Exploring the Association Between Oral Health Literacy and Oral Health‐Related Quality of Life Among Young Women in the Eastern Province of the Kingdom of Saudi Arabia: A Cross‐Sectional Study

ABSTRACT

Objectives

One approach for improving oral health is focusing on oral health literacy (OHL). In this study, we aimed to investigate the association between OHL and oral health‐related quality of life (OHRQoL) and to identify relevant OHL intervention targets that can be used to improve oral health.

Methods

We performed a cross‐sectional questionnaire study among 1889 10th grade female students in the Eastern Province of the Kingdom of Saudi Arabia (KSA). We then assessed OHL using the Arabic version of the short version of the Health Literacy in Dentistry scale (A‐HeLD‐14). OHRQoL was assessed using the 14‐item oral health impact profile (OHIP‐14). Information on socio‐demographics and oral health‐related behaviours (e.g., toothbrushing, sugary diet consumption and dental attendance) was also collected. The relationship between OHL and OHRQoL was investigated using binary logistic regression adjusted for sociodemographic information and oral health‐related behaviours.

Results

In both the adjusted and unadjusted models, poor OHL was significantly associated with OHRQoL problems (odds ration [OR] = 2.90, p < 0.001). Having any OHL problem was significantly associated with having an OHRQoL problem (p < 0.001). The ORs were highest for the domains of financial barriers (OR = 2.49, p < 0.001) and receptivity (OR = 2.16, p < 0.001).

Conclusion

Interventions targeted at individuals with low OHL are crucial to improving OHRQoL in the population. The most relevant interventions concern (a) dedicating time and attention to oral health and (b) financial barriers.

Abbreviations

A‐HeLD‐14

The Arabic version of the short version of Health Literacy in Dentistry scale

CI

confidence interval

HeLD

health literacy in dentistry

ISCED

international standard classification of education

KSA

the Kingdom of Saudi Arabia

OHL

oral health literacy

OHRQoL

oral health‐related quality of life

OR

odds ratio

RCG

random choice generator

REALD

rapid estimate of adult literacy in dentistry

1. Introduction

Oral health problems represent a major neglected health challenge worldwide [1]. Despite the high economic growth and abundant resources in the Kingdom of Saudi Arabia (KSA), oral diseases are considered a large burden, with dental caries and periodontal disease being the most widespread oral diseases in the country [2]. In the KSA, there is universal access to public dental care although it is undermined by long waiting times. Private dental services are available but require out‐of‐pocket payments or private health insurance, making access difficult for vulnerable populations. Poor oral health can affect quality of life by causing pain, discomfort, and handicaps [3]. This reality has been conceptualised through the Oral Health‐Related Quality of Life (OHRQoL), a multidimensional construct that incorporates individuals' perceptions of oral health including psychological, functional, and social implications [4] OHRQoL is considered an important aspect of assessing oral health [5], and measuring OHRQoL in epidemiological studies can help determine the treatment needs of a population [6].

Oral diseases are mostly preventable. This suggests a focus on oral health behaviours and the concept of oral health literacy (OHL), which is “the degree to which individuals have the capacity to obtain, process, and understand the basic health information and services needed to make appropriate oral health decisions.” [7] OHL has been found to influence oral health; specifically, inadequate OHL contributes to disparities in oral health [8], increased need for emergency dental treatment, and increased risk for dental caries and periodontal diseases [9, 10]. Conversely, adequate OHL plays a role in enhancing oral health and diminishing oral health inequalities [11].

Investigating the association between OHL and OHRQoL makes it possible to determine the relevance of OHL when aiming to prevent oral problems. If the OHL measure is sensitive and includes different dimensions of OHL, it can facilitate the identification of important intervention targets for use in oral health programmes within a population. Studies investigating the association between OHL and OHRQoL have been contradictory [12, 13, 14]. A single study was performed in the KSA and found a significant association between OHL and OHRQoL in one subgroup of their population, namely, undergraduate university students [13]. Unfortunately, this study, and most of the other studies, assessed OHL using the Rapid Estimate of Adult Literacy in Dentistry (REALD‐30). The REALD is a word recognition test that measures participants' abilities to read a set of dental terms [15], but the test has been criticised for not directly assessing OHL [16]. Further, the REALD does not contain any dimensions of functional OHL, so its results do not enable the identification of intervention targets. This calls for studies that are aimed at identifying OHL targets. The Health Literacy in Dentistry (HeLD) is a relevant tool for such studies because it contains information on several dimensions of functional OHL [17]. Moreover, the HeLD is regarded as an interactive OHL measure because it collects information on capacity and skills regarding tasks related to oral health [18].

