**Effects of a Mediterranean Diet Intervention on Maternal Stress, Well-Being, and Sleep Quality throughout Gestation—The IMPACT-BCN Trial**

**Irene Casas 1,†, Ayako Nakaki 1,2,†, Rosalia Pascal 1,3,4,†, Sara Castro-Barquero 1,5,6,\*, Lina Youssef 1,2,7, Mariona Genero 1, Leticia Benitez 1,2, Marta Larroya 1,2, Maria Laura Boutet 1, Giulia Casu 1, Alex Gomez-Gomez 8, Oscar J. Pozo 8, Ivette Morilla 9, Anabel Martínez-Àran 9, Eduard Vieta 9, María Dolores Gómez-Roig 1,3,4, Rosa Casas 5,6, Ramon Estruch 5,6, Eduard Gratacos 1,2,10, Fàtima Crispi 1,2,10,‡ and Francesca Crovetto 1,3,4,‡**


**Abstract:** Stress and anxiety are frequent occurrences among pregnant women. We aimed to evaluate the effects of a Mediterranean diet intervention during pregnancy on maternal stress, well-being, and sleep quality throughout gestation. In a randomized clinical trial, 1221 high-risk pregnant women were randomly allocated into three groups at 19–23 weeks' gestation: a Mediterranean diet intervention, a Mindfulness-Based Stress Reduction program, or usual care. All women who provided self-reported life-style questionnaires to measure their anxiety (State Trait Anxiety Inventory (STAI), Perceived Stress Scale (PSS)), well-being (WHO Five Well Being Index (WHO-5)), and sleep quality (Pittsburgh sleep quality index (PSQI)) at enrollment and at the end of the intervention (34–36 weeks) were included. In a random subgroup of 106 women, the levels of cortisol and related metabolites were also measured. At the end of the intervention (34–36 weeks), participants in the Mediterranean diet group had significantly lower perceived stress and anxiety scores (PSS mean (SE) 15.9 (0.4) vs. 17.0 (0.4), *p* = 0.035; STAI-anxiety mean (SE) 13.6 (0.4) vs. 15.8 (0.5), *p* = 0.004) and better sleep quality (PSQI mean 7.0 ± 0.2 SE vs. 7.9 ± 0.2 SE, *p* = 0.001) compared to usual care. As compared to usual care, women in the Mediterranean diet group also had a more significant increase in their 24 h urinary cortisone/cortisol ratio during gestation (mean 1.7 ± SE 0.1 vs. 1.3 ± SE 0.1, *p* < 0.001). A Mediterranean diet intervention during pregnancy is associated with a significant reduction in maternal anxiety and stress, and improvements in sleep quality throughout gestation.

**Keywords:** Mediterranean diet; pregnancy; anxiety; well-being; sleep quality

R.; Castro-Barquero, S.; Youssef, L.; Genero, M.; Benitez, L.; Larroya, M.; Boutet, M.L.; Casu, G.; et al. Effects of a Mediterranean Diet Intervention on Maternal Stress, Well-Being, and Sleep Quality throughout Gestation—The IMPACT-BCN Trial. *Nutrients* **2023**, *15*, 2362. https://doi.org/10.3390/ nu15102362

**Citation:** Casas, I.; Nakaki, A.; Pascal,

Academic Editor: Herbert Ryan Marini

Received: 19 April 2023 Revised: 10 May 2023 Accepted: 17 May 2023 Published: 18 May 2023

**Copyright:** © 2023 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/).

#### **1. Introduction**

The Mediterranean diet (MedDiet) has several positive effects on individual health: randomized trials demonstrated its contribution to improved cardiovascular profiles and reduced major cardiovascular events in individuals at risk of [1] diabetes, inflammatorybased disorders, cancer, and cognitive decline [2–4]. Additionally, there has been increasing interest of the effects of a MedDiet on mental health, stress, and quality of life in general [5]. The role of the diet, particularly the MedDiet, in the development of mental disorders, has become a recent research focus over the past decade [6]. Several studies evaluated the effect of a MedDiet intervention on the reduction in depressive symptoms and the improvement in quality of life in individuals with major depressive disorders [7,8]. In a secondary analysis of the PREvención con DIeta MEDiterránea (PREDIMED) study, a reduced risk in depression was observed in participants with type 2 diabetes allocated to the group receiving a MedDiet supplemented by nuts (hazard ratio 0.59 (95% confidence interval (CI) 0.36 to 0.98)) [9]. A recent review based on 37 studies confirmed the association between (poly)phenols consumption and the risk of depression, and a reduction in the severity of depressive symptoms [10]. Some authors hypothesized that a high-quality diet, rich in fiber, antioxidant dietary components and omega-3-polyunsaturated fatty acids, may be linked to a reduced risk of depression, anxiety, and stress [11], which could provide new potential methods for the treatment and prevention of mental disorders in general. Moreover, it has been described that a dysregulated redox signaling is a key factor in the pathophysiology of mental disorders, especially in depression, and increased reactive oxygen and nitrogen species were observed in these patients [12,13].

Stress and anxiety are frequent occurrences among pregnant women. Peripartum anxiety disorders are more prevalent than previously thought, as 1 in 5 women can suffer from them [14]. Mental disorders can appear before pregnancy, with a changing course during pregnancy and postpartum. These findings highlight the need for screenings for stress-related disorders and education by different health professionals from the early stages of pregnancy. Several studies have shown the effectiveness of non-pharmacological treatments in the improvement in stress and other mental disorders during pregnancy, such as mindfulness meditation, biofeedback, or exercise such as yoga [15]. However, there is paucity of data regarding the dietary approach to these conditions during pregnancy. Interestingly, a recent observational study revealed an association between the MedDiet and anxiety [16]. Moreover, the production of reactive oxygen and nitrogen species production, as well as individual antioxidant capacity, is influenced by several dietary factors. A dietary intervention promoting plant-based foods that are rich in antioxidants, such fruits, vegetables, extra-virgin olive oil, and whole-grain cereals, may modulate the individual antioxidant capacity, explaining the improvements in mental wellbeing [12]. Thus, randomized clinical trials are needed to establish the potential effects of dietary patterns on mental health, avoiding the confusion attributed to the co-occurrence of other lifestyle-related and sociodemographic factors.

During pregnancy, evidence has been provided regarding the potential beneficial effects that structured dietary interventions based on a MedDiet can have, not only on pregnant women [9,13,14], but also their offspring and the pregnancy itself. In a recent randomized clinical trial, pregnant individuals at high risk for small-for-gestational-age newborns (SGA) who followed a structured MedDiet intervention significantly reduced the incidence of newborns being born small (with birth weight below the 10th percentile) and other perinatal complications [17]. However, the influence of MedDiet on maternal wellbeing during pregnancy remains to be determined.

The present study aimed to evaluate the influence of a structured intervention during pregnancy based on a MedDiet on maternal stress and anxiety, mindful state, quality of life and sleep.

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

#### *2.1. Study Design, Population and Ethics*

Improving Mothers for a better PrenAtal Care Trial BarCeloNa (IMPACT BCN) was a parallel, unblinded randomized clinical trial conducted at BCNatal (Hospital Clínic and Hospital Sant Joan de Déu), a large referral center for maternal–fetal and neonatal medicine in Barcelona, Spain. Details of the trial are provided in the protocol of the study [18], approved by the Institutional Review Board (HCB-2016-0830) before any participant enrolment. All individuals who agreed to participate provided written informed consent before randomization. Participants were screened for eligibility during routine second trimester ultrasound scans (19–23.6 weeks of gestation) for being at high risk of developing an SGA newborn [19], and were randomly assigned 1:1:1, based on a computerized random number generator, to one of the three study groups: a MedDiet supplemented with extra-virgin olive oil and walnuts; a stress reduction intervention based on the Mindfulness-Based Stress Reduction (MBSR) program; or usual care without any intervention (control group). For this specific study, only women belonging to the group of MedDiet and usual care who provided lifestyle questionnaires were included. The trial was registered in ClinicalTrials.gov Identifier: NCT03166332.

#### *2.2. Interventions and Measurements*

#### 2.2.1. Mediterranean Diet Program

The dietary intervention, adapted from the PREDIMED trial [20], aimed to change the general dietary pattern instead of focusing on changes in single foods or macronutrients. Participants were encouraged to increase their intake of whole-grain cereals (≥5 servings/d); vegetables and dairy products (≥3 servings/d); fresh fruit (≥2 servings/d); and legumes, nuts, fish, and white meat (≥3 servings/week), as well as increasing their olive oil use for cooking and dressings. To achieve a personalized goal, personal and individual recommendations were introduced to the participant's diet according to height, weight, culture, and dietary preferences. Dieticians conducted 30 min face-to-face interviews at enrollment and monthly until the end of intervention (34–36 weeks' gestation). Two weeks following each face-to-face visit, participants underwent telephone interviews. In addition, all participants received extra-virgin olive oil (2 L every month) and 15 g of walnuts per day (450 g every month) at no cost. Additional details of the intervention are provided elsewhere [18]. No intervention or advice regarding mental health, well-being, anxiety, stress, or sleep quality were provided to the participants allocated to the Mediterranean diet group.

#### 2.2.2. Usual Care (Control Group)

Women randomized into this group received usual pregnancy care as per institutional protocols (no intervention), and lifestyle questionnaires were collected at enrollment and at the end of intervention (34–36 week's gestation). No intervention or advice regarding mental health, well-being, anxiety, stress, or sleep quality were provided to the participants allocated to the control group.

#### *2.3. Outcomes*

In this trial sub-analysis, the main aim was to investigate the influence of a Mediterranean diet intervention program during pregnancy on maternal stress, anxiety, well-being, mindful state, and sleep quality. Additionally, in a randomly selected subgroup of participants, the levels of cortisol, cortisone and other intermediate related metabolites were measured at the beginning and at the end of the intervention in 24 h urine samples.

#### *2.4. Data Collection*

The data of participants included in the study were anonymized and entered in an electronic case report form. Investigators collected maternal sociodemographic and clinical data.

All individuals included in the trial had a baseline visit (19–23 weeks of gestation) and a final visit (34–36 weeks of gestation) with a trained dietitian to assess their diet using a validated 151-item food-frequency questionnaire [21], 7-day dietary registry and the 17-item MedDiet adapted to pregnancy adherence score (score range: 0–17). All participants also provided self-report lifestyle questionnaires to measure their anxiety and stress (State-trait Anxiety Inventory (STAI) Anxiety and Personality [22], range 0–80); Perceived Stress Scale (PSS) [23], range 0–40; well-being (WHO Five Well Being Index (WHO-5) [24], range 0–100); mindful state (WHO Five Facet Mindfulness Questionnaire (FFMQ) [25], range 8–40 for the observation, description, awareness, and nonjudgmental facets, respectively, and range 7–35 for nonreactivity facet); sleep quality (Pittsburgh Sleep Quality Index (PSQI) [26], range 0–21). The questionnaires were carried out at enrollment (baseline punctuation) and at 34–36 weeks of gestation (final punctuation). Abnormal scores were considered the 75th percentile of the baseline scores of each questionnaire in the usual care group, except for the WHO-5 questionnaire, which presents a previously reported cut-off point that defines optimum mental well-being as a score greater than 52 [27].

#### *2.5. Sample Collection*

In a subgroup of randomly selected participants from each study group (excluding those receiving corticosteroid treatment), the 24 h urinary cortisone and cortisol metabolites were measured at the baseline and final assessment and analyzed by a validated method based on liquid chromatography-tandem mass spectrometry (LC-MS/MS) [28]. The activity of 11β-Hydroxysteroid Dehydrogenase Type 2 was estimated by the cortisone/cortisol ratio.

#### *2.6. Statistical Analysis*

Clinical data are presented as mean (standard deviation (SD) or standard error (SE)), median (interquartile range (IQR)) or number (percentage), as appropriate. The methods of statistical analyses used for the comparison of clinical and perinatal characteristics included Student's *t*-test, ANOVA or ANCOVA with baseline adjustments for continuous variables and X<sup>2</sup> test for categorical variables. Differences were considered significant when *p*-value < 0.05. Statistical analyses were performed using the Statistical Package for the Social Sciences statistical software package version 27 (SPSS Inc., Chicago, IL, USA).

#### **3. Results**

#### *3.1. Study Population and Pregnancy Outcomes*

Within these patients, after excluding those that did not provide lifestyle questionnaires to measure their anxiety and stress, mindful state and sleep quality, a population of 680 individuals was considered (*n* = 331 for Mediterranean diet, *n* = 349 for usual care), as reported in Figure 1.

**Figure 1.** Flowchart of participants from the IMPACT BCN trial involved in the current study.

Baseline characteristics of the study population are shown in Table 1 with no differences between study groups. Pregnancy and perinatal outcomes are shown in Supplementary Table S1, with no significant differences between groups apart from the prevalence of SGA newborns, as reported in the main outcome of the trial [17].

**Table 1.** Baseline characteristics of women included in the study according to intervention groups (*n* = 680).


BMI: Body mass index. Data are expressed as median (IQR) or mean (SD) or *n* (%). <sup>a</sup> socioeconomical status: low (never work or unemployed >2 years), medium (secondary studies and work), high (university studies and work).

#### *3.2. Effects of Mediterranean Diet on Stress, Anxiety, Well-Being, Sleep Quality and Mindful State* 3.2.1. Life-Style Questionnaires

Table 2 displays baseline and final life-style questionnaire scores on stress, anxiety, well-being, sleep quality, and mindful state between study groups, and Table 3 reports the percentage of high/poor scores at the final assessment. Perceived stress, anxiety and poor sleep quality increased throughout gestation in all study groups (Figure 2). At the end of the intervention, participants in the Mediterranean diet group showed significantly lower levels of perceived stress as compared to patients undergoing usual care, as shown in Figure 2A (mean difference −0.85 (−1.63 to −0.06), *p* = 0.035). Similarly, the Mediterranean diet group presented significantly lower final anxiety scores compared to the non-intervention group (mean 13.6 ± 0.4 SE vs. 15.8 ± 0.5, *p* < 0.004) (Figure 2B), with a lower frequency of high anxiety scores (*n* = 58, 17.9% vs. *n* = 87, 25.4%, *p* = 0.020), as reported in Table 3. Aligned with the previous findings, women's sleep quality improved following the Mediterranean

diet intervention compared to controls (PSQI mean 7.0 ± 0.2 SE vs. 7.9 ± 0.2 SE, *p* = 0.001) (see Table 2 and Figure 2C).

