Sex Differences in Health Outcomes Among Adults with ASD

Panel Presentation
Friday, May 3, 2019: 10:55 AM
Room: 517C (Palais des congres de Montreal)
L. E. Smith DaWalt1, J. L. Taylor2, A. Movaghar3, J. Hong3, M. H. Brilliant4 and M. Mailick1, (1)University of Wisconsin-Madison Waisman Center, Madison, WI, (2)Vanderbilt Kennedy Center, Nashville, TN, (3)University of Wisconsin-Madison, Madison, WI, (4)Marshfield Clinic Research Institute, Marshfield, WI

Individuals with autism spectrum disorder (ASD) face many challenges during adulthood, including elevated health problems relative to the general population. An emerging body of evidence suggests that females with ASD may be doubly-vulnerable for poor outcomes across a range of domains because of both their sex and their autism status. However, sex differences are rarely investigated in ASD studies, often because too few females are included. The present study used electronic health records (EHRs) to investigate health conditions of women with ASD compared to men with ASD as well as compared to men and women without ASD.


The present study examined (1) sex differences in health conditions and burden of disease for 2119 adults with ASD (458 females, 1661 males) and (2) if sex differences for adults with ASD were similar to or different from sex differences in the general population using a comparison group of 21,228 age- and sex-matched adults without disabilities.


We utilized EHRs from the Marshfield Clinic population (a large, private, multispecialty group practice with records for over 1 million patients) to characterize health profiles. We selected for autism cases based on a patient having an ICD-9 code of pervasive developmental disorder (299) on at least two occasions (following procedures used by Croen et al., 2015). Based on past research, our initial analyses focused on five conditions: cardiovascular disease, sleep disorders, thyroid disease, constipation, and gastroesophageal reflux. Health conditions (presence=1, absence=0) were defined based on the presence of at least two ICD-9 codes in the health record for the particular condition. For individuals with a given condition, burden of disease was determined by a count of total number of ICD-9 codes, reflecting the number of medical visits, for that condition.


Preliminary findings found differences between autism and comparison groups in all five of the conditions we investigated. Individuals with ASD had higher rates than controls of cardiovascular disease (23.5% ASD vs 10.2% controls; p=.03), sleep disorders (13.8% ASD vs 3.4% control, p=.05), constipation (5.3% ASD vs 1.0% controls, p=.05), and gastroesophageal reflux (10.3% ASD vs 3.0% control, p=.04). There was a significant group (ASD vs control) by sex interaction for thyroid disease: women with ASD (14.6%) had a higher rate of thyroid disease than men with ASD (5.4%), men without ASD (1%), and women without ASD (4.7%; p=.00). When considering the burden of disease, women with ASD had a greater number of medical visits relative to the other groups for sleep disorder (p=.002), gastroesophageal reflux (p=.047), and constipation (p=.058).


The present analyses suggest new ways in which women with ASD might be at greater risk than men (relative to their counterparts in the general population). Though both men and women with ASD were at greater risk for a number of conditions (consistent with extant research), our analyses suggest that women with ASD have more medical encounters once a condition is detected. Further analyses will examine other health conditions, as well as whether sex differences are more apparent at different ages.