Association between PM exposure and childhood obesity

1
January 2020, Volume 8, Issue 1, Number 17
Maryam Bahreynian1 , Marjan Mansourian2 , Nafiseh
Mozaffarian3 , Parinaz Poursafa4 , Mehri Khoshhali3* , Roya
Kelishadi3
Review Article:
The Association Between Exposure to Ambient Particulate
Matter and Childhood Obesity: A Systematic Review and
Meta-analysis
Context: Physical environment contamination and in particular,
air pollution might cause long-term
adverse effects in child growth and a higher risk of catching
non-communicable diseases later in life.
Objective: This study aimed to overview the human studies on
the association of exposure to
ambient Particulate Matter (PM) with childhood obesity.
Data Sources: We systematically searched human studies
published until March 2018 in PubMed,
Scopus, Ovid, ISI Web of Science, Cochrane library, and
Google Scholar databases.
Study Selection: All studies that explored the association
between PM exposure and childhood
obesity were assessed in the present study, and finally, 5 studies
were used in the meta-analysis.

Data Extraction: Two independent researchers performed the
data extraction procedure and
quality assessment of the studies. The papers were qualitatively
assessed by STROBE (Strengthening
the Reporting of Observational studies in Epidemiology)
statement checklist.
Results: The pooled analysis of PM exposure was significantly
associated with increased Body Mass
Index (BMI) (Fisher’s z-distribution=0. 028; 95% CI=0. 017, 0.
038) using the fixed effects model. We
also used a random-effect model because we found a significant
high heterogeneity of the included
studies concerning the PM (I2=94. 4%; P<0. 001). PM exposure
was associated with increased BMI
(Fisher’s z-distribution=0. 022; 95% CI=-0. 057, 0. 102).
However, the overall effect size was not
significant, and heterogeneity of the included studies was
similar to the fixed effect model.
Discussion: Our findings on the significant association between
PM10 exposure and the increased
BMI (r=0. 034; 95%CI=0. 007, 0. 061) without heterogeneity
(I2=16. 6%, P=0. 274) (in the studies with
PM10) suggest that the PM type might account for the
heterogeneity among the studies.
Conclusion: The findings indicate that exposure to ambient
PM10 might have significant effects on
childhood obesity.
A B S T R A C T
Key Words:
Air pollution, Particulate
matter, Childhood

obesity, Meta-analysis
Article info:
Received: 10 Oct 2018
First Revision: 23 Feb 2019
Accepted: 09 Mar 2019
Published: 01 Jan 2020
1. Department of Nutrition Child Growth, and Development
Research Center, Research Institute for Primordial Prevention
of Non-communicable Dis-
eases, Student Research Committee, Isfahan University of
Medical Sciences, Isfahan, Iran.
2. Department of Biostatistics and Epidemiology, School of
Health, Isfahan University of Medical Sciences, Isfahan, Iran.
3. Department of Pediatrics, Child Growth, and Development
Research Center, Research Institute for Primordial Prevention
of Non-communicable Dis-
eases, Isfahan University of Medical Sciences, Isfahan, Iran.
4. Environment Research Center, Research Institute for
Primordial Prevention of Non-communicable Diseases, Isfahan
University of Medical Sciences, Isfahan, Iran.
* Corresponding Author:
Mehri Khoshhali, MD.
Address: Department of Pediatrics, Child Growth and
Development Research Center, Research Institute for Primordial
Prevention of Non-communicable Dis-
eases, Isfahan University of Medical Sciences, Isfahan, Iran.
Tel: +98 (311) 7925215
E-mail: m. [email protected] com
Citation Bahreynian M, Mansourian M, Mozaffarian N, Poursafa
P, Khoshhali M, Kelishadi R. The Association Between
Exposure
to Ambient Particulate Matter and Childhood Obesity: A

Systematic Review and Meta-analysis. Journal of Pediatrics
Review. 2020;
8(1):1-14. http://dx. doi. org/10. 32598/jpr. 8. 1. 1
: http://dx. doi. org/10. 32598/jpr. 8. 1. 1
Use your device to scan
and read the article online
https://orcid.org/0000-0001-9832-0106
https://orcid.org/0000-0002-7217-0282
https://www.ncbi.nlm.nih.gov/pubmed/?term=Mozaffarian%20N
%5BAuthor%5D&cauthor=true&cauthor_uid=29442253
https://orcid.org/0000-0002-8113-6721
https://orcid.org/0000-0002-8067-4122
https://orcid.org/0000-0002-2883-8652
https://orcid.org/0000-0001-7455-1495
http://jpr.mazums.ac.ir/page/69/Open-Access-Policy
https://www.ncbi.nlm.nih.gov/pubmed/?term=Mozaffarian%20N
%5BAuthor%5D&cauthor=true&cauthor_uid=29442253
http://dx.doi.org/10.32598/jpr.8.1.1
https://crossmark.crossref.org/dialog/?doi=10.32598/jpr.8.1.1
http://jpr.mazums.ac.ir/page/69/Open-Access-Policy
2
1. Context
hildhood obesity is a growing public health
problem, even in developing countries (1, 2). It
is associated with several health complications
during childhood, which will usually extend to
adulthood (3). It has several underlying causes,

both genetic and environmental factors (4, 5).
Recently, researchers have paid attention to the as-
sociation between air pollution and obesity, and some
studies suggest that ambient air pollution may increase
the risk of catching Non-communicable Diseases (NCDs)
in adults, diseases such as cardiovascular diseases, dia-
betes and cancer (6-8). However, little epidemiological
evidence is available on the association of exposure to
ambient air pollution with the development of child-
hood obesity (9-11). Physical environment contamina-
tion and in particular, air pollution might cause long-
term adverse effects in child growth and a higher risk of
developing NCDs later in life. The “Obesogenic Environ-
ment” hypothesis discusses the impact of environmen-
tal chemicals with endocrine disruption properties that
can change child growth patterns and result in weight
gain, obesity, and obesity-related NCDs (12).
Some previous studies reveal a positive association
between exposure to Polycyclic Aromatic Hydrocar-
bons (PAHs) and childhood obesity (11, 13). A recent
study conducted in China reports that long-term ex-
posure to air pollutants, including Particulate Matter
(PM)
10
, NO
2
, SO
2
, and O
3

is associated with higher risk
of childhood obesity and hypertension (14). More-
over, the association of residential traffic density and
roadway proximity with rapid infant weight gain and
childhood obesity has been documented in some pre-
vious studies (15-18). A study on Latino children living
in the US shows that higher exposure to NO
2
and PM
2.
5
is related to higher Body Mass Index (BMI) at the age
of 18 (19). However, some other studies report no as-
sociation between exposure to vehicular traffic and
pollutants and the risk of obesity and dyslipidemia in
children (20, 21). Therefore, the overall evidence on
obesogenic properties of air pollutants is controversial.
2. Objectives
Due to the high prevalence of childhood obesity, its
multifactorial nature, and the importance of conducting
preventive strategies, we aimed to provide a systematic
review and meta-analysis on the association of expo-
sure to ambient PM and childhood obesity.
3. Data Sources
We performed a systematic review and meta-analysis
of human studies that explored the association be-
tween PM exposure and childhood obesity. We con-

sidered PECO as the following: Population (P): Children
and adolescents; Exposure (E): PM exposure; Compari-
son (C) (There is no comparison between exposed and
non-exposed groups because we have reported the cor-
relations of PM exposure and BMI); and Outcome (O):
Childhood obesity (BMI).
We systematically searched human studies available
on the study subject until March 2018 in PubMed, Sco-
pus, Ovid, ISI Web of Science, Cochrane library, and
Google Scholar databases. All cross-sectional and co-
hort studies were selected. We used the search terms of
“Air Pollution” OR “Pollutants” OR “Particulate Matter”
in combination with “Obesity” OR “Weight” OR “Body
Mass Index” OR “BMI” OR “Overweight” OR “Cardio-
metabolic” OR “Metabolic Syndrome” OR “Metabolic
Syndrome X” OR “Mets” OR “Adiposity” AND “Child” OR
“Adolescent” OR “School-aged” OR “Youth” OR “Teen-
ager” OR “Boy” OR “Girl” OR “Student” OR “Pediatrics”
in the form of Medical Subject Headings (MeSH) and
truncations. The relevant articles were examined with-
out any language restriction.
4. Study Selection
After removing the duplicates, the relevant papers
were selected in three phases. In the first and the second
phases, titles and abstracts of papers were screened,
and the irrelevant papers were excluded. In the third
phase, the full texts of the remaining papers were ex-
plored carefully to select only the relevant papers. To
find any additional pertinent study, the reference list of
all reviews and related papers were screened as well.
The included studies had the following criteria: 1.
Observational cross-sectional design; 2. Longitudinal

cohort studies which report the study association; 3.
Measurement of PM concentration as an index for air
pollution exposure; and 4. Reporting the Odds Ratio
(OR), Relative Risk (RR), and β-coefficient of PM with
child obesity. In the final step, all statistics were changed
to the correlation coefficient values.
5. Data Extraction
Two reviewers extracted the data independently using
a data collection form, including the first author’s name,
publication year, sample size, study design, as well as
C
Bahreynian M, et al. Exposure to Ambient Particulate Matter
and Childhood Obesity. Association Between Exposure to
Ambient Particulate. J Pediatr Rev. 2020; 8(1):1-14.
3
age, exposure measurement, statistical analysis, and
the variables adjusted in the analyses.
5. 1. Quality assessment
Two independent reviewers (MB and MKH) evalu-
ated the methodological quality of each study. The
Strengthening the Reporting of Observational studies in
Epidemiology (STROBE) checklist was used for the qual-
ity assessment of the papers. According to STROBE (22
questions), the included studies were divided into three

groups of high, medium, and low-quality. The studies
scored one to eight were ranked as low-quality studies,
9-16 as medium-quality ones, and 17-22 as high-quality
papers. The two reviewers agreed on (80%) of cases.
The remaining discrepancies were resolved by consulta-
tion and consensus.
5. 2. Statistical analysis
The effect sizes of RR, OR, and β-coefficient from all ar-
ticles were extracted directly from the original reports.
All effect sizes were transformed into (r: correlation),
and Fisher z-transformation of the r value was applied
for the pooled analysis (22, 23). The potential heteroge-
neity across studies was evaluated using the Cochran’s
Q test and was expressed using the I2 index. The pooled
results for Fisher z-transformation were calculated by
the fixed-effects model (for low heterogeneity) or the
random-effects model (for high heterogeneity). Publica-
tion bias was evaluated by the Egger’s and the Begg’s
tests. Subgroup analyses and meta-regression were per-
formed to seek the sources of heterogeneity. The sensi-
tivity analyses were performed by omitting one study at
a time to gauge the robustness of our results. All statisti-
cal analyses were conducted in STATA V. 14. 0.
6. Results
We initially retrieved 4391 articles from the databases.
Figure 1 represents the search results. After the initial
study of the titles and abstracts, the duplicate papers
were omitted, and out of 4276 papers, five articles
remained. No additional references were identified
through checking the reference lists of selected papers.
The main characteristics of the studies included in the

systematic review are presented in Appendix 1. Overall,
the studies reported data on 33825 subjects, and they
were published between 2010 and 2018.
6. 1. Meta-analysis of the correlations
Figure 2 showed the pooled results using random ef-
fect model. It showed that PM exposure was associat-
ed with the increased BMI (Fisher-z= 0. 022; 95% CI (-0.
057, 0. 102)) that overall effect size was not significant
and heterogeneity of the included studies was as same
fixed effect model.
Table 1 presents the results of the meta-regression
analysis. The univariate meta-regression analyses in-
dicated that none of the factors, including mean age,
sample size, study location (Europe, Asia, and the USA),
study type (cross-sectional and cohort), and PM type
(2. 5 and 10) contributed to the heterogeneity of meta-
analysis (P>0. 05 for all).
Table 2 presents the results of subgroup analysis ac-
cording to the study location, study type, and PM type.
We observed significant association between PM
10
ex-
posure and the increased BMI (Fisher’s z=0. 034; 95%
CI=0. 007, 0. 061) with no apparent heterogeneity
(I2=16. 6%, P=0. 274) in the studies with PM
10
. It sug-
gests that the PM type may partially account for the het-

erogeneity among the studies on BMI (Figure 3).
Begg’s test and Egger’s test revealed no obvious publi-
cation bias among these studies. The P-values for these
tests were higher than 0. 05 (P=0. 661 and 1. 0, respec-
tively). The results of sensitivity analyses showed that
with excluding the study of Fleisch AF et al. (7. 7 years),
the pooled Fisher’s z for the subgroup PM
2. 5
increased.
Although this change was not significant, it decreased
the overall heterogeneity (I2=83. 1%, P<0. 001) (Figure
4).
Figure 4 Forest plot of Fisher’s z values for the cor-
relation between PM and BMI by PM type after exclud-
ing the study of Fleisch AF et al. (7. 7 years) Table 3
presents the results of converting Fisher’s z values into
correlation values. We found a significant relationship
between PM
10
and BMI (r=0. 034, P=0. 015), but the
association of PM
2. 5
and BMI was not statistically sig-
nificant (r=0. 035, P=0. 606).
7. Discussion
This systematic review and meta-analysis revealed a weak
positive association between ambient PM

