Adult Psychiatric
Adult Psychiatric
Adult Psychiatric
Experimental statistics are official statistics which are published in order to involve
users and stakeholders in their development and as a means to build in quality at an
early stage. It is important that users understand that limitations may apply to the
interpretation of this data.
These statistics have been labelled as experimental in order to allow for further
work to be undertaken to assure users that these statistics meet the overall quality
standards necessary to be designated as National Statistics. Information on how
users should interpret these statistics are available throughout this chapter, in the
Methods chapter of this publication, and in the accompanying Background Data
Quality Statement.
NHS Digital will gather feedback to these statistics, their construction and
interpretation from relevant experts, and following this announce details of how
the assessment of these statistics will progress on the NHS Digital website during
February 2017.
All official statistics should comply with the UK Statistics Authority’s Code of
Practice for Official Statistics which promotes the production and dissemination
of official statistics that inform decision making.
Find out more about the Code of Practice for Official Statistics at:
www.statisticsauthority.gov.uk/monitoring-and-assessment/code-of-practice/
2
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Summary
• This chapter presents data on the profile of ASD among adults living in the
English household population. This is the second time such data have been
collected in England, after it was covered for the first time in the 2007 Adult
Psychiatric Morbidity Survey (APMS).
• In the phase one interview ASD was screened for using the Autism Quotient
(AQ-20). In the phase two interview, fuller assessments were carried out by
clinically trained interviewers using the Autism Diagnostic Observation Schedule
(ADOS) with a subset of participants with an AQ score of 4 or more. The results
were weighted to generate a prevalence estimate for the population as a
whole. This approach has been extensively validated. It should be recognised
however that psychiatric diagnoses tend to be reached by professionals over
multiple sessions involving probing and clinical judgement. Health surveys
are a population research tool and should not be expected to provide the
equivalent of a professional diagnosis. For further discussion of survey
limitations see Chapter 14.
• Data from the 2007 and 2014 surveys were combined to generate a larger
sample for analysis. The APMS series has been designed so that samples can
be combined in this way. Estimates based on the combined dataset are more
robust than estimates based on the 2007 or the 2014 samples separately.
31 potential cases were identified in the combined phase two samples, which
is small for subgroup analysis and means caution with interpretation is required.
Had all participants completed a phase two interview (see above), we estimate
that about 120 cases might have been identified in the sample as a whole.
3
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
• Using the combined sample, the prevalence of ASD was estimated to be
around 0.8%. Survey estimates are always subject to sampling error. Given this,
we estimate that if all adults in the population had been tested, the proportion
identified with ASD would probably be between 0.5% and 1.3% (95%
confidence interval (CI)). The size of this confidence interval is large, but similar
to that for some other low prevalence disorders considered on APMS.
• Consistent with other research, estimated rates of ASD were higher in men
(1.5%, 95% CI: 0.8% to 2.6%) than women (0.2%, 95% CI: 0.1% to 0.6%).
• ASD was associated with level of educational qualification, with rates being
higher among people with no qualifications.
• People with ASD appeared to be no more likely than other adults to make use
of treatment or services for mental or emotional problems.
6.1 Introduction
4
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
APMS 2007 was the first general population probability sample survey in any
country to have assessed ASD in adults (Brugha et al. 2009b), with APMS 2014
being the second. Rates may be different in specific adult populations, such as
among people who are homeless or living in prison. Rates were higher in men
and in those without educational qualifications.
ASD has been assessed among children and young people, and two large-scale
surveys estimated the prevalence of childhood ASD to be around 1%,1, 2 and higher
in boys than girls (Baird et al. 2006; Green et al. 2004). ASDs are more apparent and
easier to study in children, in part because the diagnosis of autism should include
presence of symptoms in childhood and parent and teacher observations of this are
more likely to be accurate and available for this group. Because these studies have
used different methodologies to APMS, the results are not directly comparable.
1 The prevalence of ASDs among children in South Thames was estimated by Baird et al. to be 116.1 per 10,000 (95% CI 90.4–141.8).
A narrower definition of childhood autism, which combined clinical consensus with instrument criteria for past and current
presentation, provided a prevalence of 24.8 per 10,000 (17.6–32.0).
2 Green et al. presented confidence intervals for the estimated prevalence of ASD among 5 to 10 year olds in England (1.13, 95%
CI 0.65–1.39) and for 11 to 16 year olds (0.76, 95% CI 0.47–1.06).
