Autism spectrum disorder (ASD) affects a person’s ability to interact socially with others and communicate using both verbal and non-verbal means and is marked by repetitive, restricted or stereotyped behaviors. This disorder has a wide range of presentations and creates a challenge not only for those who care for or support this person, but also for the person diagnosed. Persons with this disorder may experience challenges directly related to the disorder in a variety of areas (responding appropriately to common social cues or understanding another person’s point of view), often with varying levels of symptom intensity or severity (limited or constricted affect to a completely flat affect). The associated level of cognition may be very high, and the person may be academically gifted, or the person with ASD may have severe cognitive impairments and behavioral challenges such as self-injury or aggression. In addition, a wide variety of co-morbid challenges also may surface such as social isolation, attentional issues, depression and/or anxiety, vocational issues, or a sense of failure or negative self-image. Recent studies suggest that nearly 40% of persons with an ASD diagnosis will report depressive symptoms during their lifetimes (Hudson et al., 2018), and the overall prevalence rate for anxiety in ASD is between 42 and 79% (Kent & Simonoff, 2017). These varying symptoms and associated challenges are associated with a set of stressors that sometimes overlap with those experienced by typically developing persons (talking in front of a group of people, for example) and are sometimes less familiar to non-ASD populations (waiting in line, participating in a group activity, or being touched, for example). This creates a unique opportunity and need for accurate assessment of stressors in ASD, especially in lower functioning populations where the ability to communicate is severely impaired, so that targeted interventions can be generated and implemented.

Groden et al. (2001) developed the Stress Survey Schedule (SSS) based on the idea that the very characteristics of autism along with impaired the common coping strategies and supports available to typically developing populations created a high-stress situation. This situation was suggested to be present even prior to the individual interacting in the social realm, setting the stage for high levels of anxiety. The solution suggested by these researchers was to better identify stressors and improve coping so that this anxiety could be reduced. To this end, the SSS was designed to measure stressors in a wide range of presentations of persons with autism to better identify the stimuli and situations that may be causing stress and potentially leading to associated anxiety and challenging behaviors. With a better understanding of what stressors were activated in each person, staff supporting these clients may be better able to target interventions that maximized coping. Common stressors were generated, grouped into categories, and data was collected on a 49-item version that generated eight subscales: “Anticipation or Uncertainty,” “Changes and Threats,” “Unpleasant Events,” “Pleasant Events,” “Sensory or Personal Contact,” “Food-related Activities,” “Social Environmental Interactions,” and “Ritual-related.” A study by Goodwin et al. in 2007 established the validity of the scale as well as high internal consistency and found no differences for males and females. The 180 persons rated ranged in age from 3 to 41 years of age and had an average age of 17.4 years. These researchers also found an increase in severity of stress responses for three of the eight subscales as a function of age (“Anticipation or Uncertainty,” “Sensory or Personal Contact,” and “Food-related Activities”). Specifically, the 1–10 age group showed lower stress on “Anticipation or Uncertainty” and “Sensory or Personal Contact” as compared to the 21–30 and 31–41 age groups, and the 21–30 age group showed lower stress on “Food-related Activities” as compared to the 31–41 age group. This finding was suggested to be attributed to being more sheltered as a young person and decreasing levels of available support, routine, and structure as the person aged.

Woodard et al. (2020) compared the responses of a sample of students diagnosed with ASD and intellectual disability to a group of typically developing students. The purpose of the study was to expand the psychometrics of the SSS by providing information on the divergent validity of the measure, further assess the internal consistency, and examine changes in scores between pre-adolescent and adolescent sub-groups. The research supported the divergent validity of the scale, with the participants with ASD and intellectual disability scoring significantly higher than the typically developing group. The results also found high internal consistency, especially for the group with ASD and intellectual disability, and no difference in scores for male and females in either group. Further, the group with ASD and intellectual disability was divided into three age groups (ages 6–12, 13–17, and 18–21) and only the “Ritual” subscale has a positive association with age. Specifically, the difference was between the oldest group assessed (ages 18–21) and the two younger age groups. No appreciable associations were found between age and the other seven subscales. This finding was generally consistent with Goodwin (2007),

in that age-related changes were not apparent within younger (1–10 and 11–20 years of age) age groups and did not emerge until stress scores were compared to an older group (21–30 and 31–41 years of age).

