(updated 3/2019)

(Don’t do this on your own. There are risks, described below. But these risks may be smaller than those of medication approaches. )

Bottom line: light therapy for bipolar disorders is a useful tool. It’s clearly more than a placebo. Dawn simulators are the first step (much easier to use, cheaper, no risk). Unless depression is severe, where you might go straight to using a light box. Read the safety section and watch out for worsening of mood.

Light therapy for bipolar depression

Basic principles:

  1. Be careful: light boxes carry the same risks as any antidepressant for making mood worse, inducing manic and mixed states.
  2. Fortunately, it’s easy to stop! And easy to adjust the dose very carefully.
  3. Dosing is adjusted both by length of exposure and time of day of exposure.

But hold on a minute, before you go tearing off looking at light boxes (discussed below) . . . There is a far simpler way to do light therapy, which is even cheaper and so far has not been associated with hypomania or mania (which light boxes can cause).   It doesn’t work for everyone. But “treatment” is over by the time you wake up. No sitting for 30 minutes stuck in front of a light box.   If it doesn’t work, you might still like it (wouldn’t you rather wake up in room full of light than wake to an alarm and have to — whamo — turn on a bright light? ugh).

Start with a dawn simulator

What is a “Dawn Simulator”? This is simply a device to gradually increase the light in your bedroom in the morning, while you are still asleep. Try this: close your eyes and look toward the light by which you’re reading this. You can tell where the light is, even with your eyes closed. A dawn simulator gradually turns on your bedside lamp in the morning, before you wake up, so that your retina (not you — you’re still asleep!) “sees” the light show up at the time you choose, increasing gradually just as natural sunlight does, over about 30-45 minutes. It’s really nothing more than a timer and a rheostat (a device to slowly change electrical current) hooked to your bedside lamp. Note that this approach does not require a “light box”.

Why is the gradual appearance of morning light potentially “antidepressant”? Think of it this way: the dawn simulator is trying to convince your brain that it’s still July out there (even in December). It turns out that your brain knows what season it is primarily by the time at which morning light appears (okay, it’s true, you can also tell the difference between snow, sleet, ice, freezing rain, and the warm summer sun; but evolutionarily, it looks like the brain’s timing was set by factors more closely associated with light.) We think that some people are built to slow down in the winter, something akin to hibernation. Think of a hibernating bear: sleepy, slow, hungry for carbohydrates, unhappy if awakened, grouchy, grrrh. Those symptoms are pretty close to the experience of “winter blues” for some people.

So, if those peoples’ brain could be tricked into thinking that it really isn’t winter after all, might that prevent this shift toward a kind of half-hibernating way of dealing with the world? It appears that for some people, this actually works.

In the following graph, the “negative ion generator” was low-dose, the placebo. (High-dose appears to be an antidepressant of sorts Flory).  Notice how the dawn simulator, while not as good as light box, is better than a placebo.Terman (down is getting better, these are depression scores over time from left to right)

Comparison of treatments for for seasonal depression
Dawn simulator trial

Update 2015: dawn simulators appear to improve thinking ability in the morning, compared to waking and staying in dim light. Gabel2015, Gabel2013

So, to review:

  • it’s cheaper
  • it’s over by the time you wake up
  • you might like it even if it doesn’t work for mood
  • it doesn’t make bipolar disorder worse
  • it has no risks (except waking your partner)
  • and it might even make you think better!

Sold?  Here’s a page listing different versions by cost, and where to get them.  Now, on to light boxes — another way to do light therapy.

Light therapy for bipolar is different

If you already have figured out you have some degree of bipolarity, then you’ll need to be more careful with light therapy than someone using it for winter blues (Seasonal Affective Disorder, SAD).  But light therapy is probably safer to tinker with than antidepressants. I say tinker because this is not well-researched territory. You’ll have to learn what works for your bipolar depression by trying things and carefully observing the results.

A 2018 research trial put light therapy solidly on the map for bipolar depression, even though not to the point where light is replacing antidepressants (yet). That group of researchers had previously shown that morning light therapy could induce mixed states.  But when the light was used instead at midday, the same patients often got good outcomes. Their depression improved and they did not have mixed state symptoms.

Based on the work of that research group, many of us have concluded that light therapy for bipolar depression should start at midday, not in the morning.  Others disagree. Agreement is also lacking on how long one should sit in front of the box. Routine in SAD is 30 minutes. To be cautious with the initial dose, some of us think that it would be smarter to begin with just 15 minutes of exposure.

Unfortunately, the dose you get depends on the box you use, how close it is to your face, and it’s position above or below your eyes. Above is better; more on that below.  Smaller boxes may be able to deliver the same dose if you sit in front of them longer, though in my view even this is not certain.

How long, how fast?

A dawn simulator alone didn’t work? Sorry.  You could continue that while using a light box, hoping that when you’re better, the dawn simulator will help you stay well, without the light box.  Of course, don’t do this on your own: discuss the idea and how to do it with your psychiatrist or NP or primary care providers.  One way to consider:  start with 15 minutes at midday. If that works, great. If not, then after 4 days, increase to 30 minutes (which some think is where you should have started, but that’s less cautious). If after 4 days nothing at all has changed, move the timing one hour earlier in the day. Keep moving it by one hour until you’re using the light box as soon as you wake up.

