The CPQ has previously been developed and tested in a clinical convenience sample of children in Canada [
5,
6]. Every time an instrument is used in a new context or with a different group of individuals, it is necessary to reestablish its psychometric properties. In this study, the CPQ
8–10 and CPQ
11–14 were applied to a general sample of schoolchildren (8 to 14 yrs of age) in a country (Brazil) with a different cultural context. Prior to validity and reliability tests, the questionnaires were translated, back-translated, and crossculturally adapted in order to ensure their conceptual and functional equivalences.
The following subheadings discuss the results.
CPQ Pretesting
At the pretesting stage, children from 8 to 10 years old were able to answer all questions in the questionnaire, whereas in the Jokovic et al. [
5] study, 8-yr-old children did not relate to the introductory/transition statement,
"In the past 4 weeks, because of your teeth or mouth...", when responding to the questions and required either a simpler format or an interviewer-supervised/administered questionnaire. Moreover, in the present study, the children of the CPQ
11–14 group did not understand some of the questions and required some of the words to be changed to guarantee their cultural equivalence. A few problems were also encountered in the Arabic translation of CPQ
11–14 with regard to self-reporting of age, and the questionnaire was considered too long for many of the medically compromised patients [
9]. Moreover, it was suspected that quite a few of the children asked their parents for help, which probably influenced the replies [
9]. Goursand et al. [
13] chose to administer the CPQ
11–14 as an interview in order to avoid the possibility of children soliciting help from their parents when having difficulty understanding the questions. However, these authors [
9,
13] also suggested that the effect of different modes of administration on the validity and reliability of the instrument should be evaluated in population studies, as done in the present one, in which the questionnaires were self-applied on a population sample, resulting in satisfactory psychometric properties, irrespective of the mode of administration. Therefore, translating and adapting a questionnaire developed in one country for use in another usually results in some changes in the wording, format, and mode of administration, which have been facilitating the development of a culturally relevant instrument [
6,
9,
20‐
22], being a strong point of the methodology for use in a different setting.
CPQ Discriminant and construct validity
When testing discriminant validity, a clear ascending gradient was observed for oral symptoms among children aged 11–14 years, with those aged 11 years being the highest and those aged 14 years being the lowest (see Additional file
2); however, this was not observed for the CPQ
8–10 group (see Additional file
1). This reflects the fact that children's understanding of oral health and well-being is also affected by age-related experiences [
2,
21]. During mixed dentition (8–12-yr-olds), children experienced many problems related to natural processes, such as exfoliating primary teeth, dental eruption, or space due to a nonerupted permanent tooth, which simultaneously affect their QoL. On the other hand, these conditions were not reported as important causes of oral impacts in other age groups [
23]. After 12 years of age, children will move from a transitional dentition, just as they will have altered their concepts of health and probably also have different expectations [
1,
24].
While there was an apparent gender difference in the CPQ
11–14 score, it did not quite reach statistical significance (see Additional file
2). These findings suggested that girls tend to report higher impacts on QoL than boys, on average. However, in the Foster Page et al. [
3] study, the mean emotional well-being domain score was higher for girls than for boys. One explanation for these variations is related to the differences in the characteristics of selected samples between the Foster Page et al. [
3] and the present studies, and patient and general population samples, respectively.
In CPQ
8–10 group, girls had higher impacts on all CPQ
8–10 scores than boys (see Additional file
1). There are no studies in the literature that evaluated differences between genders related to oral impacts on QoL during middle childhood (6–10 yrs). Thus, further research on OHRQoL needs to be conducted using samples of this age group in order to elaborate on the findings reported here. Furthermore, these findings were similar to the results of the CPQ
11–14 group. However, the difference in the significance between the results of the two groups may be explained by the particularity in the cognitive, emotional, functional, and behavioral characteristics of each age group [
24]. This implies that the comparison between the results related to age-specific CPQ groups should be interpreted with caution, since they are heterogeneous in terms of stage of development.
