Obesity is often studied as a characteristic of an individual and discussed in relation to individual dietary patterns, physical activity levels, or genetic constitution. However, obesity can also be considered as a characteristic of populations, as well as of individuals. The term “epidemic” of obesity implies that we need to examine obesity as a population level characteristic and to develop an understanding of obesity from a population perspective.
What is an epidemic and how can we judge whether there is an epidemic of obesity? The word epidemic can be used generally to mean simply a large number of cases or more precisely to mean an increase in the number of cases over past experience or normal expectancy for a given population, time, and place (20). The term comes from the study of infectious disease but has been applied to describe issues as diverse as tobacco smoking and violence, as well as obesity (11,28).
The distribution of body mass index (BMI), like the distribution of other biological characteristics, shows a considerable degree of natural variability. Terms such as “overweight” or “obese” are applied to certain specified levels of BMI. Although many definitions have been used, the most practical current definitions of overweight and obesity for international use in adults are those of the World Health Organization (WHO), with overweight defined as a BMI of 25 or greater and obesity defined as a BMI of 30 or greater (47).
In the United States, a series of national surveys dating from 1960 includes measured height and weight for adults and children (8,43). For both children and adults, data from these surveys showed that the prevalence of overweight and obesity had increased markedly in the United States by the most recent (1988–94) survey. The age-adjusted prevalence of overweight, defined as a BMI of 25 or greater, has been high among adults at least since 1960–62, when 48% of adult men and 39% of adult women were overweight (BMI ≥ 25). By the 1988–94 survey, the age-adjusted prevalence of overweight had increased to 59% for men and 50% for women. The prevalence of obesity, defined as a BMI of 30 or greater, was lower than the prevalence of overweight but also showed an increase. The age-adjusted prevalence of obesity (BMI ≥ 30) in 1960–62 was 10% for men and 15% for women, increasing to 20% for men and 25% for women by the 1988–94 survey.
Definitions for the prevalence of overweight or obesity in children are less standardized than those for adults. Estimates of the prevalence of overweight for children from the most recent national survey in the United States, all based on the same data set, range widely, from 11 to 24%, depending on the definitions used (43,44). Using the provisional 95th percentile of BMI from the revised United States growth charts as a definition of overweight, the age-adjusted prevalence of overweight was 4% for both boys and girls ages 6–11 yr in 1963–65 and 5% for both boys and girls ages 12–17 yr in 1966–70. By 1988–94, the prevalence had increased to 11% among boys and 10% among girls in both age ranges.
To examine the current trends in overweight and obesity in more detail, a MEDLINE search for the years 1989–98 was undertaken to identify English language publications that might provide appropriate data for comparison. Additional references were gathered based on further reading. Studies were selected in which data were gathered through a systematic survey process of a defined area on a national or regional scale, weight and height were measured, and data were published using cutpoints of body mass index at least roughly similar to those used for the WHO classification. Studies from Sweden and Canada were included that used self-reported rather than measured weight or height but otherwise met these criteria. A study from Israel was included because it provided estimates of trends from repeated cross-sectional surveys of a defined area. These studies are tabulated in Table 1. Studies of smaller areas, such as those in the MONICA project, were not included.
To summarize the available information from the studies listed in Table 1 and from data published by the World Health Organization (47), the prevalence of obesity is highest in Western Samoa and other Pacific island populations, intermediate in European countries, and populations such as those in the United States with considerable European admixture and lowest in less developed countries such as Brazil and in Asian countries such as China and Japan. Large recent increases in overweight or obesity for adults (>5 percentage points) have been noted in Canada (based on self-reported data), Finland (men), New Zealand, the United Kingdom, the United States, and Western Samoa (8,24,26,29,36). Smaller increases, less than 5 percentage points, appear to have occurred in several other countries, including Australia, Brazil, China (men), Germany, Israel, Mauritius, the Netherlands, and Sweden (self-reported data) (1,12,15,17,21,22,30,33,35). For women in China and Finland, the prevalence showed no increase and even decreased for women in one area of Finland (29,30). There are fewer studies on trends among children and adolescents. Data on young men of military age (late adolescence and early adulthood) from Denmark and Singapore show considerable recent increases (23,40). Data from the United States suggest large increases among United States children (43), and a news report has suggested similar findings recently for children from the Netherlands (34).
Trend data provide the most convincing suggestion of an epidemic. For example, in the United States, the prevalences of overweight and obesity stayed fairly constant over the 20-yr period from 1960 to 1980, varying by only a few percentage points (8). Between the 1976–80 survey and the 1988–94 survey, however, the prevalence of overweight increased from 46 to 54% and the prevalence of obesity from 14.5 to 22.5%. This sudden increase after 1980 was unexpected.
