UCSF

Objective

To examine the association between physiological hyperarousal and response accuracy on reaction time tasks among individuals with insomnia.

Design and setting

This study was conducted at affiliated Veterans Administration (VA) and academic medical centers using a matched-group, cross-sectional research design.

Participants

Eighty-nine individuals (48 women) with primary insomnia, PI (MAge = 49.8 ± 17.2 y) and 95 individuals (48 women) who were well-screened normal sleepers, NS (MAge = 46.9 ± 17.0 y).

Methods and measures

Participants underwent 3 nights of polysomnography followed by daytime testing with a four-trial Multiple Sleep Latency Test (MSLT). Before each MSLT nap, they rated their sleepiness and completed computer-administered reaction time tasks. The mean number of correct and error responses made by each participant across testing trials served as dependent measures. The PI and NS groups were each subdivided into alert (e.g., MSLT mean onset latency > 8 min) and sleepy (e.g., MSLT mean onset latency ≤ 8 min) subgroups to allow for testing the main and interaction effects of participant type and level of alertness.

Results

Alert participants had longer MSLT latencies than sleepy participants (12.7 versus 5.4 min), yet both alert and sleepy individuals with PI reported greater sleepiness than NS. Alert participants also showed lower sleep efficiencies (83.5% versus 86.2%, P = 0.03), suggesting 24-h physiological hyperarousal particularly in the PI group. Individuals with PI had fewer correct responses on performance testing than did NS, whereas a significant group × alertness interaction (P = 0.0013) showed greater error rates among alert individuals with PI (mean = 4.5 ± 3.6 errors per trial) than among alert NS (mean = 2.6 ± 1.9 errors per trial).

Conclusions

Physiological hyperarousal in insomnia may lead to more apparent daytime alertness yet dispose individuals with insomnia to higher error rates on tasks requiring their attention.