PEDIA TRIC D E N TIS T R Y V 41 / NO 4 J U L /A U G 1 9

SYSTEMATIC REVIEW AND META-ANALYSIS □

Early Childhood Caries and Weight Status: A Systematic Review and Meta-Analysis
Matina V. Angelopoulou, DDS, MS, MPH' • Mitchell Beinlich2 • Alexander Crain3

Abstract: Purpose: The p u rp o s e s o f th is s tu d y w ere to : (1) e x a m in e th e a s s o c ia tio n b e tw e e n b o d y w e ig h t s ta tu s a n d caries e x p e rie n c e ; a n d (2)

d e te r m in e i f o v e rw e ig h t o r u n d e rw e ig h t c h ild re n a re a t g re a te r ris k o f d e v e lo p in g e a rly c h ild h o o d caries. Methods: A s e a rch o f s tu d ie s in P u b M e d ,
C o c h ra n e L ib ra ry, a n d G o o g le S c h o la r d a ta b a se s t h a t te s te d th e a s s o c ia tio n o f b o d y w e ig h t w ith d e n ta l caries in d e x in p re s c h o o l c h ild re n was
c o n d u c te d in A p r il 2018. S tu d ie s’ c h a ra c te ris tic s w ere e x tra c te d f o r th e q u a lita tiv e review , a n d m e a n s a n d s ta n d a rd d e v ia tio n s o f d e ca yed , m issin g ,
a n d fille d te e th /s u rfa c e s ( d m ft/d m fs ) in d e x b y b o d y m ass in d e x s ta n d a rd iz e d c a te g o rie s w ere r e trie v e d f o r q u a n tita tiv e syn th ese s. R a n d o m e ffe c ts
m e ta -a n a ly s e s w e re c o n d u c te d b y c a lc u la tin g th e m e a n d iffe re n c e a n d th e a s s o c ia te d 9 5 p e rc e n t c o n fid e n c e in te rv a ls . Results: A to ta l o f 2 9 3
s tu d ie s w ere id e n tifie d th ro u g h th e d a tab a se s’ search. T h ir ty - tw o q u a lifie d f o r q u a lita tiv e re v ie w a n d 12 o f th e m r e p o r te d d a ta to c o n d u c t a
m e ta -a n a ly s is . The m e ta -a n a ly s is s h o w e d t h a t o v e rw e ig h t/o b e s e c h ild re n a re a t a s ta tis tic a lly s ig n ific a n t g re a te r ris k o f h a v in g e a rly c h ild h o o d
caries (P<0.01). Conclusions: The re s u lts o f th e in c lu d e d s tu d ie s w ere in c o n s is te n t; h o w ever, th e m e ta -a n a ly s is s u g g e s te d t h a t o v e rw e ig h t/o b e s e
c h ild re n a re a t g re a te r ris k o f h a v in g e a rly c h ild h o o d caries. (P e d ia tr D e n t 2019;41(4):261-70.E 24-E 25) R eceived D e c e m b e r 9, 2018 I L a st R evision
A p r il 3 0 ,2 0 1 9 I A c c e p te d M a y 7 ,2 01 9

KEYW O R D S: EARLY C H IL D H O O D CARIES, BO D Y MASS IN D E X , O VERW EIG HT, U N D E R W E IG H T, O BESITY

Dental caries is the most prevalent chronic health problem in the increasing prevalence of obesity, including genetics, physical
children around the w orld.1 Caries has been associated with activity, socioeconomic status (SES), and individual behavior.9
frequent consumption of fermentable carbohydrates, and espe­ The strong association of dental caries and obesity with
cially sugars that are popular in children.2 Early childhood diet and sociopsychological factors suggests that a relationship
caries (ECC) is defined as “the presence of one or more decayed may exist between the two diseases. However, current liter­
(cavitated or noncavitated), missing, or filled tooth surface in ature is inconclusive as to whether dental caries is associated
any primary tooth in a child 71 months of age or younger.”3 with increased risk of being either overweight or underweight.
ECC is a public health concern because of both the short- Some studies suggest obese/overweight children are at greater
and long-term effects it has on the quality of life of preschool risk of having high caries than normal weight children.10' 16
children. More specifically, severe ECC has been found to This finding is based on the fact that frequent intake of sugar-
affect a child’s ability to eat, speak, grow, and socialize and sweetened drinks and sugary foods, and frequent snacking
cause varying degrees of pain.4 In addition, it has been related between meals, can be risk factors both for dental caries and
with increased utilization of the emergency room, need for obesity.17 By contrast, some other studies have shown that there
dental treatm ent under general anesthesia, frequent time of may be a relationship between high caries and being under­
missing school, and increased cost of dental treatment.3 weight.18'24 It is thought that caries may precede low weight in
Higher carbohydrate intake has also been shown to increase these situations, because of the pain these children experience
the risk for obesity.5 The Center of Disease Control and Pre­ as a result of severe decay, leading them to eat less food and,
vention estim ated that the prevalence of obesity was 18.5 as a result, weigh less.25,26 A systematic review and meta-analysis
percent, affecting approximately 13.7 million children and was recently conducted that aimed to investigate the associa­
adolescents.6 Studies show that, since the 1970s, the prevalence tion of dental caries and body mass weight in children and
of obesity in children between the ages of two and five years adolescents.27 However, this systematic review reported results
old has more than doubled,7 and approximately 10 percent of for children of all ages and not specifically in preschool years
these children were at or above the 95* percentile for body that are affected by ECC. Also, many dental caries studies in
mass index ( B M I ) . 8 There are many different risk factors for the prim ary dentition and body weight were published in
2017 and, thus, were n o t included in th a t system atic
review.22,28‘31 ECC has some unique characteristics from the
1Dr. Angelopoulou is an assistant professor, and 2Mr. Beinlich and 3Mr. Crain are dental dental perspective, such as its clinical expression and association
students, all in the Department o f Developmental Sciences, Division o f Pediatric Dentistry, with maternal oral streptococci levels and nocturnal feeding
School o f Dentistry, Marquette with sugar-containing beverages.3 Also, ECC affects preschool
University, Milwaukee, Wis., USA.
Correspond with Dr. Angelopoulou p
c
Supplemental material available
in the online version.
children whose weight status can have a significant effect in
their development, particularly in cases of obesity or failure to
at matinangelop@yahoo.gr
thrive. Thus, a separate review is necessary to investigate the
relationship of ECC to the body weight of preschool children.
HOW TO CITE: The purposes o f this study were to: (1) examine the
Angelopoulou MV, Beinlich M, Crain A. Early Childhood Caries and association between body weight status and caries experience
W eight Status: A Systematic Review and Meta-Analysis. Pediatr Dent status; and (2) determine if overweight, obese, or underweight
2019;41(4):26T70.E24-E25. children are at a greater risk of developing early childhood
caries compared to normal-weight children.

