Farwell Wildon R, Michael Gaziano J, Norkus Edward P, Sesso Howard D
Massachusetts Veterans Epidemiology Research and Information Center, VA Boston Healthcare System, Boston Division (151 MAV), 150 South Huntington Avenue, Boston, MA, USA.
Br J Nutr. 2008 Oct;100(4):883-9. doi: 10.1017/S0007114508944111. Epub 2008 Mar 12.
Individual plasma carotenoids have been associated with various chronic diseases but little is known about the relationship between total plasma carotenoids and risk factors for chronic diseases. In the Physicians' Health Study, we examined 492 men free of CVD and cancer for the relationship between total plasma carotenoids (the sum of alpha-carotene, beta-carotene, lycopene, zeaxanthin, lutein and beta-cryptoxanthin) and a wide variety of factors that predict chronic disease. Multivariate linear and logistic regression was performed to calculate parameter estimates (95% CI) and OR (95% CI) for total plasma carotenoids. In linear regression models, BMI, hypertension, alcohol intake and plasma levels of each lipid parameter and a-tocopherol significantly predicted levels of total plasma carotenoids. Upon adjustment for multiple chronic disease risk factors, the OR for levels of total plasma carotenoids greater than or equal to the median (> or=1.301 micromol/l) was statistically significant for current smoking (OR 0.21; 95% CI 0.06, 0.77), weekly alcohol ingestion (OR 2.30; 95% CI 1.06, 4.99), daily alcohol ingestion (OR 2.46; 95% CI 1.29, 4.67), each 100 mg/l increase in total cholesterol (OR 0.73; 95% CI 0.58, 0.91), LDL-cholesterol (OR 1.48; 95% CI 1.17, 1.89) and HDL-cholesterol (OR 1.58; 95% CI 1.26, 1.99), each 100 mg/ml increase in intercellular adhesion molecule-1 (OR 0.70; 95% CI 0.53, 0.93) and each 10 micromol/l increase in alpha-tocopherol (OR 1.33; 95% CI 1.12, 1.57), using logistic regression. Few lifestyle and clinical risk factors appear to be related to levels of total plasma carotenoids; however, levels of biomarkers such as plasma lipids and alpha-tocopherol may be strongly related.
个体血浆类胡萝卜素与多种慢性疾病相关，但血浆总类胡萝卜素与慢性疾病风险因素之间的关系却鲜为人知。在医生健康研究中，我们对492名无心血管疾病和癌症的男性进行了研究，以探讨血浆总类胡萝卜素（α-胡萝卜素、β-胡萝卜素、番茄红素、玉米黄质、叶黄素和β-隐黄质的总和）与多种预测慢性疾病的因素之间的关系。进行多变量线性和逻辑回归分析，以计算血浆总类胡萝卜素的参数估计值（95%置信区间）和比值比（95%置信区间）。在线性回归模型中，体重指数、高血压、酒精摄入量以及每种脂质参数和α-生育酚的血浆水平均能显著预测血浆总类胡萝卜素水平。在对多种慢性疾病风险因素进行调整后，使用逻辑回归分析发现，血浆总类胡萝卜素水平大于或等于中位数（≥1.301微摩尔/升）时，当前吸烟（比值比0.21；95%置信区间0.06，0.77）、每周饮酒（比值比2.30；95%置信区间1.06，4.99）、每日饮酒（比值比2.46；95%置信区间1.29，4.67）、总胆固醇每升高100毫克/升（比值比0.73；95%置信区间0.58，0.91）、低密度脂蛋白胆固醇（比值比1.48；95%置信区间1.17，1.89）和高密度脂蛋白胆固醇（比值比1.58；95%置信区间1.26，1.99）、细胞间黏附分子-1每升高100毫克/毫升（比值比0.70；95%置信区间0.53，0.93）以及α-生育酚每升高10微摩尔/升（比值比1.33；95%置信区间1.12，1.57）的比值比具有统计学意义。很少有生活方式和临床风险因素似乎与血浆总类胡萝卜素水平相关；然而，血浆脂质和α-生育酚等生物标志物水平可能与之密切相关。
