The study, which tracked the dietary habits of over 200,000 individuals for three decades, found that diets rich in butter but low in plant oils were linked to an increased risk of mortality.
A new study conducted by researchers from Mass General Brigham, the Harvard T.H. Chan School of Public Health, and the Broad Institute of MIT and Harvard suggests that replacing butter with plant-based oils may offer significant health benefits, including a reduced risk of premature death.
Analyzing dietary and health data from 200,000 individuals over more than 30 years, the researchers found that higher consumption of plant-based oils—particularly soybean, canola, and olive oil—was linked to lower overall mortality, as well as reduced deaths from cancer and cardiovascular disease. In contrast, butter consumption was associated with an increased risk of both total mortality and cancer-related deaths.
The findings were published in JAMA Internal Medicine and were also presented at the American Heart Association’s EPI/Lifestyle Scientific Sessions.
Significant Health Impact of Swapping Butter for Oil
“What’s surprising is the magnitude of the association that we found — we saw a 17% lower risk of death when we modeled swapping butter with plant-based oils in daily diet. That is a pretty huge effect on health,” said study lead author Yu Zhang, MBBS, research assistant at the Channing Division of Network Medicine at Brigham and Women’s Hospital, a founding member of the Mass General Brigham healthcare system. Zhang is also a student in the Department of Epidemiology at Harvard Chan School.
A key difference between butter and oil is the types of fatty acids contained in them. Butter is rich in saturated fatty acids, while plant-based oils have more unsaturated fatty acids. While there have been many studies on dietary fatty acids, fewer studies have focused on their primary food sources, including butter and oils. Many previous studies have looked at a person’s diet at a point in time and have been done in a small population, limiting their applicability to public health.
The new study analyzed dietary data from 221,054 participants in the Nurses’ Health Study (NHS), Nurses’ Health Study II (NHSII), and Health Professionals Follow-up Study (HPFS). Every four years, they answered questions about how often they consumed certain types of food. The researchers used the data to estimate how much butter and plant oils they ate.
Total butter intake included butter from butter and margarine blend, spreadable butter added to food and bread, and butter used in baking and frying at home. The intake of plant-based oils was estimated based on the reported use in frying, sautéing, baking, and salad dressing.
The researchers also identified participants who had died and their causes of death. Using statistics to compare death rates across different diet intake levels, the researchers found that participants who ate the most butter had a 15% higher risk of dying than those who ate the least. In contrast, those who ate the most plant-based oils had a 16% lower risk of death than those who ate the least.
The Benefits of a Simple Dietary Change
“People might want to consider that a simple dietary swap — replacing butter with soybean or olive oil — can lead to significant long-term health benefits,” said corresponding author Daniel Wang, MD, ScD, of the Channing Division of Network Medicine at Brigham and Women’s Hospital. Wang is also an assistant professor in the Department of Nutrition at Harvard Chan School and an associate member at the Broad Institute of MIT and Harvard. “From a public health perspective, this is a substantial number of deaths from cancer or from other chronic diseases that could be prevented.”
The researchers also did a substitution analysis, which mimics how swapping butter for plant oils would impact health in a feeding trial. They found that substituting 10 grams of butter a day (less than a tablespoon) with equivalent calories of plant-based oils could lower cancer deaths and overall mortality by 17%.
“Even cutting back butter a little and incorporating more plant-based oils into your daily diet can have meaningful long-term health benefits,” Wang said.
One limitation of the study is that the participants are mainly health professionals, so they might not represent the U.S. population as a whole, the researchers said. In the future, they’d like to study the biological mechanisms underlying why this dietary change has such a large impact.
Reference: “Butter and Plant-Based Oils Intake and Mortality” by Yu Zhang, Katia S. Chadaideh, Yanping Li, Yuhan Li, Xiao Gu, Yuxi Liu, Marta Guasch-Ferré, Eric B. Rimm, Frank B. Hu, Walter C. Willett, Meir J. Stampfer and Dong D. Wang, 6 March 2025, JAMA Internal Medicine.
DOI: 10.1001/jamainternmed.2025.0205
Funding: This study was supported by research grants from the National Institutes of Health (UM1 CA186107, P01 CA87969, R01 HL034594, R01 HL088521, U01 CA176726, U01 HL145386, U01 CA167552, R01 HL35464, R01 HL60712, P30 DK46200, R00 DK119412, R01 AG077489 and R01 NR019992).
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
23 Comments
I am a medical anthropologist researcher and author. All fats and oils are prone to contamination by herbicides, pesticides, fungicides, and other chemical contaminants. Butter may contain hormones and other biologically active compounds. Studies like this that ignore the contamination of these fats are misleading, since the effects may be more related to the contaminants than to the fatty acids. This is why so many nutrition studies are flawed and yield conflicting results. Each study’s oils and fats can be differently contaminated. And since this particular study looked at oil and fat consumption over decades, there is no way to know their quality. And no amount of computer modeling will get around this problem.
