Minimal olfactory perception during sleep: why odor alarms will not work for humans

To examine olfactory arousal threshold during sleep in comparison to an auditory tone.

On night 1, participants rated odor intensity when awake and experienced olfactory stimuli during stage 1 sleep. Night 2 involved stage 2, stage 4, and rapid-eye-movement (REM) sleep trials using the "staircase" threshold-detection method. Electroencephalogram, electrooculogram, electromyogram, electrocardiogram, and respiration were recorded along with behavioral response. An 800-Hz tone was given on trials when odors failed to arouse.

Participants slept in individual rooms. Stimulus-delivery systems were operated from a separate room, where an experimenter observed physiologic recordings and behavioral responses.

Three healthy men and 3 women aged 20 to 25 years (mean, 22 years).

Two odorants, peppermint and pyridine, at 4 concentrations were presented through nasal cannulas using an air-dilution olfactometer. Tones were played over a speaker.

Behavioral (button press and oral) responses, electroencephalographic activation, and changes in breathing and heart rate were assessed.

Participants responded to odors on 92% of stage 1 sleep trials. Peppermint was ineffective in stages 2, 4, and REM sleep. Pyridine produced behavioral threshold on 45% of stage 2 trials, none in stage 4, and one third of REM sleep trials. Tones were effective on at least 75% of trials. Heart rate increased significantly only following behavioral responses to odors or tones across sleep stages.

The data indicate that human olfaction is not reliably capable of alerting a sleeper.