Trial design

The trial was a single-masked, parallel-design, single center, randomized clinical trial that enrolled participants from outpatient medical and dental clinics in Japan following the CONSORT guidelines. The study protocol was approved by the Ethics Committee, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences and Okayama University Hospital (no. d11004). Physicians enrolled the participants during the diabetes care visits. All patients provided written informed consent. An independent center reviewed the safety data throughout the trial. Code breaking was performed after the final statistical analysis. Before and after the trial, there were no changes in eligibility criteria and outcomes.

Randomization

Randomization was stratified by levels of HbA1c (< 8% vs. ≥ 8%), insulin (use vs. no use) and the number of medications (≤ 2 vs. >2). Each selected patient received a code number and one of the study coordinators (T.M.) used a computer-generated table to randomly allocate patients to one of the two groups (control and periodontal treatment group as below) (allocation ratio; 1:1) [33].

Blinding

Study personnel, including the periodontal examiners, laboratory personnel who performed the HbA1c analyses and the investigator responsible for the data analysis were blinded to the treatment assignment.

Participants

Participants were recruited between April 2014 and March 2016 at the Nephrology, Diabetology and Endocrinology Department of Okayama University Hospital. The inclusion criteria (see below) were selected to maximize the chance that a participant would complete the 6-month trial. Men and women (≥ 30 years) were eligible if they had physician-diagnosed T2DM (diagnosed at least 2 months prior to the study), reported that they were able to make hospital visits throughout the trial, were in the care of a physician for their diabetes, and agreed to not change their diabetes medications during the trial unless medically indicated. Patients also required a diagnosis of mild to advanced chronic periodontitis, defined as ≥ 2 interproximal sites with clinical attachment level (CAL) ≥ 3 mm and ≥ 2 interproximal sites with probing pocket depth (PPD) ≥ 4 mm (not on the same tooth) or one site with PPD ≥ 5 mm [34]. Radiographs were used to confirm a diagnosis of chronic periodontitis [35]. Exclusion criteria included pregnancy, inappropriate status for the trials, such as limited life expectancy and diabetes-related emergency, and receiving periodontal treatment in the prior 6 months [35].

Intervention

Initially, all participants received one session of supra-gingival plaque removal and oral hygiene instructions. The patients of the periodontal group were treated by a protocol adapted from the Japanese Society of Periodontology. The non-surgical periodontal treatment consisted of at least 60 minutes of scaling and root planing using curettes (BioGent Curette, Hu-Friedy, Chicago, IL, USA; Scaler YB, YDM Corporation, Saitama, Japan) and an ultrasonic instrument (Varios 750, NSK, Tochigi, Japan) during 2 or more sessions completed within 42 days after the baseline visit [36]. Completeness of therapy was assessed by the study therapists and confirmed by a study dentist. During the treatment, the therapists provided oral hygiene instructions, professional toothbrushing and mechanical tooth cleaning. Three and 6 months after the baseline visit, participants in the treatment group received periodontal maintenance therapy including oral hygiene instructions and oral prophylaxis for approximately 1 hour during a single session. Participants in the control group received only oral hygiene instructions at the baseline and 3 and 6 month visits. After the study completion, the control group participants received scaling and root planing.

Outcome assessment

The primary study outcome was the change in HbA1c levels from baseline to 3 months. Secondary outcomes included changes in oxidative stress balance, glycated albumin, QOL and clinical periodontal parameters from baseline to 3 months and baseline to 6 months. Change in diabetes medications at 3 and 6 months, and the need for periodontal rescue therapy and diabetes rescue therapy were evaluated as safety outcomes. A change in medication was defined as more than 2-fold change in dosage for a hyperglycemic drug, more than 10% change in dosage for insulin, or change in an oral hyperglycemic agent or insulin [35].

Sample size calculation

Sample size was estimated assuming a 0.8% (SD, 0.8%) or greater reduction in HbA1c levels from baseline to 3 months in the periodontal treatment group compared with the control group [37]. Based on the data, it was determined that 17 subjects per group would be necessary to provide an 80% power with an alpha of 0.05, 2-tailed and unpaired t test. Assuming an attrition rate of 15% [38], the planned sample size was 40 participants (20 in each treatment group).

