**Dietary and Nutrient Intake, Eating Habits, and Its Association with Maternal Gestational Weight Gain and Offspring's Birth Weight in Pregnant Adolescents**

**Reyna Sámano 1,2,\*, Hugo Martínez-Rojano 3, Luis Ortiz-Hernández 1,4,\*, Oralia Nájera-Medina 1,4, Gabriela Chico-Barba 2,5, Estela Godínez-Martínez 2, Ricardo Gamboa <sup>6</sup> and Estefanía Aguirre-Minutti <sup>7</sup>**


**Abstract:** Pregnant adolescents' diet and eating habits are inadequate; however, their association with gestational weight gain (GWG) is uncertain. We aimed to analyze whether there is an association between dietary and nutrient intake and eating habits with GWG among pregnant adolescents and their offspring's birth weight. A longitudinal study was performed with 530 participants. We assessed GWG and applied several tools, such as a food frequency questionnaire and 24-h recall, to obtain dietary and nutrient intake and eating habits. The birth weight of adolescents' offspring was registered. Later, we performed crude and adjusted Poisson models. The mean age was 15.8 ± 1.3 years. Of all food groups, the lowest frequency of adequate intake corresponded to vegetables (7%) and legumes (10.2%). Excessive (36.8%) and insufficient (40.9%) GWG were observed. Pregnant adolescents with inadequate legumes intake increased the probability of excessive GWG: (PR 1.86 95% CI 1.00–3.44). Cereals and grains were positively associated with GWG: (PR 1.65, 95% CI 1.18–2.29). Energy, macronutrient intake, and eating habits were not associated with GWG. Offspring's small gestational age (SGA) increased when pregnant adolescents had inadequate sugarsweetened beverages intake: PR (1.58, 95% CI 1.01–2.49) and when pregnant adolescent watched television (TV). In our sample of Mexican adolescents, dietary and nutrient intake and eating habits were inadequate. Excessive dietary intake from cereals, grains, and animal-sourced foods along with insufficient legumes were associated with excessive GWG. Watching TV while adolescents ate was associated with the birth weight of the offspring.

**Keywords:** adolescent pregnancy; gestational weight gain; energy intake; food groups; dietary habits; Mexico

#### **1. Introduction**

Adolescent pregnancy represents a global public health concern. Nearly 20% of adolescents from low and middle-income countries give birth [1,2]. They have a higher frequency of adverse outcomes such as preterm birth, small-for-gestational-age (SGA), and increased neonatal and maternal mortality risk than pregnant adults [3–5]. Gestational weight gain

**Citation:** Sámano, R.; Martínez-Rojano, H.; Ortiz-Hernández, L.; Nájera-Medina, O.; Chico-Barba, G.; Godínez-Martínez, E.; Gamboa, R.; Aguirre-Minutti, E. Dietary and Nutrient Intake, Eating Habits, and Its Association with Maternal Gestational Weight Gain and Offspring's Birth Weight in Pregnant Adolescents. *Nutrients* **2022**, *14*, 4545. https://doi.org/10.3390/ nu14214545

Academic Editors: Louise Brough and Gail Rees

Received: 16 September 2022 Accepted: 21 October 2022 Published: 28 October 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

(GWG) has been associated with both short-term and long-term consequences, such as anemia and preeclampsia. In the short-term, excessive GWG is associated with adverse newborn outcomes, including preterm birth, large-for-gestational-age, and macrosomia. In the long term, it is associated with significant weight retention after pregnancy and excess body weight later in the mother's life [6]. Therefore, pregnant adolescents need more health services, which are associated with higher costs to provide them with prenatal and postnatal care [7,8]. In addition, although pregnant adolescents have a similar proportion of excessive gestational weight gain (GWG) compared to adults, the former have a higher total GWG in kilograms (kg) [6].

Several countries in sub-Saharan Africa, Latin America, and Asia have moved from low-income to middle-income status, which is accompanied by lifestyle changes, including increased food security, dietary transitions, and reduced physical activity. These changes have led to modifications in maternal diets before and during pregnancy, affecting GWG patterns and the overall pregnancy experience for women in these regions [9–11]. For example, a study from Tanzania reported that, according to Institute of Medicine (IOM) guidelines, 42.0%, 22.0%, and 36.0% of pregnant adults were characterized as having inadequate, adequate, and excessive GWG, respectively [12].

Another problem is that, as pregnant adolescents' linear growth has not reached its peak, their nutrient requirements are higher than adult women [13]. Nevertheless, studies about the dietary patterns of pregnant adolescents are scarce [14]. Pregnant adolescents tend to have low iron intake (28% for Recommended Dietary Allowances-RDA) [15,16]. Moreover, less than 30% have good adherence to folate supplementation [15]. The average intake of calcium in pregnant adolescents from the USA [16], Brazil, and Mexico [15,17,18] ranges from 400 to 900 mg/day, which does not meet the recommended intake of 1000–1300 mg/day [19]. This inadequate nutrient intake in pregnant adolescents can be linked to the low variety of food groups they consume [20,21]. At least 75% of pregnant adolescents who received antenatal care in a public hospital had low intake of vegetables and legumes, around 50% consumed more sweetened beverages than recommended, and 5–25% skipped supper or breakfast [22].

Dietary and nutrient intake and eating habits can potentially affect GWG, as energy and nutrients are necessary for tissue accretion [23]. A few studies have been conducted on adult women to analyze that relationship [24–26]. However, systematic reviews about this topic included only adult women [27–29]. We could not find any studies on the association of dietary and nutrient intake and eating habits with GWG and offspring's birth weight in pregnant adolescents [29]. Nevertheless, evidence derived from Rumanian adult women showed a positive association between a high-fat diet and excessive GWG and a negative association with a high-protein diet [26]. In addition, adult pregnant women who consume foods from the Mediterranean diet (legumes, vegetables, nuts, olive oil, and whole cereals) have high odds of having a lower [24] or adequate GWG [24,25] and a lower risk of having a small-for-gestational-age newborn when eating fruits and vegetables [30–32]. Energy intake has been associated with GWG, while macronutrients have not [28]. This paper aimed to analyze whether there is an association between dietary and nutrient intake and eating habits and GWG among pregnant adolescents and their offspring's birth weight.

#### **2. Materials and Methods**

We conducted a longitudinal study with pregnant adolescents aged 11–19 years who received antenatal care at the Instituto Nacional de Perinatología (INPer) in Mexico City. The inclusion criteria were being a woman primigravida with single pregnancy and without chronic diseases. In addition, adolescents with drug addictions, vegans or vegetarians, and those who had a newborn with congenital malformations or stillbirth were excluded.