For more about nutrition research, see my article, Why Nutrition Research is Unreliable. https://www.academia.edu/126067247/Why_Nutrition_Research_is_Unreliable
Who comes up with these theories?
I feel like this study is crap . 100 years ago and further back in history the only thing used was butter and lards . Our ancestors lived much longer and healthier lives with less sickness and diseases.
This is a government funded trial. Nothing more to say about that!
> Our ancestors lived much longer and healthier lives with less sickness and diseases.
Um, no. No they didn’t. Life span has been increasing dramatically, general overall health is better, and there is much less sickness and disease than in the past. I am 60, and in excellent health due to modern medicine and health practices. In the past, even if (and that’s a BIG “if”) you made it to 60, it was considered “old” and on death’s door.
Thank you!
Thank you for this very plausible argument. Although I have been thinking about food contamination in general, not in conjunction with studies like the one mentioned. I bet they neither corrected for supplement intake (Omega3, Vitamin C, Cholecalciferol etc.) either.
I can make no personal use of the study as long as it’s behind a paywall. The abstract gives a bare hint of the confounding factors. An earlier paper on the same study by Yu, et al. (2019) gives a fuller picture:
“This [2019] study had several limitations.
First, despite full adjustment for acknowledged confounders, we could not exclude the possibility of residual or unmeasured confounding.
Second, the causality could not be established due to the observational study setting.
Third, measurement errors may be still present in the analyses although the portion size of food was specified in the food frequency questionnaires.
Finally, only a single measurement for dietary intake…”
etc.
To cut to the old correlation v. causation chase: are people naturally, or unnaturally (eg. stressed), predisposed to an earlier demise also predisposed to eating more butter? There’s such a thing as “comfort food”. There will be a large proportion in a nursing profession to have high levels of career stress. I can say from experience that I’m not going to seek compensatory comfort frying bean-sprouts in canola oil, and I’ll prefer garlic butter with my bread to dipping breadsticks in vegetable oil at Olive Garden.
*…in t h e nursing profession w h o have high levels of career stress…
Personally, it has been revealed to me that highly processed seed oils are definitely worse than organic butter and olive oils. Basically knowing the difference between organic / first cold pressed and processed is essential.
No, the main reason butter is unhealthy is not hormones, or contaminates. It is Advanced Glycation End-products (AGEs). Butter has the most AGEs, of any food tested other than bacon (though, bacon grease would certainly be higher). AGEs are created by high temperature cooking (anything raising the temp of food much beyond boiling water), and more relevantly, in this case, by blending machines, but also some forms of fermentation (fermentation that does not also increase acidity, such as in hard cheese like Parmesan). And, there are just a few foods that are high naturally: pine nuts, and cashews. Margarine is also high, because it is also blended by machines. Oils, in general, tend to be higher in AGEs. Soybean oil, Olive oil, and Canola are much lower than butter, but are far from the lowest. The choices lowest in AGEs are Corn oil, Sunflower oil, Safflower oil, and Avocado oil. Pistachio oil should be very low, but mostly I have seen that sold for skin products rather than as food.
AGEs are molecules where a carbohydrate molecule has bonded with either a fat or protein molecule, making a molecule that is hard to get rid of. At low levels in food, healthy kidneys can remove most of it. There is one: Glucosepane, which can not be removed from the body. This one accumulates in the intercellular matrix stiffening tissues, and also causes inflammation through the RAGE pathway. AGEs are associated with every major disease and deleterious condition of aging, especially kidney disease, Alzheimer’s disease, diabetes, and atherosclerosis. They contribute to the creation of wrinkles, and cataracts.
The median increase in lifespan of mice fed a diet with 1/2 the AGEs was 15% longer, and they were much healthier through their lives than the mice fed normal AGE food. The food was identical except the AGEs. They did not put some large amount of AGEs in the food. They just reduced the AGEs in the low AGEs group. The Study: “Reduced Oxidant Stress and Extended Lifespan in Mice Exposed to a Low Glycotoxin Diet” showed many health advantages to reducing AGEs.
Virtually all processed food, and fast food, has high AGEs. In my opinion, this is the main reason processed food, and fast food, is unhealthy. And the increase of these high AGE foods is likely the reason for the increase in type 2 diabetes and metabolic syndrome in the US.
While people may challenge this study based on things that may be omitted by the lack of specifics (organic, pesticide contamination, etc…), the article offers a disclaimer that the Healthcare professionals studied, may not represent the general population. Studies done using animals can be controlled down to the tiny details, but long term studies involving thousands of people cannot. Most of us, do not eat specifically organic food, but some do, so I would posit that the information gleaned would be useful to multitudes.
