The Nutrition Transition Barry Popkin Department of NutritionThe School of Public HealthThe University of North Carolina at Chapel Hill: The Nutrition Transition Barry Popkin Department of Nutrition The School of Public Health The University of North Carolina at Chapel Hill


Presentation Outline: Presentation Outline What are our key diet-related priorities? Dietary changes are great! Physical activity is shifting toward sedentarianism! The rapidity of change has accelerated in obesity! Diabetes is an indicator of rapid NCD changes. The burden of disease is shifting to the poor! Several important biological factors and relationships may accelerate the effects of these changes on other nutrition-related noncommunicable diseases (NCD’s).


Slide3: Stages of the Nutrition Transition Urbanization, economic growth, technological changes for work, leisure, & food processing, mass media growth Pattern 3 Receding Famine Slow mortality decline increased fat, sugar, processed foods shift in technology of work and leisure Pattern 4 Degenerative Disease accelerated life expectancy, shift to increased NR-NCD, increased disability period reduced fat, increased fruit, veg,CHO,fiber replace sedentarianism with purposeful changes in recreation, other activity Pattern 5 Behavioral Change extended health aging, reduced NR-NCD MCH deficiencies, weaning disease, stunting starchy, low variety, low fat,high fiber labor-intensive work/leisure obesity emerges, bone density problems reduced body fatness, improved bone health Source: Popkin (2002). Pub. Health Nutr 5:93-103.


Slide4: From Ancient to Modern ..... Diets


Slide5: From Ancient to Modern ..… Work


Slide6: From Ancient to Modern ..... Transport


Slide7: What are the key nutrition-related health concerns? Dietary elements: Good food components : higher fiber grains, fruits and vegetables, legumes. Limited amounts of animal source foods are important for micronutrients, growth. Poorer components: increased saturated fat, trans-fatty acids, refined carbohydrates, added sugar, energy density. Physical activity: Reductions in activity at market and home production, travel and leisure.


Slide8: What are the key nutrition-related health concerns? Health outcomes: Hypertension, stroke, diabetes, cardiovascular disease, obesity Cancers: lag a generation behind heart disease. Others: bone health, poor functional status Related: rapid shift in medical care needs as shown in China and India.


Slide9: Mortality trends in China and India, 1990 and 2020 China India Popkin, Horton et al, Nutrition Reviews (2001).59:379-90 Murray and Lopez. Global Burden Disease. Harvard University Press, 1996.


Slide10: Changes in the structure of diet are very rapid Major shifts in diet are taking place: Large increases in animal food source intake, and reductions in fiber and total fruit and vegetable intake. The overall shifts in energy density are remarkable. In a decade the energy density of food consumed by Chinese adults aged 20-45 increased by over 13%. Added sugar increases across the developing world are equally dramatic, with an extra 100-300 kcals per day among all individuals in each low and moderate income country over the past 25 years.


Slide11: Vegetable Fat Intake has Increased Greatly: the Relationship Between the Percentage of Energy from Fat and GNP Per Capita, 1962 and 1990 1990 Total Fat 1962 Total Fat 1962 Animal Fat 1990 Animal Fat 1990 Vegetable Fat 1962 Vegetable Fat Source: Nonparametric regressions run with food balance data from FAOUN and GNP data from the World Bank for 134 countries. Guo et al (2000) EDCC 48:737-60


Slide12: The Chinese Diet Is Shifting From a Low to High Fat Diet Among Chinese Adults Aged 20-45, CHNS 1989-97 Source: Du Lu, Zhai, Popkin. (2002) Pub H Nutr 5:169-174. from the China Health and Nutrition Survey, 1989-1997


Slide13: Dietary Trends in Rural Areas Are Linked with Increased Animal Product and Edible Oil Dietary Intake, CHNS89-97 Meat Poultry Egg Fish/seafood Edible oil Source: Du Lu, Zhai, Popkin. (2002) Pub H Nutr 5:169-174. from the China Health and Nutrition Survey, 1989-1997


Dynamic Dietary Intake Shifts in China: Longitudinal Analysis of the Income-consumption Linkages: Dynamic Dietary Intake Shifts in China: Longitudinal Analysis of the Income-consumption Linkages Sample: adults from 1989-91-93-97 aged 20-45 in 1989. Random effects model (fixed effects results are similar but estimated with less precision). Present the income elasticities (% change in dietary intake for each % change in income). Results show a significant change in the income-fat intake relationship. We find added purchases of various products [meat, edible oil, eggs] with additional income in 1997 vs. all earlier time periods. Results show a rapid increase in total fat.