Six hundred and fifty adolescents were invited to participate in the study. Forty teenagers did not agree to participate, 38 accepted but did not arrive at any assessment, 25 did not deliver at the INPer, 15 cases were incomplete, and two neonates died at birth. There were 530 cases with complete data. During the first visit, we obtained signed

consent from adolescents and their parents/guardians as well as sociodemographic data. Anthropometric measurements and dietary assessment were conducted. We obtained maternal and neonatal outcomes from the last consultation from the medical records.

#### *2.1. Dietary and Nutrient Intake, and Eating Habits*

We assessed food group consumption to describe dietary intake using a semi-quantitative food frequency questionnaire (FFQ) [33]. Intake of nine food groups was measured. The dietary guidelines for the Mexican population were used as criteria. These guidelines present the following food groups: vegetables; fruits; legumes; cereal and grains; meat, cheese, and eggs (herein, "animal-source foods"); fats and oils; milk and yogurt; table sugar; and sweetened beverages [34]. Participants reported their frequency of intake during the last trimesters. Because it is known that macronutrient intake remains relatively stable during pregnancy [35], one measurement in the second or third trimesters was obtained. The interviewers used food replicas and standard measuring cups, spoons, and glasses to improve serving size estimation. Later, we compared the number of servings consumed with the recommendations for the Mexican population [34]. The number of servings of each food group used as a reference can be reviewed in Appendix A. Adequate consumption was defined when the number of servings was met according to the recommendation. Inadequate consumption (excessive and insufficient) was when the participants ate more or fewer servings than the recommendation range.

Three 24-h dietary recalls were applied. Two were recorded on non-consecutive weekdays and another on weekends. The 24-h recalls were administered by personnel trained in the interview technique. The nutrient and energy intake were estimated using Nutrikcal® software. Later, the mean energy intake in kilocalories (kcal) was calculated. To measure participants' energy intake adequacy, we used the reference of the IOM (2005) [36,37]. We categorized energy intake adequacy as insufficient (<80%), adequate (80–119%), or excessive (>120%). The contribution of carbohydrates, proteins, and lipids to total energy consumption was estimated. The recommendations of the IOM were used as a reference to categorize the distribution of energy contribution of macronutrients [37].

Participants were asked about the following eating habits: their number of meals; frequency of skipping meals (never, 1–3 times, 4–5 times/week); with whom they ate their foods (alone, with family, and friends); where they ate (out of home, home); and what activities they did while eating (doing homework/household chores, watching TV or using a cellphone, or just eating). In addition, we inquired as to whether participants had modified their diet during pregnancy (if it was improving, was worse, or had no change).

#### *2.2. Anthropometric Data and Gestational Weight Gain*

In the first interview, the pre-pregnancy self-reported weight was obtained. The self-reported weight is an adequate proxy for pre-pregnancy weight [38,39].

All anthropometric measurements were performed according to Lohman's techniques [40]. Height was measured at the first antenatal visit using a stadiometer (SECA, Hamburg, Germany, model 208, accuracy 0.1 cm). We estimated the pregestational body mass index (pBMI) using the pregestational weight and height. Then, we classified pBMI with AnthroPlus® (World Health Organization, Geneva, Switzerland) according to percentiles: underweight <3rd, normal weight 3–85th, overweight 85–97th, and obesity ≥97th [41].

One or two weeks before delivery, we measured and recorded participants' body weight with a digital scale (TANITA, Tokyo, Japan, model BWB-800, accuracy 0.10 kg). This measure was considered the final gestational weight. The GWG was calculated from the difference between the final gestational weight and the pregestational weight.

The expected weight gain was calculated with the following equation [42]:

Expected weight gain = recommended weight gain for the first trimester + ((gestational age final—13.86 weeks) × (recommended weight gain rate in second and third trimesters)).

The recommendation of GWG rate for the first trimester was according to pBMI: low and normal weight 2 kg, overweight 1 kg, and obesity 0.5 kg. For adolescents in the second and third trimesters, these pBMI figures were low weight 0.51 kg, normal weight 0.42 kg, overweight 0.28 kg, and obesity 0.22 kg/week [43].

The gestational weight gain adequacy percentage was estimated using the recommendations of the US Institute of Medicine [43,44]. Finally, we categorized the GWG percentage as follows: inadequate (<90%), adequate (90 to <125%), and excessive (≥125%).

#### *2.3. Neonatal Outcomes*

The sex of the newborn was obtained from the neonatal clinical record. Gestational age was obtained by ultrasound and recorded in weeks and days. If the gestational age was ≤36.6 weeks we classified it as preterm, whereas if the gestational age was between ≥37 and ≤42 weeks this was considered at term.

Standardized personnel measured and recorded birth weight (g) with calibrated equipment (SECA 374, model "Baby and Mommy"; accuracy 0.1 g) and length at birth (cm) (stadiometer SECA 416; accuracy 0.1 cm). SGA was defined when birth weight was <10 percentile, normal birth weight as the neonate being between 10–90 percentile, and large for gestational age (LGA) as >90 percentile, according to the Intergrowth-21s criteria [45].

#### *2.4. Other Variables*

In an antenatal visit, trained personnel obtained information on sociodemographic characteristics such as chronological age, marital status, education, occupation, and socioeconomic level. Age was registered at the time of the survey in years and as a dichotomous variable (≤15 or ≥16 to 19 years). In addition, marital status was classified as cohabiting or single.

Education was reported by the pregnant adolescents and was considered as elementary school or less, middle school, and incomplete high school. In addition, we created a school dropout variable according to the school grade and chronological age for adolescents who were more than two years behind in educational training.

Occupation was classified as student or housewife. A questionnaire validated for the Mexican population was used to determine socioeconomic status [46]. In our sample only middle, low–middle, and low were observed.

The initiation of antenatal care and the gestational age at delivery were obtained through ultrasound and reported in weeks. Obstetricians registered maternal adverse outcomes during prenatal visits, and the information from the clinical records was obtained. Complications were identified and recorded in the following categories: gestational diabetes, pregnancy-induced hypertension, eclampsia/pre-eclampsia, and anemia [47,48].

#### *2.5. Statistical Analyses*

A descriptive analysis was performed, including percentages for categorical variables. For continuous variables, the Kolmogorov–Smirnov test was used to assess their distribution. The mean was estimated for variables with normal distribution, and the median was obtained for those with a non-normal distribution. Next, we compared the prevalence of outcomes between the categories of nutrition, energy intake, and eating habits. The chi-square test was estimated to assess whether significant differences between categories existed. When the significance of the difference was *p* ≤ 0.250, the variable was considered for the next step.

