I tracked everything I ate for 100 days at the start of the year and I learned nothing. Nothing. I wanted to know:

  • Which food/ingredient is it that randomly gives me really bad wind?
  • When do I snack?
  • Am I getting enough protein?
  • Am I getting enough fibre?
  • Which foods/macros affect my sleep or focus?

…and after lots of tracking and a big spreadsheet I find I know nothing. The wind problem was particularly annoying. I have some candidates but nothing clear. I also couldn’t get good enough data on things like ‘sleep well’ or ‘focus’ to have anything like enough information to find evidence.

Basically I eat far too randomly to be able to make inferences. I graze a lot. I eat off the kids plates, I eat different foods at different times (sometimes we eat with the kids at 5:30 pm, sometimes together at 8:30 - sometimes I’ve been randomly snacking all evening). That’s a little worrying. If I was very much a creature of habit that ate the same breakfast, lunch, and dinner every day then I would be able to make some deductions from the random changes but I have nothing. The 100 target days were also when I was injured so it’s not like exercise interferes with the timings (and my exercise is often at random times of the day).

This is frustrating. I was very careful with my tracking - I weighed things and I recorded times, and I went back and checked things and improved my estimates (i.e - sometimes I found I logged 650g of yogurt over four days, but the pot only holds 500g), but it hasn’t given me the answers I wanted.

I also kept track of my weight - and I thought that at least I’d get a handle on my TDEE (Total Daily Energy Expenditure) - I know exactly how many calories I consumed and how my weight changed and thus I can do the maths - but it turns out that if I do it in three 30-day segments I get wildly different values.

So in general I’m quite grumpy.

Here’s what I can tell you.

  • Over the 100 days I recorded 156 different foods. Some of those are ingredients used to then calculate meals (Like, ‘flour’, ‘sugar’ and ‘olive oil’ all have a line in the library and then later on something else is a calculation based on those values), but at the same time some lines (85% dark chocolate, tofu) are covering several different brands and types so ‘about 150 in 100 days’ is probably fair. It strikes me that this is an enormously large number.
  • I’m really suspicious about how my body reacts to: resins/grapes and Love Hearts - but because all three of those are things that I might grab a couple of in passing, I don’t know how well I have recorded them.
  • In general it turns out that my understanding of things like “What has protein, what has fibre, how much caffeine do these things have” and so on had drifted significantly. Part of the point of doing a fairly complete tracking is to be able to say things like “NO wonder that breakfast didn’t fill me up”, so that was a positive

The main thing I think I got from the experience is the idea that stability might be the best thing for me for a while.

Quick FAQ below:

How do you ensure it’s accurate portion sizes?

  • I’m quite nerdy about estimation so I did some fun weighing games about “How does the weight of the veg change when oiled rather than without and how much oil is left in the pan”, and I also was careful about “I ate this packet over four days, so I’d better check that all four days add up to the correct amount” I could have been more accurate (quite often I put in, say, chocolate covered Brazil nuts as Brazil nuts and chocolate rather than create a new entry, but that’s pretty accurate.
  • Various supermarkets give the nutrients for items right on their website. It was endlessly interesting to me how much stables differed from one brand to another.
  • Something I find interesting - I put the calories and the macros (Fat, Protein, and so on) into my spreadsheet and then I have a box that works out the expected calories from the macros. This serves as a check - if there is a difference of more than about 10% I probably typed in a number wrong. But it’s interesting that the numbers aren’t particularly close. The average percentage difference of the (absolute) values was 5% and some where as high as 15%.