Okay, before we get started, you should know I’m not going to have enough time to cover everything that came out of this experiment. Needless to say, there’s lots and lots of interesting stuff in the numbers. But the good news is that I’m making the raw data available to everyone to find these things out for themselves. (See downloadable spreadsheet below)
First, let’s the stages and dates as I’ll be referencing them quite a bit:
Baseline food eaten every single day from January 30th to February 14th are the following:
- 10 eggs
- 8 oz cheddar cheese
- 60 ml Heavy Whipping Cream
- 2 sausage patties
- 43 almonds
- 15 grams of butter
- Supplements: Multivitamin, Magnesium Malate, K2 MK-7, D-3, Sports Salts
These were divided into three meals eaten at roughly 10 am, 3 pm, and 8 pm — with one floating “snack” between 3 pm and 8 pm of almonds (usually close to 5:30 pm).
- In Stage 3 and 5 I had four packets of Skittles a day. One with each meal and one with the almond snack.
- In Stage 3 the Skittles were eaten following each meal/snack.
- In Stage 5 the Skittles were eaten before each meal/snack
The Biggest Revelations
Below are the biggest findings from this experiment that I’ve noticed (so far). There are many more than this which you can discover for yourself in the downloadable Excel file at the bottom of this post.
Does the order in which I eat the Skittles matter for cholesterol? Apparently so!
As with the rest of the experiment, Stages 3 and 5 were the interventions where I ate exactly the same food over the course of the day in three time slots (again, 10 am, 3 pm, and 8 pm) but with the addition of the Skittles. The only real difference between the two stages was the order in which I ate the Skittles.
Here is a comparison of the two stage groupings of 2-3-4 and 4-5-6 so that you can see the interventions side by side. First, the Total Cholesterol:
(Note, all measurements in this second were taken with the PTS Diagnostics CardioChek Plus)
Now, unfortunately, as I had a number or readings where triglycerides (TG) fell below 50 mg/dl (more on that below), I wasn’t able to get a step-by-step accounting for LDL-C. So we’re going to do a comparison of its close cousin, Non-HDL:
All I can say is — wow! These are not inconsequential differences. Same food, same intervention, same times eaten — but that one little change in order had a fairly profound impact.
Naturally, there are already some studies on this very thing and physiologically speaking the order in which the liver will be impacted by the sugar will have measurable impacts on glucose and lipids postprandial. But again, these are effects rippling into the fasting lipid numbers of the next morning by a whole lot.
I actually have quite a bit more to say on this, but that could be a blog post by itself… and I’m a bit pressed for time right now in preparation for Low Carb Breckenridge. So on with the show…
What Blood Markers Changed the Most?
I had a total of four blood draws:
- 2/2 – Last morning of first Washout Stage
- 2/3 – Morning after 24 hours into the first Intervention Stage
- 2/5 – Morning after 72 hours into the first Intervention Stage
- 2/12 – Morning after 72 hours into the second Intervention Stage
Keep these four fresh in your mind as you look at the charts below…
Note: I had actually meant to get a lab done on 2/9 as well to precede the second Intervention Stage, but there was a problem with the lab order that I won’t get into here. Regardless, we have 2/2 serving as our “clean” washout sample along with a lot of historic lab tests I’ve done overall to compare with.
Glucose and Insulin:
So naturally, we aren’t surprised I’d have higher fasting glucose days after bringing in the sugar. This is especially true if you are fat-adapted, by the way. This is why it is generally advised you leave the diet and reintroduce carbs for a few days before taking an Oral Glucose Tolerance test (OGTT) since you will likely fail it if you don’t.
No, the really big news is how different 2/5 is from 2/12. The first is three days of Skittles after the meal, the second is three days before — and boy oh boy is that fasting insulin remarkably different! In fact, this is the highest reading I’ve ever had by far! (My average insulin is usually under 4 uiu/ml when keto)
I had only had my Cortisol checked in two tests, the before and after mornings of the first three days Skittles. Certainly, this marker isn’t one I normally watch very closely, but then, it’s usually very consistently around 10-13ish. Very first time I saw it drop this low.
