Objective: To investigate the association between total cholesterol (TC) and primary liver cancer in Chinese males. Methods: Since May 2006, all the male workers, including the employees and the retirees in Kailuan Group were recruited in the Kailuan male dynamic cohort study. Information about demographics, medical history and TC levels was collected at the baseline interview, as well as information on newly-diagnosed primary liver cancer cases during the follow-up period. A total of 110 612 males were recruited in the cohort by 31 December 2015. TC levels were divided into four categories by quartile (<4.27, 4.27-4.90, 4.90-5.56 and ≥5.56 mmol/L), with the first quartile group serving as the referent category. Cox proportional hazards regression model was used to evaluate the association between TC levels and primary liver cancer risk. Results: By December 31, 2015, a follow-up of 861 711.45 person-years was made with a median follow-up period of 8.83 years. During the follow-up, 355 primary liver cancer cases were identified. Compared with the first quartile, the HR of incident primary liver cancer among participants with the second, third and highest quartile TC levels were 0.76 (95%CI: 0.58-1.01), 0.59 (95%CI: 0.43-0.79), and 0.36 (95%CI: 0.25-0.52), respectively after adjusting for age, educational level, income level, smoking status, drinking status, body mass index, and HBsAg status (Pfor trend<0.001). Subgroup analyses found that the association between TC levels and primary liver cancer was robust (all Pfor trend<0.05). The results didn't change significantly after exclusion of newly-diagnosed cases within the first 2 years, males with history of cirrhosis or subjects who took antihyperlipidemic drugs, participants with higher TC levels had a lower risk of primary liver cancer (all Pfor trend<0.05) and HR(95%CI) of incident primary liver cancer among participants with the highest quartile TC levels were 0.41 (0.28-0.61), 0.36 (0.25-0.53) and 0.38 (0.26-0.54), respectively. Conslusion: In this large prospective study, we found that baseline TC levels were inversely associated with primary liver cancer risk, and low TC level might increase the risk of primary liver cancer.
目的： 分析我国男性总胆固醇（TC）水平与原发性肝癌发病风险的关联。 方法： 于2006年5月以开滦集团在职及离退休男性职工为研究对象，基线调查时收集人群社会人口学、个人疾病史和总胆固醇等信息，建立动态前瞻性队列，每两年进行一次随访，截至2015年12月31日，共有109 878名男性进入队列；收集研究对象原发性肝癌发病结局信息。将研究人群按照TC四分位数进行分组：Q1组（<4.27 mmol/L），Q2组（4.27~4.90 mmol/L），Q3组（4.90~5.56 mmol/L）和Q4组（≥5.56 mmol/L），并以Q1组作为参比组，采用Cox比例风险回归模型分析TC与男性原发性肝癌发病风险的关联。 结果： 累计随访861 711.45人年，中位随访时间为8.83年；收集原发性肝癌新发病例355例。调整年龄、教育程度、收入情况、吸烟状况、饮酒状况、体质指数（BMI）和空腹血糖（FBG）水平后，随着TC水平的升高，原发性肝癌的发病风险逐渐降低[Q2、Q3和Q4组HR（95%CI）值分别为0.76（0.58~1.01）、0.59（0.43~0.79）和0.36（0.25~0.52）；P趋势<0.001]。按照乙肝表面抗原（HBsAg）感染状态、吸烟状况、饮酒状况和BMI进行分层后分析发现，上述关联在各亚组人群中均未发生改变（P趋势值均<0.05）；分别剔除参加基线调查后2年内新发肿瘤病例，基线诊断有肝硬化患者以及自报有调脂药物服用史的研究对象后，结论均无显著变化，与Q1组相比，Q4组HR（95%CI）值分别为0.41（0.28~0.61）、0.36（0.25~0.53）和0.38（0.26~0.54），且随着TC水平的升高，原发性肝癌的发病风险逐渐降低（P趋势均 <0.05）。 结论： 基线TC水平与男性原发性肝癌发病风险呈负相关，低TC水平可能增加原发性肝癌发病风险。.
Keywords: Cohort studies; Male; Primary liver cancer; Total cholesterol.