Diseases of the gastrointestinal tract in individuals diagnosed as children with atypical autism: A Danish register study based on hospital diagnoses

Abstract

The purpose of this study is to compare the prevalence and types of diseases (International Classification of Mental and Behavioural Disorders, 10th Edition codes K20–K93) relating to the gastrointestinal tract in a clinical sample of 89 individuals diagnosed as children with atypical autism/pervasive developmental disorder not otherwise specified with 258 controls from the general population. All participants were screened through the nationwide Danish National Hospital Register. The average observation time was 32.9 years, and mean age at the end of the observation period was 48.5 years. Among the 89 cases with atypical autism, a total of 22 (24.7%) were registered with at least one diagnosis of any disease of the gastrointestinal tract, against 47 of 258 (18.2%) in the comparison group (p = 0.22; odds ratio = 1.5; 95% confidence interval = 0.8–2.6). Without reaching statistical significance, the rate of diseases of the gastrointestinal tract was particularly high (odds ratio = 1.2) in those with intelligence quotient < 70. Overall, people with atypical autism had about the same frequency of gastric, intestinal and hepatic diseases as had controls.

Keywords

atypical autism comorbidity gastrointestinal diseases

Introduction

In the International Classification of Mental and Behavioural Disorders, 10th Edition (ICD-10) (World Health Organization (WHO), 1992), atypical autism (AA) is classified as one of the group of disorders collectively termed as pervasive developmental disorders (PDDs) or autism spectrum disorders (ASDs). Within the autism spectrum, childhood autism (CA), AA and Asperger’s syndrome (AS) comprise the main categories. The Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; American Psychiatric Association, 1994) also uses the umbrella term of PDD, with autistic disorder (AD), AS and PDD not otherwise specified (PDD-NOS), as the main categories. The DSM-IV criteria for ASD include the same core criteria as are listed in ICD-10. In ICD-10, the term AA is used to describe individuals who do not share all the features of the classical autistic triad (impairment in social interaction, communication and a restricted range of interest and activities), or individuals where the symptoms present after 3 years of age, and who do not meet the criteria for any of the other conditions within the autistic spectrum. Although frequency estimates for AA are higher than for CA – about 3/1000 (Fombonne et al., 2006) AA is a relatively poorly researched disorder.

As there is an increasing number of people under care for ASD (Rutter, 2005), there is interest in treatment not only for the psychiatric aspects of ASD but also in the prevention and treatment of co-occurring medical conditions that might affect quality of life for the patient and the family (Bauman, 2010).

The co-occurring central nervous system (CNS) conditions that occur with sufficient frequency to be of concern in cases with CA and AA include learning disability (Baird et al., 2006) and epilepsy (Amiet et al., 2008). CA and AA have traditionally been seen as disorders primarily involving atypical development of the CNS, while less attention has been paid to other organ systems, for example, the gastrointestinal (GI) tract (Bauman, 2010), which is the topic of this study.

Over the last decade, there has been a continuing debate as to whether or not a subgroup of people with ASD may suffer from an excess of GI symptoms or diseases. A review by Buie et al. (2010a) reported that the prevalence of GI symptoms (constipation, diarrhoea, bloating, belching, abdominal pain, reflux, vomiting and flatulence) found in people with ASD range widely from 9% to 91% based on 11 studies with an average frequency of 44%. In addition, the authors concluded that there does not appear to be any digestive system diseases that are specific to ASD. In another comprehensive review of the literature relating to different aspects of GI factors in ASD, including symptoms, pathology, nutrition and treatment (Erickson et al., 2005: 725), it was concluded that ‘currently, there is no evidence that specific GI abnormalities exist or if successful treatments for these children are available. No support for the routine role of specialized GI testing in the asymptomatic autistic child exists in the literature’. We refer the interested reader to these comprehensive reviews for a more in-depth discussion of ASD and GI diseases.

