Epidemiological Study Of Uric Acid And Ankle Brachial Index In Hospitalized Patients With Coronary Heart Disease Or Diabetes Mellitus Or Ischemic Stroke

Background: Peripheral arterial disease (PAD) is a clinical manifestation of the systemic atherosclerotic process and ankle brachial index (ABI) is an ideal tool to diagnosis PAD. However, the systemic study about uric acid (UA) and ABI has not been occurred in Chinese hospitalized patients with coronary heart disease (CHD) or diabetes mellitus (DM) or ischemic stroke (IS). Objectives: The present study was to know about (1) the distribution of ABI and PAD among the hospitalized patients with CHD or DM or IS, (2) the correlation between UA and PAD in hospitalized patients with CHD or DM or IS, and (3) the predictive value of UA and ABI for mortality from all-cause and cardiovascular disease (CVD) in patients with CHD or DM or IS. Methods: Subjects with CHD or DM or IS aged≥45 years were recruited from the Department of Cardiology, Neurology or Endocrine Secretion in-patient clinic at 32 university hospitals from July 2004 to November 2005 in Beijing and Shanghai. Finally, 2509 participants, including 1361 men (54.2%) and 1446 women (45.8%) entered our study. The mean age of patients was 68.92±9.81 years and they were followed up from November 2005 to January 2006. The diagnosis of PAD was based on an ABI≤0.90 on either side of the lower extremities as described in the ACC/AHA consensus recommendations. Results: (1) The mean value of ABI was 0.96±0.24 and the prevalence of PAD was 31.1% (95% confidence interval: 29.3%~33.0%) among Chinese patients with CHD or DM or IS aged≥45 years. However, only 7.6% of PAD patients have history of disease record. The mean value of ABI among female was lower than that of male. However, the prevalence of PAD among female was 33.9%, which was obviously higher than 28.8% among male (chi-squire test, P=0.006). The prevalence of PAD dramatically increased with advancing age (P-trend<0.000l), whereas the mean value of ABI. (2) Multivariable logistic regression analysis indicates that female gender, older age, history of smoking, history of hypertension or CHD or DM or IS, higher serum level of total cholesterol (TC), lower serum level of high-density lipoprotein cholesterol (HDL-C), hyperuricemia were independently associated with PAD. The odds ratio (OR) of smoking and diabetes for PAD was 1.647 (95% confidence interval: 1.312-2.069) and 2.045 (95% confidence interval: 1.652-2.532), respectively. Prevalence of PAD for trintiles of UA level was 26.4%, 29.5% and 37.5%, respectively (P-trend<0.05). The multivariate-adjusted OR for PAD was analyzed separately with UA level as a continuous variable, binary variable and scale variable (all P value<0.05). The optimal cut-off point for UA as determined by the receiver operating characteristic curve was 229.0μmol/L. The sensitivity and specificity at this cut-off point was 0.839 and 0.219, respectively. The area under curve was 0.562 (95% confidence interval: 0.538-0.587). The multivariate-adjusted OR for PAD for UA above this level was 1.480 (95% confidence interval: 1.170- 1.871) . The results after excluding cases using diuretics were similar. (3) During the 13-months follow-up, there were 231 (9.2%) deaths, of which 130 (5.2%) died of CVD. PAD was associated with mortality from all-cause and CVD, whose adjusted relative risk was 1.485 (95% confidence interval: 1.121-1.967) and 1.970 (95% confidence interval: 1.351-2.874), respectively, in Cox regression models. By multivariable Cox regression analysis, the adjusted relative risk of hyperuricemia for mortality from all-cause and CVD was 1.429 (95% confidence interval: 1.071-1.909) and 1.500 (95% confidence interval: 1.025-2.195), respectively. The survival rate was significantly lower in group with PAD or hyperuricemia than in the group without PAD or hyperuricemia (P<0.05). In Cox regression models, the risk of all-cause and CVD mortality was increased with the decline of ABI. Low-UA group (<266.0μmol/L) and high-UA group (> 351.05μmol/L) were associated with mortality from all-cause, whose adjusted relative risk was 1.624 (95% confidence interval: 1.140-2.315) and 1.558 (95% confidence interval: 1.100-2.208), respectively, in Cox regression models. But there was no significant difference between UA levels and CVD mortality. Conclusion: The prevalence of PAD was 31.1% and many independent risk factors were associated with PAD in Chinese patients with CHD or DM or IS. Low ABI was independently associated with a high risk of all-cause and CVD mortality and ABI should be promoted as an ideal tool to predict mortality. We should strengthen the management of risk factors integration for PAD. Hyperuricemia was an independent risk factor for PAD in patients with CHD or DM or IS. Low-UA or high-UA levels can be used to predict mortality from all-cause, but there was no significant correlation between UA levels and CVD mortality in Chinese patients with CHD or DM or IS. Further study was promoted for the correlation between UA and CVD mortality.