All along, watch out for emerging mixed state symptoms. In my view (it’s complicated), these include agitation, irritability, increased and rapid thinking, difficulty concentrating, and insomnia. If any of those are starting to show up, stop using the light box entirely and wait until those symptoms have disappeared. Yes, you’ll be back to a pure bipolar depression, sorry.  Let your providers know what’s happened and re-group. In theory maybe you could back up and try lower “doses” again, progressing more slowly toward the higher doses (longer, earlier). Or maybe it’s time to try a different strategy.

Light Boxes: Safety issues

A group of  experts maintains a website (CET.org) about all aspects of light therapy. Their discussion of safety issues is thorough; and maybe a little technical, so I’ll summarize here, but if you have any further need for information, that’s the place to go.

You need to learn more about light box safety before using one if (look up your medications or ask your pharmacist; ask your eye doc’;  or at least read the CET.org site’s page about risk) :

– you have macular degeneration

– family history of macular degeneration

– you have porphyria, lupus, actinic dermatitis (seriously sun-damaged skin), or solar urticaria (sun makes your skin hurt)

– you take a medication that sensitizes you to light

– “phenothiazines” (a kind of antipsychotic; check yours out if you take one)
–  some cardiac anti-arrhythmia med’s like amiodarone
–  treatments for psoriasis
–  antimalarials
–  St. John’s Wort
–  medications for rheumatoid diseases
–  and, lower risk but still worth checking out:
tetracycline, diuretics like hydrochlorthiazide, sulfa drugs, tricyclic antidepressants

And if you have diabetes and eye problems from that — well, you get the drift. This is high intensity light. You have to be careful with it.

Which Light Box Should I Buy?

(Updated 3/2019) Whatever you buy, here are two research-based essentials:

  1. The larger the better, and 10,000 lux is good.
  2. It should sit higher than your eyeball and shine down, not up.

Light therapy as studied in research labs is extremely bright, to deliver a lot of photons in a short period of time. If you lose photons, you’ll need to sit in front of the box longer.  A larger box means you’ll lose fewer photons when you shift your gaze. A brighter box delivers more photons per minute. Smaller, dimmer boxes have not been tested as much.  Many people use a small one and leave it on their work desk for hours, but that’s not been tested either.

Above your eye? Right. The receptors that recognize light as a signal of daytime are on the bottom of your retina. To reach them efficiently through your pinhole pupil, the light source needs to be up pretty high. A light box sitting on your desk facing upward is inefficient (and it too has not been studied). More on that below (see Above my eyes?).

A. Like the research rig: Carex $115

A good light box sits above your eyes
Carex light box

Notice that the light box which comes with this stand is higher than her eyeball, and it’s LARGE, so if she moves her head a little the amount of light she sees will not change too much.  Surfing around Amazon yielded some old-style Carex boxes (leggy-lookin’, but solid) that would work fine, less expensive.


B. The cheap route
At $45 this is less than half the price of the Carex box but you’ll likely have to sit in front of it longer, especially if you can’t figure out a way to get it above your eyeballs.

Verilux HappyLight Lucent 10,000 Lux LED Bright White Light Therapy LampSmall boxes may say they put out 10,000 lux but for most of them they’d have to be about 6″ from your eye to obtain that intensity, whereas the Carex box above can yield that result at 12 inches. In either case, as you can see, they need to be close. You can also see that a small shift in the position of your head will reduce the light you get from this little one. If you must go this route, be careful to put something really interesting in front of you while using this treatment, so you won’t move around too much. The box does not need to be straight in front of you. The ideal position is shown in the Carex photo above.  It could also be in your field of vision but off to the side and up a little (e.g. on a pile of books very close to your head). You can address all these worries by spending another $70 on the big box, if you’ve got the cash.

What happened to the little blue light boxes?

As you may have read, blue light is wavelength that sets biological rhythms, including sleep/wake cycles in humans (here’s that story).  So back around 2005, some companies started making little blue light boxes. They could be smaller because they were emitting just the “active ingredient” in white light. But now:

a) they’re not the cheapest anymore

b) there has controversy as to whether the little blue boxes are as effective as large white ones, although a new 2016 study suggests they are.Meesters

c) yet still, the big white boxes are much easier to use correctly (they’re set above your eyes, so no need for piles of books or some other placement; and they have a big field so you don’t have to look in one place for 30 minutes)

If you want to be really sure you have the same box that most of the research studies have used, you need the LARGE WHITE one above.

What about a visor?


If you have $200 to drop, not just $50, you might consider a green light visor, but only if you have to be walking around at the time you need the light therapy. If you can be sitting down at the needed time, the expense is not worth it, and the data supporting it not substantial enough (yet, anyway; as of 2013). You can’t easily read or use a computer, though it’s not impossible (my experience; versus an aviation study that panned the visor approach). You definitely cannot drive with this rig on.

