Methods. The authors reviewed more than 200 articles and examined 19 studies involving more than 10,000 adults to assess any trends in dental anxiety. They investigated mean anxiety scores for college students and general adult samples using four measures of dental anxiety. Comparisons were made across publication year and location of study.

Results. Analyses of variance suggest stability in dental anxiety scores over time and region within both types of study samples. Regardless of assessment length (single item or multi-item), the authors found no significant trends suggesting an increase or decrease in self-reported anxiety levels.

Conclusions. These results suggest that despite an increase in general anxiety within the United States during the past 50 years, dental anxiety seems to have remained stable throughout the period.

Practice Implications. The authors discuss the stable trend of self-reported dental anxiety levels in the context of increasing general anxiety in the United States and the current structure of individual dental practices.

During the last three decades, researchers have shown a steady interest in dental fear. Hundreds of published studies have investigated dental fear in the United States and worldwide in such populations as dental patients, patients receving psychological treatment for dental anxiety, college students and the general public. Numerous studies have reported successful treatment of patients with dental fear using behavioral and pharmacological methods. Several books have been published on this topic.^3–5

During the same period, the practice of dentistry has evolved technologically. Lasers are used in many practices to perform certain procedures. Computers are used to provide information to guide treatment. New dental materials have proliferated and implants have become a routine part of dental therapy. National surveys have shown that between 1987 and 1999, dentists in private practice were "increasing their use of new introductions of dental equipment," obtaining the up-to-date technology necessary to increase patient comfort and decrease anxiety during treatment.⁶

While many patients clearly have been helped, from a public health standpoint, has dental fear increased or decreased during the past 50 years? Has negative media coverage about the danger of contracting AIDS in the dental office or microbes living in dental unit waterlines actually increased people’s fear of going to the dentist? Or have the hundreds of studies of dental fear and the continuing changes in dental practice decreased the level of dental fear in the U.S. population?

	METHODS

TOP ABSTRACT METHODS ANALYSES RESULTS PATTERNS OF PAIRWISE DIFFERENCES DISCUSSION CONCLUSION REFERENCES

 
We conducted a literature review to locate studies in which the level of dental fear was measured for different populations within the past 50 years. The earliest study was published in 1954 and reported the dental fear level of 30 patients in a private dental practice in New Jersey.⁷ We found more than 200 articles involving dental fear, and reviewed in detail 128 of these articles that were published from 1955 through 2000. We excluded opinion papers or discussions of treatments for dental anxiety that did not contain data (that is, sample sizes, means and standard deviations of anxiety measures).

Information recorded included article title, authors, publication title, publication date, geographic location, sample type (that is, college students, adults in the general population) and number of participants in the study. Where available, we also noted the year that data were collected, subjects’ mean age, education, income, ethnicity and sex. We recorded data collection methods and measures used to assess dental fear, as well as the statistical results of the study.

Groups studied during different periods must be relatively similar to each other and representative of a wider population.

While many important studies of dental fear have been conducted outside the United States, this study focused only on studies conducted within the United States. Attitudes often are influenced by a person’s cultural background, and it is challenging to determine the causes of different levels of dental fear across cultures.⁸ To compare changes in dental fear levels over time, a single culture must be examined.

In addition, groups studied during different periods must be relatively similar to each other and representative of a wider population. Random samples of the general population fit this criterion, and studies have used telephone surveys to reach an adequate sample of the target population. However, a limited number of studies have used this method of data collection.

Many published studies have investigated dental fear within groups of patients receiving psychological treatment for dental anxiety or in patients in general dental settings. However, these groups do not adequately represent the overall population, in that patients who are being treated psychologically for dental fear are much more fearful of dentistry than is the general population. On the other hand, regularly attending dental patients generally are less afraid of dentistry than is the general population.

Estimates of regular dental attendance vary. As mentioned above, the 2000 surgeon general’s report on oral health indicated that only two-thirds of U.S. adults had visited a dentist within the past year,¹ while a 1997 nationwide study found that 74 percent of adults surveyed reported they had seen a dentist within the past year.⁹

Patient groups also vary in age, socioeconomic status and educational levels. In addition, dental clinics (such as oral surgery, general dentistry and emergency dentistry) differ with respect to the kind of patients they treat and the patients’ frame of mind when fear is measured during treatment. Dental anxiety also varies with the type of procedure performed. Patients undergoing prosthodontic treatment have reported less dental anxiety than those receiving other forms of treatment, with periodontal patients reporting the most anxiety.¹⁰ Studies of patient groups also tend to be small (n < 100). Because patient groups differ from each other in many more features than do college students or general population groups, measurements of dental fear in patient groups are difficult to compare over time and place, and for these reasons were not examined in this study.

