Non-surgical periodontal therapy affects metabolic control in diabetics: a randomized controlled clinical trial

Authors


Dr Amir Moeintaghavi
Department of Periodontics
Dental School
Mashhad University of Medical Sciences
Vakilabad Boulevarde
Mashhad 91735-984
Iran
Email: moeentaghavia@mums.ac.ir

Abstract

Background:  Periodontal diseases and diabetes are two common diseases with high prevalence. Many clinicians have accepted the relationship between these two diseases. Some investigators have reported that periodontal treatment may enhance the metabolic control of diabetes. The effects of non-surgical periodontal treatment on metabolic control in people with type 2 diabetes mellitus (DM2) were examined.

Methods:  Forty patients with DM2 and chronic periodontitis [mean age = 50.29 years; mean glycated haemoglobin (HbA1c) = 8.72] were randomly assigned to two groups. The treatment group (n = 22) received full-mouth scaling and root planing, whereas the control group (n = 18) received no periodontal treatment. Gingival index (GI), plaque index (PI), probing pocket depth (PPD), clinical attachment level (CAL), fasting plasma glucose (FPG), HbA1c, total cholesterol (TC), triglyceride (TG) and cholesterol levels were recorded at baseline and compared to data collected three months later.

Results:  The groups did not differ in gender ratio, age or clinical parameters [PPD (p = 0.107), CAL (p = 0.888), PI (p = 0.180)] and biochemical markers at baseline [FPG (p = 0.429), HbA1c (p = 0.304), TG (p = 0.486), TC (p = 0.942), LDL (p = 0.856) and HDL (p = 0.881)]. FPG, HbA1c and clinical parameters differed between the treatment and control groups (p = 0.006, 0.003 and 0, respectively). From baseline to follow-up (after three months), HbA1c levels decreased in the treated group (p = 0.003). In the same time period, FPG, GI, PPD and CAL increased in the control group (p = 0.016, 0.0, 0.0 and 0.004, respectively) but HbA1c did not change significantly.

Conclusions:  Non-surgical periodontal therapy could improve metabolic control in diabetic patients.

Abbreviations and acronyms:
CAL

clinical attachment loss

DM2

diabetes mellitus type 2

FPG

fasting plasma glucose

HDL

high density lipoprotein cholesterol

LDL

low density lipoprotein cholesterol

PI

plaque index

PPD

probing pocket depth

TC

total cholesterol

TG

triglycerides

TNF

tumour necrosis factor

Introduction

Type 2 diabetes mellitus (DM2), the most common type of diabetes, is characterized by hyperglycaemia, hyperlipidaemia and associated complications. The classic major complications of diabetes are microangiopathy, nephropathy, neuropathy, retinopathy, macrovascular disease, delayed wound healing and periodontitis.1 The hyperlipidaemia that accompanies hyperglycaemia involves marked elevation of low density lipoprotein cholesterol (LDL), triglycerides (TG) and omega-6 free fatty acids.2,3 This abnormality in fatty acid metabolism is thought to underlie the development of many diabetic complications.4

One of the leading complications of diabetes, periodontitis, is an infection of the periodontal support tissues.5 There is an interacting, complex relationship between diabetes and periodontitis. Many studies have shown a greater incidence and a greater severity of periodontitis in diabetic patients.6–9 Meanwhile, a number of studies have suggested that periodontitis may actually be a risk factor for diabetic complications as well.1,10–15 Southerland et al.16 proposed a common pathogenesis involving an increased inflammatory response for periodontitis and diabetes. Indeed, patients with periodontitis have increased serum levels of inflammatory cytokines, while diabetic patients have hyperinflammatory immune cells that can aggravate the increased production of inflammatory cytokines.17 This exacerbation can increase insulin resistance and make it more difficult for patients to control their diabetes.17,18

Over a decade ago Christgau et al.19 demonstrated that patients’ keratin, TG and total cholesterol (TC) can be affected to a small degree by periodontal therapy. Observations that periodontal therapy appears to reduce periodontal infection and inflammation suggest that periodontal therapy may facilitate metabolic control of diabetes, improving insulin sensitivity by reducing peripheral inflammatory cytokine levels.17,20 Indeed, among the earlier studies that investigated the effects of periodontal therapy on glycaemic control, a large number reported improved glycaemic control after periodontal treatment,2,21–27 while others did not find a positive effect of periodontal therapy on glycaemic control.11,16,18,28–30 Stewart et al.21 previously reported a decrease in HbA1c level following non-surgical periodontal treatment in diabetic patients, but that study did not compare periodontal parameters to control group data and included patients whose medication regimen changed during the study period. The present study was performed to investigate the effects of non-surgical periodontal therapy on metabolic control in DM2 patients. It was a randomized, controlled clinical trial and included a group of DM2 control subjects who did not receive periodontal treatment during the study period.