You act like a larger sample is less accurate, but that is not what tens of thousands of studies of all kinds have shown. 200,000 is plenty, and can resolve small differences. But the difference found was huge. The probability that it is by chance different would be astronomical.
Grasping at straws.
“Healthy user bias,” is about the only challenge with a slim chance of being true. But this is easy to check for, so it is likely they did this.
Honestly, the number of grumbles and doubts expressed are silly, and likely motivated by people who want to eat butter without negative feelings. Rationalizations.
If the harm from butter is due to Advanced Glycation End-products (which I think is likely), there is a partial workaround. Heavy cream has 1/10 the AGEs, so it could be substituted for butter, assuming no mechanical blending or high temperature cooking is involved.
This was interesting until it got to butter/margarine products. The trans fats in margarines are a huge health risk. Also I’d like more information on the processing of plant oils. I’d eat cold pressed olive oil any day, but not ultra processed seed oils.
Agreed! How in the heck did they include freaking margarine intake the same as butter? Also missing is a separate group for Extra Virgin or cold pressed plant oils. Cheaper oils lose the healthiest part of the oils. I used Unrefined Hemp Seed Oil, which helped heart blood work and Eczema patients vs EV Olive oil. Also missing is grass-fed butter as a separate group.
I thought trans fats have been removed from margarines. The ones I have checked out have no trans fats. Is this not the case? Some are now using fully hydrogenated oils (no trans fats), various thickeners, and/or tropical oils (that add saturated fats).
I find it unbelievable that this “research” treats ALL plant based oils as one and the same. They are NOT the same!
Olive oil is good and healthy, but…
Sunflower oil is THE chief culprit in the rise of cancer, diabetes and cardiovascular diseases in the world. It can be seen when historical charts of sunflower oil consumption per country are layered over disease incidences in respective countries. Coconut oil is just as bad.
Talking about all the plant based oils in the same breath is not only unscientific, but plainly idiotic.
I strongly disagree. The harmful substances created in oil processing are Advanced Glycation End-products. These can be measured, and are much higher in olive oil, than in sunflower oil. You are getting bogus info from YouTubers appealing to the masses to get more views rather than presenting accurate info.
“Oil, olive, extra virgin, first cold pressed (Colavita, Linden, NJ) 10,040 AGE kU/100 mL ”
“Oil, sunflower (The Hain Celestial Group, Inc) 3,940 AGE kU/100 mL”
“Oil, safflower (The Hain Celestial Group, Inc, Melville, NY) 3,020 AGE kU/100 mL”
From: “Advanced Glycation End Products in Foods and a Practical Guide to Their Reduction in the Diet” 2013
Though, I do agree that coconut oil is a poor choice.
Boba, that is factually incorrect. All the “bro-science” in the world from online “influencers” cannot contradict actual science on seed oils.
The only evidence of seed oils being bad for you, are when they are highly heated and re-used (ie. in deep-frying restaurant settings) because it can change the properties of the oil. But that is also the same case with ANY fat.
There is a massive experiment regarding this phenomenon that has been taking place in Europe for 100’s if not thousands of years — there is a line that traverses Europe, running through France, north of which traditional cooking uses butter and south of which traditional cuisine uses olive oil. If there was a tremendous difference in the health outcomes of using these two ingredients, wouldn’t it likely be starkly obvious there?
“In the future, they’d like to study the biological mechanisms” – and that is the main failure of using studies like this for policy decisions or personal health decisions.
These studies look at the total dietary profile of the responders, pick a couple of items out of that list, and compare them. But without any examination of what’s actually happening to the billions of cells in each of those people’s bodies when they eat either butter or plant oil, and how that relates to all the other things they eat or absorb from their environments – this is very preliminary, and can only hint at things which might be happening.
Without any study of cellular mechanisms, the bad effects may be coming from something different in their diets, or something which is not in their diet at all, like environmental chemical pollutants which may happen to be more prevalent in countries which eat more butter.
So – this study was cheap to run, since it used someone else’s data and was just a matter of number crunching to test the researchers’ assumptions. But it was scientifically of very limited use, since it has so many methodological holes. Since cheap studies are as good for your academic standing as excellent but costly studies, that’s what we get – and journalists who don’t seem to understand the requirements needed to prove causation eat these cheap and shoddy studies up, they make for great clickbait articles. .
Advanced glycation end products (AGEs) are a heterogenous group of structures formed under conditions of high oxidative stress or hyperglycemia ( high blood sugar ).
A Keto or carnivore diet which includes butter prevents high blood sugar, so AGEs would have little effect, while a high carb diet produces high blood sugar, so AGEs would have strong effect.
I eat no fried food…. this study report was financed by the oil companies…what a scam !! All oils when fried , heated are toxic.