The prevalence of self-harm and mental disorders among individuals with intellectual disabilities

Abstract Objective Mental health disorders are prevalent among individuals with intellectual disabilities (ID). However, there is a lack of research on the impact of concomitant autism spectrum disorders (ASD) or attention deficit hyperactivity disorder (ADHD) on the mental health within this population. We aimed to investigate the prevalence of mental health disorders and registered healthcare visits due to self-harm among individuals with ID. Method We used administrative data for all healthcare with at least one recorded diagnosis of mental health disorder or self-harm during 2007–2017 among people with a diagnosis of Down syndrome (DS; n = 1298) and with ID without DS (IDnonDS; n = 10,671) using the rest of the population in Stockholm Region (n = 2,048,488) for comparison. Results The highest odds ratios for a mental health disorder were present in females with IDnonDS (9.01) followed by males with IDnonDS (8.50), compared to the general population. The ORs for self-harm among individuals with IDnonDS were high (8.00 for females and 6.60 for males). There were no registered cases of self-harm among individuals with DS. The prevalence of an anxiety or affective disorder was higher among individuals with ID including DS with concomitant ASD or ADHD. Neighbourhood socio-economic status was associated with a lower occurrence of mental health disorders and self-harm in wealthier areas for all outcomes and for all groups. Conclusions Self-harm and psychiatric comorbidities were common among individuals with ID without DS with an attenuated difference among those with concomitant ASD or ADHD, which calls for attention.


Introduction
Individuals with intellectual disabilities (ID) are for many reasons at increased risk for poor mental health, but it is important to disentangle mental health from conditions today regarded as impairments and these should be studied as conceptually distinct phenomena [1].
In addition, when studying mental health disorders, different neurodivergent conditions such as autism spectrum disorder (ASD), and attention deficit hyperactivity disorder (ADHD) should be separated from ID, as associated with a high incidence of mental health disorders [2,3].All three groups are heterogenous, but even so, within the group with ID, individuals with Down syndrome (DS) constitute a unique population and should be studied separately.It can be justified by abnormal conditions earlier reported to be present in this group and in many cases deviating from the large group with individuals with ID [4][5][6][7].In the general population there is an association of physical and psychiatric comorbidity with socio-economic status (SES), but this might not be the case for the population of individuals with ID, as multimorbidity has been reported to be on a par in the most affluent and the most deprived areas [8].
The population with ID in Sweden has little access to healthcare units specialized in disabilities such as ID, thus professionals working on all levels in the healthcare organization often have scant experience of meeting patients with ID.In addition, in Sweden, all institutions have been closed for decades, as part of a de-institutionalization process.Most adults with ID live in supported housing in the community, with a varying degree of support [9].
A population-based survey from Australia that combined registers of ID and psychiatric registers, reported that 32% of individuals with ID also had a mental disorder diagnosis [10].A national study from Scotland based on linkage of administrative data with census data reported 7-fold higher numbers comparing co-occurring mental health conditions in individuals with ID, when compared with the general population [3].Both studies excluded ASD or problem behaviours from other mental health disorders, which in other reports have been included, resulting in a risk of confusion of mental health disorders with the disability per se or with specific behaviours.A Swedish national register study on suicidal behaviours reported that as much as 60 to 68% of individuals with ID had some kind of psychiatric comorbidity, when concomitant ASD or ADHD diagnosis was present [2].
Various terms have been used to describe non-fatal suicidal behaviour (e.g.para-suicide, suicide attempt, self-injurious behaviour) and none is entirely satisfactory.In later years, self-harm has been used to refer to an intentional act of self-poisoning or self-injury, irrespective of motivation.In the context of ID, the term self-injurious behaviour (SIB) has been used since the role of emotional distress for this behaviour is uncertain in this group [11].In individuals with ID, specific syndromes are associated with prevalent SIB from early ages [12].This has been described mainly as a pathological behaviour, but during later decades more psychosocial explanations for the behaviour have emerged [13].In a review of individuals with ID, SIB was proposed to be derived from general somatic pain [12].SIB is described as a prevalent form of challenging behaviour and associated with concomitant ASD as well as severity of ID [14].
There is a growing awareness of a large variation of reasons behind self-harm among individuals with ID.The similarity and differences between individuals with or without ID who self-harm, and if self-harm equates to SIB, has been debated in the research community for decades [15][16][17].The population with ID with moderate and mild ID are increasingly involved in interview studies with descriptions of their experiences of self-harm, this including individuals with syndromes associated to SIB such as Prader Willis syndrome [17].The awareness of methods available and theories behind a better understanding of challenging behaviour among individuals with ID is being more used in practices reducing challenging behaviour including self-harm, by the use of a low arousal approach [18].Also, among individuals with more severe ID and concomitant biological or psychosocial reasons for self-harm, referred to as SIB, it is suggested that instead of referring SIB to severity of ID, the emotional development and thus emotional needs should be probed and met [19].Self-harm has been reported to occur among 95% of all individuals with specific syndromes, such as Smith-Magenis Syndrome or Cornelia de Lange Syndrome [12].However, these syndromes are rare in the total population of individuals with ID.
Self-reflection and the ability to verbalize emotions is often difficult, and an aspect of the disability [20].Concurrent mental disorders have been suggested to increase the risk for suicide among individuals with ID, however research in this field is sparse [21].A recent Swedish population-based register study reported a 9-fold increased risk of suicide among individuals with ASD, compared to the general population [22].For individuals with ID and concomitant ASD, the increased risk of suicidal behaviour was not as high, but still more than doubled the risk as in the general population [2].
We hypothesized that a variation in prevalence of mental health disorders is depending on concomitant ASD or ADHD in individuals with ID.We assumed that self-harm for at least some be a proxy for mental and emotional ill health.

