Maternal depressive symptoms not associated with reduced height in young children in a US prospective cohort study.
ABSTRACT Shorter stature is associated with greater all cause and heart disease mortality, but taller stature with increased risk of cancer mortality. Though childhood environment is important in determining height, limited data address how maternal depression affects linear growth in children. We examined the relationships between antenatal and postpartum depressive symptoms and child height and linear growth from birth to age 3 years in a U.S. sample.
Subjects were 872 mother-child pairs in Project Viva, a prospective pre-birth cohort study. The study population is relatively advantaged with high levels of income and education and low risk of food insecurity. We assessed maternal depression at mid-pregnancy (mean 28 weeks' gestation) and 6 months postpartum with the Edinburgh Postnatal Depression Scale (score > = 13 on 0-30 scale indicating probable depression). Child outcomes at age 3 were height-for-age z-score (HAZ) and leg length. HAZ was also available at birth and ages 6 months, 1, 2, and 3 years.
Seventy (8.0%) women experienced antenatal depression and 64 (7.3%) experienced postpartum depression. The mean (SD) height for children age 3 was 97.2 cm (4.2), with leg length of 41.6 cm (2.6). In multivariable linear regression models, exposure to postpartum depression was associated with greater HAZ (0.37 [95% confidence interval: 0.16, 0.58]) and longer leg length (0.88 cm [0.35, 1.41]). The relationship between postpartum depression and greater HAZ was evident starting at 6 months and continued to age 3. We found minimal relationships between antenatal depression and child height outcomes.
Our findings do not support the hypothesis that maternal depression is associated with reduced height in children in this relatively advantaged sample in a high-income country.
- SourceAvailable from: nih.gov[show abstract] [hide abstract]
ABSTRACT: Height is inversely associated with cardiovascular disease mortality risk and has shown variable associations with cancer incidence and mortality. The interpretation of findings from previous studies has been constrained by data limitations. Associations between height and specific causes of death were investigated in a large general population cohort of men and women from the West of Scotland. Prospective observational study. Renfrew and Paisley, in the West of Scotland. 7052 men and 8354 women aged 45-64 were recruited into a study in Renfrew and Paisley, in the West of Scotland, between 1972 and 1976. Detailed assessments of cardiovascular disease risk factors, morbidity and socioeconomic circumstances were made at baseline. Deaths during 20 years of follow up classified into specific causes. Over the follow up period 3347 men and 2638 women died. Height is inversely associated with all cause, coronary heart disease, stroke, and respiratory disease mortality among men and women. Adjustment for socioeconomic position and cardiovascular risk factors had little influence on these associations. Height is strongly associated with forced expiratory volume in one second (FEV1) and adjustment for FEV1 considerably attenuated the association between height and cardiorespiratory mortality. Smoking related cancer mortality is not associated with height. The risk of deaths from cancer unrelated to smoking tended to increase with height, particularly for haematopoietic, colorectal and prostate cancers. Stomach cancer mortality was inversely associated with height. Adjustment for socioeconomic position had little influence on these associations. Height serves partly as an indicator of socioeconomic circumstances and nutritional status in childhood and this may underlie the inverse associations between height and adulthood cardiorespiratory mortality. Much of the association between height and cardiorespiratory mortality was accounted for by lung function, which is also partly determined by exposures acting in childhood. The inverse association between height and stomach cancer mortality probably reflects Helicobacter pylori infection in childhood resulting in--or being associated with--shorter height. The positive associations between height and several cancers unrelated to smoking could reflect the influence of calorie intake during childhood on the risk of these cancers.Journal of Epidemiology & Community Health 03/2000; 54(2):97-103. · 3.39 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The purpose of this study was to analyze the association of adult height with cause-specific and total mortality. The study included 31,199 men and women aged 25-64 years who participated in a risk factor survey in 1972, 1977, 1982, or 1987 in eastern Finland. The cohorts were followed until the end of 1994. The relation between height and mortality was assessed by using Cox proportional hazard models. The authors found that height was associated inversely with most of the measured risk factors and directly with socioeconomic status. For both genders, height was inversely associated with cardiovascular and total mortality; the age- and birth-cohort-adjusted risk ratios per 5 cm increase in height were 0.89 and 0.91 for men and 0.86 and 0.90 for women, respectively. The inverse association also remained after adjustment for the other known risk factors. For men, an independent inverse association also was found between height and mortality from chronic obstructive pulmonary disease and from violence and accidents. Cancer mortality was not associated with height. Thus, genetic factors, and environmental factors during the fetal period, childhood, and adolescence, which determine adult height, appear to be related to a person's health later in life.American Journal of Epidemiology 07/2000; 151(11):1112-20. · 4.78 Impact Factor
- Acta medica Scandinavica. Supplementum 02/1984; 679:1-56.
