1) 84,700 (2)
North Ayrshire 11 (1) 94,350 (2)
North Lanarkshire 86 (10) 136,920 (3)
Perth and Kinross 28 (3) 337,730 (6)
Renfrewshire 12 (1) 21,570 (< 1)
Scottish Borders 7 (1) 147,750 (3)
South Ayrshire 15 (2) 173,900 (3)
South Lanarkshire 60 (6) 113,870 (2)
Stirling 19 (2) 23,200 (< 1)
West Dunbartonshire 22 (2) 112,850 (2)
West Lothian 33 (4) 314,850 (6)
Na h-Eileanan an Iar 4 (< 1) 91,260 (2)
Orkney Islands 11 (1) 89,810 (2)
Shetland Islands 6 (1) 176,140 (3)
Total 889 (100) a 5,327,700 (100)

a Note: This was the total number of individuals for whom geographic location data was available

ASD Diagnoses

7.23 Table 7.5 shows the number and percentage of individuals in the sample with each type of ASD diagnosis. In total, 217 (23%) had a diagnosis of autism, 426 (45%) had a diagnosis of Asperger’s or HFA, and 307 (32%) had other ASD diagnoses including atypical autism/ PDD-NOS (n = 9) and non-specific ASD diagnosis (n = 298).

Table 7.5 Frequency of ASD Diagnosis

Diagnosis n %
Autism a 217 23
Asperger’s/ HFA b 426 45
Other ASD diagnoses c 307 32
Total 950 100

a Including ‘Childhood Autism’ or ‘Autistic Disorder’ ;
b Including ‘Asperger’s Disorder’; c Including general/non-specific ASD diagnoses, ‘Atypical Autism’ or ‘ PDD-NOS’.

7.24 Table 7.6 describes the sample according to their age and the type of diagnosis reported. The majority of those with autism were children, with 43% of this sub-sample under the age of 10 (n = 93), and a further 29% were young adults (n = 62). Of the remaining individuals with autism diagnoses 25% were aged between 19 and 49 (n = 54) and 4% were 50 or older (n = 8).

Table 7.6 ASD diagnosis by age.

Age Group (years) ASD Diagnosis n (%) Total Sample
Autism Asperger’s/ HFA Other ASD
0 – 10 93 (43) 89 (21) 53 (50) 335 (35)
11 – 18 62 (29) 136 (32) 101 (33) 299 (31)
19 – 49 54 (25) 174 (41) 52 (17) 280 (29)
≥ 50 8 (4) 27 (6) 1 (< 1) 36 (4)
Total 217 (100) 426 (100) 307 (100) 950 (100)

7.25 In comparison with the rest of the sample, significantly fewer individuals with Asperger’s/ HFA were under the age of 10, X 2 (1, 950) = 59.39, p < .001. This finding would appear to support previous research which has indicated Asperger’s/ HFA is associated with a later age of diagnosis (Howlin & Asgharian, 1999), something that needs to be taken into consideration in planning and providing for the future. By contrast, approximately half of those with other ASD (n = 307; n = 298 with a non-specific/general ASD diagnosis, n = 9 with a diagnosis of atypical autism or PDDNOS) were under the age of 10.

7.26 Finally, turning to age of diagnosis, it is noted that few individuals in the sample were over the age of 50. It is likely that the low number of individuals within this age bracket reflects the fact that autism entered the diagnostic classifications in 1980 and consequently many born prior to this were at a higher risk of going undiagnosed. In addition diagnostic facilities have expanded very significantly in more recent years. There may therefore be many older individuals living in Scotland who would meet the criteria for ASD but have never received a diagnosis.

7.27 Table 7.7 shows the number and percentage of individuals with each type of ASD diagnosis according to their sex. As with the total sample, Asperger’s/ HFA was the most prevalent diagnoses amongst both males and females. Chi-square analysis was carried out to investigate whether significant differences existed between the number and percentage of individuals of males and females with each type of diagnosis, however no significant relationship was found, X 2 (2, 949) = 3.03 p > .05). That is to say, though ASD are considerably more prevalent in males than females, the rates of different types of ASD did not appear to be influenced by sex in this sample.

Table 7.7 ASD diagnosis and sex

Diagnosis Sex Total Sample n (%)
Male (%) Female (%)
Autism a 163 (22) 54 (25) 217 (100)
Asperger’s & HFA b 324 (44) 101 (47) 425 (100)
Other ASD diagnoses c 248 (34) 59 (28) 307 (100)
Total 735 (100) 214 (100) 949 (100)*

a Including ‘Childhood Autism’ or ‘Autistic Disorder’; b Including ‘Asperger’s Disorder’; c Including general/non-specific ASD diagnoses ‘Atypical Autism’ or ‘ PDD-NOS’;* Note 1 case missing as sex data was not provided

Intellectual Disability

7.28 As highlighted in Chapter 4 of this report, understanding the number and percentage of individuals on the spectrum with ID is of crucial importance if we are to provide appropriate levels of support for those on the spectrum. However, to date, relatively few studies have focussed on the prevalence of ID across the spectrum, and instead most of the research covering the relationship between these conditions has instead focussed the prevalence of ASD amongst those with ID.

7.29 Table 7.8 shows the number of individuals in the sample with co-occurring intellectual disabilities ( ID). The proportion of the sample for whom information on the presence or absence of intellectual disability was available was 67% (n = 649). This included 51% of those with autism (n = 110), 100% of those with Asperger’s (n = 417), and 37% of those with other ASD (n = 113).

7.30 Of this subsample of 649 individuals, 20% overall reported or were reported to have ID (n = 127), and of those with ID, 15% reported moderate or severe ID (n = 99) and 5% mild ID (n = 28). Of the individuals with autism who provided ID data (n = 110), 65% had ID (n = 72), 53% of whom had moderate and severe ID (n = 58), and 13% of whom had mild ID (n = 13). Of those with Other ASD who provided ID data (n = 113), 51% had no ID, and of the 49% with ID, 36% had moderate or severe ID, and 12% had mild ID. Finally, for those with Asperger’s Syndrome, the diagnostic criteria exclude the presence of intellectual disability.

Table 7.8 Co-occurring intellectual disability ( ID) according to ASD diagnosis

Presence and level of ID Condition n (%) Total Sample n (%) a
Autism Asperger’s/ HFA Other ASD
No ID 38 (35) 426 (100) 58 (51) 522 (80)
ID 72 (65) 0 (0) 55 (49) 127 (20)
Moderate & severe 58 (53) 0 (0) 41 (36) 99 (15)
Mild 14 (13) 0 (0) 14 (12) 28 (5)
Total 110 (100) 426 (100) 113 (100) 649 (100)

a ID data were available for 649/950 participants

7.31 Table 7.9 shows the presence and level of ID according to the age of individuals at the point of completion of the survey for whom ID data was available. Though there was some evidence from this raw data to suggest that there were slight differences in the percentage of individuals with and without ID across different age groups (with ID appearing to be slightly less prevalent in those over 50 years old and more prevalent amongst those aged 11 – 18 years), chi-square analysis confirmed these differences were not statistically significant, X 2 (1, 649) = 3.04, p > .05.

7.32 Table 7.10 shows the relationship between sex and the level and presence of ID. Chi square analysis confirmed that there were no significant differences between males and females, X 2 (1, 649) = 1.46, p > .05. The significance of the differences between the number of males and females with moderate/severe ID was also tested but again no significant relation was found, X 2 (1, 649) = 0.50, p > .05.

Table 7.9 Co-occurring intellectual disability ( ID) according to age

Age Group Presence and level of intellectual difficulties n (%) Total sample n (%) a
No ID ID
Mild Moderate/ severe Total
0 – 10 90 (80) 6 (5) 17 (15) 23 (20) 113 (100)
11 – 18 170 (78) 12 (5) 37 (17) 49 (22) 219 (100)
19 – 49 230 (82) 10 (4) 40 (14) 50 (18) 280 (100)
≥ 50 31 (86) 0 (0) 5 (14) 5 (14) 36 (100)
Total 521 (100) 28 (4) 99 (15) 127 (19) 648 (100)

a ID data was only available for 649/950 participants

Table 7.10 Co-occurring intellectual difficulty status ( ID) according to sex

Presence and level of ID Sex Total sample n (%) a
Male (%) Female (%)
No ID 386 (79) 136 (85) 522 (80)
ID 103 (21) 24 (15) 127 (20)
Moderate/Severe 76 (16) 23 (14) 99 (15)
Mild 27 (6) 1 (1) 28 (4)
Total 489 (100) 160 (100) 649(100)

a ID data was available for 649/950 participants

Table 7.11 Presence of co-occurring diagnoses (excluding ID) amongst ASD individuals ≥ 16 years

Number of Comorbidities Condition n (%) Total ≥16 years Sample n (%) (n = 404)
Autism Asperger’s / HFA Other ASD
None 48 (59) 117 (50) 43 (51) 208 (51)
One or more 34 (41) 119 (50) 41 (49) 196 (49)
1 20 (24) 66 (28) 27 (32) 113 (28)
2 12 (15) 42 (18) 10 (12) 64 (16)
3+ 2 (2) 10 (5) 4 (5) 19 (5)
Total 82 (100) 236 (100) 84 (100) 404 (100)

7.33 Though relatively few within the ASD literature have investigated sex differences in relation to ID status, these findings differ from what has previously been reported in the broader ID literature, with reports suggesting that ID (both mild and moderate/severe) tend to be more prevalent amongst males (e.g. Altarac & Saroha, 2007).

Other diagnoses

7.34 In total, 33% of the sample (n = 311) had at least one co-occurring diagnoses in addition to their ASD diagnosis (excluding intellectual disabilities dealt with earlier in this chapter). However, given that many co-occurring conditions (such as mood disorders) are more prevalent amongst older adolescents and adults, a follow-up analysis focussed specifically on the rates of comorbidities amongst those over the age of 16. Table 7.11 shows the number and percentage of individuals with co-occurring conditions in this older subsample – similar statistics for the total population have been included in Appendix C.2.

7.35 In total 49% of those aged 16 and above had at least one co-occurring condition (in comparison to 33% of the total sample). Relatively few studies in the field have previously reported overall rates of co-occurring conditions, with most instead focussing on the prevalence of specific conditions instead (this is matter discussed in more detail on the following pages).

7.36 One investigation which has covered this matter is Simonoff et al.’s (2012) study focussing on 112 ASD individuals living in London, which found that that 71% of those with ASD had at least one other co-occurring condition. One notable difference here is that the study by Simonoff et al. involved a sample involving a greater proportion of individuals who would be described as lower functioning (i.e. individuals with an IQ < 70). However, it is amongst this population that rates of comorbidities appeared lowest in our own sample.

7.37 Chi-square analyses were carried out to investigate the relationship between the type of ASD diagnosis an individual had, and the presence of at least one other co-occurring diagnosis. This analysis revealed a significant relationship between diagnosis and co-occurring conditions ( X 2 [2, 950] = 10.73, p < .01). There was some evidence from the raw data to suggest that co-occurring conditions were more prevalent amongst those with Asperger’s/ HFA. To explore the significance of this relationship in greater detail, the above data was partitioned in order to compare the presence of comorbid conditions in those with Asperger’s/ HFA in the rate of these conditions in the rest of the sample (i.e. a 2x2 contingency table was created where the initial three columns were replaced by two columns: one representing those with Asperger’s/ HFA, and the other representing everyone else in the sample). Following the partitioning, the chi-square analysis was re-run, and again there was evidence to suggest that there was a difference in the prevalence of co-occurring conditions amongst those with Asperger’s/ HFA in comparison to the rest of the sample, X 2 (1, 950) = 14.83, p < .001. Partitioning the data in this manner also allowed odds-ratio statistics to be calculated, and these calculations indicated that within our sample those with Asperger’s/ HFA were 1.7 times more likely to have a co-occurring condition in comparison to other individuals in the sample.

7.38 Table 7.12, detailing the rates of each co-occurring condition within the sample, provides further insight into the differences in the number of co-occurring conditions across each type of ASD. Notable is that of the 180 individuals with mood disorders, 122 (67%) had a diagnosis of Asperger’s/ HFA – a rate which contributes greatly to the overall differences in the number and percentage of individuals with least one co-occurring condition across the different types of ASD.

7.39 More generally, Table 7.12 also reveals mood disorders are the most prevalent co-occurring condition across the entire sample with 180 of the 950 individuals in the sample (19%) experiencing co-occurring bipolar disorder, depression or anxiety. Following on from this, 10% of the sample had co-occurring ADHD (n = 92), 6% had either OCD or Tourette’s [3] (n = 52), and 5% of the sample had a co-occurring diagnosis of epilepsy (n =45). All other co-occurring conditions affected less than 5% of the sample

7.40 In terms of the rates of co-occurring associated with each type of ASD, amongst those with autism (n = 217), 13% had mood disorders (n = 24), 8% had epilepsy (n = 17), 7% had ADHD (n = 16), and 6% had OCD or Tourette’s syndrome (n = 12). All other comorbidities were present in less than 2% of those with autism.

7.41 Of those with Asperger’s/ HFA, in addition to the previously mentioned 67% with mood disorders, 9% had ADHD (n = 38), 7% had OCD or Tourette’s syndrome (n = 28), and all remaining co-occurring conditions affected less than 3% of those with Asperger’s.

7.42 Finally, of those with other ASD diagnoses, 19% had a mood disorder (n = 34), 12% had a diagnosis of ADHD (n = 38), 7% had a diagnosis of OCD or Tourette’s (n = 12) and all other co-occurring diagnoses affected 1% of this subsample.

7.43 There was a ≥ 5% difference in the percentage of individuals with each type of ASD who had co-occurring ADHD, epilepsy and mood disorders, and these differences were explored further using chi-square analyses. These analyses revealed a significant difference in the within sample prevalence of epilepsy, X 2 (2, 950) = 9.43, p < .01, and mood disorders, X 2 (2, 950) = 47.23, p < .001, but not ADHD, X 2 (2, 950) = 4.16, p > .05.

Table 7.12 Co-occurring conditions by type of ASD

Co-occurring condition Type of ASD Total Sample (n = 950)
Autism (n = 217) Asperger’s/ HFA (n = 426) Other ASD (n = 307)
ADHD 16 (7) 38 (9) 38 (12) 92 (10)
OCD & Tourette’s a 12 (6) 28 (7) 12 (4) 52 (5)
OCD 9 (4) 21 (5) 7 (2) 37 (4)
Tourette’s 3 (1) 7 (2) 6 (2) 16 (2)
Epilepsy 17 (8) 11 (3) 17 (6) 45 (5)
Fragile X 2 (1) 1 (< 1) 2 (1) 5 (1)
Tuberous Sclerosis 1 (1) 0 (0) 0 () 1 (< 1)
Down Syndrome 3 (1) 0 (0) 3 (1) 6 (1)
Schizophrenia 3 (1) 1 (< 1) 0 (< 1) 4 (< 1)
Mood Disorder 24 (13) 122 (28) 34 (19) 180 (19)
Bipolar Disorder 1 (1) 7 (2) 2 (1) 10 (1)
Depression 10 (5) 79 (19) 14 (5) 103 (11)
Anxiety 21 (10) 85 (20) 31 (10) 137 (14)
Challenging Behaviour 4 (2) 2 (1) 0 (< 1) 6 (1)

a Group totals reflect the number of unique individuals with each of these conditions (e.g. if an individual had depression and anxiety, then they were only included once in the Mood Disorder group total).

7.44 The raw data indicated that epilepsy was least prevalent amongst those with Asperger’s/ HFA, and to test the significance of this relationship the data was partitioned to compare the prevalence of epilepsy between those with Asperger’s/ HFA and the rest of the sample. Once partitioned, the chi-square analysis was re-run and a significant difference between the groups was still found, X 2 (1, n = 950) = 7.95, p < .01. The partitioned data also allowed an odds ratio statistic to be calculated and this indicated that those with Asperger’s/ HFA were 2.6 times less likely to have epilepsy in comparison to the rest of the sample.

7.45 Further analysis also indicated that mood disorders were most prevalent amongst those in the sample with Asperger’s/ HFA, therefore again the data was partitioned to compare the prevalence of this condition amongst those with Asperger’s/ HFA in comparison to the rest of the sample. Chi-square analysis confirmed that this difference was significant, X 2 (1, 950) = 47.23, p < .001. This partitioning of the data also an odds ratio statistic to be generated, and this indicated that mood disorders were 3.23 times more prevalent amongst individuals with Asperger’s/ HFA in comparison to others in the sample.

7.46 While mood disorders were found to be prevalent amongst 28% of those with Asperger’s/ HFA, this rate is considerably lower than those which have previously been published in the literature. For example, estimates of depression in this population have previously ranged between 54% and 75% and estimates of anxiety disorders have ranged between 43% and 56% (Barnhill et al., 2001; Lugnegard et al., 2011; Sukhodolsky et al., 2008; Whitehouse et al., 2009).

Education

7.47 Table 7.13 provides a summary of the type of school that all ASD individuals aged 16 and over attended throughout their education. Complete frequency data relating to the school placement of all 950 individuals has not been provided here due to the large number of young individuals in the sample who were still at an early stage of their education, however this information has been included in Appendix C.3

7.48 The statutory school leaving age in Scotland was used as a cut-off point to determine final educational placement of individuals in the sample therefore the analysis below relates to anyone in the sample who was aged 16 and over.

7.49 Of those ≥ 16 years (n = 404), 83% had attended a mainstream school at one stage in their education, this included 61% of those with an autism diagnosis (n = 50), 93% of those with Asperger’s/ HFA (n = 220) and 76% of those with other ASD (n = 65). In addition to this 33% of the population attended a special unit within a mainstream school, including 49% of those with autism (n = 40), 25% of those with Asperger’s/ HFA (n = 60), and 37% of those with other ASD (n = 32).

7.50 A greater number of individuals had attended a general special day school (18% of those with autism, 17% of those with Asperger’s/ HFA, 14% of those with other ASD, and 17% of all those aged 16 and over), in comparison to an ASD specific special day school (18% of those with autism, 2% of those with Asperger’s/ HFA, 7% of those with Other ASD, and 6% of all those ≥ 16 years).

