The Foods You Were Told to Avoid (But Should Be Eating)
Dr. Rhonda Patrick talks with cardiologist and nutrition researcher Dr. Dariush Mozaffarian about why the whole food—and what it replaces—often matters more than a single nutrient. They discuss dairy, red meat, dietary fats, ultra-processed foods, obesity, and policies that could make nutritious food more accessible. Some of Mozaffarian's explanations, particularly those about the causes of obesity and individual responses to food, are hypotheses or interpretations of developing evidence, not settled conclusions.
1. Why a Cardiologist Focused on Food
Dr. Dariush Mozaffarian directs the Food Is Medicine Institute at Tufts University. Patrick introduces him as a researcher whose early work helped reveal the cardiovascular harms of industrial trans fats.
Mozaffarian says his interest in nutrition began partly with a gap in his medical education. Across medical school, internal medicine, and cardiology training, he learned biochemistry but very little about how people's actual diets affect chronic disease.
"Most doctors know nothing about the top cause of poor health. … That single sentence tells us so much about our current problems."
In the late 1990s and early 2000s, he began reading nutrition research independently. At a time when low-fat products dominated grocery shelves, he concluded that the evidence already favored a different approach: minimally processed dietary patterns rich in healthy fats, fiber, and plant compounds, such as traditional Mediterranean-style diets.
For him, the problem was twofold: physicians needed better nutrition training, and existing research needed to be translated into public policy and practical advice.
2. Saturated Fat, LDL, and the Whole Food
Patrick begins with a familiar puzzle: if saturated fat raises LDL cholesterol, why do saturated-fat-rich foods such as butter and cheese appear to affect LDL differently?
Mozaffarian first stresses a distinction:
"LDL cholesterol is clearly one top risk factor for heart disease—not the only one."
He does not dismiss LDL, statins, or cardiovascular risk. His argument is that judging a food solely by its saturated fat content, or by its effect on LDL alone, can miss other relevant effects.
He gives three reasons:
- Changing dietary fat can affect several markers, including triglycerides, HDL cholesterol, and lipoprotein(a), not just LDL.
- Saturated fats are a family of fatty acids, not one identical substance. Their carbon-chain lengths and biological effects differ.
- Foods contain other interacting components. Dairy fat, for example, comes packaged with compounds that differ between milk, cheese, cream, and butter.
Palmitic acid, a 16-carbon fatty acid, is a common saturated fat in foods and the body. Mozaffarian also discusses medium-chain triglycerides, or MCTs, which he says may have different metabolic effects. He notes that trials of MCT oils exist but are not large or uniformly strong.
His broader position is:
"It's not that saturated fat is inert. … But it's a distraction."
In his view, a diet can be high or low in saturated fat and still be healthy or unhealthy, depending on its overall composition. He argues for dietary guidance centered more on foods and dietary patterns than on a single fat target.
3. Why Butter and Cream May Behave Differently
One example of the food matrix—the physical structure and combination of substances in a food—is the milk fat globule membrane. This membrane surrounds fat droplets in milk and contains biologically active lipids, including phospholipids and sphingolipids.
Mozaffarian says research suggests these compounds can interact with the gut microbiome and influence cholesterol absorption. The membrane remains present in foods such as milk, cream, yogurt, and cheese, whereas much of it is separated from the fat during butter-making.
He describes an eight-week randomized trial in which participants consumed comparable amounts of fat from either butter or unsweetened cream. According to his account, LDL rose in the butter group but did not change in the cream group.
"The same fat, the same saturated fat—one's butter, one's cream."
He sees this as evidence that a nutrient's effect can depend on the food carrying it. He also notes that homogenization breaks up milk-fat droplets so the fat does not rise to the top; he does not consider that process inherently harmful.
Butter: Compared with what?
Mozaffarian calls butter a relatively neutral food compared with the average diet, while emphasizing that the replacement matters:
- Replacing olive oil or other unsaturated plant oils with butter would generally be a less favorable swap, in his view.
