Abstract
Somatic symptoms are the most common cause of outpatient medical visits in the general population, yet their presence and severity in individuals on the autism spectrum has rarely been studied. We sought to assess the prevalence, impact, and clinical correlates of fourteen commonly-reported somatic symptoms in a sample of 290 transition-aged autistic young adults (mean [SD] age: 23.10 [2.38] years, range 18–26; 76.7% diagnosed with autism before age 18) recruited from the Simons Foundation SPARK participant pool. A modified version of the Patient Health Questionnaire–15 was used to assess somatic symptom prevalence/impact, along with measures of depression, anxiety, autistic traits, and quality of life. Somatic symptom burden was much higher in autistic young adults than previously reported in the general population. The most commonly reported current symptoms were fatigue (72.8%), sleep problems (69.0%), and menstrual problems (61.4% of females). Moderate or severe symptom levels were reported by 53.9% of females and 18.75% of males in our cohort, with the odds of females endorsing any given symptom being 2–4 times greater than males. Both individual symptoms and total symptom burden were related to higher levels of depression, anxiety, and autistic traits, along with lower quality of life. Despite little research on this topic previously, somatic symptoms are highly prevalent in autistic young adults, particularly women. Future research is needed to investigate links between somatic symptoms, medical and psychiatric morbidity, and health care utilization in the autistic population.
Keywords: Autism, Somatic Symptoms, Somatization, Physical Symptoms, Bodily Distress Syndrome, Sex Differences, Prevalence
Lay Abstract
Somatic symptoms (i.e., physical symptoms such as such as pain, weakness, stomachache, or shortness of breath) are highly prevalent in the general population and account for a large proportion of health care costs. However, few studies have investigated how often these symptoms are reported by autistic adults or their associations with other clinical and demographic variables. Based on self-report data from 290 young autistic adults, we found very high rates of bothersome somatic symptoms in this population, with females endorsing all symptoms at substantially higher rates than males. Somatic symptoms were also associated with worse mental health and quality of life, suggesting that they represent an overlooked contributor to poor health outcomes in the autistic adult population.
Introduction
As the diagnostic criteria for autism have grown broader, diagnoses of autism spectrum disorder (hereafter “autism”) have increased dramatically over the last three decades. Each year in the United States, between 70,700 and 111,600 youth on the autism spectrum turn 18 years old, and approximately 1 million autistic individuals are estimated to reach adulthood over the next decade (Shattuck et al., 2020). Notably, transition-aged young adults on the autism spectrum generally utilize health care at higher rates relative to the general population and individuals with other neurodevelopmental disorders, with particularly high rates of primary care utilization (Ames et al., 2021). In the general population, physical complaints such as pain, weakness, stomachache, or shortness of breath (hereafter “somatic symptoms”) represent the most common cause of outpatient medical visits, as well as the predominant reason why individuals with psychiatric conditions present to primary care settings (Kroenke, 2003). However, with the notable exceptions of gastrointestinal symptoms (Bjørklund et al., 2020; Holingue et al., 2018; Madra et al., 2020) and sleep disturbances (Díaz-Román et al., 2018; Morgan et al., 2020), somatic symptoms are infrequently studied in the autistic population (Grant et al., 2021; Lever & Geurts, 2016; Mayes et al., 2020; Mazurek et al., 2019). Furthermore, no studies to date have investigated the prevalence of individual somatic symptoms or the natural history of these symptoms in autistic adults.
In one of the first studies exploring this topic, Lever & Geurts (2016) compared scores on the Symptom Checklist-90 Revised (SCL-90-R) between autistic adults and comparison adults from a community sample, finding that autistic adults reported much higher scores on the somatization subscale of the SCL-90-R than community controls. However, when SCL-90 scores were compared to a normative sample of adult psychiatric patients, the degree of somatic symptom burden endorsed by autistic adults was not unusual compared to individuals with other psychiatric conditions (Lever & Geurts, 2016). Thus, while it appears that autistic people are at elevated likelihood of experiencing distressing somatic symptoms, this relationship may be driven primarily by the elevated rates of psychopathology seen in autistic adults. More recent studies have also found that autistic adults report high rates of central sensitivity syndrome (CSS; Yunus, 2015) diagnoses such as fibromyalgia and irritable bowel syndrome, and CSS diagnoses in this population are associated with poorer self-reported physical and mental health, as well as lower levels of subjective well-being (Grant et al., 2021). Notably, in this sample, CSS symptoms were significantly predicted by more self-reported sensory sensitivity, higher levels of anxiety, older age, and female sex; while higher autistic trait levels originally predicted more CSS symptoms, this relationship was no longer significant after accounting for anxiety and sensory sensitivity (Grant et al., 2021). Autistic children have also been found to exhibit elevated somatic symptoms compared to same-aged general population controls, although the sex differences in symptom burden seen in older samples were largely absent in children (Mayes et al., 2020). In younger autistic children, the severity of somatic symptoms has been shown to correlate with other types of psychopathology, as well as core autism features such as repetitive behaviors and sensory hyper-reactivity (Mazurek et al., 2019). However, with the notable exceptions of sleep disturbance and GI symptoms (Leader, Barrett, et al., 2021; Leader, Dooley, et al., 2021; Mannion & Leader, 2016; Mazurek et al., 2013, 2014; Yang et al., 2018), relationships between individual somatic symptoms and other clinical outcomes have not been systematically investigated in autistic people of any age group.