In the KSA, mothers are seen as the foundation of the family and thus significantly impact the family's health behaviours [19]. Research has shown a strong link between maternal OHL and the oral health status of their children [20, 21, 22]. Therefore, focusing on women's OHL is crucial for implementing effective programmes to improve oral health. Furthermore, adolescence is a particularly relevant period for focusing on OHL because this time is critical for the development of health behaviours [23]. In this study, we aimed (a) to investigate the association between OHL and OHRQoL among female adolescents in the Eastern Province of the KSA and (b) to identify relevant OHL intervention targets for the population. This could assist healthcare professionals, researchers, and decision‐makers to better understand the role of OHL in oral health and decide on relevant strategies based on the results. Effective OHL intervention programmes could, in turn, improve the population's oral health and quality of life.

2. Materials and Methods

2.1. Design and Setting

This study was part of a larger cross‐sectional study that assessed oral health and behaviours of 10th grade female students in the Eastern Province of the KSA using questionnaires. Data were collected over a 9‐month period from January to October 2022.

The Eastern Province is the largest province in the country and is divided into three main governorates: the main Eastern area; Al‐Ahsa; and Hafar al‐Batin. These governorates formed the clusters for the current study. The main Eastern area is the largest governorate and has a population of 5 148 598 [24], followed by Al‐Ahsa with a population of 1.3 million and Hafar al‐Batin with a population of 365 000. Each governorate has rural areas (villages and hijrah) and urban areas (cities), which were considered the strata in this study. We used the stratified Random Sampling Technique to select public high schools in the province. The Random Choice Generator [25] facilitated the random selection of the schools. A total of 18 high schools were invited to take part in the study, of which nine were from the main Eastern area, six were from Al‐Ahsa and three were from Hafar al‐Batin (Figure 1).

FIGURE 1.

The distribution of schools in clusters.

2.2. Ethical Approval and Consent to Participate

The participants' data were anonymised and kept confidential. The research protocol was approved by the Institutional Review Board at King Saud University Medical City in the KSA prior to data collection (approval letter reference number 21/0309/IRB). Informed consent was obtained from all parents of the participants prior to inclusion in the study.

2.3. Questionnaire and Variables

The variables used and analysed in this study were OHRQoL, which was considered the outcome variable; OHL, which was the main explanatory variable; and socio‐demographic information and oral health‐related behaviours, which were the moderator variables. Participants complete the questionnaire electronically via Google Forms software, and the principal of each school was given a link to the questionnaire, which was distributed to the students through each school's official portal. An individual code was assigned to each study participant that matched her code in the questionnaire and informed consent form (see Appendix S1 for the Arabic version of the questionnaire).

2.3.1. Oral Health‐Related Quality of Life

Oral health‐related quality of life was measured using the Arabic version of the 14‐item Oral Health Impact Profile (OHIP‐14) [26]. This measure is divided into seven dimensions: functional limitation, physical pain, psychological discomfort, physical disability, psychological disability, social disability and handicap [27]. Each dimension contains two items that are rated in frequency on a 5‐point Likert scale as follows: never (0), hardly ever (1), occasionally (2), fairly often (3) and very often (4), with lower scores indicating better quality of life [13].

To pinpoint the participants who had oral health problems, we first dichotomised the responses to each item as (a) not a problem (frequencies of 0, 1, and 2) or (b) a problem (frequencies of 3 and 4), resulting in zero to14 problems for each participant [28]. This score was then dichotomised and each participant was categorised as either having no problems with their OHRQoL (zero problems) or problems with their OHRQoL (one to 14 problems).

