201 nights sleep data proves useful

I have results! Spreadsheeting for the win!

For the first time since starting this project, I have learned something interesting and useful from my punctilious symptom tracking. Thank goodness I wasn’t wasting my time typing fuzzy numbers into spreadsheet cells for half an hour every day for 201 days. (Was that headache a 5/10? Or more of a 6? 🤷🏻‍♂️)

The sleep data is less fuzzy, and it has coughed up a result. And I know that result will really shock you, a total surprise, no way anyone could have seen it coming…

(Sitting down?)

Every time my sleep gets particularly bad, so do all my symptoms. 😮

Amaze! Much wow. Very inform.

But I wouldn’t bother your inbox just for that. That’s not actually the useful part, or even slightly surprising. There were two actual surprises in my data — and both of them good news! If you squint a little.

And vice versa! No bad days without high sleep pressure

A strong correlation between symptoms and sleep deprivation might be at the very tippy top of my list of things I expected to confirm with my data. It is interesting to see it in hard numbers, and even a little informative to see a consistent lag time of one to two days, both on the way down and back up. It clearly takes an accumulation of “sleep pressure” (the need for sleep) to trigger symptoms.

And it takes more than one good night to recover.

But that’s still not the best part.  There is another signal that I did not see coming:

So far, I haven’t had any symptom surges without sleep loss.

This is kind of a big deal

I have always believed that my frequent bouts of horrendously bad sleep are a complication, not the main issue, just something that makes my bad situation worse.  For years I have taken it for granted that I definitely can still have a really horrible, terrible, no good day of symptoms without sleep loss.

But four months of data does not support that belief.

There’s little or no sign of symptom badness happening between sleep-deprived periods.  If the sleep pressure’s low, so are the symptoms. If I sleep okay, I feel okay. 

Who knew? Shouldn’t I have known? How could I miss that?  But it’s easy: effects that lag 1-3 days behind their causes are extremely hard for our monkey minds to perceive … and if we think we already know? If we have a clear bias, as I did? Messy contrary data probably cannot overcome that.

And that is why objective data is so useful.

It’s the narcolepsy, stupid?

Reminder: Narcolepsy is a disease of lousy sleep regulation, and narcoleptics have just as much trouble staying asleep as we do staying awake.

I have always been open to the possibility that a lifetime of serious sleep loss could actually be a casual factor in my poor health, and it’s even possible that my narcolepsy is The One True Problem here: a life wrecked by terrible sleep. But I’ve also never really taken that seriously.

This data is a strong shove in that direction.  It’s also “actionable.”

Must … protect … sleep

I already do a great deal to protect my sleep — that’s just what you learn to do when you have a serious sleep disorder.  When your sleep falls apart as easily as cotton candy in the rain, you do whatever you can to avoid rain. Anything that interferes with sleep is already my enemy, and has been for a long time.

But I could up my game.

There are ways I could get even more serious about protecting my sleep.  It’s obviously necessary if sleep deprivation is not just a major driver of my symptoms but the driver, my main demon.  But I don’t even have to be convinced of that to make that experiment worthwhile.

I wanted data driven hypotheses and experiments. Well, I just got one I didn’t have before: maybe the single best thing I can do for myself is become an obsessive sleep defender. Worth trying. It’s not like I’ll regret doing an even better job of protecting my sleep.

Some sleepy stats

• 6.1 hours average (a little better than I expected)

• 2 nights with less than one hour sleep, 7 more nights of 1-4 hours, 41 nights of 4-6 six hours.

• On the happier end of the scale, I pulled off 1 night of 9 hours! And, to my amazement, I had a dozen 8-hour nights.

• One night I got minus 1 hour of sleep. Yeah, I’m not sure what that means either…

Bonus round: does alcohol hurt my sleep? The data says…

In the last instalment, I wrote about my alcoholism lite and the need to try to quit or curtail. This seems like a good moment to check my data for a booze-insomnia link, doesn’t it? I have months of decent records on both. If booze is making me sleep badly, that should show up as a higher average alcohol intake preceding my 20 worst nights, and vice versa. Does it?

… nope!

I fully expected to get a very-bad-news result, but I was “disappointed”! There was no signal at all:

• Average of 1.58 drinks on the best 20 nights.

• Average of 1.54 drinks on the worst 20 nights.

So my sleep problems are not obviously caused by alcohol. Hallelujah. At least there’s that!