In the current study, we used data from the Simons Powering Autism Research Knowledge (SPARK; Feliciano et al., 2018) participant pool to assess self-reported current and lifetime somatic symptoms in a relatively large sample of transition-aged young adults on the autism spectrum. On the basis of previous studies of somatic symptoms in autistic individuals (Grant et al., 2021; Lever & Geurts, 2016; Mayes et al., 2020; Mazurek et al., 2019) and the general population (Hinz et al., 2017; Kocalevent et al., 2013), we hypothesized that young adults on the autism spectrum would report elevated levels of all somatic symptoms examined (Lever & Geurts, 2016; Mayes et al., 2020) compared to normative values. We also predicted that higher somatic symptom burden in our sample would be associated with older age (Grant et al., 2021; Hinz et al., 2017; Kocalevent et al., 2013), female sex (Grant et al., 2021; Hinz et al., 2017; Kocalevent et al., 2013), higher levels of autistic traits (Grant et al., 2021), elevated depression and anxiety symptoms (Grant et al., 2021; Hinz et al., 2017; Kocalevent et al., 2013; Mazurek et al., 2019), and reduced quality of life (Grant et al., 2021; Hinz et al., 2017; Kocalevent et al., 2013; Williams, Cascio, et al., 2021). In addition, we attempted to replicate results of a previous study wherein the relationship between autistic traits and somatic symptom burden was no longer significant after controlling for anxiety (Grant et al., 2021). This finding is particularly important, as it allows us to investigate whether autism itself can be thought of as an independent risk factor for high somatic symptom burden once co-occurring mood and anxiety disorders are accounted for. By describing the presence, severity, and clinical correlates of somatic symptoms in autistic individuals, the current study sought to further characterize these potentially major contributors to health care utilization and quality of life in this population.
Methods
Participants
Young adults from the United States with self-reported professional diagnoses of autism were recruited from the Simons Foundation-sponsored SPARK participant pool (Feliciano et al., 2018) as part of a larger study on repetitive thinking and its clinical correlates in autistic adults (project number RM0030Gotham; see Williams, Everaert, et al., 2021; Williams, McKenney, et al., 2021; Williams & Gotham, 2021a, 2021b for additional details). This larger study was advertised to all independent autistic adults between the ages of 18 and 45 in the SPARK participant pool (n=2887), 1014 of whom consented to complete the larger survey (35.1% response rate). As a part of this study, participants completed a series of self-report questionnaires via the SPARK platform on demographics, current and lifetime psychiatric diagnoses, autism severity, quality of life, co-occurring psychiatric symptoms, and a number of other clinical variables. Data for the current study were derived from the subset of individuals between 18 and 26 years of age who completed the portion of the study survey that included a measure of self-reported somatic symptom burden (n=290). This age range was chosen in order to decrease the likelihood of age-related medical comorbidities, thereby increasing our confidence that reported somatic symptoms were largely functional in nature (i.e., not associated with general medical conditions). All participants gave informed consent, and all study procedures were approved by the institutional review board at Vanderbilt University Medical Center.
Measures
Somatic symptom burden was measured with a modified version of the Patient Health Questionnaire–15 (Kroenke et al., 2010), a well-validated somatization measure that is widely used in both medical and psychiatric settings. In the original PHQ-15, respondents rate the impact of 15 common symptoms (e.g., joint pain, dizziness, fatigue) over the past four weeks on a three-point Likert scale from 0 (Not bothered at all) to 2 (Bothered a lot). The current study used a modified version of the PHQ-15 (M-PHQ) that queried symptom burden at multiple time points, beginning with lifetime symptom presence (e.g., Have you ever had problems with [symptom]?), then asking participants to specify the developmental periods during which this symptom was problematic (i.e., childhood [0–12 years], adolescence [13–17 years], or adulthood [18+]), and finally to report the symptom’s impact over the previous three months (rated 0–2). This modified form also omitted the item assessing pain or problems during sexual intercourse, as part of a broader effort to reduce participant burden from our lengthy battery. As a measure of total somatic symptom burden across the remaining symptoms, we took the mean score of all non-missing items (i.e., up to 14 items for females and 13 for males), multiplied that score by 15, and rounded to the nearest integer value in order to approximate the 0–30 distribution of the original PHQ-15 questionnaire. Total scores on this measure can be qualitatively interpreted as minimal (0–4), mild (5–9), moderate (10–14), and severe (≥15) (Kroenke et al., 2010).
Additional measures from the study survey battery were used in the current study to measure specific correlates of somatic symptom burden. Autistic traits were quantified using the self-reported Social Responsiveness Scale–Second Edition (SRS-2; Constantino & Gruber, 2012) total T-score. Current symptoms of depression and generalized anxiety were measured using the Beck Depression Inventory–II (BDI-II) autism-specific latent trait score (Williams, Everaert, et al., 2021) and the Generalized Anxiety Disorder–7 (GAD-7; Kroenke et al., 2010) total score, respectively. Quality of life (QoL) was measured using the four-item World Health Organization Quality of Life scale previously validated in this cohort of autistic adults (WHOQOL-4; Williams & Gotham, 2021a). For all scales except the BDI-II (which used a full-information score estimator based on complete data), composite scores were calculated as the mean of all non-missing items on each form, multiplied by the number of items on the form and rounded to the nearest integer value. The proportion of item-level missing data was low (≤2%) for all included measures, and thus the degree of added bias due to single imputation was deemed to be minimal.
Statistical Analyses
Descriptive statistics were calculated for our sample, including the lifetime prevalence of each symptom, the prevalence of these symptoms during each developmental period, and the response frequencies for items assessing symptom impact. Corrected item-total polyserial correlations (Olsson et al., 1982) were also calculated for each symptom to assess its relationship with overall symptom burden. Item-level descriptive statistics were calculated on complete rather than imputed data. Current (3-month) and lifetime endorsement frequencies for each symptom were compared between male and female participants using odds ratios (ORs), and M-PHQ total scores were compared across sexes using Cohen’s d. Associations between M-PHQ scores and other outcomes were assessed using Pearson correlations, and polyserial correlations were used to assess relationships between individual symptoms and these same outcomes. In order to determine the most important predictors of somatic symptom burden in this population, we regressed the M-PHQ onto age, sex, education, SRS-2 score, BDI-II score, and GAD-7 score, calculating partial R2 values to rank variable importance. A linear regression was used given the continuous nature of M-PHQ scores. Lastly, as an exploratory analysis, we sought to investigate the degree to which self-reported somatic symptoms independently predict lower subjective quality of life (i.e., WHOQOL-4 scores) after accounting for levels of depression, anxiety, and autistic traits. To answer this question, we conducted another linear regression, in which WHOQOL-4 scores were regressed on age, sex, education, SRS-2 score, BDI-II score, GAD-7 score, and M-PHQ score. This model was compared to a reduced model that did not include M-PHQ score as a predictor, and models were compared using Bayes factors derived from the Bayesian Information Criterion (Kass & Raftery, 1995). Missing data in all regression models were handled using 40-fold multiple imputation based on predictive mean matching. All statistical analyses were performed in R version 4.1 (R Core Team, 2021).
