Volume 11 — July 24, 2014
ORIGINAL RESEARCH
The Geography of Diabetes by Census Tract in a Large Sample of Insured Adults in King County, Washington, 2005–2006
This map of King County shows the smoothed prevalence of diabetes among adult members of the Group Health Cooperative health care system by census tract. The map has 5 categories of prevalence: the lowest category is census tracts with a prevalence of diabetes from 6.9% to 10%. These areas tend to be located in north Seattle, specifically in the following neighborhoods: Montlake, Madison Park, University District (University of Washington), Wallingford, Freemont, Ballard, Magnolia, and Queen Anne. The highest category represents census tracts with a prevalence of diabetes from 16.2% to 21.2%. Areas with a high prevalence of diabetes tended to be located in southern King County and south Seattle, specifically in the following neighborhoods or cities: Sea-Tac (location of Seattle-Tacoma International Airport), Tukwila, Renton, and Auburn.
Figure 1. Diabetes prevalence was smoothed by using an empirical Bayes tool, King County, Washington, 2005–2006. Eastern portions of census tracts in eastern King County are not shown because they are sparsely populated and consist mostly of forested land.
This map of King County shows the results of local clustering tests. Two primary clusters were identified. The first is a cluster of low diabetes prevalence, located in north Seattle, including the following neighborhoods of Seattle: Madison Park, Montlake, Eastlake, Queen Anne, Magnolia, Ballard, Freemont, Wallingford, Green Lake, University District (University of Washington), Ravenna, Laurelhurst, Windermere, and View Ridge. A primary cluster of high diabetes prevalence was observed in south Seattle and southern King County including the following neighborhoods or cities: Rainier Beach (in Seattle), Renton, Sea-Tac (location of Seattle-Tacoma International Airport), Tukwila, Kent, Auburn, Algona, and Pacific.
Figure 2. Local clusters of census-tract–level diabetes prevalence in King County, Washington, 2005–2006, as determined by the Getis-Ord* (Gi*) statistic (17). Eastern portions of census tracts in eastern King County are not shown because they are sparsely populated and consist mostly of forested land. In the key, “none” indicates no clustering; “high,” a cluster of census tracts that have a high prevalence of diabetes; “low,” a cluster of census tracts that have a low prevalence of diabetes.
This scatterplot has the smoothed diabetes prevalence on the y-axis with values ranging from about 6% to 21%. On the x-axis is the natural logarithm of median home value with a range of approximately 10.25 (~$28,300 after transformation) to 13.5 (~$730,000 after transformation). There is a negative and significant association between the natural logarithm of median home value and the smoothed prevalence of diabetes. The correlation coefficient is −0.48 (P < .001).
Supplemental Figure 1. Association between smoothed diabetes prevalence and the natural logarithm of median home value, King County, Washington, 2005–2006.
This scatterplot has the smoothed diabetes prevalence on the y-axis with values ranging from about 6% to 21%. On the x-axis is the percentage of adults with a college degree with values ranging from less than 5% to about 80%. There is a negative and significant association between the percentage of adults with a college degree and the smoothed prevalence of diabetes. The correlation coefficient is −0.60 (P < .001).
Supplemental Figure 2. Association between smoothed diabetes prevalence and the percentage of residents with a college degree, King County, Washington, 2005–2006.
This scatterplot has the smoothed diabetes prevalence on the y-axis with values ranging from about 6% to 21%. On the x-axis is the median household income ranging from $18,000 to about $140,000. There is a relatively weak negative but significant association between median household income and the smoothed prevalence of diabetes. The correlation coefficient is −0.18 (P < .001).
Supplemental Figure 3. Association between smoothed diabetes prevalence and median household income, King County, Washington, 2005–2006.
This scatterplot has the smoothed diabetes prevalence on the y-axis with values ranging from about 6% to 21%. On the x-axis is the smoothed obesity prevalence ranging from 17% to about 45%. There is a strong positive and significant association between the 2 variables. The correlation coefficient is 0.58 (P < .001).
Supplemental Figure 4. Association between smoothed diabetes prevalence and smoothed obesity prevalence, King County, Washington, 2005–2006.
The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors' affiliated institutions.