Poisson regression models were calculated to determine the association of outcomes (GWG and offspring's birth weight) with predictors of interest (nutrients, energy intake, and eating habits). We estimated separate models for inadequate and excessive GWG. For this reason, dummy variables were created for the GWG and birth-weight categories. For each outcome, three models were performed: M1, crude M2, adjusted by socioeconomic level, school drop-out, education, gynecological age, chronological age, and antenatal care; and M3, adjusted by the same variables included in M2 plus pBMI. The regression coefficients were transformed to prevalence ratios (PR).

When the cross-tabulation of two eating habits with the outcomes was estimated, the absence of any cases in certain cells was evident. Hence, these variables were not included in the regression analysis.

#### *2.6. Ethical Aspects*

This research was approved by the Institutional Ethics, Biosafety, and Research Committees from INPer (registration numbers 212250-49481, 212250-49541, and 2017-2-101, respectively). All adolescents and their guardians were informed of the study's objectives and procedures. Confidentiality was guaranteed by assigning an ID number during each participant's data collection and analysis. Written informed consent was obtained from adolescents and guardians.

#### **3. Results**

The mean age of the participants was 15.8 ± 1.3 years. Seventy percent of the adolescents were single, and the rest lived cohabiting with their partners. Most adolescents were homemakers (89%). Their socioeconomic status was low or very low. Three-quarters of the women had elementary education (74.7%). School dropout was experienced by 89.1%.

Gestational weight gain in pregnant adolescents was excessive in 36.6%, adequate in 26%, and insufficient in 37.4%. In addition, it was observed that 20.4% of newborns were SGA (<10th percentile) and 3.8% were LGA (>90th percentile).

The lowest frequency of adequate intake corresponded to vegetables, followed by legumes and animal-source foods (Figure 1). In contrast, the food groups that were eaten most frequently were table sugar, cereals and grains, and dairy foods. None of the nine food groups reached 50% recommended consumption coverage. In addition, 73% of the participants included less than three food groups in their diet.

**Figure 1.** Distribution of adequate intake from different food groups.

One-fifth of the adolescents had two or less meals. Dinner was the most skipped meal. Fifty-six percent skipped meals more than once a week. Fifty-one percent of adolescents watched TV while they ate, and 66% reported that their diet was better during pregnancy than pregestational (Appendix B).

Excessive GWG was more frequent among pregnant adolescents who did not consume legumes than those who consumed them (*p* = 0.023) (Table 1). The adolescents with high consumption of cereals and grains and animal-source foods had a higher frequency of excessive GWG (*p* ≤ 0.001) than those with low or normal consumption. Excessive and insufficient GWG were observed more frequently among pregnant adolescents who excessively consumed sugar-sweetened beverages compared to their counterparts who consumed them adequately (*p* = 0.030). The rate of small for gestational age neonates among mothers who consumed excessive sugar-sweetened beverages was higher than in those with a low intake (*p* = 0.066).


**Table 1.** Adolescents' gestational weight gain and offspring's birth weight according to dietary intake.

Percentages estimated by rows. SGA: small for gestational age. LGA: large for gestational age. *p*-value determined by Pearson's Chi-Square.

The energy intake of the participants was 2022 ± 657 kcal. The distribution of macronutrients of total energy was as follows: 102 ± 34% energy adequacy, 53 ± 8% carbohydrates, 16 ± 5% proteins, and 31 ± 8% lipids. Table 2 shows that none of the macronutrients and energy intake had statistical significance with respect to the maternal GWG and the birth weight of their offspring.


**Table 2.** Adolescent's gestational weight gain and offspring's birth weight according to energy and macronutrients intake.

Percentages estimated by rows. SGA: small for gestational age. LGA: large for gestational age. p-value determined by Pearson's Chi-Square.

The frequency of GWG and the newborn weight categories did not differ according to eating habits (Table 3).




**Table 3.** *Cont.*

Percentages estimated by rows. SGA: small for gestational age. LGA: large for gestational age. None of the variables was statistically significant. p-value determined by Pearson's Chi-Square.

Pregnant adolescents with insufficient consumption of legumes had a greater probability of excessive GWG than participants with adequate intake (Table 4). Insufficient consumption of cereals and grains was associated with a higher probability of insufficient GWG. In contrast, the excessive consumption of cereals and grains demonstrated a high probability of excessive GWG. In addition, excessive sugar-sweetened beverage consumption was associated with a higher probability of having a small-for-gestationalage newborn.

**Table 4.** Poisson regression models of adolescents' gestational weight gain and offspring's birth weight as outcome and dietary intake as predictors.



**Table 4.** *Cont.*

*p*-value determined by Poisson regression. PR: prevalence ratio; CI: confidence interval; SGA: small for gestational age. LGA: large for gestational age. M stands for Model: M1, crude; M2: adjusted by socioeconomic level, school drop-out, education, gynecological age, chronological age, and antenatal care; M3, adjusted by the same variables included in M2 plus pBMI. In bold are present the significant results.

Lipids intake and eating habits did not have any association with GWG or newborn weight (Table 5).

**Table 5.** Poisson regression models of adolescents' gestational weight gain and offspring's birth weight as outcome and lipids intake and eating habits as predictors.



**Table 5.** *Cont.*

\* *p* < 0.050. p-value determined by Poisson regression. PR: prevalence ratio; CI: confidence inter-val; SGA: small for gestational age. LGA: large for gestational age. M stands for model: M1, crude; M2, adjusted by socioeconomic level, school drop-out, education, gynecological age, chronological age, and antenatal care; M3, adjusted by the same variables included in M2 plus pBMI. In bold are present the significant results.

#### **4. Discussion**

The results of the present research show that unhealthy eating habits and nutrient intake are frequent in pregnant adolescents. The participants in our study had excessive intake of cereal and grains, animal-source foods, table sugar, and sugar-sweetened beverages, and insufficient consumption of legumes and vegetables. For example, most did not consume the recommended servings of vegetables (93.0%), legumes (89.8%), or sugar-sweetened beverages (79.8%), among other foods, and showed poor eating habits such as skipping meals (56%), eating alone (20.1%), and carrying out activities (61.3%) while they ate.

The present study showed associations between insufficient legumes and excessive cereal and grains consumption and excessive GWG. Meanwhile, sugar-sweetened beverages consumption and using cell phones/watching TV while eating had associations with birth weight.