ECC A N D W E IG H T 261
PEDIATRIC DENTISTRY V 411 NO 4 JUL I AUG 19

Methods and these criteria would all be answered with a “no” or “non-
This review was prepared following PRISMA (Preferred Re­ applicable.” However, these criteria were included in the
porting Items for Systematic reviews and Meta-Analyses) presentation of the risk of bias. Based on these criteria, two of
guidelines.32 The protocol for this systematic review was the authors rated each study as good, fair, or poor. Studies
registered in the PROSPERO international register for system­ rated as “poor” were considered to be of high risk for bias,
atic reviews (CRD42018114608). No other studies on this while “good” rated studies were considered to be of low risk
topic were registered in PROSPERO. for bias.
Eligibility criteria. Inclusion and exclusion criteria were Summary measures. The primary outcome measured was
specified prior to the study. No date or language restriction the mean difference of dmft/dmfs between the four different
was applied during the search. Studies included in this system­ BMI categories. For studies that presented results in more
atic review and meta-analysis were required to: (1) have non- BMI categories, the results were combined accordingly to the
syndromic preschool children (zero to six years old) as study four BMI categories. I2 test for homogeneity and chi-square
participants; (2) report body weight; and (3) report caries test were used to assess heterogeneity between studies at the
experience index. Studies whose participants were older children / ><0.10 level. Random-effects meta-analyses were conducted
in a mixed dentition, teenagers, and adults were excluded by calculating the mean difference and the associated 95
from analysis. Studies that were nondental caries-specific, did percent confidence intervals (Cl). All analyses were performed
not include underweight and overweight groups, or investi­ using Review Manager 5.3 software.
gated the effects of dental rehabilitation, educational programs, Risk of bias across studies. Publication bias was assessed
or malnourishment on weight were also excluded. Finally, using standard funnel plots. Quality assessment of the evidence
review articles, systematic reviews, and authors’ replies were not of the meta-analysis for each outcome was performed using
included in the analysis. the Grading of Recommendations Assessment, Development,
Search strategy. A literature search of PubMed, Google and Evaluation (GRADE) system.35 The criteria were the
Scholar, and Cochrane databases was conducted by one author study design, risk of bias, inconsistency, indirectness, impreci­
and checked by a second author in April 2018. MeSH terms sion, publication bias, and upgrading criteria such as effect
used in the search included “early childhood caries,” “body magnitude, dose response, and plausible confounders. Based
mass index,” “overweight,” “underweight,” and “obesity” All on these criteria, the quality of evidence for each one of the
articles identified in the search were included in the screening outcomes was evaluated as high, moderate, low, or very low.
process after duplicate studies were excluded. The titles and
abstracts of the articles were first screened for relevance. In Results
cases where the title and abstract failed to provide sufficient Study selection. A total of 293 articles were identified through
information, the full text was reviewed to assess for relevance. the search of the databases (Figure 1). After removing six
To ensure that no relevant studies were missed in the initial
search, the reference lists of the remaining articles were then
hand-searched and subsequently screened. Additional studies
identified through this process were added to the pool of
full-text articles to be evaluated. This pool was then assessed
for eligibility for both qualitative and quantitative review.
Data collection. Data items were extracted from each
study by two authors and consisted of: (1) publication inform­
ation (journal, title, authors, date, and country); (2) sample
characteristics (sample size, age, and gender of the participants);
(3) weight and caries indices used; and (4) qualitative and
quantitative results. A Review Manager 5.3 software (Nordic
Cochrane Centre, Copenhagen) data extraction form was used
for this purpose. For the quantitative review, a BMI was used
that standardized into categories of “underweight” (BMI-for-
age percentile less than five), “normal weight” (BMI-for-age
percentile between five and 85), “overweight” (BMI-for-age
percentile between 85 and 95), and “obese” (BMI-for-age per­
centile greater than 95).33 Regarding caries, the mean and
standard deviation (±SD) of the reported decayed, missing, and
filled teeth/surfaces (dmft/dmfs) index, was retrieved.
Risk of bias within studies. The National Institute of
Health (NIH) Quality Assessment Tool for Observational
Cohort and Cross-Sectional Studies was used to assess risk of
bias in individual studies.3<i Each study was assessed for clear
presentation of the: (1) research question; (2) study popul­
ation; (3) recruitment criteria; (4) justification of the sample’s
size; (5) different levels of the exposure of interest; (6) exposure
measures and assessment; (7) existence of repeated measure­
ments; (8) outcome measures; and (9) adjustment of co­
founders. Four criteria of the tool were not included in the
assessment, as all of the included studies were cross-sectional Figure 1. Flow diagram o f study selection.

262 ECC AND WEIGHT

 PEDIATRIC DENTISTRY V 411 NO 4 JUL l AUG 19

Table 1. CHARACTERISTICS OF INCLUDED STUDIES

Author Date Country Participants Outcome measures Results

Miller et al.26 1982 UK 1,632 children (813 Comparison of weight (Tanner’s Children needing extractions (31.3% below 25,h
boys, 817 girls), weight percentiles 1-3, 4-10, 11-25, percentile) weighted less than the group requiring
mean age=5.9 years 26-50, 51-75, over 75) with caries dental restorations (17.1% below 25lh percentile)
severity (needing extractions vs. (PcO.OOl), most significant among girls
needing restorations)

Acs et al.18 1992 USA 330 children (198 Effect of ECC (dmft=0 vs. >0) on ECC associated with lower weight (/><0.005);
boys, 132 girls), body weight (percentile categories 8.7% of ECC children were underweight vs.
mean age=3.2 years <5, 5<x<10, 10<x<25, 25<x<50, 1.7% of caries-free children; underweight chil­
50<x<75,75<x<90, >90) dren with ECC were significantly older than
normal weight children (PcO.Ol)
Ayhan et al.20 1996 Turkey 126 children (63 Effect of ECC (dmft=0 vs. >0) ECC associated with lower weight and height
boys, 63 girls), mean on body weight percentile, height (TcO.OOl); 7.1% of ECC children were under­
age=4.0 years percentile and head circumference weight vs. 0.7% of caries-free children
(cm)

Chen et al.36 1998 Taiwan 5,133 children Association between BMI per­ No statistically significant association between
(2,822 boys, 2,311 centile (>95th, 75,h-95'h, 25,h-75,h, BMI and caries prevalence; caries prevalence:
girls), 3 years old 5th-25,h, <5,h) and caries prevalence under-weight=56.8%, normal=55.4%, obese=
(dmft index) 50.9%
Reifsnider 2004 USA 104 children (59 Association between BMI and caries Children with a greater degree of caries experi­
et al.15 boys, 45 girls), mean prevalence (dmft=0 vs. >0) ence tended to have higher BMI than those
age =1.4 years with less caries (/><0.05)
Macek et al.45 2006 USA 1719 children, age Association between BMI (“under­ No statistically significant association between
range = 2-5 years weight,” “normal,” “at risk of over­ BMI-for-age and caries prevalence; caries preval­
weight,” “overweight”) and caries ence: underweight=18.0%, normal=28.1%, at
prevalence (dmft=0 vs. >0) risk of overweight=26.9%, overweight=36.1%

Oliveira et al.43 2008 Brazil 1,018 children (519 Association between nutritional Children with lower weights were more prone
boys, 499 girls), age status (WAZ, HAZ, WHZ, BMZ) to caries (OR=3.20) than higher weight children
range =1-5 years social factors and caries prevalence (OR=0.58) compared to normal weight children
(dmfs=0 vs. >0 and >6) (/)=0.007and /)=0.046, respectively)

Floyd47 2009 New 577 children (298 Association between BMI (kg/m2) ECC associated with lower BMI (Z^O.036) in the
Zealand boys, 279 girls), and height with caries prevalence low socioeconomic area, whereas no statistically
6 years old (dmft index) in two areas with significant association was found in the high
different socioeconomic levels socioeconomic area

Sheller et al.37 2009 USA 293 children (162 Association between BMI (“under­ No statistically significant association between
boys, 131 girls), weight,” “normal,” “at risk for being BMI and caries prevalence; caries prevalence:
mean age =4.1 years overweight,” “overweight”) and caries underweight= 11%, normal=69%, at risk of
prevalence and severity (dmft index overweight=9%, overweight=l 1%; underweight
and number of pulp-involved teeth) children had more pulp-involved teeth compared
to normal weight children (P=0.52)

Ismail et al.44 2009 USA 788 children (372 Association between weight per­ Caries associated with higher body weight (/*=().03)
boys, 416 girls), centile (0-26.9%, 27-56.4%, 56.7-
mean age=2.6 years 84.2%, 84.3-100%) and other
family-level factors with caries pre­
valence (dmfs=0 vs. 2<7, 3>7)