Results
Data were available for 290 young adults (61% female sex), with a mean (SD) age of 23.10 (2.38) years. Participants in the current sample were diagnosed with autism at a mean age of 12.16 (SD 6.59) years, with 76.7% of the sample being diagnosed with autism before age 18. Demographics and clinical characteristics of the participants, overall and divided according to biological sex, are presented in Table 1. The mean (SD) M-PHQ total score was 8.34 (6.02), with females reporting significantly higher symptom burden than males (d=0.906, CI95% [0.657, 1.150]). Based on original PHQ-15 cutoffs, levels of somatic symptom burden were “moderate” and “severe” in 23.4% and 16.9% of the total sample, respectively, with females reporting clinically significant (i.e., moderate or severe) symptom distress at much higher rates than males (53.9% vs. 18.75%).
Table 1.
Sociodemographic and Clinical Characteristics of the Sample
| Males (n = 112) |
Females (n = 178) |
Total (N = 290) |
d [95% CI] (F > M) |
|
|---|---|---|---|---|
|
| ||||
| Age (Years) | 22.65 (2.55) | 23.39 (2.23) | 23.10 (2.38) | 0.313 [0.074, 0.552] |
| Gender Identity | ||||
| Cisgender | 103 (92.0%) | 146 (82.0%) | 249 (85.9%) | |
| Transgender | 0 (0%) | 8 (4.5%) | 8 (2.8%) | |
| Non-binary | 9 (8.0%) | 24 (13.5%) | 33 (11.4%) | |
| Race/Ethnicity | ||||
| Asian | 5 (4.5%) | 9 (5.1%) | 14 (4.8%) | |
| Black/African American | 6 (5.4%) | 9 (5.1%) | 15 (5.2%) | |
| Native American/Alaska Native | 1 (0.9%) | 12 (6.7%) | 13 (4.5%) | |
| White | 82 (73.2%) | 145 (81.5%) | 227 (78.3%) | |
| Other Race | 7 (6.2%) | 4 (2.2%) | 11 (3.8%) | |
| Hispanic/Latino | 11 (9.8%) | 20 (11.2%) | 31 (10.7%) | |
| Education | ||||
| No High School Diploma | 5 (4.5%) | 7 (3.9%) | 12 (4.1%) | |
| High School Diploma/GED | 43 (38.4%) | 39 (21.9%) | 82 (28.3%) | |
| Vocational Certificate | 6 (5.4%) | 10 (5.6%) | 16 (5.5%) | |
| Some College | 28 (25.0%) | 61 (34.3%) | 89 (30.7%) | |
| Associate Degree | 7 (6.2%) | 15 (8.4%) | 22 (7.6%) | |
| Bachelor’s Degree | 18 (16.1%) | 35 (19.7%) | 53 (18.3%) | |
| Graduate/Professional Degree | 5 (4.5%) | 11 (6.2%) | 16 (5.5%) | |
| Age of Autism Diagnosis (Years) | 9.69 (6.32) | 13.71 (6.28) | 12.16 (6.58) | 0.637 [0.391, 0.883] |
| Ever had IEP for Autism | 79 (70.5%) | 86 (48.3%) | 165 (56.8%) | |
| SRS-2 Total T-score | 66.68 (11.14) | 70.67 (10.70) | 69.14 (11.03) | 0.367 [0.125, 0.609] |
| BDI-II Autism-specific T-score | 47.32 (9.76) | 51.13 (9.19) | 49.66 (9.58) | 0.405 [0.164, 0.646] |
| GAD-7 Total Score | 6.51 (6.07) | 8.98 (5.79) | 8.02 (6.01) | 0.419 [0.177, 0.660] |
| WHOQOL-4 Total Score (1–5) | 3.43 (0.99) | 3.22 (0.85) | 3.30 (0.91) | −0.234 [−0.474, 0.005] |
| M-PHQ Total Score† | 5.28 (4.89) | 10.27 (5.87) | 8.34 (6.02) | 0.905 [0.657, 1.154] |
| Somatic Symptom Severity | ||||
| Minimal (0–4) | 57 (50.9%) | 35 (19.7%) | 92 (31.7%) | |
| Mild (5–9) | 34 (30.4%) | 47 (26.4%) | 81 (27.9%) | |
| Moderate (10–14) | 16 (14.3%) | 52 (29.2%) | 68 (23.4%) | |
| Severe (15+) | 5 (4.5%) | 44 (24.7%) | 49 (16.9%) | |
Note. “Males” and “Females” refer to sex assigned at birth rather than gender. Values are presented as M (SD) for continuous variables and N (%) for categorical variables. d = Cohen’s d (standardized mean difference); IEP = Individualized Education Program; SRS-2 = Social Responsiveness Scale–Second Edition; BDI-II = Beck Depression Inventory–II; GAD-7 = Generalized Anxiety Disorder–7; WHOQOL-4 = Four-item World Health Organization Quality of Life Scale; M-PHQ = Modified Patient Health Questionnaire–15.
Calculated as the 15 times the mean of all 14 items (13 for males), rounded to the nearest integer value.
