Correlation between different body mass indexes and incidence of digestive carcinoma: a multicentre retrospective study (A report of 95 177 cases) / 中华消化外科杂志

ABSTRACT

Objective To explore the correlation between different body mass indexes and incidence of digestive carcinoma.Methods The retrospective cohort study was conducted.The data of 95 177 participants (75 909 males and 19 268 females) aged (51± 12)years with the range of 18-98 years who participated health examination at the Kailuan General Hospital,Kailuan Linxi Hospital,Kailuan Zhaogezhuang Hospital,Kailuan Tangjiazhuang Hospital,Kailuan Fan' gezhuang Hospital,Kailuan Jinggezhuang Hospital,Kailuan Lyujiatuo Hospital,Kailuan Linnancang Hospital,Kailuan Qianjiaying Hospital,Kailuan Majiagou Hospital and Kailuan Branch Hospital from July 2006 to December 2015 were collected.According to definition of body mass indexes from Chinese guideline for prevention and control of adult overweight and obesity,all the 95 177 participants were allocated into the 3 groups,including 37 660 with BMI＜24 kg/m2 in the normal BMI group,39 793 with with 24 kg/m2 ≤BMI＜ 28 kg/m2 in the overweight group and 17 724 with BMI≥28 kg/m2 in the obesity group.All participants received the same-order health examinations by the fixed team of doctors in 2006,2008,2010,2012 and 2014 at the same place.Epidemiological investigation,anthropometric parameters and biochemical indicators were collected.Observation indicators(1) comparisons of clinical characteristics among the 3 groups;(2) incidence of digestive carcinoma in the participants;(3) risk factors analysis affecting new-onset digestive carcinoma;(4) comparisons of the fitting degree of BMI on new-onset digestive carcinoma model;(5) stratified analysis of risk factors affecting new-onset digestive carcinoma at different locations.Measurement data with normal distribution were represented as Mean±SD,and comparisons among groups were analyzed using the one-way ANOVA.Measurement data with skewed distribution were described as M (range),and comparisons among groups were analyzed using the Kruskal-Wallis test.Count data were described as case number and percentage,and comparisons among groups were analyzed using the chi-square test.The cumulative incidence was calculated by the Kaplan-Meier method,and comparisons of incidences among groups were done by the Log-rank test.The incidences of digestive carcinomain patients with different BMI were calculated by person-year incidence (incidence density).The hazard ratio (HR) and 95％ confidence interval (CI) of different BMI (continuous variable and classification variable) on new-onset digestive carcinoma were estimated by the COX proportional hazards regression models.Restrictive cubic spline regression was used to calculate the dose-response relation between the continuous variable and the risks of digestive carcinoma.The fitting degree of BMI on new-onset digestive carcinoma model was calculated by the likelihood ratio test and akaike information criterion (AIC).Results (1) Comparisons of clinical characteristics among the 3 groupsage,sex (male),systolic pressure,diastolic pressure,waistline,total cholesterol (TC),triglyceride (TG),fasting plasma glucose (FPG),C reactive protein,cases with smoking,drinking,physical exercise,positive HBsAg,high salt intake,malignant tumor in immediate family were (51± 13)yeas,28 607,(125±20) mmHg (1 mmHg=0.133 kPa),(80± 11) mmHg,(81±9) cm,(4.9± 1.1) mmol/L,1.05 mmol/L(range,0.75-1.49 mmol/L),(5.3±1.6) mmol/L,0.58 mmol/L (range,0.20-1.60 mmol/L),11 962,6 845,5 676,711,.3 640,1 298 in the normal BMI group and (52±12)years,32 928,(133±21) mmHg,(85±11) mmHg,(89±8)cm,(5.0±1.2) mmol/L,1.39 mmol/L (range,0.99-2.08 mmol/L),(5.6± 1.7)mmol/L,0.84 mmol/L (range,0.33-2.07 mmol/L),12 364,7 413,6 322,839,4 401,1 463 in the overweight group and (51 ± 12) years,14 374,(139 ± 21) mmHg,(88 ± 12) mmHg,(96 ± 9) cm,(5.1 ± 1.2) mmol/L,1.67 mmol/L (range,1.18-2.51 mmol/L),(5.7± 1.8) mmol/L,1.22 mmol/L (range,0.53-2.82 mmol/L),5 092,2 818,2 847,355,2 235,704 in the obesity group,showing statistically significant differences among groups (F=90.60,x2 =576.34,F=2 768.38,3 570.80,22 319.30,256.99,x2 =9 108.21,F=507.11,x2 =3 219.47,52.78,64.38,13.36,0.76,130.39,9.74,P＜0.05).