Background

It has been estimated that up to 39% of the non-institutionalised older adult population suffer from xerostomia [1], with a reported overall prevalence of 21.3% in males and 27.3% in females across ages 20 – 80 years [2]. Xerostomia is the subjective perception of oral dryness, and is often caused by salivary gland hypofunction resulting in low salivary output. However, it is notable that subjective xerostomia does not always correspond with objective measures of salivary flow rate [3]. The main factors causing decreased saliva generation include natural outcomes of aging, side effects of medication or medical procedures such as head and neck radiation therapy and haemodialysis, and specific medical or psychiatric conditions such as connective tissue disorders, diabetes, anxiety and depression [4].

Xerostomia negatively affects the Oral Health Quality of Life index [5]. Symptoms include sensations of dryness or thirst, difficulty speaking, chewing and swallowing food, oral discomfort, and mouth soft tissue soreness. Xerostomia can also lead to oral infection and increased incidence of dental caries [2]. One approach to mitigate these symptoms is the use of sugar-free chewing gum which stimulates a strong flow of saliva through the separate and interactive effects of mastication and taste [6]. It has been used to provide symptomatic relief in patients suffering from xerostomia or salivary gland hypofunction. A Cochrane Collaboration review of topical therapies for dry mouth concluded that chewing gum increased saliva production in those with residual secretory capacity. Chewing gum may be preferred by patients, however, the review found no evidence to suggest the effect is greater or worse in comparison to saliva substitutes [7]. A more recent integrative review concluded that chewing gum for treatment of thirst resulted in increased salivary flow, xerostomia relief, and thirst reduction [8]. However, neither of these reviews included a meta-analysis of salivary flow rate data. Therefore, the objective of our systematic review and meta-analysis was to determine whether gum chewing leads to salivation and consequent relief from xerostomia in elderly and medically compromised people. Such evidence could support the development of interventions that address xerostomia and improve quality of life in both healthy and challenged populations.

Methods

Protocol registration

The protocol was registered with the international prospective register of systematic reviews (PROSPERO) in accordance with PRISMA-P guidelines (PROSPERO CRD42021254485). The protocol can be accessed at: https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=254485.

Defining the research question

We formulated the research question of the systematic review using PICOS (Population, Intervention, Comparison, Outcome, and Setting) [9]. The populations studied were: 1) elderly people with xerostomia (> 60 years old, any gender, and severity of xerostomia), and 2) medically compromised people with xerostomia. These populations were not restricted to any geography and included papers from all over the world. The intervention of interest was gum chewing (with or without specific ingredients designed to promote salivation). The comparisons included gum chewing vs. no gum chewing. The outcomes used for study selection included rate of salivation / salivary volume per unit time, xerostomia relief (self-reported, e.g., the Xerostomia Inventory), and thirst. Settings of any type (e.g. laboratory, clinical, nursing home, etc.), study designs of any type (e.g. RCTs including within- and between-subjects’ designs, cross-sectional, etc.) were in scope. Only fully-refereed research studies published in the English language were included. Abstracts and review papers were excluded.

Data sources and searches

We searched the Medline (1946 to 31st March 2023) and Scopus (1806 to 31st March 2023), Web of Science, Embase and Cochrane Library (CDSR and Central) databases using the syntax outlined in Table 1. Additionally, we searched for randomised controlled trials in clinical trials registries (https://clinicaltrials.gov/ and https://www.isrctn.com/), with terms relating to xerostomia, dry mouth, chewing, mastication, gum, salivation, oral hydration, and thirst. Other methods used for identifying relevant research included laddering (manual searching) from references cited in the literature obtained, identifying possible data from conferences attended, and reviewing proprietary information.

Table 1 The search syntax used to return 9,602 records in MEDLINE, Scopus, Web of Science and Cochrane Library (CDSR and Central)
Full size table

Study selection

The papers identified by the searches were independently reviewed by two researchers (JD & MD) to assess them against the inclusion and exclusion criteria of the PICOS statement. A consensus meeting was held to discuss papers where there was disagreement on whether they should, or should not, be included.

