Prevalence of mental health conditions, sensory impairments and physical disability in people with co-occurring intellectual disabilities and autism compared with other people: a cross-sectional total population study in Scotland

Objectives To investigate prevalence of mental health conditions, sensory impairments and physical disability in children, adults and older adults with co-occurring intellectual disabilities and autism, given its frequent co-occurrence, compared with the general population. Design Whole country cohort study. Setting General community. Participants 5709 people with co-occurring intellectual disabilities and autism, compared with 5 289 694 other people. Outcome measures Rates and ORs with 95% CIs for mental health conditions, visual impairment, hearing impairment and physical disability in people with co-occurring intellectual disabilities and autism compared with other people, adjusted for age, sex and interaction between age and co-occurring intellectual disabilities and autism. Results All four long-term conditions were markedly more common in children, adults and older adults with co-occurring intellectual disabilities and autism compared with other people. For mental health, OR=130.8 (95% CI 117.1 to 146.1); visual impairment OR=65.9 (95% CI 58.7 to 73.9); hearing impairment OR=22.0 (95% CI 19.2 to 25.2); and physical disability OR=157.5 (95% CI 144.6 to 171.7). These ratios are also greater than previously reported for people with either intellectual disabilities or autism rather than co-occurring intellectual disabilities and autism. Conclusions We have quantified the more than double disadvantage for people with co-occurring intellectual disabilities and autism, in terms of additional long-term health conditions. This may well impact on quality of life. It raises challenges for staff working with these people in view of additional complexity in assessments, diagnoses and interventions of additional health conditions, as sensory impairments and mental health conditions in particular, compound with the persons pre-existing communication and cognitive problems in this context. Planning is important, with staff being trained, equipped, resourced and prepared to address the challenge of working for people with these conditions.


