Abstract
Objective
Maternal psychological distress in pregnancy has been associated with both breastfeeding duration and child weight at 24 months; however, the potential that breastfeeding duration partially mediates the risk of maternal mental health problems during pregnancy on child weight classification has not been examined. The current study investigated this proposed relationship.
Methods
Data was taken from the All Our Families (AOF) cohort, an ongoing prospective pregnancy cohort located in Calgary, Canada. Psychological distress, defined as clinically significant symptoms of anxiety and depression, was assessed via self-report and measured between 34 and 36 weeks of gestation. Breastfeeding duration was assessed in the postpartum by self-report. Child overweight classification was defined as a weight-for-length/height z-score at or above the 97th percentile as per World Health Organization’s child growth guidelines.
Results
In this sample of 1582 mother−child pairs, there was no direct relationship between psychological distress and child overweight status. Both anxiety (B = − 5.40, p = 0.001) and depression (B = − 6.54, p = 0.008) were associated with decreased weeks breastfeeding. Breastfeeding duration mediated the association between maternal prenatal psychological distress and child overweight status at 24 months, for both anxiety (B(SE) = 0.10(0.05), CI 0.03–0.21) and depression (B(SE) = 0.11(0.07), CI 0.01–0.27). Covariates included maternal age, education, ethnicity, income, pre-pregnancy BMI, gestational weight gain, and infant birth weight.
Conclusions
The results of this longitudinal cohort analysis support an indirect relationship between maternal psychological distress in pregnancy and the childhood overweight/obesity at 24 months old, mediated through breastfeeding duration.
Keywords: Breastfeeding, Pregnancy, Depression, Anxiety, Child weight
Résumé
Objectif
La détresse psychologique maternelle pendant la grossesse a été associée à l’allaitement et au poids de l’enfant à 24 mois; cependant, la possibilité que la durée de l’allaitement puisse être un médiateur de la relation entre la santé mentale maternelle pendant la grossesse et la classification du poids de l’enfant n’a pas encore été examinée. La présente étude a examiné cette relation proposée.
Méthodes
Les données proviennent de l’étude All Our Families (AOF), une cohorte de grossesse prospective en cours à Calgary, au Canada. La détresse psychologique, définie comme la présence de symptômes d’anxiété et de dépression significatifs sur le plan clinique, a été évaluée à l’aide de questionnaires entre la 34ème et la 36ème semaines de gestation. La durée de l’allaitement a été mesurée pendant le postpartum en s’appuyant sur l’auto-déclaration des mères. La classification du surpoids chez l’enfant a été définie comme un rapport poids-pour-la-longueur/score z pour la hauteur qui est égal ou supérieur au 97e centile, selon les recommandations sur la croissance de l’enfant de l’Organisation mondiale de la Santé.
Résultats
Parmi les 1582 dyades mère−enfant inclus dans l’étude, il n’y avait pas de relation directe entre la détresse psychologique maternelle et le surpoids chez l’enfant. L’anxiété (B = − 5,40, p = 0,001) et la dépression (B = − 6,54, p = 0,008) maternelle étaient associés à une diminution du nombre de semaines d’allaitement. La durée de l’allaitement maternel avait un effet médiateur sur l’association entre la détresse psychologique maternelle prénatale et le surpoids de l’enfant à 24 mois, pour l’anxiété (B(SE) = 0,10(0,05), IC 0,03 – 0,21) et la dépression (B(SE) = 0,11(0,07), IC 0,01 – 0,27). Les covariables comprenaient l’âge maternel, l’éducation, l’origine ethnique, le revenu, l’IMC avant la grossesse, le gain de poids gestationnel et le poids de l’enfant à la naissance.
Conclusions
Les résultats de cette étude longitudinale soutiennent une relation indirecte entre la détresse psychologique maternelle pendant la grossesse et le surpoids/l’obésité de l’enfant à 24 mois, médiée par la durée de l’allaitement.
Mots-clés: Allaitement, Grossesse, Dépression, Anxiété, Poids de l’enfant
Introduction
According to the World Health Organization (WHO), rates of childhood and adolescent overweight and obesity in Canada rose from 23.3% in 1978 to 34.7% in 2004 (Shields and Tremblay 2010). Rising rate of overweight and obese status in childhood is concerning because it is associated with a significantly higher risk of becoming overweight as an adult, bringing with it significant health comorbidities and financial costs (Simmonds et al. 2016). Preventing overweight and obesity status in childhood is a critical public health priority.
Maternal depression and anxiety experienced during pregnancy and in the postpartum have each been associated with increased risk of their children being classified as overweight across the lifespan (Braungart-Rieker et al. 2016; Surkan et al. 2008). The mechanisms linking maternal mental health problems to increased risk of childhood overweight/obesity are multifaceted and likely include altered metabolic responses originating during gestation in addition to parenting and feeding practices in the postpartum period (Braungart-Rieker et al. 2016).
Breastfeeding behaviour is a little examined potential mechanism that may help to explain associations between maternal psychological distress and childhood overweight/obesity. Higher levels of maternal psychological distress (e.g., anxiety and/or depression) have been associated with a lower likelihood of breastfeeding initiation and shorter overall duration of breastfeeding (Dias and Figueiredo 2015; Fallon et al. 2018). Breastfeeding is also a well-established contributor to infant growth trajectories (Carling et al. 2015). Infants who are breastfed for a longer duration are less likely to be classified as obese at age 2 compared with those breastfed for shorter durations (Harder et al. 2005). Meta-analytic results suggest a dose-response relationship, with each month of breastfeeding associated with a 4% decrease in risk of a childhood overweight risk (Harder et al. 2005).
Breastfeeding may also be protective in infants at high risk of early life weight gain in excess of recommendations. For example, in children classified as high risk for developing obesity based on the fact that they were born to mothers who were measured as overweight or obese, breastfeeding acted as a positive moderator of risk (Yeung et al. 2017). Additionally, in a sample of mothers reporting high depressive symptoms, continued breastfeeding between 6 and 12 months was associated with decreased risk of obesity in their children (Surkan et al. 2008). Moderating the influence of overweight/obesity risk factors is one way that breastfeeding may offer protective benefits to developing children.
The primary goal of the current study was to investigate indirect associations between psychological distress during pregnancy and child weight at 24 months through breastfeeding duration. It was predicted that: (1) maternal psychological distress, defined as clinically significant symptoms of anxiety or depression, would be associated with an increased risk of child overweight/obesity at age 2; (2) symptoms of psychological distress would be associated with decreased duration of breastfeeding; and (3) the relationship between maternal psychological distress and the childhood overweight/obesity at 24 months would be mediated by breastfeeding duration.
