In another Finnish study, Soderling and colleagues⁸ reported that mothers who chewed xylitol-containing gum two to three times per day beginning three months after giving birth (and ending when their children were 2 years old) were less likely to transmit mutans streptococci to their offspring, and the children were less likely to have caries.⁹ Nevertheless, xylitol-containing products are not widely used in the United States, although they have the potential to help control dental caries in high-risk patients.

Xylitol has been used as a sweetening agent in foods since the 1960s. It is an odorless, white crystalline powder with the same sweetness and bulk as sucrose, but one-third fewer calories. It is found in many fruits and vegetables (such as strawberries, raspberries and plums) and is produced in the human body during normal metabolism. Xylitol currently is approved for use in foods, pharmaceuticals and oral health products in more than 35 countries including the United States.¹⁰

We do not know how xylitol works to reduce caries. Many studies report a reduction in levels of salivary S. mutans with the prolonged use (in some cases, for years) of xylitol-containing chewing gum,² indicating that xylitol may decrease the ability of bacteria to multiply in its presence. However, one clinical study reported that levels of cariogenic bacteria initially decreased and then increased during continued xylitol exposure.¹¹ In another study, Hildebrandt and Sparks¹² reported that S. mutans levels first were lowered using chlorhexidine rinses, and then xylitol gum was used for three months to maintain bacterial suppression.

An alternative explanation is that xylitol does not have long-term growth effects on cariogenic bacteria, but is noncariogenic because it is not fermented by S. mutans or S. sobrinus.¹³ Thus, laboratory results have been interpreted to mean that the bacteria become tolerant to xylitol, and are able to grow in the presence of increased concentrations of xylitol, yet do not ferment the xylitol into tooth-damaging acids.¹³ Some researchers also have suggested that xylitol-tolerant (or resistant) bacteria adhere less well to tooth surfaces and produce less acid than do xylitol-sensitive (or susceptible) bacteria.¹³ Still others have suggested that the effect is nonspecific, perhaps the result of increased salivation from the xylitol chewing gum.¹⁴

We were impressed by the clinical findings in regard to xylitol, but believe that the absence of definitive data about its mechanism of action is a barrier to more widespread use of this sweetening agent in dental practice and public health settings. Therefore, we initiated this study to examine the effect of xylitol contained in snack foods and candy on S. mutans and S. sobrinus levels in children and adults.

	SUBJECTS, MATERIALS AND METHODS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Study 1. We collected unstimulated saliva samples at baseline and after xylitol use via tongue blade from 162 children in preschool through third grade. An additional 25 subjects who received xylitol were excluded from the study, because baseline or post–xylitol-exposure salivary samples were not collected as a result of their absence from school. After baseline salivary samples were collected, the children received snack foods containing 2.5 grams of xylitol twice a day (five days per week) for four weeks. The snack foods were macaroons manufactured for the study by Grand Central Bakery Co. (Seattle) and gelatin and sorbet manufactured for the study by Darigold (Portland, Ore.), according to recipes developed in previous work.¹⁵

Because xylitol is reported to produce unpleasant gastrointestinal, or GI, effects in high doses, we assessed GI tolerance of the xylitol-based foods in these subjects. We conducted our assessment of GI symptoms daily in school by having children indicate in an age-appropriate cartoon picture booklet any occurrences of flatulence, nausea, diarrhea or stomach cramps. Before consuming snack foods containing xylitol, the children were shown how to use the cartoon book to assess GI symptoms. They completed their cartoon books at the end of each day by placing a sticker on the cartoon or cartoons that described how they were feeling that day. We performed this assessment of GI symptoms for one week before exposing the children to xylitol, and then daily in school throughout the study period.

We used repeated-measures multivariate analysis of variance, or MANOVA, for each symptom to examine differences across the five-week assessment in regard to the percentage of time that children experienced a symptom within each week. Statistical significance was set at alpha = .05.

Study 2. Two adults consumed five soft chewable candies after breakfast and after lunch and seven candies after dinner, seven days per week for four weeks. The candies were manufactured for the study by Harmony Foods (Santa Cruz, Calif.) and contained gelatin (7.3 percent), water (9.4 percent), xylitol (20.8 percent) and lycasin (62.5 percent). Each candy contained 0.6 g of xylitol, for a total of 10.2 g per day. We collected unstimulated saliva with a tongue blade in the same manner as that done for the children’s samples. Gingival samples were collected by sterile swab (Culture Swab Plus, Difco Products, Becton, Dickinson and Co., Franklin Lakes, N.J.) at baseline, and two and four weeks after xylitol use began. The gingival samples were serially diluted to determine actual numbers. Characterized isolates (that is, verified as S. mutans or S. sobrinus, tested for xylitol susceptibility and typed according to pulsed-field gel electrophoresis, or PFGE) were taken from both gingival and tongue samples. Subjects were not receiving concurrent antibiotic therapy. In addition, the subjects completed daily side-effect diaries to assess potential GI symptoms.

