Introduction

Xerostomia, the subjective experience of dry mouth, is a common and uncomfortable condition that may compromise daily functions, such as eating and speaking, and elevate risk for other adverse oral conditions, including dental caries lesions and oral fungal infections [1–3]. Xerostomia is most well studied among older adults (typically, ages 60 or above), for whom a higher burden of prescription drugs, especially medications with anticholinergic side-effects, can dramatically increase dry mouth experience [4,5]. However, growing evidence suggests that xerostomia can be common across the life course, including in young adulthood [6–8]. Investigating xerostomia across a wider age range could advance the evidence base on emerging correlates, such as nicotine and cannabis use, stress, and sleep quality, that could inform xerostomia management for affected patients. However, existing instruments to measure xerostomia have not been validated for use among younger adults.

Valid instruments to measure dry mouth experience are essential to study xerostomia clinically and in community populations. However, the most widely used xerostomia survey instruments were developed and tested near-exclusively among older adults, while few, if any, instruments have been rigorously tested among younger populations. For example, the Xerostomia Inventory (XI) was developed as an 11-item summed scale among adults aged 65 and older [9]. Later, the XI was reduced to a 5-item Shortened Xerostomia Inventory (SXI, also called the “Summated Xerostomia Inventory” or “Summated Xerostomia Inventory–Dutch Version”) and validated among six populations of older adults [10,11].

The SXI has been used to demonstrate associations between xerostomia and several other conditions, including radiotherapy for head and neck cancer [12], cannabis use [13], and symptoms of dry eye [14]. The SXI has also been translated and tested in numerous languages, such as Chinese [15], Thai [16], German [17], and Korean [18], among others. However, it is possible that the SXI might not perform adequately among younger populations, who might experience dry mouth differently, even if the etiologies of the condition are similar. For example, two SXI items relate to swallowing and eating difficulties, which may manifest only under severe dry mouth. Additionally, an SXI item related to dry lips, a highly prevalent condition, may not be sufficiently specific to xerostomia. Developing a xerostomia instrument with sufficient sensitivity and specific for use with younger adults could permit more extensive study of dry mouth among non-elder populations.

Therefore, the goal of the present study was to develop and test a brief xerostomia inventory for use with young adult populations. This goal was achieved by implementing a survey-based cross-sectional study of xerostomia among university students that was built on the infrastructure of the Economic and Educational Contributors to Emerging Adults’ Cardiometabolic and Oral Health (3E) Study.

Methods

Xerostomia measures

A single global xerostomia item asked participants, “How often does your mouth feel dry?” (options: never, occasionally, frequently, always) [19]. To construct a numeric sum scale of dry mouth experience, 14 individual xerostomia items were presented in an individually randomized order within a list (options: never, rarely, sometimes, often, very often), similarly to earlier XI and SXI development (Table 1) [9–11]. Nine of the 14 items were part of the original XI, including 5 that were retained in the SXI. Five other items were worded specifically for this analysis based on discussion with co-authors and Youth Advisory Board members, although some featured concepts, such as taste and speech disturbances, had been raised in prior literature [9,20,21].

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Participant Responses to Xerostomia Scale Items (N = 278).

https://doi.org/10.1371/journal.pone.0349869.t001

Results

Participants who completed all 14 xerostomia scale items (N = 278) ranged in age from 18 to 24 years, 67% identified as female, 44% identified as Asian ancestry and 39% as Hispanic/Latine, 9% described their overall health as excellent and 32% as very good. In this population, the overall frequency of experiencing dry mouth according to the single global xerostomia item (“How often does your mouth feel dry?”) was 33% never, 57% occasionally, 8% frequently, and 1% always. Of the 14 individual items (Table 1), those endorsed most often (i.e., prevalence of responding at least “sometimes”) were lips feel dry (67%), eyes feel dry (34%), and sip liquids all day (33%). The items endorsed least often were cotton balls in mouth (4%), difficulty swallowing certain foods (6%), and food tastes funny (7%).