- Replacing refined starches, sweets, or other low-quality foods with butter might be a favorable swap by his comparison.
- He does not recommend seeking out large amounts of butter as a health food.
His practical advice is to use plant oils for most cooking while not treating a little butter as a crisis.
4. How Nutrition Became Focused on Individual Nutrients
To explain why people became so focused on fat and saturated fat, Mozaffarian goes back well before the low-fat guidelines of the 1980s. He identifies three major influences on the modern food system: vitamin discovery, the drive to prevent famine, and ultra-processing.
Vitamin deficiency and a scientific breakthrough
He recounts the story of physician Joseph Goldberger, who investigated pellagra in the American South. Pellagra causes serious illness and can be fatal; it is now known to result from niacin deficiency.
Goldberger noticed that people confined in institutions developed pellagra while staff often did not. He suspected something missing from the patients' monotonous diets rather than an infection. His demonstrations included deliberately exposing himself and others to material from patients to challenge the infectious-disease theory. Dietary changes, including wheat germ, later helped patients.
The discovery and use of vitamins transformed public health. Mozaffarian points to the 1941 National Nutrition Conference for Defense, which contributed to flour enrichment and the development of recommended nutrient allowances.
The unintended consequence, he argues, was a habit of thinking about food primarily as a collection of measurable nutrients.
"When it comes to vitamin deficiency diseases, a single nutrient does cause or cure a disease by definition. But when it comes to complex chronic diseases … you can't break it down to a nutrient."
His point is not that vitamins are unimportant. It is that a strategy that works well for a deficiency disease may be too simple for conditions such as heart disease, diabetes, cancer, and obesity.
5. The Green Revolution and the Focus on Calories
The second historical force was fear of starvation. Following wartime famine and rapid population growth, scientists and policymakers sought ways to produce far more food.
Mozaffarian describes the work associated with Norman Borlaug and the Green Revolution: breeding high-yield crops and expanding the use of fertilizers, pesticides, and irrigation. He credits these changes with helping avert widespread famine and greatly increasing grain production.
At the same time, he argues that the system's priorities—yield, starch, and resilience—came at a cost:
- Diets became less diverse.
- Wheat, rice, and corn supplied a large share of global calories.
- Modern varieties were not necessarily selected for nutrient density or flavor.
- Industrial agriculture brought concerns about chemical residues, water use, and ecological impact.
Some of the specific historical and nutritional figures he gives are broad estimates. His central takeaway is that technologies developed to solve a real food shortage also shaped a food system that may be poorly suited to preventing today's chronic diseases.
Patrick adds that it is useful to ask not only which nutrient to avoid, but also what food supplies that the body needs: vitamins, minerals, fiber, polyphenols, and other compounds working together.
6. The "Dark Matter" of Food
Mozaffarian calls the less-understood components of food its "dark matter." He says newer measurement methods have identified millions of food compounds, many of which researchers are only beginning to study.
Examples of better-known compounds include:
- EGCG in green tea
- Cocoa flavanols in chocolate
- Polyphenols in berries and pomegranates
- Sulforaphane-related compounds in cruciferous vegetables
The number of compounds measured does not mean that each has a proven health benefit or that scientists know what they all do. His point is that food contains far more chemical complexity than appears on a nutrition label.
Patrick describes small studies from her former research environment involving a nutrient-dense bar containing vitamins, minerals, polyphenols, and a fiber-based matrix. In a follow-up experiment, separating the components did not reproduce all the observed biomarker changes of the complete bar.
She presents this as a proof-of-principle example, not evidence that the bar replaces a meal. It illustrates a challenge in nutrition research: isolating one ingredient may miss effects that depend on combinations, baseline nutritional status, and food structure.
Mozaffarian adds that convenient foods are still necessary. The goal should not be food snobbery or insisting that everyone cook everything from scratch, but making convenient options that better preserve useful food components.
7. Food Diversity and Dietary "Wars"
The conversation turns to dietary diversity. Mozaffarian argues that humans and their gut microbes developed alongside a wide range of foods, and that major changes in food production have narrowed that variety.