#### *4.1. Dietary and Nutrients Intake and Eating Habits*

Although most of our participants (67%) reported that their diet had improved during the pregnancy, they did not have adequate dietary and nutrient intake or eating habits. Our participants' dietary intake was low in legumes and vegetables and excessive in sweetenedsugar beverage consumption, which is common in most age groups [21,22,49–51]. More than 70% of pregnant adolescents did not eat more than three food groups in their meals. Only fifty percent of Mexican pregnant adolescents in the present study had adequate consumption of energy and macronutrients; similar data has been found in pregnant adults [50]. This dietary pattern could be a risk factor for developing non-transmissible chronic diseases [52,53] and micronutrients deficiencies [54].

More than half of the participants skipped meals, watched TV, or used cell phones while eating. Youth exposed to screens habitually consume ultra-processed foods [55]. Watching TV has been associated with the development of excess weight, obesity, and cardiometabolic risk in the adolescent population [55,56].

#### *4.2. Gestational Weight Gain*

Our study reported that excessive gestational weight gain in adolescent pregnant women occurred in 36.6% and was insufficient in 37.4%. This highlights that there is currently a higher probability in pregnant adolescents of not meeting the recommendations GWG of the IOM. This is similar to the findings of Santos et al. in adolescent Brazilians, which showed 37% and 33% insufficient and excessive GWG, respectively [57]. A significant rate of insufficient and excessive GWG was observed in our sample of pregnant adolescents. This population likely experiences nutritional transitions and reduced physical activity [58], which may lead to changes in maternal diets before and during pregnancy, thereby affecting GWG patterns [59].

#### *4.3. Dietary and Nutrient Intake, Eating Habits, and GWG*

Insufficient intake of legumes was associated with a higher risk of excessive GWG, even after adjusting for pBMI. Among pregnant adults from Spain and South Africa, the consumption of diets that include legumes [24,60] has been associated with lower GWG. Legumes have nutrient content (high in fiber and antioxidants but low in fat) that can help with keeping a healthy weight [60]. Nevertheless, there is little information on this topic in adolescent pregnant women.

We observed that a higher intake of cereal and grains was associated with excessive GWG. However, with animal-source foods the association was lost when the models were adjusted for pBMI, showing that GWG was affected more by pBMI than by diet in our group of pregnant adolescents. In addition, it has been documented that pBMI is a better predictor of GWG than other variables such as food consumption [61].

Watching TV is a risk factor for developing obesity [56] because it is a sedentary behavior related to higher consumption of ultra-processed foods. Our study found that adolescents who ate while watching TV were associated with LGA neonates. The mechanisms that explain this relationship may be related to maternal consumption of foods with high energy density [62].

We did not find an association between macronutrients and GWG. Our study coincides partially with a previous systematic review that reported macronutrient intake to not be associated with GWG [28]. Hence, it is challenging to estimate macronutrients, which could affect the possible association between GWG and the birth weight of adolescent's offspring. Nevertheless, the scientific evidence establishes that a whole diet and the foods that make it up can be more relevant than individual nutrients to GWG [24,25]. In this sense, in the present study, we observed that legumes, cereals, and grains were associated with GWG. However, not all foods or macronutrients were associated with GWG.

#### *4.4. Dietary Intake, Eating Habits, and Birth Weight*

Sugar-sweetened beverages consumption was associated with SGA. There is insufficient evidence to identify possible causal mechanisms to explain the association between maternal consumption of sugar-sweetened beverages and birth weight outcomes [63–65]. Therefore, our findings should be interpreted with caution. However, we believe that an inadequate maternal diet is likely to be associated with the birth weight of their offspring [66].

None of the maternal nutrient intakes were associated with birth weight in our sample, similar to GWG. Data from observational studies indicate that certain dietary habits and patterns during pregnancy have no consistent associations with birth weight. Maternal lack of all foods in their diet is relevant, as it has been demonstrated that the whole diet, beyond individual nutrients, can influence birth weight. However, if most participants do not meet a recommended diet the birth weight effect would likely be attenuated, as reported in pregnant adults [67]. Nonetheless, we did not find scientific evidence to support this hypothesis in the studied group of pregnant adolescents

#### *4.5. Limitations and Strengths*

Using the IOM references, we observed a high frequency of excessive and insufficient GWG. However, it is unknown whether the IOM reference is adequate for Mexican pregnant adolescents, which is a public health concern as we currently do not have any official parameters to evaluate GWG in adolescent pregnancy. The number of LGA neonates was small (n = 20). Therefore, certain estimates were imprecise.

Although our sample was for convenience considering the inclusion criteria, we must consider that INPer is a national reference center that provides prenatal control for women from several regions of Mexico. Moreover, our study had a prospective follow-up.

#### **5. Conclusions**

To the best of our knowledge, this is the first study to analyze the association between maternal dietary and nutrient intake and eating habits and GWG and birth weight in a sample of pregnant adolescent–baby dyads. Furthermore, we show that when certain elements of the diet are inadequate, optimal maternal and neonatal outcomes can be limited. In addition, all models were adjusted by pBMI in order to control its confounding effect to a certain extent.

Pregnant adolescents need to know the relationship between the components of the diet and GWG to improve their eating habits. Health personnel should promote the consumption of a healthy diet according to the individual requirements of pregnant adolescents and promote avoidance of inappropriate eating habits while considering sociocultural and economic characteristics.

The consumption of adequate amounts of legumes, cereals and grains, animal-sourced foods, and sugar-sweetened beverages is part of the dietary guidelines because their consumption is related to health outcomes such as weight gain and diabetes. However, our study provides evidence of other health outcomes, such as GWG and birth weight in the studied group of pregnant adolescents, which could be affected by eating habits. Our results can inform the development of clinical and nutritional guidelines for antenatal control aimed at preventing complications and promoting healthy pregnancy.

**Author Contributions:** Conceptualization, R.S. and H.M.-R.; Data curation, R.S., L.O.-H., R.G., E.A.- M. and O.N.-M.; Formal analysis, R.S., L.O.-H. and R.G; Funding acquisition, R.S.; Investigation, H.M.-R., G.C.-B.; E.G.-M., R.G. and E.A.-M.; Methodology, R.S., O.N.-M., G.C.-B., E.G.-M., E.A.-M. and H.M.-R.; Project administration, R.S.; Supervision, R.S. and L.O.-H. and H.M.-R.; Writing original draft preparation, R.S. and L.O.-H. and H.M.-R. Writing—review and editing, R.S., L.O.-H., G.C.-B., O.N.-M., E.G.-M., H.M.-R., R.G. and E.A.-M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was funded by the Instituto Nacional de Perinatología (registration numbers 212250-49541 and 2017-2-101) and by CONACyT (Registration number A3-S-40575).

**Institutional Review Board Statement:** The study was approved by the Institute National of Perinatology Ethics Committee (registration number 212250-49481 in October 2008, February 2014, 212250-49541 in February 2014, and 2017-2-101 on 10 April 2019) according to the basic principles of the Declaration of Helsinki.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available from the corresponding author upon reasonable request.

