A standard procedure enhances the correlation between subjective and objective measures of sleepiness

The goal of this study was to assess whether instituting a standard procedure to minimize transient activation prior to the subjective rating of sleepiness can improve the predictive value of the rating process.

Thirty young adults, aged 19 to 26 years, participated in the study. Subsequent to sleeping at home with bedtime restricted to 5 hours, they came to the sleep laboratory. They were instructed to rate their level of sleepiness on the Stanford Sleepiness Scale (SSS) and visual analog scales (VAS). A "calm-down" procedure, sitting quietly with eyes closed for 1 minute, was instituted prior to sleepiness ratings for half of the subjects (experimental group) but not for the other half of the subjects (control group). A nap trial with polysomnographic recording was then conducted, followed by a vigilance test.

For the experimental group, VAS results of "sleepiness" and "alertness" both correlated significantly with sleep-onset latency during the nap (SOL: r = -.62 and .64, respectively, P values < .05) and with reaction time (RT) on the vigilance test (r = .56 and -.54, P values < .05). The SSS ratings showed significant correlation with nap SOL (r = -.58, P < .05) but not with RT on the vigilance test (r = .19, p = .52). For the control group, none of the subjective ratings showed significant correlation with objective measures. The differences between the resultant correlations for the 2 groups were statistically significant for 2 sets of correlations: the correlation between VAS of "alertness" and nap SOL and the correlation between VAS of "sleepiness" and RT on the vigilance test.

The results indicate that the subjective ratings of the sleepiness state for individuals with mild sleep restriction more faithfully reflect a physiologic tendency to fall asleep as well as cognitive attentiveness when the ratings are conducted subsequent to sitting still with eyes closed for a sufficient time to minimize transient activation.