Geography matters: the prevalence of diabetes in the Auckland Region by age, gender and ethnicity

Abstract

Aim: To determine whether the prevalence of diagnosed diabetes in the greater Auckland Region varies by General Electoral District (GED). Method: Using encrypted National Health Identifiers and record linkage of routine health datasets, we identified a regional cohort of people with diagnosed diabetes in 2011 from inpatient records and medication dispensing. The geographical unit of a person's residence (meshblock) was used to determine the GED of residence. We calculated prevalence estimates and 95% confidence intervals and used binary logistic regression to map geographical variations in diabetes. Results: An estimated 63,014 people had diagnosed diabetes in Auckland in 2011, a prevalence of 8.5% of the adult population ≥30 years of age. We found significant variation in diabetes prevalence by age, gender, ethnicity and GED. There was a more than five-fold difference in the unadjusted prevalence of diabetes by GED, ranging from 3.2% (3.1 to 3.4%) in the North Shore to 17.3% (16.8 to 17.7%) in Mangere. Such variations remained after binary logistic regression adjusting for socio-demographic variables. Compared to New Zealand Europeans, Indian people had the highest odds of having diabetes at 3.85 (3.73 to 3.97), while the odds of people living in the most deprived areas having diabetes was nearly twice that of those living in least deprived areas (OR 1.93, [1.87 to 1.99]). Geographic variations in diabetes remained after adjusting for socio-demographic circumstances: people living in GEDs in south-west Auckland were at least 60% more likely than people living in the North Shore GED to have diabetes. Conclusion: There is significant variation in the prevalence of diabetes by GED in Auckland that persists across strata of age group, gender and ethnicity, and persists after controlling for these same variables. These inequities should prompt action by politicians, policymakers, funders, health providers and communities for interventions aimed at reducing such inequities. Geography and its implications on access to and availability of health resources appears to be a key driver of inequity in diabetes rates, supporting an argument for interventions based on geography, especially a public health rather than an individual risk approach.