**Acknowledgments:** We extend profound thanks to all adolescent pregnant women and their parents/guardians for their participation and cooperation in this study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A**

Recommended intake and number of servings of food groups for adolescent mothers.


Academia Nacional de Medicina 2015. México (Fernández-Gaxiola et al. 2015). \* Median (p25–75).

**Appendix B**

GWG of adolescent mother and offspring birth weights according to sociodemographic characteristics (%).



Percentages estimated by rows. SGA: small for gestational age. LGA: large for gestational age. p-value determined by Pearson's Chi-Square.

#### **References**


Panel on the Definition of Dietary Fiber; Panel on Macronutrients. *Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids*; National Academies Press: Washington, DC, USA, 2005; ISBN 9780309085250.


**Merle Roeren †, Anna Kordowski †, Christian Sina and Martin Smollich \***

Institute of Nutritional Medicine, University Hospital Schleswig-Holstein, Campus Luebeck,

University of Luebeck, 23538 Luebeck, Germany

**\*** Correspondence: martin.smollich@uksh.de

† These authors contributed equally to this work.

**Abstract:** Choline is an essential nutrient that is involved in various developmental processes during pregnancy. While the general adequate choline intake (AI) for adults has been set at 400 mg/day by the European Food Safety Authority (EFSA), an AI of 480 mg/day has been derived for pregnant women. To date, the choline intake of pregnant women in Germany has not been investigated yet. Therefore, in this survey, the total choline intake from dietary and supplementary sources in pregnant women was estimated using an online questionnaire. A total of 516 pregnant women participated in the survey, of which 283 met the inclusion criteria (13 to 41 weeks of gestational age, 19–45 years). 224 (79%) of the participants followed an omnivorous diet, 59 (21%) were vegetarian or vegan. Median choline intake was 260.4 (±141.4) mg/day, and only 19 women (7%) achieved the adequate choline intake. The median choline intake of omnivores was significantly higher than that of vegetarians/vegans (269.5 ± 141.5 mg/day vs. 205.2 ± 101.2 mg/day; *p* < 0.0001). 5% (13/283) of pregnant women took choline-containing dietary supplements. In these women, dietary supplements provided 19% of the total choline intake. Due to the importance of choline for the developmental processes during pregnancy, the study results prove the urgent need for an improved choline supply for pregnant women.

**Keywords:** choline; pregnancy; vegan; vegetarian; omnivorous; adequate intake

#### **1. Introduction**

Choline is an essential nutrient being involved in many physiological processes in the human body. It is a constituent of the neurotransmitter acetylcholine and it modulates, as a precursor to the cell membrane components phosphatidylcholine and sphingomyelin, membrane integrity, transmembrane signaling, myelination, cell growth, and cell division [1]. Moreover, choline acts as a methyl group donor via the synthesis of s-adenosylmethionine [2], thereby essentially contributing to epigenetic methylation reactions, DNA stability, and cellular metabolism [3–5].

Correspondingly, higher phosphatidylcholine intake is associated with lower risk of incident dementia and better cognitive performance [6], and dietary choline intake has been suggested to play a relevant role in the prevention of cognitive decline [7,8], poststroke cognitive impairment [9], and poststroke depression [10]. Even more, choline has recently been found to be neuroprotective against prenatal alcohol exposure-related brain structure deficits in humans [11,12]. Respective effects of choline may, at least in part, be mediated by a functional interaction with vitamin B12 [13]. Dietary choline can be converted to trimethylamine (TMA) by the colonic microbiota, with TMA being further metabolized to trimethylamine-*N*-oxide (TMAO) in the liver [14]. The role of choline-derived TMAO for cardiovascular health is subject to controversial discussions [15].

A choline deficiency has not been reported at a population level, but has been observed in experimental settings and total parenteral nutrition only [16–18]. However, inadequate choline intake has been linked to non-alcoholic fatty liver disease (NAFLD), skeletal muscle

**Citation:** Roeren, M.; Kordowski, A.; Sina, C.; Smollich, M. Inadequate Choline Intake in Pregnant Women in Germany. *Nutrients* **2022**, *14*, 4862. https://doi.org/10.3390/nu14224862

Academic Editors: Louise Brough and Gail Rees

Received: 10 October 2022 Accepted: 15 November 2022 Published: 17 November 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

atrophy, neurodegenerative diseases, and several ocular diseases including retinal hemorrhage, glaucoma, and dry eye syndrome [19,20]. Furthermore, a variety of choline-related inherited metabolic diseases has been described [21].

Even though choline, in principle, can be synthesized de novo through the methylation of phosphatidylethanolamine, the endogenous synthesis is not sufficient to meet the physiological choline requirements [22]. Therefore, choline must at least in part be acquired from dietary sources. Choline is mainly found in foods of animal origin such as eggs, poultry, and meat [23,24]. Plant-based foods such as legumes, leafy greens, and nuts, contain only small amounts of choline [25]. For an Australian cohort, it has been shown that eggs, red meat, nuts, legumes, and dairy account for 50% of the choline intake, with eggs alone contributing 17% [26]. In a randomized cross-over trial, the consumption of two eggs significantly increased the plasma choline level of adult [27]. Accordingly, it is difficult to meet the adequate choline intake especially for vegetarians and vegans [28,29].

During pregnancy and lactation, the dietary requirements for choline are substantially higher than for non-pregnant women, since the fetus and the infant accumulate choline at the expense of maternal stores [30,31]. Therefore, the adequate intake (AI) for choline in pregnant women has been set at 480 mg/day by the European Food Safety Authority (EFSA), compared to 400 mg/day for non-pregnant adults [32]. A growing body of evidence strongly suggests that choline plays a crucial role during neuronal fetal development [33], e.g., by contributing to fetal brain and memory development [34], acetylcholine biosynthesis, and neuronal cell signaling [25,35–37]. Accordingly, an increased choline intake during pregnancy probably improves the neurocognitive outcomes in the offspring [4,35,38]. Most recently, a meta-analysis found that a low maternal choline intake is not only associated with impaired child neurocognition and neurodevelopment, but also with an increased risk of neural tube defects [39]. However, substantial evidence from randomized-controlled trials investigating the prenatal effects of choline is still lacking, especially regarding both possible dose response-relationships between maternal choline intake and child neurocognitive outcomes and potential interactive effects of the two methyl-donor nutrients choline and folate [40].

