As the prevalence of obesity and overweight rises in the United States,1–5
researchers continue to investigate a range of mechanisms by which people attain excessive body weight.6–10
Agreement is growing that the source of the obesity epidemic lies in an environment that produces an energy gap,11–15
where energy intake exceeds energy expenditure even by as little as 100 excess calories per day.12,13
Yet, it is unclear whether this 100-calorie excess is a function of increased intake or decreased output (or some combination of both) in American activity and consumption behaviors over time.
Limited data exist on trends in energy intake and energy expenditure among Americans over the past 3 decades, but the data that are available are nonetheless consistent with the rise in obesity observed over the same period. Between 1977 and 1996, Americans increased their total energy consumption by about 200 kcal/day.16
This was largely a result of increased consumption of snacks and soft drinks, particularly among young adults,16,17
while vegetable and fruit consumption remained low.18–21
These consumption behaviors have all been linked to excess weight gain.22–25
The increased availability of inexpensive, energy-dense food and beverages26–30
coupled with a lack of access to fresh fruits and vegetables31,32
are some of the environmental factors that may contribute to these trends. The US population has also adopted an increasingly sedentary lifestyle33–36
in an environment that is associated with a reduction in energy expenditure, including car-dependent neighborhoods that discourage walking and biking37–41
and limited physical activity in schools.28,42
On average, American youth spend over 30 hours per week watching television,43
which is positively associated with being overweight, either through sedentary activity or through exposure to the marketing of poor-quality foods.43–46
Modern lifestyles are increasingly characterized by skipping breakfast and sleeping less,36,47–49
which have also been linked to energy imbalance.50–56
These reported trends in health behaviors, however, are based on data from repeated cross-sectional surveys18,21,57
that were often conducted up to 5 years or more apart,16,21,57
the results of which are typically reported in aggregate across a time span of 4 to 7 years.20,57
Moreover, published articles frequently focus on trends in only 1 health behavior (e.g., fruit and vegetable consumption18,20
) and not the relative practice of energy consumption and expenditure behaviors among individuals over time. As a result, more detailed trends in health behaviors, particularly as they illustrate subtle changes in the balance of energy intake and output occurring annually among American young adults over the last quarter century, are poorly understood.
Also, differences in these behaviors and their trends by gender, socioeconomic status (SES), and racial/ethnic background are not well described at a population level, even though well-documented health disparities in obesity by social position exist.4,58–61
Using data from the National Health and Nutrition Examination Survey (1988–2002), one study found that non-Hispanic Blacks, persons in poverty, and those with less than a high school education were less likely to meet US Department of Agriculture fruit and vegetable guidelines than were non-Hispanic Whites and socioeconomically advantaged individuals.57
However, these results were based on 2 cross-sectional data sets collected 5 years apart. Delva et al.10
used repeat cross-sectional data collected annually from secondary school students between 1986 and 2003 to report declining trends in the proportion of adolescents who ate breakfast or exercised regularly, with a lower prevalence among women, racial/ethnic minorities, and those with low SES. Trends in the frequency of these behaviors beyond the secondary school setting, however, remain largely undocumented.
The purpose of our study was, first, to investigate long-term patterns in weight-related health behaviors among young adults (aged 19–26 years) over the past 23 years (1984–2006) and, second, to assess how these patterns varied by social position (race/ethnicity, gender, and SES). Analyses were based on longitudinal data for multiple cohorts of individuals with frequent repeat measures to better track historical changes in weight-related health behaviors over time. By focusing on young adults, we aimed to better understand how weight-related health behaviors have changed in this early period of the adult life course, when many adult health behavior patterns show their formative roots. The transition to adulthood (sometimes referred to as emerging or early adulthood) is a period when individuals are on their own typically for the first time, when life plans are put into action, and when distinctive life paths become more manifest.62
We hypothesized that the frequency of healthy behaviors would generally decline among young adults over this period, and that the rate of decline would be greater among those in disadvantaged social positions (women, Blacks, Hispanics, and those of lower SES).