The Bryan Johnson thing collapses two things that aren’t the same, and the whole 8-10 hour number rides on the collapse. What moves when you eat your last meal at 11:30 vs 7pm isn’t primarily “sleep.” It’s autonomic tone. A meal kicks up sympathetic activity, your resting HR sits higher and your overnight HRV gets suppressed while you’re still digesting. The postprandial HRV suppression I’ve seen in the data mostly washes out a few hours after eating, call it 2-4 in most people, longer if the meal was big and fatty (gastric emptying drags the tail out). Now look at what a “sleep score” is actually built from. On every consumer device I’ve torn into, it’s some weighting of overnight RHR, HRV, and movement. It is NOT a polysomnogram. So when someone eats at 11:30 and his score climbs to 100, a big chunk of that is just the autonomic tail having fully decayed before the measurement window opens. RHR drops, HRV recovers, score goes up. That’s real, but it’s a readout of “nothing is digesting,” not necessarily “my slow-wave sleep got deeper.” You’re measuring something. I’m not sure it’s the thing he named it. The two effects also run on different clocks, which is the part nobody separates: - the thermic/digestive piece (core temp bump, autonomic arousal) decays in a handful of hours. This is the one that actually gates sleep onset, and you do not need 8-10 hours to clear it.
- the glucose/insulin excursion clears faster still.
- the circadian piece (meal timing as a zeitgeist for peripheral clocks) is real but slow and doesn’t show up as a same-night score swing. So “11:30am is optimal” is a guy who found the meal time that fully clears the autonomic tail before bed, then read his device’s autonomic proxy back as sleep quality and felt confirmed. The actual minimum is probably “stop eating long enough that your HRV has settled,” which for most people is nowhere near 8-10 hours. If you want to run this as a real n=1 instead of vibes: log last-meal clock time and meal size as their own variables, not just “ate dinner.” Then pull your HRV and RHR curve, not the composite score, because the composite is the confounded number. Watch where overnight HRV stops rising as you push the meal earlier. That plateau is your personal clearance point. Past it you’re buying nothing. (I keep last-meal time as a custom factor in CareClinic and the watch quick-log is the only reason I actually capture it before bed instead of guessing the next morning, which would defeat the whole point.) The honest version of his claim is “I found the eating window that stops my digestion from polluting my overnight autonomic readout.” Useful. Just a much smaller claim than 8-10 hours of fasting buys you better sleep.