10
and child obe-
sity. However, the results for PM
2. 5
was not significant. A few
meta-analysis or large sample size studies have explored
the association of ambient PM with adult obesity or birth
weight, but with childhood obesity (24-26).
4
NOTE: Weights are from random effects analysis
Overall (I-squared = 94.4%, p = 0.000)
Fioravanti S et al (PM2.5) (2018)
Poursafa P et al (2017)
Fleisch AF et al (3.3 years) (2016)
Study
Mao G et al (2017)

Fioravanti S et al (PM 10) (2018)
Dong GH et al (2014)
ID
0.02 (-0.06, 0.10)
0.01 (-0.08, 0.09)
0.41 (0.27, 0.56)
-0.00 (-0.05, 0.05)
0.03 (-0.03, 0.08)
-0.01 (-0.09, 0.07)
-0.20 (-0.25, -0.15)
0.04 (0.03, 0.05)
ES (95% CI)
100.00
13.80
10.56
15.20
%

15.22
13.80
15.20
16.21
Weight
0.02 (-0.06, 0.10)
0.01 (-0.08, 0.09)
0.41 (0.27, 0.56)
-0.00 (-0.05, 0.05)
0.03 (-0.03, 0.08)
-0.01 (-0.09, 0.07)
-0.20 (-0.25, -0.15)
0.04 (0.03, 0.05)
ES (95% CI)
100.00
13.80
10.56
15.20

%
15.22
13.80
15.20
16.21
Weight
0-.557 0 .557
Figure 2. Forest plot of Fisher’s z values indicating the
correlation between PM and BMI

Articles screened by
title and abstract
(n=4276)
Excluded non-
relevant articles
(n=4271)
Full text articles assessed for eligibility and studies
included in the meta-analysis (n=5)
(Two studies reported PM10 and five studies reported
PM2.5 as the indicators of air pollution.
Three studies reported PM2.5, one study reported
PM10 and one study reported both PM2.5 and PM10.
Removed duplicates
articles (n=115)
Articles identified through
electronic database search
(n=4391)
(PubMed: 662; Scopus: 2600; ISI
Web of Science: 1129)
Figure 1. The flowchart of the search results

5
Table 1. Results of meta-regression analyses for the potential
source of heterogeneity
Covariate B SE P 95% CI
Year of publication 0. 016 0. 056 0. 790 (-0. 128, 0. 159)
Mean age 0. 040 0. 030 0. 243 (-0. 038, 0. 118)
PM2. 5 (Ref, PM10) 0. 024 0. 162 0. 889 (-0. 392, 0. 440)
Sample size of study -0. 0000002 0. 000007 0. 974 (-0. 00002,
0. 00002)
Study location: (Ref. : Asia)
USA -0. 264 0. 144 0. 107 (-0. 664, 0. 136)
Europe -0. 207 0. 159 0. 109 (-0. 649, 0. 235)
Study type: Cohort (Ref: Case control) -0. 238 0. 121 0. 107 (-
0. 550, 0. 074)
SE: Standard Error; CI: Confidence Interval
Table 2. Results of subgroup analysis on the association
between PM and BMI
Variables Groups NO. of Study Effect Size (Fisher’ z) 95% CI P
Heterogeneity

I2 (%) P
PM type
10 2 0. 034 (0. 007, 0. 061) 0. 015 16. 60 0. 274
2. 5 3 0. 035 (-0. 099, 0. 169) 0. 606 95. 30 < 0. 001
Study type
Cross-sectional 2 0. 218 (-0. 148, 0. 583) 0. 243 96. 10 < 0. 001
cohort 3 -0. 037 (-0. 132, 0. 057) 0. 442 91. 60 < 0. 001
Study location
Asia 2 0. 218 (-0. 148, 0. 583) 0. 243 96. 10 < 0. 001
Europe 1 -0. 001 (-0. 06, 0. 057) 0. 961 0. 00 0. 818
USA 2 -0. 059 (-0. 2, 0. 083) 0. 416 95. 50 < 0. 001
Figure 3. Forest plot of Fisher’s z values indicating the
correlation between PM and BMI by PM type
.
.

ID
Mao G et al (2017)
PM10
Study
PM2.5
Subtotal (I-squared = 95.3%, p = 0.000)
0.02 (-0.06, 0.10)
-0.20 (-0.25, -0.15)
0.41 (0.27, 0.56)
-0.01 (-0.09, 0.07)
ES (95% CI)
0.03 (-0.03, 0.08)
-0.00 (-0.05, 0.05)

0.01 (-0.08, 0.09)
0.04 (-0.10, 0.17)
0.04 (0.03, 0.05)
0.03 (0.01, 0.06)
100.00
15.20
10.56
13.80
Weight
15.22
15.20
%
13.80
69.99
16.21
30.01
0.02 (-0.06, 0.10)
-0.20 (-0.25, -0.15)

0.41 (0.27, 0.56)
-0.01 (-0.09, 0.07)
ES (95% CI)
0.03 (-0.03, 0.08)
-0.00 (-0.05, 0.05)
0.01 (-0.08, 0.09)
0.04 (-0.10, 0.17)
0.04 (0.03, 0.05)
0.03 (0.01, 0.06)
100.00
15.20
10.56
13.80
Weight
15.22
15.20
%
13.80

69.99
16.21
30.01
0-.557 0 .557
6
These five studies investigated more than 33000 par-
ticipants. The current literature provides conflicting re-
sults on the association between air pollution and child-
hood obesity (19-21, 27). We found a relatively weak
positive relationship between exposure to PM
10
and
childhood BMI, consistent with most previous studies
findings. Five of the seven studies included in the cur-
rent meta-analysis reported the direct association of air
pollution and child weight, whereas two cohort studies
did not report such association (20, 21).
Such discrepancies among these studies results might
be due to confounding factors like age, gender, ethnicity,

physical activity, level of exposures, and some other fac-
tors. These findings might be confounded by heterogene-
ity due to multiple dispersions between studies such as
study design, different techniques to measure PM con-
centration, the way PM levels is reported, and other vari-
ous confounders which were adjusted in the analysis.
Only a few cross-sectional studies have investigated the
relation of air pollution and obesity in children (16, 17, 20,
27-29). In a longitudinal study, higher exposure to NO
2
and
.
.
Study
ID
Mao G et al (2017)

PM2.5
PM10
0.05 (-0.00, 0.10)
0.04 (0.03, 0.05)
ES (95% CI)
0.03 (-0.03, 0.08)
0.41 (0.27, 0.56)
0.03 (0.01, 0.06)
0.09 (-0.02, 0.20)
-0.01 (-0.09, 0.07)
0.01 (-0.08, 0.09)
-0.00 (-0.05, 0.05)
100.00
%
23.11

Weight
19.21
8.64
38.06
61.94
14.95
14.95
19.14
0.05 (-0.00, 0.10)
0.04 (0.03, 0.05)
ES (95% CI)
0.03 (-0.03, 0.08)
0.41 (0.27, 0.56)
0.03 (0.01, 0.06)
0.09 (-0.02, 0.20)
-0.01 (-0.09, 0.07)
0.01 (-0.08, 0.09)
-0.00 (-0.05, 0.05)

100.00
%
23.11
Weight
19.21
8.64
38.06
61.94
14.95
14.95
19.14
0-.557 0 .557
Figure 4. Forest plot of Fisher’s z values, indicating the
correlation between PM and BMI by PM type after excluding
the study of Fleisch
AF et al (7. 7 years)
Table 3. The correlation between PM exposure and BMI
Variables
Effect Size Heterogeneity
Pooled r 95% CI P I2 P τ2

PM10 0. 034 (0. 007, 0. 061) 0. 015 16. 60% 0. 0002 0. 0002
PM2. 5 0. 035 (-0. 099, 0. 167) 0. 606 95. 30% 0. 0216 0. 022
overall 0. 022 (-0. 057, 0. 102) 0. 579 94. 40% 0. 0101 0. 010
τ2: Between-studies variance
7
PM
2. 5
was associated with higher BMI, body fat percent-
age, and abdominal obesity during follow up and at the
age of 18 in children who were overweight or obese at
the study baseline (19). Another study conducted on over-
weight and obese minority youths found that higher expo-
sures to NO
2
and PM
2. 5
during one year before the study
was not associated with obesity, but it was related to low-

er insulin sensitivity and higher acute insulin response to
glucose, which might contribute to obesity (19, 30).
The mechanisms linking air pollution to obesity risk and
type-2 diabetes are not entirely determined. The effects
of air pollutants on immune response, oxidative stress,
and insulin resistance might explain the results (31).
Air pollutants such as PM might increase the infiltra-
tion and activation of immune-competent cells, includ-
ing monocyte and macrophages, in body tissues (32).
Previous findings also indicated that early life exposure
to PM
2. 5
might result in insulin resistance and obesity
later in life, through NAD(P)H oxidase-derived super-
oxide anions, which might cause changes in adipocyte
numbers and size (33).
Animal studies suggest that higher exposure to air pol-
lutants might result in increased adipose tissue inflam-
mation, accumulation of glucose in skeletal muscles,
and therefore it might contribute to metabolic dysfunc-
tion and obesity (34, 35). Furthermore, previous studies
indicate that long-term exposure to combustion-related
air pollutants can increase systemic inflammation and
oxidative stress (34).
Little information is available about the biological basis
of the relationship between exposure to air pollutants and
childhood obesity. There may be a potential for residual
confounders, including socioeconomic status and physi-
cal activity, which can be associated with both air pollu-
tion exposure and children’s weight. Therefore, residual

confounding may affect the study results due to the as-
sociations of poor diet and low physical activity with child
overweight and metabolic disruption. Also, these factors
may be related to residential proximity to sources of air
pollution (36, 37). For example, children living in areas with
higher levels of air pollution usually belong to lower socio-
economic families who often consume higher amounts of
total or saturated fats (27, 38).
Lack of physical activity among the children living in pollut-
ed regions (because of their parental control to restrict the
children’s exposure to air pollution) may be another reason
for the excess weight in children. It is documented that over-
weight children usually have less frequent and shorter peri-
ods of activities compared to their normal weight peers (39,
40). However, the findings on the associations of exposure
to air pollutants and childhood obesity are unlikely to be
confounded by these factors, because many of these studies
had adjusted these associations for socioeconomic status, as
a strong predictor of dietary intake and physical activity (36).
Furthermore, misclassification of exposure to air pollut-
ants might have occurred with residential-based estimates
of pollutant exposure, which might decrease the observed
effects (41). Some studies also lack information about oth-
er potential confounders such as active and passive smok-
ing as well as exposure to noise pollution (19). Previous
studies suggest that tobacco exposure and near roadway
air pollution contribute to synergistic effects on the devel-
opment of child obesity (18).
The findings of the current study concerning the asso-
ciation of exposure to ambient PM with childhood obesity
should be considered with caution. The cross-sectional
design of some studies used for this meta-analysis might

preclude any causality. Another limitation is the high het-
erogeneity between studies. Other potential risk factors
like child physical activity, familial socioeconomic status,
and climate conditions were not available in some studies.
8. Conclusions
This systematic review and meta-analysis indicate
that exposure to ambient PM
10
has a weak positive as-
sociation with childhood obesity. This finding should be
considered in future studies and preventive programs.
Our results are also useful for health policymakers and
health care providers to design health promotion inter-
ventions and preventive strategies. More research is
needed to clarify the effect of other ambient air pollut-
ants on child health status.
Ethical Considerations
Compliance with ethical guidelines
All ethical principles were considered in this article.
Funding
This research did not receive any specific grant from
funding agencies in the public, commercial, or not-for-
profit sectors.
Authors contribution's