Data collection

Data were collected by trained and certified study personnel. Study personnel recorded medical history, medication use, demographics, and lifestyle information including sex, age, height, weight, drinking habit, smoking habit, frequency of toothbrushing, use of interdental brush and also administered the Diabetes Therapy-Related QOL (DTR-QOL) questionnaire [39]. The DTR-QOL includes 4 factors: "burden on social activities and daily activities" (factor 1), "anxiety and dissatisfaction with treatment" (factor 2), "hypoglycemia" (factor 3), and "satisfaction with treatment" (factor 4) [39].

Oral examination

The periodontal examiners (dentists) performed the following oral examination. PPD and CAL were determined at six sites (mesio-buccal, mid-buccal, disto-buccal, mesio-lingual, mid-lingual and disto-lingual) on all teeth using a color-coded probe (CP-11 Color-Coded Probe, Hu-Friedy). Sites that bled upon gentle probing (25 g probing force) were recorded, and the proportion of sites with bleeding on probing (BOP) and the number of BOP-positive teeth were measured in each subject. Plaque Control Record was measured after erythrosine staining, and recorded with respect to their relative location to the gingival margin at four sites (mesial, distal, buccal and lingual) around each tooth [40]. All clinical procedures were performed by two calibrated dentists. To check the intra- and inter-examiner agreement, measurements of PPD and CAL were recorded and repeated within a 2-week interval in two randomly selected chronic periodontitis patients. Data were analyzed with the non-parametric κ test and intra-class correlation was determined. The κ coefficients for intra- and inter-examiner and intra-class correlation coefficients were >0.8.

Laboratory measures

At the clinical laboratory of Okayama University Hospital, venous blood samples were collected prior to periodontal measures or therapy. Fresh whole-blood samples were refrigerated and HbA1c levels were analyzed by high-performance liquid chromatography (Tosoh HLC-723G8, glycated hemoglobin analyzer, Tosoh Medics Inc., Tokyo, Japan). Serum samples were snap frozen and shipped on dry ice for analysis of levels of glycated albumin, lipids and creatinine by enzymatic methods using the automatic analyzer (JEOL BM 8040 and JCA-BM 6050, JEOL Ltd., Tokyo, Japan). High sensitive CRP was measured by an ELISA kit (High Sensitive Enzyme-Linked Immunosorbent Assay (ELISA) Kit for CRP, Cloud-Clone Corp., Houston, TX, USA). Levels of serum reactive oxygen metabolites (ROMs) (whole oxidant capacity of serum against N, N-diethylparaphenylendiamine in acidic buffer) were measured using a spectrophotometer (Diacron International, Grosseto, Italy), as reported previously [41]. Briefly, 20 μL serum samples were gently mixed with 1.2 mL acetate buffer (pH 4.8) in a cuvette, and 20 μL of chromogenic substrate (N, N-diethylparaphenylendiamine) was added. The cuvette was immediately incubated in the thermostatic block of the analyzer for 5 minutes at 37°C. Absorbance at 505 nm was automatically recorded. The Carratelli unit (CARR U), where one CARR U corresponds to 0.08 mg/dL hydrogen peroxide, was used as the measurement unit. In order to determine total serum anti-oxidant capacity, data from the OXY-adsorbent test were collected using a spectrophotometer (Diacron International) [42]. This test evaluates the capacity of serum to oppose the massive oxidative action of a hypochlorous acid (HClO) solution. Briefly, 10 μL of standards or samples, previously diluted 1:100 with distilled water, were added and mixed with 1 mL of the HClO solution. After 10-min incubation at 37°C, 10 μL of chromogenic mixture (solution provided in the kit) was added. Absorbance at 505 nm was measured immediately by the spectrophotometer. Total anti-oxidant capacity was expressed as micromoles of HClO consumed by 1 mL of sample (μmol HClO/mL). Furthermore, standardized values of the ROM and OXY-adsorbent tests were used to represent the Oxidative-INDEX as oxidative stress balance [43]. The Oxidative-INDEX was calculated by subtracting the OXY standardized variable from the ROM standardized variable. High Oxidative-INDEX values indicate high oxidative stress in the blood.

Adverse events and safety monitoring

Oral symptoms were recorded 2 weeks following periodontal treatment (periodontal treatment group) or baseline for the control group. There were no cases in which rescue therapy was administered to patients with acute changes in periodontitis. After the trial, patients were referred for follow-up periodontal care or additional treatment as needed.

