Outcomes of Periodontal Treatment in Diabetes Mellitus Patients at Kiruddu Referral Hospital in Uganda. A Prospective Cohort Study

Introduction: Periodontitis is a common complication of diabetes mellitus associated with poor glycemic control. The relationship between periodontal disease and glycemic control in patients with diabetes mellitus remains unclear. This study set out to determine the outcomes of periodontal treatment in diabetic patients in Uganda. Methods: Using a cohort study design, 41 adult diabetic patients with periodontal disease were enrolled and followed up for 3 months. Fasting blood sugar, glycated hemoglobin levels, pocketprobing depth, bleeding and clinical attachment loss at baseline and 3 months after periodontal treatment were determined. Gingival crevicular uid was assessed to detect presence of Porphyromonas gingivalis, Aggregatibacter actinomycetemcomitans, Bacteroides forsythus and Fusobacterium nucleatum. Data were analyzed using R version 4.10. Mean differences were calculated to determine effect of treatment and multiple logistic regressions was used to determine association between Hba1c and the different variables. Results: Of the41 participants, 61% were females while 39% were males. The average age was 49.2 years (S.D = 11.2)while the mean BM1 was 27.8 (S.D = 4.8). Overall, there was a reduction in the average number of teeth with bleeding sites, those with pockets and the frequencies of selected bacteria except Fusobacterium nucleatum at 3 months after periodontal treatment. Following periodontal treatment, 39% of the participants had an improvement in their glycemic control (mean HbA1c reduction of 1.86) with age < 47.1 years, diabetic duration less than 1.6 years, BM1< 27.7 and HBA1c >8.9 at baseline being positively associated with this improvement. Conclusion: Treating periodontal disease in diabetic patients especially those with HBA1c <9.0 can improve glycemic control. However, larger studies need to be conducted to ascertain why some patients fail to attain an improved metabolic control following this intervention.


Introduction
Diabetes mellitus is increasingly becoming a plague to mankind affecting about 6.8% of the population worldwide (1). In 2017, approximately 462 million people were affected by type two diabetes mellitus.
Global prevalence is projected to rise to 7079cases per 100,000 by 2030 (2). About 43.7% of diabetic patients have chronic periodontitis which is associated with poor blood glucose control in this population (3,4). Periodontal disease is more prevalent and severe in diabetic patients than the nondiabetics (5). Periodontitis a risk factor for poor glycemic control and development of diabetic complications (6). Periodontitis and diabetes establish a two way pathway, and each one, if untreated, could promote or exacerbate complications of each other (7). Treatment of periodontal disease in patients with diabetes mellitus has been shown to have an effect on glycemic control (8). Systemic reviews and meta-analysis have supported the effectiveness of periodontal therapy in improving glycemic control and periodontal status in patients with diabetes mellitus and periodontitis. patients. (9,10). Reports from one study indicated that non-surgical periodontal treatment markedly improved glycaemic control with an attributable reduction of 2.2 points of glycated haemoglobin (HbA1c) in poorly controlled T2D patients in a sub Saharan setting (8). Control of periodontal disease may enhance glycaemic control in patients with type 2 diabetes. In turn, improved glycaemic control may contribute to a better control of periodontal disease (11). However, in most studies, effects of periodontitis and its treatment on diabetes metabolic control are not clearly de ned and results of the studies remain controversial. It is recommended to carry out further studies to ascertain this effect (12). The objective of this study was to determine outcomes of periodontal treatment on periodontal status and metabolic control in diabetic patients attending an outpatient clinic at a tertiary hospital in Uganda.

Study setting
The study was conducted from Kiruddu National Referral Hospital located in Makindye division of Kampala district. The Hospital runs a number of specialized clinics, with the diabetic clinic running every Wednesday. The clinic receives about 200 to 300 diabetic patients every Wednesday.

