Intended for healthcare professionals


Xylitol chewing gum in prevention of acute otitis media: double blind randomised trial

BMJ 1996; 313 doi: (Published 09 November 1996) Cite this as: BMJ 1996;313:1180
  1. M Uhari, associate professora,
  2. T Kontiokari, research fellowa,
  3. M Koskela, head of divisionb,
  4. Marjo Niemela, research fellowa
  1. a Department of Paediatrics, University of Oulu, FIN-90220 Oulu, Finland
  2. b Division of Microbiology, University Hospital of Oulu, FIN-90220 Oulu, Finland
  1. Correspondence to: Professor Uhari.
  • Accepted 10 September 1996


Objective: To examine whether xylitol, which reduces the growth of Streptococcus pneumoniae, might have clinical importance in the prevention of acute otitis media.

Design: A double blind randomised trial with xylitol administered in chewing gum.

Setting: Eleven day care nurseries in the city of Oulu. Most of the children had had problems with recurrent acute otitis media.

Subjects: 306 day care children: 149 children in the sucrose group (76 boys; mean (SD) age 4.9 (1.5) years) and 157 in the xylitol group (80 boys; 5.0 (1.4) years).

Intervention: Either xylitol (8.4 g a day) or sucrose (control) chewing gum for two months.

Main outcome measures: The occurrence of acute otitis media and antimicrobial treatment received during the intervention and nasopharyngeal carriage of S pneumoniae.

Results: During the two month monitoring period at least one event of acute otitis media was experienced by 31/149 (20.8%) children who received sucrose compared with 19/157 (12.1%) of those receiving chewing gum containing xylitol (difference 8.7%; 95% confidence interval 0.4% to 17.0%; P = 0.04). Significantly fewer antimicrobials were prescribed among those receiving xylitol: 29/157 (18.5%) children had at least one period of treatment versus 43/149 (28.9%) (difference 10.4%; 0.9% to 19.9%; P = 0.032). The carriage rate of S pneumoniae varied from 17.4% to 28.2% with no difference between the groups. Two children in the xylitol group experienced diarrhoea, but no other adverse effects were noted among the xylitol users.

Conclusion: Xylitol seems to have a preventive effect against acute otitis media.

Key messages

  • It also inhibits the growth of S pneumoniae

  • When given to children with recurrent otitis mediain chewing gum xylitol reduced the occurrence of otitis media by about 40%

  • The daily dose required is not known, but 8.4 g given daily in a chewing gum seems to be effective

  • Xylitol had no effect on the nasopharyngeal car- riage of pneumococci


Xylitol is a polyol sugar alcohol and is referred to as birch sugar because it can be produced from birch trees. Natural sources of xylitol include plums, strawberries, raspberries, and rowan berries.1 Xylitol has the same relative sweetness as sucrose, and it has been used as a sugar substitute for dietary and medical purposes. Xylitol reduces the growth and acid production of Streptococcus mutans, which is the most important bacterium in the pathomechanism of dental caries.2 3 Regular consumption of xylitol has been shown to reduce the incidence of dental caries.4 5 6 7

We have previously found that xylitol inhibits the growth of S pneumoniae and S mutans in vitro during their logarithmic growth phase.8 This effect is dose dependent. We suspected that the growth inhibiting effect of xylitol against pneumococci could reduce the pneumococcal carriage rates and also reduce the incidence of acute otitis media. We tested this hypothesis in a double blind randomised trial using chewing gum to deliver xylitol to children.

Subjects and methods STUDY DESIGN

The study was a randomised double blind trial performed in 11 ordinary day care nurseries with healthy, normal children in the city of Oulu. Children start school at the age of 7 in Finland, and some of them attend day care nurseries for part of the day even during their first couple of school years. The study was explained to the parents during an evening session in each nursery. Each parent was aware that their child could be allocated to a group receiving either sucrose or xylitol chewing gum, and they gave informed consent. The study protocol was evaluated and approved by the ethics committee of the health centre of Oulu. The study material was donated by Leaf-Huhtamaki (Leaf-Huhtamaki, Turku, Finland). Xylitol chewing gums are commercially available in Finland, and their use is recommended by the Finnish Dental Association. A randomisation list was made by using random number tables. The cartridges were numbered accordingly, and the list was sealed in an envelope. The observers in the trial were unaware of the randomisation scheme. To ensure equal numbers of children in the two groups in each of the nurseries we used block randomisation with a block size of four.9 Each child was instructed to chew two pieces five times (one box) a day after meals or snacks, making a total dose of 8.4 g xylitol a day. The chewing lasted until there was no taste left or for at least five minutes. The parents were asked to proceed with the normal dietary routines but to avoid the use of additional xylitol during the study. Children in whom dental caries were noticed when we took the first nasopharyngeal sample were advised not to participate in the study.

