Bananas, Porridge, and The Importance of Controlled Experiments

This was meant to be a story about how bad bananas are, and how natural isn’t always better. But it’s not. It’s a story about how bananas are probably ok, and control groups are important.


For a while now, I’ve been getting occasional symptoms of hypoglycemia (low blood sugar). Shakiness, clamminess, irritability, and a fast heartbeat. It’s no big deal, but it’s annoying when it happens in certain situations, like at work.

I noticed that it seemed to happen more often after I ate certain things. I started paying attention to my diet (see this post on a few things I did, like giving up soft drinks) and figured I was probably doing ok. Sugar crashes tend to be caused by overreaction to sharp sugar spikes, and I thought I was doing a pretty good job of avoiding these.

But I was still getting the symptoms.

So maybe I was still doing something wrong?

By the book

There’s only so much you can learn by reading. Despite trawling through tables on glycemic index (and, more to the point, glycemic load) values for all sorts of foods, I was still getting these sugar crashes. Where to go next?

In The Four Hour Body, I read that Tim Ferriss used a continuous glucose monitoring device to keep track of his sugar levels for a while. The one he recommended was ridiculously expensive (£1k+!), but after a bit of research I managed to find a disposable one that would last two weeks for £60.

So I bought one and, with some help, bravely implanted it in my arm. (thanks mum!)

Then for two weeks I kept track of what I ate and how it affected my sugar levels.


Here’s one thing I noticed: when I had my usual breakfast of porridge oats with banana, my blood sugar shot straight up and crashed down quickly afterwards. (for reference, non-diabetics tend to have blood sugar levels of 4–6mmol/l before meals, and 6-8mmol/l a while after eating)

Now I knew bananas weren’t low in sugar, but I didn’t think they’d have that much of an effect! All this time I’d been having that breakfast because I thought it would give me a good mix of quick and slow release energy to keep me going all morning!

I figured it was time to change my ways. But maybe I could do a few comparisons first.

I measured my blood sugar over a few other meals, and it was stable.

I decided to see how a banana compared to pure sugar. Here’s porridge oats with three teaspoons of sugar:

Almost identical to one banana! Crazy, I might as well have been having sugar with my porridge all this time. Right?


On the very last day my sensor was active, it occurred to me that I should do one last breakfast experiment: a control group. Just to see — what would happen if I had just porridge? No banana, no sugar, nothing else. Just porridge.

Two things:

  1. It’s disgusting. Seriously, don’t do it.
  2. Surprise! The effect on my blood sugar was the same.

Here’s pure porridge (oats + semi skimmed milk):

Spike, and crash! Like the other two cases. So my conclusions were wrong!


It turns out that the assumption I’d made at the beginning was wrong*.

I’d been reading too many food and nutrition blogs, and took them at their word when they told me that “oats are a source of slow-release energy”. It turns out they’re not**.

I’d somehow selectively ignored the data that said that oats are, actually, quite high GL. One serving of oats is about halfway between a banana and a can of coke, in terms of its impact on your blood sugar.

What did I learn?

Sometimes you need to trust the way you feel over what you’re reading. Oats definitely won’t be my go-to breakfast anymore.

I also learnt that it’s easy to make conclusions based on very little data. It’s important to conduct experiments designed to disprove your hypothesis, not just to support it!

And finally… if you’re one of the people I spoke to after my first two bowls of porridge, I take back everything I said about bananas! (probably*)

* As someone pointed out to me after reading this, technically I never controlled for the milk in my porridge, which also contains some sugar. My sensor has expired now, but if I had another chance I would do another test with just oats and water. And one where I just eat a banana, to see what that does.

** Someone else pointed out that there might be a difference between different types of oats. As I was eating Tesco’s cheapest oats, it might be worth re-doing this experiment to control for that. 

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s