不同体质量人群中碳酸饮料与非酒精性脂肪肝的关系

RELATIONSHIP BETWEEN CARBONATED BEVERAGE CONSUMPTION AND NONALCOHOLIC FATTY LIVER DISEASE AMONG ADULTS WITH DIFFERENT BODY MASS INDEX LEVELS

目的在不同身体质量水平的成年人群中评价碳酸饮料饮用频度和非酒精性脂肪肝(NAFLD)之间的关联。方法对在天津医科大学总医院健康管理中心进行体检的20,020名成年人进行横断面调查。使用经信度效度检验过的食物频率问卷评估饮食摄入量,用腹部超声检查和谷丙转氨酶诊断NAFLD。按照体质指数标准将人群分为体质量正常组、超重组和肥胖组,采用logistics回归分析不同身体质量人群中碳酸饮料饮用频率和NAFLD患病率之间的关系。结果经调整混杂因素后,碳酸饮料饮用频率从低到高组中,体重正常组、超重组和肥胖组人群中NAFLD患者与非患者不同程度碳酸盐饮料暴露的比值比(95%置信区间)分别为1.00、0.93(0.68,1.25)、1.51(1.16,1.96)(趋势性P<0.01);1.00、1.16(0.99,1.37)、1.15(0.99,1.34)(趋势性P=0.04);和1.00、1.33(1.03,1.74)、1.31(1.04,1.66)(趋势性P=0.02)。结论经调整混杂因素后,在体质量正常组和肥胖组中人群中,碳酸饮料饮用频度和NAFLD患病率相关。今后的研究中,需要更多的队列研究来确认碳酸饮料和NAFLD之间的因果关系。

Objective To assess the association between carbonated beverage consumption and nonalcoholic fatty liver disease （NAFLD） among adults with different body mass index （BMI） levels. Methods A cross-sectional study was conducted in 20,020 adults who received physical examination in Health Management Centre in Tianjin Medical University General Hospital. Dietary intakes were assessed using a validated self-administered food frequency questionnaire. NAFLD was diagnosed by liver ultrasonography and ALT levels. Subjects were classified into normal weight group, overweight group and obese group based on their BMI. Multivariate logistic regression analyses were used to examine the relationship between carbonated beverage consumption and the prevalence of NAFLD. Results After adjusting for the confounding variables across the carbonated beverage frequency （low, medium and high consumption frequency）, the odds ratios （95% confidence interval） of NAFLD were 1.00, 0.93 （0.68, 1.25）, 1.51 （1.16, 1.96） （P for trend 〈0.01） in normal weight group; 1.00, 1.16 （0.99, 1.37）, 1.15 （0.99, 1.34） （P for trend = 0.04） in overweight group; and 1.00, 1.33 （1.03, 1.74）, 1.31 （1.04, 1.66） （P for trend = 0.02） in obese group. Conclusion The study demonstrates significantly associations between carbonated beverage consumption and NAFLD in normal weight and obese population groups. Randomized control studies are needed to confirm the causality.