**Within-Group Mean Changes** *p* **§ Between-Group Changes Usual Care MedDiet MedDiet vs. Usual Care** *n* **= 349** *n* **= 331 Difference (95% CI)** Perceived stress scale score Baseline † 16.3 ± 7.8 15.9 ± 7.6 Final ‡ 17.0 ± 0.4 \* 15.9 ± 0.4 0.035 −0.85 (−1.63 to −0.06) State-trait Anxiety Inventory (anxiety) Baseline † 14.1 <sup>±</sup> 8.8 12.9 <sup>±</sup> 8.3 Final ‡ 15.8 ± 0.5 \*\* 13.6 ± 0.4 \* 0.004 −1.35 (−2.28 to −0.43) State-trait Anxiety Inventory (personality) Baseline † 15.8. <sup>±</sup> 9.0 14.2 <sup>±</sup> 7.9 Final ‡ 15.8 ± 0.5 14.0 ± 0.5 0.100 −0.68 (−1.48 to 0.13) WHO Five Well-being index Baseline † 62.7 ± 17.3 67.5 ± 15.2 Final ‡ 62.9 ± 0.9 66.6 ± 0.8 0.587 0.51 (−1.32 to 2.33) Pittsburgh Sleep Quality Index Baseline † 6.7 ± 2.4 6.4 ± 2.1 Final ‡ 7.9 ± 0.2 \*\* 7.0 ± 0.2 \*\* 0.001 −0.73 (−1.15 to −0.31) FFMQ 1: Observation Baseline † 23.3 ± 5.9 24.2 ± 5.6 Final ‡ 24.0 ± 0.3 24.6 ± 0.3 0.729 0.12 (−0.57 to 0.81) FFMQ 2: Description Baseline † 32.1 ± 5.5 32.7 ± 4.8 Final ‡ 31.7 ± 0.3 32.4 ± 0.3 0.273 0.35 (−0.27 to 1.37) FFMQ 3: Awareness Baseline † 31.3 ± 6.0 31.3 ± 6.3 Final ‡ 30.6 ± 0.4 \* 30.0 ± 0.4 \*\* 0.280 −0.51 (−1.43 to 0.41) FFMQ 4: Non-judgmental Baseline † 29.9 ± 5.6 30.1 ± 5.2 Final ‡ 30.0 ± 0.3 30.0 ± 0.3 0.994 0.00 (−0.64 to 0.64) FFMQ 5: Non-reactivity Baseline † 22.5 ± 4.8 22.6 ± 4.8 Final ‡ 22.9 ± 0.2 22.5 ± 0.3 0.091 −0.55 (−1.05 to 0.08)

**Table 2.** Changes in maternal anxiety, well-being, sleep quality, and mindful state evaluated at baseline and final evaluation according to intervention groups.

MedDiet: Mediterranean diet; FFMQ Five Facet. Mindfulness questionnaire. † Baseline values are observed means ± SD. ‡ Final values are baseline-adjusted (least-squares) means ± SE and comparison among groups obtained with ANCOVA analysis. \* *p* < 0.05 and \*\* *p* < 0.001 final from baseline comparison. § ANCOVA analysis.

**Table 3.** Frequency of women high maternal stress, poor well-being and sleep quality questionnaires score at final evaluation according to intervention groups.


Data are expressed as *n* (%). High maternal stress/anxiety defined as Perceived Stress Scale and State-strait Anxiety Inventory scores above 75th percentile. Poor well-being defined as Five Well-Being Index score below 52. Poor sleep quality defined as Pittsburgh Sleep Quality score above 75th percentile. <sup>a</sup> Data available for 667 pregnancies. <sup>b</sup> Data available for 674 pregnancies. <sup>c</sup> Data available for 546 pregnancies.

Regarding the well-being questionnaire, 19.8% (*n* = 65) of women from the Mediterranean diet group presented with poor well-being as compared to 27.5% (*n* = 95) in the control group (*p* = 0.02), revealing better well-being (see Table 3 and Figure 3). No significant differences between groups were observed with the mindful state questionnaire (Table 2). Changes in key foods and nutrient intake during intervention are shown in Supplementary Tables S2 and S3.

**Figure 2.** Changes in maternal stress (**A**), anxiety (**B**) and sleep quality (**C**) at baseline (20 weeks of gestation) and final (33 weeks) evaluation according to intervention groups.

**Figure 3.** Percentage of high- vs. low-stress participants, and poor vs. good well-being (WHO-5) according to intervention groups. High stress is shown in dark grey color and defined as a State-trait Anxiety Inventory (STAI) personality score above 75th percentile in Usual care (**A**) and Mediterranean diet group (**B**). Poor well-being is shown in in dark grey color and defined as a Five Well-Being Index WHO score below 52.

#### 3.2.2. Cortisol Assessment

The baseline 24 h urinary cortisone/cortisol ratio in 106 participants was similar between groups and increased during gestation. This increase was more pronounced in the Mediterranean diet group compared to usual care (mean 1.7± SE 0.1 vs. mean 1.3 ± SE 0.1, *p* < 0.001) (Table 4). At final assessment, Mediterranean diet participants showed higher levels of total cortisone concentration (mean 134.7± SE 8.3 vs. mean 111.5 ± SE 7.7, *p* = 0.012) and percentage (mean 2.9± SE 0.1 vs. mean 2.4 ± SE 0.1, *p* = 0.002), and lower levels of the 5β-tetrahydrocortisone/Cortisone (mean 16.8 ± SE 1.2 vs. mean 21.4 ± SE 1.4, *p* = 0.032) compared to the control group.


**Table 4.** Differences in urinary 24 h cortisol, cortisone and other related metabolites at baseline and final evaluation according to intervention group (*n* = 106).

5β-THF/Cortisol: 5β-tetrahydrocortisol/Cortisol; 5β-THE/Cortisone: 5β-tetrahydrocortisone/Cortisone. † Baseline values are observed means ± SD. ‡ Final values are baseline-adjusted (least-squares) means ± SE and comparison among groups obtained with ANCOVA analysis. \* *p* < 0.05 and \*\* *p* < 0.001 final from baseline comparison. § ANCOVA analysis.

#### **4. Discussion**

In this randomized clinical trial that involved pregnant women at high risk for an SGA newborn, an intervention based on MedDiet significantly reduced maternal anxiety and stress and improved well-being and sleep quality. These effects were revealed by self-reported stress questionnaires and biomarkers, as reflected by the increased estimated activity of a cortisol-deactivating enzyme.

Interest in mental health and care has grown exponentially in recent years and associations between healthy dietary patterns and mental health parameters have been reported. Jacka et al. conducted a randomized controlled trial to investigate the efficacy of a dietary intervention based on the MedDiet for the treatment of symptoms related to major depressive episodes in subjects with Major Depressive Disorder, independently of other factors such as physical activity, smoking habit, or weight loss [29]. The MedDiet group showed significantly greater improvements in symptoms of depression compared to the control group. In addition, other studies have evidenced that a lower incidence of depression incidence was significantly correlated with increasing adherence to MedDiet [7]. Additionally, in the PREDIMED study, a preventive effect for depression was found for the MedDiet in participants with type 2 diabetes [9]. Specifically, participants with type 2 diabetes allocated to the MedDiet supplemented with nuts group showed a 40% lower risk of depression compared to the control arm.

However, the evidence about the effects of dietary interventions on mental health during pregnancy is limited. Our study reveals that following the MedDiet during pregnancy is associated with a reduction in maternal anxiety/stress, together with an increase in the cortisol-deactivating enzyme. These findings are in line with previous data. In a recent study, Papandreou et al. conducted a randomized clinical trial with 40 pregnant women incorporating MedDiet recommendations into the Clinical Decision Support Systems, showing an improvement in nutritional status and reduction in health-related anxiety and depression [30]. Similarly, a longitudinal study with 152 pregnant women showed that

higher adherence to the MedDiet was inversely associated with anxiety and directly associated with well-being [16]. Moreover, these associations were significant for some key foods of the MedDiet, specifically whole-grain cereals, fruits and vegetables, extra-virgin olive oil and nuts [16], food sources of dietary antioxidants whose consumption was encouraged during the intervention in our study. Aligned with our findings, other healthy dietary patterns promoting healthy foods not based on the MedDiet were associated with lower depression during pregnancy [31–33]. Nevertheless, in observational and cohort studies with pregnant women, some specific foods have been identified as protective against mental disorders (including depression and anxiety), including whole-grain cereals, fruits, and beans. In contrast, other foods are associated with higher risk, including ultra-processed foods such as pastries, red and processed meat, margarine, and artificial juices [16,34]. Additionally, it has been postulated that levels of depression tend to increase throughout pregnancy, highlighting the importance of structured dietary interventions to improve overall diet quality during pregnancy [33,35].

In addition to its beneficial effects on anxiety and stress, our study first demonstrates an improvement in maternal well-being and sleep quality with MedDiet. The association between higher MedDiet adherence and subjective well-being has been found in observational studies [36]. In the case of sleep quality, a longitudinal study with 150 pregnant women assessed the association between MedDiet adherence and the Pittsburgh Sleep Quality Index, showing an association between higher MedDiet adherence and better sleep quality at 16- and 34-week's gestation, results aligned with our findings [37]. It should also be considered the burden that women go through during pregnancy may affect their mental health; research often does not recognize the multiple competing demands on women, specifically during pregnancy. However, to our knowledge, the present study is the first randomized clinical trial with a structured intervention based on a MedDiet adapted to pregnancy to evaluate well-being and sleep quality.

Several biological mechanisms have been postulated regarding the relationship between diet and mental health. First, it should be noted that the MedDiet is an easy-to-follow dietary pattern and is not only a healthy diet but also promotes a healthy lifestyle, including cultural and lifestyle elements such as conviviality, seasonality, traditional recipes, physical activity, and culinary activities [38]. These behavioral changes related to lifestyle may also have a therapeutic benefit [29]. Second, the role of diet in mental health may be mediated by inflammatory and oxidative stress pathways [12,13], the modulation of gut microbiota [39] and brain plasticity [40]. A low production of brain-derived neurotrophic factor, a peptide implicated in synaptic plasticity and neuronal survival, has been observed in patients with depression [41]. Moreover, reduced brain-derived neurotrophic factor levels were observed in pregnant women with low sleep quality, as measured by the Pittsburgh Sleep Quality Index, compared to pregnant women with good sleep quality [42]. Interestingly, in a sub-group of the PREDIMED study, significantly higher plasma levels of brain-derived neurotrophic factor were observed in participants allocated to the MedDiet supplemented with nuts group compared to the control arm, whose secretion may be also modulated by diet [43]. The fatty acid profile of the MedDiet, rich in polyunsaturated fatty acids, may also promote mental health, as low polyunsaturated fatty acid intake (mainly omega-3 fatty acids) has been associated with several mental outcomes, including depression [44,45]. Thus, several dietary components, including nutrients and bioactive compounds, are required for healthy brain function and mental health, including the synergic effect between components. Therefore, dietary interventions promoting a healthy dietary pattern rather than a single nutrient may have greater benefits for mental health [46].

Important implications regarding the mental health of the mother may be expected, including a potential benefit during the postpartum period. Maternal mental health alterations, principally anxiety, are associated with several adverse outcomes for both the mother and the offspring, including postnatal depression, pre-term birth and the poor cognitive and behavioral development of the infants [47–50]. Additionally, the estimated prevalence of anxiety disorders across the perinatal period is around 21% [51]. Our results

highlight the need for anxiety and stress screenings during pregnancy, nutritional education, and referrals for evaluation and treatment if necessary. Further research is needed to characterize the impact of the MedDiet on mental health during pregnancy, including the underlying mechanisms, specifically oxidative stress, and the potential benefits for the offspring's mental health. If confirmed, the MedDiet could become an early intervention strategy for the prevention of mental disorders [52].

The major strengths of the present study include a very well-characterized population of pregnant women who followed a structured intervention in a randomized clinical trial. Moreover, the use of different validated questionnaires with clinical applicability to assess mental stress, well-being and sleep quality provided rigor and validity to the results of the study, as well as the ability to analyze various stress-related biomarkers in a subgroup of patients with the aim of measuring stress in an empirical way. The use of validated questionnaires and biomarkers may mitigate the potential misclassification of self-reported data, along with the inherent risk of inaccuracies in the measurements.

The study has some limitations. Firstly, the trial was not designed for this purpose, although maternal stress, well-being and sleep quality were prespecified in the study protocol and assessed from the beginning of the study. Secondly, we were not able to assess long-term dietary intake, including measuring diet before pregnancy or the dietary changes from the beginning of the pregnancy. Most women were of white ethnicity and middle to high socio-economical level; hence, the results should not be extrapolated to other populations with different characteristics. These findings should be considered preliminary and require replication, including reseatch involving other study populations and an evaluation of the underlying mechanisms of action.

#### **5. Conclusions**

In conclusion, a MedDiet intervention significantly reduces maternal anxiety and stress, as well as improving well-being and sleep quality during gestation. Considering the increasing importance of the role of mental health during pregnancy, these findings might imply the promotion of a pregnancy-adapted MedDiet among pregnant women as a powerful public health strategy.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/nu15102362/s1, Table S1. Pregnancy and perinatal outcome of women included in the study; Table S2. Changes in dietary key foods' intake and Mediterranean diet adherence evaluated at baseline and final visits according to intervention groups; Table S3. Changes in nutrient intake and Mediterranean diet adherence evaluated at baseline and final visits according to intervention groups.

**Author Contributions:** Conceptualization, E.G., F.C. (Fàtima Crispi) and F.C. (Francesca Crovetto); methodology, R.C., F.C. (Fàtima Crispi), F.C. (Francesca Crovetto), E.V., R.E. and E.G.; validation, I.C., A.N., R.P., S.C.-B., L.Y., M.G., L.B., M.L., M.L.B., G.C., A.G.-G., O.J.P., I.M., A.M.-À., E.V., M.D.G.- R., R.C., R.E., E.G., F.C. (Fàtima Crispi) and F.C. (Francesca Crovetto).; formal analysis, I.C., F.C. (Francesca Crovetto), S.C.-B.; investigation, R.C., F.C. (Fàtima Crispi), F.C. (Francesca Crovetto), R.E., E.V. and E.G.; resources, E.G.; data curation, I.C., S.C.-B., F.C. (Francesca Crovetto), F.C. (Fàtima Crispi), R.E. and E.G.; writing—original draft preparation, I.C., A.N., R.P., S.C.-B. and F.C. (Francesca Crovetto); writing—review and editing, L.Y., M.G., L.B., M.L., M.L.B., G.C., A.G.-G., O.J.P., I.M., A.M.- À., E.V., M.D.G.-R., R.C., R.E., E.G. and F.C. (Fàtima Crispi); supervision, F.C. (Francesca Crovetto), F.C. (Fàtima Crispi) and S.C.-B.; funding acquisition, E.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** The project was partially funded by a grant from "la Caixa" Foundation (LCF/PR/GN18/10310003 and LCF/BQ/DR19/11740018); Cerebra Foundation for the Brain Injured Child (Carmarthen, Wales, UK); AGAUR under grant 2017 SGR nº 1531; Instituto de Salud Carlos III (PI22/00684, INT21/00027, CM21/00058). S. Castro-Barquero has received support from Margarita Salas fellowship, University of Barcelona. L.Youssef was supported by the grant FJC2021-048123-I, funded by MCIN/AEI/10.13039/501100011033 and by the European Union "NextGenerationEU"/PRTR. L. Benitez was supported by a research grant from the Instituto de Salud Carlos III (CM21/00058). M. Larroya has received support from Hospital Clínic Barcelona (Contracte Clínic de Recerca Emili Letang-Josep Font). F. Crovetto has received support from Centro de Investigaciones Biomédicas en Red sobre Enfermedades Raras (CIBERER).

**Institutional Review Board Statement:** The present study was approved by the Institutional Review Board (HCB-2016-0830) before any participant enrolment.

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

**Data Availability Statement:** The datasets used and/or analyses during the current study are available from the corresponding author on reasonable request.