8
Conceptualization, methodology, and investigation: All au-
thors; Writing-original draft: Maryam Bahreynian; Writing-
review & editing: Maryam Bahreynian, Roya Kelishadi, Meh-
ri Khoshhali, and Marjan Mansourian; Supervision: Roya
Kelishadi and Mehri Khoshhali.
Conflicts of interest
The authors declared no conflict of interests.
References
1. Caballero B. The global epidemic of obesity: An overview.
Epidemiologic Reviews. 2007; 29(1):1-5. [DOI:10. 1093/
epirev/mxm012] [PMID]
2. Kelishadi R. Childhood overweight, obesity, and the meta-
bolic syndrome in developing countries. Epidemiologic Re-
views. 2007; 29(1):62-76. [DOI:10. 1093/epirev/mxm003]
[PMID]
3. Swinburn BA, Sacks G, Hall KD, McPherson K, Finegood
DT, Moodie ML, et al. The global obesity pandemic:
Shaped by global drivers and local environments. The
Lancet. 2011; 378(9793):804-14. [DOI:10. 1016/S0140-
6736(11)60813-1]

4. McAllister EJ, Dhurandhar NV, Keith SW, Aronne LJ, Barger
J,
Baskin M, et al. Ten putative contributors to the obesity epi-
demic. Critical Reviews in Food science and Nutrition. 2009;
49(10):868-913. [DOI:10. 1080/10408390903372599]
[PMID] [PMCID]
5. Keith SW, Redden DT, Katzmarzyk PT, Boggiano MM,
Hanlon
EC, Benca RM, et al. Putative contributors to the secular
increase in obesity: Exploring the roads less traveled. Inter-
national Journal of Obesity. 2006; 30(11):1585-94. [DOI:10.
1038/sj. ijo. 0803326] [PMID]
6. Wang M, Beelen R, Stafoggia M, Raaschou-Nielsen O, An-
dersen ZJ, Hoffmann B, et al. Long-term exposure to el-
emental constituents of particulate matter and cardiovas-
cular mortality in 19 European cohorts: Results from the
ESCAPE and TRANSPHORM projects. Environment Interna-
tional. 2014; 66:97-106. [DOI:10. 1016/j. envint. 2014. 01.
026] [PMID]
7. Rajagopalan S, Brook RD. Air pollution and type 2 diabetes:
Mechanistic insights. Diabetes. 2012; 61(12):3037-45.
[DOI:10. 2337/db12-0190] [PMID] [PMCID]
8. de Groot P, Munden RF. Lung cancer epidemiology, risk fac-
tors, and prevention. Radiologic Clinics. 2012; 50(5):863-76.
[DOI:10. 1016/j. rcl. 2012. 06. 006] [PMID]
9. Ino T, Shibuya T, Saito K, Inaba Y. Relationship between
body
mass index of offspring and maternal smoking during preg-
nancy. International Journal of Obesity. 2012; 36(4):554-8.
[DOI:10. 1038/ijo. 2011. 255] [PMID]

10. Bolton JL, Smith SH, Huff NC, Gilmour MI, Foster WM,
Auten RL, et al. Prenatal air pollution exposure induces
neuroinflammation and predisposes offspring to weight
gain in adulthood in a sex-specific manner. The FASEB Jour-
nal. 2012; 26(11):4743-54. [DOI:10. 1096/fj. 12-210989]
[PMID]
11. Rundle A, Hoepner L, Hassoun A, Oberfield S, Freyer G,
Holmes D, et al. Association of childhood obesity with ma-
ternal exposure to ambient air polycyclic aromatic hydro-
carbons during pregnancy. American Journal of Epidemi-
ology. 2012; 175(11):1163-72. [DOI:10. 1093/aje/kwr455]
[PMID] [PMCID]
12. Thayer KA, Heindel JJ, Bucher JR, Gallo MA. Role of
environ-
mental chemicals in diabetes and obesity: A national toxi-
cology program workshop review. Environmental Health
Perspectives. 2012; 120(6):779-89. [DOI:10. 1289/ehp.
1104597] [PMID] [PMCID]
13. Kim HW, Kam S, Lee DH. Synergistic interaction between
polycyclic aromatic hydrocarbons and environmental to-
bacco smoke on the risk of obesity in children and ado-
lescents: The US national health and nutrition examina-
tion survey 2003-2008. Environmental Research. 2014;
135:354-60. [DOI:10. 1016/j. envres. 2014. 08. 032] [PMID]
14. Dong GH, Wang J, Zeng XW, Chen L, Qin XD, Zhou Y, et
al. Interactions between air pollution and obesity on
blood pressure and hypertension in Chinese children.
Epidemiology. 2015; 26(5):740-7. [DOI:10. 1097/EDE.
0000000000000336] [PMID]
15. Fleisch AF, Rifas-Shiman SL, Koutrakis P, Schwartz JD,

Kloog
I, Melly S, et al. Prenatal exposure to traffic pollution: Asso-
ciations with reduced fetal growth and rapid infant weight
gain. Epidemiology. 2015; 26(1):43-50. [DOI:10. 1097/EDE.
0000000000000203] [PMID] [PMCID]
16. Jerrett M, McConnell R, Chang CR, Wolch J, Reynolds K,
Lurmann F, et al. Automobile traffic around the home and
attained body mass index: A longitudinal cohort study of
children aged 10-18 years. Preventive Medicine. 2010;
(Suppl. 50):S50-8. [DOI:10. 1016/j. ypmed. 2009. 09. 026]
[PMID] [PMCID]
17. Jerrett M, McConnell R, Wolch J, Chang R, Lam C, Dunton
G, et al. Traffic-related air pollution and obesity formation
in children: A longitudinal, multilevel analysis. Environmen-
tal Health. 2014; 13:49. [DOI:10. 1186/1476-069X-13-49]
[PMID] [PMCID]
18. McConnell R, Shen E, Gilliland FD, Jerrett M, Wolch J,
Chang
CC, et al. A longitudinal cohort study of body mass index and
childhood exposure to secondhand tobacco smoke and air
pollution: The Southern California Children’s Health Study.
Environmental Health Perspectives. 2015; 123(4):360-6.
[DOI:10. 1289/ehp. 1307031] [PMID] [PMCID]
19. Alderete TL, Habre R, Toledo-Corral CM, Berhane K, Chen
Z,
Lurmann FW, et al. Longitudinal associations between am-
https://doi.org/10.1093/epirev/mxm012

https://www.ncbi.nlm.nih.gov/pubmed/17569676
https://doi.org/10.1016/S0140-6736(11)60813-1
https://doi.org/10.1016/S0140-6736(11)60813-1
https://doi.org/10.1080/10408390903372599
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2932668
https://doi.org/10.1038/sj.ijo.0803326
https://doi.org/10.1016/j.envint.2014.01.026
https://doi.org/10.1016/j.envint.2014.01.026
https://doi.org/10.2337/db12-0190
https://doi.org/10.1016/j.rcl.2012.06.006
https://doi.org/10.1038/ijo.2011.255
https://doi.org/10.1096/fj.12-210989
https://doi.org/10.1093/aje/kwr455
https://doi.org/10.1289/ehp.1104597
https://doi.org/10.1016/j.envres.2014.08.032
https://doi.org/10.1097/EDE.0000000000000336
https://doi.org/10.1097/EDE.0000000000000336

https://doi.org/10.1097/EDE.0000000000000203
https://doi.org/10.1097/EDE.0000000000000203
https://doi.org/10.1016/j.ypmed.2009.09.026
https://doi.org/10.1186/1476-069X-13-49
9
bient air pollution with insulin sensitivity, β-cell function,
and adiposity in Los Angeles Latino children. Diabetes.
2017; 66(7):1789-96. [DOI:10. 2337/db16-1416] [PMID]
[PMCID]
20. Fioravanti S, Cesaroni G, Badaloni C, Michelozzi P, Forast-
iere F, Porta D. Traffic-related air pollution and childhood
obesity in an Italian birth cohort. Environmental Research.
2018; 160:479-86. [DOI:10. 1016/j. envres. 2017. 10. 003]
[PMID]
21. Fleisch AF, Luttmann-Gibson H, Perng W, Rifas-Shiman
SL,
Coull BA, Kloog I, et al. Prenatal and early life exposure to
traffic pollution and cardiometabolic health in childhood.
Pediatric Obesity. 2017; 12(1):48-57. [DOI:10. 1111/ijpo.
12106] [PMID] [PMCID]

22. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR.
Intro-
duction to meta-analysis. New York: John Wiley and Sons;
2009.
23. Peterson RA, Brown SP. On the use of beta coefficients
in meta-analysis. Journal of Applied Psychology. 2005;
90(1):175-81. [DOI:10. 1037/0021-9010. 90. 1. 175] [PMID]
24. Mazidi M, Speakman JR. Ambient particulate air pollution
(PM2. 5) is associated with the ratio of type 2 diabetes to
obesity. Scientific Reports. 2017; 7(1):9144. [DOI:10. 1038/
s41598-017-08287-1] [PMID] [PMCID]
25. Stieb DM, Chen L, Eshoul M, Judek S. Ambient air
pollution,
birth weight and preterm birth: A systematic review and
meta-analysis. Environmental Research. 2012; 117:100-11.
[DOI:10. 1016/j. envres. 2012. 05. 007] [PMID]
26. Li X, Huang S, Jiao A, Yang X, Yun J, Wang Y, et al.
Associa-
tion between ambient fine particulate matter and preterm
birth or term low birth weight: An updated systematic re-
view and meta-analysis. Environmental Pollution. 2017;
227:596-605. [DOI:10. 1016/j. envpol. 2017. 03. 055]
[PMID]
27. Dong GH, Qian ZM, Liu MM, Wang D, Ren WH, Flick LH,
et al. Ambient air pollution and the prevalence of obesity
in Chinese children: The seven northeastern cities study.
Obesity. 2014; 22(3):795-800. [DOI:10. 1002/oby. 20198]
28. McConnell R, Gilliland FD, Goran M, Allayee H, Hricko A,
Mittelman S. Does near-roadway air pollution contribute

to childhood obesity? Pediatric Obesity. 2016;11(1):1-3.
[DOI:10. 1111/ijpo. 12016] [PMID] [PMCID]
29. Vrijheid M, Casas M, Gascon M, Valvi D, Nieuwenhuijsen
M. Environmental pollutants and child health-a review of
recent concerns. International Journal of Hygiene and Envi-
ronmental Health. 2016; 219(4-5):331-42. [DOI:10. 1016/j.
ijheh. 2016. 05. 001] [PMID]
30. Toledo-Corral C, Alderete T, Habre R, Berhane K, Lurmann
F, Weigensberg M, et al. Effects of air pollution exposure
on glucose metabolism in Los Angeles minority children.
Pediatric Obesity. 2018; 13(1):54-62. [DOI:10. 1111/ijpo.
12188] [PMID] [PMCID]
31. Kelishadi R, Mirghaffari N, Poursafa P, Gidding SS.
Lifestyle
and environmental factors associated with inflammation,
oxidative stress and insulin resistance in children. Athero-
sclerosis. 2009; 203(1):311-9. [DOI:10. 1016/j. atheroscle-
rosis. 2008. 06. 022] [PMID]
32. Zou MH. Is NAD (P) H oxidase a missing link for air pollu-
tion-enhanced obesity? Arteriosclerosis, Thrombosis, and
Vascular Biology. 2010; 30(12):2323-4. [DOI:10. 1161/AT-
VBAHA. 110. 216648] [PMID] [PMCID]
33. Xu X, Yavar Z, Verdin M, Ying Z, Mihai G, Kampfrath T,
et al.
Effect of early particulate air pollution exposure on obe-
sity in mice: Role of p47phox. Arteriosclerosis, Thrombo-
sis, and Vascular Biology. 2010; 30(12):2518-27. [DOI:10.
1161/ATVBAHA. 110. 215350] [PMID] [PMCID]
34. Sun Q, Yue P, Deiuliis JA, Lumeng CN, Kampfrath T,
Mikolaj

MB, et al. Ambient air pollution exaggerates adipose in-
flammation and insulin resistance in a mouse model of
diet-induced obesity. Circulation. 2009; 119(4):538-46.
[DOI:10. 1161/CIRCULATIONAHA. 108. 799015] [PMID]
[PMCID]
35. Liu C, Xu X, Bai Y, Wang TY, Rao X, Wang A, et al. Air
pol-
lution-mediated susceptibility to inflammation and insulin
resistance: Influence of CCR2 pathways in mice. Environ-
mental Health Perspectives. 2014; 122(1):17-26. [DOI:10.
1289/ehp. 1306841] [PMID] [PMCID]
36. Hajat A, Diez-Roux AV, Adar SD, Auchincloss AH, Lovasi
GS,
O’Neill MS, et al. Air pollution and individual and neighbor-
hood socioeconomic status: Evidence from the Multi-Eth-
nic Study of Atherosclerosis (MESA). Environmental Health
Perspectives. 2013; 121(11-12):1325-33. [DOI:10. 1289/
ehp. 1206337] [PMID] [PMCID]
37. Fram MS, Ritchie LD, Rosen N, Frongillo EA. Child
experience
of food insecurity is associated with child diet and physical
activity. The Journal of Nutrition. 2015; 145(3):499-504.
[DOI:10. 3945/jn. 114. 194365] [PMID]
38. Krebs NF JM, American academy of pediatrics committee
on nutrition. Prevention of pediatric overweight and obe-
sity. Pediatrics. 2003; 112(2):424-30. [DOI:10. 1542/peds.
112. 2. 424] [PMID]
39. Bourdeaudhuij I, Lefevre J, Deforche B, Wijndaele K, Mat-
ton L, Philippaerts R. Physical activity and psychosocial
correlates in normal weight and overweight 11 to 19 year
olds. Obesity. 2005; 13(6):1097-105. [DOI:10. 1038/oby.

