Prenatal Oral Health Care and Early Childhood Caries Prevention: A Systematic Review and Meta-Analysis

Although largely preventable, early childhood caries (ECC) remains the most common chronic childhood disease, with nearly 1.8 billion new cases per year globally [Dye et al., 2007, 2012; GBD 2016 Disease and Injury Incidence and Prevalence Collaborators, 2017]. It afflicts approximately 37% of children aged 2–5 years in the USA [Dye et al., 2007, 2012] and up to 73% of socioeconomically disadvantaged preschool children in both developing and industrialized countries [Dye et al., 2015]. ECC is defined as the presence of ≥1 decayed, missing (due to caries), or filled tooth surface in primary teeth in a child 71 months of age or younger [Colak et al., 2013]. Severe ECC (S-ECC) occurs in children < 3 years of age with ≥1 decayed, missing (due to caries), or filled tooth surfaces and in children 4–6 years of age with elevated caries scores [Colak et al., 2013]. The short-term consequences of untreated ECC include pain, hospitalization, and emergency room visits due to abscess and systemic infection, and even death [American Academy of Pediatric Dentistry Council on Clinical Affairs, 2005; Casamassimo et al., 2009]. Once decay has reached this stage, children often require total oral rehabilitation (TOR) under general anesthesia [Koo and Bowen, 2014] with multiple tooth extractions and restorations/crowns, at a cost of nearly USD 7,000 per child (2009–2011 US data) [Rashewsky et al., 2012]. In the long term, there is strong evidence that children who experienced ECC are much more likely to have a diminished oral health-related quality of life and higher risk of caries lesions in permanent teeth [Powell, 1998; Heller et al., 2000].

Despite the advancement of ECC prediction and treatment strategies, ECC remains a public health burden. In the USA, more than USD 1.5 billion per year is spent on treatment. However, children remain at high risk for recurrent caries even after extensive TOR treatment. Up to 40% of children treated for S-ECC experience recurrent disease by the 6-month checkup post-TOR [Graves et al., 2004; Berkowitz et al., 2011], despite pharmacologic interventions, such as topical fluoride/antimicrobial applications and dietary counseling to alter caries-promoting eating behaviors [O’Sullivan and Tinanoff, 1996; Li and Tanner, 2015]. Hence, more effective preventive strategies are critically needed.

Pregnancy is an ideal time to promote primary prevention of ECC in children given the profound influence of maternal health and behaviors on children’s oral health outcomes [Iida, 2017]. ECC is a multifactorial bacterial disease with Streptococcus mutans as the prime cariogenic bacterium, and strongly influenced by diet [Caufield et al., 1993; Klein et al., 2004; Li et al., 2005; Kanasi et al., 2010; Slayton, 2011; Zhan et al., 2012; Klinke et al., 2014]. Studies have shown that maternal untreated caries and a greater level of salivary S. mutans increase the risk of ECC in children. Children’s dietary and oral hygiene behaviors rely on their parents’ or caregivers’ oral health knowledge, beliefs, and behaviors [Finlayson et al., 2007; Wigen et al., 2011]. By revisiting the children’s dental caries risk model described by Fisher-Owens et al. [2007] that included different levels of environmental elements, several factors that could potentially be influenced by mothers (marked with asterisks in Fig. 1) can be identified, including: (1) microflora, diet, and host in the oral health element positioning at the oral health circle; (2) health behaviors and practices, genetic endowment, demographic attributes, dental care utilization, oral health behaviors and practices, and dental insurance, that are included in the child-level influences element; (3) family position, socioeconomic status, physical safety, health status of parents, family function, family education, health behaviors, practices, and coping skills of the family, which lie in the family-level influences element. These factors in the aforementioned dental caries risk prediction model further emphasize the maternal role in ECC development. Thus, in theory, oral health care intervention during pregnancy presents an ideal entry point to preventing ECC.

Previously, studies have shown a positive ECC prevention outcome by providing prenatal oral health education or intervention [Günay et al., 1998; Nakai et al., 2010]; however, another study failed to show more effective ECC prevention when intervention during pregnancy was compared to the control group. Therefore, the aim of this study is to systematically review the scientific evidence relating to the association between prenatal oral health care, reduced carriage of S. mutans, and ECC prevention.

Database searches were conducted in May 2018 to identify published studies on prenatal oral health care and ECC-related outcome (onset of ECC and/or oral S. mutans colonization). A medical reference librarian developed individual search strategies and retrieved citations from PubMed, Embase, Scopus, Web of Science, LILACS, Cochrane Library, and ClinicalTrials.gov. A combination of text words and controlled vocabulary terms were used (Prenatal Care, Oral Health, Child, Infant, Breast Feeding, Newborn, Dental Caries). A detailed search strategy is shown in the online supplementary Appendix 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000495187).