Materials and methods

Design

This study was a randomized, controlled clinical trial conducted between June 2007 and September 2008 in Mashhad, Iran.

Patients

Based on studies by Rodrigues et al.2 and Kiran et al.37 (with α = 0.05 and β = 0.2), sample size was calculated as 20 patients in each group. The protocol employed for this study was approved by the Research Council and the Ethical Committee of Mashhad University of Medical Sciences, Iran. This trial was registered at http://www.clinicaltrial.gov and the clinicaltrial.gov identifier was NCT01252082. Eligible patients were selected and recruited from the Parsian Diabetics Clinic and the Mashhad Diabetics Center (Mashhad, Iran). They were then referred to the Periodontics Department of Mashhad Dental School in Iran. Fifty-six patients with DM2 and periodontitis were assessed, and a screening oral examination and medical history review carried out to confirm eligibility. Sixteen patients chose to leave the study; eight did not meet the inclusion criteria and the other eight refused to participate. Finally, 40 subjects gave written informed consent and completed the study. DM2 subjects were selected for this study because DM2 is highly prevalent in the general population and its incidence is increasing.31

The inclusion criteria were as follows: (1) mild to moderate periodontitis according to the criteria of the American Academy of Periodontology;32 (2) diagnosis of DM2 with glycated haemoglobin (HbA1c) values over 7%; (3) no major diabetic complications; (4) blood sugar controlled with glyben glamide and metformin, without insulin administration; and (5) no systemic antibiotic administration or periodontal treatment within the last six months.

The following exclusion criteria were applied: (1) presence of systemic diseases other than DM2 that could influence the course of periodontal disease; (2) intake of immunosuppressive drugs, steroids, hydantoin or non-steroidal anti-inflammatory drugs; (3) tobacco use; (4) pregnancy or intention to become pregnant during the study period; (5) fixed orthodontic appliances; and (6) refusal or inability to give informed consent.

Patients were randomly divided into treatment and non-treatment (control) groups by the study research assistant (KK) using a computer generated random numbers table. All the patients were under strict medical supervision and, as a prerequisite, no additional guidance in managing diabetic status or changes in diet, medication or physical therapy was given so that we could directly evaluate the effects of periodontal treatment on metabolic control.

Peridontal assessments

All subjects underwent a comprehensive clinical examination by a single examiner (AMT) who was an expert periodontist and blinded to the subjects’ group assignment. Intra-examiner reliability was tested by examining three patients in an identical manner one hour apart and observing more than 95% of recordings being within 1 mm. Periodontal parameters were recorded at baseline (day 0) and three months later in both groups. Severity of periodontal disease was evaluated using gingival index (GI), plaque index (PI), clinical attachment level (CAL) and probing pocket depth (PPD). PI was quantified using the O’Leary index in which the four smooth surfaces of all existing teeth were scored and a percentage of surfaces with plaque was recorded.33 Gingival status recordings were made for each tooth according to established GI criteria.34 PPD was measured from the gingival margin with a Williams’ periodontal probe (Hu-Friedy, Chicago, IL, USA). CAL was defined as the distance from the cemento-enamel junction to the bottom of the pocket.

Metabolic assessments

Venous blood samples were taken from each patient and analysed for fasting plasma glucose (FPG), HbA1c, TG, TC, high density lipoprotein cholesterol (HDL) and LDL. Metabolic measurements were performed at baseline and three months later in both groups. All laboratory assessments were performed in the Pars Laboratory (Mashhad, Iran) using Cobas Integra 700 apparatus (Roche Diagnostics, Germany). Pars Laboratory is a well-known private laboratory where all diabetic patients are referred to because of its high standards. Reliability of biochemical measurements were confirmed through credentialed regulations and shown to be within acceptable standards.