**Acknowledgments:** We thank the study participants for their personal time and commitment to this trial. We also thank all the medical staff, residents, midwives, and nurses of BCNatal, especially Annachiara Basso; Laura Segales; Marta Dacal; Marta Camacho; and Silvia Gomez, for their support in the recruitment, interventions, and data collection of the trial; Carlos Galante; and Tania Freitas, (Department of Internal Medicine, Hospital Clinic, Barcelona, Spain, Barcelona, Spain), for the support in the Mediterranean diet intervention; Georgina Badosa; and Amaia Helguera (Instituto esMindfulness, Barcelona, Spain), for their support in the stress-reduction intervention; we thank the Clinic-IDIBAPS Biobank for their valuable management of samples. CIBER OBN is an initiative of the Instituto de Salud Carlos III, Spain.

**Conflicts of Interest:** Ayako Nakaki reports personal fees from La Caixa Foundation (Doctoral INPhINIT—RETAINING, fellowship n LCF/BQ/DR19/11740018), during the conduct of the study. Eduard Vieta reports, outside the submitted work, personal fees from Abbott, Allergan, Angelini, Lundbeck, Sage, and Sanofi; grants from Dainippon Sumitomo, Ferrer, and Janssen. Ramon Estruch reports grants from Fundación Dieta Mediterránea, Spain and Cerveza y Salud, Spain; he reports personal fees for given lectures from Brewers of Europe, Belgium, Fundación Cerveza y Salud, Spain, Pernaud-Ricard, Mexico, Instituto Cervantes, Alburquerque, USA, Instituto Cervantes, Milan, Italy, Instituto Cervantes, Tokyo, Japan, Lilly Laboratories, Spain, and Wine and Culinary International Forum, Spain, and non-financial support to organize a National Congress on Nutrition. Eduard Gratacós reports, during the conduct of the study, grants from La Caixa Foundation, grants from Cerebra Foundation for the Brain Injured Child, and grants from AGAUR. Francesca Crovetto reports a personal fee from Centro de Investigaciones Biomédicas en Red sobre Enfermedades Raras, during the conduct of the study. The remaining authors declare no competing interests.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Article* **Dietary Intake of Pregnant Women with and without Inflammatory Bowel Disease in the United States**

**Barbara C. Olendzki 1,\*, Bi-Sek Hsiao 2, Kaitlyn Weinstein 3, Rosemary Chen 3, Christine Frisard 1, Camilla Madziar 1, Mellissa Picker 3, Connor Pauplis 4, Ana Maldonado-Contreras <sup>5</sup> and Inga Peter <sup>3</sup>**


**Abstract:** Background: Pregnancy is a vulnerable time where the lives of mother and baby are affected by diet, especially high-risk pregnancies in women with inflammatory bowel disease (IBD). Limited research has examined diet during pregnancy with IBD. Aims: Describe and compare the diet quality of pregnant women with and without IBD, and examine associations between dietary intake and guidelines during pregnancy. Methods: Three 24 h recalls were utilized to assess the diets of pregnant women with IBD (*n* = 88) and without IBD (*n* = 82) during 27–29 weeks of gestation. A customized frequency questionnaire was also administered to measure pre- and probiotic foods. Results: Zinc intake (*p* = 0.02), animal protein (g) (*p* = 0.03), and ounce equivalents of whole grains (*p* = 0.03) were significantly higher in the healthy control (HC) group than the IBD group. Nutrients of concern with no significant differences between groups included iron (3% IBD and 2% HC met the goals), saturated fat (only 1% of both groups met the goals), choline (23% IBD and 21% HC met the goals), magnesium (38% IBD and 35% HC met the goals), calcium (48% IBD and 60% HC met the goals), and water intake (49% IBD and 48% HC met the goals). Conclusions: Most pregnant women in this cohort fell short of the dietary nutrients recommended in pregnancy, especially concerning for women with IBD.

**Keywords:** diet; pregnancy; IBD; inflammatory bowel disease; dietary guidelines

#### **1. Introduction**

Pregnancy is a critical time for the intergenerational transmission of health [1–4]. Pregnant women with active inflammatory bowel disease (IBD), a chronic disease characterized by inflammation of the gastrointestinal tract [5] are considered to be at higher risk of poor pregnancy outcomes such as preterm birth, low birthweight or small for gestational age (SGA), spontaneous abortion, and stillbirth, and comprise an increased percentage of Cesarean deliveries compared to women in remission or without IBD [6–9]. The prevalence of IBD has been increasing worldwide [5]; thus, improving the health of pregnant women with IBD is essential to decreasing their risk for adverse pregnancy outcomes.

A balanced perinatal diet can support optimal health for pregnant women and have a long-term impact on their offspring [10–13]. Patients with IBD are already prone to nutrition deficiencies due to factors such as restrictive diets, nutrient loss, drug–nutrient interaction, and decreased absorption from the ileum [14]. Furthermore, reduced oral

**Citation:** Olendzki, B.C.; Hsiao, B.-S.; Weinstein, K.; Chen, R.; Frisard, C.; Madziar, C.; Picker, M.; Pauplis, C.; Maldonado-Contreras, A.; Peter, I. Dietary Intake of Pregnant Women with and without Inflammatory Bowel Disease in the United States. *Nutrients* **2023**, *15*, 2464. https:// doi.org/10.3390/nu15112464

Academic Editors: Louise Brough and Gail Rees

Received: 2 May 2023 Revised: 20 May 2023 Accepted: 23 May 2023 Published: 25 May 2023

**Copyright:** © 2023 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/).

intake and chronic inflammation increases nutrient needs among IBD patients [15,16]. Two reports have explored the diets of pregnant women in the Norwegian Mother and Child Cohort (MoBa). The first study found that compared to pregnant women without IBD, pregnant women with IBD were less likely to adhere to a traditional Norwegian dietary pattern characterized by a high intake of lean fish or fish products, potatoes, rice porridge, cooked vegetables, and gravy, and were more likely to adhere to a Western dietary pattern with higher intake of foods and beverages rich in sugar and saturated fats [17,18]. Moreover, pregnant women with IBD who did adhere to the traditional Norwegian diet had lower odds of having an SGA infant [17]. The second study found that pregnant women with IBD consumed a lower proportion of protein from dairy products compared to pregnant women without IBD. In this case, a reduced intake of protein from dairy was associated with a lower risk of having an SGA infant [18].

Maternal diet during pregnancy has also been linked to the infant microbiome composition, which is critical for the priming of a balanced immune system during early life [19]. Importantly, our team has demonstrated that infants born to women with IBD have less diverse microbiomes and higher levels of fecal calprotectin (a biomarker of intestinal inflammation) compared to the infants of women without IBD [4,20]. Along with emerging reports demonstrating the mediating role of the gut microbiota in the effectiveness of dietary interventions for IBD management [21,22], this finding suggests that improving dietary patterns during pregnancy may beneficially modify the microbiome composition, thereby promoting both maternal and infant health. This hypothesis is being explored by the MELODY (Modulating Early Life Microbiome through Dietary Intervention in Pregnancy) Trial [12].

Diet has been increasingly integrated into IBD management, and studies demonstrate the effectiveness of dietary interventions for inducing IBD remission [23–25]. In adults, the specific carbohydrate diet (SCD); the Mediterranean diet; the low fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (low FODMAP) diet; and the anti-inflammatory IBD (IBD-AID) diet are among those that have shown efficacy in reducing disease activity and symptoms [23]. Yet, informational resources on nutrition for pregnant women with IBD are sparse. The USDA MyPlate website focuses on a variety of food groups with only broad suggestions of foods and meal plans specific to pregnancy and postpartum needs [26]. The 2014 and 2017 American College of Obstetricians and Gynecologists (ACOG) guideline statements seem focused on nutrients that may be obtained by taking a prenatal vitamin, rather than on whole foods [27,28]. The 2019 American Gastroenterological Association's Inflammatory Bowel Disease in Pregnancy Clinical Care Pathway report encourages nutrition consultation for specific nutrient deficiency and weight gain patterns in this population, but with few details on compliance to guidelines [29]. In keeping with these publications, pregnant women may hear only general advice from health care providers to take a prenatal vitamin, follow a healthy diet, limit caffeine intake, avoid alcohol and tobacco, and observe caution with seafood [30,31]. However, while a prenatal vitamin may be recommended in addition to a healthy diet, it cannot supply all the nutrients that are needed to promote healthy and low-risk pregnancies [32].

While diet can support IBD management, with the potential to positively benefit perinatal as well as longer-term health outcomes, little is known about the quality of dietary patterns among pregnant women with IBD in the United States (US), a country with a high prevalence of the disease. Therefore, the objectives of the current study are to describe the dietary patterns and diet quality of pregnant women with and without IBD living in the US, and to examine the associations between dietary patterns, diet quality, and dietary guidelines for pregnancy established by the Society for Obstetricians and Gynaecologists of Canada; the American College of Obstetricians and Gynecologists; the World Health Organization Guidelines; the Academy of Nutrition and Dietetics; the Royal College of Physicians of Ireland; the National Institutes of Health Daily Recommended Intake; and UpToDate [27,30].

#### **2. Methods**

We conducted a case–control study nested into our ongoing MELODY Trial, which is a prospective non-randomized diet intervention trial testing the effects of IBD antiinflammatory diet (IBD-AID) during the third trimester of pregnancy on maternal IBD activity and microbiome composition in mothers and their babies [12]. Pregnant women with and without IBD were recruited nationwide for this trial. Study participants were identified by clinical research coordinators in outpatient gastrointestinal clinics; alternatively, pregnant women reached out if interested after seeing posts on the websites or Facebook accounts of the Crohn's and Colitis Foundation or the Center for Applied Nutrition at the University of Massachusetts. Written informed consent was obtained from all eligible participants. The current case–control study examines dietary assessments conducted at the 27th–29th weeks of gestation prior to any dietary intervention, between January 2019 and December 2022.

The study was approved by the Institutional Review Boards at each institution (IRB docket #H00016462 at the University of Massachusetts Chan Medical School and #18–01206 at the Icahn School of Medicine). The inclusion criteria included: pregnant women carrying a singleton pregnancy, and a documented IBD diagnosis or lack thereof (for healthy controls, HC). The diagnosis of IBD was based on the patient's history supported by clinical documentation. The exclusion criteria were an inability to provide informed consent, HIV/AIDS, multi-fetus pregnancy, fetal chromosomal or structural abnormalities, intrauterine growth restriction, active infection (including chorioamnionitis or sepsis), alcohol use disorder, renal disease, or a dietary regime that conflicts with the intervention diet. Additionally, pregnant IBD patients who had active perianal or extra-intestinal disease or were treated with antibiotic therapy or steroids at recruitment, as well as women scheduled for C-section prior to week 37, were excluded [12]. The final selection of participants is shown in Figure 1.

**Figure 1.** Participant Flow.

#### *2.1. Dietary Assessment*

We performed three 24 h dietary recalls (24 HRs) and a specially designed pre- and probiotic food frequency questionnaire for more detailed detection of food groups than provided by the 24 HR (IBD-AID FFQ) [33–38]. The 24 HR were performed using the University of Minnesota Nutrition Coordinating Center's (NCC) Nutrition Data System for Research (NDSR) software (current version: NDSR, 2022, updated yearly) as previously described [12,39,40]. Specifically, trained dietitians administered 24 HR on two weekdays and one weekend, by phone, between 27 and 29 weeks of pregnancy. The 24 HR also included assessments of dietary supplements. The IBD-AID FFQ was self-administered online using REDCap, as previously reported by us [12]. The dietary assessments were conducted from 2019 to 2022.

#### *2.2. Dietary Quality Assessment*

Diet quality was estimated from the 24 HR recalls using the standard Alternative Healthy Eating Index—2010 (or AHEI-2010) score (range: 0–110), with higher scores representing healthier diets [39,41,42], and the IBD-AID FFQ score (range: 0–26) [12,37,43]. The IBD-AID FFQ was developed by Barbara Olendzki and her team at the Center for Applied Nutrition, Umass Chan Medical School, and addresses a gap in the nutrition information available from the 24 HR recalls, particularly with regard to pre-and probiotic foods. We found construct validity in using the IBD-AID FFQ, as pre- and post-dietary intervention changes correlated with bacterial abundance and serum cytokine levels [43]. The beneficial foods of the IBD-AID FFQ were matched with the food categories of the validated Alternate Healthy Eating index-2010 or AHEI-2010 [44]. Namely, the IBD-AID FFQ assesses the intake of 15 food groups and components. Beneficial Nutrient Score is calculated from all components and ranges from 0 to 26. Raw Score = [prebiotic foods] + [probiotic foods] + [Beneficial Nutrient Score] − [adverse foods]. The standard score eliminates the negative values, so if the raw score is <0, then the standard score is 0. If the raw score is >0, then the standard score is the raw score. In addition, the IBD-AID FFQ measures prebiotic foods (>3 servings/day), probiotic foods (>2 servings/day), and foods associated with gastrointestinal symptoms and poor IBD outcomes, including: refined carbohydrates (<2 servings per day), lactose (0 servings), certain grains (wheat, corn; 0 servings/day), processed foods (0 servings per day) and foods high in saturated (<7% of calories) or trans fats (0 servings/day).

We scored each beneficial food component (to correlate with the AHEI) from nonadherence = 0, to perfect adherence = 26. Pre- and probiotic foods were scored separately, with a perfect score being >3 and >2 servings per day, respectively. The IBD-AID FFQ total score = (prebiotic foods + probiotic foods + beneficial foods) minus adverse foods, with higher scores representing higher servings of beneficial foods minus adverse foods.

#### *2.3. Statistical Methods*

The demographic characteristics were presented using means and standard deviations for continuous variables and compared between pregnant women with and without IBD using a two-sample *t*-test. The categorical variables were described using counts and proportions, with *p*-values calculated via a Fisher's exact test. To minimize the bias of a particular day where food intake is not typical, the reported servings were averaged across three 24 HR. The means and standard deviations summarized nutrients, components of interest from foods, and food group servings on the IBD-AID FFQ using two-sample *t*-tests quantifying between-group differences for normally distributed data and Wilcoxon rank sum tests for skewed outcome variables. The proportions and standard deviations described the proportion of participants who met the guidelines for nutrients at baseline, with chi-square tests measuring between-group differences.

#### **3. Results**

#### *3.1. Participant Characteristics*

The demographic characteristics of our study population, comprising 82 healthy controls (HC) and 88 participants with IBD (Crohn's disease (CD) = 80, and ulcerative colitis (UC) = 8), are presented in Table 1. On average, women were 34 ± 4 years of age, predominantly white (91%) and non-Hispanic (90%). Most were married (93%), had a 4-year college degree or greater (90%), were employed full-time (71%), and had a household annual income of more than USD 100,000 per year (72%). Most were non-smokers (90%) and took a daily prenatal vitamin (91%). There were no significant differences between HC and IBD participants, except in profession and religious affiliation. The proportion of

women with IBD who identified as Jewish was higher than the in the HC cohort (*p* < 0.001). The women with IBD reported working in more scientific technical professions than the HC group, while HC women reported working in more skill-, craft-, and health-based professions compared to IBD participants (*p* = 0.02). The average disease duration of the IBD participants was 14 years for CD and 10 years UC. We found 89% remission for our CD patients (three did not complete this section of the form so 4% were N/A), and 88% remission for our UC patients.