2.3.2. Oral Health Literacy

OHL was measured using the Arabic version of the HeLD‐14 (A‐HeLD‐14) questionnaire [29]. The HeLD‐14 comprises the following seven conceptual domains: communication, access, receptivity, understanding, utilisation, support and economic barriers [17]. Each HeLD‐14 domain contains a set of items (questions), and each item is rated on a 5‐point Likert scale with a score ranging from 0 to 4 ("unable to do" to "without any difficulty", respectively). The unweighted summation of the item scores is considered the total HeLD‐14 score [30], which ranges from 0 to 56. A lower score indicates a poorer level of OHL. Based on their total HeLD‐14 scores, the participants were categorised into three levels of OHL: low, medium, and high. These levels were set based on the values of the quartile limits of the total HeLD‐14 score. Initially, we categorised the participants into quartiles. Subsequently, we reorganised them into three groups: the top two quartiles indicating good OHL, the second lowest quartile indicating medium OHL, and the lowest quartile representing poor OHL. This adjustment was made to align the groups more closely with the distribution of A‐HeLD‐14 scores in the population, which tend to be high in general. Moreover, this method proved to be more sensitive in identifying participants with poor OHL, who are the primary focus group for intervention.

Oral health literacy problems were identified by dichotomising the responses to each item of the HeLD‐14 based on whether an item was considered a problem (scores of 0‐3) or not a problem (score of 4). Analysis was done based on the domains, and a domain was registered as a problem if at least one of the two items in a domain was registered as a problem.

2.3.3. Oral Health‐Related Behaviours

We included the following oral health‐related behaviours: frequency of toothbrushing, sugary diet consumption and regularity of dental attendance. The original answer choices for frequency of toothbrushing were: never, rarely, several times a month, several times a week, once a day usually in the morning, once a day usually before bed and several times a day. For analytical purposes, the frequency of toothbrushing was designated as frequent if toothbrushing was performed once a day or several times a day and infrequent if toothbrushing was performed several times a week or month, rarely, or never. Sugary diets concerned the frequency of consumption of biscuits and cake, soda and energy drinks, and chocolate or other candy. The answer choices for sugary diet consumption were several times daily, every day, several times a week, once weekly, several times a month and never. Sugary diet consumption was designated as frequent if the dietary item were consumed several times daily, every day or several times a week and infrequent if the dietary items were consumed once weekly, several times a month or never. The selected three sugary diet items were unified as one variable. Consumption of a sugary diet was designated as frequent if at least one of the dietary items was consumed frequently and infrequent when all three dietary items were consumed infrequently. The answer choices for the regularity of dental attendance were regularly, approximately once a year; regularly, approximately twice a year or even more often; when having a toothache or other acute problems; and do not visit the dentist. Dental attendance was categorised as regular if the study participants visited the dental clinic once or more a year and irregular if the reason for the dental visit was a toothache or some other acute problem or if the participant never visited a dental clinic.

2.3.4. Socio‐Demographic Information

The socio‐demographic information included the strata (city, village, hijrah), the level of parents' education and parents' occupations. The original answer categories for father's and mother's education level were as follows: no formal schooling, primary school, secondary school, high school, university, postgraduate and (specified) vocational training. For the purpose of statistical analysis, the categories were trichotomised into three levels based on the International Standard Classification of Education ‐2011‐ (ISCED): [31] high (university and postgraduate studies), medium (high school and vocational training) and low (no schooling, primary school, and secondary school). The father's and mother's education levels were unified into one variable comprising the following three levels: low (both parents have a low level of education, or one parent has a medium level of education while the other parent has a low level of education); medium (both parents have a medium level of education, or one parent has a high level of education while the other parent has a low level of education) and high (both parents have a high level of education, or one parent has a high level of education while the other parent has a medium level of education) [32].

The original answer categories for the father's and mother's occupation were as follows: not employed, employed in the public sector, employed in the private sector and retired. The original categories for each parent were trichotomised into the following: not employed, employed and retired. The father's and mother's combined occupation status was designated as follows: employed (both parents are employed, or one parent is employed), not employed (both parents are not employed) and retired (both parents are retired, or one parent is retired while the other parent is not employed) [25].

2.4. Statistical Analysis

We analysed the data using the IBM SPSS software version 28, and a statistical significance level of 0.05 was applied. We used descriptive statistics to determine the frequencies and distributions for the selected variables. We also used binary logistic regressions to determine the relationship between OHL and OHRQoL, with problem/no problem as the outcome. The models in the binary logistic regressions were made with OHL as levels based on the total score of HeLD‐14 (low, medium, high) and with individual domains in separate models. We first created unadjusted models with OHL as the sole explanatory variable, and we then adjusted those modeles with socio‐demographic and oral health‐related variables as co‐variates. Odds ratios (ORs) and confidence intervals (CLs) were extracted from all the models.

