Article Text

Original research
Association of oral health knowledge, self-efficacy and behaviours with oral health-related quality of life in Chinese primary school children: a cross-sectional study
  1. Jie Zhao1,2,
  2. Hongyan Shi3,
  3. Jingya Wang1,2,
  4. Rui Huang3,
  5. Yongyi Liu4,
  6. Yuxin Zhang1,2,
  7. Nan Jiang1,2,
  8. Ting Wang1,2,
  9. Jiwei Wang1,2,
  10. Xiaoming Xu3
  1. 1Fudan University School of Public Health, Fudan University, Shanghai, Shanghai, China
  2. 2Minhang Branch of Fudan University School of Public health, Fudan University, Shanghai, Shanghai, People's Republic of China
  3. 3Dental Disease Prevention and Treatment Clinic of Minhang District, Shanghai, China
  4. 4University of Washington School of Public Health, University of Washington, Seattle, Washington, USA
  1. Correspondence to Dr Jiwei Wang; jiweiwang{at}; Dr Xiaoming Xu; 729523188{at}; Ms Hongyan Shi; 86259225{at}


Objective Achieving good oral health-related quality of life (OHRQOL) is of particular concern in children. The inter-relations among oral health knowledge, self-efficacy, behaviours and OHRQOL in children groups remain unclear. This study aimed to explore the inter-relations between these oral health behaviour-related factors and OHRQOL in primary school children.

Methods In this cross-sectional study, 651 children in grades 2 and 3 were recruited in October 2020 from two primary schools in Minhang District, Shanghai, China. Data were collected through self-reported questionnaires, consisting of demographic characteristics, oral health knowledge, self-efficacy, oral health behaviours and OHRQOL. Pearson’s correlation analyses were used to analyse the relationship between study variables. Structural equation models were used to test the inter-relations between OHRQOL and oral health behaviour-related factors.

Results Four hypothetical structural equation models were tested and one of them was selected as the most appropriate model, which explained 15.0% of the variance in OHRQOL. This selected model showed that oral health behaviours were directly related to OHRQOL. Oral health knowledge was indirectly associated with OHRQOL through both self-efficacy and oral health behaviours. Self-efficacy was directly associated with OHRQOL or was indirectly associated with OHRQOL through oral health behaviours.

Conclusion This study revealed a pathway of association between children’s oral health knowledge and their OHRQOL, in which children’s oral health self-efficacy and behaviours had indirect effects. This provides a basis for understanding the mechanism of oral health promotion interventions to improve children’s OHRQOL and helps to identify direct or indirect intervention targets.

  • public health
  • oral medicine
  • community child health

Data availability statement

Data are available on reasonable request.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • Four hypothetical structural equation models were constructed and compared with comprehensively explore the association between research variables.

  • A pilot study was conducted and most children were all able to complete the questionnaire themselves.

  • The study population was a non-probability sample from two schools only and may cause some selection bias.

  • Some questionnaires used in this study were self-developed and may cause some measurement bias.


Oral health-related quality of life (OHRQOL) is a multidimensional construct, referring to individuals’ subjective perception of comfort while eating, sleeping and engaging in social interaction and their satisfaction with respect to their oral health.1 OHRQOL is an integral part of general health-related quality of life and is regarded by WHO as an important segment of the Global Oral Health Programme.2 As a patient-oriented outcome, OHRQOL demonstrates to clinical providers and researchers that improving the quality of a patient’s well-being is more than just treating oral diseases. OHRQOL has important implications for evaluating interventions, services and public health programmes.3–5 Besides, considering the disparities in access to healthcare and treatment rationing due to cost, comparison of OHRQOL between treatment groups may help patients, healthcare providers and policymakers make decisions.6

OHRQOL is associated with an individual’s sociodemographic, past and current experiences of oral health and dental care, and provides a vision for the future.7 Some oral health problems in childhood, such as dental caries and dental injury, may persist into adulthood.8 Achieving good OHRQOL is of particular concern in primary school children given their dynamic growth periods.9 It is noteworthy that mixed dentition period is a key period for primary school children to replace deciduous teeth with permanent teeth.10 This period is also the best time for early dental care to ensure the healthy development of teeth and the normal growth of facial bones. Consequently, children’s OHRQOL during mixed dentition period should be given special and sufficient attention.