In the present study, we expand upon the work done by Woodard et al. in 2020 and examine a larger group of students ages 6 to 19 with various forms of developmental disability. The purpose of this study is to identify the overall and sub-group stress levels for this group and to provide normative data on a population for which there is limited research to create a framework of comparison for SSS scores. Comparisons of stress scores by gender and age as well as information on commonly high and low scoring stressors for each age group will be presented. The purpose of this is to better understand how stressors may or may not change as children with ASD and intellectual disability become young adults. This will allow for more accurate identification of clinical manifestation of stress, and to plan and evaluate targeted strategies to reduce either exposure to the trigger for the stress response, or to develop strategies the individual can employ when exposed to such triggers.

Method

Participants

The total sample for this study was composed of 155 Groden Center students who were 6 to 19 years of age. Each student was diagnosed with an intellectual disability in the severe to profound range. In addition, 118 students had a primary or secondary diagnosis of autism, and 10 were diagnosed with pervasive developmental disorder (PDD) as diagnosed by a qualified clinician. The Groden Center is a treatment and educational center in Providence, RI, for students with autism, intellectual disabilities, and other developmental disorders. Students also typically present with significant behavioral challenges and comorbid diagnoses that cannot be safely managed in a typical school setting. The center focuses on positive psychology principles and the development of effective coping skills, while maintaining a behavioral psychology orientation. The students in the present study were not excluded for the presence of comorbid psychiatric diagnoses, the use of medication or for the presence of seizure disorder. Each student had the SSS completed by family members (n = 69, 46.3%) or staff member who was familiar with the person, and surveys were typically completed as a part of the intake process for admission and then as needed. Table 1 displays the characteristics of the students, who were predominantly male (77.4%). The mean age of the sample was 12.4 years (SD = 3.7), and 50 participants (32.3%) were age 6–10, 44 (28.3%) were age 11–13, and 61 (39.4%) were 14 years of age or older. Across students, 118 (76.1%) had a primary or secondary diagnosis of autism, and most were identified as having verbal ability (n = 100, 64.5%). Most students (n = 141, 82.5%) lived at home with a family member.

Table 1 Demographic characteristics of students
Full size table

Procedures

Caregivers (parents or other family members or foster parents) or center staff (direct care staff, intake staff, teachers, or other staff members) familiar with the student completed the SSS and provided ratings on their perception of the intensity of the stress response of the student to each of the events. These surveys were completed between 2002 and 2021 and were rated in paper form and then entered into each student’s electronic record. The responses by the rater were used to indicate potential areas for focused behavioral interventions typically used at the center, including relaxation procedures and cognitive picture rehearsal procedures. The SSS was completed as part of a larger intake process that includes areas such as functional behavior analysis, assessment of repetitive behavior trends, psychiatric rating scales, and others. All data was de-identified for the purposes of this study.

Measure

The Stress Survey Schedule (SSS) is a 49-item measure of events and experiences that may be associated with stress. Items are scored on a five-point Likert type scale with 1 indicating the individual exposed to the event typically experiences “None to mild” stress and 5 “Severe” stress. The development and validation of the SSS has previously been reported (Goodwin et al., 2007; Woodard et al., 2020) on. There are 8 composite sub-scale scores that compose the SSS: “Changes and Threats” (11 items), “Anticipation or Uncertainty” (7 items), “Unpleasant Events” (9 items), “Pleasant Events” (8 items), “Sensory or Personal Contact” (4 items), “Food-related Activity” (3 items), “Social or Environmental Interactions” (3 items), and “Ritual-related Stress” (4 items). Items within each sub-scale are summed and averaged to produce sub-scale scores between 1 and 5. A total stress score was also calculated from the average across sub-scales, again fixing an average overall SSS score on a 1 to 5 scale (maximum = 5). Measures of inter-item consistency and scale reliability were found to be in the acceptable to high ranges using both Cronbach’s α and as acceptable using McDonald’s ω (Cronbach’s α = 0.62 to 0.88.0, McDonald’s ω = 0.56 to 0.88). Demographic data were also collected on the students including gender, age at SSS administration, gender, relationship to the person completing the ratings, primary diagnosis, and living situation.

