Assessment of eHealth Literacy and Its Association with Oral Health Behavior Among Outpatients of a Dental College in Riyadh, Saudi Arabia—A Cross-Sectional Study

Abstract

Background/Objectives: Digital platforms have increased access to health information. The eHEALS scale evaluates individuals’ capabilities for accessing, assessing, and assimilating health information to help them make well-informed oral health decisions. It would be interesting to examine the association with oral health behavior (OHB), as digital platforms are increasingly seen as a “super social determinant”. Hence, the present study aimed to assess eHealth literacy levels and their association with oral health behaviors among dental outpatients at a dental college in Riyadh, Saudi Arabia. Methods: A cross-sectional survey using the eHEALS questionnaire was conducted after translation into Arabic, with additional questions on oral health behaviors. The internal consistency of the translated questionnaire was good. A total of 213 patients were recruited from the dental college’s outpatient department. Chi-square, followed by multinomial regression, was used in the statistical analysis. Results: The mean total eHEALS score in the sampled population was 26.17 (±7.5) of the 213 participants, with 108 (50.7%) having good oral health behavior practices. The elderly age group (OR 2.67, p = 0.01, CI 1.25–5.68), school-level education (OR 2.82, p = 0.03, CI 1.41–5.66), and low monthly family income (OR 2.53, p = 0.01, CI 1.25–5.11) were significantly associated with inadequate eHealth literacy. Participants with good oral health behavior had significantly lower odds of being categorized into inadequate (OR 0.41, p = 0.01, CI 0.20–0.81) or problematic (OR 0.43, p = 0.01, CI 0.22–0.85) levels of eHealth literacy. Conclusions: There is a significant association between eHealth literacy and individuals’ oral health behavior practices. Age, monthly family income, and education were key predictors of eHealth literacy levels.

1. Introduction

The global oral disease burden has not decreased, despite a significant increase in policy implementation and the adoption of strategies worldwide [1,2,3]. A recent study found that the Eastern Mediterranean region has the highest prevalence of oral diseases, along with associated disability adjusted life years (DALYs) due to poor oral health [4]. The Kingdom of Saudi Arabia is an integral part of this Eastern Mediterranean region, where regional oral health surveys have consistently demonstrated a high burden of oral diseases, particularly the high prevalence of dental caries [5,6].

Chronic oral diseases are multifactorial and arise due to common influences. These influences, as per Petersen et al., can be broadly classified into two domains (Proximal and Distal). Usage of oral health care services is part of the proximal domain, which is influenced by the presence of appropriate resources in society [5,7]. A defining resource in the contemporary era is access to online health care information, which plays a critical role in shaping health literacy, health care utilization, and health-related decision making [8,9,10].

Oral health literacy (OHL) has demonstrated to enhance quality of life by improving individuals’ knowledge and capacity to make appropriate treatment decisions [11]. The World Health Organization (WHO) defines oral health literacy as the “degree to which individuals can obtain, process, understand, and use oral health information to make informed decisions and take actions to improve their oral health” [12]. OHL is increasingly being recognized as a critical determinant of good oral health [13]. Enhancing oral health literacy among populations is one of the core components of oral health promotion as outlined in the Global Oral Health Action Plan (GOHAP) 2023–2030, which aims for eight percent of the global population to have access to essential oral health services by 2030 [4].

Oral health information obtained from informal sources does not necessarily translate into oral health literacy; similarly, oral health literacy does not ensure change in oral health behavior [14]. Major barriers for OHL to transform into behavior include the social identity of individuals based on social constructs in the region, along with supportive environments and access to oral care [15,16]. Evaluation of oral health behavior should precede the outcome assessment to understand the influence of social constructs in populations that are exposed to oral health information.

The internet has become a predominant platform for acquiring information related to oral health. The General Authority of Statistics (GSTAT) in Saudi Arabia, as per a 2017 survey, reported internet penetration of up to 84% among individuals ages 12 to 65 years [17]. A more recent cross-sectional study in the region by Gowdar I.M. et.al. reported that internet access is available to 98% of its participants, and more than 70% use it for oral health information [18].

Norman and Skinner, in their groundbreaking work on the electronic health literacy scale (eHEALS), highlight two crucial issues related to health information accessed over the internet. First, much of the health information available online is unregulated, increasing the risk of exposure to low-quality, misleading, or false information. Second, the stakeholder’s ability to evaluate online health information largely relies on the six core skills: (1) traditional literacy, (2) health literacy, (3) information literacy, (4) scientific literacy, (5) media literacy, and (6) computer literacy. Together, these skills increase the likelihood of an individual’s competency and confidence in using online information, thereby increasing the likelihood of informed decision making and positive behavioral change [19].

