Diet and physical activity in pregnancy and offspring’s cardiovascular health: a systematic review

Abstract There is increasing evidence linking maternal diet and physical activity before and during pregnancy with offspring’s cardiovascular health. Although many studies examined this association, the evidence has not been reviewed systematically. We therefore undertook a systematic review to synthesize evidence examining the association of maternal diet and physical activity before and during pregnancy with offspring’s blood pressure and vascular health. We systematically searched the databases MEDLINE and EMBASE from inception to June 30, 2017. Eligibility screening, data extraction and quality assessment were performed by two independent reviewers. A total of 19 articles were included comprising three randomized controlled trials and 16 observational studies. Of the studies that examined the association of interest, 60% (three out of five studies) showed that high maternal carbohydrate intake was associated with higher offspring’s blood pressure. Maternal protein intake during pregnancy was negatively associated with offspring carotid intima-media thickness in two out of two studies. No consistent findings for maternal fatty acid intake were found. There were too few studies to draw conclusions on energy intake, fibre intake, protein/carbohydrate ratio, specific foods, dietary patterns and maternal physical activity. Heterogeneity in exposure and outcome assessment hampered pooling. Also, owing to the observational nature of most studies, causality cannot be established. Harmonization of valid exposure and outcome measurements, and the development of core outcome sets are needed to enable more robust conclusions.


Introduction
Cardiovascular diseases are the number one cause of death globally. 1 Although these diseases manifest in adulthood, there is a large body of evidence to suggest that these diseases originate in early life. [2][3][4] It is thought that adaptations of the developing fetus to its environment may increase susceptibility to disease in later life. [5][6][7][8] Maternal lifestyle preceding and during pregnancy is an important contributor to early-life programming of the offspring. Inadequate maternal nutrition during pregnancy increases the risks of cardiovascular diseases. 9,10 Several studies have shown that the balance of macronutrients in the maternal diet during pregnancy is associated with offspring's blood pressure decades later. 9,11,12 Additionally, maternal exercise seems protective against the development of cardiovascular diseases in the offspring. Offspring of exercising pregnant women appear to have a lower resting heart rate, higher heart rate variability and improved vascular health. 13 However, there is a lack of knowledge regarding the type and amount of exercise needed to favourably program cardiovascular health of the offspring.
Although many studies examined the association between maternal dietary intake and physical activity in pregnancy and cardiovascular health of the offspring, the evidence has not been reviewed systematically. Therefore, we systematically reviewed all currently available evidence on the association of dietary intake and physical activity of women before and during pregnancy with offspring's blood pressure and vascular health. Secondary objectives were to study the potential modifying role of period of gestation, offspring's sex and prepregnancy body mass index (BMI) of the mother.

Methods
This systematic review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The review protocol was registered in the prospective register of systematic reviews (PROSPERO; systematic review record CRD42015020244). The PICOS criteria, used to define the research question and to select the studies, are presented in Table 1. This paper is part of a broader systematic review project regarding the association between maternal lifestyle before and during pregnancy and offspring's cardiometabolic health. In the current paper, we focussed on the cardiovascular outcomes as described in the outcome section of Table 1. A future paper will focus on the anthropometric and metabolic outcomes.

Data sources and search strategy
A clinical librarian (J.L.) performed a systematic search in OVID MEDLINE (including Epub ahead of print, in-process and other non-indexed citations) and OVID EMBASE from inception to June 30, 2017 to identify observational and experimental human studies on (pre)pregnancy maternal diet and physical activity and cardiometabolic health in the offspring. We searched for the concepts 'maternal', 'dietary intake' or 'physical activity', '(pre-)pregnancy' and 'offspring/child', using a wide variety of controlled terms, including MESH and text words. We did not search for specific outcomes as this would increase the risk of missing studies, but combined the search with a broad search filter for observational and experimental human studies. In addition, we applied a systematic review filter to check the existence of systematic reviews. No date or language restrictions were applied. We cross-checked the reference lists and the citing articles of the identified relevant papers in Web of Science and adapted the search in case of additional relevant studies. The bibliographic records retrieved were imported and de-duplicated in ENDNOTE. The complete search strategies are presented in Supplementary Table S1.