Recently, this theme has been further expanded in a population-based study, where Ibrahim et al. (2009) reported that the overall incidence of GI symptoms did not differ between participants with ASD and matched control participants. In addition, Whitehouse et al. (2011) found no association between early GI problems in the general population and autistic-like traits. In contrast, Smith et al. (2009) reported increased rate of bowel symptoms in children with ASD. However, more interestingly, the authors also found that the increase in reported bowel symptoms between the ASD group and mainstream controls was not autism specific, as there was a similar level of symptoms in a comparison group of children with heterogeneous developmental and neurological disorders.

In this article, we present results from a retrospective study where we examined the prevalence and types of diseases of the GI tract in 89 individuals diagnosed as children with AA compared with 258 controls from the general population using discharge diagnoses from the nationwide Danish National Hospital Register (DNHR) covering an observation period of 32.9 years. To our knowledge, this is the first study to systematically examine diseases of the GI tract in a cohort of people diagnosed as children with AA and in an age- and sex-matched comparison group from the general population.

Methods

Case group

The sample was taken from the patients consecutively attending the Departments of Child Psychiatry in the University Hospitals of Copenhagen and Aarhus in the years 1960–1984. The clinics provided services to the entire population of Denmark. At that time, practically all children with supposed ASD were admitted as inpatients for several months, leaving us with detailed psychiatric records. Case records of all the patients who were given a diagnosis of ‘Childhood psychosis’ – the ICD-8 (WHO, 1971) term of ASD including ‘borderline autistic condition’ – were re-diagnosed by the first two authors in 1985 in accordance with ICD-9 criteria (WHO, 1978). The diagnostic revision included a review of the patient’s child psychiatric records. Children with incomplete records were not included in the research file. A more detailed description of the case group has been published elsewhere (Mouridsen et al., 1993). Based on this diagnostic revision, 89 patients (58 males and 31 females) with a diagnosis of ‘autistic-like condition’ were identified. This category is in accordance with the later ICD-10 category ‘AA’ and DSM-IV category PDD-NOS. These 89 children make up the case group in this study.

All but one psychiatric record had information about intelligence quotient (IQ). As IQ was estimated according to the functioning on different scales (Leiter, Binet/Simon, Wechsler Preschool and Primary Scale of Intelligence (WPPSI), Wechsler Intelligence Scale for Children (WISC) or Raven was used), IQ was recorded as a categorical variable. Details of the 89 cases with AA at assessment in the 1960–1984 period are shown in Table 1.

Table 1.

Characteristics of 89 children with atypical autism at admission in the 1960–1984 period

Age (years)
Mean (SD)	8.9 (4.0)
Median	8.5
Range	3–17
Gender
Male	58
Female	31
Intelligence
<50	20 (22%)
50–69	15 (17%)
>69	53 (60%)
Unknown	1 (1%)

SD: standard deviation.

Comparison group

A comparison group was drawn from the Central Person Register (CPR). This register was established in 1968 where all persons alive and living in Denmark were registered. Since then, all persons with permanent residence in Denmark have been registered in the CPR, including every live born baby and every new inhabitant (Pedersen et al., 2006). All persons in the register are assigned a unique identification number, called the CPR number, which is used whenever a person is in contact with the Danish health-care system. This number allows for the accurate linkage at the individual level of information between registers. Wherever possible, we matched each child with AA with three (in order to increase statistical power) control children by sex, time of birth (controls were born the same day or the day before or after the patient), place of birth (region) and social group. In three cases, it was not possible to select controls, as the AA cases in question were born abroad. These 258 individuals (168 males and 90 females) make up the comparison group for our study.

Hospital register

All participants were screened through the DNHR (Andersen et al., 1999), using the CPR number. The register holds information on all persons discharged from Danish medical hospitals since 1 January 1977; information on visits to outpatient clinics and emergency rooms has been included in the register since 1 January 1995. Both groups were followed from their birthday or 1 January 1977 (whichever came later) until 24 November 2009 (the closing date of the observation period). The average observation period was 32.9 years (standard deviation (SD) = 0.3; range = 30.9–32.9 years), and the mean age at the end of the observation period was 48.5 years (SD = 7.2; range = between 30.9 and 64.1 years of age). So, the mean age at the start of the register-based observation period for a disease located to the GI tract was 15.6 years. In the calculation of mean ages at follow-up, we assume that an individual has been alive throughout the entire observation period.

