Neurodevelopment in 3–4 year old children exposed to maternal hyperglycemia or adiposity in utero

Neurodevelopment in 3–4 year old children exposed to maternal hyperglycemia or adiposity in utero

Early Human Development 125 (2018) 8–16 Contents lists available at ScienceDirect Early Human Development journal homepage:

NAN Sizes 0 Downloads 7 Views

Early Human Development 125 (2018) 8–16

Contents lists available at ScienceDirect

Early Human Development journal homepage:

Neurodevelopment in 3–4 year old children exposed to maternal hyperglycemia or adiposity in utero


John E. Krzeczkowskia, , Khrista Boylanb, Tye E. Arbucklec, Linda Doddsd, Gina Mucklee, William Fraserf, Lindsay A. Favottog, Ryan J. Van Lieshoutb, on behalf of The MIREC Study Group a

Neuroscience Graduate Program, McMaster University, Hamilton, Ontario, Canada Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, Ontario, Canada c Population Studies Division, Environmental Health Science and Research Bureau, Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, Ontario, Canada d Perinatal Epidemiology Research Unit, IWK Health Center, Dalhousie University, Halifax, Nova Scotia, Canada e Faculty of Social Sciences, Laval University, Montreal, Quebec, Canada f Centre de Recherche du Centre Hospitalier, Universitaire de Sherbrooke, Sherbrooke, Quebec, Canada g Department of Health Research, Evidence and Impact, McMaster University, Hamilton, Ontario, Canada b



Keywords: Obesity Hyperglycemia Prenatal programming Prenatal nutrition Child Wechsler Preschool and Primary Scale of Intelligence III

Background: Prenatal exposure to maternal metabolic complications has been linked to offspring neurodevelopmental problems. However, no studies investigating these links have examined the role of maternal prenatal diet. Aims: To determine if prenatal exposure to maternal adiposity or hyperglycemia is associated with neurodevelopmental problems in 3–4 year old children, and if links persist following adjustment for confounding variables, including prenatal diet. Method: 808 mother-child pairs from the Maternal-Infant Research on Environmental Chemicals-Child Development Plus cohort were used to examine associations between pre-pregnancy body mass index (BMI), hyperglycemia and offspring verbal, performance and full-scale IQ scores, as well as internalizing and externalizing problems. Associations were examined before and after adjustment for prenatal diet along with home environment, maternal depression, education and prenatal smoking. Semi-partial correlations were examined post-hoc to assess the impact of each confounder in the adjusted models. Results: In the unadjusted models, BMI and hyperglycemia predicted lower verbal and full-scale IQ. BMI was also linked to externalizing problems. However, associations were not significant after adjustment. In adjusted models, post-hoc analysis revealed that prenatal diet and home environment accounted for significant variance in verbal and full-scale IQ. The home environment and maternal depression accounted for significant variance in externalizing problems. Conclusion: In the adjusted models, maternal metabolic complications were not associated with offspring neurodevelopment. Even while adjusting for well-known risk factors for adverse offspring cognition (home environment, maternal depression), we show for the first time that maternal prenatal diet is an important confounder of the links between maternal metabolic complications and offspring cognition.

1. Introduction Nearly 40% of women are overweight or obese during pregnancy, and up to 16% of pregnancies are complicated by maternal hyperglycemia (i.e., impaired glucose tolerance or diabetes mellitus) [1–3].

Given the rapid rate of prenatal neurodevelopment, the fetal brain is particularly vulnerable to the pathological effects of prenatal adiposity and/or hyperglycemia [4,5]. Indeed, both pre-pregnancy adiposity and hyperglycemia have been linked to abnormalities of central nervous system development including defects of the neural tube [6–8]. As a

Abbreviations: BMI, body mass index; BASC-II, behavior assessment scale for children; FFQ, food frequency questionnaire; GDM, gestational diabetes mellitus; HEI2010, Healthy Eating Index (2010); IGT, impaired glucose tolerance; MIREC, Maternal-Infant Research on Environmental Chemicals; WPPSI-III, Wechsler Preschool and Primary Scale for Intelligence; 95% CI, 95% confidence interval ⁎ Corresponding author at: Neuroscience Graduate Program, McMaster University, 1280 Main St W, Hamilton, ON L8S 4L8, Canada. E-mail address: [email protected] (J.E. Krzeczkowski). Received 30 May 2018; Received in revised form 20 July 2018; Accepted 9 August 2018 0378-3782/ © 2018 Elsevier B.V. All rights reserved.

Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

2.2. Predictors

result, prenatal metabolic complications may have the potential to affect more complex neural systems underlying the development of higher mental functions such as cognition and emotion regulation. In keeping with the developmental origins of health and disease (DOHaD) hypothesis, prenatal exposure to excess maternal adiposity has been linked to poorer cognitive and language skills in children [9–11]. Higher rates of attention deficit hyperactivity disorder and externalizing problems are also seen in the children of overweight/ obese women [12,13]. Similar outcomes have been noted in children born to those with hyperglycemia during gestation including a doubling of the risk of both cognitive and behavioral problems [14,15]. Despite this accumulating evidence, the impact of potential unmeasured confounding variables on the associations between prenatal metabolic complications and offspring neurodevelopment remains a significant issue in the field. In an attempt to address these limitations, researchers have utilized twin and sibling studies, as well as paternal obesity as a negative control in order to strengthen the case for causal links between prenatal metabolic complications and offspring neurodevelopmental problems. However, while maternal adiposity had a stronger association with offspring cognitive and behavioral problems relative to paternal obesity in four studies [16–19], in three others it did not [20–22]. Prenatal metabolic complications were also linked to poorer neurodevelopment in one twin [23] and one sibling study [9], but two others using sibling designs reported null results [24,25]. Despite the advantages of sibling and twin studies in adjusting for genetic and other familial factors, current studies lack the ability to adjust for important environmental confounding variables, particularly for those that may be modifiable. Indeed, no studies have adjusted for overall prenatal diet quality when examining the associations between maternal metabolic complications and offspring neurodevelopment. Controlling for overall prenatal diet quality in pregnancy would significantly advance our understanding of these links and could help us move closer to identifying more specific targets for intervention. Given high rates of excess maternal adiposity and dysglycemia, and the importance of early cognition and behavior to health and success in life, understanding potential mechanisms underlying these links is of significant importance. Therefore, we utilized data from the panCanadian Maternal-Infant Research on Environmental Chemicals-Child Development Plus (MIREC-CD+) cohort to: a) determine if prenatal exposure to maternal overweight/obesity or hyperglycemia is associated with cognitive and behavioral problems in 3–4 year old offspring, and b) if these links persist following adjustment for confounding variables including prenatal diet quality, a previously unmeasured confounding variable of these links.