eHEALS is currently one of the most widely used instruments for assessing electronic health literacy questionnaires, despite the development of several new tools in recent years, such as the eHealth Literacy Questionnaire (eHLQ) [20]. eHEALS is supported by a strong theoretical foundation and demonstrates good internal validity. However, evidence from Saudi Arabia remains limited. Based on a literature search, only two studies have assessed health literacy using eHEALS within the Saudi population. Hakeem F.F. et al. [21] evaluated the eHEALS levels in relation to oral outcomes (tooth loss) and brushing frequency among the outpatient population, while Alhodaib H assessed the eHEALS levels among secondary school students [22]. To our knowledge, no studies in the region have explicitly examined the association between eHEALS levels and oral health behaviors such as dental visit patterns, sugar consumption, or reasons for seeking dental care. Addressing the gap in understanding eHealth literacy and its relationship with oral health behaviors may contribute to improved oral health outcomes, inform policy development for effective health information dissemination, and ultimately, reduce the burden of dental disease in the region. Accordingly, the aim of the present study was to assess eHealth literacy levels and their association with oral health behavior among dental outpatients at a dental college in Riyadh, Saudi Arabia.

2. Materials and Methods

This study followed the STROBE (Strengthening the Reporting of Observational studies in Epidemiology) guidelines for reporting the methodology.

2.1. Study Design

The present study is descriptive and cross-sectional in nature and was conducted to analyze online oral health information-seeking attitudes along with oral health behavior, utilizing the Arabic translated eHealth literacy (eHEALS) questionnaire among out-patients of a dental college in Riyadh city, a cross-sectional study.

2.2. Study Setting

Data was collected over a period of five months from April 2025 to September 2025. Prior to collecting the data, appropriate Institutional Review Board (IRB) approval was obtained (NRR25/056/3) from King Abdullah International Medical Research Center (KAIMRC). Requisite informed consents were obtained from all the participants who were willing to be enrolled in the study.

2.3. Participants

The participants were (clinic attending, urban population) recruited from the outpatient department at a public dental college in Riyadh, Saudi Arabia, who were waiting at the reception for dental screening or dental treatment procedures. A non-probability, convenience sampling technique was used while enrolling the study participants. A total of 213 participants were included based on sample size estimation performed with G power software (version 3.1.9.4, Heinrich-Heine-Universität, Düsseldorf, Germany) using the chi-square goodness of fit sample size estimation, with an effect size of 0.3, 95% confidence interval, and a power of 90%. Also, while considering the proportion of 50.4% for online dental health information-seeking behavior based on previous literature [23]. The participants in the age group of 18 to 60 years and those who were proficient in reading and writing Arabic were included in the study. The participants completed the questionnaire while they were in the waiting area at the outpatient department.

2.4. Questionnaire Translation

A self-administered Arabic translated questionnaire of the eHealth literacy scale (eHEALS) was distributed among the participants. The eight questions on eHealth literacy given by Noram, C.D. and Skinner, H.A. [19] were used for translation into Arabic. The eHEALS has previously been translated into many vernacular languages and has shown good internal consistency [24,25]. The questionnaire was forward and back translated by two independent bilingual experts. The internal consistency of the questionnaire was good (Cronbach α = 0.87). The test–retest reliability was carried out on a pilot sample of 11 (5%) participants, and the Intra Class Correlation (ICC = 0.82) value was found to be good. The responses from the pilot study were not part of the final analysis. The authors, along with translators, arrived at a consensus for certain words after the translation steps were completed. In the first two questions, which assess a skill, “I know” is used in the original questionnaire, which, when back translated from Arabic, is “I can”; as it was related to a skill, the authors deliberated and rectified accordingly. For the third question, instead of the word “internet”, “online” was used, which is understood to be the same, hence was used as it is. The psychometric analysis of this translated questionnaire will be considered in future research by the authors.

2.5. Questionnaire Components and Scoring Criteria

The questionnaire had three major components; the first component consisted of the demographic variables with five independent variables being assessed (age, gender, marital status, monthly family income, and educational qualification). The second section consisted of eight eHEALS-related questions, which were assessed on the 5-point Likert scale (strongly agree to strongly disagree). The third component was on oral health behavior practice, which consisted of five questions (frequency of brushing, do you consume biscuits/chocolates/snacks between meals, do you consume sweet drinks or bottled juice, do you utilize dental services every year, and when do you usually go to the dentist) with closed-ended multiple choice options.

The eHEALS score (continuous scale) ranged from 8 to 40. The scores were categorized into inadequate (8–20), problematic (21–26), and sufficient (27–40) based on individuals’ eHEALS scores. This categorization has been followed by most studies on eHEALS [26,27]. Likewise, the oral health behavior questions were quantified (continuous scale) for positive and negative behavior based on response, with a minimum score of 5 and a maximum score of 14. For convenience, based on the sensitivity analysis, the oral health behavior scores were dichotomized into 5–9 (good oral health behavior) and 10–14 (poor oral health behavior).

