| Sign In to gain access to subscriptions and/or personal tools. |
Prevalence and Trends in Periodontitis in the USA: from the NHANES III to the NHANES, 1988 to 2000
Department of Epidemiology, Mailman School of Public Health, Columbia University, 722 West 168th Street, 16th Floor, Room 1611, New York, NY 10032, USA; Correspondence: * corresponding author, lnb2{at}columbia.edu
Trends in periodontal diseases in the USA have been documented for years. However, the results have been mixed, mostly due to different periodontal assessment protocols. This study examined change in the prevalence of periodontitis between the NHANES III and the NHANES 1999–2000, and differences in the prevalence of periodontitis among racial/ethnic groups in the USA. Analysis was limited to non-Hispanic black, non-Hispanic white, and Mexican-American adults aged 18+ yrs in the NHANES III (n = 12,088) or the NHANES 1999–2000 (n = 3214). The prevalences of periodontitis for the NHANES III and the NHANES 1999–2000 were 7.3% and 4.2%, respectively. In multivariable analyses, blacks were 1.88 times (95%CI: 1.42, 2.50) more likely to have periodontitis than whites surveyed in the NHANES III. However, the odds of periodontitis for blacks and Mexican-Americans did not differ from those for whites surveyed in the NHANES 1999–2000. Our findings indicate that the prevalence of periodontitis has decreased between the NHANES III and the NHANES 1999–2000 for all racial/ethnic groups in the USA.
Key Words: race/ethnicity • periodontitis • prevalence • national surveys
Previous studies have documented racial/ethnic disparities in the prevalence of periodontitis in the USA (Nikias et al., 1977; Kelly and Harvey, 1979; Douglass et al., 1983; Capilouto and Douglass, 1988; Beck et al., 1990; Löe and Brown, 1991; Oliver et al., 1991, 1998; Borrell et al., 2002a). However, these studies differed in the case-definitions used, thus limiting direct comparisons. Furthermore, the earlier studies focused on non-Hispanic blacks and whites only. Periodontal data for Hispanics, the fastest-growing segment of the US population (US Department of Commerce, 2001), have been sparse. Prior to the Third National Health and Nutrition Examination Survey (NHANES III), the only data available for Hispanics came from the Hispanic Health and Nutrition Examination Survey of 1982–1984 (Ismail et al., 1987; Ismail and Szpunar, 1990). Although periodontal data for the Hispanic population surveyed in the NHANES III included only Mexican-Americans, analysis of these data showed that Mexican-Americans exhibited a prevalence of periodontitis similar to that of non-Hispanic whites and intermediate between that of African-Americans and that of non-Hispanic whites. This finding has been consistent regardless of the case-definition used (Albandar et al., 1999; Arbes et al., 2001; Borrell et al., 2002b; Hyman and Reid, 2003). The NHANES 1999–2000, the next survey in the NHANES series, also included Mexican-Americans. Because the NHANES III and the NHANES 1999–2000 used the same periodontal examination protocol, the availability of data from the NHANES 1999–2000 affords us the opportunity to investigate trends in the prevalence of periodontitis between surveys. Specifically, this study examines whether the prevalence of periodontitis has changed between the NHANES III and the NHANES 1999–2000, and whether the differences in the prevalence of periodontitis among racial/ethnic groups have increased, decreased, or remained the same between the two surveys.
Data for this study came from the NHANES III and the NHANES 1999–2000 public-use data files. The NHANES III was conducted from 1988 to 1994. The NHANES 1999–2000 collected data between March, 1999, and December, 2000. Both surveys, using the same method, assessed the health status of a nationally representative sample of the civilian non-institutionalized US population, selected through a stratified multistage probability sampling design. Full descriptions of the sample design in the NHANES III and the NHANES 1999–2000 have been reported elsewhere (National Center for Health Statistics, 1996; National Health and Nutrition Examination Survey, 2001a,b). The NHANES III yielded a sample of 33,994 persons 2 mos of age or older, of whom 31,311 were examined, while the NHANES 1999–2000 examined a total of 9956 persons 1 mo of age or older. This analysis was limited to the records of adults 18 yrs or older who self-identified as non-Hispanic black, non-Hispanic white, or Mexican-American, and who had a complete periodontal examination during the NHANES III (n = 12,088) or the NHANES 1999–2000 (n = 3214). Thus, the combined sample was 15,302.
