Effects Of Erythrocyte Parameters On Incident Hypertension And Prediction Model Of Incident Hypertension

Hypertension is the most important and common risk factor for cardiovascular disease.The incidence and prevalence of hypertension in China are severe.Identifying risk factors of incident hypertension,establishing an incident hypertension prediction model and assessing the risk of incident hypertension can help to find people who have high-risk incident hypertension and help to delay or even prevent incident hypertension onset.At present,many countries and regions have established hypertension prediction models,these models usually included age,systolic blood pressure,diastolic blood pressure,body mass index,smoke,drinking and hypertension history of family members.Pre'vious models were lack of new predictive factors,so predictive ability of models may be limited.Many studies showed that erythrocyte parameters may be predictors of incident hypertension.Therefore,we used the hospital health check-up data to build prospective cohort to study effect of erythrocyte parameters on incident hypertension,and then determined whether erythrocyte parameters could be as a predictor of incident hypertension.If so,prediction model of incident hypertension could be established in consideration of erythrocyte parameters.Methods:Based on Shandong Multi-Center Health Management Large Database we included subjects who took health check-up between 2005 and 2010 to compose the prospective cohort.Participants were eligible for inclusion if they possessed at least three year's records.We excluded participants at the first check-up who had hypertension with anti-hypertensive medication or systolic blood pressure? 140mmHg or diastolic blood pressure?90mmHg or cardiovascular disease or stroke,were younger than 18 years and eventually 12497 people(7537 men and 4960 women)entered the cohort.After adjusting other factors,we used Cox regression to assess effects of erythrocyte parameters on incident hypertension.Considering the erythrocyte parameters,we constructed incident hypertension model using multivariable backward cox analysis sex-specifically.Results:During 38958 person-years of follow-up,2785 participants(men,2021;women,764)developed hypertension.The incidence density of hypertension was 71.48 per 1000 person-year.Descriptive analysis between baseline erythrocyte parameters and study variables(erythrocyte parameters were divided into four categories according to the quartile,Q1,Q2,Q3,Q4):Regardless of men and women,most of the baseline variables increased with erythrocyte parameters.But statistically significant baseline variables were slightly different in three erythrocyte parameters.Through Cochran-Armitage trend test,only hematocrit was statically significant with the incidence of hypertension during follow-up in men(Z=-3.1628,P<0.0001).While three erythrocyte parameters all have trends with the incidence of hypertension during follow-up in women(red blood cell counts,Z=-4.2950,P<0.0001;Hemoglobin,Z=-5.8120,P<0.0001;Hematocrit,Z=-6.5504,P<0.0001).Analysis of Cox regression model on erythrocyte parameters and the risk of incident hypertension:For men,after adjusting age,the risk ratio of Q4(Q1,reference group),model trend tests and risk ratio of each standard deviation increase on erythrocyte parameters were statistically significant.Adjusting more variables,those results had no statistically significance.For women,after adjusting different covariates,the risk ratio of Q3 and Q4(Qi,reference group),model trend tests and risk ratio of each standard deviation increase on erythrocyte parameters were all statistically significant.After adjusting age,current smoking,current drinking and regular exercise,body mass index,systolic blood pressure,fasting blood glucose,high density lipoprotein cholesterol,the risk ratio of Q2,Q3,Q4 were 1.140,1.285 and 1.240 on red blood cell,1.069,1.309 and 1.311 on hemoglobin,1.019,1.263 and 1.234 on hematocrit.In multivariable backward cox analyses,age,body mass index,systolic blood pressure,diastolic blood pressure,fasting blood glucose and current drinking were retained for both men and women,while gamma-glutamyl transferase only for men,hematocrit and current smoking only for women.The interaction of diastolic blood pressure and body mass index by age were also included in incident hypertension for men.The area under receiver operating characteristic curve was 0.761(95%Cl,0.752-0.771)for men and 0.750(95%CI,0.738-0.762)for women,even after 10-fold cross-validation,the AUC was 0.760(95%CI,0.751-0.770)for men and 0.747(95%CI?0.735?0.759)for women.The O/E is 0.9561 for men and 0.9707 for women.Conclusion:Higher erythrocyte parameters are associated with higher risk of incident hypertension,especially in women.The incident hypertension prediction model for women include hematocrit.Hematocrit is an independent predictor of hypertension in women.The discrimination and calibration of incident hypertension models sex-specifically are good,and they can be used to assess the risk of incident hypertension.