7.51 A similar pattern was also found in general special needs residential schools (attended by 10% of those with autism, 25% of those with Asperger’s/ HFA, and 16% of all those over the age of 16), in comparison to those at ASD specific residential schools (attended by 5% of those with autism, 2% of those with Asperger’s/ HFA, 6% of those over the age of 16, and 4% of all those ≥ 16 years).

7.52 A number of individuals had also been educated at home at some point in their life (n = 20) including 6% of those with autism (n = 5), 4% of those with Asperger’s/ HFA (n = 9), 7% of those with other ASD (n = 6). Finally some individuals had also received an alternative form of education such as one-to-one teaching within a mainstream establishment or else were part of an ABA programme within a special need school (n = 14).

Table 7.13 Educational placement of individuals with ASD aged ≥ 16 years

School Type Condition n (%) Total ≥16 years sample n (%)

(n = 404) a
Autism (n = 82) Asperger’s/ HFA (n = 236) Other ASD (n = 86)
Mainstream School 50 (61) 220 (93) 65 (76) 335 (83)
Preschool 30 (37) 160 (68) 52 (60) 242 (60)
Primary School 25 (30) 195 (83) 51 (59) 271 (67)
Secondary School 22 (27) 186 (79) 43 (50) 251 (62)
Special Unit in a Mainstream School 40 (49) 60 (25) 32 (37) 132 (33)
Preschool 18 (22) 23 (10) 11 (13) 52 (13)
Primary School 24 (29) 27 (11) 17 (20) 68 (17)
Secondary School 13 (16) 31 (13) 17 (20) 61 (15)
Special ASD Day School 15 (18) 4 (2) 6 (7) 25 (6)
Preschool 7 (9) 0 (0) 2 (2) 9 (2)
Primary School 11 (13) 4 (2) 4 (5) 19 (5)
Secondary School 12 (15) 5 (2) 1 (1) 18 (4)
Other Special Day School 15 (18) 40 (17) 12 (14) 67 (17)
Preschool 3 (4) 10 (4) 5 (6) 18 (4)
Primary School 9 (11) 27 (11) 7 (8) 43 (11)
Secondary School 10 (12) 21 (9) 5 (6) 36 (9)
ASD Residential School 4 (5) 5 (2) 5 (6) 14 (3)
Preschool 1 (1) 3 (1) 1 (1) 5 (1)
Primary School 2 (2) 0 (0) 3 (3) 5 (1)
Secondary 3 (4) 2 (1) 4 (5) 9 (2)
Other Special Residential School 5 (10) 5 (1) 1 (0) 11 (16)
Preschool 3 (4) 2 (1) 0 (0) 5 (1)
Primary School 3 (4) 1 (0) 0 (0) 4 (1)
Secondary School 5 (6) 2 (1) 0 (0) 7 (2)
Home Education 5 (6) 9 (4) 6 (7) 20 (5)
Preschool 2 (2) 1 (0) 0 (0) 3 (1)
Primary School 1 (1) 3 (1) 2 (2) 6 (1)
Secondary School 2 (2) 5 (2) 4 (5) 11 (3)
Other 5 (6) 7 (3) 2 (2) 14 (3)
Preschool 3 (4) 1 (0) 0 (0) 4 (1)
Primary School 2 (2) 2 (1) 2 (2) 6 (1)
Secondary School 2 (2) 4 (2) 1 (1) 7 (2)

a Individuals may be represented in more than one cell in the table above; group totals reflect the number of unique individuals attending each type of school

Table 7.14 Educational placement of individuals aged ≥16 years according to ID presence and level

Type of School ID status n (%) Total Sample ≥16 years n (%) (n = 404) a
No ID (n = 328) ID
Mild (n = 15) Moderate/ Severe (n = 62) Total (n =127)
Mainstream School 308 (94) 11 (73) 32 (52) 43 (56) 351 (87)
Preschool 229 (70) 9 (60) 22 (35) 31 (40) 260 (64)
Primary School 261 (80) 7 (47) 18 (29) 25 (32) 286 (71)
Secondary School 247 (75) 4 (27) 15 (24) 19 (25) 266 (66)
Special Unit in a Mainstream School 99 (30) 9 (60) 32 (52) 41 (53) 140 (35)
Preschool 37 (11) 3 (20) 14 (23) 17 (22) 54 (13)
Primary School 51 (16) 6 (40) 17 (27) 23 (30) 74 (18)
Secondary School 47 (14) 4 (27) 12 (19) 16 (21) 63 (16)
Special ASD Day School 14 (4) 4 (27) 7 (11) 11 (14) 25 (6)
Preschool 4 (1) 1 (7) 3 (5) 4 (5) 8 (2)
Primary School 12 (4) 4 (27) 5 (8) 9 (12) 21 (5)
Secondary School 9 (3) 3 (20) 7 (11) 10 (13) 19 (5)
Special Day School (Other) 18 (5) 3 (20) 14 (23) 17 (22) 35 (9)
Preschool 15 (5) 0 (0) 6 (10) 6 (8) 21 (5)
Primary School 37 (11) 3 (20) 7 (11) 10 (13) 47 (12)
Secondary School 29 (9) 1 (7) 8 (13) 9 (12) 38 (9)
ASD Residential School 9 (3) 0 (0) 7 (11) 7 (9) 16 (4)
Preschool 3 (1) 0 (0) 1 (2) 1 (1) 4 (1)
Primary School 3 (1) 0 (0) 4 (6) 4 (5) 7 (2)
Secondary 4 (1) 0 (0) 4 (6) 4 (5) 8 (2)
Special Residential School 10 (3) 0 (0) 6 (10) 6 (8) 16 (4)
Preschool 3 (1) 0 (0) 2 (3) 2 (3) 5 (1)
Primary School 2 (1) 0 (0) 1 (2) 1 (1) 2 (0)
Secondary School 4 (1) 0 (0) 4 (6) 4 (5) 8 (2)
Home Education 17 (5) 0 (0) 4 (6) 4 (5) 21 (5)
Preschool 2 (1) 0 (0) 1 (2) 1 (1) 3 (1)
Primary School 6 (2) 0 (0) 1 (2) 1 (1) 7 (2)
Secondary School 10 (3) 0 (0) 2 (3) 2 (3) 12 (3)
Other 11 (3) 0 (0) 6 (10) 6 (8) 17 (4)
Preschool 3 (1) 0 (0) 3 (5) 3 (4) 6 (1)
Primary School 4 (1) 0 (0) 3 (5) 3 (4) 7 (2)
Secondary School 7 (2) 0 (0) 1 (2) 1 (1) 8 (2)

a Individuals may be represented in more than one cell in the table above; group totals reflect the number of unique individuals attending each type of school

7.53 In general it is clear that those in our sample with Asperger’s/ HFA are better represented at mainstream schools and less well represented at special schools in comparison to those with other forms of ASD. Both findings fit with what is typically expected in this population, in that those with the least severe social and intellectual needs are the least likely to receive additional levels of support at school.

7.54 Table 7.14 shows the number and percentage of individuals attending each type of educational establishment according to their ID status. Of those aged 16 and over without ID (n = 328) 94% had attended a mainstream school at some point in their education (n = 308), 30% had attended a special unit in a special mainstream school (n = 99), 4% had attended a special ASD day school (n = 14), 5% had attended a general special day school (n =18), 3% had attended an ASD residential school (n = 9), 3% had attended a general special needs day school (n = 10) and 5% had been educated at home (n = 17).

7.55 Of those with ID, 56% had attended a mainstream school (n = 43), 30% had attended a special unit in a mainstream school (n = 30), 8% had attended a special ASD day school (n =10), 12% had attended a general special needs day school (n = 15), 5% had attended an ASD residential school (n = 6) and 4% were educated at home.

7.56 Again, these findings are in line what would be expected within this population, with those with the majority of those without ID primarily attending mainstream schools throughout their education, while a much greater number and percentage of those with ID attended schools which provide additional levels of support.

7.57 A more in depth understanding of the educational experiences of the sample was developed by examining the highest level of educational support individuals received throughout their education. To analyse this, schools were ranked according to the level of support they are typically associated with, as shown in Table 7.15. The five main types of school were ranked so that the school associated with the lowest level of support was represented by ‘1’ and the school associated with the highest level of support was represented by ‘5’. In cases where individuals had attended more than one type of school, the school considered to be the one which provided them with the highest level of support was the one which ranked highest. Findings relating to the highest level of support received by those aged ≥ 16 years have been reported in Table 7.16, and figures for the total sample have also been included in Appendix C.4. Home education and attendance of ‘other’ types of education were not taken into consideration as part of the analysis as too little was known about the provision of support in these cases.

Table 7.15 Ranking of school type according to associated level of support

Rank Type of School
1 (lowest) Mainstream school
2 Special unit in a Mainstream school
3 Special ASD day school
4 Other ASD day school
5 (highest) Special residential school ( ASD or Other)

7.58 As shown in Table 7.16, 46% of those over the age of 16 (n = 186) received their highest level of support at a mainstream schools; this was the case for 21% of those with autism (n = 17), 56% of those with Asperger’s/ HFA (n = 133), and 42% of those with other ASD (n = 36). For 24% of the sample (n = 98), the highest level of support received was within a special unit in a mainstream school; this included 30% of those with autism (n = 25), 21% of those with Asperger’s/ HFA (n = 49), and 28% of those with other ASD (n = 24).

Table 7.16 Highest educational placement for ASD individuals aged ≥ 16 years according to ASD diagnosis a

Highest Level of Educational Support Type of ASD Diagnosis n (%) Total ≥16 years sample n (%) (n = 404)
Autism (n = 82) Asperger’s Syndrome/ HFA (n = 236) Other ASD (n =86)
Mainstream School 17 (21) 133 (56) 36 (42) 186 (46)
Special Unit in a Mainstream School 25 (30) 49 (21) 24 (28) 98 (24)
Special ASD Day School 14 (17) 8 (3) 7 (8) 29 (7)
Other ASD Day School 12 (15) 36 (16) 12 (14) 60 (15)
Special Residential School ( ASD specific or other) 13 (16) 10 (4) 7 (8) 30 (7)
Total 81 (100)* 236 (100) 86 (100) 403 (100)

a Note: One individual was not included in this analysis as their highest level of educational support was at received at home

7.59 In total 7% (n = 29) received the highest level of support at a special ASD school, including 17% of those with autism (n = 14), 3% of those with Asperger’s/ HFA (n = 8), and 8% of those with other ASD. A further 15% of individuals over 16 (n = 60) received their highest level of support at other, more general special needs schools, including 15% of those with autism (n = 12), 16% of those with Asperger’s/ HFA (n = 37), 12% of those with other ASD (n = 14). Finally, 15% of those over 16 received their highest level of support at a residential school, including 15% of those with

Table 7.17 Highest educational placement amongst individuals aged ≥ 16 years according to the presence and level of intellectual disability a

Highest Level of Educational Support Presence and Level of ID n (%) Total ≥16 years sample n (%) (n = 404)
No ID (n = 328) ID
Mild (n = 15) Moderate/Severe (n = 62) Total (n = 77)
Mainstream School 172 (52) 4 (27) 10 (16) 14 (18) 186 (46)
Special Unit in a Mainstream School 72 (22) 4 (27) 21 (34) 25 (32) 97 (24)
Special ASD Day School 16 (5) 4 (27) 9 (15) 13 (17) 29 (7)
Other Day School 49 (15) 2 (13) 10 (16) 12 (16) 61 (15)
Residential School ( ASD specific or other) 18 (5) 1 (7) 11 (18) 12 (16) 30 (7)
Total 327* (100) 15 (100) 62 (100) 77 (100) 403 (100)*

a One individual was not include in this analysis as their highest level of educational support was at received at home

autism (n = 12), 4% of those with Asperger’s/ HFA (n = 10), and 8% of those with other ASD (n = 7).

7.60 There was evidence to suggest that those with Asperger’s/ HFA were more likely to receive their highest level of educational support from a mainstream school in comparison to the rest of the sample. Chi-square analysis confirmed this X 2 (1, 404) = 23.83, p < .001, and odds ratio statistics indicated that those with Asperger’s were 3.76 times more likely, in comparison to the rest of the sample, to have received their highest level of educational support from a mainstream school. By comparison, those with autism were more 3.86 times more likely, in comparison to the rest of the sample, to have received their highest level of educational support from an additional support school ( X 2 (1, 404) = 23.71, p < .001).

7.61 Table 7.17 shows the highest level of educational support received according to the presence and level of intellectual disability. Again, these figures relate to those over the age of 16, and alternative data relating to the entire sample has been provided in Appendix C.4

7.62 In this older sub-population, 46% had received their highest level of support at a mainstream school (n = 186), including 52% of those with no ID (n = 172), 27% of those with mild ID (n = 4), and 16% of those with moderate ID (n = 10). A further 24% received the greatest level of support at a special unit in a mainstream school (n = 97), including 22% of those with no ID (n = 72), 27% of those with mild ID (n = 4), and 16% of those with moderate or severe ID (n = 34). For 7% of the sample, the highest level of support received was at an ASD specific special needs day school, including 5% of those with autism (n = 16), 27% of those with mild ID (n = 4), and 15% of those with moderate or severe ID (n = 9). A greater number of individuals (n = 61) had received the greatest level of support at a more general special needs school, including 15% of those no ID (n = 49), 13% of those with mild ID (n = 2), and 16% of those with moderate or severe ID (n = 10). Finally, 7% of this sub-population attended residential schools, including 5% of those with no ID (n = 18), 7% of those with mild ID (n = 1), and 18% of those with moderate or severe ID (n = 180).

7.63 Table 7.18 shows the type of school which provided individuals with their highest level of educational support according to their age. These data indicate that there was some influence of age on the educational experiences of the individuals in our sample in that a much lower percentage of individuals aged between 16 and 26 received their highest level of educational support from a mainstream school. This difference was confirmed as statistically significant through chi-square analysis, X 2 (1, 404) = 13.94, p < .001, and odds ratio statistics confirmed that those in the 16-26 year age band were 1.94 times less likely to have received their highest level of educational support from a mainstream school. This may indicate that individuals on the spectrum who have attended school more recently have been more likely to end up in a higher support placement, and potentially also one that more appropriately meets their needs.

Table 7.18 School providing highest level of educational support amongst individuals aged ≥ 16 years according to age a

Type of school providing highest level of educational support Age Group (years) n (%) Total ≥16 years sample n (%) (n = 404)
16 – 26 27 – 37 28 – 49 50 ≥
Mainstream School 85 (39) 37 (49) 43 (59) 22 (62) 187 (46)
Special Unit in a Mainstream School 68 (31) 17 (22) 11 (15) 1 (3) 97 (24)
Special ASD Day School 21 (10) 5 (7) 2 (3) 1 (3) 29 (7)
Other ASD Day School 32 (14) 10 (13) 13 (18) 7 (19) 62 (15)
Special Residential School
( ASD specific or other)
14 (6) 7 (9) 4 (5) 4 (11) 29 (7)
Total 219 (100) 76 (100) 73 (100) 36 (100) 404 (100)

a One individual was not included in this analysis as their highest level of educational support was at received at home

Table 7.19 School providing highest level of educational support amongst ASD individuals ≥ 16 years according to sex

Type of school providing highest level of educational support Sex n (%) Total ≥16 years sample n (%) (n = 404)
Female Male
Mainstream School 63 (54) 122 (44) 186 (46)
Special Unit in a Mainstream School 18 (15) 80 (29) 98 (24)
Special ASD Day School 5 (4) 24 (9) 29 (7)
Other ASD Day School 21 (18) 40 (14) 61 (15)
Special Residential School ( ASD specific or other) 9 (8) 20 (7) 29 (7)
Total 116 (100) 276 (100) 402 (100)

a Two individuals were not included in this analysis, one because they did not report sex data and one because their highest level of educational support was received at home

7.64 Table 7.19 shows the sex differences in the type of school providing individuals in the sample with the highest level of educational support. The data indicated that a greater proportion of females received their highest level of educational support from a mainstream school, and follow up chi-square analysis found that though small, these differences were significant, X 2 (1, 404) = 4.67, p < .05. The contrast, similar analysis indicated that males were more likely to receive their highest level of educational support from a special unit in a mainstream school, X 2 (1, 404) = 7.19, p < .01.

Table 7.20 School providing highest level of educational support according to presence of co-occurring conditions amongst individuals aged ≥ 16 years

Type of school providing highest level of educational support Co-occurring condition (%) Total ≥16 years sample n (%) (n = 381) a
ADHD (n = 29) OCD/ Tourette’s syndrome (n = 41) Epilepsy (n = 29) Mood Disorders (n = 138)
Mainstream School 18 (62) 19 (46) 12 (42) 81 (59) 130 (55)
Special Unit in a Mainstream School 2 (7) 7 (17) 8 (28) 22 (16) 39 (16)
Special ASD Day School 2 (7) 6 (15) 4 (14) 7 (5) 19 (8)
Other ASD Day School 6 (20) 6 (15) 3 (10) 21 (15) 36 (15)
Special Residential School ( ASD specific or other) 1 (3) 3 (7) 2 (7) 7 (5) 13 (5)
Home Educated 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

a This figure reflects the total number of participants who indicated (a) whether or not they had ADHD, OCD/Tourette’s syndrome, Epilepsy or mood disorders and (b) provided information relating to their educational history

7.65 Table 7.20 shows differences in the highest level of educational support received relative to the presence of other co-occurring conditions (this analysis focusses only on the four most prevalent co-occurring conditions, due to the relatively small numbers associated with the other types of co-occurring condition covered earlier in this chapter; see Table 7.12 for further details).

7.66 Of 138 individuals with co-occurring mood disorders, 59% received their highest level of educational support at a mainstream school, and follow-up analysis confirmed this to be significantly more than for other conditions, X 2 (1, 381) = 12.43, p < .001. This is likely to reflect the higher prevalence of mood disorders among those who are higher functioning.

7.67 While a similar pattern was identified in relation to ADHD, OCD/Tourette’s syndrome and epilepsy in that the majority of individuals with these conditions also received their highest level of educational support from a mainstream school, chi-square analyses confirmed that these differences were not found to be significant (all X 2 values < 3.00, all p values > .05).

7.68 Binary logistic regression analysis was used to identify factors predicting the likelihood of individuals receiving their highest level of educational support in a mainstream school. 186 individuals over the age of 16 received their highest level of educational support from a mainstream school and 332 received the highest level of support from another type of school. Exploratory analysis was carried out on candidate variables (listed in Appendix C.5) which were added to a hierarchical model in the following five blocks: (i) those relating to demographics, (ii) those relating to core diagnoses,(iii) those relating to co-occurring conditions, (iv) those relating to other outcomes and (v) those relating to service-use.