During the dental examination, dentists trained in the survey examination protocol conducted the periodontal examinations (National Center for Health Statistics, 1996; National Health and Nutrition Examination Survey, 2001a). Briefly, the periodontal examination was conducted at 2 sites, mid-buccal and mesiobuccal, for each tooth, in 2 randomly chosen quadrants, 1 maxillary and 1 mandibular, on the assumption that conditions in these 2 quadrants would represent the mouth. Third molars were excluded because of their frequent extraction in young adulthood, so a maximum 14 teeth and 28 sites per individual were examined. Previous studies used several combinations of clinical attachment loss (CAL) and pocket depth (PD) to establish periodontitis case-definitions (Beck et al., 1990; Machtei et al., 1992; Locker and Leake, 1993). For this study, the distribution of CAL and PD was evaluated in the total population as well as in each racial/ethnic group in each survey. Prior to any hypothesis testing, we tested several case-definitions before arriving at the one used in this analysis. A periodontitis case was defined as a person who had at least 3 sites with CAL The main independent variable of interest was race/ethnicity, defined in both surveys as non-Hispanic black, non-Hispanic white, and Mexican-American. For this paper, non-Hispanic black and non-Hispanic white will be referred to as black and white, respectively.
For estimation of the association between race/ethnicity and periodontitis adjusted for other factors, the following variables were included in the analysis: age at interview, gender, marital status, place of birth, education, income, presence of health insurance, history of diabetes, and tobacco use. The NHANES III and the NHANES 1999–2000 used the same or similar questions to record these covariates. Age, gender, and presence of health insurance were included in the analyses as collected in both surveys. Among participants with health insurance, a follow-up question (yes/no) was asked regarding dental coverage. Categories in the marital status question originally included married, living together with someone as married, widowed, divorced, separated, or never married. These categories were grouped into married (married or living together with someone as married), single, divorced (separated or divorced), and widowed. Place of birth was recorded as having been born in the USA or elsewhere. Education was recorded as a continuous variable, from 0 to 17 yrs of education in the NHANES III and as < 12 yrs, 12 yrs, and > 12 yrs of education in the NHANES 1999–2000. The latter categories were used in both surveys. Total family 12-month income during the preceding year was also recorded as a continuous variable and was recoded after incomes reported in the NHANES 1999–2000 were adjusted to the equivalent dollar amount in the NHANES III (US Department of Labor, 2004). Income was categorized as low, medium, or high. These categories were: The question "Have you ever been told by a doctor that you have diabetes?" was used to assess the history of diagnosed diabetes in both surveys. Diabetes that manifested in women only during pregnancy was excluded. Smoking status was derived from 2 questions in both surveys, "Do you smoke cigarettes now?" and "Have you smoked at least 100 cigarettes in your entire life?" Smoking status was defined as current smokers (subjects who answered "Yes" to both questions), former smokers (subjects who answered "No" to the first question and "Yes" to the second question), and never-smokers (subjects who answered "No" to both questions).
Statistical Analysis We used logistic regression to estimate the strength of the association between race/ethnicity and the prevalence of periodontitis in each survey. Specifically, we performed 5 sets of analyses to estimate: (1) crude odds ratios (OR); (2) ORs adjusted for age, gender, and place of birth (AOR 1); (3) ORs additionally adjusted for smoking and diabetes (AOR 2); (4) ORs additionally adjusted for health insurance and time since last dental visit (AOR 3); and (5) ORs additionally adjusted for education and income (AOR 4). Interaction terms between race/ethnicity and gender, income, and education were tested in each survey. Data management procedures were carried out with SAS (SAS Institute Inc., 2004). Statistical analyses were conducted with SUDAAN (Research Triangle Institute, 2004) to take into account the complex sampling design yielding unbiased standard error estimates. In the Tables, the sample sizes were unweighted. However, estimates for means, proportions, standard errors, and ORs with their 95% confidence intervals (CI) were weighted.