This systematic review included case-control studies, retrospective or prospective cohort studies, randomized or nonrandomized controlled trials that examined the effect of oral health care during pregnancy on the incidence of ECC and/or oral carriage of S. mutans in children under the age of 6 years. Two trained independent reviewers completed the article selection in accordance with the inclusion/exclusion criteria. The agreement between reviewers was satisfactory (K = 0.81). Disagreements were resolved by consensus between the 2 reviewers.

The following inclusion and exclusion criteria were used for literature selection.

Types of participants: pregnant women and their children under the age of 6 years. Types of intervention(s)/phenomena of interest: prenatal oral health care utilization/intervention. Types of comparisons: pregnant women who received and did not receive prenatal oral health care. Types of outcomes: reduced dental caries in children; reduced oral carriage of S. mutans. Types of studies: case-control studies; retrospective or prospective cohort studies; randomized and nonrandomized controlled trials. Types of statistical data: odds ratios (OR); relative risk; confidence intervals (CI); p values, and frequency of an absolute number of events versus total number of individuals per group.

The exclusion criteria were: in vitro studies; animal studies; papers with abstract only; literature reviews; letters to the editor; editorials; patient handouts; case report or case series, and cross-sectional studies.

Descriptive data, including clinical and methodological factors such as country of origin, study design, study site, dental examination, dental examiner calibration, age of subjects, type of prenatal oral health care intervention, outcome measures (ECC and/or oral S. mutans), as well as results from statistical analyses were obtained using an extraction form (online suppl. Appendix 2).

The quality of the selected articles was assessed using two methodological validities. (1) Cochrane Collaboration’s tool for assessing risk of bias in randomized trials [Higgins et al., 2011]. Articles were scaled for the following bias categories: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other bias. (2) Adapted Downs and Black scoring [Downs and Black, 1998] that assesses the methodological quality of both randomized and nonrandomized studies of health care interventions. A total score of 26 represents the highest study quality.

For the articles selected for quantitative analysis, the R package Metafor was used for meta-analysis (https://cran.r-project.org/web/packages/metafor/). The OR and 95% CI and p values were estimated using an unconditional generalized linear mixed effects model with random study effects. Children’s age at study endpoint was used as a covariate. Heterogeneity among the studies was evaluated using I² statistics and tested using the likelihood ratio test. A forest plot was created to summarize the meta-analysis study results.

The literature analyses identified a total of 4,956 papers from the database search (Fig. 2). A total of 787 duplicate references were removed. The remaining 4,026 studies were imported into an Endnote Library for further review. From those, 3,854 studies were excluded after title screening and 128 studies were excluded after abstract screening. The remaining 44 articles were selected for a full text review. After the full text analysis, 40 studies were eliminated based on the exclusion criteria and 5 articles were chosen for qualitative assessment. For the quantitative assessment using meta-analysis to assess the effect of prenatal oral health care intervention on the onset of ECC, 4 out of 5 articles that received qualitative assessment were included. One article that was removed from the meta-analysis only included oral S. mutans carriage in children, but not ECC as the outcome [Nakai et al., 2010]. The full list of excluded articles after the full text review is shown in Appendix 3.

The characteristics of studies included in the qualitative review are summarized in Table 1. All 5 studies were published between 1997 and 2016. One was conducted in the USA [Leverett et al., 1997], 1 in Germany [Günay et al., 1998], and 1 in Australia [Plutzer and Spencer, 2008]. Two were conducted in Japan [Nakai et al., 2010, 2016]. Among the 5 studies, 3 were randomized control trials [Leverett et al., 1997; Plutzer and Spencer, 2008; Nakai et al., 2010], 1 was a prospective cohort study [Günay et al., 1998], and 1 was a nested case-control in a cohort study [Nakai et al., 2016]. Oral health care intervention adopted in all qualitative studies extended the intervention period from the prenatal to infant stage. The interventions included: (a) fluoride-based intervention, where fluoride supplement intake was provided to pregnant women and their infant in a population that was not exposed to optimal water fluoridation [Leverett et al., 1997]; (b) primary-primary prevention originally proposed by Axelsson [1988], where all prophylactic measures were carried out in pregnant women in order to prevent the transmission of cariogenic bacteria and improve feeding behaviors after birth [Günay et al., 1998]; (c) oral health education promotion in pregnant women, which was used in the studies by Plutzer and Spencer [2008] and Nakai et al. [2016], who called it antenatal health care; (d) xylitol gum chewing in pregnant women [Nakai et al., 2010]. The intervention approaches are further detailed in Table 1.