Although there has been considerable speculation about the reasons for the increases observed in the United States and in some other countries, solid data are lacking. Two possibilities often mentioned are changes in dietary intake and changes in physical activity, both difficult to measure on either an individual or a population level. Prentice and Jebb examined this issue for the United Kingdom and concluded that sedentary lifestyles were at least as important as dietary changes and might represent the dominant factor (31). A study by Simmons and associates identified increased energy consumption as the dominant factor associated with increases in overweight in New Zealand (36). Greksa concluded that both increased energy intake and lower activity levels were factors leading to the increase in obesity in Western Samoa (13). Researchers in several countries have examined the question of the effect of smoking cessation, as tabulated in Table 2 (3,10,36,46) and reached the general conclusion that although smoking affects weight at the individual level, the impact of changes in smoking is insufficient to explain the observed changes in weight in the population.
The use of the term “epidemic” suggests a disease model, which may be considered one of several competing explanatory models for obesity (38). The term epidemic in this and other contexts not related to infectious disease (for example, tobacco or violence) might also be considered more as a rhetorical device, used to express concern and to mobilize for action, than as an exact scientific term. In the United States, the prevalence of overweight (BMI ≥ 25) was high in 1988–94 (59% for men and 49% for women) but was also quite high in 1960–62 (48% for men and 39% for women), so this is not a new phenomenon. The prevalence of obesity (BMI ≥ 30) was 20% for men and 25% for women in 1988–94 as compared with 10% for men and 15% for women in 1960–62. It is not clear whether these should be considered epidemic levels. To address the question of whether we are experiencing an epidemic of obesity also requires some consideration of what the expected numbers might be. Was there a golden age in which there was the “right” prevalence of obesity? Can obesity be prevented completely? Or is it likely that in any population where there is an adequate food supply there will be some cases of obesity and many cases of overweight? These changes in BMI perhaps should be considered to represent the expected population response to a cheap, abundant, and easily obtained food supply under current social conditions. The increasing prevalence of obesity can usefully be viewed as a shift in the distribution of a population characteristic rather than as an increase in the number of cases (9).
The net health implications of the increases are not completely clear. Three of the surveys in Table 1 included some investigations of trends in cardiovascular disease and risk factors. These surveys generally do not show increases in risk factors such as hypertension or hyperlipidemia that parallel the trends in overweight. In the United States, the prevalence of hypertension and elevated cholesterol appear to have declined over the same period as the increase in obesity (4,14,18). In Australia, cardiovascular mortality fell, average blood pressure levels declined, and the prevalence of hypertension decreased (1). There was no overall favorable trend in lipid results, but total cholesterol levels decreased significantly in younger men and older women. In Germany, cardiovascular mortality also continued a downward trend (17). There was a favorable increase in the high density lipoprotein (HDL)/total cholesterol ratio. Blood pressure showed no significant increases except a rise of 2 mm Hg in systolic blood pressure for women. Smaller studies, such as the Minnesota Heart study in the United States and several local WHO MONICA Project sites, also show that increases in BMI or overweight are only sometimes accompanied by parallel increases in obesity-related risk factors such as hypertension or hyperlipidemia and are sometimes accompanied by declines in these risk factors (6,27,41,45).
In societies such as the United States, the prevalence of such cardiovascular risk factors as hypertension or hyperlipidemia is high even among those who are not overweight or obese, particularly at older ages. For example in the United States in 1988–94, the estimated prevalence of elevated cholesterol (defined as serum cholesterol > 240 mg·dL−1 or taking cholesterol-lowering medications) for women with a BMI of 22.5–24.9 was 6% at ages 20–39 yr, 23% at ages 40–59 yr, and 49% at ages 60–79 yr. The corresponding estimates for a BMI of 30.0–32.4 were 11%, 33%, and 45%. These data show that although hypercholesterolemia generally increases with increasing BMI, particularly at younger ages, it is also at high levels at older ages even among those who are not classified as overweight or obese.
Whether or not we regard overweight as epidemic, we can use the epidemiological triad of host, agent, and environment to consider causal factors (20). The agent by definition is energy imbalance, where energy intake exceeds energy expenditure. Although energy balance can be influenced by changes in such factors as resting metabolic rate or lean body mass, the main modifiable factors affecting energy balance are dietary energy intake and energy expended through physical activity. Features of the host also affect obesity, in ways that are not well understood. Experimental studies clearly show differences between individuals in their ability to gain weight under conditions of overfeeding or to lose weight under conditions of underfeeding (2). Some individuals appear to be more susceptible than others.