Identifying various predictors that directly and/or indirectly affect primary school children’s OHRQOL is important to better design effective interventions. Most of the existing studies exploring the influencing factors of primary school children’s OHRQOL focus on clinical factors such as children’s oral conditions,11 12 family factors such as parents’ education level,13 14 parents’ oral health behaviours15 16 and parental rearing practices.17 There is evidence that behavioural variables explained as much of the variance of OHRQOL as did the clinical indicators of oral disorders.18–20 A previous study showed the link between less restricted sugar consumption and poorer OHRQOL held up in the multivariate analysis controlling for oral health status, suggesting that the students take account of preventive behaviours when judging their OHRQOL, irrespective of their oral disease status.20 Adopting preventive oral health behaviours in daily life, such as brushing teeth twice a day,21 flossing after meals22 and a healthy diet,21 23 has also become an economical and indispensable way to improve OHRQOL even in the absence of oral disease. Some studies of children aged about 9–15 years showed that supervised toothbrushing24 25 and flossing daily26 were effective in the prevention of gingivitis, while gingivitis was found to be negatively associated with children’s OHRQOL.27 Besides, another intervention study showed that oral hygiene performances were associated with the OHRQOL of Egyptian children with haemophilia aged 6–12 years.28 The relationship between oral health behaviours and OHRQOL needs further exploration among primary school children, especially those in the period of mixed dentition.

Some internal psychological constructs have been given increasing attention in research on OHRQOL. Self-efficacy refers to the set of beliefs held by an individual about his or her ability to perform a particular task or reach a certain goal.29 30 Self-efficacy beliefs affect a person’s thoughts, psychological state and behaviours.29 In an oral health-related context, people with high oral health self-efficacy may have higher motivation to adopt oral health behaviours to prevent some oral problems, thereby improving oral health.31 32 Some studies showed that oral health self-efficacy could influence OHRQOL through toothbrushing, flossing and regular dental check-ups.33 34 Besides, as a subjective experience, OHRQOL captures how persons feel about their overall oral health. Therefore, oral health self-efficacy may have a direct psychological impact on OHRQOL, and people with a higher sense of self-efficacy will be more likely to have more positive senses of OHRQOL. For example, studies showed that independently of students’ caries experience, gingival bleeding scores and sociodemographic status, higher self-efficacy was consistently associated with better OHRQOL across time, suggesting that school students with higher self-efficacy levels were more emotionally resilient to challenges caused by their oral symptoms.35–37 Oral health knowledge is also considered to be crucial for developing oral health behaviours, and it has been shown that increased knowledge is associated with better OHRQOL.21 38 However, the relationship between oral health knowledge, oral health-related self-efficacy and OHRQOL among children’s groups is still unclear.

In this study, path analyses using structural equation modelling were performed to explore the structural relationships among the direct and indirect determinants of OHRQOL, aiming to provide clues for children’s OHRQOL intervention practice.


Participants and procedures

The inclusion criteria of this cross-sectional study included: (1) children aged 7–9 years; (2) no cognitive impairment; (3) no communication disorder. To ensure the sample size required for structural equation model analysis, it is desirable to require 5–20 observations for each path or at least 200 observations (whichever is greater).39 In this study, with 9 parameters (ie, paths) in the most complex model and 20 observations per parameter, 180 participants with complete data were needed. Therefore, the final sample size of this study should be no less than 200. Due to the limited availability of resources and personal contacts, a convenient sampling method was used. There are 67 primary schools in Minhang District, with about 826 000 students. Although recruiting participants in one school may meet the sample size requirements of this study, we decided to recruit two schools to reduce selection bias. We issued recruitment invitations to five primary schools in Minhang District, and then recruited two primary schools on a first-come, first-served basis in October 2020. In each school, we adopt the cluster sampling method to recruit the class as the sampling unit. In return for taking the survey, participants were rewarded with cartoon toothbrushes or other cartoon gifts. And to avoid students in some classes feeling lost or experiencing psychological gaps when they did not receive gifts, we recruited all the second-grade and third-grade classes in these two schools, resulting in an excess of expected sample size. Finally, a sufficient sample size of 770 second-grade and third-grade students were recruited.