Study selection
The studies were independently screened by two reviewers (T.v.E. and M.K.) using the online screening and data extraction tool Exclusion: Studies solely among participants with a pre-existing chronic condition or studies solely among participants treated with medication for overweight-related health problems.

Intervention/ exposure
Inclusion: Maternal diet and/or physical activity is self-reported or objectively measured before or during pregnancy, once or multiple times. Maternal dietary intake is assessed as: Macronutrient intake in grams: carbohydrates, protein, fatty acids; Compliance to country-specific recommended daily intakes as communicated to the national population by health organisations; Diet scores/indices of diet quality; Consumption of food products (e.g., fruit intake, vegetable intake) mentioned in grams or standard portions. Maternal physical activity is assessed as total physical activity or as physical activity in at least one of the four physical activity domains (work, transport, domestic tasks and leisure time): Hours/minutes spent on activities combined with the type of activity or intensity [low, moderate, vigorous or metabolic equivalent (MET)]; Meeting or not meeting the country-specific physical activity guidelines Exclusion: Studies solely focussed on determinants of healthy lifestyle or determinants associated with a successful implementation of a healthy lifestyle. Studies examining the effects of maternal undernutrition or the effects of maternal macronutrient and/or micronutrient supplementation on offspring's health Covidence (www.covidence.org). Studies were eligible for full-text screening if they met the inclusion criteria as described in Table 1.
Full-text articles were independently read by the same reviewers (T.v.E. and M.K.) and inter-reviewer discrepancies were resolved by discussion with a third person (M.v.P., A.G. or R.G.).

Data extraction and quality assessment
The categories used for data extraction can be found in Supplementary Table S2. Data were initially extracted by the first reviewer (T.v.E.), subsequently the second reviewer (M.K.) independently extracted data for 20% of the included articles (n = 10 out of 48 articles; inter-reviewer discrepancy rate = 3.98%). The quality of the included randomized controlled trials (RCTs) was assessed using the Cochrane collaboration's tool for assessing risk of bias. 14 For observational studies the quality assessment tool for observational cohort and cross-sectional studies of the National Institutes of Health (NIH) was used. 15,16 Because of the lack of a scoring system, we did not include the quality rating of this tool in our quality assessment. The quality check was conducted by the first reviewer (Tv.E.), subsequently the second reviewer (M.K.) independently checked 20% of all included studies (n = 10 out of 48 articles: 8 out of 42 longitudinal studies and 2 out of 6 RCTs; inter-reviewer discrepancy rate = 5.69%). Because of the low inter-reviewer discrepancy and the expectation that the errors were not systematic and without influence on either the data extraction or the quality assessment, a duplicate percentage of 20% was considered sufficient.

Data analysis
Results were presented per cardiovascular outcome [blood pressure combined with heart rate, and vascular health which comprised intimamedia thickness (IMT) and pulse wave velocity (PWV)] and were grouped per child development stage. Observed associations were subdivided into positive associations: the higher the maternal diet/ physical activity exposure, the higher the offspring health outcome (▲); negative associations: the higher the maternal diet/physical activity exposure, the lower the offspring health outcome (▼); no association (▬) or other associations (as specified). When maternal exposure was reported continuously as well as categorically, we included the results from the continuous exposure assessment. We did not include results of substitution models when used additionally to study associations of maternal diet with offspring health. When multiple adjusted models were shown, we included results of the fully adjusted model. The full data extraction table is included as a supplement.