Data Analyses

Data were downloaded and cleaned and then analyzed using SAS (version 9.4, Cary NC). Descriptive analyses were conducted and used to describe the students and raters, as well as bivariate associations between variables. Descriptive analyses were reported as frequencies with 95% confidence intervals (CI) or means with standard deviation (SD). Scores across the 49 SSS items were summed to provide mean subscale scores (1–5 maximum) or total average stress (1–5 maximum). The raw scores across items were also summed to provide a total raw stress score (maximum score = 245). These score metrics were used in the analysis.

Student ages were classified into three groups: elementary school age (6–10 years), middle school (11–13 years), and high school (14–19 years). SSS scores by age group and gender were calculated for mean, SD, and 95% CI. Additionally, floor and ceiling scores were computed for each sub-scale and total average score using the cut-off suggested in prior research (1–2 = floor or low stress, 4–5 = ceiling or severe stress). For the total raw score where the maximum score = 254, the floor and ceiling values were calculated (98 = upper bounds of low stress, > 147 = lower bounds of severe stress). The percentage (with 95% CI) of students rated in the floor to ceiling category is reported for each SSS score. Differences in these score metrics by student primary diagnosis (autism vs. not) are also reported. A generalized linear model (to adjust for unequal sample sizes in the groups of interest) to examine differences by age groups, gender, and gender by age interaction across the mean subscale and total stress scores was conducted. Tukey tests to adjust for multiple pairwise comparisons were employed. To examine differences in the probability of inclusion in the floor or ceiling groups across the scores, a series of Fisher’s exact chi square test (χ2) were conducted with the frequency of being classified as ceiling stress levels versus floor as the outcome variables and gender, age-group, and autism diagnosis as predictor variables. Additionally, an item analysis was conducted to determine the five items with the greatest frequency of floor and ceiling scores by age group and autism diagnosis.

Results

SSS Scores

Across all students, the overall mean scores for the eight subscales varied, as can be seen in Table 2. Overall reported stress was significantly lower for “Positive Events” (mean = 1.6, 95% CI = 1.5, 1.7) and “Social/Environmental Interactions” (mean = 1.7, 95% CI = 1.6, 1.9) compared to “Changes and Threats” (mean = 2.6, 95% CI = 2.5, 2.7) and “Food-related Events” (mean = 2.6, 95% CI = 2.4, 2.7) sub-scale scores.

Table 2 Mean and 95% confidence interval of subscale and total stress scores
Full size table

Table 3 shows the mean (SD and CI) scores by age group and gender, with the percentage at the floor of score distribution (score = 1–2 indicating no or low stress) and ceiling (score = 4–5 indicating high stress). There was variability across age and gender groups for the sub-scale and total score metrics. Generalized linear models found no main effects of a gender and age interaction across any of the sub-scales or total stress scores, so the main effect of gender and age were examined independently. Across these scores, the mean for “Unpleasant Events” and “Pleasant Events” was significantly higher for males.

Table 3 Stress survey scale scores by gender and age
Full size table

compared to females (“Unpleasant Events” male mean = 2.4, SD = 0.70; female mean = 2.2, SD = 0.6; F [1,154] = 3.27, p = 0.048; “Pleasant Events” male mean = 1.7, SD = 0.6; female mean = 1.4, SD = 0.5; F [1,154] = 3.7, p = 0.04). This effect was independent of age, as there were no significant differences across the subscale and total stress scores for the age groups. Figure 1 shows the mean scores for males compared to females across the eight subscale and total average stress score.

