Translation and validation of the Hebrew version of the xerostomia inventory Authors: Nitzan Aizenbud, Doron J. Aframian, Gabriel Nussbaum, Galit Almoznino Published: 2026 ABSTRACT ObjectivesXerostomia significantly affects oral health and quality of life, yet no validated Hebrew assessment exists. This study aimed to translate the Xerostomia Inventory (XI) into Hebrew (HXI) and evaluate its validity and reliability.MethodsThe XI was translated into Hebrew using cross-cultural adaptation guidelines. The HXI was examined for internal consistency, using Cronbach’s alpha, as well as for construct validity (convergent and discriminant validity).ResultsThe HXI was completed by 102 xerostomia patients, 89.2% women, and the mean age was 63.5 ± 13.8 years with age range of 20–90 years. The mean HXI score was 39.9 ± 1.2, with score range 12–55. The HXI exhibited a high level of reliability with Cronbach’s α = 0.975. The convergent validity of the HXI, indicated by Spearman’s correlation between the HXI and the unstimulated salivary flow (USF) whole volume, demonstrated a strong, negative, and statistically significant correlation (r = -0.862). Strong negative correlations were also found between the total HXI score and other sialometry variables (ranging from r=-0.696 to r=-0.879). The absence of significant associations between the HXI scores and unrelated sociodemographic variables supports the discriminant validity of the HXI. Confirmatory factor analysis supported a robust unidimensional structure for the Hebrew 11-item HXI, with a single factor accounting for 80.3% of the total variance and all item loadings exceeding 0.82.ConclusionsThe HXI is a valid and reliable instrument for assessing xerostomia in Hebrew-speaking populations, demonstrating psychometric robustness consistent with international versions.Clinical relevanceThe HXI provides a valid and reliable tool for assessing patient-reported oral dryness, supporting diagnosis, treatment, and evaluation of interventions in Hebrew-speaking populations. Introduction Saliva, often referred to as the “Aqua Vita” of the oral cavity, plays an essential role in maintaining oral health [1]. Xerostomia, the subjective sensation of dry mouth, is commonly used as a patient-reported outcome measure (PROM) of oral dryness [2]. Xerostomia may arise, albeit not exclusively, from diminished salivary secretion volume and/or alterations in saliva composition [3]. PROMs capture the perspectives of patients on their health status and the impact of disease and treatment on disability, lifestyle, and quality of life (QoL) [4]. PROMs are typically standardized, validated self-report questionnaires completed by patients or, when necessary, by proxies, providing clinicians with patient-centered evaluations of treatment/intervention effectiveness. Initially developed for research, PROMs are now widely used in clinical practice to assess provider performance and evaluate health policies and health promotion interventions [5]. These instruments are often administered at multiple time points, such as pre- and post-operatively, to document changes in overall health, QoL, and condition-specific outcomes [5]. One of the most widely used instruments for assessing oral dryness is the Xerostomia Inventory (XI) questionnaire. Developed in 1999 by Thomson et al., it provides a multi-item assessment of xerostomia-related symptoms [6]. The original English version demonstrated good internal consistency reliability, with a Cronbach’s alpha of 0.84 [6]. The aim of the current study was to translate the XI and determine the validity and reliability of the Hebrew XI (HXI). The XI was selected due to its high reliability and widespread clinical use for assessing subjective oral dryness [7, 8], and it has been validated in multiple languages, including Spanish, Greek, and Portuguese [9–11]. Validating the HXI would provide a standardized tool for clinical practice and for evaluation of public health improvement programs. Methods Study design and setting This cross-sectional study analyzed medical records and response to the HXI of 102 patients suffering from xerostomia, who attended for treatment at the Salivary Glands Clinic, Department of Oral Medicine, Sedation and Maxillofacial Imaging, Faculty of Dental Medicine, The Hebrew University, Hadassah Medical Organization (HMO), Israel, between November 2021 and May 2022, and who met the inclusion and exclusion criteria. Ethical approval This study was approved by the HMO Institutional Review Board (IRB) (No. 0655-21-HMO, date of approval 08/11/2021). The HMO IRB approved an informed consent, and all patients provided a signed informed consent with their signatures after receiving detailed explanations about the research. Study sample inclusion and exclusion criteria Inclusion criteria Native Hebrew speakers ≥ 18 years who attended the Salivary Glands Clinic, HMO with a primary complaint of xerostomia for more than three months. Exclusion criteria Those under 18 years, pregnant/breastfeeding women, patients with lack of judgment or with cognitive problem that cannot answer the questionnaire coherently, and patients that are not native Hebrew speakers. The XI translation process The XI comprises 11 statements, and each one of them is scored from 1 to 5: (1) “Never”, (2) “Hardly ever”, (3) “Occasionally”, (4) “Frequently” and (5) “Always”. In this manner, the XI total score ranges from 11 (minimal XI score), to 55 (maximal XI score). The XI also has a “standard question” which gives another validity checking of the XI. The translation and validation process included the following steps: A.A “forward/backward translation” was conducted. For the “forward” translation, there were two bilingual translators with different backgrounds: the first translator was an oral medicine specialist (aware of the quantified concept and with clinical perspective), and the second translator was not aware of the quantified concept and without a clinical background. Both translators were native Hebrew speakers since the target language was Hebrew. They both made two independent “forward” translations of XI, from English to Hebrew.B.A third, bilingual, independent translator compared and synthesized the two translations.C.The united version was translated “backward” by two other independent, bilinguals, native English speakers, who have never been exposed to