For most of the twentieth century, the potato occupied a position of uncomplicated authority on the American dinner plate. It was filling, affordable, and practically a national symbol. Then nutritional science complicated the picture, and the potato became something murkier: a carbohydrate to be watched, a vegetable that some experts stopped calling a vegetable, a food that sat awkwardly between the comfort food category and the chronic disease risk category. The debate never fully resolved. It simply kept accumulating new evidence on both sides.
In August 2025, researchers at Harvard T.H. Chan School of Public Health published findings that drew attention not just for their scale – more than 205,000 participants observed across more than three decades – but for the precision with which they separated the potato question into its component parts. A regular french fry habit, defined as three servings per week, was associated with a 20 percent higher risk of developing type 2 diabetes. Baked, boiled, and mashed potatoes, by contrast, were not.
The findings showed that the cooking method and what potatoes replace in the diet matter more than potato consumption alone.
The Study: Scale, Design, and What Made It Different
The study, published in the BMJ on August 6, 2025, drew on individual participant data from three major longitudinal cohorts in the United States. All 205,107 participants were free of diabetes, cardiovascular disease, and cancer at baseline.
Participants completed validated, semi-quantitative food frequency questionnaires at baseline and updated them approximately every four years thereafter. The repeated measurement approach captured genuine long-term dietary patterns rather than snapshots that might reflect temporary changes in behavior.
Across more than five million person-years of follow-up, 22,299 participants were diagnosed with type 2 diabetes.
What Was Actually Measured
Over an average of 33 years, participants regularly reported their dietary habits, including how often they ate baked, boiled, or mashed potatoes, as well as french fries and whole grains. This distinction between cooking methods was central to the study’s design and, ultimately, to its most important finding.
The study included a substitution analysis examining what happens to diabetes risk when participants swap potatoes for other carbohydrate sources — an approach conducted at a scale and duration that few prior studies had matched.
A dose-response meta-analysis using the study data, along with data from 12 other cohort studies, gave the findings additional robustness beyond the three primary cohorts.
The French Fry Finding: 20 Percent and What It Means
Three servings of french fries weekly were associated with a 20 percent higher risk of developing type 2 diabetes. French fries are among the most widely consumed foods in the United States, and a one-in-five increase in relative risk is not a number researchers typically describe as trivial.
Relative risk figures require context to be genuinely useful. A 20 percent higher relative risk means that if a person’s baseline absolute risk of developing type 2 diabetes over a given period were, for example, 10 percent, a regular french fry habit at that frequency would raise it to approximately 12 percent. The increase is real and consistent across the study’s data, but it does not mean that eating fries three times a week guarantees a diabetes diagnosis.
The elevated risk linked to french fries remained significant even after accounting for factors like trans fat intake, polygenic risk scores, and body mass index. The association was not simply an artifact of the fact that people who eat a lot of fries also tend to have other lifestyle characteristics that independently raise diabetes risk. The fry effect held up on its own.
Why Frying Specifically Raises the Risk
Fried potatoes carry a higher glycemic index and contain added fats, sodium, and potentially harmful compounds from high-temperature cooking. When food is rapidly digested and converted to glucose, blood sugar spikes sharply, prompting the pancreas to release large amounts of insulin. Over years and decades, repeated large insulin demands can degrade the body’s ability to manage blood sugar effectively.
Potato starch and glycemic index are closely linked, and deep frying accelerates that effect rather than mitigating it. The fat content added by frying slows gastric emptying slightly but does not offset the rapid glucose response. And unlike baking or boiling, high-temperature frying generates compounds – including acrylamide – that are not present in the same quantities in other cooking methods.
Baked, Boiled, and Mashed: A Fundamentally Different Picture
French fries were associated with an increased risk of developing type 2 diabetes, while other forms of potatoes – including baked, boiled, and mashed – were not. The study also found that swapping any form of potato for whole grains may lower the risk of type 2 diabetes.
For the large portion of the population that consumes potatoes primarily in their baked, boiled, or mashed forms, the 20 percent figure is essentially not relevant to their eating habits. The risk profile of a plain baked potato with its skin intact is substantially different from a serving of restaurant fries, and the study’s data reflect that difference directly.
Potatoes are high in the kind of carbohydrates that the body digests rapidly, which can cause blood sugar to spike. That remains true for any cooking method. But the additional metabolic burden imposed by frying – the added fat, the higher glycemic load, the heat-induced chemical changes – appears to be what tips the equation from “not a meaningful risk” to “a statistically significant one.”
A medium potato with skin provides about two to three grams of fiber, though most of that fiber comes from the potato peel. Potatoes are also high in potassium, containing about 600 milligrams, though potassium levels are lower in cooked potatoes because the mineral is lost during wet cooking methods such as boiling. The nutritional case for the potato is real – it just depends heavily on what you do to it before it reaches the plate.
The Substitution Analysis: What You Eat Instead Matters as Much as What You Cut Out
The substitution component of the study examined what happens to diabetes risk not just when potato intake increases or decreases, but specifically when potatoes are replaced with other carbohydrate sources.
Choosing whole grains instead of french fries three times per week was associated with a 19 percent lower risk of type 2 diabetes. Replacing them with non-starchy vegetables was also associated with a meaningfully lower risk. These reductions carry a practical implication that the raw french-fry-risk number does not: the goal is not simply to avoid fries, but to replace them with something that actively benefits metabolic health.
