In this LAB note
Question · Hypothesis · Method · Results · What Failed · What We Learned · What Changed at MOMS
Question
Fifty-two of our 201 recipes leave the Body slot — the starchy, structural ingredient — completely empty. If the thickener is gone, what is carrying its job, and what does the finished shake actually lose?
Hypothesis
The Body slot holds the dense, low-water ingredient: banana, avocado, cauliflower, pumpkin, sweet potato, white rice, taro. Across the 149 recipes that use it, it averages 56.2 g. We expected it to be the calorie anchor — remove roughly 56 g of starch or avocado and the shake should drop 40 to 70 kcal, thin out noticeably, and those recipes should therefore pile up in the Light label. We expected the empty Body cell to be a Light-shake signature.
Method
We split all 201 production recipes by a single binary: is the Body cell populated (149 recipes) or empty (52)? We then compared mean Kcal, protein, carbohydrate, fat and fibre, the internal Score, and the gram weight and item count of the Fruit & Vegetable slot. Because Cal Label sets a calorie target, we repeated every comparison inside each label separately — Light, Standard, Filling, Meal Replacement — so the target could not do the work for us. Finally we pulled two calorie-matched pairs that share Body Type, Week, Goal and Cal Label, differing only in whether the Body slot was used. We read the result against two published sources. Flood-Obbagy and Rolls (Appetite, 2009) fed 58 adults 266 g, 125 kcal apple preloads matched for weight, energy content, energy density and ingestion rate; despite that matching, whole apple cut total lunch intake by 15 percent versus control and beat both applesauce and apple juice, and adding fibre back to the juice did not close the gap. Grundy and colleagues (British Journal of Nutrition, 2016) make the structural version of the same argument: intact plant cell walls and isolated polysaccharide solutions can report the same fibre gram count and still behave differently during digestion.
Results
The hypothesis was wrong on calories and right on nothing else. Across all 201 recipes, the 52 with an empty Body slot average 374.3 kcal and 33.0 g protein; the 149 with a Body ingredient average 375.2 kcal and 33.0 g protein. That is a gap of 0.99 kcal — less than a third of a gram of banana. The compensation is visible in one place: the produce slot. Recipes without a Body ingredient run 159.1 g of Fruit & Vegetable across 3.0 items; recipes with one run 69.0 g across 2.0 items. Roughly 56 g of dense body is being traded for roughly 90 g of watery produce and one extra ingredient, and the calorie target survives the trade. The pattern holds inside every calorie label, in both directions and always small: Light 363.7 versus 355.0 kcal, Standard 379.7 versus 372.0, Filling 379.1 versus 385.4, Meal Replacement 388.9 versus 391.5. Within-group standard deviations run 15 to 33 kcal, so every one of those gaps sits inside the noise. Two matched pairs show it at the level of a single recipe. WY-W4-RENEWAL-001-V1 "Soft Renewal" has no Body ingredient — Coconut Milk 150 ml plus Soy Milk 150 ml, Carrot 50 g, Tofu 50 g plus Greek Yogurt 220 g, Lemon 2.5 g, Monk Fruit 2.5 g plus Honey 5 g — and lands at 313.2 kcal. Its Week-4 sibling WY-W4-RENEWAL-007-V1 "Featherlight Bloom" uses Carrot 45 g as its Body and lands at 312.9 kcal. Three tenths of a calorie apart. In the Filling range, WI-W1-VITAL-011-V1 "Luminous Charge" (no Body; Oat Milk 150 ml, Beet 60 g plus Dates 21.1 g plus Papaya 70 g, Chicken 45 g plus Greek Yogurt 117 g, Cashew 11.7 g) reaches 382.4 kcal against WI-W1-VITAL-004-V1 "Vivid Rise" (Avocado 33.8 g as Body) at 382.8 kcal.
What Failed
Four things go wrong if we stop at the calorie match. First, fat is not conserved. Recipes without a Body ingredient average 8.9 g of fat against 10.7 g for those with one — 16 percent lower. The cause is not mysterious: avocado occupies 43 of the 149 Body cells, and avocado-body recipes average 13.5 g of fat against 9.5 g for every other body ingredient. Removing the slot quietly removes the main fat carrier, and the calorie total papers over it. Second, fibre moves in both directions and the group mean hides it. On average the empty-Body recipes carry slightly more fibre, 8.0 g against 7.5 g — but the Filling pair goes the other way and goes hard. Luminous Charge carries 5.9 g of fibre; Vivid Rise, with its avocado and green lentils, carries 12.0 g. Two recipes 0.4 kcal apart, differing by more than double in fibre. Anyone reading only the group average would miss that entirely. Third, and most importantly: our Score does not track any of this. Empty-Body recipes average 14.62 against 15.11 — a difference of 0.49 on an 11-to-20 scale, well inside the spread, and it points in inconsistent directions recipe by recipe. Soft Renewal, with no Body ingredient, scores 16 while Featherlight Bloom scores 13. The Filling pair ties at 17 despite the doubled fibre. We would like to tell you the Score noticed the structural difference. It did not, and we are not going to imply a causal link the numbers refuse to support. Fourth, the substitution rule breaks at the top of the calorie range. Only 3 of our 14 Meal Replacement recipes leave the Body slot empty, and their produce slot runs 92.3 g against 64.5 g — a 28 g gap, not the 90 g seen everywhere else. Three recipes is far too few to draw a conclusion from, and we are reporting it as a limit on the finding rather than as a finding.
What We Learned
The Body slot is not a calorie anchor. It is a structure slot, and our own recipe system has been treating it as optional for a while without anyone writing that down — rebuilding the same calorie and protein target out of about 90 g more produce spread across one more ingredient. That is a genuinely useful degree of freedom for a formulation team. But the swap conserves the number, not the food. What changes is the physical form the calories arrive in, and that is exactly the variable the published work says matters. Flood-Obbagy and Rolls matched their preloads on weight, energy, energy density and eating rate, and whole apple still outperformed applesauce and juice — and putting the fibre back into the juice did not rescue it. Grundy and colleagues describe the same thing at the tissue level: identical fibre grams behave differently depending on whether they arrive inside intact cell walls. A shake that matches on kcal, protein and fibre grams is not, on that basis alone, the same shake. Our fat numbers say the same thing from inside our own data.
What Changed at MOMS
We stopped treating the empty Body cell as a formatting accident and started treating it as a declared formulation choice, with two rules attached. One: when a recipe leaves the Body slot empty, the produce slot has to gain an additional item, not merely more grams of the same one. The 52 existing empty-Body recipes already average 3.0 produce items against 2.0, so this codifies what the recipes were doing anyway and makes it the floor rather than the tendency. Two: if the Body ingredient being dropped is avocado, the recipe must name a replacement fat source in the Functional slot instead of letting roughly 4 g of fat quietly disappear behind an unchanged calorie total. Luminous Charge already does this correctly with Cashew 11.7 g. Not every empty-Body recipe did, and those are the ones we are reformulating first.
References
Flood-Obbagy JE, Rolls BJ. The effect of fruit in different forms on energy intake and satiety at a meal. Appetite. 2009;52(2):416-422. | Grundy MML, Edwards CH, Mackie AR, Gidley MJ, Butterworth PJ, Ellis PR. Re-evaluation of the mechanisms of dietary fibre and implications for macronutrient bioaccessibility, digestion and postprandial metabolism. British Journal of Nutrition. 2016;116(5):816-833.
Keep reading
Next experiment
Not sure which shake fits you? We match the formulation to your body type.
Find your body type