Abstract
Previous studies have reported diurnal variations in color perception and luminance contrast sensitivity, suggesting modulation of visual processing by the circadian pacemaker. To clarify potential time-of-day effects, we examined whether cone-mediated temporal contrast sensitivity is modulated by circadian timing. We disentangled circadian and homeostatic influences on human visual performance using a 40-hour forced desynchrony protocol. Twelve participants (five female participants, 24.4 ± 2.9 years, range 19-30) followed a repeating 3.75-hour cycle comprising 2 hours 30 minutes of wakefulness and 1 hour 15 minutes of sleep under constant environmental conditions. Circadian period was estimated from up to 46 salivary melatonin samples per participant. During wake periods, participants repeatedly completed temporal contrast sensitivity tasks at 2 Hz and 8 Hz, using tasks that probed both luminance (L+M+S) and chromatic (L-M, S-cone-directed) modulations. We tested for circadian modulation of visual performance by comparing hierarchical cosinor models with intercept-only null models using Bayes factors. All participants showed robust melatonin rhythms (mean τ = 24.18 ± 0.39 hours). Bayesian inference provided decisive evidence against circadian variation in luminance or chromatic contrast sensitivity. Within this controlled protocol, we found no detectable near-24-hour modulation of cone-mediated temporal contrast sensitivity. Previously reported time-of-day effects may instead reflect homeostatic influences, pupil-linked changes in retinal illuminance, or mechanisms outside the cone-mediated pathways probed here. By separating circadian phase from sleep-wake-related influences, this study clarifies when circadian modulation of visual performance can and cannot be inferred.