Neri, D. F., Mallis, M. M., Oyung, R. L., & Dinges, D. F. (1999). Do activity breaks reduce sleepiness in pilots during a night flight? [Abstract]. Sleep 22(Suppl. #1), S150-S151.

Introduction: Vigilance lapses, slowed reaction times, and transient microsleeps have been documented in long-haul pilots operating transmeridian flights at night (Rosekind et al., 1994). These pilots are faced with the requirement to remain vigilant for long periods in an automated environment characterized by physical inactivity, reduced social and cognitive interaction, continuous monitoring for low-frequency events, low light levels, and steady background noise. Mild physical activity (e.g., getting out of the cockpit seat and walking or stretching) combined with social interaction (e.g., talking to flight attendants) is a practical countermeasure that many pilots believe help them fight drowsiness during night flights. However, this assumption has not been systematically tested. We performed the first controlled experiment investigating the effectiveness of regularly scheduled breaks (accompanied by a change from a sedentary to an upright posture, mild physical activity, and an increase in social and cognitive interaction) on vigilance and fatigue during a nighttime flight.

Methods: Using a B747-400 flight simulator, 28 male pilots flew an uneventful 6-hr night flight (~ 0200—0800 hr) from Seattle to Honolulu. They were randomly assigned in pairs either to a treatment condition (n=14), which included five regularly scheduled 7-min activity breaks (walking and social interaction) equally spaced during the cruise portion of the flight (1/hr), or to a control condition (n=14), which received a single, mid-flight, 7-min break. Breaks involved one pilot exiting the flight deck while the other remained in the cockpit seat. Four categories of data served as outcomes: (1) pilot ratings of sleepiness using a visual analog scale (VAS) at 15 and 25 min post-break and the Karolinska Sleepiness Scale (KSS) at 5 and 40 min post-break, (2) psychomotor vigilance test (PVT) performance administered pre-flight, in-flight (15 min post-break) and post-flight, (3) continuous EEG/EOG, and (4) continuous video of slow eyelid closures (PERCLOS). In the absence of breaks for the control group, sleepiness ratings and the PVT were measured at corresponding times.

Results: Analyses have been completed on PVT and subjective sleepiness rating scales (analyses on EEG/EOG and PERCLOS data are underway). As expected, sleepiness ratings and PVT performance lapses increased significantly across the night flight in both treatment (T) and control (C) conditions. ANOVA and t-tests revealed that, relative to the C condition, the T condition reduced subjective sleepiness up to 25 min post-break, especially for the breaks taken in the latter half of the flight during periods of higher sleepiness. For example, following the break at 0520 hr, the T group reported less sleepiness than the C group on the KSS 5min post-break (p=0.0001) and on the VAS 15 min (p=0.011) and again 25 min post-break (p=0.01). After the break at 0620 hr, the T group continued to report less sleepiness than the C group on the same measures 5 min (p=0.001), 15 min (p=0.005), and 25 min post-break (p=0.009). This T group improvement in subjective sleepiness was not evident on the KSS 40 min post-break. More importantly, there was no statistical evidence that breaks improved PVT performance 15—25 min post-break.

Conclusions: During a simulated, 6-hr night flight, experienced flight crews reported feeling more alert up to 25 min following hourly brief (7-min) breaks that combined getting out of the cockpit seat, walking, talking with a research monitor, and occasionally consuming non-caffeinated items. However, these benefits were not evident in objectively probed PVT performance, which showed deterioration due to elevated sleep drive and circadian time. It is possible that PVT performance was insensitive to the alertness-promoting effects of the activity breaks. However, if other objective indices of fatigue (EEG/EOG, PERCLOS) currently being analyzed confirm the PVT results, this experiment will document that breaks afford a transient "masking" effect of sleepiness without providing an enduring countermeasure to it at a level comparable to that found for a single, 26-min cockpit nap (Rosekind et al., 1994).

Rosekind, M. R., et al. (1994). Crew Factors in Flight Operations IX: Effects of Planned Cockpit rest on Crew Performance and Alertness in Long Haul Operations (NASA TM 108839). NASA Ames Research Center.