Smoking, alcohol, diabetes, obesity, socioeconomic status, and the risk of colorectal cancer in a population-based case-control study
Smoking, alcohol, diabetes, obesity, socioeconomic status, and the risk of colorectal cancer in a population-based case-control study
复制标题
DOI：
10.1007/s10552-014-0470-7
复制
发表时间：
2014-12-01
影响因子：
2.3
通讯作者：
Heyworth, Jane S.
中科院分区：
文献类型：
--
作者：
Boyle, Terry;Fritschi, Lin;Heyworth, Jane S.
虽然以前的研究已经确定了可能决定参与研究意愿的因素，但相对较少的研究试图量化不参与可能对接触-疾病关联产生的影响。本研究的目的是(a)在病例对照研究中调查吸烟、酒精、糖尿病、肥胖和社会经济地位与结直肠癌风险之间的关系（病例和对照组的应答率分别为59.7%和47.2%）；(b)进行敏感性分析，以检验不参与可能产生的影响。使用逻辑回归来估计暴露与疾病的关联。然后，我们调查了各种人口统计学和健康因素之间的关联，以及个体参与病例对照研究的可能性，然后进行了两次敏感性分析（抽样权重和多重归算），以检验不参与偏差是否可能影响暴露-疾病关联。暴露于酒精、吸烟和糖尿病与结直肠癌风险增加有关。当检查与参与研究相关的因素时，我们发现病例和对照组之间存在一些差异，并且在敏感性分析中，与原始分析相比，暴露-疾病关联略有减弱。不参与可能会使风险估计偏离零值，但通常不足以改变研究的结论。
Although previous research has identified factors that may determine willingness to participate in research, relatively few studies have attempted to quantify the impact non-participation may have on exposure-disease associations. The aims of this study were to (a) investigate the associations between smoking, alcohol, diabetes, obesity, and socioeconomic status and the risk of colorectal cancer in a case-control study (59.7 and 47.2 % response fractions among cases and controls, respectively); and (b) perform sensitivity analyses to examine the possible influence of non-participation.Logistic regression was used to estimate the exposure-disease associations. We then investigated the associations between various demographic and health factors and the likelihood that an individual would participate in the case-control study and then performed two sensitivity analyses (sampling weights and multiple imputation) to examine whether non-participation bias may have influenced the exposure-disease associations.The exposures alcohol, smoking, and diabetes were associated with an increased risk of colorectal cancer. We found some differences between cases and controls when examining the factors associated with the participation in the study, and in the sensitivity analyses, the exposure-disease associations were slightly attenuated when compared with those from the original analysis.Non-participation may have biased the risk estimates away from the null, but generally not enough to change the conclusions of the study.