Maternal Depressive Symptoms Not Associated with
Reduced Height in Young Children in a US Prospective
Karen A. Ertel1*, Karestan C. Koenen1,2,3, Janet W. Rich-Edwards2,4, Matthew W. Gillman5,6
1Department of Society, Human Development, and Health & Kellogg Health Scholars Program, Harvard School of Public Health, Boston, Massachusetts, United States of
America, 2Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts, United States of America, 3Center on the Developing Child, Harvard
University, Cambridge, Massachusetts, United States of America, 4Center for Women’s Health and Gender Biology, Brigham and Women’s Hospital, Boston,
Massachusetts, United States of America, 5Obesity Prevention Program, Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care
Institute, Boston, Massachusetts, United States of America, 6Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts, United States of America
Background: Shorter stature is associated with greater all cause and heart disease mortality, but taller stature with increased
risk of cancer mortality. Though childhood environment is important in determining height, limited data address how
maternal depression affects linear growth in children. We examined the relationships between antenatal and postpartum
depressive symptoms and child height and linear growth from birth to age 3 years in a U.S. sample.
Methods: Subjects were 872 mother-child pairs in Project Viva, a prospective pre-birth cohort study. The study population is
relatively advantaged with high levels of income and education and low risk of food insecurity. We assessed maternal
depression at mid-pregnancy (mean 28 weeks’ gestation) and 6 months postpartum with the Edinburgh Postnatal
Depression Scale (score .=13 on 0–30 scale indicating probable depression). Child outcomes at age 3 were height-for-age
z-score (HAZ) and leg length. HAZ was also available at birth and ages 6 months, 1, 2, and 3 years.
Findings: Seventy (8.0%) women experienced antenatal depression and 64 (7.3%) experienced postpartum depression. The
mean (SD) height for children age 3 was 97.2 cm (4.2), with leg length of 41.6 cm (2.6). In multivariable linear regression
models, exposure to postpartum depression was associated with greater HAZ (0.37 [95% confidence interval: 0.16, 0.58])
and longer leg length (0.88 cm [0.35, 1.41]). The relationship between postpartum depression and greater HAZ was evident
starting at 6 months and continued to age 3. We found minimal relationships between antenatal depression and child
Conclusion: Our findings do not support the hypothesis that maternal depression is associated with reduced height in
children in this relatively advantaged sample in a high-income country.
Citation: Ertel KA, Koenen KC, Rich-Edwards JW, Gillman MW (2010) Maternal Depressive Symptoms Not Associated with Reduced Height in Young Children in a
US Prospective Cohort Study. PLoS ONE 5(10): e13656. doi:10.1371/journal.pone.0013656
Editor: Landon Myer, University of Cape Town, South Africa
Received May 31, 2010; Accepted September 18, 2010; Published October 27, 2010
Copyright: ? 2010 Ertel et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Project Viva is supported by grants from the National Institutes of Health (HD 34568, HL 64925, HL 68041). The funders had no role in study design,
data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com
Extant literature largely from the United States and Western
European nations reveals that body height is strongly associated
with health outcomes. Shorter stature is associated with increased
risk of mortality,[1–5] as well as mortality due to cardiovascu-
lar[1,2] and respiratory disease.[1,4] At the other end of the
spectrum, tall individuals have greater risk for cancer and
mortality due to cancer.[4,6–9] The underlying reasons for these
associations are unknown, but two possible mechanisms are that
adult height reflects growth, nutrition, and the social environment
in early life and that height in adulthood is positively associated
with social class and its attendant exposures. Leg length, a
component of height, may be a more sensitive marker of
prepubertal environmental influences than total height because
height increases in childhood are due more to leg growth than
trunk growth. Similar to total height, leg length is inversely
associated with cardiovascular risk factors and coronary heart
disease,[6,7,11–13] but positively associated with cancer risk.