Selection of articles and study characteristics
Of the 5145 articles retrieved and screened, 19 articles were judged to be eligible for inclusion in this systematic review. Reference checking of the cited and citing articles of the included articles yielded no additional relevant articles (Fig. 1).
The studies included in this review comprised three articles about intervention studies, [17][18][19] of which two articles used data from the same RCT and 16 articles about observational studies, 20-35 covering 12 mother-child cohorts ( Table 2). During the offspring follow-up, one of the intervention studies did not examine their data as intervention v. control group but combined both groups. 17 There were 16 articles reporting on the association of maternal diet during pregnancy with offspring's cardiovascular health [17][18][19][20][21][22][23][25][26][27][28][29][30][33][34][35] and three articles reporting on the association of maternal physical activity during pregnancy with offspring's cardiovascular health. 24,31,32 No studies included both maternal diet and physical activity in one paper, although the association  arm at heart level, twice at an interval of 5-10 min; Average of the two measurements was used in the analysis.
Carotid intimal medial thickness at 20 years of age: Measured in the distal portion of the right common carotid artery using a Philip machine iE33 and a L10-4 MHz linear array transducer; The mean of six measurements was used in the analysis 1st, 2nd and 3rd trimesters were analysed separately with both maternal exposures was studied in the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort at the same child age. 29,32 In addition, there were no studies examining maternal lifestyle before conception. Three mother-child studies examined offspring health twice over time and published the results in separate articles. 17,19,23,[27][28][29]32 However, studying similar maternal exposures and offspring health outcomes at both points in time was only done in the Nutrition, Allergy, Mucosal Immunology and Intestinal Microbiota (NAMI) RCT. 17,19 In all included articles, maternal diet and physical activity were measured using self-reported questionnaires or interviews. Offspring health outcomes were all measured during clinical examination, according to standardized study protocols.

Blood pressure and heart rate
In total, 16 articles described the association of maternal lifestyle with offspring blood pressure 17,19-24,26-30,32-35 and three articles described the association of maternal lifestyle with offspring heart rate 22,31,34 (Table 3; Supplementary Table S2). Maternal carbohydrate intake during pregnancy was studied in five articles. Its association with infant blood pressure was U-shaped, 17 while it was positively linearly associated to systolic blood pressure in pre-school 19 and school-aged offspring. 28 These linear associations were not observed for diastolic blood pressure. 19,28,35 No associations of maternal carbohydrate intake with blood pressure were observed in older offspring. 29 Maternal fatty acid intake was studied in eight articles and was examined as total fat intake and/or as intake of different specific fatty acids during pregnancy. Maternal mono-unsaturated fat intake during pregnancy had an U-shaped association to infant diastolic blood pressure. 17 Maternal omega-6 and total polyunsaturated fat intake were positively linearly associated to systolic blood pressure in pre-school-aged children. 21 In addition, systolic blood pressure was lowest in offspring of mothers with a fat intake closest to the recommended intake (second tertile v. first tertile of intake). 19 These associations were not observed for diastolic blood pressure. 19,21 Five articles reported no association of maternal fat intake during pregnancy with blood pressure in older offspring (total fat intake during pregnancy, 20,28,29,35 saturated and unsaturated fatty acids 28 and marine n − 3 fatty acids 34 ). However, Adair et al. 20 observed a negative linear association of total fat intake with both systolic and diastolic blood pressure in adolescent girls.
Eight articles reported no association of maternal protein intake with offspring blood pressure. 17,[19][20][21][27][28][29]35 Hrolfsdottir et al. 26 reported a positive linear association of maternal protein intake and offspring diastolic blood pressure in young adults. In contrast, Rerkasem et al. 33 observed a negative linear association of maternal protein intake with offspring diastolic blood pressure. These associations were not observed for systolic blood pressure. 26,33 Associations of maternal energy intake, 17,20,35 protein/carbohydrate ratio (P:C ratio) 21,28,35 and fibre intake 17,21 during pregnancy with offspring blood pressure were only reported in two or three articles with contrasting results. Maternal intake of specific foods 17,22,26,28 and dietary patterns during pregnancy, 23,30 as well as maternal physical activity during pregnancy 24,32 in association with offspring blood pressure, were rarely studied. Furthermore, associations with offspring heart rate were rarely studied. 22,31,34

