Acknowledgments

To the patients. Every case in this book is a composite — Theo, Devi, Walt, the Ortiz-Lindqvists, and everyone else in these pages is assembled from many people and is not any one of them. But the composites are made of something. Twenty years of clinic hours are in here: the food diaries handed over with an apology, the supplement bottles unpacked onto a desk, the "I know I shouldn't have" that I spent two decades trying to talk people out of saying. If this book is useful, it is because a great many people were willing to be honest with a stranger about what they ate. That is harder than it sounds.

To the methodologists. Nutritional epidemiology is a field that spends most of its energy explaining why its own findings should be held loosely, which is an unusual and admirable thing for a discipline to do. The people who built the tools in Chapter 2 — who worked out healthy-user bias, who quantified how badly food-frequency questionnaires perform, who pushed the field toward substitution analysis and Mendelian randomization — did the least glamorous and most valuable work in nutrition. They rarely trend.

To the people who ran the hard studies. Controlled feeding trials are miserable to run. Long cohorts take careers. The investigators behind PREDIMED, DASH, the Diabetes Prevention Program, the Framingham cohort, the NIH ultra-processed food work, and the decades-long observational cohorts gave the field the only solid ground it has. That we argue about the edges of their findings is a tribute to them; there is nothing to argue about where no one bothered to measure.

To the standards bodies. The National Academies committees who set the Dietary Reference Intakes, the staff of the NIH Office of Dietary Supplements, the maintainers of USDA FoodData Central, the Cochrane reviewers, and the WHO and IARC working groups produce careful, slow, unexciting documents that almost nobody reads and almost everybody relies on. This book leans on their work on nearly every page.

To the dietitians. To the RDs and RDNs working in clinics, hospitals, schools, WIC offices, and community programs — usually understaffed, frequently overruled by a podcast, and paid a fraction of what the person selling the cleanse makes. You are the ones translating all of this into something a real person can do on a Tuesday. Yolanda Pierce is a composite of you, and she is the character in this book I like best.

To the critics of my own positions. I have changed my mind several times over twenty years, and never once because someone agreed with me. The researchers who publicly disagree with the consensus — carefully, with data, in the literature rather than on a podcast — are doing the thing science requires. Several 🟡 verdicts in this book are 🟡 because of them, and they are more honest verdicts for it.

To the readers who will find the errors. This book is Creative Commons for a reason. Nutrition guidance moves; some of what is written here will be wrong within a few years, and some of it is probably wrong already. Corrections, better sources, and arguments — with evidence — are welcome and will be incorporated. Being publicly wrong and then publicly fixed is a much better outcome than being quietly wrong forever.

And to anyone who has ever stood in a grocery aisle holding two boxes, trying to remember which one the internet said would kill them this month: this whole thing is for you. You were never the problem. The information environment was.

This book is released under CC BY-SA 4.0. Copy it, print it, teach from it, translate it, argue with it. Share alike.