**Table 1.** Demographic Characteristics of Healthy Controls vs. Pregnant Women with Inflammatory Bowel Disease at Baseline (*n* = 170).



**Table 1.** *Cont.*

IBD—inflammatory bowel disease. \* *p*-value < 0.05.

The list of IBD-directed medications is provided in Table 1.

#### *3.2. Nutrient Intake and Dietary Quality for Pregnant Women with and without IBD*

In total, we collected 496 24 HR at 27 and 29 weeks of pregnancy for the 170 women included in the study.

We estimated diet quality using the AHEI-2010 (scored 0–110), which incorporates components of evidence-based recommendations to identify future risk of chronic disease [40,44]. Overall, the participants in the study had a higher dietary quality (66.6 in IBD group, 67.9 in HC group) compared to the average American (47.6 ± 10.8) [44,45]. There were no differences in dietary quality between pregnant women with and without IBD.

Table 2 presents the average nutrients and components of interest from foods sources, excluding any dietary supplements. On average, the intake of nutrients was comparable between pregnant women with and without IBD, with some notable exceptions. The percentage of calories from monounsaturated fatty acids (MUFAs), and the intake of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), were significantly higher

in pregnant women with IBD than HC. Conversely, healthy control participants reported a significantly higher intake of animal protein (their total protein intake was similar), whole grains, lactose, and zinc. There were no differences in calories consumed per group (approximately 2000 kcals/day).

**Table 2.** Nutrients and Components of Interest from Foods in Pregnant Women with Inflammatory Bowel Disease vs. Healthy Controls.



**Table 2.** *Cont.*

IBD—inflammatory bowel disease; SD—standard deviation; SFA—saturated fatty acids; MUFA—monounsaturated fatty acids; PUFA—polyunsaturated fatty acids. \* *p*-value < 0.05.

The proportion of women meeting the pregnancy dietary guidelines for nutrients from food is shown in Table 3. The following dietary guideline goals were significantly lower in women with IBD than in the controls; 63% vs. 80% met the thiamine guideline (*p* = 0.01), and 38% vs. 56% met the B6 guideline (*p* = 0.02). Both the IBD and HC groups mostly met the guidelines for caffeine (94% and 96%, respectively) and taking a prenatal vitamin (93% and 90%, respectively). A total of 39% of women with IBD met the zinc guideline vs. 54% of the HCs (*p* = 0.05). Protein intake was not optimal for either group (with 66% meeting the guidelines). A total of 38% of women with IBD met the guidelines for total fat intake, and 45% met the guidelines in the HC group (*p* = 0.31). Nutrients of concern include iron (3% IBD and 2% HC met the goals), saturated fat (only 1% of both groups met the goals), choline (23% IBD and 21% HC met the goals), magnesium (38% IBD of 35% HC met the goals), calcium (48% IBD and 60% HC met the goals), and water intake (49% IBD and 48% HC met the goals), with no significant differences between the groups.

**Table 3.** Proportion of Women Meeting the Dietary Guidelines for Nutrients at Baseline (*n* = 170).



#### **Table 3.** *Cont.*

IBD—inflammatory bowel disease; SD—standard deviation. \* *p*-value < 0.05. <sup>1</sup> Guidelines were selected from a review of the following organizations: Society for Obstetricians and Gynaecologists of Canada; American College of Obstetricians and Gynecologists; World Health Organization Guidelines; Academy of Nutrition and Dietetics; Royal College of Physicians of Ireland; National Institutes of Health Daily Recommended Intake; and UpToDate.

Table 4 presents the average food group servings of the IBD-AID FFQ for the subset of women who completed the questionnaire at baseline. A lower percentage of women from both groups completed this questionnaire, as it was self-administered and not facilitated by a dietitian (68/88, 77.3% IBD and 65/82, 79.3% HC). Prebiotics were significantly higher in mothers with IBD (6.3 vs. 4.7 in HC, *p* < 0.001), as were non-wheat fiber/grains (6.3 vs. 2.8 in HC, *p* < 0.001). Servings of adverse foods (i.e., higher in sugar, wheat, lactose, and saturated and trans fats) were lower in women with IBD than in the HCs (7.1 vs. 15.7, *p* < 0.001), and lean proteins (2.4 IBD vs. 3.4 HC, *p* = 0.04) and servings of beneficial beverages (apple cider, low-sugar beverages with added probiotics, juice (no added sugar), non-dairy milk, homemade smoothies, honey tea, tomato juice, V8 juice, coconut water, tea, and coffee substitutes (chicory root)) were significantly lower in pregnant participants with IBD than those without IBD (1.8 vs. 6.6, *p* < 0.001); however, total water intake was similar between groups. The Beneficial Nutrient Score (a calculated score of prebiotics, probiotics, overall dietary quality, and intake of foods thought to be adverse) was significantly lower in HC participants than in those with IBD (15.6 vs. 14.0, *p* = 0.04), as were the IBD-AID FFQ total raw and standard scores (16.4 IBD vs. 4.6 HC, *p* < 0.001; 16.9 IBD vs. 8.4 HC, *p* < 0.001, respectively).



IBD-AID—inflammatory bowel disease anti-inflammatory diet. \* *p*-value < 0.05. <sup>1</sup> Prebiotics are foods containing fiber that feed commensal organisms. <sup>2</sup> Probiotics are fermented foods that contain live bacteria. <sup>3</sup> Adverse foods include ultra-processed foods and foods high in added sugars. N optimal adverse foods goal is zero servings per day and counts negatively toward total score. <sup>4</sup> Lean protein score includes beans/legumes, seafood, and poultry. <sup>5</sup> Fiber/grains include foods such as oats, barley, and miso. <sup>6</sup> Beneficial beverages include beverages such as those with added probiotics, non-dairy milks, homemade smoothies, no-sugar-added fruit and vegetable juices, coconut water, tea sweetened with honey, etc. <sup>7</sup> Foods with unknown effects have yet to be determined in research. <sup>8</sup> Beneficial Nutrient Score is calculated from all components and ranges from 0 to 26. <sup>9</sup> Raw Score = [prebiotic] + [probiotics] + [Beneficial Nutrient Score] − [adverse]. <sup>10</sup> The standard score eliminates the negative values, so if the raw score is <0, then the standard score is 0. If the raw score is >0, then the standard score is the raw score.

#### **4. Discussion**

The current study addresses the gap in knowledge about the diets of pregnant women with IBD through an analysis of baseline data from The MELODY Trial. We observed that pregnant women with and without IBD do not consume most of the nutrients and food components recommended during pregnancy by established government and researchbased organizations.

Specifically, we found that although most women reported taking a prenatal supplement, in hopes of supplementing the inadequate intake of nutrients from food sources, pregnant women IBD and the HC group fell short of most nutrients recommended in pregnancy from food sources alone. Of particular concern in women with IBD are the dietary micronutrients zinc, iron, calcium, magnesium, choline, folate, B6, B12, water, and fiber [46–50]. Patients with IBD require additional assistance to compensate for increased nutritional needs and poor absorption, whereby simply adding the nutrient does not guarantee that it will be well absorbed in the body [51,52].

Iron needs increase during pregnancy, especially in women with IBD, who may struggle with significant inflammation, anemia, and dysbiosis, leading to poor cardiovascular outcomes and suboptimal gestational weight gain [12,53–57]. The recommended daily allowance for pregnant women is around 27 mg per day. Many factors can influence iron

absorption in the body, such as certain nutrient–nutrient interactions, including nutrient inhibitors (such as calcium) and enhancers (i.e., ascorbic acid) [58]. Furthermore, non-heme iron, primarily found in plant sources, is less easily absorbed by the body than heme iron, primarily found in animal sources, so the recommended amount of iron for vegetarians and vegans is 1.8 times greater [59]. In this study, only 3% of women with IBD and 2% of the HCs met the dietary guideline for (animal-based) iron. It is estimated that between 36 and 90% of people with IBD have iron-deficient anemia (IDA) [47] and that 15–20% of pregnant people have IDA [60]. This can lead to worse disease outcomes for both mothers and infants [61]. It is important that pregnant women meet the dietary guidelines set for iron first through food consumption, subsequently adding supplements as the need is determined.

The adequate intake of fiber is 28 g per day [62]. Fiber is typically not found in prenatal vitamins but is an especially important dietary component during pregnancy. Adequate fiber intake during pregnancy is crucial and may help alleviate iron-induced constipation. It is also helpful for reducing certain problems during pregnancy, such as inflammation, gestational diabetes mellitus, and cardiovascular outcomes [63,64]. However, even among the general population, dietary fiber intake falls below the recommended 28 g per day [56]. A study using 2001–2014 National Health and Nutrition Examination Survey (NHANES) data found that pregnant, non-lactating women aged 20–40 (*n* = 1003) had a mean total daily intake of 17.3 g of dietary fiber [65]. Only 33% of pregnant women with IBD in our study achieved the recommended fiber intake of >28 g per day. One mechanism by which diet can provide protection from IBD is through the addition of plant-based, fiber-rich foods that promote short-chained fatty acid-producing bacteria, which have been shown to support mucosal barrier integrity [23]. Adequate fiber intake important not only for women's health during pregnancy [62,64,66–69], but also for preventing infant outcomes such as SGA, preterm birth, and fetal growth restriction [70].

Both groups were consuming an excess of foods with saturated fat, which can lead to an elevated risk of gestational diabetes [71]. Notably, the IBD group consumed less zinc and calcium than did the HC group. Low calcium and zinc intakes have been correlated with risk for poor outcomes for both mother and child [46,47,72]. Prenatal vitamins may not overcome a low dietary intake of these nutrients [73]. In our sample, predominantly white women with a college degree and who earned higher than the national average income consumes a fairly healthy diet before entering the study, which is why we do not see many significant between-group differences.

A recent study revealed that almost no supplements met nutritional needs in the doses that are required for pregnant women (without excess) [32]. Prenatal vitamins are recommended for pregnancy, especially to provide folic acid, EPA/DHA, iron, and vitamin D. Vitamins are, by definition, recommended to supplement the diet, not to replace the inclusion of the nutritious foods needed during pregnancy. Nutrients are digested and absorbed most effectively in the complex milieu of the foods themselves and the complementary enzymes and microbiota that facilitate absorption. Absorption is biologically complex, and simply adding a nutrient does not mean it will be well absorbed [51,52]. Therefore, many individuals with IBD require additional assistance to compensate for increased nutritional needs and poor absorption, as the nutritional needs of pregnancy for women with IBD may be uniquely challenging. Despite the excellent consensus recommendations by the International Organization for the Study of Inflammatory Bowel Diseases (IOIBD), there is no guidance for pregnancy with IBD or prenatal advice for the prevention of IBD. The IOIBD, based in large part on epidemiological studies, does not cover altering the textures of foods (such as pureeing fiber) for ease of absorption, one of many considerations that goes beyond the nutrients themselves and addresses malnutrition and malabsorption [74].

However, even with the limited existing evidence, healthcare providers appear to be inadequately counseling pregnant IBD patients on diet. Apart from referral to a dietitian for gestational diabetes, the prescription of dietary guidelines for pregnant women is lacking, increasing the risk for detrimental outcomes, especially for those with high-risk pregnancies. Current data show that only 37% of pregnant IBD patients reported receiving education from any physician about IBD in pregnancy [75]. Even worse, only 10% of patients reported having received pregnancy-specific information from their gastroenterologists, and of those who received information, 48% found the information to be insufficient [75]. Yet, several studies have demonstrated that those women who receive dietary counseling during pregnancy eat more fruits and vegetables, promoting the healthy growth and development of the fetus [76–79], suggesting that more dietary interventions are needed.

This apparent lack of patient education is not due to physician ignorance regarding pregnancy and IBD. In fact, when assessed with the Crohn's and Colitis Pregnancy Knowledge Score (CCPKnow), 91.8% of physicians demonstrated very good knowledge, with gastroenterologists scoring the highest [75,80]. In contrast, only 10.3% of patients exhibited very good knowledge when assessed using the CCPKnow, with 44.8% demonstrating poor knowledge levels [80]. This discrepancy between physician and patient CCPKnow scores highlights the need for increased patient counseling, particularly from gastroenterologists, who exhibit the highest CCPKnow scores [80]. Healthcare providers should first evaluate pregnant individuals at risk of nutrient deficiency and excess, and subsequently provide evidence-based suggestions for supplementation [81]. Importantly, specific suggestions and menu plans with foods that contain essential nutrients and other components, such as fiber and pre-and probiotics, should be presented in actionable formats.

We acknowledge that the assessment of diet is prone to limitations, including selfreport bias, under- or overestimation, memory bias, and weakness in the methodology. Further, we did not account for the influence of the environment, medication, dietary supplementation, or IBD activity status on nutrition, although most of our IBD patients were in remission. We conducted a large part of this study during the pandemic, when changes to food intake occurred, and this would have affected both arms of the study. This study is further limited to pregnant women in the United States of higher socioeconomic status, as they may have better access to medical care and foods that may not be generalizable to other groups and countries. However, our IBD and HC study groups were well-balanced regarding age, education, and income, suggesting that the reported differences (or lack thereof) in dietary intake are representative of this cohort.

#### **5. Conclusions**

While many women adhered to taking a prenatal supplement, both pregnant IBD participants and HC participants fell short of most dietary nutrients recommended in pregnancy through dietary sources alone, especially micronutrients and fiber. The consumption of animal protein, lactose, zinc, and whole grains was significantly lower in pregnant women with IBD compared to the HCs. Large epidemiological and dietary intervention studies are warranted to improve the nutritional recommendations for pregnant women with and without IBD while addressing malnutrition and malabsorption. Future research should consider pregnancy outcomes and the effects on offspring, and determine the causes of dietary deficiencies and excess, to ultimately inform and improve the quality of provider training and patient education, especially in the setting of IBD.

**Author Contributions:** Conceptualization, B.C.O., A.M.-C. and I.P.; methodology, B.C.O., A.M.-C., C.F. and I.P.; formal analysis, C.F.; investigation, B.C.O., B.-S.H., K.W., R.C., C.P., A.M.-C. and I.P.; resources, M.P., C.M., K.W., R.C., C.P. and B.-S.H.; data curation, C.M., M.P., K.W. and R.C.; writing—original draft preparation, B.C.O., A.M.-C. and I.P.; writing—review and editing, B.C.O., B.-S.H., K.W., R.C., M.P., C.M., C.F., C.P., A.M.-C. and I.P.; visualization, C.F.; supervision, B.C.O., A.M.-C., M.P., C.M. and I.P.; project administration, M.P.; funding acquisition, B.C.O., A.M.-C. and I.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** We are grateful to our participants. The MELODY Trial is funded by The Leona M. and Harry B. Helmsley Charitable Trust.

**Institutional Review Board Statement:** The study was approved by the Institutional Review Boards of the Icahn School of Medicine at Mount Sinai (project identification code 18–01206) and the University of Massachusetts (project identification code H00016462).

**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 on request from the corresponding author. The data are not publicly available due to data privacy restrictions.