Outcomes and variables

For the systematic review, the outcome measures were self-reported xerostomia, salivary flow rates (stimulated and unstimulated), and thirst. For the meta-analysis, the primary outcome measure was unstimulated salivary flow rate (ml / min). Since the immediate effect of chewing sugar-free gum is to increase saliva flow, we determined that unstimulated saliva flow rate would be the preferred outcome measure for the meta-analysis, since this would be more consistent across studies, and would be least influenced by different stimuli used to collect saliva. Furthermore, unstimulated, but not stimulated, salivary flow rate has been found to be significantly correlated with xerostomia symptoms [10].

Data extraction and quality assessment

Following the selection of papers, the two reviewers independently examined each paper for risk of bias using Cochrane’s RoB 2 (for individually-randomised, parallel-group trials and cross-over trials) [11] and ROBINS-I tools [12]. A second consensus meeting was held to decide on the final, agreed, risk of bias.

In addition to the systematic review, a meta-analysis was conducted on a subset of the selected papers where both a gum chewing intervention was imposed for two or more weeks, and where unstimulated salivary flow rate was measured (ml / min). Data extraction was performed independently by two reviewers (JD & MBH) and the results compared to highlight any potential errors. Data were extracted from the text, tables and figures of each of the papers. Additionally, we also recorded: 1) authors and publication year, 2) sample size (control and intervention), 3) intervention protocol (observation weeks), and 4) outcome data expressed as saliva flow rate. Where the data were reported in graphical form, WebPlotDigitizer [13] was used to extract the underlying numerical data. Where repeated measures were reported (i.e. multiple observations over a number of weeks), only the outcomes where the intervention had been imposed for two or more weeks were included.

Data synthesis and analysis

A random-effects meta-analysis using standardised mean differences (SMD) was performed using RStudio software (R version 4.1.3 (2022–03-10)) to assess the overall effect of gum chewing on unstimulated salivary flow rate. The data were converted to SMDs (Hedge’s g) and standard errors to obtain 95% confidence intervals (CIs). The following data were used for the calculation of SMDs: 1) mean ± standard deviation, and 2) sample size (n). None of the included studies reported the correlations between trials, thus a 0.5 correlation was assumed for all trials, as per recommendations [14]. Hedge’s g values of < 0.2, 0.2 ≤ 0.5, 0.5 – 0.8, and > 0.8 were considered to represent very small, small, medium, and large effects respectively [15]. The I2 statistic was used to assess the degree of heterogeneity, with values from < 50% indicating low heterogeneity, 50–75% moderate heterogeneity, and > 75% high level of heterogeneity. Heterogeneity between the studies was assessed using graphic exploration with funnel plots.

To assess the stability of the results and investigate the effects of outliers, a “leave one out” approach was conducted. In this approach, studies are removed one by one, and the random effect model is fitted on the remaining studies. Meta-regression was used to investigate whether the duration of the intervention (e.g. how many days participants chewed gum) influenced unstimulated salivary flow rates.

Results

Search results

The search returned a list of 9,602 papers. A summary of the number and types of exclusions is given in Fig. 1. The two reviewers disagreed over the classification of 40/9,602 papers in their independent reviews (0.76%). A consensus meeting was used to decide on the ultimate inclusion and exclusion of the disputed papers resulting in a final list of 25 publications for the systematic review [16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40].

Fig. 1
figure 1

Overview of the systematic review on the effect of gum chewing on xerostomia and salivary flow rate in elderly and medically compromised subjects (as per PRISMA statement)

Full size image

Characteristics of included studies

Each paper was reviewed to determine the experimental design, the numbers of participants, the outcome measures, and the key results (Table 2).

Table 2 Characteristics of the studies included in the systematic review of studies addressing the relationship between mastication, salivation and xerostomia
Full size table

In accordance with the PICOS criteria, all studies involved at least one intervention relating to the chewing of gum and a control condition (generally no chewing). Additionally, in two studies, gum chewing was also compared with sham chewing or a simple oral exercise [17, 26]. One study compared gum chewing with the sucking of a lemon-flavoured lozenge [34]. Five studies compared the effects of gum chewing with the effects of artificial saliva on associated measures of xerostomia [30, 31, 33, 34, 40].

