Potassium sleep experiments

The first night (10/12) involved falling asleep in 30 minutes rather than my usual 19.6±11.9, waking up 12 times (5.9±3.4), and spending ~90 minutes awake (18.1±16.2) The next day (10/13) I took a similar dose and double-checked the fan before bed: 25 minutes to fall asleep, 10 awakenings, 35 minutes awake, but I woke fairly rested. So it seems like the fan was only partly to blame. The third day (10/14) I omitted any potassium: 21/8/29. Fourth (10/15) on again with an evening dose: 54/7/24. Fifth (10/16), off: 16/2/6. Sixth (10/17), on with a halved dose: 33/3/6. Seventh (10/18), off: 17/6/7. Eighth (10/20), half: 33/6/15.

At this point I began randomizing consumption between on and off; since this is preliminary, I didn’t bother with blinding potassium consumption. Ninth (10/21), on: 25/7/9. Tenth (10/22), on: 18/8/10. 11th (10/23), off: 26/4/10. 12th (10/24), off: 33/7/16. 13th (10/25), on: 32/7/13. 14th (10/26), on: 21/5/8. 15th, on: 34/2/1. 16th, off: 16/7/15. 17th, on: 29/8/20. 18th, on: 17/10/17. 19th, off: 36/9/24. 20^th (11/1), on: 21/4/19. 21^st (11/2), off: 29/7/16. 22nd (11/3), on: 26/7/10. 23rd (11/4), on: 16/4/11. 24th (11/5), off: 21/4/17. 25th (11/6), on: 19/9/24.

11 Nov, on: 15/3/08. 13 Nov, off: 11/8/21. 14 Nov, off: 18/8/22. 15 Nov, on: 30/8/16. 16 Nov, off: 20/7/12. 17 Nov, on: 34/8/20. 18 Nov, on: 12/8/22. 19 Nov, off: 24/8/14. 20 Nov, on: 26/4/39. 21 Nov, off: 15/6/14. 22 Nov, on: 26/8/29. 23 Nov, on: 23/4/8. 24 Nov, off: 24/3/5. 25 Nov, on: 27/7/15. 26 Nov, on: 30/10/17. 27 Nov, off: 42/12/13. 28 Nov, off: 40/11/42. 29 Nov, off: 19/14/50. 30 Nov, off: 32/8/39. (Here I counted the sample-sizes and realized the off days were drastically under-represented, reducing statistical power; so I have eliminated randomization and gone off potassium.) 1 Dec, off: 28/10/15. 2 Dec, off: 37/8/20. 3 Dec, off: 36/6/18. 4 Dec, off: 19/9/33. 5 Dec, off: 25/8/27. 6 Dec, off: 30/13/45. (Now balanced, resuming randomization.) 7 Dec, on: 31/9/60. 8 Dec, off: 22/9/23. 9 Dec, off: 11/5/21. 10 Dec, on: 30/4/10. 11 Dec, on: 22/9/50. 13 Dec, off: 20/5/6. 14 Dec, off: 33/13/25. 15 Dec, on: 26/11/22. 16 Dec, off: 33/12/28. 17 Dec, off: 42/9/31. 18 Dec, off: 31/9/61. 19 Dec, on: 23/8/18.

If potassium was disturbing my sleep, I didn’t necessarily want to wait for any one metric of wakefulness to reach significance; rather, I wanted to combine them into a single metric of sleep problems: time to fall asleep (latency), number of awakenings, and time spent awake. (With all 3, higher is worse.) Number of awakenings tends to vary over a smaller range than time to fall asleep or time spent awake—a normal value for the former might be 5, rather than 30 for the latter; to compensate for that, we convert each metric into a standard deviation indicating how unusual eg. 10 awakenings is and whether it is more unusual than it taking 15 minutes to fall asleep. Then we can do a standard test. To graph the data at each step, starting with graphing all the data on an overlapping chart (this is not per day):

Nights off potassium are colored blue and nights on potassium are red; it looks like red dots are higher than blues, overall, but the trend is not clear. So we convert each individual datapoint to its respective standard deviation:

The trend has become much clearer, but the final step is to add each day’s scores to get an overall measure:

Now the different has become dramatic: one can almost draw a line separating both groups without any errors. As one would expect given this graphical evidence, a Bayesian two-group test reports that there is ~0 chance that the true effect size is 0, and the most likely effect size is a dismaying d = −1.1:

A two-sample test agrees: p = 0.0002168. (There is no need for multiple correction in this instance.) This confirms my subjective impression.

A secondary question is whether potassium delivered any waking benefits. I write down at the end of each day my rating 2–4 how happy and/or productive I felt that day. Does this self-rating show any effect? Here’s a plot of each day colored by whether it was a potassium day:

There is little visible effect, and the formal Bayesian analysis is as weak as the sleep disturbances are strong:

So there is no apparent benefit from the potassium.

This experiment was hastily done and has several weaknesses, some I mentioned before; in ascending order of importance:

1. Non-Uniform Dosage:

   Number of dosages varied from day to day as was convenient and doses were measured approximately with a spoon (since 4 grams is a pretty substantial amount, after all). Here is another objection I don’t think matters: lower than average doses may contribute to an underestimate of the effect size… but that implies that the effect size is even more extreme than −1.1! We are interested in problems that would shrink the effect size back to 0, not imply that it’s even worse than −1.1.

2. Incomplete Randomization:

   As covered in the data section, there was a severe imbalance in sample size for each condition, so I stopped randomization for about a week. Intuitively, I don’t think there was anything special about that week in regard to getting very good sleep (as would be necessary to contribute to an overestimated effect size), but if anyone disagreed, it would not be hard to exclude those days and use the rest.

3. No Blinding:

   I am not sure how much this matters. I had no expectation that potassium would affect my sleep at all, one user specifically denied any effect, the only study suggested I’d find improvements, I did not want to find a negative effect much less such a severe effect, and the sheer strength of the effect over a multi-month period is a bit more than I would expect from any expectancy or placebo effect.

4. Non-Uniform Timing:

   Of the issues, this is the most important. If potassium has some stimulating effects as anecdotes claim, then timing may be causing all the sleep disturbances and not potassium per se. It might be exactly like vitamin D in this respect: taken in the evening, it badly damages sleep but taken in the morning, it does nothing or it improves sleep.

If I were to do a followup experiment, it would be blinded & randomized as usual, with consistent doses (eliminating objections 1–3), but more importantly, the dose would be consumed upon awakening.

I am not sure I will bother with a followup experiment. Potassium is not of particular interest to me, my existing supply is low after months of consumption, I observed no subjective improvements on consumption, and so I am not inclined to run the risk of damaging more months of sleep. Other people can do that.