Dietary factors during pregnancy and atopic outcomes in childhood: A systematic review from the European Academy of Allergy and Clinical Immunology

Abstract

Rationale:

Allergic diseases are an increasing public health concern, and early life environment is critical to immune development. Maternal diet during pregnancy has been linked to offspring allergy risk. In turn, maternal diet is a potentially modifiable factor, which could be targeted as an allergy prevention strategy. In this systematic review, we focused on non-allergen-specific modifying factors of the maternal diet in pregnancy on allergy outcomes in their offspring.

Methods:

We undertook a systematic review of studies investigating the association between maternal diet during pregnancy and allergic outcomes (asthma/wheeze, hay fever/allergic rhinitis/seasonal allergies, eczema/atopic dermatitis (AD), food allergies, and allergic sensitization) in offspring. Studies evaluating the effect of food allergen intake were excluded. We searched three bibliographic databases (MEDLINE, EMBASE, and Web of Science) through February 26, 2019. Evidence was critically appraised using modified versions of the Cochrane Collaboration Risk of Bias tool for intervention trials and the National Institute for Clinical Excellence methodological checklist for cohort and case-control studies and meta-analysis performed from RCTs.

Results:

We identified 95 papers: 17 RCTs and 78 observational (case-control, cross-sectional, and cohort) studies. Observational studies varied in design and dietary intakes and often had contradictory findings. Based on our meta-analysis, RCTs showed that vitamin D supplementation (OR: 0.72; 95% CI: 0.56–0.92) is associated with a reduced risk of wheeze/asthma. A positive trend for omega-3 fatty acids was observed for asthma/wheeze, but this did not reach statistical significance (OR: 0.70; 95% CI: 0.45–1.08). Omega-3 supplementation was also associated with a non-significant decreased risk of allergic rhinitis (OR: 0.76; 95% CI: 0.56–1.04). Neither vitamin D nor omega-3 fatty acids were associated with an altered risk of AD or food allergy.

Conclusions:

Prenatal supplementation with vitamin D may have beneficial effects for prevention of asthma. Additional nutritional factors seem to be required for modulating the risk of skin and gastrointestinal outcomes. We found no consistent evidence regarding other dietary factors, perhaps due to differences in study design and host features that were not considered. While confirmatory studies are required, there is also a need for performing RCTs beyond single nutrients/foods.

1 |. INTRODUCTION

Allergic diseases are an increasing public health concern.¹ The most common allergic conditions are asthma, allergic rhinitis/hay fever/seasonal allergies, eczema/atopic dermatitis (AD) (from now on referred to as eczema), and food allergies. Worldwide, an estimated 300 million people suffer from asthma, and an additional 400 million have allergic rhinitis.² In addition, 570 million children and 116 million adults suffer from eczema, and 200–250 million people suffer from food allergies.³ These allergic diseases may often co-exist in one individual.⁴ Allergies can affect quality of life and reduce academic performance and income,⁵ and in rare cases can be fatal.^6,7

Maternal diet in pregnancy is an important, potentially modifiable factor in allergy prevention. Several studies have investigated the role of prenatal diet on offspring allergy, but the findings are conflicting, and there is a paucity of randomized controlled trials (RCTs). An evidence synthesis is paramount in order to adequately appreciate the underlying evidence.

A number of review papers have been published focusing on maternal diet during pregnancy, lactation, or early life^8–11 and allergy prevention.^8–12 Our review follows and extends a recent systematic review and meta-analysis by Garcia-Larsen et al,¹³ who summarized the evidence up to December 2017. Garcia-Larsen et al ¹³ considered dietary intake not only during pregnancy, but also in lactation and infancy; their main conclusions ¹³ were that maternal omega-3 fatty acid supplementation during pregnancy and lactation may reduce the risk of allergic sensitization to egg and peanut. However, only three of the interventional studies regarding omega-3 fatty acids included in that review were conducted in pregnancy only. In addition, the review¹³ showed that maternal probiotic supplementation in the postnatal period reduced the risk of offspring eczema and allergic sensitization to cow’s milk, and none of the probiotic studies were limited to pregnancy. Since the publication of that systematic review, new studies^14–40 evaluating maternal diet during pregnancy and childhood allergic outcomes have been published. We reviewed diet during pregnancy only as not all mothers breastfeed or introduce solids to their infants in a similar fashion. We therefore reviewed literature regarding the pregnancy period only, to investigate specific pregnancy-related dietary associations.

Globally, recommendations on dietary intake during pregnancy for prevention of allergic disease are sparse. The Australasian guideline recommends eating fatty fish or taking omega-3 fatty acid supplements during pregnancy to reduce eczema in the offspring.⁴¹ Other guidelines from the UK,⁴² Europe,⁴³ and the United States⁴⁴ only give recommendations pertaining to food allergen avoidance in pregnancy. They state that food allergens should not be avoided during pregnancy, but give no further dietary recommendations for allergy prevention. The German allergy prevention guidelines⁴⁵ recommend that pregnant women should follow a healthy diet, need not avoid food allergens, and should eat fish regularly. In support of this, the UK prevention guidelines state that omega-3 fatty acids (found in oily fish such as salmon, trout, mackerel, and fresh [not canned] tuna) during pregnancy may help reduce the risk of eczema and allergic sensitization (development of antibodies to allergens) in early life.

The goal of this review was to summarize studies which examined the impact of maternal diet during pregnancy on offspring’s allergic outcomes. We also compared the amount of nutrients or foods consumed and food patterns identified, which have been shown to have an allergy preventative effect to current US dietary guidelines.

2 |. METHODS

2.1 |. Studies included in this review

2.1.1 |. Types of studies

We included RCTs, observational (cross-sectional and cohort), and case-control studies, which (a) considered the time period of pregnancy alone, (b) reported allergic outcomes or allergic sensitization status in the offspring, and (c) were written in English. Studies examining food allergen intake were excluded. Studies that included intake of major allergens such as milk or nuts were included, as long as the outcome did not relate to the particular food allergen. As an example, studies on tree nut intake were included only when an association with an allergy outcome other than a tree nut was studied, such as asthma.⁴⁶ The latest data indicate that early introduction of a food allergen specifically prevents food allergy to that allergen only.⁴⁷ We were therefore confident that any outcome noted was less likely to be related to the nutritional composition, rather than the allergen content of the food. We also excluded case studies, case series, systematic reviews and meta-analyses, expert reviews and opinions, and editorials.

2.1.2 |. Types of publications

A consensus was reached on including full-text publications only, as there were discrepancies in some of the data published in abstract form (as presented at conferences) when compared to the full paper publications.

2.1.3 |. Types of participants

Pregnant women were the target group for this systematic review, provided there was allergic outcome data published for their offspring. Both general population and populations at high risk of atopy were included.