Background
People with intellectual disabilities [1][2][3][4] and people with autism 5-7 have more mental and physical health needs than other people. A whole population study of Scotland reported that 21.7% of people with intellectual disabilities also had autism, and 18.0% of people with autism also had intellectual disabilities, 8 so this dually diagnosed group warrant investigation. One would suspect that this population with co-occurring intellectual disabilities and autism is likely to have a high level of additional health needs, but this has received little previous attention. A higher number of additional health needs increases the likelihood of misdiagnosis, and treatment interactions, so requires more complex treatment plans. Hence it is important to investigate long-term additional health needs experienced by people with co-occurring intellectual disabilities and autism.
Some studies have investigated mental ill-health in people with co-occurring intellectual disabilities and autism. A small study of 149 adults with severe or profound intellectual disabilities and autism, living in state-run developmental centres in Louisiana, USA, compared comorbidity with 158 adults with intellectual disabilities without autism in the same centres. The former group had more symptomology for anxiety, mania, schizophrenia, stereotypies, selfinjurious behaviour, eating disorders, sexual disorders, and impulse control. 9 A study in Norway compared 62 adults with co-occurring autism and intellectual disabilities under the care of autism services, with 132 adults with intellectual disabilities only receiving intellectual disabilities support. 10 High levels of psychiatric disorders were reported in both groups; 53.2% in the co-occurring intellectual disabilities and autism group, and 17.4% in the intellectual disabilities only group. An English study of referrals to a specialist intellectual disabilities psychiatric service described 42% of the 137 referred adults who had autism as well as intellectual disabilities to have comorbid psychopathology, most commonly schizophrenia. 11 A study of youth aged 14-20 years age, gender matched 36 people with co-occurring intellectual disabilities and autism with 36 people with intellectual disabilities without autism. 12 They reported the former group to have more episodes of mental illhealth, most commonly depression. A study of people aged 8-29 years with intellectual disabilities and challenging behaviour living in four residential units in England included 69 who also had autism and 13 who did not. 13 They reported a higher prevalence of organic disorders, anxiety and stereotypies in the young people with co-occurring intellectual disabilities and autism. This literature is difficult to summarise overall, as, as well as having small sample sizes, the participants were not drawn from representative populations.
A further study had the advantage of being population-based, but was still small in size. 14 It compared the prevalence, and incidence, of mental ill-health in 77 adults with co-occurring intellectual disabilities and autism with 946 adults with intellectual disabilities without autism, and also with 154 individually age, gender, ability-level, and Down syndrome matched controls.
The adults with autism had a higher point prevalence of problem behaviours than the 946 without autism, but compared with the 154 matched controls there was no difference in prevalence, or incidence of either problem behaviours or other mental ill-health. 14 Three large whole population studies have reported that of people with intellectual disabilities, 21.7% reported mental health conditions; 3 and of people with autism, 33.0% of adults, 6 and F o r p e e r r e v i e w o n l y 6 7.6% of children 7 reported mental health conditions, but did not report the rates for people with co-occurring intellectual disabilities and autism.
With regards to sensory impairments, of the 36 matched youth with intellectual disabilities with and without autism, 38.9% with autism reported having visual problems compared with 50.0% without autism, and 13.9% with autism reported having hearing problems compared with 19.4% without autism. 12 An intellectual disabilities register study reported that 95 of the 368 (25.8%) adults with intellectual disabilities who had visual impairment also had markers for autism, compared with 422 of 2,674 (16%) of those who had normal vision, and that 46 of the 60 (76.7%) of the adults with intellectual disabilities and congenital blindness also had markers for autism, compared with only 36 of the 67 (53.7%) with normal vision. 15 We have not identified other papers on sensory impairments or any on physical disabilities in people with co-occurring intellectual disabilities and autism. However, previous large whole population studies have reported that of people with intellectual disabilities, 12.4% reported blindness/sight loss, 13.1% reported deafness/hearing loss, and 32.6% reported physical disability. 3 Of people with autism, 12.1% of adults 6 and 3.5% of children reported blindness/sight loss, 7 14.1% of adults 6 and 2.9% of children reported deafness/hearing loss, ,7 and 24.0% of adults 6 and 10.7% of children reported physical disability. 7 They did not, however, report the rates of these conditions for people with co-occurring intellectual disabilities and autism.
Given the frequent overlap of intellectual disabilities and autism, information on the associated comorbid conditions is important, to assist policy makers, planners, and practitioners to best adapt services for individuals with cooccurring intellectual disabilities and autism. This paper aims to investigate the prevalence of mental health conditions, sensory impairments, and physical disability in children, adults, and older adults with co-occurring intellectual disabilities and autism, compared with other people.

Approval
Approval was obtained from the Scottish Government to undertake secondary data analysis of Scotland's Census, 2011.

Data Source
Scotland's Census provides information on Scotland's population every ten years, with the most recent Census on 27 th March 2011. 16 The Census provides information on the number and characteristics of Scotland's population and households on the Census date.
It is a legal requirement to complete the census form and households were informed that failure to make a Census return, or supplying false information could result in a £1,000 fine. A very high response rate was achieved, with an estimated 94% of all of Scotland's population completing the Census. The Census team used a Census Coverage Survey with about 40,000 households, to estimate numbers and characteristics of the missing 6%. 17 The Coverage Survey and Census records were deterministically matched to check for duplicates.
Individuals estimated to have been missed were then imputed using a subset of characteristics from real individuals. The edit and imputation methodology The Census was completed by the head of each household on behalf of all occupants of the household. We consider it unlikely that people with cooccurring intellectual disabilities and autism completed the form, given the reading age required to do so. Rather, we expect that the people who completed the form on their behalf were parent-carers in family households, support workers for people living in supported accommodation, and the managers/key workers at communal establishments.