Although blacks and Mexican-Americans in general showed worse profiles across education, income, health insurance, time since last dental visit, diabetes, and smoking than did whites, there were some improvements in education, income, dental insurance, and smoking for blacks and Mexican-Americans between the NHANES III and the NHANES 1999–2000 (Table 1  ). However, the proportion of people without health insurance and the prevalence of diabetes increased between surveys for all racial/ethnic groups.
Although blacks and Mexican-Americans exhibited worse periodontal clinical conditions than whites in both surveys, the presence of bleeding and recession, the means for CAL and PD, and the extent and severity of sites with CAL and PD with 3 mm and 4 mm decreased significantly for all racial/ethnic groups between the 2 surveys (data not shown). The overall prevalence of periodontitis for the NHANES III was 7.3%, while for the NHANES 1999–2000 it was 4.2% (p < 0.001, analyses not shown). Blacks exhibited the highest prevalence of periodontitis in each survey (11.4% and 6.8% for the NHANES III and the NHANES 1999–2000, respectively), followed by Mexican-Americans (6.9% and 4.6%, respectively) and whites (6.7% and 3.8%, respectively; Table 2 ). In general, this pattern was consistent across all covariates. Although foreign-born Mexican-Americans in the NHANES III exhibited prevalence similar to that of their US-born counterparts (7.3% vs. 6.5%, p = 0.27), US-born Mexican-Americans exhibited lower prevalence than their foreign-born counterparts in the NHANES 1999–2000 (2.5% vs. 6.1%, p = 0.03).
The crude OR shows that blacks were 1.8 times more likely to have periodontitis than whites in both the NHANES III and the NHANES 1999–2000 (Table 3 ). There was no statistically significant difference in the odds of having periodontitis for Mexican-Americans vs. whites in either survey. After adjustment for all covariates, blacks were 1.88 times (95%CI: 1.42, 2.50) more likely to have periodontitis than were whites in the NHANES III. However, the fully adjusted odds of having periodontitis for blacks and Mexican-Americans was not statistically different from the odds for whites in the NHANES 1999–2000. These findings are consistent with the adjusted prevalence for periodontitis for each racial/ethnic group in each survey. For the NHANES III, the adjusted prevalences for whites, blacks, and Mexican-Americans were 6.2%, 10.1%, and 6.0%, respectively; while for the NHANES 1999–2000, the corresponding prevalences were 4.0%, 5.1%, and 2.9%, respectively. We repeated the analyses for the final model (AOR 4), including number of teeth present at examination, and the odds for blacks having periodontitis changed to 1.63 (95% CI: 1.23, 2.16) in the NHANES III but remained nearly identical in the NHANES 1999–2000 (1.31 [95% CI: 0.60, 2.87]). The results for Mexican-Americans remained unchanged (data not shown). These associations did not differ by sex or income in either survey. However, blacks and Mexican-Americans with less than a high school diploma were 3.19 and 3.39 times, respectively, more likely to have periodontitis than were highly educated whites in the NHANES 1999–2000, but not in the NHANES III (p-interaction = 0.04, data not shown).
Our study found that the overall prevalence of periodontitis decreased between the NHANES III and the NHANES 1999–2000. This change was observed for all racial/ethnic groups, with blacks exhibiting the largest absolute difference between surveys. This finding was confirmed in the multivariable analyses, where the odds of having periodontitis for blacks and Mexican-Americans did not differ from the odds for whites in the NHANES 1999–2000. Blacks exhibited higher odds of having periodontitis than did whites in the NHANES III. There was no association between race/ethnicity and periodontitis in the NHANES 1999–2000. For the NHANES 1999–2000, the association between race/ethnicity and periodontitis varied with education, with less-educated blacks and Mexican-Americans exhibiting higher odds of having periodontitis than highly educated whites.