The 25 studies included in the systematic review measured a range of different outcome variables. Twenty studies measured stimulated and/or unstimulated saliva flow rate [17,18,19, 21,22,23,24,25,26,27,28, 31, 32, 34,35,36,37,38,39,40], 17 studies measured self-reported xerostomia or symptoms of dry mouth [17,18,19, 21, 23,24,25,26,27, 29,30,31, 33, 34, 37, 38, 40], seven studies measured thirst [16, 20, 21, 26, 27, 29, 31], and one study measured mucosal moisture levels [22]. Of the 19 studies that measured stimulated and/or unstimulated saliva flow rate, 9 found a positive effect of gum chewing [17, 18, 22, 24, 27, 35, 36, 38, 39] and 10 did not detect any effects [19, 23, 25, 26, 28, 31, 32, 34, 37, 40]. Of the six studies included in the meta-analysis, three found a positive effect of gum chewing on unstimulated saliva flow rate [17, 18, 36], and three did not detect any effects of gum chewing on unstimulated saliva flow rate [26, 31, 34]. Of the 17 studies that measured xerostomia or symptoms of dry mouth, 12 found a positive effect of gum chewing [18, 19, 21, 23,24,25,26,27, 30, 31, 38, 40] and five did not detect any effects [17, 29, 33, 34, 37]. Of the seven studies that measured thirst, six found a positive effect of gum chewing [16, 20, 21, 26, 27, 31] and one did not detect any effects [29]. The single study that measured mucosal moisture found that chewing hard gum significantly increased mucosal moisture whereas chewing soft gum did not [22].

There were no results indicating that gum chewing adversely affected levels of self-reported xerostomia or symptoms of dry mouth, thirst, stimulated and/or unstimulated saliva flow rate, or mucosal moisture levels. Seven out of the 24 studies reported minor side effects associated with chewing gum use [23, 24, 29, 30, 33, 34, 36]. For the most part, the reported frequency of these events was low, and the most common symptoms were jaw pain and gastrointestinal disturbances (gas, nausea), although one study reported a high incidence of complaints of decreased appetite [29].

Risk of bias

Eleven studies were assessed for risk of bias using the RoB 2 tool for individually-randomised, parallel-group trials [11] (Fig. 2). Of these, one had a low overall risk of bias [16]; 9 had an uncertain risk of bias [17,18,19,20, 24, 27, 28, 32, 35], and one had a high risk of bias [21]. Ten studies were assessed for risk of bias using the RoB 2 tool for cross-over trials [11] (Fig. 3). All these studies had an uncertain risk of bias [22, 23, 26, 30, 31, 33, 34, 36,37,38]. Four studies were assessed for risk of bias using the ROBINS-I tool [12] (Fig. 4). Of these, one had a low overall risk of bias [39]; two had an uncertain risk of bias [25, 29], and one had a high risk of bias [40]. The most prevalent sources of bias arose through incomplete or inadequate reporting on details such as randomisation, blinding, attrition and, in some cases, selective reporting.

Fig. 2
figure 2

Risk of bias summary using the revised Cochrane risk of bias tool for randomised trials (RoB 2). Five domains are reported: D1 (randomisation process); D2 (deviations from the intended interventions); D3 (missing outcome data); D4 (measurement of the outcome) and D5 (selection of the reported result). ‘ + ’ low risk of bias; ‘!’ some concerns, and ‘- ‘ high risk of bias

Full size image
Fig. 3
figure 3

Risk of bias summary using the revised Cochrane risk of bias tool for cross-over trials (RoB 2). Six domains are reported: D1 (randomisation process); DS (bias arising from period and carryover effects); D2 (deviations from the intended interventions); D3 (missing outcome data); D4 (measurement of the outcome) and D5 (selection of the reported result). ‘ + ’ low risk of bias; ‘!’ some concerns, and ‘- ‘ high risk of bias

Full size image
Fig. 4
figure 4

Risk of bias summary using the revised Cochrane risk of bias tool for non-randomised studies—of Interventions (ROBINS-I). Seven domains are reported: D1 (confounding); D2 (selection of participants into the study); D3 (classification of interventions); D4 (deviations from intended interventions); D5 (missing data); D6 (measurement of outcomes) and D7 (selection of the reported result). ‘ + ’ low risk of bias; ‘!’ some concerns, and ‘- ‘ high risk of bias

