Nutrition science buried its most embarrassing finding for 20 years because it contradicted every dietary guideline they’d ever written. Harvard tracked 190,000+ people across three studies starting in 1986. Half a cup of ice cream per day was associated with a 20% lower risk of type 2 diabetes. A doctoral student named Ardisson Korat later found the same half-cup daily was linked to lower cardiovascular disease risk in diabetics too. The researchers tried to kill the finding. They ran every statistical test they could think of to make it disappear. Controlled for reverse causation. Stripped out data from people who changed diets after health diagnoses. The effect shrank but stayed statistically significant. Multiple independent scientists confirmed the ice cream signal was as strong as, or stronger than, the yogurt signal. Yogurt got published and promoted. Ice cream got buried. Same data quality, same cohorts, opposite treatment by the field. Food diaries expose that nutrition science operates on vibes and then reverse-engineers the statistics to match. When the data cooperates, you get a landmark study. When it doesn’t, you get a buried dissertation. Ice cream’s glycemic index is lower than brown rice. That single fact should tell you how much of what you’ve been told about “healthy eating” is evidence-based versus reputation-based.