When researchers at the Institute for Health Metrics and Evaluation set out to rank every cause of death on the planet, tobacco and alcohol topped most people’s mental lists. The answer that emerged from the data was something far more mundane: the food on our plates. A landmark 2019 analysis published in The Lancet by the GBD Diet Collaborators found that poor dietary quality was responsible for 11 million deaths in 2017 alone – more than smoking, more than any infectious disease, and more than any single clinical condition.
What made the findings particularly striking was not just the scale but the nature of the problem. The study, which tracked dietary patterns across 195 countries from 1990 to 2017, found that most of those deaths were not caused by eating too much of the wrong things. They were caused by not eating enough of the right ones. Deficits in whole grains, fruit, nuts, seeds, and vegetables drove a larger share of mortality than excess sodium, saturated fat, or sugar combined. That inversion of conventional public health messaging has reshaped how nutritionists and policymakers think about dietary intervention.
Understanding exactly which gaps cause the most harm – and in which populations – gives individuals and healthcare systems a clearer map for intervention than any blanket “eat less processed food” directive could provide.
The Three Dietary Risks Driving Half of All Diet-Related Deaths
The GBD Diet Collaborators analysed 15 dietary factors against optimal intake levels derived from meta-analyses of prospective studies. Three factors stood out as responsible for more than half of all 11 million diet-attributable deaths and 66 percent of the 255 million disability-adjusted life-years (DALYs) lost to poor diet in 2017.
High sodium intake caused approximately 3 million deaths and 70 million DALYs. The study’s optimal sodium intake was set at 3 grams per day, a level most high-sodium populations significantly exceed. In China, Japan, and Thailand, excess sodium was identified as the single leading dietary risk factor for death. In China alone, where preserved and heavily salted foods are cultural staples, the sodium burden is among the highest recorded anywhere.
Low whole grain consumption caused another 3 million deaths and 82 million DALYs – the highest DALY burden of any single dietary factor in the study. The optimal daily intake was 125 grams, yet in major economies including the United States, India, Brazil, Russia, and Germany, inadequate whole grain consumption was identified as the top dietary killer. Whole grains carry protective effects across cardiovascular disease, type 2 diabetes, and several cancers through mechanisms that include fermentable fibre feeding the gut microbiome, displacement of refined carbohydrates, and direct effects on insulin sensitivity.
Low fruit intake caused around 2 million deaths and 65 million DALYs. The study set an optimal intake at 250 grams per day – roughly two to three pieces of fruit. In Bangladesh, insufficient fruit consumption was the single largest dietary risk factor for death, reflecting a pattern common across lower-income countries where energy-dense staples crowd out micronutrient-rich produce.
Why Deficiency Outweighs Excess in Driving Mortality
The framing of diet-related mortality almost always centres on excess: too much fat, too much sugar, too much salt. The GBD analysis challenges that framing in a direct way. Across all 15 dietary factors assessed, more deaths were attributable to underconsumption of protective foods than to overconsumption of harmful ones. Nuts and seeds, for instance, were consumed at just 12 percent of their optimal level globally. Whole grains sat at roughly 23 percent of optimal intake. Milk reached 16 percent.
Christopher Murray, director of IHME and one of the lead authors, put it plainly: “Poor diet is responsible for more deaths than any other risk factor in the world.” The mechanism is straightforward but underappreciated. Foods like whole grains, fruit, nuts, and legumes do not merely add nutrients – they occupy dietary space that would otherwise be filled by energy-dense, nutrient-poor alternatives. Their absence therefore has a compounding effect: the person eating too little fruit is almost certainly eating something else instead, and that substitution adds a second layer of risk.
This matters for how dietary advice is communicated. Telling someone to “cut back on salt” or “avoid ultra-processed food” leaves an empty space. Telling someone to add 125 grams of whole grains to their daily diet, or to eat two pieces of fruit before noon, is both positive and specific – and according to the GBD data, far more effective at reducing mortality at a population level.
Country-Level Variation: From Uzbekistan to Israel
The GBD study’s country-level data reveals enormous variation in diet-attributable mortality rates. Uzbekistan recorded the highest age-standardised diet-related death rate at 892 per 100,000 population. Afghanistan, the Marshall Islands, Papua New Guinea, and Vanuatu also ranked among the worst performers. These nations share a combination of high sodium foods, low fruit and vegetable availability, and diets heavily reliant on refined grains and preserved animal products.
