Night-Shift Sleep: Melatonin vs. Evening Bright Light

Melatonin and bright evening light seem like opposites. One tells the brain that night has arrived; the other suppresses the body's own melatonin signal and pushes night later. For night-shift workers, though, both can aim the circadian clock toward the same destination: staying alert overnight and sleeping after sunrise.

The practical question is not which tool wins in a laboratory showdown. No randomized trial has directly compared evening bright light with melatonin in shift workers. The better question is whether your schedule gives your body enough consecutive nights to reset. Slow or permanent night schedules can support a gradual light-driven phase delay. Fast rotations usually cannot. Melatonin may offer more immediate help with sleep timing, but the evidence in shift workers remains limited and does not establish an ideal dose, formulation, or long-term safety profile.

Why opposite signals can pursue the same goal

Your master circadian clock sits in the suprachiasmatic nucleus, a small brain structure that uses light to set daily timing. Light received in the evening, roughly within the phase-delay portion of the body's light-response curve, can shift that clock later. Specialized retinal cells containing melanopsin carry much of that signal. They are especially sensitive to short-wavelength light near 480 nm, although the useful light spectrum changes with exposure duration.

A 2022 systematic review by Cyr et al. examined five controlled real-world or simulated night-shift trials of evening light exposure between 16:00 and 22:00. Across those trials, evening light delayed markers such as dim-light melatonin onset (DLMO), core-body-temperature phase, or sleep-propensity rhythms. The studies also reported improvements in sleep quality, sleep duration, memory, or work performance. The review rated confidence as moderate for circadian physiology but very low for mood outcomes. It did not provide pooled phase-delay estimates in hours, individual effect sizes, or a single usable light-intensity threshold.

That last limitation matters in practice. "Use bright light" is not a prescription. A controlled laboratory study by St Hilaire et al. found that the spectral sensitivity of circadian resetting and melatonin suppression changes with the duration of exposure. A light setup optimized for a brief session may not be optimal for a much longer one. Lux alone also misses the biologically relevant spectrum reaching the eye.

Melatonin works by a different route. Taken at an appropriate time, it supplies an external darkness signal that can help shift sleep timing and circadian phase without waiting for several evenings of light exposure to accumulate. That makes it attractive when a worker has only a few nights before the roster changes again. It does not make the supplement a universal fix for working at 3 a.m.

What the melatonin evidence actually supports

A 2022 systematic review of melatonin in health-care workers with shift-work sleep disorder found some evidence of improved sleep outcomes, but the trials were scarce, underpowered, and low to very low quality. The review could not establish a dose-response relationship across the commonly used 1–10 mg range using objective circadian markers. It also identified long-term safety during chronic use in shift workers, including possible effects on endogenous melatonin production, as an unanswered question.

Research in delayed sleep-wake phase syndrome (DSPS) supports a narrower point: timing may matter more than simply taking more melatonin. In a randomized clinical trial of adults with ADHD and DSPS, van Andel et al. found that melatonin chronotherapy advanced DLMO and improved sleep outcomes from the initial administration. This was an advance protocol for people whose clocks ran late, not a night-shift trial seeking a delay, so it cannot supply a shift-worker dose or schedule. It does show why melatonin can have an acute effect while light-based resetting usually requires repeated, consistently timed exposure.

A 2024 preliminary study by Swanson et al. reached a related result in adults with DSPS: low-dose melatonin combined with evening dim light and scheduled time in bed advanced circadian phase whether DLMO had been measured precisely or estimated. Again, this was not a shift-worker study, and it used dim light rather than bright light. Still, it supports a practical principle: anchor melatonin timing to the intended sleep period instead of treating dose as the only adjustable dial.

Fast rotations change the decision

Light can shift a clock only if the schedule lets those shifts build night after night. A worker on permanent nights or a schedule with one to two stable weeks has a plausible opportunity to delay circadian timing progressively. A worker who flips after two or three nights does not have the same runway.

That conclusion fits a field trial by Martin et al. in fast-rotating shift workers. Timed dim blue light used in the car, combined with morning blue-blocking glasses, did not improve circadian adaptation. The finding does not prove that every brighter, differently timed light protocol will fail: the intervention was dim and delivered in the morning rather than as bright evening exposure. But it captures the operational problem. A rotating roster can reverse direction before the clock has moved far enough to matter.