The Institutional Review Board of the University of Washington, Seattle, approved both studies. We obtained informed consent from all participants or their parent or guardian before the study began.

Microbiological analysis. For this study, we defined cariogenic bacteria as S. mutans and S. sobrinus. These were identified using specific DNA probes¹⁶ and growth on the modified mitis salivarius agar plates supplemented with kanamycin, or MSKB.¹⁷ Isolates that grew on MSKB but were not identified as either S. mutans or S. sobrinus were labeled "other streptococci."

Study 1. We coated sterile tongue blades with saliva and pressed the blades onto MSKB.¹⁸ The plates were placed in an anaerobic chamber and incubated for 72 hours at 35 C. We evaluated the cariogenic bacteria (S. mutans/S. sobrinus) under magnification and determined the colony-forming units, or CFU, of the bacteria. Children were characterized as having low (0 CFU), moderate (1 to 50 CFU) or high (> 50 CFU) bacterial counts at baseline and after xylitol exposure according to criteria previously established.¹⁷

The plates then were sealed and stored at 4 C for six months. We randomly selected a subset of 25 children who had moderate or high levels of S. mutans/S. sobrinus at baseline. We recovered S. mutans from both the baseline and post–xylitol-exposure period in seven of these children. An isolate from baseline and an isolate from the post–xylitol-exposure period were studied from each of these children. In another nine of the 25 children, S. mutans was recovered from one of the samples, but not both. For the remaining nine children, no viable bacteria were obtained.

Study 2. We diluted saliva samples from the adults and plated them onto the MSKB media incubated in 5 percent carbon dioxide at 35 C for 72 hours. To verify the phenotypes of bacteria present at each time point, we collected multiple isolates at baseline and after xylitol exposure, and identified them as S. mutans, S. sobrinus or other streptococci.¹⁶

Xylitol susceptibility. We used growth curves to determine xylitol susceptibility for four clinical isolates from children and for eight control isolates.¹⁹ S. mutans 10449^S (xylitol susceptible) and S. mutans 10449^T (xylitol tolerant), S. mutans OFMA^S (xylitol susceptible) and S. mutans OFMA^T (xylitol tolerant), S. mutans T10B ^S (xylitol susceptible) and S. mutans T10B^T (xylitol tolerant) and S. sobrinus 27352 ^S (xylitol susceptible) and S. sobrinus 27352^T (xylitol tolerant) were used as controls.²⁰

To test this method, we used the isogenic (control) pairs of xylitol-susceptible (X^S) and xylitol-tolerant (X^T) strains for xylitol growth curves. The tolerant strains grew in the presence or absence of 0.5 percent xylitol, while the xylitol-susceptible isolates were inhibited by the presence of xylitol. The figure shows a representative pair of isolates. Having confirmed that this method differentiated between the control isolates, we evaluated similar growth curves for pairs of clinical isolates, which verified tolerance in the strains collected after clinical exposure to xylitol.

View larger version (50K): [in this window] [in a new window]   Figure. Representative growth curves demonstrating xylitol susceptibility (XS) and tolerance (XT) in control organisms.

We incubated the plates for 72 hours in 5 percent CO₂ and then read them for growth. We considered the end point to be the highest concentration of xylitol at which the isolates grew. Using the agar method, we then studied the behavior of the clinical isolates (Tables 1 and 2). The two methods used (that is, growth curve and agar TYE) gave concordant results.

We found different genotype patterns in the S. mutans from each subject. Six of seven children had bacteria with the same genotype at baseline and after xylitol exposure. One child had bacteria with one genotype at baseline and a different genotype after xylitol exposure. In the adults, we found no more than two genotypes of S. mutans and one genotype of S. sobrinus in any one sampling period. However, no more than three different genotypes of S. mutans were identified during the course of the adult study, even though multiple isolates were examined at each time point (at baseline and two and four weeks after xylitol exposure).