2.1.4 |. Types of interventions in RCT

Studies that used supplementation or specific foods (eg, fatty fish) during pregnancy alone, irrespective of dose, formulation, or mode of delivery and composition (eg, oil and tablet), were included. Trials were not included if the intervention(s) had been continued beyond pregnancy through breastfeeding or given only to the infant.

2.1.5 |. Types of outcome measures

Studies were included if they reported on allergic outcomes or allergic sensitization status in the offspring, either as a primary end-point or as a secondary end-point. Allergic outcomes were defined as wheeze/asthma, allergic rhinitis/rhinoconjunctivitis/hay fever, food allergy, eczema, allergic sensitization measured by skin prick tests/specific IgE tests to food/aero/any allergen, or total IgE levels. Please see Supplementary File S1 for a list of the search terms used in respective databases. An expert panel from the United States, Europe, and Australia commented on the search strategy and the list of identified and included studies. Additional studies were located through searching the references cited in identified studies and systematic reviews, and through discussion with the expert panel.

2.2 |. Dietary guidelines for food/nutrient intake in pregnancy

To compare our study findings with current dietary guidelines, we used the following approach to calculate food or nutrient intake: For micronutrient comparisons, the US dietary reference intakes (DRI) and the United States Department of Agriculture (USDA) www.choosemyplate.gov guidelines⁴⁸ for pregnant women were used. Macronutrient requirements and number of portions for the second trimester of pregnancy were calculated, assuming a daily caloric intake of 2400 kcal, comprising 71 g of protein, 53.33–93.33gof total fat, <27 g of saturated fat, 13 g of linoleic acid (LA), 1.4 g of alpha-linolenic acid (AI), 175 g of carbohydrates, 28 g of fiber, <12 g of added sugar, 2 cups (2× approx. 125 g) of fruit, 3 cups (3 × 125 g) of vegetables, 8 oz of grains (8 × 30 g), 6.5 oz of high-protein foods (6.5 × 30 g), and 3 cups (3 × 200–250 mL) of dairy. Recommended fish intake was assumed to be 2–3 servings (in total approx. 8–12 oz/240–360 g) per week as per the United States Food and Drug Administration (FDA) (Best Choices: https://www.fda.gov/food/consumers/eating-fish-what-pregnant-women-and-parents-should-know).

2.3 |. Data collection

2.3.1 |. Ethics application/Institutional Review Board

This study is considered exempt from Institutional Review Board review by the University of Colorado School of Medicine Combined Institutional Review Board.

2.4 |. Study selection

The titles and abstracts of articles considered for inclusion were independently assessed by MBA, MP, and AM and categorized as included, not included, or potentially relevant. CV then checked each assessment. For studies considered potentially relevant, full-text copies were obtained, and inclusion of the studies was discussed by CV, MBA, MP, and AM. Any discrepancies were resolved either by consensus among the authors, or, if there was lack of consensus, by discussion with the expert panel. All the studies in this publication were reviewed and approved by both the authors and the expert panel.

2.5 |. Risk of bias assessment

Risk of bias assessment was undertaken independently by MBA and CV using a similar approach as in Garcia-Larsen et al¹³ As in the paper by Garcia-Larsen-et al,¹³ we used modified versions of the Cochrane Collaboration Risk of Bias tool for intervention trials and the National Institute for Clinical Excellence methodological checklist for cohort and case-control studies.⁴⁹ For intervention trials, risk of bias assessment included assessment of: (a) selection bias, including sequence generation and allocation concealment; (b) assessment bias, included blinding of outcome assessors and ensuring outcomes were determined by validated assessment tools; and (c) attrition bias, which was considered high when outcome data were reported in <70% of randomized participants. Risk of bias for cohort and case-control studies included assessment of: (a) selection bias, which was considered low if cases and controls were recruited from similar populations and had a similar attrition rate <20%; (b) assessment bias, which included blinding of outcome assessors and use of validated assessment tools; and (c) confounding bias (did study design and analysis account for relevant confounders?). Conflicts of interest were noted if industry was involved in any aspect of the study or if authors received funding for other activities from relevant industry partners. Using these components, each study was graded as high, medium, or low. Supplementary File S2 includes the grading for risk of bias assessment. In order to minimize publication bias, we performed a comprehensive search of the literature and included experts in the field to ensure that no relevant study was missed. We did not include gray literature (unpublished information or information that is not produced by commercial publishers) in order to avoid discrepancies with published data. None of the meta-analyses included more than 10 papers, and we therefore did not use funnel plot and Egger tests to formally test for publication bias.

2.6 |. Data extraction and reporting

Data were extracted by three authors (AM, MBA, and MP). New articles were added into the excel sheets downloaded from Garcia-Larsen et al¹³ Inclusion and exclusion criteria differed between the current review and that of Garcia-Larsen et al¹³; some entries were edited, modified, or deleted. The edited files were reviewed for quality and completeness by CV. Detailed information on study characteristics was recorded by CV, MP, and MBA and appears in Supplementary File S3. Any discrepancies regarding the classification of dietary measures and offspring allergic outcomes were discussed and resolved within the group of authors. Classification of maternal dietary measures and offspring allergic outcomes were double-checked by the expert panel. The authors’ names, institutions, journals, or data were not blinded during data extraction.

2.7 |. Data analysis

2.7.1 |. Observational studies

For wheeze, studies were split into two categories: infants and young children (0-<3 years) and older children (3 years and older). All studies reporting on allergic rhinitis included offspring with ages over 5 years. All ages were included for food allergy, eczema, and allergic sensitization.

The reported data were plotted. If multiple odds ratios (ORs) were reported for a single study, only the highest exposure compared to the lowest exposure OR was used. Meta-analysis of observational studies was not appropriate because the studies were heterogeneous in focus, design, and target populations. We used forest plots to visually summarize all significant outcomes found by the studies, that is, the adjusted OR, adjusted hazard ratio (HR), or adjusted risk ratio (RR), and their corresponding 95% confidence intervals. If more than one significant outcome was found in a manuscript, then only the largest observed difference was included in this review. For example, if authors compared quartiles, by comparing Q1 vs Q2 and Q1 vs Q4 for one nutrient in the same age-group, only the largest difference would have been included. We understand that it is better to report the estimates regardless of their statistical significance in order to give readers an appreciation of the magnitude of the effect sizes and that statistical significance is purely a function of the sample size of a study and hence provides information only of the random error. It does not give any indication of the magnitude of the estimates or potential systematic errors. However, our goal here was to provide a visual presentation of the data, and the Supplementary Files S1–S9 contain the full set of information. Therefore, findings from observational studies were synthesized narratively by grouping studies according to allergy outcome and nutrient, food, or diet pattern studied: nutrients, including vitamins and minerals, omega-3 fatty acids and fatty fish, fruit and vegetables, other foods, and food patterns.