Taking together the major relevance of a sufficient choline intake for the fetal neuronal development, the elevated choline requirements during pregnancy, and the poor choline content of plant-based foods, it can be supposed that pregnant women following a vegetarian or vegan diet have a high risk of not achieving the recommended AI for choline. Moreover, choline is absent in most dietary supplements marketed for pregnant women, with a median daily choline dose of only 25 mg [41,42].

Most surveys during pregnancy suggest that choline intakes are considerably below the AI [23], but the clinical assessment of choline status remains difficult [43]. In Germany, the choline intake of pregnant women has never been assessed systematically before. Therefore, we estimated the dietary and supplementary choline intake of pregnant women in Germany with an online survey. To detect possible subgroup differences, both omnivores and vegetarians/vegans have been included.

#### **2. Materials and Methods**

#### *2.1. Study Design and Participants*

For this online survey, pregnant women were recruited via social media in November and December 2021, using a questionnaire on the SurveyMonkey platform. The inclusion criteria were a gestational age of 13 weeks or higher, and an age between 19 and 45 years. In total, 516 subjects started the questionnaire, of whom 283 met the inclusion criteria (Figure S1). The sample size calculation was based on epidemiological data: In Germany, the population size is approx. 500,000 pregnant women in the second and third trimester of the pregnancy. With a confidence level set at 90% and a margin of error at 5%, the sample size has been calculated *n* = 273.

The participants were informed about the purpose of the study and their formal consent was collected before they started the questionnaire. Ethical review and approval were not required for the study in accordance with the local legislation and institutional requirements.

#### *2.2. Questionnaire*

The study was conducted in Germany and the questionnaire was in German language. It comprised 30 questions that were further divided into four different parts: health and pregnancy; dietary supplement use; a food frequency questionnaire (FFQ) which specified 60 choline-containing food items or groups; and questions about the sociodemographic background of the participants (Figure S2).

The FFQ was based on the National Health and Nutrition Examination Survey (NHANES) questionnaires [44] and the Project Viva FFQ [45], which have been used for the estimation of choline intake before [46]. The questionnaire used in the present study was designed to assess the choline intake from the diet within the previous week. Pictures of hand portion sizes were added to visualize the portion size and to have a standard for the subsequent evaluation. The FFQ focused on choline-containing foods only. To assess the dietary choline intake, we referred to Zeisel's measurements [47]. Using a drop-down list, respondents were able to determine the respective number of foods/food groups consumed during the previous week. Subsequently, the study population was categorized based on their dietary pattern (omnivore vs. vegetarian/vegan).

Finally, participants documented their intake of dietary supplements during the pregnancy (trademark, duration, dosing).

#### *2.3. Data Analysis*

The individual dietary choline intake was calculated by multiplying the frequency of consumption per week by consumed amounts of all the assessed food products. The concentration of choline in every respective food item was taken from previous studies [47]. To assess the daily intake, total weekly choline intake of each individual was divided by seven. The median and the interquartile range (IQR) of the entire cohort were calculated. To estimate the daily choline intake from dietary supplements, the participants answered the questions about the frequency and the dosage of the supplements they took.

For all data, statistical analyses were performed using GraphPad Prism. To test for the normal distribution, the Shapiro–Wilk test was applied. Since the normal distribution could not be assumed, non-parametric tests were used to statistically analyze the choline intake. Median values of total choline intake and dietary choline intake were calculated and presented in milligrams per day, with min to max error bars. Choline intake was compared to the adequate choline intake for pregnant women (480 mg/day).

Multivariate analysis for confounding variables was performed by linear regression. The statistical comparison for the analysis of the different groups (omnivore to vegetarian/vegan to all) was carried out using a Kruskal–Wallis test with Dunn's test for multiple analyses; when only 2 groups were compared, the Mann–Whitney was used. We adjusted for outliers in the whole population referring to the daily intake. Outliers above the 95% percentile and below the 5% percentile were excluded (in total 10 data points, 8 omnivore and 2 vegetarian/vegan).

#### **3. Results**

#### *3.1. Study Population*

The baseline characteristics of the study population (*n* = 283) are shown in Table 1. Most participants were aged between 26–35 years (234/283), lived with their partner or family (278/283), and had a university degree (144/283). 56 (20%) participants changed their diet due to the pregnancy, and 16 (6%) took choline-containing dietary supplements. Among participants taking choline-containing dietary supplements, only three received a recommendation for it. Referring to health in pregnancy, the mean number of days of feeling nauseous was 50, with 54 (19%) of participants reporting weight loss due to vomiting.


**Table 1.** Descriptive statistics of the study population. Selected characteristics of the study cohort (*n* = 283) stratified by dietary patterns. Values are absolute number and percentages unless stated otherwise within diet group according to the categories in the first column.

In the context of qualitative representativeness, our sample is representative for pregnant women in Germany regarding age distribution and living situation. Regarding education, women with a university degree (51%) are overrepresented in our sample.

59 (21%) women followed a vegetarian/vegan diet, while 224 (79%) were omnivorous. These two groups differed across health and sociodemographic characteristics. Vegetarian/vegan women were more likely to be older, more educated, and more likely to take choline-containing dietary supplements (8% vs. 5%). Moreover, they were less likely to lose weight due to vomiting (14% vs. 21%). In contrast, omnivorous women were less likely to change their diet for the pregnancy (20% vs. 76%). No-one in the vegetarian/vegan group received a recommendation to take choline-containing dietary supplements.

#### *3.2. Total Choline Intake*

For total choline intake, the estimated choline intake from both diet and dietary supplements were added. Only 7% (19/283) of participants achieved the adequate choline intake of 480 mg/day. The median choline intake was 263.5 ± 147.8 mg/day. 93% of omnivores (208/224) and 95% of vegetarians/vegans (56/59) had an inadequate choline intake when applying the choline AI (median: 274.3 ± 156 mg/day and 209.2 ± 107.7 mg/day, respectively).

After excluding outliers (choline intake > 558.70 mg/day), the median choline intake remained below the AI with 260.4 ± 141.4 mg/day for all, 269.5 ± 141.5 for omnivores, and 205.2 ± 101.2 mg/day for vegetarians/vegans (Figure 1). The difference in daily choline intake between omnivores and vegetarians/vegans was statistically significant (*p* < 0.001) (Figure 1). Calculating the odds ratio (OR), the vegetarian/vegan group had 30% lower odds of meeting the AI than the omnivorous group (95% CI 0.21–2.35, Table 2).

**Figure 1.** Total (dietary and supplementary) daily choline intake in all (*n* = 273), omnivorous (*n* = 217) and vegetarian/vegan (*n* = 56) participants (outliers excluded). The red dotted line represents the choline AI of 480 mg/day. Data are presented as median with whiskers from 'min to max' and was analyzed with Kruskal–Wallis test following the Dunn's test for multiple comparisons (\*\* adjusted *p* value < 0.01; \*\*\* adjusted *p* value < 0.001).