In exploratory factor analysis, one factor had an eigenvalue of 4.97 and explained 91.8% of the variance. No other factor had an eigenvalue greater than 0.5, suggesting that a unidimensional model is reasonable for the data (S1 Fig). Table 2 shows the factor loadings on the largest factor for the 14 individual xerostomia scale items. In IRT analysis (Table 2; S2 Fig), items ranged in difficulty (theta) lips feel dry (lowest theta; least severe xerostomia) to food tastes funny (highest theta; most severe xerostomia) and in discrimination from 0.95 (lips feel dry) to 2.93 (not enough saliva).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Factor loadings and item response coefficients.

https://doi.org/10.1371/journal.pone.0349869.t002

Five items were selected for the SXI-Y. One of the SXI-Y items (mouth feels dry) was part of the original SXI (Table 2). The five items selected for the SXI-Y were internally consistent (Cronbach alpha = 0.817). This was better internal consistency than reached in this dataset with the five items comprising the original SXI (Cronbach alpha = 0.715). Compared to the original SXI, the SXI-Y items maintained more of the IRT information provided by all 14 xerostomia scale items (S3 Fig).

SXI-Y sum scores were strongly correlated with the single global xerostomia item (polychoric rho: 0.711; Jonckheere-Terpstra P < 0.001) (S4 Fig). Based on receiver operating characteristic curves, the SXI-Y had a slightly higher but not statistically significantly different area under the curve than the original SXI for occurrences of dry mouth frequently or always (SXI-Y: 0.885; SXI: 0.875; P = 0.699). However, the SXI-Y had a statistically significantly higher area under the curve for occurrences of dry mouth occasionally, frequently, or always (SXI-Y: 0.887; SXI: 0.810; P = 0.002) (S5 Fig).

SXI-Y sum scores were correlated with self-reported oral health status (Table 3). Scores were higher (indicating more frequent dry mouth-related experiences) among participants who reported fair or poor oral health, gum bleeding, oral pain, or bad breath (Table 3). SXI-Y sum scores were also correlated with potential dry mouth risk factors, including fair/poor overall health, medication use, substance use (tobacco, nicotine, or cannabis), shorter sleep duration, and greater perceived stress (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Associations between Oral Conditions and Young Adult Shortened Xerostomia Inventory Score.

https://doi.org/10.1371/journal.pone.0349869.t003

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Associations between Potential Dry Mouth Correlates and Young Adult Shortened Xerostomia Inventory Score.

https://doi.org/10.1371/journal.pone.0349869.t004

Discussion

In this study of young adult undergraduate university students, xerostomia was a somewhat common experience, with approximately 10% of the sample frequently or always experiencing dry mouth. This prevalence was similar but slightly higher than national estimates from the USA (ages 18–24: 7%) [8] and Australia (ages 15–34: 9%) [7]. Based on this university population, a 5-item instrument, the Young Adult Shortened Xerostomia Inventory (SXI-Y), was developed, tested, and demonstrated favorable psychometric properties. While the SXI-Y is similar to the existing Shortened Xerostomia Inventory (SXI), which was developed exclusively among older adults, the SXI-Y had moderately better internal consistency and receiver operating characteristics. SXI-Y sum scores were associated with established xerostomia risk factors, such as medication use, and with emerging correlates, like substance use, stress, and sleep, suggesting its utility for further studies of dry mouth in age-diverse populations.

Among study limitations, the cross-sectional design does not allow evaluation of xerostomia over time, including in relation to potential risk factors and health-related consequences. Also, results from a sample of young adult university students may not necessarily generalize to all young adults, despite covering a range of socioeconomic backgrounds. The initial findings from this one sample, both SXI-Y properties and associations with other variables, should be confirmed in additional populations. Advantageously, this study featured a triangulated approach to instrument development, relying on insight from multiple statistical methods (e.g., factor analysis, item response theory) to build the SXI-Y scale [29]. Thus, the SXI-Y should have both strong correlation with any hypothetical underlying dry mouth factor and ability to discriminate between populations with and without xerostomia over a range of dry mouth severity.