He also discusses research suggesting that the shift from hunting and gathering to agriculture was accompanied in some populations by more nutritional deficiencies, dental problems, and skeletal issues. The historical picture is complex, but he uses it to emphasize that increased food production and better nutrition are not always the same thing.
Patrick connects the discussion to modern arguments over ketogenic and carnivore diets. Rather than treating fat, carbohydrate, animal foods, or plants as universally good or bad, she comes back to the value of a diverse selection of minimally processed foods.
The speakers also mention that plant compounds can influence cellular stress-response pathways, while periods without food activate different metabolic processes. These mechanisms are interesting, but they do not establish that any one fasting schedule or dietary pattern is right for everyone.
8. A Hierarchy of Animal Foods
Mozaffarian distinguishes between types of animal foods instead of placing them in one category. His approximate hierarchy is:
- Fermented dairy, especially yogurt, followed by foods such as cheese and kefir.
- Milk, which he says contains compounds associated with lower blood pressure and stroke risk.
- Eggs, poultry, and butter, which he places lower.
- Unprocessed red meat, including fresh or frozen beef, pork, and lamb.
- Processed meat, such as bacon, salami, sausage, and preserved deli meats—including turkey deli meats.
He considers most minimally processed plant foods particularly valuable, while saying that animal foods can also fit into a healthy diet.
For unprocessed red meat, he suggests one or two servings a week as a reasonable choice for people who enjoy it. He considers higher intake more concerning, particularly in a diet already dominated by refined and processed foods.
Processed meat, he says, is more strongly associated with adverse outcomes including diabetes, stroke, and cancer. He suspects preservation and processing play an important role.
"It's not just the saturated fat. … It's other things in the meats."
These rankings reflect his interpretation of varied evidence; observational associations cannot by themselves establish exactly how much risk an individual will experience.
9. Red Meat, Cooking, and the Background Diet
Mozaffarian says unprocessed red meat is more consistently associated with type 2 diabetes and colorectal cancer than with heart disease. He discusses several possible mechanisms.
One is heme iron, which contributes to red meat's color. Iron is essential, but he says excess iron can promote oxidative stress and inflammation. He points to animal experiments, iron-overload disorders, and associations between iron stores and diabetes as reasons to investigate its role.
The speakers also discuss compounds formed by high-heat cooking, especially when meat is burned or heavily charred. Patrick mentions heterocyclic amines, which can form during cooking, and research into whether compounds in cruciferous vegetables affect the enzymes involved in processing potential carcinogens.
Mozaffarian agrees that avoiding heavily blackened meat is sensible. He also describes a meat-industry-funded study suggesting that a Mediterranean-style background diet may blunt some changes in metabolites associated with red-meat consumption. He notes the funding source as a reason to interpret it carefully.
Patrick asks whether someone eating an otherwise excellent diet could safely increase from two to four red-meat servings a week. Mozaffarian is explicit that there is no direct science answering her exact question. He speculates that the difference may be small in that context, but it remains speculation—not a demonstrated safe threshold.
Pasture-raised meat and biomarkers
The speakers discuss whether regenerative or pasture-based farming changes meat's health effects. Mozaffarian says measured nutrient differences have generally been modest, while some early work on animal stress and inflammatory markers is intriguing. He is hopeful but says more research is needed.
Asked whether someone could test their response, he suggests TMAO, a compound related to gut-microbial metabolism, as a possible marker to observe. But he also acknowledges that there is no established blood test that can reliably capture all potential long-term effects on the pancreas or colon.
10. Omega-3s, Omega-6s, and Olive Oil
The discussion shifts to unsaturated fats. Mozaffarian describes his early research interests in omega-3s and trans fats as a kind of nutritional yin and yang: small quantities of different fats can have very different biological effects.
Omega-3 fats
He says the strongest evidence among unsaturated fats supports omega-3s, especially the long-chain forms found in seafood. Plant sources such as walnuts and flax also provide omega-3 fats, although they are not identical to the main long-chain seafood forms.