Total Fat intake : Total Fat intake For every 10% increase in income, intake of the proportion of energy from fat is increasing greater among the poor than the rich in China. Most importantly there are significant changes found in all income groups. Source: Popkin and Du (in press) J Nutr.


Slide16: The Relationships Between Income Changes and the Proportion of Energy From Fat Consumed: Income Elasticity and the Change in the Income Elasticity Between 1989 and 1997 in China 95% Confidence Interval 95% Confidence Interval A. 1989 95% Confidence Interval 95% Confidence Interval B. 1997 Source: Popkin and Du (in press) J Nutr.


The Sweetening of the World’s Diet: The Sweetening of the World’s Diet Look at the role of added sweeteners in processed foods World patterns: use FAO food disappearance data to examine time trends related to shifts in GNP per capita and urbanization Explore US individual dietary intake trends to understand food sources and impact on total energy in one country. Source for this analysis: Popkin and Nielsen (2002) unpublished paper.


Slide18: The Relationship Between Changes in Gross National Product per capita and Added Sugar Source : Popkin and Nielsen (2002) unpublished paper


Slide19: The Relationship Between Changes in GNP per Capita and Sugar as a % of Carbohydrates Source : Popkin and Nielsen (2002) unpublished paper FAOSTAT food disappearance data.


Slide20: The Relationship Between Changes in Urbanization and Added Sugar Source : Popkin and Nielsen (2002) unpublished paper


Slide21: The Effect of Shifts in the Proportion Residing in Urban Areas on Added Sugar Intake in 2000 Source: FAOSTAT food disappearance data


World Trends: What explains the trend in added sugar–changes in the composition of the population, the behavior of industry and the consumer?: World Trends: What explains the trend in added sugar–changes in the composition of the population, the behavior of industry and the consumer? Compositional changes: more persons live in urban areas (49%), higher income per capita (25%) Behavioral changes: either the food industry has changed their behaviors (e.g. greater sweetening of processed foods) or households purchase more highly sweetened foods today than earlier for the same level of income and urbanization. This component explains about 26% of the trend.


Animal Source Foods: an Equally Prominent Shift in the Composition of Diet in the Developing World—discussed by C Delgado: Animal Source Foods: an Equally Prominent Shift in the Composition of Diet in the Developing World—discussed by C Delgado Beef and Pork Poultry Fish Dairy products


Dynamics of the Food Sector: T Reardon, M Nestle, J Tillotson, C. Hawkes : Dynamics of the Food Sector: T Reardon, M Nestle, J Tillotson, C. Hawkes Food distribution and shopping options. Supermarkets are a new actor in the process! What about away-from-home consumption and the internationalization of the fast food sector? Then what about the international food companies–Nestle’s, Unilever, Kraft General Foods,etc.?


Slide25: Physical activity changes are equally rapid! Three major shifts are occurring: type of work and activities within each job; home production and leisure; and transportation. Two work shifts--the shift from agriculture and other energy-intensive occupations to service sector occupations and the reduction of the level of activity within each occupation. Changes in types of transportation used and leisure activity patterns also reflect a rapid shift toward reduced energy expenditures.


Slide26: Lower Income Countries have Experienced a Rapid Shift in the Proportion of Adults in Service Occupations, 1972 to 1993 Source: Popkin & Doak (1998) Nutr Rev: 56: 106


Slide27: Chinese Adults in the Same Occupations Use Much Less Energy: the Physical Activity Profiles of Urban Chinese Adults Aged 20-45, 1989 and 1997 Source: Du Lu, Zhai, Popkin. (2002) Pub H Nutr 5:169-174.


Slide28: TV Ownership Has Skyrocketed in China, 1989-93 (% Who Own TV Sets among Families with Children from CHNS) Source: Du Lu, Zhai, Popkin. (2002) Pub H Nutr 5:169-174.


Slide29: Odds of Becoming Overweight/obese According to Household Vehicle Acquisition: 1989 to 1997 Source: Bell et al (2002) Obesity Research 10:277_283. No change in vehicle ownership was the referent category. Adjusted for baseline age, weight, education, urban residence and change in work-related activity, energy intake, smoking status,alcohol consumption, income, and television ownership. Those who were obese in 1989 were excluded.


