diture, or both. Indeed, the relative importance of each may vary across populations and settings.
Figures 1-2 and 1-3 show recent levels of obesity in three racial/ethnic groups in the United States and projections of future prevalence under the assumption that recently observed rates of increase will continue (Wang et al., 2008). The approximately one-third of men in all three groups who were obese in 1999-2004 could increase to nearly one-half within two decades. The same could be true for non-Hispanic white women, with even higher levels for non-Hispanic black and Mexican American women, for whom the prevalence of obesity was already greater than 50 percent and 40 percent, respectively.
Prevention efforts should not only aim to improve weight levels in the population at large but also target subpopulations with higher-than-average risk. Figures 1-2 and 1-3 identify key high-risk population subgroups—Mexican Americans and African Americans (Hedley et al., 2004; Ogden et al., 2007). Other U.S. ethnic groups with above-average obesity prevalence include American Indians, Pacific Islanders, and Native Hawaiians (Anderson and Whitaker, 2009; Broussard et al., 1991; Davis et al., 2004; Kumanyika, 1993). Low-income children and youth are also at high risk for obesity.
In many cases, above-average obesity prevalence is also observed in low-income populations, or high income is associated with lower obesity prevalence. In adults, the inverse gradient in obesity levels is seen most clearly in white females (Ogden, 2009). Figure 1-4 shows data for children with both sexes combined (Freedman et al., 2007).