Endorsement frequencies for each individual symptom are displayed in Table 2. The symptoms most highly endorsed in the past three months were fatigue (72.8%), sleep problems (69.0%), and menstrual problems (61.4% of females). The majority of participants reported that their symptoms were present since adolescence or adulthood, although abdominal pain, sleep problems, dyspnea, and constipation/diarrhea had an onset in childhood in over 40% of individuals reporting those symptoms. All symptoms increased in prevalence with age, and symptoms persisted through multiple life stages in the majority of individuals. Item-total correlations were large for most symptoms, with dyspnea, chest pain, and abdominal pain showing the strongest relationships with overall symptom burden. Both current and lifetime symptom reports were higher in females for all symptoms, with confidence intervals excluding OR=1 for all but current syncope and lifetime dyspnea (Figure 1).
Table 2.
Somatic Symptom Endorsement and Impact in 290 Transition-aged Autistic Young Adults
| Symptom | Symptom Endorsement (%) |
Three-month Symptom Impact (%) |
r it | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Current | Lifetime | Childhood (0–12 yrs) | Adolescence (13–17 yrs) | Adulthood (18+ yrs) | Not Bothered | Bothered A Little | Bothered A Lot | ||
|
| |||||||||
| Fatigue | 72.8 | 80.3 | 19.0 | 58.6 | 76.9 | 27.2 | 33.4 | 39.3 | 0.572 |
| Sleep problems | 69.0 | 80.0 | 37.9 | 67.6 | 75.5 | 31.0 | 28.3 | 40.7 | 0.428 |
| Menstrual problems† | 61.4 | 82.4 | 5.2 | 44.5 | 44.8 | 23.4 | 21.4 | 15.9 | 0.151 |
| Back pain | 47.6 | 54.5 | 6.6 | 32.4 | 51.7 | 52.4 | 29.3 | 18.3 | 0.544 |
| Constipation/diarrhea | 46.6 | 56.9 | 27.9 | 46.2 | 50.7 | 53.4 | 26.9 | 19.7 | 0.567 |
| Nausea/gas/indigestion | 44.5 | 55.9 | 22.1 | 40.0 | 53.1 | 55.5 | 29.7 | 14.8 | 0.637 |
| Joint pain | 43.7 | 53.5 | 12.8 | 38.6 | 46.9 | 55.5 | 24.1 | 19.0 | 0.615 |
| Abdominal pain | 43.4 | 53.4 | 28.3 | 44.8 | 45.9 | 56.6 | 29.3 | 14.1 | 0.680 |
| Headache | 40.0 | 52.8 | 16.2 | 40.3 | 47.9 | 60.0 | 24.8 | 15.2 | 0.355 |
| Dizziness | 29.7 | 39.0 | 9.0 | 25.5 | 36.6 | 70.3 | 23.4 | 6.2 | 0.633 |
| Palpitations | 25.2 | 36.2 | 9.3 | 24.5 | 34.5 | 74.8 | 17.9 | 7.2 | 0.606 |
| Dyspnea | 24.2 | 33.9 | 15.9 | 26.2 | 29.3 | 75.5 | 17.9 | 6.2 | 0.718 |
| Chest pain | 17.6 | 30.3 | 6.6 | 19.3 | 24.5 | 82.4 | 15.5 | 2.1 | 0.687 |
| Syncope | 5.9 | 15.2 | 5.5 | 9.0 | 11.0 | 93.8 | 3.8 | 2.1 | 0.629 |
Note. Symptoms are displayed in order of current prevalence (past three months). Values indicate the percentage of the full sample endorsing each symptom or Likert scale response.during a given time period. Individuals were able to endorse a symptom at multiple time periods, and thus Childhood, Adolescence, and Adult symptom endorsement percentages do not add up to 1 or the lifetime endorsement percentage. rit = corrected item-total polyserial correlation for each symptom.
Based on results from female participants only.
Figure 1.
Sex differences in (A) current endorsement (“a little” or “a lot” bothered in past three months) and (B) lifetime endorsement (“ever” bothered by symptom) of fourteen somatic symptoms by autistic young adults. Odds ratios (Female/Male) are presented for each symptom along with 95% maximum likelihood confidence intervals.
Correlations between PHQ symptoms, total scores, and covariates of interest are displayed in Table 3. As expected, total symptom burden was strongly related to depression (r=0.522, CI95% [0.432, 0.601]) and anxiety (r=0.512, CI95% [0.421, 0.593]) symptoms, displaying slightly smaller correlations with autistic trait levels (r=0.432, CI95% [0.333, 0.522]) and QoL (r=−0.323, CI95% [−0.423, −0.216]). A small but significant correlation was also present between M-PHQ total score and age, with older individuals reporting slightly more symptoms (r=0.145, CI95% [0.030, 0.256]). Correlations at the individual symptom level were attenuated but still statistically significant for all symptoms except menstrual problems (Table 3).
Table 3.
