F.1 Measuring food and health

Non-communicable diseases such as heart disease, type 2 diabetes and many cancers, which become more common as incomes rise, diets change, activity falls and people live longer.

Why food is difficult to measure

The food a country produces or imports, divided by its population, says little about whether a particular child is well fed. Food may be too expensive for poor households, shared unequally within a family, or poorly absorbed by a body weakened by diarrhoea. Geographers therefore use several indicators, each measuring a different part of the problem.

The four dimensions of food security

Food security exists, in the FAO’s definition, when all people at all times have physical and economic access to sufficient, safe and nutritious food that meets their dietary needs and food preferences for an active and healthy life. The definition contains four dimensions: availability (enough food is present), access (people can obtain it), utilisation (the body makes good use of it) and stability (the other three hold over time). Each dimension needs its own indicators.

A food or nutrition indicator is any measure used to describe the quantity or quality of food available to, or eaten by, a population, or the nutritional status that results from it. Some indicators measure inputs (the food supply), some measure experiences (whether people worry about running out of food) and some measure outcomes in the body (a child’s height for its age).

Calories per person and undernourishment

Dietary energy supply

The most widely used measure of food availability is dietary energy supply, usually called calories per person per day. The FAO calculates it from national food balance sheets: food produced plus imports, minus exports, changes in stocks, animal feed, seed, industrial uses and losses before retail, divided by the population and by 365. The result is the food energy available at retail level, in kilocalories (kcal) per person per day.

The figure measures availability, not what people eat: it includes food later wasted in shops and homes, so supply in high-income countries (HICs) is well above consumption. In 2023 the world average supply was about 3,000 kcal per person per day, having risen from about 2,180 kcal in 1961. The USA and Ireland had about 3,940 kcal, while Somalia (about 1,810 kcal in 2022), Madagascar and Haiti (both under 2,000 kcal in 2023) had among the lowest supplies on Earth.

To judge whether a supply is enough, it is compared with a person’s energy needs. The minimum dietary energy requirement is the energy a person needs for light physical activity and a healthy body weight. It depends on age, sex, body size and activity, so the FAO calculates a separate average for every country from its age and sex structure: across countries it is typically about 1,800 kcal per person per day, lower where a large share of the population are young children.

The scatter graph shows a clear positive relationship between gross national income per capita, at purchasing power parity, and calorie supply. Below about US$5,000 most countries have fewer than 2,600 kcal per person; above US$30,000 most have more than 3,200. The relationship is not linear: on a logarithmic income scale, supply rises steeply at low incomes and levels off at high incomes, because people cannot eat endlessly more food as they become richer. Instead, richer people change what they eat, which is the nutrition transition.

Japan is the clearest anomaly: a HIC with a supply of about 2,670 kcal, below the world average, because of its rice-, fish- and vegetable-based diet, its old population and relatively low food waste. Türkiye and Egypt lie far above the trend, partly because bread and other staples have long been subsidised.

Prevalence of undernourishment

The FAO’s headline hunger indicator, used to monitor SDG 2 (zero hunger), is the prevalence of undernourishment (PoU): the percentage of a population whose habitual food consumption gives less energy than they need for a normal, active and healthy life. The FAO estimates undernourishment from the dietary energy supply, the inequality of food consumption (from household surveys) and the minimum dietary energy requirement, so it is a modelled estimate rather than a count of hungry people.

The FAO estimated that 645 million people, 7.8% of the world’s population, faced hunger in 2025, down from 8.1% in 2024 and 8.6% in 2022. The geography of hunger has shifted: Africa had about 309 million hungry people in 2025, more than Asia (292 million) for the first time, and about 20% of Africa’s population was undernourished, compared with less than 2.5% in most HICs.

Sources: FAO, The State of Food Security and Nutrition in the World 2026, via FAO Newsroom, July 2026.

The Food Insecurity Experience Scale

The PoU is modelled from national data. The Food Insecurity Experience Scale (FIES) instead asks people directly. In surveys such as the Gallup World Poll, adults answer eight yes/no questions about the previous 12 months, from worrying about not having enough food, through eating only a few kinds of food and skipping meals, to going a whole day without eating. Answers are combined into a scale of moderate and severe food insecurity that can be compared between countries.

The FIES shows that food insecurity is far wider than hunger. In 2024 about 28% of the world’s population experienced moderate or severe food insecurity, more than three times the share undernourished. It also reaches high-income countries: about 10.7% of people in North America were moderately or severely food insecure in 2024, while in Nigeria the figure was about 77%.

Sources: FAO via World Bank, World Development Indicators, 2026 (indicator SN.ITK.MSFI.ZS).

Indicators of malnutrition

Calorie measures say nothing about the quality of a diet. Malnutrition covers every deficiency, excess or imbalance in a person’s intake of energy and nutrients. It has three broad forms: undernutrition, micronutrient deficiency and overnutrition. The word does not mean hunger alone: an adult with obesity and a stunted child are both malnourished.

Undernutrition: stunting, wasting and underweight

Undernutrition results from too little energy or protein, often made worse by repeated infections. In children it is measured against the WHO Child Growth Standards, which describe how healthy, well-fed children grow anywhere: a child more than two standard deviations below the median is classed as malnourished on that measure.

