I Fixed My Blood Sugar. Then I Woke at 3am Every Night for 6 Months.

I Fixed My Blood Sugar. Then I Woke at 3am Every Night for 6 Months.

[OPENING HOOK — 0-15s] My blood sugar was back to normal. My GP was happy. And I was still waking up at 3am every single night.

[OPENING BRIDGE — 16-30s] Every blood test result was moving in the right direction — HbA1c normalising, liver enzymes dropping, inflammation coming down. The numbers said I was recovering. And then I’d lie there at three in the morning, wide awake, wondering why my body hadn’t got the memo. I’m Neil. I’ve got NHS blood test records going back to 2012 and I spent the last two tracking everything — sleep quality, energy, biomarkers, the lot. What I found was a pattern I didn’t expect: metabolic damage heals in a specific order. And sleep is always last. In the next ten minutes I’ll show you why, and what finally changed it for me. For anyone whose numbers have improved but who still doesn’t feel right — this mechanism is worth understanding.

[OPENING ROADMAP — 31-60s] First: why your sleep is still broken even when your blood tests say you’re fine — the reason is in the repair sequence, not your discipline. Second: the six-month lag I tracked in my own data — what was happening biologically while I thought I was just still broken. And then the moment it changed — a Wednesday morning in May 2024, no alarm, pitch black room, just… awake. And calm. We start with the night I realised the numbers and the sleep were on completely different clocks.

THE PUZZLE

I used to be a good sleeper. Genuinely good. Most of my adult life, head down, out in minutes, straight through till morning. I didn’t think much about it. Sleep was just something that happened.

Then I got heavier. Gradually. Over years. And I noticed — or thought I noticed — that sleep was getting worse. Lighter. More broken. More of those 3am moments where I was just… awake. For no obvious reason.

I assumed it was the weight. Made sense. More weight, worse sleep. I’d read enough to know there was a connection. So I filed it under “things that will sort itself out when I sort everything else.”

And for a while, that’s what seemed to happen. I started eating better in late 2023. Started fasting properly. Weight came off. And sleep did improve — noticeably, actually. I was dropping off faster. Staying down longer. Feeling better in the mornings. The pattern seemed clear. Fix the metabolic stuff, sleep follows.

Then I did a 21-day fast. And it wrecked it.

Not immediately. The fast itself was fine — sleep during extended fasting is complicated, I’d documented that before, and I knew some disruption was normal. But the weeks after were something else. Weeks of broken nights. That 3am wake pattern came back, harder than before. And this time, the blood tests were improving. HbA1c coming down. Liver enzymes normalising. Inflammation markers dropping.

The numbers said one thing. My sleep said another.

That was the puzzle I couldn’t figure out for months. I was doing everything right. The biology was responding. Why hadn’t my body got the memo on sleep?

THE DATA

Here’s where the tracking gets interesting. And I want to be honest upfront — what I’m about to share is N=1. This is what I observed in my own data. I’m not a researcher. I’m a fifty-three-year-old Yorkshireman with a spreadsheet and NHS blood test printouts going back to 2012.

WMTWL-5001 — my fasting and health tracking spreadsheet — includes daily sleep quality scores going back across most of the tracking period. I scored sleep out of ten, most nights.

Before the 21-day fast, in the months where I’d been eating better and losing weight, my scores were averaging around seven. Not perfect, but decent. Solid nights more often than not.

During the fast itself — weeks of fives and sixes. Some fours. Some proper disasters. Night twelve was two and a half hours, then two more, then a ninety-minute crash nap. That kind of thing. Extended fasting disrupts sleep. Not a shock.

What shocked me was after. Weeks one and two post-fast. Scores still in the fives. Sixes if I was lucky. And the pattern I kept seeing was that three in the morning thing — not struggling to fall asleep necessarily, but waking at that hour, fully alert, lying there for an hour, two hours, before drifting back.

I started cross-referencing with the biomarker data in the medical history records.

January 2025. HbA1c had dropped to 40 — back into normal range. Ferritin was low, and I was dealing with that. But the blood sugar story was moving in the right direction.

April 2025. HbA1c down again to 37. Optimal. Cholesterol still elevated, but the metabolic picture was genuinely improving.

And I was still waking at 3am.

The sleep data from WMTWL-5001 showed something I wasn’t expecting when I looked at the full year. 57 nights scoring nine or ten out of ten. Fifty-seven. That’s the number I ended 2025 with across the tracking period. Not bad. But the distribution of those nights mattered.

Almost none of them were from the months immediately after the fast. The good nights clustered towards the later part of the year. The repair wasn’t linear. It wasn’t proportional to how well the other markers were doing. Sleep lagged everything else by months.

That’s the pattern. And it took me a while to understand why.

[SUBSCRIBE PROPOSITION BEAT]

I’ll be straight with you. I’m not a doctor, I’m not a nutritionist, and I’ve never been on a morning television sofa. I’m a fifty-three-year-old Yorkshireman who got fed up and fixed it with data. If that sounds like your kind of content, subscribe. New video Monday, Wednesday and Friday.

THE INSIGHT

The thing I couldn’t work out initially was why the sleep lagged so far behind everything else.

I did what I usually do — read. Looked into what was actually happening hormonally when the body goes through significant metabolic disruption and then starts to repair. And the picture that came back was more complicated than I’d assumed.

Cortisol, ghrelin, leptin — these three hormones are all deeply tangled up with sleep. When your metabolic health is poor, they’re dysregulated. When you start making changes, they start to recalibrate. But they don’t all recalibrate at the same speed, and they don’t recalibrate in isolation from each other.