College students are a stable segment of the population, if not completely representative of the general population. Age, income and educational levels remain relatively consistent across time and location. These studies often involve large sample sizes (n > 100). Because many university courses require students to participate in research, a student sample is not entirely self-selected. We can expect that students with a wide range of dental anxiety levels will participate in these studies. We believe that the impact of efforts to change dental fear in the United States may be adequately assessed by examining changes in dental fear in college student samples.

To compare dental fear over time for a representative group in the culture, the same measure of dental fear must be used in each study. Locker and colleagues¹¹ demonstrated in a study of three measures of dental fear that "while there was a significant association between scores on pairs of measures, the agreement between them was far from perfect. Kappa values ranged from 0.37 to 0.56, indicating only fair-to-moderate agreement beyond chance."¹¹ Thus, to assess changes in dental fear levels, the same measure of dental fear must be used in each place and time.

Newton and Buck¹² reviewed 15 measures of dental care anxiety with regard to reliability, validity and other features. They stated that the Corah Dental Anxiety Scale, or DAS, is the "most widely used questionnaire measure of anxiety" and that "the internal consistency and test-retest reliability of the scale are high." The Kleinknecht Dental Fear Survey, or DFS, has been broken down into three factors that are reliable and stable across different groups of respondents.¹² Data are available on the internal consistency of the scale, test-retest reliability and validity of the questionnaire, all of which are satisfactory, according to the authors.¹²

A single-item scale used by Gatchel contains information regarding its validity, and "reflects an overall fearfulness of dental treatment, whereas the Corah DAS measures a more specific fear of the dental situation (for example, sitting in the chair, having treatment performed)."¹² A single dental anxiety question used by Milgrom also contains validity information, but "tends to overestimate severe dental anxiety."¹² Therefore, these four dental anxiety measures appear to be acceptable for the objectives of our study.

	ANALYSES

TOP ABSTRACT METHODS ANALYSES RESULTS PATTERNS OF PAIRWISE DIFFERENCES DISCUSSION CONCLUSION REFERENCES

 
Using the methods outlined above, we found 19 studies that addressed the question of whether dental fear is increasing or decreasing in the United States. We located seven studies in six published articles^13–18 that used versions of the Corah DAS to measure dental fear in college students, as well as one unpublished study (L.J. Heaton, T.A. Smith, A.S. Quevedo, C.R. Carlson, unpublished data, 2001). We reviewed four studies conducted from 1973 to 1997 that used the Kleinknecht DFS to measure college students’ dental fear levels.^16,17,19,20 In addition, we compared two random samples of the general population in Dallas evaluated by Gatchel in 1983 and 1989 using a 10-point fear scale.^21,22 Finally, we reviewed five studies of random samples of the general population in Washington and Kentucky that used Milgrom’s single item to measure dental fear.^23–27

Using means, standard deviations and sample sizes for the DAS and DFS, we performed two analyses of variance for a single-factor experiment. In addition, we performed Barlett’s test of homogeneity of variance. These analyses used the computational formulas described by Winer,²⁸ first on studies shown in Table 1 and then on studies shown in Table 2. Although we did not have access to complete study data, the information obtained from our literature search was sufficient to calculate the mean squares and perform an F test (a significant F test [typically P < .05] indicates statistically reliable differences between study groups). For the Gatchel and Milgrom single-item studies (Tables 3 and 4, page 1105), we examined ² contingency tables.

	RESULTS

TOP ABSTRACT METHODS ANALYSES RESULTS PATTERNS OF PAIRWISE DIFFERENCES DISCUSSION CONCLUSION REFERENCES

An analysis of variance of these eight means yielded an F = 19.59 (degrees of freedom = 7, 4370; P < .001). Bartlett’s test of homogeneity of variance with ² = 34.19 (df = 7, P < .005) was significant. Conditions were such that the null hypothesis for the means (that is, the mean scores for the two groups would not be statistically different in spite of Bartlett’s result) was still rejected.²⁹ For eight groups, there were 28 possible unique pairs (for example, the 1967 New York study with the 1975 New York study). Using the Tukey Honestly Significantly Different, or HSD, group comparison procedure, we found that eight (29 percent) of the 28 total mean differences between groups were significantly different at the .05 level.

The mean score from the 1975 New York study¹³ was larger than the mean scores from the 1967 New York study¹³ and the 1982 Maryland study.¹⁴ The mean score from the 2001 Kentucky study (Heaton and colleagues, unpublished data, 2001) is significantly smaller than all of the other mean scores, with the exception of the mean score from the 1997 Ohio study.¹⁷ Figure 1 (page 1106) shows the eight study mean scores.

View larger version (30K): [in this window] [in a new window]   Figure 1. Mean college students’ scores on the Corah Dental Anxiety Scale by year.

