FAOSTAT - What Is It?

FAOSTAT is one of the most widely cited agricultural databases in the world, and for good reason. It is operated by the Food and Agriculture Organization of the United Nations (FAO), and it provides global data on food production, trade, and availability. But despite its importance in agricultural and economic analysis, FAOSTAT is often misunderstood and misused when people try to apply it directly to questions of human nutrition, consumption, health outcomes, and disease. One recent trend in online nutrition debates, particularly among advocates of extreme carnivorous or meat-heavy diets, has been to cite FAOSTAT as though it were the ultimate proof that high meat intake correlates with better life expectancy, lower disease risk, or superior health. This is a misapplication of the database. FAOSTAT was never designed to provide actual consumption or dietary intake data, and it is certainly not a reliable or valid basis for comparing population health outcomes like longevity or chronic disease prevalence against so-called “meat consumption.”

To understand why, we first need to look at what FAOSTAT actually measures. The database relies on national agricultural statistics and trade records to create what are known as Food Balance Sheets (FBS). These balance sheets take the total quantity of a given food produced in a country, add imports, subtract exports, subtract animal feed, subtract industrial uses, and subtract some post-harvest and distribution losses. What remains is considered “food available for human consumption.” That figure is then divided by the national population to produce a per capita food availability number, often expressed in kilograms per person per year or grams per person per day. This is the number that many people mistakenly label as “consumption.”
The problem is clear: availability does not equal consumption. The amount of food available in a country is not the same as what individuals actually eat. These figures cannot account for retail waste, plate waste, unequal distribution of food, socioeconomic disparities, or cultural dietary patterns. For example, if FAOSTAT reports that the United States has 100 kilograms of beef per capita per year available, that does not mean every American consumes that much beef. Some people eat more, some eat none at all, and a large amount may be wasted or spoiled before reaching the plate. This gap between availability and real intake has been well documented. Dietary surveys, such as the National Health and Nutrition Examination Survey (NHANES) in the United States, consistently show lower actual intake levels than the FAOSTAT availability data would suggest.

Now, if the purpose of FAOSTAT is misunderstood, the consequences become more serious when people start trying to link FAOSTAT “consumption” numbers to health outcomes like life expectancy or disease rates. Consider what is happening in the carnivore community: proponents will point to FAOSTAT numbers showing that high-income countries have greater meat availability, and then note that those same countries have longer life expectancy than many low-income countries. From this, they claim that high meat consumption must be responsible for longevity and health. This is both scientifically inaccurate and logically flawed. High-income countries have longer life expectancy because of complex factors: access to healthcare, sanitation, clean water, antibiotics, vaccination programs, education, and wealth. To attribute longevity simply to meat intake is to commit a classic ecological fallacy, where group-level data are misused to make individual-level claims. FAOSTAT data cannot prove what individuals are eating, and it certainly cannot prove causality between meat intake and life expectancy.

Furthermore, poorer countries often have underdeveloped statistical systems, meaning their FAOSTAT data are less reliable to begin with. Many low-income nations report limited or outdated agricultural statistics, forcing FAO statisticians to use modeling, imputation, or estimates. This means that comparisons across countries are not only comparing availability rather than consumption, but also comparing availability figures with different levels of accuracy. Using these as the basis for sweeping statements about diet and disease is irresponsible.

A deeper issue is that health outcomes like cardiovascular disease, diabetes, obesity, and cancer are not determined by one food group alone. They are the product of overall dietary patterns, physical activity, environmental exposures, genetics, and healthcare access. Studies that have attempted to link red meat intake with chronic disease risk rely on carefully collected dietary intake data from cohort studies and randomized trials, not on agricultural availability figures. For example, the Nurses’ Health Study and Health Professionals Follow-up Study in the United States, which are among the most cited epidemiological datasets in nutrition science, use detailed food frequency questionnaires and follow participants over decades. These studies have shown consistent associations between higher intake of processed and red meats and increased risks of cardiovascular disease, type 2 diabetes, and certain cancers. None of this evidence comes from FAOSTAT.

So if FAOSTAT is not the right tool for this type of analysis, what are the better sources? There are several categories of more reliable data when it comes to actual human consumption and health outcomes.

First are national dietary intake surveys. In the United States, this is NHANES, which combines 24-hour dietary recalls with laboratory biomarkers to provide accurate estimates of food intake and nutrient status. In Europe, the European Food Safety Authority (EFSA) compiles national dietary survey data. Similar national surveys exist in Canada, Australia, Japan, and many other countries. These datasets measure what people actually eat, rather than what is theoretically available in the food system.

Second are large prospective cohort studies. Examples include the Nurses’ Health Study, Health Professionals Follow-up Study, EPIC (European Prospective Investigation into Cancer and Nutrition), and the PURE study (Prospective Urban Rural Epidemiology). These studies track tens of thousands to millions of participants over years or decades, collecting dietary data, lifestyle information, and health outcomes. Because of their design, they are far superior for drawing associations between specific dietary exposures, like meat intake, and health outcomes.

Third are global comparative datasets designed for health analysis. The Global Burden of Disease (GBD) study, coordinated by the Institute for Health Metrics and Evaluation (IHME), compiles data on disease, mortality, and risk factors across the world. Unlike FAOSTAT, which only provides availability, GBD models dietary risk factors using survey data, systematic reviews, and meta-analysis. This makes GBD a much more appropriate reference when discussing how diet relates to mortality and disease at the population level.