For fast rotations, attempting full circadian adaptation may therefore be unrealistic. A consistently protected morning sleep window after a night shift can at least provide a behavioral anchor. Melatonin timed to the desired sleep onset may be a more practical short-term option than a multi-night light protocol, but the supporting shift-worker evidence does not justify precise dosing advice.

Why using both is not automatically better

The obvious instinct is to combine melatonin and bright light. Current evidence does not show that this creates an extra benefit. In a three-arm, double-blind, randomized placebo-controlled trial of 49 adults aged 18 to 55 with ADHD and DSPS, van Andel et al. reported that melatonin advanced DLMO by 1.5 hours, while melatonin plus bright-light therapy advanced it by 2 hours. Adding bright light did not produce a statistically significant additional phase or sleep benefit over melatonin alone, and it did not improve ADHD symptoms.

This is the only controlled combination comparison in the evidence discussed here, so it deserves attention without being stretched beyond its limits. The participants had ADHD and DSPS, not shift-work sleep disorder. Their treatment aimed to advance a delayed clock, whereas night-shift adaptation usually aims to delay it. The timing and intensity of bright light may also differ from an evening pre-shift protocol. The result does not show that a carefully sequenced shift-worker protocol cannot work. It shows that synergy remains unproven.

Concurrent use also raises a reasonable mechanistic concern. Bright light suppresses endogenous melatonin through the same light-sensitive system used to reset the clock. Researchers have not quantified whether that meaningfully weakens an exogenous melatonin protocol in shift workers. Until they do, treating bright light and melatonin as sequential tools—light for a planned phase delay, melatonin near the chosen sleep period—makes more physiological sense than assuming that simultaneous use doubles the effect.

Personalization is the missing piece

Identical light exposure does not produce identical phase delays. Hartstein et al. documented large individual differences in evening-light phase-delay responses and found melanopic equivalent daylight illuminance to be the strongest quantitative predictor of the stimulus. The study involved young children, not working adults, so its thresholds should not be imported into an employee lighting plan. Its lesson is broader: generic light prescriptions will vary in effect from one person to another.

Measuring DLMO directly is demanding, but wearables may eventually make timing more practical. In 45 fixed night-shift workers, Cheng et al. tested wrist actigraphy and photometry against laboratory DLMO measurements and found that mathematical models could predict circadian phase. This validates phase estimation as a field approach; it does not prove that wearable-guided treatment improves sleep or health outcomes.

The reason to care extends beyond feeling less wrecked after a run of nights. In a 40-hour constant-routine study, Grant et al. found endogenous circadian rhythms in cholesterol, HDL, LDL, albumin, total protein, VLDL, and triglycerides in 16 young men. Simulated shift work reset the cholesterol rhythm. This demonstrates circadian control of metabolic biomarkers; it does not show that a light box or melatonin supplement prevents metabolic disease in shift workers.

A realistic decision rule

• Fast-rotating schedules: Prioritize a consistent post-shift sleep opportunity. Consider melatonin as a short-term, timing-based tool with advice from a clinician or pharmacist, especially if you take other medicines or have a medical condition. Do not expect a few light sessions to fully reset a clock that your roster immediately flips back.
• Slow rotations or permanent nights: Evening bright light before work has a credible physiological rationale for gradually delaying circadian timing. Expect individual variation, and avoid pretending that one lux number fits everyone.
• Melatonin plus bright light: Current controlled evidence does not show an added benefit over melatonin alone. The combination has not been adequately tested in shift workers.
• Days off after nights: Morning light may help move the clock earlier again. A 2026 review by López-Velasco et al. concluded that morning light can produce measurable melatonin phase advances within a day, but this remains indirect evidence for shift-worker recovery plans.

Evening light has moderate support for shifting circadian markers in night-shift protocols. Melatonin may help sleep timing more quickly, but its shift-worker evidence is weaker, its optimal regimen remains uncertain, and chronic safety has not been settled.

The sensible goal is not a heroic attempt to outsmart biology every week. Match the tool to the roster, protect the sleep window you can protect, and treat exact light and melatonin recipes as provisional until shift-worker trials supply the missing numbers.