2.7.2 |. Randomized clinical trial meta-analyses

For each exposure-disease outcome pair included in the RCTs, separate meta-analyses were performed. Because the number of studies was limited, several meta-analyses included fewer than four data points. All meta-analyses included at least two studies. ⁵⁰ DerSimonian-Laird random-effects models were used to perform the meta-analyses and estimate the pooled ORs using the rmeta (v 3.0)⁵¹ package in R. The heterogeneity of the studies was calculated using the Cochran χ² and I² statistics. Forest plots were created for each of the diseases using the forestplot (v1.9)⁵² package in R (v3.5.1). Where studies reported outcomes in the same children at more than one time point, we included each outcome in the meta-analysis as well as the combined outcomes. As the different analyses showed no changes in the outcomes, the overall decision was to use food allergy outcomes at 3 years; eczema at 3 years; allergic sensitization at 3 years; and asthma, rhinitis, and allergic sensitization at 6 years.

3 |. RESULTS

3.1 |. Results of literature search and characteristics of included studies

We included all the relevant published studies on the topic of maternal dietary intake during pregnancy alone identified by Garcia-Larsen et al¹³ (n = 68 papers).^{14–40,46,53–119} An additional search was performed using the same search terms to include studies published after, or potentially missed by their search, including observational studies from August 2013 to February 2019 and intervention studies from January 2017 to February 2019 (Figure 1). Based on this additional search, we identified an additional 27 publications. Of them, 25 were cohort or case-control studies ^14–38 and 2 were RCTs.^39,40 In total, our systematic review is based on 95 papers (17 RCTs and 78 cohort and case-control studies). The numbers of pregnant women and their offspring studied at different time points and presented in different publications are shown in the supplementary Files S1–S9. Supplementary File S4 has information on nutrients, including vitamins and minerals; Supplementary File S5 contains information on fats and fatty fish; Supplementary File S6 is on fruit and vegetables; Supplementary File S7 contains information on other foods; and Supplementary File S8 is on food patterns.

FIGURE 1.

Flow chart for study inclusion

Sufficient numbers (at least two) of RCTs were found for vitamin D, vitamins C and/or E, and omega-3 fatty acids. We performed a total of 9 meta-analyses.

3.2 |. Quality and risk of bias assessment

The 95 papers were generated from 50 studies. The 50 studies included 11 RCTs^{39,40,53–67} (7/11 normal risk), 2 cross-sectional^21,74 (both normal) and 30 cohort studies^{15,18–20,22–30,32–37,46,68–70,72,73,75–78,80–108,110,112,115–119} (all normal risk), 3 case-control designs nested within cohorts^{16,17,31,38,79,113,114} (all normal risk), and 4 case-control studies^{14,71,109,111} (3 normal and 1 high risk). Due to low numbers of RCTs that could be included in the meta-analysis and low numbers of studies including high-risk populations in the observational studies, these were not separated. Of the studies, 8 (7 observational and 1 RCT) (16%;) were considered to be of high quality, 10 (8 observational and 2 RCTs) (20%) were of uncertain quality, and 32 (24 observational and 8 RCTs) (64%) were of low quality. The results for the quality assessment of the studies are shown in Supplementary File S2.

3.3 |. Prevention strategies in pregnant women

Supplementary excel Files S4–S8 summarize the results of all the 95 publications. The excel files contain, if available for each paper, the last name of the first author, publication year, study design, dietary exposure, dietary intake time, outcome type, offspring’s age at assessment, and every relevant point estimate reported with its associated 95% confidence intervals and p-values, as well as the relevant comparison of interest that each point estimate describes. Due to the heterogeneity between papers on the measurement of dietary intake, the analysis of the data, and the reporting of the results, data from observational studies were not pooled in meta-analyses.

3.4 |. Dietary intake

Dietary associations measured in observational studies depend on the validity of the measures used to study dietary intake.¹²⁰ This important methodological factor has so far been omitted from past systematic reviews. Methods for measuring dietary intake differ greatly between studies. We have summarized the approaches for assessing dietary intake and adherence to study interventions used in the RCTs in Supplementary File S3. Of the 17 publications reporting on RCT outcomes, 8 papers ^{39,40,59–62,65,66} mentioned that they monitored adherence and 6 papers^{54,55,57,58,63,64} did not monitor intake. For the observational studies, instruments used to measure food or nutrient intake were validated food frequency questionnaires (FFQ_s),^{15,18,21,23,25–27,30,33,34,36,46,70,75–78,80–84,87–107,110,115,118} unvalidated FFQs,^{16,17,19,28,29,31,32,73,111–114,119} questionnaires,^{14,20,22,35,38,68,69,71,72,74,79,85,86,108,116} interviews,²⁴ information obtained from medical records,¹⁰⁹ 3/5-day food records and/or 24-hour recalls and questionnaires^37,117 These points are important to take into account when reading the data.

3.5 |. Asthma/wheeze outcomes

Early childhood wheeze may not lead to allergic asthma outcomes in later childhood.^121,122 To distinguish between transient wheeze/asthma in infancy and early childhood, and later wheeze/asthma, which may be a stronger predictor of subsequent allergic asthma outcomes, we have separately reported observational study manuscripts on maternal diet and wheeze/asthma into those which considered offspring wheeze/asthma between 0 and <3 years and for offspring 3 years and older. For the meta-analysis, we have included all ages. Allergic outcomes and how these were measured are shown in supplementary File S3.

3.5.1 |. Randomized controlled trials

We had a limited number of studies to perform meta-analysis of antioxidant (vitamins C and E; 2 studies—children of 1–2 years),^58,61 vitamin D (3 studies—children of 3 years old),^55,57,60 and omega-3 fatty acid (4 studies; 5 papers; children of age 6m, 5 – 16 years) ^{62,63,66,67,72} intake or supplementation against wheeze/asthma outcomes in the offspring (Figure 2A–C). Our meta-analyses indicate that vitamin D during pregnancy had a statistically significant protective effect on wheeze/asthma in the offspring (OR: 0.72; 95% CI: 0.56 – 0.92). Although there was a trend toward a protective effect, maternal supplementation/intake during pregnancy with omega-3 fatty acid (OR: 0.70; 95% CI: 0.45–1.08) did not reach statistical significance on wheeze/asthma outcomes in the offspring. If we remove the study by Noakes et al who reported on wheeze in the first six months of life, significance is still not reached (OR: 0.63; 95% CI: 0.4–1.00—meta-analysis not shown). Using data from children at 1–2 years, anti-oxidant did not show a protective effect against asthma/wheeze (OR: 0.67; 95% CI: 0.29–1.56).

