Strength and limitation
To our knowledge, this is the first large-scale study conducted in Japan
with meticulous attention to data collection. Therefore, the results of
this study are considered representative of the general pregnant
population in Japan.22,23 Randomised controlled
intervention trials are more likely the best study design; however, it
is impossible to conduct long-term controlled trial examining intake of
overall diet as exposure.
Although the present results were not obtained from a randomised
controlled study, a prospective birth cohort study can prove the
correlation between outcome and exposure assessed before the onset of
outcome. Nevertheless, this study also has potential limitations. First,
we do not have information on the diagnostic methods of gynaecological
complication such as endometriosis, uterine myoma, and adenomyosis.
Moreover, of the 3,249 women, the number of women who had active
endometriosis during their pregnancies and whether these complications
were treated before pregnancy were unclear, because information was
based on self-reported questionnaire. Information on these complications
in patient’s medical records was insufficient, and we did not utilise
past medical records of the participating women. In Japan,
gynaecologists usually use transvaginal ultrasonography to examine the
conditions of the uterus and ovaries at their first visit. A previous
study reported that the diagnostic accuracy of endometriosis by vaginal
ultrasonography is almost 90% regardless of its
phenotype.24 With regard to obstetric outcomes, data
were collected prospectively from the medical records by trained
research coordinators and medical doctor at the subject institution for
all puerperal patients. Therefore, we expect that the maternal
gynaecologic condition before pregnancy and outcome of delivery to be
accurate. Second, although we accounted for some confounding factors in
large portions of the questionnaire, unknown factors that might have
affected the occurrence of PTB, LBW, or HDP might have existed.
Third, the DII score of each study participant was calculated using only
the FFQ as the basis of the Japanese lifestyle, and JECS participants
were Japanese women. Therefore, the DII of this study is not validated
yet, and our results may not be applicable to other ethnicities. Fourth,
because the FFQ used for calculating the DII score in this study was
based on the self-reported information of women during their first
trimester, recall bias may be possible, as participants might have had
morning sickness and were asked to recall their diet content before
pregnancy.