At the other end of the spectrum, Israel recorded the lowest age-standardised rate at 89 per 100,000. France, Spain, Japan, and Andorra followed as the best-performing larger nations. Japan’s position is notable given its relatively high sodium burden from soy sauce and fermented fish products; its low overall mortality rate reflects strong offsetting factors including high fish intake, green tea polyphenols, and traditionally high vegetable consumption. Spain and France benefit from Mediterranean-adjacent dietary patterns with high fruit, olive oil, legume, and moderate wine consumption.
Egypt showed the highest diet-attributable death rate among the world’s 20 most populous countries, at 552 per 100,000. Japan, at 97 per 100,000, showed the lowest among the same group. The gap between those two rates – a 5.7-fold difference – demonstrates that dietary choices at a population level are among the most powerful determinants of lifespan.
The Five Dietary Shifts With the Largest Mortality-Reduction Potential
The GBD collaborators used their data to model which individual dietary changes would deliver the largest reductions in diet-attributable mortality if adopted broadly. The results point to a short, specific list of additions rather than a complex dietary overhaul.
Increasing whole grain consumption to the 125-gram daily optimum emerged as the single highest-impact shift, capable of preventing 3 million deaths annually if adopted globally. The practical form of this change is modest: replacing white rice or refined bread with brown rice, oats, whole wheat bread, or barley at one meal per day would get most people much of the way there. The second most impactful shift is reducing sodium toward the 3-gram daily target – achievable primarily by cutting reliance on processed, preserved, and restaurant-prepared foods rather than removing the salt shaker from the table, since discretionary salt typically represents only 10-15 percent of dietary sodium.
Increasing fruit intake to 250 grams daily represented the third priority. After that, the model pointed to increasing nut and seed consumption toward the 21-gram daily optimal – a small handful – and increasing legume intake. These five shifts share a common feature: they add volume and diversity to the diet rather than subtracting specific items. In trial settings, additive dietary advice consistently outperforms restrictive advice in long-term adherence, which may help explain why populations with varied, plant-inclusive diets sustain lower mortality rates decade over decade.
How Diet Compares to Other Global Risk Factors
The GBD 2017 data places diet in a striking position relative to all other modifiable risk factors. Tobacco use, long considered the leading preventable cause of death, was responsible for approximately 8 million deaths in 2017. Diet, at 11 million, exceeds it by nearly 40 percent. High blood pressure caused around 10.4 million deaths – but blood pressure is itself substantially influenced by dietary sodium, making diet a partial upstream cause of that figure too.
The disease-specific breakdown is also revealing. Cardiovascular disease accounted for 10 million of the 11 million diet-attributable deaths. Cancer claimed 913,000, and type 2 diabetes approximately 339,000. This distribution reflects the established mechanisms by which dietary quality affects vascular biology: chronic intake patterns influence LDL cholesterol, blood pressure, systemic inflammation, endothelial function, and insulin sensitivity over years and decades rather than in acute episodes. A 2019 editorial in The Lancet accompanying the study noted that “global diet and health” demands the same policy infrastructure applied to tobacco – front-of-pack labelling, taxation on harmful foods, and subsidy reform for whole grains, fruits, and vegetables.
What the Data Actually Demands of Individuals
Large-scale epidemiological studies often feel remote from individual decision-making. The GBD diet analysis is actually unusually actionable because it identifies specific food groups and specific shortfalls. For most people in high-income countries, the evidence points to a narrow set of changes: eat more whole grains, eat more fruit, reduce heavily processed and preserved foods, add a small daily portion of nuts or seeds, and include legumes regularly. None of these require expensive supplements, restrictive eating windows, or unusual ingredients.
The gap between current population intakes and optimal intakes is wide enough that even partial progress yields meaningful benefit. Moving from consuming 10 percent of optimal whole grain intake to 50 percent does not require perfection – it requires consistency. The GBD collaborators note that mortality trends improved between 1990 and 2017 in some regions, particularly where fruit and vegetable availability increased with economic development. That improvement is real and measurable, and it demonstrates that dietary patterns at a population level are neither fixed nor intractable.
This article is for general informational purposes only and is not a substitute for professional medical or nutritional advice. If you have a health condition or specific dietary needs, consult a qualified healthcare provider.