Thai Modified Dental Anxiety Scale (Thai-MDAS): cross-cultural adaptation, psychometric validation, and preliminary diagnostic accuracy

Abstract

Background

Dental anxiety is a prevalent barrier to dental care worldwide. The Modified Dental Anxiety Scale (MDAS) is a widely used, brief self-report measure with established cross-cultural validity in numerous languages, yet no validated Thai-language version currently exists. This study aimed to cross-culturally adapt the MDAS into Thai (Thai-MDAS) and to evaluate the psychometric properties and diagnostic accuracy of the resulting Thai-MDAS.

Methods

The MDAS was adapted into Thai following the five-stage cross-cultural adaptation guideline, with permission from the original developer. A cross-sectional validation study was conducted among 233 Thai adults with English proficiency. Participants completed the Thai-MDAS, a Numerical Rating Scale (NRS), and the Index of Dental Anxiety and Fear (IDAF-4 C+) in its original English form. Internal consistency was assessed using Cronbach’s alpha. Test-retest reliability was evaluated after four weeks using the intraclass correlation coefficient (ICC). Convergent validity was examined through Spearman’s correlations with the NRS and IDAF-4 C+. Known-groups validity was assessed using the Mann-Whitney U and Kruskal-Wallis tests. Receiver operating characteristic (ROC) analysis determined diagnostic accuracy and explored preliminary cut-off thresholds for this population.

Results

The Thai-MDAS demonstrated good internal consistency (α = 0.88; N = 233) and satisfactory test-retest reliability (ICC = 0.79; n = 156). Strong convergent validity was observed with the NRS (r = 0.74) and IDAF-4 C+ core module (r = 0.74). Known-groups analysis confirmed significantly higher scores among symptomatic dental attenders (p = 0.02) and participants with marked dental fear (p < 0.001), with no significant gender difference (p = 0.85). The area under the ROC curve was 0.94. At the original cut-off of ≥ 19, sensitivity was only 42.9%. Candidate thresholds in the range of ≥ 13 to ≥ 15 demonstrated more balanced diagnostic performance.

Conclusions

The Thai-MDAS is the first psychometrically validated Thai-language instrument for dental anxiety screening, demonstrating good reliability, strong validity, and preliminary outstanding diagnostic accuracy. Our findings suggest that the original cut-off of ≥ 19 may lead to the under-detection of anxious individuals in Thai populations, with lower thresholds showing promising diagnostic performance that warrants formal investigation. Validation in the general population, alongside prevalence exploration and cut-off optimisation, is recommended as a priority for future research.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12903-026-09580-0.

Introduction

Dental anxiety, according to the Diagnostic and Statistical Manual of Mental Disorders 5th edition (2013), is considered as a subtype of blood-injection-injury phobia, under an umbrella of ‘specific phobia’ [1]. It is an emotional state marked by fear of the dental environment or treatment, which is perceived as a threat [2]. If heightened, it can be a major accessibility barrier to dental services as those odontophobic persons would lack of ability to seek care [3]. Because of inattention and avoidance, people with dental anxiety would search for a dentist only when symptoms arise, which mostly ends up with more invasive dental procedures, such as local anaesthetic injection, root canal treatment, or tooth extraction. These experiences, in turn, would maintain their levels of dental anxiety as a loop, known as a vicious cycle of dental fear [4]. Globally, it has been stated that the pooled prevalence of dental anxiety in the general population is approximately 15.3% [5], while some studies may report up to or over 50% prevalence in selected populations [6, 7]. The prevalence could be higher in patients requiring Special Care Dentistry (SCD), such as those living with autism spectrum disorder [8].

Understanding the development and impacts of dental anxiety requires a biopsychosocial approach [9, 10], as its consequences extend beyond oral health to affect psychological and social well-being [11]. Biologically, patients with dental anxiety tend to present with higher caries prevalence, more severe periodontal disease, poorer oral hygiene, compromised oral functions and aesthetics, and more frequent oral symptoms compared with non-anxious peers [10, 12–14]. They also exhibit avoidance behaviours, including infrequent dental attendance and longer intervals between visits, often seeking care only when symptoms become problematic [13, 15]. This pattern contributes to treatment complexity, longer treatment duration, increased costs, and a higher likelihood of invasive procedures, which may, in turn, reinforce anxiety [4, 16].

Psychologically, increased dental anxiety has been associated with impaired oral health-related quality of life (OHRQoL), particularly in the domain of psychological discomfort, and is linked to embarrassment and the presence of psychiatric disorders, such as phobias, depression, and mood disorders [17–19]. Societal consequences including difficulties in interpersonal relationships, reduced self-esteem, and lower morale may ultimately lead to social avoidance and inhibition [20]. These problems can also extend to the occupational functioning, since increased sickness absence have been associated with dental anxiety and untreated dental issues [21, 22]. Collectively, overall poor OHRQoL has been noted, meaning that dental anxiety can negatively affect oral health outcomes across all domains, including biological, psychological, and social impacts [10, 23].