Variables
The census included a question to identify people with intellectual disabilities and autism, as well as mental health conditions, sensory impairments, and physical disabilities: 'Do you have any of the following conditions which have lasted, or are expected to last, at least 12 months? Tick all that apply'. There was a choice of ten possible responses to this question: deafness or partial hearing loss; blindness or partial sight loss; learning disability (for example, Down's syndrome); learning difficulty (for example, dyslexia); developmental disorder (for example, Autistic Spectrum Disorder or Asperger's Syndrome); physical disability; mental health condition; long-term illness, disease or condition; other condition. For "other condition" the option of providing more detail in an open text response was provided.
In Scotland, the term "learning disability" is used synonymously with that of "intellectual disabilities" used internationally. Importantly, the Census and between intellectual disabilities and autism.
During the methodology development for Scotland's Census, 2011, cognitive question testing was undertaken on the questions on long-term health conditions and disabilities. This was to determine whether the questions were answered accurately, and to identify any changes needed to improve data quality and/or the acceptability of the way questions were phrased. Cognitive interviewing is a widely used approach to critically evaluate and improve survey questionnaires. 18 This approach enables researchers to modify survey material to enhance clarity. Retrospective probing was conducted with 102 participants with a variety of sex, age, and health conditions and disabilities (including people with more than one of the conditions). They included people with autism, intellectual disabilities, dyslexia, dyspraxia, speech impairment, mental health conditions (both milder and more serious), and other long-term conditions. 19 Using the cognitive interviewing results, the question aimed to detect autism was improved and rephrased, to better capture this information.
The questions on intellectual disability, mental health condition, visual impairment, hearing impairment, and physical disability did not require any modifications.
The Census team imputed answers for the 14.7% who did not tick any of the boxes in question on long-term conditions, based on their free text answers for this question and answers to other health questions in the Census, which increased the completion rate to 97.4%. For the remaining 2.6%, the Census team assumed the most plausible explanation was that the person had no

Data Analysis
We calculated the number and rate per 1,000 population of children and adults with co-occurring intellectual disabilities and autism. We then calculated the number and percentage of people with mental health conditions, deafness or partial hearing loss, blindness or partial sight loss, and physical disability, for those with co-occurring intellectual disabilities and autism, compared with individuals who do not have co-occurring intellectual disabilities and autism using chi-squared ( 2 ) tests. For the whole population we then used logistic regression to calculate the odds ratios (OR: 95% confidence interval, 95% CI) of co-occurring intellectual disabilities and autism statistically predicting the binary response of having each of the four specific types of long-term health conditions, adjusted for age and sex. Sex was binary, with males being the reference group. Age was categorised into groups: 0-15, 16-24, 25-34, 35-44, 45-54, 55-64, 65-74, 75+, with 0-15 years as the reference group. We repeated the regressions, including the interaction term of age x co-occurring intellectual disabilities and autism, as people with the most severe disabilities die earlier, which may affect the profile of additional health problems differently to that seen in the general population. The same reference groups were used. All analysis was conducted using SPSS software version 22.

Patient and Public Involvement
The Scottish Learning Disabilities Observatory, where this research was undertaken, has a specific remit for people with intellectual disabilities and people with autism. Its steering group includes partners from third sector organisations and experts by experience, who approved the workplan for this project prior to it commencing. Results from this study will be disseminated for people with intellectual disabilities and autism in an easy-read version via the Scottish Learning Disabilities Observatory website, newsletters, and conference. Compared with the population who did not have co-occurring intellectual disabilities and autism, the population with co-occurring intellectual disabilities and autism had more males (66.0% versus 48.5%;  2 =703.5; df=1; p<0.001);

Characteristics of the Sample
were younger ( 2 =3894.7; df=7; p<0.001); were more likely to have been born in the UK rather than elsewhere ( 2 =101.9; df=1; p<0.001), revealing lesser geographic mobility; and were no different with regards to Caucasian versus non-Caucasian ethnicity ( 2 =1.1; df=1; p=0.3) (   Table 2 shows the proportion of people with co-occurring intellectual   disabilities and autism, who had each of the four additional long-term health conditions, compared to people who did not have co-occurring intellectual disabilities and autism. Some people in the sample had more than one longterm health condition.

