Bright Light Therapy: A PMHNP Prescribing Protocol

In an eight-week randomized trial published in JAMA Psychiatry in 2016, 20 mg of fluoxetine did not beat placebo for nonseasonal major depression. A light box did.

Raymond Lam and colleagues randomized 122 adults with at least moderate depression to four groups. Remission came in at 19.4% for fluoxetine alone and 30% for placebo. Thirty minutes of 10,000-lux light each morning reached 43.8%, and light plus fluoxetine reached 58.6%.

One small trial settles nothing on its own. A 2025 meta-analysis of 11 trials points the same way, though, and Carlat Publishing made light therapy the theme of the September 2026 issue of The Carlat Psychiatry Report.

Most PMHNPs learned light therapy as a footnote about winter blues, and it rarely makes it into a treatment plan with a dose attached. September is the month that costs the most. Patients with a seasonal pattern are about to hit their first low-energy weeks, and plenty of patients with bipolar depression can’t tolerate another antidepressant trial.

What follows is a working protocol covering the evidence, the dose, the bipolar adjustments, the screening, and a way of writing it that patients will actually follow.

The Evidence Is Stronger Than Its Reputation

Light therapy is a first-line treatment for seasonal affective disorder, and the evidence for other depressions is better than most prescribers assume. A 2005 meta-analysis by Golden and colleagues in the American Journal of Psychiatry, summarized in Carlat’s September 2026 review, put the effect size at 0.84 for seasonal depression. For nonseasonal depression it was 0.53, which Carlat describes as comparable to antidepressants.

The newer data are cleaner. Menegaz de Almeida and colleagues pooled 11 randomized trials with 858 patients in JAMA Psychiatry (2025). Remission was 40.7% with bright light versus 23.5% with control (odds ratio 2.42). Response was 60.4% versus 38.6% (odds ratio 2.34).

Most of those trials used 10,000 lux for 30 to 60 minutes a day, and the authors framed light as an add-on to usual treatment. The early numbers are the ones worth remembering. In trials shorter than four weeks, response was 55.6% with light versus 27.4% with control.

That speed is clinically useful. A patient who has already sat through two slow SSRI trials is often more willing to try something that can show a signal in two or three weeks.

Carlat grades the evidence by count. Six or more randomized trials support light for seasonal depression, nonseasonal depression, and delayed sleep phase. Bipolar depression, perinatal depression, and agitation and sleep problems in dementia have one to five trials each, which makes them reasonable but thinner ground.

For patients who have failed several medication trials, light belongs on the same list as other augmentation options, such as pramipexole for treatment-resistant depression. It costs one device and adds no drug interactions.

The Dose: Lux, Minutes, Distance, and Clock Time

Light therapy has a dose, and the dose has four parts. Get one of them wrong and the patient ends up with a weaker treatment than the one the trials tested.

The device matters as much as the instructions. Carlat recommends a UV-filtered unit rated at 10,000 lux with a screen at least 12 by 17 inches. It prefers LED units, because fluorescent tubes lose output as they age.

Light visors are out, because they show no dose-dependent benefit. Red-light devices don’t have the evidence their marketing implies. Insurance rarely pays for a light box, so it helps to suggest patients check whether their HSA or FSA will cover one.

Timing can flex. Carlat notes that patients whose mood drops in the late afternoon may do better with a session between 4 and 6 pm. Adherence goes up when the session is attached to something the patient already does every morning, like the first cup of coffee or a morning email check.

“Patients hear ‘light box’ and assume it’s a wellness tip. I write it the same way I’d write sertraline: a dose, a time, a duration, and a date when we check whether it’s working. Framed that way, people actually do it.”

— Lindsay Hill, DNP, PMHNP-BC

Bipolar Depression: Move It to Midday and Go Slow

Light is activating. That’s part of why it works, and it’s why bipolar depression needs its own protocol. Many of the drug options in the bipolar depression treatment algorithm carry metabolic or sedation costs, so a treatment with few systemic side effects is worth knowing how to use safely.

The best trial here is Sit and colleagues in the American Journal of Psychiatry (2018). They randomized 46 adults with bipolar depression, all already on a mood stabilizer, to 7,000-lux bright white light or 50-lux dim placebo light, delivered at midday for six weeks. Sessions began at 15 minutes and stepped up toward an hour.

At weeks 4 through 6, 68.2% of the bright-light group had remitted, compared with 22.2% of the placebo group. Nobody switched into mania or hypomania.

Carlat’s practical version is conservative. Start at 15 minutes a day, add 15 minutes each week, and schedule sessions between noon and 2 pm. The slower build gives time to catch early activation before it turns into a switch.

