Self-experiment: 7 day carb test

Guinea pig time! This n=1 experiment comes from Wired to Eat, Robb Wolf latest (and greatest IMO) piece of work. Robb Wolf is one of the most respected voices in the paleo/ancestral scene not only because he was one of the early adopters, but because he gets science, both at an academic level (he is a biochemist) and at a philosophic level (he is not afraid of changing his views when new evidence is available, which is the case with this book).

I encourage you to listen to a few of the many podcasts Robb has been interviewed in, so that you get an idea of what his book’s message is. I will just offer a very brief summary before presenting my results. After publishing his first book, The Paleo Solution, Robb realised that the paleo prescription as a blanket recommendation was not as effective as a more personalised approach. Studies like Personalized Nutrition by Prediction of Glycemic Responses, which showed wild variations in the blood sugar of people after eating the same foods. That’s why Robb recommends a 7 day carb test (following by a 30 day reset if you’re not currently eating a basic unprocessed paleo-style diet) to find out your individual tolerance to carbohydrate-containing foods.

You may be familiar with this test if you have ever done an oral glucose tolerance test (OGTT) or have served as a volunteer to find out the glycemic index of a particular food. Glycemic index is a measure of the degree that food rises blood sugar. The idea is that you eat a portion of the food (when I did it in uni was either white sandwich bread or sweetened yoghurt) that contains 50g of effective carbohydrate (i.e. total carbohydrate minus fibre) first thing in the morning (this will be your breakfast) and measure your blood glucose 2 hours later (aka 2 hour post-prandial blood glucose). Ideally, your body will release the insulin required to get the sugar out of the blood and into the cells, so that your blood sugar level will be on its way to normal (anywhere between 5.0 and 6.4 mmol/L are Robb’s recommendations based on clinical experience). If your blood sugar is higher than that, Robb recommends halving the portion so that it gives you 25g effective carbohydrate and testing another day, so that you find out the amount of that particular food you can deal with. If your blood sugar is again too high, you’d better stay away from that particular food.

So what do you need for the test?

  1. I recommend buying the book and reading it beforehand so that you get all the background and detailed instructions/handholding if required.
  2. Clean up your diet if you haven’t done so yet.
  3. Get a blood glucose monitor (aka glucometer). I went to Accu Chek’s website and found out they offer a monitor for free! You need to fill out your details so that they can suggest a model to suit your needs and send you the monitor. The catch is that you need to buy the lancets and test strips at a chemist (they are cheaper if you have a diabetes card).

    Glucometer

    Glucometer

  4. Get a kitchen scale so that you can get accurate portion sizes.
  5. Decide which foods you want to test. These have to be foods that are mainly carbohydrate, such as rice, potatoes, oats, corn, beans, bread, pasta, fruit, etc. Most importantly, these foods should be relevant to you, there’s no point in testing rice if you never eat it nor want to, or testing regular bread if you are gluten intolerant.
  6. Use the book to find out portion sizes of your test foods or calculate them using nutrition panels/databases. If you need help with this, leave a comment on this post.
  7. Start testing! For each day write down as a minimum: the food you tested, the portion size and your blood glucose after 2 hours. Robb included a more comprehensive workbook as a bonus when preordering his book, but if you don’t have it, you want to pay attention to how you feel after eating that particular food. This can give you an indication of how your body is dealing with it.

I decided to test my fasting (i.e. before eating) blood glucose as well as the 2 hour post-prandial just because I wanted to measure the difference between both and also pick up on any confounders (i.e. did I start the day with elevated blood sugar and it wasn’t entirely the foods fault?). You don’t have to do this. I have also included photos of all the foods I tested so that you can see what the portion sizes looked like. Turns out that eating a big whack of carbs with no seasoning or fat is not as enjoyable. I felt bloated and gassy with all the foods in varying degrees, but didn’t feel particularly horrible after any of them.

Day Food Weight (g) Fasting BGL (mmol/L) 2-hr BGL (mmol/L) Delta (mmol/L)
1 Potato, sebago, boiled 457 5.1 7.3 2.2
2 Sweet potato, orange, baked, peeled 290 4.8 5.4 0.6
3 Rice, basmati, boiled 216 5.7 6.2 0.5
4 Steel cut oats 90* 5.3 5.8 0.5
5 Rolled oats 87* 4.7 5.5 0.8
6 Gluten-free bread 139 4.7 4.7 0.0
7 Potato, sebago, boiled 229 5.1 5.4 0.3
8** Hot chips 177 4.9 6.2 1.3
9** Potato with avocado oil mayonnaise 396 5.0 6.1 1.1

* For the oats I measured the dry (uncooked) weight because I calculated the amount using the nutrition panel.

** Days 8 and 9 were bonus days. My friend Timmy was curious about the hot chips, so I did the test. They didn’t spike my blood sugar as bad as the plain potatoes because fat lowers the glycemic index of foods. Then my friend Sandy wondered how the hot chips would compare to cooked and cooled potatoes served with homemade avocado oil mayonnaise. Despite this option being a lot healthier and tastier, there was no real difference in glycemic response. Having said that, it’s important to note that the blood sugar rise is not the only thing to look at. Are Maccas hot chips healthier than boiled potatoes? Hell no!, for a multitude of reasons. Let’s not get blinded by a single number and always look at the full picture.

So what did I learn from my tests? One, that I had a pretty decent response to most of the foods I tested except for potatoes. Alvaro had a pretty good response to them (5.9 mmol/L) and a crappy response to rice (7.4 mmol/L), which was just fine for me. Interesting results if you consider that we should have more or less the same response given that we’re both Peruvian… BUT! I’m 1/2 Japanese and he’s only 1/8 Chinese. So it looks like my carbohydrate metabolism genes are more Asian than Peruvian and his are more Peruvian than Asian.

Alvaro’s only hurdle was rice, all other measurements were between 5.0 and 5.9. I found this interesting given that his DNAFit report said his carbohydrate tolerance is moderate and he should shoot for lower GI foods, while mine said my tolerance is high and I can get away with a bit more of the high GI foods.

Finally, I found it hilarious that my blood sugar after the gluten-free bread was stellar. I stopped eating bread in 2011 and have eaten gluten-free bread only occasionally – I’d say once a fortnight on average, mainly when eating out. Will I start eating gluten-free bread with reckless abandon? I don’t think so. I still prefer eating vegetables both for taste and health reasons. But I won’t stress too much and let the gluten-free bread happen when it happens.

Potato

Potato

Sweet potato

Sweet potato

Basmati rice

Basmati rice

Steel cut oats

Steel cut oats

Rolled oats

Rolled oats

Gluten free bread

Gluten free bread

Potato (half dose)

Potato (half dose)

Hot chips

Hot chips

Potato and avocado oil mayonnaise

Potato and avocado oil mayonnaise

If you’re interested in Robb’s new book you can find it here: Wired to Eat.