2.6. Statistical Analysis

The data was transferred to Microsoft Excel (Microsoft Corp., New York, NY, USA) from Google Forms and later analyzed using the Statistical Package for the Social Sciences (SPSS) statistical software (VERSION 20, 2011; IBM Corp., Armonk, NY, USA). Descriptive statistics were used to describe all the independent variables based on frequency and percentage. Mean (±S.D) overall eHEALS scores and oral health behavior scores were analyzed. Pearson’s chi-square analysis was used to analyze the significant independent variables that influenced the eHEALS responses. The oral health behavior (OHB) scores were dichotomized a priori, and later, a sensitivity analysis was conducted to justify the cut-off scores. Significant variables based on chi-square analysis were further introduced stepwise for multinomial regression analysis. A p-value of <0.05 was considered significant for all analyses.

3. Results

A total of 213 participants were recruited in the present study. The participants had a mean age of 33 (±12.6) years. The majority of participants, 147 (69%), were male, and the remaining 66 (31%) were female. Among them, 78 (36.6%) had a school level (elementary, preparatory, and high school) of education, whereas 135 (63.4%) had a graduate level (under- and postgraduate) of educational exposure. A majority of 73 participants (34.1%) reported that their monthly family income was in the 10,000–20,000 Saudi Arabian Riyal (SAR) category. Only a few participants, 25 (11.7%), had a monthly family income of less than SAR 5000. The marital status of participants indicated that most, 115 (53.7%), were still bachelor/ette. Total scores based on eHEALS indicated that 49 (23%), 52 (24.4%), and 112 (52.6%) belonged to the inadequate, problematic, and sufficient categories, respectively. The total oral health behavior scores showed a similar distribution among the participants, indicating good oral health behavior 108 (50.7%) and poor oral health behavior 105 (49.3%) (

Table 1. Demographic details of the participants who responded to the questionnaire.

Demographic Variable		Total (N)	Mean	Standard Deviation (±)	Frequency (n)	Percentage (%)
Age of Participants		213	33.06	12.6	-	-
Gender	Male	213	-	-	147	69
	Female				66	31
Education Level	Elementary (up to grade 5)	213	-	-	07	3.3
	Preparatory-Secondary (grades 6 to 8)				08	3.7
	High school (grades 9 to 10)				63	29.4
	University graduate				117	55.1
	Post graduate studies				18	11.7
Individual Income Per Month (GPA)	Less than SAR 5000	213	-	-	25	11.7
	SAR 5000 to 10,000				50	23.4
	SAR 10,000 to 20,000				73	34.1
	More than SAR 20,000				66	30.8
Marital Status	Single	213	-	-	115	53.7
	Married				84	39.3
	Divorced				05	7.0
eHealth Literacy	Inadequate	213	-	-	49	23.0
	Problematic				52	24.4
	Sufficient				112	52.6
eHEALS (overall score)	-	213	26.1	7.5	-	-
Oral Health Behavior	Good	213	-	-	108	50.7
	Poor				105	49.3
Oral Health Behavior (overall score)	-	213	9.5	1.9	-	-

).

Sensitivity analyses were conducted to assess the robustness of the primary cut-off (5–9 versus 10–14); the association remained significant when using a lower threshold (5–8 versus 9–14), while it was non-significant at a higher threshold (5–10 versus 11–14). Effect sizes were small to moderate and consistent in direction across cut-offs (

Table 2. Sensitivity analysis using alternative cut-off for subjective oral behavior scores reported.

Oral Health Behavior Cutt-Off Scores	OHB Comparison	Chi-Square Value	Sig	Phi/Charmer’s V (Effect Size)
5–8 versus 9–14	Good Versus Poor	6.97	0.03 *	0.18
5–9 versus 10–14		9.49	0.00 *	0.21
5–10 versus 11–14		3.70	0.15	0.13

* Significance-p value (<0.05).

).

The eHEALS-based eight questions were analyzed with each of the independent variables. This table provides insight into the significance of age, gender, and marital status. There was a significant (p = 0.03) likelihood of those participants in the age group of 18–40 years to strongly agree 38 (24.2%) with “I know how to find helpful health resources on the internet” compared to the 41–60 year age group, who were more likely to be neutral 22 (39.3%) (

Table 3. Chi-square analysis of independent variables (age, gender, and marital status) with the eHEALS questions.