Correlations of Individual Somatic Symptoms with Clinical Features
| Symptom | BDI-II (Depression) |
GAD-7 (Anxiety) |
SRS-2 (Autistic Traits) |
WHOQOL-4 (Quality of Life) |
Age |
|---|---|---|---|---|---|
|
| |||||
| Fatigue | 0.659 [0.589, 0.720] |
0.479 [0.385, 0.563] |
0.315 [0.208, 0.415] |
−0.412 [−0.503, −0.311] |
0.107 [−0.008, 0.220] |
| Sleep problems | 0.487 [0.394, 0.570] |
0.418 [0.318, 0.509] |
0.411 [0.311, 0.502] |
−0.352 [−0.449, −0.247] |
0.164 [0.050, 0.274] |
| Menstrual problems | 0.076 [−0.039, 0.190] |
0.165 [0.050, 0.275] |
0.092 [−0.023, 0.205] |
0.068 [−0.047, 0.182] |
0.040 [−0.076, 0.154] |
| Back pain | 0.454 [0.358, 0.541] |
0.390 [0.288, 0.483] |
0.299 [0.191, 0.401] |
−0.248 [−0.353, −0.136] |
0.116 [0.000, 0.228] |
| Constipation/diarrhea | 0.293 [0.184, 0.395] |
0.343 [0.237, 0.441] |
0.227 [0.114, 0.333] |
−0.123 [−0.235, −0.008] |
0.116 [0.000, 0.228] |
| Nausea/gas/indigestion | 0.237 [0.125, 0.343] |
0.291 [0.182, 0.393] |
0.232 [0.120, 0.338] |
−0.124 [−0.236, −0.009] |
0.201 [0.088, 0.309] |
| Joint pain | 0.328 [0.221, 0.427] |
0.383 [0.281, 0.478] |
0.350 [0.245, 0.447] |
−0.278 [−0.381, −0.168] |
0.026 [−0.090, 0.141] |
| Abdominal pain | 0.301 [0.192, 0.402] |
0.344 [0.238, 0.442] |
0.313 [0.205, 0.413] |
−0.149 [−0.260, −0.035] |
0.124 [0.009, 0.236] |
| Headache | 0.291 [0.182, 0.393] |
0.357 [0.252, 0.453] |
0.388 [0.286, 0.482] |
−0.180 [−0.289, −0.066] |
0.105 [−0.011, 0.217] |
| Dizziness | 0.440 [0.342, 0.528] |
0.370 [0.266, 0.465] |
0.382 [0.279, 0.476] |
−0.319 [−0.419, −0.212] |
0.068 [−0.047, 0.182] |
| Palpitations | 0.405 [0.304, 0.497] |
0.376 [0.273, 0.471] |
0.316 [0.208, 0.416] |
−0.275 [−0.378, −0.165] |
0.153 [0.038, 0.264] |
| Dyspnea | 0.265 [0.155, 0.369] |
0.257 [0.146, 0.361] |
0.246 [0.135, 0.351] |
−0.156 [−0.266, −0.042] |
−0.006 [−0.121, 0.109] |
| Chest pain | 0.233 [0.121, 0.339] |
0.238 [0.127, 0.344] |
0.194 [0.081, 0.303] |
−0.203 [−0.311, −0.090] |
−0.036 [−0.150, 0.080] |
| Syncope | 0.520 [0.431, 0.599] |
0.453 [0.356, 0.540] |
0.443 [0.346, 0.531] |
−0.356 [−0.453, −0.251] |
−0.041 [−0.155, 0.075] |
Note. Polyserial correlations are presented along with 95% confidence intervals. BDI-II = Beck Depression Inventory–II; GAD-7 = Generalized Anxiety Disorder–7; SRS-2 = Social Responsiveness Scale–Second Edition; WHOQOL-4 = Four-item World Health Organization Quality of Life Scale
Regression analyses indicated that the combination of age, sex, education, SRS-2 scores, BDI-II scores, and GAD-7 scores explained over 40% of the variance in overall somatic symptom burden (R2=0.413; Table 4). Female sex was the strongest predictor of somatic symptom burden, although SRS-2 scores (autistic traits), BDI-II scores (depression symptoms), and GAD-7 scores (generalized anxiety symptoms) each contributed significantly to the overall model. Neither age nor level of education significantly predicted overall somatic symptom burden after accounting for other predictors. In the predictive model for subjective quality of life, somatic symptom burden did not account for a significant amount of additional variance beyond that explained by demographic factors, autistic traits, anxiety, and depression (βStd=0.089, CI95% [−0.021, 0.198], R2p=0.005). When comparing the full QoL regression model to a reduced model without M-PHQ scores, a Bayesian analysis found moderate evidence against the inclusion of M-PHQ scores in the final model (ΔBIC=3.13, BF01=4.78), and of the variables examined, only BDI-II scores (βStd=−0.591, CI95% [−0.721, −0.462], R2p=0.148) and GAD-7 scores (βStd=−0.169, CI95% [−0.301, −0.037], R2p=0.012) significantly predicted lower overall QoL in the current autistic sample.
Table 4.
Regression Model Predicting Total Somatic Symptom Burden
| Predictor | βStd [95% CI] | p-value | R 2 p |
|---|---|---|---|
|
| |||
| Intercept | −0.359 [−0.510, −0.208] | < 0.001 | — |
| Age (Years) | 0.014 [−0.084, 0.113] | 0.774 | < 0.001 |
| Sex (Female > Male) | 0.600 [0.405, 0.795] | < 0.001 | 0.078 |
| Education | −0.025 [−0.122, 0.071] | 0.607 | 0.001 |
| Autistic Traits (SRS-2 Score) | 0.138 [0.023, 0.252] | 0.019 | 0.012 |
| Depression (BDI-II Score) | 0.226 0.089, 0.363 | 0.001 | 0.022 |
| Anxiety (GAD-7 Score) | 0.222 [0.083, 0.361] | 0.002 | 0.021 |
Note. All regression coefficients are based on a model wherein the dependent variable (M-PHQ scores) and all predictors other than sex were standardized to a mean of zero and standard deviation of one. Statistically significant predictors (p < 0.05) are displayed in bold. R2p = squared partial correlation (unique variance accounted for by variable).
Discussion
In this study, we examined the current and lifetime prevalence of fourteen common somatic symptoms in a cohort of transition-aged autistic young adults, finding that these individuals endorsed all somatic symptoms at much higher rates than young adults in previous general population studies using the PHQ-15 (Kocalevent et al., 2013). The majority of these somatic symptoms were reported to have an onset in adolescence or adulthood, although gastrointestinal complaints, sleep problems, and dyspnea appeared in childhood for many individuals reporting those symptoms. Covariation between these symptoms was very strong, with 11 of 14 symptoms exhibiting corrected item-total polyserial correlations greater than 0.5 and all but menstrual problems demonstrating item-total correlations of 0.3 or higher. Moreover, when considering total somatic symptom burden, 40.3% of autistic young adults (53.9% of females and 18.8% of males) reported total symptoms in the “moderate” or “severe” range, as compared to 9.3% (10.3% of females and 8.1% of males) in a German sample of young adults who completed the original PHQ-15 (Kocalevent et al., 2013).