Long-term Health Conditions
-Insert Mental health conditions, blindness or partial sight loss, deafness or partial hearing loss, and physical disability were all significantly more common in people with co-occurring intellectual disabilities and autism than people without these co-occurring conditions. The odds ratios after adjusting for age and sex and the interaction term, were substantial, being 131, 66, 22, and 158 respectively. This is important as each of these conditions are disabling and can significantly impact an individual's quality of life. They contribute to high rates of multi-morbidity, which, on top of communication and cognitive problems due to autism and intellectual disabilities, renders assessments, diagnosis, and treatment of additional health problems more complex than for other people.
Across all age groups, blindness, deafness, and physical disability were more common in females then males with co-occurring intellectual disabilities and autism, unlike the gender ratios in people without co-occurring intellectual disabilities and autism. Mental health conditions were more common in males than females with co-occurring intellectual disabilities and autism, except for the 65+ year group, contrary to the gender ratios in other people. All conditions were more prevalent with increasing age in the people with cooccurring intellectual disabilities and autism, except for physical disability which was more common in the children/youth and older people than in the adults.

Comparison with existing literature
The prevalence of these additional long-term health conditions has seldom been investigated in people with co-occurring intellectual disabilities and autism, particularly in comparison with other people, and never, to our knowledge, as a total population study. All of the long-term health conditions Smaller, less representative studies have reported a higher rate of mental health conditions in adults and youth with co-occurring intellectual disabilities and autism compared with those with intellectual disabilities and without autism, [9][10][11][12][13] but not all. 14 People with autism have been reported to have more mental health conditions than other people (OR=9 in adults and OR=16 in children), 6,7 as have people with intellectual disabilities compared with other people (OR=7), 3 using the same Scotland's Census, 2011 data as in this current paper, whereas the comparable ratio we now report for people with cooccurring intellectual disabilities and autism for mental health conditions was OR=26. Having the co-occurring conditions therefore presents a much higher risk of mental health conditions than either intellectual disabilities or autism on their own.
The previous small study of youth reported lower rates of visual and hearing impairments in those with co-occurring intellectual disabilities and autism (38.9% and 13.9%) compared with those with intellectual disabilities but without autism (50% and 19.4%). 12 This was in contrast with the larger study reporting more autistic symptoms in adults with intellectual disabilities and visual impairments than in adults with intellectual disabilities but without visual impairments. 15 Adults with autism have been reported to have more blindness or partial sight loss, and deafness or partial hearing loss than other people (12.1% and 17.5%), 6 as have children with autism (3.5% and 2.9%), 7 and people (children and adults combined) with intellectual disabilities compared with other people (13.1% and 12.4%), 3  as in this current paper. This current study found the comparable rates for people with co-occurring intellectual disabilities and autism for blindness or partial sight loss, and deafness or partial hearing loss was 21.7% and 19.3% for adults, and 16.6% and 10.3% for children. Having the co-occurring conditions of intellectual disabilities and autism therefore presents a much higher risk of sensory impairments than for children and adults with autism, and for people with intellectual disabilities (although children were not separately studied in the previous report).
Regarding physical disability, 32.6% of people with intellectual disabilities were previously reported to have physical disability. 3 Of people with autism, 24.0% of adults and 10.7% of children reported physical disability. 6,7 These rates are lower than those we report in this current study of people with co-occurring intellectual disabilities and autism -45.6% of children and 42.2% of adults.

Strengths and limitations
Strengths of the study include its large scale and general population comparison group, whole population coverage and very high response rate so the results are representative of the whole population. Intellectual disabilities, autism, and the long-term health conditions were enquired about systematically for everyone in the population. We believe the results are therefore generalisable to other high-income countries. The cognitive question testing during the design of the Census is a further strength. The Census had clear categories to distinguish between intellectual disabilities, specific learning disability (like dyslexia), and autism. Limitations include the proxy-reporting, which may, or may not reflect selfreports. However, without proxy-reports, we would have no information on people unable to self-report due to their disabilities, and a previous review on the topic concluded that overall, proxy reports are a useful addition to determine aspects of well-being in people with intellectual disabilities. 20 Additionally, people were reported who were known to have autism/Asperger's syndrome, intellectual disabilities, and the additional longterm health conditions, rather than detailed individual research assessments being undertaken which are clearly not possible in such large population studies, and may therefore be subject to a degree of error which we were not able to check.
The researchers are independent from the funders.