Three anthropometric (body measurement) indicators are used. Stunting is low height for age and shows long-term, chronic undernutrition, often from before birth; its effects on growth and brain development are largely irreversible after the age of two. Wasting is low weight for height and shows recent, acute weight loss, typical of a famine, a drought or a serious illness. Underweight is low weight for age, which combines both. The WHO estimated that 149 million children under 5 were stunted and 45 million wasted in 2022.

Wasting rates change quickly with crises, so they are used to trigger emergency food aid; stunting changes slowly, so it is used to track long-term progress.

Because underweight mixes the two, a child can be underweight because it is stunted, wasted or both.

Micronutrient deficiency: hidden hunger

Micronutrient deficiency is a lack of vitamins and minerals that the body needs only in small amounts. It is often called hidden hunger, because a person can eat enough calories and look healthy while lacking iron, vitamin A, iodine or zinc. Iron deficiency causes anaemia, which leaves people tired, reduces children’s ability to learn and raises the risk of death in childbirth. In 2023 about 31% of women of reproductive age worldwide were anaemic, and in India the figure was about 54%.

Sources: WHO via World Bank, World Development Indicators, 2026.

Overnutrition: overweight and obesity

Overnutrition is an intake of energy greater than the body uses, so that excess energy is stored as fat. In adults it is measured with the body mass index (BMI): weight in kilograms divided by the square of height in metres. The WHO classes adults with a BMI of 25 or more as overweight and 30 or more as having obesity.

In 2024, 2.6 billion adults (45%) were overweight worldwide, including 940 million (16%) living with obesity. Adult obesity more than doubled between 1990 and 2024, and the share of adults who are overweight rose from 25% to 45% over the same period.

Obesity matters as a health indicator because it raises the risk of type 2 diabetes, heart disease, stroke and several cancers.

Sources: WHO, Obesity and overweight fact sheet, October 2026; WHO, Malnutrition fact sheet, 2024.

Strengths and limitations of BMI

BMI is cheap, quick and needs only scales and a tape measure, so it allows comparisons across the world and over time. It does not measure body fat directly. It cannot tell muscle from fat, so many athletes are wrongly classed as overweight; it ignores where fat is stored, although fat around the waist is more harmful; and the same cut-offs do not fit all populations, so the WHO suggests lower action points for Asian populations. Waist circumference and waist-to-hip ratio are often used alongside it.

The double burden of malnutrition

In the past, undernutrition was treated as a problem of low-income countries (LICs) and obesity as a problem of HICs. Many middle-income countries (MICs) now have both at once. The double burden of malnutrition is the coexistence of undernutrition (including micronutrient deficiency) with overweight and obesity in the same country, community, household or even the same person over a lifetime.

India shows it at national scale: in 2020 about 35% of children under 5 were stunted and in 2023 about 54% of women aged 15–49 were anaemic, while adult obesity rose from 1.7% in 2000 to 8.0% in 2024. Egypt shows it in a single population: adult obesity reached 43.9% in 2024, among the highest rates in the world, while about 13% of children under 5 were stunted in 2021. The same household can contain a stunted child and an overweight mother, because cheap, energy-dense processed foods supply calories but not the protein and micronutrients children need to grow.

Sources: UNICEF/WHO/World Bank and WHO Global Health Observatory, via World Bank, World Development Indicators, 2026.

Composite indices: the hunger index and the food security index

Single indicators each capture one part of the picture. Composite indices combine several indicators into one score, so that countries can be ranked and changes over time tracked. Two are widely used: the Global Hunger Index and the Global Food Security Index.

The Global Hunger Index

The Global Hunger Index (GHI), published each year by Welthungerhilfe and Concern Worldwide, combines four indicators: the prevalence of undernourishment, child stunting, child wasting and under-five mortality. Undernourishment counts for one-third of the score, the two child growth indicators one-sixth each and child mortality one-third. Each country receives a score from 0 to 100, where lower is better, and a severity band: low (below 10), moderate (10.0–19.9), serious (20.0–34.9), alarming (35.0–49.9) and extremely alarming (50 or more).

The 2026 GHI scored 122 countries. Somalia had the worst score (47.0, alarming), followed by the Democratic Republic of the Congo (39.5), Chad (38.0), Madagascar (36.5), Haiti (34.4) and Nigeria (34.3). India scored 24.1 (serious, ranked 100th), Mexico 6.2 and Brazil 5.5 (both low). The global score was 18.1, "moderate", slightly higher than in 2018, which interrupted a long fall from 28.2 in 2000.

Sources: Welthungerhilfe and Concern Worldwide, Global Hunger Index 2026, ranking table (globalhungerindex.org).

Fourteen countries had too little data to be scored, including Sudan, South Sudan and Yemen, which were provisionally placed in the alarming band. The countries with the worst hunger are therefore often missing from the ranking itself, because conflict prevents the surveys on which the index depends.

The Global Food Security Index

The Global Food Security Index (GFSI), produced by Economist Impact (formerly the Economist Intelligence Unit), measures the conditions that make a country food secure rather than the outcomes in people’s bodies. It scores 113 countries from 0 to 100, where higher is better, on four categories: affordability, availability, quality and safety, and sustainability and adaptation, built from dozens of indicators such as food prices against incomes, crop storage, diet diversity, food safety rules and exposure to climate risk.