In classic epidemiological thinking, there is a third entity: the environment that facilitates the epidemic by bringing the susceptible host and the agent together. The environment has to be considered to encompass not just the physical environment, such as the layout of cities, but the environment of economic and social organization and cultural values. Developing a cheap, stable, and adequate food supply has been an important goal of human efforts for millennia. Present-day society is characterized by a highly organized system of food production and distribution. The environment in the United States, Western Europe, and other areas, in which food is widely available and which has reduced malnutrition, hunger, and the fear of starvation, apparently also results in high levels of overweight and obesity in many countries.
It is likely that research could benefit from going beyond a narrowly mechanistic focus on energy intake and physical activity and examining the economic, social, and cultural context more broadly. For example, one of the most consistently documented findings is that obesity varies by socioeconomic status among women (37,39). Many other health conditions, as well as access to health care, have also been observed to vary by socioeconomic, education or income levels (19). Research on variations in obesity by socioeconomic status may benefit from the insights and methodologies used by researchers studying other aspects of the relation of socioeconomic status to health. For another example, both in Europe and the United States, differences in the prevalence of overweight or obesity between countries or between race-ethnic groups are often more pronounced for women than for men (8,32). The reasons for this difference are unclear but are likely to be complex. The work of social scientists, including economic and social historians, sociologists and anthropologists, can help clarify the economic, social and cultural factors that may directly or indirectly affect the distribution of body mass index in a population (5,7,26,42).
Both continued monitoring of trends in overweight and obesity and additional analyses of existing data are important. At present, we do not know what further changes may be in store and whether the observed trends will continue or level off. The United States experience suggests that obesity may increase suddenly even in countries in which the prevalence has been stable over time. Existing and future survey and trend data should be exploited to the fullest extent possible. Many characteristics of the current situation are not well described. Shifts in the whole distribution of BMI should be examined, not just changes in means and prevalences. Systematic detailed comparisons between countries are infrequent but would be useful. More detailed analyses of trends by age, sex, socioeconomic status, and other variables would also be useful.
Monitoring of concurrent trends in mortality, in diseases such as diabetes, in risk factors such as blood cholesterol levels, and in other outcomes such as disability associated with obesity is also important. Clearly, obesity is linked with increased health hazards. However, the relationship of obesity to health may vary over time and between subgroups. The trends in obesity are not clearly associated with parallel increases in cardiovascular risk factors, for example. Research should also be directed at better definitions of obesity-associated risks and a better understanding of ways to identify the population characteristics associated with increased health hazards.
Better understanding of the causes of the observed increases in overweight or obesity in many countries is also an important goal. We know remarkably little about the causes of these increases. By necessity, research is limited to observational studies. However, it may be possible to take advantage of the many natural experiments being conducted all around us, as societies change and modernize. From a population perspective, it may be useful to focus less on the “agents” of diet and physical activity and more on the “environment” of social and economic organization and cultural values. The work of economic and social historians, sociologists, and anthropologists may lead to better understanding of the social forces at work.
Public health research aimed at determining what is realistically achievable on a population basis and at what economic and social cost is essential. Can we specify and defend reasonable goals? What distribution of BMI could be expected under the current circumstances? In the United States, the prevalence of overweight in middle age is greater than 60% in all sex-race-ethnic groups. For example, an estimated 73% of white men ages 50–59 yr are overweight (BMI ≥ 25), and 33% are obese (BMI ≥ 30). Even 50% of men and women aged 80 yr or above are overweight. The resources and efforts required to reduce these high prevalences appreciably in the future would be considerable, and success is by no means certain. A viable public health strategy should aim at improving health, not only at reducing the prevalence of obesity. Public health research should aim at defining realistic goals and strategies to improve health in an environment conducive to high levels of overweight and obesity.
1. Bennett, S. A., and P. Magnus. Trends in cardiovascular risk factors in Australia: results from the National Heart Foundation’s Risk Factor Prevalence Study, 1980–1989. Med. J. Aust. 161: 519–527, 1994.
2. Bouchard, C., and A. Tremblay. Genetic influences on the response of body fat and fat distribution to positive and negative energy balances in human identical twins. J. Nutr. 127(Suppl. 5): 943S–947S, 1997.
3. Boyle, C. A., A. J. Dobson, G. Egger, and P. Magnus. Can the increasing weight of Australians be explained by the decreasing prevalence of cigarette smoking? Int. J. Obes. 18: 55–60, 1994.
