Nutrition Labels are Whack
Plus the Medicare money machine, epigenetic age vs chronological age, ancient medicinal clay, analogies between deep learning and biology
Welcome back. I’m working on a few longer posts, namely speculations on FAANG’s biotech projects and another dissecting obesity research, but plan on publishing pieces like this one in the interim.
I’m also off to Berlin! I’ll be there for the better part of October until early December with stints in Portugal, Bulgaria, Denmark, and England. Email or find me on Twitter - I’d love to meet.
Weird Things I Learned about Nutrition Labels
It all started with Olipop. I’m at Nikhil Krishnan’s birthday party, considering my beverage options. Nikhil and his roommates really like Olipop apparently because there’s a formidable stack of them in the fridge. I read “Supports Digestive Health” on top of the Ginger Lemon can and end up drinking three.
As the daughter of two health pedants, I am trained to be skeptical of sodas. So when I casually flip to the Nutrition Facts panel, I’m shocked to see the dietary fiber listed in one serving meets 30%+ of daily needs. My bowels are going to be fucked.
I haven’t touched Olipop since but it did lead me to wonder what else I don’t know about nutrition labels.
1. I got 500,000+ problems
Early U.S. food-labeling requirements could be distilled to 4 attributes:
Net weight / count / measure
Address of relevant producers, manufacturers, and distributors
Artificial flavorings or colorings
A prominent statement if something was an imitation
These laws—like the 1906 Pure Food and Drug Act and 1938 Food, Drug, and Cosmetic Act—were reactions to rapid industrialization (grotesque meatpacking plants described in Upton Sinclair’s The Jungle and deaths following the consumption of a product containing anti-freeze). The acts explicitly include chewing gum as a food and refrain from defining raw and dried fruits, vegetables, and fish.
God knows why there is an exception for avocados, citrus fruits, and melons. Link
Nutritional information on food labels remained limited and inconsistent for decades. Major reform didn’t come until the Surgeon General’s report in 1988 linking poor labeling to low-quality diets (and directionally associating it with heart disease, cancer, stroke, diabetes, and atherosclerosis). Congress passed the Nutrition Labeling and Education Act (NLEA), with the official “Nutrition Facts” panel appearing in 1990.
It still took years to streamline what had exploded into 500,000+ accepted product labeling formats, but NLEA marked a solid step towards uniform, national standards.
2. Good luck to anyone calorie counting
A Youtube video from popular dietician Abbey Sharp made me aware that nutrition labels can be up to 20 percent inaccurate. I wondered if she was just tossing in the ol’ 80/20 platitude but no… Abbey was being literal. From a regulatory perspective, the FDA considers a label misleading if the nutritional content is off by more than 20% of the stated information in either direction: say 3 grams of added sugar versus 18 grams.
While this may not incentivize intentional mislabeling, it does give food companies leeway to care less about compliance. USDA Food Safety and the Food Safety and Inspection Service of the Agriculture Department are plagued by inadequate testing so beware of relying on specific numbers.
3. Why doesn’t wine have a nutrition label?
Because wine, like ~20% of other consumables, is not regulated by the FDA but falls under the Department of the Treasury. And no, it doesn’t start with Prohibition but the Revolutionary War. In 1789, Congress established a tax on imported spirits to offset debt; Alexander Hamilton suggested putting enforcement in the Treasury.
In 1886, one lone employee from the Department of Agriculture went to the Treasury to start a forensic testing lab. The adulteration and criminal evidence team grew to include chemists, document analysts, print specialists, and firearms examiners. Prohibition was in full swing by 1920 when the Department of Justice requisitioned the lab. After the repeal of the 18th Amendment, Franklin Roosevelt established an entirely new body (the Federal Alcohol Control Administration) but by the 1940s, the IRS wrested regulatory power back to the Treasury.
I also have whiplash from re-reading that paragraph.
Today the Treasury’s Alcohol and Tobacco Tax and Trade Bureau (TTB) requires importers and bottlers to obtain certificates of approval or certificates of exemption. The same group that regulates your 🥂🍷🍸🍹🍾 also regulates firearms. Only in America.
4. Lack of scrutiny on foreign-based food firms
We stock our shelves with food from tens of thousands of foreign food companies from 150+ countries (65,000 domestic food firms fall under the FDA’s jurisdiction). Just *96* foreign food products were inspected in 2007, a decrease from a pitiful 211 inspections in 2001.
Compliance rates also vary considerably between countries. From 2000 to 2006, the FDA collected samples of 783 imported products for tests of compliance with nutrition labeling regulations. Mexico, Brazil, and Thailand… not doing so hot at 40%+ of samples in violation:
But we’re barely better. Of the roughly 900 domestic samples tested in the study, 24% flouted labeling rules.