7.69 The final model shown in Table 7.21 reports only those candidate variables which improved the associated Nagelkerke R 2 by at least .02 [4] . Candidate variables excluded from the final model in this way and relevant statistics are detailed in Appendix C.5.

7.70 In Block one of the model, age was found to make a small but significant contribution to the overall model, X 2 (1, 404) = 12.39, p < .001, and explained 4% of the variance relating to whether or not individuals received their highest level of educational provision from a mainstream school (Nagelkerke R 2 for block = .04).

7.71 Exploratory analysis revealed that both autism and Asperger’s/ HFA diagnoses were significant predictors of whether or not an individual received their highest level of educational support from a mainstream school. However, as ID status was found to make a greater contribution to the overall model, it was included in the final model (note that both ID status and the ASD diagnostic categories could not be included in the same model due to the multicollinearity between the variables, this multicollinearity would have increased the likelihood of an incorrect interpretation of the data).

7.72 Block two therefore added ‘ ID status’ into the model, and this made a further significant contribution to the models, X 2 (1, 404) = 35.71, p < .001, and accounted for a further 11% of the variance in the data.

7.73 Finally, block three added ‘depression diagnosis’ to the model, which proved to be a further significant predictor of an individual’s likelihood of receiving their highest level of educational support from a mainstream school, X 2 (1, 404) = 7.03, p < .01, and explained an additional 2% of the variance in the model.

Table 7.21 Binary logistic Regression of the factors which predict mainstream school as the highest level of educational support [5]

Model β SE β Wald χ 2 df Exp β
Odds-Ratio Lower Upper
Block 1






Age *** .03 .01 12.39 1 1.03 1.01 1.05
Block: Nagelkerke R 2 = .04
Block 2






Age** .03 .10 9.74 1 1.03 1.01 1.04
ID Status*** -1.86 .67 25.06 1 .16 .04 .60
Block: Nagelkerke R 2 = .11 Model: Nagelkerke R 2 = .15
Block 3






Age* .02 .01 5.93 1 1.02 1.00 1.04
ID Status*** -1.74 .67 21.69 1 .18 .05 .68
Depression** .69 .27 6.88 1 2.00 1.17 3.42
Block: Nagelkerke R 2 = .02 Model: Nagelkerke R 2 = .17

Note: * p < .05 ** p < .01, *** p < .001

7.74 In terms of what this model is able to tell us about the factors which may influence the level and type of educational support and individual receives there were two key findings.

7.75 Firstly, there is some evidence here to suggest that older individuals in the sample were more likely to receive their highest level of educational support from a mainstream school. This is something that makes practical sense given that a) historically, provision for those with ASDs was poorer than it is now, and as such it is more likely that individuals with the condition would have received their highest level of educational support from a mainstream school due to a lack of more appropriate support and b) awareness of ASDs was also historically poorer meaning that those with ASDs were less likely to be identified and in turn less likely to receive the type of support they required.

7.76 The second key finding here is that those with intellectual disabilities were 5.55 times less likely to receive their highest level of educational support from a mainstream school. Of most interest here is that, as mentioned above, ID status was identified as a stronger predictor of highest educational support placement than either autism or Asperger’s/ HFA diagnosis, a finding which provides some evidence to suggest that school placements and educational support are more closely associated with intellectual ability than the social and behavioural symptoms which often accompany ASD and which can also be disruptive to an individual’s education.

7.77 Finally, one additional finding here was that those with a diagnosis of depression were twice as likely to receive their highest level of educational support from a mainstream school. However, it may be that depression diagnosis did not influence school placement but instead that the factors associated with diagnosis of depression are also associated with the type of abilities and requirements that make someone capable of attending a mainstream school (see para. 3.42 regarding the elevated susceptibility to depression among those who are high functioning; this relationship has also previously been reported in the ASD literature, e.g. Barnhill et al., 2001; Lugnegard et al., 2011; Sukhodolsky et al., 2008; Whitehouse et al., 2009).

7.78 Binary logistic regression analysis was used to identify factors predicting the likelihood of individuals receiving their highest level of educational support in a mainstream school. Of individuals over the age of 16, 98 received their highest level of educational support from a mainstream school and 306 received the highest level of support from another type of school. Exploratory analysis was carried out on candidate variables (listed in Appendix C.5. which were added to a hierarchical model in the following five blocks: (i) those relating to demographics, (ii) those relating to core diagnoses,(iii) those relating to co-occurring conditions, (iv) those relating to other outcomes and (v) those relating to service-use.

7.79 The final model shown in Table 7.22 reports only those candidate variables which improved the associated Nagelkerke R 2 by at least .02 [6] . Candidate variables excluded +from the final model in this way and relevant statistics are detailed in Appendix C.5.

7.80 In Block one of the model, age was found to make a small but significant contribution to the overall model, X 2 (1, 404) = 16.79, p < .001, and explained 6% of the variance relating to whether or not individuals received their highest level of educational provision from a special unit in a mainstream school.

7.81 Block two of the analysis added sex to the regression model and significantly improved the null model ( X 2 (1, 404) = 6.25, p < .05). The addition of this predictor increased the variance explained by the model to 8% (Nagelkerke R 2 for block = .02).

7.82 Finally, block three added ADHD diagnosis to the regression model, and this block was again significantly better at classifying the data than the null model ( X 2 (1, 404) = 9.33, p < .05) and yielded a further 4% improvement to the model (Nagelkerke R 2 for block = .14).

Table 7.22 Binary logistic Regression of the factors which predict a special unit in a mainstream school as the highest level of educational support [7]

Model β SE β Wald χ 2 df Exp β
Odds-Ratio Lower Upper
Block 1






Age*** -.04 .01 12.95 1 .96 .94 .98
Block: Nagelkerke R 2 = .06
Block 2






Age*** -.04 .01 12.95 1 .96 .94 .98
Sex** .70 .30 5.75 1 2.02 1.12 3.66
Block: Nagelkerke R 2 = .02 Model: Nagelkerke R 2 = .08
Block 3






Age*** -.04 .01 14.26 1 .96 .94 .98
Sex** .79 .31 7.04 1 2.19 1.21 3.99
ADHD* -1.78 .76 5.97 1 .17 .04 .75
Block: Nagelkerke R 2 = .04 Model: Nagelkerke R 2 = .12
Block 4






Age*** -.04 .01 10.72 1 .96 .94 .99
Sex* .71 .31 5.65 1 2.04 1.11 3.73
ADHD* -1.75 .76 5.73 1 .17 .04 .77
Depression* -.77 .38 4.32 1 .46 .22 .98
Block: Nagelkerke R 2 = .01 Model: Nagelkerke R 2 = .13

Note: * p < .05 ** p < .01, *** p < .001

7.83 There are four key findings from this analysis. Firstly, the final model indicated a small but significant effect of age: for each additional year of chronological age, an individual is 4% less likely to have received their highest level of educational support from a mainstream school. Again, as mentioned above, it is possible that these results reflect historical changes in ASD awareness and provision.

7.84 The second key finding from the final model is that males were twice as likely to receive their highest level of educational support from a special unit in a mainstream school in comparison to females.

7.85 The third key finding that those with ADHD were 5.88 times more likely to receive their highest level of educational support from this type of school. However, this finding must be treated with some caution given the relatively small number of individuals in the sample with ADHD (n = 30), this is something that is reflected in the associated confidence intervals reported in Table 7.22.

7.86 The fourth key finding here was that those with diagnoses of depression were 4.3 times less likely to attend a special unit in a mainstream school. There are two main ways of interpreting this result, firstly that this simply reflects the fact that depression is more prevalent amongst individuals with higher functioning variations of ASD, and these higher functioning individuals are also less likely to attend special schools. However, this finding could also provide some support for the hypothesis that those attending special schools are more likely to receive the support they require and more likely to be surrounded individuals of a similar nature, and therefore individuals attending this type of school may be less susceptible to the development of mental health issues in comparison to those attending mainstream schools.

7.87 Finally of note here is that there was no evidence to suggest that type of ASD diagnosis or ID status had a significant influence on whether or not individuals received their highest level of educational support from this type of school – as indicated in Appendix C.5, neither of these factors were found to significantly predict whether or not an individual received their highest level of educational support from this type of school or contributed significantly to the variance explained by the model. This may provide some evidence to suggest that while those who are higher functioning are more likely to receive their highest level of educational support from a mainstream school (as shown in the analysis reported in Table 7.23), educational placement may vary more amongst those with lower functioning variations of ASD. No further analyses in relation to educational placements were possible due to the small numbers receiving their highest levels of educational support.

Educational Transitions

7.88 Analyses of the trajectories of ASD individuals in regard to educational placement can be helpful for policy and planning such provision, particularly in regard to additional support and specialised placements. Throughout the course of their school education, the majority (n = 268, 66%) of ASD individuals over the age of 16 in our sample had attended more than one type of school or received more than one level of educational support in school. With the data available it is not possible to establish whether changes occurred within a sector (e.g. pre-school, primary, or secondary) but only to account for whether there was a change across a sector (e.g. from a mainstream primary school to a specialist unit in a secondary school). Further, it was not possible to account for home-schooling or for individuals who reported receiving ‘other’ types of education to those specified in the questionnaire due to lack of data regarding the levels of support provided.

7.89 Our approach to educational trajectories was thus to focus on the differences in the educational provision providing the individual with the highest level of support across preschool, primary school and secondary sectors. Accordingly, analyses were based on the highest level of support received in each sector (see Table 7.15. for details of how the support intensity of these schools was ranked). The results of these analyses have been included in Tables 7.23 and 7.24.

7.90 As reported in Tables 7.23 and 7.24, the majority (71%) of individuals who attended a mainstream preschool also attended a mainstream primary school, and the majority (71%) of individuals who attended a mainstream primary school also attended a mainstream secondary school. Changes in type of placement were however more evident amongst those who initially attended additional support schools. Of those who received the greatest level of preschool support at an additional support school (including special units in mainstream schools, special day schools or residential school), 53% moved to a primary school associated with a lower level of support, 34% attended a school associated with a similar level of support to that they had received at preschool and 53% attended a school which provided a greater level of support than that provided at preschool level.

7.91 Turning to changes between primary and secondary school placements, again the majority (78%) of those attending a mainstream primary school also went on to also receive their highest level of support at secondary school at a mainstream school. As for those who attended an additional support primary school, 52% moved to a secondary school associated with a lower level of support, 42% attended a school associated with a similar level of support to that they had received at primary school and 6% attended a school which provided a greater level of support than that provided a primary school.

7.92 There are two key findings from this analysis. Firstly, of interest here is that only a relatively small number individuals were identified as moving to a school that would provide them with a greater level of educational support, particularly as it is known that so many individuals in the spectrum begin their education in mainstream establishments. However, this could be due to the fact that changes tend to happen within educational sectors, rather than across educational sectors. Also of interest here is the number and percentage of individuals who having initially attended a special school then went on to attend a school typically associated with a lower level of educational support, shown in Table 7.23.

Table 7.23 Changes in level of support provided at preschool and primary amongst ASD individuals ≥ 16 years (n = 319) [8]

Type of school attended at pre-school level n Changes in level of support received at primary school
Decrease (%) No Change (%) Increase (%)
Mainstream 231 12 (5) 164 (71) 55 (24)
Special Unit in Mainstream 50 27 (54) 14 (28) 9 (18)
Special Day School ( ASD specific) 8 3 (38) 4 (50) 1 (12)
Special Day School (General) 21 8 (38) 12 (57) 1 (5)
Residential School (General or ASD specific) 9 9 (100) 0 (0) 0 (0)
Total 319 59 (18) 194 (61) 66 (21)

Table 7.24 Changes in level of support provided at primary and secondary school amongst ASD individuals ≥ 16 years (n = 361) [9]

Type of school attended at pre-school level n Changes in level of support received at secondary school
Decrease (%) No Change (%) Increase (%)
Mainstream 229 14 (6) 178 (78) 38 (16)
Special Unit in Mainstream 61 30 (49) 26 (43) 5 (8)
Special Day School ( ASD specific) 18 9 (50) 9 (50) 0 (0)
Special Day School (General) 44 26 (59) 14 (32) 4 (8)
Residential School (General or ASD specific) 9 3 (33) 6 (67) 0 (0)
Total 361 82 (23) 233 (65) 47 (13)

7.93 However, in interpreting this analysis we do need to be considerate of the fact that this focussed only on individuals over ≥ 16 years. The advantages of focussing on thispopulation have already been discussed; however the disadvantage here is that by focussing on those who have previously completed their education, we are not necessarily presenting an accurate representation of the current educational experiences of those on the spectrum. With this in mind, this is an issue that should be explored in more detail in the future.

Further Education

7.94 Table 7.25 shows the number of ASD individuals aged 16 and over who had attended a further education establishment according to type of ASD diagnosis. In total, 131 (33%) of individuals in this subsample did not engage in further education; this included 55% of those with autism (n = 45), 21% of those with Asperger’s/ HFA (n = 50), and 44% of those with other ASD (n = 37).

7.95 The remaining 67% of individuals had attended at least one type of further education establishment. More specifically, of those with autism 33% had attended a further education college (n = 27), 5% had attended university (n = 4), and 9% had other types of further educational. Of those with Asperger’s/ HFA, 51% had attended a further education college (n = 121), 40% had attended university (n = 94), and 8% had attended another further education establishment (n =16). Finally, of those with other ASD, 44% had attended a further education college (n = 38), 13% had attended university (n = 11), and 5% had attended been involved in an alternative form of further education such as distance learning courses or night classes (n = 4).

Table 7.25 Attendance of further education establishments according to ID status amongst individuals ≥ 16 years


Further Educational Establishments Attended
Type of ASD diagnosis n (%) Total ≥16 years Sample n (%) (n = 401)
Autism (n = 81) Asperger’s/ HFA (n = 236) Other ASD (n = 85)
None 44 (55) 50 (21) 37 (44) 131 (33)
One or More 36 (42) 186 (79) 48 (56) 270 (67)
Further Education College 27 (33) 121 (51) 38 (44) 186 (46)
University 4 (5) 94 (40) 11 (13) 109 (27)
Other 7 (9) 16 (7) 4 (5) 27 (7)

* Note: The ‘One or more Further Educational Establishment’ row relates to the number of unique individuals who attended any of the further educational establishments listed. Individuals may be represented more than once in the last three rows of this table.

Qualifications

7.96 Table 7.26 shows the qualifications achieved according to type of ASD diagnosis (again this data relates to ASD individuals aged ≥ 16 years). In total, 22% of the 404 individuals had received no qualifications at all (n = 88); this included 46% of those with autism (n = 38), 9% of those with Asperger’s/ HFA (n = 22), and 33% of those with other ASD (n = 28).

Table 7.26 Qualifications achieved by individuals with ASD according to diagnosis

Highest Qualification Achieved Type of ASD diagnosis n (%) Total ≥16 years Sample n (%) (n = 404)
Autism (n = 82) Asperger’s n = 236) Other ASD (n = 86)
None 38 (46) 22 (9) 28 (33) 88 (22)
Access or National 1 and 2 15 (18) 5 (2) 13 (15) 33 (8)
Access or National 3, or Standard Grade Foundation 3 (4) 13 (6) 11 (13) 27 (7)
Standard Grade General/National 4/O-Grade or Intermediate 1 and Above 20 (4) 178 (75) 29 (34) 227 (56)
National 4, Standard General, O-Grade or Intermediate 1 4 (5) 16 (7) 10 (12) 30 (7)
National 5, standard Grade Credit, O-Grade or Intermediate 2 4 (5) 25 (11) 7 (8) 36 (9)
Highers, Certificate of Sixth year or Advanced Highers 2 (2) 36 (15) 2 (2) 40 (10)
Higher National or Educational Certificate or Diploma 5 (6) 34 (14) 4 (5) 43 (11)
Bachelors or Master’s Degree 1 (1) 22 (9) 4 (5) 27 (7)
Bachelors or Master’s Degree with Honours 3 (4) 24 (10) 1 (1) 28 (7)
Masters (post-graduate) 2 (2) 16 (7) 1 (1) 19 (5)
Doctoral Degree 0 (0) 3 (1) 0 (0) 3 (< 1)
Other 6 (7) 19 (8) 5 (6) 30 (7)
Total 82 (100) 236 (100) 86 (100) 404 (100)

7.97 Of the remaining individuals, 8% had achieved Access, or National 1 or 2 qualifications (including 18% of those with autism, 2% of those with Asperger’s/ HFA, and 1% of those with other ASD), 7% had achieved Access, National 3 or Standard Grade Foundation Grades (including 4% of those with Autism, 6% of those with Aspeger’s, and 13% of those with other ASD), and 56% had achieved either Standard Grade General or above Grades (including 4% of those with autism, 75% of those with Asperger’s/ HFA, and 34% of those with other ASD).

7.98 Chi-square analysis was used to compare the rates of individuals achieving standard grade general qualifications or above according to the type of ASD diagnosis they had. This revealed that there was a significant relationship between diagnosis and qualification achieved, X 2 (2, 374) = 69.68, p < .001 (this analysis excluded individuals with ‘other’ qualifications, n = 30). Partitioning the data (to compare the qualifications achieved by those with autism to others in the sample) revealed that those with Asperger’s and other ASD were 5.52 times more likely to achieve standard grade general qualification or above in comparison to those with autism.

Employment

7.99 Table 7.27 shows the number and percentage of individuals in the sample who were employed, in supported employment, or unemployed according to their ASD diagnosis. These statistics relate only to those over the aged ≥ 16 years (i.e. those who were older than the minimum age of full time employment).

Table 7.27 Employment status of individuals aged ≥ 16 years with ASD

Employment Status Type of ASD diagnosis n (%) Total ≥ 16 years Sample n (%) (n = 404)
Autism Asperger’s / HFA Other ASD
In Employment 15 (18) 83 (35) 14 (16) 112 (28)
In Supported Employment 2 (2) 9 (4) 2 (2) 13 (3)
Unemployed 65 (79) 144 (61) 70 (81) 279 (69)
Total 82 (100) 236 (100) 86 (100) 404 (100)

7.100 Overall, 28% of those over 16 (n = 404) were employed, 3% were in supported employment (n = 13) and 69% were unemployed (n = 279). Of those with autism (n = 82), 18% were employed (n = 15), 2% were in supported employment (n = 2) and 79% were unemployed (n = 65). Of those with Asperger’s (n = 236), 35% were in employment (n = 83), 4% were in supported employment (n = 9), and 61% were unemployed (n = 144). Of those with other ASD, 16% were in employment (n = 14), 2% were in supported employment (n = 2), and 81% were unemployed (n = 70).