Aims of the study
We aimed to investigate the prevalence of mental health disorders among individuals with intellectual disabilities with and without concomitant ASD or ADHD.We also aimed to explore the number of individuals with ID visiting any healthcare setting for self-harm.A third aim was to explore associations between self-harm in individuals ID with area-level deprivation.

Data sources
The total number of residents in the Stockholm Region as of 1st January 2013, was 2.1 million.Administrative data for all primary healthcare, specialist outpatient care and inpatient care, was used to identify individuals with at least one recorded diagnosis of any mental disorder between 2007 and 2017.Data were drawn from VAL, the central administrative database in the Stockholm Region, covering at least 97% of all healthcare in the region, described in detail elsewhere [23].Self-harm is recorded in the patient registers for any injury (reason for the attendance) judged by healthcare personal as something caused by him or herself.

Groups
We stratified the study population (2.1 million inhabitants) into five groups according to recorded diagnosis (at least one) with priority as follow: DS group (n = 1298); IDnonDS group = ID excluding individuals with DS (n = 10,671).ASD group without DS and without ID (n = 18,625); ADHD group without DS, ID or ASD (n = 50,353) and GP group = general population without any of these diagnoses (n = 2,048,488).Individuals with ASD and ADHD were excluded from the GP group in the analyses.For the analysis of the effect from concomitant ASD or ADHD, the groups with DS and ID was combined to IDplusDS to achieve a good sample size.
Diagnoses were obtained from VAL, i.e. originally recorded in the electronic patient records.We used International Classification of Diseases, 10th revision codes for registered diagnoses for psychotic disorders F20-F29; affective disorders F30-F39; anxiety disorders F40-F43.Exposure for self-harm was defined as intentional self-harm X60-X84; or if undetermined intent of self-harm y10-y34 [25,26].In this paper, we separate certain disabilities and neurodivergent conditions (ID, DS, ASD, ADHD) from any other mental disorder in order to separate these conditions from a disease.

Socio-demography
Neighbourhood SES was classified into three levels, i.e. high, middle or low, according to the Mosaic tool, a system based on postal codes.The Mosaic tool was originally developed to categorize consumers by a marketing company (Experian) to streamline sale activities.It uses a multivariate modelling including more than 400 variables and is suitable for epidemiologic research and includes data from 23 countries [27,28].

Ethics
All data handled were anonymized and none of the individuals could be identified.Management and analysis based on the VAL database is part of continuous quality control for healthcare use in the Stockholm Region.Ethical approval was obtained from the regional ethical review board in Stockholm to study separate diagnoses and concomitant diseases using these data, Dnr 2021-05735-02.

Statistical methods
Standard descriptive statistics were used to present population characteristics.Age-adjusted odds ratios with 95% confidence intervals were calculated.Statistical analysis and data management were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC).