A single-factor analysis of variance of the four groups (using Winer’s formula²⁸) yielded F = 13.87 (df = 3 , 863;P < .005). Bartlett’s test of homogeneity of variance with ² = 2.29 (df = 3, P < .005) was significant, but conditions indicated that the null hypothesis for the means should be rejected. For four groups, there are six unique possible pairs. The Tukey HSD group comparison procedure showed that only one of the six possible mean differences was significant at the .05 level (Washington higher than Florida). Figure 2 (page 1106) shows the four mean scores and their trends over time.

View larger version (30K): [in this window] [in a new window]   Figure 2. Mean college students’ scores on the Kleinknecht Dental Fear Survey by year.

Milgrom’s single-item dental fear measure. Table 4 shows the results of five random telephone studies of the general population in Washington and Kentucky that were conducted using the one-item measure of dental fear developed by Milgrom.^{23–25,26,27} The studies took place over 11 years and involved 3,130 respondents. For these studies, ² = 160.63 (df = 8, P < .001). These results indicate that respondents reported different levels of dental fear across two states.

For the three Kentucky studies alone,^25–27 Smith³⁰ reported ² = 9.36 (df = 4, P < .10). The P value is not significant, and the differences in results were minimal. For the two Washington studies alone,^23,24 we found significant differences (² = 56.28, df = 2, P < .001), with subjects in the 1986 Washington study²³ reporting more dental fear than subjects in the 1988 study.²⁴ When we compare the 1988 Washington study²⁴ with the three Kentucky studies,^25–27 the results are significantly different from each other (² = 55.98, df = 6, P < .001). Subjects in the 1988 Washington study²⁴ reported more dental fear than did subjects in the Kentucky studies conducted in 1989,²⁵ 1992²⁶ and 1997.²⁷ Both Washington studies^23,24 clearly found more dental fear than did the three studies conducted in Kentucky.^25–27

	PATTERNS OF PAIRWISE DIFFERENCES

TOP ABSTRACT METHODS ANALYSES RESULTS PATTERNS OF PAIRWISE DIFFERENCES DISCUSSION CONCLUSION REFERENCES

 
We have assessed changes in dental fear over time and place, as reported in studies that used four measures of dental fear, two types of groups (college students and general adult samples) and 10,359 subjects in eight states over a 34-year period. Forty-five paired comparisons were made: 28 for the Corah DAS, six for the Kleinknecht DFS, one for the Gatchel scale and 10 for Milgrom’s single-item measure. Groups differed in regard to time and place for 38 comparisons and in regard to time only for seven comparisons. We found no instances of dental fear being measured in the same year in two locations.

We found three statistically significant differences among the seven comparisons made at different times in the same state. An increase in dental fear over time was found in New York from 1967 to 1975¹³ (Table 1), while a decrease in dental fear was reported in Washington from 1986 to 1988 (Table 4)^23,24 and in Kentucky from 1994¹⁶ to 2001 (unpublished data) (Table 1). Despite these differences, self-reported dental fear generally was stable over time, as shown in Figures 1 and 2.

For the comparisons that differed in regard to both time and location, we found statistically significant differences in only 13 (34 percent) of the 38 comparisons. In seven comparisons, dental fear in Washington was higher than it was in Kentucky (Tables 1 and 4). In one instance, dental fear in Washington was higher than it was in Florida (Table 2). Kentucky had lower fear in two comparisons with New York and in one comparison each with Maryland and Michigan (Table 1). Subjects in New York had greater dental fear than those in Maryland in one comparison (Table 1).

Eleven (85 percent) of the 13 differences may be summarized as subjects in Kentucky reporting somewhat lower dental fear than those in several other states. We believe that this moderate number of differences across locations may be a result of the standardization of the practice of dentistry in the United States, brought about by the National Board Examinations, accreditation of dental colleges, national professional groups and nationwide marketing of dental products.

For each of the 13 statistically significant comparisons, we found a trend in that the later of the two studies being compared reported lower dental fear scores than did the earlier study. Therefore, a moderate decrease in dental fear over time (34 percent of the comparisons) could account for these study differences. However, it is impossible to be certain whether location, time or other factors account for these moderate differences in dental fear. We found no evidence, however, of an increase in dental fear over time.

	DISCUSSION

TOP ABSTRACT METHODS ANALYSES RESULTS PATTERNS OF PAIRWISE DIFFERENCES DISCUSSION CONCLUSION REFERENCES

 
This relatively stable pattern of results is encouraging when one looks at the bigger picture regarding anxiety in the United States. Twenge³¹ conducted two meta-analyses of anxiety from 1952 to 1993. Her analysis of college students consisted of 170 studies and 21,173 women and 19,019 men. The author’s analysis of children consisted of 99 samples with 6,600 boys and 5,456 girls between the ages of 9 and 17 years.