Study outcomes were assessed when children reached 2–5 years of age. The onset of ECC and salivary S. mutans carriage are the two primary outcomes evaluated in these 5 studies. Quality and risk of bias for all 5 studies was assessed and are shown in Figure 3. Two studies with a randomized controlled trial design were of high quality based on the Cochrane risk of bias assessment tool [Higgins et al., 2011] and Downs and Black scoring system [Downs and Black, 1998]; the other 3 studies showed moderate quality.

Three studies [Günay et al., 1998; Plutzer and Spencer, 2008; Nakai et al., 2016] revealed a lower ECC incidence in the group that received oral health care intervention during pregnancy and early infancy when compared to the control group. The prenatal oral health care intervention approaches used in these 3 studies were primary-primary prevention, oral examination and cleaning, and oral health education. One study [Leverett et al., 1997] investigating fluoride supplement use during pregnancy showed no statistical difference (p > 0.05) in caries incidence in children between the intervention (8%) and control group (9%).

A meta-analysis was performed on 4 studies that assessed ECC incidence (results shown in Fig. 4). In particular, Günay et al. [1998] examined the same cohort of children at two time points, when they reached 3 and 4 years of age; their results were included as two data sets in the meta-analysis. Study heterogeneity (I² = 75.06%) and the related p value were calculated using the likelihood ratio test (p < 0.0001).

The empirical ORs and 95% CIs of the studies included in the meta-analysis are shown in Figure 4a. When compared to the control group, the empirical OR (95% CI) of ECC in children whose mothers received primary-primary prevention is 0.04 (0.00, 0.68) at 3 years of age and 0.13 (0.04, 0.42) at 4 years of age [Günay et al., 1998]. Compared to the control group, the empirical OR (95% CI) of ECC is 0.17 (0.06, 0.49) in children whose mothers received oral health education [Plutzer and Spencer, 2008], 0.36 (0.15, 0.85) in children whose mothers received antenatal health care [Nakai et al., 2016], and 0.94 (0.57, 1.56) in children whose mothers received a fluoride supplement [Leverett et al., 1997].

Based on the generalized linear mixed effects model with covariate age, the estimates of ORs and 95% CIs indicate that, regarding ECC incidence, there is a statistically significant difference between the intervention and control groups for children younger than 4 years old, regardless of intervention modalities (detailed in Fig. 4b). The odds of experiencing ECC among the children younger than 4 years whose mothers received prenatal oral health care is significantly less than those children in the control group, indicating a protective effect of prenatal oral health care against ECC development with 95% CIs whose upper bounds are smaller than 1. For instance, the estimated ORs (95% CI) are 0.12 (0.02, 0.77) for children at 1 year of age, 0.18 (0.05, 0.63) for children of 2 years of age, 0.25 (0.09, 0.64) at 3 years of age, and 0.35 (0.12, 1.00) at 4 years of age. For children 5 years of age or older, the estimated OR is still smaller than 1, but the 95% CI contains 1, indicating that the protective effect becomes insignificant.

The effect of prenatal oral health care intervention on the reduction of children’s S. mutans carriage was assessed in 2 studies [Günay et al., 1998; Nakai et al., 2010]. In the study by Günay et al. [1998], S. mutans reduction was significant between the intervention and control groups: 100% of children in the intervention group remained S. mutans free by the age of 3 years, whereas only 38.5% of children in the control group remained S. mutans free by the age of 3 years. Moreover, mothers in the intervention group also showed a significant improvement in plaque index and reduction in S. mutans score. The study by Nakai et al. [2010] showed that significantly more children in the xylitol chewing group remained S. mutans free at 9, 12, and 24 months. Furthermore, pre- and perinatal xylitol chewing by mothers delayed S. mutans carriage in children. The children’s S. mutans acquisition age in the xylitol chewing group was 8.8 months later than that of the control group (mean age 20.8 vs. 12.0 months).