Does the visor approach really work? The manufacturer offers four studies showing effectiveness versus control treatments. Unfortunately, none of the these has been published, near as I can tell ( Pub Med, 11/2014). That should be a red flag. Yet green light has been shown effective in other published work, using a light tower — much more expensive, not portable — instead of a light visor. So even though several large studies with white-light visors did not show benefit versus control treatment (Teicher; Rosenthal), we now know the active ingredient in all that light (blue or blue green; here is the whole story about blue light ). So theoretically the green visor might work, even if it does not have published evidence. Therefore you might consider it if you can’t sit for light therapy (e.g. busy parent in the morning?). You look kinda funny with this visor on your head, though: I got quite a few laughs wearing it around the hospital. So this may be a privacy-of-your-own-home thing.

Extra material for the very interested

How much light are you getting?

Light therapy can treat more than just winter depression, though that’s the main use. Here’s why: the amount of light reaching your eyeball from interior lighting is far less than the amount from the real thing. So unless you are outside much of the day in the winter, you are relying on electric light for your photons (in summer, there is so much light, most people get enough, even if they are indoors during their work hours). The following graph shows you just how much less light you receive, indoors versus outdoors (Lux is a standard unit of light flow):

Brightness Values:

Bright moonlight 1 lux
Candle light at 20 cm 10-15 lux
Street light 10-20 lux
Normal living room lighting 100 lux
Office fluorescent light 300-500 lux
Sunlight, 1 hour before sunset 1000 lux
Daylight, cloudy sky* 5000 lux
Daylight, clear sky 10,000-20.000 lux
Bright sunlight > 20,000-100,000 lux

Outdoors on a sunny day is 50-100 times brighter than office or room lighting (depends on latitude and weather, of course: a northern winter day provides less light than a southern or summer day)

Above my eyes?

Yes. The same team that developed the first blue light box also did an interesting experiment. They put a special helmet on patients undergoing light therapy, which allowed the light to hit only the top of their eyeballs, or the bottom. Why they thought this might be important, I don’t know — but they were right. It turns out you need use a light box positioned above your eyes, so that the light hits the bottom of your retina (like the top of the chicken below). Patients in the group with light hitting only the top of their retina (think chicken feet, which is what you’re doing with the light box down on your desktop) did not respond as well to the light therapy.Glickman


Think about it, this actually makes a lot of sense. Why would your body bother putting receptors for light at the top of your eyeball? The light you’re interested in is coming from the sky, not the ground, right? Evolutionarily they’d be much more useful at the bottom of the eye; why waste them at the top where they can’t “see” the light they’re supposed to be telling the brain about?

The point here: don’t put that light box on the tabletop next to your bowl of cereal and your newspaper. Put it up on something so that the light is coming down toward you (not very far away though: one foot for the research rig, and 6 inches for the little white box). I’ve been telling my patients for years to put their big suitcase-size light boxes on the table. Wrong.

Chronotherapy: timed sleep, light and darkness

Chronotherapy puts light therapy together with sleep therapies to create a complete package treatment for depression that can replace or add to medication approaches. The
ingredients include:

  1. “wake therapy”
  2. “sleep phase advance”
  3. a light box
  4. a dawn simulator

You’ve already learned about light boxes and dawn simulators above. For more on the first two ingredients, see my page on Chronotherapy.

Safety of little blue light boxes

This issue of eye safety cranked up recently with the arrival of the little blue light box. (Why blue?) Back in 1992, a team of light researchers estimated that it would take 72 winters of daily 30-minute light therapy to reach the threshold for causing eye damage.Waxler But because the blue box puts out a wavelength that is theoretically more harmful than other wavelengths, there’s been a lot of interest in the safety issue lately. Ironically, the graphs below suggests that if anything, the blue box is safer than “full spectrum” or all-white units.

What colors of light are produced by light boxes?
Spectral power distribution of light boxes and sunlight

This graph shows the distribution of light strength at different wavelengths for three light boxes (blue, red and green curves) and sunlight (black curve). From left to right on this graph are the colors of the visual spectrum of light.  Ultraviolet would be off this graph to the left; infrared off this graph to the right.  Blue is between 450 and 500 nm (nanometers).

The blue hump is the output from one of the little blue light boxes.  The red and green curves are two different “full spectrum” light boxes.  Note the big spike from the full-spectrum boxes at about 440 nm.

As you can see, all the light boxes put out a lot less energy than summer sunlight.  (In winter, if you measured outdoor light in the northern part of the United States or Europe, that black curve would be much lower on the graph — in some places near or even below the output of the light boxes).

Notice that the blue box output, the area shaded in blue, has a relatively low peak, especially compared to sunlight. Second, note that the full-spectrum light box has two peaks to the left of the blue hump. If there is any danger involved, then these blue peaks from the full spectrum boxes peaks are more dangerous to the retina for two reasons: first, these peaks are at a shorter wavelength, meaning even higher energy; and secondly, they are larger (taller), meaning more energy as well. It’s a good thing that for the vast majority of people who might consider using a light box, the safety issues described above don’t apply.