Demonstration of additivity failure in human circadian phototransduction.

OBJECTIVES: Published data, both on nocturnal melatonin suppression in humans and on widely accepted retinal structure and function, suggest that spectral opponency plays a role in human circadian phototransduction. We directly test subadditivity, implied by spectral opponency, in human circadian phototransduction in response to nearly monochromatic and to polychromatic light.

METHODS: Adult male human subjects were exposed for 60 minutes to two intensities each of two lighting conditions, during nighttime experimental sessions. One condition consisted of light from mercury vapor lamps (450 and 1050 lx), and one condition consisted of light from these lamps filtered such that only the spectral line from this lamp at 436 nm was presented to subjects (7.5 and 15 lx).

RESULTS: Melatonin suppression from the filtered illumination at 436 nm alone was greater than mercury lamp illumination (containing energy at 436 nm in addition to other wavelengths), even when the sources exposed subjects' retinae to equal amounts of irradiance at 436 nm.

CONCLUSION: This direct test of subadditivity, together with evidence from neuroanatomy, supports the inference that spectral opponency is a fundamental characteristic of human circadian phototransduction.