Competing interests
The authors declare no competing interests.

Author's contributions
KD analysed and interpreted the data, and wrote the first draft of the manuscript. ER contributed to data access, data interpretation, and drafting the manuscript. S-AC conceived and managed the project, interpreted data, and contributed to writing the manuscript. All approved the final version of the manuscript. S-AC is the study guarantor.

Data sharing
Data is available via National Records of Scotland, following project approval.

Strengths and limitations of this study
 Large scale, whole country cross-sectional study, with a high response rate (94%), so the results are representative of the whole population.
 Intellectual disabilities, autism, and additional long-term conditions were enquired about systematically for everyone in the population.
 The wording of questions was tested in advance, via cognitive question testing during the design of Scotland's Census, 2011.
 Limitations include proxy-reporting.
 People known to have autism/Asperger's syndrome, intellectual disabilities, and the four long-term conditions were reported, rather than each undergoing detailed individual research assessments which are not possible in such large population studies. intellectual disabilties and 79 with autism but no intellectual disabilities. 8 They reported no statistical difference between the two groups for claims for psychiatric disorders. Claim rates were high over all (67.2% for those who additionally had intellectual disabilties, and 75.9% for those who did not). 8 The authors state the autism only sample was probably skewed towards the lower end of socioeconomic status of all people with autism. 8 A study of people aged 8-29 years with intellectual disabilities and challenging behaviour living in four residential units in England included 69 who also had autism and 13 who did not. 14 They reported a higher prevalence of organic disorders, anxiety and stereotypies in the young people with co-occurring intellectual disabilities and autism. This literature is difficult to summarise overall, as, as well as having small sample sizes, the participants were not drawn from representative populations.
A further study had the advantage of being population-based, but was still small in size. 15 It compared the prevalence, and incidence, of mental ill-health in 77 adults with co-occurring intellectual disabilities and autism with 946 Of people with autism, 12.1% of adults 6 and 3.5% of children reported blindness/sight loss, 7 14.1% of adults 6 and 2.9% of children reported deafness/hearing loss, ,7 and 24.0% of adults 6 and 10.7% of children reported physical disability. 7 They did not, however, report the rates of these conditions for people with co-occurring intellectual disabilities and autism.
Given the frequent overlap of intellectual disabilities and autism, information on the associated comorbid conditions is important, to assist policy makers, planners, and practitioners to best adapt services for individuals with cooccurring intellectual disabilities and autism. This paper aims to investigate the prevalence of mental health conditions, sensory impairments, and physical disability in children, adults, and older adults with co-occurring intellectual disabilities and autism, compared with other people.

Approval
Approval was obtained from the Scottish Government to undertake secondary data analysis of Scotland's Census, 2011.

Data Source
Scotland's Census provides information on Scotland's population every ten years, with the most recent Census on 27 th March 2011. 17 The Census provides information on the number and characteristics of Scotland's population and households on the Census date. The Census was completed by the head of each household on behalf of all occupants of the household. We consider it unlikely that people with cooccurring intellectual disabilities and autism completed the form, given the reading age required to do so. Rather, we expect that the people who completed the form on their behalf were parent-carers in family households, support workers for people living in supported accommodation, and the managers/key workers at communal establishments.