Fish-oil trials have produced mixed results: some have found benefits for cardiovascular outcomes and others have not. His overall interpretation, combining different kinds of evidence, is that omega-3s—particularly from seafood—are beneficial.
Patrick cites an older estimate attributing a substantial number of deaths to low seafood omega-3 intake. That figure is a population-model estimate, not a count that can prove the cause of any individual death.
Omega-6 fats
Mozaffarian challenges the idea that omega-6 fats are inherently inflammatory. He says human studies generally show favorable or neutral effects on markers such as LDL cholesterol, triglycerides, and insulin-related measures, with no consistent evidence that typical omega-6 intake causes inflammation.
His advice on the omega-6-to-omega-3 ratio is:
"If we improve our ratio by eating omega-3s, we will all be healthier. … If we try to improve our ratio by decreasing omega-6s, we're just taking away another beneficial fatty acid."
Olive oil and avocados
He considers olive oil and avocados healthy, but questions whether their benefits should be credited to monounsaturated fat alone. Extra-virgin olive oil also contains compounds such as oleocanthal, which has anti-inflammatory activity.
His general recommendation is to include fats from plants and seafood, rather than fear fat as a category.
11. Fats as Biological Information
Mozaffarian argues that dietary fat should not be viewed simply as nine calories per gram.
"All food is biologic information."
He explains several ways fatty acids can affect the body:
- They help form cell membranes, influencing their properties and the behavior of receptors embedded in them.
- They can interact with cell receptors, affecting communication between cells.
- They can influence gene activity, including pathways involved in triglyceride production.
- They serve as raw materials for other signaling molecules, some involved in inflammation and its resolution.
He gives examples involving omega-3s in cardiac cell membranes and the retina. Patrick contrasts this with trans fats, which can alter membrane properties in less favorable ways.
The larger message is that equal numbers of calories from different fats do not necessarily have identical biological effects. At the same time, mechanistic effects in cells do not automatically establish the size of a clinical benefit for every person.
12. Do Seed Oils Deserve Their Reputation?
Patrick asks why seed oils have become a major target of concern. She raises several possibilities: their presence in ultra-processed foods, chemical extraction, and oxidation when oils are repeatedly heated.
Mozaffarian thinks several concerns have become mixed together:
- The real history of harmful industrial trans fats and partially hydrogenated oils.
- Unease about industrial processing and possible trace contaminants.
- The belief that omega-6 fats promote inflammation.
- Problems caused by repeatedly heating and degrading frying oils.
He argues that studies showing benefits from unsaturated oils used commercially available oils, so their overall effects should not be judged only by the possibility of trace residues. He also agrees that repeatedly overheated or burned oil is a separate concern, regardless of whether it started as a seed oil or animal fat.
"That's not the oil itself. It's how it's used."
His potato-chip example illustrates the distinction: a chip is not a health food, he says, but blaming its oil while overlooking the product's overall composition can distract from the bigger question.
What he cooks with
Mozaffarian says his household usually uses extra-virgin olive oil for its flavor and uses canola oil for deep frying. Patrick says she commonly cooks with olive or avocado oil.
He advises avoiding burned oil and says people choosing among common plant oils for home cooking are already making a reasonable choice. His preference for particular oils is practical advice, not proof that any one oil is optimal for every cooking method.
13. How Ultra-Processed Foods Expanded
Mozaffarian returns to the third major historical force: ultra-processing.
He suggests that low-fat messaging encouraged manufacturers to remove fat from products and then add ingredients to restore flavor, texture, and stability. Fat-free salad dressing is his example: remove the oil, and manufacturers may rely on starch, sugar, salt, and additives to create a usable product.
He also discusses tobacco-company ownership of major food brands in the late 20th century. He says companies brought marketing, flavor, and product-development expertise to food, helping make products especially appealing and recognizable—including to children.
In his telling, the result was a food system increasingly built around products that are convenient, heavily marketed, and easy to consume in large amounts.
"I think we've created kind of a monster."
He cites an estimate that roughly 60% of calories in the food supply or diet come from ultra-processed foods.