Methods
Procedure
This secondary analysis used data from the All Our Families study (AOF; Formerly All Our Babies, https://allourfamiliesstudy.com/), an established longitudinal pregnancy cohort in Alberta, Canada (McDonald et al. 2013; Tough et al. 2017). From 2008 to 2010, the AOF study recruited over 3000 pregnant women in Calgary, Canada. Participants were recruited through primary care facilities, laboratory services, community posters, and word of mouth. Data were collected at < 25 weeks gestation, 34–36 weeks gestation, and 4 months, 12 months, and 24 months postpartum. The goal of the larger project was to investigate maternal and infant health outcomes, infant/child development, and patterns of health care use. The current study used data from 1582 mother and child pairs who provided data about maternal education, income, age, infant/child height and weight from birth to 24 months, maternal prenatal anxiety and depression, breastfeeding behaviour, and maternal BMI. The original AOF cohort was not intended to follow women beyond 4 months postpartum. Recognizing the opportunity for long-term research analyses that the longitudinal data represented, 12- and 24-month questionnaires were developed and distributed to women who had children less than 13 months of age when ethics and funding were secured and resulted in not all of the original sample being eligible for full participation. For a full report of AOF study design, methods, and participant characteristics, see McDonald et al. (2013).
Child height/weight
Child physical measurements at birth were taken from the electronic health record. Gestational age and infant sex were also confirmed using the electronic health record. Height/length and weight at 12 months and 24 months were assessed in centimetres and kilograms respectively and obtained via parent report. z-scores for child weight status at age 2 were calculated using the gender norms provided by WHO’s growth charts (de Onis 2008). Raw physical measurements were used with the WHO Anthro syntax file for SPSS to create z-scores based on gendered norms for age (version 3.2.2) (World Health Organization 2011). In order to measure associations with overweight risk, a dichotomous overweight variable was calculated using WHO weight standards for length/height and age. Children were considered overweight at 24 months if they had a weight-for-length z-score above the 97th percentile for their age and sex, and this figure also includes those who would be considered obese. At 24 months, 84 children were missing data on weight (5.3%), and 131 children were missing data on length/height (8.3%).
Breastfeeding
Participants were asked at 4 and 12 months postpartum to report whether they had initiated breastfeeding and, if so, for how long they had breastfed in weeks. For mothers who indicated that they were still breastfeeding at the 12-month data collection point, a score of 52 weeks was assigned, and those who answered that they had not initiated breastfeeding were given a score of 0 weeks. Self-reported measurement of breastfeeding behaviour is understood to be a valid assessment technique (Ho and McGrath 2010). Breastfeeding data were not available for 221 women (14%).
Maternal prenatal psychological distress
Symptoms of maternal psychological distress were measured via self-report at 34–36 weeks gestation. To assess clinically significant psychological distress, self-reported symptoms of maternal anxiety and maternal depression were categorized to indicate whether they were above or below the respective clinical cut-off scores detailed below and then labelled as “anxiety” and “depression” throughout discussion of the results.
Maternal anxiety
The State Subscale of the Spielberger State-Trait Anxiety Inventory (STAI-S) (Spielberger et al. 1970) was used to assess maternal anxiety. The self-report STAI-S contains 20 items that assess feelings of anxiety in the current moment. Each question uses a 4-point Likert scale that ranges from “not at all” to “very much so”. Scores range from 20 to 80, with higher scores reflecting greater severity of anxious symptoms with a total score of 40 or above suggesting clinically relevant levels of anxiety. The measure has shown good reliability and validity in pregnant and postpartum populations (Benediktsson et al. 2016). Cronbach’s alpha in this sample was 0.92 (Tomfohr-Madsen et al. 2016). The clinical cut-off score was used to create a variable to separate those experiencing significant anxiety from those experiencing little or no anxiety. Anxiety data were missing for 62 women (3.9%).
Maternal depression
The Edinburgh Postnatal Depression Scale (EPDS) (Cox et al. 1987) was used to assess symptoms of prenatal depression. The EPDS is a 10-item self-report questionnaire that uses a 4-point Likert scale, with greater severity of depression symptoms represented by higher scores ranging from 0 to 30 and a recommended total score of 13 or above associated with higher likelihood of a depression diagnosis. This cut score was used to create a dichotomous variable separating those women experiencing clinically significant symptoms of depression from those experiencing non-clinically significant depression symptoms. The EPDS is considered to be valid and reliable for detection of depressive symptoms in pregnancy (Bergink et al. 2011). Cronbach’s alpha for the sample was 0.85 (Tomfohr-Madsen et al. 2016). Depression data were missing for 30 women (1.9%).
Covariates
All demographic information was collected as a part of a self-report questionnaire battery. Educational attainment and household income were used as indices of socio-economic status (SES). Maternal education was separated into high school or less, some or complete university/college, and some or complete graduate school, and annual income was separated into 3 categories as per the norms for Calgary (Statistics Canada 2014) into “below $40,000”, “$40,000–$79,999”, and “$80,000 and above”. Maternal pre-pregnancy height and weight was used to calculate pre-pregnancy BMI according to the standard formula of kg/m2 and was separated into underweight, average weight, overweight, and obese. Deliveries were classified as term (≥ 37 weeks) and preterm (< 37 weeks). Infant birth weight, maternal gestational weight gain, household income, and maternal education were included in the model as covariates based on their previously documented associations with infant weight status (Gibbs and Forste 2014; Sridhar et al. 2014). Gestational weight gain was calculated by subtracting pre-pregnancy weight from weight at 34–36 weeks gestation. Infant birth weight was taken from the medical record, with data unavailable for 212 infants (13.4%). Maternal BMI and maternal education were measured via self-report, with 96 women (6.1%) missing data on pre-pregnancy BMI and 12 (0.8%) missing data on education.
Statistical analysis
All statistical analyses were completed using SPSS 24.0 (SPSS Inc., Chicago, IL, USA) and mediation analyses were conducted using the SPSS procedure PROCESS v3.4 (Hayes 2017). First, a descriptive analysis was undertaken to assess for pre-existing patterns across groups. For the primary analysis, a logistic regression analysis was first conducted to determine whether symptoms of psychological distress in pregnancy, anxiety, and depression were associated with child overweight/obesity at 24 months. Next, a linear regression was used to investigate relationships between anxiety and depression and duration of breastfeeding.
Finally, potential mediation of the relationship between maternal psychological distress in pregnancy, anxiety, and depression and the child classification as overweight at 24 months through breastfeeding duration was investigated. The PROCESS procedure provides both direct and indirect effects for all paths included in the model, as well as 95% percentile bootstrapped confidence intervals to determine the significance of the indirect effects. Direct relationships between variables can often be obscured by other variables, and significant pathways may exist that would otherwise be missed by focusing on adhering to models of mediation that require assumptions about direct relationships between the independent and the dependent variables (for example, see Hayes 2009). An indirect effect is considered significant when the bootstrapped confidence intervals of the mediation do not cross zero, and this mediation approach does not require a significant direct effect between predictor and outcome (Hayes and Rockwood 2017). All models included maternal age, education, ethnicity, household income, pre-pregnancy BMI, term status, gestational weight gain, and infant birth weight. Bootstrapping with 10,000 samples was specified. A non-significant result was found using Little’s MCAR test (p = 0.064), suggesting that the missing data were missing at random. Therefore, all analyses were carried out using a complete case analysis method.