the original XI English version. One of them was a senior Bioscience researcher, and the second translator did not have any medical background. The two XI backward versions were performed from Hebrew back to English (the original XI language).D.Subsequently, a discussion with a multi-disciplinary committee which included the translators, health professionals and oral medicine specialists was conducted to review and discuss the translation process and conclude on a pre-final Hebrew-version of XI.E.The next step was the pilot study: the pre-final HXI was further tested on 10 xerostomia native Hebrew speaker patients, at the Salivary Glands Clinic.F.Each of these patients had completed the HXI pilot and was then interviewed by the primary investigator about their impressions. If the HXI was not clear enough, the patient was asked to share thoughts and suggestions for corrections. A.A “forward/backward translation” was conducted. For the “forward” translation, there were two bilingual translators with different backgrounds: the first translator was an oral medicine specialist (aware of the quantified concept and with clinical perspective), and the second translator was not aware of the quantified concept and without a clinical background. Both translators were native Hebrew speakers since the target language was Hebrew. They both made two independent “forward” translations of XI, from English to Hebrew. A “forward/backward translation” was conducted. For the “forward” translation, there were two bilingual translators with different backgrounds: the first translator was an oral medicine specialist (aware of the quantified concept and with clinical perspective), and the second translator was not aware of the quantified concept and without a clinical background. Both translators were native Hebrew speakers since the target language was Hebrew. They both made two independent “forward” translations of XI, from English to Hebrew. B.A third, bilingual, independent translator compared and synthesized the two translations. A third, bilingual, independent translator compared and synthesized the two translations. C.The united version was translated “backward” by two other independent, bilinguals, native English speakers, who have never been exposed to the original XI English version. One of them was a senior Bioscience researcher, and the second translator did not have any medical background. The two XI backward versions were performed from Hebrew back to English (the original XI language). The united version was translated “backward” by two other independent, bilinguals, native English speakers, who have never been exposed to the original XI English version. One of them was a senior Bioscience researcher, and the second translator did not have any medical background. The two XI backward versions were performed from Hebrew back to English (the original XI language). D.Subsequently, a discussion with a multi-disciplinary committee which included the translators, health professionals and oral medicine specialists was conducted to review and discuss the translation process and conclude on a pre-final Hebrew-version of XI. Subsequently, a discussion with a multi-disciplinary committee which included the translators, health professionals and oral medicine specialists was conducted to review and discuss the translation process and conclude on a pre-final Hebrew-version of XI. E.The next step was the pilot study: the pre-final HXI was further tested on 10 xerostomia native Hebrew speaker patients, at the Salivary Glands Clinic. The next step was the pilot study: the pre-final HXI was further tested on 10 xerostomia native Hebrew speaker patients, at the Salivary Glands Clinic. F.Each of these patients had completed the HXI pilot and was then interviewed by the primary investigator about their impressions. If the HXI was not clear enough, the patient was asked to share thoughts and suggestions for corrections. Each of these patients had completed the HXI pilot and was then interviewed by the primary investigator about their impressions. If the HXI was not clear enough, the patient was asked to share thoughts and suggestions for corrections. Recruitment and assessment of patients Patients who met the eligibility criteria for the study received an explanation of it and signed an informed consent form. The patients were recruited during the study period and all patients completed medical questioning, clinical examination, and sialometry. Collected data included age in years, sex (men/women), marital status, current work, collar, birth country, health-related habits, presence of any systemic morbidity, presence of salivary gland impairment related diseases, medications (including xerogenic) intake and liquids consumption. All co-morbidities, local conditions and medications intake were based on the medical information summary of referral letters from the general physician of the patient. Saliva collection Sialometry included unstimulated salivary flow rate (USF) and stimulated salivary flow rate (SSF; 2% citric acid) measurements [12]. Sialometry was performed in standardized conditions as described previously [13]. Patients were instructed to avoid food, mouthwash, and tooth brushing for 2 h before the salivary flow measurement and to avoid salivary stimulants (sialagogues) on the day of the salivary flow measurement. USF and SSF were collected using the spitting method for 10 min: the patient sat comfortably on the dental unit chair, with eyes open, head tilted slightly forward and mouth pointing vertically downwards. It was explained to the patients that saliva is allowed to accumulate on the floor of the mouth when lips are closed, and the accumulated saliva should be split out into the test tube [13]. To collect the SSF a sterile applicator saturated with 2% citric acid was applied to the lateral margins of the tongue 5 times in 15 s interval and after 1 additional minute the SSF measurement was obtained using the spitting method for 10 min. We recorded: The volume of USF whole salivary flow amount (USF whole amount), and its liquid fraction (USF liquid fraction), as well as foamy fraction (USF foamy fraction); and.The volume of SSF whole salivary flow amount (SSF whole amount), and its liquid fraction (SSF liquid fraction), as well as the foamy fraction (SSF foamy fraction). The volume of USF whole salivary