In the 2022 edition, the most recent set of rankings available, Finland (83.7), Ireland (81.7) and Norway (80.5) ranked highest, and Syria (36.3), Haiti (38.5) and Yemen (40.1) lowest. Because it measures national systems, the GFSI ranks HICs at the top even where some of their people are food insecure.

Sources: Economist Impact, Global Food Security Index 2022, via Wikipedia.

Evaluating composite indices

Both share the weaknesses of all composite measures. Their weightings are value judgements: giving child mortality one-third of the GHI score is a decision about what hunger means, not a fact. Their data are often old or modelled in LICs, and one number per country hides differences between regions, urban and rural areas, and men and women. Used with single indicators and maps, however, they show clearly where hunger is worst and whether it is improving.

The nutrition transition

As countries become richer and more urban, the amount of food available rises, but the bigger change is in the kind of food people eat. The nutrition transition is the shift in diet and physical activity that accompanies economic development: from diets based on cereals, roots, pulses and vegetables towards diets higher in meat, dairy, sugar, fats and oils and processed foods, combined with less physical activity at work and at home. It is also a process that changes the health of whole populations.

Popkin’s patterns

The nutritionist Barry Popkin described the transition as a series of five broad patterns of diet, each linked to a way of life. The model is descriptive: it summarises what has happened in many societies, but a country does not have to pass through every pattern in order, and different groups in the same country can be at different patterns at the same time.

Most of today’s LICs and rural areas of MICs, such as rural India and much of sub-Saharan Africa, are in pattern 3; Mexico, Egypt and urban China are in pattern 4; and only some groups in HICs, such as parts of Scandinavia, have reached pattern 5.

Why diets change: the drivers

Several processes push societies from pattern 3 to pattern 4, and they reinforce each other:

• Rising income. Households spend more on meat, dairy, oils and sugar, once luxuries; cereals become a smaller share of a larger diet.

• Urbanisation. Urbanisation moves people from physically demanding farm work to less active jobs, and city dwellers buy more processed food and eat out more often.

• Supermarkets and processed food. Supermarket chains spread rapidly in Latin America, Asia and parts of Africa, selling packaged, energy-dense foods cheaply.

• Trade and investment. Trade liberalisation and investment by transnational corporations lowered the price of imported foods and vegetable oils and spread global brands; their marketing is one driver among several.

• Women’s paid work and convenience. When both adults work outside the home, households have less time to cook and buy more ready-made food.

• Technology. Refrigeration, cheap vegetable oils and motor transport lower the cost of food and the energy needed to obtain it.

The consequences are rising rates of overweight and obesity and of the non-communicable diseases linked to diet, the "diseases of affluence" of the epidemiological transition. The nutrition transition therefore feeds the epidemiological transition: changing diets change what people die of.

The chart shows the transition in two ways. Over time, China’s food supply rose from about 2,060 to 3,390 kcal per person per day between 1980 and 2022, and its supply of energy from meat, mostly pork, rose from about 120 to about 400 kcal; cereals fell from about two-thirds of all food energy to under half. Between places, the poorer countries’ diets are dominated by staples, such as cereals in India and cassava, yams and cereals in Nigeria (each about 30% of food energy), while the USA gets only about 21% of its energy from cereals and much more from oils, sugar, meat and dairy.

Evaluating the nutrition transition model

The model has real strengths. It links diet to wider processes of development, explains why obesity and diet-related disease appear alongside economic growth, and has predictive value: it warned that MICs would face a wave of diabetes and heart disease before this was obvious in the data.

It also has clear limitations:

• It is not universal or linear. Japan and South Korea grew rich without adopting the high-fat, high-sugar diet of the USA; adult obesity in Japan was about 6% in 2024. Culture and policy can slow or redirect the transition.

• The speed has changed. HICs moved from pattern 3 to pattern 4 over a century or more; MICs such as Mexico, China and Egypt have done so in two or three decades, so undernutrition and obesity overlap (the double burden).

• It hides inequality. In HICs the poorest groups often have the worst diets and highest obesity, while pattern 5 is reached mainly by the educated and affluent. In LICs the transition starts with urban elites.

• It underplays power. The model describes changes in demand, but the food industry, trade rules and government policy shape what is cheap and available, so diets reflect the power of these stakeholders as well as consumer choice.

Regional variations in food consumption and nutrition choices

Diets vary between regions because of physical geography (what grows in a climate and soil), culture and religion, income, government policy and the degree of urbanisation. The same global processes produce different outcomes in different places.

Sources: FAO food balance sheets via Our World in Data, 2025; WHO Global Health Observatory, 2025.

Physical, cultural and economic influences

Climate and soils determine what can be grown locally: rice in the monsoon lands of South and East Asia, drought-tolerant cassava across humid West and Central Africa, teff in the Ethiopian highlands, and more fish in coastal and island societies such as Japan and Norway. These physical controls are weaker today, because trade allows diets far beyond local production.

Religion and culture shape choices strongly. Many Hindus avoid beef and many are vegetarian, Jains follow strict vegetarian diets, Muslims eat only halal food and avoid pork, and Jewish dietary law defines kosher food; fasting periods such as Ramadan change eating patterns for weeks. Income sets what households can afford: in the poorest households cheap staples supply most energy, and animal products, fruit and vegetables are cut first when prices rise.

