Abstract Background Beyond the core symptoms of autism spectrum disorder (ASD), associated symptoms of anxiety can cause substantial impairment for individuals affected by ASD and those who care for them. Methods We utilized a potentiated startle paradigm with a puff of air to the neck as the unconditioned stimulus in order to investigate differences between response to cued fear and contextual anxiety among cognitively able adolescents diagnosed with ASD and an age- and IQ-matched typically developing group. Results In a threat-modulated startle paradigm, response patterns to neutral, predictable, and unpredictable conditions were comparable across typically developing and ASD youth in terms of startle response magnitude and latency. However, the ASD group showed significantly greater absolute startle responsivity at baseline and throughout the experiment, suggesting possibly enhanced general sensitivity to threatening contexts. The ASD group, but not the control group, demonstrated moderate to strong negative correlations between psychophysiological response to unpredictable threats (uncertainty) and questionnaire measures of generalized anxiety, intolerance of uncertainty, and repetitive behavior. Conclusions Our data suggest enhanced general reactivity among the ASD group, possibly reflecting greater sensitivity to the threatening context of the startle paradigm. Associations with the response to uncertainty may help explain shared neurobehavioral mechanisms in ASD and anxiety. This task can provide useful targets for future neuroimaging and genetics studies as well as specific avenues for intervention. We emphasize the importance of further basic and clinical research into links among these important constructs.

Objective

Individuals with schizophrenia have disproportionate memory impairments when encoding relational versus item-specific information, and when using recollection versus familiarity during retrieval. It is unclear whether this pattern is unique to people with chronic schizophrenia, or if it occurs in individuals after a first episode of psychosis (FE), or when at clinical high-risk for psychosis (CHR).

Methods

We administered the Relational and Item-Specific Memory task (RiSE) to 22 CHR, 101 FE, and 58 typically developing (TD) participants. We examined group differences in item and relational encoding, and familiarity-based and recollection-based retrieval using parametric analysis and structural equation modeling (SEM). Longitudinal data allowed us to examine relations between baseline RiSE performance and change in clinical symptoms at 1-year follow-up in the FE group.

Results

Groups did not differ on familiarity. FE and CHR groups were equally impaired on overall recognition accuracy. Although recollection was impaired in both FE and CHR groups following relational encoding, only the FE group had impaired recollection following item encoding. SEM showed atypical relationships between familiarity and recollection, as well as familiarity and item recognition for both the FE and CHR groups. For FE individuals, better baseline recognition accuracy predicted less severe negative symptoms at 1-year follow-up.

Conclusions

Impaired relational and recollective memory may reflect neurodevelopmental abnormalities predating conversion to psychosis. These memory deficits appear related to negative symptom changes. In contrast, item specific recollection deficits appear to occur after the development of full psychosis. Familiarity appears to be a relatively preserved memory function across the psychosis spectrum.