Sleepy Index

Fairly Uninformative

In the last post, I looked through a load of old blog posts to see what I had thought was going on fatigue wise at the time I wrote them.

And I concluded that I’d expect a hypothetical objective measure of fatigue to have the following properties:

• Vague constant fatigue ever since 2023, which is nevertheless a noticeable improvement on my previous state, which was itself not too bad

• In summer 2023, very few afternoon naps or failed afternoon naps

• Dramatic improvements in deep ketosis, mild improvements in half-hearted lowish carb diets

• Low fatigue levels in the run up to Christmas 2024

• Increase in fatigue after Christmas 2024 to the point where I told my boss it was affecting my work (so ‘slightly disabled’ rather than just ‘limited’)

• Recovery after that

• Massive fatigue August 2025 on, probably explained by thyrotoxicity, getting worse and then better over the period I was home

• Massive fatigue early Christmas 2025, explained by sulphite poisoning, but getting better later on

• Followed by improvements to the point I felt able to write the fatal words: ‘The almost complete remission of my once-crippling chronic fatigue’

• Followed by quite bad hypo-type fatigue, improving and then getting worse as I went hyper again.

• Followed by really bad hyper-type fatigue over Easter 2026

• Followed by bad hypo-type fatigue, slowly improving until I start being a bit hyper again and get more fatigue.

• Eventually leading to my June state of being totally off the thyroid and still irritatingly hot all the time and still fatigued

• And finally now, in mild ketosis for a while, fatigue lessening notably to the point where it’s just ‘limiting’ again rather than ‘disabling’.

---

I can’t get any sort of objective measure, I don’t even know how to do that in principle, but I have about three years of daily notes in electronic form, and I’ve data-mined those records.

Wherever I’ve said “nap”, which means I’ve gone to sleep for a bit during the day, that counts 4, and “napfail”, which means I went to bed but didn’t sleep, that counts 2, and if I said “shattered”, that counts 2, and any of “sleepy”, “sluggish”, “woozy”, “tired” count 1, and “tiredish” counts 0.5.

So for instance my notes for 10/04/26 say “woke 1100, fine, achy legs in pub, 1830 tired, napfail, dinner, shattered, another nap, fine, bed 0100, headache”

That seems to have been a particularly bad day, and counts 9. Most days are quite a lot better than that.

And I’ve plotted a graph of that index, and taken a seven-day moving average as usual, and here it is:

Looking at that the first thing I notice is that it’s really very noisy! Even the seven day average is jumping around all over the place, but it’s roughly consistent with all the blog posts, in that from 2023-2025, it’s hovering around 1/day, which is to say that most days I’ve got just enough fatigue to remind me that I’m not in good health, but it’s not really a problem.

And then new year 2025 there does seem to be some increase in fatigue for some reason. But there doesn’t seem to be much evidence of recovery, it’s just getting worse. And in July that year after I left my gig at Midsummer Energy it gets much worse, then when I go home to visit Mum in August 2025 it goes away completely, and then comes back in the second half of that visit.

After that it seems to decay away, until by Christmas 2025 it’s hardly a problem at all, at which point I said the bad thing.

Then in the early part of Christmas it got bad again, then went away, and then when I came back it got really bad, and then decayed away.

And then Easter 2026 it was really really bad, and then it’s kind of decaying away for a bit, but recently it’s got worse again, and my current ex150ishsccf+low-carbs initially seemed to help, and now it’s got quite a lot worse again.

And it’s currently quite bad! Most days I need to at least try to sleep in the middle of the day, and on a typical day I have fatigue symptoms.

I really don’t trust this measure at all! Self reports of psychological symptoms are about the worst metric you can track, because the threshold where you actually decide to go and write something down is very highly variable, and it depends on what you think is going on, so you can’t help but fiddle it to conform with what you think is happening.

Still, it’s all I’ve got, and now I can’t use thyroid dose as a more objective proxy for healing (because it’s gone to zero, and I’m still a bit overheated, so I think it’s probably gone to zero for good) I think I’m going to add it to my graphs.

I think at least it’s enough to confirm my subjective impression that my tiredness was not so bad for the first couple of years, but is now becoming a real problem that is affecting my life quite negatively, and stopping me doing things I want to do, and also making me much stupider than I like to be.

I can look at it, and think, yes, all pretty stable until Christmas 2025, and then just getting steadily worse for the last eighteen months.

This is what it looks like added to the usual graph, with thyroid and waking temperature crushed down very small, because you can see everything you need to know even when they’re very shrunk.