Urban and rural diets differ too. In many MICs urban obesity rates are well above rural ones, because city dwellers buy more processed and street food and are less active. In HICs the pattern is often reversed, with the highest obesity among lower-income groups, for whom energy-dense processed food is cheaper than fresh produce.

Why health is difficult to measure

The WHO’s constitution of 1948 defines health as a state of complete physical, mental and social wellbeing, not merely the absence of disease or infirmity. Wellbeing cannot be counted directly, so geographers rely on indicators that capture parts of it. A health indicator is a measurable characteristic of a population that describes its health or the factors that affect it.

Health indicators fall into two groups. Outcome indicators measure the results: how long people live, how many babies die, how much illness people suffer. Input indicators measure the resources and conditions that produce health: doctors, hospital beds, vaccination, safe water and sanitation. Outcomes are closer to what matters, but they are harder to measure in poor countries, where many births and deaths are never registered. Inputs are easier to count but say nothing about quality: a hospital without medicines still counts as a hospital.

Health-adjusted life expectancy

Life expectancy at birth, the core unit’s measure, is the average number of years a newborn baby would live if current death rates stayed the same. It treats every year alive as equal, whether spent in good health or with a disabling illness. Health-adjusted life expectancy (HALE), the WHO’s measure, corrects for this: it is the average number of years a person can expect to live in full health, after subtracting years lived with disease and disability, weighted by how severe they are. "Healthy life expectancy" is an older name for the same idea.

In 2023 global life expectancy was 73.3 years but HALE was 62.8 years, so the average person could expect to spend about 10.5 years in poor health. The gap is larger for women: they live longer (76.1 years against 70.6 for men) but spend about 12 years in ill health against about 9 for men, partly because they live longer with chronic conditions such as arthritis and dementia.

The chart shows that HALE varies between countries even more than life expectancy suggests. Japan had the highest HALE of the countries shown, 72.2 years in 2023, while Lesotho’s was 47.9, more than 24 years less. The USA has a much lower HALE than Japan or the UK despite its wealth, because of high rates of obesity, diabetes, drug-related deaths and unequal access to health care.

Disability-adjusted life years

HALE belongs to a family of measures that count health in years. The disability-adjusted life year (DALY) measures the burden of a disease rather than the health of a population: one DALY is one year of healthy life lost, either through early death (years of life lost) or through living with illness or disability (years lived with disability). DALYs allow a disease that kills few people but disables many, such as depression or back pain, to be compared with one that kills.

The total of DALYs from all causes in a population is its disease burden, used by the WHO and by health ministries to decide where spending will save the most healthy years.

Infant, child and maternal mortality

Infant and under-five mortality

The infant mortality rate (IMR) is the number of deaths of babies under one year old per 1,000 live births in a year. The under-five mortality rate (U5MR) counts deaths before the fifth birthday per 1,000 live births. Both are highly sensitive to living conditions: most deaths in early childhood come from causes that are cheap to prevent, such as diarrhoea, pneumonia, malaria, complications at birth and undernutrition, so the rates rise and fall with sanitation, clean water, vaccination, nutrition and mothers’ education.

Global IMR fell from 64 per 1,000 in 1990 to 27.7 in 2024, and U5MR was 37.4 in 2024. The range is very wide. Japan’s IMR was 1.8 and Norway’s 2.0, while Nigeria’s was 69.8 and South Sudan’s 71.9. Under-five mortality in Nigeria (115.6 per 1,000) was more than forty times Japan’s (2.4), meaning that more than one Nigerian child in nine died before its fifth birthday.

Sources: UN Inter-agency Group for Child Mortality Estimation via World Bank, World Development Indicators, 2026.

The data have weaknesses. In many LICs births and deaths are not fully registered, so rates are estimated from household surveys, such as the Demographic and Health Surveys (DHS), and from statistical models; recent values are uncertain and are revised as new surveys arrive.

Maternal mortality

A maternal death is the death of a woman during pregnancy or within 42 days of its end from a cause related to the pregnancy. The maternal mortality ratio (MMR) is the number of maternal deaths per 100,000 live births. It is a ratio, not a rate, because the denominator is births rather than the number of women at risk. An alternative measure, the lifetime risk of maternal death, combines the MMR with the number of times a woman is likely to give birth.

About 260,000 women died from maternal causes in 2023, and the global MMR was 197 per 100,000 live births, about 40% lower than in 2000. The SDG target is a global MMR below 70 by 2030. Maternal mortality is the indicator with the widest gap between rich and poor: the MMR was 346 in LICs against 10 in HICs, and the lifetime risk of maternal death for a 15-year-old girl was 1 in 66 in LICs against about 1 in 7,900 in HICs. Sub-Saharan Africa accounted for about 70% of maternal deaths and southern Asia for about 17%. Nigeria’s MMR (993 in 2023) was the highest in the world.

Sources: WHO, Maternal mortality fact sheet, April 2025; WHO, UNICEF, UNFPA, World Bank and UN DESA estimates via World Bank, 2026.