For individuals registered in the DNHR, we ascertained the place of admission, the date of admission, together with primary and secondary discharge diagnoses. All hospital and outpatient treatment in Denmark is tax funded and free of charges for patients. All patient contacts are registered in the DNHR. The register provides diagnoses according to ICD-8 (WHO, 1971) and from 1 January 1994 according to ICD-10 (WHO, 1992). In order to clarify data presentation, all diagnoses are listed as ICD-10 diagnoses.

The diseases enquired about for this study included all diseases relating to the GI tract and in the following referred to as GI diseases, that is, ICD-8 codes 530–577 or ICD-10 codes K20–K93.

Each participant was counted when given a diagnosis; however, a participant might have received different diagnoses at the same or different hospital contacts and might therefore be counted in more than one diagnostic group.

Statistical analysis

Data handling, statistical analysis and calculation of odds ratios (ORs) were performed using the STATISTIX software packages (Analytical Software, 2003). All comparisons between the AA case group and the comparison group were based on Fisher’s exact test. All statistical tests were two tailed and were regarded as significant at p < 0.05.

Ethics

The Danish Data Inspectorate approved the study protocol.

Results

Overall, 84 of 89 (94.4%) individuals in the AA group and 241 of 258 (93.4%) in the comparison were registered in DNHR at the end of the study period (p = 0.81; OR =1.2; 95% confidence interval (CI) = 0.4–3.3). That is, a similar proportion of members from the case and comparison group had been in contact with a medical hospital during the observation period.

Prevalence and types of diseases of the GI tract

Table 2 summarizes the prevalence and types of diseases of the GI tract in the case and comparison group diagnosed any time during the 32.9-year observation period. Each participant was counted when given a diagnosis. However, a participant might have received different diagnoses as mentioned above.

Table 2.

Prevalence and types of diseases of the GI tract in a clinical sample of 89 individuals diagnosed as children with AA and 258 control cases from the general population. Different diseases of the GI tract may co-occur in the same person

	AA (n = 89)	Controls (n = 258)	p
	n (%)	n (%)
One or more disease(s) of the GI tract (K20–K93)^a	22 (24.7)	47 (18.2)	0.22
Diseases of oesophagus, stomach and duodenum (K20–K31)	3 (3.4)	11 (4.3)	0.77
Diseases of appendix (K35–K38)	2 (2.3)	10 (3.9)	0.54
Hernia (K40–K46)	10 (11.2)	12 (4.7)	0.04
Non-infective enteritis and colitis (K50–K52)	0 (0.0)	5 (1.9)	0.33
Ulcerative colitis (K51.x)	0 (0.0)	3 (1.2)	0.57
Other non-infective gastroenteritis and colitis (K52.x)	0 (0.0)	2 (0.8)	0.62
Other diseases of intestines (K55–K63)	10 (11.2)	14 (5.4)	0.08
Diseases of peritoneum (K65–K67)	0 (0.0)	1 (0.8)	1.00
Diseases of liver (K70–K77)	0 (0.0)	4 (1.6)	0.36
Disease of gall bladder, biliary tract and pancreas (K80–K87)	2 (2.2)	6 (2.3)	1.00
Other diseases of the digestive system (K90–K93)	1 (1.1)	2 (0.8)	1.00

AA: atypical autism; GI: gastrointestinal.

Conditions shown are those where at least one case or control was represented.

Among the 89 patients with AA, a total of 22 (24.7%) were known in the DNHR with at least one diagnosis of any disease of the GI tract (K20–K93). In the comparison group, the respective figure was 47 of 258 (18.2%). This difference is not statistically significant (p = 0.22; OR = 1.5; 95% CI = 0.8–2.6).