4. Burt, V. L., J. A. Cutler, M. Higgins, et al. Trends in the prevalence, awareness, treatment and control of hypertension in the adult US population: data from the Health Examination Surveys, 1960–1991. Hypertension 26: 60–69, 1995.
5. de Garine, I., and N. J. Pollock (Eds.). Social Aspects of Obesity
. Amsterdam: Gordon and Breach Publishers, 1995.
6. Dobson, A. J., A. Evans, M. Ferrario, et al. Changes in estimated coronary risk in the 1980’s: data from 38 populations in the WHO MONICA Project. World Health Organization. Monitoring trends and determinants in cardiovascular diseases. Ann. Med. 30: 199–205, 1998.
7. Ellaway, A., A. Anderson, and S. Macintyre. Does area of residence affect body size and shape? Int. J. Obes. Relat. Metab. Disord. 21: 304–308, 1997.
8. Flegal, K. M., M. D. Carroll, R. J. Kuczmarski, and C. L. Johnson. Overweight and obesity in the United States: prevalence and trends, 1960–1994. Int. J. Obes. Relat. Metab. Disord. 22: 39–47, 1998.
9. Flegal, K. M., and R. P. Troiano. Changes in the distribution of body mass index of adults and children in the U.S. Int. J. Obes. Relat. Metab. Disord.
, in press.
10. Flegal, K. M., R. P. Troiano, E. R. Pamuk, R. J. Kuczmarski, and S. M. Campbell. The influence of smoking cessation on the prevalence of overweight in the United States. N. Engl. J. Med. 333: 1165–1170, 1995.
11. Fontanarosa, P. B. The unrelenting epidemic of violence in America: truths and consequences (Editorial). JAMA 273: 1792–1793, 1995.
12. Gofin, J., J. H. Abramson, J. D. Kark, and L. Epstein. The prevalence of obesity and its changes over time in middle-aged and elderly men and women in Jerusalem. Int. J. Obes. Relat. Metab. Disord. 20: 260–266, 1996.
13. Greksa, L. P. Activity level and obesity among Samoans. In: Social Aspects of Obesity, I. de Garine and N. J. Pollock (Eds.). Amsterdam: Gordon and Breach Publishers, 1995, pp. 253–266.
14. Hickman, T. B., R. R. Briefel, M. D. Carroll, et al. Distributions and trends of serum lipid levels among united states children and adolescents ages 4–19 years: data from the Third National Health and Nutrition Examination Survey. Prev. Med. 27: 879–890, 1998.
15. Hodge, A. M., G. K. Dowse, H. Gareeboo, J. Tuomilehto, K. G. Alberti, and P. Z. Zimmet. Incidence, increasing prevalence, and predictors of change in obesity and fat distribution over 5 years in the rapidly developing population of Mauritius. Int. J. Obes. Relat. Metab. Disord. 20: 137–146, 1996.
16. Hodge, A. M., G. K. Dowse, P. Toelupe, V. R. Collins, T. Imo, and P. Z. Zimmet. Dramatic increase in the prevalence of obesity in western Samoa over the 13 year period 1978–1991. Int. J. Obes. Relat. Metab. Disord. 18: 419–428, 1994.
17. Hoffmeister, H., G. B. Mensink, and H. Stolzenberg. National trends in risk factors for cardiovascular disease in Germany. Prev. Med. 23: 197–205, 1994.
18. Johnson, C. L., B. M. Rifkind, C. T. Sempos, et al. Declining serum total cholesterol levels among US adults: the National Health and Nutrition Examination Surveys. JAMA 269: 3002–3008, 1993.
19. Kaplan, G. A., and J. E. Keil. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation. 88 (4 Pt 1): 1973–1998, 1993.
20. Kelsey, J. L., A. S. Whittemore, A. S. Evans, and W. D. Thompson. Methods in Observational Epidemiology, 2nd ed. New York: Oxford University Press, 1996.
21. Kuskowska-Wolk, A., and R. Bergstrom. Trends in body mass index and prevalence of obesity in Swedish men 1980–89. J. Epidemiol. Community Health 47: 103–108, 1993.
22. Kuskowska-Wolk, A., and R. Bergstrom. Trends in body mass index and prevalence of obesity in Swedish women 1980–89. J. Epidemiol. Community Health 47: 195–199, 1993.
23. Lim, M. K., B. L. Liam, D. Ng, and R. Teow. Twenty-five years of national service: changes in height, weight and body mass index. Ann. Acad. Med. Singapore 23: 770–774, 1994.
24. Millar, W. J., and T. Stephens. Social status and health risks in Canadian adults: 1985 and 1991. Health Rep. 5: 143–156, 1993.
