Changes in glycemic control from 1996 to 2006 among adults with type 2 diabetes: a longitudinal cohort study
1 The Diabetes Center, Massachusetts General Hospital, 50 Staniford Street, Boston MA, 02114, USA
2 Division of General Medicine, Massachusetts General Hospital, Fruit Street, Boston MA, 02114, USA
3 Harvard Medical School, Shattuck Street, Boston, MA 02115, USA
BMC Health Services Research 2010, 10:158 doi:10.1186/1472-6963-10-158Published: 9 June 2010
Our objectives were to examine temporal changes in HbA1c and lipid levels over a 10-year period and to identify predictors of metabolic control in a longitudinal patient cohort.
We identified all adults within our hospital network with T2DM who had HbA1c's measured in both 1996 and 2006 (longitudinal cohort). For patients with no data in 2006, we used hospital and social security records to distinguish patients lost to follow-up from those who died after 1996. We compared characteristics of the 3 baseline cohorts (longitudinal, lost to f/u, died) and examined metabolic trends in the longitudinal cohort.
Of the 4944 patients with HbA1c measured in 1996, 1772 (36%) had an HbA1c measured in 2006, 1296 (26%) were lost to follow-up, and 1876 (38%) had died by 2006. In the longitudinal cohort, mean HbA1c decreased by 0.4 ± 1.8% over the ten-year span (from 8.2% ± 1.7% to 7.8% ± 1.4%) and mean total cholesterol decreased by 49.3 (± 46.5) mg/dL. In a multivariate model, independent predictors of HbA1c decline included older age (OR 1.41 per decade, 95% CI: 1.3-1.6, p < 0.001), baseline HbA1c (OR 2.9 per 1% increment, 2.6 - 3.2, p < 0.001), and speaking English (OR 2.1, 1.4-3.1, p < 0.001).
Despite having had diabetes for an additional 10 years, patients in our longitudinal cohort had better glycemic and cholesterol control in 2006 than 1996. Greatest improvements occurred in patients with the highest levels in the baseline year.