Vascular health
In total, five articles described the association of maternal lifestyle with offspring vascular health 18 Table S2). PWV as well as IMT were studied in the carotid or in the aortic artery (descending thoracic aorta or proximal abdominal aorta). Two articles reported that maternal protein intake during pregnancy was negatively linearly associated to carotid IMT in school-aged offspring 25 and in young adults. 33 The association of maternal carbohydrate intake with offspring IMT was assessed inconsistently: maternal exposure was defined as glycaemic index or total carbohydrate intake, and vascular health measurements were done in the aortic or carotid artery. 18,25,33 The association of maternal energy intake during pregnancy with offspring carotid IMT was only reported once. 25 Maternal fat intake and its association with offspring carotid IMT was reported in two articles with contrasting results. 25,33 Maternal intake of specific foods 22 and dietary patterns 18,30 were rarely studied in ▲, The higher the maternal diet/physical activity exposure, the higher the offspring health outcome; ▼, the higher the maternal diet/physical activity exposure, the lower the offspring health outcome; − , there is no effect observed of maternal diet/physical activity exposure with infant health outcome. SBP, systolic blood pressure; DBP, diastolic blood pressure; MUFA, mono-unsaturated fatty acids; n − 6, omega-6 fatty acids; PUFA, poly-unsaturated fatty acids; HR, heart rate; SSDN, standard deviation of normal-to-normal inter-beat intervals; RMSSD, root mean square successive difference; LF, low frequency; HF, high frequency; ♂♀, results were stratified for male (♂) and female (♀) offspring.
association to offspring vascular health. There were no articles describing the association of maternal physical activity with offspring vascular health.

Secondary research questions
We additionally focussed on the period of gestation, offspring's sex, and obese v. normal weight mothers. In total, eight articles examined maternal lifestyle multiple times during pregnancy [17][18][19]21,22,25,27,33 (Table 2). Of those articles, four reported their results stratified for pregnancy period 22,25,27,33 with mixed results: Two articles 22,27 did not observe differences in associations stratified for pregnancy period for offspring blood pressure outcomes. However, two other articles only observed associations of maternal carbohydrate intake in late pregnancy 25 and of maternal protein intake in the first trimester of pregnancy 33 with offspring carotid IMT.
Two articles 20,24 stratified their results by sex (Table 3). Adair et al. 20 only observed a negative linear association of maternal total fat intake with blood pressure in females. Danielsen et al. 24 only observed a positive linear association of maternal daily amount of walking and bike riding with systolic blood pressure in males. In total, three articles did not take offspring's sex into account in their final analysis. 17,19,31 Two articles 23,30 added maternal pre-pregnancy BMI as an interaction term into their models, but in both studies it was a noeffect modifier. Seven articles did not take maternal (pre-pregnancy) BMI into account in their final model 17,[19][20][21]27,31,33 (Supplementary Table S2).

Quality of the included studies
The included RCTs scored generally low in the risk of bias assessment (Table 5). However, performance bias was present. ▲, The higher the maternal diet/physical activity exposure, the higher the offspring health outcome; ▼, the higher the maternal diet/physical activity exposure, the lower the offspring health outcome; ▬ , there is no effect observed of maternal diet/physical activity exposure with infant health outcome. GI, glycaemic index; HF, high fibre.
For the observational studies, all articles clearly reported the objective and recruited study participants from the same population applying clear inclusion and exclusion criteria ( Table 6). Different levels of exposure were studied in association to the outcome, with exception of May et al. 31 who used the level of exposure to define two groups for analysis. Most studies used selfreported questionnaires to determine maternal exposure, which were not always validated for the particular exposure of interest, nor a pregnant study population. All outcome assessments in the offspring were rated at low risk of bias, with exception of those in the article by Blumfield et al. 21 who measured blood pressure only once per study visit. 36 The follow-up rate of 80% was most often not reached, with exception of the follow-up rate in the article by May et al. 31 who measured the offspring at 1 month of age. The majority of studies did not correct for key potential confounding variables such as breastfeeding, maternal smoking, maternal age, maternal pre-pregnancy BMI and birth weight, with exception of Chatzi et al. 23 None of the studies gave a sample size justification. Because of the low number of RCTs and the heterogeneity in exposures, we were not able to conclude if RCTs showed different results compared with observational studies.

