Evaluating ethnic variations in the risk of infections presenting to primary and secondary care in people with pre-diabetes and type 2 diabetes: a matched cohort study

Problem

People living with type 2 diabetes have a higher infection risk. However, it is unknown how this risk varies by ethnicity, or whether this risk is similarly observed in people with non-diabetic hyperglycaemia (“pre-diabetes”). The study has examined the magnitude of infection risks among adults with type 2 diabetes or pre-diabetes in England, and whether patterns of risk were similar in people from different ethnic groups.

Approach

Using Clinical Practice Research Datalink (CPRD), patients in England aged 18-90 with pre-diabetes, and type 2 diabetes, alive on 1/1/2015 were matched to patients without diabetes. Each individual with type 2 diabetes was matched to two patients without diabetes or prediabetes on age, sex and ethnic group. Ethnicity was categorised into five broad categories: White, South Asian, Black, Mixed/Other and missing ethnicity. Infections during 2015-2019 were collated from primary care [CPRD Aurum] and linked hospitalisation records, and included any infection with a prescription in primary care for an antibiotic, antifungal or antiviral within +/- 14 days of the diagnosis or any new hospital episode where an infection was the primary diagnosis. Infection incidence rate ratios (IRR) for pre- or type 2 diabetes were estimated with 95% confidence intervals (95%CI).

Findings

Of the 527,151 people with type 2 diabetes, 69.9%, 10.4%, 4.3% and 5.8% were of White, South Asian, Black, Mixed/Other ethnicity respectively, with 9.7% missing. There was an increased risk for infections in people with type 2 diabetes presenting in primary care (IRR = 1.51, 95%CI 1.51-1.52) and hospitalisations (IRR = 1.91, 1.90-1.93). Overall, within each ethnic group, this was broadly consistent regardless of underlying differences in age and deprivation, and seen for all specific infection types considered, though younger (age <50) people with type 2 diabetes of White ethnicity experienced a greater relative risk. For pre-diabetes (n = 273,216), a significant but smaller risk was seen for primary care (IRR = 1.35, 95%CI 1.34-1.36) and hospitalisations (IRR = 1.33, 95%CI 1.31-1.35); similar within each ethnicity for primary care infections, but less consistent for infection-related hospitalisations. In the population, 5.3% of primary care and 8.9% of hospitalisation infections were attributable to pre-diabetes or type 2 diabetes.

Consequences

Our study estimated an elevated risk of infection for people with type 2 diabetes in England, that was broadly similar in each major ethnic group. An increased relative risk of infections was also seen in people with pre-diabetes compared to people without diabetes. Infections are therefore a significant cause of ill health and health service use across both primary and secondary care for people with pre-diabetes and type 2 diabetes. Given the rising prevalence of type 2 diabetes and pre-diabetes, the burden of infections will continue to have public health implications.