Results
The study population included 11,969 individuals with ID, with and without DS.These individuals were compared to the general population with 2,048,488 individuals without these diagnoses.Furthermore, a total of 68,978 individuals with ASD or ADHD were excluded from the general population comparison (Table 1).The IDnonDS group had a female/ male ratio of 0.76 and the DS group had a ratio of 0.84.In the group with ID, including individuals with DS, 22.2% had a concomitant diagnosis of ASD or ADHD.
Compared to the general population, mental disorders as well as intentional self-harm were more common in the group with ID when the DS group was excluded (Tables 2-5).In all age groups, there were only minor differences in the prevalence of psychotic disorders between females and males.However, females had a higher prevalence in all age groups and females with ID without DS the highest prevalence of affective or anxiety disorder (Tables 3 and 4).An attendance to healthcare caused by any kind of intentional self-harm was 6 times as common among women with ID compared to the general population, but there were no entries in this register of intentional self-harm for females or males with DS (Table 5).

Psychotic disorders
For psychotic disorders, the prevalence among individuals with DS was on the same level as in the general population, but with no cases in the age group 0-18 years among individuals with DS (Table 2).The odds ratio for having a psychotic disorder among individuals with IDnonDS was 9.01 for females and 8.49 for males (Table 6), and the prevalence increased with age (Table 2).Psychotic disorders increased for females with IDplusDS with concomitant ASD and for males with IDplusDS and concomitant ASD and ADHD (Table 6).

Affective and anxiety disorders
For affective and anxiety disorders, the prevalence was higher among both females and males with IDnonDS than in the general population (Tables 3 and 4).Among both females and males with DS, the prevalence of both affective and anxiety disorders was lower than in individuals with IDnonDS as well as in the general population (Tables 3 and 4).For affective and anxiety disorders, the prevalence was higher among females with IDnonDS than among males (Tables 3 and 4).The prevalence of an affective disorder was higher for females with IDplusDS with concomitant ADHD (Table 6).The prevalence of anxiety disorder was higher among both females and males with IDplusDS with concomitant ADHD (Table 6).The highest prevalence for an anxiety disorder (42%), was found among females aged 19-49 years with IDnonDS (Table 4).

Self-harm
Intentional self-harm was the reason for healthcare attendance in 4% of females aged 19-49 years old with IDnonDS compared to 0.5% in the general population (Table 5).Compared to the general population, both females and males with IDnonDS had higher ORs for healthcare attendance with intentional self-harm (5.32 and 4.42, respectively; Table 7).The ORs were even higher in those with concomitant ADHD or ASD in the IDplusDS group (Table 7).No attendances for intentional self-harm had been registered for any individual with DS (Table 5).Healthcare attendances for self-harm of undetermined intent were more common in all groups compared to the general population, with the exception of DS, in which there were less registered incidents.
Females with IDplusDS and concomitant ADHD had the highest number of recorded self-harms of undetermined intent (Table 8).

Socio-economic level
Neighbourhood SES was associated for all outcomes, with a lower occurrence of mental health disorders and self-harm in wealthier areas for all groups, for both females and males, with higher occurrence in the most deprived neighbourhoods versus both medium and high socio-economic status areas.

Discussion
We found that attendance to healthcare for problems registered as caused by self-harm was much more common among individuals with ID compared to the general population.In addition, we can confirm earlier reports on the higher prevalence of mental health disorders among individuals with ID compared to the population in general, and even more so if the individuals have a concomitant diagnosis of ASD or ADHD [2,3].Females with ID had a high prevalence of affective and anxiety disorders as well as healthcare attendance for self-harm, which recently was reported also from Canada [29].

Mental health disorders among individuals with ID
As much as 77% of our study population with ID did not have any recorded concomitant ASD or ADHD and the prevalence of mental health disorders was high within the population with ID also without ASD or ADHD, partially confirming earlier reports [10].At the same time, findings should be cautiously interpreted given the imperfect diagnostic practice regarding all these diagnoses -intellectual disability as well as ASD and ADHD and mental health disorders -made from verbal report of symptoms, or proxy-descriptions from relatives or staff, when it comes to individuals with ID [30,31].
Our results are well in line with a national register-based study on individuals with ID or ASD in Sweden [32], which however reported a much higher prevalence of psychotic disorders, possibly explained by their study population of older  individuals using disability support and service, whereas we studied the whole population.
We could see a strong increase in the prevalence of psychotic disorders from childhood to adulthood and even more so after the age of 50.In this older group, dementia could be the underlying cause especially in the population with DS [33].In particular, the high prevalence of psychotic disorders among individuals with ID was, however, consistent with what has been reported [10,34,35].
The higher prevalence of affective as well as anxiety disorders that we report, although confirmed by others [36], needs to be interpreted with some caution.A study from the uK investigated the prevalence of affective disorders in a large population with ID identified by GPs, with a comprehensive clinical examination and derived diagnoses according to three different manuals, DC-LD (Royal College of Psychiatry), ICD and DSM, performed by experienced GPs and nurses [37].They reported a higher prevalence of affective disorders in individuals with ID than in the general population, with some differences in factors associated with depression.In contrast, a study from Australia reported that depressive disorders were less common among individuals with ID and with no differences in anxiety disorders, compared to individuals without ID [35].

