Correlation between fasting blood glucose and hepatocarcinogenesis: a multicentre retrospective study (A report of 94 264 cases) / 中华消化外科杂志

ABSTRACT

Objective To explore the correlation between fasting blood glucose (FBG) and hepatocarcinogenesis.Methods The retrospective cohort study was conducted.The data of 94 264 participants who participated health examination at the Kailuan General Hospital of North China University of Science and Technology,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.There were 75 134 males and 19 130 females,aged (51±12)years,with a range of 18-98 years.All the subjects were allocated into 3 groups according to tertiles of FBG,including 31 083 with FBG ＜ 4.82 mmol/L in the T1 group,31 594 with 4.82 mmol/L≤ FBG ＜5.49 mmol/L in the T2 group and 31 587 with FBG ≥5.49 mmol/L in the T3 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) follow-up and incidence of liver cancer;(3) situations of non-liver cancer death;(4) risk factors analysis affecting new-onset liver cancer;(5) comparisons of the prognostic value of FBG on liver cancer model;(6) effects of FBG on new-onset liver cancer using competing risk model.Follow-up using physical examination was performed to detect new-onset liver cancer and survival up to December 31,2015.The start time of follow-up was the first health examination in 2016 and the terminal event was new-onset liver cancer,loss of follow-up and death.Measurement data with normal distribution were expressed 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 rank sum test.Count data were described as absolute number and percentage,and comparisons among groups were analyzed using the chi-square test.The cumulative incidence and mortality of new-onset liver cancer were calculated and incidence curve was drawn by the Kaplan-Meier method,and comparisons of incidences among groups were done by the Log-rank test.The incidence of liver cancer in patients with different levels of FBG was calculated by person-year incidence (incidence density).The hazard ratio (HR) and 95％ confidence interval (CI) of different levels of FBG (classification variable and continuous variable) on new-onset liver cancer were estimated by the COX proportional hazards regression models.Restrictive cubic spline regression was used to calculate the dose-response relation between the continuous FBG and the risks of new-onset liver cancer.The fitting degree of FBG on new-onset liver cancer model was calculated by the likelihood ratio test and akaike information criterion (AIC).The predictive power of different models was calculated using the C-statistics.The net effects of FBG on incidence of liver cancer were analyzed using cause-specific hazard function (CS) and sub-distribution hazard function (SD).Results (1) Comparisons of clinical characteristics among the 3 groupsgender (male),age,systolic pressure,diastolic pressure,waistline,body mass index (BMI),total cholesterol (TC),alanine aminotransferase (ALT),triglyceride (TG),cases with drinking,smoking,physical exercise,positive HBsAg and fatty liver were 23 567,(51±13)years,(128±21)mmHg (1 mmHg=0.133 kPa),(82±12)mmHg,(86± 10) cm,(24±3) kg/m2,(4.8± 1.2) mmol/L,17.12 U/L (range,12.21-24.01 U/L),1.18 mmol/L (range,0.82-1.75 mmol/L),5 080,9 423,4 779,724,7 591 in the T1 group,24 870,(50±12)years,(129±20)mmHg,(83±12)mmHg,(86±10)cm,(25±3)kg/m2,(4.9±l.1) mmol/L,18.31 U/L (range,13.01-24.31 U/L),1.23 mmol/L (range,0.88-1.83 mmol/L),5 448,9 397,4 570,619,9 009 in the T2 group and 26 697,(53±11)years,(135±22)mmHg,(86±12)mmHg,(89±10)cm,(26±3)kg/m2,(5.1± 1.2) mmol/L,19.00 U/L (range,13.79-26.61 U/L),1.44 mmol/L (range,1.00-2.21 mmol/L),6 354,10 292,5 369,608,13 397 in the T3 group,showing statistically significant differences among groups (x2 =761.68,F=417.84,1 010.71,747.64,702.73,1 075.06,703.83,x2=447.44,2 109.38,165.97,66.69,78.90,15.50,2 576.95,P＜0.05).