Given the multidimensional background of dental anxiety, formal assessment tools are necessary to quantify intangible constructs, promote patient’s positive experiences, improve treatment outcomes, and establish a baseline for monitoring changes over time. These tools also facilitate communication between patients and clinicians, and support research purposes [24, 25]. Three main approaches to measuring dental anxiety have been proposed: direct self-report, physiological, and behavioural [26]. Among these, direct self-report measurements are most commonly recommended due to their ability to capture the subjective experience of anxiety. A number of validated questionnaires are available for clinical and research uses [27], including Corah’s Dental Anxiety Scale (CDAS), Modified Dental Anxiety Scale (MDAS), Dental Fear Survey (DFS), Dental Belief Survey, Dental Anxiety Inventory, and Index of Dental Anxiety and Fear (IDAF-4 C+). The MDAS is widely recommended as a primary screening tool with strong reliability and validity. It is brief, easy to interpret, and imposes minimal burden on both patients and clinicians [28, 29]. The MDAS comprises of five questions, assessing anxiety across common dental-related situations, including the day prior to the dental visit, sitting in the waiting area, having a tooth drilled, having teeth scaled and polished, and receiving local anaesthetic injection. Each question is rated on a five-point Likert scale ranging from not anxious (1), slightly anxious (2), fairly anxious (3), very anxious (4), and extremely anxious (5). The total score of all questions determines the level of dental anxiety, with a score of 19 or above, indicating high dental anxiety or dental phobia [29].

The cross-cultural adaptation of a health status questionnaire for use in a new country, culture, or language requires a systematic process designed to maximise semantic, idiomatic, experiential, and conceptual equivalence between source and target versions [30]. Since the original MDAS questionnaire was developed in English, its applicability in non-English speaking countries may be limited, particularly a nature of self-administered assessment. However, the MDAS has demonstrated robust cross-cultural validity and has been translated and validated in multiple languages worldwide, including Greek, Spanish, Chinese, Japanese, Italian, and Turkish, among others [31, 32]. Notably, within Asia, validated versions are available for Chinese [33], Arabic [34–36], Indian [37, 38], Nepali [39], Japanese [40], Malay [41], and Taiwanese [42] populations. Despite these extensive adaptations, a validated Thai version has not been developed yet, highlighting a gap in the availability of standardized tools for assessing dental anxiety in Thai populations.

This study aimed to develop and validate a Thai version of the MDAS (Thai-MDAS) for use in future clinical and research purposes. The objectives of this study were to: (1) cross-culturally adapt the MDAS into Thai; (2) evaluate its psychometric properties; and (3) assess its diagnostic accuracy and preliminary cut-off thresholds for identifying clinically significant dental anxiety in the Thai population. The availability of a standardized and validated instrument is expected to facilitate the routine identification of dental anxiety in clinical practice and further support the development of targeted behavioural strategies for patients with anxiety to minimise barriers for receiving dental care.

Materials and methods

Study design

This study employed a cross-sectional design aimed at the cross-cultural adaptation and validation of the Modified Dental Anxiety Scale (MDAS), originally developed by Humphris et al. (1995) for use among Thai population [29]. The adaptation process followed established international guidelines for cross-cultural adaptation of self-report measures [30]. The resulting Thai version of the MDAS (Thai-MDAS) underwent psychometric evaluation, including assessments of its reliability, validity, and diagnostic accuracy. Following online and on-site advertisement, eligible participants were recruited between August 2025 and January 2026.

Ethical considerations

Ethical approval was granted by the Institutional Review Board for the Faculties of Dentistry and Pharmacy, Mahidol University (Certificate of Approval No. MU-DT/PY-IRB 2025/043.0107). Electronic informed consent was obtained from all participants prior to their participations in the study. Permission to work on and translate the MDAS was granted by Professor Gerald Humphris, the original scale developer.

MDAS translation and cross-cultural adaptation process

The cross-cultural adaptation of the MDAS into the Thai-MDAS followed the five-stage guideline by Beaton et al. (2000) [30]:

1. Initial translation: two professional Thai-native translators fluent in English, with different profiles and backgrounds, independently produced forward translations of the original MDAS into Thai. The first was a dental professional aware of the concept of the instrument, while the second was an experienced linguist without a healthcare background;

2. Synthesis of the translations: these translations were synthesised into a single version through discussion and resolution of discrepancies. Consensus was reached among the investigator team and the two forward translators;

3. Back translation: two professional translators, blinded to the original version and without medical or dental backgrounds, independently back-translated the synthesised Thai version into English (Supplementary file 1);

4. Committee review: the investigator team and two translators reviewed all translated versions and reached consensus on a pre-final questionnaire, ensuring semantic, idiomatic, experiential, and conceptual equivalence. Discrepancies were resolved through consensus meeting;

5. Test of the prefinal version: the pre-final Thai version was pilot-tested in 33 individuals from the target population to assess clarity, comprehension, and cultural relevance. Minor linguistic adjustments were made based on participant feedback to improve the phrasing and readability of the statements. The investigator team then reviewed, adjusted, and approved the final Thai-MDAS for subsequent validation processes (Fig. 1; Supplementary File 1).

Fig. 1.

A final Thai-MDAS questionnaire

Participants

Sample size was determined using two complementary approaches. First, for estimating the prevalence of dental anxiety, a minimum of 220 participants was calculated using the formula for estimating a single proportion [43], based on an expected prevalence of 15.3% [5], 5% precision, and a 95% confidence level (95%CI). Second, a minimum of 100 participants was recommended for psychometric validation [44], with the achieved sample of 233 yielding a subject-to-item ratio of approximately 47:1, well above the recommended minimum of 10:1 [45].