5
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
6.2 Definition and assessment
Assessment of ASD
6
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
APMS series since 1993 for the assessment of psychotic disorders. The APMS
2007 process involved a detailed validation assessment (Brugha et al. 2011b).
It includes the following stages:
7
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Details of the modelling undertaken to select the best subset of items are given
in a separate technical report (Brugha et al. 2011b). The AQ-50 questionnaire is
composed of items designed to assess five broad dimensions: social functioning,
imagination, communication, attention switching and attention to detail. The 20
adopted items selected by the modelling procedure as the best predictors of a
positive ASD assessment spread quite evenly across these categories: six were social
functioning items; four, communication; four, attention to detail; three, attention
switching; and three, imagination. The AQ-20 was discussed by an expert panel and
tested in the cognitive piloting conducted as part of the APMS 2007 development
work. Further modelling took place using a random sample of adults in contact
with mental health services. This identified the 17 most predictive AQ items used
in the 2007 survey, these were retained and three (which had performed poorly
in the 2007 data) were replaced with items with improved prediction selected
from the original AQ-50 (Tyrer et al. 2013). The revised 20 item version of the
AQ is reproduced in full in the questionnaire in Appendix D.
A score was generated for each participant based on their responses to the 17
AQ items included in both the 2007 and 2014 surveys. Each response indicative of ASD
was given one point, so that a higher score indicated greater likelihood that the person
may have ASD. The AQ-20 is a self-completion questionnaire, and it was administered
via a laptop computer in the phase one interview. Because it is a brief test and not a
diagnostic measure, a clinical assessment was included in the phase two interview.
On APMS, the AQ was used only to exclude cases with an extremely low
likelihood of having autism (those with an AQ score of between zero and three)
and to inform the selection probabilities for phase two. It was not used to
positively identify ASD.
8
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
as psychotic disorder was also covered in phase two). All men with an AQ score of
eight or more and all women with an AQ score of 11 or more were selected for
phase two (except for those interviewed in the final two months of fieldwork, who
were excluded). For men with an AQ score between four and seven and women
with a score between four and ten, a subsample of those agreeing to recontact were
selected for phase two, as outlined in the table below.
Men 0 to 3 0.000
4 to 7 0.157
8 to 10 1.000
11 or more 1.000
Women 0 to 3 0.000
4 to 7 0.045
8 to 10 0.245
11 or more 1.000
a
Of those who agreed to take part in phase two.
It was not feasible for all phase one participants to have a phase two assessment,
and this approach was designed so that those with the highest likelihood of a
having autism (or psychosis) had the highest likelihood of being assessed, combined
with being able to generate estimated rates of ASD for the population as a whole.
9
| APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
of APMS 2007). It is a widely recommended ‘gold standard’ clinical research
assessment instrument for autistic disorders that is used to collect information
on adult functioning (Lord et al. 2002). The ADOS is a semi-structured clinical
assessment of whether current behaviour is consistent with a diagnosis of an
autistic disorder. In 2014, additional questions on restricted, repetitive patterns
of behaviour and sensory differences were added to the second phase two
interview. This was done due to their increased emphasis in DSM-5 and because
the ADOS does not offer an adequate opportunity to measure restricted and
repetitive behaviours, although such behaviours are coded on the ADOS if they
occur. This additionally collected data is not reported on in the current chapter,
but will be analysed in subsequent publications.
The ADOS and its algorithm have been validated in previous clinic based
testing, but prior to APMS 2007 they had rarely been used with older adults or
in a general population setting (Gotham et al. 2008). The methods and results
of a quality assurance and validation study made use of clinician ratings and
developmental interviews with parents and other informants to inform severity and
clinical significance thresholds (Brugha et al. 2009a). That study found the ADOS
performed well, and its results have informed the case threshold used in this report.
The ADOS consists of a series of tasks that evaluate communication, reciprocal social
interaction (social functioning), creativity, imagination and stereotyped interests
and restricted interests. These tasks are rated by the trained interviewer. The ADOS
ratings that correspond to autism criteria are summed to produce an overall score.
A score of seven or more is the threshold used to identify an inclusive category of
ASD that is intended to correspond generally to the conceptualisation that underlies
the term PDD (Brugha et al 2011b). The recommended threshold of 10 or more is
applied in this report to indicate a case of autism, validated in the same population.
