den Tonkelaar I, Seidell J C, Collette H J, de Waard F
Department of Epidemiology, University of Utrecht, The Netherlands.
Br J Cancer. 1994 Feb;69(2):352-7. doi: 10.1038/bjc.1994.64.
The associations of body fat and body fat distribution with breast cancer risk were examined in a prospective study in 9,746 post-menopausal women with a natural menopause, aged 49-66 at intake, participating in a breast cancer screening project (the DOM project in Utrecht). During a follow-up period of 15 years (mean follow-up time 12.5 years) 260 women developed breast cancer. Fat distribution, assessed by contrasting groups of subcapsular and triceps skinfold thickness, was found to be unrelated to breast cancer incidence. No significant relationship between body fat, measured either by weight, Quetelet's index, triceps skinfold or subscapular skinfold, and breast cancer risk was found when analysed in quartiles. However, women in the upper decile compared with the lower decile of the distribution of Quetelet's index were found to have a 1.9 times (95% CI 1.1-3.3) higher risk for breast cancer. These results seemed to be in contrast with the significant positive association between fatness, analysed in quartiles, and breast cancer observed in a cross-sectional study, based on mammographic screening, carried out previously in the same population. Although the differences between the present, prospective, study and our cross-sectional study may be due to chance it may be that there are differences between characteristics of breast cancer detected at screening and subsequently, which influence the associations between measures of fatness and risk of breast cancer.
在一项前瞻性研究中,对9746名自然绝经的绝经后女性进行了身体脂肪及身体脂肪分布与乳腺癌风险之间关联的研究。这些女性入组时年龄在49 - 66岁,参与了一项乳腺癌筛查项目(乌得勒支的DOM项目)。在15年的随访期内(平均随访时间12.5年),有260名女性患乳腺癌。通过对比肩胛下和肱三头肌皮褶厚度分组评估的脂肪分布,发现与乳腺癌发病率无关。按四分位数分析时,无论是通过体重、奎特列指数、肱三头肌皮褶或肩胛下皮褶测量的身体脂肪与乳腺癌风险之间均未发现显著关系。然而,与奎特列指数分布的最低十分位数相比,最高十分位数的女性患乳腺癌的风险高1.9倍(95%置信区间1.1 - 3.3)。这些结果似乎与之前在同一人群中基于乳房X线筛查进行的横断面研究中观察到的按四分位数分析的肥胖与乳腺癌之间显著的正相关有所不同。尽管当前的前瞻性研究与我们的横断面研究之间的差异可能是偶然的,但也可能是筛查时及后续检测到的乳腺癌特征存在差异,从而影响了肥胖测量指标与乳腺癌风险之间的关联。