Questionnaires (online supplemental appendix 1) were distributed to students during the break time (30 min). Before the students filled in the questionnaire, the investigator explained the whole questionnaire to them from beginning to end, including the meanings of all the items in the questionnaire and how to fill out the questionnaire. Two investigators were deployed in each classroom to help answer any questions students might have while filling out the questionnaires. All the investigators majored in public health and received standardised training. In the end, 33 students failed to fill out the questionnaires due to extracurricular affairs, and 737 questionnaires were collected on site, with a response rate of 95.71%. After the questionnaires were collected, the research team conducted data quality checks on the questionnaires. If the number of missing values of oral health knowledge (8 items), self-efficacy (4 items) or oral health behaviour (7 items) was >1, or the number of missing values of OHRQOL (19 items) was 2 or above, the questionnaire was judged invalid. Finally, 86 questionnaires were judged invalid, and the remaining 651 questionnaires were all analysed in this study. In addition, to meet the requirements of readability and understandability, we conducted a pilot study with 30 second-grade students from another primary school in Minhang District to ensure that uncommon words were not used in the questionnaire and that all participants could understand the meaning of each phrase or sentence. The results of the pilot study showed that only ~two to three words were beyond a few participants’ comprehension and these words were replaced with easily understood synonyms.


Demographic characteristics

The demographic characteristics investigated in this study included age, gender, grade and only-child status.

Oral health knowledge

Referring to the existing scales of children’s oral health knowledge,40–44 the research team selected 15 items about children’s oral health knowledge according to the understanding ability and limited oral knowledge of lower grade students. The research team included two dental clinicians, a health education researcher and two primary school teachers. Subsequently, the research team organised three discussions to merge, delete and supplement the selected items based on Chinese children’s oral health education programmes. Finally, eight true-false items were determined and the face validity of the oral health knowledge questionnaire was acceptable. One point is given for the correct answer and zero is given for the incorrect answer. Besides, the ‘do not know’ option was set for oral health knowledge items. When children were unable to respond to the items, they could choose ‘do not know’ and the score was calculated as 0. The total score is the sum of the scores of the eight items. In this study, the Cronbach’s α and the Guttman Split-Half coefficients of oral health knowledge questionnaire were 0.716 and 0.655, respectively.

Oral health behaviour self-efficacy

Most existing oral health behaviour self-efficacy scales evaluate only one specific oral health behaviour36 or no more than three kinds of behaviours, such as tooth-brushing self-efficacy, dietary self-efficacy and dental visiting self-efficacy.45–47 In order to measure the self-efficacy of multiple oral behaviours, this study used four items to measure the self-efficacy of four oral health behaviours, including brushing their teeth every morning and evening, rinsing the mouth after meals, having regular oral health examinations and participating in oral health education activities at school. In the following explanatory questionnaires, we clearly explained the meaning of rinsing after meals, and also revealed the meaning of oral health education activities, such as watching oral health-related promotional videos, drawing popular science paintings of oral health knowledge and participating in oral health education lectures. There was no difference in the organisation of oral health education activities between the two schools and grades investigated in this study. The 7-Likert scale was adopted, and the score of self-efficacy in this study is the average score of four items, ranging from 1 to 7. During explaining the questionnaire, we also orally emphasised to the students that they should choose the corresponding answer options of the 7-Likert scale representing their self-confidence level rather than depending on parents or teachers in taking oral health behaviours. In this study, the Cronbach’s α and the Guttman Split-Half coefficients of oral health behaviour self-efficacy questionnaire were 0.836 and 0.756, respectively. One thing that needs to be pointed out here is that this investigation did not measure diet-related self-efficacy due to two reasons: (1) the content of ingredients in food needs to be calculated and (2) children’s perceptions of food intake vary widely.

