This one's a reader request. A subscriber wrote in about late-night snacking and the occasional drink, and honestly it's the perfect topic, because it hits every single one of us. We've done the big needle subjects: GLP-1s, peptides, TRT. This one's smaller and closer to home: what you do in the last three hours before bed. Two habits in particular, a drink and a late meal, quietly tax the thing all of it depends on, which is sleep. Here's what the data actually says, and where the popular advice is wrong in both directions.
Deep Dive: Does a nightcap actually help you sleep?
Here's the trap: it feels like it works. A drink before bed makes you fall asleep faster and pushes you into deep sleep earlier in the night. So your gut says "that helped." Your gut is reading the first half of the night and ignoring the second.
What's actually happening, measured in sleep labs and now on the wrist:
It front-loads sleep, then wrecks the back half. Alcohol shortens how long it takes to fall asleep and boosts slow-wave (deep) sleep early. But it suppresses REM in the first half of the night, and then you get a rebound: more wakefulness and fragmented, restless sleep in the back half as your body clears it. You wake up technically having "slept" but not recovered. 🟢 Strong evidence, this pattern is consistent across decades of sleep studies.
Even one or two drinks does it. This isn't a heavy-drinking problem. Moderate intake, the 1-2 drinks most of us consider harmless, still cuts REM by roughly 10-15 minutes and measurably degrades sleep quality. 🟡 Moderate evidence.
Your recovery metrics take the hit, and now you can see it. Alcohol raises your resting heart rate overnight and lowers heart rate variability (HRV), the single best wearable signal of nervous-system recovery. If you wear a Whoop, Oura, or Garmin, you've probably already seen it: a couple of drinks and your HRV craters, your resting heart rate climbs, your recovery score is in the red the next morning. That's not the app being dramatic. That's your autonomic nervous system working through the night instead of resting. 🟢 Strong evidence on the heart-rate and HRV effects.
Why this matters more at 40+. Sleep is the lever under everything we cover here: testosterone, recovery from training, next-day appetite and energy, even how well the other protocols work. Trading real sleep for a drink is a worse deal at 49 than it was at 25, because the recovery margin is thinner.
My read: the move isn't "never drink." It's "know exactly what it costs and decide on purpose." If you're going to drink, earlier and less beats later and more, and giving yourself three or four hours before bed lets your body clear most of it before you're trying to recover. The nightcap specifically, the drink closest to sleep, is the worst-timed one. Watch your own HRV for a week with and without, and you'll have your answer in your own numbers.
Not medical advice. If you're using alcohol to fall asleep regularly, that's worth a real conversation with your doctor.
Quick Hits
Late-night eating quietly works against you. 2025 research keeps landing on the same point: eating late, relative to your own body clock, is linked to lower insulin sensitivity and more fat storage, even at the same total calories. Shifting more of your food earlier in the day lines up with when your body handles glucose best. The late-night fridge raid isn't just extra calories, it's worse-timed ones. 🟡 Moderate evidence
Magnesium for sleep: real, but small. A 2025 randomized, placebo-controlled trial (155 poor sleepers) found magnesium bisglycinate improved insomnia scores more than placebo, but the effect was modest, and bigger in people who were low on magnesium to start. A reasonable, cheap thing to try, especially if your diet's thin on it, but it's a nudge, not a knockout. 🟡 Moderate evidence
If you want one lever for evenings, it's the eating window. Early time-restricted eating, front-loading calories and closing the kitchen earlier, improves glucose control and body composition in trials even without eating less. Same food, better timing. 🟡 Moderate evidence
My Experience
Late-night eating was my thing for years. Through my 20s and 30s the night didn't really start until the food showed up: instant ramen, greasy Chinese, those little slider burgers Harold and Kumar built a whole movie around. And I swear it tasted better at midnight. That part isn't just nostalgia, either. Late at night you're tired, your willpower's spent, and your brain's reward response to fat and salt runs hotter. The food genuinely did hit harder. I wasn't imagining it.
What changed wasn't the craving, it was the bill. Somewhere around 40 the next-day tax showed up: sluggish mornings, brain fog, indigestion that parked itself overnight. None of that touched me at 25. Which tracks with everything above, food eaten late lands on a body that handles glucose worse than it used to and fights its own clock. Young me got away with it. Forty-year-old me got an invoice.
Honest part: I still get the cravings. They never fully left. What worked wasn't willpower, it was changing the pattern around them. After dinner I walk instead of dropping onto the couch in front of the TV, which is exactly where the second dinner gets ordered, and that walk happens to be the single best thing going for the meal I just ate. No blue light for the hour before bed: no TV, no iPad, no laptop. Magnesium glycinate about an hour out. And I flipped the schedule, sleep early and wake early, instead of the up-late routine that basically scheduled the midnight ramen for me. Change the pattern and the craving has nowhere to land. The sliders still call. I just stopped picking up.
My protocol, my body, my doctors. Not a template.
What do you want me to dig into?
This issue exists because a reader asked for it. So I'll make it a standing invitation: hit reply and tell me what you want covered. A compound you're curious about, a claim you keep seeing, a number on your bloodwork you don't understand. I read every reply, and the good ones become issues.
Next issue (Tuesday): you sit all day. Even if you train hard a few times a week, the hours of stillness in between are quietly working against you, and the fix isn't another brutal workout. The desk-worker movement problem, what the research actually says undoes it, and the small stuff that fits into a workday. Another reader-requested one (keep them coming).
The 40+ Protocol is educational content, not medical advice. I'm a game designer reporting studies and my own n=1 experiments, not a doctor. Talk to yours before changing anything. No affiliate links in this issue.