Steven H Hendriks1, Peter R van Dijk2, Kornelis J J van Hateren2, Joost L van Pelt3, Klaas H Groenier4, Henk J G Bilo5, Stephan J L Bakker6, Gijs W D Landman7, Nanne Kleefstra8. 1. Diabetes Centre, Isala, Zwolle, the Netherlands. Electronic address: s.hendriks@isala.nl. 2. Diabetes Centre, Isala, Zwolle, the Netherlands. 3. Department of Clinical Chemistry, University Medical Center Groningen and University of Groningen, Groningen, The Netherlands. 4. Diabetes Centre, Isala, Zwolle, the Netherlands; Department of General Practice, University Medical Center Groningen and University of Groningen, Groningen, the Netherlands. 5. Diabetes Centre, Isala, Zwolle, the Netherlands; Department of Internal Medicine, University Medical Center Groningen and University of Groningen, Groningen, the Netherlands; Department of Internal Medicine, Isala, Zwolle, the Netherlands. 6. Department of Internal Medicine, University Medical Center Groningen and University of Groningen, Groningen, the Netherlands. 7. Diabetes Centre, Isala, Zwolle, the Netherlands; Department of Internal Medicine, Gelre hospital, Apeldoorn, the Netherlands. 8. Diabetes Centre, Isala, Zwolle, the Netherlands; Department of Internal Medicine, University Medical Center Groningen and University of Groningen, Groningen, the Netherlands; Langerhans Medical Research Group, Zwolle, the Netherlands.
Abstract
BACKGROUND: We aimed to investigate whether high-sensitive cardiac troponin T (hs-cTnT) is associated with all-cause and cardiovascular mortality in stable type 2 diabetes (T2D) outpatients treated in primary care. METHODS:Cardiac troponin T was measured with a high-sensitive assay at baseline in patients with T2D participating in the observational ZODIAC study. Cox proportional hazards models were used to investigate the relationship between hs-cTnT and mortality with adjustment for selected confounders. Risk prediction capabilities of hs-cTnT were assessed with Harrell C statistics. RESULTS: Complete baseline data were available for 1,133 patients. During median follow-up of 11 (7-14) years, 513 (45%) patients died, of which 218 (42%) died of cardiovascular causes. Of the patients with undetectable hs-cTnT levels (<3 ng/L), only 23% died, compared with 58% with low detectable levels (3-14 ng/L) and 84% with raised levels (≥14 ng/L). Natural log hs-cTnT was significantly associated with all-cause mortality (hazard ratio 1.30, 95% CI 1.19-1.42) and cardiovascular mortality (hazard ratio 1.33, 95% CI 1.15-1.53), independent of potential confounders. The Harrell C statistic for the crude model of hs-cTnT was 0.72 (95% CI 0.70-0.75) for all-cause mortality and 0.74 (95% CI 0.71-0.77) for cardiovascular mortality. CONCLUSIONS: Higher levels of hs-cTnT are associated with mortality in stable outpatients with T2D. The high crude Harrell C values and the excellent prognosis of patients with undetectable levels illustrate the strength of hs-cTnT as a potential marker for mortality.
RCT Entities:
BACKGROUND: We aimed to investigate whether high-sensitive cardiac troponin T (hs-cTnT) is associated with all-cause and cardiovascular mortality in stable type 2 diabetes (T2D) outpatients treated in primary care. METHODS: Cardiac troponin T was measured with a high-sensitive assay at baseline in patients with T2D participating in the observational ZODIAC study. Cox proportional hazards models were used to investigate the relationship between hs-cTnT and mortality with adjustment for selected confounders. Risk prediction capabilities of hs-cTnT were assessed with Harrell C statistics. RESULTS: Complete baseline data were available for 1,133 patients. During median follow-up of 11 (7-14) years, 513 (45%) patients died, of which 218 (42%) died of cardiovascular causes. Of the patients with undetectable hs-cTnT levels (<3 ng/L), only 23% died, compared with 58% with low detectable levels (3-14 ng/L) and 84% with raised levels (≥14 ng/L). Natural log hs-cTnT was significantly associated with all-cause mortality (hazard ratio 1.30, 95% CI 1.19-1.42) and cardiovascular mortality (hazard ratio 1.33, 95% CI 1.15-1.53), independent of potential confounders. The Harrell C statistic for the crude model of hs-cTnT was 0.72 (95% CI 0.70-0.75) for all-cause mortality and 0.74 (95% CI 0.71-0.77) for cardiovascular mortality. CONCLUSIONS: Higher levels of hs-cTnT are associated with mortality in stable outpatients with T2D. The high crude Harrell C values and the excellent prognosis of patients with undetectable levels illustrate the strength of hs-cTnT as a potential marker for mortality.
Authors: Dorijn F L Hertroijs; Arianne M J Elissen; Martijn C G J Brouwers; Nicolaas C Schaper; Sebastian Köhler; Mirela C Popa; Stylianos Asteriadis; Steven H Hendriks; Henk J Bilo; Dirk Ruwaard Journal: Diabetes Obes Metab Date: 2017-11-24 Impact factor: 6.577