**Table 2.** Participants who met the adequate choline intake (AI) of 480 mg/d.


As a result of the multivariate analysis for possible confounders (age, education, nauseous days, diet), apart from the diet, the age of 36–40 years was the only confounding factor.

#### *3.3. Dietary Choline Intake*

Applying the AI for pregnant women, only 7% of the participants achieved an adequate choline intake. The median dietary choline intake was 267.8 ± 137.7 mg/day for omnivorous women and 204.4 ± 99.5 mg/day for vegetarian/vegan women (*p* < 0.0001) (Figure 2A).

**Figure 2.** Dietary choline intake. (**A**) Daily dietary choline intake in omnivorous and vegetarian/vegan participants. The AI of 480 mg/day is represented by the red dotted line. Data are presented as median with whiskers from 'min to max' and was analyzed with Mann–Whitney test (\*\*\*\* *p* < 0.0001). (**B**) Dietary choline sources in omnivorous and vegetarian/vegan participants. Data are shown as parts of 100, numbers in the graph indicate the mean contribution of the given food to the total dietary choline intake.

The minimal dietary choline intake was 49 mg per day, being higher among vegetarians/vegans than among omnivores (60.06 mg/day vs. 48.69 mg/day). Diet contributed most to total choline intake in both groups, with differences between omnivores and vegetarians/vegans. The main sources of choline in omnivores were eggs (56.7 mg/day), red meat (48.2 mg/day), and white meat (25.1 mg/day); whereas, in the vegetarians/vegans, eggs, green kale, and fruit juice contributed the most with 42.7 mg/day, 19.1 mg/day, and 11.8 mg/day, respectively (Figure 2B). 3–4% of the total choline intake via food was provided by milk and potatoes in both groups.

#### *3.4. Choline Intake from Dietary Supplements*

In total, 13/283 (5%) participants reported to take choline-containing dietary supplements (omnivores: *n* = 10; vegetarians/vegans: *n* = 3, Figure 3A). In these women, dietary supplements accounted for 19% of total choline intake. Differentiating between dietary habits, choline-containing supplements contributed 16% to the total choline intake of omnivores (mean supplementary choline intake: 100.94 mg/day), but 34% (mean supplementary choline intake: 126.67 mg/day) to the total choline intake of vegetarians/vegans (Figure 3B).

**Figure 3.** Intake of dietary supplements. (**A**) Proportion of participants taking choline-containing dietary supplements in all (13/267), omnivorous (10/213), and vegetarian/vegan (3/54) participants (**B**) Percentage of choline coming from food sources and dietary supplements in participants supplementing choline (all *n* = 13; omnivore *n* = 10, vegetarian/vegan *n* = 3) (**C**) Intake of dietary supplements during pregnancy. Percentage of participants taking given supplement alone or as a part of a prenatal vitamin complex supplement.

#### *3.5. Dietary Supplement Use*

274/283 (97%) participants took any dietary supplement during pregnancy. 90% of the women took folic acid alone or as part of a prenatal vitamin complex (Figure 3C). Vitamin D was supplemented by 52%, iodine by 50%, and magnesium and vitamin B12 by 49% each.

#### **4. Discussion**

#### *4.1. Main Finding*

Our study was the first to estimate the dietary and supplemental choline intake of pregnant women in Germany, demonstrating that 93% of pregnant women do not meet the adequate choline intake, with vegetarian/vegan women having an even lower chance of achieving the adequate intake. Moreover, taking dietary supplements does not substantially improve the situation.

#### *4.2. Previous Findings on Choline Intake*

As early as 1998, choline was recognized as an essential nutrient, and the adequate choline intake has been set for both the general population and pregnant women by the Institute of Medicine (IOM) [48]. In contrast, until now, no respective recommendations have been published by the German Society of Nutrition.

Data on the dietary and supplementary choline intake of healthy adults are only available for North America and few European countries; respective data for Germany are lacking. Most surveys during pregnancy suggest that the AI of choline is met by few women only [23]. Accordingly, the choline intake of healthy, non-pregnant women has recently been estimated at 291 mg/day (France), 285 mg/day (Greece), 334 mg/day (The Netherlands), 294 mg/day (UK), and 362 mg/day (Australia) [26,49]. For Germany, total choline intake estimates have only been published for children and adolescents with an average intake in females ranging from 151–295 mg/day [50]. With the estimated median choline intake in our survey being 260.4 mg/day, our results are consistent with previous findings as they are both close to the estimated choline intake from neighboring countries with similar dietary habits (291–374 mg/day) [49], and very similar to the estimated choline intake of German female adolescents (295 mg/day) [50]. The result of the multivariate analysis indicating that the age of 36–40 years is the only confounding factor (apart from the diet) can be interpreted as a statistical artefact, as it cannot be plausibly explained by physiological/psychological hypotheses and previous studies.

Investigating the choline intake during pregnancy, systematic data for European countries are lacking almost completely. The only data published so far come from a Latvian survey showing an average choline intake among pregnant women of 336 mg/day [51], which too is in line with our results. Moreover, our finding that 93% of pregnant women in Germany do not reach the adequate choline intake is in line with similar data from the US with 91% of the pregnant women not meeting the AI for choline [52].

#### *4.3. Studies on Choline-Containing Dietary Supplements*

A randomized controlled trial assessed the effect of third trimester maternal choline supplementation (930 mg/day vs. 480 mg/day) on child memory at 7 years of age. Both groups were above the adequate intake levels in Germany [32]. Children of higher supplemented mothers scored better results than children in the control group with lower choline doses. Another study investigating the effect of prenatal choline supplementation (500 mg/day vs. 25 mg/day) on maternal and fetal biomarkers of choline metabolism measured higher plasma concentrations of free choline, betaine, dimethylglycine, phosphatidylcholine, and sphingomyelin among higher supplemented women [53]. Moreover, pregnancy-related metabolic adaptions were supported in this trial. These findings indicate that even the choline AI set for pregnant women by the EFSA may not be sufficient for optimal offspring neurodevelopment. However, in both studies the sample size was relatively small which makes it difficult to draw generalized conclusions. Furthermore, compared to the supplements used in our study, the dietary supplements used in these trials by far exceeded the choline doses used in Germany.

