The present study was approved by the Regional Committee for Medical and Health Research Ethics, (REC- South-East (#2010/401) and conducted from October 2011 until December 2012 at the Faculty of Dentistry, University of Oslo and the Dermatology Department, Oslo University Hospital-Rikshospitalet, Norway.
Study participants
Sample size was determined based on the assumption that the expected prevalence of periodontitis is 30% among individuals with psoriasis compared to 10% among otherwise healthy controls [23]. The calculations were performed using Fleiss method at OpenEpi open source software [24]. Based on a significance level of 0.05, power of 90% and 1:2 ratio of cases to controls, it was decided to include 120 healthy controls and 60 individuals with psoriasis in the study. A higher number of controls compared to psoriasis individuals were selected in order to increase the study power.
Individuals with psoriasis were consecutively recruited from the Department of Dermatology, Oslo University Hospital-Rikshospitalet, Norway. Sixty five individuals received invitation to participate during their regular visit to the dermatological department or from the inpatient unit. Their psoriasis diagnosis was verified by a resident dermatologist (EMS) using Psoriasis Area and Severity Index (PASI) [25]. This index is based on the degree of erythema, infiltration and scaling (respectively: 0 = none, 1 = mild, 2 = moderate, 3 = severe, 4 = very severe) and the extent of involvement of four body skin areas (head, trunk, arms and legs). According to European consensus definitions, plaque psoriasis is graded into mild and moderate to severe disease [26]. Psoriasis is classified as mild if the PASI is below 10, and moderate to severe if it is 10 or above. Patients with a PASI score ≥10 or ongoing systemic or biologic treatment were considered to have moderate to severe psoriasis and received an invitation to attend for clinical and radiographic periodontal examination at the Faculty of Dentistry, University of Oslo.
The inclusion criteria were individuals 18 – 65 years of age, who had suffered from moderate to severe psoriasis for more than five years (see above definition). Persons with known familiar hypercholesterolemia, concomitant inflammatory diseases such as infections, autoimmune disorders, malignancies and pregnancy were excluded. After subsequent invitation by telephone, 51 accepted the invitation and attended the study (79% response rate). During the clinical examination one of the patients was found to be edentulous and was excluded, leaving 50 individuals with moderate or severe psoriasis for further analysis.The control sample was obtained from a list provided by the Norwegian National Population Register containing one thousand randomly selected 35-65-yr-old individuals, from the greater Oslo area. A smaller random sample of 400 individuals was selected in a two stage procedure (200 individuals at each stage) and invited to participate in the study. Only participants born in Norway or a Western country were selected. Of the 242 eligible individuals, 125 attended the study (response rate 52%) (Figure 1). After excluding three participants because of self reported psoriasis as well as one edentulous individual, the final study sample comprised 121 controls. Of those who refused to participate, 56% of controls and 50% of psoriasis individuals agreed to be telephone interviewed about their education, dental health related behavior, smoking habits and reason for not attending the study. Twenty per cent of the non-responders among controls reported that they had been diagnosed with gum disease. Comparisons of the participants and the non-attenders in relation to available background characteristics revealed that there were more current smokers among the non-responders in the control group. In contrast, a lower number of the non-attenders reported that they were current smokers (14% versus 32%) compared with the participants in the psoriasis group.
Data collection
All participants received written information about the study and signed an informed consent. Data collection included self-administered questionnaires as well as clinical and radiographical examination at the Institute of Clinical Dentistry, Faculty of Dentistry. The structured questionnaire included background information (age, gender and education), information on dental care and dental attendance, smoking, self reported general diseases, and medicament use. All items in the questionnaire had closed ended questions, except for age, self reported diseases and medicament use, weight and height, which were open ended. Educational level was recorded as 1) lower secondary education or less, 2) upper secondary or 3) higher education. For regression analyses, the education variable was dichotomized into higher education versus upper secondary or lower education. Smoking status was recorded as never-, former- or current smokers. For regression analyses, the variable was dichotomized into current versus former and never smokers.