7.101 The data from this analysis indicated that those with Asperger’s were more likely to be in employment in comparison to those with autism and other ASD. Therefore the data was partitioned to carry out a chi-square analysis comparing the employment status of those with Asperger’s to the employment status across the rest of the sample. As the number of individuals in supported employment was so low in comparison to those who were employed or unemployed, those in supported employment were grouped with those in employment for the purposes of this analysis. This chi-square analysis confirmed that there were significant differences between the employment of those with Asperger’s diagnoses in comparison to the rest of the sample X 2 (1, 404) = 17.18, p < .001. Odds ratio statistics were also calculated which indicated that those with Asperger’s were 2.61 times more likely to be in employment in comparison to the rest of the sample.

7.102 Overall these results fit with previous findings in this area which indicate that amongst those aged 16 and over the unemployment rate sits at between 25% and 50% (Cedurland et al., 2008, Helles et al., 2016, Howlin et al, 2004a). That said, the majority of studies in investigating this matter have focussed on a relatively small sample size of 70 or less, and this is investigation is one of the first to collect employment data from an ASD sample of this size.

7.103 Tables 7.28 and 7.29 show the differences in employment rates in those with autism and other ASD according to the presence and level of ID (there are no similar statistics for those with Asperger’s/ HFA as there were no recorded cases of ID within this subsample). In both cases this analysis found some evidence to indicate that a higher proportion of those without ID were in employment in comparison to those with ID although the small n should be noted.

7.104 Table 7.30 shows the employment status of individuals aged ≥ 16 years according to their age. There was some evidence to suggest that the number and percentage of individuals involved in employment was highest amongst those who were middle aged and lowest amongst the youngest and oldest individuals. These differences were explored further using chi-square analysis comparing employment rates across the different age groups (again, this analysis grouped those in employment and supported employment together).

7.105 Significant differences in employment were found when comparing employment rates amongst those aged 16-26 years and 27-49 years, X 2 (1, 404) = 19.17, p < .001, odds ratio indicating that those aged 16-26 years were 2.74 times less likely to be in employment in comparison to those who were middle aged. These results could indicate that even amongst those in this population who are capable of gaining and maintaining employment, it may take longer to find suitable employment.

7.106 Though data was only available from a small number of individuals ≥ 50 years, there also evidence to suggest that employment rates were similarly low amongst individuals in this age group, X 2 (1, 404) = 4.48, p < .05, odds ratio statistics indicated that those aged 50 or older were 2.45 times less likely to be in employment in comparison to those who were middle aged.

Table 7.28 Employment status amongst individuals aged ≥ 16 years with autism according to ID status

Presence and level of ID Employment Status n (%) Total ≥16 years Sample n (%) (n = 82)
In Employment (n = 15) In Supported Employment (n = 2) Unemployed (n = 65)
Autism + No ID 9 (25) 1 (3) 26 (72) 36 (100)
Autism + ID 6 (13) 1 (2) 39 (85) 46 (100)
Mild 1 (20) 0 4 (80) 5 (100)
Moderate & Severe 5 (12) 1 (2) 35 (85) 41 (100)

Table 7.29 Employment status amongst individuals aged ≥ 16 years with other ASD according to ID status

Presence and level of ID Employment Status n (%) Total ≥16 years Sample n (%) (n = 86)
In Employment (n = 14) In Supported Employment (n = 2) Unemployed (n = 70)
Other ASD + No ID 11 (20) 2 (3) 43 (77) 56 (100)
Other ASD + ID 3 (10) 0 (0) 27 (90) 30 (100)
Mild 1 (13) 0 (0) 7 (87) 8 (100)
Moderate & Severe 2 (9) 0 (0) 20 (91) 22 (100)

Table 7.30 Employment Status amongst individuals aged ≥ 16 years according to age

Age (years) Employment Status n (%) Total ≥16 years sample n (%) (n = 404)
In Employment In Supported Employment Unemployed
16 – 26 47 (21) 3 (1) 170 (77) 220 (100)
27 – 37 30 (39) 6 (8) 40 (53) 76 (100)
38 – 49 28 (38) 2 (3) 43 (59) 73 (100)
≥ 50 8 (22) 1 (3) 27 (75) 36 (100)
Total 113 (28) 12 (3) 280 (69) 405 (100)

7.107 Table 7.31 shows the employment status of individuals aged ≥ 16 years according to their sex. A total of 27% of males were in employment (n = 288), 2% were in supported employment (n = 6), and 70% were unemployed (n = 203). By comparison, 29% of females were in employment (n = 34), 6% were in supported employment (n = 7) and 65% were unemployed (n = 76). Chi-square analysis confirmed that these differences were non-significant, X 2 (1, 404) = 1.19, p > .05.

Table 7.32 shows the employment status of individuals aged ≥ 16 years according to the presence of co-occurring conditions (note: this excludes intellectual difficulties covered earlier in this section). From this analysis there appeared to be some evidence to suggest that those with co-occurring conditions were less likely to be in employment, however in the case of ADHD, OCD, epilepsy, and Tourette’ syndrome, chi-square analysis failed to show that these differences were significant (all X 2 values < 2, all p values > .05). There were too few individuals with schizophrenia for a statistical analysis, but all four were unemployed.

7.108 Similarly, no significant relationship was found between employment status (when full-time and supported employment were combined) and the presence of a mood disorder, X 2 (1, 404) = 2.75, p > .05. However, of interest here was the proportion of individuals in employment who experienced depression, which may be seen as high given that the prevalence of the condition across the general population is estimated at around 5% (Kessler et al., 2010). This finding would therefore provide some evidence to support the hypotheses that although employment may offer individuals on the spectrum with an opportunity to live independently and to socialise with others on a regular basis, it may not serve as a protective factor against the development of mental health issues, compared to the negative impact of unemployment.

Table 7.31 Sex differences in employment amongst individuals aged ≥ 16 years

Employment Status Sex Total ≥16 years sample n (%) (n = 404)
Male (n = 288) Female (n = 117)
In Employment 79 (27) 34 (29) 113 (28)
In Supported Employment 6 (2) 7 (6) 13 (3)
Unemployed 203 (70) 76 (65) 279 (69)
Total 288 (100) 117 (100) 405 (100)*

*Note that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

Table 7.32 Co-occurring neurological and mental health conditions and employment amongst individuals aged ≥ 16 years a

Employment Status ADHD OCD & Tourette’s Syndrome Epilepsy Schizophrenia Mood Disorders
Bipolar Depression Anxiety
In Employment 9 (30) 11 (22) 5 (17) 0 (0) 3 (33) 37 (41) 32 (33)
In Supported Employment 0 (0) 5 (10) 1 (3) 0 (0) 0 (0) 3 (3) 4 (4)
Unemployed 21 (70) 33 (67) 23 (79) 4 (100) 6 (67) 51 (56) 61 (63)
Total 30 (100) 49 (100) 29 (100) 4 (100) 9 (100) 91 (100) 97 (100)

a Percentages reported here are relative to the total number of individuals ≥ 16 years (n = 404)

Table 7.33 Employment status and ability to travel independently amongst individuals aged ≥ 16 years

Employment Status Ability to Travel Independently n (%) Total ≥16 years sample n (%) (n = 404)
Able Unable
In Employment 85 (38) 28 (15) 113 (28)
In Supported Employment 9 (4) 4 (2) 13 (3)
Unemployed 126 (57) 153 (83) 279 (69)
Total 220 (100) 185 (100) 405 (100) a

a The arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.109 Table 7.33 shows the number and percentage of individuals in employment according to their ability to travel. There was evidence from the data to indicate that those who were able to travel were more likely to be in employment and this relationship was investigated further using chi-square analysis (again this analysis combined those who were in employment and supported employment). The analysis revealed that these differences were significant, X 2 (1, 404) = 27.55, p < .001, and follow up odds ratio statistics indicated that participants who were able to travel independently were 3.57 times more likely to be in employment in comparison to those who were unable to travel.

7.110 Table 7.34 shows the number and percentage of individuals in employment according to the highest level of educational support they received. Of most interest here is that the percentage of individuals in employment was fairly consistent across the different types of school providing individuals with their highest level of educational support with the exception of special units within mainstream schools. Chi-square analysis confirmed that individuals who received their highest level of educational support at a special unit within a mainstream school were less likely to be in employment, X 2 (1, 404) = 4.34, p < .05, with odds ratio statistics indicating that individuals attending this type of school were 1.75 times less likely to be employed.

Table 7.34 Employment status according to school providing individual with highest level of educational support amongst individuals aged ≥ 16 years

School providing highest level of educational support In Employment In Supported Employment Unemployed Total ≥16 years sample n (%) (n = 404)
Mainstream School 60 (32) 6 (3) 120 (65) 186 (100)
Special Unit in a Mainstream School 18 (18) 4 (4) 76 (78) 98 (100)
Special ASD Day School 8 (28) 0 (0) 21 (72) 29 (100)
Special Day School (Other) 21 (34) 2 (3) 39 (63) 62 (100)
Special Residential School 5 (17) 1 (3) 23 (79) 29 (100)
Total 112 (28) 13 (3) 279 (69) 404 (100)

7.111 Table 7.35 reports the number and percentage of ASD individual ≥ 16 years according to the qualifications they had achieved throughout their education. There was some evidence from the data to suggest that the likelihood of employment increased according to the level of qualification that an individual achieved. Chi-square analysis was run in order to test whether this relationship was significant. After running a series of chi-square analyses the most significant difference was found in the employment status of those who had achieved above and below standard grade general qualifications, X 2 (1, 404) = 15.18, p < .001.

Table 7.35 Employment status amongst individual ≥ 16 years, according to qualifications achieved

Highest Qualification Achieved Type of ASD diagnosis n (%) Total ≥16 years sample n (%) (n = 404)
In Employment In Supported Employment Unemployment
None 14 (16) 3 (3) 72 (81) 89 (100)
Access or National 1 and 2 4 (13) 1 (3) 27 (84) 32 (100)
Access or National 3, or Standard Grade Foundation 6 (22) 1 (4) 20 (74) 27 (100)
Standard Grade General/National 4 or Intermediate 1 and above 79 (35) 5 (2) 143 (63) 228 (100)
National 4, Standard General, or Intermediate 1 8 (28) 1 (3) 20 (69) 29 (100)
National 5, standard Grade Credit, or Intermediate 2 9 (24) 0 (0) 28 (76) 37 (100)
Highers, Certificate of Sixth year or Advanced Highers 10 (25) 1 (3) 29 (73) 40 (100)
Higher National or Educational Certificate or Diploma 13 (31) 1 (2) 28 (67) 42 (100)
Bachelors or Master’s Degree 11 (41) 1 (4) 15 (56) 27 (100)
Bachelors or Master’s Degree with Honours 17 (59) 0 (0) 12 (41) 29 (100)
Masters (post-graduate) 8 (44) 0 (0) 10 (56) 18 (100)
Doctoral Degree 2 (67) 0 (0) 1 (33) 3 (100)
Other 11 (37) 2 (7) 17 (57) 30 (100)
Total 114 (28) 12 (3) 279 (69) 406 (100)

a The arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

Predictors of Employment

7.112 Binary logistic regression analysis was used to identify the factors which predicted the likelihood an individual being in employment. As with other analysis in this section those in supported employment (n = 13) were grouped with those who were in full time employment (n = 112) and compared to those who were unemployed (n = 279). Exploratory analysis was carried out to identify candidate variables (listed in Appendix C.6.) which were added to a hierarchical model in the following five blocks: (i) those relating to demographics, (ii) those relating to core diagnoses, (iii) those relating to co-occurring conditions, (iv) those relating to other outcomes and (v) those relating to service-use.

7.113 As before, the final model shown in Table 7.36 reports only those candidate variables which improved the associated Nagelkerke R 2 by at least .02. Candidate variables excluded from the final model in this way and relevant statistics are detailed in Appendix C.6.

Table 7.36 Logistic regression analysis testing the factors predicting employment status amongst ASD individuals aged ≥ 16 years [10]

Model β SE β Wald χ 2 Df Exp β
Odds-Ratio Lower Upper
Block 1






Aged 27 – 49 *** 1.46 .28 32.59 1 4.33 2.62 7.17
Block: Nagelkerke R 2 = .13
Block 2






Aged 27 – 49 *** 1.33 .29 24.20 1 3.81 2.23 6.49
Ability to Travel*** 2.06 .38 35.40 1 7.95 4.01 15.75
Block: Nagelkerke R 2 = .16 Model: Nagelkerke R 2 = .29
Block 3






Aged 27 – 49 *** 1.33 .30 23.22 1 3.81 2.21 6.55
Ability to Travel*** 1.92 .38 29.57 1 6.87 3.43 13.77
Relationship Status*** .95 .31 9.85 1 2.59 1.43 4.69
Block: Nagelkerke R 2 = .03 Model: Nagelkerke R 2 = .32

Note: * p < .05 ** p < .01, *** p < .001

7.114 The model does not contain reference to type of ASD diagnosis or ID status, though both were considered as part of the development of the model. As described in Appendix C.6, Asperger’s/ HFA was identified as the strongest predictor of employment amongst the three main types of diagnosis, and depression was also found to be a stronger predictor of employment than mood disorders in general, however as part of a more complex model these factors were found to be highly non-significant and unreliable predictors (see Appendix C.6 for more details).

7.115 In Block 1 of the model, the factor of age was entered. Initially age was entered as a continuous variable, and was not found to be significant predictor. However, given that there was evidence from the raw data and follow-up chi-square analysis to indicate that those aged between 26 and 49 were more likely to be employment, this age group was included in the model instead. The logistic regression analysis indicated that those who were in this ‘middle-aged’ group were 3.81 times more likely to be in employment in comparison to the rest of the sample, and this variable accounted for 13% of the variance, X 2 (1, 398) = 35.81, p < .001.

7.116 In Block 2 of the analysis ‘ability to travel’ was added to the regression model and made a significant contribution to the null model, X 2 (1, 398) = 14.90, p < .001, and increased the variance explained by the model by 16% (Nagelkerke R 2 for this block = .29).

7.117 Finally, in Block 3, ‘relationship status’ was added to the regression model and made a further significant contribution to the model, X 2 (1, 398) = 6.45, p < .05, and increased the variance explained by the model by a further 3% (Nagelkerke R 2 for this block = .29

7.118 Of greatest interest, here are the first two variables included in the model, each of which explained around 15% of the variance in those who were and were employed and unemployed. As indicated in relation to the raw data and follow-up chi-square analysis there was evidence to suggest that in our sample there was a relationship between age and employment status. More specifically, the regression analysis indicated that those who were middle aged were 3.81 times more likely to be in employment in comparison to those under the age of 26 or over the age of 50, indicating that the youngest and oldest individuals in the ASD population were more likely to struggle to find and maintain employment.

7.119 The second variable of interest was ‘ability to travel’ which relates to individual’s ability to travel independently. The analysis indicated that those capable of travelling independently were 6.87 times more likely to be in employment (note: while this result was associated with relatively broad confidence intervals, the magnitude of the lower confidence interval indicated that individuals in this population would be at least 3 times more likely to be in employment if they could travel independently).

7.120 The final factor in this model, relationship status, was also found to be associated with employment status in that those involved in a long-term relationship were 2.59 times more likely to be in employment in comparison to the rest of the sample. This result could be interpreted in one of two ways. Firstly it may simply indicate that characteristics and skills that enable someone to engage in and maintain a long term relationship may be the same as those which increase the likelihood of employment. The second interpretation here could be that involvement in a long-term relationship provides a level of support which helps an individual gain and maintain employment.

7.121 Finally, of interest here, qualifications were not found to be a significant predictor of an individual’s employment status. In modelling the factors that predicted employment our team considered those with no qualifications, those with above and below the standard grade general level of qualification, those above and below the certificate of sixth year studies, higher or advanced higher level of qualification, and finally those with and without university or college degrees. In each of these cases there was no evidence to suggest that inclusion in one of these categories increased the likelihood that an individual would be employed or unemployed.

Relationships

7.122 At the time they completed the survey, 18% of the 404 individuals aged ≥ 16 years, were involved in a long-term relationship which had lasted 2 years or longer, as shown in Table 7.37 (information about relationships was only collected in relation to ASD individuals aged ≥ 16 years).

7.123 Overall, 72 individuals within the sample were involved in long-term relationships and as shown in Table 7.37 there was evidence from the raw data to suggest that long-term relationships were more prevalent amongst those with Asperger’s/ HFA. This matter was investigated further through the use of chi-square analysis which partitioned the data, comparing the rates of long-term relationships amongst those with Asperger’s/ HFA to the rates across the rest of the sample, and these results indicated that these rates differed considerably, X 2 (2, 404) = 29.20, p < .001. Follow-up odds ratio statistics indicated that within our sample those with Asperger’s/ HFA were 5.29 times more likely to be involved in a relationship in comparison to the rest of the sample.

Table 7.37 Relationship status amongst ASD individuals aged ≥ 16 years according to type of diagnosis.

Relationship Status a ASD diagnosis n (%) Total ≥16 years sample n (%) (n = 404) b
Autism Asperger’s/ HFA Other ASDs
In a long-term relationship 5 (6) 62 (27) 5 (6) 72 (18)
Not in a long-term relationship 78 (94) 173 (73) 81 (94) 332 (82)
Total b 83 (100) 235 (100) b 86 (100) 404 (100)

a Long-term relationships here were defined as relationships lasting ≥ 2 years; bNote that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.124 While the issue of long-term relationships is something that has previously been covered in the literature, most investigations in this area have either focussed specifically on those with Asperger’s or else have investigated this matter using relatively small sample sizes. Of the research focussing on Asperger’s, findings have tended to indicate that between 30% and 50% of individuals are involved in long-term relationships (e.g. Helles, Gillberg, Gillberg & Billstedt, 2017; Strunz, Schermuck, Ballerstein, Ahlers, Dziobek & Roepke, 2017) – a rate markedly different from our own. One study which did focus on a somewhat more representative ASD sample was carried out by Eaves and Ho (2008), and found a much lower rate of long-term relationship involvement in their sample, with only 10% of the 48 individuals included in their sample reporting being involved in long-term relationship (this sample included 26 individuals with autism, hence the study focussed on a sample that was much lower-functioning overall in comparison to the research described above). Therefore, while in comparison to the pre-existing literature we report lower rates of long-term relationships amongst those with Asperger’s/ HFA, our findings to comply with the overall trends in the ASD literature which indicate that involvement in long-term relationships is associated with the type of the severity and type of symptoms an individual has. It is worthy of note that the raw data relating to long-term relationship status and ID status revealed that only 1 individual with ID was involved in a long-term relationship. This is compatible with the outcome literature (Howlin et al., 2004).