Full size image

One study did not counterbalance the order in which treatments were imposed [22], and six studies had some concerns over selection bias [17, 25, 30, 31, 36, 40] as they reported insufficient detail to fully explain how participants were selected for inclusion in the trial (Table 2). Ten studies were judged to have some concerns over the blinding of participants and personnel [17, 19, 22, 25, 27, 29,30,31, 36, 40]. Whilst the blinding of participants is not possible with a gum chewing intervention, study personnel should have been blinded where possible. In many cases, blinding of study personnel was not reported which resulted in some concerns over bias. Thirteen studies were judged to have some concerns over bias due absent or unreported blinding of outcome assessment [17, 19, 22, 25,26,27, 29,30,31, 34, 36, 37, 40]. Three studies were judged to have some concerns over bias due to incomplete outcome data [22, 26, 38]. In these cases it was not possible to determine whether all the participants recruited had completed the study. Levels of attrition (if present) were not reported, and degrees of freedom were not quoted to indirectly ascertain the sample sizes present in the analysis. Nine studies were judged to have some concerns over bias due to selective reporting [17, 22, 23, 25, 27, 29, 36, 37, 40]. In most cases, this risk arose through the partial reporting of either the objective (e.g. saliva flow rates) or subjective (e.g. self-reported relief from xerostomia) outcome measures. Four studies were judged to have some concerns over bias due to other sources [17, 19, 23, 25]. These sources included imbalanced groups [19], potentially confounded treatments [17], unclear statistical analysis [23] and saliva collection protocols [23, 25].

Table 3 Data used for the meta-analysis. Data are mean rate of unstimulated salivation (ml / min)
Full size table

Effect of gum chewing on unstimulated rate of salivation

Six studies contained data suitable for inclusion in the meta-analysis (Table 3). Three of these studies were of patients undergoing dialysis [18, 26, 31], one was of an elderly population [17], one was of patients with rheumatism [36], and one was a population of patients with diagnosed xerostomia / hyposalivation [34].

Each of the studies had some overall concerns over bias. It was decided to include these studies because the primary outcome measure (salivary flow rate) was less likely to be biased by factors such as blinding of participants and personnel than subjective measures such as self-reports. However, this assumption was tested through sensitivity analysis and measures of heterogeneity.

The meta-analysis results are presented in the forest plot in Fig. 5. There was a significant overall effect of gum on saliva flow outcomes compared to control (SMD = 0.44, 95% CI: 0.22—0.66; p = 0.00008; I2 = 46.53%). The effect is small but certain, heterogeneity is average but less than 50%.

Fig. 5
figure 5

Forest plot

Full size image

Heterogeneity

Heterogeneity between the studies was assessed using graphic exploration with funnel plots in Fig. 6.

Fig. 6
figure 6

Funnel plot

Full size image

The results show that nearly all the studies are inside the funnel. One study, however, has a large SMD and lies outside the triangle as it has a large mean difference.

Sensitivity analysis

To assess the stability of the results and outlier analysis using a “leave one out” approach was conducted. In this approach studies were removed one by one, and the random effect model was fitted on the remaining studies. Figure 7 shows that the results remain stable for both the effect direction and magnitude.

Fig. 7
figure 7

Sensitivity analysis

Full size image

Effect of duration of gum chewing on unstimulated salivary flow rate

Because of the moderate heterogeneity, we used meta-regressions to investigate how the duration of the chewing intervention influenced unstimulated salivary flow rate. The results show significant decrease in heterogeneity when I2 is less than 1%. The results show also that the rate of salivation increased by 0.06 standard deviations per week (p < 0.001).

The bubble plot shows the estimated regression slope, as well as the effect size of each study (Fig. 8). To indicate the weight of a study, the bubbles have different sizes, with a greater size representing a higher weight.