	RESULTS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Study 1. At baseline, 58, 63 and 41 children were in the low, moderate and high categories, respectively, in regard to S. mutans/S. sobrinus levels. After four weeks of xylitol use, 55, 52 and 55 of the children, respectively, were in the low, moderate and high categories. This suggests that no significant overall changes in S. mutans/ S. sobrinus levels occurred (² test, P >.05). Eighteen (17 percent) of 104 children moved from the high or moderate categories at baseline to the low category after xylitol exposure. A total of 34 (28 percent) of 121 children moved from the low or moderate categories to the high category. Eight (19.5 percent) of 41 children moved from the high category to the low category, while only one (1.7 percent) of 58 children switched from the low category to the high category. One hundred ten children did not change categories in regard to S. mutans/S. sobrinus levels.

In five of seven children whose bacteria could be recovered after storage, S. mutans at baseline failed to grow when xylitol levels in the agar medium exceeded 10 percent. The S. mutans in all seven children became more xylitol tolerant after the children consumed xylitol snacks, and the bacteria grew on agar containing 15 percent xylitol (Table 1). Six of these children had the same S. mutans genotype at both time points.

Study 2. One of the two adult subjects experienced a reduction in S. mutans and S. sobrinus levels with xylitol exposure; the bacterial isolates from this subject also became more xylitol tolerant, with growth occurring when 15 to 25 percent xylitol was added to the agar (Table 2). In the second adult, S. mutans/S. sobrinus levels increased fivefold to 10-fold during the four-week study, while the proportion of susceptible and tolerant strains sampled from both the gingiva and tongue remained the same.

At baseline, the proportion of children who reported experiencing at least one symptom ranged from 12.7 percent (21 of 165 children) for nausea/vomiting to 40 percent (66 of 165 children) for stomach cramps. However, the percentage of children with symptoms did not increase during the xylitol consumption period (P > .05). Stomach cramps decreased during xylitol exposure (MANOVA analysis: F(4, 388) = 2.87, P = .02). Neither of the adult subjects reported experiencing GI side effects.

	DISCUSSION

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
The results of our study show that four weeks of exposure to xylitol snacks or candies did not reduce S. mutans levels in children and adults. However, we found that xylitol exposure did increase the tolerance of the cariogenic bacteria. Xylitol-tolerant isolates grow in the presence of xylitol and are thought to be less able to produce caries than xylitol-sensitive isolates.¹³ In contrast to antibiotic-resistant bacteria, xylitol does not confer an ecological advantage to S. mutans or suppress the normal background flora of the mouth. Our interpretation of these clinical studies is limited by the absence of study groups of children and adults who did not receive xylitol-containing products and the small number of subjects for whom microbiological testing was conducted.

We found that identification of xylitol-susceptible and xylitol-tolerant bacteria was comparable using the agar dilution and growth curve methods. The agar dilution method was easier to perform and took less time to complete. We used both of these methods to characterize S. mutans isolated at baseline and during xylitol exposure. Although only a small number of subjects were studied, they varied by age, as well as by the source and frequency of the xylitol ingested. Five of seven children and one of two adults did not experience a reduction in S. mutans levels, but the related isolates changed from xylitol susceptible to xylitol tolerant. We found one to three S. mutans genotypes among all subjects throughout the study, and no more than two S. mutans genotypes at any one time point.

The children tolerated the xylitol-containing snacks well, with no increase in GI symptoms from baseline. Future studies should assess more children over a longer period and obtain clinical outcome measures, such as caries incidence.

	CONCLUSION

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Many dentists and dental hygienists recommend xylitol-containing chewing gum as an adjunct to topical fluorides and proper oral hygiene for caries control. Xylitol snacks and confections might be a vehicle for delivering xylitol when chewing gum is not desirable. A recent NIH consensus conference on the diagnosis and management of dental caries identified xylitol-containing products as effective and promising as a caries-preventive measure.¹ The results of this study provide a basis on which xylitol-containing products can be recommended and xylitol’s mechanism of action explained to patients.

View larger version (104K): [in this window] [in a new window]   Dr. Roberts is a professor, Department of Pathobiology, University of Washington, Seattle.

View larger version (86K): [in this window] [in a new window]   Dr. Kaakko is a resident in Pediatric Dentistry, University of Washington, Seattle.

View larger version (92K): [in this window] [in a new window]   At the time of this study, Dr. Castillo was a graduate student in Orthodontics, University of Washington, Seattle. He now is in private orthodontic practice in Lima, Peru.

View larger version (88K): [in this window] [in a new window]   Dr. Milgrom is a professor and director of the Northwest/Alaska Center to Reduce Oral Health Disparities, Department of Dental Public Health Sciences, University of Washington, Suite B509, Box 357475, Seattle, Wash. 98195-7475, e-mail "dfrc{at}u.washington.edu". Address reprint requests to Dr. Milgrom.