Most maternal deaths come from severe bleeding, infection, high blood pressure in pregnancy (pre-eclampsia), complications of delivery and unsafe abortion, and most are preventable with skilled care. Public health researchers explain why women still die with the three delays model: delay in deciding to seek care (because of cost, distance, lack of knowledge or a woman’s lack of power over household decisions), delay in reaching a health facility (roads, transport) and delay in receiving adequate care once there (staff, blood supplies, equipment).

Access to sanitation

Access to sanitation measures the share of a population using facilities that safely separate human excreta from human contact. It is a health indicator because faeces carry the pathogens that cause diarrhoea, cholera, typhoid and intestinal worms. The WHO estimates that unsafe sanitation causes about 564,000 deaths a year, mostly from diarrhoeal disease; how these diseases spread is explained under water-borne diseases.

The WHO/UNICEF Joint Monitoring Programme (JMP) measures sanitation on a ladder of five service levels, so that progress can be tracked even before a household reaches the top.

The bottom rung, open defecation, is the most dangerous for health, because faeces are left where people walk, children play and water is collected. It fell from 22% of the world’s population in 2000 to 4.3% in 2024. India accounts for much of that fall: open defecation dropped from about 73% of its population in 2000 to about 7% in 2024, helped by the Swachh Bharat (Clean India) Mission launched in 2014, which subsidised millions of household toilets. In Niger and Chad more than 60% of people still practised open defecation in 2024.

Sources: WHO/UNICEF Joint Monitoring Programme via World Bank, World Development Indicators, 2026; WHO, Sanitation fact sheet, 2024.

The indicator has limits: a household counts as having a toilet whether or not every member uses it, many countries lack data on where waste goes after it leaves the toilet, and urban averages hide informal settlements where many households share a few latrines.

Physicians per 1,000 people

The number of physicians per 1,000 people is the most common input indicator of a health system’s capacity. The WHO publishes it as medical doctors per 10,000 people; the World Bank as physicians per 1,000. In 2023 Norway had about 5.0 physicians per 1,000 people, Germany 4.5 (2022) and the UK 3.3, while Nigeria had 0.38, Chad 0.09 and Niger 0.04: one doctor for every 25,000 people. Cuba, an upper-middle-income country, had about 9.5 per 1,000 in 2021, the highest in the world, the result of a state policy of training large numbers of doctors.

Sources: WHO Global Health Workforce Statistics via World Bank, World Development Indicators, 2026.

The scatter graph shows a strong negative relationship: under-five mortality falls as the number of doctors rises. The relationship is not proof that doctors alone save children’s lives. Countries with many doctors are also richer, with better sanitation, nutrition and education, which reduce child deaths too. Some anomalies stand out: Nigeria has more doctors per person than many neighbours but one of the highest under-five mortality rates, reflecting conflict in the north-east, poor sanitation and the concentration of doctors in cities.

Limitations of doctors as a health indicator

• It ignores other health workers. Nurses, midwives and community health workers deliver most primary care in LICs; Ethiopia’s Health Extension Programme, for example, trains community health workers to provide vaccination and basic maternal care in rural areas.

• It ignores distribution. Doctors cluster in capital cities and private hospitals, so a national average can hide rural areas with almost none.

• It ignores quality and access. A doctor whom a family cannot afford, or who works in a hospital without medicines, does little for that family’s health.

• It is affected by migration. The brain drain of doctors and nurses from LICs and MICs, such as Nigeria, Ghana and the Philippines, to the UK, the USA and the Gulf states lowers the ratio in the countries that trained them.

Because of these limits, the WHO also uses the Universal Health Coverage service coverage index, which combines coverage of essential services such as immunisation and care in childbirth into one score from 0 to 100.

Comparing places with a development diamond

A development diamond compares places on four indicators at once. Each axis runs from the centre (worst) to the tip (best), so a large, even diamond shows a high level of development on all four measures and a small or lopsided one shows where a place lags.

The epidemiological transition

As societies develop, not only do people live longer, but they die of different things. In 1971 the epidemiologist Abdel Omran described this change as the epidemiological transition: the long-term shift in a population’s main causes of death and illness from infectious diseases, malnutrition and famine to chronic, non-communicable diseases, as death rates fall and life expectancy rises.

Communicable and non-communicable diseases

The transition is defined by two groups of disease. A communicable disease (infectious disease) is caused by a pathogen, such as a bacterium, virus or parasite, and can pass from one person to another directly, through water or food, or through a vector such as a mosquito: examples are tuberculosis (TB), malaria, HIV, cholera and measles.

A non-communicable disease (NCD) cannot be passed between people. NCDs are usually chronic, developing slowly over many years from a combination of genetic, physiological, behavioural and environmental causes. The four main NCDs are cardiovascular diseases (heart attacks and strokes), cancers, chronic respiratory diseases and diabetes.

Omran’s stages

Omran’s model has three stages, to which later researchers added two more:

1. The age of pestilence and famine. Death rates are high and fluctuate with epidemics, famines and wars; infectious diseases and malnutrition dominate; life expectancy at birth is between about 20 and 40 years.

2. The age of receding pandemics. Better food supplies, clean water, sanitation and housing, then vaccination and medicine, cut deaths from infection; epidemics become rarer; life expectancy rises towards 50 or more years and population grows quickly.

