When to Start Using a SAD Lamp

Short answer: start in the morning, start daily, and start before the worst of the dark season sets in. Everything after that sentence is about what the trials did and did not test, because the calendar question you are really asking, "which week of autumn?", has never been answered by a randomized trial.

What the trials tell us about timing

The evidence for SAD lamps comes with a built-in timing instruction. In the Can-SAD trial, 96 patients with winter seasonal affective disorder received 30 minutes of 10,000-lux light each morning for 8 weeks, and light matched fluoxetine on clinical response (67 percent in both groups) with earlier onset of improvement at week one. In the Eastman trial, 96 patients were randomized to 1.5 hours per day of morning light starting around 6 AM, evening light starting around 9 PM, or a morning placebo, over 4 weeks; morning light produced the most complete or near-complete remissions, 61 percent versus 50 percent for evening light and 32 percent for placebo.

Notice what was tested: morning versus evening, and light versus dim light or a drug. No trial randomized "start in October" against "start in November." The timing evidence is about the time of day, not the time of year. That is an important gap, because the calendar is the question most shoppers actually have.

So a fair statement of the evidence is this: morning sessions are supported by head-to-head data, while the season-start date is a judgment call you make against your own history, not a protocol you can copy from a paper.

Morning light produced more full remissions than placebo or evening light in a randomized trial

Why earlier in the season beats waiting

Seasonal affective disorder follows the photoperiod: symptoms typically build as days shorten through autumn and peak in midwinter. Nobody needs a citation for the general shape of that curve, and more importantly, no study says you should wait for a full episode before starting a therapy with a slow onset.

The onset data is the key point. Eastman found it took at least 3 weeks of daily sessions before light separated from placebo on strict response criteria. The Can-SAD trial ran 8 weeks. In both, the treatment that worked was measured in weeks of daily use, not in rescue sessions once symptoms peaked. Starting while your symptoms are mild or even absent gives the protocol its full runway: if the benefit takes 3 weeks to show up, starting 3 weeks before you expect trouble is not superstition, it is arithmetic.

That is also why "I will start when I feel bad" is the one strategy the evidence argues against. A treatment with a multi-week onset cannot be started at the moment you most need it and still be on time. If your pattern in past winters was symptoms by late October, a sensible plan is to have the lamp positioned and the morning slot established weeks before that, not to order the lamp in late October.

None of this is a universal medical schedule. Past years are your best predictor, and they are still only a predictor. Some winters arrive early; some years the pattern skips you entirely. The evidence tells you the treatment needs weeks, and your own history tells you which weeks.

The morning slot matters more than the month

If the calendar question has no trial behind it, the clock question has two. Both trials above used morning sessions: the Can-SAD protocol ran light between 7:00 and 8:00 AM, and the Eastman trial found morning light, average start around 6 AM, outperformed both evening light and a morning placebo on remission counts. Morning light is not a habit recommendation; it is the condition under which the evidence was gathered.

Practical translation: the session goes within an hour of waking, before the day pulls you in other directions. For most people that means shortly after getting out of bed, with the lamp at the distance the manufacturer pairs with its lux figure. The Day-Light Classic Plus from Carex Health Brands is specified at 10,000 lux at 12 to 14 inches with a 16 by 13 inch screen, and is priced at 169.99 USD. At that distance it reproduces the exposure level the trials tested. Move the chair back and the delivered light falls off fast, which is exactly the positioning error that turns a tested protocol into an untested one.

Evening sessions are not forbidden by the evidence, they are just unsupported as an equal choice: evening light reached 50 percent response against 61 percent for morning light in the Eastman trial, with placebo at 32 percent. If your schedule forces an evening slot, you are still probably above placebo, but you should know you are no longer inside the best-tested version of the treatment.

Consistency is the treatment

Across both trials, the active ingredient that gets the least marketing attention is the daily repetition. Sessions ran every day for 4 to 8 weeks. Nothing in the evidence supports the common pattern of using the lamp hard on weekends, skipping weekdays, and expecting the weekly total to average out.

Think of the protocol the way the researchers did: a dose, in the pharmacological sense. 10,000 lux for 30 minutes, each morning, for weeks. Missed doses reduce exposure the same way skipped pills do. The Can-SAD light arm showed earlier improvement than fluoxetine at week one, but the separation between treatments was measured after 8 weeks of accumulated sessions. There is no shortcut variant of the tested protocol, so there is no shortcut version of its results.

If you travel or miss days, the honest response is to resume the daily routine rather than double a session. Doubling does not reproduce the protocol, and the trials give you no reason to believe the body treats it as equivalent.

When your sleep timing shifts

Life does not keep trial hours. Shift work, a new schedule, daylight saving changes, or a late-night phase can move your wake time, and the protocol keyed to "morning" has to move with it. The principle that survives is relative, not absolute: the session follows your wake time, not the clock on the wall.

A few practical cases:

• If you wake at 5 AM for a new shift, the session moves to shortly after 5 AM. The 7:00 to 8:00 window in the Can-SAD protocol reflected the patients' mornings, not a law of nature.
• After a daylight saving transition, keep the session anchored to wake time and you will have made the only adjustment the evidence supports.
• If your sleep schedule is drifting later through the winter, which happens when light exposure drops, holding a fixed morning session can act as the anchor that pulls it back. This is a practical use of the intervention, not a claim the trials tested.

One caution that does come from the safety literature: people with bipolar disorder can be pushed into mania or hypomania by bright light, and light-sensitizing medications change the risk picture. If your sleep timing is shifting because of a medication change, a mood episode, or a new diagnosis, that is a clinician conversation, not a lamp-adjustment conversation. Starting a SAD lamp in those circumstances without medical input trades an evidence-backed routine for an unexamined risk.

A start-of-season checklist

Distilled, without pretending any of this was randomized:

1. Pick a start date a few weeks before symptoms usually appear in your winters, because the treatment needs weeks to show its effect.
2. Put the lamp at the distance its lux rating assumes, 12 to 14 inches for the Carex unit above, not across the room.
3. Attach the session to waking: 30 minutes of 10,000 lux each morning, daily.
4. Run it every day through the dark season. Boring and daily was the protocol that produced the trial numbers.
5. Reassess if symptoms or sleep timing change materially, and bring a clinician in if medication, mood stability, or eye health are part of the picture.

The evidence table

Note the gap the table makes visible: the trials pin down the time of day and the weeks of use, but none randomized the calendar start date. The seasonal timing advice above is reasoned from the onset data, not copied from a protocol.

The bottom line

Start morning sessions a few weeks before your usual symptom window, keep them daily, and keep the distance and lux figures the trials assumed. The evidence supports the clock and the calendar shape; it does not hand you a date on the wall. Treat the first weeks as the loading phase the trials ran, not as a verdict on whether the lamp works for you.

For more on the condition itself, see our overview of blue light therapy for seasonal depression.