Most women in our study supplemented folic acid (FA) during their pregnancy. Recent studies suggest that imbalances between FA and other methyl-donor nutrients involved in one-carbon metabolism can determine the pregnancy outcomes [54] and metabolic adaptions [55]. Folic acid and choline play critical roles in the production of S-adenosylmethionine (SAM), a key modulator of DNA methylation [56]. However, in contrast to FA, choline is absent in most prenatal dietary supplements in Germany. This aspect is also observed in our study, since most of the women took folic acid-containing dietary supplements while most of these supplements did not contain any choline.

#### *4.4. Implications*

Our results for the first time demonstrate that the choline intake of pregnant women in Germany generally does not meet the recommendations for an adequate intake, with vegetarian and vegan women having an even lower chance of achieving the AI for choline. Moreover, taking dietary supplements does not improve the situation. Thus, it can be concluded that currently neither the majority of pregnant women, nor health care professionals, nor manufacturers of dietary supplements are aware of choline being a critical

nutrient in pregnancy. Furthermore, our results underline the imbalance of folic acid and choline intake in pregnant women in Germany. Several implications arise from our results.

First, obviously, it is very difficult to meet the recommended AI for choline with a regular, omnivorous diet, and it is even more so for pregnant women following a vegetarian or vegan diet. Therefore, it might be suggested to advice pregnant women to change their dietary habits, thereby improving their choline intake. This approach, however, is hardly practicable considering the low success rates of general dietary recommendations. Even more, pregnant women sticking to a vegetarian or vegan diet due to ethical reasons are very unlikely to switch to an omnivorous, egg and meat-containing diet just to increase their choline supply. On the other hand, small dietary changes even of omnivorous women will not be able to substantially increase their choline intake.

Second, the awareness that choline might be an essential nutrient in pregnancy should be raised, both among pregnant women and healthcare professionals, including gynecologists, midwives, general practitioners, and pharmacists. The urgent need for improved choline provision for pregnant women, both through individual counselling and public health interventions, has been emphasized by other authors before [57]. This goal, however, is unlikely to be achieved through general information campaigns, but rather through targeted training and continuing medical education programs. If choline-specific dietary recommendations prove to be insufficient or unsuccessful, the intake of choline-containing dietary supplements might be considered. Particularly, respective supplements may be useful in vegetarian/vegan women and in women suffering from nausea and vomiting during pregnancy. Since vegetarian and vegan diets are increasingly common in pregnant women, there is an increased risk that the maternal choline supply will deteriorate as a result.

Third, the results of our survey suggest that it might be useful to add sufficient amounts of choline to products that are advertised for pregnant women. As shown here, most dietary supplements used do not contain any choline. If they do contain choline, the respective concentrations are too low to substantially improve the choline supply of the pregnant women.

Finally, fourth, more randomized controlled trials are needed to further specify the health benefits for the offspring resulting from an improved maternal choline intake. The adequate choline intake recommendations are not derived from randomized controlled trials but estimated from epidemiological data only. Therefore, it must be kept in mind that not meeting the recommended AI not necessarily means that the respective individual (or her offspring) is insufficient of choline or even suffering from clinically relevant deficiency.

#### *4.5. Strengths and Limitations*

It is a particular strength of our study that it makes an important contribution to putting the focus on the choline supply of pregnant women. The results presented here are the first to estimate the total choline intake of pregnant women in Germany, both from dietary and supplementary sources and differentiating between omnivorous and vegetarian/vegan diets. A major strength of our methodology is the semiquantitative questionnaire with hand portion pictures that enabled improved estimation of dietary intakes.

Regarding the limitations, it has to be considered that our data are based on a nonprobability convenience sample rather than a representative population-based sample. With this type of sampling, the generalizability of our results is limited to populations that share similar characteristics with our sample. Therefore, it remains questionable whether the results would be similar in a representative sample. Due to the case number calculation, the quantitative representativeness of our results is given. In terms of qualitative representativeness, our sample is representative for pregnant women in Germany regarding age distribution and living situation. However, regarding education, women with a university degree (51%) are overrepresented in our sample, as due to census data only 28% of women < 60 years hold a university degree. Thus, the choline intake of pregnant women without a university degree might differ from the results presented here.

In this context, a selection bias might be relevant, too. Recruitment was done through social media, so pregnant women without appropriate media use were not reached. Women with a heightened interest in nutritional issues and dietary supplements probably preferentially participated in the survey. When interpreting the data, it must be noted that the number of subjects in some subgroups was too small to obtain statistically meaningful results. This problem must be addressed with appropriately powered follow-up studies.

Additionally, the choline content of several foods was unknown since it has never been analyzed and published. As a result, some dietary choline sources were not named so that the final result might be lower. Specifically, the FFQ used in our study did not include vegan and vegetarian meat or dairy alternatives which might have affected the estimated choline intake especially in vegetarians and vegans.

Since the FFQ method is widely used in nutrition surveys, its inherent limitations are well-known and have been discussed in detail elsewhere [58]. Additionally, some participants obviously misread the instructions and filled in the questionnaire not for a week but rather for a whole month. In order to attenuate this error, we excluded outliers as indicated. An alternative approach for food intake assessment might have been repeated dietary recalls or records; however, an FFQ is more achievable in a large cohort and within the given time frame [59], even more, it is less prone to over- or underestimating the food intake than other methods [60].

Of course, the results may not be transferred to other countries without further ado, since not only dietary habits may differ, but also the medical counselling of pregnant women, the market situation of dietary supplements, the health policies and public opinion regarding food fortification, and the women's attitude towards taking dietary supplements.

Finally, any evaluation of choline intake must be done with caution, as intake below the AI not necessarily indicates a health-affecting deficiency [49].

#### **5. Conclusions**

Due to the relevance of choline for fetal development, and considering our results that suggest an inadequate choline intake in pregnant women in Germany, efforts to encourage the increased intake of choline-rich foods and/or choline-containing dietary supplements during pregnancy might be useful. This is especially true for pregnant women who follow a vegetarian or vegan diet. Moreover, further research is necessary to define optimal choline requirements in pregnancy.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/nu14224862/s1, Figure S1: Flow Chart. Figure S2: Original questionnaire in German language with an example of the drop-down list used for the food frequency questionnaire.

**Author Contributions:** Conceptualization, M.S. and M.R.; methodology, M.R.; formal analysis, M.R. and A.K.; writing—original draft preparation, M.R., A.K. and M.S.; writing—review and editing, C.S. and M.S.; supervision, M.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Ethical review and approval were waived for this study due to local legislation and institutional requirements. Approval by the local ethics committee is not required if the study is non-interventional, does not involve biological samples from humans, the data analysis does not lead to a medical guarantee position and the personal data are anonymous and non-traceable.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Data available upon request from the corresponding author.

**Acknowledgments:** We would like to thank all the participants for taking part in this study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

