Association Between Body Mass Index And Mortality In Adults:a Prospective Study In China

Backgrounds and objectiveOver the past few decades, there has been a dramatic increase in the prevelance of obesity in many countries. Obesity is associated with multiple chronic diseases, including type2diabetes, hypertension, coronary heart disease, stroke, and several cancers.Substantial epidemiology studies commonly assess weight by using body mass index (BMI)(the weight in kilograms divided by the square of the height in metres), a form of weight adjusted for height. BMI is often categorized for purpose of analysis and presentation.Most of the studies have been conducted in populations of European origin, however, the relationship between BMI and the overall risk of death among Asians, who account for more than60%of the world population, remains unclear. The optimal weight range associated with a minimal risk of death in Asian populations remains controversial. To address these unresolved issues, we evaluated the relationship between BMI and the risk of death using data from the Electronic Health Records of Yinzhou District between2005and2014, a cohort involving more than100thousand participants.Material and methodsThe data were collected from the Electronic Health Records of five towns in Yizhou District between January2005and July2014. A total of111894residents aged18years and older were recruited, of which103976participants were available to be followed, including49528men and53223women, with a response rate of92.9%. All deaths that ocurred between study entry and July2014were included for analysis.To define BMI groups for the analysis, we used the BMI cutoff points of less than18.5for underweight, more than24.0for overweight and more than28.0for obesity. We then established9BMI levels (i.e.,<18.5.18.5-19.9,20.0-21.9.22.0-23.4,23.5-24.9,25.0-27.4、27.5-29.9、30.0-34.9, and>35.0). The BMI range of23.5to24.9was used as the reference.Cox proportional hazard regression analysis was used to estimate the hazard ratios (HR) of mortality for different BMI categories during a maximum follow-up of9years. Stratified analyses and sensitivity analyses were performed to minimize the influence of possible "reverse causation".All the statistical analyses were finished by SAS Version9.2(SAS Institute, Gary, NC, USA). The significance level was P<0.05.ResultsFor the102751participants included in our analysis,3297people (1683men and1614women) died during a mean follow-up period of7.07years. The all-cause mortality was453.77per100000person-years.Our analyses showed an increased risk of all-cause mortality for the lowest category (<18.5kg/m2) and highest category (>35.0kg/m2) of BMI among both men and women. After adjustment for age and sex, the lowest BMI category (<18.5kg/m2) and the highest BMI category (>35.0kg/m2) were strongly associated with increased mortality in this cohort compared with the mid-range group (BMI:23.5-24.9); the HRs of the lowest BMI category and the highest BMI category were in1.31(95%confidence interval:1.11-1.54) and3.71(95%confidence interval:2.22-6.21), respectively. After multi-factor adjustment, the lowest BMI category (<18.5kg/m2) and the highest BMI category (>35.0kg/m2) were also strongly associated with increased mortality compared with the mid-range group, with HRs of1.52(95%confidence interval:1.24-1.86) and3.18(95%confidence interval:1.50-6.74), respectively. We observed a U-shaped association between BMI and all-cause mortality.Similar U-shaped associations were observed in the following stratified analysis by sex, age (18-60or≥60years) and follow-up time (<7years or≥7years). The residents whose BMI between25.0and27.4in this cohort showed the lowest all-cancer mortality risk, while the highest BMI category (>35.0kg/m2) showed the highest all-cancer mortality risk.There was a similar association between BMI and mortality from cardiovascular disease, respiratory disease, and infectious disease. The U-shaped association existed even after excluding participants who were current smokers, or who died during the first year of follow-up, or the participants in the first three years of follow-up.ConclusionThere was a U-shaped or J-shaped curve association between body mass index (BMI) and mortality among Chinese adults in our study. The lowest risk of deaths was among adults who had a BMI of23.5-27.4. Underweight is a related factor of mortality in the Chinese population. Overweight was not associated with increased mortality, however, the level of related risk factors such as glucose and hypertension will increase. This study suggests the need for public health policies focusing on reducing under-nutrition and obesity.