2005. 128] [PMID]
40. Trost SG, Kerr L, Ward DS, Pate RR. Physical activity and
determinants of physical activity in obese and non-obese
children. International Journal of Obesity. 2001; 25(6):822-
9. [DOI:10. 1038/sj. ijo. 0801621] [PMID]
41. Nerriere É, Zmirou-Navier D, Blanchard O, Momas I, Lad-
ner J, Le Moullec Y, et al. Can we use fixed ambient air
monitors to estimate population long-term exposure to
air pollutants? The case of spatial variability in the Geno-
tox ER study. Environmental Research. 2005; 97(1):32-42.
[DOI:10. 1016/j. envres. 2004. 07. 009] [PMID]
https://doi.org/10.2337/db16-1416
https://doi.org/10.1111/ijpo.12106
https://doi.org/10.1037/0021-9010.90.1.175
https://doi.org/10.1038/s41598-017-08287-1
https://doi.org/10.1038/s41598-017-08287-1
https://doi.org/10.1016/j.envpol.2017.03.055

https://doi.org/10.1002/oby.20198
https://doi.org/10.1016/j.ijheh.2016.05.001
https://doi.org/10.1016/j.ijheh.2016.05.001
https://doi.org/10.1016/j.atherosclerosis.2008.06.022
https://doi.org/10.1016/j.atherosclerosis.2008.06.022
https://doi.org/10.1161/ATVBAHA.110.216648
https://doi.org/10.1161/CIRCULATIONAHA.108.799015
https://doi.org/10.3945/jn.114.194365

https://doi.org/10.1542/peds.112.2.424
https://doi.org/10.1542/peds.112.2.424
https://doi.org/10.1038/oby.2005.128
https://doi.org/10.1038/oby.2005.128
10
A
ppendix 1. Sum
m
ary of exposure and outcom
e assessm
ent strategies and an estim
ate of effect size in included studies
Author(s)/
D
ate
Study
Location
Study
D

esign
Follow
U
p
D
ura-
tion
Sam
ple
Size
Age (y)
O
utcom
e
Exposure Assess-
m
ent M
ethod
Exposure
G
roup /
Subgroups

Effect Size
CI (95%
)
Adjustm
ent Factors
M
ichael
Jerrett
,
2010
Southern
California,
U
SA
Cohort
8-year
follow
-
up
2889
10–18 years
Body M
ass
Index (BM

I)
Traffi
c-related air
pollution
Annual
average daily
traffi
c (AADT)
150 m
B (SE):
0. 0039 (0. 0008)
G
ender, cohorts variables
parental education,
personal w
eekly sm
oking,
second hand sm
oke (cur-
rent + past), ever asthm
a,
buffer population, gam
m
a

index, proportion of below
poverty people w
ithin
census block, norm
alized
difference vegetation
index (N
DVI), foreign-born,
com
m
unity-level violent
crim
e rate, and having no
food stores w
ithin 500-m
road netw
ork buffer w
ith
random
com
m
unity effects

AADT 300 m
0. 0013 (0. 0008)
Pei, 2013
G
erm
any
Cohort
10-year
follow
-
up
3121
Fem
ales
(N
=1114),
M
ales
(N
=1158)
10 (W
e predicted
standardized body
m

ass index (BM
I) at 10
years of age using stan -
dardized BM
Is from
birth to 5 years. )
M
aternal sm
oking dur-
ing pregnancy
β (CI):
0. 13
(0. 03, 0. 22)
Parental education, fam
-
ily incom
e, and m
aternal
sm
oking during pregnancy
M
ichael Jer-

rett
(2014)
Southern
California,
U
SA
Cohort
4-year
follow
-
up
N
=4550
5–11 years
Β (SE)
Having asthm
a,
parental education,
im
m
igrant status,
the m
easure of green

cover, street connectivity,
recreational program
m
ing
w
ithin 5 km
of the hom
e,
and fast food access w
ithin
500 m
of the hom
e
Traffi
c density
0. 0002 (0. 0001)
N
on-Freew
ay
N
O
x
0. 0861 (0. 0255)

11
Author(s)/
D
ate
Study
Location
Study
D
esign
Follow
U
p
D
ura-
tion
Sam
ple

Size
Age (y)
O
utcom
e
Exposure Assess-
m
ent M
ethod
Exposure
G
roup /
Subgroups
Effect Size
CI (95%
)
Adjustm
ent Factors
G
uang-Hui
Dong, 2014
China
Cross-
sectional

-
30056
2-14
O
besity
BM
I>
95th per -
centile
M
easurem
ents of am
bi -
ent PM
10 , SO
2 , N
O
2, and
O
3 concentrations
from
2006 to 2008 w
ere

obtained at m
unicipal
air pollution m
onitoring
stations.
PM
10 (μg/m
3)
O
R(CI):
1. 19
(1. 11–1. 26)
Age, gender, parental
education, breast
feeding, low
birth w
eight,
area of residence per
person, house decorations,
hom
e coal use, ventilation
device in the kitchen, air
exchange in w
inter, passive

sm
oking exposure,
and districts
SO
2 (ppb)
1. 11
(1. 03–1. 20)
N
O
2 (ppb)
1. 13
(1. 04–1. 22)
O
3 (ppb)
1. 14
(1. 04–1. 24)
Rob M
cCon -
nel,
2015
Southern
California,
U
SA

Cohort
8-year
3318
10-18
BM
I
Residential near-road-
w
ay pollution exposure
(N
RP) w
as estim
ated
based on a line source
dispersion m
odel
accounting for traffi
c
volum
e, proxim
ity, and
m
eteorology
Secondhand

sm
oke
BM
I grow
th
(95%
CI):
Ethnicity, sex, com
m
unity,
year of enrollm
ent, and
age
O
ne sm
oker at
hom
e: 0. 48 (0. 16,
1. 12)
Tw
o or m
ore
than 2 sm

okers at
hom
e: 1. 08 (0. 19,
1. 97)
M
aternal sm
oking
during pregnancy:
0. 72 (0. 14, 1. 31)
N
RP: 1. 13 (0. 61,
1. 65).
The difference in
the attained BM
I
(95%
CI):
O
ne sm
oker at
hom
e: 0. 95 (0. 42,
1. 47)
≥ 2 sm
okers at

hom
e: 1. 77 (1. 04,
2. 51
M
aternal sm
oking
during pregnancy;
1. 14 (0. 66, 1. 62
N
RP: 1. 27 (0. 75,
1. 80)
12
Author(s)/
D
ate
Study
Location
Study
D

esign
Follow
U
p
D
ura-
tion
Sam
ple
Size
Age (y)
O
utcom
e
Exposure Assess -
m
ent M
ethod
Exposure
G
roup /
Subgroups
Effect Size

CI (95%
)
Adjustm
ent Factors
Fleisch,
2016
Boston,
U
SA
Cohort
Early
child-
hood
(m
e-
dian: 3.
3 years
of age)
m
id-
child-
hood
(m
e-
dian: 7.

7 years
of age)
1418
M
ean (standard
deviation) of age at
early childhood visit 3.
3 (0. 4),
and at m
id-childhood
visit 8. 0 (0. 9)
BM
I z-score
Spatiotem
poral m
odels
to estim
ate prenatal
and early life residential
PM
2. 5 and black carbon
exposure as w
ell as traf-
fic density and

roadw
ay proxim
ity.
PM
2. 5 (μg/m
3)
3. 3 years
Child (age, sex and race/
ethnicity), m
other (age,
education, and sm
ok-
ing during pregnancy),
neighborhood (census
tract m
edian incom
e),
season and date of health
outcom
e
Third trim
es -
ter

0. 0
(-0. 1, 0. 1)
Year prior to
early child-
hood visit
-0. 0
(-0. 1, 0. 1)
N
ear-residence
traffi
c density
Birth address
0. 0
(-0. 0, 0. 1)
Early child -
hood address
0. 0
(-0. 0, 0. 1)
Proxim
ity to
m
ajor roadw
ay,
birth address

Reference:≥200m
<50m
0. 3
(0. 0, 0. 7)
[50, 100m
]
-0. 0
(-0. 4,0. 3
[100, 200m
]
0. 4
(0. 1, 0. 6)
Proxim
ity to
m
ajor roadw
ay,
early child -
hood address
Reference:≥200m
<50 m
0. 1
(-0. 2, 0. 5)

[50, 100 m
]
-0. 0
(-0. 4, 0. 3)
[100, 200 m
]
0. 1
(-0. 2, 0. 3)
Yueh-
HsiuM
athil -
daChiua
2017
Boston,
U
SA
Cohort
4. 0±0.
7 years
239
BM
I, z-
score

Prenatal daily PM
2.
5 exposure w
as esti-
m
ated using a validated
satellite-based spatio-
tem
poral
m
odel.
Prenatal PM
2. 5 level
(μg/m
3
M
edian IQ
R
10. 7(9. 9─11. 4)
PM
2. 5 (μg/m
3)
G
irls

-0. 12
(-0. 37,-0. 03)
M
aternal age, race/
ethnicity, education, pre-
pregnancy BM
I, and child’s
age at anthropom
etric
m
easurem
ent
Boys
0. 21
(0. 003,-0. 37)
13
Author(s)/
D

ate
Study
Location
Study
D
esign
Follow
U
p
D
ura-
tion
Sam
ple
Size
Age (y)
O
utcom
e
Exposure Assess-
m
ent M

ethod
Exposure
G
roup /
Subgroups
Effect Size
CI (95%
)
Adjustm
ent Factors
Tanya. Alde-
rete, 2017
Los Ange-
les, U
SA
Cohort
3. 4
years
314
overw
eight
and obese
children
8-15 years
BM

I
N
O
2 and PM
2. 5 w
ere
m
odeled as long-term
exposure using cum
ula-
tive
12-m
onth averaged
exposure during the
follow
-up.
Estim
ated effect esti-
m
ates w
ere reported
for a

A 5-ppb difference in
N
O
2 and a 4-μg/m
3
difference in PM
2. 5
N
O
2
(ppb)
2. 1
(0. 1, 4. 1)
Sex, Tanner stage, the
season of testing
(w
arm
/cold), prior year
exposure at each follow
-up
visit, social position, body
fat percentage (w
here
appropriate), study w

ave,
and study entry year
PM
2. 5 (μg/m
3)
3. 8
(1, 6. 6)
Poursafa,
2017
Iran
Cross-
sectional
---
186
6-18
BM
I
The air quality index
(AQ
I) is used to describe
the level of air pollution
w
ith adverse health

effects. W
e used PM
2.
5 data.
Β=0. 34
Age and gender
Sara Fiora-
vant, 2018
Italy
Cohort
8-year
follow
-
up
581
Birth-8 years
Prevalence
of over-
w
eight/
obesity w
as
9. 3%
and

36. 9%
Air pollution w
as as-
sessed at the residential
address
N
O
2 (per 10 μg/
m
3)
0. 99
(0. 86, 1. 12)
M
aternal and paternal
education, m
aternal pre-
pregnancy BM
I, m
aternal
sm
oking during pregnancy,
gestational
diabetes, m
aternal age at

delivery, gestational age,
childbirth w
eight, breast-
feeding duration, age (in
m
onths) at w
eaning
N
O
X (per 20 μg/
m
3)
0. 98
(0. 86, 1. 12)
PM
10 (per 10 μg/
m
3)
0. 971
(0. 77, 1. 23)
PM
2. 5 (per 5 μg/
m
3)

