norswap

aka Nicolas Laurent

The Big Nutrient Spreadsheet

28 Jan 2020

As part of my journey into weightlifting (1, 2, 3, 4), I've been increasingly interested in nutrition.

I thought to push this a bit further and try to figure out which nutrients I'm getting in my diet, which I'm lacking, and how I can remedy that. I didn't quite get around to that last part, but here are my results so far, as well as the tech I built to help me (spoiler: it's a spreadsheet).

(Essential) Nutrients

What does our body actually need?

Additionally, some substances can be synthesized by our bodies but can be (should be?) ingested directly, for instance choline.

Some other substances are non-essential but you really want to have them, most notably fibers.

And finally, yet other substance are non-essential but might have beneficial effects, such as various antioxidants.

And, it goes without saying, adequate calorie intake and hydration.

In this article I'm going to focus on the vitamins and the minerals

Why? First, it's more likely to have deficiency in those. It's also much harder to reach the recommended daily intake values. Nevertheless, I've included some remarks on fats in an Appendix 1.

The Nutrient Spreadsheet

The goal was simple. First, to make a spreadsheet were each row was a food, and each column was a nutrient. Second, to have a way to autofill the nutrient values from some online database to avoid tediously copying ~30 values into a spreadsheet for each food item I was interested in.

You can view the resulting spreadsheet with some food items here.

I used the same setup as when I made a Fitbit spreadsheet: there is an Observable notebook with controls that help filling the spreadsheet.

In particular, you get a food item you're interested in from Food Data Central (or FDC for the hip kids — it's the best nutrient database I could find, maintained by the US Department of Agriculture) and then instruct the notebook to export the nutrient values to a row in the spreadsheet.

Unfortunately, FDC does not track every essential nutrients. I reviewed the essential nutrients it does not cover, to make sure nothing too important was left out — the result of that analysis is in Appendix 2 (It's quite interesting, I might extend the analysis to all essential nutrients sometime.) Basically, you should be fine for all the nutrients not covered, especially if you drink milk regularly. If not, maybe just check your iodine intake.

If you want to make your own spreadsheet, it's also possible! Fork this Observable notebook and follow the instructions under the "Setup" heading to create your own (for posterity, the code is also archived here).

Misc Remarks: Nutrients & Bio-Availability

An obvious question about this kind of analyzis is: are nutrients the only thing that matters? To which I'd tend to answer "mostly, but".

First, some non-essential nutrients are quite important (such as fibers, as we mentionned before). And because we can subsist on essential nutrients doesn't mean we can't benefit from non-essential nutrients, or even non-nutrients such as antioxidants.

Second, the co-occurence of nutrients and other substances matters a great deal. In particular, it determines bioavailability: how much of the nutrients present in food we can actually absorb, as well as how fast they can be absorbed (the two being sometimes interrelated). For instance, liquid foods will be absorbed much faster.

This is a factor to keep in mind when considering supplements, though a bit of careful googling will help you a great deal. [For instance][curcumine], you'll learn never to buy a curcumin supplement that doesn't also include piperine or some lipids, because otherwise you won't actually absorb most of it.

(An important point to make is that you don't actually absorb all the nutrients in real food either, and the intake recommendations do take this into account.)

In general, it's hard to know what matters and how (much) it will impact health. I belabored this point in my article about red meat. If you have no deficiencies, do not consume lots of well-known harmful substances (e.g. trans fats) and keep your bodyfat in check, it's going to be very hard to ascertain if (for instance) eating whole foods is going to be much better than drinking Soylent in the long run.

And that's it for today!

I still have to actually analyze what nutrients I am getting and not getting and how I could fix the situation. I'm hoping to report on that at some point, but it's not currently super high in my list of priorities.

Appendix 1 - Notes on (The Many Kinds of) Fats

Disclaimer: These are just compiled notes from my own googling and wikipedia'ing. They mostly reflect "the generally accepted idea" — I wouldn't even go as far as to say "the scientific consensus", which I haven't research (because it's harder).

There is a complex hierarchy of different kinds of fat:

Each category further includes multiple kinds of fatty acids.

Despite this variety, only two kinds of fatty acids are essential: alpha-linolenic acid (ALA, an Omega-3) and linoleic acid (LA, an Omega-6 — note the missing "n" when compared to ALA). We can synthesize the rest from those two.

The consensus regarding fat is that the unsaturated fats are the "healthy" fats. Saturated fats are not inherently bad, but increased consumption is linked with heart disease (though the studies are sometimes contradictory). Trans fats are mostly artificial, and those artificial trans fats should be avoided as their consumption is strongly linked to heart disease.

In general, Omega-3 are regarded as having a variety of health benefits, though (a) a lot of these claims are shaky, and consuming a lot of omega-3 does not seem to procure additional advantages.