flow amount (USF whole amount), and its liquid fraction (USF liquid fraction), as well as foamy fraction (USF foamy fraction); and. The volume of USF whole salivary flow amount (USF whole amount), and its liquid fraction (USF liquid fraction), as well as foamy fraction (USF foamy fraction); and. The volume of SSF whole salivary flow amount (SSF whole amount), and its liquid fraction (SSF liquid fraction), as well as the foamy fraction (SSF foamy fraction). The volume of SSF whole salivary flow amount (SSF whole amount), and its liquid fraction (SSF liquid fraction), as well as the foamy fraction (SSF foamy fraction). Statistical analysis IBM SPSS software No. 28.0 was used. Descriptive statistics were presented by means and standard error (continuous variables), and absolute numbers and percentages (categorical variables). Explanatory statistics included reliability tests. The internal consistency was evaluated by calculation of Cronbach’s alpha. Cronbach’s alpha is a measure to evaluate reliability and is presented as a number between 0 and 1. A higher value of α indicates that the different items in the tested scale are measuring the same idea/concept. Cut-offs for α were as follows: α ≥ 0.9= “excellent”, 0.9 > α ≥ 0.8= “good”, 0.8 > α ≥ 0.7= “acceptable”, 0.7 > α ≥ 0.6= “questionable”, 0.6 > α ≥ 0.5= “poor”, 0.5 > α= “unacceptable” [14]. The construct validity of the HXI was evaluated by two sub-types of validity: the convergent validity and the discriminant validity. The convergent validity expresses the extent of two factors measuring the same construct [15]. In this study, it was expressed by the correlation between the HXI total score and USF, and also between the standard question and USF, that were both analyzed by Spearman’s correlation coefficient. The cut-offs of the Spearman’s correlation coefficient were as follows: −0.1 < r < 0.1: “none or very weak” correlation, −0.3 < r<−0.1 or 0.1 < r < 0.3: “weak” correlation, −0.5 < r<−0.3 or 0.3 < r < 0.5: “moderate” correlation, −1 < r<−0.5 or 0.5 < r < 1: “strong” correlation [16]. The Discriminant validity expresses low value of correlation between two measures of distinct contrast [17]. In this study, it was expressed by the correlation between the HXI total score and sociodemographic variables and was calculated by ANOVA (Analysis of variance) test. The content validity represents the extent of relevance of items of an assessment tool in estimating a particular area. In this study it was evaluated by “content validity ratio” (CVR). CVR was measured by the experts rating each item based on its relevance: “Essential”, “Useful, but not essential” and “Not necessary”. The CVR was calculated using the formula: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:CVR=\frac{\left({n}_{e}-\frac{N}{2}\right)}{\left(\frac{N}{2}\right)}$$\end{document} (ne is the number of experts indicating an item is “Essential” and N is the total number of panel experts). CVR values can range between 1 (perfect agreement) and − 1 (perfect disagreement). A value of 0 means that half of the experts in the panel agree that a certain item is essential. The two-tailed statistical significance (α) level was defined as p < 0.05. A Confirmatory Factor Analysis (CFA) was conducted to evaluate the underlying factor structure of the 11-item HXI. The analysis used Principal Axis Factoring (PAF) as the extraction method. Sampling adequacy was assessed using the Kaiser–Meyer–Olkin (KMO) measure, and Bartlett’s test of sphericity was used to evaluate the suitability of the correlation matrix. Factor extraction was determined by the Kaiser criterion (eigenvalues > 1) and visual inspection of the scree plot. Item retention was based on a factor loading cut-off of 0.40, following the recommendations of Hair et al. [18], to ensure significant contribution to the construct. While a Promax rotation was pre-specified, it is only applicable if multiple factors are identified. Results Results of the XI translation process We performed the backward forward translation as depicted in steps A-C of the methods section. Subsequently, the multi-disciplinary committee (step D) reviewed and discussed all the translations versions: the original version, forward translation 1 version, forward translation 2 version, united version of forward translations 1 and 2, first and second backward translation, including semantic translation of the different rating options of XI and idiomatic phrases of the different 11-items and the standard question of XI. A consensus on pre-final Hebrew-version of XI was achieved regarding all questionnaire items and rating options by all committee members, in order to measure the different aspects of xerostomia. In the next step the pilot study (step E) followed by an interview (step F) were both conducted. Only one patient of the pilot pre-test group suggested corrections for one component of the HXI. The final approved Hebrew version of the Xerostomia Inventory (HXI) is presented in the supplementary file. One hundred and two xerostomic patients were recruited and assessed. Table 1 presents the socio-demographic, health-related habits, clinical and medical characteristics of the study sample. The study sample was predominantly female and middle-aged to old. The clinical profile was characterized by a high prevalence of systemic conditions, most notably Sjögren’s disease and rheumatoid arthritis, alongside common non-rheumatological comorbidities such as hypertension and hyperlipidemia. Consequently, the majority of the study sample were polymedicated, with a high frequency of xerogenic agent use.Table 1Sociodemographic, health-related habits, clinical and medical characteristics of the study sample (N=102)ParameterVariableNumber (%)SexWomen91 (89.2)Men11 (10.8)Marital statusMarried69 (67.7)Divorced13 (12.7)Widow11 (10.8)Single9 (8.8)Current workNo58 (56.8)Yes44 (43.2)CollarWhite22 (21.6)Blue22 (21.6)Not working58 (56.8)Birth countryIsrael55 (53.9)Africa15 (14.7)Asia11 (10.8)Western Europe7 (6.9)North America7 (6.9)Eastern Europe6 (5.9)South America1 (1.0)Health-related habitsAlcohol consumption25 (24.5)Past smoking21 (20.6)Current smoking1 (1.0)Rheumatological or autoimmune conditionsSjögren’s disease53 (52.0)Rheumatoid arthritis23 (22.5)Fibromyalgia18 (17.6)SLE3 (2.9)Autoimmune hepatitis2 (2.0)Osteoarthritis1 (1.0)Other autoimmune disease19 (18.6)Other medical