Prognostic Role Of Clinic And Ambulatory Blood-Pressure Measurements:10-Year Mortality In Elderly

BACKGROUNDEvidence for the influence of ambulatory blood pressure on prognosis derives mainly from population-based studies and a few relatively clinical investigations in young adults.Moreover,Prognosis of masked and white coat uncontrolled hypertension(MUCH and WCUH,respectively)detected by ambulatory blood pressure(BP)monitoring is incompletely clear in elderly treated hypertensive patients.AIMS1.This study examined the associations of blood pressure measured in the clinic(clinic blood pressure)and 24-hour ambulatory blood pressure with all-cause mortality in a cohort of old patients in hospital.2.To select the blood pressure parameters with the most predictive value to build a death prediction model.3.We evaluate prognosis of MUCH and WCUH identified by ambulatory BP monitoring in elderly treated hypertensive patients.METHODS1.To examine the associations of blood pressure measured in the CBP and 24 hABPM with all-cause mortality in a cohort of old patients in hospital and select the blood pressure parameters with the most predictive value to build a death prediction model,data on a total of 795 people(all aged more than 65 years)underwent clinic and ambulatory blood-pressure measurements in Guangzhou military region were collected between2003 and 2008.Analyses were conducted with Cox regression models,adjusted for clinic and 24-hABPM and for confounders.The nomogram was used to build the death prediction model.The discrimination of the model was evaluated according to the time-dependent AUC fitting chart and C-index,the calibration was evaluated according to Calibration curve,the clinical utility was evaluated by clinical Decision Curve Analysis(DCA),and the model was internally verified by resampling(bootstrap)1000 times.2.To evaluate prognosis of MUCH and WCUH identified by ambulatory BP monitoring in elderly treated hypertensive patients.611 primary hypertensives from the population with complete blood pressure data in part 1 were selected in the study.Clinic and 24-hour ambulatory blood-pressure data were examined in the following categories:Controlled hypertension(CH)was defined as clinic BP <140/90 mmHg and 24-hour BP <130/80 mmHg,MUCH as clinic BP <140/90 mmHg and 24-hour BP ≥130 and/or ≥80 mmHg WCUH as clinic BP ≥140 and/or≥90 mmHg and 24-hour BP < 130/80 mmHg and sustained uncontrolled hypertension(SUCH)as clinic BP ≥140and/or ≥90 mmHg and 24-hour BP ≥130 and/or ≥80 mmHg.Analyses were conducted with Cox regression models,adjusted for clinic and 24-hour ambulatory blood pressures and for confounders.RESULTS1.During the follow-up of 10 years,241 patients died from any cause.In a model that included both 24-hour and clinic measurements,24-hour systolic pressure was more strongly associated with all-cause mortality(HR 1.13 per 1-SD increase in pressure;95% CI,1.02-1.33,P=0.019,after adjustment for clinic blood pressure)than the clinic systolic pressure(HR 1.06;95%CI 1.01-1.26,P=0.046,after adjustment for 24-hour blood pressure).Corresponding hazard ratios per 1-SD increase in pressure were 1.18(95% CI 1.03-1.51,P=0.02,after adjustment for clinic and daytime blood pressures)for nighttime ambulatory systolic pressure and 0.97(95% CI 0.75-1.26,P=0.85,after adjustment for clinic and nighttime blood pressures)for daytime ambulatory systolic pressure.2.The nomogram was constructed with nocturnal ambulatory blood pressure and various cardiovascular risk factors.The time-dependent AUC curve fitting increased steadily from 0.782 to 0.796 and C index0.7311(95%CI 0.6974-0.7647)with good discrimination.Calibrationevaluation: when the incidence rate of observed death events is 10%-30%,the prediction is consistent with the actual risk value;The DCA of nighttime blood pressure prediction model has obvious net benefits.AUC of the complete model after 1000 resampling is 0.777,which is not significantly different from that of the constructed model and still has good prediction value.3.Among the treated elderly hypertensive population,MUCH was identified in 142 patients(20.79% of all the population)and WCUH in230 patients(9.66% of all the population).During the follow-up of 10 years,192 events occurred.The risk of MUCH increased by 64%(OR1.64,95%CI 1.05-2.56,P=0.0290)for every unit of age.Compared with the control group,the risk ratio of MUCH was1.04(95%CI 1.02-1.06,P<0.0001)for every unit of nighttime SBP.With the increase of nighttime SBP,the risk of WUCH decreased by 4%(OR 0.96,95%CI 0.93-0.99,P=0.0108).4.In stratified analysis of the effects of various risk factors on hypertension phenotype and all-cause death,it was found that there was significant interaction between statins administration history(P interaction =0.0234).After adjustment for various covariates,patients with MUCH(HR 1.56,95%CI 1.04-2.35,P=0.032)and WCUH has a tendency of poor prognosis.CONCLUSIONS1.Ambulatory blood-pressure measurements were a stronger predictor of all-cause mortality than clinic blood-pressure measurements.Especially,the nighttime blood pressures were the most significant prognostic marker of death risk.2.The nomogram prediction model based on nocturnal ambulatory blood pressure,has good discrimination,calibration and clinical effectiveness,and has good prediction value.3.The prevalence of masked uncontrolled hypertension increases with age and nighttime SBP.Nighttime SBP may not be a risk factor for white-coat uncontrolled hypertension.4.Among the elderly treated hypertensive patients,masked uncontrolled hypertension is more significantly related to all-cause mortality than controlled hypertension,increased risk of death by 56%.