Question (eHEALS)	Age/ Gender/ Marital Status (Variable)	Strongly Agree	Agree	Neutral n (%)	Disagree	Strongly Disagree	Total	df	Chi- Square Test p Value
I know how to find helpful health resources on the internet.	18–40 years	38 (24.2)	56 (35.7)	40 (25.5)	16 (10.2)	7 (4.5)	157 (100)	4	0.03 *
	41–60 years	5 (8.9)	17 (30.4)	22 (39.3)	6 (10.7)	6 (10.7)	56 (100)
	Total	43 (20.2)	73 (34.3)	62 (29.1)	22 (29.1)	13 (6.1)	213 (100)
I know where to find helpful health resources on the internet.	18–40 years	27 (7.0)	50 (31.8)	43	26	11	157	4	0.03 *
				(27.4)	(16.6)	(7.0)	(100)
	41–60 years	5	12	25	6	8	56
		(8.9)	(21.4)	(44.6)	(10.7)	(14.3)	(100)
		32	62	68	32	19	213
	Total	(15.0)	(29.1)	(31.9)	(15.0)	(8.9)	(100)
I have the skills I need to evaluate the health resources I find on the internet.	18–40	21	58	39	27	12	157	4	0.03 *
	years	(13.4)	(36.9)	(24.8)	(17.2)	(7.6)	(100)
	41–60	5	8	25	9	9	56
	years	(8.9)	(14.3)	(44.6)	(16.1)	(16.1)	(100)
		26	66	64	36	21	213
	Total	(12.2)	(31.0)	(30.0)	(16.9)	(9.9)	(100)
I can tell high quality from low quality health resources on the internet.	18–40	28	54	37	23	15	157	4	0.04 *
	years	(17.8)	(34.4)	(23.6)	(14.6)	(9.6)	(100)
	41–60	4	13	17	11	11	56
	years	(7.1)	(23.2)	(30.4)	(19.6)	(19.6)	(100)
		32	67	54	34	26	213
	Total	(15.0)	(31.5)	(25.4)	(16.0)	(12.2)	(100)
I feel confident in using information from the internet to make health decisions.	18–40	17	54	40	26	20	157	4	0.01 *
	years	(10.8)	(34.4)	(25.5)	(16.6)	(12.7)	(100)
	41–60	6	6	27	5	12	56
	years	(10.7)	(10.7)	(48.2)	(8.9)	(21.4)	(100)
		23	60	67	31	32	213
	Total	(10.8)	(28.2)	(31.5)	(14.6)	(15.0)	(100)
I know how to find helpful health resources online.		Strongly	Agree	Neutral	Disagree	Strongly	Total	4	0.02 *
	Gender	Agree				Disagree
		10	21	21	5	9	66
	Female	(31.8)	(7.6)	(31.8)	(7.6)	(13.6)	(100)
		34	52	41	17	4	147
	Male	(23.0)	(35.1)	(27.7)	(11.5)	(2.7)	(100)
		44	73	62	22	13	213
	Total	(20.6)	(10.3)	(29)	(20.6)	(6.1)	(100)
I feel confident using information from the internet to make health decisions.	Marital Status	Strongly	Agree	Neutral	Disagree	Strongly	Total	8	0.04 *
		Agree				Disagree
	Single	14	39	30	21	11	115
		(12.2)	(33.9)	(26.1)	(18.3)	(9.6)	(100)
	Married	8	17	34	9	16	84
		(9.5)	(20.2)	(40.5)	(10.7)	(19.0)	(100)
	Divorced	1	4	3	2	5	15
		(6.7)	(26.7)	(29.0)	(13.3)	(33.3)	(100)

* Significance-p value (<0.05), df (degree of freedom), and n—frequency.

).

Likewise, the younger age group (18–40 years) responded more favorably with regard to “where to find the health resources on the internet” (p = 0.03), “possess the skills to evaluate the resource” (p = 0.03), “differentiate between low and high quality resource” (p = 0.04), and subsequently feel “confident using the resource to arrive at health decisions” (p = 0.01), compared to the elderly age group (41–60 years).

Variables such as gender and marital status significantly influenced response to the questions “I know how to find helpful health resources online” (p = 0.02) and “I feel confident using information from the internet to make health decisions” (p = 0.04), respectively (Table 3).

The education category was dichotomized into two (school-level and graduate-level education) for convenience of analysis, as there were not many respondents in the elementary and preparatory school categories. Those with graduate level of education were significantly positive (strongly agree/agree) in their response compared to participants who had school level of education for the following eHEALS questions: “how to find helpful health resources on the internet” (p = 0.00), how to use the internet to answer my health questions (p = 0.04), where to find helpful health resources on the internet (p = 0.03), how to use the health information (p = 0.00), and confidence in using the information to arrive at health decisions (p = 0.05). Five of the eight questions in eHEALS were significantly influenced by individuals’ educational level of attainment (

Table 4. Chi-square analysis of educational level as independent variable with the responses to eHEALS questions.