conditionsHypertension38 (37.3)Hyperlipidemia35 (34.3)Gastroenterological disease28 (27.5)Heart Disease19 (18.6)I131 treatment11 (10.8)Liver disease9 (8.8)Diabetes7 (6.9)General radiation4 (3.9)Head and neck radiation3 (2.9)Hepatitis C2 (2.0)Other disease58 (56.8)Oncological conditionsSolid carcinoma15 (14.7)Lymphatic cancer3 (2.9)Hematological cancer1 (1.0)Salivary gland impairment related diseasesDrug induced xerostomia53 (52.0)Chronic obstructive sialadenitis19 (18.6)Sialadenosis2 (2.0)MedicationsSystemic medications use90 (88.2)ParameterMean±SEmedianMinimum-maximumAge (years)63.51±1.3665.5020–90Liquids consumption (ml)1836.46±96.521500.00100–6000Number of medications5.34±0.434.000–25Number of xerogenic medications0.7±0.11.00–5USF (Whole volume, mL)1.9±0.21.10–18.0USF (Liquid fraction, mL)1.3±0.20.90–12.0USF (Foamy fraction, mL)0.6±0.10.30–9.2SSF (Whole volume, mL)2.6±0.31.80–18.0SSF (Liquid fraction, mL)1.8±0.21.20–12.0SSF (Foamy fraction, mL)0.8±0.10.20–9.4SE standard error of the mean Table 1Sociodemographic, health-related habits, clinical and medical characteristics of the study sample (N=102)ParameterVariableNumber (%)SexWomen91 (89.2)Men11 (10.8)Marital statusMarried69 (67.7)Divorced13 (12.7)Widow11 (10.8)Single9 (8.8)Current workNo58 (56.8)Yes44 (43.2)CollarWhite22 (21.6)Blue22 (21.6)Not working58 (56.8)Birth countryIsrael55 (53.9)Africa15 (14.7)Asia11 (10.8)Western Europe7 (6.9)North America7 (6.9)Eastern Europe6 (5.9)South America1 (1.0)Health-related habitsAlcohol consumption25 (24.5)Past smoking21 (20.6)Current smoking1 (1.0)Rheumatological or autoimmune conditionsSjögren’s disease53 (52.0)Rheumatoid arthritis23 (22.5)Fibromyalgia18 (17.6)SLE3 (2.9)Autoimmune hepatitis2 (2.0)Osteoarthritis1 (1.0)Other autoimmune disease19 (18.6)Other medical conditionsHypertension38 (37.3)Hyperlipidemia35 (34.3)Gastroenterological disease28 (27.5)Heart Disease19 (18.6)I131 treatment11 (10.8)Liver disease9 (8.8)Diabetes7 (6.9)General radiation4 (3.9)Head and neck radiation3 (2.9)Hepatitis C2 (2.0)Other disease58 (56.8)Oncological conditionsSolid carcinoma15 (14.7)Lymphatic cancer3 (2.9)Hematological cancer1 (1.0)Salivary gland impairment related diseasesDrug induced xerostomia53 (52.0)Chronic obstructive sialadenitis19 (18.6)Sialadenosis2 (2.0)MedicationsSystemic medications use90 (88.2)ParameterMean±SEmedianMinimum-maximumAge (years)63.51±1.3665.5020–90Liquids consumption (ml)1836.46±96.521500.00100–6000Number of medications5.34±0.434.000–25Number of xerogenic medications0.7±0.11.00–5USF (Whole volume, mL)1.9±0.21.10–18.0USF (Liquid fraction, mL)1.3±0.20.90–12.0USF (Foamy fraction, mL)0.6±0.10.30–9.2SSF (Whole volume, mL)2.6±0.31.80–18.0SSF (Liquid fraction, mL)1.8±0.21.20–12.0SSF (Foamy fraction, mL)0.8±0.10.20–9.4SE standard error of the mean Sociodemographic, health-related habits, clinical and medical characteristics of the study sample (N=102) SE standard error of the mean The difference between the USF and the SSF was statistically significant (Related-Samples Wilcoxon Signed-Rank Test: Z = 5.232, p value < 0.001, data are not shown in Table). The high range is attributed to a single woman participant diagnosed with chronic obstructive sialadenitis who exhibited an unusually high flow rate of 18 mL/10 min and was retained in the analysis to reflect the full clinical spectrum of the recruited patients. In the USF and in the SSF, most of the saliva was liquid. The HXI was fully completed by all participants, with no exclusions required for missing data. The study sample exhibited a high symptomatic burden, as evidenced by a response distribution heavily weighted toward the higher end of the scale for both the total score and the global standard question (Table 2). While participants identified the physical sensation of oral dryness as the most severe symptom, the use of confectionery for relief was the least frequent behavior reported. Across the entire instrument, “Always” was the most prevalent selection while “Never” was the least utilized, reflecting the severe clinical nature of this specific sample. Specific mean values and response percentages for each item are detailed in Table 2.Table 2Descriptive statistics of the Hebrew Xerostomia Inventory (HXI)QuestionOptionsNumber (%)Mean±SE1. I sip liquids to aid in swallowing food1-Never9 (8.8)3.6±0.12-Hardly ever17 (16.7)3-Occasionally15 (14.7)4-Frequently23 (22.5)5-Always38 (37.3)2. my mouth feels dry when eating a meal1-Never9 (8.8)3.6±0.12-Hardly ever19 (18.6)3-Occasionally14 (13.7)4-Frequently23 (22.5)5-Always37 (36.3)3. I get up at night to drink1-Never7 (6.9)3.6±0.12-Hardly ever15 (14.7)3-Occasionally24 (23.5)4-Frequently21 (20.6)5-Always35 (34.3)4. my mouth feels dry1-Never1 (1.0)4.0±0.12-Hardly ever9 (8.8)3-Occasionally23 (22.5)4-Frequently24 (23.5)5-Always45 (44.1)5. I have difficulty in eating dry foods1-Never10 (9.8)3.6±0.12-Hardly ever15 (14.7)3-Occasionally19 (18.6)4-Frequently21 (20.6)5-Always37 (36.3)6. I suck sweets or cough lollies to relieve dry mouth1-Never11 (10.8)3.3±0.12-Hardly ever25 (24.5)3-Occasionally15 (14.7)4-Frequently22 (21.6)5-Always29 (28.4)7. I have difficulties swallowing certain foods1-Never10 (9.8)3.5±0.12-Hardly ever13 (12.7)3-Occasionally24 (23.5)4-Frequently24 (23.5)5-Always31 (30.4)8. The skin of my face feels dry1-Never7 (6.9)3.6±0.12-Hardly ever15 (14.7)3-Occasionally21 (20.6)4-Frequently22 (21.6)5-Always37 (36.3)9. My eyes feel dry1-Never6 (5.9)3.7±0.12-Hardly ever12 (11.8)3-Occasionally24 (23.5)4-Frequently17 (16.7)5-Always43 (42.2)10. My lips feel dry1-Never5 (4.9)3.6±0.12-Hardly ever15 (14.7)3-Occasionally22 (21.6)4-Frequently28 (27.5)5-Always32 (31.4)11. The inside of my nose feels dry1-Never9 (8.8)3.5±0.12-Hardly ever20 (19.6)3-Occasionally12 (11.8)4-Frequently28 (27.5)5-Always33 (32.4)Standard question: How often does your mouth feel dry?1-Never0 (0.0)3.9±0.12-Hardly ever12 (11.8)3-Occasionally22 (21.6)4-Frequently26 (25.5)5-Always42 (41.2)XI total score39.9±1.2 Table 2Descriptive statistics of the Hebrew Xerostomia Inventory (HXI)QuestionOptionsNumber (%)Mean±SE1. I sip liquids to aid in swallowing food1-Never9 (8.8)3.6±0.12-Hardly ever17 (16.7)3-Occasionally15 (14.7)4-Frequently23 (22.5)5-Always38 (37.3)2. my mouth feels dry when eating a meal1-Never9 (8.8)3.6±0.12-Hardly ever19 (18.6)3-Occasionally14 (13.7)4-Frequently23 (22.5)5-Always37 (36.3)3. I get up at night to drink1-Never7 (6.9)3.6±0.12-Hardly ever15 (14.7)3-Occasionally24 (23.5)4-Frequently21 (20.6)5-Always35 (34.3)4. my mouth feels dry1-Never1 (1.0)4.0±0.12-Hardly ever9 (8.8)3-Occasionally23 (22.5)4-Frequently24 (23.5)5-Always45 (44.1)5. I have difficulty in eating dry foods1-Never10 (9.8)3.6±0.12-Hardly ever15 (14.7)3-Occasionally19 (18.6)4-Frequently21 (20.6)5-Always37 (36.3)6. I suck sweets or cough lollies to relieve dry mouth1-Never11 (10.8)3.3±0.12-Hardly ever25 (24.5)3-Occasionally15 (14.7)4-Frequently22 (21.6)5-Always29 (28.4)7. I have difficulties swallowing certain foods1-Never10 (9.8)3.5±0.12-Hardly ever13 (12.7)3-Occasionally24 (23.5)4-Frequently24 (23.5)5-Always31 (30.4)8. The skin of my face feels dry1-Never7 (6.9)3.6±0.12-Hardly ever15 (14.7)3-Occasionally21 (20.6)4-Frequently22 (21.6)5-Always37 (36.3)9. My eyes feel dry1-Never6 (5.9)3.7±0.12-Hardly ever12 (11.8)3-Occasionally24 (23.5)4-Frequently17 (16.7)5-Always43 (42.2)10. My lips feel dry1-Never5 (4.9)3.6±0.12-Hardly ever15 (14.7)3-Occasionally22 (21.6)4-Frequently28 (27.5)5-Always32 (31.4)11. The inside of my nose feels dry1-Never9 (8.8)3.5±0.12-Hardly ever20 (19.6)3-Occasionally12 (11.8)4-Frequently28 (27.5)5-Always33 (32.4)Standard question: How often does your mouth feel dry?1-Never0 (0.0)3.9±0.12-Hardly ever12 (11.8)3-Occasionally22 (21.6)4-Frequently26 (25.5)5-Always42 (41.2)XI total score39.9±1.2 Descriptive statistics of the Hebrew Xerostomia Inventory (HXI) The internal consistency presented by Cronbach’s alpha was 0.975 showing “excellent” reliability of the HXI. Table 3 presents the inter-item correlation matrix in responses to the different questions of the HXI, including the standard question. There was a strong correlation observed among all the questions, with correlation coefficients ranging from 0.694 to 0.881.Table 3Inter-item correlation matrix between the different questions of the HXI, including the standard question1. I sip liquids to aid in swallowing food2. My mouth feels dry when eating a meal3. I get up at night to drink4. My mouth feels dry5. I have difficulty in eating dry foods6. I suck sweets or cough lollies to relieve dry mouth7. I have difficulties swallowing certain foods8. The skin of my face feels dry9. My eyes feel dry10. My lips feel dry11. The inside of my nose feels dryStandard question: How often does your mouth feel dry?1. I sip liquids to aid in swallowing food1.0000.8740.8040.7510.8570.6940.8240.7910.7950.7400.8680.7692. my mouth feels dry when eating a meal0.8741.0000.8230.8070.8630.7730.8570.8280.7670.7560.8410.8383. I get up at night to drink0.8040.8231.0000.7100.8100.7690.8230.7650.7750.7440.8490.7884. my mouth feels dry0.7510.8070.7101.0000.7770.7300.7460.7840.7510.7790.7360.8815. I have difficulty in eating dry foods0.8570.8630.8100.7771.0000.6990.8340.7930.7590.7540.8130.7876. I suck sweets or cough lollies to relieve dry mouth0.6940.7730.7690.7300.6991.0000.7310.7260.7270.7790.7020.7657. I have difficulties swallowing certain foods0.8240.8570.8230.7460.8340.7311.0000.7720.7650.7650.8080.8128. The skin of my face feels dry0.7910.8280.7650.7840.7930.7260.7721.0000.7820.8370.7920.7619. My eyes feel dry0.7950.7670.7750.7510.7590.7270.7650.7821.0000.8000.8190.81310. My lips feel dry0.7400.7560.7440.7790.7540.7790.7650.8370.8001.0000.7350.80011. The inside of my nose feels dry0.8680.8410.8490.7360.8130.7020.8080.7920.8190.7351.0000.773Standard question: How often does your mouth feel dry?0.7690.8380.7880.8810.7870.7650.8120.7610.8130.8000.7731.000 Table 3Inter-item correlation matrix between the different questions of the HXI, including the standard question1. I sip liquids to aid in swallowing food2. My mouth feels dry when eating a meal3. I get up at night to drink4. My mouth feels dry5. I have difficulty in eating dry foods6. I suck sweets or cough lollies to relieve dry mouth7. I have difficulties swallowing certain foods8. The skin of my face feels dry9. My eyes feel dry10. My lips feel dry11. The inside of my nose feels dryStandard question: How often does your mouth feel dry?1. I sip liquids to aid in swallowing food1.0000.8740.8040.7510.8570.6940.8240.7910.7950.7400.8680.7692. my mouth feels dry when eating a meal0.8741.0000.8230.8070.8630.7730.8570.8280.7670.7560.8410.8383. I get up at night to drink0.8040.8231.0000.7100.8100.7690.8230.7650.7750.7440.8490.7884. my mouth feels dry0.7510.8070.7101.0000.7770.7300.7460.7840.7510.7790.7360.8815. I have difficulty in eating dry foods0.8570.8630.8100.7771.0000.6990.8340.7930.7590.7540.8130.7876. I suck sweets or cough lollies to relieve dry mouth0.6940.7730.7690.7300.6991.0000.7310.7260.7270.7790.7020.7657. I have difficulties swallowing certain foods0.8240.8570.8230.7460.8340.7311.0000.7720.7650.7650.8080.8128. The skin of my face feels dry0.7910.8280.7650.7840.7930.7260.7721.0000.7820.8370.7920.7619. My eyes feel dry0.7950.7670.7750.7510.7590.7270.7650.7821.0000.8000.8190.81310. My lips feel dry0.7400.7560.7440.7790.7540.7790.7650.8370.8001.0000.7350.80011. The inside of my nose feels dry0.8680.8410.8490.7360.8130.7020.8080.7920.8190.7351.0000.773Standard question: How often does your mouth feel dry?0.7690.8380.7880.8810.7870.7650.8120.7610.8130.8000.7731.000 Inter-item correlation matrix between the different questions of the HXI, including the standard question We examined mean HXI scores across the four ordinal response categories of the standard dry-mouth question. The one-way ANOVA shown in Table 4 revealed a highly significant difference in HXI scores between these groups (F (3, 98) = 126.18, p < 0.001). Mean HXI scores were progressively higher in groups reporting greater frequencies of dryness, ranging from 19.6 for the “Hardly ever” group to 51.0 for the “Always” group (Table 4). Bonferroni-adjusted post-hoc tests demonstrated that all pairwise differences between categories were statistically significant (all p < 0.001).Table 4Mean HXI score by response to the standard questionStandard question: How often does your mouth feel dry?N (%)Mean HXI score (SD and 95% 95% Confidence Interval for Mean)Post‑hoc BonferroniMeanStd. Deviation95% Confidence Interval for MeanLower BoundUpper Bound2-Hardly ever12 (11.8)19.65.616.023.1Differs from 3, 4, 5 (p<0.001)3-Occasionally22 (21.6)28.35.225.930.6Differs from 2, 4, 5 (p<0.001)4-Frequently26 (25.5)41.28.037.944.4Differs from 2, 3, 5 (p<0.001)5-Always42 (41.2)51.04.6349.552.4Differs from 2, 