Two months ago I hit a wall. I was at 216 pounds, down from 232, and for three straight weeks the scale didn’t move. My food logging was tight — I was hitting my protein most days, staying inside my calorie range, not sneaking bites off my kids’ plates. By every metric I usually watch, I was doing everything right. So I did something I never do: I stopped staring at the food data and looked at my sleep data instead.

My phone had been quietly tracking it the whole time. Average sleep over those three weeks: 5 hours and 50 minutes. I’d told myself for years that 6 hours was fine for me. Turns out 6 hours was not fine — it was the plateau.

The pattern was hiding in my own meal log

Once I started paying attention, I went back through three weeks of entries in ZeroToHero (zerotoherofit.com) — the app I built after getting sick of guessing what I was eating. The pattern was almost embarrassing in hindsight. On nights I slept under 6 hours, I logged an extra 300 to 500 calories the next evening, almost always between 9 and 11 pm, almost always something crunchy or sweet. Cereal standing over the sink. Half a sleeve of crackers while half-watching TV. None of it showed up as a conscious decision. It showed up in my head as “I’m just hungry tonight.”

That’s not a willpower story. It’s biology. Short sleep pushes ghrelin — the hormone that makes you hungry — up, and pulls leptin — the one that tells you you’re full — down. At the same time, the part of your brain that normally talks you out of the second bowl of cereal is running on less fuel itself. You’re hungrier and worse at saying no, at the exact same time. I didn’t need a study to believe that once I saw three weeks of my own timestamps lining up with it.

What I actually changed

I didn’t do a sleep overhaul. I made three small changes and gave them three weeks:

• Phone charges in the kitchen after 9:30 pm, not next to the bed. This alone killed most of the mindless late scrolling that was pushing my bedtime back 45 minutes some nights.
• I used the Meal Planner in ZeroToHero the night before instead of the morning of. Breakfast decided at 8 pm meant one less decision waiting for me at 6 am, which meant I wasn’t dragging on 5 hours of sleep and reaching for whatever was fastest.
• I moved my workout from 6:30 pm to 6:15 am. Small shift, but it meant I wasn’t wired from exercise at 9 pm, and getting up earlier made falling asleep earlier actually happen instead of just being a good intention.

None of that is groundbreaking advice. I’d read versions of it before and shrugged it off as wellness-blog filler. What made it stick this time was seeing my own late-night calorie spikes lined up against my own sleep numbers, in my own data, for three weeks straight. That’s a different kind of convincing than a headline.

The result

Average sleep went from 5:50 to just over 7 hours across the next three weeks. The late-night logging mostly disappeared on its own — I wasn’t white-knuckling it, I just wasn’t as hungry at 10 pm anymore. The scale broke the plateau and I dropped another 2.5 pounds, landing at what’s now 208. Nothing about my calorie target or my protein goal changed in that window. The only variable that moved was sleep.

I’m not saying sleep is the whole game — the first 24 pounds came almost entirely from fixing what and how much I ate, and I still track that closely through ZeroToHero (zerotoherofit.com) every day. But I’d been treating sleep like a nice-to-have, a wellness bonus round after the real work of nutrition was handled. My own three weeks of data said otherwise. It’s not separate from the diet. At 10 pm, tired, it basically is the diet.

If you’re doing everything right on paper and still stuck, it might be worth checking the number you’re not tracking instead of tightening the ones you already are.