Question (eHEALS)	Educational Level (Variable)	Strongly Agree	Agree	Neutral n (%)	Disagree	Strongly Disagree	Total	df	Chi-Square Test p Value
I know how to find helpful health resources on the internet.	School	8	25	26	9	10	78	4	0.00 *
	Level	(10.3)	(32.1)	(33.3)	(11.5)	(57.1)	(100)
	Graduate Level	35	48	36	13	3	135
		(25.9)	(35.6)	(26.7)	(9.6)	(2.2)	(100)
	Total	43	73	62	22	13	213
		(20.2)	(34.3)	(29.1)	(10.3)	(6.1)	(100)
I know how to use the internet to answer my health questions	School	6	26	27	9	10	78	4	0.04 *
	Level	(7.7)	(33.3)	(34.6)	(11.5)	(12.8)	(100)
	Graduate Level	30	48	32	15	10	135
		(22.2)	(35.6)	(23.7)	(11.1)	(7.4)	(100)
	Total	36	74	59	24	20	46
		(16.9)	(55.6)	(27.7)	(11.3)	(9.4)	(100)
I know where to find helpful health resources on the internet.	School	5	22	26	14	11	78	4	0.03 *
	Level	(6.4)	(28.2)	(33.3)	(17.9)	(14.1)	(100)
	Graduate Level	27	40	42	18	8	135
		(20.0)	(29.6)	(31.1)	(13.3)	(5.9)	(100)
	Total	32	62	68	32	19	213
		(15.0)	(29.1)	(31.9)	(15.0)	(8.9)	(100)
I know how to use the health information I find on the internet to help me.	School	6	23	27	9	13	78	4	0.00 *
	Level	(7.7)	(29.5)	(34.6)	(11.5)	(16.7)	(100)
	Graduate Level	24	51	33	22	5	135
		(17.8)	(37.8)	(24.4)	(16.3)	(3.7)	(100)
	Total	30	74	60	31	18	213
		(14.1)	(34.7)	(28.2)	(14.6)	(8.5)	(100)
I feel confident in using information from the internet to make health decisions.	School	5	16	27	13	17	46	4	0.05 *
	Level	(6.4)	(20.5)	(34.6)	(16.7)	(21.8)	(100)
	Graduate Level	18	44	40	18	15	38
		(13.3)	(32.6)	(29.6)	(13.3)	(11.1)	(100)
	Total	23	60	67	31	32	35
		(10.8)	(28.2)	(31.5)	(14.6)	(15.0)	(100)

* Significance-p value (<0.05), df (degree of freedom), and n—frequency.

).

Similar to education, monthly family income was dichotomized into low and high; the low category constituted individuals from the less than SAR 5000 and SAR 5000 to 10,000 monthly family income category, and the high category constituted individuals with household monthly income between SAR 10,000 and 20,000 or above. Those within the higher family income category were significantly more likely to respond that they know “how to find helpful health resources on the internet” (p = 0.00), “how to use the internet to answer my health questions” (p = 0.03), “how to use the health information to help themselves” (p = 0.04), and “do possess the skills they need to evaluate the health resources” (p = 0.01) (

Table 5. Chi-square analysis of monthly family income as independent variable with the responses to eHEALS questions.

Question (eHEALS)	Monthly Family Income Level	Strongly Agree	Agree	Neutral n (%)	Disagree	Strongly Disagree	Total	df	Chi-Square Test p Value
I know how to find helpful health resources on the internet.	Low	6	23	27	13	6	75	4	0.00 *
		(8.0)	(30.7)	(36.0)	(17.3)	(8.0)	(100)
	High	37	50	35	9	7	138
		(26.8)	(36.2)	(25.4)	(6.5)	(5.1)	(100)
	Total	43	73	62	22	13	213
		(20.2)	(34.3)	(29.1)	(10.3)	(6.1)	(100)
I know how to use the internet to answer my health questions.	Low	7	21	26	11	10	75	4	0.03 *
		(9.3)	(28.0)	(24.7)	(14.7)	(13.3)	(100)
	High	29	53	33	13	10	138
		(21.0)	(38.4)	(23.9)	(9.4)	(7.2)	(100)
	Total	36	74	59	24	20	213
		(16.9)	(34.7)	(27.7)	(11.3)	(9.4)	(100)
I know how to use the health information I find on the internet to help me.	Low	6	22	30	11	6	75	4	0.04 *
		(8.0)	(29.3)	(40.0)	(14.7)	(8.0)	(100)
	High	24	52	30	20	12	138
		(17.4)	(37.7)	(21.7)	(14.5)	(8.7)	(100)
	Total	30	74	60	31	18	213
		(14.1)	(34.7)	(28.2)	(14.6)	(8.5)	(100)
I have the skills I need to evaluate the health resources I find on the internet.	Low	4	17	28	17	9	75	4	0.01 *
		(5.3)	(22.7)	(37.3)	(22.7)	(12.0)	(100)
	High	22	49	36	19	12	138
		(15.9)	(35.5)	(26.1)	(13.8)	(8.7)	(100)
	Total	26	66	64	36	21	213
		(12.2)	(31.0)	(30.0)	(16.9)	(9.9)	(100)

* Significance-p value (<0.05), df (degree of freedom), and n—frequency.

).