Results
Sample characteristics
The final sample included in this analysis was 1582 Canadian mother−child pairs from the AOF cohort. Mothers were an average of 30.95 years old, and 97.08% were married or common law. The participants primarily reported household income above the median income for Calgary with 72.90% reporting a household income greater than $80,000 per year. Participants were also highly educated with over 90% reporting at least some post-secondary education experience and 15.86% reporting a graduate-level education. Based on maternal reports of weight and height, 35.41% of women were classified as overweight/obese prior to pregnancy. On average, women gained 12.04 kg during gestation. The sample was predominately white (81.54%), with the remaining 18.46% representing a diverse subsample: Black (n = 22), First Nations (n = 4), Métis (n = 4), Chinese (n = 61), South Asian (n = 43), Filipino (n = 31), Latin American (n = 31), Southeast Asian (n = 20), Arab (n = 13), West Asian (n = 4), Japanese (n = 7), Mixed/Other (n = 43).
The self-reported prevalence of anxiety and depression in the sample was within the normal range for a pregnant population (Dennis et al. 2017; Woody et al. 2017), with 6.78% of pregnant participants scoring above the clinical cut scores for depression and 18.03% above the clinical cut score for anxiety. The correlation was moderate (r = .46) between women scoring above the cut-off on anxiety and those scoring above the cut-off on depression. Breastfeeding was initiated by nearly all participants (97.91%). The majority of participants (73.30%) reported breastfeeding for at least 6 months as per WHO recommendations and the average duration of the entire sample was 35.15 weeks. The average birthweight of infants was 3.36 kg. Over 12% of children were classified as overweight at 24 months. Full descriptive statistics for the entire sample, for women scoring above the cut score on the EPDS and STAI, and for women with children classified as overweight are presented in Tables 1 and 2.
Table 1.
Characteristics of sample according to self-reported mental health
| Variable | M (SD) or n (%) | ||
|---|---|---|---|
| Full sample, N = 1582* | EPDS > 13, N = 105* (6.8) | STAI > 40, N = 274* (18.0) | |
| Maternal age, years | 30.95 (4.44) | 30.88 (4.88) | 31.12 (4.40) |
| Education | |||
| High school or less | 130 (8.28) | 9 (8.74) | 31 (11.44) |
| Some or complete university | 1191 (75.86) | 72 (69.90) | 201 (74.17) |
| Some or complete graduate school | 249 (15.86) | 22 (21.36) | 39 (14.39) |
| Annual income, $ Canadian | |||
| $0–$39,999 | 89 (5.88) | 13 (15.66) | 27 (12.27) |
| $40,000–$79,999 | 321 (21.22) | 23 (27.71) | 71 (32.27) |
| ≥ $80,000 | 1103 (72.90) | 47 (56.63) | 122 (55.46) |
| Race | |||
| Caucasian | 1290 (81.54) | 77 (73.33) | 211 (77.01) |
| Other | 292 (18.46) | 28 (26.67) | 63 (22.99) |
| Marital status | |||
| Married/common law | 1275 (97.08) | 76 (91.57) | 213 (95.09) |
| Single/divorced/separated/widowed | 28 (2.92) | 7 (8.43) | 11 (4.91) |
| Any breastfeeding | |||
| Yes | 1548 (97.85) | 98 (93.33) | 266 (97.08) |
| No | 34 (2.15) | 7 (6.67) | 8 (2.92) |
| Breastfeeding duration, weeks | 35.15 (18.08) | 28.18 (20.05) | 29.59 (18.47) |
| Pre-pregnancy BMI | |||
| Underweight (< 18.5) | 62 (3.99) | 5 (4.90) | 11 (4.09) |
| Normal weight (18.5–24.9) | 943 (60.60) | 62 (60.78) | 148 (55.02) |
| Overweight (25–29.9) | 356 (22.88) | 17 (16.67) | 64 (23.79) |
| Obese (≥ 30) | 195 (12.53) | 18 (17.65) | 46 (17.10) |
| Gestational weight gain, kg | 12.04 (5.45) | 12.39 (6.71) | 12.12 (5.23) |
| Infant birth weight, kg | 3.36 (0.53) | 3.29 (0.62) | 3.30 (0.68) |
| Term status | |||
| Term (≥ 37 weeks) | 1425 (93.38) | 82 (83.67) | 233 (87.92) |
| Preterm (< 37 weeks) | 101 (6.62) | 16 (16.33) | 32 (12.08) |
| Child weight-for-length > 97th percentile | |||
| Yes | 177 (12.60) | 13 (14.94) | 32 (14.48) |
| No | 1228 (87.40) | 74 (85.06) | 189 (85.52) |
*Numbers presented in table represent available data, excluding those with missing values
Table 2.
Characteristics of sample by child weight
| Variable | M (SD) or n (%) | |
|---|---|---|
| Full sample, N = 1582* | Child overweight/obese, N = 177* | |
| Maternal age, years | 30.95 (4.44) | 31.43 (4.53) |
| Education | ||
| High school or less | 130 (8.28) | 16 (9.20) |
| Some or complete university | 1191 (75.86) | 125 (71.84) |
| Some or complete graduate school | 249 (15.86) | 33 (18.97) |
| Annual income, $ Canadian | ||
| $0–$39,999 | 89 (5.88) | 12 (7.95) |
| $40,000–$79,999 | 321 (21.22) | 40 (26.49) |
| ≥ $80,000 | 1103 (72.90) | 99 (65.56) |
| Race | ||
| Caucasian | 1290 (81.54) | 142 (80.23) |
| Other | 292 (18.46) | 35 (19.77) |
| Marital status | ||
| Married/common law | 1275 (97.08) | 146 (96.05) |
| Single/divorced/separated/widowed | 28 (2.92) | 6 (3.95) |
| Child sex | ||
| Male | 798 (52.00) | 99 (55.93) |
| Female | 738 (48.00) | 78 (44.07) |
| Any breastfeeding | ||
| Yes | 1548 (97.85) | 174 (98.31) |
| No | 34 (2.15) | 3 (1.69) |
| Breastfeeding duration, weeks | 35.15 (18.08) | 29.12 (18.49) |
| Maternal psychological distress | ||
| Depression (EPDS > 13) | 105 (6.78) | 13 (7.43) |
| Anxiety (STAI-S > 40) | 274 (18.03) | 32 (18.71) |
| Pre-pregnancy BMI | ||
| Underweight (< 18.5) | 62 (3.99) | 2 (1.16) |
| Normal weight (18.5–24.9) | 943 (60.60) | 89 (51.74) |
| Overweight (25–29.9) | 356 (22.88) | 48 (27.91) |
| Obese (≥ 30) | 195 (12.53) | 33 (19.19) |
| Gestational weight gain, kg | 12.04 (5.45) | 11.90 (7.73) |
| Infant birth weight, kg | 3.36 (0.53) | 3.55 (0.51) |
| Term status | ||
| Term (≥ 37 weeks) | 1425 (93.38) | 164 (93.71) |
| Preterm (< 37 weeks) | 101 (6.62) | 11 (6.29) |
*Numbers presented in table represent available data, excluding those with missing values
Psychological distress during pregnancy and child overweight at 24 months
Logistic regression was used to investigate the associations between maternal anxiety and depression in pregnancy and child overweight at 24 months. Covariates included maternal age, education, ethnicity, household income, pre-pregnancy BMI, gestational weight gain, term status, breastfeeding duration, and infant birth weight. Results indicated that higher maternal age (anxiety only), lower breastfeeding duration, and higher infant birthweight were associated with increased risk of childhood overweight/obesity at 24 months. Neither anxiety nor depression were directly related to child weight classification at 24 months. Results of the logistic regression analyses are reported in Table 3.
