Plant oils outperformed butter in a decades-long mortality analysis
A new Harvard analysis adds fresh weight to a long-running nutrition debate: butter tracked with higher mortality, while plant oils tracked with longer survival. The findings come from a large observational study published in JAMA Internal Medicine that followed dietary patterns across decades and compared death rates among adults with the highest and lowest reported intakes of these fats.
The signal was not subtle. Participants with the highest butter intake had a 15% higher risk of death than those with the lowest intake, while those consuming the most plant oils had a 16% lower risk of death than those consuming the least. In a substitution analysis, replacing about 10 grams of butter a day with certain plant oils was associated with fewer total deaths and fewer cancer deaths. Ten grams is a little less than a tablespoon, a small change on a plate that may matter over years of daily eating.
The oils that tracked with benefit were not all the same
Five plant oils were examined, but the survival association was strongest for soybean, canola, and olive oil. That detail matters because it keeps the finding from being reduced to a vague endorsement of “more oil” in general. Food fats differ in fatty acid profile, oxidation behavior, and their effects on lipids and inflammation, and those differences can shape cardiovascular and cancer risk over time.
Butter contains almost no essential fatty acids and a modest amount of trans fat, a lipid pattern long linked to worse cardiovascular risk. Plant oils, by contrast, are rich in unsaturated fats and often provide antioxidants and essential fatty acids. Those nutrients can improve cholesterol and triglyceride profiles, reduce insulin resistance, and dampen chronic inflammatory pathways that contribute to atherosclerosis and other long-horizon disease processes.
One sentence captures the biological logic plainly: saturated fat pushes risk in one direction, while unsaturated fat pulls it the other way.
Cancer and cardiovascular deaths moved in the same direction
The mortality pattern extended beyond all-cause death. Higher butter intake was associated with greater cancer mortality, while higher plant oil intake was associated with lower risk of dying from cancer or from cardiovascular disease such as stroke or heart attack. That convergence is clinically important because it suggests the diet change may influence more than one disease pathway at once.
Cardiovascular biology offers a plausible explanation. Replacing saturated fat with unsaturated fat can improve lipid handling in the liver, lower atherogenic particles in circulation, and reduce the inflammatory burden that destabilizes arterial plaque. Cancer biology is less direct, but long-term diet patterns can influence insulin signaling, adiposity, inflammatory tone, and cellular growth pathways, all of which may help explain why a more plant-forward fat pattern aligned with better survival.
The finding is still observational, so it cannot prove that butter caused harm or that plant oils directly caused benefit. It does, however, fit closely with established cardiometabolic physiology and with prior dietary guidance.
Why the substitution analysis matters at the bedside
A substitution model is more useful than a simple comparison of high versus low intake because patients rarely ask whether to eliminate one food in isolation. They ask what to use instead. Here, the answer was concrete: a modest daily exchange of butter for soybean, canola, or olive oil tracked with fewer deaths.
For clinicians counseling patients, that is a practical food swap rather than a sweeping diet overhaul. A patient who cooks eggs in butter every morning or spreads butter at the table could shift part of that habit toward liquid plant oils without changing the entire menu. The study also leaves room for taste and culinary reality. A small amount of butter for flavor is not the same as making butter the default fat used for most cooking.
Dr. Walter C. Willett summarized the logic directly: use liquid plant oils whenever possible instead of butter for cooking and at the table. In practical terms, that recommendation aligns with American Heart Association guidance and the Dietary Guidelines for Americans.
The cohort was large, but the design still has limits
The analysis drew on questionnaire data collected every four years from more than 221,000 adults in the Nurses’ Health Study, Nurses’ Health Study II, and Health Professionals Follow-Up Study. Follow-up extended for up to 33 years, and investigators adjusted for many factors that could affect health, including age, physical activity, smoking status, and family history of illnesses such as cancer and diabetes.
Those features strengthen the analysis. The large sample, long follow-up, and repeated dietary measures reduce some of the noise that often weakens nutrition research. Even so, observational nutrition studies carry a familiar constraint: they can show association, not causation. People who choose more plant oils may differ from people who rely on butter in other ways that are difficult to fully measure, even after statistical adjustment.
Another limitation is the population itself. Most participants were white health care professionals, so the results may not generalize cleanly to broader communities with different diets, incomes, cultural food patterns, or baseline disease risk.
What the study adds to day-to-day prevention counseling
Despite those limits, the study adds a strong, duration-based argument for replacing butter with plant oils where possible. The result is especially useful because it does not ask patients to abandon familiar foods altogether. It asks them to change the fat source used most often, and the direction of the association is consistent across total mortality, cancer death, and cardiovascular death.
For patients trying to make one realistic dietary change, the message is uncomplicated: liquid plant oils appear to be the better default fat than butter. Olive oil, canola oil, and soybean oil carried the most favorable signal in this analysis, and the estimated benefit emerged from a swap small enough to fit into everyday cooking. The research will not settle every debate about dietary fat, but it does sharpen the clinical argument for trading some butter for unsaturated oils when the goal is longer survival.