Study population
To describe periodontal conditions in the Sámi population in Northern Norway, data from the “Dental Health in the North” study was used. This study and its methodology is described in detail in Brustad et al. [22]. The study was a cross-sectional study of adults 18–75 years old in Finnmark County in Northern Norway. Data was collected between February 2013 and May 2014. All patients attending public dental care services in five municipalities (Tana, Nesseby, Porsanger, Karasjok and Kautokeino) during the study period were invited to participate in the study regardless of the reason for their appointment at the clinic. Of a total of 2520 persons invited to participate, 285 persons declined (crude response rate at 88.7%) and 157 participants were not included in the final sample. Reasons for non-inclusion were the following: missing questionnaire, missing clinical data or both, unknown target age, missing written consent, or not accounted for and thus given missing unknown status. The final sample consisted of 2078 participants.
The regional committee for medical and health research ethics of the University of Tromsø, Norway, approved the study (2012/1902/REK Nord). All participants provided written informed consent.
Questionnaire
Population characteristics were collected by self-reported questionnaire. The questionnaire covered information about ethnicity, household income, education, smoking habits, use of dental health care services, and oral hygiene related behaviours. Self-reported ethnicity was based on three questions: 1) Which language do/did you/your parents/grandparents speak at home? 2) What is your/ your parents’ ethnic background, and 3) What ethnicity do you consider yourself to be? The response options were ‘Norwegian’, ‘Sámi’, ‘Kven’ and ‘other’. For a more thorough description of questions included in the questionnaire, including ethnic categorization, see Brustad et al. [22]. Brustad et al. describe the ethnicity as a complex phenomenon, were both the objective factors (parents and grandparents being Sámi) and the subjective feelings of belonging to the Sámi culture have to be taken into account when creating the ethnicity variable [22]. In this study, ethnic affiliation was categorised as ‘Sámi’ and ‘non-Sámi’. The ‘Sámi’ category represented those who answered ‘Sámi’ on at least one of the two questions about language and ethnic background, in addition, reported that they consider themselves as Sámi. All other respondents were categorised as ‘non-Sámi’ even though some of the participants reported that they had Sámi heritage but did not consider themselves as Sámi. The ‘non-Sámi’ group were mainly Norwegians, Kven (without Sámi affiliation, n = 99) and Sámi with some affiliation, but without subjective Sámi criteria (n = 165).
Education in Norway is mandatory for all children aged 6–16 years. The education system is made up of primary school (6 years), secondary school (4 years), High school (3 years) and Higher education (University level). Participants’ educational level was assessed with one question that elicited responses in number of years, and grouped into three categories: [1] 1–9 years, [2] 10–13 years and [3] ≥14 years. Tooth brushing habits were assessed with one question with four response options from two or more times per day to never. These four options were merged into three categories: [1] less than daily, [2] 1 time/day, and [3] ≥ 2 times/day. Smoking habits were assessed with two questions 1) Do you smoke daily? 2) How many cigarettes do you smoke per day? Age was divided into four age groups: 18–34, 35–49, 50–69 and 65–75 year olds.
Clinical dental examination
Nine dentists and six dental hygienists with assisting nurses, in six separate dental offices, carried out the clinical examination. Data on a participant’s periodontal condition was collected from a clinical examination including four bitewing radiographs. Periodontal probing depth (PD) at six sites per tooth was measured to the nearest millimetre with a periodontal probe with single millimetre graduations (WHO- probe LM555B). Clinical attachment level (CAL) was not assessed, so the alveolar bone level (ABL) based on the radiographs was used as main criteria classifying prevalence and severity of periodontitis. Third molars and implants were examined but excluded from analysis.
Calibration of the examiners were done as follows: First, the examiners had a workshop regarding the diagnostic criteria and examination procedures. Secondly, all examiners were trained and calibrated towards an experienced periodontist who was the gold standard (NO). This calibration included radiographic examination technique and periodontal pocket probing on one patient each. Third, the examiner (A-KSB) was trained by an experienced periodontist (NO). ABL based on radiographs of randomly chosen participants were measured and an inter-examiner reliability was conducted, κ = 0.97. Examiner (A-KSB) measured the ABL twice, and intra-examiner reliability for the ABL measurements was conducted, κ =0.95.
The clinical dental examination procedure and the post clinical measurements of bone level, including validity, is described in details elsewhere [22].
Classification of periodontitis
A modified classification system based on the new AAP/EFPFootnote 1classification system of periodontal disease was used to present prevalence of periodontitis [23, 24]. In this study, stages of periodontitis were classified by radiographic bone loss (RBL) and PD. Missing teeth and furcation involvement were not included. Stage I is the borderline between gingivitis and periodontitis and represents the early stages of attachment loss. Because RBL < 15% can be difficult to measure on radiographs without having the exact root length, Stage I was not included as a periodontal case. A patient was classified as a Stage II periodontitis case if the RBL was between 15 and 33% and as Stage III/IV if the RBL was extending to middle or apical third of the root in two or more non-adjacent teeth. The complexity factor (PD) was included and may shift the stage to the higher level; PD 4–5 mm classified to Stage II and PD ≥ 6 mm classified to stage III/IV. Cases with no periodontitis and early stage of periodontitis were classified as ‘Non-severe periodontitis’ (NSP).
Statistical analysis
Missing data occurred at a low frequency (0.1–5.0%). There were no internal loss in regards to ethnicity. The greatest proportion of loss was for household income followed by education (3.3%).
Differences between the Sámi- and the non-Sámi groups, as well as classification of periodontal disease, were calculated for demographic and socioeconomic status (age, sex, household income and education), smoking habits and tooth brushing frequency. Prevalence of periodontitis was presented as the frequency distribution for AAP/EFP classification method and presented as ‘Non-severe periodontitis’, ‘Stage II’ and Stage ‘III/IV’. Differences in prevalence of stages of periodontitis between Sámi and non-Sámi was stratified by age group, and assessed with z-test and analysed by univariate regression analysis.
Differences in background characteristics between Sámi and non-Sámi group and between classifications of periodontitis were assessed with Pearson χ2 test, and differences between groups were assessed with z-test. Age and number of teeth were presented as means and standard deviation (SD).
RBL and PD are presented as percent and proportions (SE) of affected sites and teeth for the total study population, stratified by age group and ethnicity. Differences between groups were assessed with χ2-test and t-test.
Multinomial logistic regression was performed to determine the relationship between stages of periodontitis in relation to ethnicity, socio-demographic and behavioural factors. Ethnicity, sex, age, education, smoking and dental service attendance were used as independent variables. The logistic regression model was done in two steps: 1) associations for each variable with the odds of having different stages of periodontitis were studied in a univariate model. 2) Multivariate models were used to study the adjusted associations. The analysis in the univariate and the multinomial regression were done first in the total population, were the ethnicity was one of the confounding variables and secondly in Sámi population and non-Sámi population separately. ‘NSP’ and ‘Stage II’ were used as reference categories. Differences were assessed using Odds Ratio and 95% confidence intervals. In all analyses the significance level was set at 0.05. Data were analysed using the IBM® SPSS® Statistics, version 25.