We examined the independent impact of pre-pregnancy adiposity and hyperglycemia separately on each of our outcomes since not all overweight women develop hyperglycemia, and not all women with hyperglycemia are overweight. 2.2.1. Pre-pregnancy adiposity-body mass index (BMI) Maternal pre-pregnancy BMI was calculated at the first trimester visit by dividing self-reported pre-pregnancy weight (kg) by height (m2), obtained by research staff. BMI was considered both a continuous and a dichotomous variable (normal: 18.5 < BMI < 25 vs. overweight/obese: BMI > 25) [27]. Underweight cases (BMI < 18.5, n = 20) were excluded since complications in pregnancy associated with being underweight are different from complications due to maternal overweight/obesity [28,29]. 2.2.2. Hyperglycemia (GDM/IGT) Of the women who participated in MIREC-CD Plus, 594 were assessed for hyperglycemia based on medical chart review. A dichotomous variable for hyperglycemia (gestational diabetes mellitus (GDM) or impaired glucose tolerance (IGT)) was created where GDM or IGT status = 1 and no GDM or IGT status = 0. GDM/IGT was determined based on the results of a glucose challenge test (GCT), or a 50 g or 100 g oral glucose tolerance test (OGTT). GDM was diagnosed if 1) fasting glucose levels after a 1 h 50 g GCT result exceeded 10.3 mmol/L, or 2) if two or more of the following cut-off values were exceeded following either a 50 g or 100 g OGTT: [50 g OGTT criteria: a) 1 h post > 10.6 mmol/L, b) 2 h > 8.9 mmol/L, or c) fasting plasma glucose level > 5.3 mmol/L; 100 g OGTT criteria: a) 1 h post > 10.6 mmol/L, b) 2 h > 9.2 mmol/L, c) 3 h > 8.0 mmol/L, or d) fasting plasma glucose level > 5.8 mmol/L]. IGT was diagnosed if 1 of these OGTT levels was exceeded. GDM and IGT were combined because adverse health outcomes in offspring have been observed following exposure to maternal glycemic levels below diagnostic criteria for gestational diabetes mellitus [30]. 2.3. Outcomes 2.3.1. Cognition: Wechsler Preschool and Primary Scale of Intelligence-III (WPPSI-III) Qualified research staff assessed child cognitive functioning using the WPPSI-III, a gold standard measure of intellectual function in children aged 2½ to 7½ years. Age-corrected verbal IQ, performance IQ and full-scale IQ scores served as our cognitive outcomes. The verbal IQ scale measures acquired knowledge, verbal comprehension and reasoning. The performance scale measures spatial abilities including visual and motor skills. Full-scale IQ score measures general intelligence. Software provided by the test publishers was used to calculate these three scales. Composite scales have a mean of 100, a standard deviation of 15, and a maximum possible score of 160.

2. Methods 2.1. Subjects MIREC is a longitudinal pregnancy cohort that recruited women from ten Canadian cities during their first trimester (< 14 weeks gestation) between 2008 and 2011 [26]. Eligibility criteria included fluency in English or French, maternal age > 18 years, plans to deliver locally, and an agreement to provide a cord blood sample. Women were excluded if there were abnormalities or malformations of the fetus in the current pregnancy, a history of medical complications such as heart disease, or a history of illicit drug use. Women provided sociodemographic information over three prenatal visits (one per trimester), and clinical information was obtained from medical charts following delivery. The current study utilized data from the MIREC-CD Plus cohort (a sub-study of the original MIREC cohort) designed to assess cognitive and behavioral outcomes in 808 3–4 year old children (mage = 40.7 months, SD = 3.73). Parents provided written consent prior to participation and study procedures were approved by research ethics boards at Health Canada and all recruitment sites.

2.3.2. Behavior: the Behavioral Assessment System for Children-Second Edition (BASC-II) Mothers completed the 134-item BASC-II which measures emotional and behavioral problems in 2 to 5 year old children. Individual items are scored on a 4-point Likert scale (0 = never to 3 = almost always) with higher scores indicating more problems. T-scores on the internalizing and externalizing composite scales were utilized. Each t-score has a mean of 50 and a standard deviation of 10. The externalizing scale is comprised of 22 items from the aggression and hyperactivity subscales (α = 0.93), and the internalizing problems scale consists of 37 items drawn from the anxiety, depression, and somatization subscales (α = 0.90). 9

Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

2.4.5. Prenatal smoking Self-reported maternal smoking behavior was captured at the third prenatal visit categorically: never smoked, former smoker, smoke currently/quit in pregnancy.

2.4. Confounding variables Confounding variables were selected based on previous longstanding evidence of associations between the confounder and both our predictors and outcomes.

2.5. Statistical analysis 2.4.1. Prenatal diet quality Prenatal diet was assessed using a 46-item food frequency questionnaire (FFQ) to assess the frequency and serving sizes of foods consumed in the past month. The FFQ was administered to mothers between 16 and 21 weeks of pregnancy and assessed intake of foods in 8 subgroups (vegetables, fruits, meat, poultry, fish and alternatives, dairy products, grain produces and other foods) [31]. The Healthy Eating Index-2010 (HEI) was used to convert FFQ data into a measure of total prenatal diet quality (see reference [32] for a step by step guide on the methods used to convert FFQ data to HEI scores). The HEI assesses diet quality according to national dietary guidelines and has been validated in pregnant women [33]. It is comprised of 8 “adequacy” components including total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and fatty acids. Higher scores indicate better diet quality. The HEI also contains 3 “moderation” components including refined grains, sodium, and empty calories (from fats and added sugars). The moderation components are reverse-scored, therefore higher scores indicate better diet quality. The sum of the adequacy and moderation components yields an HEI total score with a maximum score out of 100. The HEI total score was used in our statistical analyses. HEI total scores in our sample were consistent with scores in other samples of pregnant women [34,35].

2.5.1. Statistical models Descriptive statistics. Between-group comparisons (normal vs. overweight/obese BMI and normal glucose vs. hyperglycemia) were examined using independent samples t-tests and chi-squared tests for continuous and categorical variables, respectively. Unadjusted associations. A series of linear regression models were used to examine the links between each of our predictors (overweight/obesity and hyperglycemia) and outcome variables (verbal IQ, performance IQ, full-scale IQ, internalizing behavior, externalizing behavior). Adjusted associations. Multiple linear regression models were used to examine the same associations observed in our unadjusted models, but while adjusting for each of our confounding variables (BMI was also added as a confounder in the hyperglycemia model). The variance inflation factors in the adjusted models were below 10, therefore no evidence for multi-collinearly was detected. Adjusted associations—post-hoc assessment of confounding variables. Finally, a post-hoc analysis of the semi-partial correlations in the adjusted models was completed to determine which variables accounted for the most unique variance in our outcomes. For all results, unstandardized regression coefficients (B) values and 95% confidence intervals (CI) are presented, given that these B values display the change in the outcome per one unit change in the predictor. Statistical tests were 2-tailed with significance set at α = 0.05 and carried out in SPSS v23.

2.4.2. The home environment Trained research staff assessed the quality of the child's home environment using The Home Observation for Measurement of the Environment (HOME) scale [36]. The HOME scale is comprised of both an observation of the family's home environment and a semi-structured interview with the mother. Research staff used a binary (yes/no) scale to indicate the presence of stimulating factors in the home environment and adaptive maternal parenting behaviors, as well as an absence of harmful factors. Stimulating factors include the presence and variety of educational toys, games and books (e.g., child has: 3 or more puzzles, toys that help them learn animal names, toy or real musical instruments) and adaptive parenting behaviors (e.g., parent converses with child, reads stories, sets routines) [37]. The absence of harmful factors is also measured (e.g., house and surrounding environment are free of hazards, parent does not yell or scold child during the visit) [37]. This assessment was carried out when children were 3–4 years of age and in the home in which the child spends the most time. Administration of this scale takes approximately 1 h. Total HOME scale scores were calculated and higher scores (maximum of 55 points) indicate more adaptive parenting behaviors and a more stimulating home environment [37].