Gaur and 2011 India 100 children (50 Effect of ECC (dmft=0 vs. >6) on ECC associated with lower weight (75=0.011);
Nayale21 boys, 50 girls), mean growth parameters (BMI, weight, mean weight of ECC children was 15.5 vs. 16.3
age =5.4 years height, head circumference) and of caries-free children; ECC children also had a
QoL in children from low socio­ lower QoL score, which improved after dental
economic families rehabilitation

Campos et al.39 2011 Brazil 491 children (232 Spatial dependence between BMI No statistically significant association between
boys, 259 girls), age (“underweight,” “normal,” “over­ caries and BMI; average dmft was 1.22±2.23,
range = 5 months- weight,” “obese”) and caries preval­ 9.4% were underweight, 59.5% were normal,
6 years ence (dmft index) 17.5% were overweight, 13.7% were obese

ECC AND W EIGHT 263

PEDIATRIC DENTISTRY V 41 / NO 4 JUL AUG 19

Table 1. CONTINUED

A u th o r D a te C o u n tr y P a rtic ip a n ts O u tc o m e m e a su re s R esu lts

Trikaliotis 2011 Greece 361 c h ild re n (1 8 3 A ssociation betw een B M I (“u n d e r­ O verw eight children had statistically significantly
et al.16 b o y s , 1 7 8 g ir l s ) , w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” larger dm fs th a n n o rm a l w e ig h t (PcO .O O l) and
m ean age=4 years “o b e s e ” ) a n d c a r ie s p re v a le n c e u n d e rw e ig h t c h ild re n (7J= 0 .0 1 5 ); m e a n d m fs:
(dmfs index) u n d e rw e ig h t= 1 .0 2 ± 2 .4 1, n o r m a l= 0 .7 4 ± 2 .2 4 ,
overw eight=1.88±4.28, obese=0.80±2.53

Vania et al.24 2011 Italy 8 3 0 c h ild re n (4 3 5 A ssociation betw een B M I (“u n d e r­ S ignificantly, m o re c h ild re n in th e E C C g ro u p
b o y s , 3 9 3 g ir l s ) , w eight,” “norm al,” “a t risk for being w ere u n d e rw e ig h t th a n in th e c o n tr o l g ro u p
m ean age=4.6 years o v e rw e ig h t,” “o v e r w e ig h t”) a n d (10% vs. 4.94% )
caries p re v a le n ce (d m ft= 0 vs. >0,
E C C categorized into non-cavitated
lesions, cavitated lesions, and severe
ECC)

N orberg 2012 Sweden 9 2 0 c h ild re n (4 6 6 A ssociation betw een B M I (“u n d e r­ H ig h e r caries prevalence in u n d e rw e ig h t (m ean
e t al.23 boys, 4 5 4 g irls), 5 w e ig h t,” “ low w e ig h t,” “n o rm a l d m ft= 2 .0 0 ) a n d lig h t w e ig h t c h ild re n (m e a n
years old w eight,” “high w eight,” “obese”) and d m ft= 1 .2 7 ) th a n n o rm al w eig h t ch ild re n (m ean
caries prevalence (dm ft index) dm ft= 0.65; .P=0.010 and P - 0.025, respectively)

Yen a n d H u 38 2013 Taiwan 3 2 9 c h ild re n (1 7 5 A ssociation b etw een B M I (“ lean,” T h e B M I ( 1 5 .8 8 ± 1 .8 3 vs. 1 5 .7 3 ± 1 .6 7 ) , W H I

b o y s , 1 5 4 g ir l s ) , “u n d e r w e ig h t,” “n o r m a l,” “o v e r­ in d e x ( 0 .9 9 ± 0 .1 4 vs. 0 . 9 7 ± 0 .1 3 ) , b o d y fa t
m ean age=4.8 years w eight,” “obese”), w eig h t-to -h eig h t (1 6 .8 7 ± 4 .5 0 vs. 1 6 .4 4 ± 4 .5 1 ), a n d trice p s s k in ­
in d e x ( W H I ) , tr i c e p s s k in f o ld fo ld (1 0 .2 0 ± 3 .3 4 vs. 9 .9 9 ± 2 .7 7 ) w ere n o t sig­
thickness, body fat (% ), and caries n ificantly different betw een the caries-free group
prevalence (dm ft index) and the caries group

Powell et al.14 2013 USA 2 1 5 c h ild re n (1 1 9 A ssociation betw een B M I (“under- O v e rw eig h t c h ild re n (PcO .O Ol) had h ig h e r p re­
boys, 96 girls), m ean w e ig h t/n o r m a l” vs. “o v e rw e ig h t/ valence o f caries th a n n o rm a l w e ig h t o r u n d e r­
age=4.2 years o b e s e ” ) a n d c a r ie s p r e v a le n c e w eight children
(dm ft=3-6, 7-10, >10)

Bagherian 2013 Iran 4 0 0 c h ild re n (211 A ssociation betw een B M I (“u n d e r­ H ig h e r caries p re v a le n ce in o v e rw e ig h t (m ean
and Sadeghi10 b o y s , 1 8 9 g ir l s ) , w eight,” “norm al,” “a t risk o f over­ dm fs= 10.39) th a n norm al w eight children (m ean
m ean age=4.6 years w e ig h t,” “o v e rw e ig h t”) a n d caries d m fs = 8 .8 4 ) o r u n d e r w e ig h t c h ild r e n (m e a n
prevalence (dmfs index) dm fs=4.89; T^O.001)

B hoom ika 2013 India 2 0 0 c h ild re n (1 0 0 A ssociation betw een B M I (“u n d e r­ E C C a ss o c ia te d w ith h ig h e r B M I (7, < 0 .0 5 );
et al.11 boys, 100 girls), age w eight,” “n orm al,” “a t risk o f over­ 10% o f E C C c h ild re n w ere o v e rw e ig h t vs. 3 %
range=3-6 years w e ig h t,” “o v e rw e ig h t”) a n d caries o f caries-free children
prevalence (dm ft=0 vs. >0)

C osta et al.40 2013 Brazil 3 0 3 c h ild re n (1 3 9 Association betw een BM I (“severely N o s ta tistic a lly s ig n ific a n t a sso c ia tio n betw een
b o y s , 1 6 4 g ir l s ) , th in ,” “th in ,” “a d e q u a te ,” “ over­ caries a n d B M I; average d m ft w as 2 .5 ± 3 .2 , and
m ean age=5.7 years w eight,” “obese”) and caries preva­ m ean B M I was 15.9±2.2
lence (dm ft=0 vs. >0 and >6)

dos Santos 2014 Brazil 3 2 0 ch ild ren , m ean A ssociation betw een B M I (“u n d e r­ P rev a le n c e o f E C C w as re la te d to low fa m ily
e t al.42 age=3.6 years w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” in c o m e , low b irth w e ig h t, in f a n t ob esity , an d
“o b e se ”), p e rin a ta l v a ria b le s an d shorter gestational age (P<0.05).
family incom e w ith caries prevalence
(dm ft=0 vs. >0)

R odriguez 2015 A rgentina 60 c h ild re n (3 0 A ssociation betw een B M I (“u n d e r­ N orm al w eight children m ore likely to have caries
e t al.46 boys, 30 girls), m ean w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” than overw eight children (56.7% vs.37% )
age=4.9 years “obese”), salivary flow, a n d caries
risk (dm ft=0 vs. >0)

K hanh et al.48 2015 V ietnam 5 9 3 c h ild re n (3 0 4 A ssociation betw een w eight, height E C C associated w ith lower B M I (P=0.006). M ean
b o y s , 2 9 0 g ir l s ) , a n d B M I w ith c arie s p re v a le n c e B M I o f carie s-fre e c h ild re n w as 1 .2 7 vs. 0 .8 2
m ean age= 4.1 years (d m ft= 0 vs. 1<5, >5) a n d m o u th o f th e E C C c h ild re n a n d 0 .6 2 o f th e severe
pain E C C children