Eligible participants were Thai nationals aged 18 years or older who were proficient in English, as the comparator instrument (IDAF-4 C+) was administered in its original form due to the absence of a Thai version. English proficiency requirement was defined as having an IELTS score of at least 6.0 or equivalent, prior work or study experience in English-speaking countries for at least one year, or regular use of English by working in English-speaking countries or in the international companies or organisations. English proficiency was self-assessed by participants against the above criteria. Formal English proficiency testing was not performed as part of this study. Participants who withdrew, or provided incomplete responses on key items of the Thai-MDAS or comparator instrument, were excluded from the study. In this study, no participants were excluded due to incomplete data on key items, and no imputation was performed.

Data collection procedures

Eligible participants were invited to complete a self-administered questionnaire electronically via Google Forms. The questionnaire comprised:

1. Demographic questionnaire: Age, sex, educational level, occupation, and dental attendance pattern.

2. Thai-MDAS: Thai-translated five questions assessing dental anxiety across different dental situations, each rated on a 5-point Likert scale (1 = not anxious to 5 = extremely anxious), yielding total scores ranging from 5 to 25. Based on two most commonly used cut-off scores of its original form, three groups were classified as low (5–9), moderate (10–18), and high (19–25) [31].

3. Numerical Rating Scale (NRS): An 11-point segmented rating tool used to quantify subjective dental anxiety, ranging from 0 (least anxiety) to 10 (maximum anxiety).

4. Index of Dental Anxiety and Fear (IDAF-4 C+) [46]: The original English version was used to measure dental anxiety and fear across three modules:

   1. Core module (IDAF-4 C): Eight items rated on a 5-point Likert scale (1 = disagree, 2 = agree a little, 3 = somewhat agree, 4 = moderately agree, 5 = strongly agree) assessing cognitive, behavioural, emotional, and physiological components of dental fear.
   2. Phobia module: Five dichotomous (yes/no) items assessing features associated with the diagnosis of dental phobia.
   3. Stimulus module: Ten items rated on a 5-point Likert scale (1 = not at all, 2 = a little, 3 = somewhat, 4 = moderately, 5 = very much) identifying specific stimuli that contribute to dental fear.

MDAS reliability and validity assessment

Eligible participants completed the Thai-MDAS together with a demographic questionnaire and comparative dental anxiety measures, including a Numerical Rating Scale (NRS) and the Index of Dental Anxiety and Fear (IDAF-4 C+) The questionnaire was self-administered, with investigators available to address any queries and ensure clarity and completeness of responses. For test-retest reliability assessment, all participants were invited to complete the Thai-MDAS again after a minimum of four-week. This interval was considered sufficient to minimise recall bias while maintaining relatively stable anxiety levels in the absence of intervening dental treatment [47].

Statistical analysis

All statistical analyses were performed using Stata (version 18.0; StataCorp LLC, College Station, Texas, USA). A two-tailed significance level of p < 0.05 was applied throughout. Continuous variables are reported as mean with standard deviation (SD) and median with interquartile range (IQR) for categorical variables as frequency and percentage. Normality of continuous outcome variables was assessed using the Shapiro-Wilk test, which indicated that the Thai-MDAS scores, IDAF-4 C+ scores, and NRS scores were not normally distributed (p < 0.05); therefore, non-parametric methods were used for all comparative and correlational analyses.

Internal consistency was assessed using Cronbach’s alpha coefficient, with values 0.70 considered acceptable, ≥ 0.80 good, and ≥ 0.90 excellent. Corrected item-total correlations were calculated using the Spearman’s ρ between each item and the sum of the remaining four items, and the alpha-if-item-deleted statistic was calculated for each item to ensure no item adversely affected the overall scale reliability. Test-retest reliability was assessed in the retest using the intraclass correlation coefficient (ICC: two-way mixed-effects model, absolute agreement, single measures) [48]. The Wilcoxon signed-rank test was used to assess systematic shift between occasions, and the Bland-Altman analysis was performed to graphically assess agreement between test and retest scores (mean difference ± 1.96 SD of differences).

Construct validity was assessed using Spearman’s ρ correlation between the Thai-MDAS total score and existing dental anxiety measures, including NRS, IDAF-4 C+ mean core score, IDAF-4 C+ total phobia score, and IDAF-4 C+ total stimulus score. Correlation strength was interpreted as: <0.30 weak, 0.30–0.49 moderate, 0.50–0.79 strong, ≥ 0.80 very strong [49]. Known-groups validity was examined by comparing Thai-MDAS total scores across defined groups expected to differ in dental anxiety levels. The Mann-Whitney U test was used for two-group comparisons, while the Kruskal-Wallis test was used for three or more groups. For this sample population, the preliminary discriminative ability of the Thai-MDAS was assessed using receiver operating characteristic (ROC) curve analysis with the IDAF-4 C+ marked dental fear criterion (mean core score ≥ 3) as the reference standard. The area under the ROC curve (AUC) was the primary measure of discriminative performance, interpreted as acceptable (0.7–0.8), excellent (0.8–0.9), outstanding (≥ 0.90) [50]. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the Youden Index were calculated at variable cut-off scores, derived from 2 × 2 contingency tables.

Results

A Thai version of MDAS was professionally developed and approved following the established cross-cultural adaptation guideline [30]. Forward translation was independently performed by two bilingual translators; one of whom is a healthcare professional. Two translations were then synthesised into a single version, which underwent independent back translation by other two professional translators without medical or dental background. Consensus was reached among the investigators for the pre-final version, which was then pilot-tested to assess clarity, translation accuracy, and comprehension. Minor adjustments were made, resulting in the final version of the Thai-MDAS (Fig. 1). The final Thai-MDAS and its English back-translation are provided as Supplementary File 1.

