The Study On Influencing Factors And Risk Assessment Of Hypertension Among The Elderly In Nanjing

Objective:(1)This study was carried out to investigate and understand the health status and prevalence of hypertension and prehypertension among elderly residents in a community in Nanjing,and to analyze the influencing factors of hypertension and prehypertension in the elderly.(2)The risk assessment model of hypertension was constructed according to the risk factors related.It provides a basis for primary prevention and health management of hypertension for the elderly residents in community.Methods:(1)In this study,cluster sampling method was conducted for the elderly residents in a community hospital in Nanjing from March 2016 to December 2016.Elderly residents over 60 were investigated by questionnaire survey and physical examination.After investigation,data were entered via Epi Data3.2.Statistical methods included descriptive analysis andX~2test were analyzed by SPSS22.0,and a multivariate Logistic regression model was established to analyze the influencing factors of hypertension and prehypertension in elderly population in Nanjing.(2)Based on the baseline data of the questionnaire survey on elderly population in the community in 2016,a modeling group(70%)and a verification group(30%)were established according to the random sampling method.Through the analysis of risk factors of hypertension according to data of modeling group,the hypertension risk assessment models were established by Harvard cancer index and risk factor score method,and verified the model by data of verification group.The area under the ROC curve(AUC)was used as the evaluation index to evaluate the prediction effectiveness of the model,and Hosmer-Lemeshow(H-L)was used as index to test the fitting degree of the evaluation model.Result:(1)In this survey,2492 eligible elderly people in the community were included,with an average age of 72.01±6.73 years old,including 1,255 males(50.36%)and 1,237 females(49.64%).The majority of them were aged between 60 and 69 years old(41.33%).There were 1029 patients with prehypertension(41.29%)and 767 patients with hypertension(30.78%).The prevalence of hypertension in elderly male residents(33.15%)was higher than that in women(28.38%),and the prevalence of prehypertension in men(41.67%)was higher than that in women(40.91%).With the age increased,the incidence of hypertension and prehypertension increased in men and women.The prevalence of hypertension and prehypertension was high in the elderly with lower education level.The prevalence of hypertension and prehypertension was higher among the widowed.The prevalence of hypertension and prehypertension in family history of hypertension was higher than that in non-family history of hypertension.There were differences in hypertension and prehypertension among people with different BMI and waist-to-hip ratio.The difference of hypertension and prehypertension was statistically significant among the elderly residents who had different habits of smoking,drinking,pickled food and moderate physical activity(P<0.05).The results of multivariate analysis showed that male,age≥70-year-old,primary school education or below,widowed,family history of hypertension,overweight and obesity,central obesity,smoking,drinking,regular consumption of pickled food and lack of moderate physical activity were positively correlated to hypertension(P<0.05).Male,age≥70-year-old,primary school education or below,widowed,family history of hypertension,overweight and obesity,central obesity,smoking and lack of moderate physical activity were positively correlated to prehypertension(P<0.05),and abstinence was negatively correlated to prehypertension(P<0.05).(2)Binary Logistic regression showed that gender,age,education,marital status,family history of hypertension,obesity and overweight,central obesity,smoking,drinking,pickled food,moderate physical activity are risk factors for hypertension.Harvard cancer index model and risk factor score method based risk factors were established.Using Harvard cancer index model,among 548 patients in the modeling group,58(10.58%)was rated as"higher"risk,312(56.93%)as"high"risk,69(12.59%)as"moderate"risk,61(11.13%)as"low"risk and 48(8.76%)as"lower"risk.Among the 219 patients in the verification group,24(10.96%)were rated as"higher"risk,123(56.16%)as"high"risk,21(9.59%)as"moderate"risk,29(13.24%)as"low"risk and 22(10.05%)as"lower"risk.When the optimal cut-off values was 1.2,the Harvard cancer index model had the largest Yordan index in modeling and validation group(0.244 and 0.275,respectively),and the area under the ROC curve(AUC)was 0.686(95%CI=0.659～0.712)and 0.691(95%CI=0.649～0.733),respectively,with sensitivity of 57.5%and 55.7%,respectively,and specificity of 66.9%and 71.8%,respectively.When the optimal cut-off values were 25 and 24,the risk factor score model had the largest Yordan index in modeling and validation group(0.280 and 0.341,respectively),and AUC were 0.715(95%CI=0.690～0.740)and 0.738(95%CI=0.700～0.776),with sensitivity of 72.3%and 78.1%respectively,and specificity of 55.7%and 56.0%respectively.The model had good prediction ability.In this study,the H-L test of the risk factor score model in modeling group wasX~2=3.371,P=0.909;and that in validation group,wasX~2=7.974,P=0.436,indicating a good fitting degree of the model.Conclusion:(1)In this survey,the prevalence of hypertension and prehypertension among the elderly residents in Nanjing community are relatively high.Therefore,health intervention should be carried out according to relevant risk factors.Moreover,health publicity and education should be strengthened to promote the elderly residents to change their bad living habits and prevent the occurrence of hypertension and prehypertension.(2)The hypertension risk assessment model based on the Harvard cancer index method and risk factor scoring method can assess the risk of hypertension,identify high-risk groups,and conduct targeted health interventions for high-risk groups to improve the quality of life,which is conducive to the rational allocation of health resources and improve medical security system.The predictive power of the risk factor score model and the ability to identify hypertensive patients were higher than that of the Harvard cancer index model,and the ability of the Harvard cancer index model to identify non-hypertensive patients was higher than that of the risk factor score model.