Periodontal treatment

All periodontal treatments were performed by a general practitioner (YB) under the supervision of an expert periodontist (HRA). Prior to the first treatment session, patients in both groups received standard oral hygiene instructions, placement of emergency restorations and extraction of unsalvageable teeth. In the treatment group, patients were given local anaesthetic (Lidocaine HCl 2% and Epinephrine 1:100 000 injection, 3–5 single dose cartridges of 1.7 mL; Daroopakhsh, Iran) and then subjected to full-mouth scaling and root planing by using an ultrasonic device (VGE 302k, Juya Electric Co., Tehran, Iran) and standard periodontal curettes with no time limitation. Patients did not receive any further periodontal treatment for three months and medical treatment for diabetes remained unchanged. Control group patients received no periodontal treatment during the study period but after completion of the study, for ethical reasons, they received a full non-surgical and supportive periodontal treatment. Three months after the baseline examinations, all 40 subjects were recalled for a second clinical examination and all parameters that were assessed at baseline were reassessed.

Statistical analysis

Statistical analyses were performed using SPSS 11.5. In each of the analyses, p < 0.05 was regarded as significant. There were no dropouts in both groups so we used per protocol analysis. Data from the baseline and three-month follow-up evaluations were subjected to the Kolmogorov-Smirnov test to determine whether they were normally distributed. Means and standard deviations were calculated for each variable at each time point for each group. Significant group differences over time were calculated using Student’s t-test analysis for continuous data (PPD, CAL, FBS, TG, TC, LDL, HDL) and the Mann-Whitney test for non-parametric data (PI, GI and HbA1c). Differences between baseline and three-month follow-up data for each group were calculated using paired t-tests for continuous data. Wilcoxon’s rank sum test was used to compare non-parametric data values between baseline and the three-month follow-up time point. To make an association measurement between periodontal therapy and metabolic control, we calculated number of well-controlled unchanged and poorly controlled patients in both groups after three months. We considered patients with HbA1c <7% as well-controlled and HbA1c ≥7% as poorly controlled.17 Chi-square test was used to show this association and logistic regression test was performed to evaluate the effect of baseline periodontal parameter (PD) and periodontal therapy on metabolic control.

Results

The subject cohort comprised of 20 (50%) females and 20 (50%) males, with an overall mean age of 50.29 ± 3 years. The mean age of females (48.1 ± 3 years) was similar to that of males (52.48 ± 3 years; p = 0.9). The distribution of patients with respect to gender was similar between the groups, with the treatment group consisting of 13 (59.1%) females and 9 (40.9%) males and the control group consisting of 7 (38.9%) females and 11 (61.1%) males (p = 0.341). In 2 patients we had to extract hopeless teeth; 1 in the treated group with an unrestorable third molar and 1 in the control group with a periodontally hopeless premolar. One patient in the treatment group suffering from an endodontic problem received required treatment.

Baseline assessments

All assessed variables were similar between control and treated groups at baseline: PPD (p = 0.107), CAL (p = 0.888), PI (p = 0.180), FPG (p = 0.429), HbA1c (p = 0.304), TG (p = 0.486), TC (p = 0.942), LDL (p = 0.856) and HDL (p = 0.881).

Effects of treatment on periodontal assessments

Periodontal therapy significantly improved all periodontal parameters in the treated group (Table 1). For the treated group, from the baseline assessment to the three-month follow-up assessment, PPD decreased from 2.31 ± 1.3 to 2.21 ± 1.2 (p = 0.012), CAL decreased from 3.14 to 2.8 (p = 0.00), PI decreased from 88.9% to 63.22% (p = 0.005) and GI decreased from 1.86 ± 0.83 to 1.24 ± 1.03 (p = 0.00). In the same time period, PPD (p = 0.00), GI (p = 0.04) and CAL (p = 0.004) all increased in the control group, while PI remained similar between the two time points for controls (p = 0.06).

Table 1.   Comparison of periodontal parameters (mean ± SD and mean change) in treated and control groups
GroupnPeriodontal parameterPre-txPost-txP vs. pre-txMean changeP vs. control
  1. tx = treatment; P = p-value.

