● Food Intelligence · Dairy
Across 87,237 logged butter meals, members averaged +30.8 mg/dL — but butter itself has a glycemic index of 0. That cohort signal is the bread, potato, popcorn, and pancakes underneath: pancake-and-waffle pairings ran +16 mg/dL above the no-carb baseline, baked potato +14, rice and corn +13, popcorn +12. Butter is a fat-vehicle story — the carb it rides on is the entire glucose response.
Tap any ingredient below to swap it for a glucose-friendlier alternative. Your Signos Glucose Score, predicted curve, and nutrition update in real-time. Every swap is data-backed by real CGM response measurements.
No — butter alone is glycemically inert (GI=0, zero carbs). The +30.8 mg/dL cohort average across 87,237 logs is driven entirely by the carb underneath the butter. Butter contributes essentially zero carbohydrate, which is why every form anchor in our recipe builder — salted, unsalted, grass-fed, ghee, compound, whipped — is set to 0 mg/dL. The cohort mean of +30.8 mg/dL exists because members log butter alongside toast, pancakes, baked potato, popcorn, and pasta. The carb-pair table makes the structure visible: pancake or waffle pairings averaged +42.3 mg/dL (+16 mg/dL delta, n=1,350), baked potato +40.9 mg/dL (+14, n=1,698), rice and cornbread +40.1 mg/dL (+13, n=4,217), popcorn +38.2 mg/dL (+12, n=4,603), and toast or bread +34.7 mg/dL (+8, n=25,488). Counterintuitively, butter can blunt a high-carb spike: in the modifier table, ≥15g fat pairings cut peaks by ~11% (n=69,988, p<0.001). The fat slows gastric emptying — but only when there's a carb to slow down.
Some members spike +26.4 mg/dL. Others spike +49. The only way to know how you'll spike is to measure with a Signos CGM.
Butter contributes zero glucose; the cohort mean of +30.8 mg/dL across 87,237 logs reflects the carb underneath, not the fat. Counterintuitively, butter often LOWERS the curve when added to a high-carb meal because dietary fat slows gastric emptying and delays starch absorption — an effect visible in our modifier table as a ~11% reduction in peak with fat pairings ≥15g (CI: −4.54 to −2.67 mg/dL, n=69,988, p<0.001). The headline story is the carb-pair slot: pancake/waffle pairings averaged +42.3 mg/dL (+16 mg/dL delta), baked potato +40.9 (+14), rice/corn +40.1 (+13), popcorn +38.2 (+12), bagel/muffin +37.8 (+11), pasta +37.8 (+11), sautéed vegetables +35.5 (+9), toast +34.7 (+8). Plain butter — off a knife, with no carb — anchors at +26.7 mg/dL across 48,116 baseline logs, almost certainly reflecting trace co-logged items rather than the butter itself.
This page draws on 87,237 butter-tagged logged meals from 14,122 unique Signos members, collected through the Signos production CGM platform. Data were filtered to postprandial glucose responses (PPGR) between 0 and 100 mg/dL to exclude sensor artifacts and implausible readings. Modifier effects were estimated using Welch's t-test on matched pairs against a defined baseline; all reported effects carry p<0.001 unless noted. Because butter has a published glycemic index of 0 and glycemic load of 0, the recipe-builder form anchors are overridden to 0 mg/dL — the underlying cohort means (preserved in `option_mean_full_cohort`) reflect the carbs members ate alongside each form, not the butter itself. The food was identified in member logs using a regex that matched "butter" as a primary ingredient, with extensive exclusions for nut butters (peanut butter, almond butter, cashew butter), butter-flavored compounds (butter chicken, butterscotch, butterfly), and plant spreads. Any carb-pair pairing slot requires at least 100 matched meals before it appears in the recipe builder.