Predictive value of ankle-brachial index to all-cause mortality and cardiovascular mortality in Chinese patients with chronic kidney disease.

Background: To investigate the predictive value of ankle-brachial index (ABI) for all-cause mortality and cardiovascular mortality in Chinese patients with chronic kidney disease (CKD).
Patients and Methods: 1563 CKD patients were enrolled in the cohort and were followed up for about 3 years in China. CKD was defined as an eGFR less than 60 ml/min/1.73m(2). 573 participants were diagnosed with PAD using ABI <= 0.90. Their average age was 73.4 ±8.2 years.
Results: During a median follow-up of 38 months, there were 1353 CKD patients with complete data. A total of 313 patients (161 with and 152 without PAD) died during follow-up. 184 deaths (99 with and 85 without PAD) were caused by cardiovascular disease (CVD). All-cause and CVD mortality of CKD patients with PAD was increased 2.2-fold and 2.4-fold compared with CKD patients without PAD. The hazard ratio (HR) of PAD for all-cause and CVD mortality was 2.15 (95 % CI: 1.66 - 2.79) and 2.51 (95 % CI: 1.80 - 3.50) respectively. Mortality of CKD patients significantly increased with decreasing ABI. That of CKD patients with ABI <= 0.4 was the highest (42.9 % and 28.6 %, respectively) in different ABI categories. Relative risks of all-cause and CVD mortality of CKD patients with ABI <= 0.4 were increased 3.479-fold (95 % CI: 2.076 - 5.830) and 4.960-fold (95 % CI: 2.644 - 9.302) respectively compared with those of patients with ABI > 1.0 and <= 1.4. Special models to evaluate the predictive value of ABI to mortality of CKD patients suggested that addition of ABI significantly increased the predictive value of the model for 3-year mortality compared with a model including conventional risk factors alone.
Conclusions: Low ankle-brachial index can predict increased mortality of chronic kidney disease patients. Addition of ankle-brachial index can significantly improve the prediction of 3-year mortality compared with conventional risk factors alone.