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

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

**Anna Demuth 1, Joanna Ratajczak 1,\*, Urszula Czerniak <sup>1</sup> and Katarzyna Antosiak-Cyrak <sup>2</sup>**

	- 61-871 Pozna ´n, Poland; antosiak-cyrak@awf.poznan.pl

**\*** Correspondence: jratajczak@awf.poznan.pl

**Abstract:** Health education (HE), an educational process that leads to increased nutritional awareness and improved health, is one of the factors influencing diet quality (DQ) during pregnancy. The aim was to evaluate the DQ of pregnant women and its determinants considering their HE. The study included 122 pregnant women aged 20–40 years. DQ was assessed using the Kom-PAN® questionnaire and the Pro-Healthy Diet Index (pHDI). Data collected included dietary habits, sociodemographic data, education level, place of residence, and maternal lifestyle-related characteristics, namely, pre-pregnancy weight, trimester of pregnancy, and pre-pregnancy and pregnancy physical activity (PA). Weekly energy expenditure was determined using the Polish version of the PPAQ questionnaire. HE at school more than tripled the odds of a higher DQ. Women in their second trimester were 54% more likely to have a higher DQ than women in their third trimester of pregnancy. Undertaking pre-pregnancy PA increased the odds of a higher DQ 2.5 times. Comparative analyses performed in a group of women with HE (HEG, n = 33) and without HE (nHEG, n = 89) showed better DQ in the former, but this was still unsatisfactory in health-promoting properties. The results obtained showed that the HE and trimester of pregnancy and pre-pregnancy Pa influenced DQ in pregnant women.

**Keywords:** pregnancy; health education; diet quality; diet quality determinants

#### **1. Introduction**

Health and dietary behaviors before and during pregnancy consistently remain an important and ongoing area of research [1–3]. Epidemiological studies have highlighted the significance of assessing diet quality and its determinants as the consequences of inadequate nutrition expose not only women but also their children to poorer health outcomes for the rest of their lives [4].

Among the factors influencing diet quality in pregnant women age, socioeconomic and lifestyle variables are the most commonly reported factors [5,6]. Other factors include pre-pregnancy BMI, physical activity, smoking, and alcohol consumption [7–12] but also nutritional knowledge, which has been reported to play an important role in pregnancy and influencing dietary choices. Obesity and overweight are currently a serious problem among women of reproductive age. The Central Statistical Office in Poland shows that the percentage of women of reproductive age (20–39 years) with excessive body weight (BMI > 25 kg/m2) increased from 25.8% to 31.3% between 2009 and 2019 [13]. Efforts to provide appropriate health education and care for pregnant women should be intensified due to the fact that almost one in three Polish women of reproductive age has problems with maintaining a healthy body weight.

Despite the proven link between maternal nutrition and pregnancy outcomes, many pregnant women do not follow the dietary recommendations. Moreover, behaviors such as a sedentary lifestyle and unhealthy eating habits are common among pregnant women

**Citation:** Demuth, A.; Ratajczak, J.; Czerniak, U.; Antosiak-Cyrak, K. Is Health Education among the Decisive Factors for the Diet Quality of Pregnant Women in Poland? *Nutrients* **2023**, *15*, 2627. https://doi.org/10.3390/ nu15112627

Academic Editors: Louise Brough and Gail Rees

Received: 28 April 2023 Revised: 1 June 2023 Accepted: 2 June 2023 Published: 4 June 2023

**Copyright:** © 2023 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/).

worldwide, including Poland [14–19]. It has been indicated that such behaviors during pregnancy are caused by both non-adherence to the recommendations and insufficient health education and health promotion [20].

In Poland, health education is understood as a didactic and educational process in which pupils starting from primary school learn how to maintain and improve their own and other people's health, how to create a health-favorable environment, and, in the case of illness or disability, how to actively participate in its treatment, cope with its negative effects, and reduce its consequences [21]. An important part of health education in schools is the development of appropriate eating behaviors, by which pupils acquire competences in the knowledge of basic nutrients and their role in the body, the preparation and storage of food, the knowledge of diseases related to poor nutrition, the knowledge of labeling food packages, and the ability to prepare menus for different groups of people [22]. Health education also plays an important role in shaping health-promoting attitudes by practicing hygienic behaviors that are safe for health, as well as the use of prevention, the practice of physical activity, and the consolidation of knowledge about its benefits.

Researchers emphasize that the level of knowledge and awareness may influence the level of acceptance of educational messages, and, therefore, their effectiveness, which is why health education may be particularly important [23,24]. As schools play an important role in meeting the nutritional needs of children and adolescents and in shaping appropriate behavior [21], health education should already start at an early stage of education. Unfortunately, this type of education has not always been included in the compulsory school curriculum, Poland being an example. In Poland, health education was included in the core curriculum of general education for all types of schools only in 1997 [25]. However, it is not a separate school subject, but its content has been included in many subjects, e.g., biology, family life education, social studies, and safety education. An important step in the Polish education system was linking health education with physical education 2013. Since then, physical education has been playing a leading role in health education. According to the Education Law in Poland [26], a child's compulsory education starts at the beginning of the school year in the calendar year in which the child turns 7 and lasts until the end of primary school, no longer than until the age of 18. Therefore, all people who started primary school in 1997 or later have achieved the expected learning outcomes for health education. In contrast, people who started school before 1997 did not receive health education classes at school, and their knowledge about health behavior comes from a variety of sources.

So far, little attention has been paid to the impact of early health education on diet quality in pregnant women. Considering the importance of diet in pregnant women and studies assessing diet quality, the aim of this study was to evaluate the diet quality of pregnant women and identify its determinants with particular attention to health education.

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

#### *2.1. Survey Design and Sample*

The survey was conducted in Pozna ´n in free birthing schools, i.e., where access is universal, and there is no extra cost for parents to attend. Due to the lack of official data on the percentage of pregnant women attending antenatal classes, the sample size was estimated based on the list of women attending the classes in three randomly selected birthing schools between September and December 2019. The total number of the women attending antenatal classes was 170, all of whom were asked to participate in the study. Slovin's formula (see below) was used to calculate the sample size with a 5% margin of error and 95% confidence interval [27]. The minimum number of the necessary sample size to meet the criteria listed above was 119. Of the total number of 170 women, 129 (75.9%) agreed to participate in the study, and 41 (24.1%) refused. In addition, seven of the women were excluded from the study due to incomplete questionnaire answers. Finally, 122 women were included in the study.

$$n = \frac{N}{1 + \text{Ne}^2}$$

n—sample size; N—population size; e—margin error.

The study took the form of a direct, individual questionnaire survey. The participants completed the questionnaire on their own. In case of any problems with understanding the questions, the interviewer was helpful in explaining the inaccuracies. The collected information included eating habits, sociodemographic data, i.e., age, level of education (university/secondary/vocational/primary), presence of health education in school, place of residence (urban/rural), and maternal lifestyle-related characteristics, i.e., pre-pregnancy weight, trimester of pregnancy, and pre-pregnancy and pregnancy physical activity (PA). In addition, the respondents were asked about the year in which they started primary school. This made it possible to distinguish two groups of women: with health education (i.e., women who started primary school education after 1997; HEG; n = 33) and without health education (i.e., women who started primary school education before 1997; nHEG; n = 89). The distinguished groups were further used for comparative analyses. The study was approved by the Bioethics Committee of the Poznan University of Medical Sciences (reference no. 878/19, 12 September 2019). All women gave written consent to participating in the study.

The KomPAN® questionnaire provided data on eating habits and enabled the calculation of the Pro-Healthy Diet Index (pHDI), which gave information on diet quality [28]. The index was the sum of the daily intakes (times/day) of 10 food groups with potentially beneficial outcomes: 1. wholemeal bread; 2. grains and coarse-ground groats; 3. milk (including flavored milk, cocoa, coffee with milk); 4. fermented milk beverages; 5. curd; 6. white meat; 7. fish; 8. legumes; 9. fruits; and 10. vegetables. Each respondent reported habitual consumption of the above-mentioned products by indicating one of the six frequency categories: never, 1–3 times a month, once a week, a few times a week, once a day, and a few times a day. Those categories were converted to daily frequency expressed as times/day: never (0), 1–3 times a month (0.06), once a week (0.14), a few times a week (0.5), once a day (1.0), and a few times a day (2.0). The pHDI values ranged from 0 to 100 points and were calculated using the formula below. The pHDI values in the range of 0–33 points were defined as low, in the range of 34–66 points as moderate, and in the range of 67–100 points as high. The higher the value, the greater the intensity of health-promoting properties in the diet and, therefore, the better quality of the diet [28].

pHDI in points <sup>=</sup> <sup>100</sup> <sup>20</sup> <sup>×</sup> sum of the consumption of 10 food groups (times/day)

The Polish version of PPAQ questionnaire enabled us to determine the weekly energy expenditure (MET hour/week<sup>−</sup>1) [29]. The respondents self-assessed their physical activity levels by filling in a questionnaire consisting of 33 items grouped into the following activity categories: household/caregiving (15 items), occupational (5 items), sports/exercises (7–9 items), transportation (3 items), and inactivity (3 items). The declared duration of performance of particular tasks was assigned fixed numbers of minutes (0; 0.12; 0.50; 1.0; 2.0; 3.0) and then multiplied by the number of days of performance of the tasks per week. The obtained values were then multiplied by intensity (MET) in accordance with the guidelines in "Compendium of Physical Activities: an update of activity codes and MET intensities" [30], thus obtaining the energy expenditure measured in Metabolic Equivalent of Task (MET). The following levels of intensity were assigned to the different activities: sedentary < 1.5 METs; light 1.5–<3.0 METs; moderate ≥3.0–≤6.0 METs; and vigorous > 6.0 METs. In addition, the respondents were asked if they had undertaken physical activity before pregnancy. The participants could choose between yes/no answers.

#### *2.2. Analysis*

All statistical analyses were performed using STATISTICA 13 (Dell Inc.; Tulsa, OK, USA, StatSoft Polska, Cracow, Poland, 2017). The threshold of statistical significance was set at *p* ≤ 0.05. The distribution of the variables was tested using the Shapiro–Wilk test. For quantitative variables, arithmetic means and standard deviations (SD) were

calculated. The median, lower, and upper quartiles were calculated for the frequency of consumption of 10 product groups. The Mann–Whitney (Z) test was used to test the significance of differences between the distinguished groups. The Chi-square test (χ2) was used for comparative analysis of categorical variables. The Spearman's rank correlation coefficients (r) were used to assess the presence and strength of the associations between diet quality and consumption of selected food products, as well as sociodemographic data and maternal lifestyle-related variables. The interpretation of the correlation coefficients was as follows: weak (<0.3), moderate (0.3 to <0.5), strong (0.5 to <0.7), and very strong (≥0.7) correlation [31]. To identify the determinants of diet quality, multiple regression models were run with diet quality as the dependent variable. Only factors that were significantly correlated with diet quality were included in the models. Logistic regression analysis was used to assess the odds of having a higher-quality diet. The dependent variable was diet quality as assessed by the Pro-Healthy Diet Index (pHDI). The categorization of the two groups for the dependent variable in the logistic regression was based on pHDI values. Values ≤ 33 points were assigned to the "lower quality diet" category, whereas values > 33 points were assigned to the "higher quality diet" category. Odds ratios (ORs) and 95% confidence intervals (95% CI) were calculated.

#### **3. Results**

#### *3.1. Group Characteristics*

The characteristics of the participants are shown in Table 1. The mean age was 27.7 ± 3.7 years. The women from the health education group were younger than women in the group without health education (23.4 ± 1.5 vs. 29.3 ± 2.9; Z = −8.45; *p* < 0.001). The percentage of the women with a higher level of education was greater in nHEG than in HEG (76.4% vs. 54.6%; χ<sup>2</sup> = 6.58; *p* = 0.037). The vast majority of participants had a higher level of education (70.5%) and lived in urban areas (68.9%). Of the participants, 49.2% were in their third trimester of pregnancy. The mean pre-pregnancy weight was 66.3 ± 14.3 kg. Undertaking physical activity before pregnancy was declared by 58.2% of the respondents. An assessment of the physical activity levels of the pregnant women showed that the highest energy expenditure was recorded for light and moderate intensity efforts (<sup>−</sup> x =71.7 MET hour/week; <sup>−</sup> x =72.9 MET hour/week), accounting for 37.5% and 38.2% of total physical activity, respectively.

#### *3.2. Diet Characteristics*

In the entire study group, the mean value of the pHDI was 26.3 ± 13.0 points. The women with health education had a higher value of the pHDI than the women without health education (HEG = 28.3 ± 12.7 points vs. nHEG = 20.9 ± 12.3 points; Z = 2.99; *p* = 0.002). There were no women with a high-quality diet in the whole study group; however, a moderate-quality diet was noted in 30.3% of the participants.

In general, fruit and vegetables were consumed with the greatest frequency (on average once a day), whereas fish and legumes were consumed least frequently (on average 1–3 times a month). The remaining products were consumed with an average frequency of once to several times a week (see Supplement Table S1). A comparative analysis showed significant differences in the frequency of consumption of wholemeal bread (Z = 2.72; *p* = 0.007), grains and coarse-ground groats (Z = 2.43; *p* = 0.02), legumes (Z = 1.97; *p* = 0.049), and fruits (Z = 2.21; *p* = 0.03). Each time, the nHEG group was characterized by a lower frequency of consumption of the above-mentioned products.

The correlations of health education with the pHDI and ten food products with beneficial health outcomes are shown in Table 2. Positive and significant correlations were found for all variables except milk consumption, fermented milk beverages, curd, and white meat.


**Table 1.** Characteristics of surveyed group.

HEG—group with health education; nHEG—group without health education; *p* ≤ 0.05—a statistically significant value.

**Table 2.** Correlation coefficients of the health education with the pHDI and 10 food products with beneficial health outcomes.


*p* ≤ 0.05—a statistically significant value.

#### *3.3. Food Determinants of Diet Quality*

Before testing the hypothesis concerning the food correlates of diet quality, the correlations between diet quality as the dependent variable and ten food products with a potentially beneficial effects on health were analyzed (Table 3). In each group, moderate but significant, strong, and very strong correlations between diet quality and the mentioned independent variables were found. In the group with health education, all variables were positively correlated with the diet quality. They were as follows: wholemeal bread (r = 0.58; *p* ≤ 0.001); grains and coarse-ground groats (r = 0.74; *p* ≤ 0.001); milk (r = 0.60; *p* ≤ 0.001); fermented milk beverages (r = 0.62; *p* ≤ 0.001); curd (r = 0.62; *p* ≤ 0.001); white meat (r = 0.35; *p* = 0.001); fish (r = 0.45; *p* ≤ 0.001); legumes (r = 0.47; *p* ≤ 0.001); fruits (r = 0.75; *p* ≤ 0.001); and vegetables (r = 0.81; *p* ≤ 0.001). In a group with no health education, no correlation was found for milk and white meat, whereas positive correlations for other food products were as follows: wholemeal bread (r = 0.50 *p* = 0.002); grains and coarse-ground groats (r = 0.42; *p* = 0.009); fermented milk beverages (r = 0.60; *p* ≤ 0.001); curd (r = 0.37; *p* = 0.02); fish (r = 0.33; *p* = 0.041); legumes (r = 0.53; *p* ≤ 0.001); fruits (r = 0.61; *p* ≤ 0.001); and vegetables (r = 0.73; *p* ≤ 0.001).