Oral health behaviours

With reference to the Chinese National Oral Health Epidemiology Questionnaire48 and the relevant literature,40–42 we used seven most applicable items to measure children’s oral health behaviours, including frequency of brushing, the number of surfaces brushed, time of brushing, frequency of changing toothbrushes or replacement head, frequency of rinsing after meals, frequency of flossing and frequency of eating sweet foods. Considering that fluoridated toothpaste was purchased and provided by parents, the use of fluoridated toothpaste was not included in the investigation of children’s own oral health behaviours in this study. As for the investigation of sweet food intake, desserts and sweets, sweet drinks and other sugary drinks (some may be unpopular among children) were all included in the questionnaire. A few examples of each kind of sweet food were listed to help children to fill out the questionnaire, but children were not limited to these examples. During explaining the questionnaire, we explained the differences in frequency options to children. For example, in the frequency option of flossing, ‘sometimes (frequency <50%)’ meant that flossed for total 3 days or less in a week, and ‘often (frequency >50%)’ meant that flossed for total 4 days or more in a week.

Because the number of options for different oral health behaviour items varied, we assigned weighted scores to the different options for each item. The weighted scores of some options under the same item were the same, which made the maximum score of some items with fewer options equal to that of items with more options. We referred to the method of assigning weight to the options of some oral health behaviour items in previous research,44 while the weighted scores of the options of some other items were assigned by the research team according to the relative importance of the options. For example, the answer (and its weighted score) for the item ‘the daily frequency of brushing’ is: zero (weighted score=0); 1 time (weighted score=1); 2 times or more (weighted score=2). Then we organised two expert symposiums and eight experts in the field of children’s oral health were invited to discuss the rationality of weight and put forward suggestions. Finally, the experts reached a consensus on the allocation of weights. Three items were related to the behaviour of sweet food intake, so the average score of these three items was calculated as the score of dietary behaviour. The total score for the oral health behaviour was added up by a weighted score for each kind of behaviour, ranging from 0 to 13.

Oral health-related quality of life

OHRQOL was measured by the Child Oral Health Impact Profile-Short Form 19 (COHIP-SF 19). The COHIP-SF 19 is a simplified version of the 34-item COHIP scale, which has been widely used in school children aged 7–17 years.49 50 COHIP-SF 19 contains three dimensions, including oral health (5 items), functional well-being (4 items) and a combined subscale named socio-emotional well-being (10 items). All items were rated on a 5-point Likert scale, and scores of the dimensions are the average scores of the corresponding items, ranging from 1 to 5. In addition, the COHIP-SF 19 adopts a 3-month period as a recall period. Prior to the study, we obtained the authorisation to carry out the Sinicization and the research use of COHIP-SF 19. The Sinicization process is as follows: first, the COHIP-SF 19 was independently translated into Chinese by two bilingual experts. Their translations were checked and coordinated by a third bilingual expert, who merged the mixed version into the original Chinese version. The original Chinese version was then translated back into English by two other bilingual experts who did not know the original English version of COHIP-SF 19. Finally, the research team checked and proofread the items of the translation and the reverse translation one by one. If the translations do not conform to the Chinese context or are not easily understood by readers, they would be modified without changing their original meaning. After repeated discussions, the final Chinese version was determined, and the consensus was reached on the translation version. In this study, the Cronbach’s α and the Guttman Split-Half coefficients of Chinese COHIP-SF 19 were 0.740 and 0.618, respectively.

