It was 3 AM and I couldn’t sleep. Again.

I’d been lying in bed since midnight, staring at the ceiling, my Galaxy Watch glowing softly on the nightstand showing 3:14 AM in crisp green digits. My heart rate was elevated at 67 bpm—higher than it should be for someone supposedly resting. The Xiaomi scale downstairs had weighed me at 183 pounds this evening, but anxiety weighs nothing on a scale. It feels like gravity doubled.

I was 40 years old. I had a mortgage, a solid job, responsibilities. I should be able to sleep like a normal person. Instead, here I was, for the third night in a row, wide awake in the darkness.

That’s when I did something I’d been contemplating all week: I decided to test something crazy. Something that breaks every sleep hygiene rule ever written.

I got up and went downstairs to make coffee.

Not decaf. Not tea. Full-strength, dark roast, pour-every-ounce-of-caffeine-into-my-bloodstream coffee.

The Logic That Makes No Sense

Here’s where most people would argue with me. At 3 AM, when you can’t sleep, the solution is NOT coffee. The solution is melatonin. Or warm milk. Or meditation. Or literally anything except the thing that literally keeps you awake.

But I had a theory, and I was tired enough (literally) to test it.

I’d been reading about ultradian rhythms—the 90-minute cycles our bodies go through. Supposedly, if you’re already awake and can’t get back to sleep, sometimes you’re in a natural wake phase that won’t pass for another 45 minutes anyway. So what if, instead of lying in bed fighting it, you leaned into it? What if you did something engaging and stimulating, and rode out the peak of the cycle?

So at 3:17 AM (I checked the timestamp), I measured out a full 12-ounce mug of French roast coffee. Dark enough to barely see through. Strong enough to wake you up from a coma.

I sat at my kitchen island with my laptop, my Galaxy Watch tracking everything, and I did something productive while drinking it.

The 15 Minute Window

The first 15 minutes were normal. I drank, I worked, I felt the familiar warmth of coffee in my stomach. The caffeine hadn’t hit yet—that takes about 10-20 minutes. So far, no revelation.

But then something shifted.

I was on my spreadsheet, logging my sleep data as I always do (yes, I’m that person), when I noticed my heart rate had jumped to 82 bpm. My palms were getting warm. The caffeine was kicking in, and I was bracing for full wakefulness.

But instead of feeling more awake, I felt something else entirely.

I felt calm.

No, not just calm. I felt focused. Like the noise in my head—the anxious spiral that had kept me awake the previous two nights—had been volume-adjusted down to manageable levels. The caffeine, which should have amplified my anxiety, seemed to have short-circuited it somehow.

I worked for another 20 minutes. My productivity was genuinely good. I was making decisions. I was solving problems. My watch showed me in my optimal state—heart rate steady at 79 bpm, breathing slow and regulated.

Then, something impossible happened.

At 3:52 AM, I got tired.

Genuinely, authentically tired. Not the kind of tired you feel when you force yourself to go to bed. The kind of tired where your eyes start to feel heavy, your thoughts slow down, and your body practically demands sleep.

I went back upstairs. I was asleep by 4:07 AM.

I woke up at 8:31 AM. I slept for over four hours straight without waking up once. My Galaxy Watch showed a sleep score I hadn’t seen in months: 72 out of 100.

This Should Not Have Happened

I tested this theory for seven consecutive nights.

Night 1 (The Original): 3 AM coffee → productive focus → asleep by 4:07 AM → 4 hours 24 minutes of uninterrupted sleep

Night 2: Tried it again. Same result. Asleep by 3:58 AM, slept until 8:15 AM.

Night 3: Still working. Started to wonder if this was placebo or actual biology.

Night 4: Failed. I drank the coffee but drank it too quickly. Stayed awake until 5:30 AM, felt jittery, only got 2 hours and 47 minutes of sleep. The Xiaomi scale the next morning showed I’d lost two pounds of water weight.

Night 5: Tried again with slower coffee consumption. Back to working. Back to being asleep by 4:12 AM.

Nights 6-7: Successful both nights. The pattern was consistent.

I have no explanation for this. I was breaking every rule. I was consuming 200+ milligrams of caffeine in the middle of the night. According to everything I’ve read, according to every sleep doctor, according to basic science, this should have kept me awake.

Instead, it allowed me to sleep better than I had in years.

The Mechanism (Probably)

I did some research after realizing I’d stumbled onto something interesting.

The theory I landed on: Maybe for some people, nighttime insomnia isn’t about being overaroused. Maybe it’s about being stuck in a loop of rumination and anxiety. Maybe the anxious spiral creates a feedback loop where trying to sleep makes the anxiety worse, and the anxiety prevents sleep.

Caffeine, being a stimulant, forced my brain to engage with concrete tasks instead of abstract worries. For those 35-40 minutes of productivity, my brain couldn’t spiral into anxiety because it was busy working on real problems.

Then, when the caffeine wore off and my body’s natural sleepiness kicked in (around the time my ultradian cycle would have shifted back into sleep phase anyway), I was actually ready to sleep. The anxiety had been metabolized away through engagement instead of festering in bed.

Did the coffee directly cause sleep? Probably not.

Did the coffee reframe the 3 AM hour from something to struggle against into something productive and therefore less stressful? Definitely.

Why I’m Telling You This

I’m not recommending everyone drink coffee at 3 AM. That’s ridiculous. You’ll get different results. You might stay awake until 7 AM. You might feel sick. You might just waste a good cup of coffee.

But I am saying that sometimes the solution to a problem is counterintuitive. Sometimes fighting what’s happening makes it worse. Sometimes leaning into the wake cycle, making use of it, and treating it like a feature instead of a bug can actually reset the system.

My insomnia isn’t completely gone. I still have bad nights. But the nights I tested this theory, I slept like I hadn’t slept in five years.

At 40, with a Galaxy Watch tracking my sleep and a spreadsheet documenting my patterns, I learned that sometimes the crazy experiment is the one that works. Even if nobody else would ever think to try it.

Especially then.

My morning Xiaomi scale reading after the best night of sleep: 181 pounds. Turns out sleeping well affects hydration and weight distribution more than I expected. My Galaxy Watch heart rate that morning: 58 bpm. Calm, rested, and—against all conventional wisdom—caffeinated at 3 AM.

I’m never going back to conventional sleep advice.