7.125 Table 7.38 shows the age distribution of the individuals who were involved in a long-term relationship according to their ASD diagnosis. As might be expected, the data collected suggested that the percentage of ASD individuals involved in a relationship is a figure which increases with age, indicating that, as with employment, long-term relationships may be something that those on the spectrum are less likely to engage in until they are slightly older.

Table 7.38 Long-term relationship status amongst ASD individuals aged ≥ 16 years according to age

Relationship Status Age Group n (%) Total ≥16 years sample n (%) (n = 404)
16 – 26 27 – 37 38 – 49 ≥ 50
In a long-term relationship 14 (6) 17 (22) 23 (32) 19 (53) 73 (18)
Not in a long-term relationship 206 (94) 59 (78) 50 (68) 17 (47) 332 (82)
Total 220 (100) 76 (100) 73 (100) 36 (100) 405 (100) a

a The arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.126 Table 7.39 reports the number and percentage involved in relationships according to their sex. Chi-square analysis confirmed that these differences were non-significant, X 2 (1, 404) = 1.32, p > .05.

Table 7.39 Sex differences in relationship status amongst ASD individuals aged ≥ 16 years

Relationship Status Sex n (%) Total ≥16 years sample n (%) (n = 404)
Male Female
In a long-term relationship 45 (16) 28 (24) 73 (18)
Not in a long-term relationship 243 (84) 89 (76) 332 (82)
Total 288 (100) 117 (100) 405 (100)

Table 7.40 Long-term relationship status amongst individuals aged ≥ 16 years and co-occurring conditions

Relationship Status ADHD OCD & Tourette’s Epilepsy Schizophrenia Mood Disorders
Bipolar Depression Anxiety
In a long-term relationship 9 (30) 6 (14) 2 (7) 1 (25) 4 (44) 35 (38) 25 (26)
Not in a long-term relationship 21 (70) 36 (86) 27 (93) 3 (75) 5 (56) 56 (62) 72 (74)
Total 30 (100) 42 (100) 29 (100) 4 (100) 9 (100) 91 (100) 97 (100)

a Percentages reported here are relative to the total number of individuals ≥ 16 years (n = 404)

Table 7.41 Long-term relationship status amongst ASD individuals aged ≥ 16 years according to highest level of educational provision

School Provision

In Long-term Relationship

Not in Long-Term Relationship

Total ≥16 years sample n (%)

(n = 404)

Mainstream School

51 (70)

135 (41)

186 (46)

Special Unit in a Mainstream School

6 (8)

92 (28)

98 (24)

Special ASD Day School

9 (12)

52 (16)

61 (15)

Special Day School (Other)

2 (3)

27 (8)

29 (7)

Special Residential School

4 (5)

25 (8)

29 (7)

Home Educated

0 (0)

1 (0)

1 (0)

Total

73 (100)

332 (100)

405 (100) a

a The arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.127 Table 7.40 shows the long-term relationship status of individuals in our sample according to the presence of co-occurring conditions. Again, the small n associated with the majority of the conditions described in this table meant that it was not possible to draw any strong and reliable inferences about the influence of these conditions on the likelihood of an individual being involved in a long-term relationship.

7.128 That said, it was notable that almost one-third of those with ADHD were involved in long-term relationships, indicating that this diagnosis in combination with an ASD diagnosis does not preclude an individual from being involved in a long term relationship. In contrast, around 90% of those with OCD and epilepsy were not involved in long-term relationships, indicating that these conditions may have more of a negative impact on an individual’s ability to engage in and maintain a relationship. Also of note here is the number of individuals involved in long-term relationships who also had a diagnosis of a mood disorder, indicating that while often loneliness and social isolation may be at the root of these conditions amongst individuals with ASD, ASD individuals may experiences these symptoms even when they are involved in close social relationships.

7.129 Table 7.41 shows the long-term relationship status of participants according to the highest support school they attended. Of most interest here is that 70% of those involved in a long-term relationship received their highest level of educational support from a mainstream school. This provides some evidence to suggest that the majority of those who are involved in long-term relationships are the individuals with the least severe social, communication and intellectual difficulties. While a minority of individuals attending other special schools were involved in long-term relationships, another finding of interest here was that around 15% of those attending special ASD day schools were involved in relationships, which given that these schools typically provide services for individuals with greater needs, may indicate that there is a long-term benefit (in terms of relationships) of an individual attending a school which caters to individual’s with similar needs to their own.

Table 7.42 Long-Term Relationship Status amongst individuals aged ≥ 16 years according to employment status

Relationship Status Employment Status n (%) Total ≥16 years sample n (%) (n = 404)
In Employment In Supported Employment Unemployed
In a long-term relationship 36 (49) 1 (2) 36 (49) 73 (100)
Not in a long-term relationship 77 (23) 11 (4) 244 (73) 332 (100)
Total 112 (28) 13 (3) 280 (69) 405 (100) a

a Note that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.130 Table 7.42 shows the number and percentage of individuals involved in long-term relationships according to their employment status. There was evidence to suggest that a greater number of employed individuals were involved in long-term relationships in comparison to those who were unemployed; therefore these were tested using chi-square analysis (again the categories of employment and supported employment were collapsed for the purposes of this analysis). These differences were found to be significant, X 2 (1, 404) = 19.23, p < .001, and follow-up odds ratio statistics indicated that those in employment were 2.97 times more likely to be in a long-term relationship in comparison to those who were unemployed. This may provide some evidence to suggest that (a) within this population individuals are more likely to be involved in a relationship if they are able to support themselves financially and live independently (a) matter explored further in the next section of this chapter), and (b) some individuals on the spectrum may struggle to form close relationships simply as a result of missing out on the social opportunities that are available in the work place.

Predictors of Relationship Status

7.131 Binary logistic regression analysis was used to identify the factors which predicted the likelihood an individual being long-term relationship. As with other analyses in this section exploratory analysis was carried out to identify candidate variables (listed in Appendix C.7.) which were added to a hierarchical model in the following five blocks: (i) those relating to demographics, (ii) those relating to core diagnoses, (iii) those relating to co-occurring conditions, (iv) those relating to educational, health and social variables and (v) variables relating to service-use.

7.132 As before, the final model shown in Table 7.43 reports only those candidate variables which improved the associated Nagelkerke R 2 by at least .02. Candidate variables excluded from the final model in this way and relevant statistics are detailed in Appendix C.7.

7.133 In block one of the model age was introduced, and identified as a significant predictor, X 2 (1, 398) = 85.60, p < .001, which could account for 35% of the variance in individuals who were an were not engaged in long-term relationships.

7.134 In block two of the model, depression was added, and again this was found to be a significant predictor, X 2 (1, 384) = 17.22, p < .001, which could explain a further 6% of the variance in the data.

7.135 Finally in block three of the model employment status was introduced. This was also found to be a significant predictor of relationships status, X 2 (1, 384) = 22.56, p < .001, and explained 19% of the variance in the data, raising the total variance explained by the model to 49%.

7.136 There were three main findings from this regression analysis. The first of these was that for every year older an individual was they were 1.12 times more likely to be involved in a long-term relationship. This provides further support for the idea proposed earlier in this section that even those on the spectrum who experience positive social outcomes may experience them at a later stage in life in comparison to typically developing individuals.

Table 7.43 Logistic regression analysis testing the factors predicting relationship status amongst individuals with ASD aged ≥ 16 years [11]

Model β SE β Wald χ 2 df Exp β
Odds-Ratio Lower Upper
Block 1






Age*** .10 .01 61.73 1 1.11 1.08 1.14
Block: Nagelkerke R 2 = .35
Block 2






Age*** .10 .01 50.49 1 1.10 1.07 1.12
Depression*** 1.45 .35 17.24 1 4.28 2.13 8.57
Block: Nagelkerke R 2 = .06 Model: Nagelkerke R 2 = .41
Block 3






Age*** .11 .02 30.77 1 1.12 1.08 1.15
Depression*** 1.28 .38 5.06 1 3.61 1.73 7.52
Employment Status*** 1.77 .41 9.22 1 5.84 2.64 12.94
Block: Nagelkerke R 2 = .19 Model: Nagelkerke R 2 = .49

Note: * p < .05 ** p < .01, *** p < .001

7.137 The second key finding here was that individuals with depression were 3.61 times more likely to be involved in a long-term relationship in comparison to the rest of the sample. In interpreting this result, it is first important to acknowledged the relatively broad confidence intervals associated with this finding, indicating that this finding should be treated with some caution. However, this point aside, while this finding may at first appear counter-intuitive, it is most likely that it reflects the number of high functioning individuals with mental health issues, as it is these high functioning individuals who, in comparison to the rest of the spectrum, are the most likely to be involved in long-term relationship.

7.138 Finally, this analysis provided evidence to suggest that an individual’s relationship status may be underpinned by their employment status, as those in employment were 5.84 times more likely to be involved in a long-term relationship in comparison to the rest of the sample (though again this result should be treated with some caution given the range of confidence intervals associated with this analysis). This result may be seen to give support to the hypothesis that (a) individuals in this population are more likely to be involved in relationships if they are financially independent and (b) that being in employment may give an individual the opportunity to socialise and meet with people with whom they could engage in a relationship.

Residential Status

7.139 Table 7.44 shows the residential status of participants. In total 87% (n = 352) lived in a private household (with their parents, partners, friends or on their own), while a further 8 lived in supported accommodation (n = 32), and 5% lived in another form of accommodation (n = 20; e.g. some in this category were students at residential schools or and others were in hospital accommodation).

Table 7.44 Residential status of ASD individuals aged ≥ 16 years (n = 404)

Residential Status n (%)
In Private Household 352 (87)
With Parents 226 (56)
With Partner or Friends 55 (14)
Alone 71 (18)
In Supported Living 32 (8)
Other b 20 (5)

a Percentages reported here are relative to the total number of individuals ≥ 16 years (n = 404) b Includes individuals staying in hospital accommodation, or attending residential school

7.140 The primary interest in this data was to establish the number and percentage of ASD individuals who were living independently from their parents. Therefore the data shown in Table 7.42 was re-categorised to group together those who were living independently in this way and those who were living in a situation where they were mood disorder supported by someone else (those in the ‘other’ category above were not included in this further analysis due to a lack of information regarding the day to day support provided/available to these individuals; this resulted in all subsequent analysis being based on 384 adults rather than 404). These adjusted categories, described in Table 7.45, were subsequently used to explore the data relating to residential status further.

Table 7.45 Re-categorisation of residential status

Living independently Not living independently
Individuals living alone Individuals living with parents
Individuals living with partner or friends Individuals in supported accommodation

7.141 Table 7.46 shows the number and percentage of individuals living independently according to their ASD diagnosis. Evidence from the data indicated that those with Asperger’s/ HFA were more likely to live independently in comparison to the rest of the sample and chi-square analysis confirmed that this difference was significant, X 2 (1, 386) = 36.79, p < .001. Follow-up odds ratio statistics also indicated that in comparison to the rest of the sample, those with Asperger’s/ HFA were 4.58 times more likely to live independently in comparison to the rest of the sample. This is consistent with the outcomes literature (see Howlin et al., 2004).

Table 7.46 Residential status of ASD individuals aged ≥ 16 years according to type of ASD diagnosis

Residential Status ASD Diagnosis n (%) Total ≥16 years sample n (%) (n = 386)
Autism Asperger’s/ HFA Other ASDs
Living independently 11 (15) 103 (45) 12 (15) 126 (33)
Not living independently 62 (85) 128 (55) 69 (85) 259 (67)
Total 73 (100) 231 (100) 81 (100) 385 (100) a

a Complete data on residential status was available for 385 of the 404 adults with ASD.

7.142 Table 7.47 shows the number and percentage of individuals living independently according to the level and presence of ID. Of most interest here is that only 4% (n = 3) of those with mild or moderate/severe ID were living independently with all other individuals with a diagnosis of ID either in supported accommodation or else living with their parents or guardians (see Howlin et al., 2004). Differences in the number of individuals with and without ID who lived independently were confirmed as significant by chi-square analysis, X 2 (1, 386) = 37.14, p < .001. Follow-up odds ratio statistics indicated that those without ID were 14.2 times more likely to be living independently in comparison to those with ID.

Table 7.47 Residential status of ASD individuals aged ≥ 16 years (n =386) according to ID status and level.

Residential status ID status and level n (%) Total ≥16 years sample n (%) (n = 386)
No ID (n = 328) ID
Mild (n = 15) Moderate/Severe (n = 62) Total (n = 77)
Living independently 124 (39) 1 (7) 2 (4) 3 (4) 127 (33)
Not living independently 192 (61) 14 (93) 52 (96) 66 (96) 258 (67)
Total 316 (100) 15 (100) 54 (100) 69 (100) 385 (100) a

a Complete data on residential status was available for 385 of the 404 adults with ASD.

7.143 Table 7.48 shows the number and percentage of individuals living independently according to their age. Most notable here is that considerably fewer individuals in the 16 – 26 age bracket were living independently in comparison to older individuals (while this may be expected to some extent, as many typically developing individuals live with their parents until their mid-twenties, follow-up analysis focussing on a slightly older age group of 22-26 revealed similar results, in that only 22% of those within this age range were living independently). Follow-up chi-square analysis indicated that these differences were significant, X 2 (1, 386) = 103.98, p < .001, and odds ratio statistics confirmed that those aged 16 – 26 were 12.26 times less likely to be involved in a long-term relationship in comparison to the rest of the sample. As with other findings in this chapter, these results provide some evidence to suggest even amongst those on the spectrum who are capable of achieving positive life outcomes, in comparison to those in the typically developing population these positive outcomes are likely to be achieved later in life.

Table 7.48 Residential status amongst ASD individuals aged ≥ 16 years (n = 386) according to age

Residential Status Age Group n (%) Total ≥16 years sample n (%) (n = 386)
16 – 26 27 – 37 38 – 49 ≥ 50
Living independently 22 (11) 39 (53) 41 (59) 26 (72) 128 (33)
Not living independently 186 (89) 34 (47) 28 (41) 10 (28) 258 (67)
Total 208 (100) 73 (100) 69 (100) 36 (100) 386 (100) a

a Complete data was available for 385 of the 404 adults with ASD, however, arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.144 Table 7.49 shows the sex differences in the number of ASD individuals who lived independently. Chi-square analysis confirmed that these differences were not significant, X 2 (1, 386) = 3.09, p > .05.

Table 7.49 Sex differences in residential status amongst ASD individuals aged ≥ 16 years

Residential Status Sex n (%) Total ≥16 years sample n (%) (n = 386)
Male Female
Living independently 82 (30) 47 (41) 129 (33)
Not living independently 189 (70) 69 (59) 258 (67)
Total 271 (100) 116 (100) 387 (100) a

a Complete data was available for 385 of the 404 adults with ASD, however, arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.145 Table 7.50 shows the number and percentage of ASD individuals who were living independently according to the presence of co-occurring conditions. There was some evidence to suggest that the presence of these conditions could have an influence on the likelihood of an individual living independently. However, again chi-square analysis did not reveal significant differences in residential status in relation to ADHD, OCD/Tourette’s, Epilepsy and Schizophrenia (all X 2 values < 2, all p values > .05). Significant differences in residential status were however identified amongst those with and without depression, X 2 (1, 386) = 43.64, p < .001, with follow-up odds ratio statistics indicating that those living independently were 5.05 times more likely to experience depression [12] . There was no evidence to suggest that amongst those who were living independently the prevalence of depression diagnoses differed significantly between those who lived independently, and those who lived with friends or a partner, X 2 (1, 128) = .95, p > .05.

7.146 There are two potential interpretations of these results. On the one hand, these results, could reflect the fact that those with high functioning variations are both more likely to have diagnoses of depression and more likely to be living independently from their parents in comparison to others on the spectrum. However, on the other hand, this finding may also indicate that those who live independently are more likely to experience depression due to the difficulties they experience in everyday life – this is an issue discussed in more detail in relation to the logistic regression analysis reported at the end of this chapter.

Table 7.50 Residential status amongst individuals aged ≥ 16 years and co-occurring conditions

Residential Status ADHD OCD/ Tourette’s syndrome Epilepsy Schizophrenia Mood Disorders
Bipolar Depression Anxiety Mood Disorder Total
Living independently 11 (40) 23 (66) 6 (21) 1 (33) 5 (63) 55 (62) 47 (50) 70 (53)
Not living independently 17 (60) 12 (34) 23 (79) 2 (67) 3 (37) 34 (38) 47 (50) 62 (47)
Total 28 (100) 35 (100) 29 (100) 3 (100) 8 (100) 89 (100) 94 (100) 132 (100)

a Complete data was available for 385 of the 404 adults with ASD, the percentages reported here are relative to the number of available data

Table 7.51 Residential status amongst individuals aged ≥ 16 years (n = 386) according to employment status a

Residential Status Employment Status n (%) Total ≥16 years sample n (%) (n = 384) b
In Employment In Supported Employment Unemployed
Living independently 51 (40) 8 (6) 67 (53) 126 (100)
Not living independently 59 (23) 4 (2) 195 (76) 258 (100)
Total 110 (29) 12 (3) 262 (68) 384 (100)

a Complete data was available for 385 of the 404 adults with ASD, however, arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.147 Table 7.51 shows the number of individuals living independently and dependently according to their employment status. There was evidence to suggest that a greater proportion of those living independently were also in employment, X 2 (1, 386) = 13.08, p < .001, and that those in employment were 2.72 times more likely to be living independently. While these findings are to some extent to be expected, this does provide some evidence to suggest that being able to gain and maintain employment is an outcome which underpins an individual’s overall ability to live independently without support from parents, carers or professionals.