A luckal et al.19 2016 India 4 3 3 c h ild re n (2 1 8 A ssociation betw een B M I (“u n d e r­ H ig h e r caries prevalence in u n d e rw e ig h t (m ean
b o y s , 2 1 5 g ir l s ) , w eight,” “n orm al,” “a t risk o f over­ d m ft= 2 .5 5 ) th a n n o rm al w e ig h t c h ild re n (m ean
m ean age=2.8 years w e ig h t,” “o v e rw e ig h t”) a n d caries d m f t = 1 .7 2 ) o r o v e r w e ig h t c h ild r e n ( m e a n
prevalence (dm ft index) d m ft= 1 .8 6 ;/>=0.0035)

264 ECC AND W EIGHT

 P E D IA TR IC D E N TIS T R Y V 4 1 1 NO 4 JUL / A U G 19

Table 1. CONTINUED

A u th o r D a te C o u n tr y P a rtic ip a n ts O u tc o m e m e a su re s R esu lts

da Silva et al.41 2016 Brazil 65 c h ild r e n , age A ssociation betw een B M I (“u n d e r­ N o s ta tistic a lly s ig n ific a n t a sso c ia tio n betw een
range=2-5 years w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” caries an d B M I; average d m ft was 6 .2 ± 4 .4 , and
“o b e s e ”) a n d c a rie s p r e v a le n c e m ean B M I was 16.5±2.5
(d m ft index)

D avidson 2016 C anada 2 3 5 c h ild re n (1 1 8 A ssociation betw een B M I (“u n d e r­ E C C a ss o c ia te d w ith h ig h e r B M I (P = 0 .0 3 8 ) ;

et al.12 b o y s , 1 1 7 g ir l s ) , w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” 6 9 .0 % o f E C C c h ild re n w ere o v e rw e ig h t vs.
m ean age=3.6 years “o b e s e ”) a n d c a rie s p re v a le n c e 56.8% o f caries-free children
(dm ft=0 vs. >0)

P ikram enou 2016 Greece 2 ,1 8 0 c h ild re n A ssociation betw een B M I (“u n d e r­ O v e rw eig h t c h ild re n (P < 0 .0 0 1 ) a n d obese c h il­
e t al.13 (1 ,1 7 3 boys, 1,0 0 7 w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” dren (P = 0 .0 0 8 ) w ere m o re likely to have h ig h er
girls), m ean age=4.2 “obese”) and caries prevalence (dmfs dm fs th a n n o rm a l w e ig h t o r u n d e rw e ig h t c h il­
years index) dren.

A ntunes et al.28 2017 Brazil 4 8 8 c h ild re n (2 4 7 A ss o c ia tio n o f M T R a n d M T R R H ig h e r caries prevalence in u n d e rw e ig h t (m ean

b o y s , 2 4 1 g ir l s ) , g en es w ith B M I (“u n d e rw e ig h t,” d m ft= 2 .1 9 ) th a n n o rm a l w e ig h t c h ild re n (m ean
m ean age=3.5 years “norm al,” “overw eight,” “obese”) and d m f t = 1 .3 3 ) o r o v e r w e ig h t c h ild r e n ( m e a n
caries prevalence (dm ft index) d m f t= 0 .9 0 ; P < 0 .0 5 ) ; a lso , M T T R w as c o r ­
related w ith b oth E C C and low BM I

K rishna e t al.22 2017 India 3 5 0 c h ild re n (1 8 8 A ssociation betw een B M I (“u n d e r­ T r e n d o f h ig h e r c a rie s p re v a le n c e in u n d e r ­
boys, 162 girls), age w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” w e ig h t g ro u p (P = 0 .0 6 6 ) ; m e a n d m ft: u n d e r-
range=3-6 years “o b e s e ” ) a n d c a r ie s p r e v a le n c e w e ig h t= 4 .9 6 ± 4 .0 9 , n o r m a l= 4 .6 2 ± 3 .3 5 , o v e r-
(d m ft index) w eight=2.73± 1.90, obese=4.49±3.08

M adsen et al.29 2017 G reenland 3 7 3 c h ild re n (1 8 5 A sso ciatio n betw een B M I (“th in ,” T re n d o f h ig h e r caries prevalence in o verw eight
boys, 188 g irls), 6 “norm al,” “overw eight,” “obese”) and a n d obese c h ildren (P = 0 .0 6 3 ); caries prevalence:
years old caries prevalence (d m ft/D M F T = 0 t h i n = 5 0 . 0 % , n o r m a l = 5 5 .5 % , o v e r w e ig h t =
vs. >0) 64.1% , obese=73.7%

M itrakul 2017 T h ailan d 1 0 0 c h ild r e n (51 A ssociation betw een B M I (“u n d e r­ N o a ss o c ia tio n b e tw e e n d m f t scores a n d B M I
e t al.30 boys, 49 girls), m ean w e ig h t,” “n o rm a l,” “o v e rw e ig h t,” ( P = 0 .1 5 7 ) ; m e a n d m ft: u n d e r w e ig h t= 3 .0 0 ±
age=4.2 years “obese”) and caries prevalence (dm ft 4.24, norm al= 5.77±4.87, overw eight=4.20±4.43
index)

Soares et al.31 2017 Brazil 2 8 5 c h ild re n (131 E f fe c t o f m a s t ic a to r y f u n c tio n , C hildren w ith a greater degree o f caries experience
b o y s , 1 5 4 g ir l s ) , so c io -e c o n o m ic sta tu s a n d d e n ta l te n d e d to have low er B M I th a n th o se w ith less
m ean age=4.2 years caries (dm ft index) o n B M I (“u n d e r­ caries (P=0.04)
w eight,” “ideal,” “overw eight/obese”)

* ECC=early childhood caries; dmft=decayed, missing, fdled teeth; dmfs=decayed, missing, filled surfaces; BMI=body mass index; OR=odds ratio, W AZ-weight
for age; H A Z =height for age; W H Z = w eig h t-h eig h t for age; B M Z=body mass index for age; Q oL= quality o f life; W H I= w eight-height index; M TR =5-
methyltetrahydrofolate-homocysteine methyltransferase; MTRR=5-methyltetrahydrofolate-homocysteine methyltransferase reductase.

duplicate studies and 149 studies irrelevant to the topic or 84 Brazil,28,31,39"43 In d ia,11,19,21,22 and the U nited States14,15,18,37,44,45;
related to children over the age o f six years and adults, a total European countries have also published data on this topic13,16,
o f 55 studies were considered for full-text assessment. Few 23,24,26 (Table 1).
studies were published in a language other than English and The m edian sample size o f the included studies was 355,
were excluded after translation o f their title or abstract, as they an d m ost had a sam ple o f 100 to 500 ch ild ren (Table 1).
did n ot fulfill other criteria. A nother 13 studies were added Five studies had a sample o f 100 children or less,11,21,30,41,46 six
from hand searching the reference lists o f the full-text articles. studies had a sample o f 500 to 1,000 children,23,24,29,44,47,48 and
Follow ing full-text evaluation, 32 studies were included for four studies had m ore th an 1,000 children included in the
qualitative assessm ent b u t only 12 provided sufficient data study.13,24,36,43 T h e total num ber o f participants in all studies
in th e article th a t qualified for q u an tita tiv e synthesis. O f included in the systematic review was 21,351 patients; 11,272
th ese 12 stu d ies, th ree p rese n ted results u sing th e dm fs o f them were included in the m eta-analysis. All studies in ­
in d ex '0,13,16 and were analyzed separately from the rem aining cluded children o f both genders, 47.13 percent o f the partici­
nine studies that used the dm ft index.19,22,23,28,30,31,36"38 pants were boys an d 43.01 percent o f them were girls. T he
Q ualitative analysis: study characteristics. All included percentages including only data from the studies o f the m eta­
studies were cross-sectional and published between 1982 and analysis were 53.47 percent and 46.53 percent, respectively.
2017 (Table 1). A trend for more studies on the topic in the T he age range o f the participants varied from five m onths to
last 10 years was noted, as 27 o f the 32 selected studies were six years old, and the mean age was 4.22 years.
published from 2008 onw ard. M ost studies originated from