FIGURE 2.

A-C, Randomized controlled trials with asthma/wheeze as an outcome

3.5.2 |. Observational studies

Many different outcomes have been reported as summarized in supplementary File S9. Sixty-two papers from observational studies^{14–16,19–26,28,30–33,35–38,46,68–70,74–80,82–93,95,96,98–101,103–111,113,117–119} reported on the association between maternal diet during pregnancy and asthma/wheeze as an outcome (Table 1; Figures 3A and B and 4A and B). We summarized all the significant outcomes reported in Figure 3 to give a visual presentation of the outcomes rather than making a statistical conclusion. Figure 4 indicates that a large number of foods/nutrients/food patterns were associated with an increased or reduced risk of asthma/wheeze in children 3 years and over. The most significant protective factors were foods associated with a Mediterranean diet, meat, vitamin D, and fatty fish. The most significant factors associated with an increased risk were pasta, fish sticks, and arachidonic acid. From the summarized data, no firm conclusion can be drawn in regard to the protective effect of nutrients/foods/food patterns studied on allergic disease. However, when considering asthma, reduced offspring allergic outcomes were associated with intake of copper, vitamin D, Mediterranean diet and Western diet, fish/fatty fish, tree nuts, fruit, and meat (type of meat not specified) in the observational studies included.

TABLE 1.

Correct ref numbers

Food and dietary patterns	Asthma/wheeze	Hayfever/rhinitis	Eczema	Food allergy	Sensitization
Total	62 papers14,16,19–26,28,30–33,35–38,46,68–70,74–80,82–93,95,96,98–101,103–111,113,117–119	19 papers15,16,18,21,24,25,31,32,46,70,90–93,103,104,106,107,113	38 papers15,16,19,20,24,26,29,31,37,68,70,72,76,77,81,82,84–88,90,94–101,103,104,107,108,112,113,116,118	5 papers34,71,77,115,118	19 papers16–18,24,27,31,68,73,75–77,82,86,102,110,112–114,118
Fruit and/or vegetables (Supplementary excel file 6 Fruits and Vegetables)	11 papers20,28,36,69,75,78,90,99,106,108 Fruit, vegetables, leafy vegetables, roots and potatoes, citrus fruits, Malaceous fruits, berries, fruit and berry juices, Green and yellow vegetables, Cruciferous vegetables, Folate vegetables, apples	1 paper106 classified as fruits, vegetables, leafy and root vegetables, berries, malicious, citrus fruit and juices	4 papers20,99,108,112 fruits, vegetables, citrus fruits, apples, green leafy vegetables, green and yellow vegetables, carrots, spinach, cabbage, celery, tomato, sweet pepper, salad, and juice	1 papers115 fruit and berries, vegetables and roots	3 papers17,102,112 fruits, vegetables, malicious fruits, citrus fruits, berries, apples, bananas, strawberry, exotic fruit, root vegetables, carrots, spinach, cabbage, celery, tomatoes, sweet pepper, salad, and fruit and vegetable juices
Fats and fatty fish (Supplementary excel file 5 Fish and Fats) Fat and fatty acid intake (4 papers), classified as total fat intake, omega 3 or omega 6 poly-unsaturated fatty acids or other types of fatty acids, and specific fatty acids (we also included fish intake and fish oil supplementation in this group);	12 papers20,32,36,74,78,85,95,96,104–106,111	432,90,104,106	920,29,85,95–97,104,112,116	1	4 papers27,73,102,112
Vitamin and mineral intake (Supplementary excel file 4 vitamins and minerals)	33 papers14,15,22,23,25,26,28,30,33,35,36,38,68,77,79,83,84,86–89,91,98–100,107,110,117–119 multivitamin, pre-natal vitamins, folic acid, vitamin A, vitamin C, vitamin D, vitamin E, vitamin K,B-vitaminsand their precursors, zinc, calcium, selenium, copper, magnesium, and manganese	6 papers15,18,25,103,107 folate, riboflavin, retinol, vitamin A, vitamin D, vitamin E, vitamin K, and their precursors, zinc, copper, selenium, magnesium, calcium, manganese	14 papers15,26,68,77,84,87,88,98–100,103,107,116,118 vitamins A, C, E, and their precursors, vitamin D, vitamin B and folic acid, mineral intake zinc, calcium, copper, magnesium, manganese and selenium	4 papers71,77,115,118 antioxidant vitamins C, E, and B-carotene, vitamin D and folic acid, mineral intake, copper and zinc	9 papers17,18,68,77,86,102,110,114,118 Vitamin A, Vitamin C, Vitamin D, Vitamin E, Folic acids, Prenatal vitamins, Multivitamins, Anti-oxidants and their precursors, Zinc, Selenium
Other dietary exposures (Supplementary excel file 7 other foods)	25 papers16,19,20,26,31,32,37,46,69,70,75,76,80,85,92,93,95,96,98,105,106,108,109,111,117 sugar, milk/dairy products, grains (pasta), nuts, meat, fish, chocolate candy, soft drinks/artificial sweeteners, egg, alcohol, glucose and dietary water	8 papers16,31,32,46,70,92,93,106 sugar, milk/dairy products, cereals/grains, fish, seafood, meat, chocolate candy, nuts, soft drinks/artificial sweetener, alcohol.	22 papers16,19,20,26,29,31,37,70,72,76,81,85,90,94–98,108,112,116 alcohol, probiotic drinks, fast foods, meat, dairy, cereals, fish and shellfish, chocolate, cheese, natto, eggs, cream, seeds and nuts.	1 papers115 Cereal and dairy	7 papers16,27,31,73,76,102,112 sugar, fish, nuts, legumes, seeds, mild/dairy products, cereals/grains, egg, chocolate candy, juices and alcohol
Food patterns (Supplementary excel file 8 dietary patterns)	8 papers20,21,24,74,75,82,101,113 Mediterranean, Health conscious, Vegetarian, Prudent, Western, Confectionery, Traditional, Japanese, Vegetable, Fruit and white Rice, Seafood and Noodles Pasta, Cheese and Processed meat, Alternative Healthy Eating index	3 papers21,24,113 Mediterranean, Health conscious, Traditional, Processed, Vegetarian, Confectionery, Japanese, Vegetable, Fruit and white Rice, Seafood and Noodles Pasta, Cheese and Processed meat Vegetable, Fruit and white Rice as well as the Alternative Healthy Eating index	520,24,82,101,113 Mediterranean, Health conscious, Vegetarian, Prudent, Western, Confectionery, Traditional, Japanese, Vegetable, Fruit and white Rice, Seafood and Noodles Pasta, Cheese and Processed meat Vegetable, Fruit and white Rice (VFR) as well as the Alternative Healthy Eating index during pregnancy	0	5 papers17,24,27,75,82,102,112,113 Mediterranean diet, Alternate Healthy Eating Index modified for Pregnancy score, Prudent diet, Western diet, Vegetable, Fruit and white Rice, Seafood and Noodles, Pasta, Cheese and Processed meat, Vegetable, Fruit and white Rice, Seafood and Noodles, Pasta, Cheese and Processed meat, Healthy Conscious, Traditional, Confectionery, Traditional and Vegetarian

FIGURE 3.

