Increased Rate of Hospitalization for Diabetes and Residential Proximity of Hazardous Waste Sites
Kouznetsova, Maria, Huang, Xiaoyu, Ma, Jing, Lessner, Lawrence, Carpenter, David O., Environmental Health Perspectives
BACKGROUND: Epidemiologic studies suggest that there may be an association between environmental exposure to persistent organic pollutants (POPs) and diabetes.
OBJECTIVE: The aim of this study was to test the hypothesis that residential proximity to POP-contaminated waste sites result in increased rates of hospitalization for diabetes.
METHODS: We determined the number of hospitalized patients 25-74 years of age diagnosed with diabetes in New York State exclusive of New York City for the years 1993-2000. Descriptive statistics and negative binomial regression were used to compare diabetes hospitalization rates in individuals who resided in ZIP codes containing or abutting hazardous waste sites containing POPs ("POP" sites); ZIP codes containing hazardous waste sites but with wastes other than POPs ("other" sites); and ZIP codes without any identified hazardous waste sites ("clean" sites).
RESULTS: Compared with the hospitalization rates for diabetes in clean sites, the rate ratios for diabetes discharges for people residing in POP sites and "other" sites, after adjustment for potential confounders were 1.23 [95% confidence interval (CI), 1.15-1.32] and 1.25 (95% CI, 1.16-1.34), respectively. In a subset of POP sites along the Hudson River, where there is higher income, less smoking, better diet, and more exercise, the rate ratio was 1.36 (95% CI, 1.26-1.47) compared to clean sites.
CONCLUSIONS: After controlling for major confounders, we found a statistically significant increase in the rate of hospitalization for diabetes among the population residing in the ZIP codes containing toxic waste sites.
KEY WORDS: behavior, diabetes mellitus, dioxins, negative binomial regression, PCBs, persistent pesticides, polychlorinated biphenyls, SES, socioeconomic status, ZIP codes. Environ Health Perspect 115: 75-79 (2007). doi:10.1289/ehp.9223 available via http://dx.doi.org/ [Online 18 August 2006]
Diabetes is one of the leading causes of death and one of the most costly diseases in developed countries. During 1980-2002 the number of people with physician-diagnosed diabetes in the United States increased more than 2-fold, from 5.8 million to 13.3 million. An estimated 5.2 million cases remain undiagnosed. In 2002, total direct and indirect health care costs for people with diabetes amounted to $132 billion [Centers for Disease Control and Prevention (CDC) 2003]. The prevalence of diabetes of all types was 6.3% in the United States in 2002, of which approximately 90-95% of cases is adult-onset, type 2 diabetes (CDC 2003).
Established risk factors for diabetes include age, hyperinsulinemia (a marker for insulin resistance), obesity, genetic factors, and a sedentary lifestyle [Haffner 1998; World Health Organization (WHO) 1994]. Socioeconomic status (SES) is also a risk factor, in that lower income is associated with an increased risk of obesity and sedentary life style (Brancati et al. 1996). The National Health Interview Survey (National Center for Health Statistics 2003) found race, sex, obesity, and age to be effect modifiers for the prevalence of diabetes. Diabetes generally increased more rapidly with obesity among women than among men, but there was no other consistent sex difference. African-American race was a strong risk factor for diabetes, especially among individuals of low SES. After adjustments for racial differences in age, SES, weight, and central adiposity, African Americans remained over twice as likely to have diabetes as whites [odds ratio (OR) = 2.35; 95% confidence interval (CI), 1.49-3.73; p = 0.0003] (Brancati et al. 1996).
In addition, recent epidemiologic evidence suggests associations between diabetes and several environmental exposures, including cigarette smoke (Will et al. 2001) and arsenic (Tsai et al. 1999). Dioxin-exposed populations have been found to be at increased risk of diabetes (Bertazzi et al. 1998; Cranmer et al. …