ECC A N D W E IG H T 265
P E D IA TR IC D E N TIS T R Y V 41 / NO 4 JUL I A U G 19

Regarding the outcom e tested, most studies examined the o f the p articip a n ts in total were co n sidered u n d erw eig h t,
association between BMI and caries experience (Table 1). Five 68.09 percent were considered norm al, 15.05 p ercent were
studies tested the association o f different or additional growth considered overw eight, and 6 .28 p ercen t were co n sidered
param eters besides B M I.38,43,44,47,48 F our studies tested the obese. Based on the results o f the nine studies that used the
effect o f ECC on body weight and growth param eters,1820,21,31 dm ft index,19,22,23,28,30,31,36'38 overweight/obese children are at a
and six studies included the SES level o f the fam ily in the statistical significant greater risk of having ECC (Figure 3). No
analysis.21,31,42'44,47 Fifteen studies com pared growth
param eters in children w ith caries and caries-free
ch ildren,11,12,15,18,20,21,24,29,40,42'46,48 one study compared Research question

children in need o f extractions versus dental rehab­ Study population

ilitation due to caries,26 and the remaining 16 studies Participation rate

used the dm ft/dm fs index for caries experience. Eight Recruitment criteria
studies took into consideration the severity o f dental Sample size justification
caries in their analysis.l4-24-26.37.4°.43.44,48 Exposure assessed prior to outcome measurement
Risk o f bias w ithin studies. Overall, 14 studies Sufficient timeframe to see an effect
were assessed as good w ith low risk o f bias, 16 as Different levels o f exposure of interest
fair with moderate risk o f bias, and two as poor with Exposure measures and assessment
high risk o f bias based on the N IH Q uality Assess­ Repeated exposure assessment
m e n t Tool for O b serv atio n al C o h o rt an d C ross- Outcome measures
Sectional Studies. The percentages o f compliance for
Blinding o f outcome assessors
each item of the quality assessm ent tool across all Follow up rate
in clu ded studies are p resented in Figure 2. M ore Statistical analyses
specifically, the assessment on three criteria (exposure
assessed prior to outcom e m easurem ent, sufficient
timeframe to see, and effect and blinding o f outcome r~f Low risk of bias | | Unclear risk of bias | High risk of bias
assessors) indicated high risk o f bias, as all o f the in­
cluded studies were cross-sectional and these criteria
Figure 2. Risk of bias within studies. Percentages o f compliance for each item o f the N ' H
were answ ered w ith a “n o ” (i.e., 100 p ercent red
quality assessment tool across all included studies. The criteria answered with a “no” are
columns, Figure 2; one criterion (follow-up rate) was presented with red indicating high risk o f bias. The criteria answered as “nonapplicable,”
answered as “non-applicable,” indicating unclear risk are presented with yellow indicating unclear risk of bias.
o f bias (i.e, 100 percent yellow, Figure 2).
O f the other criteria, all o f the studies pre­
Mean D ifference Mean D ifference
sented the research question and the study S tu d y o r S u b g ro u p W e ig h t IV, R andom , 95% Cl IV, R andom , 95% Cl
p o p u la tio n , m ost o f them presented the Chen e ta l., 1998 24.1% 0.60 [0.20. 1.00]
recruitm ent criteria, different levels of expo­ Shelter e ta l., 2008 1.7% 0.50 [-1.55, 2.55]
Norberg et al., 2012 13.0% 0.89 [0.24, 1.54] — —
su re, ex p o su re m easu res, an d o u tc o m e Yen & Hu, 2013 2.1% -0.70 [-2.56, 1.16]
measures decreasing the risk o f bias. T he Aluckal et al.. 2016 3.8% 1.85 [0.51 ,3.19]
Soares et al.. 2017 47.1% 0.80 [0.73, 0.87] ■
criteria that increased the risk o f bias o f the Antunes et al., 2017 4.4% 1.29 [0.04, 2.54]
included studies were the lack o f presen­ Mitrakul et al., 2017 0.4% -1.20 [-5.70, 3.30]
ta tio n o f th e p a rtic ip a tio n rate, sam ple Krishna et al., 2017 3.5% 2.23 [0.82, 3.64]

justification, repeated exposure assessment, T o ta l (95% Cl) 100.0% 0.83 [0 .5 6 ,1 .1 1 ] ♦

and statistical analyses used. H eterogeneity Tau* = 0.04; Chi’ 11.30, d f - 8 (P = 0.19); l? - 29% -----------1---------- 1---------- ---------- 1---------- 1----------------
Test for o ve ra l effect: Z = 5.95 (P < 0.000 01)
Results o f individual studies. Eleven Underweight Overweight

stu d ies in d icated th a t u n d erw eig h t chil­

Figure 3. Forest plot comparing the decayed, missing, and filled teeth (dmft) index between overweight
d ren have m ore c a rie s,18'24,26,28,31,43,48 1 1
and underweight zero- to six-year-olds.*
studies found a nonsignificant difference in
* df=degrees of freedom; CI=confidence interval.
caries experience between overw eight and
underw eight children,29,30,36'41,45'47, and nine
stu d ies in d ic ated th a t o v erw eight/obese Mean D ifference Mean D ifference
S tu d y o r S u b g ro u p W e ig h t IV, R andom , 95% Cl IV, R andom , 95% Cl
children have more caries10' 16,42,44 (Table 1).
Chen et al.. 1998 23.1% 0.45 [0.09. 0.81]
Also, the association between E C C preval­ Shelter et al.. 2008 2.3% -0.301-1.88, 1.28]
ence and being u n d e rw e ig h t was fo u n d Norberg et al., 2012 13.3% 0.99 [0.43, 1.55]
Yen & Hu, 2013 4.8% -0.65 [-1.71,0.41]
m ore sig n ific an t in old er c h ild re n 18 and Aluckal et al., 2016 3.6% 0.83 [-0.41,2.07]
girls.26 Krishna et al.. 2017 5.9% 0.34 [-0.60, 1.28]
Mitrakul et al.. 2017 0.3% -2.77 [-7.08. 1.54]
Quantitative analysis: synthesis o f the
Antunes et al.. 2017 3.9% 0.86 [-0.32, 2.04]
results. The outcom e o f interest was tested Soares et al., 2017 42.8% 0.42 [0.34, 0.50]
using the d m f index in two different ways
Total (95% Cl) 100.0% 0.45 [0.21, 0.70]
o f assessment: that o f the num ber o f teeth Heterogeneity: Tau* = 0.03; Chi? = 11.73, d f = 8 (P = 0. = 32%
(dm ft) affected and that o f the num ber of -1 0
Test for overa'I effect: Z = 3.61 (P = 0.0003)
Underweight Normal
carious surfaces (dmfs). Both syntheses con­
sisted o f comparisons between the different Figure 4. Forest plot comparing the decayed, missing, and filled teeth (dmft) index between normal
BMI groups. O f all the studies th a t were and underweight zero- to six-year-olds.*
included in the meta-analysis, 10.57 percent * df=degrees of freedom; CI=confidence interval.

