Study population and study design
This study was conducted among preschool children below 6 years of age in Lagos, Nigeria, from March 2018 to October 2019. Details of study design has been reported in previous study [22]. Briefly, the study was conducted in five Local Government Areas (LGAs) of Lagos State (Mushin, Surulere, Lagos Mainland, Ikorodu and Epe). Lagos State is in the South-Western geopolitical zone of Nigeria with an estimated population of 17 million and it is the economic centre of the country. It shares boundaries with Ogun State both in the North and East and on the West by Benin Republic. In the South, it stretches for 180 km along the coast of the Atlantic Ocean and occupies an area of 3577 km2, 22% or 787 km2 which consists of lagoons and creeks. It consists of 20 LGAs.
Ethics consideration
Ethical approval for the study was obtained from the Health Research Ethics Committee of the Lagos University Teaching Hospital (LUTH), Lagos, Nigeria (ADM/DCST/HREC/APP/1669). The study was conducted according to the guidelines laid down in the Declaration of Helsinki for all procedures involving human subjects/patients. Written informed consent was obtained from the parents or legal guardian of the study participants.
Inclusion and exclusion criteria
Children younger than 6 years of age whose parents agreed to their children partaking in the research and who provided informed consent were included in the study. Children with physical or mental incapacity, those with disorders that made routine oral hygiene measures difficult, and hence are more predisposed to dental caries, those with any developmental dental anomaly and those with ongoing dental treatment were excluded.
Sampling procedure
A multistage probability sampling technique was adopted for this study [22]. At the first stage, five LGAs (Mushin, Surulere, Lagos Mainland, Ikorodu and Epe) were selected out of the 20 LGAs in Lagos state, using simple random sampling. In the second stage, one public and one private school was selected by the simple random sampling method from each LGA. The lists of government registered public and private nursery and primary schools [23, 24] was obtained from the Lagos State Universal Basic Education Board (SUBEB) and it served as the sampling frame. Systematic random sampling method was used to select infants, toddlers and children in the nursery and primary sections of the schools, who were eligible by set criteria for the study, with the nominal roll of pupils in the selected schools serving as the sampling frame. A consent letter was sent to the parents of the children in the schools explaining the aim, scope, and importance of the study and asking for parental approval. A proforma with the study questionnaire was attached to the consent form. The consenting parents completed the questionnaire and signed the consent forms for their children to be examined and to have salivary samples taken. Children without signed parental consent forms or whose parents expressed objection to participate were excluded from the study.
Questionnaire
The interviewer-administered questionnaire was adapted from previously validated questionnaires [5,6,7]. It consisted of five main parts. The first part included demographic information (child’s age, gender, tribe, mother’s education, father’s occupation). The second part assessed behavioural characteristics such as child’s feeding patterns while the third part assessed the participants utilization of dental services. The fourth part included child’s oral hygiene practices and the fifth part of the questionnaire assessed nature of child’s diet and snacking habits (Additional file 1).
Clinical examination
Detailed enrolment procedures, inclusion and exclusion criteria are reported in a previous study [7]. In brief, dental caries diagnosis was done using the World Health Organization (WHO) diagnostic criteria. The participants were examined sitting, under natural light, using sterile dental mirrors and probes by trained dentists who were calibrated. Caries was assessed by means of visual/tactile examination without radiographs. Intra examiner and extra examiner were assessed by examining 10 children with dental caries on two separate occasions with two weeks interval. The result was coded and entered into the computer. The data were then subjected to a Cohen’s kappa scores analysis, to determine the intra-examiner and inter-examiner variability. The intra-examiner variability scores ranged from 0.80 to 0.85 while the inter-examiner variability scores ranged from 0.81 to 0.84 for the dmft indices. Children under 6 years of age, having one or more decayed (non-cavitated or cavitated), missing (due to caries), or filled tooth surfaces (dmfs) in any primary tooth were considered to have ECC. A child younger than three years of age, with one or more cavitated, missing (due to caries), or filled smooth surfaces in primary maxillary anterior teeth and from ages three through five, a decayed, missing, or filled score of greater than or equal to four (age 3), greater than or equal to five (age 4), or greater than or equal to six (age 5) was considered to have S‑ECC. Dental caries was recorded using decay, missing, filled/teeth (dmft) and surface (dmfs) index for each participating child. Children with caries were allocated into cases while children without caries were allocated into the control group. After the screening procedure, children with untreated caries were referred to the Department of Child Dental Health, LUTH for treatment and follow-up.