3. The age of degenerative and human-made diseases. Death rates are low and stable; most people live past 70 and die of heart disease, cancer, stroke and diabetes, linked to ageing, diet, tobacco, alcohol and pollution.

4. The age of delayed degenerative diseases (added in the 1980s). Medical advances and healthier lifestyles delay death from NCDs to older ages, so life expectancy goes on rising into the 80s, as in Japan.

5. The age of emerging and re-emerging infections (a later suggestion). New and returning infectious threats such as HIV, drug-resistant TB, antimicrobial resistance and pandemic diseases such as COVID-19 show that the transition can be reversed.

The epidemiological transition runs alongside the demographic transition model. Stage 1 of Omran’s model matches the high and fluctuating death rates of DTM stage 1; the age of receding pandemics matches the falling death rates of DTM stages 2 and 3; and the age of degenerative diseases matches the low death rates of DTM stages 4 and 5, where an older population means more deaths from NCDs. The nutrition transition is linked to both, because changing diets drive much of the rise in NCDs.

Real data show the transition at work and also its reversals. In India the share of deaths from communicable, maternal, perinatal and nutritional causes fell from 53% in 2000 to 22% in 2019, so that by 2019 NCDs caused about two-thirds of Indian deaths. Nigeria remains in the age of receding pandemics, with 68% of deaths from these causes in 2019. COVID-19 then pushed the communicable share back up: in Mexico it rose from 11% in 2019 to 39% in 2021, a temporary reversal that the fifth stage of the model anticipates.

Strengths and limitations of the model

The model works best as a description of a broad global trend; as a forecast for one country it is weaker, because many LICs and MICs now take on NCDs before they have controlled infectious disease.

The disease continuum: diseases of poverty to diseases of affluence

The disease continuum arranges diseases along a line from those most common where people are poor to those most common where people are rich. At one end are diseases of poverty: communicable diseases such as diarrhoeal disease, TB, malaria, HIV and pneumonia, together with undernutrition and maternal and newborn conditions. At the other are diseases of affluence: NCDs such as heart disease, stroke, type 2 diabetes and many cancers.

Diseases of poverty spread where water is unsafe, sanitation poor, housing crowded and health care hard to reach; most are cheap to prevent.

Diseases of affluence rise with longer lives, diets high in sugar, fat and salt, inactivity, tobacco, alcohol and air pollution.

Sources: WHO, Noncommunicable diseases fact sheet, September 2025.

The double burden of disease

Many MICs and LICs now face the double burden of disease: a high level of communicable disease at the same time as a fast-rising level of NCDs. The double burden stretches health systems, because they must keep running vaccination, TB and maternal care programmes while building services for long-term conditions such as diabetes, cancer and heart disease, which need regular monitoring and expensive drugs for life.

South Africa is a clear example. It has one of the world’s largest HIV epidemics and a high rate of TB, and in 2019 36% of its deaths still had communicable, maternal, perinatal or nutritional causes. Yet adult obesity reached about 32% in 2024, and NCDs caused more than half of all deaths in 2019. India shows the same pattern at a larger scale: diarrhoeal disease and TB remain major killers in rural areas while diabetes and heart disease rise quickly in the cities.

Sources: WHO Global Health Estimates and Global Health Observatory, via World Bank, 2026.

Multimorbidity

As people live longer with NCDs, many have more than one at the same time, such as diabetes, high blood pressure and arthritis together. Multimorbidity makes treatment more complex and costly, because conditions and their medicines interact, and it is the main reason why years lived in ill health (the gap between life expectancy and HALE) grow as populations age.

An uneven transition within societies

The epidemiological transition does not happen evenly inside a country. Groups with lower incomes, less education, poorer housing or less access to health care can remain at an earlier stage, or face both burdens, while others in the same country move ahead. National averages therefore hide large internal differences, and those differences are a matter of equity.

• Income. In HICs life expectancy rises with income, and the poorest areas have the highest rates of smoking, obesity and early death from NCDs.

• Indigenous peoples. In Australia in 2020–2022, life expectancy at birth for Aboriginal and Torres Strait Islander males was 71.9 years, 8.8 years less than for non-Indigenous males (80.6); for females the gap was 8.1 years. In remote and very remote areas the gap widened to 12.4 years.

• Regions and cities. In the Calton area of Glasgow, Scotland, the WHO reported male life expectancy of 54 years in 1998–2002, far below the Scottish average; a later estimate for Calton and Bridgeton in 2008–2012 was 67.8 years for males, still well below greater Glasgow’s 71.6 and Scotland’s 78.2.

• Urban and rural areas. In LICs, cities often have better access to clean water and doctors but also more processed food and air pollution, so urban residents may have lower infectious disease but higher NCD rates than rural ones.

Sources: Australian Bureau of Statistics, Aboriginal and Torres Strait Islander life expectancy, 2020–2022 (2023); Glasgow Centre for Population Health and WHO Commission on Social Determinants of Health (2008), via Wikipedia.

Global ageing and the burden of disease

The world’s population is ageing, the result of falling fertility and rising life expectancy in almost every country. The share of people aged 65 and over rose from 5.0% in 1960 to 10.4% in 2025, and the WHO expects the number of people aged 60 and over to rise from about 1 billion in 2020 to 2.1 billion in 2050, with the number aged 80 and over tripling to 426 million. An ageing population changes the burden of disease as well as the size of the workforce.