1. 02
(0. 75, 1. 40)
PM
coarse (per 5
μg/m
3)
0. 91
(0. 77, 1. 09)
PM
2. 5 abs (per 1
μg/m
3)
1. 10
(0. 88, 1. 37)
G
uangyun
M
ao, 2017
Boston,
U
SA
Cohort
2-year
follow

-
up
1,446
first 2 years of life
Com
paring the
highest and low
-
est quartiles of
PM
(2:5 μg/m
3)
The adjusted Rela-
tive Risks (RRs)
1. 3
(1. 1, 1. 5)
M
aternal age at delivery,
race/ ethnicity, education
level, sm
oking status,
diabetes, m
arriage status,
household incom

e per
year, the season of child-
birth, preterm
birth, birth
w
eight, and breastfeeding
This Page Intentionally Left Blank
Vision board
prosperity
Graduate college with my Bachelors degree in Psychology
In the next 3 years

aspiration
Growing my start up business
Build more clientele
Pay off all my school loans in the next 4 years
family
Huge supporters
Mother and father
My brother constantly helps me when I get stuck and do not
know where my next move is
motivators
My family
Huge financial supporters, but constantly motivates to get to the
finish line(graduation)
(my best friend)
Keeps me level headed and grounded
(my boyfriend)
Reminds me to stay focus on the big picture, and to never give
up on your hope and dreams.

goals
College tuition paid off in the next 4 years
Have my start up business flourishing and open a new location
Skills knowledge
Bachelors degree in Psychology
Internships
Background working in child care
Career/ life path
Land My Dream Job
Child psychologist
RESEARCH ARTICLE Open Access
Salivary epigenetic biomarkers as

predictors of emerging childhood obesity
Amanda Rushing1, Evan C. Sommer2, Shilin Zhao3, Eli K.
Po’e4 and Shari L. Barkin5*
Abstract
Background: Epigenetics could facilitate greater understanding
of disparities in the emergence of childhood
obesity. While blood is a common tissue used in human
epigenetic studies, saliva is a promising tissue. Our prior
findings in non-obese preschool-aged Hispanic children
identified 17 CpG dinucleotides for which differential
methylation in saliva at baseline was associated with maternal
obesity status. The current study investigated to
what extent baseline DNA methylation in salivary samples in
these 3–5-year-old Hispanic children predicted the
incidence of childhood obesity in a 3-year prospective cohort.
Methods: We examined a subsample (n = 92) of Growing Right
Onto Wellness (GROW) trial participants who were
randomly selected at baseline, prior to randomization, based on
maternal phenotype (obese or non-obese).
Baseline saliva samples were collected using the Oragene DNA
saliva kit. Objective data were collected on child
height and weight at baseline and 36 months later. Methylation
arrays were processed using standard protocol.
Associations between child obesity at 36 months and baseline
salivary methylation at the previously identified 17
CpG dinucleotides were evaluated using multivariable logistic
regression models.
Results: Among the n = 75 children eligible for analysis,
baseline methylation of Cg1307483 (NRF1) was significantly
associated with emerging childhood obesity at 36-month follow-
up (OR = 2.98, p = 0.04), after adjusting for child
age, gender, child baseline BMI-Z, and adult baseline BMI. This

translates to a model-estimated 48% chance of child
obesity at 36-month follow-up for a child at the 75th percentile
of NRF1 baseline methylation versus only a 30%
chance of obesity for a similar child at the 25th percentile.
Consistent with other studies, a higher baseline child
BMI-Z during the preschool period was associated with the
emergence of obesity 3 years later, but baseline
methylation of NRF1 was associated with later obesity even
after adjusting for child baseline BMI-Z.
Conclusions: Saliva offers a non-invasive means of DNA
collection and epigenetic analysis. Our proof of principle
study provides sound empirical evidence supporting DNA
methylation in salivary tissue as a potential predictor of
subsequent childhood obesity for Hispanic children. NFR1
could be a target for further exploration of obesity in this
population.
Keywords: Obesity, Hispanic children, Epigenetics,
Methylation, Saliva
Background
The prevalence of pediatric obesity has been increasing
at an alarming rate in the last forty years [1, 2]. Although
pediatric obesity prevalence is a global issue, the United
States is facing epidemic levels of pediatric obesity [3, 4].
The Center for Disease Control and Prevention indicates
that the prevalence of obesity among children aged 2–
19 years old has risen from 13.9% in 2000 to 18.5% in
2016 [5]. However, some ethnic groups have an even
higher obesity prevalence [1, 6]. For example, the 2015–
2016 National Health and Nutrition Examination Survey
(NHANES) reported 25.8% of Hispanic 2–19 year-olds
were obese compared to 14.1% of their non-Hispanic
white counterparts [7]. Identifying what influences dif-

ferent populations is critical to successfully reducing
obesity-related health disparities.
Childhood is a particularly sensitive period for neuro-
logical, endocrine, and metabolic development. For ex-
ample, obesity at a young age contributes to an
© The Author(s). 2020 Open Access This article is distributed
under the terms of the Creative Commons Attribution 4.0
International License
(http://creativecommons.org/licenses/by/4.0/), which permits
unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate
credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were
made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to
the data made available in this article, unless otherwise stated.
* Correspondence: [email protected]
5Department of Pediatrics, Vanderbilt University School of
Medicine, 2200
Children’s Way, Doctor’s Office Tower 8232, Nashville, TN
37232-9225, USA
Full list of author information is available at the end of the
article
Rushing et al. BMC Medical Genetics (2020) 21:34
https://doi.org/10.1186/s12881-020-0968-7
http://crossmark.crossref.org/dialog/?doi=10.1186/s12881-020-
0968-7&domain=pdf
http://orcid.org/0000-0003-1585-3818
http://creativecommons.org/licenses/by/4.0/
http://creativecommons.org/publicdomain/zero/1.0/
mailto:[email protected]

increased risk of diabetes, hypertension, and cardiovas-
cular disease in adulthood [8–10]. Recent literature indi-
cates that susceptibility to obesity within an
“obesogenic” environment differs among individuals [11,
12]. It is not clear what mechanisms are responsible for
obesity variation, but many studies identify a dynamic
interaction of genetic and environmental exposures at
sensitive periods of development [13, 14]. While mono-
genic DNA mutations exist and are associated with
obesity, common forms of childhood obesity have frus-
trated the scientific community with the so-called prob-
lem of missing heritability. It appears that obesity is a
multi-trait, multi-state phenotype. The field of epigenet-
ics, modifications that affect transcriptional plasticity,
might offer insights into the emerging phenotype of
childhood obesity. Epigenetic patterns, often measured
by DNA methylation, change rapidly in response to en-
vironmental factors such as nutrition and physical activ-
ity and are specifically vulnerable to changes during
early childhood development. Moreover, epigenetic pat-
terns vary between ethnic groups and could explain dif-
fering susceptibility to early emerging obesity and its
commonly associated later chronic diseases [15–18].
Epigenetic patterns are tissue-dependent. While blood
is a common tissue used in studies of human epigenetic
changes, saliva is also a promising tissue. Saliva could be
particularly valuable in studying pediatric populations
given the ease of tissue access, cost-efficiency, and the
ability to collect it in multiple settings [19, 20]. Abraham
and colleagues illustrated that when comparing DNA
fragmentation, quality, and genotype concordance, saliva
is comparable to blood samples [21]. When examining
methylation patterns, both saliva and blood reliably as-

sess epigenetic modifications [22]. In comparing the col-
lection of blood and saliva samples, saliva collection is
associated with lower infection rates, decreased cost, in-
creased patient acceptance, and higher participant com-
pliance [23]. Saliva also has the advantage of offering
insight into the gastrointestinal tract, which could be
useful when examining obesity. The ease of saliva collec-
tion coupled with DNA fidelity could allow for a more
practical source of DNA collection for children. Given
that salivary tissue is used less often for epigenetic stud-
ies, this approach is novel.
Recently, Oelsner et al. examined 92 saliva samples
from 3 to 5-year-old Hispanic children who were at risk
for obesity but not yet obese and analyzed 936 genes
previously associated with obesity [24]. The cross-
sectional study identified 17 CpG dinucleotides that
demonstrated an association between baseline differen-
tial child DNA methylation and maternal BMI (obese
versus non-obese). While this analysis was conducted on
baseline saliva samples, these children subsequently par-
ticipated in a three-year longitudinal study where more
than a third of children became obese. The current study
investigates to what extent baseline child salivary DNA
methylation patterns were associated with the emerging
incidence of childhood obesity in this 3-year prospective
cohort of young Hispanic children [25].
Methods
Informed consent
Trained, bilingual study staff administered written
consents to the parent or legal guardian of the child
of interest in the language of their choice (English or
Spanish). The parent or legal guardian provided con-
sent for both themselves and their child. Because this

was a low health literacy population, the consenting
process utilized specific measures to ensure partici-
pant understanding including a “teach-back” method
and protocol visual aids [26]. The study was approved
by Vanderbilt University Review Board (IRB No.
120643).
Sample population study subjects
We examined a subsample (n = 92) of Growing Right
Onto Wellness (GROW) trial participants [24]. They
were randomly selected at baseline, before
randomization, based on maternal phenotype. One
group of children had obese mothers (BMI ≥30 and
waist circumference ≥ 100 cm), and the other had non-
obese mothers (BMI < 30 and waist circumference < 100
cm). The two groups were matched on child age and
gender. The current study examines this subsample as a
prospective cohort, after a 3-year follow-up.
Child participant eligibility criteria in the original
GROW study included: 3–5 years old, no known medical
conditions, and being at risk for obesity (high normal
weight or overweight) but not yet obese (BMI ≥50th
and < 95th percentile). Three children were excluded
due to being obese at baseline. Parent eligibility criteria
included: ≥18 years old, signed written consent to par-
ticipate in a 3-year trial, consistent phone access, spoke
English or Spanish, and no known medical conditions
that would preclude routine physical activity. Families
were recruited from East Nashville and South Nashville.
All parents self-reported that at least one person in their
household was eligible to participate in a program that
qualified them as underserved. The underserved pro-
grams included but were not limited to TennCare (Me-
dicaid), Special Supplemental Nutrition Program for
Women, Infants, and Children (WIC), CoverKids, Food

Stamps, and/or reduced-price school meals [25]. To be
eligible for analysis in the epigenetic study reported here,
child BMI must have been collected at the 36-month
follow-up (n = 75).
Rushing et al. BMC Medical Genetics (2020) 21:34 Page
2 of 9
Saliva collection and assay method
Baseline saliva samples from children were collected vol-
untarily from interested participants using a separate
consenting form in the participant’s language of choice.
Saliva was collected from children at baseline using the
Oragene DNA saliva kit. Children were asked to fast for
30 min and rinse their mouths with water immediately
before collection. Two mL of saliva was collected from
children using saliva sponges, inserted between the
cheek and gums in the upper cheek pouch without
swabbing the buccal mucosa. Samples were collected at
home or in community centers. To ensure safety and de-
crease contamination, trained sample collectors wore
gloves and capped the samples as soon as the saliva was
collected. After samples were properly collected and la-
beled, the samples were sent to the Vanderbilt Technol-
ogy for Advanced Genomics (VANTAGE) core at
Vanderbilt University. DNA was then extracted from the
saliva using the PrepIT L2P reagent with guidance from
DNA Genotek’s recommendations and stored at − 80
degrees Centigrade.
Anthropometric data
Objective height and weight for parent-child pairs were
collected and used to calculate BMI (kg/m2) at baseline
and 36 months. Trained research staff collected weight