If I score my desiderata:

Vague constant fatigue ever since 2023, which is nevertheless a noticeable improvement on my previous state, which was itself not too bad

TRUE (although I can’t see the previous state)

In summer 2023, very few afternoon naps or failed afternoon naps

FALSE (fatigue actually seems worse then than before or after, all through the mysterious honeymoon period I was apparently quite tired, and occasionally napping)

Dramatic improvements in deep ketosis, mild improvements in half-hearted lowish carb diets

UNKNOWN (I can read it that way if I stare at it, but it’s not obvious, and in my memory it was obvious)

Low fatigue levels in the run up to Christmas 2024

TRUE (That does appear to be the least tired I’ve been since i started keeping records)

Increase in fatigue after Christmas 2024 to the point where I told my boss it was affecting my work (so ‘slightly disabled’ rather than just ‘limited’)

TRUE (but it’s only gone back up to what it had been previously)

Recovery after that

FALSE (again, that’s a clear memory and I wrote about it at the time, but it’s not visible on the graph)

Massive fatigue August 2025 on, probably explained by thyrotoxicity, getting worse and then better over the period I was home

FALSE (I went home, and the fatigue went away completely for a couple of weeks, and then came back stronger. After that it decayed steadily away for a couple of months until it was nothing by Christmas)

Massive fatigue early Christmas 2025, explained by sulphite poisoning, but getting better later on

TRUE (A big spike in fatigue while sulphited)

Followed by improvements to the point I felt able to write the fatal words: ‘The almost complete remission of my once-crippling chronic fatigue’

TRUE ( followed by recovery to almost nothing again)

Followed by quite bad hypo-type fatigue, improving and then getting worse as I went hyper again.

TRUE (steady improvement even though I was losing weight, then it starts getting worse)

Followed by really bad hyper-type fatigue over Easter 2026

TRUE (most tired I’ve ever been by the look of it)

Followed by bad hypo-type fatigue, slowly improving until I start being a bit hyper again and get more fatigue.

TRUE (gets better and then gets worse)

Eventually leading to my June state of being totally off the thyroid and still irritatingly hot all the time and still fatigued

And finally now, in mild ketosis for a while, fatigue lessening notably to the point where it’s just ‘limiting’ again rather than ‘disabling’.

TRUE (although that improvement seems to have reversed since I wrote it, and in this case I think that’s because it was Kate’s birthday and that involved a startling amount of whisky and a fairly predictable hangover, and I am still feeling a bit out of sorts two days later)

That’s nine TRUEs, three FALSES, and one UNKNOWN, which is really not very impressive since I’d already seen this graph when I was reading through my notes.

---

Going forward, I think this measure will be even worse as a metric, because now I know it’s going to get data-mined and graphed, I’m not going to be able to avoid unintentionally biasing my records.

I think I’m going to try to aim for ‘every time you feel even a bit tired, write it down’.

That will probably make this metric jump much higher, but it should still be true that if I ever get back to what I’d consider good health, it should go to zero. I don’t think it’s at all normal to be tired enough to make a note of it if you’re in good health and you haven’t been doing some kind of hard physical labour for a few hours at least.

One thing that’s really conspicuously missing from this graph is my very clear impression that when I’m in deep ketosis¹, I get hardly any tiredness.

And that really should be there, so I’m going to stare at everything very hard and work out if I’ve made some sort of terrible mistake. But at the moment I can’t see it. I think it’s more likely that this is just a really bad measure to be looking at².

But it’s all I’ve got, so I’m going to add it on a permanent basis, and maybe try to come up with similar things for hypo- and hyper- thyroid symptoms, which might make what’s going on a bit clearer.

My suspicion that this recent tiredness is somehow related to recent massive weight yo-yos doesn’t seem borne out here either!

It’s definitely true that since I wrote https://theheartattackdiet.substack.com/p/yo-yo-theory in October 2025, my weight has been yo-yoing much more than before³, and it does kind of seem that periods of large weight loss were followed by increases in fatigue levels, whereas long periods of rising weight seem to also be periods of slow decay of fatigue, but it’s not at all jumping out at me from this graph, and there are two great fat obvious counterexamples.

The first is the six-month honeymoon period after I first gave up all PUFAs, when I just lost weight effortlessly for four months. At that point I think I must have had more linoleic acid in my body fat than ever before or after, and must have been dumping quite a lot of it into my blood, and yet it doesn’t seem to have made much difference fatigue wise. That was also a period of steady decrease in thyroid dose, which is also weird if PUFAs in the blood are slowing my metabolism. I wonder if that was a period where I was just usually slightly hyper?

And the second is the recent Easter 2026 Mom Test, where I gained a ton of weight (so probably very little PUFA coming from stores), and yet the fatigue was really bad, worst it’s been since I started keeping records. In that case, that probably was a period of hypermetabolism.

But there’s really nothing clearly going on here, even the things I expected to see like ‘ex150ish fixes my tiredness’, just aren’t obviously there, even though you can sort of see it if you squint.

---

The only thing I think I’ve learned from doing this is that my subjective strong impression that the fatigue has been much worse recently is probably true.

But I’m not even sure about that. I may have just been paying it more attention...

I am not happy! And I notice that I am as usual, very confused. And so at least one of the things I believe must be fiction.

1. Blue bars on the graph, not eating carbohydrates at all except for what’s in the tomato passata I add to my everlasting stew.

2. On the other hand, keto entry often means ‘keto flu’, and maybe that’s showing up and cancelling later lack of tiredness?

3. And that’s weird in itself! Why?