Personal light dosimetry in permanent night and day workers.

Light exposure was measured in six day and six night watches (working 12-hour shifts five days in a row) during 48 h on work days and 48 h on days off using a photocell with a sensitivity corresponding to photopic vision. The photocell was mounted on a frame of spectacles, thus measuring in viewing direction. Light exposure was low both in night and day watches; however, in night watches exposures were significantly lower: On work days, night watches spent a mean of 13 min above 1,500 lx, day watches 52 min; on days off, night watches spent 3 min above 1,500 lx but day watches 89 min. Unexpectedly, night watches had no higher exposure during days off. We suspect that this is due to a light avoidance tendency in permanent night workers. High negative correlations between the acrophases of subjective state (e.g., alertness and mood) and light exposure in night watches indicate that bright light would probably increase desynchroniza-tion between subjective state, sleep, and activity.