Does anxiety sensitivity correlate with startle habituation? An examination in two independent samples

Abstract

Individuals with anxiety disorders have previously demonstrated abnormal habituation to aversiveness over time. As anxiety sensitivity (AS), or an individuals' propensity to fear of anxiety-related sensations, has been shown to be a risk factor for anxiety disorders (particularly panic disorder), the present study examined whether AS was also associated with abnormal habituation. This association was examined in two independent samples of undergraduates (N_total=178). Habituation was operationalised as the reduction in startle response to multiple startle probes presented over 2.5 minutes and three definitions of this reduction were employed. Results indicated that individuals with higher levels of AS evidenced deficits in startle habituation, but the strength of this relationship was somewhat dependent on the definition of startle habituation, with the most robust definition being an analysis of participants' individual slopes across all nine blinks. The present findings suggest that startle habituation is a key mechanism underlying AS, and may help elucidate the role this risk factor plays in the pathogenesis of anxiety disorders.

Keywords:

• Anxiety sensitivity
• Startle habituation
• Startle reduction

Notes

¹ In Sample 1, seven were excluded because of mechanical error; one due to fewer than six of nine usable blinks; five due to an unusable first blink; and six due to a missing ninth blink, resulting in 89 usable subjects. In Sample 2, three were excluded because of mechanical error; five due to fewer than six of nine usable blinks; five due to an unusable first blink; and three due to a missing ninth blink, resulting in 53 usable subjects.

² Analyses with the ASI-R yielded a similar (although slightly weaker) pattern of results. None of our analyses were significant using the ASI-O.

³ We also examined the relationship between each of our personality variables and startle onset latency across each of the nine blinks, using an analyses strategy identical to what was used when startle amplitude was our dependent variable. None of these analyses yielded significant findings. These null findings are consistent with studies that have failed to find a relationship between anxiety-related traits and startle latency in non-clinical populations (Grillon, Ameli, Foot, & Davis, 1993).

⁴ This discrepancy between the two samples may be due to the fact that ASI was significantly associated with the first blink in Sample 2 (r=−.404, p=.003), but not Sample 1 (r=−.037).

Additional information

Funding

This project was supported by an NIH R21 grant MH08069, the University of Illinois at Chicago Chancellor's Fund, and a Center for Clinical Translational Science (CCTS) grant UL1TR000050 awarded to the University of Illinois at Chicago.