3, 4 (p<0.001)Total102 (100)39.912.8037.442.4-Oneway ANOVA Result: F (3, 98) = 126.18, p < 0.001; Effect size: η² = 0.794 (95% CI 0.716–0.835). Category 1 (“Never”) is excluded because no patients in this dataset selected that response Table 4Mean HXI score by response to the standard questionStandard question: How often does your mouth feel dry?N (%)Mean HXI score (SD and 95% 95% Confidence Interval for Mean)Post‑hoc BonferroniMeanStd. Deviation95% Confidence Interval for MeanLower BoundUpper Bound2-Hardly ever12 (11.8)19.65.616.023.1Differs from 3, 4, 5 (p<0.001)3-Occasionally22 (21.6)28.35.225.930.6Differs from 2, 4, 5 (p<0.001)4-Frequently26 (25.5)41.28.037.944.4Differs from 2, 3, 5 (p<0.001)5-Always42 (41.2)51.04.6349.552.4Differs from 2, 3, 4 (p<0.001)Total102 (100)39.912.8037.442.4-Oneway ANOVA Result: F (3, 98) = 126.18, p < 0.001; Effect size: η² = 0.794 (95% CI 0.716–0.835). Category 1 (“Never”) is excluded because no patients in this dataset selected that response Mean HXI score by response to the standard question Oneway ANOVA Result: F (3, 98) = 126.18, p < 0.001; Effect size: η² = 0.794 (95% CI 0.716–0.835). Category 1 (“Never”) is excluded because no patients in this dataset selected that response The construct validity of the HXI was measured by convergent validity and discriminant validity. Figure 1 shows that the convergent validity of the HXI, indicated by Spearman’s correlation between the HXI and USF whole volume, demonstrated a strong and negative correlation. Strong correlations were also found between the total HXI score and other sialometry variables.Fig. 1Convergent validity presented by correlation between the total HXI and the standard question scores and sialometry variables using Spearman’s correlation (*p<0.001) Convergent validity presented by correlation between the total HXI and the standard question scores and sialometry variables using Spearman’s correlation (*p<0.001) A strong, negative correlation was observed between the standard question and the USF. Additionally, strong, negative correlations were found between the standard question and all other sialometry variables. As can be seen in Table 8 (Appendix), the discriminant validity confirmed there were no sociodemographic differences in the mean HXI scores, including all tested sociodemographic variables: sex, current work, collar, birth country, and parental birth countries. Regarding the content validity, all the three experts in the panel, including the two oral medicine specialists and the senior bioscience researcher, agreed that all of the 11 items and the standard questions in the HXI were “essential”. According to the formula: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$CVR=\frac{\left(3-\frac{3}{2}\right)}{\left(\frac{3}{2}\right)}=1$$\end{document}, the HXI has perfect agreement (CVR=1). The results of the factor analysis are shown in Table 5; Fig. 2. The sampling adequacy for the 11-item HXI was confirmed by a Kaiser–Meyer–Olkin (KMO) measure of 0.957, indicating an excellent distribution of values for factor analysis. Bartlett’s test of sphericity was highly significant (ꭓ2 (55) = 1385.419, p < 0.001), rejecting the null hypothesis that the variables were uncorrelated, and confirming the suitability of the data for factor extraction.Table 5Factor Loadings and Communalities for the HXIItem numberhxi itemFactor loadingCommunality (h2)2My mouth feels dry when eating a meal0.9300.8661I sip liquids to aid in swallowing food0.9070.82311The inside of my nose feels dry0.9020.8145I have difficulty in eating dry foods0.9020.8147I have difficulties swallowing certain foods0.8970.8053I get up at night to drink0.8900.7938The skin of my face feels dry0.8900.7929My eyes feel dry0.8730.76310My lips feel dry0.8660.7504My mouth feels dry0.8520.7266I suck sweets or lollies to relieve dry mouth0.8220.676Fig. 2Scree plot of the Exploratory Factor Analysis for the HXI Table 5Factor Loadings and Communalities for the HXIItem numberhxi itemFactor loadingCommunality (h2)2My mouth feels dry when eating a meal0.9300.8661I sip liquids to aid in swallowing food0.9070.82311The inside of my nose feels dry0.9020.8145I have difficulty in eating dry foods0.9020.8147I have difficulties swallowing certain foods0.8970.8053I get up at night to drink0.8900.7938The skin of my face feels dry0.8900.7929My eyes feel dry0.8730.76310My lips feel dry0.8660.7504My mouth feels dry0.8520.7266I suck sweets or lollies to relieve dry mouth0.8220.676 Factor Loadings and Communalities for the HXI Scree plot of the Exploratory Factor Analysis for the HXI Confirmatory Factor Analysis (CFA) via Principal Axis Factoring (PAF) identified a clear unidimensional structure. Based on the Kaiser criterion, only one factor yielded an eigenvalue greater than unity (initial eigenvalue = 8.834), accounting for 80.3% of the total variance. As can be seen in Fig. 2, the Scree Plot exhibited a definitive “elbow” after the first component, with subsequent components contributing negligibly to the variance (all eigenvalues < 0.47). Detailed factor loadings and communalities are presented in Table 6. All 11 items displayed robust factor loadings, ranging from 0.822 (Item 6) to 0.930 (Item 2), well exceeding the conservative threshold of 0.40. Communalities (h2) ranged from 0.676 to 0.866, suggesting that the single-factor model explains a high proportion of the variance for each individual item and confirming the structural integrity of the Hebrew translation.Table 6Validity and reliability tests of translated XI to various languagesLanguageNumber of patientsType of patientsXI score (Mean ±standard deviation)Cronbach's αICCConstruct validityFactor analysisOtherHebrew-current version102Xerostomia patientsMean ±SE:39.9±1.20.9750.78· Spearman: r= −0.862, P value <0.001.· Correlation between the standard question and USF: r=−0.765, P value <0.001.1-Factor(80.4% variance explained)"Cronbach's Alpha if Item Deleted" showed very high values (0.971–0.974), which were also similar to the original α, meaning that they were all required for the "excellent" reliability of the HXI.German[22]39Radiation-induced xerostomia44.2 ± 5.00.920.85-Confirmatory factor analysis showedgood to moderate values-Spanish[10]41Xerostomia patients40.8±10.00.890.59–0.91Pearson: r= −0.32, P value= 0.050Not provided-Portuguese[9]30Primary Sjögren's disease42.2±8.90.900.79–0.94Pearson: r= −0.781, P value= 0.010 for resting saliva flow and r=−0.757, P value= 0.010 for stimulated saliva flowNot provided-Korean [25]194Primary Sjögren's diseasemedian 37; IQR 30–440.8680.48–0.82· Spearman: r= −0.447, P value<0.001 for resting salivary flow and r=−0.515, P value<0.001 for stimulated salivary flow.