Fig. 8
figure 8

Meta-regression bubble plot

Full size image

Discussion

The objective of this systematic review and meta-analysis was to investigate if gum chewing is associated with objective improvements in saliva output and subjective relief from xerostomia. Xerostomia is a condition with multiple possible aetiologies, including use of prescribed and over-the-counter medications, recreational drug use, rheumatic or autoimmune conditions, such as Sjögren’s syndrome, and radiation therapy for head and neck cancers [4]. Other associated causes include primary and secondary effects of aging [41], stress [42], and the multifactorial effects of chronic haemodialysis [43]. The physiological, neurological and cellular control of saliva secretion is complex and therefore susceptible to disruption at various levels of control. In addition, there is no consistent correlation between the subjective symptoms of xerostomia and objective measures of salivary gland function (i.e., saliva flow rates [3]). Owing to the heterogeneity of aetiological factors and the disconnect between objective and subjective measures, xerostomia is not a diagnosis or single disease entity, and certain therapeutic approaches may not be applicable to all sufferers. Both chronic haemodialysis and aging can result in salivary gland hypofunction and xerostomia due to multiple factors, including glandular fibrosis, medication use and fluid restriction [41, 43]. In contrast, Sjögren’s syndrome and radiation therapy are conditions in which the cellular secretory mechanism is irreversibly damaged and likely not responsive to stimuli such as chewing gum. However, it is apparent from several studies included in this review that stimulation of salivary flow using chewing gum has the potential to improve either subjective symptoms or objective measures of salivary output regardless of aetiology, so we elected not to constrain the breadth of the review by limiting the analysis to specific known aetiologies of xerostomia. The meta-analysis allowed assessment of the objective outcome of unstimulated whole saliva flow rate in six studies, including three conducted on kidney dialysis patients [18, 26, 31], one on subjects with self-reported xerostomia due to different medical conditions [34], one on older adults over 65 [17], and one on rheumatic patients with dry mouth [36]. Although we could not confirm that the effect of mastication or chewing gum is independent of the aetiology of the condition, we would argue the results of the meta-analysis suggests that where the effect is multifactorial (dialysis, aging), the effect of chewing was more apparent than in the studies that included patients with autoimmune conditions [34, 36] and/or post-radiation therapy [34].

Systematic review

In accordance with the commentaries made by other reviewers [8], a large proportion of the studies reviewed included patients under dialysis [18, 21, 23, 26, 27, 29,30,31], or with specific conditions such as cancer [19, 25, 33], heart disease [16], or rheumatism [36]. The extent to which the aetiology of xerostomia in such studies can be compared with other studies where participants were recruited due to a history of dry mouth [20, 24, 34, 37,38,39,40] or by age [17, 22, 28, 32, 35] remains uncertain. However, chewing gum is unlikely to have an impact on xerostomia where salivary gland tissue has been ablated by radiation therapy or other catastrophic medical/surgical causes. In accordance with this proposition, the two studies of patents being treated for head, neck or oral cancers in the present review were unable detect a significant effect of gum chewing on unstimulated saliva flow rate [19, 25]. In these cases, emerging treatments such as salivary gland regeneration, repair, or replacement may be more appropriate therapies [44, 45].

Historically, a number of questionnaires have been developed to measure the presence and severity of xerostomia [46]. Initially Fox et al. (1987) defined nine questions related to xerostomia which predicted low saliva flow rates [3]. Later, the 11-item Xerostomia Inventory (XI) was proposed by Thomson et al. (1999) to develop valid multi-item method of measuring the symptoms of xerostomia which includes the wide range of xerostomia symptoms in a single quantitative measure [47]. An associated instrument was also proposed by Torres et al. (2002) which was based on 10 questions [48]. Other instruments have been developed for specific populations. For example, to measure quality of life in patients with head and neck cancer, measurements related to dry mouth / xerostomia are also included in the EORTC QLQ-H&N35 questionnaire developed by Bjordal et al. [49].

Of the studies that measured saliva flow rates, 13 studies also collected self-reported xerostomia data (e.g. xerostomia inventory [21, 26, 27, 31], EORTC QLQ-H&N35 questionnaires [19, 25], symptoms of dry mouth questionnaire [17, 24] or other questionnaire-based instruments [34, 40], visual analogue scales [18, 23, 38], and interviews [37]). There was insufficient data to conduct a meta-analysis to investigate whether increases in the rate of salivation were associated with self-reported relief from xerostomia. We had hoped to conduct a meta-analysis on self-reported data, however, the types of index used across the studies were inconsistent. It is recommended that experts align on a common instrument for future studies to allow direct comparison and meta-analysis.