2.5.2. Sensitivity analysis The interaction between BMI and hyperglycemia was assessed using linear regression in order to determine whether having both hyperglycemia and overweight BMI was linked to greater offspring problems. 2.5.3. Missing data Both complete case and imputed data for all variables were examined using the fully conditional specification multiple imputation method in SPSS 23. We report only the results for subjects with complete predictor and outcomes data since there were no significant differences in results between complete and imputed data. A comparison of characteristics between women with and without BMI and hyperglycemia data is presented in Supplemental Table S1 (online). 3. Results

2.4.3. Maternal depression Symptoms of maternal depression were assessed when offspring were 3 years of age using the 10-item Center for Epidemiological Studies Depression (CES-D) scale. Mothers rated depressive symptoms experienced during the previous week using a four-point scale (1 = “rarely” to 4 = “all of the time”) with higher scores representing more depressive symptoms [38].

3.1. Descriptive statistics Mothers were on average 32.80 (SD = 4.81) years of age at enrollment and had an average pre-pregnancy BMI of 25.10 (SD = 5.63) (Table 1). At enrollment, 37% of women were overweight, and 8% had hyperglycemia. Of the women that had data for both BMI and hyperglycemia (n = 531), 26 (12.5%) of the overweight women, and 13 (4%) of the normal weight women were hyperglycemic. Fifteen (2.80%) women of the 531 women were diagnosed with GDM and 32 (6.03%) had IGT. The average child full-scale IQ in the sample was 106.90 (SD = 13.48). The mothers of children that participated in the MIREC-CD Plus sub-study were not different from mothers that participated only in the original MIREC cohort on each of our predictors and

2.4.4. Education Maternal education (high school or less vs. post-secondary education or greater) was used to as a measure of maternal cognitive function since education is considered the most efficient substitute for maternal IQ [39]. 10

Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

Table 1 Demographic information of the MIREC-CD Plus sample. Demographic variablesa

Body mass index

Maternal education (n, %) High school or less College educated/university degree Marital status (n, %) Married/common law Divorced/separated Single Maternal age (M. SD) Household income (n, %) < 50,000 50,001–100,000 > 100,000 Smoking (n, %) Never Former Current/quit in pregnancy Birth country (n, %) Canada Elsewhere HOME total (M, SD) Maternal depression (M, SD) Maternal diet quality (M, SD)b Breastfeeding (M, SD) Presence of hyperglycemia (n, %) GDM IGT Infant sex (n, %) Male Female Gestational age (M, SD) Birth weight (g) (M, SD)


Normal (n = 464)

Overweight/obese (n = 273)


Normal glucose (n = 547)

GDM/IGT (n = 47)


37 (8.0) 425 (92.0)

35 (12.8) 238 (87.2)


60 (11) 486 (89.0)

3 (6.4) 44 (93.6)


448 (96.5) 3 (0.6) 13 (2.8) 32.9 (4.8)

260 (95.2) 1 (0.4) 12 (4.4) 32.9 (4.8)


527 (96.4) 4 (0.7) 16 (2.9) 32.7 (4.8)

46 (97.8) 0 1 (2.2) 33.7 (4.5)


51 (11) 183 (39.4) 230 (49.6)

46 (16.8) 133 (48.7) 94 (34.4)

< 0.01

78 (14.3) 226 (41.3) 243 (44.4)

10 (21.3) 23(48.9) 14 (29.8)


293 (63.1) 131 (28.2) 40 (8.6)

185(67.8) 68 (24.9) 20 (7.3)


349 (63.8) 147 (26.9) 51 (9.3)

31 (66.0) 13 (27.7) 3 (6.4)


370 (79.7) 94 (20.3) 48.0 (3.7) 5.0 (3.9) 73.5 (7.4) 5.6 (2.0)

237 (86.8) 36 (13.2) 46.7 (4.9) 6.2 (4.5) 70.4(8.4) 5.3 (2.1)


445 (81.4) 102 (18.6) 47.4(4.3) 5.4 (4.2) 71.9 (8.0) 5.4 (2.2)

37 (78.7) 10 (21.3) 46.0 (4.42) 7.9(5.7) 69.7 (7.5) 4.9 (2.9)


4 (1.2) 9 (1.9)

17 (6.2) 9 (3.3)

< 0.01

220 (47.4) 244 (52.6) 39.1 (1.53) 3419.2 (489.2)

147 (53.8) 126 (46.2) 38.7 (1.84) 3486.8 (579.4)



0.001 0.001 < 0.01 0.14

< 0.01 0.11


0.09 0.01 0.07 0.27

15 (32) 32 (68) 279 (51) 268 (49) 38.9(1.7) 3438.9 (531.4)

24 (51.1) 23 (48.9) 38.2 (1.7) 3355.9 (608.1)

0.99 < 0.01 0.31

M = mean. SD = standard deviation. a Any discrepancies in n for demographic variables are due to missing data (e.g., two women in the normal BMI category did not have information on education). b Diet quality scores between 60 and 79.99 are considered ‘average’ in pregnant women [34].

(Table 2). Maternal BMI in the overweight/obese category was associated with significantly lower verbal BMIDi B = −3.41, 95% CI [−5.69;−1.12]) and full-scale IQ scores BMIDi B = −3.21, 95% CI [−5.21;−0.88]) compared to women with a normal pregnancy BMI. Maternal BMI was also a risk factor of increased levels of externalizing problems (BMIcont B = 0.10, 95% CI [0.001;0.21], BMIDi B = 1.87, 95% CI [0.72;1.72]), but not performance IQ or internalizing difficulties (Table 3). In the unadjusted hyperglycemia models, GDM/IGT was associated

covariates (however, MIREC-CD Plus mothers were more likely to have pursued education beyond high school).

