Rebecca S Gold1, Jonathan T Unkart2, Robyn L McClelland3, Alain G Bertoni4, Matthew A Allison2. 1. School of Medicine, University of California San Diego, La Jolla, California, USA. Electronic address: rsgold@health.ucsd.edu. 2. Department of Family Medicine and Public Health, University of California San Diego, La Jolla, California, USA. 3. Department of Biostatistics, University of Washington, Seattle, Washington, USA. 4. Department of Epidemiology and Prevention, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA.
Abstract
AIMS: To investigate associations of health insurance with measures of glucose metabolism, and whether associations vary by diabetes status or insurance type. METHODS: Cross-sectional analysis of baseline data from the Multi-Ethnic Study of Atherosclerosis. Cohort a priori stratified by age <65 (N = 3,665) and ≥65 years (N = 2,924). Multivariable linear and logistic regression assessed associations between insurance and fasting glucose, HOMA-IR, and prevalent diabetes, controlling for relevant confounders, including age, sex, race/ethnicity, income, and education. RESULTS: In participants <65, compared to uninsured, having any insurance was associated with lower fasting glucose in participants with diabetes (Mean Difference = -20.4 mg/dL, P = 0.01), but not in participants without diabetes. Compared to Private insurance, uninsured participants had higher fasting glucose (Mean Difference = 3.8 mg/dL, P = 0.03), while participants with Medicaid had higher HOMA-IR (Mean Difference = 3.5 mg/dL, P < 0.01). In participants ≥65, compared to Private insurance, uninsured participants (Mean Difference = 7.5 mg/dL, P = 0.02), and participants with Medicaid only (Mean Difference = 19.9 mg/dL, P < 0.01) or Medicare + Medicaid (Mean Difference = 5.2 mg/dL, P = 0.03) had higher fasting glucose. CONCLUSIONS: In this large multiethnic cohort, having any insurance was associated with significantly lower fasting glucose for individuals with diabetes. Levels of fasting glucose and insulin resistance varied across different insurance types.
AIMS: To investigate associations of health insurance with measures of glucose metabolism, and whether associations vary by diabetes status or insurance type. METHODS: Cross-sectional analysis of baseline data from the Multi-Ethnic Study of Atherosclerosis. Cohort a priori stratified by age <65 (N = 3,665) and ≥65 years (N = 2,924). Multivariable linear and logistic regression assessed associations between insurance and fasting glucose, HOMA-IR, and prevalent diabetes, controlling for relevant confounders, including age, sex, race/ethnicity, income, and education. RESULTS: In participants <65, compared to uninsured, having any insurance was associated with lower fasting glucose in participants with diabetes (Mean Difference = -20.4 mg/dL, P = 0.01), but not in participants without diabetes. Compared to Private insurance, uninsured participants had higher fasting glucose (Mean Difference = 3.8 mg/dL, P = 0.03), while participants with Medicaid had higher HOMA-IR (Mean Difference = 3.5 mg/dL, P < 0.01). In participants ≥65, compared to Private insurance, uninsured participants (Mean Difference = 7.5 mg/dL, P = 0.02), and participants with Medicaid only (Mean Difference = 19.9 mg/dL, P < 0.01) or Medicare + Medicaid (Mean Difference = 5.2 mg/dL, P = 0.03) had higher fasting glucose. CONCLUSIONS: In this large multiethnic cohort, having any insurance was associated with significantly lower fasting glucose for individuals with diabetes. Levels of fasting glucose and insulin resistance varied across different insurance types.
Authors: Quyen Ngo-Metzger; Dara H Sorkin; John Billimek; Sheldon Greenfield; Sherrie H Kaplan Journal: J Gen Intern Med Date: 2011-10-18 Impact factor: 5.128