· Correlation between the standard question with USF: r= 0.772, P value<0.001Not provided-Turkish [26]69Primary Sjögren's disease36.41 ± 7.670.8690.993Spearman: r= −0.611, P value<0.001 for resting salivary flow and r= −0.674, P value<0.001 for stimulated salivary flow.Not provided"Cronbach's Alpha if Item Deleted" showed similar values to the Cronbach's α value, for all 11 XI questions, meaning no question can be deleted.Greek[11]100Older greeksNot provided0.78Not providedSignificant correlations with self-perceived dryness7-item scale identified4 items were deletedFrench [23]65Older adults20.0±7.40.790.83No relation with salivary flow1-Factor(32.7% variance explained)-Persian [27]100-0.840.95---Thai [24]200Thai adults (50+)Median 21.5, IQR 16–270.875Not providedModerate to strong correlation with comparator instruments1-Factor Table 6Validity and reliability tests of translated XI to various languagesLanguageNumber of patientsType of patientsXI score (Mean ±standard deviation)Cronbach's αICCConstruct validityFactor analysisOtherHebrew-current version102Xerostomia patientsMean ±SE:39.9±1.20.9750.78· Spearman: r= −0.862, P value <0.001.· Correlation between the standard question and USF: r=−0.765, P value <0.001.1-Factor(80.4% variance explained)"Cronbach's Alpha if Item Deleted" showed very high values (0.971–0.974), which were also similar to the original α, meaning that they were all required for the "excellent" reliability of the HXI.German[22]39Radiation-induced xerostomia44.2 ± 5.00.920.85-Confirmatory factor analysis showedgood to moderate values-Spanish[10]41Xerostomia patients40.8±10.00.890.59–0.91Pearson: r= −0.32, P value= 0.050Not provided-Portuguese[9]30Primary Sjögren's disease42.2±8.90.900.79–0.94Pearson: r= −0.781, P value= 0.010 for resting saliva flow and r=−0.757, P value= 0.010 for stimulated saliva flowNot provided-Korean [25]194Primary Sjögren's diseasemedian 37; IQR 30–440.8680.48–0.82· Spearman: r= −0.447, P value<0.001 for resting salivary flow and r=−0.515, P value<0.001 for stimulated salivary flow.· Correlation between the standard question with USF: r= 0.772, P value<0.001Not provided-Turkish [26]69Primary Sjögren's disease36.41 ± 7.670.8690.993Spearman: r= −0.611, P value<0.001 for resting salivary flow and r= −0.674, P value<0.001 for stimulated salivary flow.Not provided"Cronbach's Alpha if Item Deleted" showed similar values to the Cronbach's α value, for all 11 XI questions, meaning no question can be deleted.Greek[11]100Older greeksNot provided0.78Not providedSignificant correlations with self-perceived dryness7-item scale identified4 items were deletedFrench [23]65Older adults20.0±7.40.790.83No relation with salivary flow1-Factor(32.7% variance explained)-Persian [27]100-0.840.95---Thai [24]200Thai adults (50+)Median 21.5, IQR 16–270.875Not providedModerate to strong correlation with comparator instruments1-Factor Validity and reliability tests of translated XI to various languages Hebrew- current version Mean ±SE: 39.9±1.2 · Spearman: r= −0.862, P value <0.001. · Correlation between the standard question and USF: r=−0.765, P value <0.001. 1-Factor (80.4% variance explained) German [22] Confirmatory factor analysis showed good to moderate values Spanish [10] Portuguese [9] · Spearman: r= −0.447, P value<0.001 for resting salivary flow and r=−0.515, P value<0.001 for stimulated salivary flow. · Correlation between the standard question with USF: r= 0.772, P value<0.001 Greek [11] 1-Factor (32.7% variance explained) Discussion The findings of this study confirm that the Hebrew version of the XI is a robust and reliable instrument for assessing patient-reported oral dryness. The high internal consistency as indicated by Cronbach’s alpha, coupled with successful construct and discriminant validity, align with existing international versions. Furthermore, factor analysis supports a unidimensional structure, with a high explained variance of 80.3%, reinforcing the structural integrity of the HXI for use in Hebrew-speaking populations. The findings from the pilot pre-test, in which only one patient suggested a correction, represent a feedback rate of less than 20%. This ensures that the HXI remains within the clarification guidelines established in the methodology [8]. Xerostomia has a significant impact on oral health, caries development, and QoL related to comfort, social interaction, swallowing and speech. However, objective measurement of salivary flow rate do not always correlate with the subjective xerostomia symptoms [19]. PROMs are utilized to evaluate health status of patients from their own perspective at a specific moment [4]. Therefore, they provide clinicians with an insight on the subjective general health and QoL perception of the patients, thus supporting a patient-centered approach to care [20]. Despite xerostomia being a reasonably prevalent condition [2, 21], no validated assessment exists in the Hebrew language. Linguistic validation of the Hebrew version of the XI would align with established international standards for the cross-cultural adaptation of PROMs. By documenting this validation in the international literature—joining existing adaptations such as the German [10, 22], Portuguese [9], Greek [11], French [23] and Thai [24] versions—we ensure the accessibility of the XI instrument to the global scientific and clinical communities. Such dissemination is fundamental for several reasons. First, it facilitates multicenter international trials by providing a standardized, validated instrument for diverse geographic samples. Second, it promotes methodological consistency, enabling researchers to benchmark data against global standards. Ultimately, the publication of this Hebrew version contributes to the broader psychometric evidence base, confirming robustness of the instrument across varied linguistic and cultural contexts. Table 6 provides a comparative analysis of the validity and reliability metrics obtained in this study for the HXI, juxtaposed with those from similar assessments conducted across various language translations, including