Last time you heard from her, Krinn had just put out a tumblr post titled An Ad-Hoc, Informally-Specified, Bug-Ridden, Single-Subject Study Of Weight Loss Via Potassium Supplementation And Exercise Without Dieting. After losing 6 lbs in our Low-Dose Potassium Community Trial, she decided not to stop as planned but instead to keep going, and in fact go even harder. Eventually she ramped up to around 10,000 mg potassium a day, and lost even more weight.
Krinn also added an exercise habit that she described as a “naïve just-hit-the-treadmill exercise regimen”. Even with this in mind, her results still seem remarkable, because most people do not lose 50 lbs from starting a moderate treadmill habit:
We published a short review of that original post on this here blog of ours. That was in July 2023. Now, Krinn is back, and more powerful than ever, with an untitled post we’ll call A Year And Change After The Long Post About The Potassium Experiment (AYACATLPATPE).
The long and short of it is that Krinn kept taking high doses of potassium and kept losing weight, eventually reaching her goal of 200 lbs. There was a long plateau in the middle after she first brushed up against her goal, but she maintained the original weight loss and eventually lost the remaining weight:
In personal communication (see very bottom of this post), Krinn noted that:
One of the few things the graphs say really, really, really loudly is “Krinn lost 30+ pounds _and stayed that way for at least a year._” … one of the overwhelmingly common failure modes of existing interventions: people lose some weight and then gain some weight and end up fairly close to where they started. Whatever else happened in my experiment, it sure wasn’t that: I lost a significant amount of weight and then _stabilized._ That seems important.
This time we don’t have much to add, but as before we wanted to reproduce her post for posterity. And we do have a few thoughts, mainly:
This seems like more evidence that high doses of potassium cause weight loss. It suggests that potassium is probably one of the active ingredients, maybe the only active ingredient, in the weight loss caused by the potato diet. Krinn was taking about as much potassium as you would get if you were eating 2000 calories of potatoes per day, and experienced similar weight loss.
It’s good to be skeptical of single case studies, however rigorous and careful they may be, but here are a few things to keep in mind:
Remember that participants in the Low-Dose Potassium Community Trial lost a small but statistically significant amount of weight (p = .014) on a dose much lower than what Krinn was taking — only about 2,000 mg of potassium a day on average, compared to Krinn’s ~10,000 mg per day. This can’t confirm the effects of the higher dose, but it is consistent with Krinn’s results, and the final sample size was 104 people.
There’s also at least one successful replication. Inspired by Krinn’s first report, Alex Chernavsky did a shorter potassium self-experiment and lost about 4 pounds over a two-month period, otherwise keeping his diet and exercise constant. He also provided this handy table:
Finally, we know of two other people who are losing weight on high-potassium brines, at least one of them without any additional exercise. They’re both interested in publishing their results, probably in early 2025. So watch this space. :)
As before, we want to conclude by saying that Krinn is a hero and a pioneer. She is worth a hundred of the book-swallowers who can only comment and couldn’t collect a data point to save their life. If you want to do anything remotely like what Krinn did, please feel free to reach out, we’d be happy to help.
Here’s a reproduction of Krinn’s full report as it appears in her tumblr post:
A Year And Change After The Long Post About The Potassium Experiment
A year and change after the long post about the potassium experiment, I reached my weight-loss goal. This is a quick, minimally-structured thought-dump about it. As before, this is part of a wider conversation that starts with A Chemical Hunger.
Methodology: I mostly kept doing what I’d been doing. Turned up the exercise dial a bit, turned down the potassium dial a bit. Both still, AIUI, quite high compared to American baseline. Some bad news — in addition to whatever confounding factors were present last time, there’s a few extra ones now from my life in general going very poorly. As before, here’s the data, Creative Commons Zero, good luck with whatever you try on it. After making it to one year of being fairly diligent, I decided to let things vary and see what happened — on the one hand, I’d gotten far enough towards my personal goal that I wasn’t too fussed about the last 10%, and on the other hand, if this works in general and even work when you’re kinda half-assing it, that too is great news.
Interpretations: There’s multiple ways this could go. Here are a few that were easy to think of.
- Potassium or potassium-plus-exercise caused me to lose weight
- Exercise caused me to lose weight and potassium was irrelevant
- Something else caused me to lose weight
I would prefer to believe that potassium-plus-exercise caused me to lose weight. The data I have and my experience of gathering/being that data, to some extent support that conclusion. Flipping that around, if I ask “does that data rule out this conclusion?” no it absolutely does not. But it’s important to note that the exercise-only conclusion is only slightly less-well-supported and the none-of-the-above explanation is much-less-well-supported but certainly not ruled out. I have a preferred explanation, but all three of these explanations are live.
My subjective experience of the thing was that there was an easy part and a hard part. In the easy part I lost weight at a pretty rapid and consistent pace. In the hard part, my weight changed less and went back and forth more than it went down. If you buy into SMTM’s “something is screwing up people’s lipostats” theory, this is very consistent with that theory: potassium reduced or removed the something, my weight briskly dropped back to a healthy range (the first 9 months of the graphs) and then stabilized. However, the competing theory of “Krinn was super out of shape and then she started exercising” is also supported by the graphs (not shown on the graphs: my fairly poor 2022 exercise habits — my long-term exercise habits have had some good stretches, but the plague years did not do good things for me there!). I’m not sure whether it matters that I shifted from mostly treadmill time to having a couple of walks around the neighborhood that I can do pretty much on autopilot (shout-out to Mike Duncan’s Revolutions, this show is the first time podcast as a medium has clicked for me and it’s a great show). I do think, though, that exercise is a bit more complicated than I was really grasping. That, in turn, makes me glad that I’m tracking three exercise metrics rather than just one — if I was going to track only one, it’d be exertion, but exertion, exercise minutes, and step count, together make it possible to at least take a guess at what qualities a day’s exercise had.
