Dietary guideline adherence during preconception and pregnancy: A systematic review

1Dietary guidelines exist to promote health and well‐being through the prevention of diet‐related disorders and nutritional deficiencies.Non‐adherence during the preconception period and pregnancy may have a negative impact on fertility, pregnancy and birth outcomes, and the health of the offspring.Sociodemographic factors are associated with compliance to dietary guidelines.Understanding the factors associated with adherence and non‐adherence can inform future policy and practice aimed at improving overall dietary guideline compliance, during the preconception period and pregnancy.Nutrition education and counselling may improve compliance; however, further studies are needed to determine how the intervention translates to maternal and infant health outcomes.

2Dietary guidelines exist to promote health and well‐being through the prevention of diet‐related disorders and nutritional deficiencies.

3Non‐adherence during the preconception period and pregnancy may have a negative impact on fertility, pregnancy and birth outcomes, and the health of the offspring.

4Sociodemographic factors are associated with compliance to dietary guidelines.

5Understanding the factors associated with adherence and non‐adherence can inform future policy and practice aimed at improving overall dietary guideline compliance, during the preconception period and pregnancy.

6Nutrition education and counselling may improve compliance; however, further studies are needed to determine how the intervention translates to maternal and infant health outcomes.

7Dietary guidelines and nutritional recommendations aim to provide guidance on dietary intake/composition and macronutrient/micronutrient intake, respectively. These guidelines serve to maintain health and well‐being and to prevent diet‐related disorders and nutritional deficiencies (World Health Organization [WHO], 2015). The World Health Organization (WHO) and the Food and Agricultural Organization (FAO) of the United Nations developed the estimated average requirement (EAR) and recommended nutrient intake (RNI) as targets for human nutrient intake. EARs meet the average daily nutritional needs of 50% of healthy individuals by age and gender, and RNIs meet the average nutritional needs of nearly all healthy individuals by age and gender and are set as the EAR plus two standard deviations (WHO and Food and Agriculture Organization of the United Nations, 2004). Some countries describe values specific to their populations, such as Recommended Dietary Allowance (RDA; United States, Canada, and United Kingdom) or Recommended Dietary Intake (RDI; Australia; National Health and Medical Research Council [NHMRC], 2013) that are equivalent to the WHO/FAO RNI. Some nutrient intake targets are described as either an acceptable intake (AI) instead of a RNI/RDA/RDI where an EAR has not been established or an upper limit (UL) where excessive intake of the nutrient may cause adverse effects on health. Many countries utilize nutrient EAR, RNI, AI, and UL as a foundation to develop national food‐based dietary guidelines, where foods are grouped and recommendations are made with reference to daily food group servings in an attempt to meet estimated nutrient EAR, RNI, AI, and UL from the diet. Although food group serving suggestions only provide an estimate of EAR, RNI, AI, and UL due to their variant composition, they enable a simplified approach to providing populations with food‐based dietary guidelines that attempt to meet daily nutritional needs (World Health Organization and Food and Agriculture Organization of the United Nations, 2004). Although WHO recommends that all countries adhere to these guidelines (WHO, 2015), non‐adherence remains an ongoing issue across the globe (Development Initiatives, 2017; WHO, 2017) with one in three persons experiencing undernutrition, malnutrition, nutritional deficiencies, or obesity (WHO, 2017).