Control PPD (mm)2.06 ± 0.242.33 ± 0.30.000.27 
18CAL (mm)3.1 ± 1.053.47 ± 1.440.0040.37 
 PI (%)94.44 ± 6.6287 ± 18.70.06−7.44 
 GI1.15 ± 0.511.723 ± 0.480.0490.57 
Treatment PPD (mm)2.31 ± 0.652.21 ± 0.60.012−0.10.00
22CAL (mm)3.14 ± 1.082.8 ± 1.090.00−0.340.00
 PI (%)88.9 ± 17.3863.22 ± 21.130.005−25.680.002
 GI1.867 ± 0.831.24 ± 1.030.00−0.620.00

Effects of treatment on metabolic assessments

As shown in Table 2, patients in the treatment group exhibited a decrease in FPG from the baseline assessment (170.95 ± 52.4) to the three-month follow-up assessment (153.45 ± 44.44; p = 0.006). HbA1c levels decreased from 8.15 ± 1.18 to 7.41 ± 1.18 after treatment (p = 0.003). TG, TC, LDL and HDL were not altered by the treatment (p = 0.49, p = 0.33, p = 0.30 and p = 0.08, respectively). In the control group, FPG increased from 159.22 ± 37.01 to 169 ± 38.95 (p = 0.016), while HBA1c, TG, TC, HDL and LDL values remained not significantly changed after three months. Our data analysis showed that periodontal therapy was associated with decreased FPG and HbA1c levels, but no changes in plasma lipid levels (Table 2).

Table 2.   Comparison of metabolic parameters (mean ± SD and mean change) in treated and control groups
GroupnMetabolic parameterPre-txPost-txP vs. pre-txMean changeP vs. control
  1. tx = treatment; P p-value.

Control FPG (mg/dl)159.22 ± 37.01169 ± 38.950.0169.78 
 HBA1c (%)8.72 ± 2.228.97 ± 1.820.2630.25 
18TG (mg/dl)150.94 ± 27.21147.44 ± 29.180.706−3.5 
 TC (mg/dl)192.44 ± 27.2197.16 ± 27.10.464.72 
 LDL (mg/dl)117.11 ± 34.17114.55 ± 38.190.776−2.56 
 HDL (mg/dl)46.33 ± 12.3943.88 ± 9.710.55−2.45 
Treatment FPG (mg/dl)170.95 ± 52.4153.45 ± 44.440.006−17.50.00
 HBA1c (%)8.15 ± 1.187.41 ± 1.180.003−0.740.003
22TG (mg/dl)137.81 ± 81.67129.95 ± 57.880.495−7.860.773
 TC (mg/dl)191.81 ± 26.81185.63 ± 30.920.337−6.180.231
 LDL (mg/dl)118.81 ± 24.9112.68 ± 31.610.309−6.130.73
 HDL (mg/dl)45.81 ± 9.1844.04 ± 6.70.083−1.770.861

In the control group after three months, 5 patients were considered as well-controlled, 9 patients were poorly controlled and in 4 patients HbA1c remained unchanged (unchanged metabolic control). In the treatment group, 11 patients were well-controlled and 11 patients remained unchanged. Chi-square test showed a significant association (p = 0.001) between periodontal therapy and metabolic control (Table 3). Logistic regression analysis showed that regardless of pre-treatment pocket depth, periodontal therapy was associated with a decrease in HbA1c (p = 0.005).

Table 3.   Association measurement between periodontal therapy and metabolic control
GroupNMetabolic control
Well-controlledUnchangedPoorly controlled
  1. Well-controlled: HbA1c <7%.

  2. Poorly controlled: HbA1c ≥7%.

  3. Pearson Chi-square: 14.259, p = 0.001.

Control18549
Treatment221111 

Discussion

The present study showed significant decreases in HbA1c and FPG three months after non-surgical periodontal therapy in patients with DM2, while blood lipid parameters remained unchanged. These findings suggest that a reduction of periodontal infection can reduce HbA1c levels within a short period of time and thus may improve metabolic control in DM2 patients. The outcome of this study corroborates prior evidence supporting an interaction between periodontal status and diabetic metabolic control,4,11,35 and supports the hypothesis that a successful periodontal treatment can improve glucose metabolism. These findings document a direct interrelationship between periodontal conditions and metabolic parameters in DM2 patients, extending earlier studies reporting a similar relationship that did not consider DM2 specifically.28,36 Moreover, our findings are consistent with prior positive responses to non-surgical periodontal therapy in persons with DM2 reported by Westfelt et al.28 and Kiran et al.37 Our findings contradict those of Rodrigues et al.2 who reported no difference in periodontal parameters in DM2 patients and healthy controls three months after periodontal therapy.

