Epidemiological Evaluation Of The Incidence Rate Of Hypertension Based On The Mean Of 3 Measurements Taken At One Epidemiological Investigation

Objective:The epidemiological estimation of the incidence and prevalence of hypertension are mostly based on the mean of 3measurements taken at one epidemiological investigation.Although a large number of clinical studies have shown that if we use the blood pressure readings measured at one visit to make the diagnosis of hypertension,the false positive rate of hypertension will increase,that is to overestimate the prevalence of hypertension.However,most epidemiological study still estimate hypertension rate according to the value of blood pressure measured at a single visit due to the limited time and funds.This leads to the risk for overestimation of the prevalence and incidence of hypertension in epidemiological surveys,while the degree of overestimation has not been reported so far.This study aims to estimate the diagnostic significance of blood pressure measured at one visit by comparing the variation of blood pressure at two consecutive visits to provide basis of accurate estimation of hypertension rate in epidemiological investigation.Methods:Data were collected from 1989 to 2011 by the China Health and Nutrition Survey（CHNS）and focused on adults aged 18 or older.This research is divided into two stages:Firstly,we followed up the participants who entered the cohort from 1989 to 2006 with normal blood pressure and calculated the incidence rates（IRs）of one visit hypertension and two visits hypertension according to the number of person-years of follow-up;all the incidence rates of hypertension were adjusted by direct method according to the 2010 census of the Chinese adult population using the corresponding age groups.Secondly,we followed up the incident hypertension based on one visit from 1991 to 2009 and estimated the positive predictive values of one visit hypertension and 95%confidence interval according to their BP values at next survey;analyses were stratified by gender,age,BMI,type and degree of elevations of blood pressure,using Cox proportional hazards regression model to explore the influencing factors of PPV.Mann-Whitney test was used to compare the value of continuous variables between subgroups,and chi-square was used to compare the value of categorical variables between subgroups,the trend test was conducted by linear-by-linear trend test.All statistical analyses were performed by using SPSS software for Windows version 21.0.The statistical significance was at a two-tailed P<0.05.Results:1.Normotensive participants（n=14057）at baseline were included.Until 2009,the follow-up time was 113,848 person-years（average was8.10 years）,the adjusted IR of one visit hypertension was 4.80（95%CI,4.68-4.93）per 100 person-years,the adjusted IR of two visits hypertension was 1.82（95%CI,1.74-1.90）per 100 person-years.The adjusted IR of one visit hypertension and two visits hypertension all increased with age and BMI(P _trend<0.001).Relative risk of IR of one visit hypertension to two visits hypertension was 2.64（95%CI,2.50-2.77）.2.There were 2987 participants with new-onset ISH,IDH and SDH based on one visit and with final follow-up outcomes as study subjects.The positive predictive value of the new-onset hypertension based on one visit was 41.0%（95%CI,39.3-42.8%）for hypertension.The positive predictive values of men and women were 42.5%（95%CI,40.1-45.0%）and 39.3%（95%CI,36.7-41.8%）（P=0.071）,respectively.3.Overall,the positive predictive value of the new-onset hypertension based on one visit increased significantly with age,from26.4%in those aged 18-39 years to 50.1%in those aged≥60(P _trend<0.001).The PPV of the new-onset hypertension based on one visit increased significantly with BMI,from 31.5%in those with BMI<18.5kg/m² to 47.6%in those with BMI≥28.0 kg/m²(P _trend<0.001).4.When stratifying for type of BP,the PPVs of the new-onset hypertension based on one visit in ISH,IDH and SDH were 42.3%（95%CI,38.6-45.9%）,33.8%（95%CI,31.4-36.2%）,54.1%（95%CI,50.5-57.6%）（P<0.001）,respectively.Increased level of elevated BP was associated with an increased PPV(P _trend<0.001).5.Exploring the factors associated with the PPV of the new-onset hypertension based on one visit by Cox proportional hazards regression model,age,BMI and level of blood pressure measured at first visit were associated with PPV.Conclusions:1.The incidence of hypertension based on one visit is 3.08 times of that based on two visits,and the epidemic of hypertension was overestimated to a certain degree by estimating the incidence of hypertension relied on measurement of BP taken at one visit.2.The positive predictive value of the new-onset hypertension based on one visit was as low as 41.0%,and was not associated with sex,smoking,alcohol drinking,region and period-effect.3.The positive predictive value of the new-onset hypertension based on one visit was associated with age,BMI and the increased level of blood pressure.