Results from previous studies investigating trends in the prevalence of periodontal diseases in the USA have been mixed. For example, Douglass et al.(1983), comparing data from the Health Examination Survey (HES, 1960–62) and the NHANES I (1971–74), found that the Periodontal Index mean values remained nearly unchanged between the HES (1.13) and the NHANES I (1.09). However, the proportion of adults with periodontal pockets decreased between surveys (from 30.1% to 26.6%). Although the NHANES I and the then-National Institute of Dental Research (NIDR) National Survey of Oral Health in US Employed Adults and Seniors (1985–86) used a different periodontal examination protocol, Capilouto and Douglass (1988) found a decrease in periodontal diseases between the NHANES I and the NIDR adult survey, as measured by disease with pockets in the NHANES I (23.4%) and the presence of PD Our study shows a lower prevalence of periodontitis for all racial/ethnic groups in the NHANES 1999–2000 relative to the NHANES III. Possible explanations for the observed difference in periodontitis between surveys could be attributed to awareness and changes in health risk behaviors such as smoking. The prevalence of current smokers decreased between the 2 surveys for all racial/ethnic groups. Although the NHANES 1999–2000 sample size is small relative to the NHANES III, and issues related to sampling variation could be raised, estimates for diabetes and smoking status obtained from the NHANES 1999–2000 for each racial/ethnic group were comparable with estimates obtained from the National Health Interview Survey 1999 and 2000, an annual survey of approximately 31,000 adults aged 18 yrs and older (1999 National Health Interview Survey, 2002; 2000 National Health Interview Survey, 2002). Similarly, the education and income estimated by the US Census Bureau indicated a significant change in educational attainment and median income for blacks between 1994 and 1999 (Educational attainment, 2005; Historical Income Tables, 2005). Thus, despite the NHANES 1999–2000 sample size, the estimates obtained from the NHANES 1999–2000 are representative of the US population. Therefore, the observed difference in the prevalence of periodontitis between surveys could be an early sign of a downward trend in periodontitis among US adults. Our analyses showed that the effect of race/ethnicity on periodontitis was modified by education in the NHANES 1999–2000 only after adjustment for all covariates in the model: Less-educated blacks and Mexican-Americans exhibited higher odds of having periodontitis than did highly educated whites. This finding underscored the disease burden of disadvantaged groups in our society. Among the strengths of this study are: the use of 2 surveys with representative national samples; the use of the same case-definition for periodontitis in both surveys, allowing for a direct comparison; and the large sample size, which allows us to control for numerous potential confounders and examine interactions. Important limitations are the cross-sectional nature of the data, which precludes the drawing of inferences regarding cause and effect, and the small sample size for the NHANES 1999–2000 relative to the NHANES III. However, because the 95%CIs observed for the estimates for the NHANES 1999–2000 were reasonably narrow, we doubt that sample size, after weighting, represented a threat to our results. Future research, including additional years of the ongoing NHANES data collection, will help address this issue further. Finally, a limitation inherent to the collection of periodontal data via national surveys is the use of partial-mouth recording, with only 2 sites (mid-buccal and mesiobuccal) in 2 randomly selected quadrants examined, under the assumption that these measurements are representative of the full mouth (Albandar et al., 1999). Because the NHANES III and the NHANES 1999–2000 used partial-mouth examination (Kingman and Albandar, 2002), the results presented here could have underestimated the prevalence of periodontitis in both surveys. Our findings indicate that the prevalence of periodontitis has decreased between the NHANES III and the NHANES 1999–2000. This change was observed across all racial/ethnic groups, with blacks exhibiting the greatest absolute difference. The fact that the change in periodontitis was observed across all covariates is encouraging and could indicate a true decline in periodontitis over time. However, the findings need to be interpreted with caution and confirmed with further data from the ongoing NHANES survey as data become available. Further, given the growing diversity of our society, the periodontal health of other racial/ethnic groups, such as Asians, should be considered in the current discourse of racial/ethnic disparities.
This work was supported by the National Institute of Dental and Craniofacial Research Grant K22DE15317 and by the Robert Wood Johnson Foundation Health & Society Scholars Program (both to LNB). Received for publication October 11, 2004. Revision received February 23, 2005. Accepted for publication May 20, 2005.
Journal of Dental Research, Vol. 84, No. 10,
924-930 (2005) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
4 mm and at least 2 sites with PD 
$14,999, $15,000 to $24,999, and 