Clinical examination included numbers of missing teeth, assessment of probing pocket depth (PPD) and clinical attachment level (CAL) in mm, performed with an LM 52B probe (LM-Instruments Oy, Finland). PPD, CAL as well as the presence of dental plaque (yes/no) and bleeding on probing (BOP) (yes/no) were recorded at four surfaces (MB, B, DB, P/L) of all teeth present, except for the third molars. For statistical analyses, the recordings were computed into the proportion of sites with plaque or BOP. The clinical examination was performed by two trained examiners. The first examiner (RSR) recorded presence of plaque, BOP and PPD, whereas the second (BFH) recorded CAL.
Periodontitis was defined according to the case definitions for surveillance of periodontitis proposed by the Centers for Disease Control and Prevention (CDC) and the American Academy of Periodontology (AAP) [27]. Moderate periodontitis was defined as having ≥2 interproximal sites with AL ≥4 mm (not on same tooth), or ≥2 interproximal sites with PPD ≥5 mm (not on same tooth). Severe periodontitis was defined as having ≥2 interproximal sites with AL ≥6 mm (not on same tooth) and ≥1 interproximal site with PPD ≥5 mm. Individuals who had moderate or severe periodontitis according to the definition, were considered to have the disease.
Six intra-oral digital radiographs, two bitewings on each side and two periapical radiographs of the incisors of both jaws, were taken using Kwik-Bite bite-wing holders (Kerr Corporation, West Collins Orange, CA, USA) and Eggens (Eggen X-ray, Lillehammer, Norway) film holders. Radiographs were obtained by the Digora Digital imaging plate system film (Soredex, Finland ) and a Planmeca intra (Planmeca Oy, Finland) X-ray unit. The images were transferred and stored using Sectra PACS (Sectra AB, Linköping, Sweden) image server digital storage system. The radiographs were magnified using a computer, and radiographic bone loss was registered by one examiner (BFH) using the public domain digital imaging program Image J (Image J, National Institutes of Health, Bethesda, Maryland, USA). Bone loss was registered when the distance from the cemento-enamel junction (CEJ) to the alveolar crest (AC) exceeded 2 mm, in sites where CEJ and AC could be identified clearly. The measurements were made in mm in the interproximal areas of fully erupted canines, premolars, molars, upper central incisors and lower central and lateral incisors. Participants with radiographic bone loss ≥3 mm at one or more sites were considered as having the condition.
The radiographic examination was masked, i.e. the examiner was unaware of if participants had psoriasis or not, whereas the clinical examination could not be regarded as masked since some patients harbored psoriasis lesions visible to the examiner.
The intra-examiner reproducibility of radiographic bone loss was evaluated on the basis of repeat recordings radiographs from 20 randomly selected individuals within a time frame of 14 days , the K-value being recorded to 0.61 for recordings of bone loss in mm and 0.75 for dichotomized outcome (bone loss present/absent) and deemed as good [28].
Statistical analyses
When comparing differences between the psoriasis and the control groups, the Chi-square test was used for categorical variables, and the independent samples t-test for continuous variables. In order to balance for differences in baseline characteristics between the psoriasis individuals and the controls, a propensity score was calculated for each person using logistic regression analysis, regressing psoriasis status (0 = no psoriasis; 1 = psoriasis) on the background characteristics (age, gender and education) [29]. The propensity score was considered as individuals’ probability of having psoriasis, conditional on the background covariates.
A subgroup of control individuals was then selected based on the propensity score values and 1:1 matched to individuals in the psoriasis group. This would allow to compare individuals who based on observables have more similar background characteristics (similar propensity score), except for their psoriasis status.
The association between psoriasis and periodontitis was investigated by using logistic regression analysis for the initial eligible study sample and for the propensity score matched sample. Variables significantly associated with the outcome in bivariate analyses at p-value <0.05, were entered into the regression analyses as independent variables. Multicollinearity diagnostics for independent variables was calculated using the Variance Inflation Factor (VIF) [30]. The VIF was <5 for all associations, indicating absence of multicollinearity which may invalidate the regression analysis. Results were reported as odds ratios (OR) with 95% confidence intervals (CI). The level of significance was set at 5%. Data were analyzed using the SPSS statistical program package (IBM SPSS 20.0, SPSS Inc., Chicago, IL, USA).