Twenge reported that "both college student (adult) and child samples increased almost a full standard deviation in anxiety between 1952 and 1993."³¹ Also, "the average American child in the 1980s reported more anxiety than child psychiatric patients in the 1950s. Correlations with social indices (e.g., divorce rates, crime rates) suggest that decreases in social connectedness and increases in environmental dangers may be responsible for the rise in anxiety. Economic factors, however, seem to play little role."³¹ The birth cohort (that is, the generation in which an individual comes of age) explained considerably more variance in anxiety than did family environment in most studies.

In light of this reported increase in more general forms of anxiety, we might expect a similar increase in dental anxiety. Previous studies have found a direct link between general anxiety and dental fear.^32–34 In a mail survey of 350 university employees, dentally fearful respondents reported a higher proportion of symptoms of general anxiety and depression than did employees who did not report high levels of dental fear.³²

In a sample of subjects reporting significant fears of dental injections, Kaakko and colleagues³³ found that 64 (54 percent) of the 118 subjects met Diagnostic and Statistical Manual of Mental Disorders—4th Edition criteria for at least one current or past Axis I diagnosis (the most common diagnoses included generalized anxiety disorder, major depressive disorder and specific phobias [aside from the fear of dental injections]). In a telephone survey, subjects reporting intense levels of one or more common fears were more than twice as likely to report experiencing high levels of dental fear.³⁴ Owing to this strong direct relationship between general anxiety and dental fear, we might expect a rise in the latter to follow an increase in the former. However, dental anxiety does not seem to follow the trend of increasing general anxiety in the United States.

These differences in anxiety may be due to trends in dental practice that run counter to the societal trends of a decrease in social connectedness and an increase in environmental dangers. Books written about dental anxiety urge the practitioner to build a trusting relationship with the patient, and provide specific detailed strategies for accomplishing this. Dental staff members who treat fearful patients are urged to provide strong social support for these patients.

A 1997 national survey reported that 55 percent of adults rate dentists as excellent in demonstrating a caring attitude toward patients, in contrast to 32 percent of respondents to a similar 1989 survey.⁹ In 1997, 51 percent of those surveyed rated dentists as excellent in their ability to explain to patients what they are doing, and 48 percent rated dentists as excellent in listening to what patients tell them. This contrasts with 30 percent and 23 percent of respondents, respectively, who rated dentists as excellent in these behaviors in 1989.

Dentistry has changed within the last three decades by adopting lasers, bonding procedures and other less technically threatening methods of accomplishing tasks.⁶ This may be why 63 percent of adults surveyed in 1997 felt that less pain was involved during a dental visit as an adult than it was as a child.⁹

Another possible reason why dentistry does not appear to be generating increased anxiety in the United States is the fact that its setting—the private dental practice—still follows the cottage industry model rather than the factory model. Most care is delivered in an outpatient setting, in practices that are small compared with a hospital or health maintenance organization. Of private dental practitioners in the United States, 64 percent are in a practice with no other dentists, and 22 percent work in a practice with only one other dentist.³⁵ These small traditional settings allow for social connectedness because of the continuity of staff, which is not evident in many other areas of modern medical care.

	CONCLUSION

TOP ABSTRACT METHODS ANALYSES RESULTS PATTERNS OF PAIRWISE DIFFERENCES DISCUSSION CONCLUSION REFERENCES

 
In general, comparisons of studies conducted during the past 50 years provide no clear evidence of dental fear either increasing or decreasing. Although several pairwise comparisons between study groups across locations suggest a decrease in dental fear over time, results of comparisons made within the same location are mixed and do not provide enough information to draw definitive conclusions regarding changes in dental fear over time.

One limitation inherent in this type of literature review is the lack of a common measure by which most researchers assess people’s fear of dental care. Future research should aim to use consistent measurements to allow more comparisons across sample groups.

Nevertheless, while dental fear is not dropping dramatically, it also is not rising as dramatically as is general anxiety in the United States. The fact that dental anxiety is not rising when dentists are treating patients who are increasingly anxious in general is—at least in part—a tribute to advances made in dental technology and patient management skills. In comparison with the rising tide of general anxiety in the United States, this relative reduction in dental anxiety is encouraging. The efforts of the last three decades appear to have had an impact on the problem of dental fear in our society.

View larger version (135K): [in this window] [in a new window]   Dr. Smith is a professor, Behavioral Science Department, College of Medicine, University of Kentucky, 117 College of Medicine Office Building, Lexington, Ky. 40536-0086, e-mail "tasmit01{at}uky.edu". Address reprint requests to Dr. Smith.

View larger version (131K): [in this window] [in a new window]   Ms. Heaton is a graduate student, Behavioral Science Department, College of Medicine, University of Kentucky, Lexington.