Once nuance is that they seem to have grouped together all ultra-processed plant foods into one category, everything from tomato sauce to tofu to sodas (see supplement: https://www.thelancet.com/journals/lanepe/article/PIIS2666-7...). The nutritional profile of tofu and soda are quite different, but the study measures total consumption in calories, so sodas are going to heavily skew that number.
I wonder if that makes enough of a difference (e.g. is someone who eats a lot of ultraprocessed noodles with tofu and tomato sauce going to have similar cardiovascular outcomes as someone who chugged a 2L of coke a day?)
I'm someone who's been vegan for 15+ years, but lately I shifted about 50%-70% of my meals to a super-artificial ultra-processed food product (https://huel.com/products/huel-instant-meal-pots) that is supposed to "healthy" in that it contains all the macro and micro nutrients a person supposedly needs, in a relatively tasty and satiating 400kcal meal. I wonder if that would still have the same risk as drinking 400kcal of soda or eating potato chips? Heh, maybe it'll save me from diabetes only to kill me from heart disease. Alas.
They consider tofu as ultra-processed food? That doesn't make much sense - it's just boiled soybeans which are then ground and mixed with coagulants (such as magnesium chloride, which you can get from seawater). It's about as much "processed" as cheese.
Hmm. I can't tell whether this is a mistake, disagreement, or ambiguity in the NOVA classification system that the paper claims to use. Their Appendix A, Table S1 clearly lists as ultra-processed foods: "Meat alternatives: Vegetarian sausage/burger; Tofu/tempeh/TVP/soya mince; Quorn". I would think all those things are quite different to each other, but not to this study's classifications.
I was unable to find an authoritative source on what the NOVA classification of tofu actually is, or how the NOVA classification system methodology even works. Secondary and tertiary sources on the internet disagree with each other, and ChatGPT cites Brazilian institutions like "NUPENS" whose websites are on Google but seem to be either dead or inaccessible from the US.
Edit: There seems to be a third-party database, https://world.openfoodfacts.org/cgi/search.pl?search_terms=t..., that lists many tofus in category 3, "processed" (because they have salt and oil added) or 4, "ultraprocessed" (because of gelling agents).
-------------
But even excluding tofu, they also consider things like bread (yeast and wheat?) to be ultra-processed, or potato wedges, or fruit juices, or canned vegetable soups, along with ketchup.
To be sure, many of those CAN be commercially processed, but I think this particular study might be trying a bit too hard to lump together a bunch of ambiguous foods in order to try to find power in their stats? It's hard to trust a dietary analysis that treats sugary sodas the same as pasta sauce or whole-grain bread...
Comments
This is super interesting. Thanks for sharing!
Once nuance is that they seem to have grouped together all ultra-processed plant foods into one category, everything from tomato sauce to tofu to sodas (see supplement: https://www.thelancet.com/journals/lanepe/article/PIIS2666-7...). The nutritional profile of tofu and soda are quite different, but the study measures total consumption in calories, so sodas are going to heavily skew that number.
I wonder if that makes enough of a difference (e.g. is someone who eats a lot of ultraprocessed noodles with tofu and tomato sauce going to have similar cardiovascular outcomes as someone who chugged a 2L of coke a day?)
I'm someone who's been vegan for 15+ years, but lately I shifted about 50%-70% of my meals to a super-artificial ultra-processed food product (https://huel.com/products/huel-instant-meal-pots) that is supposed to "healthy" in that it contains all the macro and micro nutrients a person supposedly needs, in a relatively tasty and satiating 400kcal meal. I wonder if that would still have the same risk as drinking 400kcal of soda or eating potato chips? Heh, maybe it'll save me from diabetes only to kill me from heart disease. Alas.
They consider tofu as ultra-processed food? That doesn't make much sense - it's just boiled soybeans which are then ground and mixed with coagulants (such as magnesium chloride, which you can get from seawater). It's about as much "processed" as cheese.
Hmm. I can't tell whether this is a mistake, disagreement, or ambiguity in the NOVA classification system that the paper claims to use. Their Appendix A, Table S1 clearly lists as ultra-processed foods: "Meat alternatives: Vegetarian sausage/burger; Tofu/tempeh/TVP/soya mince; Quorn". I would think all those things are quite different to each other, but not to this study's classifications.
I was unable to find an authoritative source on what the NOVA classification of tofu actually is, or how the NOVA classification system methodology even works. Secondary and tertiary sources on the internet disagree with each other, and ChatGPT cites Brazilian institutions like "NUPENS" whose websites are on Google but seem to be either dead or inaccessible from the US.
Edit: There seems to be a third-party database, https://world.openfoodfacts.org/cgi/search.pl?search_terms=t..., that lists many tofus in category 3, "processed" (because they have salt and oil added) or 4, "ultraprocessed" (because of gelling agents).
-------------
But even excluding tofu, they also consider things like bread (yeast and wheat?) to be ultra-processed, or potato wedges, or fruit juices, or canned vegetable soups, along with ketchup.
To be sure, many of those CAN be commercially processed, but I think this particular study might be trying a bit too hard to lump together a bunch of ambiguous foods in order to try to find power in their stats? It's hard to trust a dietary analysis that treats sugary sodas the same as pasta sauce or whole-grain bread...