| Sign In to gain access to subscriptions and/or personal tools. |
Fluorosis Development in Seven Age Cohorts after an 11-month Break in Water Fluoridation
1 Department of Epidemiology, School of Public Health, University of Michigan, 109 Observatory Street, Ann Arbor, MI 48109-2029; Correspondence: *corresponding author, bburt{at}umich.edu
This study used an 11-month break in water fluoridation to identify the time when developing incisors are most sensitive to fluorosis development. The study was based in Durham, NC, where an interruption to water fluoridation occurred between September, 1990, and August, 1991. A total of 1896 children was dentally examined. Fluorosis was measured by the TF index, and parents or guardians completed a questionnaire on demographics and fluoride history. Age cohorts ranged from those born 5 years before the break, to those born 1 year after the resumption of fluoridation. Fluorosis prevalence for seven age cohorts whose birth years ranged from 1985–86 to 1991–92 was 57.1, 62.3, 33.0, 32.3, 39.8, 30.2, and 36.8%, respectively. Children aged from birth to 3 years at the break, and those born 1 year after it, had less fluorosis than those aged 4–5 years at the break.
Key Words: fluoridation • fluorosis • epidemiology • caries • children
In the United States, the main policy for ensuring widespread fluoride exposure is through water fluoridation, which has been categorized as one of the major public health achievements of the 20th century (US Public Health Service, 1999). Fluoridation’s benefits have been accompanied by an increased prevalence of dental fluorosis (Szpunar and Burt, 1987; Pendrys and Stamm, 1990; Williams and Zwemer, 1990), though fluorosis in the United States is usually seen only in the mild to very-mild categories. Children are thought to be at particular risk of fluorosis if fluoride is ingested during the "critical period", the developmental time around late secretion/early maturation at which the unerupted tooth appears to be especially sensitive to fluoride (Evans and Stamm, 1991b; DenBesten, 1999). In view of these "critical periods", a more precise definition of the ages at which fluoride most affects both caries and fluorosis will expand the science base for the use of fluoride. An opportunity to study this issue came from Durham, North Carolina, where the water supply has been fluoridated at 1.0 mg/L since the 1960s. Records from the water treatment plant in Durham show that fluoridation ceased between September, 1990, and August, 1991, because of technical problems. The fluoride level in Durham water during September, 1990, dropped sporadically to 0.1 mg/L. In October, the mean level was 0.5 mg/L, then it dropped to zero until August, 1991, when it was 1.1 mg/L. Fluoride remained around that level for the rest of 1991, and fluoridation has not been interrupted since. The stoppage was not publicized at the time, and later discussions with health professionals indicate that dentists and physicians did not try to compensate for it by prescribing fluoride supplements during the no-fluoridation period. In an earlier report from Phase 1 of this project (Burt et al., 2000), we found no caries effects, but we did find that fluorosis was less prevalent in children aged 1 to 3 years at the break than in those aged 4 or 5 years. We concluded that this was because the maxillary permanent incisors, in which fluorosis was measured, were at a more critical stage of development in the younger children at the time of the break. If that conclusion were correct, we would expect that children born after the break would have a fluorosis prevalence approximating that seen in those aged 4 and 5 when it occurred. In Phase 2 of the project, described in this paper, we tested that hypothesis by adding fluorosis data from children born 1990–92 (the same year as the break and 1 year after it) to the results for the five age cohorts seen in Phase 1.
Details of the study methods were described in the Phase 1 report (Burt et al., 2000). In brief, the study design was a follow-up cohort comparison. The time interval between the age cohorts was one year. Since fluoridation ceased on September 17, 1990, children included in a cohort were those born on September 17 ± six months (e.g., the 1985–86 cohort was comprised of children born between March 18, 1985, and March 17, 1986). Although the data were cross-sectional, examinations were spread over 6 years so that children were as close as possible to the same age at the time of examination. Phase 1 (1996–98) assessed children born between 1985 and 1989, and Phase 2 (2000–01) saw the addition of data from children born between 1990 and 1992. Informed written consent was received from the parents or guardians of all subjects who participated in the study. The research protocol was approved by the University of Michigan Health Sciences Institutional Review Board (IRB), and by the Duke University Medical Center IRB before the study began.
Recruitment of Participants
Clinical Examinations
Questionnaire Information
Statistical Analysis
Examiner Reliability Each examiner independently re-examined the same group of 23 children during the 2001 clinical examinations. Comparison of results showed 79% agreement for fluorosis, with a kappa score of 0.55. The analytic unit in each case was the tooth.
Demographics and Fluoride History
Fluorosis
Table 1  shows the bivariate relationships between fluorosis prevalence and the demographic and fluoride exposure variables from the questionnaire. Cohort effects are evident, and toothbrushing frequency and use of fluoride supplements also were significantly associated with fluorosis.