Height in childhood is strongly correlated with adult height: at
age 2 years, length or height has a correlation of 0.8 with adult
height. Similar to attained height in adulthood, accelerated
linear growth in childhood has been positively associated with
breast cancer risk.[15,16] Additionally, a recent analysis of over
3,000 young girls in the United States revealed that greater height
in childhood (ages 4–5 years) predicted early menarche, which
The current literature regarding determinants of height focuses
on developing countries. This is critically important, as nutritional
PLoS ONE | www.plosone.org1 October 2010 | Volume 5 | Issue 10 | e13656
status, disease, and infrastructure have significant influences on
attained height in this context. However, the evidence reviewed
above for the associations between height and adult health
outcomes comes largely from the United States or Western
European nations, thus we believe there is value in continued
investigation of determinants of linear growth in the context of
developed countries, such as the United States. Though many
factors contribute to final height, including genetics, nutrition, and
disease, less attention has been paid to potential social and
psychosocial influences on linear growth. Maternal mental health is
a primary determinant of the child’s psychosocial environment.
Maternal depression is common in the postpartum period, with an
estimated 14%of new mothers experiencing depressionin the first 6
months after delivery. Given its prevalence and its documented
effect on child environment,[22,23] mother-child interaction,
[24–26] and child physiology,[27–29] maternal depression may
be an important determinant of linear growth in young children.
Previous studies using underweight or non-organic failure to
thrive as indicators of poor physical growth have suggested an
association between maternal depression and inadequate growth,
but results have not consistently supported this association
[30–37](for a review of recent findings, see Stewart et al,
2007). The literature specific to height is smaller and largely
in developing countries; these studies provide mixed support for an
association between postpartum depression and reduced child
linear growth. Studies in India, Bangladesh, Brazil,
and Nigeria have shown an association between maternal
depressive symptoms and child stunting, while studies in South
Africa and Brazil have shown no significant difference in
child height according to maternal depressive symptoms. Similar
methodology was used in four developing countries to examine
maternal mental health (not depression per say) and child stunting;
they found an independent, cross-sectional association between
maternal ‘common mental disorder’ and stunting in India, but not
Ethiopia, Peru, or Vietnam. It is difficult to compare across
studies due to different measures of maternal depression and ages
of assessment of children. Further complicating interpretation is
the possibility of reverse causation, as most of these studies were
Antenatal depression, through its effects on fetal growth, may
also influence child stature. Antenatal depression, with a period
prevalence of 18% during pregnancy, is associated with low
birth weight, preterm birth, and small for gestational age at
birth. Infants born small tend to remain small in height as well
as weight through childhood, suggesting that an adverse
intrauterine environment may contribute to short stature. A
prospective cohort study in India reported 3.2 and 2.8 increased
odds of stunting at 6 months and 12 months, respectively, for
children whose mothers were depressed in the 3rdtrimester of
pregnancy compared to children whose mothers were not
depressed at that time.
Based on this prior literature, we hypothesized that children
whose mothers experienced antenatal or postpartum depression
would be shorter than children of mothers without depression.