The children were assessed by a paediatrician on the occasions when the nasopharyngeal samples were taken. Compliance was assessed by asking the parents to report the actual number of pieces of chewing gum used at the end of each month of the trial. Dental health of 225 children was examined by a dental nurse four months after the trial.


Nasopharyngeal samples were taken before the sugar challenge, at two weeks, and at the end of the study. Pneumococci were identified by (alpha) haemolysis on sheep blood agar plates, colony morphology, and optochine sensitivity.


The symptoms were followed with symptom sheets filled in by the parents. The time and reasons for being absent from day care and the use of additional xylitol products and possible medications were recorded. When a child had an appointment the physicians were asked to register their clinical diagnosis and the drugs prescribed. The criteria for diagnosis of acute otitis media were symptoms and signs of acute respiratory infection and simultaneous signs of middle ear effusion: a cloudy tympanic membrane or impaired tympanic membrane motility in pneumatic otoscopy.10


As the sample size required for the establishment of a clinically measurable effect in the carriage of pneumococci was greater than that for the determination of the occurrence of acute otitis media, the calculations of sample size were based on the carriage rates. The estimates of sample size were made on the basis of the knowledge that about 30% of the children in child day care nurseries in our area carry S pneumoniae in their nasopharynx. We postulated that a reduction in the carriage figure down to 15% would be clinically important. A reduction of this size with a two tailed P of 0.05 and a power of 0.9 required a sample size of 152 for each group.9 To ensure this sample size in each group at the end of the follow up we recruited 11 child day care nurseries in which a total of 336 children were willing to participate in our study.


The differences between the groups were tested by calculating χ2 values, with the Mann-Whitney U test or the t test, depending on the variable tested. Standard normal deviates were calculated for the differences between the proportions, and 95% confidence intervals were calculated accordingly. Logistic regression analysis was used to control for the risk factors for acute otitis media. The occurrences of the first episodes during follow up in the sucrose and xylitol groups were compared with Kaplan-Meier curves and the log rank test.


A total of 336 children were enrolled in March 1995. There were 30 drop outs, which left 306 children—149 in the sucrose group (mean (SD) age 4.9 (1.5) years) and 157 in the xylitol group (5.0 (1.4))—eligible for analysis. The baseline characteristics were similar in the two randomised groups (table 1).

Table 1

Baseline characteristics of subjects and reasons for dropping out by study group. Figures are numbers (percentages) of children unless stated otherwise

View this table:

Rhinitis and cough were the most common symptoms. There were no differences in the duration of cough or other symptoms of infections between the groups (table 2). The pneumococcal carriage rates varied from 17.4% to 28.2% during the study, and there was no difference between the groups (table 3).

Table 2

Symptoms of children as reported by parents. Figures are mean durations in days (ranges)

View this table:
Table 3

Numbers (percentages) of children with positive results for nasopharyngeal cultures of pneumococci

View this table:

The numbers of upper respiratory tract infections without acute otitis media, acute bronchitis, sinusitis, and conjunctivitis that resulted in visits to 43 different physicians were nearly equal in the two groups (table 4). The total number of attacks of acute otitis media was 43/149 among the children in the sucrose group compared with 22/157 among the children who received xylitol. The number of children with at least one episode of acute otitis media was 31/149 (20.8%) in the sucrose group and 19/157 (12.1%) among those who received xylitol (difference 8.7%; 95% confidence interval 0.4 to 17.0%; P = 0.040). The occurrence of the first attack differed significantly between the groups (log rank test P = 0.036) (fig 1) and was associated with the amount of xylitol used: those who experienced acute otitis media had forgotten their chewing gum significantly more often than those who had not had any acute otitis media, the means of forgotten xylitol being 48.8 g versus 22.4 g (difference 26.4 g; 5.7 g to 47.1 g; P = 0.024), respectively. Within the sucrose group the mean amount of gum forgotten by those with acute otitis media was 25.9 g compared with 28.0 g of those who had not had any attacks (difference −2.2 g; −27.0 g to 22.7 g; P = 0.86). The occurrence of acute otitis media in the sucrose group did not differ from the occurrence before the trial.