Looking at specific diseases, we found that only hernia (K40–K46) was significantly associated with AA (10 of 89 (11.2%) vs 12 of 258 (4.7%); p = 0.04; OR = 2.6; 95% CI = 1.08–6.2). Other GI diseases occurred with relatively low frequency in both groups. There were no inflammatory GI diseases, such as Crohn’s disease or ulcerative colitis in the AA group.

Diseases of the liver were reported only in participants from the comparison group: 0.0% versus 1.6%. This difference is not statistically significant (p = 0.36).

Gender, IQ and rate of GI diseases

Among the 58 males with AA, 15 (25.9%) were known in the DNHR with a GI diagnosis. In the comparison group, the respective figure is 26 of 168 (15.5%). The difference is not statistically significant (p = 0.11; OR = 1.9; 95% CI = 0.93–3.91). Among the 31 females with AA, 7 (22.6%) against 21 of 90 (23.3%) in the comparison group were known in the DNHR with a GI diagnosis (p =1.00; OR = 0.96; 95% CI = 0.36–2.53). So gender was not a statistically significant risk factor with respect to being diagnosed with a GI disease. Stratification of patients according to the intellectual level, IQ < 70 (n = 35) and IQ > 69 (n = 53) (IQ in one case was unknown), shows no statistically significant difference between the two groups with respect to a GI diagnosis in the DNHR (9 of 35 (25.7%) against 12 of 53 (22.7%); p = 0.80; OR = 1.2; 95% CI = 0.44–3.20), although there was a tendency for those individuals with IQ < 70 to have more GI diagnoses than participants with an IQ > 69.

Discussion

During the last 10 years, several researchers have called attention to the possibility that diseases of the GI tract may relate to ASD, even though the majority of people with ASD do not have clinical evidence of significant GI diseases such as malnutrition, failure to thrive, short stature, anaemia or signs indicating a chronic inflammatory process (Smith et al., 2009). A recent review by Buie et al. (2010a) reported that the published frequencies of GI symptoms range from 9% to 91%, averaging 41%. However, interpretation of the findings published so far is hampered by limitations in study design, including small sample sizes, proband heterogeneity (in most studies, participants were not analysed with respect to ASD diagnostic subgroup), varied ages of studied subjects (many studies only included participants under 18 years of age), limited or no case–control matching and reliance on parent-reported GI disease history or symptoms in some studies. In addition, the publications we identified lacked uniform definitions for the GI symptoms and diseases under investigation. There is no generally accepted, validated questionnaire that is available to screen or diagnose for bowel diseases based on bowel symptomatology in children (Smith et al., 2009). Smith et al. (2009) based their study on symptoms present for the previous 2 weeks, whereas Ibrahim et al. (2009) based their study on all GI diagnoses or abnormal GI symptoms reported in the medical charts of autism incident cases and the matched control participants before 21 years of age. Such differences in study design make it difficult to make conclusions across published studies.

In this study, we based our findings on hospital records of participants with DNHR diagnoses systematically collected during a study period of average 32.9 years. GI diseases of concern were defined and classified using ICD-10 (WHO, 1992).

In our retrospective study of 89 cases with AA and 258 control persons from the general population, we found that more participants from the AA case group than the comparison group had a GI diagnosis in the DNHR (24.7% against 18.2%). However, the difference is not significant (p = 0.22). Our prevalence rate is higher than those of Taylor et al. (2002), who, without including a comparison group from the general population, reported that 32 of 195 children (16%) with AA had bowel problems.

Stratifying individuals with AA by gender and level of IQ may help identifying those most at risk of experiencing GI diseases. Stratifying in terms of gender and GI morbidity shows that without reaching statistical significance, the rate of GI diseases was particularly high in males from the AA case group; 25.9% relative to 15.5% in males from the comparison group. Risk of receiving a GI diagnosis was slightly elevated (OR = 1.2) among the members of the AA group with an intellectual disability (IQ < 70). The difference did not reach statistical significance.