3.2. Unadjusted associations In unadjusted statistical models containing BMI as a predictor, a per unit increase in maternal pre-pregnancy BMI was associated with significantly lower verbal (BMIcont B = −0.30, 95% CI [−0.48;−0.11]) and full-scale IQ scores (BMIcont B = −0.24, 95% CI [−0.44;−0.05])

Table 2 Unadjusted and fully adjusted associations between maternal metabolic risk factors and cognitive outcomes (WPPSI-III) in children at 3–4 years of age. Predictor β, 95% CI

Verbal IQ Unadjusted

BMIcontinuous BMIdichotomousc Hyperglycemiad

Performance IQ a

−0.30⁎⁎ (−0.48;−0.11) −3.41⁎⁎ (−5.69;−1.12) −5.17⁎⁎ (−8.67;−1.67)



−0.16 (−0.35;0.03) −1.70 (−4.01;0.61) −2.09 (−6.15;1.96)

Full-scale IQ





−0.11 (−0.32; 0.10) −1.79 (−4.4; 0.78) −0.70 (−4.47; 3.14)

−0.13 (−0.23;0.21) −0.43 (−3.11;2.24) 2.29 (−1.82;7.20)

−0.24⁎ (−0.44;−0.05) −3.21⁎⁎ (−5.21;−0.88) −3.48⁎ (−7.0; 0.00)

−0.12 (−0.31;0.07) −1.48 (−3.81;0.85) −0.34 (−3.63;4.32)


Unadjusted associations. Adjusted for gestational diet quality, home environment, maternal depression, prenatal smoking and maternal education, (and BMI in the hyperglycemia models). c Normal (18.5–24.9 kg/m2) vs. overweight or obese (≥25 kg/m2). d GDM and IGT vs. No-GDM. ⁎ p < 0.05. ⁎⁎ p < 0.01. b


Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

Table 3 Unadjusted and fully adjusted associations between maternal metabolic risk factors and behavioral outcomes (BASC-II) in children at 3–4 years of age. Predictor B, 95% CI

Externalizing Unadjusted

BMIcontinuous BMIdichotomousc Hyperglycemiad




0.10⁎ (0.001; 0.21) 1.87⁎⁎ (0.72; 3.02) −0.09 (−1.90; 1.72)


0.01 (−0.11; 0.13) 0.54 (−0.90; 1.98) −1.03 (−3.56; 1.50)



−0.02 (−0.12; 0.09) 0.43 (−0.79; 1.66) 0.27 (−1.67; 2.21)

−0.04 (−0.16; 0.08) 0.08 (−1.40; 1.56) −0.19 (−2.73;2.30)


Unadjusted associations. Adjusted for gestational diet quality, home environment, maternal depression, maternal education, prenatal smoking (and BMI in the hyperglycemia models). c Normal (18.5–24.9 kg/m2) vs. overweight or obese (≥25 kg/m2). d GDM and IGT vs No-GDM. ⁎ p < 0.05. ⁎⁎ p < 0.01. b

with reduced verbal (B = −5.17, 95% CI [−8.67;−1.67]) and fullscale IQ scores (B = −3.48, 95% CI [−7.00; −0.01]), but not performance IQ, or behavior problems (Tables 2 and 3). A sensitivity analysis examining the interaction between maternal BMI and hyperglycemia was tested but was not statistically significant. Finally, there were no significant differences in our results when stratifying by offspring sex.

Following adjustment for confounding variables, no significant associations persisted between our predictors and outcomes.

and full-scale IQ (rpart = 0.38 p < 0.01). Maternal prenatal diet also accounted for variance in verbal (rpart = 0.17, p < 0.01) and full-scale IQ (rpart = 0.11, p < 0.05) (Table 5). Although not selected a-priori as confounding variables, maternal income, birth country and infant gestational age differed significantly between BMI groups (Table 1). Therefore, we ran additional analyses with these variables added as confounders to our adjusted models and results were unchanged (diet and home environment still accounted for the most unique variance in verbal and full scale IQ and home environment and maternal depression accounted for unique variance in offspring externalizing problems).

3.4. Adjusted associations — post-hoc assessment of confounding variables

4. Discussion

Semi-partial correlations (rpart) were used to examine the variables that remained statistically significant in the fully adjusted models. Significant variance in verbal IQ was accounted for by the home environment (rpart = 0.28, p < 0.001) and prenatal diet (rpart = 0.12, p < 0.01). A significant amount of variance in full-scale IQ was also accounted for by the home environment (rpart = 0.31, p < 0.01) and prenatal diet (rpart = 0.08, p < 0.05). The home environment (rpart = −0.20, p < 0.01) and symptoms of maternal depression (rpart = 0.18, p < 0.01) accounted for variance in externalizing problems (Table 4). In the fully adjusted hyperglycemia model, the home environment accounted for significant variance in verbal IQ (rpart = 0.34, p < 0.01)

In a large sample of children, prenatal exposure to maternal adiposity and hyperglycemia were associated with lower verbal and fullscale IQ scores at 3–4 years of age. Exposure to maternal adiposity was also associated with more externalizing problems in offspring. However, after adjustment for confounding variables, these associations were no longer statistically significant. Associations between prenatal maternal metabolic problems and offspring cognitive outcomes appeared to be due to confounding variables, particularly home environment, maternal depression symptoms and overall maternal diet quality reported during pregnancy. Our primary objective was to examine the associations between maternal metabolic complications and offspring neurodevelopment,

3.3. Adjusted associations

Table 4 Confounding effect of variables in the fully adjusted BMI model. BMI model



Verbal IQ

HOMEb Depression symptoms Education Smoking Maternal diet qualityc

Full-scale IQ


B (95% CI)


B (95% CI)


B (95% CI)


0.91⁎⁎ (0.65;1.18) 0.06 (−0.21;0.33) 2.43 (−1.54;6.40) −0.70 (−2.46;1.06) 0.20⁎⁎ (0.06;0.35)


1.02⁎⁎ (0.75;1.29) 0.14 (−0.13;0.41) 5.04⁎⁎ (1.02;9.06) −0.49 (−2.27;1.29) 0.14⁎ (0.01;0.28)


−0.40⁎⁎ (−0.57;−0.24) 0.37⁎⁎ (1.98;0.56) 0.77 (−1.73;3.28) 0.69 (−0.39;1.79) 0.05 (−0.04;0.14)


0.02 0.05 −0.03 0.12


0.04 0.10 −0.02 0.08

Semi-partial correlation: the amount of unique variance each variable accounts for in the fully adjusted models. Home Observation for Measurement of the Environment Scale. c Healthy Eating Index 2010. ⁎ p < 0.05. ⁎⁎ p < 0.01. b


0.18 0.03 0.05 0.05

Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

Table 5 Confounding effect of variables in the fully adjusted hyperglycemia model. Hyperglycemia model



Verbal IQ

BMIa HOMEb Depression symptoms Education Smoking Maternal diet quality

a b c ⁎


Full-scale IQ


B (95% CI)


B (95% CI)


B (95% CI)


−0.001 (−0.22;0.22) 1.10⁎⁎ (0.79;1.4) 0.09 (−0.23;0.4) 0.33 (−4.37;5.02) −0.96 (−3.0;1.08) 0.28⁎⁎ (0.12;0.44)


−0.004 (−0.22;0.21) 1.2⁎⁎ (0.92;1.54) 0.19 (−0.12;0.49) 4.07 (−0.53;8.68) 0.44 (−2.4;1.55) 0.18⁎ (0.03;0.34)


0.05 (−0.91;0.18) −0.46⁎⁎ (−0.65;−0.26) 0.35⁎⁎ (0.15;0.55) 0.79 (−2.20;3.77) 0.73 (−0.53;1.99) 0.05 (−0.05;0.15)


0.34 0.03 0.01 −0.05 0.17

0.38 0.06 0.08 −0.02 0.11

−0.23 0.17 0.03 0.06 0.05

Continuous body mass index. Home Observation for Measurement of the Environment Scale. Healthy Eating Index 2010. p < 0.05.

