In this LAB note
Question · Hypothesis · Method · Results · What Failed · What We Learned · What Changed at MOMS
Question
Raw sweet potato carries about three times the carbohydrate of raw pumpkin per gram. So why do our 10 sweet-potato shakes carry only 2.3 grams more carbohydrate than our 12 pumpkin shakes, when each group uses roughly 50 to 56 grams of the vegetable?
Hypothesis
We expected a gap close to the arithmetic. USDA lists raw pumpkin at 6.5 g of carbohydrate per 100 g and raw sweet potato at 20.1 g per 100 g. At about 50 g of pumpkin and 56 g of sweet potato, that is roughly 3.2 g versus 11.3 g from the Body slot alone, a gap of about 8 g. We expected the finished shakes to show most of it.
Method
Published research: Hätönen et al. (2011) fed 11 healthy adults mashed potato alone and with chicken breast, rapeseed oil and salad. Mashed potato alone had a glycaemic index of 108. Combined with chicken breast, rapeseed oil and salad it was 54, against a predicted 103 based on the individual components. The lesson we took from it: a single ingredient does not predict the response of the meal around it. That study measured blood glucose, not label carbohydrate, so we use it only as a reason to check the rest of the recipe, not as a mechanism. Internal analysis: from our 201-recipe database we kept only shakes whose Body slot holds exactly one ingredient, either pumpkin (12 shakes) or sweet potato (10 shakes). We compared their listed macros, then rebuilt each shake's carbohydrate slot by slot using approximate USDA per-gram values for every ingredient. Those slot-level figures are our own estimates, not label numbers. Two matched examples at 50 g: EI-W1-CLEANSE-001 (Quiet Cleanse; Pumpkin 50 g, Carrot 42.5 g, Pineapple 20 g, Chickpeas 30 g, Lentils 30 g, Egg White 240 g) and WI-W1-VITAL-002 (Electric Bloom; Sweet Potato 50 g, Blueberry 35 g, Green Apple 22.5 g, Lentils 30 g, Greek Yogurt 203 g).
Results
Pumpkin-body shakes use 49.9 g of pumpkin on average and list 32.8 g of carbohydrate. Sweet-potato-body shakes use 56.3 g and list 35.1 g. The gap is 2.3 g, not the roughly 8 g the Body slot alone predicts. Calories are 356 versus 381, protein 32.2 g versus 35.0 g, fat 11.2 g versus 11.6 g, and fiber 6.7 g versus 6.9 g. The estimated slot-by-slot breakdown shows where the missing carbohydrate went. Body slot: about 3.2 g (pumpkin) versus 11.3 g (sweet potato). Fruit and Vegetable slot: about 9.2 g versus 12.4 g. Liquid: about 3.8 g versus 4.4 g. Protein slot: about 13.1 g versus 5.7 g. The protein slot runs 7.4 g in the opposite direction, almost cancelling the produce-side gap of about 11.9 g. The reason is who gets which protein. Seven of 12 pumpkin shakes use Greek yogurt and only 1 of 12 uses chicken, which averages 5 g across the group. Three of 10 sweet-potato shakes use yogurt and 7 of 10 use chicken, which averages 60 g across the group. Chicken adds essentially no carbohydrate; yogurt and legumes add some. Body type is tangled into this. Pumpkin appears in EY, EI, WY and WI shakes; sweet potato appears in WY, WI, WAY and WAI shakes. Restricting pumpkin to its 7 WY and WI shakes, the carbohydrate gap shrinks to 0.4 g (34.7 g versus 35.1 g).
What Failed
The result is small-sample and confounded. Twenty-two shakes cannot separate the vegetable from the body type and protein choices that travel with it, and we did not run a controlled swap. The carbohydrate ranges overlap heavily: pumpkin shakes span 22.7 g to 52.6 g and sweet-potato shakes span 27.8 g to 44.5 g, so the 2.3 g average gap is smaller than the spread inside either group. The counterexample is in the data itself. EY-W2-DETOX-007 (Pure Ease; Soy Milk 150 ml, Pumpkin 45 g, Carrot 55 g, Egg 27.5 g, Chickpeas 60 g, Egg White 150 g) lists 37.6 g of carbohydrate. WAI-W3-NUTR-001 (Antioxidant Glow; Soy Milk 195 ml, Sweet Potato 58.5 g, Apple 55 g, Carrot 25 g, Chicken 90 g) lists 27.8 g, despite carrying 13.5 g more of the starchier vegetable. A pumpkin shake out-carbed a sweet-potato shake by nearly 10 g. Our slot-level estimates also miss: they total 29.3 g for the average pumpkin shake against a listed 32.8 g, so they are good for direction, not for decimals. We also did not test what drives the internal Score here. Pumpkin shakes average 14.75 and sweet-potato shakes 15.90, but we do not claim that difference comes from the carbohydrate swap.
What We Learned
A per-gram ingredient fact is a good first guess and a poor forecast. In a finished shake the carbohydrate is the sum of four slots, and the slot you did not touch can cancel the one you did. Swapping a high-carbohydrate Body ingredient for a low one only lowers the label if the protein, fruit and liquid slots hold still.
What Changed at MOMS
When we substitute one Body ingredient for another to move carbohydrate, we now check the protein slot first, because it can quietly refill what the swap removed. In review, we compare the full slot-by-slot carbohydrate of the old and new versions instead of relying on the per-gram figure of the swapped ingredient. Our estimates stay internal checks; the listed macros remain the only numbers on the label.
References
Hätönen KA, Virtamo J, Eriksson JG, Sinkko HK, Sundvall JE, Valsta LM. Protein and fat modify the glycaemic and insulinaemic responses to a mashed potato-based meal. Br J Nutr. 2011;106(2):248-253. | U.S. Department of Agriculture, Agricultural Research Service. FoodData Central, SR Legacy: Pumpkin, raw (FDC ID 168448); Sweet potato, raw, unprepared (FDC ID 168482).
Keep reading
Previous experiment
Not sure which shake fits you? We match the formulation to your body type.
Find your body type