We found no inflammatory GI diseases such as Crohn’s disease or ulcerative colitis in the AA case group. This is consistent with the result from other studies of people with ASD (Erickson et al., 2005). Hernia was significantly more common (p = 0.04) in the AA case group than in the comparison group. The mechanisms of this association remain unclear. This may be a coincidental finding that needs to be assessed in future studies. Noteworthy is that no patients with AA had received a diagnosis of a liver disease against four individuals (1.6%) in the comparison group. This non-significant difference may reflect that people with ASD have a low consumption of alcohol (Santosh and Mijovic, 2006). We ourselves found a similar result in an earlier published study on diseases of the GI tract in people with infantile autism (Mouridsen et al., 2010).

Methodological considerations

The strengths and limitations of this study deserve mention. Our study has several strengths, including its relatively large size and linkage to nationwide registers, ensuring a complete follow-up. Equally, we consider it a strength that recall bias is omitted because data are registered as a part of clinical routine and independent of researchers and specific studies. As the incidence of many of the studied diseases cumulates throughout life, our long observation period is another important strength. On average, 32.9 years, which is longer than in any other published study of cases with ASD as far as we know.

Limitations of the study are related to the fact that the original case notes were re-diagnosed by clinicians rather than by structured research assessment, and that only the most severe cases of GI diseases are included, since we investigated only those individuals who were referred to a medical hospital for assessment and treatment. Overall, receiving inpatient or outpatient treatment in a medical hospital is a valid indicator of significant pathology that required care. In general, studies relying on medical records to identify GI diseases report lower frequencies of GI problems than studies relying on parental report or clinical examinations of referred samples (Maenner et al. 2012). Finally, it is important to note that our results mainly apply to GI diseases in young adults and adults, as the participants were on average 15.6 years old when they entered the observation period for GI diseases. Equally, that our study design did not allow us to determine at what age the participants were first given a diagnosis of the GI tract.

The reliability of the medical data is unknown. However, a validation study dealing with a number of diagnoses in the DNHR found validity ranging from 75% to 90% (Mosbech et al., 1995). For ulcerative colitis and type 1 diabetes, the figures were 90% and 96%, respectively (Nickelsen, 2002). Finally, it is important to emphasize that all follow-up information about the participants is register data. We have no clinical data, apart from the discharge diagnoses, and no information about relevant exposures, for example, smoking, alcohol consumption, dietary intake habits or any information about medical treatment, for example, antipsychotic medication. However, it is interesting to note that medications or dietary habits, commonly thought to contribute to GI in ASD (Ibrahim et al., 2009), were not associated with GI dysfunction in a recently published study (Gorrindo et al., 2012).

At the present time, evidence-based guidelines for the evaluation of GI symptoms are not yet available for individuals with ASD. However, recently, eight experienced gastroenterologists reviewed the current literature and published guidelines regarding GI symptoms that frequently occur in the general paediatric population. Based on this information and their collective expertise, they adopted guidelines for the current best practice related to the diagnosis and treatment for children with ASD (Buie et al., 2010b). The authors emphasized that individuals with ASD deserve the same thoroughness and standard of care in diagnostic workup and treatment of GI symptoms as should occur for other children.

To conclude, studies of the prevalence and types of GI diseases in people with ASD so far have shown inconsistent results. Our findings, contrary to most other studies, that deal with the AA subcategory suggest no association between AA and diseases of the GI tract.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association.

Amiet

Gourfinkel-An

Bouzamondo

. (2008) Epilepsy in autism is associated with intellectual disability and gender: evidence from a meta-analysis. Biological Psychiatry 64: 577–582.

Analytical Software (2003). STATISTIX 8 User’s Manual. Tallahassee, FL: Analytical Software.

Andersen

Madsen

Jørgensen

. (1999) The Danish National Hospital Register. A valuable source of data for modern health sciences. Danish Medical Bulletin 46: 263–268.