The overall score of oral health behavior was assessed on a scale of 5 to 14, where scores between 5 and 9 were categorized as good oral health behavior practices, and scores between 10 and 14 were categorized as poor oral health behavior practices. The oral health behavior was assessed against the categorized (inadequate, problematic, and sufficient) overall score of eHEALS. Good oral health behavior practice was significantly (p = 0.00) associated with those categorized as having sufficient levels of 68 (63.0%) of eHealth literacy, whereas higher frequency of participants who indicated poor oral health behavior also had inadequate 30 (28.6%) or problematic 31 (29.5%) levels of eHealth literacy scores (

Table 6. Chi-square analysis of oral health behavior practice with that of overall eHEALS score, as well as individual eHEALS questions.

Variable (Oral Health Behavior)		Inadequate		Problematic n (%)		Sufficient	Total	df	Chi-Square Value (p Value)
Good		19		21		68	108	2	0.00 *
		(17.6)		(19.4)		(63.0)	(100)
Poor		30		31		44	105
		(28.6)		(29.5)		(41.9)	(100)
Total		49		52		112	213
		(23.0)		(24.4)		(52.6)	(100)
Question	Variable (Oral Health Behavior)	Strongly Agree	Agree	Neutral n (%)	Disagree	Strongly Disagree	Total	df	Chi-Square value (p Value)
I know where to find helpful health resources on the internet.	Good	30	40	24	9	5	108	4	0.01 *
		(27.8)	(37.0)	(22.2)	(8.3)	(4.6)	(100)
	Poor	13	33	38	13	8	105
		(12.4)	(31.4)	(36.2)	(12.4)	(7.6)	(100)
	Total	43	73	62	22	13	213
		(20.2)	(34.3)	(29.1)	(10.3)	(6.1)	(100)
I know how to use the internet to answer my health questions.	Good	27	34	28	9	10	108	4	0.02 *
		(25.0)	(31.5)	(25.9)	(8.3)	(9.3)	(100)
	Poor	9	40	31	15	10	105
		(8.6)	(38.1)	(29.5)	(14.3)	(9.5)	(100)
	Total	36	74	59	24	20	213
		(16.9)	(34.7)	(27.7)	(11.3)	(9.4)	(100)
I know what health resources are available on the internet.	Good	21	41	23	15	8	108	4	0.00 *
		(19.4)	(38.0)	(21.3)	(13.9)	(7.4)	(100)
	Poor	8	24	42	22	9	105
		(7.6)	(22.9)	(40.0)	(21.0)	(8.6)	(100)
	Total	29	65	65	37	17	213
		(13.6)	(30.5)	(30.5)	(17.4)	(8.0)	(100)
I know where to find helpful health resources on the internet.	Good	21	40	30	9	8	108	4	0.00 *
		(19.4)	(37.0)	(27.8)	(8.3)	(7.4)	(100)
	Poor	11	22	38	23	11	105
		(10.5)	(21.0)	(36.2)	(21.9)	(10.5)	(100)
	Total	32	62	68	32	19	213
		(15.0)	(29.1)	(31.9)	(15.0)	(8.9)	(100)
I know how to use the health information I find on the internet to help me.	Good	20	46	22	16	4	108	4	0.00 *
		(18.5)	(42.6)	(20.4)	(14.8)	(3.7)	(100)
	Poor	10	28	38	15	14	105
		(9.5)	(26.7)	(36.2)	(14.3)	(13.3)	(100)
	Total	30	74	60	31	18	213
		(14.1)	(34.7)	(28.2)	(14.6)	(8.5)	(100)
I can tell high quality from low quality health resources on the internet.	Good	18	45	22	11	12	108	4	0.00 *
		(16.7)	(41.7)	(20.4)	(10.2)	(11.1)	(100)
	Poor	14	22	32	23	14	105
		(13.3)	(21.0)	(30.5)	(21.9)	(13.3)	(100)
	Total	32	67	54	34	26	213
		(15.0)	(31.5)	(25.4)	(16.0)	(12.2)	(100)
I feel confident in using information from the internet to make health decisions.	Good	17	37	25	15	14	108	4	0.01 *
		(15.7)	(34.3)	(23.1)	(13.9)	(13.0)	(100)
	Poor	6	23	42	16	18	105
		(5.7)	(21.9)	(40.0)	(15.2)	(17.1)	(100)
	Total	23	60	67	31	32	213
		(10.8)	(28.2)	(31.5)	(14.6)	(15.0)	(100)

* Significance-p value (<0.05), df (degree of freedom), and n—frequency.

).

Those with good oral health behavior were significantly more likely to strongly agree/agree with knowing “where to find helpful oral health resources on the internet” (p = 0.01), “know how to use internet to answer their health questions” (p = 0.02), “know the available resources ”(p = 0.00), know what (p = 0.00) and where (p = 0.00) helpful resources are available on the internet, “use the resource to help self” (p = 0.00), and “differentiate between quality of resource and felt confident using the information to arrive at health decisions” (p = 0.01) (Table 6).

