The Spread of the Obesity Epidemic in the United States, 1991-1998 FREE

Although attempts to lose weight are common in the United States,^1- 2 the prevalence of obesity has increased since the 1980s.^3- 4 Such increases will tremendously affect public health since obesity is strongly associated with several chronic diseases, such as cardiovascular diseases and diabetes.^5- 6 Recent estimates suggest that obesity-related morbidity may account for 6.8% of US health care costs.⁷

Recently published trend data from the National Health and Nutrition Examination Surveys (NHANES) show the percentage of obese persons has increased from 14.5% in the years 1976-1980 to 22.5% in 1988-1994.³ To monitor obesity trends since 1994 and to present state-specific results, we used data from a large population-based survey to examine changes in the prevalence of obesity among adults in the United States from 1991 to 1998.

We analyzed data from all states that participated in the Behavioral Risk Factor Surveillance System (BRFSS). The BRFSS, a cross-sectional telephone survey of noninstitutionalized adults aged 18 years or older, is conducted by the Centers for Disease Control and Prevention and state health departments. The BRFSS questionnaire concerns personal behaviors that increase risk for 1 or more of the 10 leading causes of death in the United States.⁸

The BRFSS uses a multistage cluster design based on random digit dialing methods to select a representative sample from each state's noninstitutionalized residents.^9- 10 Data collected from each state are pooled to produce nationally representative estimates.⁹ Further details about the BRFSS have been published.^9- 10

We calculated body mass index (BMI) (weight in kilograms divided by the square of height in meters) based on self-reported weight and height. Participants were classified as obese if their BMI was greater than or equal to 30 kg/m².¹¹ Questions on leisure-time physical activity were included in the BRFSS only in 1991, 1992, 1994, 1996, and 1998 and were used to create a leisure-time physical activity score: inactive, irregularly active, regular, not intense, and regular, intense.¹² We used SAS and SUDAAN statistical software in the analyses and to account for the complex sampling design.^13- 14 Because of the large sample size (more than 100,000 participants each year), we have not emphasized statistical testing.

We excluded from our trend analyses 6 states because they did not collect weight and height for 1 or more years: Arkansas (1992), the District of Columbia (1995), Kansas (1991), Nevada (1991), Rhode Island (1994), and Wyoming (1991, 1992, and 1993). However, we included the data for all available years for these and all other states in our maps. We used the US Bureau of Census method of grouping states into regions.

The prevalence of obesity increased from 12.0% in 1991 to 17.9% in 1998 (Table 1). Obesity increased in men and women and across all sociodemographic groups (Table 2), with the highest increase among the youngest ages and higher education levels. Among Hispanic men, the prevalence of obesity increased from 10.0% in 1991 to 18.3% in 1997 and for Hispanic women from 13.2% to 23.4%. The prevalence of obesity increased steadily from 1991 to 1998 in all states (Figure 1).

In 1991, 4 of the 45 participating states had obesity rates of 15% or higher (Table 3). By 1998, 37 states had rates higher than 15%. The magnitude of the increase varied by region (ranging from 31.9%-67.2% increase in the mid Atlantic and South Atlantic regions, respectively) and by state, ranging from 11.3% in Delaware to 101.8% in Georgia.

In 1991, the level of leisure-time physical activity was 29.7% inactive, 28.4% irregularly active, 33.2% regular not intense, and 8.7% regular intense. In 1998, they were 28.6% inactive, 28.2% irregularly active, 29.6% regular not intense, and 13.6% regular intense.

To exclude the possibility that demographic differences accounted for the variation in obesity prevalence between the states, we computed the age-, sex-, and race-adjusted prevalence of obesity. Although the adjusted rates were higher than the unadjusted rates, similar patterns were observed among states and over time. Therefore, we only report the unadjusted estimates.

These data show that obesity increased in every state, in both sexes, and across all age groups, races, educational levels, and smoking statuses. Rarely do chronic conditions such as obesity spread with the speed and dispersion characteristic of a communicable disease epidemic. However, this rapid trajectory of obesity may present both clues to origin and measures for efficient attempts to control its spread.

Since overweight participants in self-reported studies tend to underestimate their weight and all participants tend to overestimate their height; true rates of obesity are likely underestimated.^15- 16 Moreover, people without telephones are not surveyed through BRFSS, and such individuals are likely to be of lower socioeconomic status, a factor that is associated with obesity.^17- 18 The net effect of these limitations is that the prevalence of obesity reported herein is likely a conservative estimate. In fact, the prevalence of obesity from NHANES III (1988-1994) in which weight and height were measured by health professionals was 22.5% in adults, more than a third higher than the rates reported in our survey.³

The BRFSS data provide states with unique population-based estimates of self-reported obesity against which prevention efforts may be evaluated. State-level population-based estimates of obesity should be used to provide each state with a basis for setting priorities for public health interventions.

This rapid increase in obesity in all segments of the population and regions of the country implies that there have been sweeping changes in US society that are contributing to weight gain by fostering energy intake imbalance. Such changes are unlikely to be due to diminished individual motivation to maintain weight or in genetic or other biological changes in the population.

We focused on describing the changes in patterns of obesity instead of its contributory factors, such as alterations in diet, activity patterns, or other behaviors that affect energy balance. However, our data demonstrate that a major contributor to obesity—physical inactivity—has not changed substantially at the population level between 1991 and 1998. By focusing on the challenge of stopping the obesity epidemic and the profound negative health consequences of obesity, it is important to increase the awareness and involvement of health professionals in dealing with the epidemic.

Our data suggest that the development of strategies and programs for weight maintenance as well as weight reduction must become a higher priority. Public health messages should focus increasingly on balancing energy intake with physical activity. To control the obesity epidemic, a wide range of population groups, including physicians and other health care professionals, public health professionals, legislators, communities, work sites, and organizations, must become engaged in working toward a solution.