and our unadjusted findings are consistent with the majority of studies that have observed these links. Numerous studies have observed poorer cognitive performance in children born to obese women [9–11,16,40–45], and in children exposed to prenatal hyperglycemia [14,46–48]. These problems are generally in the language domain [9,14,16,41,46]. Elevated levels of behavior problems, particularly externalizing problems, have also been observed in offspring as young as two [49–52]. However, since most of these studies were not designed to specifically examine links between maternal metabolic complications and offspring neurodevelopment, many fail to contain data on key confounding variables. Indeed, to our knowledge, no studies have considered the confounding effect of prenatal diet quality and just 5 have objectively assessed stimulation of the home environment [11,41–43,53]. A novel finding of this study is that total maternal diet quality during pregnancy remained significant in both our adjusted adiposity and hyperglycemia models for verbal and full-scale IQ, suggesting that overall prenatal diet quality is an important confounding variable in associations between excess adiposity/hyperglycemia in mothers and offspring cognitive problems. Maternal diet is a potentially modifiable factor known to play a significant role in fetal brain development [54,55]. Indeed, prenatal exposure to a nutrient-dense diet (high in fruits, vegetables, and seafood) has been linked to higher offspring IQ [56] and may be particularly important for the development of verbal IQ in children [57]. Conversely, consumption of a low nutrient diet, high in processed, energy-dense foods has been linked to reduced cognitive function in 8 year old offspring [57], as well as a 3-point decrease in IQ relative to children whose mothers consumed a healthier diet [56]. Experimental studies using non-human animal models have noted that prenatal exposure to high-fat, high-sugar diets results in worse performance on tasks assessing learning and memory, reduced levels of brain derived neurotropic factor in the hippocampus and altered expression of genes involved in hippocampal synaptic plasticity [58]. Observational studies have also noted associations between maternal diet and behavioral problems in offspring [59,60]; however, in the current study, diet was not significant in our models predicting child behavior. This could be due to the age at which behavior problems were assessed, maternal reporting bias, or a true lack of effect. Women of lower socioeconomic status (SES) are more likely to consume an unhealthy diet (e.g., [61]). Therefore, child cognitive development may be more of a reflection of the family SES rather than prenatal diet quality. However, by including both prenatal diet quality

and the assessment of the child's home environment together in each adjusted model, we observed that both variables accounted for significant unique variance in offspring cognitive function. Additionally, we did not find any evidence of multicollinearity between these two variables. Therefore, despite the greater amount of variance accounted for by the home environment relative to prenatal diet, our results indicate that these are disparate constructs and that prenatal diet also has an impact on offspring cognitive function. In each of the adjusted models, the home environment accounted for the most variance in offspring cognitive function and offspring behavior, and maternal depression accounted for significant unique variance in offspring behavior. It is well documented that the home environment [62,63] and maternal depression [64] play a significant role in child development, therefore, it is not surprising that these variables accounted for a significant amount of variance in our models. Despite the potential importance of these findings, this study must be examined in the context of the following limitations. First, since some MIREC centers only screened participants at high risk for GDM, of the 808 subjects participating in MIREC-CD Plus, 214 were missing hyperglycemia data which may partially explain why our prevalence of hyperglycemia lower than other studies. Second, women in our sample had higher levels of education and income relative to the Canadian population [65], which needs to be taken into account when generalizing to other populations. Additionally, a greater proportion of women of low income were missing information on BMI (Supplementary Table 1). Since it is well understood that higher BMI in pregnancy is linked to lower SES, this may have contributed to the slightly lower prevalence of overweight women observed in our sample relative to that seen in other studies. Third, we were unable to adjust for maternal IQ; however, we did measure maternal education and evidence suggests that HOME scale scores at least partially mediate associations between maternal and offspring IQ [62]. Fourth, while the FFQ we used covered a wide range of foods commonly consumed by Canadians, it was comprised of only 46 items. This may have limited the scope of our diet assessment; however, other studies calculated the HEI using an FFQ with a similar number of items [66] and previous research suggests that short FFQs are valid for assessing overall diet quality [67]. Future studies could extend upon these findings by using other diet quality indices (e.g., alternative health eating index for pregnancy). We also did not consider intake of multivitamins or other dietary supplements. Fifth, reports of maternal BMI and child behavior problems were made only by mothers and therefore could be affected by response bias. Sixth, despite only occurring in a small percentage of the population, potential 13

Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

Statement of financial support

prenatal confounding factors that lead to neonatal hospitalizations, such as failure to thrive, could be adjusted for in future studies. Additionally, we excluded women with a fetus with any known abnormalities or malformations and future studies could also consider these factors. Finally, we were unable to adjust for child diet; however, the home environment has been associated with access to healthy food and healthy eating behaviors in children [68]. Future studies should aim to extend follow-ups of offspring born to women that participated in prenatal healthy lifestyle randomized controlled trials (e.g., the Maternal Obesity Management Trial [69]) to elucidate potential causal associations between prenatal maternal diet and offspring cognition [70,71]. This is important since factors that optimize cognitive function may have a significant impact at the population level. Since a 1-point increase in the national average IQ is associated with a 0.11% increase in gross domestic product per person annually [72], optimizing cognitive development represents a tremendous opportunity to improve the health and success of the population.

The MIREC Research Platform is supported by the Chemicals Management Plan of Health Canada, the Canadian Institutes for Health Research (grant # MOP-81285) and the Ontario Ministry of the Environment. JEK is Funded by the Vanier Graduate Scholarship Program. Funder role The funders did not play a role in study design, collection, analysis and interpretation of data, writing of the report nor in the decision to submit this manuscript to Early Human Development. Competing interests None. Disclosure statement

5. Conclusion

None, all authors have approved the final article.

Numerous studies have reported on associations between prenatal maternal metabolic complications and offspring neurodevelopmental problems. However, none of these studies have examined the role prenatal maternal diet quality might play in these links. We show for the first time that prenatal diet quality is a significant confounder of these associations, even in models that also adjusted for traditional risk factors for adverse cognitive development (e.g., maternal depression and the home environment). Due to the preventative potential of early life interventions, assessing the mechanisms that underlie links between maternal prenatal metabolic complications and offspring neurodevelopment is of particular interest; therefore, future studies must adjust for prenatal diet quality in order to fully elucidate the mechanisms underlying these links. Supplementary data to this article can be found online at https://