Fourth are controlled intervention trials. While more limited in size and duration, randomized trials provide the strongest evidence of causality. For example, the PREDIMED trial in Spain showed that a Mediterranean diet rich in plant foods and moderate in fish and poultry reduced cardiovascular events compared to a control diet. Trials like this are the gold standard for demonstrating causal links between diet and health outcomes.

Finally, systematic reviews and meta-analyses of observational and clinical studies provide the best summary of the available evidence. For instance, a systematic review and meta-analysis in the journal Circulation (Micha et al., 2017) concluded that higher intake of processed meat is associated with increased coronary heart disease, stroke, and diabetes risk. These types of analyses synthesize thousands of data points collected directly from individuals, not national food availability estimates.

Given all this, it becomes clear why FAOSTAT cannot and should not be used as a proxy for human consumption when making claims about health. To recap, FAOSTAT reports availability, not intake. It is dependent on the quality of national agricultural statistics, which vary greatly by country. It does not account for food waste, unequal access, or dietary preferences. And its data cannot be causally linked to health outcomes like life expectancy or disease. Trying to do so is a misuse of the data and leads to misleading conclusions.

Yet, the misuse persists. Carnivore advocates often cherry-pick FAOSTAT graphs showing high meat availability in wealthy countries and juxtapose this with longevity data, suggesting meat explains the difference. This is a dangerous oversimplification. Wealthier countries not only have more meat, but also more fruits, vegetables, medical infrastructure, antibiotics, vaccines, clean water, and education. FAOSTAT alone cannot separate these variables. Claiming that meat availability explains longevity is like claiming that smartphone ownership explains life expectancy — both are more common in wealthy countries, but correlation does not prove causation.

If the goal is to understand how meat intake affects health, far better data sources exist. NHANES in the United States provides detailed intake and biomarker data. EPIC in Europe tracks diet and disease over time in hundreds of thousands of participants. The Global Burden of Disease study integrates worldwide data and provides estimates of the health impacts of dietary risk factors, including high red and processed meat consumption. Randomized trials like PREDIMED provide causal evidence for the benefits of plant-rich diets. These are the types of studies and datasets that nutrition scientists rely on. FAOSTAT, by contrast, is a useful tool for agricultural economists, policymakers, and trade analysts, but it is not a nutrition survey.

Another layer of complexity is that meat itself is not a single homogeneous food. Processed meats (like bacon, sausage, and hot dogs) carry higher health risks than unprocessed red meats, and poultry and fish show different associations altogether. FAOSTAT cannot make these distinctions. It simply reports categories like “beef and buffalo meat” or “poultry meat” by weight, without any information about processing, preparation methods, or portion sizes. In nutrition science, these distinctions matter enormously. Observational studies consistently show that processed meat is associated with greater health risks than unprocessed meat, and that replacing red meat with plant proteins, fish, or legumes reduces risk of chronic disease. These insights cannot be gleaned from FAOSTAT availability data.

It is also worth noting that nutrition science has moved away from focusing on single nutrients or single food groups in isolation. Modern research emphasizes dietary patterns, such as the Mediterranean diet, DASH diet, or plant-forward diets, as better predictors of health outcomes. These patterns consider the balance of foods consumed, not just the quantity of meat. FAOSTAT does not capture patterns; it only tallies kilograms of food available. This makes it unsuitable for pattern-based analysis, which is what health research now relies on.

So, when evaluating the trust value of FAOSTAT in the context of human health, the verdict is clear. It is highly trusted for what it is meant to do: provide standardized agricultural data on production, trade, and availability. But it is not trusted, nor should it be, as a measure of dietary intake or as a dataset for correlating with life expectancy or disease risk. The scientific community uses more precise tools for those questions, ranging from national nutrition surveys to longitudinal cohort studies and randomized trials.

The next time someone tries to cite FAOSTAT as “proof” that meat consumption leads to longer life or lower disease risk, the response should be simple: FAOSTAT is not measuring consumption. It cannot account for distribution, waste, or cultural dietary habits. It cannot be linked directly to health outcomes. The only thing it measures is food availability at the national level. To understand how diet affects health, we need to rely on data sources designed for that purpose, such as NHANES, EPIC, the Global Burden of Disease study, and randomized clinical trials. Anything else is misuse.

For those who want to verify or explore further, the FAOSTAT database itself is freely available at https://www.fao.org/faostat). The Global Burden of Disease project can be accessed at https://www.healthdata.org/gbd). NHANES data are available at https://www.cdc.gov/nchs/nhanes. EPIC data and publications are available at https://epic.iarc.fr. These are the kinds of sources that nutrition and health scientists turn to when studying diet and disease. Using FAOSTAT in this way is like trying to use a trade ledger to measure calorie intake at the dinner table: the numbers simply do not mean what advocates claim they mean.

In conclusion, FAOSTAT is an invaluable agricultural and economic resource, but it is not a nutrition survey and not a health database. Its numbers cannot be used to claim that meat availability equals consumption, nor can they be used to link meat with life expectancy or disease risk. For those discussions, we must rely on direct intake surveys, cohort studies, clinical trials, and global health databases designed for the task. Anything else is not science; it is distortion. And when we see such distortion being used to promote extreme carnivore diets, we should call it what it is: misuse of data to support a preconceived belief. Real science, grounded in the best available evidence, tells a far more nuanced and accurate story about diet, health, and longevity.