Fig. 1
figure 1

Female and male average sub-scales and total average stress scores

Full size image

A series of χ2 examining the frequency of being in the floor versus ceiling groups across the subscale scores was conducted. For the “Changes and Threats” subscale, a significantly greater proportion of students ages 6–10 were at the ceiling (37.5%) compared to the other two age groups (12.5% in each, χ2 [2] = 7.21, p = 0.03). There were two significant differences by gender; “Sensory/Personal Contact” subscale had a significantly greater proportion of males being at the ceiling than females (11.4% vs, 0%, χ2 [1] = 3.74, p = 0.05) and for the ceiling raw total stress score (males = 32% vs. females = 10% χ2 [1] = 4.25, p = 0.04). There were no significant differences across the ceiling versus floor groups for students who had an autism diagnosis versus those who did not.

Autism

There was no main difference by autism diagnosis by age group or gender across the subscale and total stress scores. An age group by gender interaction found that among the older students (age ≥ 14 years) with autism the mean “Changes and Threats” subscale score was significantly higher (mean = 2.57, SD = 0.75) compared to students in this age group not diagnosed with autism (mean = 2.08, SD = 0.74; F (5,154) = 2.60, p = 0.03).

Item Analysis

Table 4 shows the items with the lowest and highest stress scores by age and autism diagnosis. There were two items that were identified as low stress for all age groups including “Receiving a present” and “Having something marked as correct.” “Someone else making a mistake” was a low stress item for the younger two groups, and some type of reinforcement (verbal, activity or tangible) was low stress for the older two age groups. Items that were related to high stress were identical for the two youngest age groups and are listed in Table 4. In contrast, the items that represented high stress were mostly different for the oldest age group and included items related to the physical environment and the presence of others, rituals being prevented, and having challenges related to communication.

Table 4 Highest and lowest stress items by age and autism diagnosis
Full size table

Discussion

The purpose of the present study was to provide normative data on overall scores and sub-group scores for the Stress Survey Schedule (SSS) for persons with autism and intellectual disability under the age of 19 and to examine trends in the participant sample related to age and gender. The overall mean of the sample indicated moderate levels of stress (see Table 2), while averages of subscale scores ranged from low to moderate levels of stress. The sub-group averages for the entire sample suggested that this sample of children, adolescents, and young adults has the least stress related to “Pleasant Events” and “Social/environmental Conditions” and that these are significantly lower than the higher stress sub-groups of “Changes and Threats” and “Food-related Activities.” With the exception of a statistically significant difference on the subscales of “Pleasant Events” and “Unpleasant Events,” there were no differences between males and females on sub-scale scores for the total sample, and there were no differences on average subscale scores as a function of age. These gender-related findings are generally consistent with the findings of Woodard et al. (2020) and Goodwin et al. (2007), which both found no differences in sub-scale scores between males and females. The age-related findings are also generally consistent with previous research, in that no age differences were found in the Goodwin et al. (2007) sample under age 21, and only one sub-group score in the Woodard et al. (2020) study (“Ritual-related”) was lower in the 6–17 age group as compared to the 18–21 age group.

Further, with the exception of the “Changes and Threats” subscale scores which were higher for participants diagnosed with autism, there were no differences in subscale scores related to gender or age for those diagnosed with autism and intellectual disability compared to related developmental disability diagnoses and/or intellectual disability alone. These findings as a group suggest that for a population with autism or other developmental disability and intellectual disability, overall stress is rated as moderate by persons familiar with the individuals assessed, and averaged scores of identified sub-scales are generally consistent across gender, as well as childhood, adolescent, and young adult age groups. The data in Table 3 provides a point of reference for each subscale score by age. Deviations from these values may be clinically significant and indicate a need for further assessment, investigation, intervention, or treatment. Additionally, the estimates of proportions of the population expected to be in the ceiling range of the stress subscale scores by age group have the potential to be used for clinical-cut-off guidance.