Replacing any form of potato with whole grains was tied to a lower risk of type 2 diabetes, while swapping them for white rice was linked to a higher risk. White rice has long occupied a complex position in dietary guidance, and this data suggests that substituting it for potatoes does not produce the health benefit that many might assume.
The substitution analyses relied on statistical modeling rather than actual dietary replacements. Real-world dietary substitutions are not controlled experiments; people who swap french fries for brown rice on a Tuesday may also make dozens of other dietary choices that day that the model cannot capture. But the statistical approach used here is standard in nutritional epidemiology and, across a dataset of this size and duration, produces findings that are genuinely informative.
Historical and Dietary Guideline Context
Potatoes have occupied an uncertain position in American dietary guidance for more than two decades, and the tension between their nutritional value and their glycemic profile predates this research by a significant margin.
In early 2024, there was active discussion about whether potatoes might be reclassified from a vegetable to a grain in the federal government’s Dietary Guidelines for Americans 2025-2030, because of the way their nutrient profile differs from that of other vegetables. That debate reflected an ongoing institutional ambivalence: potatoes deliver real nutrients, but they behave metabolically more like a starchy grain than like broccoli or spinach.
Potatoes do not count as vegetables on Harvard’s Healthy Eating Plate.
A 2024 study in the Journal of Nutrition, conducted in Norway where boiled potatoes are a dietary staple, offered a counterpoint worth considering: researchers found that people who ate around 35 ounces of boiled potatoes per week had a 12 percent lower risk of dying from any cause compared to those who ate about half that amount. The Norwegian data does not contradict the Harvard findings. It reinforces them, since the Norwegian population consumes potatoes predominantly in boiled form. What it adds is evidence that, in the right preparation context, potato consumption can be associated with a mortality benefit.
The cumulative picture that emerges across these studies is consistent: the potato itself is not the primary variable. The cooking method is.
Expert Response and Practical Implications
Corresponding author Walter Willett, professor of epidemiology and nutrition, said that “limiting potatoes – especially limiting french fries – and choosing healthy, whole-grain sources of carbohydrate could help lower the risk of type 2 diabetes across the population.”
Willett added that for policymakers, the findings highlight the need to move beyond broad food categories and pay closer attention to how foods are prepared and what they’re replacing, noting that “not all carbs – or even all potatoes – are created equal.”
Lead author Seyed Mohammad Mousavi, a postdoctoral research fellow in Harvard’s Department of Nutrition, described the study as offering “deeper, more comprehensive insights by looking at different types of potatoes, tracking diet over decades, and exploring the effects of swapping potatoes for other foods.”
Mousavi encouraged patients to “prepare potatoes in healthier ways, like baking or boiling, and to prioritize whole grains as their main carbohydrate source.”
For the average adult managing diet and metabolic risk, those recommendations have the merit of specificity. They do not require eliminating a beloved food category. They require preparation choices and substitution awareness. Metabolic risk is rarely about single foods in isolation; it is about patterns, frequencies, and the cumulative effect of what those foods replace. The link between diet quality, blood sugar regulation, and long-term metabolic conditions extends across multiple lifestyle domains.
Study Limitations and What Remains Unanswered
The Harvard team was explicit about several limitations. The study population was predominantly U.S.-based, predominantly health-focused (nurses and health professionals), and predominantly white. Whether the findings translate with the same magnitude to populations with different genetic risk profiles, different baseline diets, or different potato-consumption habits is not definitively established.
The substitution analyses relied on statistical modeling rather than actual dietary replacements. No participant was randomized to swap their fries for brown rice and then monitored prospectively. That limitation is standard for long-duration nutritional research, where controlled feeding trials over 30-year periods are not feasible.
What the study does not address: the role of portion size in modifying risk within the french fry category, the specific contribution of frying oils versus the starch itself, and whether emerging potato preparations – such as air frying, which substantially reduces added fat – produce a different risk profile than traditional deep frying. Virginia Polytechnic Institute and State University registered a randomized controlled feeding study in 2025 specifically examining the cardiometabolic effects of potato-rich diets in older adults at elevated diabetes risk, which may provide more controlled data in the years ahead.
What This Means for Your Plate
The Harvard potato study – the most comprehensive potato health study conducted to date – does not deliver a verdict that potatoes are dangerous. It delivers something more precise: french fries, consumed three or more times per week, carry a 20 percent elevated risk of type 2 diabetes; baked, boiled, and mashed potatoes do not produce the same signal; and replacing any form of potato with whole grains is associated with meaningfully lower diabetes risk.
The study’s substitution analysis moves the policy conversation away from categorical food bans and toward preparation-method literacy and meal-substitution awareness.
Several questions remain open, including the specific role of frying oils, the effects of lower-fat cooking methods like air frying, and whether the findings generalize across more diverse global populations. A new randomized controlled trial registered in 2025 may begin to address the controlled-feeding gap in the evidence base over the coming years. In the meantime, the core implication of this body of research is durable: preparation method is not a footnote to the potato question. It is the central variable.
How a potato arrives on your plate matters considerably more than whether it does.
Disclaimer: This information is not intended to be a substitute for professional medical advice, diagnosis, or treatment and is for information only. Always seek the advice of your physician or another qualified health provider with any questions about your medical condition and/or current medication. Do not disregard professional medical advice or delay seeking advice or treatment because of something you have read here.
AI Disclaimer: This article was created with the assistance of AI tools and reviewed by a human editor.