For three weeks in October, I was on a story that took me through Toronto → Reykjavik → London → Lisbon → Berlin → Dubai → Singapore → Tokyo → home. Eight time zones in twenty-two days. The kind of itinerary that makes your circadian clock put up a white flag and your editor start to worry.

By the end, I had tried everything — the no-caffeine advice, the no-alcohol advice, the chug-water-on-the-plane advice, the don't-sleep-on-the-plane advice, the supplement-that-your-friend-sold-you advice. Some of it worked. Most of it didn't. Here is what actually did.

What didn't work

Melatonin. The 3mg doses that are sold over the counter in the US did almost nothing for me. The 10mg doses sold at the airport pharmacy in Dubai put me in a state that I can only describe as "a thick, warm fog, but with dreams." I woke up groggy and worse than before. Most of the research is on 0.5mg, which is roughly what's available in the EU but not in the US. If you try melatonin, try the small dose.

No alcohol on the plane. I did this, dutifully, for the first four flights. Then I had a glass of wine on the way to Tokyo, slept the best I slept all month, and gave up the experiment. Moderate alcohol on a long flight is fine. Heavy drinking on a long flight will wreck you.

Forcing myself to stay awake until local bedtime. This is the standard advice, and it is correct in theory and miserable in practice. On a 14-hour flight from New York to Tokyo, the trick is to stay awake through your home morning, sleep through the middle, and stay awake through the destination evening. In practice, my body did not consult this advice before falling asleep at hour four.

What worked

Light, on the schedule of the destination. This is the boring one that always works. As soon as you arrive, get outside. If it's morning, get sun on your face. If it's evening, get whatever ambient light is there. If it's the middle of the night, do not get sun on your face. The light is the signal. The light is what re-anchors your circadian clock to the new phase. No app, no supplement, no $400 sleep tracker does this better than a real sky.

Short exercise, hard. Twenty minutes of real exertion — a run, a hard walk, a swim — within an hour of landing, in daylight. Not later. Not in the hotel gym under fluorescent lights. Outdoors, in real light, while the destination's morning light is still strong.

Consistent wake time on the destination, not on the home. The advice I'd always been given was to stick to my home wake time, even when travelling. This is, in retrospect, the worst advice I have ever received. If you are in Tokyo and you wake up at 6am Toronto time, you have woken up at 7pm Tokyo time, and you will be unable to fall asleep until 5am the next morning, and you will be useless for two days. Wake up on the destination's clock. Let the light do the rest.

The sunrise trick

The single most useful thing I packed was a small sunrise alarm clock. Specifically, the Dreamegg Sunrise Beam, which is small enough to fit in a hotel-safety-box and runs on a USB-C cable that I was already carrying.

Here is the trick. When I arrived in a new time zone, the first morning I set the alarm for whatever time I wanted to wake up — destination local time. The Dreamegg then did the sunrise simulation for 30 or 60 minutes before that, even if the actual sun in the new city wasn't going to rise for another two hours. The brain received the signal "morning is here" and started the wake-up sequence on local time, not on home time. By the second morning in a new city, I was waking up naturally. By the third, I was on local time.

It is not a magic device. It is a 30-minute, warm, dawn-simulation light. It does what real light does, in places where the real light isn't yet cooperating.

The takeaway

If you travel a lot — even one time zone a month — a small sunrise alarm is the single best $83.00 you will spend on travel gear. Better than melatonin, better than compression socks, better than the special neck pillow. The boring light-based stuff works. The fancy stuff usually doesn't.