266 ECC A N D W E IG H T
 PEDIATRIC DENTISTRY V 411 NO 4 JUL I AUG 19

statistically significant heterogeneity was found on this syn­ significant statistical heterogeneity was evident on the synthesis
thesis (overweight: I2 equals 29 percent, / >=0.19; obese: I2 equals of the overweight versus underweight comparison (I2 equals
zero percent, P = 0.96; Figure 3). The com parison between 54 percent, /V 0 .11; Figure 8). Flowever, statistically signifi­
children with normal BMI and underweight or overweight cant heterogeneity was detected on the synthesis between the
children was also found statistically significant (Figures 4 obese and underweight group (I2 equals 82 percent, /V 0.05;
and 5), indicating that children with higher weight are at see Electronic Appendix: Figure 9). The comparison between
greater risk of having ECC. More specifically, underweight the normal weight children and the other three groups was
children ( n equals 806) have a significantly lower dm ft not statistically significant either. More specifically, the dmfs
index than normal weight children ( n equals 5,653; mean index did not differ between underweight children ( n equals
difference equals 0.45, 95 percent C l [95% C l] equals 385) and normal weight children ( n equals 2019; mean dif­
0.21,0.70, /VO.001, I2 equals 32 percent; Figure 4). Overweight ference equals -0.23, 95 percent Cl equals -1.15 to 0.69, /V 0.62,
children ( n equals 1,338) have a significantly higher dmft I2 equals 73 percent; see Electronic Appendix: Figure 10),
index than normal weight children ( n equals 5,653; mean dif­ overweight children ( n equals 358), and normal weight chil­
ference equals -0.39, 95% Cl equals -0.64 to -0.14, P = 0.002, dren ( n equals 2019; mean difference equals 0.14, 95% Cl
I2 equals 62 percent; Figure 5). Obese children ( n equals 529) equals -0.12 to 0.41, /V 0 .28, I2 equals zero percent; see
have similar dmft index to normal weight children ( n equals Electronic Appendix: Figure 11), or obese children ( n equals
5,486; mean difference equals -0.07, 95% Cl equals -0.31 to 179) and normal weight children ( n equals 2019; mean dif­
0.17, P = 0.57, I2 equals zero percent; Figure 6). Underweight ference equals 0.35, 95% Cl equals -0.09 to 0.79, P = 0.12,
children ( n equals 806) have a significantly lower dmft index I2 equals zero percent; see Electronic Appendix: Figure 12).
than overweight children ( n equals 1,338; mean difference Also, dmfs did not differ between underweight ( n equals 385)
equals 0.83, 95% C l equals 0.56 to 1.11, /VO.01, I2 equals and overweight children ( n equals 358) (Mean difference
29 percent; Figure 3). Finally, underweight children (n equals equals -0.77, 95% Cl equals -2.14 to 0.60, /V 0.27, I2 equals
707) have a significantly lower dm ft index than obese chil­ 54 percent), or between underweight ( n equals 385) and obese
dren ( n equals 529; mean difference equals 0.62, 95% Cl children ( n equals 179; mean difference equals -1.73, 95%
equals 0.27 to 0.97, /V 0 .001, I2 equals zero percent; see C l equals -3.57 to 0.83, P = 0.22, I2 equals 82 percent; see
Electronic Appendix: Figure 7). Electronic Appendix: Figures 8 and 9).
Considering the results of the three studies assessing caries Risk o f bias across studies. Statistically significant hetero­
with the dmfs index,10,13,16 no difference was found on the geneity on the syntheses related to the research questions was
ECC prevalence between overweight/obese and underweight only found for the dmfs studies com paring the u n d er­
children (see Electronic Appendix: Figures 8 and 9). No weight and obese groups (I2 equals 82 percent, /V 0 .05; see
Electronic Appendix: Figure 9). However, publication bias
could not be evaluated, as there were only
three studies combined in this synthesis,
Mean Difference Mean Difference
Study or Subgroup Weight IV, Random, 95% Cl IV, Random, 95% Cl and for this reason, no funnel plots are
Chen et al., 1998 23.5% -0.151-0.37,0.07) presented.
2.6% -0.80 [-2.29, 0.69)
Shelter et a l..2008 Based on the GRADE assessment, the
Norberg et al., 2012 16.9% 0.10 (-0.29, 0.49]
Yen & Hu, 2013 2.0% 0.05 1-1.67. 1.77] quality of evidence for the meta-analysis
Aluckal et al., 2016 9.7% -1.02 [-1.67,-0.37] using dmft index was found to be moder­
Soares et al., 2017 28.2% -0.38 [-0.45, -0.31] ■

Antunes et al., 2017 12.0% -0.43 [-0.98.0.12]

ate (see Electronic Appendix: Table 2),
Mitrakul et al.. 2017 1.4% -1.57 [-3.64,0.50] while the evidence for the dmfs index was
Krishna et al., 2017 3.8% -1.89 [-3.09,-0.69]
assessed to be o f very low quality (see
Total (95% Cl) 100.0% -0.39 [-0.64, -0.14] ♦ Electronic Appendix: Table 2).
Heterogeneity: Tau2 = 0.06; Chi7 21.17, df = 8 (P = 0.007); I2 = 62%
-2 -1 \ 2
Test for overall effect: Z = 3.05 (P = 0.002) Overweight Normal
Discussion
T he results of the present meta-analysis
Figure 5. Forest plot comparing the decayed, missing, and filled teeth (dmft) index between normal and
suggest that overweight and obese preschool
overweight zero- to six-year-olds.*
children are at greater risk of having early
* df=degrees o f freedom; CI=confidence interval.
childhood caries. However, this finding
should be interpreted with caution due to
Mean Difference Mean Difference the risk of bias of cross-sectional studies.
Study or Subgroup W eight IV, Random , 95% Cl IV, Random , 95% Cl The present study followed the PRISMA
-0.15 [-0.46, 0.16]
Chen e ta l., 1998 63.0% guidelines, in contrast to previous system­
Sheller et al., 2008 2.7% 0 .2 0 [-1 .2 9 , 1.69]
Norberg et al.. 2012 15.9% 0 .1 8 [-0 .4 3 . 0.79]
atic reviews, and was the only one registered
Yen & Hu, 2013 2.3% -1.25 [-2.84, 0.34] in the PROSPERO database on this topic.
Aluckal et al., 2016 6.0% 0.14 [-0.85, 1.13]
Studies that follow the PRISMA guide­
Antunes et al . 2017 8.9% 0.13 [-0.68. 0.94]
Krishna et al.. 2017 1.2% -0.13 [-2.31. 2.05] lines and are registered in PROSPERO
have been found to be of higher quality,32,49
Total (95% Cl) 100.0% -0.07 [-0 .3 1 ,0 .1 7 ] ►

3.55, df = 6 (P = 0.74); I2 = 0%
which is one of the strengths of the present
Heterogeneity: Tau2 = 0.00; Chi2
-2 -1 l 2
Test for overall effect: Z = 0.57 (P = 0.57) Obese Normal
study.
The purpose of the study was to test
Figure 6. Forest plot comparing the decayed, missing, and filled teeth (dmft) index between normal and the association of caries and body weight
obese zero- to six-year-olds.* in preschool children. Previous systematic
* df=degrees o f freedom; CI=confidence interval. reviews have tested this association in older