8Multiple factors can impact guideline adherence, including income, food availability and affordability, individual beliefs and preferences, cultural traditions, and educational, social, geographical, and environmental aspects (WHO, 2015). This sociocultural context can exert considerable influence on a diet at different stages of life; in particular, during conception and pregnancy (Withers, Kharazmi, & Lim, 2018). Poor adherence to dietary guidelines or nutritional recommendations during the preconception period (i.e., the weeks to months before pregnancy occurs) and throughout pregnancy can have a negative impact on fertility, pregnancy and birth outcomes, as well as the future health of the offspring (Craig, Jenkins, Carrell, & Hotaling, 2017; Lane, Robker, & Robertson, 2014; Stephenson et al., 2018). For example, in women, adequate intake of folate during the preconception period may help to prevent neural tube defects in the foetus (De‐Regil, Peña‐Rosas, Fernández‐Gaxiola, & Rayco‐Solon, 2015; WHO, 2017). Parental dietary factors are also implicated in the development of disease in the offspring (Fleming et al., 2018). Preventable non‐communicable diseases in the offspring, including cardiovascular diseases, diabetes, cancer, chronic respiratory diseases, atopic disease, and neurological disorders, have been attributed to parental health (physiology, metabolic factors, and body composition) and nutrition status at the time of conception (Fleming et al., 2018). During pregnancy, a healthy diet may prevent excessive weight gain and the associated risks of gestational diabetes, hypertension, or pre‐eclampsia (Schoenaker, Soedamah‐Muthu, Callaway, & Mishra, 2015; Stephenson et al., 2018). A healthy diet also may assist in the prevention of low birth weight infants, macrosomia, preterm birth, stillbirth, and maternal anaemia (World Health Organisation, 2016).

9Although non‐adherence to national dietary guidelines and nutritional recommendations is a global issue (WHO, 2015), it is not clear to what extent non‐adherence applies to individuals in the preconception period and during pregnancy. Understanding the level of concordance with dietary guidelines in these populations, as well as the factors impacting adherence, may help to inform the development of future policies and/or practices specifically targeted at improving dietary practices during the preconception period and throughout pregnancy.

10This systematic literature review was conducted and reported in accordance with the preferred reporting items for systematic reviews and meta‐analysis: the PRISMA statement (Moher, Liberati, Tetzlaff, & Altman, 2009).

11The aim of this systematic review was to determine the level of adherence to national/international dietary guidelines in their entirety, among adult men and women during the preconception period, and in pregnant women, and the factors associated with adherence.

12Literature searches were conducted in five electronic databases from their inception to March 22, 2018. Table 1 shows the search terms used and the number of articles retrieved in PubMed [NCBI]. A similar search was carried out in CINAHL Plus with Full Text [EBSCOhost], AMED [EBSCOhost], EMBASE [Ovid], and Maternity and Infant Care [Ovid], with some variations to search terms required to accommodate for differences in keywords or subject headings. The reference lists of included papers were also manually searched to detect relevant studies not identified in the electronic search. There were no restrictions on date of publication, study setting, or language.

13As the review focussed on usual dietary intake and habits, it only considered for inclusion observational studies. Excluded from the review were qualitative and experimental studies.

14Studies involving (a) adult men and/or women, aged 18 years or older, that identified as trying to conceive or intending to conceive (i.e., during the preconception period), or (b) pregnant women aged 18 years or older, were eligible for inclusion. Studies involving participants with known preexisting conditions such as polycystic ovary syndrome, obesity, and gestational diabetes were excluded as these populations were likely to be receiving dietary interventions to manage their condition. Obese populations were also likely to have experienced difficulties adhering with dietary guidelines. Also excluded were studies specifically examining participants from known only low or only medium‐high socioeconomic cohorts. These studies were excluded to mitigate contributions to the data with potential to result in skewed effects. Studies examining women's dietary intake throughout the post‐partum period or during lactation were excluded as these groups were not the focus of this review; the exception being if these participants were clearly distinguishable from preconceptual or pregnant populations if reported within the same study.

15Studies observing the usual dietary and/or nutritional intake, and/or dietary habits of participants, the components of which were derived from foodstuffs and beverages, were eligible for inclusion. Studies assessing nutrient intakes arrived from supplement use were excluded, as were studies examining isolated or limited numbers of nutrients (i.e., omega‐3 fatty acids, iodine, calcium, or iron) or limited components of the diet (i.e., total energy or fibre), as these studies cannot be compared with dietary guidelines to determine overall adherence. Also excluded were studies involving participants receiving parenteral nutrition, feeds via percutaneous gastrostomy tubes, and those with special dietary needs (i.e., due to allergy).

16The primary outcome was the level of adherence to national/international dietary guidelines and/or recommendations in full; this could be reported using any measure of adherence. The secondary outcome was the factors associated with the level of guideline adherence; this was not limited to any type of factor or association.