Discussion
To our knowledge this is the first systematic review on the association of dietary intake and physical activity of pregnant women with offspring's cardiovascular health, including both observational and experimental human studies. In total, we included 19 studies with over 29,000 participants. High maternal carbohydrate intake in pregnancy was consistently associated with higher blood pressure of the offspring. Less consistent associations were observed for high maternal intake of unsaturated fatty acids and low total fat intake with higher offspring blood pressure. There was no evidence for a programming effect of maternal protein intake on offspring blood pressure. Maternal protein intake during pregnancy was negatively associated to carotid IMT in school-aged and young adult offspring. We were unable to assess the potential modifying role of period of gestation, offspring's sex or BMI of the mother, because of the small number of studies reporting stratified results.

Underlying mechanism
We speculate, in line with the results of previous studies, 11,12,37 that the observed associations between offspring blood pressure with maternal carbohydrate intake can be explained by the ratio between maternal protein to carbohydrate intake (P:C ratio). Maternal energy and protein needs increase during pregnancy, which enables the fetus and placenta to grow. 38 A low intake of maternal protein and an increased intake of carbohydrates are associated with reduced placental weight. 12,39 Reduced placental size might induce increased placental flow with lasting consequences for the pressure against which the fetal heart develops. 40 Such increased levels of pressure may have lasting effects for the physiology of heart and blood vessels and might increase later blood pressure. Indeed, there is evidence that reduced placental size is linked to increased risks of hypertension in later life. 41 This also explain the observed association of a lower maternal protein intake with a higher offspring's carotid IMT. But lower overall maternal energy intake altering endothelium-dependent responses in the offspring's aorta 42 could also explain this association, as there is evidence for a negative linear association of adequate maternal energy intake with carotid IMT in school-aged offspring. 25 We observed weak evidence for a programming effect of maternal fat intake with offspring blood pressure. 17,[19][20][21] This is in line with evidence from animal studies, showing that high fat diets before and during pregnancy induced high blood pressure through endothelial dysfunction, including reduced endotheliumdependent vasodilatation in both small and large vessels and increased aortic stiffness. 43