ECC AND WEIGHT 267

PEDIATRIC DENTISTRY V 41 / NO 4 JUL I AUG 19

children and adolescents.27,50,51 However, ECC has unique The meta-analysis of the three studies10,13,16 that used the
characteristics both from a clinical and etiologic perspective dmfs index did not find any differences in ECC prevalence
and may have severe consequences in the child’s development. between overweight and underweight children. This finding
More specifically, ECC has been associated with an increased could be attributed to the different confounding factors, as
risk for caries in the permanent dentition, increased need for explained earlier. However, this analysis presented some hetero­
treatment under general anesthesia, and failure to thrive.3 geneity, which could not be explored further as only three
For this reason, only studies reporting separate data for pre­ studies were included. Also, based on the GRADE assess­
school children were included in this systematic review, ment, the evidence of this synthesis was found of low quality
whereas, studies that included older children, teenagers, and and, thus, cannot be trusted to derive any conclusions.
adults were excluded from analysis. Also, many studies have The results of the meta-analysis of the nine studies19,22-23,28,
tested the association of being obese/overweight with caries51 30,31,36-38 t}iat u s e d the dmft index found that overweight/
or compared underweight or overweight children with normal obese children are at a significantly greater risk statistically of
weight children individually.27 However, the purpose of this having ECC. Based on the GRADE assessment, the evidence
review was to test if children with ECC are at higher risk of of this synthesis was found to be of moderate quality, sug­
being overweight or underweight. Thus, only studies that gesting this outcome more trustworthy. Similar results have
included both underweight and overweight groups were in­ been found in the secondary analyses of a previous systematic
cluded in the present review. review.27 ECC has been strongly correlated with specific
For assessment of the risk of bias, we used the NIH feeding practices, such as nocturnal bottle feeding with sugar-
Quality Assessment Tool for Observational Cohort and Cross- containing drinks such as milk, formula, and juice.3 Obesity
Sectional Studies34 as all of the included studies had a cross- has also been associated with consumption of sugar-sweetened
sectional design. Although this is an accredited tool to assess beverages in addition to high fat diet and less exercise.7 In
risk of bias of cross-sectional studies, this design by default addition, both obesity and ECC are considered chronic diseases
has an increased risk of bias.52 Also, quality assessment of the of multifactorial etiology that have severe public health effects.56
outcomes was performed using the GRADE tool, which is More specifically, they are both associated with low SES level
considered important to evaluate the quality of the evidence and result in increased medical costs.3,53,56 Besides the common
presented in each meta-analysis.35 ground in the etiology of these two diseases, the association
The results of the included studies in the qualitative found in this study can help develop public health programs
analysis were inconsistent, as 11 studies indicated that under­ and strategies that will target both ECC and obesity. This
weight children have more caries,18'24,26,28,31,43,48 1 1 studies way, the cost can be diminished while the effectiveness and
found a nonsignificant difference in caries experience between target population increases.
overweight and underweight children,29,30,36'41,45'47 and nine One of the limitations of the present study is the in­
studies indicated that overweight/obese children have more clusion of cross-sectional studies, which are considered of lower
caries,10'16,42,44 (Table 1). This heterogeneity between studies quality as they present higher risk of bias and lower the quality
has also been reported in previous systematic reviews50,51 and of evidence. Another limitation is the fact that studies varied
could be associated with the multiple confounding factors in regard to the sample’s SES, age, culture, and other aspects
that influence both ECC and body weight. that could be considered confounding factors. Also, the sample
Although this systematic review included only studies with size varied significantly among the studies as well as the index
preschool children, the age of the participants varied between used for reporting body weight, BMI, and caries.
them. One study reported that older children with high caries The results of this systematic review and meta-analysis
experience were underweight.18 In addition, other studies suggest that longitudinal studies that will examine the growth
also reported that children with severe ECC and pulp-involved parameters and caries in different intervals are necessary in
teeth were at higher risk of being underweight.26,37 It is order to draw stronger conclusions. Also, rhe association that
known that caries experience increases with age and can be was found between overweight children and ECC suggests that
assumed that, if left untreated, could result in pain and less public health programs targeting ECC and obesity are neces­
food consumption. Thus, studies that included older pre­ sary and can help prevent these two diseases with less cost.
schoolers may report higher association of ECC with being
underweight whereas studies with younger participants may Conclusions
report an association of ECC with being overweight or obese. 1. The results of the included studies from the current
Another confounding factor that can explain the discre­ literature were inconsistent.
pancies between the studies is the SES level of the participants. 2. The present meta-analysis suggests that overweight
It has been found that both ECC and obesity increase in and obese preschool children are at greater risk of
children with low SES family level.53 Thus, based on the SES having early childhood caries.
level of the participants in different studies, the results may 3. However, this finding should be interpreted with
vary. Most of the studies that included the SES in their anal­ caution, as all of the included studies were cross-
ysis found that children of low SES were underweight and sectional, presented moderate quality of evidence, and
had higher levels of ECC.21,31,47 Also, culture and race can have have a relatively high risk of bias.
an effect on the body weight and caries as a result of genetic
and environmental factors.54,55 The studies selected in this References
systematic review originated from 15 different countries and, 1. Bagramian RA, Garcia-Godoy F, Volpe AR. The global
as a result, the findings for each study may vary due to cultural increase in dental caries: a pending public health crisis.
and racial differences. Am J Dent 2009;22(l):3-8.
2. Touger-Decker R, Van Loveren C. Sugars and dental
caries. Am J Clin Nutr 2003;78(4):881S-892S.