3. Results

3.1. Participants

A total of 1922 female students were invited to participate in this study. After the exclusion of some potential participants, 1889 (98.3%) were included in the analysis (Figure 1). Among the 33 excluded students, the majority were excluded because they transferred to other schools during the data collection period or because of missing questionnaires as a result of a technical issue (11 and 17 students, respectively). One participant did not respond to the informed consent, one gave a negative informed consent and three students refused to take part in the study.

Cities were the predominant residential living area for the study participants, and most participants' parents were employed and had a medium education level (Table 1). The most common behaviours of the participants were frequent toothbrushing (78.2%) along with a high intake of sugary diets (89.4%) and irregular dental visits (70.9%; Table 1).

TABLE 1.

Distribution of participants according to oral health‐related behaviours and socio‐demographic information (n = 1889).

Variable	Categories	n (%)
Toothbrushing	Frequent	1477 (78.2)
	Infrequent	412 (21.8)
Sugary diets	Frequent	1689 (89.4)
	Infrequent	200 (10.6)
Dental attendance	Regular	550 (29.1)
	Irregular	1339 (70.9)
Place of residence (stratum)	City	1349 (71.4)
	Village	401 (21.2)
	Hijrah	139 (7.4)
Parents' level of education	Low	683 (36.2)
	Medium	934 (49.4)
	High	272 (14.4)
Parents' occupation	Not employed	152 (8)
	Employed	1377 (72.9)
	Retired	360 (19.1)

3.2. Oral Health Literacy

The levels of OHL were as follows: 0–42 (low OHL), 43–50 (medium OHL) and 51–56 (high OHL). The mean HeLD‐14 score was 45.83, and the most frequently reported problems concerned the communication and receptivity domains followed by the utilisation and financial barriers domains (Table 2).

TABLE 2.

Distribution of participants according to oral health literacy (n = 1889).

OHL	Level	n (%)
A‐HeLD‐14 total score	Low	516 (27.3)
	Medium	477 (25.3)
	High	896 (47.4)
Domain	Problem/no problem
Communication	Problem	892 (47.2)
	No problem	997 (52.8)
Access	Problem	804 (42.6)
	No problem	1085 (57.4)
Receptivity	Problem	888 (47)
	No problem	1001 (53)
Utilisation	Problem	838 (44.4)
	No problem	1051 (55.6)
Understanding	Problem	623 (33)
	No problem	1266 (67)
Support	Problem	766 (40.6)
	No problem	1123 (59.4)
Financial barriers	Problem	835 (44.2)
	No problem	1054 (55.8)

3.3. Oral Health‐Related Quality of Life

The mean OHIP‐14 score was 12.45, and the participants reported few problems overall (Table 3). The items most frequently reported as problems were discomfort in eating (18.6%) and embarrassment (16.6%). The items least frequently reported as problems were feelings of worsening of taste and lack of satisfaction regarding diet (8.3% and 8.5% respectively).

TABLE 3.

Distribution of participants according to oral health‐related quality of life (n = 1889).

Dimension	Item	Problem/no problem	n (%)
Functional limitation	Pronunciation	Problem	165 (8.7)
		No problem	1724 (91.3)
	Sense of taste	Problem	156 (8.3)
		No problem	1733 (91.7)
Physical pain	Oral pain	Problem	261 (13.8)
		No problem	1628 (86.2)
	Uncomfortable with eating	Problem	352 (18.6)
		No problem	1537 (81.4)
Psychological discomfort	Worry	Problem	166 (8.8)
		No problem	1723 (91.2)
	Tenseness	Problem	256 (13.6)
		No problem	1633 (86.4)
Physical disability	Unsatisfactory diet	Problem	160 (8.5)
		No problem	1729 (91.5)
	Meal interruption	Problem	289 (15.3)
		No problem	1600 (84.7)
Psychological disability	Difficulty relaxing	Problem	289 (15.3)
		No problem	1600 (84.7)
	Embarrassment	Problem	313 (16.6)
		No problem	1576 (83.4)
Social disability	Irritable with others	Problem	265 (14.0)
		No problem	1624 (86.0)
	Difficulties doing usual job	Problem	191 (10.1)
		No problem	1698 (89.9)
Handicap	Less satisfaction towards life	Problem	184 (9.7)
		No problem	1705 (90.3)
	Inability to function	Problem	204 (10.8)
		No problem	1685 (89.2)

3.4. Association Between OHL and OHRQoL

All the binary logistic regression models showed good fit to the models based on the Omnibus Tests of Model Coefficients and the Hosmer‐Lemeshow test.