This study is among the first to examine these associations in a
high-income country. Additionally, the sample population for this
study differs from those in which an association has been found
between maternal depression and stunting in several important
ways, including higher levels of family income, access to
sophisticated health services, low prevalence of food insecurity
and hunger, low rates of infection, high levels of education, and
low rates of both depression and stunting. Thus, this analysis may
represent a rather stringent test of the association between
maternal depression and child height.
Sample and data collection
Subjects were from Project Viva, a prospective cohort study of
pregnant women and their singleton children. Participants were
recruited at their first prenatal visit from eight obstetric practices of
Harvard Vanguard Medical Associates, a large group practice in
the greater Boston area. Exclusion criteria included multiple
gestation, inability to answer questions in English, plans to move
out of the area before delivery, and gestational age more than 22
weeks at first prenatal visit. 2670 pregnant women (64% of those
eligible) were enrolled between April 1999 and July 2002; 329
subsequently became ineligible (60% because they were no longer
pregnant). Of the remaining 2341 enrolled women, 195 withdrew
and 18 were lost to follow-up before delivery. 2128 participants
delivered a live infant and were eligible for inclusion in the current
analysis. 1635 (70%) completed a mid-pregnancy questionnaire
that assessed depressivesymptoms and 1422 (60% of thoseenrolled)
completed a second questionnaire that assessed depressive symp-
toms at 6 months postpartum. 1249 women (53% of those enrolled)
had data on both antenatal and postpartum depression. Data were
and 3 years after delivery. Among participants with depressive
symptom data at mid-pregnancy and 6 months postpartum, 928
had height data for children at age 3 years. Of these, we excluded
56 (6.0%) due to missing covariate information for a final sample
size of 872. Comparison of the 872 subjects included in this analysis
to those excluded revealed some differences: included mothers were
more likely to be white (79.6% vs. 74.3%), more highly educated
(38.4% vs. 22.1% with a graduate degree), slightly older (mean age
33 years vs. 31 years), and slightly taller (mean height 65.1 vs. 64.7
inches). The Project Viva study and consent procedures were
approved by human subjects committees of Harvard Pilgrim
Health Care, Brigham and Women’s Hospital, and Beth Israel
Deaconess Medical Center. Written informedconsent wasobtained
from all mothers in the study.
with the 10-item Edinburgh Postnatal Depression Scale (EPDS) in
approximately 6 months after birth. We chose the EPDS
because it has been validated for antenatal and postpartum
use. We used the cutpoint of 13 or more (on the 0–30 point
scale) to indicate probable depression, consistent with previous
work in our cohort [50,51] and in other large cohorts that
collected EPDS data antenatally and postnatally. The cut-
point at 13 or more indicates probable depression with a sensitivity
of 86% and specificity of 78% in the postnatal period. The
optimal cut-off for probable major depression in the antenatal
period may be higher (15 or more),[53,54] thus we report
multivariable results using this alternate cut-off. It should be noted
that the EPDS is a screening tool that measures probable
depression, and is not a clinical diagnosis of depression;
however, to be succinct, we refer to an EPDS score $13 as
antenatal or postpartum depression.
Child anthropometric measurements.
height was assessed at birth, 6 months, 1 year, 2 years, and 3 years
measurements at birth, 6 months, and 3 years. We obtained
measurements for one and two years of age from clinical records.
In a previous study of patients in this group practice, Rifas-Shiman
and colleagues found that clinical assessment of length or height in
children under age 2 years systematically underestimated height
We assessed depressive symptoms
weeks’gestation) and at
Child length or
research staff completed
Depression and Child Height
PLoS ONE | www.plosone.org2October 2010 | Volume 5 | Issue 10 | e13656
compared to research standard methods, and they developed a
regression equation to correct for this bias. In the analyses
presented here, we employed the corrected lengths for clinical
measurements of children at 1 and 2 years. At age 3, staff also
measured sitting height. We calculated leg length at age 3 as the
difference between standing height and sitting height. Length/
height was measured to the nearest 0.1 centimeter. Details
regarding the age-three assessment have been published. We
calculated age- and sex-specific height-for-age z-score (HAZ)
according to US national reference data for birth and ages six
months, 1, 2, and 3 years. We also calculated age- and sex-
specific height-for-age z-score for age 3 years according to World
Health Organization (WHO) Child Growth Standards, using
the SAS program provided by WHO (available at http://www.
who.int/childgrowth/software/en/, accessed 9/2/2010).