**Table 3.** Food correlates of diet quality in surveyed groups.


HEG—group with health education; nHEG—group without health education. \* a statistically significant correlation coefficient.

These significant variables were then included in the multiple regression model in order to assess which of them contributed most to explaining the variability in the diet quality in separate groups. According to the results obtained (Table 4), diet quality in the group with health education was determined by eight variables, i.e., vegetables, fermented milk beverages, milk, wholemeal bread, fruits, grains, coarse-ground groats, curd, and white meat. The model was significant and explained 99.6% of the variance in the diet quality F(8.75) = 2297.2; *<sup>p</sup>* ≤ 0.001). The consumption of vegetables (R<sup>2</sup> = 0.607; *<sup>p</sup>* ≤ 0.001) and fermented dairy drinks (ΔR<sup>2</sup> = 0.223; *<sup>p</sup>* ≤ 0.001) made the greatest contribution to the prediction of the dependent variable.

**Table 4.** Regression analysis of food determinants of diet quality in distinguished groups.


*p* ≤ 0.05—a statistically significant value.

In the group with no health education, six variables were included in the final diet quality model, i.e., vegetables, fermented milk beverages, grains and coarse-ground groats, fruits, legumes, and wholemeal bread. The model was significant and explained 93.9% of the variance in the dependent variable (F(6.31) = 79.5, *p* ≤ 0.001). As in the previous model, vegetables and fermented dairy drinks had the largest contribution to the prediction of the dependent variable (respectively: R<sup>2</sup> = 0.594; *<sup>p</sup>* ≤ 0.001; <sup>Δ</sup>R2 = 0.179; *<sup>p</sup>* ≤ 0.001). The smallest, however still significant, contribution to explaining the variability of diet quality was made by wholemeal bread (ΔR<sup>2</sup> = 0.038; *<sup>p</sup>* ≤ 0.001).

#### *3.4. Sociodemographic and Maternal Lifestyle-Related Determinants of Diet Quality*

The first step in assessing the sociodemographic and maternal determinants of diet quality was to examine the correlations between diet quality as the dependent variable and all the variables listed in Table 1. Significant correlations were found for variables such as age (r = 0.20; *p* = 0.026), health education (r = 0.27; *p* = 0.002), educational level (r = 0.20; *p* = 0.025), trimester of pregnancy (r = 0.31; *p* ≤ 0.001), pre-pregnancy PA (r = 0.26, *p* = 0.003), moderate PA (r = 0.19; *p* = 0.039), and vigorous PA (r = 0.19; *p* = 0.035). A regression model was then run with diet quality as the dependent variable. According to the results obtained (Table 5), diet quality was predicted by four variables, i.e., health education, trimester of pregnancy, moderate PA, and pre-pregnancy PA. The greatest contribution to the prediction of the dependent variable was made by health education (ΔR2 = 0.069; *p* = 0.003), followed by the trimester of pregnancy (ΔR2 = 0.063; *p* = 0.028). Then, moderate PA was added (ΔR2 = 0.044; *p* = 0.013), and, in the last step, pre-pregnancy PA was included (ΔR2 = 0.032; *p* = 0.032). The final model was significant and explained 20.8% of the variance of the diet quality (F(4.17) = 7.69; *p* ≤ 0.001).

**Table 5.** Regression analysis of socio-demographic and maternal lifestyle-related determinants of diet quality.


*p* ≤ 0.05—a statistically significant value.

#### *3.5. The Odds Ratio of Higher-Quality Diet*

A logistic regression analysis was performed to assess how the sociodemographic and maternal lifestyle-related predictors from Table 5 affected the odds of achieving a higherquality diet (Figure 1). Unfortunately, due to lack of standards and cut-off points, a similar analysis could not be performed for moderate PA during pregnancy. The results showed that the presence of health education in the educational history of the surveyed participants more than tripled the odds of a higher-quality diet (OR = 3.14; 95% CI: 1.09–7.03; *p* = 0.032). The women in their second trimester were 54% more likely to have a higher-quality diet than the women in their third trimester of pregnancy (OR = 1.54; 95% CI: 1.23–2.17; *p* = 0.046). Undertaking PA before pregnancy increased the odds of a higher-quality diet by 2.5 times (OR = 2.51; 95% CI: 1.08–5.88; *p* = 0.032).

**Figure 1.** The odds ratio of a higher-quality diet.

#### **4. Discussion**

The literature indicates that pregnancy is an important time in a woman's life, contributing to changes in both her dietary habits and other health-related behaviors that are undertaken out of concern for her life and health and that of her baby [1,3,32]. Previous studies have shown a wide variation in the determinants of diet quality among pregnant women. In addition to social and cultural factors [24,33–36], nutritional knowledge and health education have also been indicated as factors influencing diet quality in pregnancy [35]. Therefore, the aim of this study was to assess the dietary quality of pregnant women and its determinants, with attention to health education as possible one.

Among the most commonly reported factors influencing the quality of pregnant women's diets are age and socioeconomic variables, including the education level, which is considered to be an awareness variable that significantly influences dietary decisions [8,9,12,37–40]. In turn, the presented study highlighted the particularly important role of health education, trimester of pregnancy, moderate PA, and pre-pregnancy PA in shaping dietary habits and diet quality. According to the literature, younger mothers have poorer diet quality because they have lower levels of education, lower socioeconomic status, and less life experience, unlike older women [1,37,41–47]. However, our own results show that women in the no health education group, despite being older and having achieved a university degree, had poorer diet quality than younger women without a higher education but with health education in the core curriculum. This suggests that diet quality does not depend as much on age and educational attainment but, to a large extent, on the health education provided as part of compulsory schooling for children up to the age of 18. Our further analysis showed that participation in compulsory health education more than tripled the odds of having a better diet. Sedentary lifestyles and unhealthy eating habits are known to be common among pregnant women [17], but our results show that women with healthier pre-pregnancy behaviors were also those with better diets during pregnancy. In contrast to McGowan and McAuliffe [48], our study showed a significant positive influence of pre-pregnancy and pregnancy PA on diet quality, with the pre-pregnancy PA increasing the odds of a higher-quality diet during pregnancy by a factor of 2.5. This confirms that physical activity is an important target for nutrition and health interventions. In the presented study, women in the second trimester of pregnancy had a healthier dietary profile than women in the third trimester. This is on the contrary to Fernández-Gómez et al. [49], but consistent with McGowan and McAuliffe [42], who reported the odds in predicting the likelihood of following a healthy dietary pattern in each trimester. In their study, higher levels of maternal education together with normal maternal BMI as well as the nationality were important predictors of following a healthy diet in the second trimester. This indicates that women with higher levels of education also are more likely to make positive changes in their diet. Although awareness of the positive effects of a healthy diet and physical activity on pregnancy outcomes has been reported to be a strong motivator for changing dietary behaviors [50,51], it is not always sufficient to maintain changes until the end of pregnancy. As shown by McGowan and McAuliffe [48], 69 out of 95 women continued

the healthy dietary pattern into the third trimester. Therefore, there is a strong need for research to investigate the reasons why healthy dietary behaviors are not maintained during pregnancy.

A positive contribution of health education to dietary behaviors was also shown in the case of the Pro-Healthy Diet Index, which provides information on diet quality. The diets of women who received counselling and education on healthy eating and lifestyles were of better quality than those of women who did not receive adequate substantive support. In addition, health education was positively associated with the intake of wholemeal bread, grains and coarse-ground groats, fish, legumes, fruit, and vegetables but not with intakes of milk, fermented milk beverages, curd, or white meat. Our results differ from those obtained by Goodarzi-Khoigani et al. [52], who showed that health education was positively associated with the intake of vegetables and fish but not bread, legumes, dairy products, or fruit in the Japanese population.

Unfortunately, despite the positive contribution of health education to dietary behaviors and noticeable differences in the level of DQ and the frequency of consumption of selected groups of products, the diets of women with nutrition education were not in accordance with nutritional recommendations [53,54]. In the surveyed groups, the consumption of products with beneficial health effects was insufficient, which corresponds with the findings of other authors [14,16,18]. In general, the respondents consumed fruit and vegetables most frequently (once a day on average), which is significant, as they are the basis of a healthy diet in many nutritional recommendations, mainly because of the vitamins, minerals, and antioxidants they contain [55,56]. The remaining food products were consumed with an unsatisfactory frequency, and the identified dietary errors were particularly related to insufficient consumption of whole grain products (wholemeal bread, groats, oatmeal), fish, and legumes. Given the fact that whole grain products are a good source of fiber and have a positive impact on the prebiotic index [57,58], a well-balanced diet should be rich in these products. Unfortunately, only 16% of the women with health education met the recommendations of several servings of whole grain per day [57], compared to 9% of the women without health education. The recommended intake of 2–3 portions of fish per week was reported by 10% of the women with health education and only 3% of those without health education. The consumption of legumes was also low. However, this can be regarded as a positive outcome, especially if they had been eaten as raw sprouts (e.g., beansprouts). Similar to legumes, sprouts are a good source of protein [59] and also of health-maintaining nutrients such as glucosinolates, phenolics, and isoflavones [60]. However, it should be noted that sprouts also belong to a group with a high risk of Listeria monocytogenes infection [61], and, unlike maternal listeriosis infection, fetal or neonatal infection carries a high risk of fatal complications [62]. Therefore, pregnant women should limit their consumption of sprouts.

The results obtained indicate the positive impact of educational programs conducted in Polish schools aimed at implementing the principles of proper nutrition described by the healthy eating pyramid [54]. In the group of the women with health education, eight out of ten groups of products with potentially health-promoting properties determined the quality of the diet (i.e., vegetables, fermented milk beverages, milk, wholemeal bread, fruit, grains and coarse-ground groats, curd, and white meat). In turn, in the group without health education, the variety of food determinants of diet quality was smaller. Only six out of ten recommended products explained the variance in diet quality, i.e., vegetables, fermented milk beverages, grains and coarse-ground groats, fruit, legumes, and wholemeal bread. However, it should also be noted that the consumption of vegetables and fermented dairy drinks was one of the determining factors of diet quality in each studied group.

Previous studies have shown that women's compliance with recommendations increased when they were given detailed explanations on the importance of the recommended food products [23,24]. On the other hand, the lack of adequate knowledge about nutritional recommendations of those responsible for developing nutritional awareness have been identified as one of the barriers to changing dietary behaviors [63–65]. Therefore, it is

possible that the results obtained in the present study are caused by inadequate health education in Polish schools, e.g., the content provided may be insufficient or not adapted to the age of the recipients, yet the individual non-adherence to the recommendations cannot be excluded.

Important clues for nutrition education also come from studies that indicate the preferred form of knowledge transfer. As was shown by Wise and Arcamone [66], among adolescents, the best way to learn about nutrition was to listen to teachers and health professionals. Unfortunately, it is not appreciated in Poland. Here, health education is provided only by schoolteachers, nor is it not a separate school subject, but it is implemented in a number of different subjects in the form of selected individual class topics. Crucially, partners and relatives are an important source of nutritional support for mothers and mothers-to-be [67]; in order to improve the quality of pregnant women's diets, it is also necessary to educate and increase knowledge about the positive or reinforcing effects of healthy nutrition also in the woman's immediate environment.

#### *Limitations*

This study has some strengths and limitations. The study included a group of women attending childbirth school, and access to health education was taken into account. The KomPAN® and PPAQ (Polish version) questionnaires used in the study have good relevance, and acceptable test–retest reliability of the test–retest, therefore, represent a reliable set of data. The PPAQ questionnaire has been adapted to the cultural conditions of many countries, including Poland, allowing international comparisons to be made regarding the level of AP of pregnant women. Furthermore, the study was conducted in the form of a direct questionnaire interview, which allowed us to better understand the questions and obtain more complete and reliable information about the dietary habits of the women surveyed. However, we are aware of some limitations. It was a cross-sectional study, in which diet quality was analyzed based on questions about general food consumption rather than questions about specific dietary components. Future research should include this type of data to gain full insight into the complex model of determinants of dietary quality. The Pro-Healthy Diet Index (pHDI) used to assess diet quality is based on the consumption of health-promoting products recommended in the Mediterranean diet and included in the healthy eating pyramid.

#### **5. Conclusions**

The present study highlighted the particularly important role of health education, trimester of pregnancy, moderate PA, and pre-pregnancy PA in shaping dietary habits and diet quality. We recommend that the proposed interventions for the nutritional education of women of reproductive age include not only nutritional aspects but also physical activity adapted to the gestational age and capabilities of the pregnant women. Appropriate adaptation of the interventions to the individual needs of the woman, her preferences, and, above all, her knowledge and health habits can effectively influence the modification of her dietary behavior during pregnancy. The present study also has practical implications. The results obtained can be used by institutions providing health education to preconceptional and pregnant women to develop an appropriate strategy aimed at raising awareness of the importance of proper nutrition during pregnancy and possibly changing inappropriate eating habits.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/nu15112627/s1, Table S1. Dietary characteristics of surveyed women.

**Author Contributions:** Conceptualization: A.D.; methodology: A.D., K.A.-C., J.R. and U.C.; formal analysis: A.D. and J.R.; investigation: K.A.-C.; writing—original draft preparation: A.D. and K.A.-C.; writing—review and editing: J.R., U.C., A.D. and K.A.-C.; project administration: A.D. and K.A.-C. All authors have read and agreed to the published version of the manuscript.

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

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki. The study protocol was reviewed and approved by bioethics committee at the Medical University of Karol Marcinkowski in Pozna ´n (reference no. 878/19, 12 September 2019).

**Informed Consent Statement:** All participants provided written informed consent prior to their participation in this study.

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

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

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Article* **Exposure to Phosphates and Nitrites through Meat Products: Estimation of the Potential Risk to Pregnant Women**

**Danijela Vrani´c 1, Jelena Mileševi´c 2, Dejana Trbovi´c 1, Mirjana Gurinovi´c 2, Vladimir Kori´canac 1, Milica Zekovi´c 2, Zoran Petrovi´c 1, Slavica Rankovi´c <sup>2</sup> and Dragan Mili´cevi´c 1,\***


**Abstract:** Diet during pregnancy is one of the most important nutritional challenges associated with some risks for the mother and the fetus. For the first time, the study aims to estimate long-term (2018–2022) exposure to nitrate and phosphates in Serbian pregnant women, based on individual consumption data and accurate values measured in frequently consumed meat products. For this purpose, seven types of meat products, consisting of 3047 and 1943 samples, were collected from retail markets across Serbia, to analyze nitrites and phosphorus content, respectively. These data were combined with meat product consumption data from the Serbian National Food Consumption Survey to assess dietary intake of nitrites and phosphate. The results were compared with the acceptable daily intake (ADI) proposed by the European Food Safety Authority. The average dietary exposure (EDI) to phosphorus ranged from 0.733 mg/kg bw/day (liver sausage and pate) to 2.441 mg/kg bw/day (finely minced cooked sausages). Considering nitrite intake, the major sources were bacon (0.030 mg/kg bw/day) and coarsely minced cooked sausages (0.0189 mg/kg bw/day). In our study, average nitrite and phosphorus exposure in the Serbian pregnant women population are far below the EFSA recommendations (ADI 0.07 mg/kg bw/day and 40 mg/kg bw/day, respectively).