The shifts in patterns of diet, physical activity and body composition seem to be occurring more rapidly: The shifts in patterns of diet, physical activity and body composition seem to be occurring more rapidly The obesity patterns are much higher for the level of development than heretofore found The rates of change are very rapid or at least the data we have seems to lead to that conclusion child trends-comparison (will not present here: see Wang et al, 2002: AJCN 75: 971) adult patterns and trends


Slide31: Overweight almost tripled among Chinese men and doubled among women over eight years. The 8-Year Change in the BMI Distribution for a Cross-Section of Chinese Adults 20-45. Source: Bell et al, Int’l Jour. Obes 2001.25:1-8


Obesity Patterns Among Adults in Latin America: Obesity Patterns Among Adults in Latin America Cuba Havana 1998 Mexico 1999 GNP 3840 Brazil 1996/97 GNP 4630 Source: Popkin (2002). Pub. Health Nutr 5. 93-103.


Slide33: Obesity Patterns Among Adults in Asia Kyrgystan 1993 GNP 380 25

Slide34: Obesity Patterns Among Adults in N Africa /Middle East Jordan 1994-96 GNP 1150 Morocco 1998-99 GNP 1240 Egypt 1998 GNP 1290 Tunisia 1990 GNP 2060 Saudi Arabia 1996 GNP 6910 M Male F Female Iran 1999 GNP 1650 Bahrain 1991-92 Kuwait 1993-94 Source: Popkin (2002). Pub. Health Nutr 5: 93-103.


Slide35: Obesity Patterns Among Adults in Sub-Sahara Africa Mali 1991 GNP 250 South Africa (black) 1998 GNP 3310 Mauritius 1992 GNP 3730 Source: Popkin (2002). Pub. Health Nutr 5: 93-103.


Slide36: Prevalence of Obesity among Females Age 20-49, in Sub-Saharan Africa BMI>30 BMI 25-29.9 GNP 141 GNP 177 GNP 317 GNP 229 GNP 253 GNP 297 GNP 369 GNP 378 GNP 537 GNP 627 GNP 1245 GNP 2108 GNP 3833 Source: Woman of Child-bearing Age With a Child. Latest DHS survey in each country with maternal anthropometry data. Monteiro &Popkin (2002)


Slide37: Obesity Trends Among Adults in the U.S. And Europe (The Annual Percentage Point Increase in Prevalence) U.S. 1960-94 BMI>30 25

Slide38: Obesity Trends Among Adults in Latin America (The Annual Percentage Point Increase in Prevalence) Cuba 1982-98 BMI>30 25

Slide39: Prevalence of Obesity among Females Age 20-49, in Sub-Saharan Africa BMI>30 BMI 25-29.9 GNP 141 GNP 177 GNP 317 GNP 229 GNP 253 GNP 297 GNP 369 GNP 378 GNP 537 GNP 627 GNP 1245 GNP 2108 GNP 3833 Source: Woman of Child-bearing Age With a Child. Latest DHS survey in each country with maternal anthropometry data. Monteiro &Popkin (2002)


Slide40: Obesity Trends Among Adults in Asia (The Annual Percentage Point Increase in Prevalence) China 1989-97 GNP 750 Thailand 1991-96 GNP 2160 Korea 1995-98 GNP 8600 BMI>25 BMI>30 BMI 25-30 M Male F Female Annual percentage point change Source: Popkin (2002). Pub. Health Nutr 5: 93-103.


Slide41: Obesity Trends Among Adults in N. Africa/Middle East (The Annual Percentage Point Increase in Prevalence) Mauritius 1987-92 GNP 3730 BMI>30 25

Slide42: South Atlantic Ocean South Pacific Ocean North Atlantic Ocean Indian Ocean Arctic Ocean Arctic Ocean North Pacific Ocean Brazil Cuba South Africa Egypt Tunisia Morocco Saudi Arabia Iran China S. Korea Mauritius The Obesity Epidemic among Adults, Males Aged 20 and Older (BMI>30) Kyrgyzstan Canada Finland France Spain England Germany Australia Neth. Arctic Ocean United States of America Jordan Sweden Philippines Malaysia New Zealand Kuwait 20% Bahrain


Slide43: South Atlantic Ocean South Pacific Ocean North Atlantic Ocean Indian Ocean Arctic Ocean Arctic Ocean North Pacific Ocean Brazil Cuba South Africa Egypt Tunisia Morocco Saudi Arabia Iran China India Malaysia Philippines S. Korea Mauritius Jordan Kyrgyzstan Bahrain Mexico Canada Finland France Spain Germany Australia England Arctic Ocean The Obesity Epidemic among Adults, Females Aged 20 and Older (BMI>30) United States of America 20% Neth. New Zealand Kuwait Bahrain