(2) Follow-up and incidence of liver cancerall 94 264 participants were followed up for 817 475 person-year,with a total person-year incidence of 3.71/10 000 person-year,1.13/10 000 person-year in the female participants and 4.37/10 000 person-year in the male participants.The incidence density of liver cancer was 2.84/10 000 person-year,3.64/10 000 person-year,4.64/10 000 person-year in the T1,T2,T3 groups,respectively.The cumulative incidence was 2.76‰,3.90‰,4.90‰ in the T1,T2,T3 groups,respectively,showing statistically significant differences among groups (x2=11.95,P ＜ 0.05),showing no statistically significant difference between T1 and T2 groups (x2 =2.73,P＞0.05),showing statistically significant differences between T1 and T3 groups,between T2 and T3 groups (x2=11.56,4.10,P＜0.05).(3) Situations of non-liver cancer deathduring the follow-up,6 880 of 94 264 participants had of non-liver cancer related death,with a non-liver cancer death intensity of 84.16/10 000 person-year.The non-liver cancer death intensity was 79.19/10 000 person-year,68.17/10 000 person-year,105.32/10 000 person-year in the T1,T2,T3 groups.The accumulative mortality was 78.90‰,67.80‰,104.40‰ in the T1,T2,T3 groups,respectively,showing a statistically significant difference among groups (x2 =1 231.46,P ＜ 0.05),showing statistically significant differences between T1 and T2 groups,between T1 and T3 groups (x2 =5.29,4.36,P＜0.05),showing no statistically significant difference between T2 and T3 groups (x2 =0.09,P＞ 0.05).(4) Risk factors analysis affecting new-onset liver cancer.Results of COX proportional hazards regression model analysis showed that continuous FBG was a related factor affecting new-onset liver cancer after adjustment of gender,age,BMI,ALT,drinking,smoking,physical exercise,positive HBsAg,fatty liver,liver cirrhosis,malignant tumor in immediate family (HR =1.06,95％ CI1.01-1.12,P＜0.05).After ln transformation of FBG,ln FBG was a related factor affecting new-onset liver cancer (HR=1.81,95％ CI1.21-2.70,P＜0.05).Results of restrictive cubic spline regression showed that continous FBG and ln FBG were nonlinear correlated with incidence of liver cancer (RCS_ S1_x2 =7.21,4.36,P＜0.05).After adding FBG as classification variable in the COX model,risk of new-onset liver cancer in the T2 and T3 groups was increased compared with the T1 group (HR=1.45,1.67,95％ CI1.07-1.95,1.25-2.22,P ＜ 0.05).(5) Comparisons of the prognostic value of FBG on liver cancer modelmultivariate model was constructed after adding risk factors of gender,age,BMI,ALT,drinking,smoking,physical exercise,positive HBsAg,fatty liver,liver cirrhosis,malignant tumor in immediate family,and C-value,-2Log L and AIC were 0.79,6 313.30 and 6 345.30 for the multivariate model.Then FBG variable was added into the multivariate model,and the C-value,-2Log L and AIC of the multivariate model + FBG model were 0.80,6 300.48 and 6 336.48,respectively,showing statistically significant differences compared with the T1 group (x2 =12.82,P＜0.05).(6) Effects of FBG on new-onset liver cancer using competing risk model.Results of competing risk model showed that the risk of new-onset liver cancer in the T2 group was not affected compared with the T 1 group (HR =1.42,95％CI0.98-1.97,P＞0.05) and risk of new-onset liver cancer in the T3 group was increased compared with the T1 group with the SD model (HR=1.63,95％ CI1.16-2.26,P＜0.05),after adjustment of gender,age,BMI,ALT,drinking,smoking,physical exercise,positive HBsAg,fatty liver,liver cirrhosis,malignant tumor in immediate family.In the CS model,the risk of new-onset liver cancer in the T2 group was not affected compared with the T1 group (HR=1.43,95％ CI0.99-1.97,P＞0.05) and risk of new-onset liver cancer in the T3 group was increased compared with the T1 group (HR=1.65,95％ CI1.18-2.23,P＜ 0.05).Conclusions The elevated FBG is an independent risk factor for the incidence of liver cancer.After considering the competitive risk of death,the risk effect of high-level FBG on the liver cancer still exists.