Regarding my own questions from the first post:
• How safe is this? When I made the first post I was antsy about “adding this much potassium to your diet is probably safe for people in generally good health” but now I’m pretty sure it’s true. Some health problems can take a long time to present themselves, but adding this much of something to your diet for two years and having it be fine, is pretty persuasive evidence that the thing is probably fine. It could still easily turn out to have negative health impacts that are important, but a huge swath of the things you’d be worried about, are vanishingly unlikely once you’ve hit the point of “I’ve been taking this for two years and I’m fine.”
• Does this replicate? Well, it’s self-consistent for me, and I don’t want to gain 50 pounds and try again. I did not like the shape of my body at +50 pounds from where I am now! So this is a question for others.
• How much do other nutrients matter? I don’t know. Mostly not equipped to rigorously check.
• Does HRT matter? I’ll let you know if I can get back on HRT. I would definitely like to investigate this.
• Does dieting matter? Probably: my diet changed involuntarily over the course of two years and that certainly matters to some extent, but one of my ground rules is that I’m focusing on controlling exercise and potassium, the things I can control. Diet is far more complex and also in my life particularly, more susceptible to unplanned, involuntary change, so I’m writing it off as a factor.
• Does this help with cannabis-induced hunger? I think I was off-base/over-optimistic with this one and it either doesn’t matter or matters a small amount.
• Is there a point where I get really hungry/tired or start accidentally starving? I did not reach such a point. I felt basically fine the whole time.
I was cooking with this though:
If you tell someone you want to lose weight and would like their advice, it is overwhelmingly likely that the advice will involve exercising more. Everyone has heard this advice. And yet, as Michael Hobbes observes in a searing piece for Highline, “many ‘failed’ obesity interventions are successful eat-healthier-and-exercise-more interventions” that simply didn’t result in weight loss. Even if we as a society choose to believe “more exercise always leads to weight loss, most people just fuck up at it,” that immediately confronts us with the important question, why do they fuck up at it? and its equally urgent sibling, what can we learn from those who succeed at it to give a hand up to those who have not yet succeeded?
Conclusion: I’m gonna keep writing things down in my spreadsheet for the same reasons as last time. I’m not sure what exactly I’m going to do as far as twiddling the factors, because now my main goal is somewhere between “don’t gain weight again” and “see what happens,” but I do know that writing down what happens is Good Actually, so I’m going to keep doing that.
Slightly after publication, Krinn sent us these comments, which she agreed we could publish:
Personal Communication
Dangit now I’m having the first draft effect: writing the first draft and sleeping on it tells me things I should have written. In this case, I think there’s a plausible reading that my experience supports the “potassium does something good at a high enough effect size to care about” line of argument because while the peaks of how much effort I put in were fairly high — the periods of combined high exercise and high potassium intake — the most noticeable effect was when I was ramping up on both of those in the first 9 months, and when I was in just-bumbling-through-like-an-average-human mode, the effect didn’t reverse itself. There were plateau periods and there were slow-reversion periods, but there was definitely no “you slacked off and now there’s rapid weight gain mirroring the rapid weight loss” effect. I think that’s positive? I think it’s plausible to read it as “once I got the majority of the weight loss effect, locking in that benefit was easy.”
In any case one of the questions I was interested in was “if this works, does it work well enough that an average person can successfully implement it?” and I am now convinced that that’s a clear “Yes”.
I wouldn’t say there’s any part of this experiment that I’m actively unhappy about, but I do find it a little frustrating that this is basically just another piece of evidence on the pile of “here’s something that is consistent with the lithium/potassium hypothesis, but that is also consistent with some other stuff, and my main observation is that Something Happened” — intellectually I feel sure that much solid science is built by assembling big enough piles of such evidence and then distilling it into “now we know Why Something Happened,” but putting one single bit of evidence on the pile is still something where I need to make my own satisfaction about it rather than having a well-established cultural narrative rushing to bring me “yes! you did the thing! Woohoo!”
Also thinking more about the potassium experiment I’m having one of those “hold on a minute, this should have been obvious to me” moments — one of the few things the graphs say really, really, really loudly is “Krinn lost 30+ pounds and stayed that way for at least a year.” That’s one of the crucial parts of the whole obesity thing, that second half, right? That’s one of the overwhelmingly common failure modes of existing intervention: people lose some weight and then gain some weight and end up fairly close to where they started. Whatever else happened in my experiment, it sure wasn’t that: I lost a significant amount of weight and then stabilized. That seems important.
Yessssss I get the smug clever-kitty feeling, this is exactly why I have that “ratchet” column in the spreadsheet: the last ratchet-tick day from more than a year ago (i.e. it’s locked in) was July 10th 2023, on which day my week-average weight was 212.4lbs, down 33.6lbs from the start of the year.
So that early period of dramatic weight loss is noteworthy because we can be confident that whatever the cause was — potassium, exercise, or something else — it caused durable weight loss, which is exactly the thing we are looking for.
This is a conclusion we couldn’t have reached in July 2023, with the major writeup I did, because at that point “something else happens and Krinn gains the weight back” was very possible, was one of the likely answers to “what comes next?”
SLIME MOLD TIME MOLD 