Interpretation of the results
Most of the included studies assumed a linear association of maternal lifestyle with offspring cardiovascular health, or did not report whether assumptions for linearity were justified. U-shaped or trends towards (reversed) U-shaped relationships were also observed. 17 It could be that associations went undetected by using inappropriate statistical models.
Associations may also have gone undetected since most studies failed to report stratified analyses for sex. There is evidence for sex differences in the programming of cardiovascular diseases 20,24 and although the underlying mechanism is unclear, it seems that male offspring are more sensitive to their prenatal environment. [44][45][46] For example, intrauterine growth restriction caused by placental insufficiency resulted in a significant increase in blood pressure in young adulthood in male offspring, whereas female offspring were normotensive. 44,47 Table 5. Quality assessment of the included randomized controlled trial studies according to the Cochrane collaboration's tool for assessing risk of bias a Aaltonen, 2008 17 Kizirian, 2016 18 Normia, 2013 19 Random sequence generation (selection bias) + ? + Allocation concealment (selection bias) + ? + Blinding of participants and personnel (performance bias) + /-a c + /-b Blinding of outcome assessment (detection bias) ? + ?
Incomplete outcome data (attrition bias) + + + Selective reporting (reporting bias) + + + Other sources of bias (other bias) + + + a + = low risk of bias; − = risk of bias; ? = unclear. b + /-for performance bias was given, because all three studies had a partly blinded design (two groups double-blind, one group single-blind). c Blinding of the participants and personnel was not possible due to the nature of the intervention. Table 6. Quality assessment of the included longitudinal studies according to the quality assessment tool for observational cohort and cross-sectional studies of the NIH a Adair, 2001 20 Blumfield, 2015 21 Bryant, 2015 22 Chatzi, 2017 23 Danielsen, 2013 24 Gale, 2006 25 Hrolfsdottir, 2017 26 Huh, 2005 27 Leary, 2005 28 Leary, 2013 29 Leermakers, 2017 30 May, 2014 31 Millard, 2013 32 Rerkasem, 2012 33 Rytter, 2013 34 van den Hil, 2013 35   Questionnaires used to measure maternal exposure were not always validated for the particular exposure of interest, or they were not validated for use in a pregnant population. This makes it questionable whether the observed (non)associations are due to measurement error, and also therefore associations may have gone undetected. Additionally, the majority of the included observational studies did not correct for key potential confounders such as breastfeeding, maternal smoking, maternal age, maternal pre-pregnancy BMI and birth weight. Therefore, residual confounding could have influenced the results and made results less reliable.
In view of the overwhelming amount of evidence, we a priori decided to study the programming effects of maternal diet and physical activity in pregnancy and offspring's cardiovascular health up to the age of 25 years. Nevertheless, there is evidence that the associations between maternal diet and offspring blood pressure persist into adulthood and may increase over time. 9,11,12

Strengths and limitations
The strength of this review is the systematic approach in finding and summarizing the available evidence on the association of maternal diet and physical activity with offspring cardiovascular health including both observational and experimental human studies. We were therefore able to give a comprehensive overview of the available literature. Owing to the heterogeneity in the assessment of maternal dietary intake and physical activity, the vascular outcomes, and the differences in offspring age, a metaanalysis was not possible. Associations of maternal lifestyle with offspring cardiovascular health were rarely studied using an RCT design. Therefore, we were not able to infer causality. There are, however, indications for causality from the UPBEAT trial, showing that a lifestyle intervention targeting maternal diet and physical activity during pregnancy had the potential to reduce infant adiposity. 48 Also, animal studies convincingly show that maternal lifestyle in pregnancy causes lasting changes to the offspring cardiovascular system. 13,49,50 Recommendations for further research In order to optimally use the information from studies on maternal lifestyle and offspring health, harmonization of valid exposure and outcome measurements and the development of core outcome sets would reduce research waste and speed up scientific progress in this field. 51,52 Since there is evidence from animal studies that maternal exercise can abolish the negative effects of maternal diet, 53 more research should focus on the programming effect of maternal physical activity in combination with maternal diet, which both should be examined validly and consistently across studies. Moreover, studying both maternal diet and physical activity at the same time could give more insight in the role of maternal energy balance on offspring cardiovascular health, with the ultimate goal to gain knowledge on how to help women to provide their child with the best start in life through an optimal lifestyle before and during pregnancy. In order to establish causality, experimental studies of lifestyle interventions before and during pregnancy should include follow-up of the offspring.

Conclusion
Currently there is a lack of consistent evidence to be able to draw robust conclusions on the association of women's dietary intake and physical activity before and during pregnancy with offspring's blood pressure and vascular health. We did find evidence for an association of high maternal carbohydrate intake with higher offspring blood pressure, and a negative linear association of maternal protein intake with offspring carotid IMT. We hypothesize that the macronutrient composition of the diet underlies these associations. However, no consistent findings for maternal fatty acid intake were found. There were too few studies to draw conclusions on energy intake, fibre intake, P:C ratio, specific foods, dietary patterns and maternal physical activity. Harmonization of valid exposure and outcome measurements, and the development of core outcome sets are needed to enable more robust conclusions.
Supplementary material. To view supplementary material for this article, please visit https://doi.org/10.1017/S204017441800082X