The binary logistic regression overall indicated that poor OHL was significantly associated with having an OHRQoL problem in both the unadjusted and adjusted models (Table 4).

TABLE 4.

Binary logistic regression to predict the oral health‐related quality of life from the oral health literacy (n = 1889).

Explanatory variable	Oral health‐related quality of life (OHIP‐14)
	Unadjusted				Adjusted a
	OR	CI		p	OR	CI		p
OHL (HeLD‐14)
High	Ref.	—			Ref.	—		< 0.001
Medium	2.59	2.04	3.28	< 0.001	2.43	1.91	3.09	< 0.001
Low	3.33	2.64	4.19	< 0.001	2.90	2.26	3.72	< 0.001
Communication
No problem	Ref.	—			Ref.	—
Problem	1.92	1.59	2.32	< 0.001	1.75	1.44	2.13	< 0.001
Access
No problem	Ref.	—			Ref.	—
Problem	1.84	1.52	2.22	< 0.001	1.59	1.31	1.94	< 0.001
Receptivity
No problem	Ref.	—			Ref.	—
Problem	2.48	2.05	3.01	< 0.001	2.16	1.76	2.64	< 0.001
Utilisation
No problem	Ref.	—			Ref.	—
Problem	1.54	1.28	1.86	< 0.001	1.38	1.13	1.67	< 0.001
Understanding
No problem	Ref.	—			Ref.	—
Problem	2.17	1.78	2.64	< 0.001	1.88	1.53	2.32	< 0.001
Support
No problem	Ref.	—			Ref.	—
Problem	2.11	1.74	2.55	< 0.001	1.90	1.56	2.32	< 0.001
Financial barriers
No problem	Ref.	—			Ref.	—
Problem	2.76	2.28	3.34	< 0.001	2.49	2.03	3.05	< 0.001

^a Adjusted for stratum, parents' education level, and parents' occupation, oral health‐related behaviours (frequency of toothbrushing, regularity of dental attendance and consumption of sugary diets).

The ORs for both the adjusted and unadjusted models were significant and rather similar, but all were relatively higher in the unadjusted models. In the adjusted models, the OR for having an OHRQoL problem was highest in the low‐level OHL group (OR = 2.90, CI = 2.26–3.72) even though the OR in the medium‐level OHL group was also high (OR = 2.43, CI = 1.91–3.09). Having a problem in the OHL domains was significantly associated with having an OHRQoL problem (Table 4). The ORs were highest for the financial barriers domain (OR = 2.49, CI = 2.03–3.05) followed by the receptivity domain (OR = 2.16, CI = 1.76–2.64). The domain with the lowest OR was utilisation (OR = 1.38 and CI = 1.13–1.67).

4. Discussion

In this study, we investigated the relationship between OHL and OHRQoL among young women in the KSA. The most obvious finding to emerge from the study was that poor OHL was significantly associated with OHRQoL problems. The OHL problems most related to having OHRQoL problems were financial barriers and receptivity.

The current study confirmed the significant association between OHL and OHRQoL found in other studies exploring the relationship between adolescent and adult populations [33, 34, 35]. The agreement with these studies is not surprising, especially since two of the previous studies were also conducted in a female population. The OHIP‐14 was the most common instrument used in other studies to evaluate OHRQoL, but only one study used the HeLD‐14 as a measure for OHL. This makes the comparison between these studies and ours somewhat difficult. In contrast to our study, a Turkish study performed in young people between 18 and 30 years of age found no correlation between OHL and OHRQoL [12]. That study used the REALD‐30 to assess OHL, which makes a comparison to our study difficult. Moreover, that study did not find a significant association between OHL and oral health status, which contradicts the findings of multiple other studies [36, 37, 38].

The study that is most relevant to compare with our study, due to its setting and study population, is a prospective cross‐sectional study conducted in the Southern Province of the KSA, which assessed the association between OHL and OHRQoL among 394 undergraduate students (82.3% of whom were females). The study reported that the association was significant but only among those enrolled in health‐related colleges [13]. OHL was assessed using an interview‐based REALD‐30 instrument, and the mean of its overall score indicated low OHL for the study participants, which calls into question the validity of the measure, when taking the study population into account.