Variables
The census included a question to identify people with intellectual disabilities and autism, as well as mental health conditions, sensory impairments, and physical disabilities: 'Do you have any of the following conditions which have lasted, or are expected to last, at least 12 months? Tick all that apply'. There was a choice of ten possible responses to this question: deafness or partial hearing loss; blindness or partial sight loss; learning disability (for example, In Scotland, the term "learning disability" is used synonymously with that of "intellectual disabilities" used internationally. Importantly, the Census differentiated between intellectual disabilities and specific learning disabilities; and between intellectual disabilities and autism. During the methodology development for Scotland's Census, 2011, cognitive question testing was undertaken on the questions on long-term health conditions and disabilities. This was to determine whether the questions were answered accurately, and to identify any changes needed to improve data quality and/or the acceptability of the way questions were phrased. Cognitive interviewing is a widely used approach to critically evaluate and improve survey questionnaires. 19 This approach enables researchers to modify survey material to enhance clarity. Retrospective probing was conducted with 102 participants with a variety of sex, age, and health conditions and disabilities (including people with more than one of the conditions). They included people with autism, intellectual disabilities, dyslexia, dyspraxia, speech impairment, mental health conditions (both milder and more serious), and other long-term conditions. 20 Using the cognitive interviewing results, the question aimed to detect autism was improved and rephrased, to better capture this information.
The questions on intellectual disability, mental health condition, visual of age x co-occurring intellectual disabilities and autism, as people with the most severe disabilities die earlier, which may affect the profile of additional health problems differently to that seen in the general population. The same reference groups were used. All analysis was conducted using SPSS software version 22.

Patient and public involvement
The Scottish Learning Disabilities Observatory, where this research was undertaken, has a specific remit for people with intellectual disabilities and people with autism. Its steering group includes partners from third sector organisations and experts by experience, who approved the workplan for this project prior to it commencing. Results from this study will be disseminated for people with intellectual disabilities and autism in an easy-read version via the Scottish Learning Disabilities Observatory website, newsletters, and conference.  Table 2 shows the proportion of people with co-occurring intellectual disabilities and autism, who had each of the four additional long-term health conditions, compared to people who did not have co-occurring intellectual disabilities and autism. Some people in the sample had more than one longterm health condition.

Long-term Health Conditions
-Insert table 2 about here -

Mental Health Condition
Adjusting for age and sex, given the different distributions compared with the general population, having co-occurring intellectual disabilities and autism had an OR=25.553 (23.933-27.282, 95% CI) for mental health conditions (table 3).
When the interaction term was added (age x co-occurring intellectual

Principle findings
Mental health conditions, blindness or partial sight loss, deafness or partial hearing loss, and physical disability were all significantly more common in people with co-occurring intellectual disabilities and autism than people without these co-occurring conditions. The odds ratios after adjusting for age and sex and the interaction term, were substantial, being 131, 66, 22, and 158 respectively. This is important as each of these conditions are disabling and can significantly impact an individual's quality of life. They contribute to high rates of multi-morbidity, which, on top of communication and cognitive problems due to autism and intellectual disabilities, renders assessments, diagnosis, and treatment of additional health problems more complex than for other people.
Across all age groups, blindness, deafness, and physical disability were more common in females then males with co-occurring intellectual disabilities and autism, unlike the sex ratios in people without co-occurring intellectual disabilities and autism. Mental health conditions were more common in males than females with co-occurring intellectual disabilities and autism, except for the 65+ year group, contrary to the gender ratios in other people. All conditions were more prevalent with increasing age in the people with cooccurring intellectual disabilities and autism, except for physical disability which was more common in the children/youth and older people than in the adults.
There are likely to be many biological, social, and environmental reasons accounting for these results. The prevalence of these additional long-term health conditions has seldom been investigated in people with co-occurring intellectual disabilities and autism, particularly in comparison with other people, and never, to our knowledge, as a total population study. While prevalence rates of mental health conditions and impairments in a full country population for individuals with intellectual disabilities, and separately for individuals with autism have been compared to the general population, no such study has been conducted on the prevalence of mental health conditions and impairments for those with co-occurring intellectual disabilities and autism. We found all of the long-term health conditions were more common than in those without co-occurring intellectual disabilities and autism.