Baird

Simonoff

Pickles

. (2006) Prevalence of disorders of the autism spectrum in a population cohort of children in South Thames: the Special Needs and Autism Project (SNAP). Lancet 368: 210–215.

Bauman

(2010) Medical comorbidities in autism: challenges to diagnosis and treatment. Neurotherapeutics 7: 320–327.

Buie

Campbell

Fuchs

III . (2010a) Evaluation, diagnosis, and treatment of gastrointestinal disorders in individuals with ASDs: a consensus report. Pediatrics 125(Suppl. 1): S1–S18.

Buie

Fuchs

III Furuta

. (2010b) Recommendations for evaluation and treatment of common gastrointestinal problems in children with ASDs. Pediatrics 125(Suppl. l): S19–S29.

Erickson

Stigler

Corkins

. (2005) Gastrointestinal factors in autistic disorder: a critical review. Journal of Autism and Developmental Disorders 35: 713–727.

10.

Fombonne

Zakarian

Bennett

. (2006) Pervasive developmental disorders in Montreal, Quebec, Canada: prevalence and links with immunizations. Pediatrics 118: 139–150.

11.

Gorrindo

Williams

Lee

. (2012) Gastrointestinal dysfunction in autism: parental report, clinical evaluation, and associated factors. Autism Research 5: 101–108.

12.

Ibrahim

Voigt

Katusic

. (2009) Incidence of gastrointestinal symptoms in children with autism: a population-based study. Pediatrics 124: 680–686.

13.

Maenner

Arneson

Levy

. (2012) Association between behavioral features and gastrointestinal problems among children with autism spectrum disorder. Journal of Autism and Developmental Disorders 42: 1520–1525.

14.

Mosbech

Jørgensen

Madsen

. (1995) The national patient register: evaluation of data quality. Ugeskrift for Laeger 157: 3741–3745.

15.

Mouridsen

Rich

Isager

(1993) Parental age in infantile autism, autistic-like conditions, and borderline childhood psychosis. Journal of Autism and Developmental Disorders 23: 387–396.

16.

Mouridsen

Rich

Isager

(2010) A longitudinal study of gastrointestinal diseases in individuals diagnosed with infantile autism as children. Child: Care, Health and Development 36: 437–443.

17.

Nickelsen

(2002) Data validity and coverage in the Danish National Health Register. Ugeskrift for Laeger 164: 33–37.

18.

Pedersen

Gøtzsche

Møller

JØ

. (2006) The Danish Civil Registration System. Danish Medical Bulletin 53: 441–449.

19.

Rutter

(2005) Incidence of autism spectrum disorders: changes over time and their meaning. Acta Paediatrica 94: 2–15.

20.

Santosh

Mijovic

(2006) Does pervasive development disorder protect children and adolescents against drug and alcohol use? European Child & Adolescent Psychiatry 15: 183–188.

21.

Smith

Farnworth

Wright

. (2009) Are there more bowel symptoms in children with autism compared to normal children with other developmental and neurological disorders? A case control study. Autism 13: 343–355.

22.

Taylor

Miller

Lingam

. (2002) Measles, mumps, rubella vaccination and bowel problems or developmental regression in children with autism: population study. BMJ 324: 393–396.

23.

Whitehouse

AJO

Maybery

Wray

. (2011) No association between early gastrointestinal problems and autistic-like traits in the general population. Developmental Medicine and Child Neurology 53: 457–462.

24.

World Health Organization (WHO) (1971). International Classification of Diseases. Manual of the International Statistical Classification of Diseases, Injuries and Causes of Death (ICD-8) (Danish version). Copenhagen: Sundhedsstyrelsen.

25.

World Health Organization (WHO) (1978). International Classification of Diseases. Mental Disorders: Glossary and Guide to Their Classification (9th Revision). Geneva: World Health Organization.

26.

World Health Organization (WHO) (1992). The ICD-10 Classification of Mental and Behavioural Disorders. Clinical Descriptions and Diagnostic Guidelines. Geneva: World Health Organization.