The ultimate goal of therapy for type 2 diabetes mellitus (T2DM) is prevention of microvascular and macrovascular complications. Multiple studies have demonstrated that treating HbA1c below 7% reduces long-term microvascular problems.1 Additionally, meta-analyses have shown an association between reducing cardiovascular events and lowering HbA1c.2,3 Long-term follow-up of the DCCT study and UKPDS showed that early treatment of HbA1c to less than or near 7% was associated with a reduction in macrovascular complications. In contrast, the ADVANCE, ACCORD, and VADT trials showed no association between macrovascular outcomes and intensive glycemic control.
A recent retrospective cohort study by Currie et al. evaluated the association between all-cause mortality and HbA1c in patients with T2DM in the primary care setting.4 Data were collected from the General Practice Research Database (GPRD) in the UK from November 1986 to November 2008. GPRD is a proprietary health data resource that gathers information from the computerized daily records of general practitioners. The database includes demographic information, diagnoses, test results, smoking status, drug treatments, mortality, and additional health-related data. Patients with a diagnosis of T2DM whose treatment course showed escalation of therapy for diabetes, received oral hypoglycemic agents or insulin, were older than 50, and had a case history of at least 6 months were included in the analysis. Patients were divided into two cohorts depending on the type of therapy escalation. Those with a newly-identified switch from oral hypoglycemic monotherapy to combination therapy with a sulfonylurea and metformin were included in cohort 1 (n = 27,965). Cohort 2 (n = 20,005) included patients who switched from oral therapy to insulin with or without concomitant oral hypoglycemic agents. Exclusion criteria included diabetes secondary to other causes (gestational or drug-induced) and patients with less than 12 months of exposure to intensified therapy required for cohort inclusion.
The post-index HbA1c was used to evaluate glycemic control. This variable was calculated using the mean of all HbA1c values recorded between index date (date of the initial prescription for intensified diabetes therapy) and outcome event (death or large vessel disease) or censoring point (last recorded value or when an additional change in therapy was made). The primary outcome was all-cause mortality and the secondary outcome was occurrence of major cardiovascular events in patients who had no record of cardiovascular disease prior to the index date. Cohorts were divided into deciles by the rank of the mean post-index HbA1c values. There were observed differences in gender, age, systolic blood pressure, smoking history, total cholesterol, weight, previous history of large vessel disease, duration of diabetes, and evidence of microvascular complications (elevated SCr and visual problems) between each intra-cohort decile. An analysis to assess the significance of these differences was not reported. Survival was assessed with Cox proportional hazards models and the covariates included were age at index date, sex, smoking, mean total cholesterol, BMI, and general comorbidity (measured using the Charleson index) in addition to the HbA1c decile.
The decile with a mean HbA1c of about ...