Sociodemographic factors and potential confounders.
chose variables that have been previously linked with perinatal
depression or child height as covariates, see Table 1. Maternal
age, race/ethnicity, household income, education, partnership
status, height, pre-pregnancy BMI, and maternal report of the
father’s height were recorded at enrollment. We assessed
maternal smoking during pregnancy by questionnaire and
categorized responses into former smoker, never smoker, and
smoked during pregnancy. We calculated pregnancy weight gain
as pre-pregnancy weight subtracted from the last clinically
recorded weight before delivery and categorized it according to
the Institute of Medicine’s gestational weight gain guidelines.
Presence of gestational diabetes or impaired glucose tolerance
during pregnancy was collected from medical records. We
identified women who used medication for depression during
pregnancy by searching prescription data for the 90 days before
the last menstrual period through delivery for 29 commonly used
antidepressants. We calculated gestational age at delivery from
the last menstrual period; if estimated gestational age based on last
menstrual period differed by more than 10 days from the 2nd
trimester ultrasound estimate, we used the estimate from the
ultrasound. We calculated birthweight for gestational age z-scores
with use of US national reference data.
We considered the following to be
potential mediators of antenatal and/or postpartum depression’s
impact on child height: subsequent depression (at 6 months and 1
year for antenatal depression and at 1 year for postpartum
depression, each assessed by the EPDS), duration of breastfeeding,
and age at which solid foods were introduced to the child, all of
which were assessed by questionnaire. There was little overlap
between depressive symptoms occurring in the perinatal period
with depressive symptoms at 1 year postpartum: 18 subjects had
elevated depressive symptoms at both 6 months and 1 year
postpartum and 11 at both antenatal and 1 year postpartum.
Additionally, we considered gestational age at birth and
birthweight for gestational age potential mediators of the
influence of antenatal depression on child stature.
We examined two sets of outcomes: HAZ and leg length at age 3
years, and change in HAZ from birth through age 3. In both cases,
we separately modeled the impact of antenatal (EPDS.=13 and
EPDS.=15) and postpartum depression. For HAZ and leg length
at age 3, we used multivariable linear regression models. In the base
model (Model 1) for each exposure, we adjusted for child sex and
age in months at assessment. In Model 2, we adjusted for
sociodemographic and physiological factors related to maternal
depression and/or child height. In Model 3, we additionally
adjusted for gestational age at birth and birthweight for gestational
age z-score; these variables may be mediators of the association of
antenatal depression with height as well as confounders of the
association of postpartum depression with height. To increase
precision of the estimate of interest,if removal of a covariate did not
substantially alter the effect estimate of the primary exposure, that
covariate was not included in the models presented here.
Covariates tested and not included were maternal marital status,
education, smoking during pregnancy, gestational diabetes or
impaired glucose tolerance during pregnancy, paternal height, and
anti-depressant use during pregnancy. Our primary outcomes were
HAZ accordingto CDC referencestandardsand leg length;we also
ran our models using HAZ according to WHO reference standards
because the CDC and WHO standards are based on different
reference populations. If results are consistent using both standards,
we have some assurance that estimated associations are not due to
the choice of reference population.
Seventy mothers experienced depression during pregnancy and
64 at 6 months postpartum. Only 19 mothers experienced
depression at both time points. Due to the relative lack of overlap
in these exposures, we examined antenatal and postpartum
depression separately. Estimated effects for antenatal and postpar-
tum depression were similar whether we included these 19 subjects
or not; thus they are included in both exposure categories.