SXI-Y scores in the present study were associated with use of nicotine and/or cannabis; however, the number of participants using any particular tobacco, nicotine, or cannabis product was too small to examine each product separately. This finding is consistent with recent work from a national study of adults that found each of cigarette, e-cigarette, and cannabis use to be associated with xerostomia [8]. Additionally, greater perceived stress was also associated with xerostomia in the present sample. Similar relationships were reported in studies of dental students and patients visiting a saliva clinic [30,31]. It is possible that xerostomia may engender stress or that chronic stress may impact salivary function and/or dry mouth perception [32]; prospective studies are needed to evaluate this relationship further.

In the present study, according to SXI-Y score, xerostomia was more severe among participants who slept fewer hours per night and reported worse sleep quality (albeit not statistically significantly for sleep quality). This mirrors a prior study of male military service members in Greece, which reported associations of xerostomia with poor sleep quality and excessive daytime sleepiness [33]. Salivary flow follows a circadian pattern, with less salivation overnight [34]. Speculatively, overnight dry mouth symptoms that cause waking (for example, to drink water [33]) or disrupt breathing (for example, by increasing surface tension in the upper airway [35]) could negatively impact sleep quality. Confirming this association and further examining the direction of the sleep-xerostomia relationship merits additional study.

In this analysis, participants taking any prescription medication had higher SXI-Y scores than those taking none. We did not ask participants which medications they were prescribed. Many types of medications have been associated with xerostomia, particularly those with anticholinergic effects, such as antihistamines, antidepressants, antipsychotics, antianxiety medications, and decongestants [1,2]. Some of these medication classes are commonly used among young adults, however, the relatively small number of participants prescribed medications in this study sample did not permit an in-depth analysis by medication type. Similarly, given their low prevalence, systemic conditions, such as diabetes, were not examined in this investigation. Availability of a valid xerostomia instrument for young adults could permit larger future studies of xerostomia risk factors, including systemic diseases.

Treatment approaches for xerostomia largely aim to manage symptoms, such as encouraging hydration, use of chewing gum or lozenges to stimulate saliva flow, or over-the-counter saliva stimulants or substitutes [3]. Patients taking xerogenic medications may discuss with their healthcare team substitute medicines or dose reduction [3]. Prescribing daily topical fluoride rinses or gels, is one approach to reduce caries risk among patients with medication-induced dry mouth [1,2]. Particularly relevant for younger adults but virtually unexamined, are the long-term implications of xerostomia over the life course, including the physical and psychological impacts on quality-of-life and the most effective and sustainable management strategies. The SXI-Y proposed here offers of tool for such investigations.

Other questions needing further investigation include affirming SXI-Y validity in additional populations, such as those with lower literacy levels than university students. Also, prospective studies are required to examine how SXI-Y scores change over time, which would allow calculation of a minimally important difference in SXI-Y scores that aligns with meaningful improvement in dry mouth experience. How the psychometric performance of the SXI-Y compares to the SXI in populations of middle-aged adults is another question worthy of future study. A prior Rasch analysis of the SXI suggested similar instrument performance with fewer response options [36]; a similar analysis could be carried out for the SXI-Y. Future studies that incorporate additional clinical factors could also provide further insight about xerostomia etiology or the causes of potentially related conditions, such as halitosis. For example, the present study did not measure objective salivary flow rate, water intake, or clinical outcomes, like caries lesions. Such studies could relate these factors to the subjective experience of dry mouth, leveraging the SXI-Y to extend these investigations to younger populations.

In summary, the results of this study suggest that the SXI-Y has utility as a brief, valid xerostomia scale among young adults. Further testing should be conducted in additional populations, both to test and potentially refine the instrument and to confirm emerging associations with sleep and stress, which could be important elements in managing dry mouth and improving quality-of-life among xerostomia patients. The availability of xerostomia measurement tools that have been tested in age-diverse samples can expand research on this condition to wider populations.

Supporting information

Acknowledgments

Thank you to the 3E Study collaborators, research staff, and volunteers for logistical and administrative support.