WBC (White Blood Count):
Another interesting surprise. While I’ve had my WBC dip below range before, my overall average is 4.2, this is certainly unusual.
I also got NMRs in my blood draws for advanced cholesterol testing:
Okay, this one I’m super excited about. A small group of you uber-geeks are familiar with my theories know why I’m calling out the change in HDL-C and triglycerides. But for fun, I’m just going to pass the ball back to you in the comments below. Why, pray tell, would you think HDL-C would go up and triglycerides would come down in the wake of all that added sugar? (Some of you already know my answer)
Does the order matter in personal experience? Absolutely!
Once I decided I was going to add the extra intervention stage, I figured it would pretty much feel like the first one. Nope! That one tiny difference of order made the entire experience very different.
In the first intervention, I ate the fatty meal, then ate the skittles for around 20-30 minutes following. I didn’t feel especially hungry for the Skittles, but I didn’t mind eating them either. It simply felt like it was a simple dessert.
On the second intervention stage, I tried to time eating the skittles for about the same amount of time (20-30 minutes), but this time before the meal. And each time I just wanted to stop right afterward. I looked at my prepared meal with dread as I had no appetite at all for it.
It’s hard to explain just how radically different both stages were in spite of everything else that was intentionally the same.
Here’s the kicker — I had notably more gut pain in the first stage relative to the second one (again, I have Fructose Malabsorbtion, so this was somewhat anticipated during the experiment). This difference could just be my having acclimated to the diet better by the second sprint, or perhaps the order did really matter in this case. I can’t say for sure.
I’m stopping here and just posting the raw data for everyone. Note I don’t have all the numbers entered yet, but I hope to add these at a later point.
Skittles? You ate Skittles? Your dedication to the cause is awe inspiring.
Thanks! Note I’m doing these experiments so you don’t have to! 😉
Does the TG get low because glycogen levels rose telling the body that enough energy is present so that it reduces global fat transport?
And since CETP lowers HDL when TG are high, HDL will go up when TG are low?
PS: Please reveal the answer 😀
Not bad — that’s really close. 😀
Maybe TG is lower because BG is still hanging around in the blood instead of going into TG production. Lower TG will result in higher HDL.
Kinda close. But you and Dominik are on the right track in that my theory involves energy economizing.
Hi Dave, Great experiment, quite a effort to maintain the pattern for this many days! I was just reading a newly published research article in the journal Cell Metabolism regarding processing of dietary fructose that may be relevant to your pre/post meal Skittles intervention. The article demonstrated that when feeding mice a mix of glucose/fructose, the small intestine is the first organ that deals with fructose metabolism (conversion to mostly glucose). Only when the small intestine’s capacity is overloaded does the liver and microbiota of the large intestine get involved. An interesting finding of the study that seems relevant to your latest experiment is the finding that the small intestine’s capacity to metabolism fructose could be increased if feeding preceded the fructose/glucose challenge.
I should note the physiology and anatomy of mice is similar to humans, but the conversion of macronutrient dosages between mice and humans is tricky. The authors perform this conversion based on relative caloric intakes. Using this approach, the limit of the human small intestine to metabolism fructose may be equivalent to only a few ounces of soda (in a simulated fasted state). These are just estimates and would have to be verified experimentally, but interesting nonetheless. If anyone is curious, here is a link that summarizes the key findings: http://www.cell.com/cell-metabolism/abstract/S1550-4131(17)30729-5
Great tip, Todd. Indeed I have read a little on this from before giving my fructose malabsorption problem, but it would be exciting if the small intestine played a much larger role than we first believed as that would almost certainly be impacted by the order of food eaten.
Dave, would you be interested in presenting these data at the Ancestral Health Symposium (AHS18) in July? We’ll be in Bozeman, Montana, which should be gorgeous in the summer. The Call for Proposals is about to go out and it would be fabulous if you’d consider AHS. As you may know, AHS provides online videos of its talks, available to the world for free, so its reach and influence is great. Thank you for your consideration, and please feel free to contact me privately for more information.