Multicollinearity among independent variables was assessed using collinearity diagnostics (conditioned index and variance proportions). The maximum conditioned index was 12.95, suggesting no multicollinearity. Multinomial regression analysis of outcome variable of eHealth literacy levels was assessed against the variables, which were significant on chi-square analysis (to reduce multiple testing inflation). Those in the elder age group (41–60 years) were more likely to have inadequate (OR 2.67, p = 0.01, CI 1.25–5.68) and problematic (OR 2.43, p = 0.02, CI 1.15–5.14) eHealth literacy skills compared to the younger age group (18–40 years). Gender was seen to have no significant influence on the eHEALS levels. Those with school level of education were significantly (OR 2.82, p = 0.03, CI 1.41–5.66) more likely to have inadequate eHealth literacy skills compared to those with graduate-level and higher education. Monthly family income was a significant predictor; the low monthly family income (<SAR 10,000) category had higher odds of having inadequate (OR 2.53, p = 0.01, CI 1.25–5.11) and problematic (OR 2.27, p = 0.02, CI 1.13–4.53) levels of eHealth literacy skill compared to the high monthly family income category (>SAR 10,000). Finally, those with good oral health behavior were less likely to be categorized as having inadequate (OR 0.41, p = 0.01, CI 0.20–0.81) or problematic (OR 0.43, p = 0.01, CI 0.22–0.85) eHealth literacy skills (

Table 7. Assessment of independent variables with the overall eHEALS score by multinomial regression analysis.

Independent Variable		Outcome Variable (eHEALS Level)	B	df	Sig.	Exp(B)	95% Confidence Interval for Exp(B)
							Lower Bound	Upper Bound
Age	41–60 years	Inadequate	0.98	1	0.01 *	2.67	1.25	5.68
		Problematic	0.89	1	0.02 *	2.43	1.15	5.14
Reference: 18–40 years		Reference: Sufficient	0 b	-	-	-	-	-
Gender	Female	Inadequate	0.24	1	0.51	1.27	0.62	2.59
		Problematic	0.06	1	0.86	1.06	0.52	2.17
Reference: Male		Reference: Sufficient	0 b	-	-	-	-	-
Qualification	School Level	Inadequate	1.03	1	0.03 *	2.82	1.41	5.66
		Problematic	0.44	1	0.20	1.56	0.78	3.12
Reference: Graduate Level		Reference: Sufficient	0 b	-	-	-	-	-
Monthly Family Income	Low (<SAR 10,000)	Inadequate	0.92	1	0.01 *	2.53	1.25	5.11
		Problematic	0.82	1	0.02 *	2.27	1.13	4.53
Reference: High (>SAR 10,000)		Reference: Sufficient	0 b	-	-	-	-	-
Oral Health Behavior	Good	Inadequate	−0.89	1	0.01 *	0.41	0.20	0.81
		Problematic	−0.82	1	0.01 *	0.43	0.22	0.85
Reference: Poor		Reference: Sufficient	0 b	-	-	-	-	-

b—Reference Category. * p < 0.05—Significant. Exp B—Odds.

).

eHEALS questions that did not demonstrate statistically significant association with the independent variables on chi-square analysis have been presented in the Supplementary Materials. Gender and marital status were significant with only one eHEALS-related question, hence they have been omitted from the detailed tabular presentation.

4. Discussion

Oral health behavior has multiple determinants, of which social determinants are of profound importance. In the present digital era, access to the internet and digital literacy have been labelled as a “super social determinant” [28]. Most health information, support, and services are increasingly available exclusively online. The present study found that the majority of participants had sufficient levels of health literacy, as defined by the eHEALS total score categorization. This finding corroborates two studies conducted among the Arabic population, one by Bergman et al. [29], who demonstrated a mean eHealth literacy of 28.1 out of 40 in their surveyed population, whereas a nationwide population survey of eHealth literacy levels and associated factors by Al-Ruzzieh MA et al. [8] had a mean eHealth literacy of 28.9 out of 40.

A review article on eHealth literacy by Suyeon Ban et al. suggests the influence of urban or metropolitan settings on the levels of overall digital health literacy [30]. While the mean eHEALS score observed in this population was lower than [8,31], similar to [32], and higher than [33] studies of a similar nature.

All the independent variables in the study had a certain degree of significant influence in determining the individual’s approach towards online health information. Age of the participants in the present study indicated having a critical role in determining their confidence in accessing, assimilating, differentiating (low- vs. high-quality resources), and finally utilizing online information in making informed decisions. A similar finding has been reported in a recent study by Cheng Yuan et al., who surveyed the eHealth literacy levels among the middle-aged and elderly population in Shanghai City, China [34]. This may be influenced by the fact that the younger generation is better adept at navigating complex online interaction and has better awareness of online resources for seeking health information through applications (apps), social media, and websites.