Overall somatic symptom burden was strongly associated with female sex, more severe symptoms of anxiety and depression, higher levels of autistic traits, and lower overall QoL. These relations were individually replicated for each of the symptoms except menstrual problems, which appeared to be largely unrelated to the underlying multi-system symptom distress represented by the remainder of the M-PHQ items. Somatic symptom burden was slightly related to older age, but this relationship disappeared in regression analyses, indicating it was likely due to the slight age difference between males and females in our sample. Additionally, the positive relationship between autistic traits and somatic symptom burden remained after controlling for age, sex, depression, and generalized anxiety. Contrary to prior findings in autistic adults (Grant et al., 2021), the current study suggests that autism may indeed be an independent risk factor for elevated somatic symptom burden, even after co-occurring psychopathology is accounted for. However, there were several notable differences between our study and that of Grant et al., including a much narrower (and younger) age range in the current sample, inclusion of self-reported sensory sensitivities in only one of the studies, and different measures of autistic traits, somatic symptom burden, and anxiety being used in each investigation. Lastly, additional regression analyses failed to find evidence that somatic symptom burden uniquely predicts overall quality of life in autistic adults once co-occurring depression and anxiety have been accounted for. Additional studies are necessary to further investigate the clinical significance of somatic symptoms in autistic people, including their effects on health care utilization, relationships with core features of autism, and quantification of associated functional impairment.
As with prior general population (Hinz et al., 2017; Kocalevent et al., 2013; Lieb et al., 2000) and autistic adult (Grant et al., 2021) samples, we found large sex differences in somatic symptom reporting, with the odds of reporting a given symptom in the past three months being 2–4 times higher in female compared to male participants. These results are consistent with the higher rates of co-occurring physical health conditions seen in autistic females compared to males (Croen et al., 2015; Davignon et al., 2018; Kassee et al., 2020). Individuals identifying as transgender or non-binary made up a sizable portion of the sample, and although this group was too small to allow for systematic examination of symptom endorsement patterns by gender identity, it is notable that individuals identifying as a gender minority on average endorsed numerically more somatic symptoms than their cisgender counterparts. The sex difference in total symptom burden was also very large (d>0.9), with a magnitude greater than double that seen in the general population (Hinz et al., 2017). This sex difference also remained large in the multiple regression analysis, which controlled for observed sex differences in age, education, autistic traits, depression, and anxiety symptomatology. These findings stand in contrast to the lack of sex differences in somatic complaints reported in a primarily prepubescent sample of children on the autism spectrum (Mayes et al., 2020). As the majority of somatic symptoms in the current sample were reported to have arisen in adolescence or adulthood, it is therefore possible that sex differences in somatic symptom burden may only emerge after puberty, as has been observed in the general population (Campo et al., 1999). Longitudinal studies in autistic youth are necessary to accurately characterize the onset, persistence, and developmental trajectory of multiple somatic symptoms in this population, additionally determining whether sex differences in these symptoms are limited to certain developmental periods.
Limitations of our study include a female-predominant and therefore unrepresentative sample of autistic adults, a relatively low response rate to the study survey, no neurotypical control participants who completed the same M-PHQ, a lack of determination regarding which symptoms were attributable to diagnosed somatic conditions, and an inability to relate symptoms to health anxiety or health care utilization. In particular, given the large sex differences in somatic symptom burden, the 40% rate of clinically significant somatization in our sample is likely an overestimate of the true population prevalence, with sex-weighted prevalence estimates of 25.8% and 27.5% based on our data using 4:1 and 3:1 male to female ratios, respectively (Loomes et al., 2017). Notably, these prevalence values still estimate that rates of clinically significant somatization are nearly three times higher than those reported in a similarly-aged general population sample of young adults (Kocalevent et al., 2013). An additional limitation concerns the interpretation of scores from the M-PHQ, which included a longer recall period compared to the original PHQ-15 and may thus overestimate current symptom endorsement slightly. Moreover, as normative values for the M-PHQ were not investigated in a general population sample, it is possible that the prevalence of some somatic symptoms in this sample does not substantially exceed that reported by general population controls. Somatic symptom questionnaires such as the PHQ-15 also do not assess the etiology of the experienced symptoms, and thus we were unable to determine the degree to which these symptoms are medically explainable. As autistic individuals have increased risk of medical conditions such as inflammatory bowel disease (Bjørklund et al., 2020; Croen et al., 2015) and sleep-disordered breathing (Díaz-Román et al., 2018; Morgan et al., 2020), it is important that practitioners do not simply forego appropriate medical screenings because symptoms in this population are presumed to be functional in origin. Future work is therefore necessary to investigate the degree to which organic pathology contributes to the burden of somatic symptoms in autistic individuals. Lastly, the current study was limited to young adults on the autism spectrum; much less work has been conducted with middle-aged and older autistic adults (Lever & Geurts, 2016), and it is currently unclear whether the prevalence, burden, and correlates of somatic symptoms in the autistic population change over the adult lifespan.
Conclusions
In the first study to systematically examine individual and aggregate somatic symptoms in a large sample of transition-aged autistic young adults, we found that autistic individuals endorse somatic symptoms at much higher rates than individuals in the general population. Total symptom burden was significantly associated with female sex, higher autistic trait levels, increased symptoms of anxiety and depression, and reduced quality of life. However, after accounting for levels of anxiety, depression, and autistic traits, total somatic burden did not significantly predict subjective quality of life in the current sample. Although the effects of somatic symptoms on health care utilization and overall morbidity in the autistic population are still undetermined, our findings indicate that consideration of somatic symptoms is warranted when investigating the health status of individuals on the autism spectrum.
Acknowledgements
This study was supported by National Institute of Mental Health grant R01-MH113576 (KG), National Institute on Deafness and Other Communication Disorders grant F30-DC019510 (ZJW), National Institute of General Medical Sciences grant T32-GM007347 (ZJW), and the Nancy Lurie Marks Family Foundation (ZJW). The authors are grateful to all of the individuals and families enrolled in SPARK, the SPARK clinical sites and SPARK staff. They appreciate obtaining access to demographic and phenotypic data on SFARI Base. Approved researchers can obtain the SPARK population dataset described in this study (project number RM0030Gotham) by applying at https://base.sfari.org.