German [22], Spanish [10], Portuguese [9], Korean [25], Turkish [26], Greek [11], French [23], Persian [27], and Thai [24]. To the best of our knowledge, this represents a comprehensive comparison of full-length (11 or 12-item) versions available in the English-language literature. The data on Table 6 represent specific study samples and are intended to show instrument consistency rather than representative population prevalence. This analysis explicitly distinguishes between the full-length XI used in this study and the 5-item Summated version (SXI-5) derived from the original instrument [28]. The SXI-5 has been also validated in various languages such as Chinese [29], Indonesian [30], Arabic [31], Hindi [32], Dutch [28], German (validated alongside XI) [22] and Thai [24] (validated alongside the XI). While SXI instrument offers acceptable reliability and is useful for screening or large-scale studies, they differ in item composition, scoring, and dimensional structure compared with the full XI, and therefore does not capture the full range of xerostomia symptoms assessed by the original instrument. For this reason, our study focused exclusively on the full XI, including all 11 original items and the standard xerostomia question, to ensure methodological consistency and allow direct comparison with previous full-length validations. The mean HXI score on the HXI was 39.9 ± 1.2, which is comparable to the scores obtained with the Portuguese (42.2 ± 8.9), Germen (44.2 ± 5.0), Spanish (40.8 ± 10.0), Korean (median 37; IQR 30–44), and Turkish (36.4 ± 7.7) versions. These versions also included patients with autoimmune conditions such as Sjögren’s disease. Scores obtained from the French (20.0 ± 7.4) and the Thai versions (median 21.5) that included clinical convenience samples of older adults but without autoimmune conditions were lower. The robust inverse correlation between subjective data and sialometry (r = −0.862) establishes superior convergent validity, as it confirms the capacity of the instrument to yield scores that align significantly with an independent, objective criterion—sialometry—measuring the identical physiological construct. The high correlation between subjective data and sialometry is a direct result of the clinical profile and was also found in the Portuguese (r= −0.781), Korean (r = 0.772 with the standard question), and Turkish (r= −0.611) versions. In accordance with the validation framework of the original 1999 study by Thomson et al.[6], the Hebrew HXI demonstrated excellent convergent validity (specifically concurrent validity) when tested against a single-item global standard. Our analysis revealed a progressive gradient in mean HXI scores across the four ordinal response categories of the standard dry‑mouth question (see Table 3B). In our data, all pairwise contrasts were statistically significant, and the effect size was large (η² = 0.794), indicating that the HXI discriminates well between individuals reporting different levels of xerostomia. This reinforces the utility of the HXI as a continuous, sensitive measure of xerostomia severity. The HXI demonstrated exceptional internal consistency with a Cronbach’s alpha of 0.975. This value is the highest among all compared international versions, exceeding the original Thomson et al. (1999) validation (0.84) [6] as well as the German (0.92) [22], Portuguese (0.90) [9], and Spanish (0.89) [10] adaptations. The “Cronbach’s Alpha if Item Deleted” of the HXI remained stable between 0.971 and 0.974, proving that every item in the 12-item inventory is essential for maintaining its high reliability. Such high Cronbach’s alpha values imply that the items on the XI are highly interrelated and consistently measure the underlying construct of xerostomia symptoms. This robustness in internal consistency across different language adaptations strengthens the validity of the HXI as a reliable tool for assessing xerostomia. As can be seen in Table 6, most studies did not report detailed factor analysis, which highlights the thoroughness of the validation process of the HXI. In line with the original factor analysis conducted by Thomson et al. [6], the HXI demonstrated a robust unidimensional structure. The HXI explained 80.3% of the variance, greater than the significantly lower variance explained in other full-length validations like the French version (32.7%) [23]. This high level of explained variance in the HXI underscores its strong unidimensional structure and its potential robustness in capturing the construct of xerostomia symptoms effectively. Strength and limitations The main strength of the study includes the strict protocol utilized to collect and analyze the data. A limitation of this study is the use of a clinical convenience sample from a tertiary center, which may not be fully representative of the general population and could limit the generalizability of the findings to the entire population. Furthermore, it must be noted that the average HXI scores reported in this study are specific to our study cohort, and as such like other reported XI averages in the literature, these values may not be generalizable to the broader population or different clinical subgroups. Our stimulated salivary flow results were obtained via a single stimulus rather than the continuous, interval-based stimulation described by Navazesh et al.13 This alternative approach was chosen to prioritize patient comfort and prevent irritation of the dry oral mucosa, consistent with our previously published methodology [33]. We acknowledge that this results in lower flow rates than with continuous stimulation, and comparisons should be made with studies using similar methods. Conclusions The overall findings of reliability (Cronbach’s α) and validity (construct and discriminant validity), along with the comparative data for the HXI for other languages, confirm that the HXI is a valid and reliable instrument for assessing patient-reported oral dryness in Hebrew-speaking populations. Factor analyses, explaining 80.4% of the variance, confirm its unidimensional structure. The HXI can be used to support diagnosis, treatment, and evaluation of interventions in these populations. The HXI validation enables global multicenter trials, ensures methodological consistency, and facilitates scientific benchmarking across diverse linguistic backgrounds.