Regarding the validation process, a total of 233 participants were enrolled. The sample was predominantly female (n = 142, 60.9%), with 91 male participants (39.1%). The mean age was 32.68 years (SD = 8.20), with the majority aged 30–44 years (n = 134, 57.5%). Participants aged 18–29 years accounted for 34.8% (n = 81), while those aged 45 years and older comprised 7.7% (n = 18). Detailed demographic and clinical characteristics are presented in Table 1.

Table 1.

Sociodemographic and clinical characteristics of participants (N = 233)

Characteristics	Total (N = 233)	Low Thai-MDAS (n = 108, 46.4%)	Moderate Thai-MDAS (n = 113, 48.5%)	High Thai-MDAS (n = 12, 5.2%)
Age, years – mean (SD)	32.68 (8.20)	32.40 (8.40)	32.70 (7.96)	35.00 (8.83)
Sex, n (%)
Male	91 (39.1%)	39 (33.3%)	49 (43.4%)	3 (25.0%)
Female	142 (60.9%)	69 (63.9%)	64 (56.6%)	9 (75.0%)
Medical History, n (%)
None	187 (80.3%)	94 (87.0%)	87 (77.0%)	6 (50.0%)
Non-communicable diseases	12 (5.2%)	6 (5.6%)	5 (4.4%)	1 (8.3%)
Respiratory	17 (7.3%)	5 (4.6%)	10 (8.8%)	2 (16.7%)
Other	17 (7.3%)	3 (2.8%)	11 (9.7%)	2 (16.7%)
Smoking Status, n (%)
Non-smoker	212 (91.0%)	96 (88.9%)	104 (92.0%)	12 (100.0%)
Active smoker	12 (5.2%)	7 (6.5%)	5 (4.4%)	0 (0.0%)
Past smoker	9 (3.9%)	5 (4.6%)	4 (3.5%)	0 (0.0%)
Alcohol Consumption Frequency, n (%)
None	79 (33.9%)	39 (36.1%)	32 (28.3%)	8 (66.7%)
Social drinker	139 (59.7%)	62 (57.4%)	73 (64.6%)	4 (33.3%)
Active drinker	15 (6.4%)	7 (6.5%)	8 (7.1%)	0 (0.0%)
Dental Attendance Pattern, n (%)
Regular attender	190 (81.5%)	96 (88.9%)	87 (77.0%)	7 (58.3%)
Irregular attender	18 (7.7%)	6 (5.6%)	11 (9.7%)	1 (8.3%)
Symptomatic attender	25 (10.7%)	6 (5.6%)	15 (13.3%)	4 (33.3%)
Most Recent Dental Visit, n (%)
Routine/check-up	149 (63.9%)	75 (69.4%)	69 (61.1%)	5 (41.7%)
Symptomatic treatment	53 (22.7%)	19 (17.6%)	27 (23.9%)	7 (58.3%)
Orthodontic/cosmetic	15 (6.4%)	6 (5.6%)	9 (8.0%)	0 (0.0%)
No recent visit	16 (6.9%)	8 (7.4%)	8 (7.1%)	0 (0.0%)

The majority of participants reported no significant medical history (n = 187, 80.3%). Among those with comorbidities, respiratory conditions were most common (n = 17, 7.3%), followed by non-communicable diseases, such as cardiovascular disease, diabetes mellitus, and dyslipidaemia (n = 12, 5.2%). Most participants were non-smokers (n = 212, 91.0%) and identified as social drinkers, with alcohol consumption less than twice weekly (n = 139, 59.7%). The majority of participants were regular dental attenders (n = 190, 81.5%), with 7.7% (n = 18) attending irregularly and 10.7% (n = 25) visiting only when symptomatic. The most common reason for the most recent dental visit was routine check-up or cleaning (n = 149, 63.9%), followed by symptomatic treatment (n = 53, 22.7%).

Using the two most common cut-off scores of ten and nineteen [31], 108 participants (46.4%) were classified as having low dental anxiety (Thai-MDAS scores 5–9), 113 (48.5%) as moderate anxiety (Thai-MDAS scores 10–18), and 12 (5.2%) as high anxiety (Thai-MDAS scores 19–25). The mean Thai-MDAS total score was 10.56 (SD = 4.10), with a median of 10 (range: 5–25). All five items spanned the full responses ranging from 1 (not anxious) to 5 (extremely anxious). The item reporting the greatest anxiety was a question about local anaesthetic injection (mean = 2.82, SD = 1.15, median = 3), followed by tooth drilling (mean = 2.40, SD = 1.09, median = 2), tooth scaling and polishing (mean = 2.03, SD = 1.03, median = 2), waiting area anxiety (mean = 1.73, SD = 0.89, median = 2), and anticipatory anxiety for a dental visit on the next day (mean = 1.58, SD = 0.88, median = 1) (Fig. 2).

Fig. 2.

Mean Thai-MDAS score for each item (N = 233)

Thirty-six participants (15.5%) reported absolutely no anxiety (NRS score 0) whereas 97 participants (41.6%) with mild anxiety (NRS score 1–3), 50 participants (21.5%) on moderate anxiety (NRS score 4–6), and 50 participants (21.5%) severe anxiety (NRS score 7–10). For the IDAF-4 C+, the mean core score was 1.57, with an SD of 0.78. A total of 21 participants (9.0%) met the DSM-IV criterion for marked dental fear defined as the mean core score ≥ 3, and were classified as the high dental anxiety reference group for subsequent validity analyses.

