Pyramid Schemes

One of the first government graphic on recommended food groups was published by the Swedish government in 1972. It featured a seven slice pizza-shaped graphic, including three separate groups for vegetables (leafy/green/yellow, citrus fruit/tomatoes/raw cabbage, and potatoes/other fruit), a milk/cheese/ice cream group, and a group just for butter/fortified margarine! The Swedes apparently found this confusing, and two years later the Kooperativa Förbundet (a retail chain) released the first food pyramid, to better represent how much of each food group should be eaten, like not eating butter in equal portions to whole grains. The Swedish government never accepted the commercial version1.

Sweden’s Basic Seven food groups

The decision of what kinds of foods to fill your plate is a key factor for determining how healthy your diet is. Unfortunately, in the plans presented by governments and companies, science often comes second to the preferences from their sponsors and stakeholders. This is why the US Department of Agriculture recommendations feature milk as its own category, despite questionable health impacts and other easy protein substitutes.

In 2017, the American Journal of Clinical Nutrition (AJCN)2 released a systematic review and meta-analysis of the studies that have been published on how much of every food group (they didn’t provide butter/margarine as its own food group, though) and how that affects all-cause mortality. I consider all-cause mortality to be the gold-standard of whether something is healthy. The body is complex, and it can be easy, especially when working pieces and parts of food in a component model, to show that something is good or bad for you because of this one part of it that does one thing to the body. Often these studies about this chemical being good or bad are often disproven later, when they extract the chemical, and do a randomized, controlled trial. In comparison, death as an endpoint is pretty unambiguously bad. I’m fairly certain that later studies will not find it is actually good for you. 😉

The AJCN study provides a solid baseline of how each food group affects your health. In my previous post, I talked about how nuts are good for you, but only up to 20 grams. Many other healthy foods also show limits to their benefits. Vegetables provide longevity benefits up to 300 g per day. After this, no additional benefit. Fruits improve health up to 250 g per day, but no extra credit is available for eating more.

Several foods show benefits that keep growing all the way through the testing range, so we don’t know the “maximum beneficial dose” because not enough people eat enough to show statistical results for them. Beans (150+ g/day), whole grains(150+ g/day), and fish (100+ g/day) all show this pattern.

Other foods increase all-cause mortality, and also have no upper limit to how much harm they can do from this study. Red meat, processed meat (even worse), eggs, and sugar-sweetened beverages all show this pattern.

Milk and dairy are in their own category. No harm (or benefit) up to 750 ml/day, but continued increased mortality after that point.

Refined grains also stand out as unique. No increase or decrease shown at any amount per day, up to the tested amount of 150 grams.

I haven’t made any pyramids yet or plate diagram to show how to get this amount of food, but I have made a calculator based on US actuarial tables to calculate how your diet affects your effective real age. Give it a try!


  1. Wikpedia Food Pyramid nutrition
  2. Food groups and risk of all-cause mortality: a systematic review and meta-analysis of prospective studies