Need for Valid, Reliable Gastrointestinal Symptoms Measurement Tool for Autism Spectrum Disorder, a Review of the Literature

Thursday, May 11, 2017: 5:30 PM-7:00 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
C. Holingue1, C. A. Newill2, L. C. Lee2,3, P. Pasricha4 and M. D. Fallin2,3, (1)Wendy Klag Center for Autism and Developmental Disabilities, Department of Mental Health, Johns Hopkins University, Baltimore, MD, (2)Wendy Klag Center for Autism and Developmental Disabilities, Department of Mental Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, (3)Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, (4)Center for Neurogastroenterology, Department of Gastroenterology and Hepatology, Johns Hopkins School of Medicine, Baltimore, MD

Gastrointestinal symptoms (GI) have been implicated in Autism Spectrum Disorder (ASD), and the purported association between GI symptoms and ASD has fueled the development of dietary interventions that claim to repair a damaged gut, and to improve GI and core ASD symptoms. However, there is no validated tool for measuring GI symptoms in this population and therefore no valid estimate of GI symptoms.


The purposes of this review were to 1) describe approaches used to measure GI symptoms, 2) summarize GI prevalence estimates across studies, and 3) discuss how the measurement variation affects GI symptom estimates.


PubMed was used to find ASD studies published between 1980-2016 that measured GI symptoms/diagnoses. For inclusion, studies were required to have 10 or more individuals with ASD; review articles, meta-analyses, hypotheses papers, narratives or editorials, and animal studies were excluded. From these studies, we extracted study characteristics as well as information on how GI symptoms/diagnoses were measured. To ascertain a range of estimates of GI symptom prevalence, we further refined this set of studies by excluding those in which sampling was based on GI symptoms/diagnoses, that included or excluded participants based on diet, or that were experimental.


Of the 357 studies returned from PubMed, 132 met our inclusion criteria for description of GI measurement. The publication dates ranged from 1986 to 2016. Fifty-five studies were clinic-based, 18 were population-based, 2 were enriched-risk, and 57 studies did not have enough information to assess the type of sample. There were 55 case-control studies, 32 cross-sectional studies, 15 cross-sectional studies with a comparison group (such as history of regression), 24 experimental studies, 1 study with both a case-control and experimental sample, and 5 cohort studies. Studies used the following methods to assess GI symptoms/diagnoses: parents/caregivers questionnaires (n=73), medical records (n=15), parents/caregivers as well as medical records, physician, teacher, or self-report (n=27). Seventeen studies did not specify who reported GI symptoms/diagnoses. Of the 110 studies that used a questionnaire for assessment, 61 used their own tool, while 49 studies used or modified an existing tool. Popular questionnaire options included the Rome III criteria (n=11), the GI severity index (n=5), and the Autism Treatment Network GI Symptom Inventory (n=3). Of the 78 studies that were kept for the prevalence estimates, the range of estimates for variables including an aggregate of GI symptoms or counting any GI symptom was 4.2-96.8% (median 47.1%). The range for estimates for diarrhea specifically was 2.3-75.6% (median 14.6%), and constipation 4.3-45.5% (median 21.4%). GI symptom estimates were significantly associated with the type of study sample and the method for reporting symptoms/diagnoses (p<0.05).


This study reviewed the breadth of approaches to measuring GI symptoms in studies of ASD. The type of study sample and approaches used to measure GI symptoms were significantly associated with the estimated GI symptom prevalence. The range of GI symptom estimates was wide, highlighting the need for a reliable, valid tool to be used consistently across studies measuring GI symptoms in ASD studies.