Self-harm among individuals with ID
As many as four percent of adult females (age 19-49 years old) with ID in this studied population, compared to only a half percent of women without ID, had attended healthcare at least once for a reason defined by the physician as self-harm during these years.Regardless of the reason for self-harm, research reports lack in describing even basic physical examinations in connection with records of self-harm when the population of ID is studied [12].Examples of common physical causes of self-harm such as constipation, Table 6.odds ratios of psychiatric disorders in individuals with intellectual disability with and without including down syndrome and with and without concomitant autism spectrum disorders (asd), and attention deficit hyperactivity disorder (adhd), and common combinations of these diagnoses, when using the general population as referents, presented for females and males.Presented as odds ratios (or;95% confidence intervals).
figures in bold means ors significantly over 1.0.Table 7. occurrence of at least one attendance to any healthcare unit for an event recorded as being caused by intentional self-harm or an event of undetermined intent in individuals with intellectual disability with and without including down syndrome and with and without concomitant autism spectrum disorders (asd) and attention deficit hyperactivity disorder (adhd), and common combinations of these diagnoses, when using the general population as referents, presented for males and females.gastro-esophageal reflux disease, dental problems, urinary tract infection, otitis media, presence of foreign body or fracture may be missed, due to communication difficulties.
Although the research literature since decades added psychosocial factors explaining the self-harm, such as poor communication or lack of support, this has not always reached clinicians in primary healthcare [31].
A strength of our study is the inclusion of the whole population of individuals with diagnosed ID, including all ages, both with and without access to special support and service.
Many of these individuals have some degree of communication difficulties caused by their intellectual impairment, which is a challenge for health and medical care.As suggested by Maiano, potential inadequacy regarding current diagnostic criteria and procedures could be the reason behind some of the attendances to healthcare caused by self-harm Maïano [38].Self-harm as well as anxiety and depressive symptoms, has also been reported to be common among individuals with ID exposed to sexual abuse [38,39].In our view, healthcare needs to pay special attention to these patients, their function and life situation.However, this register study does not specify if diagnoses of self-harm represent SIB associated to certain syndromes, or what causes the self-harm.We can only report that the studied population with ID had an increased risk of self-harm and that further studies exploring the causes are warranted.

Down syndrome, co-occurring mental health disorders and self-harm
We report a prevalence of mental health disorders in individuals with DS comparable to that seen in the general population, but lower than in individuals with ID without DS.Some of these cases may be unidentified cases of dementia if misinterpreted as a mental health disorder instead of brain dysfunction caused by dementia (Holland 2000).The literature on psychiatric comorbidities among individuals with DS warrants further studies [40] and results are contradictory except for the increased prevalence of dementia [41].A review of the literature could not support the earlier assumption of a higher prevalence of depression among individuals with DS, and that is confirmed by our results and others [42,43].Moreover, our study reported null registrations of self-harm among individuals with DS.Our results are supportive of earlier findings that DS might involve a protective factor for mental illness [6,7,10] as well as SIB, which Cooper and colleagues have suggested [44].