Research-quality measurements of birth length were available in
a subset of the study population (n=516). In this subset, we found
that controlling for birth length z-score in place of birthweight for
gestational age z-score (available for the whole cohort) did not
affect results, so we present results for the larger cohort only, and
use birthweight for gestational age z-score as a control variable.
Also in this subset of the study population, we conducted a direct
test of reverse causation for the association between postpartum
depression and height outcomes at age 3: we examined whether
length at birth predicted postpartum depression, controlling for
maternal age, race/ethnicity, height, pregnancy weight gain,
household income, and child gender.
Potential mediators of the effect of antenatal depression were
gestational age at birth, birthweight for gestational age,
breastfeeding duration, age of introduction of solid foods, and
maternal depression at 6 months and 1 year postpartum. Potential
mediators of the association of postpartum depression with height
were breastfeeding duration, age of introduction of solid foods,
and depression at 1 year postpartum. To test for mediation, we
added potential mediators to the fully-adjusted models and
assessed the magnitude of attenuation of the estimated effect of
exposure to maternal depression.
For the outcome change in HAZ from birth to 3 years, we used
longitudinal models allowing a random intercept, random slope,
and an unstructured covariance matrix. Covariates were based on
the fully-adjusted models presented in Tables 2 and 3. In models
with antenatal depression as primary predictor, we modeled the
change in HAZ from birth through age 3. In models with
postpartum depression as primary predictor, we modeled the
change in HAZ from age 6 months to 3 years. We entered time as
a categorical variable, with each outcome assessment (birth, 6
months, 1, 2, and 3 years) as a category and tested for an
interaction between maternal depression and time. All analyses
employed SAS version 9.1 (SAS Institute, Cary, NC), proc glm for
age 3 outcomes and proc mixed for longitudinal analyses.
Mothers in this sample were predominantly white, married, and
had an annual household income over $70,000 (Table 1). Seventy
(8.0%) participants experienced antenatal depression and 64
Depression and Child Height
PLoS ONE | www.plosone.org3October 2010 | Volume 5 | Issue 10 | e13656
(7.3%) experienced postpartum depression. Women with antenatal
depression were less likely to be married and more likely to have
lower income, inadequate weight gain during pregnancy, and
gestational diabetes or impaired glucose tolerance. Women with
postpartum depression were younger, less likely to be married, had
lower income, and their children were taller and had longer leg
length at age 3.
In multivariable models predicting age 3 outcomes, there was
some indication that exposure to antenatal depression was
associated with greater leg length but not HAZ (Table 2). In a
model controlling for potential confounders (Model 2), antenatal
depression (EPDS.=13) was associated with (0.45 (20.07, 0.97))
greater leg length. Use of the more stringent definition of antenatal
depression (EPDS.=15) revealed a stronger association between
antenatal depression and leg length in children (0.74 (0.07, 1.40)).
Further control for depression at 6 months and 1 year postpartum
did not materially alter results.
Postpartum depression was associated with greater HAZ and leg
length. As Table 3 displays, after controlling for potential
confounders (Model 3), compared to children not exposed to
postpartum depression, exposed children had HAZ 0.37 (0.16,
0.58) and leg length 0.88 cm (0.35, 1.41) larger. Maternal height,
birthweight for gestational age, non-white race/ethnicity, and
lower household income were other factors associated with taller
height in children (Table 4). The finding that black children of this
age are taller is consistent with findings from a nationally-
Table 1. Characteristics of Project Viva Participants According to Maternal Antenatal and Postpartum Depression. Data from 872
mother-child pairs participating in Project Viva.