14. The GRAS Loophole
Mozaffarian criticizes the U.S. GRAS pathway: substances can be considered "generally recognized as safe" under specified conditions without going through the same premarket food-additive approval process.
He says companies can make their own GRAS determinations and are not always required to notify the FDA, creating potential conflicts of interest and gaps in the agency's knowledge. He cites a 99% figure for additives entering without FDA review and argues that some ingredients are difficult for consumers to identify under broad labeling terms.
Those claims should be understood as his characterization; the precise percentage depends on what substances and time period are being counted. A lack of premarket FDA review also does not, by itself, prove that a particular additive is harmful.
At the time of the conversation, he says a proposed rule intended to address the loophole was under federal review. That is a time-specific policy update, not a guarantee that a rule has been finalized or what its final terms will be.
He also argues that the FDA has inadequate resources for nutrition and for reviewing ingredients already on the market. He cites the slow response to industrial trans fats and discusses state-level action on additives, warning labels, and school meals. His policy preference is a federal safety floor that does not prevent states from adopting stronger protections.
15. What Makes a Food Ultra-Processed?
Mozaffarian explains the NOVA classification, developed by Brazilian researcher Carlos Monteiro. NOVA groups foods according to the nature and extent of processing; NOVA 4 is the ultra-processed category.
He contrasts foods such as yogurt, canned fish, or traditionally made bread with products made using industrial processes and ingredients not typically used in a home kitchen. A practical clue, he says, is a long ingredient list containing substances added mainly to create flavor, color, or texture.
However, he stresses that processing is one dimension, not the only way to judge a food. Not all ultra-processed foods are equally harmful, and not every food outside that category is automatically nutritious.
He proposes several possible reasons why highly processed diets may be harmful:
- Altered structure: Breaking down a food's natural matrix may change how quickly it is digested.
- Loss of useful compounds: Processing may reduce some fibers, polyphenols, and other food components.
- Additives: Some may affect metabolism, the gut, or other systems, though effects differ by ingredient and dose.
- Hyper-palatability and marketing: Products can be designed to be especially appealing and easy to overconsume.
- Contaminants: Processing, packaging, or agricultural production can introduce unwanted substances.
He is clear that researchers do not yet know how much each factor contributes.
16. Food Structure, Digestion, and Carb Quality
To explain food structure, Mozaffarian compares steel-cut oats with a highly processed whole-grain cereal. Both may contain whole-grain ingredients, but their physical structures differ. He says the more intact oats generally digest more slowly, while a processed cereal may release starch more quickly.
His proposed "triple hit" from highly disrupted food structure is:
- Faster absorption of nutrients and a larger glucose response.
- Less material reaching microbes in the large intestine.
- A possible shift in how much energy is absorbed by the person versus used by microbes.
The last point is an emerging hypothesis. The exact shares of calories handled by humans and microbes vary, and the percentages he mentions should not be treated as established values for every diet.
Patrick connects this to products marketed as high-protein, low-sugar, keto, or paleo. A favorable claim about one nutrient does not necessarily make a heavily processed snack a nutritious food.
Two practical checks
Mozaffarian offers two rough shopping rules:
- Can you recognize actual foods? For example, can you see nuts, seeds, coconut, or dates?
- For grain-based packaged foods, look for at least 1 gram of fiber per 10 grams of carbohydrate. A product with 30 grams of carbohydrate would therefore have at least 3 grams of fiber.
He considers a higher fiber-to-carbohydrate ratio useful, but it remains a screening heuristic, not a complete assessment of every food.
17. Obesity and the Energy-Balance Puzzle
Patrick asks what has driven the obesity epidemic. Mozaffarian separates the story into two periods.
For the period from roughly the 1980s to 2000, he points to increased refined starch, larger portions, more food eaten away from home, and the growth of ultra-processed products. He estimates that refined starch consumption increased substantially and considers it a major contributor.
For the period after 2000, he describes what he sees as a paradox: estimates of average calorie intake flattened or declined slightly, some measures of physical activity increased, yet obesity continued to rise.
