● Food Intelligence · Legume
Across 15,450 logged tofu meals, Signos members averaged +32.9 mg/dL — but tofu itself has a glycemic index of 0. That cohort signal is the carb tofu sits next to: pad-thai builds ran +16 mg/dL above the plain-tofu baseline, curry-and-rice +10, stir-fry +6, miso soup +2. Tofu is a protein-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 — tofu alone is glycemically inert (GI=0, zero net carbs). The +32.9 mg/dL cohort average across 15,450 logs is driven entirely by the carb underneath the tofu. Tofu contributes essentially zero carbohydrate, which is why every form anchor in our recipe builder — firm/extra-firm, silken, fried, smoked/marinated — is set to 0 mg/dL. The cohort mean of +32.9 mg/dL exists because members log tofu alongside rice, noodles, curry sauce, and stir-fry vegetables. The pairing table makes the structure visible: pad-thai builds averaged +46.2 mg/dL (+16 mg/dL delta, n=386), curry-with-rice +40.7 mg/dL (+10, n=563), stir-fry over rice +35.9 mg/dL (+6, n=562), miso soup +32.4 mg/dL (+2, n=507), Asian sesame salad +31.5 mg/dL (+1, n=1,703). At the meal level, ≥60g total carbs drove a 59.7% peak increase across 3,329 matched logs (p<0.001) — the single largest effect measured. Tofu is a protein-vehicle; whatever it sits next to is the actual glucose driver.
Some members spike +27.4 mg/dL. Others spike +47.7. The only way to know how you'll spike is to measure with a Signos CGM.
Tofu contributes essentially zero glucose; the cohort mean of +32.9 mg/dL across 15,450 logs reflects the carb underneath, not the soy protein. Firm tofu carries only about 1–2g net carbs per 100g, which is why the published GI is 0 and our recipe-builder form anchors are overridden to 0 mg/dL. The headline story is the pairing slot: pad thai averaged +46.2 mg/dL (+16 mg/dL delta), curry-with-rice +40.7 (+10), stir-fry +35.9 (+6), miso soup +32.4 (+2), sesame salad +31.5 (+1). Plain tofu cubes anchored at +30.2 mg/dL across 10,200 baseline logs — almost certainly reflecting trace co-logged items rather than the tofu itself. Morning timing layers on a secondary effect: breakfast tofu meals averaged +26.7 mg/dL vs. +34.4 mg/dL at dinner (n=2,223 vs. n=6,284), a 22% gap consistent with diurnal insulin sensitivity. These are observational patterns, not causal claims.
This page is grounded in Signos CGM data collected between March 2025 and April 2026, covering 15,450 tofu-tagged logged meals from 3,303 unique members (the broader cohort that powers the modifier and pairing analysis), with a single-item sub-cohort of 898 meals (521 members) where tofu was logged alone. Because tofu 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 tofu itself. Modifier effects were estimated using Welch's t-test on matched pairs; all reported effects carry p<0.001 unless noted. Only meals with a measured glucose rise between 0 and 100 mg/dL were included (ppgr_case='regular'). The reported broader-cohort mean is +32.9 mg/dL; the single-item mean is +28.1 mg/dL — the gap reflects the carb context that typically accompanies tofu in real meals. A minimum of 30 matched meals is required for any pairing slot to appear in the recipe builder.