Table 7.52 Residential and relationship status amongst ASD individuals aged ≥ 16 years a

Residential Status Relationship status n (%) Total ≥ 16 years sample n (%)
(n = 384) b
In a long-term relationship Not in a long-term relationship
Living independently 56 (81) 70 (22) 126 (33)
Not living independently 13 (19) 245 (78) 258 (67)
Total 69 (100) 315 (100) 384 (100)

a Complete data was available for 385 of the 404 adults with ASD, however, arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.148 Table 7.52 shows the number of individuals living independently and dependently according to their relationship status. There was evidence to suggest that a greater proportion of those living independently were also in a long-term relationship X 2 (1, 386) = 90.83, p < .001, and that those in relationships were 15.05 times more likely to be living independently. This finding could indicate one of two things. Firstly it may indicate that in comparison with those who are unemployed, those who are in employment are more likely to socialise with others on a day-to-day basis and as a result are more likely to encounter individuals with whom they can develop long-term relationships. However, again it is possible to hypothesise that these results are indicating that those in employment are most likely to also be individuals who are higher functioning and have fewer social impairments, and naturally these individuals are also more likely to be involved in long-term relationships for this reason. This is an issue that has been explored in more detail in relation to the main logistic regression analysis reported in this chapter.

Predictors of independent residential status

7.149 Binary logistic regression analysis was used to identify the factors which predicted the likelihood an individual living independently, either on their own or with friends or a partner. As with other analyses in this section exploratory analysis was carried out to identify candidate variables (listed in Appendix C.8.) which were added to a hierarchical model in the following five blocks: (i) those relating to demographics, (ii) those relating to core diagnoses, (iii) those relating to co-occurring conditions, (iv) those relating to other outcomes and (v) those relating to service-use.

7.150 As before, the final model shown in Table 7.53 reports only those candidate variables which improved the associated Nagelkerke R 2 by at least .02. Candidate variables excluded from the final model in this way and relevant statistics are detailed in Appendix C.8.

7.151 Block 1 of the model introduced age as a predictor of residential status, and this analysis revealed that for each additional year of age individuals in the sample were 11% more likely to be living independently, X 2 (1, 378) = .119.32, p < .001. This variable alone explained 38% of the variance in the data.

7.152 Block 2 of the model introduced ‘Mood disorder diagnosis’, which significantly explained a further 8% of the variance in the data, X 2 (1, 378) = .28.74, p < .001 (raising the total variance explained to 46%). It should be noted that there were similar variables, namely ‘depression diagnosis’ and ‘anxiety diagnosis’ which were also found to be significant predictors of residential status, however in this case ‘mood disorder’ diagnosis was selected as the Wald value associated with each of the factors was fairly similar, but ‘mood disorder’ diagnosis was the term that applied to the broadest number of individuals.

7.153 Block 3 of the model added in ‘ability to travel independently’, which significantly explained a further 10% of the variance, raising the total variance explained to 56%, X 2 (1, 378) = .45.35, p < .001

Table 7.53 Logistic regression analysis testing the factors predicting residential status amongst individuals with ASD aged ≥ 16 years [13]

Model β SE β Wald χ 2 df Exp β
Odds-Ratio Lower Upper
Block 1






Age*** .11 .01 82.90 1 1.11 1.09 1.14
Block: Nagelkerke R 2 = .38
Block 2






Age*** .11 .01 75.13 1 1.11 1.09 1.14
Mood Disorder Diagnosis*** 1.49 .29 27.13 1 4.43 2.51 7.82
Block: Nagelkerke R 2 = .08 Model: Nagelkerke R 2 = .46
Block 3






Age*** .10 .01 56.18 1 1.10 1.08 1.13
Mood Disorder Diagnosis*** 1.33 .31 18.34 1 3.76 2.03 6.97
Ability to travel independently*** 2.27 .41 34.47 1 9.64 4.33 21.46
Block: Nagelkerke R 2 = .10 Model: Nagelkerke R 2 = .56
Block 4






Age*** .09 .02 42.16 1 1.10 1.07 1.13
Mood Disorder Diagnosis*** 1.12 .37 53.97 1 3.08 1.59 5.94
Ability to travel independently*** 2.10 .43 39.11 1 8.20 3.56 18.89
Relationship status *** 2.05 .45 50.38 1 7.77 3.18 18.93
Block: Nagelkerke R 2 = .06 Model: Nagelkerke R 2 = .62

Note: * p < .05 ** p < .01, *** p < .001

7.154 Finally, block 4 of the model entered ‘relationship status’ into the model, which related to those who were and were not in a long-term relationship lasting 2 years or longer. This predictor was also found to be significant, X 2 (1, 378) = .26.28, p < .001, and could explain a further 6% of the variance, raising the total variance explained by the model to 62%.

7.155 There are four key findings from this analysis. The first was that for every year older an individual in our sample was they were 10% more likely to be living on their own or with a partner or friend. Again this provides further evidence to suggest that even those who experience positive life outcomes are likely to experience them at an older age.

7.156 The second key finding was that those with mood disorders were three times more likely to live independently in comparison to those without mood disorders. This finding could be interpreted in two ways. Firstly, again this may simply reflect the fact that those with mood disorders tend to be higher functioning, and it is also those who are higher functioning who tend to be capable of living independently without support. However, this finding could also indicate that rates of mood disorders are higher amongst those who live independently as they struggle to cope with the everyday stresses associated with a condition like ASD and are in need of support with some aspect of their life.

7.157 The third key finding is that those who are able to travel independently were over eight times more likely to be living independently in comparison to those who were not able to travel independently. Overall this result provides some evidence to support the hypothesis that many on the spectrum may live with parents or carers because of the stress, anxiety, or challenges associated with travelling independently.

7.158 The fourth key finding was that those who were involved in long-term relationships were over seven times more likely to be living independently. This is a result that makes which underpins the fact that those engaged in long-term relationships will live together and therefore independently.

Independent Living

7.159 One of the reasons for collecting data on adult outcomes in this population is to allow us to learn more about the number and percentage of individuals on the spectrum living independently. Therefore while the majority of analysis so far has concentrated on the influences on specific adult outcomes, the data analysis in Table 7.53 brings together data relating to the some of the outcomes considered of most relevance to independent living, namely, ability to travel independently, employment status and residential status.

7.160 The data presented here highlight a number of points for consideration, some of which have previously been covered in this chapter. First and foremost this data clearly indicated that only a relatively small proportion of the adult population involved in our study could be described as ‘living independently’, as defined by their ability to travel independently, to be in employment and to live independently in their own accommodation (i.e. individuals who live independently of parents, carers or guardians). Of the 404 adults involved in our study, only 12% met this criterion.

7.161 One of the factors that should however be taken into consideration in interpreting this result is age. As shown in Table 7.54, there is again some evidence to suggest that though many on the spectrum are capable of achieving positive outcomes, it may be the case that, in comparison to the typically developing population, these outcomes are less likely to be achieved when the individual is in early adulthood. Indeed, the data presented here suggest that outcomes were most positive amongst adults who were middle aged between 27 and 49, and poorest amongst those who were under the age of 16.

7.162 There is also evidence here to suggest that in our sample positive outcomes were considerably more prevalent amongst those with Asperger’s/ HFA, with 18% of this subsample living independently, in comparison to the 3% of those with autism and 5% of those with other ASDs. However, the data also indicates that differences in outcomes are potentially more a product of an individual’s intellectual disabilities, in that of the 48 individuals identified as living independently, none had intellectual disabilities.

7.163 Finally, of interest here is that there is some evidence to suggest that in this sample, long-term relationships are not necessarily something reserved for those considered to live independently. Indeed, while 38% of the 72 individuals involved in long term relationship did live independently (n = 27), the majority of those in long-term relationships did not.

Table 7.54 Independent living amongst ASD individuals aged ≥ 16 years

Demographics and Outcomes Total n a Individuals able to travel independently, in employment and are living independently n (% of subsample)
Age

16 – 26 219 3 (2)
27 – 37 76 19 (25)
38 – 49 73 18 (25)
≥ 50 36 7 (20)
Sex

Male 288 32 (12)
Female 117 16 (14)
ASD Diagnosis

Autism 82 2 (3)
Asperger’s/ HFA 236 42 (18)
Other ASDs 86 4 (5)
ID Status

No ID 277 48 (18)
Mild ID 14 0 (0)
Moderate/Severe ID 62 0 (0)
Co-occurring conditions*

ADHD 30 4 (14)
OCD/Tourette’s syndrome 35 1 (3)
Epilepsy 29 6 (21)
Mood Disorders 138 30 (22)
Relationship Status

Involved in long-term relationship 72 27 (38)
Not involved in long-term relationship 331 21 (7)
Total adult (≥ 16 years) population 404 48 (12)

a Note that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

Table 7.55 Service use by ASD individuals and the parents of ASD individuals in the last 6 months

Demographics and Outcomes Total n (%)
Mental Health Services 243 (26)
Psychiatrist 120 (13)
Psychologist 146 (15)
Group Counselling 4 (0)
Individual Counselling 11 (1)
GH Services 83 (9)
GP Visits ( ≥ 3 visits) 83 (9)
ID & PD Services 232 (24)
Child Developmental Paediatrician 60 (6)
Occupational Therapist 75 (8)
Speech Therapist 98 (10)
Physiotherapist 28 (3)
Community LD Nurse 31 (3)
Other Community Nurse 34 (4)
Other Community LD Member 18 (2)
Challenging Behaviour Team Member 13 (1)
Employability Services 5 (1)
Sheltered Workshop 2 (0)
Individual Placement 5 (1)
Social Engagement Services 198 (21)
Befriending Service 26 (3)
Social Club 89 (9)
After School Club 59 (6)
Play-schemes 63 (7)
Care & Respite Services 116 (12)
Day care 25 (3)
Babysitter 23 (2)
Holiday Scheme 56 (6)
Home Help 22 (2)

Service Use

7.164 Table 7.55 shows the number and percentage of individuals in the sample who reported using each type of service in the 6 months prior to completing the survey. In Tables 7.56 and 7.57 the issue of service use is further explored in relation to several key factors relating to demographics, diagnosis and life outcomes. It should however be noted that this analysis did not take into account the use of employability services given that only 5 individuals reported using this type of service. The analysis here focussed on those aged ≥ 16 years given that many of these services are not relevant until adolescence or adulthood. However alternative statistics for the complete sample have been included in Appendix C.9.

7.165 In relation to age, our analysis suggested that though this factor did not appear to have an influence on the use of care and respite, social engagement or ID and PD services there was some evidence to indicate that mental health and general health service use were associated with age. For example, aged 27 – 49 years were significantly more likely to use mental health services in comparison to the rest of the sample, X 2 (1, 404) = 15.52, p < .001, and general health service use was greater amongst adults aged > 38 years), X 2 (1, 404) = 15.52, p < .001.

7.166 Analysis focussing on sex indicated that significant differences only existed in the use of general health services, X 2 (1, 404) = 21.20, p < .001, which were more frequently used by females.

7.167 There were however a greater range of differences in service use between those with different types of ASD diagnosis. More specifically, those with Asperger’s/ HFA were significantly less likely to use ID & PD services, X 2 (1, 404) = 40.79, p < .001, social engagement, X 2 (1, 404) = 2.47, p < .001, and care and respite services, X 2 (1, 404) = 11.50, p < .001. While there was some evidence to suggest that those with Asperger’s/ HFA had used general health services more in the 6 months prior to completing the survey, in comparison to the rest of the sample these differences were not found to be significant X 2 (1, 404) =1.78, p > .05.

7.168 Similar differences were identified when investigating the relationship between ID status and service use. Those without ID were significantly less likely to use ID & PD, X 2 (1, 404) = 14.40, p < .001 and social engagement services X 2 (1, 404) = 21.40, p < .001.

7.169 In terms of the relationship between co-occurring condition and service-use there was evidence to suggest that individuals were significantly more likely to use mental health services if they had OCD or Tourette’s, X 2 (1, 404) = 20.06, p < .001 or a mood disorder, X 2 (1, 404) = 27.46, p < .001. General health service use was also found to be more frequently used by those with mood disorders, X 2 (1, 404) = 50.65, p < .001.

Table 7.56 Service use amongst ASD individuals ≥ 16 years according to age, sex, ASD diagnosis and ID status

Demographics and Diagnoses n a Use of support services n (% of subsample) b
MH Services GH Services ID & PD Services Employability Services Social Engagement Services Care and Respite Services
Age (years)
16 – 26 219 58 (26) 20 (9) 38 (17) 4 (2) 39 (18) 23 (11)
27 – 37 76 28 (37) 7 (9) 15 (20) 0 (0) 8 (11) 9 (12)
38 – 49 73 27 (37) 15 (21) 9 (12) 0 (0) 6 (8) 6 (8)
≥ 50 36 10 (28) 6 (17) 3 (8) 1 (3) 3 (8) 2 (6)
Sex
Male 288 84 (29) 23 (8) 45 (16) 4 (1) 42 (15) 26 (9)
Female 117 40 (34) 26 (22) 20 (17) 1 (1) 14 (12) 14 (12)
ASD Diagnosis
Autism 82 27 (33) 6 (7) 25 (30) 2 (2) 13 (16) 16 (20)
Asperger’s/ HFA 236 73 (31) 35 (15) 19 (8) 0 (0) 26 (11) 11 (5)
Other ASDs 86 23 (27) 7 (8) 21 (24) 3 (3) 17 (20) 13 (15)
ID Status
No ID 328 76 (23) 3 (1) 23 (7) 3 (1) 14 (4) 18 (5)
ID 77 22 (29) 3 (4) 23 (30) 3 (4) 14 (18) 18 (23)
Mild ID 15 5 (33) 0 (0) 1 (7) 1 (7) 4 (27) 3 (20)
Moderate/Severe ID 62 17 (27) 3 (5) 22 (35) 1 (2) 10 (16) 15 (24)

a Reflects number of people for whom data was available, not the total number of people meeting this description in the sample. b Participants may be included in more than one column as they may have used more than one type of service

Table 7.57 Service use amongst ASD individuals ≥ 16 years according to co-occurring conditions, employment status, relationship status and residential status

Demographics and Diagnoses n a Use of support services n (% of subsample) b
MH Services GH Services ID & PD Services Employability Services Social Engagement Services Care and Respite Services
Co-occurring conditions c
ADHD 30 7 (23) 4 (13) 3 (10) 0 (0) 5 (17) 1 (3)
OCD/Tourette’s syndrome 35 17 (49) 6 (17) 8 (23) 1 (3) 3 (9) 4 (11)
Epilepsy 29 6 (21) 1 (3) 6 (21) 0 (0) 5 (17) 4 (14)
Mood Disorders 138 65 (47) 34 (25) 21 (15) 1 (1) 10 (7) 10 (7)
Employment Status
In Employment 112 43 (38) 16 (14) 12 (11) 2 (2) 14 (13) 6 (5)
Unemployed 292 81 (28) 32 (11) 54 (18) 3 (1) 43 (15) 35 (12)
Relationship Status
Involved in long-term relationship 71 22 (31) 16 (23) 4 (6) 0 (0) 3 (4) 1 (1)
Not involved in long-term relationship 310 101 (33) 32 (10) 61 (20) 5 (2) 53 (17) 39 (13)
Residential Status
Living Independently 126 44 (35) 21 (17) 12 (10) 0 (0) 8 (6) 10 (8)
Dependent on Others 237 75 (32) 25 (11) 50 (21) 5 (2) 45 (19) 29 (12)

a Reflects number of people for whom data was available, not the total number of people meeting this description in the sample b Participants may be included in more than one column as they may have used more than one type of service c Only the 4 most prevalent co-occurring conditions are mentioned here. It should also be noted that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.170 Analysis of the raw data relating to employment status and service use indicated that there was potentially a relationship between employment status and mental health service use, a greater proportion of individuals who were in employment using mental health services. However, follow up chi-square analysis indicated that these differences were not significant, X 2 (1, 404) = 3.15, p > .05.

7.171 As may be expected, those who were not involved in long-term relationships were more likely to have used ID and PD, X 2 (1, 404) = 7.05 , p < .01,as well as care and respite services, X 2 (1, 404) = 7.06 , p < .01 (both of which primarily provide for those with greater needs). Also of note here is that those who were not involved in long-term relationships were significantly more likely have used social engagement services in the 6 months prior to completing the survey, X 2 (1, 404) = 6.54 , p < .01.

7.172 Similar findings were also identified in relation to residential status in that individuals who were living with their parents or caregivers were significantly more likely to be have used ID and PD services, X 2 (1, 404) = 6.13 , p < .01, and social engagement services, X 2 (1, 404) = 8.81 , p < .01 both services that would typically be used by those with greater needs, and in turn those who are more likely to be living with their parents or caregivers.

Parental and familial impact of ASD

7.173 The final section of the survey focussed on gathering information about the parental and familial impact of ASD and Tables 7.58 and 7.59 report a summary of the data relating to five statements that parents and carers were asked to respond to (this section of the survey was completed by parents and carers, respondents with ASD were asked to leave this section of the survey blank). Parents were asked to rate these statements on a 4-point scale where ‘1’ indicated ‘no impact and ‘5’ indicated ‘major impact’.

7.174 In response to the first statement, the majority of participants (49%, n = 410) indicated that caring for an individual with ASD had had a ‘major’ impact on their ability to engage in work, training or employment, and a further 30% (n = 251) reported that the impact was ‘moderate’. However, the number and percentage of individuals reporting ‘major impact’ was significantly lower amongst individuals who cared for those with Asperger’s/ HFA, X 2 (1, 404) = 56.27 , p < .001. Only 8% of the sample (n = 71) indicated that caring for an individual with ASD had ‘no impact’ on their ability to be employment, training or education.