Garcia et al. [8] reviewed 12 studies on the effect of chewing gum on thirst in healthy and unhealthy adults. Five of these studies found that chewing gum increased salivary flow [25, 27, 37, 38, 50]. Seven studies found that chewing gum increased xerostomia relief [26, 27, 30, 31, 33, 34, 37], and four studies found that chewing gum increased thirst reduction [26, 27, 30, 31]. Garcia et al. [8] concluded that gum chewing resulted in increased salivary flow, xerostomia relief, and thirst reduction.

Meta-analysis

The reported effects of gum chewing on saliva flow rate across the individual studies of the systematic review were ambiguous. Nine of the 19 studies that measured stimulated and/or unstimulated saliva flow rate found a positive effect whereas 10 studies did not detect an effect (see Results section). In the absence of reported power analyses it is uncertain whether the numbers of participants were sufficient to detect an effect. However, meta-analysis can provide a more precise estimate of the effect of treatment or outcomes than any individual study contributing to the pooled analysis [51].

To the best of our knowledge, this is the first time meta-analysis has been applied to assessment of the effect of chewing gum as an intervention for dry mouth. A meta-analysis was neither included in the integrative review conducted by Garcia et al. [6], nor the Cochrane Collaboration review [7]. A recent review on the efficacy of malic acid mouth sprays on xerostomia and salivary flow rates included meta-analysis of flow rate data and, showed a significant effect of the treatment on unstimulated saliva flow rates [52] and, in the present review, our meta-analysis confirmed a small, statistically significant effect of mastication on unstimulated rate of salivation in challenged populations. An average level of heterogeneity was present in the results and the funnel plot suggested some publication bias might have been present. A meta-regression found that the duration of the intervention influences the changes in salivation rate, with longer periods of chewing being associated with higher rates of salivation (in the range 2 – 12 weeks).

Our meta-analysis could be criticised for including studies that had some concerns over the overall risk of bias. However, in many cases the potential sources of bias were a lack of reporting on issues such as blinding (which is acknowledged to be difficult / impossible in studies of gum chewing) and randomisation. To address these issues, we investigated the potential impacts of any potential biases through sensitivity analysis and by examining the levels of heterogeneity. The sensitivity analysis confirmed that no one study was leveraging the outcome and that the results remained stable for both the effect direction and magnitude.

Variation in the methodologies used in the individual studies included in the meta-analysis could have influenced the direction and effect size detected. For example, it is possible that differences in how the chewing intervention was imposed (e.g. type and flavour of gum, frequency and duration of chewing, etc.). In addition, the methodologies used to measure saliva flow rates could have influenced the outcome. A range of different chewing gums were used across the studies included in the systematic review. However, in the meta-analysis, the gum chewing interventions were more consistent. Five studies used commercially available gums sweetened with aspartame or sorbitol or xylitol [7, 32,33,34, 53], and one used a prototype gum produced specifically for the study (no medicinal additives) [17]. Ozen et al. (2021), Bots et al. (2005b) and Fan et al. (2013) [18, 26, 30] required participants to chew for 10 min, six times per day and when their mouth felt dry or they were thirsty. Kim et al. [17] required participants to chew for 10 min, two times per day. Stewart et al. [34] instructed participants to chew ad libitum in accordance with the manufacturer’s instructions, and Risheim & Arneberg [36] instructed participants to chew for 30 min, two times per day on days 1–4 of the intervention, and five times per day on days 5–14 of the intervention.