Besides ALA, the other two kinds of Omega-3 deserve a mention, namely EPA and DHA (I'll spare you the full names). Both can be synthesized from ALA though at low rates (8%) and the ability decreases with age, so it might be a good idea to get them from your diet.

There is also some talk of optimal ratios between Omega-3 and Omega-6, though without much substantiation. A typical western diet has 14-25 more Omega-6 compared to Omega-3, while some people advocate a 4:1 or even 1:1 ratio.

The 1:1 ratio seems crazy to me, as only some fishes and seafood (such as salmon and mackerel) as well as some seeds (chia, flax) have more Omega-3 than Omega-6. There are a couple others (big list here).

To top it off, it's quite hard to find precise ratios for animal sources, because it depends on nutrition. For instance, beef will vary between a 2:1 ratio for grass-fed beef to 6:1 for grain-fed beef (I also found the alternatives figures of 1:1 and 4:1, respectively).

In Belgium at least, finding what your beef is being fed is also quite hard. I researched nutrition for the "Belgian blue bull" breed, and they seem to be fed mostly beetroot and hay, depending on the seasons. This opens even more questions: what is the nutritive impact of beetroots? Of hay ("grass-fed" beef is allowed hay, but is it different from the stuff the cows graze from the ground?).

As a side note, while finding omega-3 and omega-6 values is hard, it's almost impossible to find usable nutritive values for omega-7 and omega-9.

It's probably a good idea to get more Omega-3, but there is scant evidence and little urgence here.

Finally, we can talk about Cholesterol. Cholesterol is a lipid, but is not a "fat" (which are chemically triglycerides).

There's HDL cholesterol ("good cholesterol") and LDL cholesterol ("bad cholesterol", whose excess is likely to cause arterial narrowing). Like omega-3/6, the key here might be in the ration between both. It also seems that omega-3 and omega-7 could help with lowering LDL / increasing HDL.

Again, I haven't researched this. Someone on the LessWrong slack made the case that high LDL wasn't bad if other markers (like omega-3/6 ratio) were okay. Just googling for it you'll find this:

Consistent evidence from numerous and multiple different types of clinical and genetic studies unequivocally establishes that LDL causes atherosclerotic cardiovascular disease.

and that:

High LDL-C is inversely associated with mortality in most people over 60 years. This finding is inconsistent with the cholesterol hypothesis (ie, that cholesterol, particularly LDL-C, is inherently atherogenic). [...]

At least, this suggests to me that maybe the story is not as simple as "LDL = bad". Other factors may be at play, and correlations may pollute the studies.

Appendix 2 - Vitamins & Minerals Not Covered by FDC

FDC tracks most of the essential nutrients (and then some), but not all. Let's talk quickly about these forgotten children.

Chloride (Cl)

Yes, it's the stuff they put in swimming pool, but it's also about 60% of what table salt (NaCl) is made of by weight. The recommended daily intake is 2.3g and you probably get that by salt only.

Chromium (Cr)

The recommended daily intake is 35 mcg (or μg — a microgram). The estimated intake for European adults is 57-83 mcg/day while another study found that the average for Mexican adults was 30 mcg/day. It's hard to get accurate estimate of this element for food items because it depends very strongly on the soil used to grow plants and the feed given to animals. Anyhow, your intake of this mineral should be fine.

Fluoride (F-)

Yes, that's the stuff they put in toothpaste (or sometimes in the water supply in the US). It's good for your teeth and bones... and after a cursory search I couldn't really find what else it's good for. The recommended daily intake is 4 mg, and toothpaste will have at least 1000 ppm (part per million) of it — hence about 1 mg of fluoride per gram of toothpaste. Black tea has about 0.8 mg/cup. It's otherwise pretty rare and maybe that's good because this stuff isn't good in excess. Just wash your teeth at least once a day and you'll be fine.

Iodine (I−)

The recommended daily intake is 150 mcg. Salmon has 28 mcg/100g. Seaweed contains tons of it — kombu kelp up to 3k mcg/g, while wakame seaweed (the neon green one you can get on the side at the Japanese) contains 40-60 mcg/g. Yes, those figure are per gram. Milk also has a lot, about 250 mcg per liter. Iodine is needed by the thyroid (at least 70 mcg to avoid problems).

Molybdenum (Mo)

Average daily intake varies between 120 and 240 mcg/day, which is higher than dietary recommendation of 45 mcg, so don't worry too much about it. It's in small amount (5-10 mcg) in many, many things, but there's a ton in liver (150 mcg / 100g) and a lot in cereals, bread, rice & almonds (20-30 mcg/100g).

Vitamin B12 (Cobalamine)

The recommended daily intake is 2.4 mcgg. Milk has 0.5 mcg/100mL. It's also found in an energy drink I favor.