Slide44: South Atlantic Ocean South Pacific Ocean North Atlantic Ocean Indian Ocean Arctic Ocean Arctic Ocean North Pacific Ocean Brazil Cuba South Africa Egypt Tunisia Morocco Saudi Arabia Iran China India Malaysia Philippines S. Korea Mauritius Kyrgyzstan Bahrain Canada France Thailand Neth. Arctic Ocean The Obesity Epidemic among Adults, Males Aged 20 and Older (BMI>25) United States of America New Zealand Italy Kuwait


Slide45: South Atlantic Ocean South Pacific Ocean North Atlantic Ocean Indian Ocean Arctic Ocean Arctic Ocean North Pacific Ocean Brazil Cuba Egypt Tunisia Morocco Saudi Arabia Iran China India Malaysia Philippines S. Korea Mauritius Kyrgyzstan Bahrain South Africa Canada France Thailand Arctic Ocean The Obesity Epidemic among Adults, Females Aged 20 and Older (BMI>25) United States of America 10-20% 31-40% 21-30% 41-50% >51% Neth. New Zealand Italy Kuwait


Slide46: South Atlantic Ocean South Pacific Ocean North Atlantic Ocean Indian Ocean Arctic Ocean Arctic Ocean Arctic Ocean North Pacific Ocean Brazil Tanzania Kenya Egypt Mali China India Ghana Kyrgyzstan Bolivia Madagascar Namibia Zimbabwe Malawi Cameroon C. A. R. Côte D’Ivoire Turkey Vietnam Kazakhstan Uzbekistan Uganda Peru Colombia Dominican Republic Guatemala Haiti Niger Niger Nigeria Senegal Benin Zambia Chad Guinea Yemen Nepal Bang. Togo Patterns of Obesity among Women of Child-bearing Age from the DHS (BMI>30, Ages 20-49, Age-Standardized, Weighted) * 10-14% Burkina Faso Comoros Mozambique Eritrea South Africa Jordan Mexico * This sample includes only women with preschoolers in 41 of the 44 countries


Slide47: South Atlantic Ocean South Pacific Ocean North Atlantic Ocean Indian Ocean Arctic Ocean Arctic Ocean Arctic Ocean North Pacific Ocean Brazil South Africa Tanzania Kenya Egypt Mali China India Ghana Kyrgyzstan Bolivia Madagascar Namibia Zimbabwe Malawi Cameroon C. A. R. Côte D’Ivoire Turkey Vietnam Kazakhstan Uzbekistan Uganda Peru Colombia Dominican Republic Guatemala Haiti Niger Nigeria Senegal Zambia Benin Chad Guinea Yemen Nepal Bang. Togo Patterns of Overweight and Obesity among Women of Child-bearing Age from the DHS (BMI>25, Ages 20-49 , Age-Standardized, Weighted)* 10-20% 31-40% 21-30% 41-50% >51% <10% Burkina Faso Mozambique Comoros Eritrea Jordan Mexico * This sample includes only women with preschoolers in 41 of the 44 countries


The Prevalence of Diabetes is increasing rapidly: The Prevalence of Diabetes is increasing rapidly Previous projections of the prevalence of diabetes world-wide have, in general, been underestimates compared with new data. Over two-thirds of diabetes cases are in the developing world. In absolute numbers over half of the new cases in the world are found in India and China. The age-specific burden is relatively much higher at a younger age in the developing world.


Slide49: Prevalence of Diabetes Mellitus in the Adult Population (Aged 20 Years and Over) by Year and Region Source: King H, et al. Diabetes Care 1998 21: 1414-1431.


Slide50: Number of Persons With Diabetes in the Adult Population (Aged 20 Years and Over) by Year and Region Source: King H, et al. Diabetes Care 1998 21: 1414-1431.


Slide51: Number of Persons With Diabetes by Age Group, Year and Region Source: King H, et al. Diabetes Care 1998 21: 1414-1431.


The burden of disease is shifting rapidly towards the poor.: The burden of disease is shifting rapidly towards the poor. By burden, we refer to the greater prevalence of poor diets, sedentarianism, obesity, NR-NCD’s among the poor. Evidence from Brazil points to a clear shift in obesity and other risk factors. Studies on the shifts in diet among various income groups in China point towards a similar shift occurring there in 10-15 years.