The results of this review have shown a reduced ECC incidence in children whose mothers received prenatal oral health care. ECC is a multifactorial disease with complex socioeconomic, genetic, oral hygiene behaviors, and bacterial and diet factors that affect its risk [Ruby and Goldner, 2007; Wang et al., 2012]. S. mutans and, more recently, Candida species have been implicated as potential major etiological microorganisms that may be involved in the initiation and development of ECC [Tanzer et al., 2001; Gross et al., 2012; Xiao et al., 2018]. Studies have shown an association between maternal poor oral health and increased risk for ECC [Chaffee et al., 2014]. The association between mother’s and child’s oral health could possibly be explained by: (1) the mothers’ oral health behavior, e.g., perception and knowledge influences the dental health of her children [Saied-Moallemi et al., 2008; Goettems et al., 2012; Olak et al., 2018]; (2) the mother might be a main source of her children’s acquisition of oral S. mutans and Candida sp. [Waggoner-Fountain et al., 1996; Caufield et al., 2005; Bliss et al., 2008; Xiao et al., 2016; Childers et al., 2017].

The following points should be considered when interpreting the results of this review. (1) Various intervention modalities and frequencies were used across the 5 studies, which produced challenges for data analysis, e.g., the heterogeneity of studies included in the meta-analysis is significant (p < 0.01). (2) The timing of the main outcome measurement (ECC incidence) with respect to children’s age lacks consistency throughout the 5 studies. The peak of ECC onset is 3 years of age, and there is a significant increase in incidence between the age of 2 and 3 years. Kopycka-Kedzierawski et al. [2008] reported a 26% ECC prevalence among 2-year-old children in Rochester, NY, USA; Quiñonez et al. [2001] reported a 20% ECC prevalence in children aged 18–36 months in North Carolina, USA; Rosenblatt and Zarzar [2002] reported a 46% S-ECC prevalence rate among Brazilian children aged 25–36 months. Two studies included in the quantitative analysis only monitored study children until the age of 2 years, which might have underestimated the preventive effect of prenatal oral health care on ECC. (3) As we were not able to collect study subjects’ data on other caries determinants, e.g., demographic, socioeconomic, sugar consumption, etc., the meta-analysis performed in this review did not use multivariate analyses to consider the potential confounders mentioned above. Given the multifactorial nature of ECC, the ORs calculated might have under- or overestimated the effectiveness of prenatal oral health care. (4) For the strategies that used prenatal oral health education or primary-primary prevention, it was not clear to what degree the prenatal oral intervention had improved or restored pregnant women’s oral health. Therefore, it is challenging to make recommendations on how much oral health care a pregnant woman needs to receive and how much oral health education is needed to demonstrate effective ECC prevention in children. Taking the aforementioned limitations into account, future randomized clinical trials are desired to test prenatal oral health care strategies that maintain or restore an expectant mother’s oral health and that measure improvements in oral health knowledge.

Moreover, another dilemma that needs to be considered is that, although routine oral care during pregnancy has been demonstrated to be safe, and recommendations for prenatal oral care have been disseminated globally, utilization of prenatal oral health care is limited in both developed and developing countries [Rocha et al., 2018]. In contrast to the limited utilization of prenatal dental care, over 76% of US women admitted to suffering from oral health problems (pain, bleeding gums, and oral infection) during pregnancy, while more than 43% did not have a dental checkup during pregnancy [DentistryIQ Editors, 2015]. Furthermore, dental care utilization during pregnancy was lower among black women [Thompson et al., 2013], ethnic minorities [Marchi et al., 2010], and women with socioeconomic disadvantages [Singhal et al., 2014]. Thus, oral health represents an important often-neglected heath disparity during pregnancy among minority women and women who are socioeconomical ly disadvantaged [Guarnizo-Herreño and Wehby, 2012; Azofeifa et al., 2014]. In order to successfully use prenatal oral health care to prevent ECC, future efforts need to gain a better understanding of the factors that enable or inhibit the use of prenatal dental care at both the community and individual levels. Effective strategies might derive from collaborations among dental and medical providers involved in women’s and children’s dental and medical health, policy makers, and community social workers.

This review reports a reduced ECC incidence and S. mutans carriage in children whose mothers received prenatal oral health care. Maintaining oral health and improving oral health care knowledge during pregnancy is a critical and promising step towards ECC prevention. Future studies should consider testing strategies that maintain an expectant mother’s oral health or restore an expectant mother’s oral health to a disease-free state during pregnancy.

This study was supported in part by Jin Xiao’s faculty start-up funds from the Eastman Institute for Oral Health, University of Rochester, and the National Institute for Dental and Craniofacial Research/National Center for Advancing Translational Sciences grant KL2 TR001999 and K23 DE027412.

The funding agencies had no role in the study design, data collection, analyses, decision to publish, or preparation of the manuscript.

J.X. contributed to the study design; J.X., N.A., D.A.C., T.T.W. performed the data acquisition and analysis; J.X., D.A.C., D.T.K., R.J.B., L.R., H.M., E.E., and Y.R. contributed to the data interpretation, manuscript writing, and critical revision of the manuscript.