The research I came across — and the sleep science community has been fairly consistent on this point — is that when sleep drops below six hours, ghrelin spikes and leptin drops. That’s not controversial. It’s the physiology. Your brain gets mixed signals. Thinks you’re starving. Cravings go up, self-control goes down. EXP 796 documents this mechanism clearly: it’s not willpower failure, it’s biology.

But here’s what I think I’d missed. That relationship runs both ways. Poor metabolic health disrupts sleep hormones. But the sleep hormones themselves need the metabolic system to have settled before they can properly recalibrate.

What I found in my reading — and this isn’t clinical data from my own tests, it’s my interpretation of the research — was that cortisol dysregulation, specifically around that cortisol awakening response in the early hours, seemed like the most plausible explanation for the 3am pattern. Your body starts preparing you to wake an hour or two before you actually need to. When the cortisol system has been stressed, that timing gets disrupted. And sorting your blood sugar doesn’t automatically reset your cortisol rhythm. They’re connected, but they’re on different clocks.

Prof. Debra Skene — chronobiologist at Surrey, 190-plus publications — puts it plainly: peripheral clocks in the liver, the pancreas, the muscle, adapt at different speeds. The body isn’t one clock. It’s a collection of clocks, and when they’ve been running badly for years, they don’t all synchronise at once just because you’ve started eating well.

That’s the frame that made sense of what I was seeing. The biomarker data was improving. The metabolic system was repairing. But the sleep clock — the cortisol rhythm, the ghrelin-leptin balance — was still catching up. Lagging behind. Finishing last.

The analogy that best fitted what I was experiencing — from the WMTWL library, ANA-034 — is the council bin lorry.

Deep sleep is when the brain’s bin lorry does its rounds. All day, your brain accumulates rubbish — metabolic waste, cellular debris, the proteins that shouldn’t build up. Deep sleep is when the lorry drives through and clears the lot. When the sleep is broken and shallow, the lorry either doesn’t come, or it only does half the route. Week after week. The rubbish piles up. And when your body has been in poor metabolic shape for years, the bin lorry schedule has been a mess for a long time. Fixing the conditions that allow the lorry to come back doesn’t happen overnight. The routes have to be re-established.

That felt like exactly what I was experiencing. Not broken sleep in a permanent sense. Just a bin lorry that was slowly getting its schedule back.

And then, gradually — not on any particular day, not with any intervention I consciously chose — it did.

WMTWL-MOM-298 gave me a data point that stuck. October 2025. One evening with cake, chocolate, a bit of wine. By Sunday morning my sleep was disrupted. Monday through Wednesday: low-grade brain fog, restless nights. Thursday morning I woke up clear — eight hours, no alarm. Reset complete. And I noted at the time: that’s the third time I’ve tracked this exact pattern. One off-plan evening with sugar and alcohol, six days to get back to baseline sleep quality.

Six days. That’s how long the sleep system held onto the effects of one meal.

That told me two things. One: the sleep system was working again, because I could measure the disruption and the recovery. Two: the sleep system is the most sensitive indicator of metabolic state that I track. More sensitive than the bloods, in some ways, because it responds in real time.

THE LESSON

Beat 1 — What I’d do differently

If I were starting again, knowing what I know now, I’d track sleep quality from day one. Not obsessively. Just a simple score out of ten, every morning. Because without that, you can’t see the lag. You just feel like things aren’t working. With the data, you can see the direction — even when the direction is slow.

I’d also expect sleep to be last. Not worry about it. Just expect it. The biomarkers improving, the energy coming back, the weight moving — those come first. Sleep is the last one to the party. If I’d known that, I’d have been less confused in those months when everything else was improving and sleep was still patchy.

Beat 2 — What to watch for

If you’re already tracking, or already dealing with a health challenge, and you’re in the recovery phase — improving biomarkers, better energy, but still waking in the small hours — I’d say this: don’t assume you’ve broken something else. Don’t assume the protocol isn’t working. The sleep lag is real, and it’s predictable.

In my case it seems to have been somewhere between six weeks and six months after the main metabolic changes, depending on how significant the disruption had been. The 21-day fast pushed it toward the longer end. More modest changes, shorter lag.

If I were looking at my own data and seeing improving HbA1c, dropping inflammation, better energy — but still waking at 3am — I’d want to ask my GP whether cortisol rhythm and adrenal function were worth a conversation. Not a diagnosis. Just a conversation.

Beat 3 — The early warning signal

For anyone who’s noticing their sleep has quietly gotten worse over the last few years — not dramatically, not insomnia exactly, just lighter, more broken, less restorative — and they’re not sure why: this pattern is worth being curious about.

The sleep deterioration that I noticed as I got heavier wasn’t a sleep problem. It was a metabolic problem that showed up in sleep first. By the time the blood tests flagged anything, the sleep had already been telling me for years.

The thing is — cortisol regulation, ghrelin and leptin balance, the hormone systems that govern sleep quality — these start drifting before any GP test calls it anything. The sleep signal shows up in the data first. That’s worth knowing, whether you’ve had a diagnosis or not. If your sleep has quietly been getting worse and nothing obvious explains it, it may be worth looking at what else your body is telling you.

CTA + SAFETY

If this kind of pattern tracking is something you want to do for yourself, I’ve got a free sleep and metabolic tracking template — the same format I use for my own data. Link’s in the description.

And for anyone in the early stages of a recovery journey, or at the start of extended fasting — give sleep its own column in whatever you’re tracking. It’ll be the most informative thing you log.

One important note: everything you’ve heard today is educational content. It’s not medical advice. If you’re experiencing significant sleep disruption, or if you’re considering extended fasting or making changes to your diet or medication, speak to your GP first. All my biomarker changes are on my NHS record.