Harlap S
Department of Medical Ecology, Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY 10021.
J Reprod Med. 1991 May;36(5):374-95.
Case-control studies dealing with breast cancer and oral contraceptives (OCs) were reviewed against a background of currently accepted rules for conducting and interpreting epidemiologic research. There is a strong consensus that OCs have not increased the risk of breast cancer in women over age 45, even when they have been used for long periods. The controversy is confined to studies of cancer diagnosed in young women. Although an effect of duration of OC use has been observed in some studies, there is neither a latency period nor an effect from recent or current use. There is no obvious relationship to the type of pill and no consistent effect of the dose of estrogen. The effect, if it exists at all, is not confined to any specific subgroup; it has not been related to age at starting OCs. It has been observed in only a few studies, some of which were subsequently discounted. Others have shown no effect or have suggested a reduced risk of breast cancer with increasing duration of use. None of the studies purporting to show an effect of OCs have controlled for the potential confounding effects of duration of lactation, induced abortion, recent pregnancy or a history of diseases that are associated both with reduced use of OCs and reduced risk of breast cancer. Most have inappropriately controlled for benign breast disease. There is an urgent need for a reanalysis of existing studies to take these factors into account. The lack of consistency and the potential for bias and confounding argue strongly against concluding, at this time, that there is a causal relationship between OCs and breast cancer.
在当前公认的流行病学研究开展及解读规则背景下，对有关乳腺癌与口服避孕药（OCs）的病例对照研究进行了综述。人们已形成强烈共识，即OCs不会增加45岁以上女性患乳腺癌的风险，即便长期使用亦是如此。争议仅限于对年轻女性诊断出的癌症研究。尽管在一些研究中观察到了OCs使用时长的影响，但既不存在潜伏期，近期或当前使用也无影响。与避孕药类型无明显关联，雌激素剂量也无一致影响。该影响（若确实存在）并不局限于任何特定亚组；与开始使用OCs时的年龄无关。仅在少数研究中观察到该影响，其中一些研究随后被否定。其他研究则未显示出影响，或表明随着使用时长增加，患乳腺癌风险降低。所有声称显示OCs有影响的研究均未控制哺乳期时长、人工流产、近期怀孕或与OCs使用减少及乳腺癌风险降低均相关的疾病史等潜在混杂效应。大多数研究对良性乳腺疾病的控制都不恰当。迫切需要重新分析现有研究以考虑这些因素。缺乏一致性以及存在偏差和混杂的可能性，有力地反驳了目前得出OCs与乳腺癌之间存在因果关系的结论。