Footnotes
Conflict of Interest Statement
Zachary Williams has served as a consultant for Roche. He also serves as a family partner of the Autism Speaks Autism Care Network Vanderbilt site and member of the Autistic Researchers Review Board of the Autism Intervention Research Network on Physical Health (AIR-P). Katherine Gotham has no conflicts of interest to disclose.
References
- Ames JL, Massolo ML, Davignon MN, Qian Y, & Croen LA (2021). Healthcare service utilization and cost among transition-age youth with autism spectrum disorder and other special healthcare needs. Autism, 25(3), 705–718. 10.1177/1362361320931268 [DOI] [PubMed] [Google Scholar]
- Bjørklund G, Pivina L, Dadar M, Meguid NA, Semenova Y, Anwar M, & Chirumbolo S (2020). Gastrointestinal alterations in autism spectrum disorder: What do we know? Neuroscience & Biobehavioral Reviews, 118, 111–120. 10.1016/j.neubiorev.2020.06.033 [DOI] [PubMed] [Google Scholar]
- Campo JV, Jansen-McWilliams L, Comer DM, & Kelleher KJ (1999). Somatization in Pediatric Primary Care: Association With Psychopathology, Functional Impairment, and Use of Services. Journal of the American Academy of Child & Adolescent Psychiatry, 38(9), 1093–1101. 10.1097/00004583-199909000-00012 [DOI] [PubMed] [Google Scholar]
- Constantino JN, & Gruber CP (2012). Social Responsiveness Scale–Second Edition (SRS-2): Manual (2nd ed.). Western Psychological Services. [Google Scholar]
- Croen LA, Zerbo O, Qian Y, Massolo ML, Rich S, Sidney S, & Kripke C (2015). The health status of adults on the autism spectrum. Autism, 19(7), 814–823. 10.1177/1362361315577517 [DOI] [PubMed] [Google Scholar]
- Davignon MN, Qian Y, Massolo M, & Croen LA (2018). Psychiatric and Medical Conditions in Transition-Aged Individuals With ASD. Pediatrics, 141(Suppl 4), S335–S345. 10.1542/peds.2016-4300K [DOI] [PubMed] [Google Scholar]
- Díaz-Román A, Zhang J, Delorme R, Beggiato A, & Cortese S (2018). Sleep in youth with autism spectrum disorders: Systematic review and meta-analysis of subjective and objective studies. Evidence-Based Mental Health, 21(4), 146–154. 10.1136/ebmental-2018-300037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feliciano P, Daniels AM, Snyder LG, Beaumont A, Camba A, Esler A, Gulsrud AG, Mason A, Gutierrez A, Nicholson A, Paolicelli AM, McKenzie AP, Rachubinski AL, Stephens AN, Simon AR, Stedman A, Shocklee AD, Swanson A, Finucane B, … Chung WK (2018). SPARK: A US Cohort of 50,000 Families to Accelerate Autism Research. Neuron, 97(3), 488–493. 10.1016/j.neuron.2018.01.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grant SL, Norton S, Weiland RF, Scheeren AM, Begeer S, & Hoekstra RA (2021). Autism and Chronic Ill Health: An Observational Study of Symptoms and Diagnoses of Central Sensitivity Syndromes in Autistic Adults. Research Square. 10.21203/rs.3.rs-587669/v1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hinz A, Ernst J, Glaesmer H, Brähler E, Rauscher FG, Petrowski K, & Kocalevent R-D (2017). Frequency of somatic symptoms in the general population: Normative values for the Patient Health Questionnaire-15 (PHQ-15). Journal of Psychosomatic Research, 96, 27–31. 10.1016/j.jpsychores.2016.12.017 [DOI] [PubMed] [Google Scholar]
- Holingue C, Newill C, Lee L-C, Pasricha PJ, & Daniele Fallin M (2018). Gastrointestinal symptoms in autism spectrum disorder: A review of the literature on ascertainment and prevalence: Gastrointestinal Symptoms in Autism Spectrum Disorder. Autism Research, 11(1), 24–36. 10.1002/aur.1854 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kass RE, & Raftery AE (1995). Bayes factors. Journal of the American Statistical Association, 90(430), 773–795. 10.1080/01621459.1995.10476572 [DOI] [Google Scholar]
- Kassee C, Babinski S, Tint A, Lunsky Y, Brown HK, Ameis SH, Szatmari P, Lai M-C, & Einstein G (2020). Physical health of autistic girls and women: A scoping review. Molecular Autism, 11(1), 84. 10.1186/s13229-020-00380-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kocalevent R-D, Hinz A, & Brähler E (2013). Standardization of a screening instrument (PHQ-15) for somatization syndromes in the general population. BMC Psychiatry, 13(1), 91. 10.1186/1471-244X-13-91 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kroenke K (2003). Patients presenting with somatic complaints: Epidemiology, psychiatric co-morbidity and management. International Journal of Methods in Psychiatric Research, 12(1), 34–43. 10.1002/mpr.140 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kroenke K, Spitzer RL, Williams JBW, & Löwe B (2010). The Patient Health Questionnaire somatic, anxiety, and depressive symptom scales: A systematic review. General Hospital Psychiatry, 32(4), 345–359. 10.1016/j.genhosppsych.2010.03.006 [DOI] [PubMed] [Google Scholar]
- Leader G, Barrett A, Ferrari C, Casburn M, Maher L, Naughton K, Arndt S, & Mannion A (2021). Quality of life, gastrointestinal symptoms, sleep problems, social support, and social functioning in adults with autism spectrum disorder. Research in Developmental Disabilities, 112, 103915. 10.1016/j.ridd.2021.103915 [DOI] [PubMed] [Google Scholar]
- Leader G, Dooley E, Whelan S, Gilroy SP, Chen JL, Farren Barton A, Coyne R, & Mannion A (2021). Attention-Deficit/Hyperactivity Disorder Symptoms, Gastrointestinal Symptoms, Sleep Problems, Challenging Behavior, Adaptive Behavior, and Quality of Life in Children and Adolescents with Autism Spectrum Disorder. Developmental Neurorehabilitation, 1–12. 10.1080/17518423.2021.1964005 [DOI] [PubMed] [Google Scholar]