Naomi Norwood, Trustee and Program Committee Member
Hi Naomi — thanks for contacting. I’ve responded you on the email address you provided. 🙂
Hi dave.Hope u doing good.I stared keto diet on 21st may 2017.on 13th dec 2017 my cholestrol was 216.triglyceride 49.hdl 100.ldl 144 and on 14th feb 2018 it was 254,53,100 and 97.Are my results ok?my serum calcium was9.3 and then 9.6 .Kindly do reply.thx
It’s not for me to say if it is “ok” per se as I’m not a medical professional. But what I can say is that I’d be very happy with those numbers personally, assuming all other blood markers look good as well. 🙂
Cholesterol figures are in mg/dl.doctor gave me XOEAT 100MG for cholestrol.Should i take it as i dont want to use medicine.usually on LCHF diet cholesterol is high.do reply.thx
Its XOWAT 100MG
I think you’re talking about XOWAT 100mg as a medication, but I’m not familiar with it.
Hi dave.I request you to kindly let me know the REFERANCE RANGE for the following.In pakistan the Labs and doctors use these terms. 1.SERUM CHOLESTROL 2.SERUM TRIGLYCERIDES 3. HDL CHOLESTROL(Direct) 4.LDL CHOLESTEROL(Calc) 5.VLDL CHOLESTEROL(Calc) 6.CHOLESTEROL/HDL Ratio 7.SERUM HDL CHOLESTEROL 8.SERUM LDL CHOLESTEROL I Will be waiting for your reply.Thanks a lot
Reference ranges should be listed on the sheet. It differs per country and it depends on what the units of measure are (most common are mg/dL and mmol/L)
What is meant by TOTAL CHOLESTEROL? Is it HDL+LDL = Total Cholesterol ?? Thanks
No — Total Cholesterol reflects HDL, LDL, and another kind called Remnant Cholesterol. See more here: http://cholesterolcode.com/remnant-cholesterol-what-every-low-carber-should-know/
Fascinating. I have a question. Would you suggest I reintroduce carbs (the thought of it as I’m on a zero carbs diet) prior to a regular glucose blood test?
Why? If it’s a regular glucose test, it is likely you’ll have higher fasting glucose numbers from adding in carbs than you would if you didn’t.
Thanks for the clarification. I just got my A1C result and it’s down to 4.3 from 4.6. What I don’t understand, is why my triglycerides went from 63 to 114 in the test prior to my A1C. I recently switched to strict Keto AIP (meat and fat) and it could be that my body is adjusting
I don’t consider 114 TG that high. It’s the “noisest” marker I track and swings of 40-50% aren’t uncommon depending on physical activity of the previous days and (of course) being water fasted 12-14 hours for the blood draw.
Can you explain the TGL noisy comment? Curious which way you see it change based on physical activity on the proceeding days.
He means it’s the marker to shift the most frequently even with everything else being held steady. I’ve had mine go from upper 60s to 110ish then back down to 80 within a week of each other.
According to his endurance exercise experiment it dropped after exercising, you can also check out his TG readings from his resistance training experiment as well.
After watching your “Incredible Changes in LDL” video, I did a mini experiment, where I increased my carbs while reducing fat for 4 days before testing. Predictably, similar to your results, my TC & LDL decreased, while TG increased,, and HDL kinda stayed within a couple points. I had assumed that with the increased in carbs, there is less need for fat transport since there are more glucose available, the liver produce less LDL, but the TG that ride share with the lipoprotein is not being uptaken by the tissues, hence a higher reading.
Comparing to this experiment, I’d have expected a similar trend… Your results seems to suggest LDL does get impacted by the timing of carb intakes…but, I can’t seem to think of a theory that would explain all 3 scenarios with regards to TG and HDL.
Please do reveal your theory on this because the suspense is killing me… 😀
This experiment is different than the “swap” experiments as it was an *addition* of carbs (mostly sugar) instead. As with some other commenters who only added (not swapped) carbs, total and LDL cholesterol maintained or increased.
