We report here results from the first study which have measured maternal intakes of dietary fat (including n-3 PUFA and n-6 PUFA) using food diaries in all trimesters in relation to maternal plasma estradiol and progesterone levels in a group of Caucasian women. No association between dietary intakes of total fat or fat subtypes (including n-3 PUFA and n-6 PUFA) and plasma estradiol or progesterone levels were found. Gestational weight gain was not associated with plasma estradiol levels, while a positive association was found for progesterone.
To our knowledge this is the first study to evaluate n-3 PUFA and n-6 PUFA in relation to maternal progesterone levels. Further, only one earlier study has investigated the impact of n-3 PUFA and n-6 PUFA on pregnancy estradiol levels. In 189 Japanese pregnant women dietary intake was measured on gestation week 29 using a food diary and pregnancy estrogen levels were measured on gestation week 10 and 29. No association was found between maternal dietary intakes of n-3 PUFA or n-6 PUFA and serum estradiol levels. However, a positive association between n-6 PUFA and umbilical cord estriol levels as well as an inverse association between n-3 PUFA with umbilical cord estradiol levels was found [14
]. In our study, conducted in a Caucasian population with a high risk of breast cancer, did not provide any support for an opposing effect of n-3 and n-6 PUFA on pregnancy estradiol levels. The somewhat inconsistent results may be due to several reasons. First, the associations for n-3 and n-6 PUFA and estradiol in Japanese women were only detected in umbilical cord, and we measured estradiol in circulating plasma. Second, the populations differ in several respects. Estradiol levels were higher in the Japanese women, which in agreement with earlier reports comparing estradiol levels in Asian and Caucasian pregnant women [9
]. Furthermore, the dietary intake of long n-3 PUFA was twice as high in the Japanese women.
Three earlier studies have evaluated the associations between pregnancy estradiol levels and maternal dietary intakes of total fat, MUFA, PUFA and SFA [14
], and one of these studies also investigated the corresponding associations for progesterone [18
]. Their results are consistent with our findings. The first one was conducted in 141 pregnant Greek women and the authors found no association between pregnancy estradiol levels and maternal dietary fat (total, MUFA, PUFA or SFA), measured using a food frequency questionnaire (FFQ) in pregnancy week 26 [21
]. The second study which was conducted in 270 American women, did not find any association between pregnancy estradiol or progesterone levels and maternal dietary fat (total, MUFA, PUFA or SFA) estimated using a FFQ either on gestation weeks 16 and 27 [18
]. The third study is the Japanese study referred to above and they also reported no association between total fat, MUFA, PUFA or SFA and estradiol levels on gestation week 29 [14
To our knowledge no other study has evaluated the impact of gestational weight gain using repeated measurements of body weight during pregnancy on maternal plasma estradiol or progesterone levels. Still, in accordance to our findings, Lagiou et al reported no effect of total gestational weight gain on estradiol levels on gestation week 16 and 27 in a group of American women [18
], and a very small effect of progesterone [18
]. Furthermore, a recent study reported no association between total pregnancy weight gain and maternal umbilical cord estrogen levels [22
]. Thus, although pregnancy weight gain has been linked to an increased breast cancer risk [15
] the limited available data does not support a major role for maternal estradiol levels for this association. However, pregnancy progesterone levels were positively associated with weight gain during pregnancy. Still, this finding should be confirmed in other populations since we made multiple comparisons and our finding may be due to chance. Our animal studies indicate that other hormones e.g. leptin may be interesting in this context [15
Two earlier studies have reported a positive associate between maternal alcohol intake and estradiol levels [14
], while another study reported no association [24
]. We did not adjust for alcoholic beverages in our study, since none of our women reported that they drank alcohol during pregnancy.
The strengths of our study are the prospective design and our repeated measurements of estradiol and progesterone levels and dietary fat intake from all three trimesters. Furthermore, in comparison to earlier studies in Caucasian women [18
] we used food diaries to assess dietary fat intake. The food diaries may be advantageous for our purpose since they are developed to capture "current diet" compared to food frequency questionnaires which often estimate "usual diet" i.e. dietary habits during a longer time period, for instance one year before the dietary assessment. Body weight was also measured during standardised conditions. Limitations are that we cannot exclude some misclassification of dietary fat intake due to the fact that dietary intake was self-reported. Still such errors are likely not associated with estradiol or progesterone levels. Furthermore, we only measured the pregnancy hormone estradiol, and thus we do not know if our associations described are similar for the other estrogens such as estriol and estrone.
Our women were not randomly recruited. However, we have earlier concluded that this cohort is likely not very different regarding characteristics such as gestational weight gain and infant characteristics from other healthy pregnant women in Sweden and other Western countries [19
]. The magnitudes of the increase in pregnancy estradiol and progesterone levels were also similar to earlier reports [10
], and in our study nulliparous women had higher estradiol levels than parous women which is in agreement with findings from other populations [25
]. Furthermore, our women's diet consisted of 31%, 11%, 4.5%, 13%, 1% and 3.5% of energy from total fat, MUFA, PUFA, SFA, n-3 PUFA and n-6 PUFA, respectively, and these figures agree very well with a large Swedish cohort of 49 000 pre-menopausal women for total fat, MUFA, PUFA, and SFA [26
] as well as for n-3 PUFA and n-6 PUFA (unpublished data).