A and B, Observational studies with asthma/wheeze (age 0-<3 y) as an outcome

FIGURE 4.

A and B, Observational studies with asthma/wheeze (age 3 y and above) as an outcome

3.6 |. Allergic rhinitis outcomes

Figures 5 and 6 summarize allergic rhinitis outcomes in the offspring for both observational and randomized controlled studies, respectively. All observational studies with significant outcomes are shown in Figure 6A and B. The studies shown in both these figures included results for offspring aged 3 years and above.

FIGURE 5.

Randomized controlled trials with allergic rhinitis/hay fever as an outcome

FIGURE 6.

A and B, Observational studies with allergic rhinitis/hay fever as an outcome

3.6.1 |. Randomized controlled trials

We were able to perform meta-analysis of omega-3 fatty acid supplementation (2 studies, 3 papers—all children over 5 years) ^39,54,66,67 against allergic rhinitis outcomes in the offspring (Figure 5). Although we found a potential protective effect based on the meta-analysis combining two studies of omega-3 supplementation during pregnancy on allergic rhinitis in the offspring, the effect showed a trend toward statistical significance (OR: 0.76; 95% CI: 0.56–1.04).

3.6.2 |. Observational studies

Nineteen papers from observational studies^{15,16,18,21,24,25,31,32,46,70,90–93,103,104,106,107,113} reported an association between maternal diet during pregnancy and offspring allergic rhinitis/rhinoconjunctivitis as an outcome. Different studies reported different terms to describe outcomes and definitions (ie, some studies referred to ever suffering from hay fever and others reported on current allergic rhinitis) were not always similar, though the same disease group was being studied. The results are described and summarized in both Table 1 and Figure 6.

Figure 6A and B indicates that vitamin D, vitamin A, fatty fish, polyunsaturated fatty acids (PUFA), alpha-linolenic acid, peanut, pistachio, and probiotic milk products were associated with reduced risk of allergic rhinitis. Fish (unspecified type), the ratio of omega-6 to omega-3 fatty acids, butter, and fruit were all associated with increased risk of allergic rhinitis. As above, due to heterogeneity between studies, these data were summarized narratively and not pooled for meta-analysis. Vitamin A, vitamin D, dairy, nuts, and probiotic-containing foods showed associations with reduced allergic rhinitis outcomes.

3.7 |. Eczema outcomes

Eczema usually develops in infancy¹²³ prior to food allergy and other allergic manifestations, making it a target for allergy prevention studies. Data presented in this systematic review include offspring data from 6 months onwards.

3.7.1 |. Randomized controlled trials

We were able to perform meta-analysis on the effects of vitamin D^55,57,60 (3 studies) and omega-3 fatty acid intake⁶² or supplementation^{39,53,54,56,65–67} (4 studies, 8 papers; 4 used in meta-analysis) on offspring eczema outcomes (Figure 7A,B). None of the dietary exposures during pregnancy had a significant effect on offspring eczema (vitamin D [OR: 0.90; 95% CI: 0.71 – 1.15] and omega-3 [OR: 1.07; 95% CI: 0.82–1.40]).

FIGURE 7.

A and B, Randomized controlled trials with atopic dermatitis as an outcome

3.7.2 |. Observational studies

Thirty-eight papers from observational studies^{15,16,19,20,24,26,29,31,37,68,70,72,76,77,81,82,84–88,90,94–101,103,104,107,108,112,113,116,118} reported the association between maternal diet during pregnancy and eczema in their offspring (Table 1; Figure 8). Due to heterogeneity between studies, these data were not pooled for meta-analysis.

FIGURE 8.

A and B, Observational studies with atopic dermatitis as an outcome

Figure 8A and B summarize the significant associations published between various components of maternal diet during pregnancy and offspring eczema that were reported from observational studies. In terms of vitamins and minerals, studies presented significant associations for a decreased risk of offspring eczema with maternal intake of beta-carotene, vitamin E, zinc, calcium, magnesium, and copper during pregnancy. Paradoxically, studies showed both positive and protective associations between maternal intake of vitamins C and D and offspring eczema. Intake of fatty fish, margarine, vegetable oil, total fat, omega-6 fatty acid, omega-3/omega-6 ratio, linoleic acid, alpha-linoleic acid, natto, and MUFAs all was positively associated with offspring risk of eczema. Cholesterol and arachidonic acid showed a decreased association with eczema. Looking at other foods, fast foods, shellfish, alcohol, and meat (not defined) were associated with an increased risk of offspring dermatitis. Maternal intake of milk, fish, vegetables, and apples was associated with a reduced risk of developing AD. Intake of copper, vitamin E, calcium, zinc, beta-carotene, magnesium, dairy/probiotic foods, vegetables, and fruit was associated with reduced eczema outcomes.

3.8 |. Food allergy outcomes

3.8.1 |. Randomized controlled trials

There were two (three papers)^56,65,66 RCTs that studied the effect of maternal intake of omega-3 fatty acid during pregnancy and food allergy outcomes in the offspring. The meta-analysis is shown in Figure 9, revealing that maternal intake/supplementation of omega-3 fatty acids during pregnancy had no significant effect on food allergy in offspring (OR: 1.18; 95% CI: 0.56 – 2.46).

FIGURE 9.

Randomized controlled trials with food allergy as an outcome

3.8.2 |. Observational studies

Five publications from observational studies^{34,71,77,115,118} discussed associations between maternal diet during pregnancy and food allergy in their offspring (Table 1). Figure 10 summarizes significant results from observational studies regarding the association between various components of maternal diet during pregnancy and food allergy in offspring. The studies reported protective associations between maternal intake of copper and vitamin C and risk of food allergy outcomes in the offspring. Maternal intake of vegetables and vitamin D was associated with increased risk of food allergy in the offspring.

FIGURE 10.

Observational studies with food allergy as an outcome

3.9 |. Allergic sensitization outcomes

Allergic sensitization in the included studies measured both allergic sensitization to food and/or aero-allergens, using skin prick test (SPT) and/or specific IgE levels. Different outcome measures were used, including increased total IgE, a range of specific IgE or SPT to foods and aero-allergens, and amalgamation using the term atopy to indicate any level of allergic sensitization.

