Butter vs Plant Oils, What That JAMA Study Missed

You are halfway through a meal, maybe eggs, toast, stew, or vegetables, and you add a pat of butter because it tastes good.

Then a headline pops up, “Butter is deadly.”

This is the kind of nutrition whiplash the video pushes back against. The discussion is not that diet does not matter. It is that the way some studies are reported, and sometimes the way they are interpreted, can turn a complex, messy topic into a simple villain.

The specific “new study” behind the headlines was published in JAMA and analyzed butter, plant-based oils, and mortality. The expert’s unique perspective is blunt: this is nutritional epidemiology, not a trial that can prove cause and effect, and the groups being compared differ in major lifestyle ways that can swamp the butter signal.

Important: Observational nutrition studies can be useful for generating hypotheses, but they are not designed to prove that one food causes (or prevents) death. If you have heart disease, diabetes, cancer, or are making major diet changes, it is smart to talk with a clinician who knows your history.

The headline you saw, and the lunch you were eating

The media framing tends to be dramatic because it gets clicks.

But the video frames this as a common health puzzle: when a headline says “swap X for Y and reduce death risk,” what is actually being measured?

In this case, the claim was that swapping butter for plant-based oils may reduce risk of dying from all causes. The speaker repeatedly puts “study” in air quotes, not because the paper is fake, but because the design has limits that are easy to miss in a headline.

The conversation also points out something that feels counterintuitive to many people. Butter intake has generally gone down over time, and vegetable oil intake has generally gone up, yet obesity and cardiometabolic disease have climbed. That does not prove vegetable oils are the cause, but it does challenge the idea that “less butter” automatically equals “better population health.”

What the JAMA butter and oils study actually did

This analysis used data from three large, long-running cohorts: the Nurses’ Health Study, Nurses’ Health Study II, and the Health Professionals Follow-up Study.

That sounds impressive, and in some ways it is. Large cohorts can detect patterns that small studies miss.

But it also means the study is observational. It estimates associations between reported intake and outcomes over time, rather than assigning people to eat butter or oils under controlled conditions.

The core question and the headline results

The paper’s research question was essentially: what are the associations of long-term butter and plant-based oil intake with mortality?

The reported findings, as repeated in the video, were:

This is where many articles stop.

The study also modeled a substitution, replacing about 10 grams of butter per day (roughly close to a tablespoon) with an equivalent amount of plant-based oils. The headline-friendly result was a sizable relative risk reduction, including a notable reduction in cancer mortality.

What the research shows: The published analysis in JAMATrusted Source reported associations between higher butter intake and higher mortality, and between higher plant oil intake and lower mortality, using long-term cohort data.

Why the American Heart Association highlighted it

The video references an American Heart Association write-up that emphasized the “swap butter for plant oils” message.

From a public health communications standpoint, this is simple and actionable.

From a scientific standpoint, the video argues it may be oversimplified, because the model is only as good as the data and the assumptions behind it.

For broader context, the American Heart Association has long recommended limiting saturated fat and replacing it with unsaturated fats, especially polyunsaturated fats, as part of a heart-healthy pattern. You can see their framing in their dietary guidance on saturated fatTrusted Source.

The video’s central critique, apples-to-oranges groups

The key insight in the video is not a biochemical argument about butter versus canola oil.

It is a baseline characteristics argument.

When the expert looks at the tables comparing people who ate the most butter versus those who ate the most plant oils, the groups do not look like the same “type” of people.

They look like different lifestyles.

The two differences the video keeps coming back to

The video highlights two standout differences:

Those are not small details.

Smoking and physical activity are among the strongest predictors of mortality we know. If one group smokes more and moves less, it becomes extremely difficult to conclude that butter is the main driver of the outcome differences.

This is the “apples-to-oranges” point. The comparison is not butter versus oil in a vacuum. It is butter plus a cluster of behaviors versus oil plus a different cluster of behaviors.

