RESULTS

3.1 Study population

The flowchart of participant selection is depicted in Figure S2. A total of 93 women were included in the analyses. Table S2 shows participant characteristics at baseline and periconceptional maternal dietary intake of total energy, PEI-UPF, macronutrients and its compounds.

3.2 Ultra-processed foods and total energy

First, we investigated associations between periconceptional maternal intake of UPF and first-trimester imaging markers of utero-placental vascular development. All models showed higher PEI-UPF is associated with increased density of vascular branching for bifurcation points and crossing points in the uPVS but not with absolute morphologic development, see Table 2. Model 3 shows a statistically significant association between PEI-UPF and density of bifurcation points [β=0.465, 95%CI=0.148;0.782, p-value=0.006].
Next, we investigated associations between the periconceptional maternal intake of total energy (kcal/day) and first-trimester imaging markers of utero-placental vascular development. Total energy intake was negatively associated with density of bifurcation points in the uPVS [β=-0.053, 95%CI=-0.101;-0.004, p-value=0.039], see Table S3. We found no associations with the imaging markers of absolute morphologic utero-placental vascular development.

3.3 Macronutrients

We investigated the intake of separate macronutrients and observed positive associations between the intake of carbohydrates (g/day) and first-trimester development of uPVV [β=0.017, 95%CI=0.001;0.032] and imaging markers of absolute morphologic development: uPVS end points [β=0.286, 95%CI=0.062;0.511], bifurcation points [β=0.004, 95%CI=0.003;0.006], vessel points [β=0.772, 95%CI=0.137;1.408] and total length [β=0.700, 95%CI=0.106;1.295], all p-values<0.05, see Table 3 (only model 3 is shown). No associations were found between carbohydrate intake and imaging markers representing density of vascular branching. Further, we observed positive associations between the periconceptional maternal intake of mono-/disaccharides and first-trimester development of uPVV [β=0.021, 95%CI=0.003;0.039] and imaging markers of absolute morphologic development: uPVS end points [β=0.336, 95%CI=0.072;0.603], bifurcation points [β=0.380, 95%CI=0.057;0.703], crossing points [β=0.257, 95%CI=0.011;0.502], vessel points [β=0.916, 95%CI=0.170;1.662] and total length [β=0.856, 95%CI=0.158;1.554], all p-values<0.05, but not for the intake of polysaccharides, see Table 3. We found no associations between carbohydrate intake and imaging markers representing density of vascular branching.
These analyses were repeated for the total intake of fats and proteins and their respective compounds, see Table S4 (only model 3 shown). We observed no associations between the periconceptional maternal intake of fats, proteins and their compounds and first-trimester imaging markers of utero-placental vascular development in any of our models.

3.4 Dietary patterns

Using the PCA analysis we identified three distinct dietary patterns. The first dietary pattern is associated with higher intake of fresh and processed meats, cereal products, potatoes, eggs, cakes and sauces, and is therefore referred to as a Western dietary pattern, which explains 16.2% of the variance. The second dietary pattern is associated with higher intake of vegetables, fruits, fish and shellfish, nuts and seeds, vegetable oils and soy- and other plant-based meat-/dairy substitutes. Accordingly, this dietary pattern is referred to as a Mediterranean dietary pattern, which explains 11.7% of the variance. The third dietary pattern is associated with higher intake of sugar and confectionary, savoury snacks, milk and dairy products, soft drinks, fruit/vegetable juices and butter. We refer to this last dietary pattern as a Snack dietary pattern, which explains 8.4% of the variance. A nutritional overview of the dietary patterns is depicted in Table S5.
We investigated associations between periconceptional maternal adherence to the three dietary patterns and imaging markers of utero-placental vascular development. There are no associations between adherence to the Western or Mediterranean dietary patterns and imaging markers of utero-placental vascular development in all three models, see Table 4. Model 2 shows (borderline-) statistically significant positive associations between adherence to the Snack dietary pattern and first-trimester development of uPVV and imaging markers of absolute morphologic development: uPVS end points, bifurcation points, crossing points, vessel points and total length. Model 3 shows similar effect estimates with wide confidence intervals and no statistically significant associations, see Table 4.