Gordon-Dseagu Vanessa L Z, Thompson Frances E, Subar Amy F, Ruder Elizabeth H, Thiébaut Anne C M, Potischman Nancy, Stolzenberg-Solomon Rachael
Am J Epidemiol. 2017 Aug 1;186(3):305-317. doi: 10.1093/aje/kwx036.
Given the long latency period of pancreatic cancer, exploring the influence of early and midlife exposures will further advance our understanding of the disease. We assessed associations between diet and pancreatic cancer incidence in the National Institutes of Health (NIH)-AARP (formerly American Association of Retired Persons) Diet and Health Study. In 1996, a total of 303,094 participants completed 2 food frequency questionnaires that assessed diet at ages 12-13 years and 10 years previously. We used Cox proportional hazards regression to estimate adjusted hazard ratios and 95% confidence intervals. Through the end of 2006, a total of 1,322 pancreatic cancer cases occurred (average follow up time = 10.1 years). When comparing the highest tertiles with the lowest, carbohydrate intake during adolescence (hazard ratio (HR) = 0.87, 95% confidence interval (CI): 0.76, 0.99), but not 10 years before baseline, was inversely associated with pancreatic cancer risk. Total fat intake 10 years before baseline was significantly associated with increased risk (HR = 1.17, 95% CI: 1.02, 1.34), while risk was higher for high fat intake during both adolescence and midlife. Calcium intake 10 years before baseline was associated with reduced risk (HR = 0.87, 95% CI: 0.76, 0.99), as was a change from low intake in adolescence to high intake in midlife (HR = 0.71, 95% CI: 0.54, 0.93). Our study found a number of dietary factors present during adolescence and midlife to be associated with pancreatic cancer.
鉴于胰腺癌的潜伏期较长，探索早期和中年时期的暴露因素影响将进一步加深我们对该疾病的理解。我们在美国国立卫生研究院（NIH）-美国退休人员协会（AARP，前身为美国退休人员协会）饮食与健康研究中评估了饮食与胰腺癌发病率之间的关联。1996年，共有303,094名参与者完成了两份食物频率问卷，这些问卷评估了他们在12 - 13岁时以及10年前的饮食情况。我们使用Cox比例风险回归来估计调整后的风险比和95%置信区间。到2006年底，共发生了1322例胰腺癌病例（平均随访时间 = 10.1年）。将最高三分位数与最低三分位数进行比较时，青春期碳水化合物摄入量（风险比（HR）= 0.87，95%置信区间（CI）：0.76，0.99）与胰腺癌风险呈负相关，但基线前10年的碳水化合物摄入量则无此关联。基线前10年的总脂肪摄入量与风险增加显著相关（HR = 1.17，95% CI：1.02，1.34），而青春期和中年时期高脂肪摄入量的风险更高。基线前10年的钙摄入量与风险降低相关（HR = 0.87，95% CI：0.76，0.99），从青春期低摄入量转变为中年高摄入量时也是如此（HR = 0.71，95% CI：0.54，0.93）。我们的研究发现青春期和中年时期存在的一些饮食因素与胰腺癌有关。
