DISCUSSION

4.1 Main findings

This study shows a 10%/day higher maternal UPF intake is positively associated with first-trimester density of vascular branching (bifurcation points: β = 0.465 √n, p = 0.006). The intake of carbohydrates, in particular of mono-and di-saccharides, is associated with increased absolute morphologic development of the first-trimester utero-placental vasculature, but not with density of vascular branching. We established positive associations between adherence to a Snack dietary pattern and absolute morphologic development of the utero-placental vasculature in model 2, but no clear associations in model 3. A graphic summary of the results is depicted in Figure 4.

4.2 Strengths and Limitations

Strengths of our study are the use of validated FFQ’s to study periconceptional nutritional intake, application of the validated Goldberg cut-off to exclude cases with unreliable nutritional intake and the use of EPIC project-based food groups to categorize food items for the identification of the dietary patterns 24-26.
The prospective observational study design allowed us to collect a comprehensive set of patient characteristics, which we used as covariates in advanced statistical models to minimize the effect of confounding bias. Although we cannot exclude the possibility of residual bias from any unknown factors, the direction of the effect estimates is similar in all three models. Therefore we consider it unlikely our findings result from residual confounding.
We used a unique longitudinal data collection of power Doppler ultrasounds in the first trimester of pregnancy to perform validated uPVV measurements (ICC above 0.80 and relative differences of less than 20%) 19 and subsequently generate the uPVS, our imaging markers of utero-placental vascular development20.
Recruitment from an academic hospital might affect the generalizability of our results. Although our study population contains a relatively high number of IVF-ICSI pregnancies and participants have a higher risk of developing pregnancy complications 21, we believe it is unlikely the direction of the associations will be different in the general population.
We performed multiple statistical analysis, which raises some concern about multiple testing. However, previously uPVV and uPVS characteristics were highly correlated and are inversely correlated with the density of vascular branching20. Therefore, the positive correlations between the effect estimates of uPVV and uPVS characteristics and the inverse correlation with density of vascular branching found in the present study can be interpreted as internal validation.

4.3 Interpretation (in light of other evidence)

4.3.1 Ultra-processed foods

So far, no studies have investigated associations between the consumption of UPF and placental development. High UPF intake is associated with increased total energy intake, an overall unhealthy dietary pattern and an increased risk of obesity 15, 29-31. UPF intake is negatively associated with embryonic growth and most studies on the effect of UPF-rich diets on birthweight suggests high UPF intake is associated with lower birth weight14, 32-35. In the present study, UPF is positively associated with first-trimester density of vascular branching, which has previously been associated with decreased embryonic and foetal growth and lower birth weight percentiles36, 37. Associations with UPF persist in model 3, which implies our findings are not solely mediated by higher energy content in UPF. We propose a high UPF exposure induces oxidative stress resulting in aberrant first-trimester vascular development and a compensatory increased density of vascular branching. This way the first-trimester utero-placental vascular development might act as a mediator in the relation between periconceptional maternal intake and pregnancy outcomes. This mechanism is substantiated by previous research on associations between UPF intake, oxidative stress, first-trimester placental development and prenatal growth21, 22, 38-42.

4.3.2 Carbohydrates

In our study, periconceptional maternal carbohydrate intake, more specifically of mono-and disaccharides, is positively associated with volumetric and absolute morphologic utero-placental vascular development in the first trimester. Carbohydrates seem to affect a different aspect of the utero-placental vascular development than UPF, which likely results from the high heterogeneity of the type and amounts of carbohydrates in UPF-rich foods.
During the first trimester, the placental microenvironment is hypoxic43. Rather than on oxidative phosphorylation, the placenta relies on glycolysis for energy production, supported by the rich supply of glucose from the endometrial glands 44. With increased carbohydrate intake maternal serum glucose concentration rises and subsequently the expression of glucose transporter 1 (GLUT1) increases. GLUT1 is an important cargo molecule of extracellular vesicles that augment decidualization, stimulate angiogenesis, and modulate trophoblast differentiation in the endometrial stroma45. The physiological interplay between serum glucose and placental development is substantiated by previous research indicating an insufficient glucose metabolism during gestation results in placental aberrations and foetal growth restriction (FGR)46-50.

4.3.3 Fats and Proteins

Multiple studies suggest fatty acids to play a regulatory role in the angiogenesis in tumours and in various organs, including the placenta51-55. The contradictory pro- and antiangiogenic effects of individual PUFAs might explain why we did not find any associations between the periconceptional intake of fatty acids and utero-placental vascular development 51, 52.
Not much is known about dietary intake of proteins and placental development. Recent studies suggest low-protein diet is associated with placental morphologic disruption and foetal growth restriction and found maternal protein intake positively associated with prenatal growth56, 57. Our findings suggest it is unlikely associations between protein intake and prenatal growth are mediated by utero-placental vascular development.

4.3.5 Dietary patterns

The Snack dietary pattern in this study is characterized by excessive consumption of food items with high carbohydrate content. This might explain why we found a (borderline-) statistically significant positive associations between adherence to the high snack pattern and first-trimester density of vascular branching in model 2 similar to the associations with carbohydrates. However, the snack dietary pattern also includes food items with high energetic value but relatively low carbohydrate content, like butter. The heterogeneity in the Snack dietary pattern and the small sample size might explain why the positive associations in model 3 were not statistically significant.

4.4 Implications

The intake of UPF seems to impair first-trimester utero-placental vascular development. These findings are not substantiated by associations with related macronutrients and dietary patterns as increased maternal intake of carbohydrates, within healthy quantities, seems to promote first-trimester utero-placental vascular development. Associations between UPF intake and placental vascular development could reflect the effects of ‘ultra-processing’ techniques in the food industry, but likely result from a coinciding lack of micronutrients in these food items.58, 59 UPF-rich diets are associated with deficiencies in folate and vitamin B12, important micronutrients involved in the 1-carbon metabolism. Derangements in the 1-carbon metabolism are associated with hyperhomocysteinemia and oxidative stress, which causes aberrant development of the utero-placental vasculature, associated with placenta-related complications.60