31982
Feasibility of Measuring Physiological Reactivity in More Severely Affected Youth with Autism in a Psychiatric Inpatient Setting
Previous literature demonstrates that individuals with Autism Spectrum Disorder (ASD) can present with arousal dysregulation and difficulties with arousal modulation; however, objective physiological measures of arousal are seldom used in these studies, especially in youth with ASD who are severely affected (i.e., minimally-verbal and/or co-morbid intellectual impairments).
Objectives:
Assess the feasibility of exposing more severely affected youth with ASD from an inpatient psychiatric unit to experimental tasks designed to elicit frustration while wearing a wireless biosensor. We hypothesized that there would be significantly more physiological reactivity during demand tasks than at rest.
Methods:
Forty-five participants drawn from a larger sample (mean age = 12.8, range 6-20yrs; 74.1% male; 20.2% racial or ethnic minority; non-verbal IQ mean = 76.5, sd = 23.7; ADOS Module distribution: 28.9% Module 1 non-verbal, single words; 10.7% Module 2 phrase speech; 44.6% Module 3 fluent speech, child; 17.9% Module 4 fluent speech, adolescent/adult) of the Autism Inpatient Collection (AIC) were administered a subset of tasks designed to elicit negative affect (an inhibitory control task: Tower and two frustration/anger tasks: Locked Box, End of the Line) from the Laboratory Temperament Assessment Battery (Lab-TAB) that began with and interspersed baseline rest conditions. A wrist-worn wireless biosensor (Q Sensor, Affectiva, Inc.) was used to record electrodermal activity (EDA) and resulting data was parameterized within each segment (i.e. task) as normalized mean skin conductance level (SCL), linear trend change, and skin conductance response (SCR) as measured by coefficient of variation (CV).
Results:
An exploratory repeated-measures ANOVA with 7 factors (4 baselines and 3 tasks) revealed a main effect of task on CV (F(6)=17.5, p < 0.0001) and all Bonferroni-adjusted t-tests showed statistically significant differences between sequential resting baselines and demand tasks (Table 1, Figure 1). Log-adjusting the data to achieve assumptions of sphericity did not change results. The normalized SCL (mean) and trend were not significantly different across demand tasks, seemingly due to similarly escalating arousal during each frustration-inducing condition.
Conclusions:
The current study is the first we are aware of to measure physiological reactivity during standardized tasks in more severely affected youth with ASD in a psychiatric inpatient setting. SCR increases during demand tasks vs. antecedent baselines and an overall increase in SCL and trend throughout the Lab-TAB suggest that the tasks effectively increase physiological arousal associated with distress/frustration/negative affect. We conclude that ambulatory EDA is a feasible and useful objective measure of arousal for those with severe ASD who are unable to provide reliable self-reports, potentially enhancing future studies of temperament, emotion regulation, and emotion expression in this understudied segment of the autism population. Future work is underway to code these sessions for affect/emotion expression and explore relations between arousal, frustration, and emotional response.