Background: Recent studies suggest an inverse relation between breast cancer and osteoporosis. Oestrogen is important in the pathophysiology of both breast and bone, and although cumulative exposure to oestrogen may explain the link between breast cancer and bone mass, this has never been proved. The Marburg breast cancer and osteoporosis trial (MABOT) aimed to elucidate the relation between breast cancer and bone mass ascertained by ultrasonometry measurement and to investigate whether endogenous and exogenous exposure to oestrogen and reproductive correlates has a role in this association.
Methods: We performed a case-control study including 2492 women (mean age+/-S.D., 54.4+/-10.3 years) in whom diseases and drug treatments known to affect bone metabolism, except for HT, had been excluded. All women underwent ultrasonometry measurement at the heel; 242 of the women had an incident breast cancer without a prior, specific pharmacological breast cancer treatment. The ultrasonometry variables - speed of sound (SOS), broadband ultrasound attenuation (BUA) and the stiffness index (SI) - were calculated and compared in women with and without breast cancer. Because of significant intergroup differences in factors such as age, body mass index and exposure to oestrogen, a multiple linear regression analysis as well as a second analysis of ultrasonometry variables was undertaken using a randomly selected sample of 242 healthy women post-matched with the breast cancer group for possible confounding variables. Odds ratios were used to compare the relation between breast cancer risk and ultrasonometry heel measurements.
Results: Women with breast cancer were significantly older, weighed more, had a higher body mass index, were more likely to be parous and to have breast fed, were older at the menopause and had been exposed to oestrogen for longer than control women. In addition, the ultrasonometry variables speed of sound and the stiffness index T- and Z-score were significantly higher in women with breast cancer even after a matched pair analysis was performed (p<0.001). Additionally, results of a multiple linear regression showed that women with breast cancer had a significantly higher SOS (p<0.001), body weight (p<0.05) and duration of breast feeding (p<0.05) while osteoporotic fracture were reduced (p<0.001). When women with breast cancer and their matched controls were finally grouped according to SOS and T-score quartiles, the odds ratios (95% confidence intervals) for breast cancer risk in the second, third and fourth quartiles compared with the lowest quartile were 2.5 (1.4-4.3), 3.1 (1.8-5.3) and 4.7 (2.7-8.2) as well as 1.9 (1.1-3.2), 2.3 (1.3-3.9) and 2.9 (1.7-5.0), respectively.
Conclusions: The ultrasonometry variables speed of sound, stiffness index, T- and Z-score are higher in women with an incident breast cancer than in healthy controls, even after post-matching for possible confounding variables. This association was confirmed in a multiple linear regression model. Women with SOS and T-score values in the higher quartiles have a greater risk of breast cancer than women in the lowest quartile. We found no association between the higher ultrasonometry variables and cancer specific characteristics or reproductive correlates such as age at menarche and menopause or cumulative oestrogen exposure. Although the biological mechanisms linking bone mass and the risk of breast cancer are not fully understood, factors other than reproductive correlates, endogenous and exogenous exposure to oestrogen must play a part.