Statistical analysis

Data analysis was performed using the Statistical Package for the Social Sciences (SPSS version 19, IBM, Tokyo, Japan). Data were summarized using means (± SDs) for continuous variables and frequencies (percentages) for categorical variables. We used an analysis of covariance (ANCOVA) for primary outcome and for secondary outcomes using the intention-to-treat analysis. The ANCOVA model included use of insulin, number of diabetes medication and HbA1c as covariate. Adjusted differences and 95% confidence intervals (CIs) were determined. P values were based on t-tests comparing mean changes between the control and periodontitis groups with Bonferroni’s method from ANCOVA. Missing data were managed according to the last observation carried forward analysis in accordance with the guidelines [44].

A per-protocol analysis based on data available both at baseline and the 3/6-month visits was also performed without imputation. Subgroup analyses were preplanned for HbA1c levels. In the subgroup analysis, we focused on moderately poor glycemic control (HbA1c; 7.0 to 10.0%) [35, 45]. To check the attrition bias, we compared the baseline characteristics between retained and dropout groups using chi-squared test and t-test. The significance level was p < 0.05.

Retention of participants

As shown in Fig 1, 46 individuals were screened and 40 were randomized between April 2014 and March 2016. When the number of participants reached 40, the recruitment was ended.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of the study design.

https://doi.org/10.1371/journal.pone.0188171.g001

Of the 40 randomized participants, 37 (92.5%) were included in the primary analysis, and 28 (70.0%) completed the study; retention in the periodontal treatment group was 15/20 and 13/20 in the control group. Baseline characteristics were similar between the two groups (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of participants at baseline.

https://doi.org/10.1371/journal.pone.0188171.t001

Primary outcome

The target 3-month reduction of HbA1c levels of 0.8% or greater was not achieved. In the intention-to-treat analysis of the primary outcome, change in HbA1c levels at 3 months did not differ significantly between the periodontal treatment and control groups (Table 2). A per-protocol analysis evaluating the change in HbA1c levels also did not reveal between-group differences at 3 months (S1 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. General conditions and periodontal parameters at follow-up in the primary analysis.

https://doi.org/10.1371/journal.pone.0188171.t002

Secondary outcomes

A per-protocol analysis evaluating the change in HbA1c levels also did not reveal between-group differences at 6 months (S1 Table).

Change in glycated albumin levels also did not reveal between-group differences at 3 months in the intention-to-treat analysis (Table 2). Results at 6 months and in the per-protocol analysis were also similar (Table 2 and S1 Table).

In the intention-to-treat analysis of periodontal parameters, changes in mean PPD and mean CAL at 3 and 6 months significantly differed between the periodontal treatment and control groups (Table 2). Changes in the percentage of CAL≥4mm and BOP significantly differed between the periodontal treatment and control groups at 6 months (Table 2).

In the intention-to-treat analysis of oxidative stress, changes in oxidative-INDEX significantly differed between the periodontal treatment and control groups at 3 months (Table 2). Results in the per-protocol analysis were also similar (S1 Table).

At 3 months, the change in factor 4 (satisfaction with treatment of diabetes) in DTR-QOL significantly differed between the periodontal treatment and control groups (Table 2). The result in the per-protocol analysis was also similar (S1 Table).

We also compared the baseline characteristics between retained and dropout groups. There were significant differences in some parameters between the retained and dropout groups (S2 Table).

Of the 30 participants with medication data available at all study visits, 14 of 15 (93.3%) in the periodontal treatment group and 13 of 15 (86.7%) in the control group had no protocol-defined changes in diabetes medications during the study.

Sub-group analysis

Change in HbA1c levels at 3 months in the sub-group [moderately poor glycemic control (HbA1c; 7.0 to 10.0%)] tended to show the difference between the periodontal treatment and control groups but did not reach the significant level (P = 0.070) (Table 3). Changes in mean PPD at 3 and 6 months significantly differed between the periodontal treatment and control groups (Table 3). The results of counterpart were shown in the S3 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. General conditions and periodontal parameters at baseline and follow-up in the sub-group analysis [moderately poor glycemic control (HbA1c; 7.0 to 10.0%)].

https://doi.org/10.1371/journal.pone.0188171.t003

Harms

The non-surgical periodontal treatment was a low-risk study, and no study-related serious adverse events occurred in all patients. No patients in the two groups required generalized periodontal rescue therapy during the study. In the control group, there were no patients who received the non-surgical periodontal treatment during the study.