Gann P H, Ma J, Giovannucci E, Willett W, Sacks F M, Hennekens C H, Stampfer M J
Department of Preventive Medicine, Northwestern University Medical School, Chicago, Illinois 60611, USA.
Cancer Res. 1999 Mar 15;59(6):1225-30.
Dietary consumption of the carotenoid lycopene (mostly from tomato products) has been associated with a lower risk of prostate cancer. Evidence relating other carotenoids, tocopherols, and retinol to prostate cancer risk has been equivocal. This prospective study was designed to examine the relationship between plasma concentrations of several major antioxidants and risk of prostate cancer. We conducted a nested case-control study using plasma samples obtained in 1982 from healthy men enrolled in the Physicians' Health Study, a randomized, placebo-controlled trial of aspirin and beta-carotene. Subjects included 578 men who developed prostate cancer within 13 years of follow-up and 1294 age- and smoking status-matched controls. We quantified the five major plasma carotenoid peaks (alpha- and beta-carotene, beta-cryptoxanthin, lutein, and lycopene) plus alpha- and gamma-tocopherol and retinol using high-performance liquid chromatography. Results for plasma beta-carotene are reported separately. Odds ratios (ORs), 95% confidence intervals (Cls), and Ps for trend were calculated for each quintile of plasma antioxidant using logistic regression models that allowed for adjustment of potential confounders and estimation of effect modification by assignment to either active beta-carotene or placebo in the trial. Lycopene was the only antioxidant found at significantly lower mean levels in cases than in matched controls (P = 0.04 for all cases). The ORs for all prostate cancers declined slightly with increasing quintile of plasma lycopene (5th quintile OR = 0.75, 95% CI = 0.54-1.06; P, trend = 0.12); there was a stronger inverse association for aggressive prostate cancers (5th quintile OR = 0.56, 95% CI = 0.34-0.91; P, trend = 0.05). In the placebo group, plasma lycopene was very strongly related to lower prostate cancer risk (5th quintile OR = 0.40; P, trend = 0.006 for aggressive cancer), whereas there was no evidence for a trend among those assigned to beta-carotene supplements. However, in the beta-carotene group, prostate cancer risk was reduced in each lycopene quintile relative to men with low lycopene and placebo. The only other notable association was a reduced risk of aggressive cancer with higher alpha-tocopherol levels that was not statistically significant. None of the associations for lycopene were confounded by age, smoking, body mass index, exercise, alcohol, multivitamin use, or plasma total cholesterol level. These results concur with a recent prospective dietary analysis, which identified lycopene as the carotenoid with the clearest inverse relation to the development of prostate cancer. The inverse association was particularly apparent for aggressive cancer and for men not consuming beta-carotene supplements. For men with low lycopene, beta-carotene supplements were associated with risk reductions comparable to those observed with high lycopene. These data provide further evidence that increased consumption of tomato products and other lycopene-containing foods might reduce the occurrence or progression of prostate cancer.
饮食中摄入类胡萝卜素番茄红素（主要来自番茄制品）与降低前列腺癌风险相关。关于其他类胡萝卜素、生育酚和视黄醇与前列腺癌风险的证据一直存在争议。这项前瞻性研究旨在探讨几种主要抗氧化剂的血浆浓度与前列腺癌风险之间的关系。我们进行了一项巢式病例对照研究，使用1982年从参与医师健康研究的健康男性中采集的血浆样本，该研究是一项关于阿司匹林和β-胡萝卜素的随机、安慰剂对照试验。受试者包括578名在随访13年内患前列腺癌的男性以及1294名年龄和吸烟状况相匹配的对照者。我们使用高效液相色谱法对五个主要血浆类胡萝卜素峰（α-和β-胡萝卜素、β-隐黄质、叶黄素和番茄红素）以及α-和γ-生育酚和视黄醇进行了定量。血浆β-胡萝卜素的结果单独报告。使用逻辑回归模型计算每个血浆抗氧化剂五分位数的比值比（OR）、95%置信区间（CI）和趋势P值，该模型允许对潜在混杂因素进行调整，并通过在试验中分配到活性β-胡萝卜素或安慰剂来估计效应修饰。番茄红素是唯一在病例组中平均水平显著低于匹配对照组的抗氧化剂（所有病例P = 0.04）。所有前列腺癌的OR随着血浆番茄红素五分位数的增加而略有下降（第五五分位数OR = 0.75，95% CI = 0.54 - 1.06；趋势P = 0.12）；对于侵袭性前列腺癌，存在更强的负相关（第五五分位数OR = 0.56，95% CI = 0.34 - 0.91；趋势P = 0.05）。在安慰剂组中，血浆番茄红素与降低前列腺癌风险密切相关（侵袭性癌症第五五分位数OR = 0.40；趋势P = 0.006），而在分配到β-胡萝卜素补充剂的人群中没有明显趋势。然而，在β-胡萝卜素组中，相对于番茄红素水平低的男性和安慰剂组，每个番茄红素五分位数的前列腺癌风险都有所降低。唯一另一个值得注意的关联是较高的α-生育酚水平与侵袭性癌症风险降低有关，但无统计学意义。番茄红素的所有关联均未受到年龄、吸烟、体重指数、运动、饮酒、使用多种维生素或血浆总胆固醇水平的混杂影响。这些结果与最近的一项前瞻性饮食分析一致，该分析确定番茄红素是与前列腺癌发生最明显负相关的类胡萝卜素。这种负相关在侵袭性癌症和未食用β-胡萝卜素补充剂的男性中尤为明显。对于番茄红素水平低的男性，β-胡萝卜素补充剂与高番茄红素水平者观察到的风险降低相当。这些数据进一步证明，增加番茄制品和其他含番茄红素食物的摄入量可能会降低前列腺癌的发生或进展。