3.9.1 |. Randomized controlled studies

Three^55,57,60 RCTs reported results on the effects of vitamin D on allergic sensitization in the offspring, and five reported on the effects of omega-3 fatty acid intake⁶² or supplementation^{39,54,56,65–67} on the same outcome. The meta-analysis is shown in Figure 11A, B. The meta-analyses showed no effects for either maternal intake of vitamin D (OR: 0.91; 95% CI: 0.68 – 1.21) or intake/supplementation of omega-3 fatty acid (OR: 01.02; 95% CI: 0.77 – 1.36) during pregnancy with offspring allergic sensitization outcomes.

FIGURE 11.

A and B, Randomized controlled trials with “any” sensitization as an outcome

3.9.2 |. Observational studies

Nineteen papers from observational studies ^{16–18,24,27,31,68,73,75–77,82,86,102,110,112–114,118} studied the association between maternal diet during pregnancy and allergic sensitization as an outcome.

Data shown in Figure 12 and B and summarized in Table 1 indicate that vitamin C, vitamin D, copper, omega-6 PUFAs, PUFA, butter, fish, the Mediterranean diet, and the seafood and noodle diet pattern were associated with reduced allergic sensitization. Fruit, celery, sweet pepper, free sugar, fatty fish, vegetable fat, and vegetarian and health-conscious diet were associated with increased prevalence of allergic sensitization. Intake of vitamin D, copper, vitamin C, and diet containing soy, fish, and nuts was associated with reduced prevalence of sensitization.

FIGURE 12.

A and B, Observational studies with “any” sensitization as an outcome

3.10 |. Dietary intake and its association with allergy outcomes

Where data allowed, we summarized the dose of nutrients/food intake associated with allergy outcomes in observational studies (Tables 2 and 3). We find no clear association between dose of nutrient/food required to reduce or increase any specific allergy outcomes. Of note is that the dose/amount showing an association frequently does not comply with national dietary recommendations from the United States.¹²⁴ The following associations between diet pattern, food and/or nutrient intake and allergic outcomes was reported in the observational studies: Vitamin A (allergic rhinitis), vitamin D (asthma, allergic rhinitis, and sensitization), copper (food allergy, eczema, and allergic sensitization), vitamin C (food allergy and allergic sensitization), vitamin E (eczema and asthma), calcium (eczema), zinc (eczema and asthma), beta-carotene (eczema), magnesium (eczema), dairy/probiotic-containing foods (eczema and allergic rhinitis), Mediterranean diet and Western diet (asthma/ wheeze), Mediterranean and soy/fish/nuts diet (allergic sensitiza- tion), fruit and vegetables (asthma and eczema), fish (sensitization), nuts (asthma and allergic rhinitis), and meat (asthma). We suggest that researchers use nutrition guidelines to set predetermined cutoffs for food intake to (a) compare outcomes in observational studies and (b) determine dose given during RCTs. This will help ensure safe dietary interventions and will also allow comparison of future studies.

TABLE 2.

Comparison of food and nutrient (where available) against RDA or AI

Wheeze/asthma
Food/nutrient intake	Referencea	Intake	RDI or AI	% RDI or AI
Alpha-tocopherol28	Increase	6.5 mg	15 mg	43%
Calcium98	Reduce	571.1 mg	1000 mg	57.1%
Folate30	Increase	>578	600 μg	>96.3%
Vitamin D15	Reduce	4.69–35 μg	15 μg	31.2%–233%
Vitamin D15	Reduce	3.51–4.68 μg	15 μg	23.4%–31.2%
Vitamin D84	Reduce	724 IU (18.1 μg)	15 μg	120%
Vitamin D23	Reduce	4.86–17	15 μg	32.4%–113%
Vitamin D23	Reduce	3.44–4.85	15 μg	22.9–32.3
Vitamin D91	Increase	8.2, 13, 16.5	15 μg	54.7%, 86.7%, 110%
Vitamin D98	Reduce	5.1, 6.4	15 μg	34%, 42.7%
Vitamin E15	Reduce	6.22–7.08 mg	15 mg	41.5%, 47.2%
Vitamin E88	Reduce	7.07–30.30	15 mg	47.1%–202%
Vitamin E23	Reduce	10.3–29.4	15 mg	68.7%–98%
Vitamin E99	Reduce	7.3 mg	15 mg	48.6%
Zinc88	Reduce	14.25–30.30 mg	11 mg	126%–268%
Zinc99	Reduce	7, 8.5 mg	11 mg	61.9%–75.2%
Omega 6105	Reduce	<8.03 g	13 g	61.8%
Alpha-Linolenic acid105	Increase	<7.78 g	13 g	58.9%
N-3 poly-unsaturated fatty acid105	Increase	<2.24 g	1.4 g	160%
n-3 poly-unsaturated fatty acid96	Reduce	2.2 g	1.4 g	157%
Alpha-Linolenic96	Reduce	2.3 g	1.4 g	164%
Vegetables106	Increase	<149.1	375 g	39.7%
Fruit106	Increase	326.2	250 g	130.5%
Vegetables28	Reduced	206–286 g/d	375 g	54.9%–76.3%
Free sugar16	Increase	82.4–345.1 g/d	<12 g	687%–2875.8%
Meat96	Reduce	63.6 g/d	71 g	89.6%
Dairy products98	Reduce	280.7 g/d	250 g	115.%
Allergic rhinitis
Vitamin D - food107	Reduce	4.4–5.72 mg	15 μg	29%–38.1%
Alpha-linolenic acid104	Reduce	6.4–8.9 g	1.4 g	457%–635.7%
Fruit106	Increase	145.7–1700 g/d	250 g	58.3%–680%
Atopic dermatitis
Beta-carotene99	Reduce	1923.9 and 4218.0 mg	770 μg	41.5% and 91.3%
Vitamin D26	Increase	8.6 μg/d	15 μg	57%
Vitamin E15	Reduce	2.32–2.9 mg	15 mg	15%–0.19%
Vitamin E99	Reduce	86.7 mg/d	15 mg	578%
Zinc88	Reduce	10.81, 13.4, 22.27 mg/d	11 mg	95%, 117% 194%
Calcium26	Reduce	677.6 mg/d	1000 mg	67.8%
Omega - 696	Increase	14.1	13 g	107%
Linolenic acid96	Increase	10.4–11.3 g	1.4 g	743%–807%
Omega 697	Increase	10.6 g/d	13 g/d	81.5%
Total fat97	Increase	58.6/d	80 g/d (range 53–93 g)	3.25%
Alpha Linolenic acid97	Increase	1.9	1.4 g/d	136%
Vegetables99	Reduce	Total Veg 144.4 g	3 × 125 g	39%
Total dairy products98	Reduce	120.8 g/d	500 ml/g/d	24%
Total dairy products26	Reduce	255.3 g/d	500 ml/g/d	50%
Total fish90	Reduce	≥1×/wk vs no intake	2–3 times per week	33%–50%
Food allergy
Vitamin C118	Increase	102–201 (mg/d)2	85 mg	120%–236%
Copper118	Increase	>2.6 (mg/d)2	1 mg	260%
Vegetable115	Reduce	>324 g/d (21/2 cups) vs 153–324 g/d (1–2.5 cups)	3 cups	75% to >100%
Cereal115	Reduce	Both having less than 143 g/d (5 × 1 oz) and more than 230 g/d (8 × 1 oz)	240 g	Decrease if intake is <60% and >95%
Sensitization
Vitamin C118	Increase	130	85 mg	216.7%
Copper118	Increase	1.6	1 g	160%
Free sugar16	Increase	82.4–345.1 g/d	<12 g	687%–2875.8%