Internal consistency of the Thai-MDAS was assessed using Cronbach’s alpha (α), resulting in alpha coefficient of 0.88 (95%CI: 0.85–0.90), indicating good internal consistency. All corrected item-total correlations exceeded the minimum acceptable threshold of 0.30, ranging from 0.54 (Q5, LA injection) to 0.78 (Q2, waiting room) (Supplementary Table 1). No item, if deleted, would substantially improve the overall alpha (alpha-if-item-deleted: 0.84–0.89). In terms of inter-item correlations, all pairwise correlations displayed positively and statistically significant Spearman’s ρ values (r) (p < 0.001). The highest correlation (r = 0.83) was observed between anticipatory anxiety (Q1) and waiting area anxiety (Q2, whereas the lowest correlation (r = 0.41) was between Q1 and Q5 (LA injection) (Supplementary Table 2).

Convergent validity was assessed by examining Spearman’s ρ correlations between the Thai-MDAS and standardised dental anxiety measures (Table 2). The Thai-MDAS demonstrated strong, statistically significant correlations with the NRS (r = 0.74; 95%CI: 0.68–0.79, p < 0.001) and the IDAF-4 C+ core module (r = 0.74; 95%CI: 0.68–0.79, p < 0.001). Moderate positive correlations were also observed with the IDAF-4 C+ stimulus module (r = 0.56; 95%CI: 0.47–0.65, p < 0.001) and phobia module (r = 0.41; 95%CI: 0.30–0.51, p < 0.001).

Table 2.

Convergent validity: spearman’s ρ correlations between Thai-MDAS and dental anxiety measures (N = 233)

Measures	Spearman ρ (95%CI)	p-value
NRS (Numeric Rating Scale)	0.74 (0.68–0.79)	< 0.001*
IDAF-4 C+ Mean Score of Core Module	0.74 (0.68–0.79)	< 0.001*
IDAF-4 C+ Total Score of Phobia Module	0.41 (0.30–0.51)	< 0.001*
IDAF-4 C+ Total Score of Stimulus Module	0.56 (0.47–0.65)	< 0.001*

For known-groups validity (Table 3; Fig. 3), no statistically significant difference (p = 0.85) in Thai-MDAS was observed between male (n = 91, mean = 10.47, SD = 3.70, median = 10, IQR: 8–12) and female participants (n = 142, mean = 10.69, SD = 4.38, median = 10, IQR: 7–13; p = 0.847). However, it differed significantly across dental attendance groups (p = 0.02). Participants who attended the dentist only when symptomatic reported the highest anxiety scores (n = 25, mean = 12.60, SD = 5.30, median = 11, IQR: 10–14), followed by irregular attenders (n = 18, mean = 11.50, SD = 4.13, median = 11, IQR: 9–14), and regular attenders had the lowest scores (n = 190, mean = 10.20, SD = 3.84, median = 9, IQR: 7–12). Furthermore, strong discrimination was demonstrated between participants with and without marked dental fear as defined by the IDAF-4 C + DSM-IV criterion. Participants meeting the criterion for marked fear (n = 21, mean = 18.38, SD = 3.97, median = 18, IQR: 17–21) scored significantly (p < 0.001) higher than those who did not (n = 212, mean = 9.78, SD = 3.21, median = 9, IQR: 7–11.5). Conversely, IDAF-4 C+ mean core scores also differed significantly across Thai-MDAS severity groups (p < 0.001).

Table 3.

Known-groups validity (N = 233)

Groups	n	Mean (SD)	Median	IQR (Q1–Q3)	p-value
Sex (Mann-Whitney U)
Male	91	10.47 (3.70)	10	8–12	0.847
Female	142	10.69 (4.38)	10	7–13
Dental Attendance Pattern (Kruskal-Wallis)
Regular attender	190	10.20 (3.84)	9	7–12	0.020*
Irregular attender	18	11.50 (4.13)	11	9–14
Symptomatic only	25	12.60 (5.30)	11	10–14
Most Recent Dental Visit (Kruskal-Wallis)
Routine/check-up	149	10.14 (3.76)	9	7–12	0.061
Symptomatic treatment	53	11.89 (5.05)	11	9–14
Orthodontic or cosmetic	15	11.47 (3.60)	12	8–14
IDAF-4 C+ Marked Fear Criterion (Mann-Whitney U)
No marked fear	212	9.78 (3.21)	9	7–11.5	< 0.001*
Marked fear	21	18.38 (3.97)	18	17–21

Fig. 3.

Known-group validity by sex (top-left), marked fear criteria (top-right), dental attendance pattern (bottom-left), and most recent dental visit (bottom-right) (N = 233)

Test-retest reliability was assessed in a subsample of participants who completed the Thai-MDAS on two occasions separated by at least four weeks (n = 156, 67.0%). To assess potential attrition bias, baseline characteristics of retest responders (n = 156) and non-responders (n = 77) were compared. No statistically significant differences were observed in age (p = 0.367), sex (p = 0.083), dental attendance pattern (p = 0.317), or the type of most recent dental visit (p = 0.655). Importantly, baseline Thai-MDAS total scores were virtually identical between responders (mean = 10.55, SD = 4.10) and non-responders (mean = 10.58, SD = 4.12; p = 0.945). No statistically significant systematic shift between test and retest total scores was confirmed (p = 0.16). Intraclass correlation coefficients (ICC) were computed using a two-way mixed-effects model with absolute agreement (Supplementary Table 3). The ICC for the total Thai-MDAS score was 0.79 (95% CI: 0.72–0.84, p < 0.001), indicating good test-retest reliability [48]. The item-level ICCs ranged from 0.70, anxiety at tooth scaled and polished, to 0.84, anticipatory anxiety. The Bland-Altman plot, available as a Supplementary Fig. 1, further confirmed acceptable agreement between test and retest administrations. The mean difference was − 0.22 points (SD = 3.28), with 95% limits of agreement ranging from − 6.66 to + 6.21. The narrow, symmetrically distributed band around zero indicated no proportional bias across the score range.