Table 7.58 Number and percentage of responses to rating scale statements assessing parental and familial impact associated with caring for individuals with ASD according to diagnosis of ASD individual

Area of parental or familial impact Type of ASD Diagnosis Total a
Autism Asperger’s/ HFA Other ASDs
To what extent has caring for an individual with ASD affected…
Your ability to be in employment, training or education
No Impact 17 (8) 40 (12) 14 (5) 71 (8)
Little Impact 23 (11) 48 (15) 34 (11) 104 (12)
Moderate Impact 58 (27) 113 (34) 79 (27) 251 (30)
Major Impact 114 (54) 127 (39) 169 (57) 410 (49)
Total* 212 (100) 328 (100) 296 (100) 836 (100)
The quality of your relationship with a partner or spouse
No Impact 28 (13) 47 (14) 32 (11) 107 (13)
Little Impact 34 (16) 64 (19) 49 (17) 147 (18)
Moderate Impact 64 (30) 113 (34) 100 (34) 276 (33)
Major Impact 86 (41) 105 (32) 115 (39) 306 (37)
Total* 212 (100) 329 (100) 296 (100) 836 (100)
Your ability to pursue social and leisure activities
No Impact 10 (5) 23 (7) 13 (4) 46 (6)
Little Impact 18 (8) 51 (16) 26 (9) 95 (11)
Moderate Impact 60 (28) 123 (38) 86 (29) 268 (32)
Major Impact 124 (58) 131 (40) 172 (58) 427 (51)
Total* 212 (100) 328 (100) 297 (100) 836 (100)

a Note that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

Table 7.59 Number and percentage of responses to Likert scale statements assessing parental and familial impact associated with caring for individuals with ASD according to diagnosis of ASD individual

Area of parental or familial impact ASD Diagnosis Total a
Autism Asperger’s/ HFA Other ASDs
To what extent has caring for an individual with ASD affected…
Your mental health
No Impact 21 (10) 26 (8) 23 (8) 70 (8)
Little Impact 32 (15) 84 (26) 56 (19) 171 (20)
Moderate Impact 88 (41) 118 (36) 115 (39) 321 (38)
Major Impact 72 (34) 100 (30) 102 (34) 274 (33)
Total* 213 (100) 328 (100) 296 (100) 836 (100)
Your physical health
No Impact 26 (12) 54 (16) 42 (14) 122 (15)
Little Impact 62 (29) 109 (33) 82 (28) 252 (30)
Moderate Impact 70 (33) 99 (30) 104 (35) 272 (33)
Major Impact 55 (26) 66 (20) 69 (23) 190 (23)
Total* 213 (100) 328 (100) 297 (100) 836 (100)
Other Family Members
No Impact 12 (6) 28 (9) 10 (3) 50 (6)
Little Impact 29 (14) 67 (20) 48 (16) 144 (17)
Moderate Impact 72 (34) 121 (37) 133 (45) 326 (39)
Major Impact 98 (46) 112 (34) 106 (36) 316 (38)
Total* 211 (100) 328 (100) 297 (100) 836 (100)

a Note that the arithmetic total values reported here were calculated through rounding following multiple imputation analysis and so may not always reflect the exact number of individuals involved in the analysis.

7.175 The second statement related to personal relationships amongst parents and carers, and 37% of respondents (n = 306) indicated that caring for an individual with ASD had had a ‘major impact’ on the quality of their relationship with their spouse, and a further 33% indicated that the impact was ‘moderate’. As above there was also evidence to suggest that the rate of individuals reporting ‘major impact’ was lower amongst those who cared for individuals with Asperger’s/ HFA in comparison to the rest of the sample, X 2 (1, 404) = 20.72, p < .001. Again, a relatively small percentage (13%, n = 107) reported that caring for an individual with ASD had ‘no impact’ on this aspect of their life.

7.176 The third statement related to the leisure time of parents and carers and in total 83% indicated that their ability to pursue social and leisure activities was either ‘moderately’ or ‘majorly’ impacted by caring for someone with an ASD. However, again the rate of those reporting ‘major impact’ was significantly lower amongst those caring for individuals with Asperger’s/ HFA, X 2 (1, 404) = 62.27, p < .001. In this case a very small percentage of individuals (6%, n = 46) indicated that caring for someone with ASD had no impact on their leisure time.

7.177 The fourth statement related to mental health, and this was the first case in which a greater number of individuals reported a ‘moderate’ impact as opposed to a ‘major’ impact. That said, in total the number and percentage of individuals reporting that caring for someone with ASD had had a ‘moderate’ or ‘major’ influence on their mental health (n = 595, 71%) was still relatively high. Also notable in this case that there were no significant differences in the experiences of individuals caring for people with different types of ASD.

7.178 Similar results were identified in relation to physical health in that the majority (63%, n = 524) of individuals reported that in caring for an individual with ASD there was ‘little impact’ or a ‘moderate impact’ on their mental health. Again, no significant differences were found between the experiences of those caring for different types of ASD.

7.179 Finally, parents and carers were asked to comment on the extent to which they felt that caring for an individual with ASD had impacted other family members. The majority of respondents (77%, n = 642) responded that they felt that this had had a ‘major’ or ‘moderate’ influence on other family members. Again it was notable here that only a very small number of individuals (6%, n = 50), indicated that caring for someone with ASD had no impact on other family members. Also of interest here is that the rate of individuals reporting ‘major impact’ was significantly higher amongst those who cared for individuals with autism, X 2 (1, 404) = 18.69, p < .001.

Predictors of Parental Impact

7.180 As with other areas of analysis reported in this section the authors carried out exploratory analysis to investigate whether there were factors which could explain the variance in responses to the statements described above. However, this was only possible in the case of one of the statements ( ‘To what extent has caring for an individual with ASD impacted your ability to be in employment, training or education’), in all other cases no relevant predictors identified.

7.181 The analysis exploring this statement was carried out using multiple linear regression. This type of analysis was selected in line with the recommendations made by Byrne (2000) that data of this nature may be analysed using multiple linear regression when the analysis involves four or more ranked categories.

7.182 As with other sections in this chapter analysis began by identifying relevant candidate variables (listed in Appendix C.10) which were subsequently added into a hierarchical model in the following five blocks relating to co-occurring conditions, (iv) those relating to other outcomes and (v) those relating to service-use.

7.183 As before, the final model shown in Table 7.60 reports only those candidate variables which resulted in an R 2 change of greater than .02. Candidate variables excluded from the model for this reason are detailed in Appendix C.10.

Table 7.60 Linear regression model testing the factors which predict parent and carer likert scale responses to the statement ‘To what extent does caring for an individual with ASD influence the extent to which you can be in employment, training or education’. [14]

Variable Β t R R 2 ∆R 2
Block 1

.17 .03 .03
Age -.02 - 2.47***


Block 2

.30 .09 .09
Age -.01 -1.85***


Ability to travel independently*** -.53 -3.79***


Note: * p < .05 ** p < .01, *** p < .001

7.184 Block 1 of the model introduced age (of the ASD individual) as a predictor of the level of impact that caring for an individual with an ASD had on a parent or carer’s ability to be in employment, training or education. Age was found to be a significant predictor F (2, 836) = 13.59, p < .001, and explained 3% of the variance in the data.

7.185 In block 2 of the model ‘ability to travel independently’ (note this related to the ability of the ASD individual rather than parents or carers) was introduced and was also found to be significant predictor F (2, 836) = 15.42, p < .001. This block explained a further 9% of the variance in the data, raising the total variance explained by the model to 12%.

7.186 In terms of what this model tells us, firstly it provided some evidence to suggest that for every year older an individual with ASD was, the less likely it was that their parent or carer would indicate that caring for someone with ASD has had a moderate or major impact on their ability to engage in employment, training or education. This is to some extent to be expected, and also fits with some of the other analysis in this chapter, in that ultimately this result indicates that the older an individual is, the less likely it is that they will be dependent on their parent or carer (though it should be stressed that this is not always the case), and that in turn the more likely it is that a parent can engage in other activities.

7.187 Secondly, this analysis also revealed that the parents and carers of those ASD individuals with ASD who could travel independently were also more likely to report a lower level of impact. Again, this is to some extent to be expected for two reasons. Firstly, it is likely that those who are unable to travel independently are more likely (but not exclusively) to be individuals who are lower functioning. Secondly, given that being able to travel independently is an important aspect of everyday life, it may be the case that the parents of ASD individuals who are unable to travel independently are more likely to feel restricted in engaging in other activities if a significant portion of their time is spent ensuring that their child is able to travel safely from place-to-place.

Summary of Findings from Statistical Modelling Analyses

7.189 This section provides a summary of the key findings from the statistical analyses of the responses to the questionnaire. Table 7.61 summarises statistically-significant findings from chi-square analyses and Table 7.62 the statistically-significant findings from linear and logistic regression analyses reported in this chapter. To take account of multiple testing (Tabachnick & Fidell, 2007) we report in the summary tables relationships which are statistically significant at p < .001. Rounded, this p-value equates to Bonferroni correction (Tabachnick & Fidell, 2007) for the 55 comparisons reported in each of the two tables.

7.190 In both tables, following an approach used by Morton and Frith (1995) in modelling autism, outcome variables are grouped at the levels of either ‘biological’, ‘cognitive’, ‘social’, ‘affective’ or ‘behavioural’ (Morton & Frith, 1995, pp. 357-358) to integrate the findings. Here we assigned sex and age and co-occurring conditions (excluding mood disorders) to the ‘biological’ level, together with type of ASD diagnosis as a proxy ‘biological’ variable; intellectual disability status to the ‘cognitive’ level; ‘relationships’ to the ‘social’ level; mood disorders to the ‘affective’ level; and highest level of educational support, employment, ability to travel independently and residential to a broader ‘behavioural and other’ level.

7.191 The findings summarised in Table 7.61 at the ‘biological’ level highlight the salience of age and ID status for type of ASD diagnosis, with an Asperger’s/ HFA diagnosis in the sample more likely over the age of 10 where the individual did not have ID. In turn, individuals with Asperger’s/ HFA over 16 years of age were more likely to have co-occurring diagnoses, including mood disorders, be involved in a long-term relationship of two years or more, to have been educated in a mainstream school or unit in a mainstream school and achieved higher levels of qualification, and be in employment and able to travel and live independently. Those with co-occurring conditions but excluding mood disorders were also more likely to require higher levels of educational support. There were further effects of sex and age in regard to educational support. With regard to sex, males were more likely to have their highest level of educational support in special units in mainstream, and in regard to age, individuals in the 16-26 years age-range were less likely than older individuals to have received their highest level of support in mainstream school. This latter finding indicates that individuals in the current ASD population have greater access to educational support in comparison with previous generations of individuals on the spectrum. There was a further association between age and ability to travel independently, and in addition, those in the 16-26 years age-range were less likely to be living independently than their older counterparts.

7.192 At the ‘cognitive’ level, mirroring the findings from type of ASD diagnosis above, those with higher ID status were more likely to be in a relationship, experience mood disorders, require lower levels of educational support, be in employment and be able to travel and live independently.

7.193 At the ‘social and ‘affective’ levels respectively, those in a long-term relationship were more likely to be able to travel independently and often had a diagnosis of mood disorder, both of which are associated with type of ASD diagnosis. In a similar vein, those with a mood disorder were also more likely have attended a mainstream school, be in employment and to live independently. At the ‘behavioural and other outcomes’ level, those in employment were more likely to be living independently, able to travel independently, have a co-occurring mood disorder, and higher educational qualifications, all again associated with type of ASD diagnosis. Turning to service use, age was a significant predictor of use of mental health and general health services, with females making more use of the latter than males. Those with an Asperger’s/ HFA diagnosis were also less likely to use care and respite services and specialist services for those with ID. Health services, both general and mental health, were more likely to be used by those with OCD, Tourette’s and mood disorder in the case of the former, and by those with mood disorder in the case of the latter. Finally, fewer parents and carers of those with an Asperger’s/ HFA diagnosis reported a major impact upon their own employment, training or education or upon family life.

7.194 Table 7.62 reports significant findings from the final models from the linear and logistic regression analyses. To correct for multiple testing as above, only significant findings with p < .001 are reported. Regression analyses are multivariate analyses in which the effects of specific variables can be examined while controlling for the effects of the other variables in the regression model. These models are more complex than those of the chi square analyses above and

Table 7.61 Summary of significant relationships ( p < .001) emerging from chi square analyses

Variables Biological Cognitive Social Affective Behavioural & Other Outcomes
Type of ASD diagnosis Co-occurring conditions (excl. Mood Disorders) Sex Age ID Status Relationships Mood Disorders Highest level of educational support Employment Ability to Travel Independently Residential Status
Biological
Type of ASD diagnosis --- ***
*** *** *** *** *** *** *** ***
Co-occurring conditions (excl. Mood Disorders)
---








Sex

---



***


Age


---


***
*** ***
Cognitive
ID Status



--- *** *** *** *** *** ***
Social
Relationships




---


***
Affective
Mood Disorders





--- ***
*** ***
Behavioural & Other Outcomes
Highest level of educational support






--- ***

Employment







--- *** ***
Ability to Travel Independently








---

Table 7.62 Summary of significant relationships ( p < .001) emerging from regression analyses (final models)

Variables Biological Cognitive Social Affective Behavioural & Other Outcomes
Type of ASD diagnosis Co-occurring conditions (exc. Mood Disorders) Sex Age ID Status Relationships Mood Disorders Highest level of educational support Employment Ability to Travel Independently Residential Status
Biological
Type of ASD diagnosis ---
Co-occurring conditions (exc. Mood Disorders) ---
Sex ---
Age --- *** *** *** ***
Cognitive
ID Status --- ***
Social
Relationships
--- *** *** ***
Affective
Mood Disorders

--- *** *** ***
Behavioural & Other Outcomes
Highest level of educational support

---
Employment

--- ***

Ability to Travel Independently

--- ***

also reflect any strong correlations between predictor variables which can result in some variables being excluded from the final regression model due to ‘multicollinearity’, where the variables are too similar. In particular dealing with multicollinearity in the regression analyses yields findings which differ in some respects with from those reported in Table 7.61. For example, at the ‘biological’ level, only age was a significant predictor, here of being in a long-term relationship, of receiving the highest level of educational support in a mainstream school, of being in employment and living independently. At the ‘cognitive’ level, ID status was a predictor of level of educational support. Key findings are at the ‘social’ level are that being in a long-term relationship of over two years duration is associated with not only with being in employment and living independently, but also with having a diagnosis of mood disorder, as those in a relationship are more likely to have an Asperger’s/ HFA diagnosis and less likely to have an ID. Similarly, at the ‘affective’ level, those with a mood disorder diagnosis are more likely to live independently, and to have attended mainstream school as their highest level of educational support as they are more likely to have an Asperger’s/ HFA diagnosis. Finally, at the ‘behavioural’ level, those able to travel independently were more likely to be in employment and to live independently.

Discussion

7.195 The findings above highlight the direct and indirect effects of type of ASD diagnosis and ID status upon a range of outcomes captured by the questionnaire. While the associations between age, education, independent living, independent travel, employment, relationships and residential status on the surface may appear unsurprising, they flag up the fact that there are those in this population who experience positive life outcomes in these areas despite experiencing depression and anxiety, which in turn highlights questions about the nature and provision of mental health services for this population.

7.196 The relatively low uptake of service use in our sample is interesting given the many problems experienced by those on the spectrum and their parents and families. This raises questions about access and availability of service.

Limitation to the modelling analyses

7.197 There are a number of limitations to this part of the study including the cross-sectional design, self-report measures and a self-selected sample, which pose problems for representativeness. Further, some of the analyses, for example of the effects of co-occurring conditions, were constrained by small numbers and regression analyses were further constrained by the effects of multi-collinearity. In addition, additional data about the nature of relationships would have been helpful.

Future Research

7.198 Areas for future research include exploring specific support arrangements and their impact upon outcomes in greater depth and together with investigating service use in more detail. For example, given the levels of mood disorder reported in the sample, why was there a relatively low take-up of befriending schemes and therapy services? Does this reflect acceptability of these services or availability? Further information about engagement with employment schemes and support into work would also be important areas to research.

Analysis of Free-Text Comments

7.199 Comments from the individuals with ASD and parents/carers were analysed separately to capture any distinctive comments from the two groups of respondents. The comments and associated themes/sub-themes may be found in Tables 11.24 and 11.25.

Comments from individuals with ASD

7.200 Nine individuals with ASD provided additional comments, some 8% of the 114 who completed the questionnaires themselves. The three themes and constituent sub-themes which emerged from the analysis of the free-text comments from individuals with ASD are shown in Figure 7.3, together with a network analysis of the relationships between themes. The themes and sub-themes provide an indication of the range of views expressed by the respondents. To each theme and sub-theme we added the number of respondents who mentioned them. Thus, for example, concerns about services or support for older adults were raised by three of the individuals with ASD.

7.201 At the level of themes, however, the ‘number of mentions’ is not the cumulative total of mentions of the relevant sub-themes. Rather, any mention of any constituent sub-themes counts here as only one mention at the level of the theme. For example, a mention by a respondent of concerns at the sub-theme level about stress and anxiety linked to day-to-day life or care and a mention also by the same respondent of concerns about stress and anxiety linked to employment would count as mentions at the level of both sub-themes but as only one mention at the level of the theme. This approach helps to provide information about the both range of views expressed at the level of sub-themes and the distribution of opinions across individual respondents at the level of themes.

7.202 Most of the respondents linked sub-themes and themes (see Table 11.25 for full details), with the network indicating underlying relationships between these views and concerns. The relationships between themes are represented in the network analysis in Figure 7.3 by bi-directional arrows which make no assumptions about whether the relationships are causal given the cross-sectional nature of the survey but merely denote linkage between the themes reported by at least one participant.

7.203 As the network analysis reveals, the three themes of concerns about support and service provision, diagnosis and stress and anxiety were linked by respondents, with the most prevalent theme, concern about support and service provision (particularly the availability and quality of support and services), mentioned by 8 respondents.

Figure 7.3 Thematic Network and Summary of All Themes and Sub-Themes from Free Comments Provided by Individuals with ASD.

Figure 7.2 Thematic Network and Summary of All Themes and Sub-Themes from Free Comments Provided by Individuals with ASD.

Support and service provision

7.204 The following comments illustrate the concerns expressed in regard to support and service provision:

‘I am finding that there is not much support for people in my situation - I do not need much day-to-day help but I could do with a regular opportunity to talk about how/how not to deal with things. Services seem to be focussed upon more immediate needs.’

I feel …that if you need support because you have an ASD you have to really, really fight for it. I now have the right support but it was not easy getting it.’

‘Autism services in the area are a disaster.’

‘[Charity] services require funding, but the majority of us have no access to this and do not have a social worker, nor have we ever been assessed for what help/support we need.