Table 3.
Logistic regression for maternal prenatal depression and anxiety predicting child classification as overweight at 24 months
| Predictor | B | SE, B | OR | 95% CI |
|---|---|---|---|---|
| Depression (N = 897) | ||||
| Maternal age | 0.05 | 0.03 | 1.05 | 0.99–1.10 |
| Household income | − 0.22 | 0.20 | 0.81 | 0.54–1.19 |
| Maternal education | 0.17 | 0.23 | 1.18 | 0.76–1.85 |
| Maternal ethnicity | − 0.43 | 0.40 | 0.65 | 0.30–1.42 |
| Pre-pregnancy BMI | 0.23 | 0.14 | 1.26 | 0.95–1.66 |
| Gestational weight gain | − 0.00 | 0.02 | 1.00 | 0.97–1.04 |
| Term status | 0.16 | 0.30 | 1.18 | 0.66–2.10 |
| Infant birthweight | 0.79*** | 0.22 | 2.21 | 1.44–3.39 |
| Breastfeeding (weeks) | − 0.02*** | 0.01 | 0.98 | 0.97–0.99 |
| EPDS > 13 | −0.80 | 0.55 | 0.45 | 0.15–1.31 |
| Anxiety (N = 885) | ||||
| Maternal age | 0.05* | 0.03 | 1.05 | 1.00–1.10 |
| Household income | − 0.21 | 0.20 | 0.81 | 0.55–1.21 |
| Maternal education | 0.16 | 0.23 | 1.18 | 0.75–1.84 |
| Maternal ethnicity | − 0.38 | 0.40 | 0.69 | 0.31–1.50 |
| Pre-pregnancy BMI | 0.23 | 0.14 | 1.26 | 0.95–1.66 |
| Gestational weight gain | 0.00 | 0.02 | 1.00 | 0.97–1.04 |
| Term status | 0.18 | 0.30 | 1.20 | 0.67–2.15 |
| Infant birthweight | 0.81*** | 0.22 | 2.25 | 1.46–3.46 |
| Breastfeeding (weeks) | − 0.02*** | 0.01 | 0.98 | 0.97–0.99 |
| STAI-S > 40 | − 0.40 | 0.31 | 0.67 | 0.37–1.23 |
*p < 0.05
**p < 0.01
***p < 0.001
EPDS, Edinburgh Postnatal Depression Scale; STAI, Spielberger State-Trait Anxiety Inventory
Psychological distress during pregnancy and breastfeeding duration
Hierarchical regression analyses were used to investigate associations between psychological distress during pregnancy and breastfeeding duration. Estimates for anxiety and depression were conducted separately. In step 1, covariates were entered into the model, including maternal age, education, ethnicity, household income, pre-pregnancy BMI, gestational weight gain, term status, and infant birth weight. In step 2, anxiety or depression was entered. Results of the regression analyses are presented in Table 4. Anxiety during pregnancy was associated with decreased breastfeeding duration (B = − 5.23, p = 0.001), such that mothers who reported symptoms of anxiety above the cut-off for clinical significance breastfed an average of 5.23 weeks less than those who did not report significant anxiety. When anxiety was added to the model, the variance explained by the model improved (F(1, 932) = 11.87, p = 0.001, R2 = .07 (R2 change = .01)).
Table 4.
Maternal prenatal depression and anxiety predicting weeks breastfeeding
| Regression Step | R2 | B | SE | β |
|---|---|---|---|---|
| Depression | ||||
| Step 1 | .05 | |||
| Maternal age | 0.33* | 0.13 | 0.08 | |
| Household income | − 0.58 | 1.04 | − 0.02 | |
| Maternal education | 3.52** | 1.21 | 0.10 | |
| Maternal ethnicity | 0.75 | 1.75 | 0.01 | |
| Pre-pregnancy BMI | − 04.09*** | 0.78 | − 0.18 | |
| Gestational weight gain | − 0.02 | 0.11 | − 0.01 | |
| Term status | − 0.49 | 2.61 | − 0.01 | |
| Infant birthweight | 1.91 | 1.19 | 0.06 | |
| Step 2 | .06*+ | |||
| Maternal age | 0.33* | 0.13 | 0.08 | |
| Household income | − 0.70 | 1.04 | − 0.02 | |
| Maternal education | 3.64** | 1.21 | 0.10 | |
| Maternal ethnicity | 0.85 | 1.74 | 0.02 | |
| Pre-pregnancy BMI | − 03.95*** | 0.78 | − 0.17 | |
| Gestational weight gain | − 0.01 | 0.11 | − 0.00 | |
| Term status | − 0.00 | 2.61 | 0.00 | |
| Infant birth weight | 1.89 | 1.19 | 0.06 | |
| EPDS > 13 | − 5.82** | 2.35 | − 0.08 | |
| Anxiety | ||||
| Step 1 | .05 | |||
| Maternal age | 0.32* | 0.13 | 0.08 | |
| Household income | − 0.52 | 1.04 | − 0.02 | |
| Maternal education | 3.58** | 1.21 | 0.10 | |
| Maternal ethnicity | 0.87 | 1.78 | 0.02 | |
| Pre-pregnancy BMI | − 4.14*** | 0.78 | − 0.18 | |
| Gestational weight gain | − 0.01 | 0.11 | − 0.00 | |
| Term status | − 1.05 | 2.67 | − 0.01 | |
| Infant birth weight | 1.91 | 1.19 | 0.06 | |
| Step 2 | .07***+ | |||
| Maternal age | 0.33** | 0.13 | 0.08 | |
| Household income | − 0.89 | 1.04 | − 0.03 | |
| Maternal education | 3.54** | 1.21 | 0.10 | |
| Maternal ethnicity | 0.97 | 1.76 | 0.02 | |
| Pre-pregnancy BMI | − 3.93*** | 0.78 | − 0.17 | |
| Gestational weight gain | 0.01 | 0.11 | 0.00 | |
| Term status | − 0.25 | 2.66 | − 0.00 | |
| Infant birth weight | 1.92 | 1.19 | 0.06 | |
| STAI-S > 40 | − 5.23*** | 1.52 | − 0.11 | |
*p < 0.05
**p < 0.01
***p < 0.001
+Denotes significant R2 change
Similarly, inclusion of depression into the model resulted in a significant improvement in explained variance (F(1, 945) = 6.12, p = 0.014, R2 = .06 (R2 change = .01), with classification of depression being associated with decrease in weeks spent breastfeeding (B = − 5.82, p = 0.014). Said another way, women who were classified as depressed during pregnancy reported 5.82 fewer weeks of breastfeeding.
