Blue Light, S-Cones, and the Real Limits of Sleep Glasses

Blue light near 480 nm is the main wavelength that suppresses melatonin and shifts the human body clock. That makes it a sensible target for evening lighting and blue-blocking glasses. But it does not settle every question about a pair of glasses with a "blue-light" label.

Human photobiology suggests that cones—especially S-cones, which respond most strongly around 420–440 nm—may add to melatonin suppression at the beginning of an exposure. If so, a lens designed mainly around the melanopsin peak near 480 nm could leave some active short-wavelength light behind. The catch is substantial: a 2025 paper disputes that cones make a meaningful contribution at all. The practical advice is therefore less dramatic than the marketing: reduce bright evening light first, use amber or blue-depleted light if it helps you, and inspect a lens's transmission spectrum before assuming it blocks every relevant wavelength.

Melanopsin is the main circadian sensor

Specialized retinal cells called intrinsically photosensitive retinal ganglion cells, or ipRGCs, contain melanopsin. They send light information to brain regions that regulate melatonin and circadian timing. Their peak sensitivity sits close to 480 nm, which explains why blue-enriched light is unusually effective at disturbing an evening melatonin signal.

The largest human action-spectrum study available, St Hilaire et al. (2022), tested 99 participants with six monochromatic wavelengths from 420 to 620 nm. Participants received 6.5-hour exposures under pharmacological pupil dilation, across photon densities from 2.52×10¹¹ to 1.53×10¹⁴ photons/cm²/sec. Wavelengths around 460–480 nm required the lowest photon densities to reach half-maximal melatonin suppression. A single-opsin model placed peak sensitivity at roughly 481–482 nm.

That is strong support for melanopsin as the dominant target of evening light management. At the highest intensities in the study's ED90 analysis, ipRGC signaling alone accounted for the response, with no detectable cone contribution. A blue-filtering strategy is therefore aimed at a real biological target, not a wellness-store ghost story.

The spectrum changes with exposure time

The more interesting result is that spectral sensitivity did not stay fixed. For the 6.5-hour exposures, St Hilaire and colleagues found that a three-channel model fit the melatonin-suppression data best: about 81% melanopsin, 12% S-cone input, and 7% combined L- and M-cone input. Those percentages should not be treated as design specifications for consumer products. They came from estimates across only five analyzable wavelengths, and some wavelength groups were small: nine participants at 420 nm, 11 at 480 nm, eight at 507 nm, and four at 620 nm, where the model did not converge.

Still, the study's broader conclusion matters: several photoreceptor systems may shape the response early in an exposure, while melanopsin becomes increasingly dominant as the exposure continues. Shorter action-spectroscopy work discussed by the authors estimated a roughly 2:1 melanopsin-to-S-cone contribution after 30 minutes—about 67% versus 33%. By 90 minutes, melanopsin-only models became more adequate. No study has yet measured the same participants at enough intermediate time points to identify a precise 30-, 45-, 60-, or 90-minute crossover. The direction of the change is supported; the exact clock time is not.

For a 20-to-60-minute evening screen session, the S-cone question is therefore biologically plausible rather than settled. It becomes less relevant for a long, continuous exposure, where melanopsin appears to dominate.

Do ordinary blue-light glasses miss the S-cone window?

Possibly, but no human trial has directly tested that claim. Many products emphasize attenuation near 480 nm. S-cones peak at shorter wavelengths, around 420–440 nm. A lens that sharply reduces 480 nm light but transmits much of that shorter band could, in theory, leave part of a short-exposure melatonin-suppression signal intact.

That inference depends on the cone model being correct and on the actual spectrum of the lens. Product names, tint color, and a tiny blue-light badge do not answer either question. A useful specification shows transmission across wavelengths, including the 415–440 nm range as well as the melanopsin-sensitive region near 460–480 nm.

Amber glasses can still be a reasonable practical option. A 2021 systematic review by Hester et al. reported sleep-quality and mood benefits from evening blue-blocking glasses in clinical studies. However, the review did not characterize the lenses' transmission spectra or compare glasses that covered the putative S-cone window with glasses that did not. It supports a clinical effect; it does not prove which wavelengths caused it.

One 2025 study challenges the cone story directly

The uncertainty is not a footnote. Fazlali et al. (2025) found no strong evidence that cone photoreceptors contribute meaningfully to human melatonin suppression or the alerting response to light. If later work confirms that conclusion, melanopsin-focused filtering near 480 nm may be sufficient, and concern about an S-cone gap would have little practical value.

The disagreement cannot be resolved by averaging the studies into a comforting middle. The papers address related but distinct models and conditions, while action-spectrum experiments must infer receptor contributions from limited wavelengths, intensities, and exposure durations. The St Hilaire study also used dilated pupils, whereas normal pupils constrict under melanopic light and reduce retinal irradiance. In Schöllhorn et al. (2024), 72 young men viewed metameric screen light for 3.5 hours at 7–89 lux; the high-melanopic conditions produced smaller steady-state pupils. That feedback loop means fixed-pupil laboratory thresholds may overstate suppression under everyday viewing conditions.

Population averages also hide real variation. Najjar et al. (2024) reported different photoreceptor contributions to melatonin suppression in young versus older adults. A 2024 study of 323 adults linked variation in ipRGC-driven pupil responses with circadian timing and sleep measures across seasons and depression severity. Advice derived largely from young-adult laboratory samples should not pretend to fit every retina.

Melatonin suppression is not the whole outcome

Melatonin is a useful circadian biomarker, but suppressing it does not automatically produce worse sleep that night. In a roughly 60-lux, one-hour metameric screen study, Blume et al. (2022) found melanopsin-driven melatonin suppression while holding cone excitation constant, yet did not find automatic changes in sleepiness, vigilance, sensory processing, or sleep architecture. A lower melatonin level is not a guaranteed miserable night.

Clock resetting deserves separate attention. Blume et al. (2024) used calibrated blue-yellow changes that varied melanopsin stimulation while controlling L-, M-, and S-cone excitation, and found melanopsin-dependent circadian phase resetting. Melanopsin also helps explain why light timing can matter even when an acute melatonin effect tells only part of the story.

Brief light can count, too. In a 34-hour laboratory study, Joyce et al. (2022) tested flash intensity at a fixed two-millisecond duration in 28 participants and flash duration from 10 microseconds to 10 seconds at fixed 2,000 photopic lux in 31 participants. Phase shifts increased with flash intensity and showed little dependence on duration across the tested range. Melanopsin-containing cells continue signaling after a light stimulus ends, so total photon exposure is not the whole accounting system.

What to do tonight

• Get consistent morning bright light. A 2026 narrative review concluded that morning light may advance melatonin rhythms within the same day, although the evidence it reviewed used varied protocols.
• Reduce evening light intensity and duration before obsessing over spectral perfection. In a randomized crossover hospital study, Vethe et al. (2021) found that 12 healthy adults spending five days in blue-depleted evening lighting had higher melatonin and better sleep than during standard lighting.
• If you buy filtering glasses, favor products that publish full transmission data. Coverage through roughly 415–440 nm is a reasonable precaution, but the benefit over melanopsin-focused filtering remains untested and disputed.
• Keep nocturnal light interruptions brief and dim. A phone check at full brightness is a poor circadian habit even if its exact cost cannot be calculated from a flash study.

Blue light remains the primary spectral target because melanopsin remains the primary circadian photoreceptor. The proposed S-cone gap is worth checking when choosing a lens, especially for short evening exposures, but it is not established enough to outrank the basics: bright mornings, dimmer evenings, and a sleep schedule you can actually keep.