Overall (n=872) Antenatal DepressionPostpartum Depression
No (n=802)Yes (n=70)No (n=808) Yes (n=64)
mean/n SD/% mean/nSD/% mean/n SD/%mean/nSD/%mean/nSD/%
Age33.04.533.04.432.1 5.033.1 4.431.6 5.4
Black 73 8.464 8.0912.9 65 8.0812.5
Other 10512.0 9211.5 1318.6 9411.6 1117.9
White 69479.6 64680.6 4868.6 64980.3 45 70.3
Less than BA/BS18921.7 16921.120 28.6169 20.920 31.3
BA/BS34839.9 32240.226 37.132740.5 2132.8
Graduate degree33538.4 31138.8 2434.3 31238.6 2335.9
Married * 78490.0 72890.9 5680.0 73390.8 5179.7
,=$40,000 778.8 61 7.616 22.963 7.81421.9
40,001 to $70,00019522.4 176 22.0 19 27.117621.8 1926.7
.=$70,001 60068.8565 70.5 3550.0 56970.4 3148.4
Pre-pregnancy BMI .=25288 33.0 259 32.32941.4 260 32.2 2843.8
Height (inches) 65.12.7 65.1 2.764.62.4 65.12.7 64.82.9
Paternal height (inches) 70.73.0 70.7 3.070.6 3.070.72.9 70.53.3
BF duration (months) * 6.554.5 6.50 4.57.11 4.76.57 4.5 6.314.3
Introduction of solid foods *
before 4 months 130 14.9 12515.65 7.1 11914.8 11 17.2
4–5 months 62371.5566 70.757 81.457871.6 4570.3
6 months or after 11813.611013.7811.4 11013.6812.5
Girl 456 52.341752.039 55.7 42752.92945.3
GA at delivery (weeks) 39.61.739.61.7 39.5 1.539.6 1.739.5 1.6
BW for GA (z value)0.23 0.930.25 0.93 0.04 0.950.240.92 0.131.07
Child characteristics at 3 yrs
Age (months) 39.0 3.239.03.238.8 2.839.03.238.6 2.6
Height (cm)97.24.297.2 4.2 97.14.697.14.298.6 4.0
HAZ0.20.9 0.20.9 0.21.0 0.20.9 0.60.9
Leg Length (cm) 41.6 2.6 41.62.542.02.8 41.62.6 42.4 2.3
BA/BS, Bachelor of Arts or Science; BMI, body mass index (kg/m2); BF, breastfeeding; GA, gestational age; BW, birth weight; HAZ, height-for-age z-score.
*Sample sizes are slightly different due to missing data: breastfeeding duration n=869; marital status n=871; introduction of solid foods n=871.
Depression and Child Height
PLoS ONE | www.plosone.org4 October 2010 | Volume 5 | Issue 10 | e13656
representative sample of young children. Restriction to
individuals without antenatal depression did not materially alter
results, nor did further control for maternal depression at 1 year
postpartum. There was no evidence for mediation by breastfeed-
ing duration or age of introduction of solid foods, though longer
duration of breastfeeding was an independent predictor of reduced
HAZ and leg length. Each month of longer duration of
breastfeeding was associated with a 20.02 (20.03, 20.01)
difference in HAZ and a 20.05 cm (20.08, 20.01) difference in
leg length. Additionally, there was no evidence of reverse
causation: birth length was not a significant predictor of
postpartum depression (OR=1.39, 95% CI: 0.85, 2.27) in a
logistic regression model controlling for maternal age, race, height,
pregnancy weight gain, household income, and child gender.
Use of WHO reference standards revealed similar results as
those displayed in Table 2 and Table 3. The estimated association
between antenatal depression and WHO height-for-age z-score
was 20.01 (20.27, 0.24) in a model controlling for maternal age,
race/ethnicity, household income, height, and pre-pregnancy
weight gain (the same covariates as in Model 2 of Table 2). The
estimated association between postpartum depression and WHO
HAZ was 0.28 (0.02, 0.54) controlling for the same covariates as in
Model 3 of Table 3.
In longitudinal models, the association between postpartum
depression and child HAZ from 6 months through age 3 was
similar for each age (Figure 1), thus we did not include a
postpartum depression by age interaction term in the final model.
The association between postpartum depression and child HAZ
was 0.29 (0.11, 0.47) at each time point. Similar to age 3 findings,
antenatal depression was not associated with HAZ from birth
through age 3: the estimated association over all time points was
20.01 (20.19, 0.17).
