Association of Dietary Factors With Gastrointestinal Symptoms During Pregnancy and Pregnancy Outcome: The role of dietary fiber and vegetarian and gluten-free diet

This thesis was aimed to investigate associations (Studies I and II) between dietary factors, especially dietary fiber (DF), and other modifiable lifestyle factors with common gastrointestinal symptoms, nausea or vomiting in pregnancy (NVP), heartburn, and constipation. Associations with pregnancy outcome were moreover investigated (Study II). The associations of special diets (vegetarian or gluten-free) with pregnancy and pregnancy outcome were investigated in Studies III and IV.

Gastrointestinal symptoms, such as NVP, heartburn, and constipation are common during pregnancy. Nausea affects up to 85% and vomiting approximately 50% of pregnant women. These symptoms impair quality of life and have a negative impact on the daily family and social lives of pregnant women. The pathophysiology of NVP is unknown, but it has been suggested to be multifactorial and to include hormonal changes. The incidence of another common gastrointestinal symptom, heartburn, has been reported to vary between 17% and 80%. Pregnancy-related risk factors for heartburn have been found to be multiparity, pre-existing heartburn, and advanced gestational stage. Constipation symptoms are most prevalent in the first and second trimesters and decrease in the third. The prevalence rate of constipation has been reported to vary between 11% and 40%. Risk factors for constipation include multiparity, lack of exercise, low fiber intake, insufficient fluid intake, iron supplementation, and previous cesarean section (CS). Dietary fiber (DF) is known to alleviate constipation and heartburn in non-pregnant population, but in pregnancy the evidence of these is scanty, likewise the research on the connection between DF and NVP. Ascertaining the associations of DF and other lifestyle factors with gastrointestinal symptoms (NVP, constipation and heartburn) during pregnancy would be important to alleviate symptoms in pregnant women and to provide suitable nutritional support.

Special diets e.g. gluten-free diet (GFD) and vegetarian diets are becoming increasingly popular, especially among the young and among females. GFD is also becoming popular among those without diagnosed celiac disease (CD). There are studies on CD patients, but fewer on those following a gluten-free diet without a CD diagnosis. Whether GFD is appropriate during pregnancy remains unclear. The recommendations regarding vegetarian diet during pregnancy are contradictory. GFD has been suggested to be associated with disturbed fetal growth, but according to most studies, pregnancy outcome has been favorable.

The primary aims of Studies I and II were to detect associations of DF intake and other lifestyle factors with NVP, constipation, and heartburn, and to test the hypothesis that higher DF consumption is associated with fewer gastrointestinal symptoms. As a secondary aim the association of DF intake with pregnancy outcome was studied. In Studies III and IV the association of special diets, vegetarian diet, and GFD, pregnancy and pregnancy outcome were evaluated. In Study III, the hypothesis of an association between lower birthweight of the newborn and vegetarian diet was tested.

For Studies I (n=188) and II (n=173), participants completed self-administered questionnaires before pregnancy, at trimesters I-III, and after delivery. Data on bowel function, occurrence of NVP, DF intake and lifestyle characteristics were collected from questionnaires. Data on deliveries and perinatal outcomes were collected from the Tampere University Hospital records. In Studies III and IV the data were collected retrospectively from the Tampere University Hospital records. The study period was from January 2015 to April 2021. Diet was a self-reported variable in the records. In Study III 150 women following a vegetarian diet and 300 omnivores constituted a control group. In Study IV all women following a GFD were included (n=370). Of these 21% had CD and 79% followed a GFD without CD diagnosis. The control group included 456 omnivores. Outcomes of interest were compared between special diet groups (GFD and vegetarian) and omnivores.

In Study I women suffering from NVP (n = 91, 48 %) consumed significantly more fiber derived from cereal products (insoluble dietary fiber=IDF) and total fiber prior to pregnancy than did women without NVP (n=97, 52 %). No significant differences were found between the study groups in the first trimester of pregnancy. In both groups, total fiber intake and intake of fiber derived from fruit and vegetables (soluble dietary fiber=SDF) increased during the first trimester. Thus, DF intake did not protect against NVP, but women with NVP had greater pre-pregnancy intake of fiber than those without NVP. However, women suffering from NVP were able to maintain their fiber intake. Women of normal weight (BMI <25 kg/m2) had significantly more NVP than did overweight women (BMI ≥25 kg/m2) (p=.007).

In Study II the women were grouped according to intake of fiber and liquids during pregnancy. A combination of low fiber and low fluid intake increased the risk of constipation during pregnancy. Heartburn increased significantly as pregnancy advanced, with the highest prevalence during the third trimester, but no associations with DF were seen. Low fiber intake (<22g/day) was associated with risk of combined adverse outcome (cesarean section (CS), premature delivery, and/or small newborn for gestational age (SGA).

In Study III the rate of SGA was lower and median birthweight was higher in the vegetarian group. This finding was contrary to the original hypothesis. Gestational diabetes (GDM) was more common in the vegetarian group, however the numbers of newborns large for gestational age (LGA) were comparable between the study groups. Induction of labor was more common, and the second stage of labor was longer in the vegetarian group. Preterm births (< 32+0 and < 37+0 gestational weeks) were more common in the control group. No differences were found in mean umbilical artery pH value, 1- and 5- minute Apgar scores or in number of newborns transferred to the neonatal intensive care unit and neonatal ward.

In Study IV, incidence of pregnancy complications did not differ between the groups. Induction of labor rate was higher, and the duration of labor was longer in the GFD group, especially among those on GFD without no CD diagnosis. However, no differences in the CS rates were found between the groups. Median birth weight was higher in the GFD group than among controls, but the incidences of SGA or LGA did not differ between the study groups.

In summary: DF was not associated with reduced NVP frequency and did not seem to protect against it. However, DF was tolerated well during NVP, and this finding could be of use when diet counseling women suffering from NVP. Sufficient fiber and liquid intake may protect against pregnancy-associated constipation, and to some extent, DF was associated with improved combined pregnancy outcome (CS, SGA and preterm delivery). According to the patient material of this thesis, both vegetarian and gluten-free diets may be considered safe during pregnancy; they showed no association with increased risk of pregnancy- or neonatal complications.