• For men with a phase one AQ score of four or more and women with a phase
one AQ score of eight or more, and who had an ADOS assessment, the results
of the ADOS were used.
10 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
• Men with a phase one AQ score of three or less and women with a phase two
ADOS score of seven or less were designated ASD negative, regardless of whether
or not an ADOS assessment was completed.
• Men with a phase one AQ score of four or more and women with a phase one
AQ score of eight or more who did not have an ADOS assessment (e.g. due to
non-selection, refusal or non-contact) were excluded from the analysis, and a
weighting strategy was applied to take account of their absence and to address
non-response biases.
For the analyses presented in this report the 2007 and 2014 samples were
combined to increase the sample size available for subgroup analysis. The survey
series has been designed with the intention that samples should be combined,
especially for analyses of low prevalence disorders or subgroups. This approach
is also taken in the chapter on psychotic disorder.3
6.3 Results
3 Several papers published in peer-reviewed medical journals have analysed samples combined from across the survey series. See
Appendix A for examples of these.
11 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
A total of 12 probable cases were identified in the 2014 sample, because a
sub-sample of respondents was selected for a phase two interview. This small base
means that great caution is required in interpreting the population distribution of
autism. To improve how robust the estimates are, the 2007 and 2014 samples have
been combined, yielding 31 participants identified with autism. Estimates based
on the combined sample are more robust than those based on the separate 2007
and 2014 samples.
Estimated rates of ASD were higher in men (1.5%, 95% CI: 0.8% to
2.6%) than women (0.2%, 95% CI: 0.1% to 0.6%). Tables 6.1, 6.2
Men Women
% 2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
2007 2014 Combined
Year
4 For example, the estimated proportion of the population with signs of dependence on drugs other than cannabis is 0.8%, with a
95% confidence interval of 0.6% to 1.2%.
12 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Some variation in prevalence of autism was evident with age, although there was
not a clear pattern. Tables 6.1, 6.2
% 2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
2007 2014 Combined
Year
Ethnic group
None of the 31 adults identified with autism in the APMS 2007 and 2014
samples was a participant from a minority ethnic group. However, due to the
small number of minority ethnic respondents in the sample as a whole, caution
is required in interpreting whether or not autism is associated with ethnic group.
No table is presented for this analysis.
13 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Region
There was no significant variation in the prevalence of autism across the four NHS
England regions. Table 6.3
Education
Presence of autism was associated with the highest educational qualification that
people had achieved. Overall, the rate was lowest among those with a degree level
qualification (0.2%) and highest among those with no qualifications (1.5%). 3.2%
of men without qualifications were identified with autism. Table 6.4
% 1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Degree Other qualifications No qualifications
Employment status
Analysis by employment status was run on those aged 16–64, to exclude people
who are retired constituting most of the economically inactive group. There was no
significant variation in the proportion of adults identified with autism according to
whether they were employed, unemployed or economically inactive. Table 6.5
14 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Figure 6D: Autism, by employment status: 2007 and 2014 combined
Base: adults aged 16-64
% 1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
In employment Unemployed Economically inactive
Employment status
People identified with autism were no more likely than those without autism to
use any of a range of different types of treatment or services for a mental health
reason. In fact, in terms of use of health services for a mental health reason, it
even appears that people with autism were less likely. Healthcare services included
use of inpatient or outpatient health services within the last three months for
a mental health reason or discussing a mental or emotional problem with a
GP within the last year. 3.7% of adults identified with autism reported this,
compared with 11.6% of those who without autism. Table 6.6
15 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
6.4 Discussion
There was no clear pattern in the distribution of autism by age. Rates were
higher in men than women, as found in most research on autism (Brugha 2011a).
However, it has been suggested that assessments for autism may draw more on
how the condition manifests in men, and this may lead to under identification
of autism in women (Trubanova et al. 2014). Autism was much more common
among people, especially men, without any qualifications, and rates were lower
in those with a university degree. No one from an ethnic minority group was
identified with autism in either the 2007 or the 2014 APMS. This is likely to be
due to the small sample size.
Adults in the survey identified with autism were no more likely than adults
without autism to use treatment or services for a mental health reason. And in
fact, adults with autism appeared even less likely than those without to use health
services for a mental or emotional reason. In contrast, every other mental health
condition examined in this and previous APMS has been shown to be associated
16 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
with increased use of treatment and health services. Other work shows that
adults with mental disorders receive attention from services because mental
health problems are recognised needs (Spiers et al. 2016). This recognition of
need does not appear to extend to adults with ASD.
