Prevalence trends tell us what did not precipitate the US obesity epidemic
Hypotheses abound concerning the origins of the global obesity epidemic. Genetic predisposition, maternal obesity, excessive maternal weight gain, diabetes, tobacco use during pregnancy, and prenatal exposure to obesogens or endocrine disruptors have all been implicated.1Gluckman P Hanson M Mis-match. Why our world no longer fits our bodies. Oxford University Press, Oxford2008Google Scholar, 2McAllister EJ Dhurandhar NV Keith SW et al.Ten putative contributors to the obesity epidemic.Crit Rev Food Sci Nutr. 2009; 49: 868-913Crossref PubMed Scopus (513) Google Scholar An international survey3European Association for the Study of ObesityObesity perception and policy, multi-country review and survey of policymakers.http://easo.org/wp-content/uploads/2014/05/C3_EASO_Survey_A4_Web-FINAL.pdfDate: May 2014Google Scholar of more than 300 policymakers reported that more than 90% believed personal motivation was a strong or very strong influence on the rise of obesity. We describe how simple descriptive statistics can explain a great deal about what factors may, and may not, have caused the rapid rise in prevalence of obesity in the USA, where this phenomenon was first apparent. Since 1960, the National Health Examination Study and National Health and Examination Surveys (NHANES) have been done on a regular basis, by the US Centers of Disease Control, among representative samples of the general population (non-institutionalised population).4Centre of Disease Control and PreventionNational Health and Nutrition Examination Survey.www.cdc.gov/nchs/nhanes/index.htmGoogle Scholar Height and weight are measured in a standardised fashion, and are used to calculate body mass index. Results are shown for men and women of different ages (figure). Similar patterns to those shown in the figure were seen when trends were graphed separately for non-Hispanic white, non-Hispanic black, and Mexican American men and women.4Centre of Disease Control and PreventionNational Health and Nutrition Examination Survey.www.cdc.gov/nchs/nhanes/index.htmGoogle Scholar The increases in the prevalence of obesity began in the late 1970s across the whole US population.5Reither EN Hauser RM Yang Y Do birth cohorts matter? Age-period-cohort analyses of the obesity epidemic in the United States.Soc Sci Med. 2009; 69: 1439-1448Crossref PubMed Scopus (208) Google Scholar The speed and extent of weight gain varied somewhat by age, sex and ethnicity5Reither EN Hauser RM Yang Y Do birth cohorts matter? Age-period-cohort analyses of the obesity epidemic in the United States.Soc Sci Med. 2009; 69: 1439-1448Crossref PubMed Scopus (208) Google Scholar but for all subgroups most people became heavier at about the same time. This simple observation indicates something important about factors that did not precipitate the US obesity epidemic. We believe it is implausible that each age, sex and ethnic group, with massive differences in life experience and attitudes, had a simultaneous decline in willpower related to healthy nutrition or exercise. If intrauterine exposures played a major causative role,1Gluckman P Hanson M Mis-match. Why our world no longer fits our bodies. Oxford University Press, Oxford2008Google Scholar, 2McAllister EJ Dhurandhar NV Keith SW et al.Ten putative contributors to the obesity epidemic.Crit Rev Food Sci Nutr. 2009; 49: 868-913Crossref PubMed Scopus (513) Google Scholar one would have to hypothesise that there was a time-lag of about 70 years for babies born in 1910, 60 years for those born in 1920, and so on. This is also clearly implausible. Changes in genetic predisposition do not occur over the period of a few years, nor do they affect all age groups simultaneously. Indeed, it is unlikely that any factor with a long induction period had a major role in precipitating the US obesity epidemic. Nor can quick-acting exposures have been important if they only affect certain subgroups of the population, because whole population distributions of bodyweight shifted to the right. Rather, the epidemic must have been caused by factors that led to rapid population-wide changes. In the search for factors that can plausibly explain the precipitation of this epidemic, the question is not whether the factors in question cause obesity, or even if they affect all population subgroups and have been increasing over time. Instead, the question is, are these factors a cause of obesity that increased substantially in all major US population subgroups in the late 1970s? One candidate is the change to US farm bills in the 1970s, which led to a rapid increase in food production and thus an increase in food portion sizes;6Young LR Nestle M Expanding portion sizes in the US marketplace: implications for nutrition counseling.J Am Diet Assoc. 2003; 103: 231-240Summary Full Text Full Text PDF PubMed Scopus (238) Google Scholar accelerated marketing, availability, and affordability of energy dense foods;7Bleich SN Cutler D Murray C Adams A Why is the developed world obese?.Annu Rev Public Health. 2008; 29: 273-295Crossref PubMed Scopus (255) Google Scholar and widespread introduction of cheap and potent sweetening agents, such as high-fructose corn syrup, which infiltrated the food system and affected the whole population simultaneously.8Bray GA Nielsen SJ Popkin BM Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity.Am J Clin Nutr. 2004; 79: 537-543Crossref PubMed Scopus (1474) Google Scholar Although other countries that are experiencing a sharp rise in obesity rates might not have the equivalent of the US farm bill, most have been exposed to similar substantial changes in food supplies, with consequences for dietary patterns, such as increased portion sizes. By contrast, there have been marked differences between age groups in the slowing of the epidemic—eg, obesity rates have plateaued or reversed in children.9Rose G Sick individuals and sick populations.Int J Epidemiol. 1985; 14: 32-38Crossref PubMed Scopus (2262) Google Scholar These cohort effects suggest a more heterogeneous set of explanations, less ubiquitous than those that applied at the onset of the epidemic. More than 30 years ago, Geoffrey Rose noted that the determinants of population-level changes in incidence could be missed if one only focused on the causes of individual cases.9Rose G Sick individuals and sick populations.Int J Epidemiol. 1985; 14: 32-38Crossref PubMed Scopus (2262) Google Scholar But the tendency continues to confuse factors that act on individuals with the fundamental, underlying causes of population epidemics. Many factors, such as genetic predisposition, help explain where individuals are placed on the distribution of individual weights. However, they do not explain why the whole distribution of body mass has shifted towards heavier weights so rapidly in many populations. The search for the causes of such epidemics requires consideration of factors that have a mass exposure, are widely distributed, and act with short time-lags. It is not necessarily simple to identify causes of period effects (ie, those that act on the whole population at the same time) and then to intervene. It should also not be forgotten that immediate benefits can accrue from addressing factors with long induction periods—for example, gains due to smoking cessation. But when strong period effects are evident, as they are here, they signal quick potential wins in population health.10Woodward A Blakely A The healthy country? A history of life and death in New Zealand. Auckland University Press, Auckland2014Google Scholar WHD receives consulting fees as a member of the scientific advisory committee for Weight Watchers, as a member and chair of the JPB Foundation's Poverty Advisory Board, and as a consultant for the RTI: Feeding Infants and Toddlers Study. WHD received a grant from Bridgespan to analyse NHANES data on obesity in young adults. All other authors declare no competing interests. Tackling obesity seriously: the time has comeAccording to WHO data, worldwide, more than 1·9 billion adults are overweight—of these over 650 million are obese. 41 million children under the age of 5 years and 340 million children and adolescents aged 5–19 years are overweight or obese. Obesity is a major risk factor for non-communicable diseases, such as cardiovascular disease, diabetes, and several cancers. Crucially, obesity is preventable. So how is it that obesity has nearly tripled worldwide since 1975? Full-Text PDF Open Access