Level of education had the highest impact on overall eHealth literacy levels in the present population. Diverse populations have been analyzed over the years using the eHEALS or similar scales [35]. Education is one of the two predictor variables (the other being age) that have consistently influenced overall eHealth literacy scores, irrespective of the population being assessed [36]. A review by Ronald W. Berkowsky et al. on factors predicting eHealth literacy in California, USA, validates the role of educational background, with lower levels of education leading to insufficient skills to access information through digital platforms [37]. On similar lines, a cross-sectional study by Robin Milne et al. [38] on lung cancer survivors and a study by Abdullah Alhewiti et al. [39] on dental outpatients reported that higher education was associated with better eHealth literacy levels in their respective study populations.

Socioeconomic status (SES), as measured by income or wages, has been shown to be another key predictor of eHealth literacy [40]. In the present study, those with low monthly family income were more likely to belong in the category of insufficient and problematic levels of eHEALS overall scores. This finding of our study on the role of monthly income is consistent with a study by Kristjánsdóttir Ó et al., who explored the role of socioeconomic status in terms of monthly wages. The study was conducted on parents of children requiring pediatric surgery in Sweden [41]. The finding is further strengthened by observations in a systematic review by Amy Chesser et al., who explain eHealth literacy through the concept of the “digital divide” among underserved and low-income populations. The digital divide no longer restricts its definition to access and availability of the internet but encompasses the varying patterns of access and online skills [42]. The influence of socioeconomic status on oral health behavior was evident in the present study. Individuals from families with a monthly income of less than SAR 5000 (USD 1333) were more likely to have poor oral health behavior compared to individuals with a family income of more than SAR 10,000 (USD 2666). SES as a variable in oral health behavior is well established in the literature.

Interestingly, not many studies have assessed oral health behavior along with eHealth literacy scores to ascertain the association between them. A study by Hakeem F.F. et al. in Saudi Arabia assessed only brushing frequency as part of oral health behavior and primarily focused on oral health outcomes [21]. Studies have looked for an association between eHealth literacy and oral health outcomes [21,43]. Our study tries to highlight the missing link in the chain of association between eHealth literacy and oral health outcomes by assessing the population’s oral health behaviors. As information does not always translate into behavior, those with a strong internal locus of control exhibit key skills that may be common to good oral health behavior, as well as knowledge of how to navigate digital platforms to find reliable oral health information. In the present study, good oral health behavior was associated with sufficient eHealth literacy scores. In a survey by Purcell, Donrie J. et al., good eHealth literacy skills, along with an appropriate internal locus of control, have been shown to mediate optimal oral health behavior and outcomes [44]. On similar lines, a study by de Oliveira Collet, Giulia et al. evaluated the psychometric properties of the eHEALS scale on Serbian adolescent children as the target population. The study suggests “higher eHealth literacy is linked with enhanced self-management which further is associated with improved oral health behavior” [45].

Oral health information is obtained from various platforms, with a growing number of individuals relying on Artificial Intelligence (AI) and large language models (LLMs) to answer their oral health queries and acquire information. AI-driven tools can simplify complex oral health terminology (jargon), which could benefit individuals with insufficient or problematic eHealth literacy. Within this evolving digital ecosystem, AI-enabled systems function as interactive information intermediaries that summarize and personalize oral health information. With these developments, parallelly there is a need to integrate assessment of eHealth literacy levels among the high-risk population, while accounting for the key determinants of the digital divide. Those at high risk of being left out of access to relevant preventive and therapeutic oral health information are the elderly, individuals from low-income households, individuals with lower levels of educational attainment, and those exhibiting poor oral health behavioral practices, as determined in this study. A targeted approach to these groups could help them achieve optimal oral health behaviors.

There are several limitations to the present study. The study has been conducted in a dental hospital setting, which may influence the eHEALS scores due to Berksonian bias. The study is cross-sectional and questionnaire-based, which may inherently introduce recall bias. The study has been conducted in an urban population, and the study findings are from participants attending the dental clinic, limiting its generalizability, as urban residents have greater exposure, access, and penetrability to eHealth information. Although the multinomial regression analysis was used to reduce the effects of multiple testing, the exploratory nature of the study should be interpreted cautiously, given the risk of type I errors. The oral health behavior practice assessment was subjective (a proxy index), without assessing oral hygiene outcomes using indices. Though we assessed a few aspects of oral health behavior, it was not exhaustive in itself.

5. Conclusions

The present study diligently identified the sociodemographic variables associated with sufficient and inadequate levels of eHealth literacy in the population of interest. The study also found that eHealth literacy is associated with individuals’ oral health behavior practices. There is a need for multicentric studies involving populations (with community-based sampling) from urban and rural regions to further understand the association between eHealth literacy and oral health behavior. Insights from the studies can be used to identify and support individuals lacking the skills to use eHealth resources to achieve optimal oral health.