**Keywords:** food additives; nitrites; phosphates; pregnant women; meat products; exposure assessment

#### **1. Introduction**

Even though meat and meat products are one of the most important contributors of the modern diet, it is well known that the nutritional profile of processed meat has been perceived as unhealthy due to the high levels of saturated fatty acids, cholesterol [1], or components that could be considered with negative health impacts (sulfites, nitrites, and sodium). Moreover, elevated consumption of processed meat and red meat has been associated with cardiovascular diseases, colorectal, stomach, prostate, and pancreatic cancer [2]. According to an epidemiological study, processed meat has been classified as carcinogenic to humans (Group 1) while red meat is probably carcinogenic to humans (Group 2A) [3]. Because of this, there is growing interested in the processed meat industry to reduce food additives that could be considered unhealthy [4].

Nitrates and nitrites (E249–E252) are food additives of concern for humans' health because they may interact with secondary amines in the stomach, producing nitritesNnitroso compounds (*N-NAs*), which could play a role in the carcinogenicity of processed meat [5,6]. Among meat products, cured meats often contain detectable levels of *N-NAs* mainly due to the use of nitrites as a preserving agent, additionally influenced by several

**Citation:** Vrani´c, D.; Mileševi´c, J.; Trbovi´c, D.; Gurinovi´c, M.; Kori´canac, V.; Zekovi´c, M.; Petrovi´c, Z.; Rankovi´c, S.; Mili´cevi´c, D. Exposure to Phosphates and Nitrites through Meat Products: Estimation of the Potential Risk to Pregnant Women. *Nutrients* **2023**, *15*, 2777. https:// doi.org/10.3390/nu15122777

Academic Editor: Shannon L. Kelleher

Received: 19 April 2023 Revised: 1 June 2023 Accepted: 14 June 2023 Published: 16 June 2023

**Copyright:** © 2023 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/).

processing factors (i.e., temperature, pH, and storage conditions) strongly linked by the presence of free amines, particularly biogenic amines [7].

Besides meat products, the occurrence of *N-NAs* has also been reported in other foods, such as processed vegetables, cereals, milk and dairy products, and alcoholic and nonalcoholic beverages, among others. Regarding other sources of *N-NAs* exposure, tobacco products (cigarettes, cigars) followed by products used in personal hygiene (cosmetics, hair products, lotions, shampoos, soaps, etc.) represent the important non-dietary exposure sources to *N-NAs* [6].

The high incidence of gastrointestinal cancer reported in the United Kingdom, Canada, Colombia, Chile, Japan, Denmark, and Italy has been correlated with elevated nitrite intake from food [8]. Moreover, nitrites and nitrates may cause methemoglobinemia, a blood disorder in which hemoglobin can carry oxygen but is unable to release it effectively to body tissues. It is also known that nitrites cross the placenta in pregnancy, causing methemoglobin formation in fetuses [9]. Some earlier studies have demonstrated the teratogenic effect of nitrites, emphasizing their toxicity and severe developmental defects on embryos or even spontaneous abortions [10]. An ADI is established for the additives that represent a concern for the consumers' health. The European Commission, according to the Scientific Opinion of the Panel on Food Additives and Nutrient Sources added to Food (ANS) using a Benchmark Dose (BMD) approach, recommended an ADI of 0.07 mg nitrite ion/kg bw per day [7].

On the other hand, phosphates are used as food additives (E338–E341, E343, E450–E452) to improve food quality. Excessive intake of phosphates via consuming processed meat products can lead to various adverse effects on human health inducing anionic imbalance. An association between high serum phosphate levels and cardiovascular morbidity and mortality in patients with chronic kidney disease and bone health complications has long been known [11]. Therefore, high phosphorus intake from additives should be considered as a potential public health concern. For this purpose, the Scientific Committee for Food [11] derived a group acceptable daily intake (ADI) for phosphates expressed as phosphorus of 40 mg/kg bw/day and concluded that this ADI does not have adverse effects on human health.

Maximum permitted levels of food additives are set at the international level by the WHO-FAO JECFA and the European Food Safety Authority (EFSA) with the aim to ensure that additives are used properly to minimize potential risks to human health. Furthermore, under the European Directive [12], all Member States are obligated to monitor intakes to ensure that consumers do not have an excessive intake of a given food additive, which could lead to a health hazard. The current Serbian legislation has restricted the concentration of residual NaNO2 in processed meat to 100 and 150 mg/kg depending on the type of product [13,14], whereas regulations in Europe state that the maximum residual level (expressed as NaNO2) amount of nitrites that may be added to the processed meat during manufacturing should be from 50 to 180 mg/kg, particularly for dry-cured meat products such as bacon (175 mg/kg), for dry non-heat-treated meat products (50 mg/kg), and for other dry-cured meat products such as dry-cured ham (100 mg/kg), with several exemptions [12,15]. In terms of phosphorus used as a food additive, the Serbian standard maximum limit for total phosphorus expressed as P2O5 in meat products is less than 8 g/kg [14] or ≤5 g/kg of added phosphorus [13].

Diet during pregnancy is one of the most important nutritional challenges that may be associated with some risks for the health of mothers and the development of the fetus. In this context, a healthy and balanced diet is of the utmost importance during pregnancy and is an ongoing task for health care. Although nitrates and nitrites alone are considered to have no or limited carcinogenic potential [16], there are major human health concerns raised regarding nitrite intake, due to their potential conversion to form *N-NAs*. Based on the literature data, on associations between dietary intake of meat products, nitrite content, and cancer, the genotoxic properties of the *N-NAs* have been extensively investigated [6,17]. The high and frequent consumption of meat products, containing harmful substances such as nitrites, increases the risk of colorectal cancer and thyroid tumor promotion and adversely affects reproductive outcomes (e.g., fetal loss, reduced number of litters and live births, and neonatal mortality). Moreover, some studies reported a correlation between excessive dietary nitrite intake and a higher risk of development of neural tube defects [18,19] or even pediatric brain tumors in offspring [20].

Considering above mentioned rational and following our previous nitrites and phosphate studies [21–23], this study objective was to, for the first time, estimate dietary intake of nitrate and phosphates in Serbian pregnant women, based on individual consumption data and accurate values measured in most consumed groups of meat products. In addition, as a predictive model, i.e., "worse-case" scenario, values at the MPL was used in order to determine the level of reaching or exceeding ADI values for these two additives in meat products as a measure of identifying potential risk.

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

#### *2.1. Meat Products and Sample Preparation*

In the present study, 3047 meat product samples obtained from different regions of the Serbian retail market for the purposes of official controls by veterinary inspectors or for self-monitoring purposes of the meat producers during 2018–2022 were analyzed for nitrite content. Samples were divided into seven groups, including 381 bacon, 244 dry meat, 406 coarsely minced cooked sausages, 822 dry fermented sausages, 747 finely minced cooked sausages, 87 liver sausage and pate, and 423 smoked meat products, produced by the Serbian meat industry or imported.

In the same period of investigation, a total of 1943 meat product samples were categorized into five groups including bacon (298), coarsely minced cooked sausages (405), finely minced cooked sausages (718), liver sausage and pate (86), and smoked meat products (436) were analyzed for phosphorous content.

All samples of meat products were kept at refrigeration temperature and analyzed within 48 h. If the analyses were not conducted within the same day, the samples were stored in a refrigerator at 4 ◦C until required for testing.

The analyzed samples were thawed and blended in a commercial kitchen blender unit (Homogenizator Blixer 2, Robot Coupe, Vincennes, France (2.9 L) 700 w, 3000 rpm). For each sample, two composite samples were prepared. All samples were then analyzed in duplicate.

#### *2.2. Determination of Sodium Nitrite Content*

The content of sodium nitrites-NaNO2, which is usually added to meat products—was examined in meat products according to the standard ISO procedure [24]. A representative sample amount (~10 g) was measured in a 300 mL flask using an analytical balance (Mettler, AE 200, Columbus, OH, USA), followed by the addition of a solution of hydrous sodium borate, Na2B4O7·10H2O (50 g/L) and 100 mL deionized water at 70.0 ± 0.2 ◦C. Residual nitrite extraction was achieved by keeping the samples in a hot water bath, at the temperature of boiling, for 15 min, and every 5 min, flasks were shaken vigorously. After cooling, 2 mL of each Carrez solution (Carezz reagent I and Carezz reagent II) was added and mixed thoroughly. Samples were then diluted to 200 mL with deionized water. Samples were filtered through quantitative cellulose filters (pore size < 5 μm). Color generation was achieved by transferring an aliquot of the filtrate (25 mL) to a 100 mL volumetric flask and adding 10 mL of the sulfanilamide solution and then 6 mL conc. HCl. Flasks were stored in the dark for 5 min. Subsequently, 2 mL solution of *N*-naftil-1-ethylenediamine-chloride (0.25 g/250 mL) was added to each flask and moved to the dark for 3 min. Thereafter, samples were diluted to 100 mL. Absorbance was measured at 538 nm using a spectrophotometer (UV/VIS Spectrophotometer, Jenway 6405, East Lyme, CT, USA). A procedural blank was run with every batch of samples.

Calibration curves were generated using concentration levels ranging from 2.5 to 10 NaNO3 μg mL−1, Y = 0.0669X + 0.024: R2 = 0.999. A recovery study of the analytical procedure was carried out by spiking several already analyzed samples with standard solutions, and recovery rates were found to be between 87% and 94%. The nitrite content is expressed as NaNO2 (mg·kg<sup>−</sup>1), following c × 2000/m × V, where c is the concentration of NaNO2 (μg/mL) from the calibration curve, m is the mass of sample (g) for analysis, and V is a volume of an aliquot of the filtrate used for spectrometric determination.

The limit of detection (LOD) was considered to take the same value as the limit of quantification (LOQ) (0.03 mg/kg).

#### *2.3. Determination of Phosphorus Content*

The total phosphorus content, expressed as P2O5 (g/kg), in examined meat products was determined according to the standard ISO procedure [25]. The total phosphorus content, expressed as P2O5 (g/kg), in examined meat products was determined according to the standard ISO procedure [25]. In brief, a ~5 g portion of samples (measured using an analytical balance (Mettler, AE 200, USA)) was ashed at the maximum temperature of 500 ◦C in a muffle furnace (LE 14/11/R7, Nabertherm, Lilienthal, Germany). On completion of the digestion, the white ash was dissolved by heating with dilute nitric acid (1 + 1, *v*/*v*) and quantitatively transferred to a 100 mL flask. Then, made up by the addition of deionized water, and after mixing, the solution was then filtered, and the first 5 to 10 mL were discarded.

Aliquots (20 mL) of the treated solution were pipetted into 100 mL volumetric flasks and mixed thoroughly with 30 mL ammonium heptamolybdate solution 50 g/L. The resulting solution was then diluted to the volume with deionized water. After 15 min at room temperature, the absorbance was read against a reagent blank at 430 ± 2 nm using a UV-visible spectrophotometer (UV/VIS Spectrophotometer, Jenway 6405).

The standard curve was determined under the same conditions as those for the samples using potassium dihydrogen phosphate as a standard (10–60 P2O5 μg/mL; Y = 0.0187X − 0.0096: R2 = 0.9999). A recovery study of the analytical procedure was carried out by spiking several already analyzed samples with standard solutions, and recovery rates were found to be between 89% and 95%. The total phosphorus content is expressed as P2O5 (g/kg), following c/20 m, where c is the concentration of P2O5 (μg/mL) from the calibration curve and m is the mass of the sample (g) for analysis.

The LOD was estimated at 0.0024 g/kg, while the LOQ for phosphorus as P2O5 was 0.081 g/kg.

#### *2.4. Meat Products Consumption Data*

The National Food Consumption Survey on adults including pregnant women, in compliance with the EFSA EU Menu methodology [26], was conducted between 2017 and 2022 and included a total of 145 pregnant women. EFSA EU Menu methodology considers the use of set of questionnaires: a general questionnaire on sociodemographic and anthropometric characteristics of the participants, an age-appropriate Food Propensity Questionnaire (FPQ), that is used to determine the frequency of food groups' consumption in a year, and a twice-repeated 24 h dietary recall. All the data are collected in the required format following the EU Menu framework, to provide harmonious and standardized data collection in all countries in Europe [27]. The consumed portion sizes were estimated based on natural units, household measures, packaging information, and a validated national Food Atlas for Portion Size Estimation [28]. The study was conducted in four geographical regions of Serbia (Belgrade, Vojvodina, Southeast Serbia, and West Serbia).

In this study, the following data were used: anthropometric characteristics of the participants, i.e., age, body weight, and height measurements, and intake data of meat products. This study assessed the consumption of meat products (per meat product type and on average) in a pregnant population. Consumed meat products were categorized into seven categories which were defined according to the actual Serbian Regulation on the quality of meat products [14]. The study group age is divided into two groups, to better describe characteristics of the population and age distribution, but is later not correlated in the exposure assessments, as both age groups belong to the same, adult population groups in the reference values—ADI and EDI—by the EFSA.

#### *2.5. Exposure Assessment and Risk Characterization*

According to the European Commission [29], there are three types of approaches to estimate the dietary exposure from food additives that pose a concern to human health:


The estimated daily intake (EDI) of nitrite and phosphate additives from processed meat by pregnant women included in this study was calculated using the Tier 3 approach, by combining data on individual food consumption patterns in pregnant women (g/day) with data on the levels of this type of additive in the investigated samples and division by the population's average body weight (Table 1). Additionally, the mean value regarding the body weight of the investigated population of pregnant women obtained in our study is in accordance with EFSA recommendations, where a body weight of 70 kg should be used as the default for the European adult population [30].


**Table 1.** Baseline characteristics of study participants.

*N*—number of participants; Means with different superscripts in the same column are significantly different (*p* < 0.05).

In addition to the abovementioned method, the exposure assessment included certain assumptions of the worst-case scenario, so two levels of consumption were considered mean and high consumer (P95 percentile)—assuming the maximum use level of these compounds defined by Serbian regulation [13,14] in meat processing combined with individual consumption data (Tier 2). For risk characterization, obtained results were then compared with the ADI values established by the European Union [7,11]. Relative contributions of processed meat products to the dietary intake of nitrites and phosphorus for pregnant women was expressed as a percentage of the ADI established at 0.07 mg/kg body weight/day and 40 mg/kg body weight/day, for nitrites and phosphorus, respectively. Taking into consideration adaptive changes in phosphorus metabolism that occur during pregnancy and lactation, the ADI for adults (40 mg/kg bw/day) could be also applied to pregnant and lactating women [31].

As international guidelines recommend [32] when calculating dietary exposure, all non-detected results, i.e., below the LOD or the LOQ, are known as left-censored. According to this guidance, for dietary exposure assessments where less than 60% of the results were left-censored, middle-bound (all non-detected results to the LOD/2) exposure scenarios were considered [32].

#### *2.6. Statistical Analysis*

Data were analyzed using Minitab statistical software version 17 (Minitab Ink., Coventry, UK). The results are presented in the form of descriptive statistics (mean ± standard deviation—SD) and their distribution (percentiles, and ranges). The normality of the distribution of the of the data were checked using by Kolmogorov–Smirnov normality test. One-way analysis of variance (ANOVA) followed by Tukey's test was used to compare the dietary intake of phosphorous and nitrites among different meat products. The level of significance was set at *p* < 0.05.