Indirect (mediated) associations between psychological distress during pregnancy and child overweight at 24 months through breastfeeding duration
We next tested whether there was an indirect effect of psychological distress (X) on child risk of overweight at 24 months (Y), mediated by breastfeeding duration (M). The conceptual mediation model is presented in Fig. 1. All models included maternal age, education, ethnicity, household income, pre-pregnancy BMI, gestational weight gain, term status, and infant birth weight as covariates.
Fig. 1.
Conceptual models of breastfeeding as a mediator between maternal prenatal anxiety and depression and child weight-for-length at 24 months. Note: The pathways presented combine OLS regression and logistic regression. To better interpret the mediated relationship presented, we describe the pathways from anxiety to weight as an example. Path A1 from anxiety to breastfeeding is an OLS linear regression with a coefficient of − 4.86. In the sample, scoring in the clinically significant range on the STAI (anxiety symptoms) was associated with approximately 4.86 weeks less spent breastfeeding. Path B1 from breastfeeding to child weight-for-length above the 97th percentile at 24 months was conducted as a logistic regression due to the dichotomous nature of the weight outcome. Regardless of anxiety status, the B1 pathway coefficient describes the relationship between breastfeeding and child weight in our sample and produces a log odds value of − 0.02. Odds ratio = elog odds; therefore, taking the exponent of this value (e−0.02) provides the odds ratio for the effect of breastfeeding on child weight. The odds ratio = 0.98, indicating a protective effect of breastfeeding such that for each additional week of breastfeeding, there is a 2% decrease in odds of child overweight at 24 months. Altogether, the cumulative result is 4.86 weeks less breastfeeding together with 2% additional risk with each week reduction in breastfeeding. This pathway represents a modifiable behavioural pathway that is associated with non-insignificant increases in childhood classification of overweight/obese at 24 months
Anxiety
Significant pathways were observed between anxiety and weeks spent breastfeeding (path A1; B(SE) = − 4.86 (1.60), p = 0.003), as well as between weeks spent breastfeeding and child weight classification (path B1; B(SE) = − 0.02 (0.01), p < 0.001). The indirect effect of maternal prenatal anxiety on child weight classification at 24 months through breastfeeding duration was significant (path A1B1, B(SE) = 0.10 (0.04), 95% CI 0.03–0.21). There was no direct effect of maternal anxiety on child weight classification at 24 months (path C1′; B(SE) = − 0.40 (0.31), p = 0.197).
Depression
Significant pathways were also observed between depression and weeks spent breastfeeding (path A2; B(SE) = − 5.20 (2.43), p = 0.033), as well as between weeks spent breastfeeding and child weight classification (path B2; B(SE) = − 0.02 (0.01), p < 0.001). The indirect effect of maternal prenatal depression and child weight classification at 24 months through breastfeeding duration was significant (path A2B2, B = 0.11 (0.06), 95% CI 0.00–0.25). There was no direct effect of maternal depression on child weight classification at 24 months (path C2′; B(SE) = − 0.80 (0.54), p = 0.145).
Discussion
The results of this longitudinal cohort analysis support an indirect relationship between maternal psychological distress in pregnancy and childhood overweight/obesity at 24 months, mediated through breastfeeding duration. Symptoms of psychological distress were negatively associated with duration of breastfeeding, and shorter duration of breastfeeding increased risk for a child measuring above the 97th percentile in weight-for-length. Importantly, the indirect relationship between maternal psychological distress and child weight classification was significant over and above inclusion of variables previously determined to be important contributors to early life weight gain, including: socio-economic status (SES) indicators (Wang and Lim 2012), maternal weight (Gaillard et al. 2013), and infant birth weight (Gaudet et al. 2014). We did not detect a direct association between maternal psychological distress and child weight. This is not unexpected, as there is little evidence of a direct physiological mechanism between maternal anxiety or depression and child weight. It is more likely, therefore, that there are multiple indirect behavioural pathways through which maternal distress influences child weight. In this paper, we focused on breastfeeding due to its established protective effects on child growth and development and its modifiable nature.
Breastfeeding and child weight classification
Meta-analytic studies suggest that breastfeeding for any amount of time can significantly reduce the risk of obesity in childhood (Horta et al. 2015). The protective effects of breastfeeding remain even after adjusting for common confounding variables such as SES and parental physical measurements (Harder et al. 2005). Indeed, breastfeeding acts in a protective capacity against several specific risk factors for obesity in childhood, such as high maternal BMI, higher than recommended gestational weight gain, high birth weight, and indicators of low SES (Wang and Lim 2012; Yeung et al. 2017; Zhu et al. 2015). There also appears to be a dose-response relationship, suggesting reductions in risk of a childhood overweight with each additional month of exclusive breastfeeding (Horta et al. 2007). There are several possible explanations for the protective effect of breastfeeding, including endocrine response differences between breastmilk and formula, better milk intake control associated with breastfeeding, and macronutrient differences between breastmilk and formula (e.g., protein) (Bartok and Ventura 2009).
Mental health difficulties, along with structural and social barriers, represent factors associated with higher likelihood of parents engaging in behavioural practices that increase the risk of their child being classified as overweight. For instance, mothers experiencing depression and mothers who report higher levels of stress are more likely to engage in early introduction of solid foods (before 4 months) and juices and allow their infants to sleep with a bottle, both of which are associated with infant weight (Gibbs and Forste 2014). Overall maternal psychological distress has been associated with increased energy intake in children less than 6 months old (Thompson and Bentley 2013). Fortunately, it appears that breastfeeding may act in a protective capacity. One recent study illustrated that breastfeeding may contribute to a slower rate of eating by the child, thereby protecting them from overindulgent eating patterns (Rogers and Blissett 2017); other research has suggested the possibility that breastfed children more readily adopt a wider variety of foods (Birch and Fisher 1998). Additionally, children in lower SES brackets who are primarily breastfed in the first year rather than fed formula are half as likely to have obesity at age 2 (Gibbs and Forste 2014). Not only may breastfeeding protect against the risk of unhealthy infant weight gain, but longer duration of exclusive breastfeeding can delay the introduction of complementary foods such as juices or solid foods that may confer additional risk of overweight classification (Armstrong et al. 2014).