Unfortunately, we lack studies investigating the association among specific aspects of OHL and oral health and OHRQoL. Therefore, it is difficult to discuss our findings regarding the HeLD‐14 items based on the current literature. One study, though, found a significant association between OHRQoL and all the HeLD domains [39], which is consistent with our findings. With regard to our findings, it seems reasonable that financial barriers would be associated with poor OHRQoL. Other studies also have found that low income [40] and cost barriers [41] are related to poor OHRQoL. It also makes sense that not being able to pay attention to one's oral health care needs would be associated with poor OHRQoL. Inattention to oral health can lead to more oral disease and thus have a higher impact on OHRQoL.

OHL interventions aimed at improving OHL or accommodating people with low OHL are worth considering in order to improve OHRQoL and thus limit the oral health care needs in the population. Although the need for general OHL interventions seems obvious, focusing on reducing financial barriers and helping people to make time for and to pay attention to oral health appear to be important places to start and could result in effective oral health improvement programmes. If one considers the most affected dimensions of OHIP‐14 in our study population, it seems that such interventions could have a positive effect on eating in particular, as well as on pain and embarrassment. These were also the most frequent problems found in other studies that assessed OHRQoL in young adults in the KSA [13, 42].

Our large sample size that was randomly selected from both urban and rural areas of the Eastern Province of the KSA and thus represent a spectrum of socio‐economic characteristics adds to the generalisability of the study's findings. In this regard, the sample reflects the diversity in socio‐economic characteristics within the Eastern Province. However, because we did not include private schools, we may have missed some representativity of the population with the highest socio‐economic status. Because the socio‐economic profile of the Eastern Province does not differ significantly from other regions of the KSA, the transferability and generalisability of the study findings across the KSA are likely feasible. Another strength is the robust analyses, resulting from the high response rate. Sampling bias was limited because the data were collected from a school setting and not in the usual health‐care service, dental clinic, or university setting. In addition, the study used valid instruments for assessing OHL [43] and OHRQoL [26], which adds to the robustness of the findings.

However, the study has several limitations. For one, the study targeted only young females, who generally had good OHL and OHRQoL, which limits the transferability of the research results to the entire population. It could also be argued that the overall good oral health of our sample implies that this may not have been the most relevant population for this study. However, it is important to bear in mind that a focus on OHL is a preventive focus, making it worthwhile to investigate this association in a young population. Furthermore, the study population was characterised by high consumption of sugary diets and irregular dental attendance, which indicates that there is room for improvement in the oral health behaviours of this population. Another limitation is that the questionnaire was distributed and completed electronically, which eliminated the investigators' opportunity to explain or clarify any essential matters or unclear questions while participants were completing their questionnaires. This might have resulted in flawed answers. Additionally, even though the moderator variables used in this study are considered relevant, they had little effect on the OR when comparing the unadjusted and adjusted binary logistic regression models. This could imply either that their is a true association between OHL and OHRQoL or that other variables impacting OHRQoL were not included. In this regard, it is important to highlight that this study was limited by the absence of intra‐oral examinations to investigate the oral health status in the study sample. Such data could have strengthened the precision of the investigation.

5. Conclusion

We thus conclude that poor OHL is significantly associated with OHRQoL problems among young women in the KSA and that the OHL problems most associated with OHRQoL problems were financial barriers and making time for and paying attention to oral health. This knowledge is important for oral care workers, health authorities, and decision‐makers. Future studies should investigate the effect of OHL interventions on oral health and quality of life.

Author Contributions

Muneera Essa Alzeer as the principal investigator: conceptualised the study and served as the primary writer of the manuscript; collected, curated, and analysed data. Esben Boeskov Øzhayat as the main supervisor: supervised the research, took part in conceptualisation, writing, reviewing, and editing; provided detailed feedback on each part of the manuscript; and added important information to it. AlBandary Hassan AlJameel: participated in the direct supervision of the data collection process and provided detailed feedback on the entire manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Peer Review

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/joor.13880.

Supporting information

Appendix S1

Data Availability Statement

The datasets generated and analysed during the current study are not publicly available due to the confidentiality of research data but can be obtained from the corresponding author upon reasonable request.