and height of participants using standard anthropomet-
ric procedures, and participants wore only light clothing
and no shoes. Height was measured to the nearest 0.1
cm using a stadiometer (Perspective Enterprise, Portage,
MI), and weight was measured to the nearest 0.1 kg
using a calibrated scale. BMI measurements were col-
lected using the trial protocol [27]. BMI-Z was calcu-
lated based on each child’s BMI, gender, and age, and
BMI categories were defined using CDC guidelines: nor-
mal weight (<85th percentile); overweight (≥85th and <
95th percentile); and obese (≥95th percentile).
Statistical analysis
Categorical variables were summarized using frequencies
and percentages, and Pearson’s chi-squared test was
used to evaluate baseline differences. Continuous vari-
ables were summarized using means and standard devia-
tions, and the Wilcoxon rank-sum test was used to
evaluate baseline differences.
Genome-wide DNA methylation was conducted on
the 92 saliva samples using the Infinium Illumina
HumanMethylation 450 K BeadChip (Illumina, San
Diego, CA, USA). Methylation arrays were processed
using a standard protocol [24, 28] and quality control
was done using the Methylation module (V1.9.0). Sam-
ples with a call rate lower than 98% were excluded,
resulting in the removal of one sample (total baseline eli-
gible sample n = 91). The Background Subtraction
method [29] was used for methylation array
normalization. In this method, the average signals of
built-in negative controls represent background noise
and are subtracted from all probe signals to make unex-
pressed targets equal to zero. Outliers were removed
using the median absolute deviation method. Lastly, the

normalized values were log-transformed and multiplied
by 10 to put degree of methylation on a continuous scale
from 0 to 10 for statistical analysis.
Associations between child obesity at 36 months and
baseline methylation levels were evaluated using multi-
variable logistic regression models. Other variables were
included in the models as covariates, including child age,
child baseline BMI-Z, child gender, and parent baseline
BMI. Statistical significance was defined as p < 0.05. All
analyses were performed using R software (www.r-pro-
ject.org) version 3.5.0.
Results
Of the original 92 participants in the baseline subsample,
75 met quality control and eligibility requirements and
were included for analysis. The mean age was 4.3 years
(SD = 0.8), and the mean baseline BMI was 16.7 (SD =
0.8). Within the analytic sample, at baseline, 64.0% (n =
48) were normal weight, and 36.0% (n = 27) were over-
weight, and 73.3% (n = 55) were Hispanic-Mexican.
Among parents, 48.0% (n = 36) were obese (stratified by
design for this subsample). Refer to Table 1 for further
baseline demographic descriptions of the sample. At the
study’s conclusion, 37% (n = 28) of children were obese.
Comparing children who were non-obese at 36 months
(n = 47) to those who were obese at 36 months (n = 28),
there were no statistically significant differences in base-
line child characteristics, although, descriptively, baseline
weight-related characteristics appeared to be lower in
children who were not obese at follow-up. Parents did
not have any significantly different baseline characteris-
tics between the two groups, although mean age was de-
scriptively slightly younger in parents of children who
became obese at 36 months (33.0 vs. 30.4) (Table 2).

Table 3 describes the associations between children’s
continuous degree of baseline methylation at each CpG
dinucleotide and childhood obesity at 36 months for the
75 children with follow-up data. The multivariable logis-
tic regression model was adjusted for child gender, base-
line age, baseline BMI-Z, and adult baseline BMI. After
accounting for these covariates, higher baseline methyla-
tion of cg01307483 (NRF1) was significantly associated
with a higher probability of childhood obesity at 36
months (odds ratio = 2.98, 95% CI = [1.06, 8.38], p =
0.04). PPARGC1B methylation was potentially associated
with decreased obesity at 36 months but was not statisti-
cally significant. SORCS2 methylation was not statisti-
cally significant for cg03218460 or cg18431297 [30, 31].
3 of 9
http://www.r-project.org
http://www.r-project.org
In the logistic regression model analyzing the NRF1 di-
nucleotide, child gender, child baseline age, and baseline
parent BMI, were not significant predictors of childhood
obesity at 36 months (Table 4). However, in this model,
child baseline BMI-Z and baseline differential methyla-
tion of NRF1 were significant predictors of childhood
obesity at 36 months. Figure 1 displays the model-
predicted probability of child obesity at 36-month
follow-up as a function of increasing NRF1 methylation.
Child baseline BMI-Z was a significant predictor in all
but two CpG dinucleotide models (odds ratio = 3.09–
4.08, p < 0.05). Median degree of baseline methylation at
each CpG dinucleotide examined in this study is shown

in Additional file 1: Table S1.
Discussion
To our knowledge, this is the first prospective cohort
study that investigates DNA methylation collected via
salivary samples as a predictor of childhood emerging
obesity among 3–5-year-old Hispanic children. Although
other studies have investigated DNA methylation pat-
terns in children who are already obese, our prospective
study investigated how these patterns might be used to
predict the future emergence of obesity in non-obese
preschool-aged children above and beyond what is pro-
vided by their age, gender, baseline BMI-Z, and their
mother’s BMI. After adjusting for these covariates, base-
line methylation of Cg1307483 (NRF1) was significantly
associated with emerging childhood obesity at 36-month
follow-up with a significant positive odds ratio (OR =
2.98, p = 0.04). To place this odds ratio finding into con-
text and enhance interpretation, the model estimated a
Table 1 Demographics of Sample Populationa
Child Characteristics Total (n = 75)
Gender
Male 35 (46.7%)
Female 40 (53.3%)
Age at anthropometry collection (years) 4.3 (0.8)
Age category (years)
3 34 (45.3%)

4 26 (34.7%)
5 15 (20.0%)
BMI (kg/m2) 16.7 (0.8)
BMI-Z 0.9 (0.5)
BMI category
Normal weight 48 (64.0%)
Overweight 27 (36.0%)
Waist circumference (cm) 53.4 (3.0)
Race/Ethnicity
Hispanic Mexican 55 (73.3%)
Hispanic non-Mexican 20 (26.7%)
Parent Characteristics
Age (years) 32.0 (5.7)
BMI (kg/m2) 29.4 (7.3)
BMI category
Normal weight 28 (37.3%)
Overweight 11 (14.7%)
Obese 36 (48.0%)

Waist circumference (cm) 97.9 (16.1)
Abbreviations: BMI body mass index, BMI-Z BMI z-score
a Values are mean (SD) or frequency (percent)
Table 2 Baseline Participant Characteristics by Child Obesity
Status at 36 monthsa
Child Characteristics Child not obese at
36 months (n = 47)
Child obese at 36
months (n = 28)
P
valueb
Gender 0.32
Male 24 (51.1%) 11 (39.3%)
Female 23 (48.9%) 17 (60.7%)
Age at
anthropometry
collection (years)
4.3 (0.8) 4.3 (0.8) 0.90
Age category (years) 0.93
3 21 (44.7%) 13 (46.4%)
4 17 (36.2%) 9 (32.1%)

5 9 (19.1%) 6 (21.4%)
BMI (kg/m2) 16.6 (0.7) 16.9 (0.8) 0.06
BMI-Z 0.8 (0.5) 1.0 (0.4) 0.06
BMI category 0.05
Normal weight 34 (72.3%) 14 (50.0%)
Overweight 13 (27.7%) 14 (50.0%)
Waist circumference
(cm)
52.9 (2.6) 54.3 (3.6) 0.13
Race/Ethnicity 0.41
Hispanic Mexican 36 (76.6%) 19 (67.9%)
Hispanic non-
Mexican
11 (23.4%) 9 (32.1%)
Parent Characteristics
Age (years) 33.0 (5.2) 30.4 (6.0) 0.06
BMI (kg/m2) 29.2 (7.4) 29.7 (7.3) 0.88
BMI category 0.67
Normal weight 18 (38.3%) 10 (35.7%)

Overweight 8 (17.0%) 3 (10.7%)
Obese 21 (44.7%) 15 (53.6%)
Waist
circumference
(cm)
97.0 (16.2) 99.3 (16.1) 0.47
Abbreviations: BMI body mass index, BMI-Z BMI z-score
a Values are mean (SD) or frequency (percent)
b Wilcoxon rank-sum test used for continuous variables, and
Pearson’s chi-
squared used for categorical variables
4 of 9
48% chance of child obesity at 36-month follow-up for a
child at the 75th percentile of NRF1 methylation versus
only a 30% chance of obesity for a similar child at the
25th percentile.
Consistent with other studies, a higher baseline child
BMI-Z during the preschool period was associated with
the emergence of obesity 3 years later, but baseline
methylation of NRF1 was associated with later obesity
even after adjusting for baseline BMI-Z. NRF1 is associ-
ated with the innate immune response governing adipo-
cyte inflammation and cytokine expression, as well as
brown adipose tissue thermogenic adaption. It also plays
a role in insulin resistance [34–36]. The current results
build on the existing literature by demonstrating that

DNA methylation of a critical CpG dinucleotide within
the NRF1 gene in 3–5-year-old non-obese children is
associated with the emergence of obesity 3 years later.
This provides a potential target of further investigation
and suggests that adipocyte inflammation might already
be affected before the phenotypic emergence of child-
hood obesity in Hispanic children. Other studies demon-
strate that early life exposures can affect later health and
disease outcomes. This life-course understanding of
emerging phenotypes might contribute to health
disparities.
It is important to note that prior studies implicated
NRF1 associated with existing obesity and young His-
panic children using blood and skeletal muscle samples
as well [36]. Comuzzie and colleagues investigated
chromosome 7q in Hispanic children and found unique
loci contributing to pediatric obesity [33]. As in our
study, the genes significantly associated with obesity
Table 3 Association of Baseline Methylationa at Each CpG
Dinucleotide With Child Obesity Status at 36-Month Follow-Up
Unique CpG
Dinucleotide
Associated
Gene
Odds
Ratio
95% CI p-
value

Biological Relevance
cg21790991 FSTL1 1.35 [0.88,
2.06]
0.16 Regulate endothelial cell function and vascular remodeling
in response to hypoxic ischemia
[30, 40]
cg03218460 SORCS2 2.08 [0.83,
5.21]
0.12 Functions to regulate fasting insulin levels and secretion of
insulin, diabetes susceptibility
[41, 42]
cg23241637 ZNF804A 1.47 [0.61,
3.49]
0.39 Schizophrenia [43, 44]
cg04798490 SHANK2 1.51 [0.6,
3.8]
0.38 Autism [45, 46]
cg01307483 NRF1 2.98 [1.06,
8.38]
0.04* Innate immune response governing adipocyte
inflammation, cytokine expression,
and insulin resistance [34–36]
cg19312314 CBS 0.75 [0.27,
2.13]

0.59 Catalyzes the conversion of homocysteine to cystathionine,
associated with homocystinuria
and hydrogen sulfide production [47]
cg14321859 DLC1 1.82 [0.69,
4.81]
0.23 Regulates Rho GTP-binding proteins, cytoskeletal
signaling, tumor suppressor, adipocyte dif-
ferentiation [48, 49]
cg03067613 ATP8B3 1.24 [0.4,
3.8]
0.71 Reproduction [37]
cg11296553 CEP72 0.99 [0.03,
32.38]
0.99 Ulcerative colitis [38]
cg16509445 CRYL1 1.04 [0.23,
4.63]
0.96 Heptocellular carcinoma [50]
cg16344026 PPARGC1B 0.27 [0.04,
2.04]
0.21 Fat oxidation, non-oxidative glucose metabolism, and
energy regulation, ubiquitous in duo-
denum and small intestines [51, 52]
cg15354625 ODZ4 0.78 [0.12,
4.94]

0.79 Bipolar disorder [53]
cg23836542 CHN2 1.13 [0.14,
9.07]
0.91 Encodes GTP-metabolizing protein that regulates cell
proliferation and migration, insulin re-
sistance [54, 55]
cg07511564 NXPH1 1.02 [0.34,
3.03]
0.97 Forms a tight complex with alpha neurexins, promoting
adhesion between dendrites and
axons, diabetic neuropathy [56, 57]
cg18799510 GRIN3A 1.02 [0.32,
3.28]
0.98 Schizophrenia [31, 58]
cg14996807 UNC13A 3.17 [0.64,
15.76]
0.16 ALS [59, 60]
cg18431297 SORCS2 1.01 [0.51,
1.99]
0.98 Neuropeptide receptor activity, strongly expressed in the
central nervous system, acts with
IGF1 in the setting of cardiovascular disease [42, 61]
a Degree of methylation was a continuous variable calculated by
log-transforming the normalized values and multiplying by 10
to put it on a scale from 0 to 10