Note: Italic = below recommended intake; bold = above recommended intake; normal – stretch across lower than higher than recommended intake.

^aListing only significant OR/RR/HR.

TABLE 3.

Summary table for systematic review

	Decreased				Increased
Observational trials	Vit/Min	Fats/fatty fish	Foods	Diet Pattern	Vit/Min	Fats/fatty fish	Foods	Diet Pattern
Food allergy	Copper Vitamin C	NA	NA	NA	Vitamin D	NA	Vegetables	NA
Atopic dermatitis	Vitamin E Copper Calcium Zinc Beta-carotene Magnesium	Cholesterol Arachidonic acid	Green vegetables Dairy Fruit (apples) Fish Probiotic milk	NA	Vitamin D Vitamin C	Vegetable oil Fatty fish Margarine Linoleic acid Omega-6 fatty acid Ratio of omega-3/omega-6 Alpha-linolenic acid Total fat Monounsaturated fatty acids	Alcohol Shellfish Meat Fast food Natto	NA
Asthma/wheeze	Vitamin E Zinc Copper Manganese Folic acid Vitamin D Vitamin C Alpha-tocopherol Calcium Vitamin K	DHA Fatty fish Palmitic acid Arachidonic acid Alpha-linolenic acid Saturated fatty acid Olive oil Omega-3 fatty acid	Vegetables Fruit (apples)	Mediterranean Western	Folic acid Vitamin D Vitamin A Vitamin K Alpha-tocopherol Vitamin C	Omega-3 fatty acid Alpha-linolenic acid	Fruit Vegetables	NA
Rhinoconjunctivitis	Vitamin D Vitamin A	Fatty fish Polyunsaturated fatty acid Alpha-linolenic acid	Alcohol Fish Probiotic foods Peanut and pistachio	NA	Vitamin D	Butter Omega-6/omega-3 ratio	Fruit	NA
Sensitization	Vitamin C Copper Vitamin D	Butter Omega-6 fatty acids Polyunsaturated fatty acid	fish	Mediterranean Soy, Fish, nuts	NA	Vegetable fat Oily fish	Fruit Vegetable (celery and sweet pepper) Free sugar	Health-conscious Vegetarian
	Decreased				Increased
RCTs	Vit/Min	Fats/fatty fish	Foods	Diet Pattern	Vit/Min	Fats/fatty fish	Foods	Diet Pattern
Food allergy	NA	No effect	NA	NA	NA	No effect	NA	NA
Atopic dermatitis	No effect	No effect	NA	NA	No effect	No effect	NA	NA
Asthma/wheeze	Vitamin D	No effect	NA	NA	NA	No effect	NA	NA
Rhinoconjunctivitis	NA	No effect	NA	NA	NA	NA	NA	NA
Sensitization	No effect	No effect	NA	NA	No effect	No effect	NA	NA

Note: Italic = only associated with reduced outcomes.

4 |. DISCUSSION

4.1 |. Summary of findings

We find insufficient evidence to provide guidance on diet diversity (no published reports), diet patterns, diet indices or specific foods, food groups, and macro- or micronutrients that should be consumed or avoided during pregnancy for the prevention of allergic diseases in the offspring. However, based on 3 RCTs, we did find that intake of vitamin D, higher than EFSA and IOM recommendations, was significantly associated with a reduced risk of wheeze/asthma. This needs to be further studied before guidelines are changed, particularly as intake of vitamin D in both lactation¹²⁵ and early life is associated with increased food allergy in the offspring.

4.2 |. Strengths and limitations

We included the most up-to-date studies on dietary intake during pregnancy and prevention of allergic disease in their infants. Studies from Europe, North America, Asia, and Australia were included. We used the stringent standard employed by Garcia-Larsen et al¹³ to conduct the search and perform the risk of bias assessment and data extraction. We brought together a panel of international experts to review the data and conclusions of this review.

The review was limited by the quality of studies included. Some 64% of the studies included were deemed to be of low quality. The pool of data is also relatively small with 50 studies contributing to 95 papers. The observational studies were highly heterogeneous and were therefore not considered to be suitable for meta-analysis. There is a paucity of RCTs in this field; only four meta-analyses were performed using data from RCTs. Finally, none of the studies looked at diet diversity during pregnancy, this may be partly due to the uncertainties in defining diet diversity, which has now been addressed in an EAACI position paper.⁹

4.3 |. Comparison of findings to previous studies

The systematic review by Vadhaninia et al¹²⁶ based on RCTs assessed the impact of omega-3 fatty acid supplementation solely in pregnancy and found that based on pooled estimates, omega-3 fatty acid intake significantly reduces allergic sensitization to egg and peanut in the offspring. We did not find a significant effect on any allergic sensitization outcome. The discrepancy may be attributed to the fact that one of the studies included by Vadhaninia et al¹²⁶ was excluded from our review, as maternal supplementation continued up to 2.5 months post-delivery.¹²⁷ Our findings are also different from the conclusions drawn by Garcia-Larsen et al.¹³ The difference probably lies in the choice of studies included in each systematic review. Garcia-Larsen et al¹³ included studies in pregnancy, lactation, and early life. We considered studies in pregnancy alone. It is therefore possible that supplementation with omega-3 fatty acid, which continues during breastfeeding and early life, may be more effective than just during pregnancy alone to prevent allergic outcomes in the offspring.

