Pregnancy weight gain and breast cancer risk – part 3

Results
In the cohort, 123 (5.9%) women developed breast cancer during the mean follow-up of 38.9 years. The mean age at diagnosis was 56.0 years (range 35-74). Background characteristics and index pregnancies are described by tertiles of pregnancy weight gain in Table 1 . Low pregnancy weight gain (<11 kg) was associated to slightly higher ages during index pregnancy, during first pregnancy and at menarche, and to lower height, higher BMI before pregnancy, higher gestation weeks at delivery, lower weight of the placenta, smaller infant and lower proportion of users of estrogen drugs compared to women with higher pregnancy weight gain (11-15 kg or >15 kg).

Weight development during and after pregnancy is presented by tertiles of pregnancy weight gain in Table 2 . Higher weight gain during pregnancy was associated to higher weight loss after delivery, but also to higher weight retention and BMI at the postpartum check-up visit.

Breast Cancer Incidence per 100,000 Person Years
The mean BMI before pregnancy was 21.8 kg/m2 and the mean total extrapolated weight gain during pregnancy was 13.3 kg (range -5.0-33.1 kg) in our cohort. Average pregnancy weight gain (and range) was 13.1 kg (-3.0-33.1) among primiparas, 13.5 kg (1.9-30.7) among women who gave birth to their second child and 13.2 kg (-5.0-32.4) among women who gave birth to at least their third child. The incidence of breast cancer by 5 kg categories of total pregnancy weight gain is shown in Table 3 . Higher pregnancy weight gain was associated with a higher incidence of breast cancer. However, the number of women in some of the weight gain categories was small, and therefore the statistical analyses were carried out in tertiles of total pregnancy weight gain ( Table 4 ). The incidence of breast cancer was significantly higher in mothers in the highest tertile of pregnancy weight gain (15-33 kg), when compared to the middle tertile (11-15 kg) (p = 0.04). Breast cancer incidence was lowest in the middle tertile, but no differences in the risk were seen in the mothers of the lowest tertile of weight gain (less than 11 kg), when compared with the other two categories.

All analyses were initially carried out separately for pre- and postmenopausal breast cancers. The results for postmenopausal women were similar to the results for the whole cohort (results not shown). The incidence of premenopausal breast cancer was too low for statistical analysis and pre- and postmenopausal breast cancers were not separated further in the analyses. When these analyses were restricted to mothers who delivered after 39th gestation week, results were similar than in the whole cohort (results not shown).

The incidence of breast cancer was calculated separately for early (0-15th gestation weeks) and later pregnancy weight gain (15-40th gestation weeks). Early pregnancy weight gain was not associated with breast cancer risk. The impact of later pregnancy weight gain was similar to the impact of total weight gain, but more modest (results not shown).

Multivariate Analysis
Both unadjusted and multivariate adjusted rate ratios and confidence intervals for the risk of breast cancer are presented in Table 4 . In the Cox regression model, mothers in the highest tertile of pregnancy weight gain (>15 kg) had a 1.62-fold higher risk for breast cancer compared to mothers in the middle tertile (average weight gain 12.9 kg), when age at menarche, age at first birth, age at index pregnancy, BMI before pregnancy and parity at index birth were included in the model. To assess the sensitivity of these analyses, the lowest and highest weight gain groups were used as reference groups. When the lowest weight gain group was the reference group, no differences among the groups were seen. However, when the highest weight gain group was the reference group, women with average weight gain had significantly lower risk of breast cancer (multivariate adjusted RR 0.62, 95% CI 0.40-0.97).

When the middle tertile of weight gain was again used as the reference group and the analysis was restricted to mothers who delivered after 39th week of gestation, the results were essentially similar although statistically not significant (data not shown). The results did not either change when adjusted additionally for the year of index birth. The increased breast cancer risk in the highest tertile of pregnancy weight gain was found only for postmenopausal breast cancer (relative risk, RR = 1.80, 95% confidence interval, CI 1.05-3.07, p = 0.03). The RR for premenopausal cancer was 1.00 (95% CI 0.40- 2.48, p = 0.99). However, the number of premenopausal breast cancer cases with the information on all variables in the model was too low (n = 25) to yield sufficient power.

No statistically significant differences in breast cancer risk were observed between the tertiles of postpartum weight retention, determined approximately 51 days after delivery ( Table 4 ).

Other Results
Later age at menarche was marginally related to a decreased risk of breast cancer (adjusted RR = 0.99, 95% CI 0.97-1.00). Mother’s age at the time of first pregnancy or at the index pregnancy, parity at index birth or BMI before pregnancy were not statistically significantly associated with the risk of breast cancer. The results were similar when height and weight were used as separate variables in the model, instead of BMI.

Lower pre-pregnancy BMI was associated with higher weight gain during pregnancy (p < 0.001) and higher postpartum weight retention (p = 0.003), but not with postpartum weight loss. The differences in the incidence of breast cancer were not statistically significant between the pre-pregnancy BMI-categories.

The Case-control Study
Women who gained at least 15 kg weight during pregnancy had a higher BMI at the time of later hospital visit (29 years after pregnancy in average) than women who gained <11 kg weight during pregnancy (p = 0.021) ( Table 2 .). Changes in body weight (p < 0.001) and BMI (p < 0.001) were also higher in women who gained 11-15 kg or >15 kg compared to women who gained <11 kg during pregnancy. These findings are in agreement with earlier reports showing a link between excessive pregnancy weight gain and becoming overweight/obese later on.[26,27]