#### **3. Results**

The mean and range of baseline characteristics of participants included in this study are presented in Table 1. The mean weight of the pregnant women included in this survey ranged from 70.34 ± 11.21 kg to 72.92 ± 13.93 kg (average 71.85 ± 12.90 kg). No statistically significant difference (*p* > 0,05) was observed between these two ages group of pregnant women in body weight. Regarding meat consumption, based on 145 participants interviewed, the highest average value of meat product consumption obtained in our research was for finely minced cooked sausages (84.83 ± 46.33 g/day), followed by dry fermented sausages (65.44 ± 64.22 g/day), while the lowest consumption was for bacon (30.84 ± 21.77 g/day). A statistically significant difference (*p* < 0.05) was found between the daily intake of bacon and finely minced cooked sausages and between the daily intake of finely minced cooked sausages and smoked meat products.

The mean, median, and 95th percentile levels of residual nitrites (NaNO2 and NO2 −) and phosphorus (P2O5 and P−) in examined processed meat products over the period of 2018–2022 are summarized in Tables 2 and 3. Nitrites were detected in 2443 (80%) of the total of 3047 analyzed meat product samples (Table 2). The results obtained in our study reveal that nitrite concentration varied with the type of meat product. The highest level of occurrence (99%) and mean residual level of nitrites, expressed as NaNO2, was detected in finely minced cooked sausages (38.72 ± 20.52 mg/kg), followed by coarsely minced cooked sausages (31.86 ± 23.30 mg/kg), while the lowest incidence (45%) and mean residual level of nitrite, as NaNO2, was detected in dry fermented sausages (1.44 ± 2.35 mg/kg). The average concentration of nitrites in the analyzed meat products was 19.56 ± 22.83 mg/kg. These results are far below the national Serbian or EU-regulated limit of 150 mg/kg [12,13]. In the current study, only one sample of smoked meat products exceeded the maximum permitted level of nitrites (data not shown). There were no statistically significant differences (*p* > 0.05) between the mean residual level of nitrite in bacon and liver sausage and pate, between dry meat and liver sausage and pate, and between dry meat and dry fermented sausages.

Phosphorus was detected in all analyzed meat product samples (1943) (Table 3). The average concentration of phosphorus, expressed as P2O5, in the analyzed meat products was 5.03 ± 1.37 g/kg within the range of 0.27 to 10.64 g/kg. The highest mean concentration and level of phosphorus, as P2O5, was found in smoked meat products (6.16 ± 1.38 g/kg and 10.64 g/kg, respectively), followed by coarsely minced cooked sausages (5.23 ± 1.14 g/kg and 9.92 g/kg, respectively), while the lowest mean concentration was found in liver sausage and pate (2.87 ± 0.95 g/kg). The results obtained in this study imply that the level of phosphorus in a total of 34 (1.7%) of the examined samples (except bacon and liver sausages and pate) exceeded the maximum permitted limit (MPL) of <8 g/kg as defined by the Serbian regulation [14] (Table 3).


**Table 2.** Ranges of residual nitrite levels expressed as NaNO2 and nitrite ion (NO2 −), consumption of processed meat products (g/day), dietary exposure to nitrite (mg/kg bw/day), and relative contributions of processed meat products to nitrite exposure.

*N*—total number of analyzed samples; *n*-number of samples that contained nitrites (%); Nitrite ion content (66.65% of NaNO2); Means values with different superscripts in the same column are statistically significantly different (*p* < 0.05); ADC–average daily consumption of meat products (g/day); EDI—estimated daily intake (mg/kg bw/day); ADI—acceptable daily intake of nitrite ion (NO2 −) (0.07 mg/kg bw/day) [7]; LOQ—limit of quantification = 0.03 mg/kg.

**Table 3.** Range of phosphorus levels (P2O5 and P), consumption of processed meat products (g/day), dietary exposure to phosphorus (mg/kg bw/day), and relative contributions of processed meat products to phosphorus exposure.


*N*—total number of analyzed samples; P content was 43.64% of P2O5; MPL—maximum permitted level (≤8 g/kg); Means with different superscripts in the same column are significantly different (*p* < 0.05); ADC average daily consumption of meat products (g/day); EDI—estimated daily intake (mg/kg bw/day); MTDI maximum tolerable daily intake of phosphorus (P) (40 mg/kg bw/day) [11].

Exposure (mean, median, and 95th percentile) and the contribution of meat products to the daily nitrite intake of the pregnant women considered in this study are presented in Table 2 and Figure 1. Overall, dietary nitrite exposure at the mean and 95th percentile did not exceed the ADI for nitrite (0.07 mg/kg bw/day) [7] (Table 2). The main contributors to dietary exposure to nitrites were finely minced cooked sausages (43.54%), followed

by coarsely minced cooked sausages (28%) and smoked meat products (12%), while the contribution of other groups was less than 10% (Figure 1).

The main meat product contributing to dietary exposure to phosphates in our study was found to be finely minced cooked sausages, accounting for 33%, followed by coarsely minced cooked sausages (27%), smoked meat products (19%), bacon (11%), and liver sausage and pate (10%) (Figure 2). Hence, mean and 95th percentile exposure to phosphates in our study is far below this ADI (40 mg/kg bw/day) [11] (Table 3).

**Figure 2.** Relative contributions (%) of processed meat products to phosphorus daily intake; CMCS coarsely minced cooked sausages; FMCS—finely minced cooked sausages; SMP—smoked meat products.

The results for the estimated daily intake and the relative percent contribution of each meat product included in this study to nitrite and phosphate exposure, combining individual consumption data with the MPL of the nitrite and phosphate additives (Tier 2 approach), are presented in Tables 4 and 5. The major contributors to excess nitrite ADI are finely minced cooked sausages, followed by dry fermented sausages and coarsely minced cooked sausages, at 168.62%, 130.11%, and 127.25%, respectively (Table 4).


**Table 4.** Scenario 2. Dietary exposure to nitrites by using actual national food consumption data and MPLs (Tier 2).

\* MPL—maximum permitted level of NaNO2 (150 mg/kg); EDI—estimated daily intake (mg/kg bw/day); ADI—acceptable daily intake of nitrite ion (NO2 −) (0.07 mg/kg bw/day) [7].



\* MPL—maximum permitted level for phosphorus (≤8 g/kg); MTDI—maximum tolerable daily intake of phosphorus (P) (40 mg/kg bw/day) [11].

Concerning exposure to phosphates, in the worst-case scenario (Tier 2 approach), the meat products identified as the main contributor to phosphate intake were finely minced cooked sausages (10.30%), followed by coarsely minced cooked sausages (7.77%) and liver sausage and pate (5.10%) (Table 5). This is because these meat products were consumed in large quantities.

#### **4. Discussion**

The present study provides new information about pregnant women's exposure to food-grade additives—nitrites and phosphorus via meat products. Pregnant women are considered more vulnerable to chemicals, particularly to ones which have carcinogenic and teratogenic properties because exposure occurs during the development of an embryo or fetus. Although nitrate and nitrites alone are considered to have no or limited carcinogenic potential [16], there are major human health concerns raised regarding nitrite intake, due to their potential conversion to form *N-NAs*. Based on the literature data, on associations between dietary intake of meat products, nitrite content, and cancer, the genotoxic properties of the *N-NAs* have been extensively investigated [6,17]. Although the primary sources of dietary nitrates and nitrites are vegetables, nitrates/nitrites from animal sources were attributable to an increase in cancer risk for the presence of amines, amides, and amides

and heme iron that favor the increased production of *N-NAs* carcinogens. Consequently, there is a trend to reduce or eliminate these compounds in meats [33].

The present study showed a wide range of nitrite levels within and between the meat products at 0.05–180.25 mg/kg and is comparable to those reported by Nurul Farhanah Haji Abd Hamid [34] at 0.5–140.6 mg/kg. However, the mean and P95 residual nitrite levels in analyzed samples were below the maximum permitted limit specified by Serbian or EU regulations (150 mg/kg) [12,13]. These findings are consistent with previously reported nitrites content in sausages by Bajˇci´c et al. [35] and Vrani´c et al. [22] at 0.65–36.60 mg/kg and 1.86–40.35 mg/kg (mean 12.96 mg/kg), respectively.

The dry fermented sausage samples were found to have the lowest amount of nitrites at 1.44 ± 2.35, followed by dry meat (4.86 ± 10.96 mg/kg), and these values were much lower compared to the other types of sausages. These findings are unexpected because the shorter shelf life was valid for coarsely or finely minced cooked sausage products, which were mostly less than 90 days, hence a lower amount of nitrate and nitrite additives were necessary to add. In this study, the highest health risks regarding nitrite intake by consuming meat products are in finely minced cooked sausages followed by coarsely minced cooked sausages (mean 0.0305 mg/kg bw/day and 0.0189 mg/kg bw/day, respectively). Consumption of finely minced cooked sausages at 152 g/day recorded in our study is of high concern, contributing to 0.054 mg/kg bw/day or 78.02% of the nitrite ADI (0.07 mg/kg bw/day), while the least risk was from dry fermented sausages with a level of 0.0009 mg/kg bw/day or 1.26% of the nitrite ADI.

Phosphorus is an essential nutrient, occurring in foods of animal origin as a natural component and an approved ingredient added during food processing. Thus, JECFA proposed to assign a "maximum tolerable daily intake" (MTDI) rather than an ADI. The phosphorus EDI ranged from 0.733 to 2.445 mg/kg bw/day, representing 1.83 and 6.10% of the ADI specified by the EFSA [11]. The major contributor to phosphorus intake for pregnant women was finely minced cooked sausage (33% of phosphorus intake) and coarsely minced cooked sausage (27% of phosphorus intake) consumption. In both scenarios, the exposure does not exceed the ADI of 40 mg/kg bw per day (Tables 3 and 5). However, ADI did not apply to populations with chronic kidney disease (CKD) or cardiovascular disease (CVD), considered a vulnerable population group. Thus, assessment of the EDI for those who consume phosphorus-rich food products regularly was important.

Although most authorized food additives are used at a lower level than the MPL, to ensure a high level of consumer protection, in addition we created a worst-case scenario for our risk assessment. The Tier 2 approach included certain assumptions of the worstcase scenario assuming the maximum use level of these food additives defined by the EU Regulation [12] in meat processing and the mean and highest percentile (95th percentile) of food-intake consumers. The Tier 2 intake estimates for nitrites and phosphorus are presented in Tables 4 and 5. The differences between the results of nitrites and phosphates exposure obtained with two different exposure scenarios (Tier 2 and Tier 3 approaches) were significant. As expected, in the Tier 2 approach, exposure was considerably above the ADI. The major contributors to exceeding the ADI of nitrites and phosphates in this approach were finely minced cooked sausages, dry fermented sausages, and coarsely minced cooked sausages. These results could be explained by the fact that they were consumed in high quantities.

The strength of this study is in the fact that this exposure assessment determines a realistic dietary intake of nitrites and phosphorus additives based on data from national food consumption surveys and the concentrations of nitrites and phosphorus in each meat product measured analytically as practiced by EFSA [29]. From a broader perspective, these findings could be accepted as the most accurate reflection of current industry practices in Serbia. Furthermore, they complement and confirm the findings on nitrites and phosphorus content, obtained from laboratory analysis of meat products previously reported by the authors [21–23,35]. Bearing in mind that this provides new information about the dietary intake of nitrate and phosphates in Serbian pregnant women, using the method proposed by EFSA [29], the present survey has a lot of strengths and emphasizes the importance of monitoring the added amounts of food additives and why dietary exposure assessment must be continued.

Considering the wide range of nitrites and phosphorus concentration levels within the meat products, we are aware of some limitations of our study. Thus, further study is necessary to consider the brand loyalty scenario. Very comprehensive studies revealed that consumers always tend to buy products of a given brand, which could have a higher concentration of additives than others [36]. Another limitation of this study lies in the circumstance that data on pregnancy state (trimesters) was not collected and correlation with exposure to observed additives could not be performed. This should be considered in the design of exposure studies in pregnancy in the future.

The results of our study are not easily compared with others. To the best of our knowledge, so far, no studies on the exposure to nitrites and phosphorus by meat products in pregnant women were identified. In our study, the consumption of several meat product types exceeds the recommended intake (≤50 g/day) [37]. Consumption of industrially processed meat products, high in calories, fats, and salt with additives such as nitrites, has a cumulative detrimental effect on the overall health of pregnant women, i.e., an unnecessary increase in weight, swelling, water retention in the body, and can increase the risk of high blood pressure in pregnancy and the occurrence of preeclampsia. Balancing the diet with a wider variety of (less processed) foods could help consumers of this kind reduce their intake of nitrosamines.

#### **5. Conclusions**

Our study revealed that the population of pregnant women in Serbia is not at risk of exceeding the ADI for nitrites or phosphates from the consumption of processed meat. Furthermore, food-grade additive nitrites and phosphates as currently used in industry practices in Serbia do not result in excessive exposure to the populations of pregnant women, even at the highest food consumption level (95%). Despite this, these results should be interpreted with caution, as other dietary sources of nitrites and phosphorus must be considered. Results in our study confirm that the Tier 2 approach can lead to overestimated exposure to additives, because the measured level of nitrites and phosphates was far below the MPL in meat products. Although we used the representative National Food Consumption Database, it is reasonable to assume that eating habits tend to change over the years. Therefore, it is mandatory to establish monitoring systems for the use and intake of food additives to ensure that the ADI is not exceeded. The application of nitrites should be decreased and controlled. For this reason, a further investigation into the presence of *N-NAs* in food of animal origin will be of great interest. Besides this, the study demonstrates the need for community work on raising awareness and constant education on healthy nutrition during pregnancy that includes information on the detrimental effects that additives can have on infants and offers advice on alternative healthier dietary options.

**Author Contributions:** Conceptualization, D.M.; methodology, D.M., J.M., M.G. and D.V.; software, D.M., J.M. and M.Z.; validation, D.V., D.T. and Z.P.; formal analysis, D.V., J.M. and V.K.; investigation, D.V., J.M., M.Z. and S.R.; data curation, D.V. and Z.P.; writing—original draft preparation, D.M. and J.M.; writing—review and editing, D.M., J.M. and M.G.; visualization, D.M.; supervision, M.G.; project administration, M.G.; funding acquisition, D.M.; J.M. and M.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by the Ministry of Science, Technological Development and Innovation, Republic of Serbia, Grant No. 451-03-47/2023-01/200050 and 451-03-47/2023-01/200015 from 03.02.2023.

**Institutional Review Board Statement:** The study was conducted in accordance with the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Institute for Medical Research Ethics Committee in Serbia on 8 December 2017 (EO 123/2017).

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

**Data Availability Statement:** Results attained in this study are included in the manuscript. Individual data are not available due to official legal, organizational and data security policies, and ethical restrictions.

**Acknowledgments:** Authors declare gratitude to Agnes Kadvan for the support in data processing and other IT issues.

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

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

MDPI St. Alban-Anlage 66 4052 Basel Switzerland www.mdpi.com

*Nutrients* Editorial Office E-mail: nutrients@mdpi.com www.mdpi.com/journal/nutrients

Disclaimer/Publisher's Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Academic Open Access Publishing

mdpi.com ISBN 978-3-0365-9573-3