How ageing changes the burden of disease

The risk of almost every NCD rises sharply with age, so an older population means more cases of heart disease, stroke, cancer, chronic lung disease, arthritis, hearing and sight loss and dementia, even if age-specific rates do not change. In 2021 about 57 million people were living with dementia, more than 60% of them in LICs and MICs, with nearly 10 million new cases each year; dementia was the seventh leading cause of death and cost economies about US$1.3 trillion in 2019, around half of it in unpaid care by family members.

Sources: WHO, Dementia fact sheet, 2025; WHO, Ageing and health fact sheet, 2025.

The implications reach well beyond hospitals:

• More years lived with disability. The gap between life expectancy and HALE widens, so more health spending goes on long-term care rather than cures.

• Multimorbidity and costs. Older patients often need several treatments at once, making care more complex and expensive per person.

• Long-term and social care. Care homes, home carers and nurses are needed in large numbers; much care falls on family members, mostly women.

• Workforce and pensions. A shrinking working-age population pays the taxes that fund health care and pensions, raising the dependency ratio.

• Prevention. Healthy ageing programmes aim to keep older people active and independent for longer and narrow the gap between life expectancy and HALE.

Ageing in different places: "old before rich"

Ageing began earliest in Japan and Europe, where it happened slowly and after these countries had become rich. Japan’s share aged 65 and over reached 14.7% in 1995, when its gross national income per capita (Atlas method, current US$) was about US$43,600. China reached almost the same share (14.9%) in 2025 with a GNI per capita of about US$13,700 in 2024.

Countries such as China, Thailand and Brazil are therefore growing old before they are rich: they must fund care for a rapidly ageing population with a smaller tax base and a health system still dealing with infectious disease. In sub-Saharan Africa, by contrast, the share aged 65 and over is only about 3%, so its main health challenges remain those of young populations.

Sources: World Bank, World Development Indicators, 2026 (SP.POP.65UP.TO.ZS; NY.GNP.PCAP.CD).

Case study: Mexico’s nutrition and epidemiological transitions

Mexico is an upper-middle-income country of about 130 million people, with a gross national income per capita of about US$12,800 (Atlas method) in 2024. Its share of people living in urban areas rose from 51% in 1960 to 80% in 2024. Within two generations it has moved from a population whose main nutritional problem was undernutrition to one with some of the highest rates of obesity and diabetes in the world, while undernutrition has not disappeared. It shows both transitions at work, the speed at which they can happen in a MIC, and how a government can respond.

A changing diet

The traditional Mexican diet was based on maize (eaten as tortillas), beans, chillies and vegetables, with relatively little meat. FAO food balance sheets show how this changed. Between 1961 and 2022 the food supply rose from about 2,400 to 3,230 kcal per person per day. The supply of energy from pulses (mainly beans) fell from about 180 to 76 kcal, while energy from vegetable oils rose from about 140 to 370 kcal, from poultry from about 14 to 145 kcal and from sugar from about 260 to almost 400 kcal.

Maize remained the largest single source of energy, but a much larger share of the diet came from fats, sugar and animal products.

The causes are the drivers of the nutrition transition: rapid urbanisation, rising incomes, the spread of supermarkets, convenience stores and fast food, closer food trade with the USA after the North American Free Trade Agreement (NAFTA) came into force in 1994, and cheap, widely available soft drinks and snacks. The shift to processed food began among affluent urban groups and spread to poorer urban households, who often choose cheap, energy-dense foods low in vitamins and minerals.

Health outcomes: obesity, diabetes and the double burden

Adult obesity in Mexico rose from 8.7% in 1980 to 21.5% in 2000, 28.0% in 2010 and 37.4% in 2024. Women are more affected than men (41.4% against 32.7% in 2024), and about 73% of adults were overweight or obese in 2022. Diabetes, closely linked to obesity, had become the largest single cause of death in Mexico by 2012.

The epidemiological transition followed. NCDs caused 70% of deaths in Mexico in 2000 and 79% in 2015, while the share from communicable, maternal, perinatal and nutritional causes fell from 18% to 10%. In 2023 life expectancy was 75.3 years but HALE only 63.1, so Mexicans could expect to spend about 12 years in poor health. Obesity and diabetes then increased the risk of severe illness from COVID-19, and in 2021 communicable causes, almost all COVID-19, made up 39% of deaths.

Mexico also shows the double burden of malnutrition. In 2022, 12.5% of children under 5 were stunted, a sign of chronic undernutrition, concentrated in poorer and rural communities, at the same time as most adults were overweight. Researchers have also linked stunting in early childhood to a higher risk of overweight later in life, so the two burdens may be connected rather than separate.

Sources: FAO food balance sheets via Our World in Data, 2025; WHO Global Health Observatory, 2025; World Bank, World Development Indicators, 2026; Wikipedia, "Obesity in Mexico", 2026.

Responses: taxes and warning labels

Mexico has become a leading example of government action against diet-related disease. In January 2014 it introduced an excise tax of one peso per litre on sugar-sweetened beverages, alongside a tax on energy-dense non-essential foods. A study of 6,253 households in 53 cities found that purchases of taxed drinks in 2014 were 6% lower than pre-tax trends predicted, falling to 12% lower by December 2014. Low-income households reduced purchases most (9% on average and 17% by December), and purchases of untaxed drinks, mainly bottled water, rose by 4%.

