Two-thirds (66.3%) of all 798 deaths after a median 2.1 years of follow-up were caused by one of five cardiovascular (CV) conditions, with sudden cardiac death accounting for the largest share (30.1%) of the total, Ilaria Cavallari, MD, PhD, and associates said in the Journal of the American College of Cardiology.
Most common among the non-CV causes was malignancy at 13.9% of all deaths in a T2DM population at high/very high risk for CV disease (n = 16,492), followed by infection (9.3%), the members of the TIMI Study Group noted.
After variables independently associated with overall mortality were identified, a subdistribution of competing risks was constructed using a competing-risk analysis based on the proportional hazards model, they explained.
Prior heart failure was the clinical variable most associated with CV death and could, along with older age, worse glycemic control, prior CV events, peripheral artery disease, and kidney complications, “identify a subgroup of T2DM patients at high risk of mortality who are likely to achieve the greatest benefit from aggressive management of modifiable risk factors and newer glucose-lowering agents,” the investigators wrote.
It was a pair of laboratory measurements, however, that had the largest subdistribution hazard ratios. “Interestingly, the magnitude of associations of abnormal N-terminal pro–B-type natriuretic peptide [sHR, 2.82] and high-sensitivity troponin T [sHR, 2.46] measured in a stable population were greater than clinical variables in the prediction of all causes of death,” Dr. Cavallari and associates said.