Periodontal disease may affect insulin signalling through pro-inflammatory mediators. The highly vascularized inflamed periodontium can be a source of inflammatory mediators, such as tumor necrosis factor (TNFα), which can affect glucose and fat metabolism.38,39 The pro-inflammatory cytokine TNF impairs insulin signalling by increasing adipose secretion of free fatty acids. There is a consensus that this process strengthens glycaemic control in diabetic patients by raising insulin resistance. Accordingly, periodontal therapy might improve glycaemic control by decreasing pro-inflammatory mediators.

There is growing evidence in the literature supporting the notion that non-surgical periodontal therapy can improve glycaemic control. Kiran et al.37 observed an improvement in periodontal parameters and a decrease in HbA1c levels three months after mechanical periodontal therapy, administered with no changes in medical therapy or diet, in DM2 patients with periodontitis. They suggested that the improvement in the HbA1c values could have been due to reduction in GI and bleeding on probing. Yang et al.40 found decreases in TNF, HbA1c and periodontal parameters in DM2 patients after periodontal therapy, providing support for the possibility that periodontal therapy may reduce HbA1c values by reducing TNF concentrations in DM2 patients with periodontitis.

On the contrary, Promsudthi et al.41 did not observe a reduction in HbA1c levels following mechanical periodontal treatment combined with systemic doxycycline. In contrast, Grossi et al.10 and Iwamato et al.22 reported decreases in HbA1c levels which they attributed to reduced TNF levels in circulation in response to systemic doxycycline, which decreased periodontal infection and inflammation.

In this study, we observed decreased HbA1c values in patients whose diabetes was moderate to poorly controlled. Ryan42 suggested that more extensive periodontal treatment, such as surgery or local/systemic antibiotic therapy, may be needed in order to improve glycaemic control in poorly controlled diabetic patients. Indeed, many factors may influence the short-term glucose level of diabetic patients, including their medical care. To determine the relative contribution of periodontal therapy to glycaemic control, our study as a prerequisite included only patients who did not have any change in their diabetic control regimen during the three-month study period. We also found a significant association between periodontal therapy and metabolic control.

In the present study, all probing depths in both groups were measured at <5 mm. Therefore, unlike many other studies, highly variable periodontitis severity was not a confounding factor. Our analysis showed that the positive effect of periodontal therapy on metabolic control was irrespective of pre-treatment probing depth. Hence, the improved HbA1c measures seen after treatment can be attributed to diminished gingivitis.

There were two limitations in this study worth noting. Firstly, while the strict inclusion and exclusion criteria helped minimize confounding factors, it also limited the patient sample to a relatively small size. Secondly, we only examined subjects at a single three-month follow-up time point. Researchers have proposed different periods for reassessing periodontal parameters following non-surgical treatment. Morrison et al.43 and Lowenguth et al.44 suggested a one-month period while Badersten et al.45 believed that maximum tissue reattachment in deepened periodontal pockets (>4 mm) occurs 4–5 months after treatment and can continue to increase for 12 months in 12 mm-deep periodontal pockets. Clarifying the effect of non-surgical periodontal therapy on glycaemic control will require further studies with larger sample sizes and more longitudinal data.

Conclusions

The present study demonstrated that non-surgical periodontal therapy can effectively decrease FPG and HbA1c levels in moderate to poorly controlled diabetic patients. Periodontal therapy and follow-up might be considered in the treatment plan of DM2 patients, especially those who have poor metabolic control despite implementation of numerous medical interventions. Preventive periodontal regimens for diabetic patients should be sufficiently intense and sustained as to eliminate periodontal inflammation and should be closely coordinated with the patient’s overall clinical diabetic management.

Acknowledgement

This study was supported by a grant from the Vice Chancellor for Research, Mashhad University of Medical Sciences, Iran.

Ancillary