PDF(Pubmed)
Abstract：
BACKGROUND: Previous literature on dairy products and risk of breast cancer is inconsistent, and the relationship may depend on the life-period of dietary assessment.
OBJECTIVE: We examined dairy consumption from adolescence through later adulthood and incidence of breast cancer by menopausal status and tumor molecular subtypes in the Nurses\' Health Study (NHS), a prospective cohort study.
METHODS: We analyzed data from 63,847 females in the NHS collected from 1980 to 2018. Average intake of dairy products during adulthood was assessed by validated semiquantitative food frequency questionnaires throughout follow-up. Participants recalled adolescent dietary intake in 1986. Multivariable Cox proportional hazards models were used to estimate hazard ratios (HRs) relating dairy product consumption to breast cancer risk overall, by menopausal status, and by subtypes.
RESULTS: We documented 5733 incident cases of invasive breast cancer during 32 y of follow-up (n = 5298 postmenopausal). Lifetime, adolescent, adulthood, and postmenopausal total dairy and milk intakes were not associated with overall breast cancer risk (nonsignificant HRs comparing highest with lowest quintile range = 0.97-1.08), although there was a suggestive positive association between adolescent milk intake and breast cancer risk (HR: 1.09; 95% CI: 1.00, 1.18). Higher lifetime and premenopausal cheese intakes were associated with modestly lower risks of breast cancer (comparing highest with lowest quintile, HR for lifetime cheese intake: 0.90; 95% CI: 0.82, 0.98; HR for premenopausal cheese intake: 0.89; 95% CI: 0.79, 1.00). Results varied by tumor subtype and some evidence for heterogeneity was observed for an association between premenopausal milk intake and breast cancer (HR for estrogen receptor [ER]-positive: 0.84; 95% CI: 0.72, 0.99; ER-negative: 1.36; 95% CI: 1.00, 1.84; P heterogeneity = 0.04).
CONCLUSIONS: These findings suggest that overall dairy consumption was not associated with risk of breast cancer. However, heterogeneity was observed for type of dairy food, period of life, and tumor subtypes.
OBJECTIVE: We examined dairy consumption from adolescence through later adulthood and incidence of breast cancer by menopausal status and tumor molecular subtypes in the Nurses\' Health Study (NHS), a prospective cohort study.
METHODS: We analyzed data from 63,847 females in the NHS collected from 1980 to 2018. Average intake of dairy products during adulthood was assessed by validated semiquantitative food frequency questionnaires throughout follow-up. Participants recalled adolescent dietary intake in 1986. Multivariable Cox proportional hazards models were used to estimate hazard ratios (HRs) relating dairy product consumption to breast cancer risk overall, by menopausal status, and by subtypes.
RESULTS: We documented 5733 incident cases of invasive breast cancer during 32 y of follow-up (n = 5298 postmenopausal). Lifetime, adolescent, adulthood, and postmenopausal total dairy and milk intakes were not associated with overall breast cancer risk (nonsignificant HRs comparing highest with lowest quintile range = 0.97-1.08), although there was a suggestive positive association between adolescent milk intake and breast cancer risk (HR: 1.09; 95% CI: 1.00, 1.18). Higher lifetime and premenopausal cheese intakes were associated with modestly lower risks of breast cancer (comparing highest with lowest quintile, HR for lifetime cheese intake: 0.90; 95% CI: 0.82, 0.98; HR for premenopausal cheese intake: 0.89; 95% CI: 0.79, 1.00). Results varied by tumor subtype and some evidence for heterogeneity was observed for an association between premenopausal milk intake and breast cancer (HR for estrogen receptor [ER]-positive: 0.84; 95% CI: 0.72, 0.99; ER-negative: 1.36; 95% CI: 1.00, 1.84; P heterogeneity = 0.04).
CONCLUSIONS: These findings suggest that overall dairy consumption was not associated with risk of breast cancer. However, heterogeneity was observed for type of dairy food, period of life, and tumor subtypes.
摘要：
背景：以前关于乳制品和乳腺癌风险的文献不一致，这种关系可能取决于饮食评估的生命周期。
目的：我们在护士健康研究（NHS）中通过绝经状态和肿瘤分子亚型检查了从青春期到成年后期的乳制品消费量以及乳腺癌的发病率。前瞻性队列研究。
方法：我们分析了1980年至2018年收集的NHS中63,847名女性的数据。在整个随访过程中，通过经过验证的半定量食物频率问卷评估了成年期乳制品的平均摄入量。参与者回顾了1986年青少年的饮食摄入量。多变量Cox比例风险模型用于估计乳制品消费与乳腺癌总体风险的风险比（HR）。根据更年期状态，和亚型。
结果：我们在32年的随访中记录了5733例浸润性乳腺癌事件（n=5298例绝经后）。寿命,青春期,成年,和绝经后乳制品和牛奶的总摄入量与总体乳腺癌风险(比较最高和最低的5分率范围=0.97-1.08的不显著HR)无关,尽管青少年牛奶摄入量与乳腺癌风险之间存在暗示性正相关(HR=1.09,95%CI=1.00,1.18).较高的寿命和绝经前奶酪摄入量与乳腺癌风险较低相关(比较最高和最低的五分之一，终身奶酪摄入量的HR=0.90，95%CI=0.82，0.98；绝经前奶酪摄入量的HR=0.89，95%CI=0.79，1.00）。结果因肿瘤亚型而异，绝经前乳汁摄入与乳腺癌之间存在一定的异质性(ER阳性的HR=0.84,95%CI=0.72,0.99;ER阴性=1.36,95%CI=1.00,1.84;P异质性=0.04).
结论：这些研究结果表明，总体乳制品消费与乳腺癌风险无关。然而,观察到乳制品类型的异质性，生命的时期,和肿瘤亚型。
目的：我们在护士健康研究（NHS）中通过绝经状态和肿瘤分子亚型检查了从青春期到成年后期的乳制品消费量以及乳腺癌的发病率。前瞻性队列研究。
方法：我们分析了1980年至2018年收集的NHS中63,847名女性的数据。在整个随访过程中，通过经过验证的半定量食物频率问卷评估了成年期乳制品的平均摄入量。参与者回顾了1986年青少年的饮食摄入量。多变量Cox比例风险模型用于估计乳制品消费与乳腺癌总体风险的风险比（HR）。根据更年期状态，和亚型。
结果：我们在32年的随访中记录了5733例浸润性乳腺癌事件（n=5298例绝经后）。寿命,青春期,成年,和绝经后乳制品和牛奶的总摄入量与总体乳腺癌风险(比较最高和最低的5分率范围=0.97-1.08的不显著HR)无关,尽管青少年牛奶摄入量与乳腺癌风险之间存在暗示性正相关(HR=1.09,95%CI=1.00,1.18).较高的寿命和绝经前奶酪摄入量与乳腺癌风险较低相关(比较最高和最低的五分之一，终身奶酪摄入量的HR=0.90，95%CI=0.82，0.98；绝经前奶酪摄入量的HR=0.89，95%CI=0.79，1.00）。结果因肿瘤亚型而异，绝经前乳汁摄入与乳腺癌之间存在一定的异质性(ER阳性的HR=0.84,95%CI=0.72,0.99;ER阴性=1.36,95%CI=1.00,1.84;P异质性=0.04).
结论：这些研究结果表明，总体乳制品消费与乳腺癌风险无关。然而,观察到乳制品类型的异质性，生命的时期,和肿瘤亚型。