Pregnancy is a time of increased stress, and during pregnancy, a substantial number of women experience anxiety disorders and/or depressive disorders (Dennis et al. 2017). One possible pathway from psychological distress to child overweight classification may be through parallel behaviours of increased tendency to introduce complementary foods earlier than recommendations and decreases in both overall length of breastfeeding and duration of exclusive breastfeeding. By treating psychological distress during pregnancy, an additional benefit may be increases in breastfeeding and decreases in behavioural actions that promote early life overweight risk. The current investigation lacks the necessary design to explore this relationship, but it represents a rich area for future research.
Limitations
Findings from this study should be interpreted in light of several limitations. First, on average, the sample was highly educated and reported a high household income. The relationships observed in these findings may not generalize outside of well-educated two-parent families with average to high income. Therefore, the results may differ in families with greater socio-economic risk factors. While the final sample size used in this analysis was large, there were several cases where breastfeeding, child weight, or maternal psychological distress data were not available for all time points. Attrition is common in longitudinal research and is unlikely to represent a systematic difference in those participants who were missing at follow-up in this study. While there was evidence of randomness in our missing data, there remains the possibility that completers and non-completers differ on non-measured variables. Our attrition rates were within the expected range for a study of this duration and magnitude (Gustavson et al. 2012). Despite the large sample, in this analysis, we were unable to control for all delivery-level maternal factors that may have influenced breastfeeding rates and potentially altered the results (e.g., gestational diabetes, hypertensive disorders of pregnancy). Similarly, child diet, timing of physical activity, and screen time were not included in this analysis.
Another limitation is the high level of breastfeeding seen in this sample. Our sample reported high rates of breastfeeding initiation and duration, which may be specific to the local health care system and resources available in the area and limit generalizability to other populations. Women self-reported their pre-pregnancy weight, which may be less accurate than physician-measured weight. These self-reports were then used to calculate gestational weight gain. However, previous research has suggested that self-reported and measured pre-pregnancy weight do not significantly differ with regard to perinatal outcomes (Han et al. 2016). Finally, we did not examine or adjust for other pathways through which psychological distress might influence child weight, such as screen time or early introduction of solid foods. Previous investigations have provided some evidence for an association between timing of solid food intake and child weight classification (Schack-Nielsen et al. 2010), although literature reviews have criticized the overall body of evidence for lacking the high-quality research that provides evidence for how and when to introduce complementary foods (Daniels et al. 2015). Nevertheless, it remains possible that there are behavioural factors other than breastfeeding contributing indirectly to child weight.
Recommendations and future directions
The results of the present study add to an already established literature pointing to the importance of screening and treating psychological distress in pregnancy, and highlight the need to provide support for women with mental health concerns in relation to breastfeeding. These findings are significant because they identify specific risk factors that negatively influence child weight and increase the likelihood of unhealthy weight status. Importantly, the pathway to overweight examined in this paper highlights modifiable behavioural factors that can be targeted in order to improve child weight outcomes. Several studies have suggested that pregnant women prefer psychotherapy to medication for both depression and anxiety (Arch 2014; Goodman 2009), and fortunately, several efficacious treatments exist, including cognitive behaviour therapy, interpersonal therapy, and family therapy (Cluxton-Keller and Bruce 2018; Marchesi et al. 2016). WHO provides exhaustive guidelines on breastfeeding recommendations (Kramer and Kakuma 2012), and interventions aimed at improving breastfeeding adherence have shown promise (Kramer et al. 2007). Ultimately, our finding of a mediation effect of breastfeeding on child weight in mothers experiencing psychological distress suggests that it may be most effective to implement interventions that target mental health and breastfeeding simultaneously in order to achieve meaningful change in child weight. Future research should focus on this possibility.
Rates of breastfeeding initiation were high in this cohort, as was duration. This above-average uptake of recommended breastfeeding practices could be due to the demographic factors of the participants, as higher education and older age are positively associated with breastfeeding practices. Lack of sufficient information and structural supports on breastfeeding behaviour is one risk factor that has been cited as a possible explanation for the steep drop-offs in breastfeeding behaviour after initiation (Chalmers et al. 2009). Thus, it is possible that the efforts of the breastfeeding-friendly messaging adopted by Alberta Health Services (AHS) in the last 16 years are responsible for a cultural shift that encourages positive breastfeeding messaging. AHS was quick to adopt the recommended breastfeeding practices released by WHO, and as such, women get consistent messaging on the importance of breastfeeding throughout their prenatal care. This cohort provides promising data that increases in breastfeeding behaviour are possible within a health care system that provides the necessary resources and support. This raises the possibility that targeting breastfeeding alone or in combination with traditional psychotherapeutic interventions can have a greater positive downstream effect on child overweight than focusing on maternal mental health alone.
Conclusions
Weight classification of overweight or obesity in early life is associated with an increased risk of higher BMI in adolescence and adulthood and higher rates of obesity-related chronic health conditions (Llewellyn et al. 2016; Simmonds et al. 2016). National public education and health programs are showing signs of reducing obesity rates in young Canadians (Rodd and Sharma 2016). However, a significant number of Canadian children remain in higher weight classifications, suggesting that the identification of unique and treatable risk factors is still necessary (Rodd and Sharma 2016). Future research is needed to determine whether improvements in mental health during pregnancy may lead to improvements in breastfeeding initiation and duration and whether these combined increases result in lower risks of childhood weight-related health consequences.
Acknowledgements
Special thanks to the participant mothers and children of the AOF cohort, as well as all the investigators, research assistants, administrators, and volunteers who helped bring such a large-scale project to life. Thank you to Meredith Brockway, who was consulted on breastfeeding information in Calgary, and to Anna L. MacKinnon for providing the French language translation of the abstract.
Funding information
The All Our Babies (AOB) study was funded by Alberta Innovates Health Solutions Interdisciplinary Team grant #200700595, Calgary, Alberta, Canada. Matthew Shay was supported by grants from the Alberta Children’s Hospital Research Initiative, Talisman Energy Fund for Healthy Living and Injury Prevention Studentship Support, and internal awards through the University of Calgary.
Compliance with ethical standards
Ethical approval was provided by Research Ethics Board Certification #REB16-0345 through University of Calgary.