17The title and abstract of all articles were screened by the lead reviewer (CC) against the predefined inclusion and exclusion criteria (Figure 1). Potentially eligible articles were retrieved as full‐text and screened by CC. Two second reviewers (ML or AS), with methodological and content expertise, used the liberal accelerated approach (Khangura, Konnyu, Cushman, Grimshaw, & Moher, 2012) to randomly review 10% of included studies and 10% of excluded studies. Any disagreements were discussed until consensus was reached. If unresolved, another reviewer (ML or AS) was invited to adjudicate (Figure 1).

18Data from each eligible study were extracted by CC using a customized data extraction form. Data extracted from a random selection (10%) of included studies were cross‐checked by a second reviewer (ML or AS) using the liberal accelerated approach, as described previously. The data extracted included author, setting, location, sample characteristics, eligibility, duration of study, study design, exposure, instrument used, main outcome measures, and results (Table 2).

19The lead reviewer (CC) and a second reviewer (ML or AS) critically appraised each included study, independently, using the National Institutes of Health Quality Assessment Tool for Observational Cohort and Cross‐sectional studies (National Heart, Lung and Blood Institute (NHBLI), 2018). The appraisal tool comprised 14 items, with each item rated as either yes (criterion met), no (criterion not met), not applicable, cannot determine, or not reported. An overall quality rating (i.e., good, fair, and poor) was then subjectively determined by the reviewer (CC; Table 3).

20Given the descriptive nature of this review, the analysis of included studies was undertaken in a narrative manner.

21Database searches retrieved 722 articles, and an additional five articles were found by manually searching the reference lists of included studies. After the removal of duplicates, 565 articles remained. Of these, 524 were removed at the title/abstract screening stage, and 23 removed at the full‐text screening stage as they did not meet the eligibility criteria (Figure 1). A total of 18 studies met the inclusion criteria for this review.

22The included studies comprised 15 cross‐sectional and three case–control studies. These studies were conducted in Australia (n = 4) (Bookari, Yeatman, & Williamson, 2017; Hure, Young, Smith, & Collins, 2009; Malek, Umberger, Makrides, & Zhou, 2016; Mishra, Schoenaker, Mihrshahi, & Dobson, 2015), China (n = 4) (Gao et al., 2013; Liu et al., 2015; Yang et al., 2017; Zhang, Zhou, Perkins, Wang, & Sun, 2017), India (n = 3) (Dahiya, 2002; Jood, Bishnoi, & Khetarpaul, 2002; Panwar & Punia, 1998), Pakistan (n = 1) (Sajjad & Khan, 2012), Poland (n = 1) (Bojar, Wdowiak, Humeniuk, & Blaziak, 2006), Canada (n = 1) (Pick, Edwards, Moreau, & Ryan, 2005), Spain (n = 1) (Olmedo‐Requena et al., 2017), Japan (n = 1) (Okubo et al., 2011), Portugal (n = 1) (Pinto, Barros, & Santos Silva, 2009), and the Netherlands (n = 1) (de Weerd et al., 2003). Seventeen studies reported the dietary intake of pregnant women, and five reported the dietary intake of women in the preconception period. No study reported the dietary intake of men in the preconception period. The majority of included studies (n = 16) compared the dietary intakes of participants to national dietary guidelines. Only two studies compared dietary intakes to international dietary guidelines. Nine studies reported on factors predicting guideline adherence (Table 2). The dietary recall method was utilized in seven studies; four of these studies conducted a 24‐hr recall, with the remaining assessing dietary recall over three or four consecutive days. Eleven studies utilized food frequency questionnaires (FFQ) assessed over periods ranging from 1 week to 12 months.

23The quality of the included studies was fair (44% of studies) to good (56% of studies; Table 3). Six studies (Bojar et al., 2006; Bookari et al., 2017; Gao et al., 2013; Malek et al., 2016; Okubo et al., 2011) reported on sample size justification, and six studies (Bojar et al., 2006; Jood et al., 2002; Olmedo‐Requena et al., 2017; Panwar & Punia, 1998; Pick et al., 2005; Pinto et al., 2009) assessed dietary intake more than once over time. Two studies (Bojar et al., 2006; Panwar & Punia, 1998) did not specify the age of the participants. The authors of these studies were contacted for further information; however, no response was received from any author.