(see Methods section)
5 of 9
suggest a strong inflammatory influence. While our
study does not confirm this hypothesis, it does provide
supplemental data supporting this theory. Because NRF1
is a major transcription factor in metabolic regulation
and stimulates the expression of PPARGC1B, these look
like promising targets for further research and interven-
tion approaches. Although direct comparisons between
saliva and blood cannot be made in the current study,
the fact that we noted the same association in saliva as
found in blood and skeletal muscle further bolsters sup-
port for the use of saliva as a useful tissue for epigenetic
inquiry.
Other genes could also play an important role in both
predicting later obesity and understanding the pathways
that lead to obesity. For example, PPARGC1B methyla-
tion had a potentially strong association with decreased
obesity at 36 months but was not statistically significant.
PPARGC1B is associated with fat oxidation, non-
oxidative glucose metabolism, and energy regulation [37,
38]. Similarly, SORCS2 methylation, which functions to
regulate fasting insulin levels and secretion of insulin,
was potentially associated with increased obesity at 36
months but was not statistically significant in this rela-
tively small sample [31, 39]. Repeating this work in a lar-
ger sample is necessary for further understanding these
and other epigenetic contributions to the early emer-
gence of obesity in populations who experience higher
health disparities associated with obesity. Moreover,

while the small analytic sample size precluded moder-
ation analysis in the current study, it would be interest-
ing for future research to explore whether the potential
relationships between methylation and subsequent obes-
ity status depend on initial BMI status or other potential
moderators of interest (e.g., gender, ethnicity, income,
etc.).
To-date, many epigenetic studies have focused on the
exploration of molecular pathways. While it is not yet
known if these DNA methylation patterns can be used
as biomarkers, our study provides a proof-of-principle
demonstrating that even in non-obese Hispanic children,
Fig. 1 Model Predicted Probability of Child Obesity at 36-
month Follow-up as a Function of NRF1 Methylation. Figure 1
displays the logistic
regression model-predicted probability of child obesity at 36
months as a function of the degree of methylation of
cg01307483 (NRF1). The solid
line indicates the predicted probability, and the gray shaded
region represents the 95% confidence interval. As the degree of
methylation of
NRF1 increases, the probability of child obesity at 36 months
increases significantly
Table 4 Association of baseline differential DNA methylationa
with obesity at 36 months, adjusted for co-variates
Odds Ratio 95% CI P Value
Child
Baseline BMI-Z 3.25 [1.00, 10.50] 0.049

Baseline age 1.50 [0.76, 2.94] 0.24
Gender (male) 0.59 [0.21, 1.63] 0.31
Parent
Baseline BMI 0.99 [0.92, 1.07] 0.86
CpG Baseline Methylation
Cg10307483 (NRF1) 2.98 [1.06, 8.38] 0.04
a Degree of methylation was a continuous variable calculated by
log-
transforming the normalized values and multiplying by 10 to put
it on a scale
from 0 to 10 (see Methods section)
6 of 9
some differential methylation patterns are associated
with the later emergence of obesity. While it is clear that
susceptibility to obesity within an “obesogenic” environ-
ment varies among individuals, it is not clear why. This
line of epigenetic inquiry using saliva as an accessible tis-
sue for pediatric study holds promise for guiding further
exploration in both understanding and intervening be-
fore the emergence of childhood obesity.
Although NFR1 was significantly related to child obes-
ity at 36-month follow-up, the relatively small sample
size analyzed in this study might have contributed to a
failure to detect important relationships for the other
CpG dinucleotides. Expanding the current analysis to in-

clude larger sample sizes would help to confirm and val-
idate the findings. While there was a strict collection
protocol for saliva collection, contamination and human
collection error are possible when collecting salivary
DNA. Although previous literature indicates DNA
methylation in saliva and blood samples are similar, the
current study only investigated methylation patterns in
saliva and cannot be used to make direct comparisons to
blood. Furthermore, although this sample yields insight
into Hispanic 3–5-year-olds, DNA methylation patterns
should be studied in children of various ages and race/
ethnicities.
Conclusions
Saliva offers a non-invasive means of DNA collection
and epigenetic analysis. This proof of principle study
provides empirical evidence supporting the idea that
DNA methylation assessed using salivary tissue collected
in non-obese children could be used as an important
predictor of childhood obesity 3 years later. NFR1 could
be a target for further exploration of obesity in Hispanic
children.
Supplementary information
Supplementary information accompanies this paper at
https://doi.org/10.
1186/s12881-020-0968-7.
Additional file 1: Table S1. CpG Dinucleotide Methylation.
Abbreviations
BMI: Body Mass Index; BMI-Z: Body Mass Index Z-score;
DNA: Deoxyribonucleic Acid; GROW: The Growing Right Onto
Wellness Trial;
NHANES: National Health and Nutrition Examination Survey;
VANTAGE: The

Vanderbilt Technologies for Advanced Genomics; WIC: The
Special
Supplemental Nutrition Program for Women, Infants, and
Children
Acknowledgements
We are grateful for the participation of the Hispanic families
involved in this
study.
Authors’ contributions
Conceived and designed experiment: SB, AR; Analyzed the
data: SZ, ES;
Wrote the paper: SB, AR, SZ, ES, EP. All authors edited and
proofed the
paper. All authors read and approved the final manuscript.
Funding
This research was supported by grants (U01 HL103620) with
additional
support for the remaining members of the COPTR Consortium
(U01HL103622, U01HL103561, U01HD068890, U01HL103629)
from the
National Heart, Lung, and Blood Institute and the Eunice
Kennedy Shriver
National Institute of Child Health and Development and the
Office of
Behavioral and Social Sciences Research. The content is solely
the
responsibility of the authors and does not necessarily represent
the official
views of the National Heart, Lung, And Blood Institute, the
National Institutes
of Health, or the National Institute of Child Health and
Development. This
research was also supported by grants 5P30DK092986–03 from

the National
Institute of Diabetes and Digestive and Kidney Diseases
(NIDDK) and
5UL1TR0045 from the Vanderbilt Institute for Clinical and
Translational
Research (VICTR). The NHLBI and NICHD played an advisory
role in all phases
of the study, including the design and conduct of the study;
collection,
analysis, and interpretation of the data; and in writing the
manuscript.
Availability of data and materials
The genetic dataset supporting the conclusions of this article are
available in
NCBI’s Gene Expression Omnibus and are accessible through
GEO Series
access number GSE72556
(https://www.ncbi.nlm.nih.gov/geo/query/acc.
cgi?acc=GSE72556).
Ethics approval and consent to participate
Data were collected after written informed consent was obtained
by parent/
legal guardian. This study was approved by the Vanderbilt
University
Institutional Review Board (IRB No. 120643).
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Author details
1Louisiana State University Health Sciences Center, School of

Medicine, 1901
Perdido Street, New Orleans, LA 70112, USA. 2Department of
Pediatrics,
Vanderbilt University Medical Center, 2146 Belcourt Ave,
Nashville, TN
37232-9225, USA. 3Department of Biostatistics, Vanderbilt
University Medical
Center, 571 Preston Research Building, 2220 Pierce Ave,
Nashville, TN
37232-6838, USA. 4Department of Pediatrics, Vanderbilt
University Medical
Center, 2146 Belcourt Ave, Nashville, TN 37232-9225, USA.
5Department of
Pediatrics, Vanderbilt University School of Medicine, 2200
Children’s Way,
Doctor’s Office Tower 8232, Nashville, TN 37232-9225, USA.
Received: 5 March 2019 Accepted: 6 February 2020
References
1. Ogden CL, Fryar CD, Hales CM, Carroll MD, Aoki Y,
Freedman DS.
Differences in obesity prevalence by demographics and
urbanization in US
children and adolescents, 2013-2016. JAMA.
2018;319(23):2410–8. https://doi.
org/10.1001/jama.2018.5158.
2. Ogden CL, Carroll MD, Lawman HG, Fryar CD, Kruszon-
Moran D, Kit BK, et al.
Trends in obesity prevalence among children and adolescents in
the United
States, 1988-1994 through 2013-2014. JAMA.
2016;315(21):2292–9. https://
doi.org/10.1001/jama.2016.6361.

3. Singh AS, Mulder C, Twisk JW, van Mechelen W, Chinapaw
MJ. Tracking
of childhood overweight into adulthood: a systematic review of
the
literature. Obes Rev. 2008;9(5):474–88.
https://doi.org/10.1111/j.1467-
789X.2008.00475.x.
4. Wang Y, Lobstein T. Worldwide trends in childhood
overweight and
obesity. Int J Pediatr Obes. 2006;1(1):11–25 Epub 2007/09/29.
5. Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence of
obesity among
adults and youth: United States, 2015-2016. NCHS Data Brief.
2017;288:1–8
PubMed PMID: 29155689.
6. Taveras EM, Gillman MW, Kleinman K, Rich-Edwards JW,
Rifas-Shiman SL.
Racial/ethnic differences in early-life risk factors for childhood
obesity.
Pediatrics. 2010;125(4):686–95. Epub 2010/03/03.
https://doi.org/10.1542/
peds.2009-2100.
7 of 9
https://doi.org/10.1186/s12881-020-0968-7
https://doi.org/10.1186/s12881-020-0968-7
https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE72556
)
https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE72556
)

https://doi.org/10.1001/jama.2018.5158
https://doi.org/10.1111/j.1467-789X.2008.00475.x
https://doi.org/10.1111/j.1467-789X.2008.00475.x
https://doi.org/10.1542/peds.2009-2100
7. CDC National Center for Health Statistics. National Health
and Nutrition
Examination Survey: questionnaires, datasets, and related
documents.
Available from: http://www.cdc.gov/nchs/nhanes. Accessed 1
Jan 2019.
8. Barker DJ, Osmond C, Forsen TJ, Kajantie E, Eriksson JG.
Trajectories of
growth among children who have coronary events as adults. N
Engl J Med.
2005;353(17):1802–9 Epub 2005/10/28. doi: 353/17/1802.
9. Kit BK, Kuklina E, Carroll MD, Ostchega Y, Freedman DS,
Ogden CL.
Prevalence of and trends in dyslipidemia and blood pressure
among US
children and adolescents, 1999-2012. JAMA Pediatr.
2015;169(3):272–9.
https://doi.org/10.1001/jamapediatrics.2014.3216.
10. Chen X, Wang Y. Tracking of blood pressure from
childhood to adulthood:
a systematic review and meta-regression analysis. Circulation.
2008;117(25):

3171–80.
https://doi.org/10.1161/CIRCULATIONAHA.107.730366.
11. Dubois L, Ohm Kyvik K, Girard M, Tatone-Tokuda F,
Perusse D, Hjelmborg J,
et al. Genetic and environmental contributions to weight,
height, and BMI
from birth to 19 years of age: an international study of over
12,000 twin
pairs. PLoS One. 2012;7(2):e30153. Epub 2012/02/22.
journal.pone.0030153.
12. Bouchard L, Rabasa-Lhoret R, Faraj M, Lavoie ME, Mill J,
Perusse L, et al.
Differential epigenomic and transcriptomic responses in
subcutaneous adipose
tissue between low and high responders to caloric restriction.
Am J Clin Nutr.
2010;91(2):309–20. Epub 2009/11/27.
https://doi.org/10.3945/ajcn.2009.28085.
13. Herrera BM, Keildson S, Lindgren CM. Genetics and
epigenetics of obesity.
Maturitas. 2011;69(1):41–9. Epub 2011/04/07.
https://doi.org/10.1016/j.
maturitas.2011.02.018.
14. Godfrey KM, Sheppard A, Gluckman PD, Lillycrop KA,
Burdge GC, McLean C,
et al. Epigenetic gene promoter methylation at birth is
associated with
child's later adiposity. Diabetes. 2011;60(5):1528–34.
db10-0979.