Conflict of interest
The authors declare that they have no conflict of interest.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Arch JJ. Cognitive behavioral therapy and pharmacotherapy for anxiety: treatment preferences and credibility among pregnant and non-pregnant women. Behav Res Ther. 2014;52(1):53–60. doi: 10.1016/j.brat.2013.11.003. [DOI] [PubMed] [Google Scholar]
- Armstrong J, Abraham EC, Squair M, Brogan Y, Merewood A. Exclusive breastfeeding, complementary feeding, and food choices in UK infants. J Hum Lact. 2014;30(2):201–208. doi: 10.1177/0890334413516383. [DOI] [PubMed] [Google Scholar]
- Bartok CJ, Ventura AK. Mechanisms underlying the association between breastfeeding and obesity. Int J Pediatr Obes. 2009;4(4):196–204. doi: 10.3109/17477160902763309. [DOI] [PubMed] [Google Scholar]
- Benediktsson I, McDonald S, Tough S. Examining the psychometric properties of three standardized screening tools in a pregnant and parenting population. Maternal and Child Health Journal. 2016;21:1–7. doi: 10.1007/s10995-016-2128-4. [DOI] [PubMed] [Google Scholar]
- Bergink V, Kooistra L, Lambregtse-van den Berg MP, Wijnen H, Bunevicius R, van Baar A, Pop V. Validation of the Edinburgh Depression Scale during pregnancy. J Psychosom Res. 2011;70(4):385–389. doi: 10.1016/j.jpsychores.2010.07.008. [DOI] [PubMed] [Google Scholar]
- Birch LL, Fisher JO. Development of eating behaviors among children and adolescents. Pediatrics. 1998;101(3):539–549. [PubMed] [Google Scholar]
- Braungart-Rieker JM, Lefever JB, Planalp EM, Moore ES. Body mass index at 3 years of age: Cascading effects of prenatal maternal depression and mother-infant dynamics. Journal of Pediatrics. 2016;177(037224):128–132.e1. doi: 10.1016/j.jpeds.2016.06.023. [DOI] [PubMed] [Google Scholar]
- Carling SJ, Demment MM, Kjolhede CL, Olson CM. Breastfeeding duration and weight gain trajectory in infancy. Pediatrics. 2015;135(1):111–119. doi: 10.1542/peds.2014-1392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chalmers B, Levitt C, Heaman M, O’Brien B, Sauve R, Kaczorowski J, et al. Breastfeeding rates and hospital breastfeeding practices in Canada: a national survey of women. Birth. 2009;36(2):122–132. doi: 10.1111/j.1523-536X.2009.00309. [DOI] [PubMed] [Google Scholar]
- Cluxton-Keller F, Bruce ML. Clinical effectiveness of family therapeutic interventions in the prevention and treatment of perinatal depression: a systematic review and meta-analysis. PLoS One. 2018;13(6):e0198730. doi: 10.1371/journal.pone.0198730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression: development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150(6):782–786. doi: 10.1192/bjp.150.6.782. [DOI] [PubMed] [Google Scholar]
- Daniels L, Mallan KM, Fildes A, Wilson J. The timing of solid introduction in an “obesogenic” environment: a narrative review of the evidence and methodological issues. Aust N Z J Public Health. 2015;39(4):366–373. doi: 10.1111/1753-6405.12376. [DOI] [PubMed] [Google Scholar]
- de Onis M. The new WHO child growth standards. Paediatria Croatica Supplement. 2008;52(SUPP.1):13–17. doi: 10.4067/S0370-41062009000400012. [DOI] [Google Scholar]
- Dennis CL, Falah-Hassani K, Shiri R. Prevalence of antenatal and postnatal anxiety: systematic review and meta-analysis. Br J Psychiatry. 2017;210(5):315–323. doi: 10.1192/bjp.bp.116.187179. [DOI] [PubMed] [Google Scholar]
- Dias CC, Figueiredo B. Breastfeeding and depression: a systematic review of the literature. J Affect Disord. 2015;171:142–154. doi: 10.1016/j.jad.2014.09.022. [DOI] [PubMed] [Google Scholar]
- Fallon V, Halford JCG, Bennett KM, Harrold JA. Postpartum-specific anxiety as a predictor of infant-feeding outcomes and perceptions of infant-feeding behaviours: new evidence for childbearing specific measures of mood. Archives of Women’s Mental Health. 2018;21(2):181–191. doi: 10.1007/s00737-017-0775-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gaillard R, Durmus B, Hofman A, MacKenbach JP, Steegers EAP, Jaddoe VWV. Risk factors and outcomes of maternal obesity and excessive weight gain during pregnancy. Obesity. 2013;21(5):1046–1055. doi: 10.1002/oby.20088. [DOI] [PubMed] [Google Scholar]
- Gaudet L, Ferraro Z, Wen S, Walker M. Maternal obesity and occurrence of fetal macrosomia: a systematic review and meta-analysis. BioMed Research Internation. 2014;2014:1–22. doi: 10.1155/2014/640291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gibbs BG, Forste R. Socioeconomic status, infant feeding practices and early childhood obesity. Pediatric Obesity. 2014;9(2):135–146. doi: 10.1111/j.2047-6310.2013.00155.x. [DOI] [PubMed] [Google Scholar]
- Goodman JH. Women’s attitudes, preferences, and perceived barriers to treatment for perinatal depression. Birth. 2009;36(1):60–69. doi: 10.1111/j.1523-536X.2008.00296.x. [DOI] [PubMed] [Google Scholar]
- Gustavson, K., Von Soest, T., Karevold, E., & Roysamb, E. (2012). Attrition and generalizability in longitudinal studies: findings from a 15-year population-based study and a Monte Carlo simulation study. BMC Public Health, 12(1). 10.1186/1471-2458-12-918. [DOI] [PMC free article] [PubMed]
- Han E, Abrams B, Sridhar S, Xu F, Hedderson M. Validity of self-reported pre-pregnancy weight and body mass index classification in an integrated health care delivery system. Paediatr Perinat Epidemiol. 2016;30(4):314–319. doi: 10.1111/ppe.12286. [DOI] [PubMed] [Google Scholar]
- Harder T, Bergmann R, Kallischnigg G, Plagemann A. Duration of breastfeeding and risk of overweight: a meta-analysis. Am J Epidemiol. 2005;162(5):397–403. doi: 10.1093/aje/kwi222. [DOI] [PubMed] [Google Scholar]
- Hayes AF. Beyond Baron and Kenny: statistical mediation analysis in the new millennium. Commun Monogr. 2009;76(4):408–420. doi: 10.1080/03637750903310360. [DOI] [Google Scholar]
- Hayes AF. In: Introduction to mediation, moderation, and conditional process analysis. Second. Little T, editor. New York: The Guildford Press; 2017. [Google Scholar]
- Hayes AF, Rockwood NJ. Regression-based statistical mediation and moderation analysis in clinical research: observations, recommendations, and implementation. Behav Res Ther. 2017;98:39–57. doi: 10.1016/j.brat.2016.11.001. [DOI] [PubMed] [Google Scholar]
- Ho YJ, McGrath JM. A review of the psychometric properties of breastfeeding assessment tools. JOGNN - Journal of Obstetric, Gynecologic, and Neonatal Nursing. 2010;39(4):386–400. doi: 10.1111/j.1552-6909.2010.01153.x. [DOI] [PubMed] [Google Scholar]
- Horta, B. L., Bahl, R., Martines, J. C., & Victora, C. G. (2007). Evidence on the long-term effects of breastfeeding: systematic reviews and meta-analyses. World Health Organization, 1–52.