Acknowledgements The MIREC Research Platform is supported by the Chemicals Management Plan of Health Canada, the Canadian Institutes for Health Research (grant # MOP-81285) and the Ontario Ministry of the Environment Vanier Graduate Scholarship. We would like to acknowledge the MIREC Study Group as well as the MIREC study participants and staff for their dedication. Conflict of interest statement Authors have no conflicts of interest to disclose. References [1] I. Guelinckx, R. Devlieger, K. Beckers, G. Vansant, Maternal obesity: pregnancy complications, gestational weight gain and nutrition, Obes. Rev. 9 (2008) 140–150. [2] E. Ashwal, M. Hod, Gestational diabetes mellitus: where are we now? Clin. Chim. Acta 451 (2015) 14–20, [3] IDF, Diabetes Atlas, 7th edition, (2015), i03. [4] W. Gao, W. Lin, K. Grewen, J.H. Gilmore, Functional connectivity of the infant human brain: plastic and modifiable, Neuroscientist (2016), 1177/1073858416635986. [5] G.Z. Tau, B.S. Peterson, Normal development of brain circuits, Neuropsychopharmacology 35 (2010) 147–168, 2009.115. [6] A. Bharadwaz, G. Madhab, Review article burden of maternal obesity on congenital anomalies : implications and future trend, Ann. Women Child Heal. 2 (2016) 1–8. [7] A. Ornoy, E.A. Reece, G. Pavlinkova, C. Kappen, R.K. Miller, Effect of maternal diabetes on the embryo, fetus, and children: congenital anomalies, genetic and epigenetic changes and developmental outcomes, Birth Defects Res. C. Embryo Today 105 (2015) 53–72, [8] E. Villamor, K. Tedroff, M. Peterson, S. Johansson, M. Neovius, G. Petersson, S. Cnattingius, Association between maternal body mass index in early pregnancy and incidence of cerebral palsy, Jama 317 (2017) 925–936, 1001/jama.2017.0945. [9] L. Huang, X. Yu, S. Keim, L. Li, L. Zhang, J. Zhang, Maternal prepregnancy obesity and child neurodevelopment in the collaborative perinatal project, Int. J. Epidemiol. 43 (2014) 783–792, [10] E. Basatemur, J. Gardiner, C. Williams, E. Melhuish, J. Barnes, A. Sutcliffe, Maternal prepregnancy BMI and child cognition: a longitudinal cohort study, Pediatrics 131 (2013) 56–63, [11] W.Y. Craig, G.E. Palomaki, L.M. Neveux, Maternal body mass index during pregnancy and offspring neurocognitive development, Obstet. Med. (2013) 20–25. [12] A. Rodriguez, J. Miettunen, T.B. Henriksen, J. Olsen, C. Obel, A. Taanila, H. Ebeling, K.M. Linnet, I. Moilanen, M.-R. Järvelin, Maternal adiposity prior to pregnancy is associated with ADHD symptoms in offspring: evidence from three prospective pregnancy cohorts, Int. J. Obes. (Lond.) 32 (2008) 550–557, https:// [13] A. Rodriguez, Maternal pre-pregnancy obesity and risk for inattention and negative emotionality in children, J. Child Psychol. Psychiatry 51 (2010) 134–143, https://

Contributors statement Mr. John Krzeczkowski conceptualized the idea for this study and its design, analyzed data and interpreted findings, wrote the first draft of the manuscript and incorporated subsequent edits and revisions and approved the final manuscript as submitted. Dr. Boylan aided in conceptualizing the idea for the study, supported data extraction procedures, provided feedback and interpretation of the data. Dr. Boylan reviewed and critically evaluated the intellectual content of the manuscript and approved the final manuscript as submitted. Dr. Arbuckle contributed to design of the MIREC and cohort study, provided support in the interpretation of findings and data analysis, reviewed and revised the manuscript, and approved the final manuscript as submitted. Drs. Dodds, Muckel and Fraiser contributed the design of the MIREC cohort study and the selection of study instruments. Each contributed to data interpretation, provided feedback and critically evaluated subsequent drafts of the manuscript, and approved the final manuscript as submitted. Ms. Lindsay Favotto Analyzed the data and interpreted the findings, provided support for the intellectual content of the article and supported edits and subsequent drafts of the manuscript, and approved the final draft of the manuscript for submission. Dr. Van Lieshout aided in the conceptualization of the idea for the project, the design of the study, data analysis and interpretation, critically reviewed and revised the manuscript and approved the final manuscript form submission. 14

Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.