In terms of probiotic supplementation, published literature suggests that probiotic supplementation during late pregnancy, breastfeeding, and early life^{13,43,128–135} may be associated with reduced risk of the development of eczema. Substantial uncertainties remain regarding the specific strain required and the optimal timing of supplementation. We were unable to find any RCTs using probiotic supplementation in pregnancy only. We did find only one observational study that examined the association between maternal intake of probiotics and offspring allergy, but the authors found no association with allergic outcomes.¹⁰⁶

We were able to demonstrate that vitamin D supplementation in pregnancy may reduce the risk of wheeze/asthma.^55,57,60 Vadhadania et al¹³⁶ also concluded that prenatal supplementation of vitamin D might be associated with reduced odds of recurrent wheezing in children. We were unable to find any RCT on folic acid supplementation in pregnancy only and observational studies provided conflicting evidence.^{14,22,28,30,33,35,38,68,71,77,79,83,100,103,114,115,119} There is lack of evidence on the effect of other vitamins for the prevention of respiratory and/or allergic outcomes.

Many systematic reviews have made recommendations, which relate to maternal diet in pregnancy and allergy prevention in the offspring. The review by Lodge et al¹³⁷ did not identify any nutritional factors during pregnancy that reduced the risk of food allergy in the offspring. Four reviews recommended omega-3 fatty acid intake in pregnancy, vitamins A, D, and E; zinc; fruits and vegetables; and a Mediterranean diet for the prevention of asthma.^43,137–140 Reviews focusing on allergy prevention in general concluded that pregnant women should consume their normal diet and that maternal intake of Mediterranean dietary patterns, diets rich in fruits and vegetables, fish, and vitamin D-containing foods may be associated with reduction of offspring allergic disease.^{10,43,131,132,137,140–146} We were unable to make generalized recommendations from the current review. Another factor that needs to be considered is whether family, paternal, and/or specifically maternal history of allergy play a role in the outcomes measured.

Since the review was performed, one study has shown an association between Diet inflammatory index (DII) scores in pregnancy and childhood asthma outcomes over 10 years.¹⁴⁷ Another study found an association with wheeze trajectories in the child, but not asthma up to 7.5 years of age¹⁴⁸ indicating that perhaps, the inflammatory potential of the diet needs to be assessed when focusing on asthma and/or wheeze outcomes.

4.4 |. Interpretation of the evidence

The heterogeneity of the results from observational studies yields little guidance regarding changes in the maternal diet that may be associated with reduction in offspring allergic diseases. Several nutrients, foods and food patterns such as vitamin A, vitamin D, copper, vitamin C, vitamin E, calcium, zinc, beta-carotene, magnesium, dairy/probiotic-containing foods, Mediterranean diet and Western diet, Mediterranean and soy/fish/nuts diet, fruit and vegetables, fish, nuts, fish/fatty fish, and meat showed associations with reduced allergy outcomes in the observational studies. But amounts consumed do not relate to current nutrition guidelines. Yet, our meta-analysis of RCTs conducted in pregnancy does show some significant and replicable results. Randomized intervention trials in pregnancy show a preventative effect of vitamin D supplementation of 800,⁵⁷ 2400,⁵⁵ or 4000⁶⁰ IU per day on wheeze/asthma; though, EFSA and the IOM recommend only 600 IU of vitamin D per day during pregnancy.

4.5 |. Clinical, policy, and research implications of the findings

The current systematic review does not lead to any changes in clinical practice or policies but clearly highlights the research need for more RCTs in this field, particularly focusing on diet patterns or total dietary intake. There may be a particular need to review vitamin D recommendations during pregnancy.

5 |. CONCLUSIONS

The lack of consistent results across the different studies presented in this review may largely be influenced by the lack of a standardized approach to supplementation in terms of dose and duration, disease definition/outcomes used, and individual host features that are difficult to compare across studies. Polymorphisms in genes associated with catabolism and utilization will influence nutrient requirements and function. GWAS-led prevention and intervention studies, including functional microbiome, immunologic, metabolomic, and lipidomic assessments, are required and will increase our understanding of the importance of specific nutrients in the natural course of allergies and asthma. It is likely that a custom-individual-tailored approach to nutrition, focusing on overall dietary intake which may include supplementation, is required to observe the optimal benefits that can potentially be derived from dietary factors in the prevention and treatment of allergies and asthma. Future research and clinical efforts should be focused on large, adequately powered human studies of rigorous methodological design. These studies should focus on identifying the key host characteristics (ie, genetics, environmental factors, microbiome, biochemical and inflammatory parameters, and functional clinical characterization) that influence responses, while also taking levels of nutrient intake, history of allergic diseases, and the composition of the total underlying diet and nutrient interactions into account. Improving our understanding of potential strategies to prevent the development of allergic diseases would significantly reduce morbidity, mortality, and costs from allergic diseases. The current evidence mainly focused on single dietary factors, and the data from observational studies were very heterogeneous. Meta-analysis based on 3 RCTs showed some benefits of vitamin D supplementation on asthma/wheeze, but doses used were higher than recommended by EFSA and the IOM. Before routine supplementation during pregnancy for the prevention of allergic disease can be advised, more studies are required.

Key Message.

The systematic review focused on diet during pregnancy only, as not all mothers breastfeed or introduce solids to their infants in a similar fashion. Thus, we hence provide here pregnancy-specific dietary recommendations. Based on RCTs, we found that prenatal supplementation with vitamin D in doses higher than recommended by most countries may have beneficial effects for prevention of asthma in the offspring. Doses used in RCTs were 800 IU, 2400 IU, and 4000 IU. The European Food Safety Authority (EFSA) (https://www.efsa.europa.eu/en/press/news/161028) and the Institute of Medicine (https://www.aafp.org/news/health-of-the-public/20101201iomrpt-vitdcal.html) recommend 600 IU per day for pregnant women. Summarizing the data from observational studies, we found that the following nutrients, foods, and diet patterns appear to be associated with reduced allergic outcomes: vitamin A (allergic rhinitis), vitamin D (asthma, allergic rhinitis, and sensitization), copper (food allergy, eczema, and allergic sensitization), vitamin C (food allergy and allergic sensitization), vitamin E (eczema and asthma), calcium (eczema), zinc (eczema and asthma), beta-carotene (eczema), magnesium (eczema), dairy/probiotic-containing foods (eczema and allergic rhinitis), Mediterranean and Western diet (asthma/wheeze), Mediterranean and soy/fish/nuts diet (allergic sensitization), fruit and vegetables (asthma and eczema), fish (sensitization), nuts (asthma and allergic rhinitis), fish/fatty fish (asthma), and meat (asthma). However, the doses of foods or nutrients studied were heterogeneous and we were unable to relate these to current nutrition guidelines. We found no consistent evidence regarding other dietary factors, perhaps due to differences in study design and host features such as nutrient intake/levels at the start of the study that were not considered. There is a need for confirmatory studies and RCTs beyond single nutrients.

Abbreviations:

AD

Atopic dermatitis

AI

Adequate intake

FA

Food allergy

PC

Prospective cohort study

RDI

Recommended dietary intake