Questionnaire adjustment

According to the Chinese curriculum standard of the Shanghai primary school, students of the second-grade are required to recognise about 2000 commonly used Chinese words and be able to correctly write >1000 Chinese words, while students of the third-grade are required to recognise and correctly write about 2500 Chinese words. In order to ensure the readability and understandability of the questionnaire, when we referred to the items in the existing literature, we just chose the items suitable for the reading and understanding ability of second-grade children. Besides, we referred to the vocabulary required by the second-grade Chinese textbooks in Shanghai, and modified some items appropriately without changing the original meaning of the items.

Statistical analysis

Descriptive statistics were used to summarise the demographic characteristics of participants. Pearson’s correlation analyses was used to preliminarily analyse the relationship between study variables. Structural equation models were used to test the inter-relations between OHRQOL and its associated variables, including the direct or indirect effects of the variables on OHRQOL. OHRQOL was modelled as a latent variable with the oral health, functional well-being and socio-emotional well-being as indicators; all other variables were modelled as observed variables. We used t-test and Pearson’s correlation analysis to test whether age, gender, grade and only-child were the confounding factors in this study. Variables tested as confounding factors were adjusted in the structural equation model. Demographic variables, as categorical variables, were set as dummy variables for correlation analysis. The grouping of demographic variables is as follows: ‘age (1=‘≤7 years’, 2=‘8 years’, 3=‘≥9 years’), gender (1 =‘male’, 2=‘female’), grade (1=‘grade 2’, 2=‘grade 3’), Only-child or not (1=‘yes’, 2=‘no’). The statistical analyses were performed using the SPSS V.22.0 and the AMOS V.22.0. Two-tailed statistical significance level was set at 0.05. The missing value was replaced by the continuous mean value method.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.


Demographic characteristics and descriptive results

A summary of participants’ demographic characteristics is provided in table 1. The participants in this study included 241 students of grade 2 (37.0%) and 410 students of grade 3 (63.0%). Nearly half of the participants were male and the only child. The average age of the participants was 8.14 years.

Table 1

Summary of basic information for children

Table 2 shows that the participants had a better understanding of the symptoms and risk factors of gum bleeding, gum inflammation and tooth decay, and the correct rates of answers to these items ranged from 51.6% to 66.0%. They had poor knowledge of the protective effect of fluoride and pit and fissure sealing on oral health, and the correct rates of answers were 32.1% and 38.8%, respectively. In addition, the mean score of the four oral healthcare self-efficacy items ranged from 5.86 to 6.31, showing that the participants of this study had good oral healthcare self-efficacy.

Table 2

The proportion of correctly response to oral health knowledge and the level of self-efficacy

Table 3 presents that among the participants, 70% brushed their teeth twice a day or more, and 54% brushed four surfaces when brushing teeth, and 48% brushed their teeth for >2 min each time and 58.2% changed their toothbrushes or brush heads more frequently than once every 3 months. It could be noted that 32.0% of the participants never gargled after meals and 64.5% never used dental floss. Most of the participants ate sweet food once a week or less.

Table 3

The weight scores and descriptive results of each oral health behaviour item

Preliminary correlation analysis and t-test

Table 4 shows that the understanding of the effect of fluoride on teeth was significantly related to oral health-related self-efficacy (p=0.025) and behaviours (p=0.025), while the understanding of gingival bleeding was significantly related to the oral health (p=0.017) and socio-emotional well-being (p=0.013) dimensions of OHRQOL. In addition, the understanding of the effect of brushing on gum inflammation was significantly related to oral health-related self-efficacy (p=0.001).

Table 4

The association between different oral health knowledge items and other research variables

Table 5 shows the results of correlation analysis among variables. Children’s age, gender, grade and only-child status had no significant correlation (p>0.05) with the three factors of OHRQOL. However, the three factors of OHRQOL had significant correlations with oral healthcare self-efficacy (Pearson’s r ranged from 0.105 to 0.227) and had more strong significant correlations with oral health behaviours (Pearson’s r ranged from 0.160 to 0.253). The correlation between oral health knowledge and OHRQOL is relatively weak, in which oral health knowledge is significantly correlated with functional well-being factor (Pearson’s r=0.108) and socio-emotional well-being factor (Pearson’s r=0.112), but not with oral health factor (Pearson’s r=0.050).