Table 2 gives results from the logistic regression model for the case-control analysis. The cohort effects were confirmed, and reported toothbrushing frequency and use of fluoride supplements in infancy were also significant predictors of fluorosis. Not included in this model, because of weak relationships in Table 1 , were gender, formula feeding in infancy, use of fluoride gels and rinses, tap or bottled water, race/ethnicity, and educational level of the female head of household.
In Phase 1 of this study (Burt et al., 2000), we examined the impact of the fluoridation break on caries as well as fluorosis. No caries effects could be discerned, a finding we attributed to continuing exposure to fluoride from other sources as well as to generally low levels of caries. We looked for caries effects in these additional two cohorts and again could find none, so caries data are not presented. A continuing decrease in caries severity was found following the cessation of fluoridation in Finland (Seppä et al., 1998) and in Germany (Künzel et al., 2000), when children were examined several years after fluoridation ceased permanently. However, in both instances there was continuing fluoride exposure from toothpaste, and Finland has comprehensive school dental services. In the German communities, social change following German re-unification included increased use of dental services and dietary improvements. But in Durham, the 11-month interruption in water fluoridation is most likely not long enough to have adverse caries effects in this generally low-caries population, and, as in the European examples, exposure to other forms of fluoride did not cease during the break. With fluorosis, the Hong Kong studies found that a permanent reduction in fluoridation level from 1.0 mg/L to 0.7 mg/L led to a reduction in fluorosis prevalence from 64% to 47% (Evans and Stamm, 1991a). Based on this finding, our hypothesis was that the children born in 1990–92 would show the same fluorosis prevalence as those children born in 1985–87 (whose anterior teeth were reasonably well-developed by the time of the fluoridation break in 1990). This hypothesis could not be supported. Fluorosis prevalence in the age cohorts born in 1990–92 was similar to the prevalence in those born in 1987–90, mostly children for whom the fluoride exposure from drinking water was interrupted at an earlier developmental stage.
In trying to explain this unexpected result, the first possibility we checked was whether the children born in 1985–87 were younger at the time of examination than those born in 1990–92. The significance of age in fluorosis measurement is that newly erupted maxillary incisors can frequently exhibit fluorosis only in the mammelons (the "snowcapping" effect), and these mammelons are worn away a few years later. Table 1
Another possibility was examiner error. This would account for our results if the examiners were using more stringent criteria for fluorosis when they examined those born in 1985–87 than they did for those born in 1990–92. The examinations were conducted over 6 years, and "examiner drift" with respect to criteria is always a possibility in studies running for that period of time. Inter-examiner consistency scores for fluorosis (kappa) were 0.67 in Phase 1 and 0.55 in Phase 2, scores which demonstrate a moderate level of inter-examiner agreement, and which are similar to the level of intra-examiner agreement reported from the Hong Kong studies (Evans and Stamm, 1991a). However, low precision in inter-examiner scores biases toward the null, so that a lower level of inter-examiner agreement would tend to make a significant difference between cohort scores less, rather than more, likely. We found in Phase 1 that one examiner scored higher than the other, so we looked at the fluorosis distribution for each examiner separately. These distributions are shown in Fig. 2
If our results are not due to age bias or examiner error, the observed pattern of high fluorosis in children born in 1985–87 could be real. That could occur if children in those cohorts were exposed to unusually high levels of ingested fluoride during their early years, though there is nothing in our data to support such a happening. Formula feeding in infancy, which has been cited as a risk factor for fluorosis (Pendrys et al., 1994), was not associated with fluorosis in our study (Table 1 ), and in any event formula feeding was distributed evenly throughout the cohorts. In addition to cohort effects, fluorosis was associated with more-frequent toothbrushing and the use of fluoride supplements in infancy. We checked for the possibility that these behaviors were concentrated in those born in 1985–87, but they were not: Both were distributed evenly across all cohorts. If the high prevalence of fluorosis in those born in 1985–87 is real, the reason would have to be some excess ingestion of fluoride in the early years. However, the actual source remains unexplained. The fluorosis effects of this 11-month break in fluoridation, following these Phase 2 results, remain uncertain. We think it most likely that the break was not long enough to lead to a reduction of fluorosis prevalence, and that the higher prevalence levels seen in children born in 1985–87 are the result of some unexplained exposure to excess fluoride in their infant years. After Phase 1 of this study, we concluded that even the short cessation of exposure to water-borne fluoride seemed to reduce fluorosis prevalence in those cohorts whose teeth were in mid-development. Our results from Phase 2, however, do not permit this conclusion to be supported.
This study was supported by NIDCR grant DE-11240. The authors extend their gratitude to the administration, principals, and teachers of Durham Public Schools, as well as to the participating children and their parents. We also gratefully acknowledge the contributions of Karen Carter, coordinator and interviewer, and the other members of the field crew: Valerie Abbott, Lydia Ruth Easter, Kelli Poteat, and Jennifer Stout. Received for publication April 12, 2002. Revision received August 26, 2002. Accepted for publication October 24, 2002.
Journal of Dental Research, Vol. 82, No. 1,
64-68 (2003) This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