Adults with autism have enduring problems with communication and social
understanding. There are no effective medical treatments for autism in adulthood.
However care services for identifying and supporting them are being developed lead
by local authorities throughout England.5 These services are based on the principle
that carers and health and social care staff can recognise and accept the presence
of the condition, and learn how to understand and communicate with those who
have it. Clinical experience of providing informed social care of this kind to adults
given a diagnosis of ASD could lead to real improvements in quality of life.
6.5 Tables
Prevalence
Table 6.1 Autism in 2007 and 2014, by age and sex
Table 6.2 Autism (2007 and 2014 combined), by age and sex
Characteristics
Table 6.3 Autism (observed and age-standardised), by region and sex
5 www.improvinghealthandlives.org.uk/projects/autsaf2014results
17 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
6.6 References
18 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Brugha T, Cooper SA, McManus S, Purdon S, Smith J, Scott FJ, Spiers N, Tyrer F
(2012) Estimating the Prevalence of Autism Spectrum Conditions in Adults:
Extending the 2007 Adult Psychiatric Morbidity Survey. NHS IC. Leeds.
http://digital.nhs.uk/pubs/autism11
Brugha TS, Tyrer F, Scott, FM, Bankart J, Cooper SA, McManus S. (2014)
‘The Epidemiology of Autism Spectrum Disorders in Adulthood’ in Autism
in adolescents and adults. Yale: Springer.
Brugha TS, Spiers N, Bankart J, Cooper SA, McManus S, Scott FJ, Smith J, Tyrer F.
Epidemiology of autism in adults across age groups and ability levels. Br J
Psychiatry, 2016; Jul 7. pii: bjp.bp.115.174649. [Epub ahead of print]
Buescher AV, Cidav Z, Knapp M, Mandell DS. Costs of autism spectrum disorders
in the United Kingdom and the United States. JAMA Pediatr, 2014; 168(8):
721–728.
Constantino JN, Davis SA, Todd RD, Schindler MK, Gross MM, Brophy SL,
Metzger LM, Shoushtari CS, Splinter R, Reich W. Validation of a brief
quantitative measure of autistic traits: comparison of the social responsiveness
scale with the autism diagnostic interview-revised. J Autism Dev. Disord, 2003;
33(4): 427–433.
19 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Green H, McGinnity A, Meltzer H, Ford T & Goodman R. (2005) Mental Health
of Children and Young People in Great Britain, 2004. Hampshire: Palgrave
McMillan. http://digital.nhs.uk/catalogue/PUB06116/ment-heal-chil-youn-peop-
gb-2004-rep2.pdf
Howlin P, Goode S, Hutton J & Rutter M. Adult outcome for children with autism.
Journal of Child Psychology and Psychiatry, 2004; 45(2): 212–229.
Lord C, Risi S, Lambrecht L, Cook EH Jr, Leventhal BL, DiLavore PC, Pickles A,
Rutter M. The autism diagnostic observation schedule-generic: a standard
measure of social and communication deficits associated with the spectrum of
autism. Journal of Autism and Developmental Disorders, 2002; 30: 205–223.
Rai D, Kerr MP, McManus S, Jordanova V, Lewis G, Brugha TS. Epilepsy and
psychiatric comorbidity: a nationally representative population-based study.
Epilepsia, 2012; 53(6): 1095–103.
Ritvo RA, Ritvo ER, Guthrie D, Yuwiler A, Ritvo MJ, Weisbender L. A scale to assist
the diagnosis of autism and Asperger’s disorder in adults (RAADS): a pilot study.
J Autism Dev. Disord, 2008; 38(2): 213–223.
Rutter M. Incidence of autism spectrum disorders: changes over time and their
meaning. Acta Paediatr, 2005; 94(1): 2–15.
Silberman,S. (2015). NeuroTribes: The legacy of autism and how to think smarter
about people who think differently. Allen & Unwin: London.
20 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre
Tyrer F, Leaver A, Lewis S, Lovett C, van Rensburg K, Seaton S, Smith J, Brugha T.
(2013) Testing adults for possible social and communication disorders, including
autism spectrum disorders. p.83 Leicester: University of Leicester.
21 | APMS 2014 | Chapter 6: Autism spectrum disorder | Copyright © 2016, Health and Social Care Information Centre