jand.2012.12.016. [34] M. Tsigga, V. Filis, K. Hatzopoulou, C. Kotzamanidis, M.G. Grammatikopoulou, Healthy Eating Index during pregnancy according to pre-gravid and gravid weight status, Public Health Nutr. 14 (2011) 290–296, S1368980010001989. [35] M.E. Pick, M. Edwards, D. Moreau, E.A. Ryan, Assessment of diet quality in pregnant women using the Healthy Eating Index, J. Am. Diet. Assoc. 105 (2005) 240–246, [36] B.M. Caldwell, R.H. Bradley, HOME Inventory and Administration Manual, University of Arkansas for Medical Sciences and University of Arkansas at Little Rock, 2001. [37] V. Totsika, K. Sylva, The home observation for measurement of the environment revisited, Child Adolesc. Ment. Health 9 (2009) 25–35, 1475-357X.2003.00073.x. [38] K.L. Smarr, A.L. Keefer, Measures of depression and depressive symptoms: Beck Depression Inventory-II (BDI-II), Center for Epidemiologic Studies Depression Scale (CES-D), Geriatric Depression Scale (GDS), Hospital Anxiety and Depression Scale (HADS), and Patient Health Questionnaire, Arthritis Care Res. 63 (2011) 454–466, [39] J.D. Matarazzo, D.O. Herman, Relationship of education and IQ in the WAIS-R standardization sample, J. Consult. Clin. Psychol. 52 (1984) 631–634, https://doi. org/10.1037/0022-006X.52.4.631. [40] U. Heikura, A. Taanila, A.L. Hartikainen, P. Olsen, S.L. Linna, L. Von Wendt, M.R. Jorvelin, Variations in prenatal sociodemographic factors associated with intellectual disability: a study of the 20-year interval between two birth cohorts in northern Finland, Am. J. Epidemiol. 167 (2008) 169–177, 1093/aje/kwm291. [41] R. Tanda, P.J. Salsberry, P.B. Reagan, M. Fang, The impact of prepregnancy obesity on children's cognitive test scores, Matern. Child Heal. J. 17 (2013) 222–229, [42] S.J. Pugh, G.A. Richardson, J.A. Hutcheon, K.P. Himes, M.M. Brooks, N.L. Day, L.M. Bodnar, Maternal obesity and excessive gestational weight gain are associated with components of child cognition, J. Nutr. (2015) 1–8, jn.115.215525. [43] Y.H. Neggers, R.L. Goldenberg, S.L. Ramey, S.P. Cliver, Maternal prepregnancy body mass index and psychomotor development in children, Acta Obstet. Gynecol. Scand. 82 (2003) 235–240, [44] K.O. Duffany, K.H. McVeigh, T.S. Kershaw, H.S. Lipkind, J.R. Ickovics, Maternal obesity: risks for developmental delays in early childhood, Matern. Child Health J. 20 (2016) 219–230, [45] S.N. Hinkle, L.A. Schieve, A.D. Stein, D.W. Swan, U. Ramakrishnan, A.J. Sharma, Associations between maternal prepregnancy body mass index and child neurodevelopment at 2 years of age, Int. J. Obes. 36 (2012) 1312–1319, 1038/ijo.2012.143. [46] A. Fraser, S.M. Nelson, C. MacDonald-Wallis, D.A. Lawlor, Associations of existing diabetes, gestational diabetes, and glycosuria with offspring iq and educational attainment: the Avon longitudinal study of parents and children, Exp. Diabetes Res. 2012 (2012) 1–7, [47] A. Ornoy, A. Wolf, N. Ratzon, C. Greenbaum, M. Dulitzky, Neurodevelopmental outcome at early school age of children born to mothers with gestational diabetes, Arch. Dis. Child. Fetal Neonatal Ed. 81 (1999) F10–F14, fn.81.1.F10. [48] M.C. Robles, C. Campoy, L.G. Fernandez, J.M. Lopez-Pedrosa, R. Rueda, M.J. Martin, Maternal diabetes and cognitive performance in the offspring: a systematic review and meta-analysis, PLoS One 10 (2015) 1–16, 1371/journal.pone.0142583. [49] R.J. Van Lieshout, Role of maternal adiposity prior to and during pregnancy in cognitive and psychiatric problems in offspring, Nutr. Rev. 71 (2013) 95–101, [50] P. Krakowiak, C.K. Walker, A.A. Bremer, A.S. Baker, S. Ozonoff, R.L. Hansen, I. Hertz-Picciotto, Maternal metabolic conditions and risk for autism and other neurodevelopmental disorders, Pediatrics 129 (2012) e1121–e1128, https://doi. org/10.1542/peds.2011-2583. [51] M. Li, M.D. Fallin, A. Riley, R. Landa, S.O. Walker, The association of maternal obesity and diabetes with autism and other developmental disabilities, Pediatrics 137 (2016) e20152206. [52] K.D. Getz, M.T. Anderka, M.M. Werler, S.S. Jick, Maternal pre-pregnancy body mass index and autism spectrum disorder among offspring: a population-based casecontrol study, Paediatr. Perinat. Epidemiol. (2016), 12306. [53] S. Pugh, J. Hutcheon, G. Richardson, M. Brooks, K. Himes, N. Day, L. Bodnar, Gestational weight gain, prepregnancy body mass index and offspring attentiondeficit hyperactivity disorder symptoms and behaviour at age 10, BJOG (2016) 1–10, [54] B.M.Y. Leung, K.P. Wiens, B.J. Kaplan, Does prenatal micronutrient supplementation improve children's mental development? A systematic review, BMC Pregnancy Childbirth 11 (2011) 12, [55] A. Nyaradi, J. Li, S. Hickling, J. Foster, W.H. Oddy, The role of nutrition in children's neurocognitive development, from pregnancy through childhood, Front. Hum. Neurosci. 7 (2013) 97, [56] A.A. Freitas-Vilela, R.M. Pearson, P. Emmett, J. Heron, A.D.A.C. Smith, A. Emond, J.R. Hibbeln, M.B.T. Castro, G. Kac, Maternal dietary patterns during pregnancy and intelligence quotients in the offspring at 8 years of age: findings from the ALSPAC cohort, Matern. Child Nutr. (2017) 1–11, 12431. [57] E.D. Barker, N. Kirkham, J. Ng, S.K.G. Jensen, Prenatal maternal depression [14] G. Dionne, M. Boivin, J.R. Séguin, D. Pérusse, R.E. Tremblay, Gestational diabetes hinders language development in offspring, Pediatrics 122 (2008) e1073–e1079, [15] R. Perna, A.R. Loughan, J. Le, K. Tyson, W.R. Hospital, Gestational diabetes: longterm central nervous system developmental and cognitive sequelae, HAPC 4 (2015) 217–220, [16] M. Casas, L. Chatzi, A.E. Carsin, P. Amiano, M. Guxens, M. Kogevinas, K. Koutra, N. Lertxundi, M. Murcia, M. Rebagliato, I. Riaño, C.L. Rodríguez-Bernal, T. Roumeliotaki, J. Sunyer, M. Mendez, M. Vrijheid, Maternal pre-pregnancy overweight and obesity, and child neuropsychological development: two Southern European birth cohort studies, Int. J. Epidemiol. 42 (2013) 506–517, https://doi. org/10.1093/ije/dyt002. [17] M. Robinson, S.R. Zubrick, C.E. Pennell, R.J. Van Lieshout, P. Jacoby, L.J. Beilin, T.A. Mori, F.J. Stanley, J.P. Newnham, W.H. Oddy, Pre-pregnancy maternal overweight and obesity increase the risk for affective disorders in offspring, J. Dev. Orig. Health Dis. 4 (2012) 1–7, [18] V. Daraki, T. Roumeliotaki, K. Koutra, V. Georgiou, M. Kampouri, A. Kyriklaki, M. Vafeiadi, S. Papavasiliou, Effect of parental obesity and gestational diabetes on child neuropsychological and behavioral development at 4 years of age: the Rhea mother–child cohort, Crete, Greece, Eur. Child Adolesc. Psychiatry (2017), https:// [19] S.H. Mikkelsen, L. Hohwü, J. Olsen, B.H. Bech, Z. Liew, Original Contribution Parental Body Mass Index and Behavioral Problems in Their Offspring: A Danish National Birth Cohort Study, (2017), pp. 1–10, kwx063. [20] M. Brion, M. Zeegers, V. Jaddoe, F. Verhulst, D.A. Lawlor, G.D. Smith, Maternal prepregnancy overweight and child cognition and behavior: exploring intrauterine effects in two pregnancy cohorts, Pediatrics 127 (2011) 1–18, 1542/peds.2010-0651.Maternal. [21] M. Bliddal, J. Olsen, H. Støvring, H.L.F. Eriksen, U.S. Kesmodel, T.I.A. Sørensen, E.A. Nøhr, Maternal pre-pregnancy BMI and intelligence quotient (IQ) in 5-year-old children: a cohort based study, PLoS One 9 (2014) 3–9, journal.pone.0094498. [22] P. Surén, N. Gunnes, C. Roth, M. Bresnahan, M. Hornig, D. Hirtz, K.K. Lie, W.I. Lipkin, P. Magnus, T. Reichborn-Kjennerud, S. Schjølberg, E. Susser, A.S. Oyen, G.D. Smith, C. Stoltenberg, Parental obesity and risk of autism spectrum disorder, Pediatrics 133 (2014) e1128–e1138, [23] E.E. Antoniou, T. Fowler, K. Reed, T.R. Southwood, J.P. McCleery, M.P. Zeegers, Maternal pre-pregnancy weight and externalising behaviour problems in preschool children: a UK-based twin study, BMJ Open 4 (2014), bmjopen-2014-005974 no pagination. [24] A. Fraser, C. Almqvist, H. Larsson, N. Långström, D.A. Lawlor, Maternal diabetes in pregnancy and offspring cognitive ability: sibling study with 723,775 men from 579,857 families, Diabetologia 57 (2014) 102–109, s00125-013-3065-z. [25] Q. Chen, A. Sjolander, N. Langstrom, A. Rodriguez, E. Serlachius, B.M. D'Onofrio, P. Lichtenstein, H. Larsson, Maternal pre-pregnancy body mass index and offspring attention deficit hyperactivity disorder: a population-based cohort study using a sibling-comparison design, Int. J. Epidemiol. 43 (2014) 83–90, 1093/ije/dyt152. [26] T.E. Arbuckle, W.D. Fraser, M. Fisher, K. Davis, C.L. Liang, N. Lupien, S. Bastien, M.P. Velez, P. Von Dadelszen, D.G. Hemmings, J. Wang, M. Helewa, S. Taback, M. Sermer, W. Foster, G. Ross, P. Fredette, G. Smith, M. Walker, R. Shear, L. Dodds, A.S. Ettinger, J.P. Weber, M. D'Amour, M. Legrand, P. Kumarathasan, R. Vincent, Z.C. Luo, R.W. Platt, G. Mitchell, N. Hidiroglou, K. Cockell, M. Villeneuve, D.F.K. Rawn, R. Dabeka, X.L. Cao, A. Becalski, N. Ratnayake, G. Bondy, X. Jin, Z. Wang, S. Tittlemier, P. Julien, D. Avard, H. Weiler, A. Leblanc, G. Muckle, M. Boivin, G. Dionne, P. Ayotte, B. Lanphear, J.R. Séguin, D. Saint-Amour, É. Dewailly, P. Monnier, G. Koren, E. Ouellet, Cohort profile: the maternal-infant research on environmental chemicals research platform, Paediatr. Perinat. Epidemiol. 27 (2013) 415–425, [27] World Health Organization, WHO: Global Database on Body Mass Index, World Heal. Organ., 2012 doi: papers3://publication/uuid/DF45B6E6-94B7-4B8A-97753A1313BA45EC. [28] R.J. Van Lieshout, M. Robinson, M.H. Boyle, Maternal pre-pregnancy body mass index and internalizing and externalizing problems in offspring, Can. J. Psychiatr. 58 (2013) 151–159. [29] C. Buss, S. Entringer, E.P. Davis, C.J. Hobel, J.M. Swanson, P.D. Wadhwa, C.A. Sandman, Impaired executive function mediates the association between maternal pre-pregnancy body mass index and child ADHD symptoms, PLoS One 7 (2012), [30] B. Metzger, L. Lowe, A. Dyer, E. Trimble, U. Chaovarindr, D. Coustan, Hyperglycemia and adverse pregnancy outcomes, N. Engl. J. Med. 358 (2008) 1991–2002. [31] A. Morisset, H.A. Weiler, L. Dubois, J. Ashley-Martin, G.D. Shapiro, L. Dodds, I. Massarelli, M. Vigneault, T.E. Arbuckle, W.D. Fraser, Rankings of iron, vitamin D, and calcium intakes in relation to maternal characteristics of pregnant Canadian women, Appl. Physiol. Nutr. Metab. 757 (2016) 749–757. [32] Overview & Background of The Healthy Eating Index, NIH National Cancer Institute. Division of Cancer Control and Population Sciences, 2015 (accessed 10 January 2018), [33] P.M. Guenther, K.O. Casavale, J. Reedy, S.I. Kirkpatrick, H.A.B. Hiza, K.J. Kuczynski, L.L. Kahle, S.M. Krebs-Smith, Update of the Healthy Eating Index: HEI-2010, J. Acad. Nutr. Diet. 113 (2013) 569–580,


