So I started out this experiment with a different design. But along the way, I made a few different adjustments which I’ll list below and why.
- Changed baseline food. I wasn’t feeling very well when I started out and this could have been for a number of reasons I detail in the second post of this series. As such, I decided to switch baseline food from the keto shake to a combination of eggs, cheese, heavy whipping cream, and some almonds. I further supplemented as I always do with magnesium, K2, D3, and a multivitamin.
- I effectively made this three day period of food reset the new Washout Stage, pushing the Intervention Stage with the Skittles back to the weekend.
- Given the new timeline, I decided to leave open the possibility of cutting the Aftermath Stage short if my numbers appeared to “snap” back before Friday. This would leave open the unlikely but possible scenario that I introduce a second intervention while still having this level of dietary control. (More on that in a moment…)
Technically, the experiment completed as of today given I’d speculate my numbers have indeed returned to baseline. Here’s the raw data:
Date | 1/28/18 | 1/29/18 | 1/30/18 | 1/31/18 | 2/1/18 | 2/2/18 | 2/3/18 | 2/4/18 | 2/5/18 | 2/6/18 | 2/7/18 | 2/8/18 |
Stage | Washout | Washout | Washout | Washout | Washout | Washout | Intervention | Intervention | Intervention | Aftermath | Aftermath | Aftermath |
Shake | Shake | Shake | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | |
+ Skittles | + Skittles | + Skittles | ||||||||||
CC TC | 263 | 270 | 261 | 246 | 259 | 232 | 247 | 245 | 245 | 245 | 198 | 286 |
CC HDL | 70 | 73 | 68 | 64 | 64 | 62 | 63 | 70 | 70 | 73 | 75 | 66 |
CC TG | 55 | 52 | 62 | 76 | 67 | 70 | <50 | 58 | <50 | 51 | 51 | 59 |
CC Glu | 95 | 97 | 108 | 90 | 102 | 96 | 117 | 116 | 111 | 111 | 98 | 105 |
CC LDL | 182 | 186 | 181 | 167 | 182 | 156 | 163 | 163 | 113 | 208 | ||
CC Non-HDL | 193 | 197 | 193 | 182 | 195 | 170 | 184 | 175 | 175 | 173 | 124 | 220 |
Test
Okay, so technically, we can’t judge the three day average of both the Washout and Aftermath stages because TG got <50 for two of the Intervention days. Meh! Thus, we can’t effectively see if I met the original hypothesis that the LDL of the Washout Stage would be higher than that of the Aftermath Stage.
We could compare Non-HDL, and those numbers look pretty interesting. But before going any further, I need to bring something else to your attention.
The Adjustment
There are two days where I didn’t feel I got the right amount of blood onto the CardioChek device strip, Feb 2nd and 7th.
On the 2nd, I was recording from all my devices as normal – Keto Mojo glucose, then ketones, eGlu, then lipid strip. But when releasing all the blood from the capillary tube for the CarioChek, a tiny bit sucked back up instead of emptying completely. I scrambled to squeeze the tube harder to release it, but there was no use and the device already started its analysis. I figured at the time it was probably so small it wouldn’t matter, but couldn’t be sure.
One other problem with the 2nd is that I had just enough stips and tubes to carry me to the next delivery date, so I couldn’t have retested anyway without risking a blind spot in my continuum.
On the 7th, the same thing happened, only this time the amount reabsorbed back into the tube was far more substantial. But this time around I had enough supplies to do a second test. The second test yielded numbers that I felt were more likely.
(Side note: please don’t ask why I don’t just do multiple CardioChek tests every time I check. This test is $11 a pop and I’m already out of pocket for very expensive lab tests ($100-450) throughout this experiment that together runs higher than my mortgage. I’m saying this preemptively because some assume I’m somehow funded by research grant or company. No, I allow only funding from the awesome people who support my Patreon or contribute directly. Thanks!)
So given this, I’m going to provide an adjusted table that uses the second CardioChek test here:
Date | 1/28/18 | 1/29/18 | 1/30/18 | 1/31/18 | 2/1/18 | 2/2/18 | 2/3/18 | 2/4/18 | 2/5/18 | 2/6/18 | 2/7/18 | 2/8/18 |
Stage | Washout | Washout | Washout | Washout | Washout | Washout | Intervention | Intervention | Intervention | Aftermath | Aftermath | Aftermath |
Shake | Shake | Shake | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | Baseline Food | |
+ Skittles | + Skittles | + Skittles | ||||||||||
CC TC | 263 | 270 | 261 | 246 | 259 | 232 | 247 | 245 | 245 | 245 | 257 | 286 |
CC HDL | 70 | 73 | 68 | 64 | 64 | 62 | 63 | 70 | 70 | 73 | 68 | 66 |
CC TG | 55 | 52 | 62 | 76 | 67 | 70 | <50 | 58 | <50 | 51 | 52 | 59 |
CC Glu | 95 | 97 | 108 | 90 | 102 | 96 | 117 | 116 | 111 | 111 | 95 | 105 |
CC LDL | 182 | 186 | 181 | 167 | 182 | 156 | 163 | 163 | 178 | 208 | ||
CC Non-HDL | 193 | 197 | 193 | 182 | 195 | 170 | 184 | 175 | 175 | 173 | 189 | 220 |
So yes, I do think had I gotten the second test on the 2nd, it would have been higher and more in line with the days before it.
And yes, I’m advocating to use the second test on the 7th, even though (ironically) it pushes the trendline away from my original hypothesis. But I’m not doing this to appear right, I’m doing this to get to the truth.