Concerning dental caries experience, it was hypothesized that children with more severe caries would have higher impacts on their QoL, corroborating recent studies [
3,
4,
9,
25]. However, only primary dentition showed significant correlation with both CPQ scores (see Additional files
1 and
2). There was an ascending difference between dmft and all CPQ
11–14 domains, except for functional limitations. Such a gradient was also observed with respect to the CPQ
8–10 social well-being domain but not with the others. These findings may be explained by the fact that adolescents had experienced untreated disease for longer than the younger participants, also reflecting the health view as a multidimensional concept during early adolescence [
26].
Analysis within DMFT was not statistically significant but also provided some evidence to suggest that the CPQ
8–10 scores were associated with the severity of this clinical condition in an expected direction (see Additional file
1). Furthermore, no clear statistically significant gradient was observed with respect to the CPQ
11–14 scores and DMFT categories (see Additional file
2). It is known that caries progresses more rapidly in primary teeth than in permanent teeth, supporting the hypothesis that deciduous enamel is more susceptible to caries than permanent enamel [
27]. Consequently, although dental caries was relatively prevalent in permanent dentition, it did not affect the child's ability to perform daily activities.
No clear gradients were observed in both mean CPQ scores across the categories of malocclusion severity (see Additional files
1 and
2). The results of other studies conducted to date are equivocal [
3,
28,
29]. While some studies indicated good discriminant validity between children with different levels of malocclusion severity [
3,
29], others did not [
28]. Despite the small sample sizes involved with a handicapping malocclusion, which could be considered as a limitation into this context, the lack of marked difference is also consistent with the contemporary models of disease/disorder and its consequences. The model by Wilson and Cleary [
30] indicates that health outcomes experienced by an individual are not determined only by the nature and severity of the disease/disorder but also by the personal and environmental characteristics. Moreover, different meanings of QoL vary between and within groups of individuals [
31] according to culture and education [
23], contributing for distinct impacts of malocclusion on QoL.
Although CPQ
8–10 and CPQ
11–14 scores tended to be lowest for the 'fluorosis ≥ 1' category and highest for children without dental fluorosis, differences were not significant (see Additional files
1 and
2). A potential explanation may be the low disease levels in the sample. However, although the levels of fluorosis were low in the Robinson et al. [
4] study, the Ugandan children experienced appreciable impacts on OHRQoL. These contradictory outcomes suggest that cultural norms and expectation influence children's perception of their oral health and its effect on their QoL, as considered, since causal pathways between clinical variables may include individual and environment variables as both moderators and mediators [
30].
It was observed that children without gingivitis would have higher CPQ
8–10 scores (see Additional file
1), contrasting with other studies [
2,
25]. The following explanations may account for the present findings: the clinical instrument was not performed as a discriminant measure, there was oral disease in the small sample size, or the impacts were mediated by a variety of factors, such as relevance. Moreover, while there was an ascending difference between preadolescents without and with gingivitis, it did not quite reach statistical significance (see Additional file
2). The lack of marked difference may be due to the low disease levels in the sample, which caused immeasurably low levels of impact. Furthermore, the way people feel about their QoL also needs to be considered, since it does not develop in isolation from their existing expectations (that constrain what is relevant) as well as the environment in which the margins of relevance are constructed [
31].
Finally, the results of this study suggested that both questionnaires have good construct validity (Table
2). Significant correlations were shown between global rating of oral health and overall well-being and the total scale and all subscales, indicating also that children are able to give psychometrically acceptable accounts concerning their health status and its overall effects on their lives [
32].
CPQ Internal Consistency and Test-retest Reliability
Both questionnaires have acceptable reliability with the internal consistency [
33] and test-retest reliability [
34] (Table
3). Cronbach's alpha and ICCs found in this study were similar to the results from Canada [
5,
6]. However, in the Jokovic et al. [
6] study, the ICC for the social well-being subscale was low at 0.16, suggesting that children are more likely to experience variability over time in social functioning and experiences than in physical and emotional effects of oral and orofacial conditions. An alternative explanation for these contradictory outcomes is that enjoying contact with people might be an inherently unstable construct to children, which varies with time [
35].
In addition, children are, in a sense, 'moving targets' not just because childhood is a period with immense changes in psychosocial awareness but because the children's dental and facial features change rapidly [
36]. Furthermore, children's cognitive development varies such that the wording of items, specific dimensions, and their relevance and meaning to children of similar ages can differ, and the changes in a child over time can make repeated measurements difficult to compare [
37].