Navarro Sandi L, Kantor Elizabeth D, Song Xiaoling, Milne Ginger L, Lampe Johanna W, Kratz Mario, White Emily
Fred Hutchinson Cancer Research Center, Division of Public Health Sciences, Seattle, Washington.
Fred Hutchinson Cancer Research Center, Division of Public Health Sciences, Seattle, Washington. Memorial Sloan Kettering Cancer Center, Department of Epidemiology and Biostatistics, New York, New York.
Cancer Epidemiol Biomarkers Prev. 2016 Mar;25(3):521-31. doi: 10.1158/1055-9965.EPI-15-0956. Epub 2016 Jan 11.
While much is known about correlates of C-reactive protein (CRP), little is known about correlates of other inflammation biomarkers. As these measures are increasingly being used in epidemiologic studies, it is important to determine what factors affect inflammation biomarker concentrations.
Using age, sex, and body mass index (BMI) adjusted linear regression, we examined 38 exposures (demographic and anthropometric measures, chronic disease history, NSAIDs, dietary factors, and supplement use) of 8 inflammation biomarkers [CRP, IL1β, IL6, IL8, TNFα, and soluble TNF receptors (sTNFR) in plasma; and prostaglandin E2 metabolite (PGE-M) in urine] in 217 adults, ages 50 to 76 years.
Increasing age was associated with higher concentrations of all biomarkers except IL1β. BMI was positively associated with CRP and sTNFR I and II. Saturated fat intake was associated with increased CRP, sTNFRII, TNFα, and IL1β, whereas eicosapentaenoic acid + docosahexaenoic acid (EPA+DHA) intake (diet or total) was associated with decreased CRP, TNFα, and IL1β. Results for sex were varied: CRP and IL6 were lower among men, whereas PGE-M and sTNFRI were higher. Higher CRP was also associated with smoking, hormone replacement therapy use, and γ-tocopherol intake; lower CRP with physical activity, and intakes of dietary vitamin C and total fiber.
Although the associations varied by biomarker, the factors having the greatest number of significant associations (P ≤ 0.05) with the inflammation biomarkers were age, BMI, dietary saturated fat, and EPA+DHA omega-3 fatty acids.
Our results suggest that potential confounders in epidemiologic studies assessing associations with inflammation biomarkers vary across specific biomarkers.
虽然关于C反应蛋白（CRP）的相关因素已有很多了解，但对于其他炎症生物标志物的相关因素却知之甚少。由于这些指标越来越多地用于流行病学研究，因此确定哪些因素会影响炎症生物标志物浓度非常重要。
我们使用年龄、性别和体重指数（BMI）调整的线性回归，研究了217名年龄在50至76岁之间的成年人的38种暴露因素（人口统计学和人体测量指标、慢性病病史、非甾体抗炎药、饮食因素和补充剂使用情况）与8种炎症生物标志物的关系，这8种生物标志物分别为血浆中的CRP、IL1β、IL6、IL8、TNFα和可溶性TNF受体（sTNFR）；以及尿液中的前列腺素E2代谢物（PGE-M）。
年龄增加与除IL1β外的所有生物标志物浓度升高有关。BMI与CRP以及sTNFR I和II呈正相关。饱和脂肪摄入量与CRP、sTNFRII、TNFα和IL1β升高有关，而二十碳五烯酸+二十二碳六烯酸（EPA+DHA）摄入量（饮食或总量）与CRP、TNFα和IL1β降低有关。性别的结果各不相同：男性的CRP和IL6较低，而PGE-M和sTNFRI较高。较高的CRP还与吸烟、使用激素替代疗法和摄入γ-生育酚有关；较低的CRP与身体活动以及饮食中维生素C和总纤维的摄入量有关。
尽管不同生物标志物的关联有所不同，但与炎症生物标志物具有最多显著关联（P≤0.05）的因素是年龄、BMI、饮食饱和脂肪和EPA+DHAω-3脂肪酸。
我们的结果表明，在评估与炎症生物标志物关联的流行病学研究中，潜在的混杂因素因特定生物标志物而异。