24Of the eight studies reporting dietary intake (as food groups) among pregnant women, most indicated that women during pregnancy were least adherent with dietary recommendations for the daily intake of vegetables (Bookari et al., 2017; Dahiya, 2002; Jood et al., 2002; Malek et al., 2016; Mishra et al., 2015), cereals and grains (Bookari et al., 2017;Dahiya, 2002; Malek et al., 2016; Mishra et al., 2015), and most frequently adherent with recommendations for dairy (Bojar et al., 2006; Bookari et al., 2017; Dahiya, 2002; Malek et al., 2016; Pinto et al., 2009) and fruit intake (Dahiya, 2002; Malek et al., 2016; Mishra et al., 2015). Of the eleven studies reporting dietary intake (as nutrients) among pregnant women, the consumption of iron (Dahiya, 2002; Gao et al., 2013; Hure et al., 2009; Jood et al., 2002; Liu et al., 2015; Okubo et al., 2011; Panwar & Punia, 1998; Pick et al., 2005; Pinto et al., 2009; Sajjad & Khan, 2012), folate (Dahiya, 2002; Hure et al., 2009; Okubo et al., 2011; Panwar & Punia, 1998; Pick et al., 2005; Pinto et al., 2009; Sajjad & Khan, 2012; Yang et al., 2017), and calcium (Gao et al., 2013; Liu et al., 2015; Panwar & Punia, 1998; Sajjad & Khan, 2012; Yang et al., 2017; Zhang et al., 2017) were considered by most studies as being inadequate. By contrast, fat intake frequently exceeded daily recommended levels (Dahiya, 2002; Gao et al., 2013; Jood et al., 2002; Liu et al., 2015; Panwar & Punia, 1998; Yang et al., 2017).

25Of the two studies reporting dietary intake as food groups, evidence indicated that most women in the preconception period did not achieve the recommended daily intake of vegetables (Bojar et al., 2006; Okubo et al., 2011) and cereals (Bojar et al., 2006; Okubo et al., 2011), but most frequently met requirements for dairy intake (Bojar et al., 2006). Of the three studies reporting dietary intake as nutrients, the consumption of folate (Hure et al., 2009; Pinto et al., 2009) and vitamin E (Hure et al., 2009; Pinto et al., 2009) was most frequently reported as being inadequate in the diets of preconceptual women. On the other hand, protein intake was more likely to exceed daily recommendations (de Weerd et al., 2003; Olmedo‐Requena et al., 2017).

26Nine studies reported on factors predicting guideline adherence in pregnancy. Most of these studies reported a statistically significant association between higher level of education (Bojar et al., 2006; Okubo et al., 2011; Panwar & Punia, 1998; Yang et al., 2017), older age (Bojar et al., 2006; Okubo et al., 2011; Olmedo‐Requena et al., 2017), or non‐smoking status (Malek et al., 2016; Okubo et al., 2011; Olmedo‐Requena et al., 2017) and adherence to dietary guidelines or nutritional recommendations. The impact of residential location on guideline adherence was contradictory; two studies (Bojar et al., 2006; Malek et al., 2016) reported a statistically significant positive association between level of guideline adherence and living in non‐metropolitan residences, whereas two studies (Gao et al., 2013; Yang et al., 2017) found a statistically significant positive association between level of adherence and living in urban residences. Only one study (Olmedo‐Requena et al., 2017) reported on factors associated with guideline adherence among women during the preconception period. This study found a statistically significant association between guideline adherence and higher social class, non‐smoking status, increased physical activity, and older age.

27This review synthesized the evidence from observational studies reporting the level of adherence to national/international dietary guidelines and/or nutritional recommendations, and the factors impacting adherence, among women in the preconception period, and during pregnancy. The eighteen studies included in this review were conducted across 10 countries, thus representing a broad perspective on the issue. The review found that overall adherence to dietary guidelines during the preconception period and pregnancy was low, with a number of demographic factors shown to have an impact on guideline adherence.