Table 5

The correlation analysis between the study variables

Structural equation model comparisons of the inter-relationships between OHRQOL and its associated variables

Based on a review of the literature,29–37 an appropriate hypothetical model is constructed, namely model 1 in figure 1. The literature suggests that oral health knowledge may be directly associated with OHRQOL,38 and there is some doubt as to whether self-efficacy was directly related to OHRQOL.35–37 Therefore, three alternative models, models 2–4 in figure 1, were also applied. The results show that, compared with model 3 and model 4, the fitting of model 1 and model 2 are better, indicating that the path in which self-efficacy indirectly affects the OHRQOL through oral health behaviours is critical. Compared with model 1, model 2 has a larger interpretation variance of OHRQOL and a better fitting, so it was selected as the final model in this study. Model 2 could explain 1.0% variance of self-efficacy, 16.6% variance of oral health behaviour and 15.0% variance of OHRQOL. The results of model 2 showed that both self-efficacy and oral health behaviour directly affected OHRQOL, while oral health knowledge had no significant direct effect on OHRQOL (p=0.086). Besides, oral health knowledge and self-efficacy could indirectly affect OHRQOL through oral health behaviours.

Figure 1

The four structural equation models tested in this study. B represents the standardised parameter estimates for *p<0.05; **p<0.01; ***p<0.001. CFI, comparative fix index; GFI, goodness-of-fit index; AGFI, adjusted goodness-of-fit index;NFI, normed fit index; OHRQOL, oral health-related quality of life; RMSEA, root mean square error of approximation.


In this study, four hypothetical structural equation models were constructed to explore the relationship between children’s oral health behavior-related variables and OHRQOL. The final selected model showed that oral health behaviour was directly related to OHRQOL. Oral health knowledge was indirectly associated with OHRQOL through both self-efficacy and oral health behaviour. Self-efficacy was directly associated with OHRQOL or was indirectly associated with OHRQOL through oral health behaviour.

At present, there were many valid scales measuring children’s OHRQOL, among which COHIP-SF 19 (suitable for children aged 7–17 years),51 Child Oral Health Quality of Life Questionnaire (suitable for children aged 8–14 years)52 and Early Childhood Oral Health Impact Scale (suitable for children aged 3–5 years)53 were the most commonly used.54 COHIP-SF 19 was used in this study and it showed good internal reliability, which ensured the effective measurement of the outcome variable OHRQOL to a certain extent. The results of this study showed that the oral health behaviours of children aged 7–9 years significantly related to their OHRQOL. This was consistent with a previous study which found that dental flossing frequency was the strongest predicting factor correlated with gingivitis in children aged 9–13 years,26 while gingivitis was found to be negatively associated with children’s OHRQOL.27 Another study showed that children (aged 14.32±0.68 years) who brushed their teeth less than once a day and consumed chocolate or biscuits regularly reported a statistically worse OHRQOL.21 Some intervention studies showed that supervised toothbrushing was effective in improving the overall oral health status of children aged about 12–15 years.24 25 Oral health behaviours may prevent the progression of oral diseases such as dental caries, prevent the onset of these oral diseases and, in addition, improve normal oral health status, thereby improving OHRQOL. Besides, the results of this study showed that the oral health behaviours in children were significantly correlated with oral health knowledge and self-efficacy, which indicated that promoting each of the two predisposing factors could contribute to the improvement of the oral health behaviours among children.