Comparison with existing literature
Smaller, less representative studies have reported a higher rate of mental health conditions in adults and youth with co-occurring intellectual disabilities and autism compared with those with intellectual disabilities and without autism, 9- of intellectual disabilities and autism therefore presents a much higher risk of sensory impairments than for children and adults with autism, and for people with intellectual disabilities (although children were not separately studied in the previous report).
Regarding physical disability, 32.6% of people with intellectual disabilities were previously reported to have physical disability. 3 Of people with autism, 24.0% of adults and 10.7% of children reported physical disability. 6,7 These rates are lower than those we report in this current study of people with co-occurring intellectual disabilities and autism -45.6% of children and 42.2% of adults.
Cerebral palsy is associated with intellectual disabilities, but the extent of this association does not appear to account for the differences found in physical disability: cerebral palsy has been reported to occur in 13% of children and Limitations include the proxy-reporting, which may, or may not reflect selfreports. Without proxy-reports, we would have no information on people unable to self-report due to their disabilities, and a previous review on the topic concluded that overall, proxy reports are a useful addition to determine aspects of well-being in people with intellectual disabilities. 23 Additionally, people were reported who were known to have autism/Asperger's syndrome, intellectual disabilities, and the additional long-term health conditions, rather than detailed individual research assessments being undertaken which are clearly not possible in such large population studies, and may therefore be subject to a degree of error which we were not able to check.
Scotland's Census 2011 was administered 8 years ago, and so any potential changes in prevalence of conditions since then are not captured by our analyses. People with intellectual disabilities and autism are both higher health care users than other people and so may receive more diagnoses, but they are also subject to "diagnostic overshadowing". We do not know the extent to which these factors may impact on reporting of mental health conditions, sensory impairments, and physical disability at Scotland's Census, 2011.
However, given the long-term nature of these conditions, any impact is likely to be less than it would be for acute conditions." The study is cross-sectional, rather than longitudinal. We acknowledge that the use of odds ratios may overestimate the strength of associations in crosssectional studies where the prevalence in the general population is very low; calculating odds ratios has enabled us to draw comparisons with previously published results on mental health conditions and impairments of people with intellectual disabilities, and of people with autism. We also present example prevalence differences. It is also important to note that whilst our regressions adjusted for sex and age, the effect sizes for sex and age shown in tables 3-6 might not be the total effect of sex and age on the four outcomes: they show the proportion of the sex, age effect on the odds ratio for the four outcomes that are not mediated through any sex, age effect on co-occurring intellectual disabilities and autism. 24 Hence in our discussions of sex and age, we referred to prevalence rather than odds ratios. The study sponsor and funders had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and

Implications
in the decision to submit the article for publication.
The researchers are independent from the funders.

Competing interests
The authors declare no competing interests.

Author's contributions
KD analysed and interpreted the data, and wrote the first draft of the manuscript. ER contributed to data access, data interpretation, and drafting the manuscript. MF provided advice on response to reviewer comments regarding choice of statistical methods and contributed to the writing of the manuscript. S-AC conceived and managed the project, interpreted data, and contributed to writing the manuscript. All approved the final version of the manuscript. S-AC is the study guarantor.             . These ratios are also greater than previously reported for people with either intellectual disabilities or autism rather than co-occurring intellectual disabilities and autism.

Conclusions:
We have quantified the more than double disadvantage for people with co-occurring intellectual disabilities and autism, in terms of additional long-term health conditions. This may well impact on quality of life.