Receiver operating characteristic (ROC) analysis was performed to evaluate the ability of the Thai-MDAS to identify included participants with marked dental fear (IDAF-4 C+ mean core score ≥ 3). The area under the ROC curve (AUC) was 0.94 (95%CI: 0.87-1.00). The diagnostic performance at four candidate cut-off scores is available in Table 4. The highest Youden Index (0.73) was observed at three candidate thresholds: ≥13 (sensitivity 90.5%, specificity 82.5%), ≥ 15 (sensitivity 81.0%, specificity 91.5%), and ≥ 17 (sensitivity 76.2%, specificity 96.7%). At the original cut-off score ≥ 19, the Youden index decreased substantially to 0.42 (sensitivity 42.9%, specificity 98.6%). Negative predictive values exceeded 94% at all candidate thresholds.

Table 4.

Discriminative performance and candidate cut-off scores

Cut-off (≥)	Sensitivity (95%CI)	Specificity (95%CI)	PPV (95%CI)	NPV (95%CI)	Youden Index
13	90.5% (69.6–98.8)	82.5% (76.8–87.4)	33.9% (21.8–47.8)	98.9% (96.0-99.9)	0.73
15	81.0% (58.1–94.6)	91.5% (86.9–94.9)	48.6% (31.4–66.0)	98.0% (94.9–99.4)	0.73
17	76.2% (52.8–91.8)	96.7% (93.3–98.7)	69.6% (47.1–86.8)	97.6% (94.5–99.2)	0.73
19 (original)	42.9% (21.8–66.0)	98.6% (95.9–99.7)	75.0% (42.8–94.5)	94.6% (90.7–97.2)	0.42

Discussion

This study presents the cross-cultural adaptation and psychometric validation of the Thai-MDAS, the first validated Thai-language dental anxiety screening instrument (Fig. 1). The Thai-MDAS demonstrated good internal consistency, satisfactory test-retest reliability over a four-week interval, and strong convergent validity with both the NRS and IDAF-4 C+ core module, indicating that it measures the same construct as established instruments. Preliminary outstanding discriminative performance was also observed; however, the original UK cut-off of ≥ 19 reflected poor sensitivity in this English-proficient Thai sample, suggesting that an adapted threshold warrants further investigation in the general Thai population.

The development of Thai-MDAS rigorously followed five-stage adaptation process [30] with permission obtained from the original developer. An adequate sample size was achieved, with a requirement of language proficiency in order to obtain effective validation process with an established tool in its original language. Our study employed not only a single reference standard, but two, of established anxiety screening tools, including NRS and the IDAF-4 C+. Each tool has well-established external criteria, with a score of ≥ 7 for the NRS [27] and a mean score of ≥ 3 for the core module of IDAF-4 C+ [46]. Multiple validity approaches were performed, including convergent, known-groups, and discriminative validations, providing a comprehensive evaluation of the Thai-MDAS performance.

Regarding the choice of convergent measure, NRS was used in the present study rather than the VAS as a convergent measure of dental anxiety. The NRS was chosen because the online, self-administered format makes a true 100-mm VAS impractical due to screen size and resolution variability across devices. Systematic reviews have demonstrated that NRS and VAS produce corresponding scores, with NRS showing better compliance and ease of use [51, 52]. The strong correlation between Thai-MDAS and NRS scores supports the convergent validity of the Thai-MDAS, and the use of NRS in place of VAS is considered methodologically acceptable and unlikely to have influenced the validity findings. Selection of the IDAF-4 C + as another reference standard was based on three considerations. First, it assesses dental anxiety and fear across cognitive, behavioural, emotional, and physiological dimensions, providing a broader construct assessment than situational measures such as the MDAS or CDAS [27]. Second, it has been used as the criterion standard in a comparable MDAS validation study [42]. Third, using the MDAS as its own reference standard would introduce circularity. Furthermore, test-retest reliability with over a minimum of four-week interval displayed no significant difference in a substantial and comparable subsample of over two-thirds of all first rounders, indicating good temporal stability of the tool.

On the other hand, this study has limited generalisability as recruited participants were Thai natives with good English proficiency. The selected samples, with a mean age 33 years old, tended toward individuals who were well-educated, urbanised, or likely higher socioeconomic status, which may not represent general Thai population. Additionally, no clinical data such as dental caries or periodontal indices were collected, limiting the ability to examine correlations between the level of dental anxiety and oral health outcomes. Although the IDAF-4 C + is a well-established multidimensional measure of dental anxiety and fear, it was administered in its original English form, and no validated Thai version currently exists. Classification of marked dental fear in this bilingual sample may therefore be subject to misclassification. Furthermore, the reliance on a single reference standard for ROC analysis, rather than multiple independent criteria, should be acknowledged. Additionally, the small number of participants meeting the marked dental fear criterion (n = 21) resulted in wide confidence intervals for diagnostic accuracy estimates, further supporting the preliminary nature of these findings. With respect to test-retest reliability, every effort was made to retain participants, with 98% contacted by email and invited for the follow-up. The retest response rate of 67.0% is comparable to or exceeds those reported in similar MDAS validation studies, such as 20% in the Nepali version [39]. Comparison of baseline characteristics between retest responders and non-responders revealed no statistically significant differences, suggesting that attrition bias is unlikely to have substantially influenced the reliability findings.

