OBJECTIVE:To examinetrendsinlipidmanagement(cholesterol testing, treatment, and goal attainment) amongpatientswithdiabetesand to analyze the factors associated with initiation oflipid-lowering therapy.METHODS:We conducted a longitudinal, retrospective study ofpatientswithdiabetesidentified during a 24-month baseline period (January 1, 1995, to December 31, 1996) and for whom follow-up was continued for 5 years (1997 to 2001). Generalized estimating equations were used to test for time trend effects inlipidmanagement. We modeled the days from baseline to the firstlipid-lowering prescription fill date with a multivariate Cox proportional hazards regression model.RESULTS:Rates oflipidtesting, treatment, and goal attainment significantly improved (P<0.001) during the 5-year study period: from 37% to 67% forlipidtesting; from 19% to 41% for treatment with alipid-lowering agent; from 22% to 37% for achievement of low-density lipoprotein cholesterol (LDL-C) levels < 100 mg/dL; and from 54% to 75% for achievement of LDL-C levels < 130 mg/dL. The relative likelihood (hazard rate) of treatment withlipid-lowering agents was greater forpatientswith LDL-C levels > or = 100 mg/dL relative topatientswith LDL-C concentrations < 100 mg/dL. Treatment withlipid-lowering agents ofpatientswith a cardiovascular event during follow-up was approximately 3 times more likely relative to those without such an event.CONCLUSION:We found that rates oflipidtesting, treatment, and goal attainment improved significantly between 1997 and 2001. Nevertheless, ample room for improvement of these rates continues to exist. Particular attention may be warranted to ensure thatpatientswithdiabetesreceivelipid-lowering agents not only after a cardiovascular event but also before such an event occurs.