- Lever AG, & Geurts HM (2016). Psychiatric co-occurring symptoms and disorders in young, middle-aged, and older adults with autism spectrum disorder. Journal of Autism and Developmental Disorders, 46(6), 1916–1930. 10.1007/s10803-016-2722-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lieb R, Pfister H, Mastaler M, & Wittchen H-U (2000). Somatoform syndromes and disorders in a representative population sample of adolescents and young adults: Prevalence, comorbidity and impairments. Acta Psychiatrica Scandinavica, 101(3), 194–208. 10.1034/j.1600-0447.2000.101003194.x [DOI] [PubMed] [Google Scholar]
- Loomes R, Hull L, & Mandy WPL (2017). What Is the male-to-female ratio in autism spectrum disorder? A systematic review and meta-analysis. Journal of the American Academy of Child and Adolescent Psychiatry, 56(6), 466–474. 10.1016/j.jaac.2017.03.013 [DOI] [PubMed] [Google Scholar]
- Madra M, Ringel R, & Margolis KG (2020). Gastrointestinal issues and autism spectrum disorder. Child and Adolescent Psychiatric Clinics of North America, 29(3), 501–513. 10.1016/j.chc.2020.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mannion A, & Leader G (2016). An investigation of comorbid psychological disorders, sleep problems, gastrointestinal symptoms and epilepsy in children and adolescents with autism spectrum disorder: A two year follow-up. Research in Autism Spectrum Disorders, 22, 20–33. 10.1016/j.rasd.2015.11.002 [DOI] [Google Scholar]
- Mayes SD, Castagna PJ, & Waschbusch DA (2020). Sex Differences in Externalizing and Internalizing Symptoms in ADHD, Autism, and General Population Samples. Journal of Psychopathology and Behavioral Assessment, 42(3), 519–526. 10.1007/s10862-020-09798-4 [DOI] [Google Scholar]
- Mazurek MO, Dovgan K, Neumeyer AM, & Malow BA (2019). Course and Predictors of Sleep and Co-occurring Problems in Children with Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 49(5), 2101–2115. 10.1007/s10803-019-03894-5 [DOI] [PubMed] [Google Scholar]
- Mazurek MO, Keefer A, Shui A, & Vasa RA (2014). One-year course and predictors of abdominal pain in children with autism spectrum disorders: The role of anxiety and sensory over-responsivity. Research in Autism Spectrum Disorders, 8(11), 1508–1515. 10.1016/j.rasd.2014.07.018 [DOI] [Google Scholar]
- Mazurek MO, Vasa RA, Kalb LG, Kanne SM, Rosenberg D, Keefer A, Murray DS, Freedman B, & Lowery LA (2013). Anxiety, Sensory Over-Responsivity, and Gastrointestinal Problems in Children with Autism Spectrum Disorders. Journal of Abnormal Child Psychology, 41(1), 165–176. 10.1007/s10802-012-9668-x [DOI] [PubMed] [Google Scholar]
- Morgan B, Nageye F, Masi G, & Cortese S (2020). Sleep in adults with Autism Spectrum Disorder: A systematic review and meta-analysis of subjective and objective studies. Sleep Medicine, 65, 113–120. 10.1016/j.sleep.2019.07.019 [DOI] [PubMed] [Google Scholar]
- Olsson U, Drasgow F, & Dorans NJ (1982). The polyserial correlation coefficient. Psychometrika, 47(3), 337–347. 10.1007/BF02294164 [DOI] [Google Scholar]
- R Core Team. (2021). R: A Language and Environment for Statistical Computing (4.1.0) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/ [Google Scholar]
- Shattuck PT, Garfield T, Roux AM, Rast JE, Anderson K, Hassrick EM, & Kuo A (2020). Services for Adults With Autism Spectrum Disorder: A Systems Perspective. Current Psychiatry Reports, 22(3), 13. 10.1007/s11920-020-1136-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams ZJ, Cascio CJ, & Woynaroski TG (2021). Measuring Subjective Quality of Life in Autistic Adults with the PROMIS Global-10. ResearchGate. 10.13140/RG.2.2.19753.72804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams ZJ, Everaert J, & Gotham KO (2021). Measuring Depression in Autistic Adults: Psychometric Validation of the Beck Depression Inventory–II. Assessment, 28(3), 858–876. 10.1177/1073191120952889 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams ZJ, & Gotham KO (2021a). Assessing general and autism-relevant quality of life in autistic adults: A psychometric investigation using item response theory. Autism Research, 14(8), 1633–1644. 10.1002/aur.2519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams ZJ, & Gotham KO (2021b). Improving the measurement of alexithymia in autistic adults: A psychometric investigation of the 20-item Toronto Alexithymia Scale and generation of a general alexithymia factor score using item response theory. Molecular Autism, 12(1), 56. 10.1186/s13229-021-00463-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams ZJ, McKenney EE, & Gotham KO (2021). Investigating the structure of trait rumination in autistic adults: A network analysis. Autism, 25(7), 2048–2063. 10.1177/13623613211012855 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang X-L, Liang S, Zou M-Y, Sun C-H, Han P-P, Jiang X-T, Xia W, & Wu L-J (2018). Are gastrointestinal and sleep problems associated with behavioral symptoms of autism spectrum disorder? Psychiatry Research, 259, 229–235. 10.1016/j.psychres.2017.10.040 [DOI] [PubMed] [Google Scholar]
- Yunus MB (2015). An update on central sensitivity syndromes and the issues of nosology and psychobiology. Current Rheumatology Reviews, 11(2), 70–85. 10.2174/157339711102150702112236 [DOI] [PubMed] [Google Scholar]