Socio-economy, sex, age and healthcare utilization
Surprisingly, in our study, socio-economic status influenced the occurrence of mental disorders as well as attendances caused by self-harm, despite that a large part of the studied population are expected to live under similar living conditions, concerning those getting service and support provided in the community by the Swedish social security system and the law 'Support and Service for Persons with Certain Functional Impairments law' (Swedish acronym: LSS) [45,46].A recent report from Canada reported females with ID compared to females with other disabilities to be the most socioeconomically vulnerable [29].
Prior studies report self-harm to be slightly more common among females than males with ID [14].This is similar to our findings but not present in all age groups.We found that the largest differences between females and males were in those having an affective or anxiety disorder.However, many studies do not report on sex differences, as observed in a recent review and meta-analysis on the prevalence of anxiety and depression in younger individuals with ID [38].
We report results for a young population with ID compared to the general population.Co-morbidity increase with age among individuals with ID similar to the general population [3,8].There were fewer individuals above 50 years of age, in the groups DS as well as IDnonDS for several reasons; one is that individuals with DS have a shorter life span.Thus, as life expectancy for individuals with ID have increased in recent years, the prevalence of mental health disorders in this group is expected to increase in the future.

Limitations and strengths
We have defined our exposures based on diagnostic codes on single occasions, which is a limitation with regard to known difficulties with the stability of diagnoses such as ID, ASD as well as ADHD and their combinations, especially since the reporting of these diagnoses have changed over time.Another limitation is the clinical difficulties with psychiatric diagnostics in this population, as they derive from verbally reported Table 8. occurrence of at least one attendance to any healthcare unit for an event recorded as being caused by an undetermined event a in the population of females and males in stockholm region in groups with at least one recorded diagnosis of intellectual disability or down syndrome during 2007-2017 in all levels of care, and the general population without any of these diagnoses.symptoms.This includes mental health disorders as well as ID, ASD or ADHD.In addition to this, the lack of specialist clinics may be another reason for possible misdiagnosing.A further limitation is that we, due to small numbers, had to combine ID and DS when dividing the studied population in groups with concomitant ASD or ADHD.This probably attenuated the result in the IDplusDS with ASD or ADHD groups, for both mental health disorders as well as for self-harm.
A strength of this study was the thorough inclusion of different diagnoses representing ID and the separation of DS and when concomitant ASD and ADHD were present.Another strength was access to the Swedish database VAL with registrations from all levels of healthcare within the Stockholm Region.Furthermore, the Stockholm Region is heterogeneous with regard to socio-economic conditions and comprises more than 20% of the total Swedish population.

Conclusions
Self-harm and psychiatric comorbidities were common among individuals with ID without DS with an attenuated difference among those with concomitant ASD or ADHD.This calls for attention and actions to improve care for this population.
Especially women with ID need extra awareness.
and delusional disorders icd codes f20-f29.b ds: down syndrome.c idnonds: intellectual disabilities non down syndrome.d Numbers given with percentage in each age group within brackets.

Table 1 .
Population of females and males in stockholm region by 1 Jan 2013, within groups with at least one recorded diagnosis of intellectual disability or down syndrome during 2007-2017 in all levels of care, and the general population without any of these diagnosis.
a ds: down syndrome.b idnonds = intellectual disabilities non down syndrome.c Numbers given with percentage in each age group within brackets.

Table 2 .
Psychotic disorders a in the population of females and males in stockholm region in groups with at least one recorded diagnosis of intellectual disability or down syndrome during 2007-2017 in all levels of care, and the general population without any of these diagnosis.

Table 3 .
affective disorders a in the population of females and males in stockholm region in groups with at least one recorded diagnosis of intellectual disability or down syndrome during 2007-2017 in all levels of care, and the general population without any of these diagnosis.ds: down syndrome.c idnonds: intellectual disabilities non down syndrome.d Numbers given with percentage in each age group within brackets. b

Table 4 .
anxiety disorders a in the population of females and males in stockholm region in groups with at least one recorded diagnosis of intellectual disability or down syndrome during 2007-2017 in all levels of care, and the general population without any of these diagnoses.
a Neurotic, stress-related and somatoform disorders icd codes f40-f43.b ds: down syndrome.c idnonds: intellectual disabilities non down syndrome.d Numbers given with percentage in each age group within brackets.

Table 5 .
occurrence of at least one attendance to any healthcare unit for an event recorded as being caused by intentional self-harm a in the population of females and males in stockholm region in groups with at least one recorded diagnosis of intellectual disability or down syndrome during 2007-2017 in all levels of care, and the general population without any of these diagnoses.
a intentional self-harm icd-codes X60-X84.b ds: down syndrome.c idnonds: intellectual disabilities non down syndrome.d Numbers given with percentage in each age group within brackets.
Event of undetermined intent icd-codes y10-y34.b ds: down syndrome.c idnonds: intellectual disabilities non down syndrome.d Numbers given with percentage in each age group within brackets. a