28Evidence from this review suggests that women in the preconception period, and during pregnancy, may not be meeting the minimum dietary requirements for vegetable and cereal grain intake. According to WHO, vegetables and cereal grains should form a part of a healthy diet for the prevention of diet‐related non‐communicable disease and nutritional deficiencies (WHO, 2015). Vegetables and cereal grains are low in fat and key sources of vitamins, minerals, protein, and dietary fibre (National Health and Medical Research Council, 2013). Diets high in grains and vegetables may help reduce the risk of obesity, type 2 diabetes, cardiovascular disease, and cancer (National Health and Medical Research Council, 2013). For preconceptual and pregnancy health, vegetables are an important source of folate, and cereal grains a valuable source of folate and iron (National Health and Medical Research Council, 2013).

29The insufficient intake of certain food groups by preconceptual and pregnant women may also impact nutrient intake, with findings of the review suggesting that dietary folate, dietary iron, and dietary calcium intakes may be inadequate in these populations. This is concerning given the body of evidence supporting an association between adequate preconceptual folate intake and lowered risk of neural tube defects (De‐Regil et al., 2015; WHO, 2017). Further, adequate folate intake during pregnancy may reduce the risk of megaloblastic anaemia and help increase birthweight (Lassi, Salam, Haider, & Bhutta, 2013). Adequate iron intake during pregnancy is important for the prevention of adverse pregnancy and birth outcomes, such as maternal anaemia, low birth weight infants, and preterm births (Peña‐Rosas, De‐Regil, Garcia‐Casal, & Dowswell, 2015; World Health Organisation, 2017). The findings of this review are consistent with that of other reports, which indicate that iron deficiency is a global problem (with two billion people worldwide failing to meet the required dietary intake of iron), and the prevalence of anaemia in women of reproductive age is increasing (Development Initiatives, 2017). Maintaining an adequate intake of dietary calcium during pregnancy is also of importance as it may help reduce the risk of pre‐eclampsia (Hofmeyr & Manyame, 2017).

30Our findings also indicate that fat intake among pregnant women may exceed recommended levels stipulated in national/international dietary guidelines. However, it is important to note that the included studies assessing fat consumption only reported total fat intake (Liu et al., 2015; Panwar & Punia, 1998; Yang et al., 2017) or the combination of fats and oils (Dahiya, 2002; Gao et al., 2013; Jood et al., 2002). In other words, these studies did not differentiate between the types of fats consumed in the diet, such as saturated fats, trans fats, and omega‐3/omega‐9 fatty acids. In light of emerging clinical evidence linking omega‐3 fatty acid intake in pregnancy to improvements in infant cognitive development (Braarud et al., 2018), and the prevention of allergic disease (Best, Gold, Kennedy, Martin, & Makrides, 2016), this information would have been beneficial. Notwithstanding, the finding that total dietary fat intake exceeded recommendations (i.e., 30% of total energy) is still clinically significant given that this may potentially contribute to unhealthy maternal weight gain (World Health Organisation, 2015). The impact of this weight gain for both the mother and infant can be considerable, with maternal obesity shown in multiple systematic reviews (Marchi, Berg, Dencker, Olander, & Begley, 2015) to be associated with an increased risk of pre‐eclampsia, gestational diabetes, macrosomia, congenital abnormalities, still birth, low birthweight infants, and maternal mortality (Marchi et al., 2015).

31The included studies reported three main predictors of adherence to dietary guidelines or nutritional recommendations. Among pregnant women, a higher level of education was the principal predictor of adherence, and among preconceptual and pregnant women, non‐smoking status, and older age were the main predictors. A potential implication of this finding is that consideration needs to be given to how dietary guidelines and recommendations are communicated to different subgroups of preconceptual and pregnant women. Beyond the development of clear and accessible dietary guidelines and nutritional recommendations, consideration could be given to the role of nutrition education and counselling in clinical settings.