The internal consistency of the Thai-MDAS (α = 0.88) falls within the range reported in several validated MDAS language versions (0.70–0.93) and is comparable to most existing translations (Table 5). Notably, it is virtually identical to the Chinese Beijing [33], Japanese [40], and Taiwanese formats [42], suggesting that the MDAS construct maintains robust internal reliability across East and Southeast Asian cultural contexts. The similarity with the Taiwanese version is particularly noteworthy, as both studies employed the IDAF-4 C + as an external criterion, indicating methodological as well as psychometric consistency.

Table 5.

Previously established translated and validated MDAS

Language (Country of study)	Year	n	Comparator(s)	Cronbach’s α
English (United Kingdom) [29]a	1995	1,392	CDAS	> 0.70
Norwegian (Norway) [53]	2000	~ 350	Self-reported Dental visit pattern	0.89
English (United Kingdom) [54]	2005	135	STAI	0.93
Turkish 1 (Türkiye) [55]	2005	294	CDAS	0.81
Turkish 2 (Türkiye) [56]	2005	557	DFS	0.91
Greek (Greece) [57]	2008	402	DFS	0.90–0.92
Spanish (United States) [58]	2008	344b	1. DFS	0.80–0.85
			2. NS
Chinese Beijing (China) [33]	2008	783	HADS	0.88–0.89
Hindi (India) [37]	2008	482	MDBS	0.78
Arabic (Jordan) [35]	2011	1,602	Dental visit experience questions	0.82
Arabic (Saudi Arabia) [34]	2014	474	VAS	> 0.90
Malay (Malaysia) [41]	2014	455	N/A	0.85
Italian (Italy) [59]	2015	230c	N/A	0.92
Kannada (India) [38]	2015	301	VAS	0.83
Nepali (Nepal) [39]	2017	150	VAS	0.78
Japanese (Japan) [40]	2020	275	DFS	0.88
Arabic (Lebanon) [36]	2021	451	VAS	0.91
Taiwanese (Chinese Taipei) [42]	2021	402	IDAF-4 C+	0.88
Hebrew (Israel) [60]	2022	553	N/A	0.93
Portuguese (Brazil) [61]	2022	486d	VAS	0.79–0.87
Swedish (Sweden) [62]	2024	246	1. VAS	0.91
			2. Dental phobia diagnosis
Thai (Thailand)	2026	233	1. IDAF-4 C+	0.88
			2. NRS

^a Original study; ^b Spanish-speaking populations in the USA; ^c Oral surgery patients only; ^d Pregnant women only

CDAS Corah’s Dental Anxiety Scale, DFS Dental Fear Survey, HADS Hospital Anxiety and Depression Scale, IDAF-4 C + Index of Dental Anxiety and Fear, MDBS Modified Dental Beliefs Scale, NRS Numerical Rating Scale, NS Needle Survey, STAI Spielberger State-Trait Anxiety Inventory, VAS Visual Analogue Scale, N/A No external comparator used

Regarding the item hierarchy in the Thai-MDAS, the strong correlation between anticipatory and waiting room anxiety items may reflect their shared pre-treatment nature, while the weaker association with the injection item suggests it captures a somewhat distinct facet of dental anxiety. Local anaesthetic injection elicited the greatest anxiety (mean = 2.82), followed by tooth drilling (2.40), scaling (2.03), waiting (1.73), and anticipatory anxiety (1.58). The pattern is largely consistent with established cross-cultural contexts. A similar ranking order was observed in the Turkish validation, where injection scored highest (2.6) and anticipatory anxiety lowest (2.0) [56]. The original MDAS also identified injection and drilling as the two most anxiety-provoking items [29]. However, in the Helsinki and Belfast populations, drilling was ranked as the highest, whereas injection did not consistently dominate [32]. Despite these differences, the consistency of injection-related anxiety as the highest-ranked item across Thai, Turkish, English, and Japanese populations suggests a near-universal stimulus hierarchy for this item, although the relative ranking and magnitude of other items appear to vary by cultural context, potentially reflecting differences in patient expectations, dental care systems, and prior treatment experiences.

The absence of a statistically significant gender difference in Thai-MDAS scores (p = 0.85) contrasts with the majority of previous MDAS studies, in which females consistently score higher [5, 56, 63]. This finding may be partly explained by the bilingual nature of the sample, which may have introduced homogeneity in anxiety expression, as English-proficient Thai adults likely share more similar educational and socioeconomic backgrounds than the general population. The nature of self-reported questionnaire may be influenced by local sociocultural response pattern, notably “Kreng-Jai” which is an emotional restraint acted by a positive reluctance to impose or express discomfort to others [64]. This cultural norm may play a role toward emotional moderation which could contribute to the considerable false negative rates of the original MDAS cut-off of ≥ 19. This remains a hypothesis and warrants further investigation through qualitative or mixed-methods approaches.