Did you know? Smoking is strongly linked to heart disease, stroke, COPD, and many cancers. The CDC summarizes the broad health impacts of smoking in its overview of health effectsTrusted Source.

Why “adjusting for” smoking and exercise is not magic

The paper reportedly adjusted for many factors, including smoking and exercise.

Statistical adjustment is valuable, but the video argues it is not a cure-all.

A short way to say this is: you can adjust numbers, but you cannot perfectly adjust reality.

Residual confounding, the problem that never fully goes away

In observational nutrition research, residual confounding is the idea that even after adjustment, some differences between groups remain, either because they were not measured well or not measured at all.

Smoking is a good example. “Smoker” is not one thing. It can mean pack-years, current versus former, cigar versus cigarette, quitting during follow-up, secondhand smoke exposure, and more.

Exercise is similar. “Exercise” can mean intensity, consistency, strength training, daily steps, occupational movement, and changes over time.

If the study relies on broad categories, or self-report that changes across decades, the adjustment can be incomplete.

That is why the video questions whether it is plausible to attribute a large mortality difference primarily to butter when smoking and exercise differ so sharply between groups.

Pro Tip: When you read a nutrition headline, look for baseline tables. If the “high intake” group also smokes more, sleeps less, drinks more, or exercises less, treat the food claim as a hypothesis, not a verdict.

Why a randomized controlled trial would answer a different question

The video repeatedly contrasts this analysis with a randomized controlled trial.

In a trial, people would be assigned to butter or plant oils, ideally with similar baseline characteristics, and followed for outcomes. That design can better address causality, although long-term diet trials are expensive and difficult.

Observational studies can still be informative, but they cannot fully disentangle whether butter is the cause, or whether butter is a marker for a broader pattern.

Food frequency questionnaires, the measurement problem

Another major theme in the video is that the dietary data came from food frequency questionnaires.

These tools ask people to recall how often they eat certain foods. They are widely used in large cohorts because they are practical.

They are also imperfect.

People forget. People misestimate portion sizes. People change their diets. People report what they think they should eat.

Measurement error tends to be non-random, too. Health-conscious participants may report differently than less health-conscious participants.

This matters because if intake is misclassified, the model may link “butter” to outcomes that are actually driven by other behaviors that travel with butter intake in that population.

A broader scientific discussion of the strengths and limits of nutritional epidemiology has been published over the years, including critiques of how hard it is to infer causality from diet questionnaires. One accessible overview of the challenge of diet measurement and confounding is discussed in reviews like the NIH’s general primer on observational study designsTrusted Source, which explains why associations do not necessarily imply cause.

Butter is down, vegetable oils are up, so why are we sicker?

The video shows a trend chart: butter and animal fats down, vegetable oils up.

Then it asks the uncomfortable question.

If butter is the villain, why did population health not improve as butter fell?

This is not a clean scientific proof, because many things changed at the same time. Food processing increased. Portion sizes grew. Sugar intake patterns shifted. Daily movement decreased. Sleep changed. Stress changed.

The expert’s framing is that obesity and metabolic disease are multifactorial. Focusing on one food, especially in the context of observational associations, can distract from bigger levers.

A strong example is physical activity. Public health agencies consistently emphasize that regular movement reduces risk of cardiovascular disease and improves metabolic health. The CDC summarizes benefits and targets in its physical activity guidelinesTrusted Source.

A second example is ultra-processed foods. Many people are not swapping butter for olive oil in home cooking. They are eating more restaurant and packaged foods, which often include refined carbohydrates, added sugars, and repeatedly heated oils.

The video also calls out repeatedly fried oils as a real-world concern, especially in deep-fried foods eaten frequently.

Single sentence truth: the context matters.

Real-world fat choices without fear or extremes

The video’s tone is skeptical of “butter is deadly” messaging, but it is not a call to eat unlimited butter.

In fact, the speaker explicitly says it is possible to overdo butter, and even agrees with the critique that some low-carb communities may overuse it.