Genetic analysis
Saliva samples were collected from 338 children with early childhood caries and 353 children without caries. Saliva samples were collected using Oragene-DNA Collection Kits (http://www.dnagenotek.com), labeled and shipped to the laboratory at the University of Iowa for extraction, processing and analyses. DNA was extracted from samples using Qiagen DNA Extraction Kits and sample concentration was quantified using Qubit Assays and the Qubit 2.0 Fluorometer (http://www.invitrogen.com/site/us/en/home/brands/Product-Brand/Qubit.html). XY genotyping was also used as a quality control measure to confirm that sample sex matched clinical records from collection. SNP assays for the 14 SNPs (included in a 24 assay panel required for fluidigm) of interest were designed based on human genome assembly GRCh38/hg38, 2013 and purchased through ABi/Life Technologies (www.lifetechnologies.com).
All DNA samples were diluted to an identical concentration of 1 ng/ul, and underwent preamplification processes as per Fluidigm manufacturer recommendations due to the nanoscale volume of Fluidigm Integrated Fluidic Circuits (IFCs). Following preamplification, the samples were diluted to decrease reaction component concentration, to prevent reaction inhibition which may affect quality of results in subsequent steps. Additional reagents are then added in preparation for Fluidigm. In this study, we used the 192.24 Dynamic Array IFC Fluidigm chip, which allowed for 4,608 simultaneous reactions using 192 samples and 24 assays (14 of these assays are relevant and were used for the current study). Samples were randomly assigned in labeled 96-well plate maps. Each Fluidigm chip had at least two “No Template Controls” (NTCs), which consisted of TE Buffer only. NTCs were added to ensure the presence of negative control conditions for comparison, and to establish a point of origin in scatter plot analyses later in the study.
Prior to loading samples, the chip was primed with control line fluid, helping to maintain pressure and facilitate the closing of chip valves. Samples were then pipetted into the sample inlets and assays were placed into detector inlets. Once all samples had been introduced to the chip and all air bubbles had been eliminated, the chip was placed onto the IFC Controller RX machine for loading and mixing. The machine then applied pressure, pushing the sample and assay fluid into their respective fluid lines. Using a carry-over slug design, sample mixing was carried out as partial volumes of the first assay and sample were pushed into and combined in isolated reaction chambers. The load-mixing process occurred on the first machine over the span of 45 min, after which the chip, with all samples amply mixed with their corresponding assays, underwent PCR on the FC1 Cycler in order to induce isolated reactions.
Once both machine cycles had been completed, a fluorescent image was taken of the endpoint state of the genotyping chip using the BioMark System for Genetic Analysis. The fluorescent signals were obtained through FAM and VIC channels, with FAM having an excitation peak of 495 nm and an emission peak 520 nm, and VIC having an excitation peak of 538 nm and an emission peak of 554 nm. The final image with fluorescent signals in all 4,608 reaction chambers was then plugged into Fluidigm SNP Genotyping Analysis software to generate genotype calls. Fluorescent signal brightness levels in both FAM and VIC channels were calculated and plotted on a scatter plot in the software based on relative intensity: Wang et al. [25]. With FAM on the X-axis and VIC on Y-axis, the fluorescent signals fell into four different genotype groups, homozygous FAM, homozygous VIC, heterozygous brightness between the two, and no signal negative controls.
Data analysis
The data was entered and analyzed using IBM Statistical Package for the Social Sciences (SPSS) Version 22.0 (IBM Corp, Armonk, NY). Bivariate comparisons (t Test, ANOVA, and Chi-square tests) and Multivariate logistic regression analyses were conducted to assess associations between ECC and predictor variables. Case–control analyses of the genotype data was done using the additive model in PLINK. Nominal significance was set at p < 0.05 and Bonferroni corrected p value was p < 0.004 i.e. 0.05 divided by 14 (number of SNPs and tests, where every SNP is a test and 0.05 is significant for just one SNP/ test).