Table 4

Numbers of respiratory infections recorded by treating physician

View this table:
Fig 1
Fig 1

Cumulative occurrence of first attack of acute otitis media during two month monitoring period. Difference between groups is significant, log rank test P = 0.036

The total number of antimicrobial drugs prescribed in the sucrose group was 60 compared with 34 in the xylitol group. At least one episode on antimicrobial drugs was experienced by 43/149 (28.9%) children in the sucrose group and 29/157 (18.5%) children in the xylitol group (difference 10.4%; 0.9% to 19.9%; P = 0.032).

We controlled for parental education and smoking, breast feeding, use of pacifier, sibling prone to otitis, previous history of acute otitis media, previous use of xylitol, and nasopharyngeal carriage of pneumococci by using a logistic multivariate analysis. When at least one attack of acute otitis media was the dependent variable of the model, the type of sugar given to the child was significantly associated with the occurrence of acute otitis media in such a way that the children receiving xylitol had fewer attacks (P = 0.045).

Of the children who underwent a dental examination, dental decay was found in 23/114 of those who had used sucrose compared with 21/111 who had received xylitol chewing gum. The compliance with the instructions was similar in the two groups: 38% in the xylitol and 32% in the sucrose group forgot to take one or more pieces of gum, and 62% and 67%, respectively, used other xylitol products despite the instructions, although the amounts used were minor compared with the amount of xylitol used in the study.


We were able to show a clear decrease in the occurrence of acute otitis media among the children who used chewing gum containing xylitol. As our trial was a double blind study and the control group used chewing gum sweetened with sucrose, the decrease cannot be due to any mechanical factors. Also, the observed dose dependent response among those receiving xylitol and not among those receiving sucrose supports the conclusion that there was a decrease due to xylitol and not an increase due to sucrose chewing gum. This was also supported by the finding of a similar occurrence of acute otitis media before and during the trial in the sucrose group.

Our finding is best explained by the efficacy of xylitol in reducing the growth of S pneumoniae and thus preventing the attacks of acute otitis media caused by pneumococci.8 Pneumococci are the major cause of acute otitis media, causing about 30% or more of such attacks.11 12 A preceding viral infection is needed before the nasopharyngeal bacteria can rise to the middle ear and cause acute otitis media.13

On the basis of our in vitro studies we expected xylitol to reduce the growth of pneumococci so that even the nasopharyngeal carriage would have decreased. But there was no decrease in the carriage rate over time nor was there any difference between the xylitol and sucrose groups. Xylitol may still have some effect on the number of pneumococci growing in the nasopharynx, which we were not able to estimate in this study.

Being a polyol, xylitol is slowly absorbed by the gut wall and may cause osmotic diarrhoea when ingested in large amounts. Children can tolerate daily doses up to 45 g of xylitol without gastrointestinal symptoms.14 Two children in our trial stopped using xylitol because of complaints of diarrhoea. Yet the recorded number of episodes of diarrhoea and the mean duration of diarrhoea were similar in both groups.

The use of xylitol in the prevention of dental caries is widely accepted in Scandinavia. There are day care nurseries where chewing gums containing xylitol are given regularly after each meal. We thus did not change the current Finnish habits of using xylitol sweets in our trial. The chewing gum was given for only two months and always immediately after a meal or a snack. We found no difference in dental caries between the treatment groups four months after our trial. A large 40 month follow up trial showed that sucrose chewing gum increases the risk of dental caries only marginally.15

The episodes of acute otitis media were diagnosed by several doctors who used pneumatic otoscopy in diagnosis. The accuracy of pneumatic otoscopy in detecting middle ear effusion in a child with respiratory symptoms has been found to be over 80%.16 Inaccuracy in diagnosis would cause estimates of the efficacy of xylitol in our trial to be too low.

The mean age of 5 years of the children in our study was older than that of most patients with acute otitis media as the peak incidence occurs at the age of 6–15 months.17 18 The children participating in our trial had to be able to chew gum without swallowing it, which resulted in the selection of older children. Most of the children had previously had several episodes of acute otitis media and many had undergone adenoidectomies. Thus, our estimate of the efficacy of xylitol may be too optimistic and not directly applicable to unselected series of younger children.

Respiratory infections, especially acute otitis media, in children are the main reason for the use of oral antimicrobials. The emergence of multidrug resistant strains of S pneumoniae substantiates the need for new approaches in the prevention of bacterial infections. We have been able to show that xylitol, a well tolerated food additive, is an effective approach.

We thank Huhtamaki-Leaf for providing the chewing gums.


  • Funding The Emil Aaltonen Foundation, the Maud Kuistila Foundation, the Foundation for Paediatric Research in Oulu, and the Oulu University Central Hospital.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.