‘From my perspective, as a late-diagnosis adult, the system as regards those of us with Asperger’s syndrome is a complete mess.

Older adults

7.205 Concerns were also linked to the availability and quality of provision of support and services for older adults, reflecting the demographic of the respondents, as illustrated by the following comments:

The vast majority of people with ASD in Scotland are adult males and we are being pushed to the side-lines and not having our needs met while smaller groups within the ASD community are having huge amounts of attention paid to them. This situation is ridiculous and needs to urgently be addressed. No one is suggesting that children and young people should not receive good services, but this has to be proportionate.’

‘I am tired of seeing questionnaires like this which clearly focus on the needs of children and younger people.’

Stress and anxiety: older adults

7.206 Three of the older adult respondents also linked adequacy of support and service provision with reported experiences of stress and anxiety, including mental health problems, impacting upon everyday life, employment and post-secondary education:

‘There is no point in providing a Rolls-Royce service to children and young people who are then going to have to spend their adult lives receiving a second-hand Skoda service. The result of the inadequacy of service provision for adult males is to condemn them to increasing and debilitating mental health problems which could easily have been averted with relatively little investment.’

‘Older adults may have managed to cope with hidden difficulties for most of their life but the ageing process severely curtails both the ability to cope and the resilience needed to overcome the daily problems caused by lack of motivation, inability to make decisions, lack of ability to plan and the tendency to be impulsive. Together these difficulties make self-management of one's personal environment extremely difficult and there is currently no support service available to provide appropriate support at the appropriate time according to individual needs.’

Associations with co-occurring conditions

7.207 Comorbid or co-occurring conditions were associated in turn with diagnosis and with reported experiences of stress, anxiety and mental health in day-to-day life:

‘Too often services have only been made available if there is evidence or diagnosis of a learning disability or mental illness together with autism, but not for people with autism alone.’

‘I was recently freed from a diagnosis of Emotionally Unstable Personality Disorder, after I pointed out that the symptoms are more consistent with the result of living in neurotypical society with an undiagnosed (until recently) ASD.’

ASD and employment issues

7.208 Some respondents also expressed concerns linking ASD, diagnosis and a lack of support with stress in turn associated with employment:

There is …a cost to the Scottish Government where lack of appropriate support for adults of working age who have had to withdraw from meaningful employment because of the stress associated with both diagnosed and undiagnosed autism.’

ASD and education issues

7.209 Some of the respondents also expressed concerns linking ASD, diagnosis, and a lack of support to post-secondary education:

‘Unfortunately, as soon as I start studying formally, even under these conditions, [Benefit System] would conclude that this means I am fit for work and able to handle their emotional thuggery. The current social insecurity system is thus designed to keep me down.’

Comments from parents/carers

7.210 The five themes and constituent sub-themes which emerged from the analysis of the free-text comments from the 68 parents and carers of individuals with ASD, some 10% of the 705 parents and carers who completed the questionnaire, are shown in Figure 7.4. As before, the number of participants who mentioned each of the sub-themes is also indicated and in the case of themes, any mention of any constituent sub-themes counts here as only one mention at the level of the theme. Again, the network of links between themes is represented in Figure 7.3 by bi-directional arrows which make no assumptions about whether the relationships are causal given the cross-sectional nature of the study, but merely denote linkage between the themes from the comments of at least one participant.

7.211 At the level of the five themes, the analysis revealed links between concerns about support and service provision, diagnosis, and stress and anxiety experienced by both individuals with autism and their parents/carers, which in some cases had an impact also on family life. Interestingly, concerns about social issues (mentioned also by eight respondents and relating to difficulties in socialisation, maintaining employment, or to criminal justice issues) were linked only to concerns about support and service provision. Full details of the links between sub-themes and themes may be found in Table 11.26, with the network indicating underlying relationships between these views and concerns. We consider these relationships below, together with illustrative comments from the respondents.

Support and service provision

7.212 Sixty two respondents commented on support and service provision. These included four respondents who reported positively on outcomes or on the support received, as the following comments illustrate:

‘My son has the best support we can hope for at our local primary school and has moved from having to have a SLA to now coping with all the work he is set just with the help of his teacher. His school always have great transition between years and choose his class teachers carefully! I couldn't ask for better.’

‘The services and support that [Scottish City] Autism Support provides are invaluable to us. They provide services and activities that no one else does and without them my son would not be as able to socialise with his peers in a variety of environments nor have opportunity to learn skills.’

7.213 Some parents and carers also commented on the positive experiences of parenting and caring for an individual with ASD:

‘He's worth every stress-filled, pull your hair out, penny pinching moment of it.’

Figure 7.4 Thematic Network and Summary of All Themes and Sub-Themes from Free Comments Provided by Parents and Carers of Individuals with ASD.

Figure 7.3 Thematic Network and Summary of All Themes and Sub-Themes from Free Comments Provided by Parents and Carers of Individuals with ASD.

7.214 However, 58 respondents (85% of the parents/carers who provided additional free comments) expressed concerns about the support and service provision. Many of these were in regard to education provision:

Large mainstream primary schools are not equipped to deal with ASD/Asperger's: dumping these kids into a class of 27 other kids with no classroom assistance is not inclusion, the amount of phone calls, notes and issues coupled with meetings, IEPs, child planning meetings is soul destroying especially when often the people who are meant to be there to help don't seem to grasp the basics about Autism and have to be reminded continually, to look for the triggers and not just the undesired behaviour itself. My son is intelligent and would not be put into a special school. The autism units locally are full but would be a better option as the staff know what they are doing. In his mainstream school the teachers have 45 mins of optional info. What on earth can they gain from that to prepare them for 6 hours a day with our kids? If they chose to do it. We have a long way to go in society before people with autism and their carers are treated equally. There is a consultation in [Local Authority Council] over local strategy and not one person on the consultation is an expert in autism.

Diagnosis of ASD

7.215 Concerns in this area, particularly in regard to education provision, were also linked by some to problems in obtaining a diagnosis of ASD:

‘Professionals did not realise he had difficulties. When I raised the issue with an educational psychologist I was made to feel stupid and was told he definitely did not fit the criteria. After pushing for assessment other professionals were more helpful. He has been diagnosed but this took a year due to waiting list at [Diagnostic Service]’.

‘My youngest has a working diagnosis of ASD and possible ADHD. We are now going into P4. The time taken to reach a diagnosis and the support my child needs has I feel taken a lifetime to come. This needs to be addressed.’

Stress and anxiety in day-to-day life of parents and carers

7.216 There were links also between support and service provision and stress and anxiety in the day-to-day life of parents/carers, with financial concerns a contributory issue for some, linked to parental employment issues:

‘No matter what age a person with ASD is they will always need some form of help. The change over from DLA to PIP is causing so much stress for carers that have to apply for the ASD sufferer. We have had to phone every week to see if my son’s DLA was going to be extended. We applied for his PIP in February this year we have been told it will be January… before we find out if he will get it or not. He hasn't changed in the 16 years since his diagnosis and things get harder for him every year not easier so why should his claim for DLA or PIP need to take so long. This causes stress to the person and their carer.’

‘My son is not able to travel on the school transport without it causing him great anxiety. When I am not in work, I have to take my son to and from school myself which is a mileage of 32 miles per day as we live in a rural area. I often try to take him in, even when I am working which requires me to request a late start at work which does cause my employers some difficulties. I have to juggle the need to keep my job for financial reasons and not letting my son get too anxious.’

‘We have had to go to some extraordinary lengths to secure our son's future...it has exhausted our health & finances. There should be more support for parents dealing with such a severe condition that seems to be on the increase. Most parents won't know how to access the help or even have the energy to go out and get it. Social services are stretched to the limit but there should be a hub of information. Once they leave school it is a mine field...most parents I know are not given enough options for their young adult child moving into the adult world. There seems to be no provision of continued education after they leave school...they may be 18 by age and legally they're seen as an adult but they are leaving at a different mental age and I have found their education ceases. If they were tested to establish their mental age it would be noticed that they should still be getting educational input or at least some input. It's a bit like taking a 10 year old out of school and expecting them to just get on with it in the world. People continue to learn no matter what conditions they have; they shouldn't just stop getting support and learning input.’

‘I would like the education system to review their summer holiday schedule. Seven weeks over the summer is too long for everyone. Even those who have normally developing children, say it is too long for the children to have no structure in their lives. It’s a financial drain, but most importantly, it simply is not good for the children. In England the holidays are six weeks. This is quite long enough. Also, there seem to be a constant stream of holidays over the year. In fact there isn't one single month in the whole year, where there are no days off from one holiday or another. Added to the volume of training days for the teachers, it is a constant strain on our resources; mentally, physically and financially. My partner is so tired he is dropping to part-time work next month so things are just going to get harder. Also, summer support is lacking. [Charity] provided some summer camps but they were not suitable for a severely autistic boy - mainly high functioning. We tried one day and it was not possible for my son to attend further. We do get Direct Payment and pay for cover for him, but managing the Direct Payment is also a bit draining. I think what I’m saying is we don't feel we can go on much longer with the situation we live in.’

‘We have managed because one of us has always been at home. This makes caring for all our children manageable. Financially it was tight at times, but it meant minimal childcare costs except in emergencies. But it also meant we knew someone was there for our son’.

Stress and anxiety in day-to-day life of individuals with ASD

7.217 Five parents/carers also reported links between level of support and service provision particularly in regard to education and stress and anxiety on the part of individuals with ASD:

‘Support in school tailored for young people is so difficult to access. Our daughter was treated very badly in her first secondary school which resisted in mental and physical problems, and her not being in school for several months. Her new school have been amazing and it proves what can happen if the will is there. Not enough support available to parents.’

‘He needed extra support around school as school was very stressful especially up till P5. He still needs emotional support around the more difficult days and having a parent at home helps immensely.’

‘I feel mainstream schools have a long way to go before they really understand children with ASD. I am hoping he will get the support he needs in high school as on days he was not coping he was sent home, which made my life very stressful as he then learned if he didn’t feel like being in school he let them think he wasn’t coping, so he was sent home. This has left him with no education over the last two years which I found very hard as he is a bright boy who will have to work really hard to catch up. This will put too much stress on him and he then shuts down.’

Family life

7.218 The impact of availability/quality of support and service provision and financial pressures upon family life was also noted by some respondents:

‘Support for siblings is also very poor. They need more support to understand why their brother behaves the way he does.’

‘I worry also for the mental health of my other son.’

‘Caring has impacted on all the family’

‘[Scottish City Education Department], Social Work and the NHS completely fail in their 'duty of care' for ASD children and adults. The stress this is putting on ASD sufferers and their families is intolerable and an utter disgrace!’

Social issues

7.219 Finally, social issues (including employment and criminal justice issues) were linked to service provision and also to financial concerns by some parents and carers:

‘Fascinated that you aren't questioning the single biggest stressor: the manic dance we are tortured through with the benefits system which fails to provide ANY support for intelligent Aspies to get into work.’

‘The person I care for has had many jobs but has walked out of almost every one because of nastiness expressed in the workplace and although the human resources staff have asked him to return he would not and, in discussion with other carers I find that this is common amongst people on the autism spectrum who are employable.’

‘I am one of 10 families whose children attended a special school who have been restrained and ill-treated by staff. There seems to be no accountability where children are hurt in council schools. We have fought long and hard and are prepared to campaign the government if needs be. Police Scotland have no experience in disability and have no idea how to deal with autistic children or people with any kind of communication difficulty when there are allegations of abuse. This needs to change.’

Discussion

7.220 The level of concern expressed here by the individuals with ASD and the parent/carers in regard to support and service provision and its relationship with mental health and well-being and family life is consistent not only with the rating scale data from the full-sample of parents reported in Tables 7.59a and 59b relating to the impact of ASD but also with recently published studies.

7.221 The individuals with ASD who responded flagged up concerns about support and service provision, including provision for older adults, as well as the associations between co-occurring conditions, mental health, employment and post-secondary education. Gillott and Standon (2007), for example, in a study of 35 adults with ASD in England found that adults with ASD were three-times more likely to have elevated anxiety levels associated with coping with change than a matched control group of 20 adults with intellectual disabilities. With regards to employment and post-secondary education, poor outcomes for adults with ASD have been identified by recent studies carried out in the US by Roux et al. (2013) and Gelbar, Shefyck, and Reichow (2015) respectively. These studies note the need for comprehensive support in social and emotional domains and also the importance of self-advocacy in regard to post-secondary education. The importance of informal social support from family, friends and acquaintances for adults with ASD is also highlighted by a study carried out in Belgium by Renty and Roeyers (2007).

7.222 Turning to the responses from the parents and carers, positive experiences of parenting children with ASD have been reported in a recent study of 56 parents in the US carried out by Altiere and von Kluge (2009). However, concerns regarding variability in provision of services, delays in diagnosis, and reductions in contact with multi-agency services as children with ASD become older are confirmed by recent studies in the UK (Bebbington & Beecham, 2007; McConachie & Robinson, 2006) and elsewhere (Sun et al., 2013). Concerns expressed by parents and carers regarding the importance of peer relationships have been reported in the literature (Lindsay, Ricketts, Peacey, Dockrell, & Charman, 2016). Concerns about the provision of programmes of social activities for children and continuity in the support and services provided have also been identified in other studies, notably by Canadian researchers (Brown, Ouellette-Kuntz, Hunter, Kelley, & Cobigo, 2012; Brown et al., 2011; Hodgetts, McConnell, Zwaigenbaum, & Nicholas, 2017).

7.223 The links between parenting a child with ASD and parental mental well-being identified by the parents and carers are well-established in the literature (Barker et al., 2011; Hodgetts et al., 2017; Lai, Goh, Oei, & Sung, 2015; Smith, Seltzer, Tager-Flusberg, Greenberg, & Carter, 2008). The quality and range of service provision, financial pressures including employment difficulties (Hill, Jones, Lang, Yarker, & Patterson, 2014), problems in engaging with the benefits system, and also concerns about education provision can all be sources of stress for parents and carers leading to problems with anxiety and depression.

7.224 The parents and carers also highlighted pressures from schools’ ability to cope with the social and emotional needs of pupils with ASD, social relationships, employability and the youth justice system as sources of stress and anxiety for individuals with ASD, but also note the effects upon the siblings of those with ASD. Tsai, Cebula, and Fletcher-Watson (2016), for example, carried out a cross-sectional survey of 155 mother and typically-developing sibling dyads (75 in the UK and 80 in Taiwan) which revealed the importance of parents’ coping style upon the adjustment of the typically-developing siblings in the UK.

Limitations to the thematic analyses

7.225 There are limitations to the thematic analyses reported here. Firstly, only a relatively small proportion of those who completed the survey, 8% and 10% of individuals and parents/carers respectively, elected to provide and share additional comments. We cannot claim therefore that the views expressed are representative of the sample as a whole.

7.226 Further, the views were not obtained by means of individual interviews or focus groups, which would have yielded a richer data set and permitted exploration and follow-up of comments and views made by the respondents.

7.227 Finally, as a cross-sectional survey, we cannot draw inferences regarding underlying causal relationships, but can only report associations and links. However, with these caveats notwithstanding, this part of the questionnaire provided the parents/carers and individuals with ASD themselves with a voice, and their comments illuminate key issues regarding the impact of ASD upon individuals, carers and families and of the provision both formal and informal available by way of support.

Comments on autism and sex (male/female) and on ID

7.228 Throughout this analysis we have presented data separately for males and females with ASD. In summary, the data from this sample has comprised a significantly larger number of males than females with ASD, in line with the established literature ( para. 7.19); there have been no significant differences in the figures in terms of type of ASD diagnosis received ( para. 7.27), in numbers with intellectual disability, including numbers separately for moderate/severe ID ( para. 7.32), in those in employment compared with those not in employment ( para. 7.107), in those in a long-term relationship compared with those not in such a relationship ( para. 7.126), in those living independently compared with those not living independently ( para. 7.144), or in patterns of service use, other than in use of general health services, which were used more by females ( para. 7.166). We also found that more males in our sample had their highest level of educational support in a special school or unit, while more females were in mainstream school ( para. 7.64).

7.229 Regarding the significantly higher number of males than females diagnosed with ASD, it is not known to what extent this reflects actual differences in prevalence or to what extent is represents under-diagnosis of women and girls. Baron-Cohen and others have argued for higher real prevalence of ASD in males from a neuropsychological standpoint (Baron-Cohen, 2002, 2009). Others have suggested that females have superior ability to cope with ASD deficits (Kreiser & White, 2014; Dworzynski, Ronald, Bolton, Happé, 2012), that they are more likely to be quiet and compliant in school (Lai et al., 2011), or that they are more able to imitate appropriate social behaviour (Gould & Ashton-Smith, 2011), thus leading to reduced rates of referral and diagnosis. Some have hypothesised a female ‘phenotype’ for ASD (see Kirkovski, Enticott and Fitzgerald, 2013, for a review), while others have proposed no significant gender differences in ASD symptoms (see, for example, May, Cornish and Rinehart, 2014). As a general statement, males and females in the general population differ in many aspects of their presentation, and it has not been established that any differences in ASD presentation between males and females are anything other than a reflection of this.

7.230 In the overall sample reported in this chapter, the data for males and females showed almost no significant differences beyond prevalence. Regarding the higher use of general health services by females it is difficult to comment, since the literature on use of general health services by males and females in the general population is unclear. Regarding the fact that more females remained in mainstream while more males were educated in special provision, this reflects more general patterns in the distribution of additional support needs between males and females, with males over-represented in special schools. This pattern has been clearly established in Scottish special educational statistics for a very long time, with historically higher numbers of boys than girls in provisions such as schools for moderate learning difficulties and schools for emotional and behavioural difficulties (see, for example, Clark and MacKay, 1976).

7.231 Turning to ID, previous research reviewed above consistently indicates that this is a strong predictor of a broad range of outcomes for both children and adults alike. However, our findings reported here reveal that type of ASD diagnosis was a stronger predictor of outcomes than ID. As type of ASD diagnosis is partly dependent upon ID, the two variables could not both be included in the same model due to marked multi-collinearity. Details of ID were available for only 649 of the 950 participants, however, whereas type of ASD diagnosis was available for all. This increased statistical power, and the level of prediction of the type of ASD diagnosis, which accounts for the elimination of ID from the hierarchical regression models reported here.


Contact