- Horta B, Loret De Mola C, Victora C. Long-term consequences of breastfeeding on cholesterol, obesity, systolic blood pressure and type 2 diabetes: a systematic review and meta-analysis. International Journal of Paediatrics. 2015;104:30–37. doi: 10.1111/apa.13133. [DOI] [PubMed] [Google Scholar]
- Kramer MS, Kakuma R. Optimal duration of exclusive breastfeeding. Cochrane Database of Systematic Reviews (Online) 2012;8(4):CD003517. doi: 10.1002/14651858.CD003517.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kramer MS, Matush L, Vanilovich I, Platt RW, Bogdanovich N, Sevkovskaya Z, et al. Effects of prolonged and exclusive breastfeeding on child height, weight, adiposity, and blood pressure at age 6.5 y: evidence from a large randomized trial. Am J Clin Nutr. 2007;86(6):1717–1721. doi: 10.1093/ajcn/86.5.1717. [DOI] [PubMed] [Google Scholar]
- Llewellyn A, Simmonds M, Owen CG, Woolacott N. Childhood obesity as a predictor of morbidity in adulthood: a systematic review and meta-analysis. Obes Rev. 2016;17(1):56–67. doi: 10.1111/obr.12316. [DOI] [PubMed] [Google Scholar]
- Marchesi C, Ossola P, Amerio A, Daniel BD, Tonna M, De Panfilis C. Clinical management of perinatal anxiety disorders: a systematic review. J Affect Disord. 2016;190:543–550. doi: 10.1016/j.jad.2015.11.004. [DOI] [PubMed] [Google Scholar]
- McDonald SW, Lyon AW, Benzies KM, McNeil DA, Lye SJ, Dolan SM, et al. The All Our Babies pregnancy cohort: design, methods, and participant characteristics. BMC Pregnancy and Childbirth. 2013;13(Suppl 1):S2. doi: 10.1186/1471-2393-13-S1-S2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rodd, C., & Sharma, A. K. (2016). Recent trends in the prevalence of overweight and obesity among Canadian children. Can Med Assoc J. 10.1503/cmaj.150854. [DOI] [PMC free article] [PubMed]
- Rogers SL, Blissett J. Breastfeeding duration and its relation to weight gain, eating behaviours and positive maternal feeding practices in infancy. Appetite. 2017;108:399–406. doi: 10.1016/j.appet.2016.10.020. [DOI] [PubMed] [Google Scholar]
- Schack-Nielsen L, Sørensen TIA, Mortensen EL, Michaelsen KF. Late introduction of complementary feeding, rather than duration of breastfeeding, may protect against adult overweight. Am J Clin Nutr. 2010;7:1–9. doi: 10.3945/ajcn.2008.27078.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shields M, Tremblay MS. Canadian childhood obesity estimates based on WHO, IOTF and CDC cut-points. Int J Pediatr Obes. 2010;5(3):265–273. doi: 10.3109/17477160903268282. [DOI] [PubMed] [Google Scholar]
- Simmonds M, Llewellyn A, Owen CG, Woolacott N. Predicting adult obesity from childhood obesity: a systematic review and meta-analysis. Obes Rev. 2016;17(2):95–107. doi: 10.1111/obr.12334. [DOI] [PubMed] [Google Scholar]
- Spielberger CD, Gorsuch RL, Lushene RE, Vagg PR, Jacobs GA. STAI: manual for the State-Trait Anxiety Inventory. Palo Alto: Consulting Psychologist’s Press; 1970. [Google Scholar]
- Sridhar SB, Darbinian J, Ehrlich SF, Markman MA, Gunderson EP, Ferrara A, Hedderson MM. Maternal gestational weight gain and offspring risk for childhood overweight or obesity. American Journal of Obstetrics and Gynecology. 2014;211(3):259.e1–259.e8. doi: 10.1016/j.ajog.2014.02.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Statistics Canada. (2014). Family income and income of individuals, related variables: sub-provincial data, 2012, 2010–2013. Retrieved from http://www.statcan.gc.ca/daily-quotidien/140723/dq140723c-eng.pdf
- Surkan PJ, Kawachi I, Peterson KE. Childhood overweight and maternal depressive symptoms. J Epidemiol Community Health. 2008;62(5):e11. doi: 10.1136/jech.2007.065664. [DOI] [PubMed] [Google Scholar]
- Thompson AL, Bentley ME. The critical period of infant feeding for the development of early disparities in obesity. Soc Sci Med. 2013;97:288–296. doi: 10.1016/j.socscimed.2012.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tomfohr-Madsen LM, Bayrampour H, Tough S. Maternal history of childhood abuse and risk of asthma and allergy in 2-year-old children. Psychosom Med. 2016;78(9):1031–1042. doi: 10.1097/PSY.0000000000000419. [DOI] [PubMed] [Google Scholar]
- Tough, S. C., McDonald, S. W., Collisson, B. A., Graham, S. A., Kehler, H., Kingston, D., & Benzies, K. (2017). Cohort profile: the All Our Babies pregnancy cohort (AOB). Int J Epidemiol. 10.1093/ije/dyw363. [DOI] [PubMed]
- Wang Y, Lim H. The global childhood obesity epidemic and the association between socio-economic status and childhood obesity. International Review of Psychiatry (Abingdon, England) 2012;24(3):176–188. doi: 10.3109/09540261.2012.688195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Woody CAA, Ferrari AJJ, Siskind DJJ, Whiteford HAA, Harris MGG. A systematic review and meta-regression of the prevalence and incidence of perinatal depression. J Affect Disord. 2017;219(April):86–92. doi: 10.1016/j.jad.2017.05.003. [DOI] [PubMed] [Google Scholar]
- World Health Organization. (2011). WHO anthro.
- Yeung H, Leff M, Rhee KE. Effect of exclusive breastfeeding among overweight and obese mothers on infant weight-for-length percentile at 1 year. Breastfeed Med. 2017;12(1):39–47. doi: 10.1089/bfm.2016.0071. [DOI] [PubMed] [Google Scholar]
- Zhu Y, Hernandez LM, Dong Y, Himes JH, Hirschfeld S, Forman MR. Longer breastfeeding duration reduces the positive relationships among gestational weight gain, birth weight and childhood anthropometrics. J Epidemiol Community Health. 2015;69(7):632–638. doi: 10.1136/jech-2014-204794. [DOI] [PubMed] [Google Scholar]