Despite these limitations, the Thai-MDAS potentially addresses a longstanding gap as no validated Thai-language instrument for assessing dental anxiety has previously been available, despite the well-documented global burden of dental anxiety and its impacts on oral health and quality of life. Although a study in visually impaired paediatric populations employed the MDAS in Thai language [65], the questionnaire was not publicly available and had never undergone formal psychometric evaluation against established criterion standards. The present study provides a validated, freely available, self-administered screening tool that can be completed in under two minutes and requires no specialist training to administer or score, supporting its practical use in both clinical and research settings.

Regarding preliminary diagnostic accuracy, lower thresholds demonstrated substantially improved diagnostic balance which is consistent with those reported in other non-English MDAS versions such as the Norwegian [53], Turkish [56], Saudi Arabic [34], and Lebanese Arabic [36] versions. These findings collectively suggest that the original threshold may not be directly transferable across cultures and that a lower cut-off is likely necessary for the Thai population. However, as noted above, the IDAF-4 C + was administered in English, and the true diagnostic performance may differ when evaluated against a culturally adapted criterion standard. Formal determination of a Thai-specific cut-off in the general population is therefore a priority for future research. Nonetheless, these preliminary findings indicate that the Thai-MDAS may offer clinicians a practical means to identify patients who could benefit from Dental Behaviour Support strategies, with the aim of promoting positive dental attitudes and interrupting the vicious cycle of dental fear [66, 67].

From a clinical and policy perspective, further validation of the Thai-MDAS in a representative Thai population may support several developments at both service and national levels. Routine dental anxiety screening could be integrated into Thai clinical practice, similar to recommendations already advocated in the UK [28]. The Thai-MDAS also enables national-scale epidemiological studies to generate prevalence data that are currently lacking but essential for workforce planning, training curricula, and resource allocation in dental anxiety management. Such screening may be particularly important in the field of Special Care Dentistry, where patients already face multiple barriers to accessing care. Early identification of anxiety can facilitate individualised support and improve clinical outcomes. Future research in the general Thai population is required to confirm the optimal cut-off and to establish nationally generalisable prevalence estimates.

Several areas warrant further investigations. The immediate priority is validation of the novel Thai-MDAS in a general Thai adult population without an English-proficiency requirement. This would allow determination of a definitive population-specific cut-off scale and the generation of nationally representative dental anxiety prevalence estimates, which are currently unavailable. Confirmatory factor analysis should also be performed to test whether the Thai-MDAS is best represented by a single-factor structure or a two-factor model distinguishing anticipatory from treatment-related anxiety [33, 68]. Future research should examine the correlations between Thai-MDAS scores and clinical oral health outcomes such as dental caries experience, periodontal status, and objectively recorded dental attendance patterns, which were not captured in the present study. Evaluating the responsiveness of the Thai-MDAS to detect changes in anxiety following behavioural or pharmacological interventions would further support its use as an outcome measure in clinical trials. In addition, implementation studies examining routine integration of the Thai-MDAS into clinical workflows are needed to assess feasibility and impact, particularly its potential to support patient-centred care by identifying and addressing barriers associated with dental anxiety.

Conclusion

The Thai-MDAS demonstrated good internal consistency (α = 0.88), satisfactory test-retest reliability (ICC = 0.79), strong convergent validity with both the NRS (r = 0.74) and IDAF-4 C+ core module (r = 0.74), and preliminary outstanding discriminative performance (AUC = 0.94), establishing it as the first psychometrically validated Thai-language dental anxiety screening instrument. Preliminary analysis suggests that the original cut-off of ≥ 19 substantially underdetects anxious individuals in Thai cultural context, with lower thresholds showing improved diagnostic balance. These findings were derived from a bilingual sample using an English-language criterion standard, and future validation in the general Thai population using a culturally adapted reference instrument is warranted to confirm the proposed cut-off and to support patient-centred dental care for those living with dental anxiety.

Abbreviations

AUC

Area under the receiver operating characteristic curve

CDAS

Corah’s Dental Anxiety Scale

CI

Confidence interval

DFS

Dental Fear Survey

DSM

Diagnostic and Statistical Manual of Mental Disorders

ICC

Intraclass correlation coefficient

IDAF-4C +

Index of Dental Anxiety and Fear

IQR

Interquartile range

MDAS

Modified Dental Anxiety Scale

NPV

Negative predictive value

NRS

Numerical Rating Scale

OHRQoL

Oral health-related quality of life

PPV

Positive predictive value

ROC

Receiver operating characteristic

SCD

Special Care Dentistry

SD

Standard deviation

Thai-MDAS

Thai version of the Modified Dental Anxiety Scale

UK

United Kingdom

USA

United States of America

VAS

Visual Analogue Scale

Authors’ contributions

Author Credit contributions: Conceptualisation: PK PH KR RK SS SO; Methodology: PK PH KR RK; Formal analysis: PK KR RK SO; Investigation: PK PH KR RK SS SO; Data curation: PK PH KR RK SS SO; Writing — original draft: PK KR RK SS SO; Writing — review and editing: PK PH KR RK SS SO; All authors read and approved the final manuscript.

Funding

This research’s publishing cost was funded by Mahidol University.

Data availability

The datasets generated and analysed during the current study are not publicly available due to ethical restrictions and participant confidentiality but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval was granted by the Institutional Review Board for the Faculties of Dentistry and Pharmacy, Mahidol University (Certificate of Approval No. MU-DT/PY-IRB 2025/043.0107). Electronic informed consent was obtained from all participants prior to their participation in the study. The study was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.