5. Infant formula, not even once
4 in 10 infant formula products violate labeling rules. Per the FDA: for “lack [of] vitamins, minerals, or other nutrients required by law.” Nutritional negligence that affects babies is unconscionable and I don’t think I’ll _ever_ use infant formula. I wonder how this will affect the crop of synthetic breast milk startups…
The most common mislabeling slots into the generic finding that the “label is false or misleading in any particular” (25%). Other violations include failure to meet the requirements to bear health claims (19%) and failure to declare all ingredients (15%).
6. The FDA and USDA do not agree on the term “natural”
Organic and natural are meaningless terms but it’s interesting to note that production practices are the principal disagreement between the two agencies. The USDA, which regulates meat, poultry, egg products, and catfish (lol), allows food manufacturing companies to use the term natural if the product contains no artificial ingredient or added color and is only minimally processed. The FDA gives more leeway: the use of pesticides, antibiotics, and other production practices do not prohibit a product from labeling as natural.
7. Why mayo died… and aioli took its place
I already tweeted this but the FDA recently required anything labeled “mayonnaise” contain a minimum of 65 percent vegetable oil by weight. Aioli is not subject to the same stipulations.
US Google Trends confirm the aioli surge
Restaurants, however, are not subject to any labeling requirements unless they have more than 20 locations. I asked my local Asian-Cajun spot Crabby Shack if their chipotle mayo is actually mayo. The “uhh.. why? does it matter?” response from the bewildered woman behind the register IS the right mindset. Does it matter?
Interesting Tweets and Links
In a step back for entrepreneurship, Biden is repealing fast-tracked Medicare coverage for brekathrough medical devices, a rule announced under the Trump administration.
Chris Olah details unusual analogies between biology and deep learning. Worth reading in full as it’s both poetic and critical but particularly the section on evolvability ↔ metalearning (analogy: model=organism, evolution=learning). I’ve noticed that similarity in medical metaphors:
H/t from my former Oscar colleague Shohini Gupta that Medicaid is now reimbursing transportation costs for non-emergency vehicles. The thread goes into potential opportunities and limitations for Uber and Lyft:
A new MIT Press publication by engineer and writer Ritu Raman on the current and future uses of biological materials (tissue engineering, lab grown meat and fabrics, “live” building materials, biohybrid machines). Definitely picking up a copy to furnish the brain and the overgrown bookshelf.
My brilliant friend Michael Rechtin of Petri Bio is helping launch Nucleate, giving infrastructure and funding to students and postdocs (à la YC) without any dilution or fees. They’ve tried the model on some test cohorts. Manifold Bio, a George Church lab spinout in Cohort 2, is building a barcoding platform for protein therapeutics; they raised $5.4M last year.
Michael also linked me to COVID Moonshot, a promethean coordination between leading institutes to develop a patent-free antiviral therapy: “If successful, we believe this would be the first time in history that a clinical drug candidate has been crowdsourced from scratch.”
An evidenced diatribe against Medicare Advantage as a distorted money machine incentivizing insurers to code patients with as many diagnoses as possible. MedPAC and the HHS Inspector General identify the rise in home visits as a major driver of upcoding and overpayments.
Social prescribing is a growing phenomenon in Europe (particularly the UK). NHS doctors and nurses refer patients to non-clinical community activities and programming like gardening and volunteering. Emily, of behavioral health startup Alma, comments on museum visits as self-care. While in need of more clinical proof points, I love social prescribing for endowing individuals, families, and communities autonomy in their own care as an alternative to predominant medicalized health care.
Josiah Zayner, PhD (founder of in-home genetic engineering company, ODIN) was banned from Youtube for documenting his process of creating a COVID vaccine. His essay is a clarion call to democratizing the self and science.
Author of Behave, neuroscientist Dr. Robert Sopolsky, discuss stress testosterone and free will on Andrew Huberman’s podcast. 22:30 has a section on the importance of estrogen for brain development and health. The soy boys might be onto something.
Science in disguise: thanks to DNA sequencing, we now know that ancient ritual treatment with clay introduced a fungus called Talaromyces containing antibacterial and antimalarial properties.
Thank you to Pamela Mishkin (who is joining me in Berlin!) for her thoughtful comments on this post.
Disclosure: I don’t have a copywriter so yolo on the typos. This post does not represent the views of my employer. I am a scout for a16z’s healthcare and bio team.
If you liked this post from Waiting Room, subscribe for more.