Data Sources, Study Population, and Sample Size

NHANES is a cross-sectional survey designed to monitor the health and nutritional status of the resident, civilian, noninstitutionalized US population.22 It is administered by CDC’s National Center for Health Statistics. Survey participants are selected by using a complex, multistage probability design, which includes oversampling of selected population subgroups to improve reliability of the estimates. NHANES collected the data during standardized oral health examinations of participants in mobile examination centers and by interviews in participants’ homes. Additional details of the surveys are available on the NHANES website.

This report used NHANES data for two 6-year periods: 1999– 2004 (combining 2-year cycles 1999–2000, 2001–2002, and 2003–2004) and 2011–2016 (combining 2-year cycles 2011–2012, 2013–2014, and 2015–2016). Although NHANES data are released in 2-year cycles, this report used 6 years of data to be consistent with data reporting in the last oral health surveillance report, which used 6 years of data to compare changes in oral health status between NHANES 1988–1994 and 1999–2004.10 In addition, because we reported findings for individual age groups stratified by sociodemographic characteristics, we combined cycles as recommended by NHANES guidelines to improve the statistical reliability of our estimates.23

Our study population included participants aged 2 years or older who completed an oral health assessment. During 2011–2016, among 27,925 participants interviewed, 26,799 were selected for an oral health examination (96%), and of these, 25,566 (95.4%) completed the examination. Hispanics, non-Hispanic blacks, and Asians were oversampled. During 1999–2004, among 28,661 participants interviewed, 27,032 were selected for an oral health examination (94.3%), and of these, 25,364 (93.8%) completed the examination. Adolescents aged 12–19 years, adults 60 years or older, non-Hispanic blacks, and Mexican Americans were oversampled.22

Oral Health Assessment

Oral health assessments in both survey periods used the same protocols, allowing comparisons of estimates for outcomes, including dental caries, dental sealants, tooth retention, and edentulism. Visual oral health assessments were conducted by previously trained dentists in a mobile examination center, and these dentists were monitored during NHANES data collection to ensure consistent assessment standards. The examinations were used to detect the presence or absence of permanent or primary teeth. In this report, all dental measures of permanent teeth are based on 28 teeth, excluding 3rd molars. Each tooth was assessed for caries and restorations. Caries assessment was based on Radike’s criteria with minor modifications.24,25 Specific surfaces of primary and permanent molars, premolars, and upper permanent lateral incisors were assessed for dental sealants. A sealant was considered present if any portion of the surface was covered by sealant. Further descriptions of the clinical procedures for assessing dental caries and dental sealants are described elsewhere.6,10,17,24

Case definitions for dental caries were based on those proposed by Klein and colleagues.26 This report examines coronal caries, which is decay located on part of the crown of the tooth above the gum line. Assessment of root caries was not included because data were unavailable during 2011–2014 cycles. Prevalence of dental caries was defined as the percentage of the population having at least one tooth with untreated decay (d/D) or a restoration or filling (f/F). (Note: In the abbreviations used for categories of teeth affected by caries in this report, lowercase letters denote primary teeth, while uppercase letters denote permanent teeth.) Prevalence of untreated tooth decay was defined as the percentage of the population with at least one tooth with untreated, cavitated carious lesions. Among children and adolescents aged 6–19 years, prevalence of caries in permanent teeth also included permanent teeth missing (M) from disease (i.e., caries).

In this report, caries severity is measured as the mean number of teeth affected by caries among children, adolescents, or adults younger than 65 years with caries. Caries severity is reported for all older adults aged 65 years or older because more than 95% of older adults have caries.7 Previous reports have measured caries severity as the mean number of affected teeth or surfaces for all children or adults.6,10 In this report, caries severity also is described by permanent tooth type (i.e., anterior teeth [incisors and canines], premolars, and first and second molars) for children and adolescents aged 6–19 years.

For sealants, overall prevalence of having one or more sealants on permanent teeth (including permanent molars, premolars, or upper lateral incisors) was reported for children and adolescents. The mean number of sealants—overall and on molars and premolars specifically—was reported only for those with one or more sealants.

Age categories for the study population were 2–5 and 6–8 years for primary teeth in children; 6–11 and 12–19 years for permanent teeth in children and adolescents; 20–34, 35–49, and 50–64 years for adults; and 65–74 years and 75 years or older for older adults. We adjusted all age-specific estimates to the US 2000 standard population to control for potential differences in age distribution.27 Data for children and adolescents were adjusted by using single years of age. Data for adults were adjusted by using 5-year age groups, with the maximum age group set to age 80 years or older.

We used the same selected sociodemographic characteristics used in previous surveillance summary reports.6,10 Educational attainment was classified as less than high school, high school graduate or equivalent, and more than high school. Poverty status was reported as living in a household with income below 100% of the federal poverty level, 100% or higher but below 200% of the federal poverty level, or 200% or higher of the federal poverty level and categorized as poor, near-poor, or not-poor. Findings are also presented for poor and near- poor children combined (less than 200% of the federal poverty level) to more closely align with federal thresholds used to determine eligibility for public health insurance and free and reduced price meal programs within schools.28,29 Similar income categorizations, such as 185% of the federal poverty level or less, have been used elsewhere to define low- and higher- income children for analyses of untreated tooth decay and sealant use among school-aged children in the United States.16

Information on self-reported race or ethnicity was stratified into three categories to be comparable between 1999–2004 and 2011–2016: non-Hispanic white, non-Hispanic black, and Mexican American. Although NHANES 2011–2016 oversampled the entire Hispanic population, NHANES 1999–2004 only oversampled Mexican Americans.22 Cigarette smoking status was classified into three categories: current smoker, former smoker, and never smoked. Current smokers were defined as respondents who reported having smoked at least 100 cigarettes during their lifetime and currently smoking cigarettes every day or some days. Former smokers were defined as respondents who reported having smoked at least 100 cigarettes during their lifetime but not currently smoking. Never smokers were defined as respondents who reported having smoked fewer than 100 cigarettes during their lifetime.

When comparing outcomes among different categories within the same sociodemographic characteristic, we used reference categories used in previous surveillance reports. The reference category for educational attainment was more than high school; for income, 200% of the federal poverty level or higher; for race or ethnicity, non-Hispanic white; and for smoking status, never smoked.

All statistical analyses were performed with SUDAAN software Version 11.0.0 (RTI International, Research Triangle Park, North Carolina) and SAS software Version 9.3 (SAS Institute Inc., Cary, North Carolina). Age-adjusted oral health outcomes, standard errors, and differences in oral health outcomes between 1999–2004 and 2011–2016 are presented overall and stratified by sociodemographic characteristics. Because of rounding, the differences between the two periods presented may be slightly different from what would have been obtained if we had subtracted the 2011–2016 estimates from the 1999–2004 estimates.

All reported differences are statistically significant at P < 0.05 (2-sided t-test) unless otherwise noted. When a range of differences for a specific outcome was reported in the current analysis, we only presented differences that were significantly different at P < 0.05. For the 2011–2016 data, t-tests were also performed to detect significant differences (P < 0.05) in oral health outcomes between categories within each sociodemographic characteristic. Estimates with relative standard error higher than 30% were considered unreliable and not displayed in the tables.30