That nuance is the practical takeaway: avoid extremes, and focus on patterns.

How to think about butter in a balanced way

Butter is calorie-dense.

It also contains saturated fat, which can raise LDL cholesterol in many people, although individual responses vary. Major guidelines still recommend limiting saturated fat and replacing it with unsaturated fats in many cases, especially for people at higher cardiovascular risk, as summarized by the American Heart AssociationTrusted Source.

At the same time, butter is also a traditional cooking fat, and for many people it is used in small amounts to improve taste and satiety.

The video’s practical stance is not “never butter.” It is more like: use it intentionally, not as a default calorie add-on.

A simple, realistic approach (without turning meals into math)

If you are trying to decide what to do tomorrow, these are the kinds of choices consistent with the video’s framing.

Short closing thought: dietary fat is not a morality test.

What about specific oils like canola and soybean?

The AHA summary highlighted canola and soybean oils.

From a mainstream nutrition perspective, these oils are often recommended because they are low in saturated fat and contain unsaturated fats. Some evidence suggests replacing saturated fat with polyunsaturated fats can reduce LDL cholesterol and cardiovascular risk markers, which is part of why guidelines emphasize substitution patterns rather than single foods.

From the video’s perspective, the concern is not only chemistry, it is also how these oils show up in the modern diet. They are common in ultra-processed foods and restaurant frying, and intake can become very high without people noticing.

If you are choosing oils for home cooking, many clinicians and dietitians suggest focusing on minimally processed options when possible, and using appropriate heat-stable methods. For example, olive oil is often discussed in the context of Mediterranean dietary patterns, and the evidence base around the Mediterranean diet and cardiovascular outcomes is summarized by sources like the AHA Mediterranean eating pattern overviewTrusted Source.

»MORE: If you want a practical “kitchen audit,” write down every fat source you use for one week, including restaurant meals. Many people discover that most of their “oil intake” is not from home cooking, it is from packaged snacks and fried foods.

Expert Q&A box, “Does this mean butter is safe?”

Q: If this study is observational, does that mean butter is totally safe and plant oils are bad?

A: Not necessarily. Observational findings can still flag potential risks, but they cannot prove cause and effect, especially when lifestyle factors like smoking and exercise differ between groups.

A more useful question is, “What does my overall pattern look like?” If your diet is mostly minimally processed foods, you are active, you do not smoke, and your lab markers are monitored with a clinician, a modest amount of butter may fit for many people. If you have high LDL cholesterol, known cardiovascular disease, or other risk factors, it is worth discussing fat choices with your healthcare team.

Health education summary based on the video’s critique of observational causality

Expert Q&A box, “How much butter is too much?”

Q: What is a healthy amount of butter?

A: There is no single amount that fits everyone, and the video avoids prescribing a number. The practical point raised is that it is easy to overconsume butter because it is calorie-dense and can be added on top of meals.

A reasonable approach is to use butter mainly for cooking and flavor in small amounts, rather than eating large quantities daily. If you are unsure, or if you have heart disease risk factors, consider reviewing your overall fat intake with a registered dietitian or clinician.

General nutrition guidance consistent with the video’s “avoid extremes” message

A brief note on supplements mentioned in the video

Near the end, the speaker mentions a berberine product marketed for metabolic health and suggests using 2 to 3 capsules before mealtime, often in the morning or evening.

Berberine is a bioactive compound that has been studied for effects on blood glucose and lipids, but it can interact with medications and may not be appropriate for everyone. If you have diabetes, take glucose-lowering drugs, are pregnant, or have liver or kidney disease, consult a clinician before using it. For an overview of berberine’s potential effects and safety considerations, you can review summaries like the NIH’s information on supplements via the National Center for Complementary and Integrative HealthTrusted Source.

Single sentence bottom line: supplements should not be used to compensate for smoking, inactivity, or a highly processed diet.

Key Takeaways