Primary school children should be also educated and empowered to develop various oral health behaviours, not just one single behaviour. The results of this study showed that more than half of primary school students had good tooth brushing habits, including brushing their teeth twice or more a day, brushing enough surfaces and changing the brush head in time. In addition, the frequency of sweet food intake is low among the primary school children in this study. However, it is worth noting that 64.5% of primary school children in this study had never used dental floss. This is consistent with the result of a previous study, which found that 90.41% of children aged 12 years in China had never used dental floss.55 Therefore, it was suggested that the use of dental floss should be given more attention in future children’s oral healthcare projects.

In this study, compared with model 1 and model 2, the fitting results of model 3 and model 4 were poor, which fully demonstrated the importance of the pathway in which self-efficacy affects OHRQOL through oral health behaviours. Individuals with a high sense of oral health-related self-efficacy beliefs might believe that they can do much to prevent poor oral conditions and would have a high motivation for oral health-promoting behaviours.31 32 This study also showed that self-efficacy was directly associated with OHRQOL. A cross-sectional study showed that the more positive the general self-efficacy of adult dental patients, the better their subjective oral health.56 In addition, a 2-year follow-up of students (15.2±1.3 years) in Tanzania showed that general self-efficacy was significantly positively associated with OHRQOL.57 Individual self-efficacy beliefs affect a person’s cognition and psychological states. Therefore, positive self-efficacy beliefs have a direct impact on a person’s OHRQOL, which is essentially a subjective perception of oral health status.

Self-efficacy develops in the course of life, and the transition from pre-adolescence to adulthood is considered to be a critical period for the growth and maintenance of self-efficacy. This period of life is also critical in adopting healthy behaviours that are long lasting and difficult to change beyond adolescence. A study adopting the health belief model showed that self-efficacy was the most important factor influencing oral health behaviour of Korean elementary school students.58 Toothbrushing and dietary self-efficacy are shown to be strongly associated with corresponding behaviours and oral health status among Turkish pre-adolescents aged 10–12 years.45 Self-efficacy has been accepted as an important component of skills-based health education in health promotion schools,59 and it plays an important role in developing effective preventive oral health interventions among pre-adolescent children.

The structural equation model results of this study suggest that oral health knowledge can indirectly, but not directly, affect children’s OHRQOL by influencing oral health-related self-efficacy and behaviours. As yet, there is still a lack of studies on the relationship between children’s oral health knowledge and their OHRQOL. Most existing studies have explored the correlation between children’s oral diseases and their own oral health knowledge. For example, a study showed that lack of appropriate oral health knowledge was the main risk factor for dental decay among children aged 10–14 years in Saudi Arabia.60 The results of this study also show that the association between individual oral health knowledge and oral health-related self-efficacy, behaviour and OHRQOL were quite different, and all of the associations were poor, which also supported that it was more suitable to use the total score of oral health knowledge for structural equation model analysis.

In this study, the primary school children did not know much about the protective effect of fluoride and pit and fissure sealing on oral health. Pits and fissures of the first permanent molar of children aged 6–7 years in mixed dentition period are most vulnerable to dental caries.43 Pit and fissure sealing of school children has been included in the basic public health project since 2011 in Shanghai, and some primary school children in Shanghai have heard of or received pit and fissure sealing. Therefore, the awareness of pit and fissure sealing was included in this investigation. Although the purchase and use of fluoride toothpaste and treatment of pit and fissure sealing are all decided by children’s parents, children’s understanding of this oral health knowledge helps to increase their enthusiasm and compliance with oral health behaviours, especially in the absence of parental supervision.