"Calorie intake is pretty flat, physical activity's going up, and we're getting more obese."
He points to research suggesting that resting energy expenditure—energy used while the body is at rest—may have fallen over time. He cites an estimated decline of around 220 calories per day in the populations studied.
This is an important area of research, but his conclusion about why the decline occurred and how much it explains is not established. Population trends and measurements of intake and expenditure have limitations, and they do not identify a single cause of obesity.
Both speakers agree on an important distinction: energy balance still applies, but it is more complicated than deliberately counting food calories and exercise calories. Digestion, metabolism, activity, body composition, hormones, and other factors all influence the balance.
18. Brown Fat, Thermogenesis, and Genetics
Mozaffarian's leading hypothesis is that changes in diet and the microbiome have reduced thermogenesis, the production of heat by the body.
He explains that:
- White fat mainly stores energy.
- Brown fat can use energy to produce heat.
- Some cells can take on intermediate, "beige" characteristics.
- Muscle can also contribute to heat production through metabolic processes.
Adults retain some brown fat, and its activity is associated with leanness in research. He says food components and microbial metabolites—including compounds studied in berries, turmeric, green tea, and omega-3 sources—can influence thermogenic pathways.
"We're gaining weight because we're cold."
That is his memorable summary of the hypothesis, not a proven explanation for the obesity epidemic. Evidence that a compound affects a pathway, particularly in laboratory or animal research, does not establish that eating it produces meaningful weight loss in humans.
Patrick notes that cold exposure can increase measures of brown fat, but some studies do not show substantial weight loss. Mozaffarian agrees that brown-fat quantity and activity may both matter and says combining dietary changes with cold exposure would be interesting to study—not an established treatment.
He also discusses the FTO gene, a common genetic region associated with obesity risk. He notes research linking it to the development of fat cells with different energy-handling properties. Genetics can affect susceptibility, he says, but population-wide changes over a few decades also point to major environmental influences.
19. Weight Loss, Metabolic Slowdown, and the Microbiome
Mozaffarian describes the common observation that energy expenditure often falls after weight loss. Some decrease is expected when a smaller body requires less energy, and additional metabolic adaptation can also occur.
He highlights studies by researcher David Ludwig comparing diets after weight loss. In his account, lower-carbohydrate, higher-fat diets preserved more energy expenditure than lower-fat diets in those trials. He interprets this as evidence that food composition can affect metabolic adaptation.
The size, causes, and generalizability of these differences remain debated. The findings do not prove that everyone needs a ketogenic diet or that calorie intake is irrelevant.
His personal advice is to prioritize food quality, pay attention to hunger and fullness, and aim for gradual, sustainable change rather than trying to lose weight as rapidly as possible.
"Don't try to lose 10 pounds a month. … Aim for a pound a month."
He also speculates that gut-microbial composition may influence how energy is used or stored. When Patrick asks which foods would produce a particular desirable microbial response, he says researchers do not yet have a reliable one-to-one answer. Broadly, he favors a varied diet containing nuts, seeds, legumes, minimally processed grains, fruits, vegetables, fermented foods, and seafood.
20. GLP-1 Drugs, Exercise, and Health Beyond Weight
The speakers discuss GLP-1-based medications, now used for diabetes and obesity. Mozaffarian says these drugs can be valuable and that he prescribes them for appropriate patients. But he argues they work best alongside nutrition and lifestyle support, rather than as a substitute for it.
He notes that many people regain weight after stopping treatment and raises concern that rapid weight loss can include loss of muscle and bone. His proposal is for medication to come with structured dietary and lifestyle counseling, particularly to help preserve health and support longer-term outcomes.
Patrick returns to exercise. Mozaffarian says exercise alone often produces less weight loss than people expect, but that does not make it unimportant. It can improve fitness, metabolic health, muscle, and body composition, including reductions in visceral fat even when the scale changes little.
His wider point applies to food as well:
"Think of it not as a tool for weight loss, but think of it as a tool for good health."