Bring the family in on that last instruction. Relatives often notice early activation before the patient does.

Screen the Eyes, the Skin, and the Med List First

Light therapy is low-risk, but it isn’t risk-free. A two-minute screen at the visit where it’s recommended catches nearly everything that should change the plan.

Start with the eyes. Carlat advises ophthalmologist supervision for patients with glaucoma, cataracts, or retinopathy, and every device should filter UV. That covers a lot of older adults, whose other prescribing considerations appear in the guide to geriatric psychopharmacology. Diabetic patients deserve a direct question about retinal exams.

Then the skin and the med list. Carlat flags systemic lupus erythematosus and photosensitizing medications, and it names lamotrigine specifically. Plenty of bipolar patients take it, and it’s already a drug where any skin change gets attention, as the lamotrigine rash titration protocol lays out. Photosensitivity doesn’t automatically rule light out. It does call for a slower start, a skin check at follow-up, and a conversation before the first session.

For the medication side of that review, the free PMHNP prescribing cheat sheet keeps dosing and monitoring details for common psychotropics in one place.

The side effects Carlat lists are headache, eyestrain, and nausea. They usually settle with a shorter session or a little more distance from the screen. Insomnia or jitteriness points to timing, so the session should move earlier in the day.

Finally, screen for mania history in every patient, including those who carry a unipolar diagnosis. A patient whose first hypomanic episode shows up on a light box is still a patient with bipolar disorder, and the plan changes right away.

How to Write the Light Prescription

A verbal suggestion to “get a SAD lamp” almost guarantees the wrong device, used at the wrong time, for the wrong length. A written plan fixes most of that. Something like this belongs in the note and in the patient’s after-visit summary:

For bipolar depression, the same template changes three lines: sessions at midday, 15 minutes to start, and 15-minute weekly increases.

The two-week follow-up gives the visit a job. Compare the PHQ-9 with baseline, look at the sleep log for later bedtimes or shorter nights, and ask how many sessions actually happened. A patient who managed three sessions a week hasn’t had a trial yet.

Response timing differs by diagnosis. Carlat notes that seasonal depression can respond within days, and a clear lift after the first hour can predict how the rest of the week will go. Nonseasonal depression builds over weeks, closer to an antidepressant.

For patients with a known seasonal pattern, timing the start is most of the work. Carlat advises beginning at the first prodromal signs, such as lower energy, less enthusiasm, or a bigger appetite, and not waiting for a full episode. Plan to continue through the darker months and review the plan in spring.

What Most People Get Wrong

The common assumption is that light therapy is a gentle seasonal extra, something to mention for mild winter blues after the real treatment is in place. Almost every part of that sentence runs against the data.

Seasonal? The nonseasonal effect size in Golden’s meta-analysis was 0.53. Mild? Lam’s trial required at least moderate depression, and Carlat says severity doesn’t disqualify a patient. An extra? In that same trial, light alone produced more than twice the remission rate of fluoxetine alone, 43.8% against 19.4%.

The second mistake is reading a failed trial as proof that light doesn’t work for that patient. When someone says the light box did nothing, the first questions are about dose. How bright is it at the distance they actually sit? How long are the sessions, and at what time of day?

Light intensity drops quickly with distance. A lamp that reaches 10,000 lux only a few inches from the face delivers far less at arm’s length, and a ten-minute session at 9 pm is a different treatment altogether.

“Before I call light therapy a failure, I want to see the device and hear the routine. A small lamp used at night is not the treatment the trials tested.”

— Lindsay Hill, DNP, PMHNP-BC

The third mistake runs the other way: assuming light is harmless enough to skip screening. It is an activating treatment with real ocular and bipolar precautions. That’s why it deserves the same structure as any prescription.

Frequently Asked Questions

A Light Box Is a Prescription

Bright light therapy has a dose, a timeline, contraindications, and side effects. That’s enough to earn it a real line in the treatment plan. For unipolar and seasonal depression, the starting point is 10,000 lux for 20 to 30 minutes soon after waking, raised to 60 minutes if needed. In bipolar depression, move it to midday, start at 15 minutes, and titrate weekly with a mood stabilizer already on board.

The concrete next step takes one afternoon. Before September ends, pull every patient with a documented seasonal pattern and offer a written light prescription ahead of their usual onset. Starting in early fall costs little, and waiting until December means treating an episode that may have been preventable.

Prescribers who want this kind of protocol thinking built into their first years of practice work through it with mentors inside the Psych NP Fellowship.

This content is for educational purposes and does not replace individualized clinical judgment or supervision.