ABSTRACT

Objectives: Recent downward trends in breast cancer incidence have been attributed to declining use of hormone replacement therapy (HRT). To determine whether this is a credible conclusion, this study calculated population attributable risk (PAR) for HRT and other modifiable breast cancer risk factors.

Methods: PAR calculation needs both the prevalence of a risk factor, and the relative risk (RR) for breast cancer incidence for that risk factor. Prevalences were calculated for Canadian women, aged 50-69, participating in the National Population Health Survey, 1994-2006. RR were derived from published research: 1.4 for HRT use, 1.4 for excessive alcohol use, 1.15 for physical inactivity, 1.25 for smoking, 1.4 for BMI over 30 kg/m2. Trends for PAR were calculated for the risk factors separately, as well as combined. Age-adjusted breast cancer incidence rates were calculated for Canadian women aged 50-69 for the years 1994-2004.

Results: Between 1998 and 2004, PAR for HRT decreased by 50%. PAR for other risk factors showed only small changes, and the combined PAR decreased by 18.6%. Age-adjusted breast cancer incidence for women aged 50-69 peaked in 2000 at 330.0/100,000, then dropped by 17.2% by 2004.

Conclusion: Patterns of PAR for HRT use in Canada are consistent with the noticeable decrease in breast cancer incidence observed for women of the same age group. Combining PAR for all risk factors indicated that changes in HRT use overpowered any trends of other risk factors. The combined PAR suggest that alterations in lifestyle could have considerable impact on breast cancer incidence.

Key words: Breast cancer incidence; population attributable risk; hormone replacement therapy; alcohol; physical activity; obesity

La traduction du résumé se trouve à la fin de l'article. Can J Public Health 2010;101(4):405-9.

Many menopausal women have unpleasant symptoms which can be alleviated by hormone replacement therapy (HRT).1-3 For decades, the putative protective effects against heart disease and osteoporosis were also considered incentives for HRT use.4,5 When results from the Women's Health Initiative (WHI) and the Heart and Estrogen/progestin Replacement Study (HERS) provided no evidence for heart disease protection - in fact, even suggesting an increased risk6-8 - dramatic decreases in HRT use ensued.9-11

HRT use is also known to increase the risk of breast cancer. In many countries, breast cancer incidence appears to be on a downward trend in recent years. In Canada, age-adjusted incidence rates for breast cancer changed from a high of 105.1/100,000 in 1999 to a low of 96/100,000 after 2002.12 Similar decreases were seen for other countries.9,13,14 Because of HRT's known breast cancer risk, researchers were quick to attribute such declines in breast cancer rates to lower HRT use.15,16 Not all agreed.17,18

However, it is not clear whether a roughly 4% drop in overall breast cancer incidence is a drop of reasonable magnitude to be attributable to the observed decrease in HRT use. Population attributable risk (PAR) can be used to estimate the proportion of breast cancer incidence attributable to risk factors such as HRT. Accordingly, the objective of this study is to estimate trends in breast cancer PAR due to five modifiable breast cancer risk factors: HRT use, excessive alcohol consumption, obesity, lack of physical activity and smoking. Smoking is not as generally accepted as a breast cancer risk factor, but a recent Canadian expert panel concluded that there is a small, but real, increase.19

In order to estimate the PAR, risk measures for each of the risk factors are needed. Increases in breast cancer risk for HRT use varies with type of HRT used. For overall HRT use, breast cancer risk increased about 50%.15,16,20 For women taking the estrogen and progestin (E&P) combination, breast cancer risk doubled.15,16 The increased risk for unopposed estrogen varied from a high of an 80% increase to no effect at all. …