Early Human Development 125 (2018) 8–16

J.E. Krzeczkowski et al.








0080-1. [65] Statistics Canada, Education and Labour, 2011/as-sa/99-014-x/99-014-x2011003_2-eng.cfm, (2011) accessed on July 18th 2018. [66] P. Clarys, T. Deliens, I. Huybrechts, P. Deriemaeker, B. Vanaelst, W. De Keyzer, M. Hebbelinck, P. Mullie, Comparison of nutritional quality of the vegan, vegetarian, semi-vegetarian, pesco-vegetarian and omnivorous diet, Nutrients 6 (2014) 1318–1332, [67] C.L. Cleghorn, R.A. Harrison, J.K. Ransley, S. Wilkinson, J. Thomas, J.E. Cade, Can a dietary quality score derived from a short-form FFQ assess dietary quality in UK adult population surveys? Public Health Nutr. 19 (2016) 2915–2923, https://doi. org/10.1017/S1368980016001099. [68] R.R. Rosenkranz, D.A. Dzewaltowski, Model of the home food environment pertaining to childhood obesity, Nutr. Rev. 66 (2008) 123–140, 1111/j.1753-4887.2008.00017.x. [69] K.B. Adamo, Z.M. Ferraro, G. Goldfield, E. Keely, D. Stacey, S. Hadjiyannakis, S. Jean-Philippe, M. Walker, N.J. Barrowman, The Maternal Obesity Management (MOM) Trial Protocol: a lifestyle intervention during pregnancy to minimize downstream obesity, Contemp. Clin. Trials 35 (2013) 87–96, 1016/j.cct.2013.02.010. [70] R.J. Van Lieshout, J.E. Krzeczkowski, Just DO(HaD) It! Testing the clinical potential of the DOHaD hypothesis to prevent mental disorders using experimental study designs, J. Dev. Orig. Health Dis. (2016) 1–9, S2040174416000441. [71] R.C. Richmond, A. Al-Amin, G. Davey Smith, C.L. Relton, Approaches for drawing causal inferences from epidemiological birth cohorts: a review, Early Hum. Dev. 90 (2014) 769–780, [72] G. Jones, W.J. Schneider, Intelligence, human capital, and economic growth: a Bayesian averaging of classical estimates (Bace) approach, J. Econ. Growth 11.1 (2006) 71–93.

symptoms and nutrition, and child cognitive function, Br. J. Psychiatry 203 (2013) 417–421, K.C. Page, E.K. Jones, E.K. Anday, Maternal and postweaning high-fat diets disturb hippocampal gene expression, learning, and memory function, Am. J. Phys. Regul. Integr. Comp. Physiol. 306 (2014) R527–R537, 00319.2013. F.N. Jacka, E. Ystrom, A.L. Brantsaeter, E. Karevold, C. Roth, M. Haugen, H.M. Meltzer, S. Schjolberg, M. Berk, Maternal and early postnatal nutrition and mental health of offspring by age 5 years: a prospective cohort study, J. Am. Acad. Child Adolesc. Psychiatry 52 (2013) 1038–1047, 2013.07.002. J. Steenweg-De Graaff, H. Tiemeier, R.P.M. Steegers-Theunissen, A. Hofman, V.W.V. Jaddoe, F.C. Verhulst, S.J. Roza, Maternal dietary patterns during pregnancy and child internalising and externalising problems. The Generation R Study, Clin. Nutr. 33 (2014) 115–121, S.K. Ng, C.M. Cameron, A.P. Hills, R.J. McClure, P.A. Scuffham, Socioeconomic disparities in prepregnancy BMI and impact on maternal and neonatal outcomes and postpartum weight retention: the EFHL longitudinal birth cohort study, BMC Pregnancy Childbirth 14 (2014) 1–15, L. Ronfami, L. Vecci-Brumatti, M. Mariuz, V. Tognin, M. Bin, V. Ferluga, A. Knowles, M. Montico, F. Barbone, The complex interaction between home environment, socioeconomic status, maternal IQ and early child neurocognitive development: a multivariate analysis of data collected in a newborn cohort study, PLoS One (2015) 1–13, S. Tong, P. Baghurst, G. Vimpani, A. McMichael, Socioeconomic position, maternal IQ, home environment, and cognitive development, J. Pediatr. 151 (2007), https:// S.H. Goodman, M.H. Rouse, A.M. Connell, M.R. Broth, C.M. Hall, D. Heyward, Maternal depression and child psychopathology: a meta-analytic review, Clin. Child. Fam. Psychol. Rev. 14 (2011) 1–27,