Melatonin suppression is exquisitely sensitive to light and primarily driven by melanopsin in humans.

INTRODUCTION: Light elicits a range of non-visual responses in humans. Driven predominantly by intrinsically photosensitive retinal ganglion cells (ipRGCs), but also by rods and/or cones, these responses include melatonin suppression. A sigmoidal relationship has been established between melatonin suppression and light intensity; however, photoreceptoral involvement remains unclear. METHODS AND RESULTS: In this study, we first modelled the relationships between alpha-opic illuminances and melatonin suppression using an extensive dataset by Brainard and colleagues. Our results show that (a) melatonin suppression is better predicted by melanopic illuminance compared to other alpha-opic illuminances, (b) melatonin suppression is predicted to occur at levels as low as ~1.5 melanopic lux (melanopsin-weighted irradiance 0.2 µW/cm2 ), (c) saturation occurs at 305 melanopic lux (melanopsin-weighted irradiance 36.6 µW/cm2 ). We then tested this melanopsin-weighted illuminance-response model derived from Brainard and colleagues' data and show that it predicts equally well melatonin suppression data from our laboratory, although obtained using different intensities and exposure duration. DISCUSSION: Together, our findings suggest that melatonin suppression by monochromatic lights is predominantly driven by melanopsin and that it can be initiated at extremely low melanopic lux levels in experimental conditions. This emphasizes the concern of the non-visual impacts of low light intensities in lighting design and light-emitting devices.