A nutritious diet can matter for cardiovascular health, immune function, mood, sleep, and other outcomes regardless of whether a person loses weight.
21. Dietary Guidelines, Protein, and Dairy
Mozaffarian reviews what he describes as major changes in the then-new U.S. dietary guidelines. He welcomes greater emphasis on real foods, healthy fats, and less processed food, while noting that the value of guidelines depends on whether they are communicated and applied in programs such as school meals.
He also recalls that earlier guidelines had already moved away from a limit on total fat and had changed their treatment of dietary cholesterol.
One concern is that a recommendation to consume adequate protein can be turned into a marketing opportunity. Adding protein to candy, chips, pastries, or sugary drinks, he says, can give an otherwise low-quality product a health halo.
Whole-fat versus low-fat dairy
Mozaffarian says the evidence does not consistently show that whole-fat dairy is worse than low-fat dairy for major health outcomes. He places particular value on yogurt and cheese, especially without added sugar, and suggests choosing the fat level that makes a nutritious dairy food enjoyable and sustainable.
He cites evidence on children and whole-fat dairy, but some of those comparisons are observational; they do not prove that switching a child's milk fat content will cause a particular weight outcome. Families should follow individualized pediatric advice where relevant.
Patrick asks whether unsweetened plant milk is preferable to dairy milk. Mozaffarian says both can be reasonable choices depending on taste and tolerance. Their nutrient profiles are not interchangeable in every case, so the type of plant milk and whether it is fortified matter.
What about APOE4?
Patrick asks whether people carrying APOE4, a genetic variant associated with increased risk of Alzheimer's disease and other conditions, should avoid whole-fat dairy.
Mozaffarian says research has not produced a sufficiently clear, reproducible gene–diet finding to make a universal dairy rule. He recommends focusing on the overall diet and including unsaturated fats. His further suggestion about when to choose low-fat dairy is an individualized judgment, not a proven APOE4-specific prescription.
22. Making Nutritious Food More Accessible
The conversation ends with policy. Mozaffarian argues that the current food system emerged through a combination of science, government priorities, and private-sector innovation. He believes the same forces can be redirected toward healthier food.
Rather than relying only on education or comparing all food policy to tobacco policy, he uses clean-energy policy as an analogy: public incentives can help businesses develop alternatives and make them more accessible.
His proposals include:
- Healthcare spending: Redirect a portion toward nutrition support and programs in which clinicians can prescribe or help provide healthy food.
- Federal nutrition programs: Improve the nutritional impact of programs such as SNAP and school meals while preserving recipients' dignity and food security.
- Research: Invest more of the National Institutes of Health budget in meaningful nutrition research.
- Regulation: Strengthen premarket oversight and review of ingredients already in the food supply.
- Business incentives: Use grants, loans, and tax incentives to support healthier manufacturing, farming, restaurants, and supermarkets in underserved areas.
He cites large figures for healthcare and federal nutrition spending to show the scale of the opportunity. His suggested reallocations are policy proposals, not descriptions of enacted budgets.
Patrick is enthusiastic about food prescriptions, public education, and incentives that make healthier choices easier rather than simply placing the burden on individuals.
23. What "Food Is Medicine" Means
Patrick asks an essential closing question: if food is medicine, how is it different from actual medical treatment?
Mozaffarian is clear that food is not a replacement for appropriate care. Eating broccoli does not mean someone with cancer should skip chemotherapy. Rather, nutritious food can support health and may complement treatment.
"I don't lead the Food Is Only Medicine Institute."
Food, he adds, is also culture, flavor, dignity, and joy. The phrase "food is medicine" is meant to draw attention to food's biological importance, not reduce every meal to a prescription.
He closes by introducing his book, Food Is Medicine, described in the conversation as coming out in October 2026. It is intended for general readers and covers nutrition science, practical food choices, and policy.
The final message returns to the beginning: look beyond isolated nutrients, eat a varied diet of nourishing foods when possible, and build systems that make those foods easier for everyone to access.
Need a summary like this?
Get instant summaries with Harvest