32Nutrition education and counselling has been shown in a meta‐analysis of 37 studies to be an effective approach for improving maternal nutrition in pregnancy, particularly in reducing maternal anaemia, increasing infant birth weight, and decreasing preterm births (Girard & Olude, 2012). Findings from a randomized controlled trial also indicate that dietary counselling may be effective in decreasing the proportion of women with excessive weight gain during pregnancy through improvements in food knowledge, diversity of food choices, and attitudes to diet (Abdel‐Aziz, Hegazy, Mohamed, Abu El Kasem, & Hagag, 2018). Strategies aimed at improving nutrition and health behaviours before conception also have been explored. One framework recently described by Barker et al. (2018) acknowledges the importance of identifying the age and life phase of an individual; this may help to identify the individual's motivations and likely receptiveness to different intervention types, thus enabling a clinician to deliver a more targeted approach to preconception management (Barker et al., 2018). Additionally, Barker et al. (2018) suggest that an intervention framework should combine public health policy with strategic engagement from the private sector in order to further improve health outcomes (Barker et al., 2018). Of course, further work is needed to determine whether such an approach actually translates into improvements in maternal and infant outcomes.

33An important knowledge gap identified through this review is the absence of any studies examining dietary intake in males during the preconception period (i.e., the male partner of a woman trying to conceive), and despite having an important role to play in preconception, men have been largely overlooked in the literature to date (Kotelchuck & Lu, 2017). This represents a much needed area for future research, for several reasons. First, epigenetic changes to the DNA of spermatozoa may be accelerated through nutrition and lifestyle factors, and these changes may have long‐lasting transgenerational effects (Lane et al., 2014). Second, paternal obesity has implications for fertility, in particular, an association with disturbed endocrinology, erectile dysfunction, and increased sperm DNA fragmentation (El Salam, 2018). Indeed, reduced rates of clinical pregnancies in assisted reproductive cycles utilizing sperm of obese men have been observed (Mushtaq et al., 2018). Evidence also suggests that the offspring of obese fathers may result in metabolic changes within the offspring and the subsequent inheritance of obesity (Craig et al., 2017). A better understanding of the nutritional status of males during the preconception period, as well as the factors impacting nutrition in this population, would help to provide a targeted approach for couples experiencing infertility.

34Several limitations exist within this review. One limitation is the heterogeneity of instruments used to assess dietary intake. Some studies used dietary recall methods and others FFQs; there was also great variability in reported time frames, ranging from 24 hr to 12 months. These differences in outcomes and outcome measures, and the disparities in how data were reported, meant the pooling of data for meta‐analysis was not possible. Lastly, the dietary intakes reported in the included studies were assessed for adherence against 13 various national food‐based dietary guidelines (i.e., food groups) and/or national or international nutritional recommendations (i.e., RNI/RDA/EAR/AI/UL), dependent upon the country the study was conducted. Given that the recommendations reported in each of these guidelines may differ to some extent, it is possible that adherence (or non‐adherence) to one guideline may not be necessarily transferable to all dietary guidelines. Although beyond the scope of this review, comparisons between countries/regions that utilize RNI/RDA/EAR/AI/UL for nutritional recommendations and food‐based dietary guidelines would be an interesting area for future research to determine how well the two approaches align. For example, a sensitivity and specificity analysis could explore whether populations meeting food group serving recommendations also meet RNI/RDA/EAR/AI/UL recommendations.

35The consumption of a healthy diet during the preconception period and pregnancy can be achieved by adhering to national/international dietary guidelines and/or nutritional recommendations. Concordance with these guidelines should provide some assurance that energy, macronutrient, and micronutrient intake are adequate to support fertility, pregnancy, and positive birth outcomes, as well as the future health of the offspring. The findings from this review indicate that women both in the preconception period and throughout pregnancy may be falling short of targets stipulated in dietary guidelines and nutrient recommendations. In particular, preconceptual and pregnant women may not be consuming enough vegetables, cereal grains, folate, iron, and calcium and may be consuming excess fat. A number of demographic factors were shown to impact guideline adherence, which could have implications for both future policy and practice. However, further research is needed to assess the predictors of guideline adherence for both men and women in the preconception period, as well as the quality of diets of men during preconception.

36The authors declare that they have no conflicts of interest.

37None.

38CC formulated the research question and undertook the database searches, study selection, data extraction, and critical appraisal. AS and ML contributed to the design of the research question, served as second reviewers at the study selection, data extraction, and critical appraisal phases and contributed to manuscript drafting. All authors approved the final manuscript before submission.