The systematic review found that the protocols used to measure the unstimulated rate of salivation were reasonably consistent across the studies included in the meta-analysis [17, 18, 26, 30, 34, 36]. All of the studies, with the exception of one [36], reported that participants were requested to refrain from eating and drinking pre-sampling. However, the time specified ranged from 30 min [18], to either one hour [17, 26, 30] or two hours [34] pre-sampling. In addition to restricting the consumption of food and drink, four studies also required participants to refrain from smoking [17, 26, 30, 34], and four studies required participants refrain from any oral hygiene activity such as tooth brushing [18, 26, 30, 34]. Three studies used a total saliva collection period of 5 min [17, 26, 30], whereas one used a period of 10 min [34], and another 3 min [36]. In Bots et al. [30], participants were required to rinse their mouth with tap water to alleviate thirst and xerostomia. The temperature of the water was not specified. Measurements of saliva flow rate appear to be largely unaffected by collection time, but Gill et al. (2016) found 60% of the participants had a higher saliva flow rate after rinsing with water at a temperature of 10 °C compared with water at 20 and 30 °C [54]. In many cases it was not clear at which time of day the saliva samples were collected or whether this was standardised between collection sessions. It has been found that there are significant circadian rhythms in unstimulated saliva flow rate [55]. It is possible that the small variations in the saliva collection protocols may have introduced variability into the data which, in turn, may have influenced the size of effect detected. Navazesh and Kumar (2008) described techniques for measuring unstimulated and stimulated salivary flow, including a five-minute collection time for unstimulated saliva and refraining from eating, drinking, smoking and chewing gum for one hour before collection [56]. Consistent use of these methods in future studies would facilitate easier comparisons between studies.

In a study of 191 participants (aged 18–65 years) it was found that a history of frequent gum chewing was associated with higher unstimulated salivary flow rate [57] and, in an earlier study, healthy subjects instructed to chew gum regularly for eight weeks showed increased unstimulated saliva flow rates [58]. Gueimonde et al. (2016) found twice daily gum chewing progressively increased unstimulated saliva flow rates in 52 subjects over three months, becoming significant after two months compared to baseline [24]. This makes gum chewing an attractive alternative to pharmaceutical options, either prescription or ‘over the counter’, that may be less palatable, have side-effects, or are simply not as convenient to use. There also remains the intriguing possibility that functional stimulation of the salivary glands by increased mastication over time can increase either the basal (unstimulated) saliva flow rate or stimulated flow rate, at least in subjects without physical damage to, or loss of, the acinar cells. This was suggested by the results of the meta-analysis in the present study but is also in other studies not included in this analysis. For example, Guiemonde et al. [24] reported that the mean unstimulated saliva flow rate in healthy subjects with hyposalivation increased steadily in all subjects for three months of twice-daily chewing gum use and remained at the higher level for a month following cessation of the chewing regimen. Similarly, in a study of institutionalised, frail elderly subjects who chewed gum two times a day for 12 months, Simons et al. [35] demonstrated significant incremental increases in stimulated salivary flow rates in gum chewing subjects.

Limitations of the review and meta-analysis

A key limitation of any systematic review and meta-analysis is that of publication bias. There is a preference to publish studies that have statistically significant results despite the clinical significance of studies that do not detect significant results. In addition heterogeneity in study design, population, intervention, or outcome measures can make it challenging to draw a definitive conclusion about the effectiveness of the intervention being studied. The quality of studies included can vary, which can affect the overall quality of the review (although this can be managed to an extent through the use of risk of bias tools). Meta-analyses are typically based on summary data, making it challenging to control for confounding factors that may influence the results. Therefore, it is essential to interpret the results with caution, considering the potential sources of bias and confounding that may have affected the results.

Conclusion

Chewing sugar-free gum is one of many options available to manage xerostomia and symptoms of dry mouth. Our review and that of other authors suggests that chewing sugar-free gum provides relief from xerostomia, and that chewing gum daily over a period of two or more weeks increases the rate of unstimulated salivation. As xerostomia negatively affects a large proportion of both the non-institutionalised older adult population (39%), and of the general population (21.3% in men and 27.3% in women), interventions involving gum chewing offer potential to improve Oral Health Quality of Life (especially in challenged populations). Sugar-free gum is low cost, readily available, safe, is not a drug, has minimal side-effects, and is generally preferred by users to other options, such as artificial salivas. However, additional work is required to unequivocally define the relationships between gum chewing, increased salivation rate and self-reported relief from xerostomia. Progress in this regard has been hampered by a lack of standardisation on the instruments used to measure self-reported xerostomia. We also recommend that future studies clearly differentiate chronic effects of chewing on salivary flow rates from the acute effects of chewing, and if possible measure both stimulated and stimulated salivary flow rates using standardised techniques as described by Navazesh and Kumar [56].