Susan T. Reisine, Ph.D., and Walter Psoter, D.D.S.:
The Surgeon General�s report (U.S. DHHS, 2000) and other reviews (Burt, Eklund, 1999) conclude that oral health is significantly related to socioeconomic status (SES), with those in the low-income segments of society being at greatest risk for dental caries. This premise is said to hold for caries incidence and prevalence among both children and adults. However, no systematic review of this relationship has been conducted, and the premise is based largely on selective reviews of the literature.
This paper presents the results of a systematic review, based on pre-established criteria, of 299 scientific papers that were deemed relevant to the topic. These 299 were selected from a total of 3,135 initially thought to be relevant. The paper also evaluates the literature on two risk factors that may partly explain SES differences in caries risk, namely, toothbrushing and infant feeding practices.
Improved prevention and management of dental caries among children and adults is the primary objective of this analysis. The results can be used to evaluate how SES serves as a risk factor for caries, and how knowledge of this risk factor can influence management of disease. The results can also be used as the basis for a research agenda on how to intervene to reduce the effects of SES on caries incidence and prevalence. Finally, results on the relationship of toothbrushing and infant feeding practices to caries risk can be integrated into an evidence-based approach to clinical management of caries.
This review focuses on eight questions:
Search Strategy
A consultant was hired by the Institute of Dental and Clinical Research (NIDCR) to construct search terms and search in two databases, MEDLINE and EMBASE, on the subjects of the study. Because of limitations in resources, we did not conduct hand searches or search unpublished studies. This is a limitation, in that it is possible that only studies showing significant effects for the risk factors of interest have been published. This review may therefore have a bias toward showing more significant relationships than are warranted.
Selection Criteria
The selection of papers on the relation of caries to SES was limited to papers in English published in 1990 or after with 100 subjects or more in more than one SES classification. Investigations of the relation between caries and behavior were limited to studies involving toothbrushing and use of the baby bottle published in 1975 or later with 25 subjects per group. The toothbrushing studies had to include at least one of the following measures of brushing: plaque scores, calculus scores, self-reports of brushing frequency, or use of fluoride toothpaste. The baby bottle studies had to include at least one of the following measures: use of a bottle past the age of 12 months, use of a bottle when the baby was put to bed at night or at nap time, frequency of bottle use during the day, or the contents of the bottle (milk, juice, etc.). Data on breastfeeding was included where reported.
SES and Caries Among Children
The quality of the evidence demonstrating a significant inverse relationship between SES and caries among young children and adolescents was moderate. Relatively few longitudinal studies were found that assessed this relationship, but many cross-sectional studies did so. Bivariate analyses generally found a strong inverse relationship between SES and caries prevalence measured by DMFS/T indices, but few studies made a distinction between occlusal and smooth surface caries. About half of the studies used multivariate analysis to adjust for confounding variables but did not consistently find that SES had a significant effect on caries prevalence. Some of the evidence suggests that the effects of SES on caries risk are attenuated in fluoridated communities.
The evidence on the relationship of SES to caries among adults was weaker, with a smaller number of studies of only moderate quality. The problem of defining caries in adult is more difficult than for children, since the most widely used measures of caries (DMFS/T indices and the root caries index) represent accumulated years of disease. Studies that reported the number of carious lesions present in adults did not provide information on the length of time that individual lesions were present or the severity of the lesions. SES was not consistently related to caries among adults, either in bivariate or multivariate analyses.
Toothbrushing and Caries
Although there were a large number of studies on toothbrushing and caries among children, there were relatively few longitudinal studies and a limited number of multivariate analyses. The results of our review were equivocal: some studies found a strong and consistent relationship between brushing and/or other measures of oral hygiene and caries incidence/prevalence, while others did not. Some studies, in fact, found that more brushing was associated with higher rates of caries. The results of multivariate analyses, where available, also were inconsistent. Other variables significantly related to caries prevalence/incidence included the use of fluoride mouth rinses, regular dental visits, SES, and snacking.
Unlike the literature on the relationship between caries and toothbrushing among children, that on adults was quite small. Only 20 papers met our inclusion criteria. Their quality was poor, and the few longitudinal cohort studies used samples of convenience rather than representative community samples. The indicators of caries were measures of disease over a lifetime. A few included new carious lesions and recurrent decay as caries measures, but those were in the minority. It is therefore not surprising that the data on the association between caries and toothbrushing among adults is equivocal, given the limited evidence.
Baby Bottle Use and Caries
The quality of the 42 papers reviewed on this topic was generally weak; only 23 percent reported multivariate analyses. Most were cross-sectional surveys that relied on retrospective reports of bottle use, making them subject to recall bias. In addition, the majority of the studies used samples of convenience. The studies did not consistently demonstrate that prolonged bottle use, use of the bottle at bed time, or contents of the bottle significantly affect caries risk.
Conclusions
There is considerable evidence that SES may be related to caries risk. The studies in general showed that those in the lower SES groups, particularly young children, demonstrate elevated risk for caries prevalence. But the quality of the data was not strong, and the association between SES and caries risk among adults was inconsistent. Further, the studies did not provide insight into how SES influences caries risk.
Toothbrushing seems to have a protective effect against caries risk, although the quality of the studies (particularly among adults) was poor. Toothbrushing as a strategy for managing caries is not well supported by the literature.
The literature on baby bottle use in relation to caries risk was weak, and no recommendations can be made about either limiting bottle use to prevent caries or altering the current recommendations about prolonged bottle use or putting a child to bed with a bottle.
Recommendations
Longitudinal studies of socioeconomic status in relation to caries risk are needed, particularly among adults. This would require additional discussion of how to define caries as well as how to measure SES in a way that would provide a better understanding of how it contributes to poor health. Likewise, longitudinal studies of toothbrushing and baby bottle use in relation to dental caries are needed to assess the role of these risk factors in caries incidence and prevalence.
References
Burt B, Eklund S (eds): Dentistry, dental practice and the community. Chapter 19, Dental caries. Philadelphia: Saunders, 1999:212�36.
U.S. Department of Health and Human Service. Oral health in America: a report of the Surgeon General. Rockville, MD: U. S. Department of Health and Human Service, National Institute of Dental and Craniofacial Research, National Institutes of Health; 2000.