Although many studies had shown that clinical dental treatment can significantly improve children’s OHRQOL,61–63 the development and practice of oral health education programmes are also of significant concern. OHRQOL intervention strategies that combine clinical dental treatment with oral health education approaches64 65 generated a greater positive effect than those related to only clinical dental treatment64 or to isolated educational practices.65 Many studies also showed interventions that coupled traditional oral health education with a variety of other interventions (including supervised toothbrushing, fluoride supplement, clinic-based professional preventive oral care, etc) were effective in improving children’s oral condition, specifically in reducing plaque66 and dental caries,67 but most studies lasted <1 year and only a few studies looked at long-term sustainability of the effects.68–70 More long-term evidence is needed to assess the sustainability of these interventions in improving children’s oral health. With the intensification of children’s education in China, the literacy level and comprehension ability of primary school children in China have also been greatly improved. In this context, oral health education programmes aiming at children’s groups directly to improve their own oral health knowledge, self-efficacy and behaviours are feasible and necessary. It is not enough to rely solely on parental or teachers’ supervision and control of children’s oral health activities to improve children’s OHRQOL. In addition, it should be noted that the key period of oral healthcare for children is the mixed dentition period, that is why children aged 7–9 years were selected as the research population in this study. Due to the poorer reading ability of the children of first grade and the fact that children at the age of 10–12 years were about to complete teeth replacement, this population of children were not included in this study.

There are some limitations in this study. First, the results of this cross-sectional study presented the associations rather than the causality. Although the design of a prospective study has the advantage of indicating the direction of a relationship over time, it nevertheless may also fail to draw a definitive conclusion of causality.71 In this study, the directions of causality between oral health-related knowledge, self-efficacy and behaviours and OHRQOL were supported by the existing literature,29–38 and the impacts of some possible confounding factors were also analysed and excluded to ensure rigour in drawing causal conclusions. Based on these considerations, this study would give some clues for possible causal relations and provide ideas for intervention studies which are needed to further draw firmer causal conclusions. Second, the sample is a non-probability sample from only two schools in Minhang District and may cause some selection bias. Due to the high level of urbanisation, the differences between various communities in Shanghai are small, and the primary schools in Minhang District are also well represented in Shanghai to a certain extent. Besides, 33 children refused to participate in the survey, which may lead to non-respondent bias. Third, this study did not include clinically related variables, which may act as mediating variables16 in the relationship between oral health behaviours and OHRQOL. The dentist visiting (such as frequency of dental visits and reasons for dental visits) was also not included in the analysis, but it might be a moderator for oral health behaviours and have an effect on oral health knowledge. The relationships between oral health behaviours, clinical characteristics, dentist visiting and OHRQOL in middle childhood children need to be further explored. Finally, the questionnaires of oral health behavior-related variables used in this study were self-developed, which may cause certain measurement biases, and the test-retest reliability of these questionnaires needs further verification. In addition, the Likert scales were used, while some studies pointed out that children’s understanding of the answer options of the Likert scale was limited.72 Therefore, before the children filled in the questionnaire, we introduced the meaning of answer options of the Likert scale to students by drawing corresponding emoticons on the blackboard. Less than 10% of the children asked questions when they filled out the questionnaire, but they were all able to complete the questionnaire successfully with the help of the investigator.


This study revealed a pathway of association between children’s oral health knowledge and their OHRQOL, in which children’s oral health self-efficacy and behaviours had indirect effects. This provides a basis for understanding the mechanism of oral health promotion interventions to improve children’s OHRQOL and helps to identify direct or indirect intervention targets.

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by Minhang District Health Academic Ethics Institute with the approval number of 2019-03-02. Prior to the investigation, the informed consent form was taken home by the children and was approved by their legal guardians.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • JZ and HS are joint first authors.

  • Contributors HS, JW (Jingya Wang), RH and XX worked on the data collection. JZ did the statistical analysis and wrote the manuscript. YL, YZ, NJ and TW revised the paper and improved the technical quality of the manuscript. JW (Jiwei Wang), HS and XX were the project coordinator and participated in all parts of the work. JW (Jiwei Wang), HS and XX are responsible for the overall content of the manuscript as guarantors. All authors approved the final version of the paper.

  • Funding This work was supported by 2019 Fudan-Minhang Health Consortium Cooperation Project (2019FM04).

  • Disclaimer The results presented here are solely the responsibility of the authors and do not necessarily represent the views of the study funders.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.