Past research suggests that differences in what are considered high stressors for this population may begin around the age of 21 (Goodwin et al. (2007)) or 18 to 21 Woodard et al. (2020), in that differences in averaged sub-scale scores emerge. For this reason, we investigated differences at the item level in stressors endorsed for the different age groups in the present sample. Based on past findings, we would expect consistency among the younger age groups and shifting high stress items for the older group. Table 3 provides information on items commonly endorsed as low stress (ratings < 3) and high stress (ratings > 3) for the identified age groups. It is notable that across all age groups, “Receiving a present” is consistently scored as low stress. This item and others consistently scored as low stress experiences would also likely be considered low stress, neutral, or even enjoyable to typically developing persons. For example, “Having something marked as correct” and receiving variable forms of reinforcement were frequently endorsed as low stress items across age groups. However, “Receiving hugs and affection” was only low stress for the youngest group, suggesting that as this population ages through adolescence and young adulthood, physical contact may become increasingly stressful. “Losing at a game” and “Having something marked as incorrect” became more frequently endorsed as low stress as the population moved into adolescence and young adulthood. This may reflect a declining emphasis or concern for the perceptions or evaluations by others in the person’s environment.

At the other end of the stress spectrum, as expected, there were a number of individual items consistently marked as high stress across the two youngest age groups. These included “Waiting for preferred events” “Transitioning from preferred to non-preferred activities,” “Being told ‘no’,” and “Having to engage in a not-liked activity.” All of these items represent external demands or limits, expectations inconsistent with the desires of the individual, shifts or changes, or the ability to be flexible, all of which are areas of challenge inherent in an autism diagnosis. Alternately, this may reflect an inability to understand the perspectives of others or an inability to comprehend environmental conditions or limitations. The high stress endorsement of these items may be compounded or exacerbated by “Not being able to communicate needs,” as the person may lack the ability to effectively communicated desires such as wanting to remain engaged in a preferred activity.

Also as expected, there were significant differences in which items were frequently endorsed as high stress for the oldest age group in this study compared to the two younger groups. “Being unable to communicate needs” is the only items shared with the younger groups. “Noise or disruption,” Uncomfortable environments,” and “Crowding” all suggest and increased sensitivity to external or environmental physical conditions, perhaps partially related to the presence of people. This would be consistent with the preference for solitary activities often present in persons with autism. Relatedly, “Blocking rituals” may demonstrate an extension of the inflexibility present in the younger sub-groups, which perhaps becomes more focused into a specific behavioral pattern or ritual as the person ages. It is notable that three of these four items load on the “Sensory or Personal Contact” or “Ritual-related” sub-groups, which were the sub-groups found to be higher for older groups in previous research. This suggests that examination at the group average as well as individual item levels may prove fruitful, as averages may obscure consistent and important findings.

As with typically developing persons, persons with autism experience stress that is likely worsened by needing to function in a very social world with coping skills that are limited by features of the autism diagnosis. Being able to identify specific stressors in persons with autism assists the clinician in developing targeted treatment interventions that may reduce the level of stress experienced or conversely increase the ability to cope. This may, in turn, decrease challenging behaviors that result from high stress experiences and/or increase a person’s ability to actively take part in the world around them in a more positive way. The Stress Survey Schedule was developed with this purpose in mind, and the present study provides reference points for comparison, so that persons who make use of this scale can quickly and easily identify items that diverge from the norm. The results of the present study also indicate that while stressors may be consistent for younger persons with autism, as a person becomes a young adult, specific areas of high stress tend to change in important ways. A general shift in stress experiences from internally directed desires to externally experienced environmental conditions, as well as ritual-related stress is evident. It is important to note these shifts and focus on the idiosyncratic features of each person assessed, as grouped or averaged data may tend to “wash out” important information. Only by assessing each possible area of stress on a person by person, item-by-item basis, can those who support persons with autism generate effective supports.

Limitations and Future Research

Although the sample allows for some conclusions to be drawn about the overall sample, it is insufficient in age representation and non-ASD diagnosis to allow more robust estimates of differences to be drawn. With larger samples of participants in each age group, additional differences may emerge. Further, we would hypothesize that extending the age ranges would produce more information on how stressors change as a person ages. Currently, information in aging and stressors in persons with intellectual disabilities and autism is limited and needs more exploration. A third possible area of inquiry related to the present research involves the observable behavioral profiles related to certain sub-groups of stressors. Specifically, are there groups or types of behavior challenges that can be associated with certain sub-groups of stressors? If so, what is the hypothesized reason for such a relationship? This line of inquiry might shed light on behavioral repertoires by better understanding stress.