Risk Of Diabetes Among Hypertention Patients In Community Of Shanghai Under Antihypertensive Therapy With Valsartan Or Non RAS Inhibitors

Objective:This research investigates the new onset of diabetes between valsartan and non-RAS inhibitors on Chinese hypertension patients in real world setting.Methods:Real world setting was established on an Electric Health Recording (eHR) system from Minhang District of Shanghai in China, covering follow-up data of hypertension patients from January1,2005to July31,2011. There were159,775hypertension adult patients (age>18y), out of who selected patients at least12months continuous mono-therapy anti-hypertension treatment and established the corhort. The cumulative incidence of diabetes in each group was examined at the end of24th month. There were two series comparisons, including valsartan and non-RAS inhibitors (CCBs, β-blockers and diuretics), valsartan and representatives of the CCB (nifedipine, amlodipine and felodipine).Results:(1) Comparisons between valsartan and non-RAS inhibitors. A total of29295patients were selected, including2107(7.19%) in valsartan group,4094(13.96%) in β-blockers group,21397(73.04%) in CCB group and1697(5.79%) in diuretics group. At the end of the24th month, the new onset of diabetes in the valsartan-therapy group (5.437per thousand person-year (p. r)) was significantly (p=0.010) lower than that in the CCB group (7.088p. r). The multi-variable adjusted hazard ratio (HR) and95%confidential interval (CI) were0.722(0.601-0.867, Log rank p=0.001). Comparing with CCB, valsartan decreases the incidence of diabetes by27.8%. The incidence rate of β-blockers group (5.201p. r) and the diuretics group (4.683p. r) were not statistically significant comparing with valsartan group (Log rank p=0.715, Log rank p=0.350). Multi-ad justed HR and95%CI were1.027,(0.826-1.277, p=0.811)、1.074,(0.832-1.387, p=0.584) by adjusted factors of age, gender, BMI, baseline blood pressure, risk factors, complications and target organ damage. There were no differences in the comparisons of valsartan and β-blockers group, valsartan and diuretics group. The results of propensity score matching were consistent. (2) Comparisons between valsartan and CCB. A total of20097patients were selected. There included2107(10.80%) in valsartan group,3921(19.51%) in nifedipine group,7167(35.66%) in amlodipine group and6839(34.03%) in felodipine group respectively. The incidence rates at the end of24th months were valsartan5.596p. r, nifedipine group,7.310(Log rank p=0.001), amlodipine group6.944(p=0.007) and felodipine6.418(Log rank p=0.041). Comparing with valsartan the multi-adjusted HR and95%CI were (0.573-0.865, Log rank p=0.001),0.767(0.631-0.932, p=0.008)和0.748(0.614-0.911, Log rank p=0.040). Relative reducing rate were29.6%,23.3%and25.2%. The results were confirmed by propensity score matching. The subgroup Cox analysis also showed the consistent protective effects of valsartan on diabetes incidence.Conclusion:Our study thus corroborates the finding from randomized clinical trials that diabetes risk is lower for hypertensive patients who received valsartan versus CCB and extends this finding to a’real-world’ setting in Chinese hypertensive patients.