25. Monteiro, C. A., L. Mondini, A. L. de Souza, and B. M. Popkin. The nutrition transition in Brazil. Eur. J. Clin. Nutr. 49: 105–113, 1995.
26. Offer, A. Epidemics of abundance: overeating and slimming in the USA and Britain since the 1950’s. University of Oxford Discussion Papers in Economic and Social History, Number 25, Nov. 1998.
27. Peltonen, M., F. Huhtasaari, B. Stegmayr, V. Lundberg, and K. Asplund. Secular trends in social patterning of cardiovascular risk factor levels in Sweden. The Northern Sweden MONICA Study 1986–1994. Multinational Monitoring of Trends and Determinants in Cardiovascular Disease. J. Intern. Med. 244: 1–9, 1998.
28. Peto, R., Z. M. Chen, and J. Boreham. Tobacco: the growing epidemic (News). Nat. Med. 5: 15–17, 1999.
29. Pietinen, P., E. Vartiainen, and S. Mannisto. Trends in body mass index and obesity among adults in Finland from 1972 to 1992. Int. J. Obes. Relat. Metab. Disord. 20: 114–120, 1996.
30. Popkin, B. M., S. Paeratakul, K. Ge, and F. Zhai. Body weight patterns among the Chinese: results from the 1989 and 1991 China Health and Nutrition Surveys. Am. J. Public Health 85: 690–694, 1995.
31. Prentice, A. M., and S. A. Jebb. Obesity in Britain: gluttony or sloth? Br. Med. J. 311: 437–439, 1995.
32. Seidell, J. C., and K. M. Flegal. Assessing obesity: classification and epidemiology. Br. Med. Bull. 53: 238–252, 1997.
33. Seidell, J. C., W. M. M. Verschuren, and D. Kromhout. Levels and trends of obesity in The Netherlands, 1987–1991. Int. J. Obes. Relat. Metab. Disord. 19: 924–927, 1995.
34. Sheldon, T. Dutch are tallest in the world (News). Br. Med. J. 316: 1929, 1998.
35. Sichieri, R., D. C. Coitinho, M. M. Leao, E. Recine, and J. E. Everhart. High temporal, geographic, and income variation in body mass index among adults in Brazil. Am. J. Public Health 84: 793–798, 1994.
36. Simmons, G., R. Jackson, B. Swinburn, and R. L. Yee. The increasing prevalence of obesity in New Zealand: is it related to recent trends in smoking and physical activity? N. Z. Med. J. 109: 90–92, 1996.
37. Sobal, J. Obesity and socioeconomic status: a framework for examining relationships between physical and social variables. Med. Anthropol. 13: 231–247, 1991.
38. Sobal, J. The medicalization and demedicalization of obesity. In: Eating Agendas: Food and Nutrition as Social Problems, D. Maurer and J. Sobal (Eds.). New York: Aldine De Gruyter, 1995, pp. 67–90.
39. Sobal, J., and A. J. Stunkard. Socioeconomic status and obesity: a review of the literature. Psychol. Bull. 105: 260–275, 1989.
40. Sorensen, H. T., S. Sabroe, M. Gillman, et al. Continued increase in prevalence of obesity in Danish young men. Int. J. Obes. Relat. Metab. Disord. 21: 712–714, 1997.
41. Sprafka, J. M., G. L. Burke, A. R. Folsom, R. V. Luepker, and H. Blackburn. Continued decline in cardiovascular disease risk factors: results of the Minnesota Heart Survey, 1980–1982 and 1985–1987. Am. J. Epidemiol. 132: 489–500, 1990.
42. Stearns, P. N. Fat History: Bodies and Beauty in the Modern West. New York: New York University Press, 1997.
43. Troiano, R. P., and K. M. Flegal. Overweight children and adolescents: description, epidemiology, and demographics. Pediatrics 101: 497–504, 1998.
44. Troiano, R. P., and K. M. Flegal. Overweight prevalence: why so many different numbers? Int. J. Obes. Relat. Metab. Disord. 23(Suppl. 2): S22–S27, 1999.
45. Wilhelmsen, L., S. Johansson, A. Rosengren, I. Wallin, A. Dotevall, and G. Lappas. Risk factors for cardiovascular disease during the period 1985–1995 in Goteborg, Sweden: the GOT-MONICA Project. J. Intern. Med. 242: 199–211, 1997.
46. Wolk, A., and S. Rossner. Effects of smoking and physical activity on body weight: developments in Sweden between 1980 and 1989. J. Intern. Med. 237: 287–291, 1995.
47. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO Consultation on Obesity, Geneva, June 3–5, 1997. Geneva: World Health Organization, 1998.