Discussion
The results obtained in our study indicate that higher than recommended pregnancy weight gain increased mothers’ risk of developing breast cancer. Thus, women who gained more than 15 kg during pregnancy had a 62% increase in breast cancer risk, compared to those who gained between 11-15 kg. The Institute of Medicine (IOM) published their most recent recommendations for pregnancy weight gain in 1990.[28] The recommended pregnancy weight gain in the USA is 11.5-16 kg for women with normal pre-pregnancy BMI; i.e., they are not obese or underweight. Pregnancy weight gain recommendations are lower (7-11.5 kg) for overweight women and higher (12.5-18 kg) for underweight women. As seen in Table 3 , the incidence of breast cancer in our study was highest among women who gained more than 20 kg during pregnancy, suggesting that the increase in risk may apply primarily to women at the most extreme range of pregnancy weight gain.
An increase in breast cancer risk was seen mostly in women who were diagnosed with this disease after age 50 and thus were postmenopausal. However, the number of premenopausal breast cancers was low in the cohort, and we cannot exclude the possibility that pregnancy weight gain may also increase the risk of premenopausal breast cancer.
Data generated in epidemiological studies rarely provide causal relationships. We propose four different mechanisms that may link high pregnancy weight gain to a later increase in breast cancer risk. First, weight retention in women who gained excessive amounts of weight during pregnancy may have persisted into their postmenopausal years. Women prone to postpartum weight retention might also be prone to long-lasting weight gain after pregnancy,[29] and high BMI during postmenopausal years increases breast cancer risk.[26] To examine this possibility, information on body weight at the time of breast cancer diagnosis was obtained. If the association between pregnancy weight gain and breast cancer risk was affected by later weight development, breast cancer cases should have had higher BMI at the time of diagnosis. As this was not the case, we propose that high pregnancy weight gain increases breast cancer risk independently from body weight at the time of diagnosis.
Another alternative is that women who gained an excessive amount of weight during pregnancy may have had higher pregnancy hormone and growth factor levels than women who gained within the recommended range, stimulating the growth of existing malignant cells in the breast, leading to development of a detectable tumor. Several studies have shown that markers of high pregnancy estrogen levels increase mother’s breast cancer risk.[4-6,9-12,16] Estrogen levels may correlate with high pregnancy weight gain,[19] but two recent studies have not confirmed this observation.[20,21] Other possible hormones that could be mediating the effect of pregnancy weight gain on breast cancer risk include leptin. Leptin levels correlate strongly with BMI,[27] also during pregnancy,[30] and leptin is suggested to increase breast cancer risk.[31] We did not have any biological samples available for hormone measurements.
It is known that high hormone levels increase the proliferation of normal breast cells that then is accompanied by increased genomic instability and accumulation of DNA adducts.[16,32] Thus, the third explanation is that high pregnancy weight gain increased the likelihood of DNA damage and mutations in genes that initiate breast cancer. Since the window between index pregnancy and diagnosis of breast cancer was approximately 30 years, there was enough time for the initiation to have taken place during pregnancy.
Finally, known and unknown causes of breast cancer may have confounded the results. For example, these causative factors might be more common in women who gain excessive amounts of weight during pregnancy or they caused women to gain excessive amounts of weight during pregnancy. A theoretical example is a gene mutation/polymorphism that could both make a woman more prone to gain weight during pregnancy and increases breast cancer risk.
Methodological limitations have to be considered when interpreting the results, and they include high rate of exclusion and an exposure to estrogenic drugs during pregnancy. Of the 4,090 women available for the study, 48.9 % were excluded for reasons listed in Figure 1 (109 of which were diagnosed with breast cancer). Total pregnancy weight gain could be extrapolated only for 2,089 women, of which 123 had developed breast cancer. Other information on background and index pregnancies indicated that the excluded women might have gained less weight during pregnancy than the final study population (see chapter Inclusions and exclusions). However, we found no evidence that breast cancer incidence was different between the women excluded and included to the study.
Some women in our cohort had been exposed to synthetic estrogens during pregnancy to avoid a threatening miscarriage, and this exposure might have affected the results. However, it was the initial reason for obtaining information from pregnant women, and we found no effect of the drug exposure on the incidence of breast cancer. We are not aware of any other cohort that could be used to assess the link between pregnancy weight gain and breast cancer risk, but if such a cohort becomes available, and it is not potentially compromised by high rate of exclusion of subjects or an exposure to drugs during pregnancy, the present results can be either confirmed or nullified. Follow-up of “old” cohorts similar to ours is rarely possible, making our study relatively unique.
Another area of potential source for errors is the variability in time period between the weight measurements during pregnancy (range 3-295 days), requiring us to extrapolate the pregnancy weight gain for each woman. This step also has been successfully used in other studies.[33] A further weakness of the study was that no information on weight gain in previous and subsequent pregnancies was available. Therefore, we cannot exclude the possibility that a woman who did not develop breast cancer and during the index pregnancy gained less than 15 kg, might have had subsequent pregnancies that were characterized by excessive weight gain.
Finally, in the case-control study that determined the impact of body weight at the time of diagnosis on breast cancer risk, information on this weight was obtained only for 53% of the cases and 50% of the controls. However, the direction of bias rising from exclusion may have diluted the effect, rather than caused it.
In conclusion, our findings suggest that excessive pregnancy weight gain increased later risk of developing breast cancer. This association needs to be further confirmed in prospective studies.
The work was supported by grants to LH-C from NCI (5 RO1 CA89950-02), the Susan G. Komen Breast Cancer Research Foundation (9847), Breast Cancer Research Foundation (BCRA-01) and Department of Defense (DAMD17-99-1-9196).
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
DES = diethylstilbestrol; BMI= body mass index