268 ECC A N D W E IG H T
 PEDIATRIC DENTISTRY V 411 NO A JUL / AUG 19

3. American Academy of Pediatric Dentistry. Policy on early 22. Krishna H, Manaswini E, Kumar VY, Bellamkonda P,
childhood caries (ECC): classifications, consequences, and Bhargava ASK, Jaidupally RR. Association between nutri­
preventive strategies. Pediatr Dent 2017;39(6):59-61. tional status and early childhood caries in Indian children.
4. Colak H, Dulgergil CT, Dalli M, Hamidi MM. Early J Int Soc Prev Community Dent 2017;7(3):131-5.
childhood caries update: a review of causes, diagnoses, 23. Norberg C, Hallstrom Stalin U, Matsson L, Thorngren-
and treatments. J Nat Sci Biol Med 2013;4(l):29-38. Jerneck K, Klingberg G. Body mass index (BMI) and
5. Ferretti F, Mariani M. Simple vs. complex carbohydrate dental caries in 5-year-old children from southern Sweden.
dietary patterns and the global overweight and obesity Community Dent Oral Epidemiol 2012;40(4):315-22.
pandemic. Int J Environ Res Public Health 2017; 14( 10): 24. Vania A, Parisella V, Capasso F, et al. Early childhood caries
1174-85. underweight or overweight, that is the question. Eur J Paediatr
6. Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence Dent 2011; 12(4):231-5-
of obesity among adults and vouth: United States, 2015- 25. Acs G, Shulman R, Ng MW, Chussid S. The effect of
2016. NCHS Data Brief 2017;(288): 1-8. dental rehabilitation on the body weight of children with
7. Vann WF Jr, Bouwens TJ, Braithwaite AS, Lee JY. The early childhood caries. Pediatr Dent 1999;21(2):109-13.
childhood obesity epidemic: a role for pediatric dentists? 26. Miller J, Vaughan-Williams E, Furlong R, Harrison L.
Pediatr Dent 2005;27(4):271-6. Dental caries and children’s weights. J Epidemiol Com­
8. Ogden CL, Carroll MD, Flegal KM. Epidemiologic trends munity Health 1982;36(l):49-52.
in overweight and obesity. Endocrinol Metab Clin North 27. Chen D, Zhi Q, Zhou Y, Tao Y, Wu L, Lin H. Association
Am 2003;32(4):74l-60. between dental caries and BMI in children: a systematic
9. Hruby A, Hu FB. The epidemiology of obesity: a big review and meta-analysis. Caries Res 2018;52(3):230-45-
picture. Pharmacoeconomics 2015;33(7):673-89. 28. Antunes LA, Machado CM, Couto AC, et al. A polymor­
10. Bagherian A, Sadeghi M. Association between dental phism in the MTRR gene is associated with early child­
caries and age-specific body mass index in preschool chil­ hood caries and underweight. Caries Res 2017;51(2): 102-8.
dren of an Iranian population. Indian J Dent Res 2013; 29. Madsen SS, Wetterstrand VJR, Pedersen ML. Dental
24(l):66-70. caries and weight among children in Nuuk, Greenland, at
11. Bhoomika W, Ramakrishna Y, Munshi AK. Relationship school entry. Int J Circumpolar Health 2017:76(1):
between severe early childhood caries and body mass 1311535.
index. J Clin Pediatr Dent 2013;37(3):235-42. 30. Mitrakul K, Arunakul M, Asvanund Y, Laisirireoungrai T,
12. Davidson K, Schroth RJ, Levi JA, Yaffe AB, Mittermuller Praneechotiros T, Tevavichulada P. Diet, body mass index
BA, Sellers EAC. Higher body mass index associated with and dental caries among Thai children aged 3 to 5 years.
severe early childhood caries. BMC Pediatr 2016; 16:137. Southeast Asian J Trop Med Public Health 2017;48(2):
13. Pikramenou V, Dimitraki D, Zoumpoulakis M, Verykouki 466-72.
E, Kotsanos N. Association between dental caries and 31. Soares ME, Ramos-Jorge ML, de Alencar BM, Oliveira SG,
body mass in preschool children. Eur Arch Paediatr Dent Pereira LJ, Ramos-Jorge J. Influence of masticatory func­
2016; 17 (3): 171-5- tion, dental caries and socioeconomic status on the body
14. Powell JC, Koroluk LD, Phillips CL, Roberts MW. mass index of preschool children. Arch Oral Biol 2017;
Relationship between adjusted body mass index percentile 81:69-73.
and decayed, missing, and filled primary teeth. J Dent 32. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA
Child 2013;80(3):115-20. statement for reporting systematic reviews and meta­
15. Reifsnider E, Mobley C, Beckman-Mendez D. Childhood analyses of studies that evaluate health care interventions:
obesity and early childhood caries in a WIC Population. J explanation and elaboration. PLoS Med 2009;6(7):
Multicult Nurs Health 2004;10:24-31. elOOOlOO.
16. Trikaliotis A, Boka V, Kotsanos N, Karagiannis V, Hassapidou 33. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth
M. Short communication: dmfs and BMI in preschool charts for the United States: methods and development.
Greek children. An epidemiological study. Eur Arch Vital Health Star 11 2002;(246):1-190.
Paediatr Dent 2011; 12(3): 176-8. 34. National Institutes of Health. Study Quality Assessment
17. Costacurta M, DiRenzo L, Sicuro L, Gratteri S, De Lorenzo Tools. Available at: “https://www.nhlbi.nih.gov/health-
A, Docimo R. Dental caries and childhood obesity: anal­ topics/study-quality-assessment-tools”. Accessed March
ysis of food intakes lifestyle. Eur J Paediatr Dent 2014; 14,2019.
15(4):343-8. 35. Ryan R, Hill S. How to GRADE the quality of the evi­
18. Acs G, Lodolini G, Kaminsky S, Cisneros GJ. Effect of dence. Cochrane Consumers and Communication Group,
nursing caries on body weight in a pediatric population. Version 3.0; 2016. Available at: “http://cccrg.cochrane.org/
Pediatr Dent 1992;l4(5):302-5. author-resources” Accessed March 14, 2019.
19. Aluckal E, Anzil K, Baby M, George EK, Lakshmanan S, 36. Chen W, Chen P, Chen SC, Shih WT, Hu HC. Lack of
Chikkanna S. Association between body mass index and association between obesity and dental caries in three-
dental caries among Anganwadi children of Belgaum city, year-old children. Zhonghua Min Guo Xiao Er Ke Yi Xue
India. J Contemp Dent Pract 20l6;17(10):844-8. Hui ZaZhi 1998;39(2): 109-11.
20. Ayhan H, Suskan E, Yildirim S. The effect of nursing or 37. Sheller B, Churchill SS, Williams BJ, Davidson B. Body
rampant caries on height, body weight and head circum­ mass index of children with severe early childhood caries.
ference. J Clin Pediatr Dent 1996;20(3):209-12. Pediatr Dent 2009;31(3):216-21.
21. Gaur S, Nayak R. Underweight in low socioeconomic 38. Yen CE, Hu SW. Association between dental caries and
status preschool children with severe early childhood obesity in preschool children. Eur J Paediatr Dent 2013;
caries. J Indian Soc Pedod Prev Dent 2011;29(4):305-9. 14(3): 185-9.

ECC AND W EIGHT 269

PE D IA T R IC D EN TIS T R Y V 41 / NO 4 JUL I A U G 19

39. Campos JA, Melanda EA, Antunes Jda S, Foschini AL. 48. Khanh LN, Ivey SL, Sokal-Gutierrez K, et al. Early child­
Dental caries and the nutritional status of preschool chil­ hood caries, mouth pain, and nutritional threats in
dren: a spatial analysis. Cien Saude Colet 2011; 16( 10): Vietnam. Am J Public Health 2015;105(12):2510-7.
4161-8. 49. Sideri S, Papageorgiou SN, Eliades T. Registration in the
40. Costa LR, Daher A, Queiroz MG. Early childhood caries international prospective register of systematic reviews
and body mass index in young children from low in­ (PROSPERO) of systematic review protocols was asso­
come families. Int J Environ Res Public Health 2013; 10 ciated with increased review quality. J Clin Epidemiol
(3):867-78. 2018;100:103-10.
41. da Silva RA, Barreiros D, Oliveira S, da Silva LA, Nelson- 50. Hooley M, Skouteris H, Boganin C, Satur J, Kilpatrick N.
Filho P, Kiichler EC. Association between body mass Body mass index and dental caries in children and adoles­
index and caries experience in Brazilian children and cents: a systematic review of literature published 2004 to
adolescents. J Dent Child 2016;83(3): 146-51. 2011. Syst Rev2012;l:57.
42. dos Santos Junior VE, de Sousa RM, Oliveira MC, de 51. Hayden C, Bowler JO, Chambers S, et al. Obesity and
Caldas Junior AF, Rosenblatt A. Early childhood caries dental caries in children: a systematic review and meta­
and its relationship with perinatal, socioeconomic and analysis. Community Dent Oral Epidemiol 2013;41(4):
nutritional risks: a cross-sectional study. BMC Oral 289-308.
Health 2014;14:47. 52. Von Elm E, Altman DG, Egger M, et al. The Strength­
43. Oliveira LB, Sheiham A, Bonecker M. Exploring the ening the Reporting of Observational Studies in Epide­
association of dental caries with social factors and nutri­ miology (STROBE) statement: guidelines for reporting
tional status in Brazilian preschool children. Eur J Oral observational studies. Int J Surg 2014; 12(12): 1495-9.
Sci 2008;ll6(l):37-43. 53. Marshall TA, Eichenberger-Gilmore JM, Broffitt BA,
44. Ismail AI, Sohn W, Lim S, Willem JM. Predictors of Warren JJ, Levy SM. Dental caries and childhood obesity:
dental caries progression in primary teeth. J Dent Res roles of diet and socioeconomic status. Community Dent
2009;88(3):270-5. Oral Epidemiol 2007;35(6):449-58.
45. Macek MD, Mitola DJ. Exploring the association between 54. Sharifi M, Sequist TD, Rifas-Shiman SL, et al. The role
overweight and dental caries among US children. Pediatr of neighborhood characteristics and the built environment
Dent 2006;28(4):375-80. in understanding racial/ethnic disparities in childhood
46. Rodriguez PN, Martinez Reinoso J, Gamba CA, et al. obesity. Prev Med 2016;91:103-9.
Association among salivary flow rate, caries risk and nu­ 55. Matsuo G, Rozier RG, Kranz AM. Dental caries: racial
tritional status in pre-schoolers. Acta Odontol Latinoam and ethnic disparities among North Carolina kindergarten
2015;28(2): 185-91. students. Am J Public Health 2015; 105( 12):2503-9.
47. Floyd B. Associations between height, body mass, and 56. Werner SL, Phillips C, Koroluk LD. Association between
frequency of decayed, extracted, and filled deciduous childhood obesity and dental caries. Pediatr Dent 2012;
teeth among two cohorts of Taiwanese first graders. Am 34(l):23-7.
J Phys Anthropol 2009; 140( 1): 113-9.

270 ECC A N D W E IG H T
Copyright of Pediatric Dentistry is the property of American Society of Dentistry for Children
and its content may not be copied or emailed to multiple sites or posted to a listserv without
the copyright holder's express written permission. However, users may print, download, or
email articles for individual use.