A Cohort Study On Body Mass Index And The Risk Of All-Cause Mortality In Adults With And Without Hypertension

In recent years,more and more studies have found that compare with normal weight group,those who was overweight/obese had a lower risk of mortality when obesity was measured by body mass index?BMI?.Overweight/obesity has a protective effect on mortality which called"obesity paradox".Previous studies focused on the phenomenon of"obesity paradox"were mostly restricted to hypertensive populations.Few studies have investigated whether baseline blood pressure status?hypertensive vs non-hypertensive?affects the relationship between BMI and all-cause mortality.The study is based on the established rural Chinese cohort population to explore this phenomenon.ObjectiveTo investigate the association between BMI and all-cause mortality among subjects with and without hypertension in rural Chinese cohort population.MethodsIn this study,a prospective cohort study design was used to conduct a baseline survey of 20 194 rural community population aged above 18 years old using cluster sampling method from 2007 to 2008.Residents was conducted baseline surveys including questionnaires,physical examinations,fasting blood glucose testing,and lipid profile testing.Follow-up surveys were completed in 2013-2014.The response rate was 85.5%and the average follow-up time was 6.0 years.Excluding 2 cases with height/weight information loss and 1 case who unmeasured systolic blood pressure?SBP?/diastolic blood pressure?DBP?during the baseline period,a total of 17 262participants were included in the analysis.We estimated hazard ratios and 95%CI for the association between different levels of BMI stratification and all-cause mortality among subjects with and without hypertension by Cox regression,taking attained age as the underlying time variable,and adjusting for potential confounders.Restricted cubic spline models were performed to model the dose-response association.In addition,sensitivity analyses were used to assess the robustness of the main results:?1?excluded baseline smokers;?2?excluded baseline smokers and those who died in the first two years of follow-up;?3?excluded baseline smokers,deaths within 2 years of follow-up,and those who had cardiovascular diseases?stroke,heart failure and myocardial infarction?and cancer.Results1.A total of 17 262 subjects with 6 814 males and 10 448 females were included in this follow-up program?101 581.7 person-years and median 6 years?.Among these,11 691 were non-hypertensive and 5571 were hypertensive.During follow-up period,the death density?/1000 person-years?was 7.137 and 19.265 for non-hypertensive and hypertensive patients,respectively.In non-hypertensive population,as the weight increases,the proportion of drinkers,SBP,DBP,Total cholesterol?TC?,Triglyceride?TG?,Low-density lipoprotein cholesterol?LDL-C?and FPG all increase but age and high-density lipoprotein cholesterol?HDL-C?decreased?P<0.01 for trend test?.In hypertensive patients,with the increase of weight,the values of DBP,TG,and FPG increased,but age,the proportion of males,smokers,and HDL-C values all decreased?P<0.01 for trend test?.2.The Kaplan-Meier survival curve showed that the risk of all-cause mortality was highest in underweight group?BMI<18.5 kg/m²?while the risk of all-cause mortality in overweight?24.0 kg/m²?BMI<28.0 kg/m²?and obese groups?BMI?28.0 kg/m²?was lower than normal weight group?18.5 kg/m²?BMI<24.0 kg/m²?for hypertensiveandnon-hypertensivepopulations.Aftercontrollingpossible confounding factors,the multivariable-adjusted HRs for all-cause mortality associated with BMI levels of non-hypertensive population?<18.5,18.5-24[reference group],24-28,and?28 kg/m²?were 1.52?95%confidence interval[CI],1.11-2.08?,1.00,0.84?95%CI,0.67-1.05?,and 1.07?95%CI,0.75-1.51?,respectively.Compared with the normal weight group,participants with hypertension who was overweight or obese significantly reduced the risk of all-cause mortality,corresponding HR values?95%CI?were 0.79?0.66-0.95?and 0.71?0.55-0.90?,respectively.After stratification by gender and age,results in the male subgroup and age<60 years subgroup were similar to those in the general population.3.After adjusting for possible confounding factors,restricted cubic spline models showed that the dose-response relationship between BMI and all-cause mortality risk in non-hypertensive patients was“U”-shaped(p_{non-linearity}<0.001)and the lowest mortality risk was at BMI of 23.0-24.0 kg/m²,the group with a BMI of<22.0 kg/m² could significantly increase the risk of mortality while the result was not statistically significant in the group with a BMI of>24.0 kg/m².The dose-response curve between BMI and all-cause mortality risk presented nonlinear"L"-shaped(p_{non-linearity}<0.001)in the hypertensive group,and the risk of all-cause mortality gradually decreased with the BMI increased.4.After stratified by gender?male;female?,the results did not change much in non-hypertensive populations while the results become statistically insignificant in hypertensive female subgroups;After stratified by age?<60 years and?60 years?,the results did not change substantially in both hypertensive and non-hypertensive populations After adjusting for potential confounding factors,no significant effect modification according to age or gender was observed in the present study?P>0.05for all interactions?5.After excluding baseline smokers,deaths within 2 years of follow-up,and those who had cardiovascular diseases?stroke,heart failure and myocardial infarction?and cancer,the results of multivariable cox regression analysis and restricted cubic spline model analysis did not change obviously in non-hypertensive population.However,among hypertensive patients,the cox regression analysis and restricted cubic spline model analysis both showed that the risk of all-cause mortality in the overweight and obese groups became statistically insignificant compared with the normal weight group.Conclusions1.Among non-hypertensive patients,underweight significantly increased the risk of all-cause mortality,and"obesity paradox"phenomenon was not found.The results did not change after excluding baseline smokers,deaths within 2 years of follow-up,and those who had cardiovascular diseases and cancer.2.Among hypertensive patients,underweight significantly increased the risk of all-cause mortality while overweight/obesity significantly reduces the risk.Overweight or obesity may have a protective effect on all-cause mortality among hypertensive population,which may support the"obesity paradox"phenomenon.However,after excluding baseline smokers,deaths within 2 years of follow-up,and those who had cardiovascular diseases and cancer,the relationship between BMI and all-cause mortality risk became statistically insignificant.