15. Hochberg Z, Feil R, Constancia M, Fraga M, Junien C,
Carel JC, et al. Child
health, developmental plasticity, and epigenetic programming.
Endocr Rev.
2011;32(2):159–224. https://doi.org/10.1210/er.2009-0039.
16. Mamtani M, Kulkarni H, Dyer TD, Goring HH, Neary JL,
Cole SA, et al.
Genome- and epigenome-wide association study of
hypertriglyceridemic
waist in Mexican American families. Clin Epigenetics.
2016;8:6. https://doi.
org/10.1186/s13148-016-0173-x.
17. Kulkarni H, Kos MZ, Neary J, Dyer TD, Kent JW Jr, Goring
HH, et al. Novel
epigenetic determinants of type 2 diabetes in Mexican-American
families.
Hum Mol Genet. 2015;24(18):5330–44.
https://doi.org/10.1093/hmg/ddv232.
18. King K, Murphy S, Hoyo C. Epigenetic regulation of
Newborns' imprinted
genes related to gestational growth: patterning by parental
race/ethnicity
and maternal socioeconomic status. J Epidemiol Community
Health. 2015;
69(7):639–47. https://doi.org/10.1136/jech-2014-204781.
19. Yoshizawa JM, Schafer CA, Schafer JJ, Farrell JJ, Paster
BJ, Wong DT. Salivary
biomarkers: toward future clinical and diagnostic utilities. Clin
Microbiol Rev.
2013;26(4):781–91. https://doi.org/10.1128/CMR.00021-13.
20. Bonne NJ, Wong DT. Salivary biomarker development using

genomic,
proteomic and metabolomic approaches. Genome Med.
2012;4(10):82.
https://doi.org/10.1186/gm383.
21. Abraham JE, Maranian MJ, Spiteri I, Russell R, Ingle S,
Luccarini C, et al. Saliva
samples are a viable alternative to blood samples as a source of
DNA for
high throughput genotyping. BMC Med Genet. 2012;5:19.
https://doi.org/10.
1186/1755-8794-5-19.
22. Langie SAS, Moisse M, Declerck K, Koppen G, Godderis L,
Vanden Berghe W,
et al. Salivary DNA methylation profiling: aspects to consider
for biomarker
identification. Basic Clin Pharmacol Toxicol. 2017;121(Suppl
3):93–101.
https://doi.org/10.1111/bcpt.12721.
23. Pfaffe T, Cooper-White J, Beyerlein P, Kostner K,
Punyadeera C. Diagnostic
potential of saliva: current state and future applications. Clin
Chem. 2011;
57(5):675–87. https://doi.org/10.1373/clinchem.2010.153767.
24. Oelsner KT, Guo Y, To SB, Non AL, Barkin SL. Maternal
BMI as a predictor of
methylation of obesity-related genes in saliva samples from
preschool-age
Hispanic children at-risk for obesity. BMC Genomics.
2017;18(1):57. https://
doi.org/10.1186/s12864-016-3473-9.
25. Po'e EK, Heerman WJ, Mistry RS, Barkin SL. Growing right

onto wellness
(GROW): a family-centered, community-based obesity
prevention randomized
controlled trial for preschool child-parent pairs. Contemp Clin
Trials. 2013;36(2):
436–49. Epub 2013/09/10.
https://doi.org/10.1016/j.cct.2013.08.013.
26. Heerman WJ, White RO, Hotop A, Omlung K, Armstrong S,
Mathieu I, et al.
A tool Kit to enhance the informed consent process for
community-
engaged pediatric research. IRB. 2016;38(5):8–14.
27. Barkin SL, Gesell SB, Po'e EK, Escarfuller J, Tempesti T.
Culturally tailored,
family-centered, behavioral obesity intervention for Latino-
American
preschool-aged children. Pediatrics. 2012;130(3):445–56. Epub
2012/08/08.
https://doi.org/10.1542/peds.2011-3762.
28. Bibikova M, Fan JB. GoldenGate assay for DNA
methylation profiling.
Methods Mol Biol. 2009;507:149–63. Epub 2008/11/07.
https://doi.org/10.
1007/978-1-59745-522-0_12.
29. Dedeurwaerder S, Defrance M, Bizet M, Calonne E,
Bontempi G, Fuks F. A
comprehensive overview of Infinium HumanMethylation450
data
processing. Brief Bioinform. 2014;15(6):929–41. Epub
2013/08/31. https://doi.
org/10.1093/bib/bbt054.

30. Wu Y, Zhou S, Smas CM. Downregulated expression of the
secreted
glycoprotein follistatin-like 1 (Fstl1) is a robust hallmark of
preadipocyte to
adipocyte conversion. Mech Dev. 2010;127(3–4):183–202.
https://doi.org/10.
1016/j.mod.2009.12.003.
31. Greenwood TA, Lazzeroni LC, Calkins ME, Freedman R,
Green MF, Gur RE,
et al. Genetic assessment of additional endophenotypes from the
consortium on the genetics of schizophrenia family study.
Schizophr Res.
2016;170(1):30–40.
https://doi.org/10.1016/j.schres.2015.11.008.
32. Bradfield JP, Taal HR, Timpson NJ, Scherag A, Lecoeur C,
Warrington NM,
et al. A genome-wide association meta-analysis identifies new
childhood
obesity loci. Nat Genet. 2012;44(5):526–31.
https://doi.org/10.1038/ng.2247.
33. Comuzzie AG, Cole SA, Laston SL, Voruganti VS, Haack
K, Gibbs RA, et al.
Novel genetic loci identified for the pathophysiology of
childhood obesity
in the Hispanic population. PLoS One. 2012;7(12):e51954.
https://doi.org/10.
1371/journal.pone.0051954.
34. Bartelt A, Widenmaier SB, Schlein C, Johann K, Goncalves
RLS, Eguchi K,
et al. Brown adipose tissue thermogenic adaptation requires
Nrf1-mediated
proteasomal activity. Nat Med. 2018;24(3):292–303.

nm.4481.
35. Wang W, Chan JY. Nrf1 is targeted to the endoplasmic
reticulum
membrane by an N-terminal transmembrane domain. Inhibition
of nuclear
translocation and transacting function. J Biol Chem.
2006;281(28):19676–87.
https://doi.org/10.1074/jbc.M602802200.
36. Patti ME, Butte AJ, Crunkhorn S, Cusi K, Berria R,
Kashyap S, et al.
Coordinated reduction of genes of oxidative metabolism in
humans with
insulin resistance and diabetes: potential role of PGC1 and
NRF1. Proc Natl
Acad Sci U S A. 2003;100(14):8466–71. Epub 2003/07/02.
https://doi.org/10.
1073/pnas.1032913100.
37. Atikuzzaman M, Alvarez-Rodriguez M, Vicente-Carrillo A,
Johnsson M, Wright
D, Rodriguez-Martinez H. Conserved gene expression in sperm
reservoirs
between birds and mammals in response to mating. BMC
Genomics. 2017;
18(1):98. https://doi.org/10.1186/s12864-017-3488-x.
38. McGovern DP, Gardet A, Torkvist L, Goyette P, Essers J,
Taylor KD, et al.
Genome-wide association identifies multiple ulcerative colitis
susceptibility
loci. Nat Genet. 2010;42(4):332–7.
https://doi.org/10.1038/ng.549.

39. Wu JN, Nguyen N, Aghazarian M, Tan Y, Sevrioukov EA,
Mabuchi M, et al.
Grim promotes programmed cell death of Drosophila
microchaete glial
cells. Mech Dev. 2010;127(9–12):407–17.
https://doi.org/10.1016/j.mod.2010.
06.001.
40. Mattiotti A, Prakash S, Barnett P, van den Hoff MJB.
Follistatin-like 1 in
development and human diseases. Cell Mol Life Sci.
2018;75(13):2339–54.
https://doi.org/10.1007/s00018-018-2805-0.
41. Kato H, Nomura K, Osabe D, Shinohara S, Mizumori O,
Katashima R, et al.
Association of single-nucleotide polymorphisms in the
suppressor of
cytokine signaling 2 (SOCS2) gene with type 2 diabetes in the
Japanese.
Genomics. 2006;87(4):446–58.
https://doi.org/10.1016/j.ygeno.2005.11.009.
42. Paterson AD, Waggott D, Boright AP, Hosseini SM, Shen E,
Sylvestre MP,
et al. A genome-wide association study identifies a novel major
locus for
glycemic control in type 1 diabetes, as measured by both A1C
and glucose.
Diabetes. 2010;59(2):539–49. https://doi.org/10.2337/db09-
0653.
43. Chow TJ, Tee SF, Loh SY, Yong HS, Abu Bakar AK, Tang
PY. Variants in
ZNF804A and DTNBP1 assessed for cognitive impairment in
schizophrenia

using a multiplex family-based approach. Psychiatry Res. 2018.
https://doi.
org/10.1016/j.psychres.2018.04.051.
44. Porcelli S, Lee SJ, Han C, Patkar AA, Albani D, Jun TY, et
al. Hot genes in
schizophrenia: how clinical datasets could help to refine their
role. J Mol
Neurosci. 2018;64(2):273–86. https://doi.org/10.1007/s12031-
017-1016-8.
45. Bai Y, Qiu S, Li Y, Li Y, Zhong W, Shi M, et al. Genetic
association between
SHANK2 polymorphisms and susceptibility to autism spectrum
disorder.
IUBMB Life. 2018;70(8):763–76.
https://doi.org/10.1002/iub.1876.
8 of 9
http://www.cdc.gov/nchs/nhanes
https://doi.org/10.1001/jamapediatrics.2014.3216
https://doi.org/10.1161/CIRCULATIONAHA.107.730366
https://doi.org/10.1371/journal.pone.0030153
https://doi.org/10.3945/ajcn.2009.28085
https://doi.org/10.1016/j.maturitas.2011.02.018
https://doi.org/10.1016/j.maturitas.2011.02.018
https://doi.org/10.2337/db10-0979
https://doi.org/10.2337/db10-0979
https://doi.org/10.1210/er.2009-0039
https://doi.org/10.1186/s13148-016-0173-x
https://doi.org/10.1186/s13148-016-0173-x
https://doi.org/10.1093/hmg/ddv232
https://doi.org/10.1136/jech-2014-204781

https://doi.org/10.1128/CMR.00021-13
https://doi.org/10.1186/gm383
https://doi.org/10.1186/1755-8794-5-19
https://doi.org/10.1186/1755-8794-5-19
https://doi.org/10.1111/bcpt.12721
https://doi.org/10.1373/clinchem.2010.153767
https://doi.org/10.1186/s12864-016-3473-9
https://doi.org/10.1186/s12864-016-3473-9
https://doi.org/10.1016/j.cct.2013.08.013
https://doi.org/10.1007/978-1-59745-522-0_12
https://doi.org/10.1007/978-1-59745-522-0_12
https://doi.org/10.1093/bib/bbt054
https://doi.org/10.1093/bib/bbt054
https://doi.org/10.1016/j.mod.2009.12.003
https://doi.org/10.1016/j.schres.2015.11.008
https://doi.org/10.1038/ng.2247
https://doi.org/10.1038/nm.4481
https://doi.org/10.1038/nm.4481
https://doi.org/10.1074/jbc.M602802200
https://doi.org/10.1073/pnas.1032913100
https://doi.org/10.1073/pnas.1032913100
https://doi.org/10.1186/s12864-017-3488-x
https://doi.org/10.1038/ng.549
https://doi.org/10.1007/s00018-018-2805-0
https://doi.org/10.1016/j.ygeno.2005.11.009
https://doi.org/10.2337/db09-0653
https://doi.org/10.1016/j.psychres.2018.04.051
https://doi.org/10.1016/j.psychres.2018.04.051
https://doi.org/10.1007/s12031-017-1016-8
https://doi.org/10.1002/iub.1876

46. Kim R, Kim J, Chung C, Ha S, Lee S, Lee E, et al. Cell-
type-specific Shank2
deletion in mice leads to differential synaptic and behavioral
phenotypes. J
Neurosci. 2018;38(17):4076–92.
https://doi.org/10.1523/JNEUROSCI.2684-17.
2018.
47. Sabin MA, Werther GA, Kiess W. Genetics of obesity and
overgrowth
syndromes. Best Pract Res Clin Endocrinol Metab.
2011;25(1):207–20. https://
doi.org/10.1016/j.beem.2010.09.010.
48. Sim CK, Kim SY, Brunmeir R, Zhang Q, Li H,
Dharmasegaran D, et al.
Regulation of white and brown adipocyte differentiation by
RhoGAP DLC1.
PLoS One. 2017;12(3):e0174761.
https://doi.org/10.1371/journal.pone.
0174761 PubMed PMID: 28358928; PubMed Central PMCID:
PMCPMC5373604.
49. Liao YC, Lo SH. Deleted in liver cancer-1 (DLC-1): a tumor
suppressor not
just for liver. Int J Biochem Cell Biol. 2008;40(5):843–7.
https://doi.org/10.
1016/j.biocel.2007.04.008.
50. Huang Y, Zheng J, Chen D, Li F, Wu W, Huang X, et al.
Transcriptome
profiling identifies a recurrent CRYL1-IFT88 chimeric
transcript in
hepatocellular carcinoma. Oncotarget. 2017;8(25):40693–704.

Association between PM exposure and childhood obesity

Association between PM exposure and childhood obesity

Recommended

Recommended

More Related Content

Similar to Association between PM exposure and childhood obesity

Similar to Association between PM exposure and childhood obesity (20)

More from aulasnilda

More from aulasnilda (20)

Recently uploaded

Recently uploaded (20)

Association between PM exposure and childhood obesity