In other words, it sounds like you emulated a swap experiment (like the ones I’ve done since May) instead of an addition experiment like this one.
Curious on the WBC count. Have you seen this before? I’ve been LCHF (generally <20 g carb) for almost 2 years. Just got blood checked in February and WBC was 3.6 (range 4.5-11). Doc had me do a follow up in March with a WBC of 4.0. Wondering if this is just me or some side effect of LCHF.
I don’t see it very often. Ironically, I’ve had it happen to me recently in my Added Sugar experiment here: http://cholesterolcode.com/sugar-and-cholesterol-experiment-findings/
I met you at the Gary Taubes event in London a few weeks back. You suggested I could contact you with a query regarding my lipid profile. I am on a LCHF diet for years. I am a 46 yr old healthy female. My cholestoral (fasting for 9 hours) was 10.2, Triglycerides 1.50, HDL 2.23 and LDL 7.3.
As I mentioned when I spoke with you, I’m not worried about the high cholesterol reading, although my GP wants to send me to a specialist lipid clinic. She mentioned my LDL and triglycerides are both high and that this should be cause for concern. I have a family history of high cholesterol readings and both parents on statins, which I am not prepared to do myself.
Hope you are well,
Hi dave.Are my result fine?Kindly do reply.thx. CholesterolCode.com/Report v0.9.2 ====–
… on – ::: …
Total Cholesterol:150 mg/dL3.88 mmol/LLDL Cholesterol:70 mg/dL1.81 mmol/LHDL Cholesterol:50 mg/dL1.29 mmol/LTriglycerides:70 mg/dL0.79 mmol/L
Remnant Cholesterol:30 mg/dL0.78 mmol/L >>> Medium-High Risk QuintileRemnant Chol to HDL:0.6 >>> Medium-High Risk Quintile
Go to https://tinyurl.com/y8hokam2 for more on Cholesterol Remnants
–ATHEROGENIC INDEX OF PLASMA (AIP)–
AIP: -0.213 >>> Lowest Risk Third
Go to https://tinyurl.com/ycccmmnx for more on Atherogenic Index of Plasma
–CONVENTIONAL MARKERS AND RATIOS–
Friedewald LDL-C: 86 | Iranian LDL-C: 79
Total/HDL Ratio: 3
TG/HDL Ratio in mg/dL: 1.4 | in mmol/L: 0.61
Dave is on a bit of a hiatus for a few weeks, but I’ve taken a look meanwhile.
The remnant cholesterol is a little on the higher side (as indicated by the medium-high risk) but your Triglycerides are <100 and HDL is >40 (these are at the recommended levels)
Do you mind if I ask a few follow up questions to get some more context?
1) Are you on any medications?
2) What did you diet look like in the days leading up to the test?
3) How long did you water fast before the test?
4) Did you do any high intensity exercise before the test?
This is just to give us a better view of what your environment looks like, as this can impact what results you get.
[…] But that said, I had already done an experiment that added carbs to a keto diet known as the Added Sugar Experiment. The net effect wasn’t a significant drop in […]
Dave, Curious if you continue to have lower than standard WBC levels that were noted in this post? I have been keto/ low carb for about 3 years and consistently have WBC counts below the “normal” range. I am otherwise healthy, no increased colds or sickness. Thanks!
This is actually something that’s noted a couple times in the Lean Mass Hyper-responder facebook group – that sometimes WBC will be a touch below range. None of us here are doctors, nor can we say whether this is of concern or not, but there is this article from the mayoclinic that says that WBC that’s ‘slightly below the cutoff’ doesn’t necessarily imply concern (but of course this should always be discussed with your doctor). So, perhaps what we’re seeing for those who are just a touch under is just normal variation around the average – but double checking context is something I’d likely want to do anyway.
The link to spreadsheet is not found. Does it have a different location?
Hi Mickey – thanks for pointing out the link isn’t working… the file might have been moved. I’ll ping Dave on it to make sure the link gets updated. 🙂