Current Thoughts
Funny as this may sound, this data was exciting precisely because it didn’t change that much! I know you, dear reader, are used to seeing my dramatic shifts in cholesterol with the carb swap experiments of Phase II in my research. But again, those were removing of fat and replacing with carbs. As I said at the beginning of this experiment, this is the first really controlled experiment where I added carbs.
Now that said, I did introduce several new things at once, which is a little uncharacteristic of me.
- This was the first time I experimented with high fructose corn syrup. And I had hoped it would be the last. Naturally, many will point out fructose is metabolized differently via the liver of course, and there are many studies stratifying how this plays into lipogenesis, adipose storage, NAFLD, etc. But I won’t get too deep in that rabbit hole here.
- Obviously, there are other ingredients in Skittles providing more potential confounding, such as hydrogenated palm kernel oil. (Which I’m sure Tucker will tell me is A-Ok)
- And finally, while I didn’t think of it until halfway though my first Intervention day, I was eating the Skittles at the end of the fatty meal, whereas I usually ate my carbs at the beginning in prior experiments.
On this last point, I should emphasize this was entirely by accident. I’m used to thinking of sweets as naturally following a meal — you know, dessert! Once I realized I was doing that, I decided not to change gears and instead keep the pattern the same.
Now, this order change may be indeed relevant to my resulting blood glucose following the meal; this wouldn’t surprise me. But would it actually make a difference to the lipid scores of the following day? That question has been haunting me the last couple days.
Experiment Extention!
Alas, this is where I reveal I’ll be prolonging my misery and doing a second round intervention followed by a second round washout!
Why? Because if there’s even a tiny chance the order in which I ingest the sugar component of the meal could have a longer-term impact in changing lipids for the morning of the next day, we need to know!!!
So without further adu, here is the new, extended schedule:
Thus, this is not a postmortem, it’s a “midmortem”.
I won’t lie — these experiments are very taxing not just for the money spent, but because I likewise eat to the schedule (as shown in the first post) of 10 AM, 3 PM, and 8 PM to keep consistent with all prior experiments and equalize against other possible confounders. So I don’t relish this extra seven days added, especially given right now I’d kill for a fat, juicy steak.
Yet there’s no getting around it, I’m now very uniquely well placed in controlled, personal lipid data now and I might as well take this opportunity to find this out for all of us as well. The things I do for science!
Keep up the good work … I love this kind of stuff. Partly because it shows that the so called experts do not really know what is going on and partly because you are finding out what is really going on.
Thx! I too am amazed at how much is left to be discovered on this subject given how many billions and billions of dollars have been already spent researching it.
I’m looking forward for the results. It does seem that the order matters. https://www.economist.com/news/science-and-technology/21736499-it-not-supposed-leave-small-intestine-how-too-much-fructose-may-cause
What is interesting, how much is ok? Judging by your data, it looks like a whole packet of skittles was not so bad if taken after a good meal?
I definitely *feel* different with the reordering. But more on that in a coming blog post. 🙂
Liver Fat and Fructose this is compelling.
http://www.gastrojournal.org/article/S0016-5085(17)35685-8/pdf
Great stuff — and not to give anything away… but I’m sure you’ll be interested the results of this experiment especially. 😉
As always I’m on the edge of my seat! I would like a blog post about big picture. Your most pressing questions
Implications of your findings so far
Most significant findings so far
LDLc high for bad reasons high for not bad reasons
Waiting on your N=1 with Alcohol! Guess it would have to occur on vacation! LOL
I hope to do just such a blog post soon. It might need to be broken up into several parts.
That said, I am working on a book now that will do that in full. 🙂
I probably won’t do one with alcohol as I generally don’t drink. When I do, I’m actually a huge lightweight and have been my whole life, so I wouldn’t be that representative anyway.
Response to 15-Feb twitter out of lab-range low White Blood Cell count: https://twitter.com/DaveKeto/status/964356043886116865
At 63 yo, LCHF/keto for 6 years. Three consecutive CBCs over 5 months in 2016 with below lab-range WBC and other blood measurements, without noticeable health impact. As many suggest, ‘ancestral’ dieters may have different healthy ranges. My NMR panel results from this past week coincidentally are all within 3% of the five-year average from all test values:
–==== CholesterolCode.com Report v0.8 ====–
Total Cholesterol: 358 mg/dL 9.26 mmol/L [5 yr Range 303 to 424]
LDL Cholesterol: 230 mg/dL 5.95 mmol/L [Range 191 to 266]
HDL Cholesterol: 108 mg/dL 2.79 mmol/L [Range 78 to 127]
Triglycerides: 100 mg/dL 1.13 mmol/L [Range 71 to 153]
–CHOLESTEROL REMNANTS–
Remnant Cholesterol: 20 mg/dL 0.52 mmol/L >>> Medium-Low Risk Quintile
Remnant Chol to HDL: 0.19 >>> Lowest Risk Quintile
Go to https://tinyurl.com/y8hokam2 for more on Cholesterol Remnants
–ATHEROGENIC INDEX OF PLASMA (AIP)–
AIP: -0.393 >>> Lowest Risk Third
Go to https://tinyurl.com/ycccmmnx for more on Atherogenic Index of Plasma
–CONVENTIONAL MARKERS AND RATIOS–
Friedewald LDL-C: 230 | Iranian LDL-C: 217
Total/HDL Ratio: 3.31
TG/HDL Ratio in mg/dL: 0.93 | in mmol/L: 0.41
Interesting — thanks for the added data! 🙂