(2) Incidence of digestive carcinoma in the participants:all the 95 177 participants were followed up for 845 085 person-year,1 215 were diagnosed as new-onset digestive carcinoma,with a total person-year incidence of 1.44 thousand person / year.Of 1 215 patients,413 had colorectal-anal cancer,306 had liver cancer,234 had gastric cancer,113 had esophageal cancer,91 had the pancreatic cancer,36 had gallbladder carcinoma or cholangiocarcinoma,25 had intestinal cancer.Three patients had intestinal cancer complicated with colorectal-anal cancer.The person-year incidence of digestive carcinoma was 1.46 thousand person / year,1.37 thousand person / year and 1.53 thousand person / year in the normal BMI group,overweight group and obesity group,respectively.The cumulative incidences of digestive carcinoma in the normal BMI,overweight,obesity group were respectively 11.8‰,10.1‰ and 12.1‰,showing a statistically significant difference among 3 groups (x2=6.13,P＜0.05).There was no statistically significant difference between the normal BMI group and obesity group (x2 =1.07,P＞0.05),and statistically significant differences between the overweight group and normal BMI group and obesity group,respectively (x2=3.90,4.10,P ＜ 0.05).(3) Risk factors analysis affecting new-onset digestive carcinoma.Results of COX proportional hazards regression models showed that continuous BMI was not related factor affecting new-onset digestive carcinoma after adjustment of age,gender,systolic pressure,TC,TG,FPG,smoking,drinking,physical exercise,positive HBsAg,high salt intake,malignant tumor in immediate family (HR=0.99,95％CI0.98-1.01,P＞0.05).After adding BMI as classification variable in the COX model,risk of new-onset digestive carcinoma in the overweight group was reduced compared with normal BMI group (HR =0.88,0.88,95％CI0.78-1.01,0.77-0.98,P＜0.05) and risk of new-onset digestive carcinoma in the obesity group was not affected (HR=1.03,1.04,95％CI0.88-1.20,0.89-1.22,P＞0.05).Results of restrictive cubic spline regression showed a "U" shaped relationship between BMI and incidence risk of digestive carcinoma and the lowest incidence of digestive carcinoma in patients with BMI as 25-27 kg/m2.(4) Comparisons of the fitting degree of BMI on new-onset digestive carcinoma modelmultivariate model was constructed after adding risk factors of age,gender,systolic pressure,TC,TG,FPG,smoking,drinking,physical exercise,positive HBsAg,high salt intake,malignant tumor in immediate family,and-2Log L and AIC were 27 175.05 and 27 203.05 for the multivariate model.Then BMI variable was added into the multivariate model,and the-2Log L and AIC of the multivariate model+BMI model were 27 169.53 and 27 201.53,respectively,with a statistically significant difference compared with normal BMI group (x2 =5.52,P＜0.05).(5) Stratified analysis of risk factors affecting new-onset digestive carcinoma at different locations.Results of COX proportional hazards regression models showed risks of new-onset digestive carcinoma in the overweight and obesity groups were reduced compared with normal BMI group (HR=0.57,0.42,95％CI0.38-0.84,0.23-0.79,P＜0.05) in the esophageal cancer model.Risks of new-onset digestive carcinoma in the overweight group were reduced compared with normal BMI group (HR=0.72,95％CI0.55-0.93,P＜0.05) and risk of new-onset digestive carcinoma in the obesity group was not affected (HR=1.10,95％CI0.82-1.47,P＞0.05) in the liver cancer model.Conclusions Participants in the overweight group have the lowest incidence of digestive carcinoma,especially in the esophageal cancer and liver cancer model.Incidence of digestive carcinoma is the lowest with BMI as 25-27 kg/m2.