In this LAB note
Question · Hypothesis · Method · Results · What Failed · What We Learned · What Changed at MOMS
Question
When tofu sits in a shake's protein slot, what does it change on the label: the protein, or something else?
Hypothesis
We expected tofu to behave like a protein ingredient. It sits in the protein slot, so shakes that contain it should show more protein than shakes that do not, and recipes with more tofu should show more protein than recipes with less. We also expected the effect to be small. Tofu is rarely the only source in the slot, and a published review of soy protein quality ranks tofu at the low end of the soy foods it examined (see Method).
Method
Published research. van den Berg and colleagues pooled published digestibility-based protein quality scores for soy products and reported mean scores for tofu of 63.7 (DIAAS) and 78.9 (PDCAAS), the lowest of the soy product groups reviewed. They attribute this to the coagulation step raising endogenous ileal amino acid loss. Mariotti and Gardner reviewed protein in vegetarian diets and report that measured true digestibility differs between plant and animal protein sources by only a few percent, and that concern about plant protein quality has been overstated for varied diets. Our label numbers are grams, not quality-adjusted grams, so this analysis asks only what tofu does to the printed protein figure. Internal analysis. We parsed the Protein slot of all 201 recipes in our recipe database and flagged the 49 that include tofu (firm or silken); the other 152 are the comparison group. We compared protein (g), fat (g) and calories, and we parsed tofu grams as a share of total protein-slot grams. Tofu appears in six of our eight body types and in none of the other two, so the main comparison is made within each body type and weighted by the number of tofu shakes in it. Evidence recipes: WY-W1-DETOX-010-V1, WY-W1-DETOX-014-V1, MI-W2-IMMUNE-005-V1 and WAY-W1-HYDRATE-011-V1.
Results
Unadjusted, the 49 tofu shakes average 33.4 g protein, 11.0 g fat and 373 kcal. The 152 others average 32.9 g protein, 9.9 g fat and 376 kcal: a 0.5 g protein gap and a 1.1 g fat gap. The dose is small and never leads the slot. Tofu ranges from 18 to 70 g per recipe, 48 of the 49 recipes use 50 g or less, and tofu averages 22% of the protein slot by weight (range 7% to 45.5%). Within body type, weighted by tofu-shake count, tofu shakes carry +0.8 g protein, +2.1 g fat and +1.0 kcal. Fat is higher in all six body types that contain tofu. Protein is higher in all six too, but by 1.2 g or less in five of them; the exception is a two-recipe group (EI) at +3.1 g. Across the 49 tofu shakes, tofu grams correlate with protein at r = 0.13 and with fat at r = 0.44. A matched pair from the same body type, week and goal: WY-W1-DETOX-010-V1 (Tofu 50 g + Chickpeas 32.5 g + Greek Yogurt 136.1 g) has 32.3 g protein, 13.9 g fat and 391.4 kcal. WY-W1-DETOX-014-V1 (Kidney Beans 35 g + Greek Yogurt 180 g) has 32.0 g protein, 6.2 g fat and 319.7 kcal. Protein is 0.3 g apart; fat is 7.7 g apart.
What Failed
The fat effect is not universal. MI-W2-IMMUNE-005-V1 (Chickpeas 40 g + Tofu 30 g + Egg White 240 g) has 36.5 g protein and only 5.9 g fat, the lowest fat of all 49 tofu shakes. The fat gap is 0.5 g in the MI group and 0.03 g in the two-recipe EI group, and egg-white-based tofu shakes in general show little added fat. Tofu is also not what produces our fattiest shakes. The highest-fat recipe in the sheet, WAY-W1-HYDRATE-011-V1 (Chicken 80 g + Egg 28 g), has 22.0 g fat and no tofu. Co-sources such as egg and Greek yogurt may carry part of the fat we are attributing to tofu, and the matched pair above is only one comparison. The calorie direction is inconsistent: tofu shakes are +6.6 kcal in WY but -7.3 kcal in MY and -6.4 kcal in WAY. Controlling for body type does not control for week, goal or the other slots, so this is an association in a designed recipe set, not an effect of tofu. We also did not adjust label protein for digestibility, and we are not claiming the printed numbers are wrong.
What We Learned
In our recipes tofu is a co-source, not a lead source. Its measurable footprint on the label is fat (+2.1 g within body type) more than protein (+0.8 g). The protein figure of a tofu shake is set mostly by what sits beside the tofu, which is about four-fifths of the slot by weight.
What Changed at MOMS
We no longer describe tofu as a protein booster in recipe notes. When a draft uses tofu in the protein slot, we size the protein target on the other sources first, then read tofu's fat and calorie change alongside it. Recipe review now lists tofu grams next to the fat and calories they travel with.
References
van den Berg LA, Mes JJ, Mensink M, Wanders AJ. Protein quality of soy and the effect of processing: A quantitative review. Frontiers in Nutrition. 2022;9:1004754. | Mariotti F, Gardner CD. Dietary Protein and Amino Acids in Vegetarian Diets—A Review. Nutrients. 2019;11(11):2661.
Keep reading
Previous experiment
Not sure which shake fits you? We match the formulation to your body type.
Find your body type