This study did not find evidence for the hypothesis that
maternal depression is associated with reduced height in children.
Contrary to expectations, maternal postpartum depression was
associated with greater child total height and leg length in this
relatively privileged sample in a high income country. Postpartum
depression was associated with 0.37 higher HAZ, which translates
into an approximately 1.5 centimeter difference in 3-year-olds of
average height (94.5 cm) in the US. This finding was
independent of potential confounders, including maternal socio-
demographic factors and height as well as child size at birth. We
found no evidence for mediation by duration of breastfeeding or
age of introduction of solid foods, nor any evidence for reverse
causation. Additional control for depression at 1 year postpartum
did not diminish the estimated effect for postpartum depression at
6 months, nor was there a main effect of depressive symptoms at 1
year postpartum on height outcomes, suggesting that exposure to
maternal depression earlier in life is more important.
Antenatal depression when defined by the more stringent
criteria of EPDS.=15 was associated with greater leg length in
children. Comparing results using alternate cut-off values for the
EPDS suggests a dose-response relation between antenatal
depressive symptoms and linear growth outcomes in children.
Our results contrast with several studies in developing countries
that report either a cross-sectional or prospective association
between maternal depressive symptoms or clinical depression and
reduced child height or increased risk of stunting.[39–42] However,
not all studies in developing countries support an association
Table 2. Difference in 3-year height outcomes among children exposed versus not exposed to antenatal depression, using 2
alternate cut-off values for the Edinburgh Postnatal Depression Scale (EPDS).
EPDS. .=13 EPDS. .=15
Estimate (95% confidence interval) Estimate (95% confidence interval)
Height-for-age z-scoreLeg Length (cm)Height-for-age z-score Leg Length (cm)
Model 10.002 (20.22, 0.23) 0.48 (20.08, 1.04)0.07 (20.21, 0.36)0.63 (20.08, 1.34)
Model 20.01 (20.20, 0.22)0.45 (20.07, 0.97)0.13 (20.14, 0.39) 0.74 (0.07, 1.40)
Model 30.03 (20.17, 0.24)0.48 (20.03, 1.00)0.16 (20.10, 0.42)0.79 (0.14, 1.44)
20.04 (20.25, 0.17) 0.32 (20.20, 0.84)0.06 (20.21, 0.32)0.58 (20.09, 1.24)
20.02 (20.23, 0.19)0.36 (20.16, 0.89) 0.10 (20.17, 0.37)0.68 (0.01, 1.35)
Model 1 covariates: child sex and age at 3-year assessment.
Model 2: Model 1+ maternal: age, race/ethnicity, household income, height, pregnancy weight gain.
Model 3: Model 2+ gestational age at birth and birthweight for gestational age z-value.
Model 4: Model 3+ postpartum depression.
Model 5: Model 4+ breastfeeding duration and age of introduction of solid foods (n=869).
Data from 872 mother-child pairs participating in Project Viva.
Table 3. Difference in 3-year height outcomes among
children exposed versus not exposed to postpartum
Estimate (95% confidence interval)
Height-for-age z-score Leg Length (cm)
Model 10.39 (0.16, 0.62)0.97 (0.39, 1.56)
Model 20.35 (0.13, 0.57)0.85 (0.31, 1.39)
Model 30.37 (0.16, 0.58)0.88 (0.35, 1.41)
Model 40.37 (0.16, 0.58)0.89 (0.36, 1.41)
Model 1 covariates: child sex and age at 3-year assessment.
Model 2: Model 1+ maternal: age, race/ethnicity, household income, height,
pregnancy weight gain.
Model 3: Model 2+ gestational age at birth and birthweight for gestational age
Model 4: Model 3+ breastfeeding duration and age of introduction of solid
Data from 872 mother-child pairs participating in Project Viva.
Depression and Child Height
PLoS ONE | www.plosone.org5 October 2010 | Volume 5 | Issue 10 | e13656