Strengths and limitations of this study
 Large scale, whole country study, with a high response rate (94%), so the results are representative of the whole population.
 Intellectual disabilities, autism, and additional long-term conditions were enquired about systematically for everyone in the population.
 The wording of questions was tested in advance, via cognitive question testing during the design of Scotland's Census, 2011.
 Limitations include proxy-reporting.
A further study had the advantage of being population-based, but was still small in size. 14 It compared the prevalence, and incidence, of mental ill-health in 77 adults with co-occurring intellectual disabilities and autism with 946 adults with intellectual disabilities without autism, and also with 154 individually age, gender, ability-level, and Down syndrome matched controls.
This high prevalence of health conditions among people with intellectual disabilities and people with autism spectrum disorders is partly attributable health inequalities 17,18 , and to the fact that certain conditions, such as cerebral palsy, are associated with both intellectual and physical disabilities.
Additionally, the socioeconomic status of individuals within these populations is typically lower than for members of the general population 19,20 . While prevalence rates of health conditions in a full country population for individuals with intellectual disabilities and individuals with autism have been compared to the general population in previous work using the Scotland Census 2011 3,4,7 , no such study has been conducted on the prevalence of health conditions for those with both intellectual disabilities and autism.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Given the frequent overlap of intellectual disabilities and autism, information on the associated co-existing conditions is important, to assist policy makers, planners, and practitioners to best adapt services for individuals with cooccurring intellectual disabilities and autism. This paper aims to investigate the prevalence of mental health conditions, sensory impairments, and physical disability in children, adults, and older adults with co-occurring intellectual disabilities and autism, compared with other people.

Data Source
Scotland's Census provides information on Scotland's population every ten years, with the most recent Census on 27 th March 2011. 23 The Census provides information on the number and characteristics of S11cotland's population and households on the Census date.
It is a legal requirement to complete the census form and households were informed that failure to make a Census return, or supplying false information In Scotland, the term "learning disability" is used synonymously with that of "intellectual disabilities" used internationally. Importantly, the Census differentiated between intellectual disabilities and specific learning disabilities; and between intellectual disabilities and autism.

Data Analysis
We calculated the number and rate per 1,000 population of children and adults with co-occurring intellectual disabilities and autism. We then calculated the number and percentage of people with mental health conditions, deafness or partial hearing loss, blindness or partial sight loss, and physical disability, for those with co-occurring intellectual disabilities and autism, compared with individuals who do not have co-occurring intellectual disabilities and autism using chi-squared ( 2 ) tests. For the whole population we then used logistic regression to calculate the odds ratios (OR: 95% confidence interval, 95% CI) of

Patient and Public Involvement
The Scottish Learning Disabilities Observatory, where this research was undertaken, has a specific remit for people with intellectual disabilities and people with autism. Its steering group includes partners from third sector organisations and experts by experience, who approved the workplan for this project prior to it commencing. Results from this study will be disseminated for people with intellectual disabilities and autism in an easy-read version via the Scottish Learning Disabilities Observatory website, newsletters, and conference.  Table 2 shows the proportion of people with co-occurring intellectual disabilities and autism, who had each of the four additional long-term health conditions, compared to people who did not have co-occurring intellectual disabilities and autism. Some people in the sample had more than one longterm health condition.

Long-term Health Conditions
-Insert

Principle findings
Mental health conditions, blindness or partial sight loss, deafness or partial hearing loss, and physical disability were all significantly more common in people with co-occurring intellectual disabilities and autism than people without these co-occurring conditions. The odds ratios after adjusting for age and sex and the interaction term, were substantial, being 131, 66, 22, and 158 respectively. This is important as each of these conditions are disabling and can significantly impact an individual's quality of life. They contribute to high rates of multi-morbidity, which, on top of communication and cognitive problems due to autism and intellectual disabilities, renders assessments, diagnosis, and treatment of additional health problems more complex than for other people.
Across all age groups, blindness, deafness, and physical disability were more common in females then males with co-occurring intellectual disabilities and autism, unlike the gender ratios in people without co-occurring intellectual disabilities and autism. Mental health conditions were more common in males conditions were more prevalent with increasing age in the people with cooccurring intellectual disabilities and autism, except for physical disability which was more common in the children/youth and older people than in the adults.

Comparison with existing literature
The prevalence of these additional long-term health conditions has seldom been investigated in people with co-occurring intellectual disabilities and autism, particularly in comparison with other people, and never, to our knowledge, as a total population study. All of the long-term health conditions were more common than in those without co-occurring intellectual disabilities and autism.
The researchers are independent from the funders.

Data sharing
Data is available via National Records of Scotland, following project approval.