From 1 October 2020 Mexico also required black octagonal front-of-pack warning labels on packaged foods and drinks high in sugar, saturated fat, salt or energy, under its labelling standard NOM-051, following a model first used in Chile in 2016. The labels are designed to be understood quickly, including by people with low literacy.

How effective have the responses been?

Mexico’s experience shows the key concepts clearly. The transitions are processes set in motion by development, urbanisation and trade, and they unfolded much faster in Mexico than they did in HICs. Place matters: affluent urban groups changed their diets first, while poorer rural communities still have stunting. Questions of power run through the response, because taxing food sets the government against the interests of large food and drinks companies and affects poor consumers most. The possibility of reversing obesity remains uncertain: the tax changed buying habits, but adult obesity was still rising in 2024.

Sources: Colchero, M.A. et al., BMJ 352:h6704 (2016); USDA Foreign Agricultural Service report on NOM-051 (2023); WHO Global Health Observatory, 2025.

• To what extent does Mexico’s experience support Popkin’s model of the nutrition transition?

• Explain why the double burden of malnutrition is concentrated in particular regions and groups within Mexico.

• Evaluate the view that taxing sugary drinks is a fair way to reduce obesity.

Exam practice

Paper 1 tests this content through a structured question built on a stimulus, often a choropleth map, a scatter graph, a table of indicators or a transition graph, and through extended essays. Structured questions reward accurate reading of the stimulus, with figures and units; essays reward precise use of indicators, located examples and evaluation of how well each indicator measures what it claims to.

Indicator questions also test judgement about what a measure can and cannot show.

Structured question

The table shows health indicators for five countries (real data).

Sources: WHO Global Health Observatory, 2025; UN IGME and WHO via World Bank, World Development Indicators, 2026.

Part (b) asks for one reason with its development; part (c) asks for two separate reasons, each explained.

In part (c), each reason needs its own causal chain: name the reason, explain how it raises infant deaths, and support it with a figure or place.

Extended response

The answer evaluates as it goes, paragraph by paragraph, rather than leaving all the evaluation to the conclusion.

Summary

• Food security has four dimensions (availability, access, utilisation, stability), and each needs its own indicators.

• Calories per person measure food availability, not consumption; the prevalence of undernourishment (7.8% of the world in 2025) and the FIES measure access.

• Malnutrition includes undernutrition (stunting, wasting, underweight), micronutrient deficiency and overnutrition (BMI 25+ overweight, 30+ obese); many MICs face a double burden of malnutrition.

• The GHI (lower is better) and the GFSI (higher is better) are composite indices whose weightings are value judgements and whose national averages hide inequality.

• The nutrition transition shifts diets from staples to meat, oils, sugar and processed food as incomes rise and societies urbanise; it is faster in today’s MICs and varies by region and culture.

• Health indicators include outcomes (HALE, IMR, U5MR, MMR) and inputs (sanitation, physicians per 1,000), each with its own unit and limitations.

• The epidemiological transition shifts the main causes of death from communicable diseases to NCDs; the labels diseases of poverty and of affluence are simplifications, since most NCD deaths are in LICs and MICs.

• Global ageing raises the burden of NCDs, dementia and multimorbidity; countries such as China are growing old before they are rich.

Self-check

• To do:

  I can explain the four dimensions of food security and name an indicator for each.

• To do:

  I can describe the global pattern of undernourishment and calorie supply, with figures and named countries.

• To do:

  I can define stunting, wasting, underweight, micronutrient deficiency, overweight and obesity, and calculate a BMI.

• To do:

  I can evaluate the Global Hunger Index and the Global Food Security Index.

• To do:

  I can explain the nutrition transition and give located examples of regional variations in diet.

• To do:

  I can define HALE, IMR, U5MR and MMR with their units, and describe their global patterns.

• To do:

  I can explain why access to sanitation and physicians per 1,000 people are used as health indicators, and their limitations.

• To do:

  I can draw and explain the epidemiological transition and the disease continuum.

• To do:

  I can examine how global ageing changes the burden of disease, using named countries.

• To do:

  I can use Mexico as a case study of the nutrition and epidemiological transitions and the responses to them.

Good answers may be well structured (AO4) and may additionally offer a critical evaluation (AO3) which weighs up whether the simplicity of a composite index is worth the detail lost. Usefulness may depend on purpose and scale: a GHI score helps international organizations target aid between countries, but local planners need disaggregated data. Another approach might be to compare composite indices with single indicators, arguing that each answers a different question. Perspectives may differ on whose priorities a weighting reflects.

For 5–6 marks, expect some weakly-evidenced outlining of what one or more composite indices measure, with at least one strength or limitation.

For 7–8 marks, expect a structured account which includes:

• either well-evidenced explanation of both strengths and limitations of composite indices

• or a discursive conclusion (or ongoing evaluation) grounded in geographical concepts and/or perspectives, such as scale or purpose.

For 9–10 marks, expect both of these traits.

NB. Maximum [4] if the answer only discusses single indicators (eg calories per capita) and no composite index.

Markbands (10 marks, 2 marks per band)