Slide53: Dietary Patterns by Income Brazil Metropolitan Households, 1996


Slide54: The Proportion of Adults with any Leisure-time Physical Activity by Income, Brazil, 1997


Slide55: The Prevalence of Smoking for Each Income Group in Brazil, 1989


Slide56: The BMI Distribution in Brazilian Women From the Southeastern Region: Lower Versus Higher Income Group in 1997 Source: Monteiro et al (2002) Public Health Nutrition 5


Slide57: The Odds Ratios of High Education vs. Low Education Women Being Obese (or BMI >25) Show the Burden of Disease Shifts, Women Aged 20-49: Sub-saharan Africa and North Africa/west Asia/europe Source : Woman of Child-bearing Age With a Child. Latest DHS survey in each country with maternal anthropometry data. Nigeria 1999 GNP 229 S Africa 1998 GNP 3833 Tanzania 1996 GNP 177 Zimbabwe 1994 GNP 627 Egypt 1995 GNP 1036 Turkey 1998 GNP 3264 Jordan 1997 GNP 1586 Sub-Saharan Africa N Africa/West Asia/Europe


Slide58: The Odds Ratios of High Education vs. Low Education Women Being Obese (or BMI >25) Show the Burden of Disease Shifts, Women Aged 20-49 : Central Asia and South/southeast Asia Source : Woman of Child-bearing Age With a Child. Latest DHS survey in each country with maternal anthropometry data. Kyrgz Repub 1997 GNP 798 Uzbekistan 1996 GNP 712 Kazakhstan 1999 GNP 1323 Vietnam 1997 GNP 291 China 1997 GNP 667 India 1999 GNP 448 Central Asia South/Southeast Asia


Slide59: The Odds Ratios of High Education vs. Low Education Women Being Obese (or BMI >25) Show the Burden of Disease Shifts, Women Aged 20-49 : Latin America/caribbean/south America Source : Woman of Child-bearing Age With a Child. Latest DHS survey in each country with maternal anthropometry data. Guatemala 1998 GNP 1517 Bolivia 1998 GNP 952 Dominican Republic 1996 GNP 1521 Columbia 2000 GNP 2268 Mexico 1999 GNP 3515 Peru 2000 GNP 2278 Brazil 1996 GNP 4430


Slide60: Biological Differences May Accentuate and Speed up the Effects of Nutritional Changes There are important body composition differences that lead to shifts in BMI-disease patterns. The Asian recommendation to reduce the BMI cutoff for overweight and obesity is an example. The rapid shift in the stage of the nutrition transition enhances the effects of fetal and infant insults. Stunting may also affect fat metabolism (Hoffman et al, AJCN. 72: 702–7) Might also be inflammatory burden issues that are yet to be explored.


Slide61: Compared to US-White males, the odds of prevalent hypertension were significantly higher for Chinese men at every level of BMI above the range 18.5-22.9 kg/m2. Adjusting for waist:hip ratio attenuated the ethnic differences but did not eliminate them. Source: Bell et al, AJE (in press) * p < 0.05 from US-White men


Slide62: Fetal Insults: Systolic Blood Pressure Among Cebu Male Adolescents According to BMI at Birth and Age 14-15 Source: Adair .2001 Logit regression adjusted for current age, height, maturation status, energy& fat intake, SES.


Slide63: Classic Fetal Insults Results: The Probability of Elevated Blood Pressure Among Cebu, Philippines Adolescent Males, According to Birth Weight. Source: Adair , 2001 Logit regression adjusted for current age, height, BMI, maturation status, energy&fat intake, SES.


Summary: Summary Clearly rapid shifts in diet, activity, and non-communicable disease patterns are occurring around the world. Elsewhere we present in depth the effects of these changes on the nutrition-related chronic diseases. A few results from the Bellagio conference on the Nutrition Transition and its Health Implications in the developing world, August 20-24, 2001. The pdf files of the articles from Public Health Nutrition Feb. 2002 can be obtained on our website for free. See www.nutrans.org and Bellagio papers.


Summary: Summary Prevention is the only feasible approach to nutrition-related chronic diseases. The cost of their treatment and management imposes an intolerable economic burden on developing countries. The challenge is to find environmental changes that might slow down or stop these changes. Food policies changes represent a critical aspect of this agenda.