Study design
This was a cohort study involving 59 participants with diabetes mellitus and periodontal disease. 18 participants were lost to follow up. Reasons for loss to follow up included the corona virus pandemic that led to some participants migrating to rural areas, the uncertain political environment as this was period for presidential elections and the covid -19 SOPs like a ban on public transport.
Participant recruitment and study procedure Inclusion criteria were; participants with con rmed diabetes mellitus diagnosis, aged 18 years and above diagnosed with periodontal diseaseand provision of written informed consent.All participantswho were pregnant, those receiving broad spectrum antibiotics or had a history of periodontal treatment in the last 6 months were excluded from the study.
Apre tested study tool developed by the investigators was administered to record socio-demographic characteristics, diabetichistory like duration of diabetes and diabetic medication and factors associated with periodontal disease which included oral hygiene practices, history of smoking. Weight in Kilograms and height in m 2 were measured to calculate the BMI. The values for BMI were recorded as <18: underweight, 18.5-24.9: normal, 25-29.9: overweight and >30: obese (13). Venous blood was collected and analyzed to determine fasting blood sugar levels andHbA1c levels. The cut off value for good glycemic control was 6.5 and values above this were recorded as poor glycemic control (14).
An oral examination was then carried out by a dental surgeon using a dental mirror and a calibrated periodontal probe to determine the Community Periodontal Index value (CPI) and the Clinical Attachment Loss (CAL) from 6 sites namely upper central incisor (11), lower central incisors (31), upper right rst or second molar (16/17), lower right rst or second molar (46/47), upper left rst or second molar (26/27) and lower left rst or second molar (36/37). The data wererecorded in the Oral Health Assessment form for Adults (WHO, 2013). The CPI values were recorded as 0 for normal, 1 for bleeding, 2 for calculus, 3 for pockets less than 4mm and 4 for pockets greater than 4 mm. The CAL levels were recorded as distance from the cemento-enamel junction to the gingival margin and values were assigned as follows 0: 0-3mm, 1: 4-5mm, 2: 6-8mm, 3: 9-11mm and 4: 12mm or more. Samples of gingival crevicular uid were collected using a paper point and transferred to a Cryo vial container containing a buffer. These were then transported in ice packs to the laboratory where DNA extraction was carried out using the Quick-gDNA method by ZymoResearch. PCR detection was then carried out in three stages: Ampli cation, Annealing and Elecrophoresis to determine the presence of four common periodonto-pathogenic bacteria namely Porphyromonas gingivalis, Aggregatibacter actinomycetemcomitans, Bacteroides forsythus and Fusobacterium nucleatum.

Data Analysis
Data was entered in Microsoft excel, cleaned and exported to R (version 4.10) for analysis. The results obtained at baseline and after 3 months of follow up were compared. Mean differences in HbA1c levels, number of teeth with bleeding sites, number of teeth with pockets and four common periodontopathogenic bacteria were determined. The Paired t-test was used to determine the association of the different outcomes with HbA1c levels following periodontal treatment and a p-value of <0.05 was considered statistically signi cant. Multi-level regression analysis was used to determine the factors associated with improvement in glycemic control following periodontal treatment.

Results
Baseline characteristics of the study participants Of the 41 participants, sixty-nine percent were female and the average age was 49.2 years (range = 39.0 years, S.D =11.2 years). Over two thirds of the participants were either overweight or obese. Only one person reported cigarette smoking and he had been doing so for more than 10 years. Seventy-ve percent of the participants had attended either primary or secondary education with only fourteen-point six percent reporting no school attendance. Ninety-seven-point seven percent had periodontitis and the mean HbA1c was 9.0 (SD=2.9). (See Table 1 in appendix i).  Effect of periodontal treatment on selected periodonto-pathogenic bacteria Following periodontal treatment, there was a reduction in proportion of participants with P. gingivalis from 48.7-31.7%, those with A. actinomycemcomitans from 36.6-24.4%, those with B.forsythus from 43.9-12.2% but no change in the proportion of those with F.nucleatum. There was thus a general reduction in the subgingival microbiota following periodontal treatment. (Table 3) Association between the different variables with HbA1c Table 4 provides a summary of the nal model multilevel regression modeling containing the set of variables that were retained in relation to their effect of the HbA1c levels controlling for the participant. What stands out in this nal model was that the presence of Fusobacterium nucleatum (Fn) was associated with a signi cant 2.14 reduction in HbA1c levels keeping all other values in the model constant. Overall, the presence of Porphyromonasgingivalis (Pg) and Bacteroides forsythus (Bf) were each associated with an increase in the HbA1c levels. On the other hand, Aggregatibacter actinomycetemcomitans (Aa) was associated with a decrease in the levels of HbA1c. For these other bacteria it is important to note that these effects were not signi cant. It is also important to note that there was an interaction between the presence of Aggregatibacter actinomycetemcomitans (Aa) and having high random sugar associated with a -0.75 reduction in HbA1c levels (p value = 0.70). On average the intervention led to a 0.23-unit reduction in the levels of HbA1c compared to baseline for all participants. This was not signi cant. Respondents with a high random blood sugar levels had on average 3.00 units more than those with normal random blood sugar.

Discussion
This study reports the outcomes of periodontal treatment in diabetic patients in Uganda namely number of teeth with pockets, those with bleeding sites, frequencies of four periodonto-pathogenic bacteria and metabolic control determined by hbA1c levels. Generally, there was a drastic decline in number of teeth with pockets and those with bleeding sites. In addition, the frequencies of the bacteria decreased as well as an improvement in metabolic control in thirty-nine percent of the participants, three months after periodontal treatment.
Periodontal treatment has been reported to improve periodontal status (15). Brauchle et al (2013) reported a positive effect of periodontal treatment on oral health quality of life with a decrease in probing pocket depth of >3mm from 76% at baseline to 27.6% after treatment (16). Huang et al, 2016 also reported that patients on comprehensive periodontal treatment exhibited shallow pockets, less bleeding on probing and improved PCR (17). Periodontal treatment involves removal of plaque, calculus deposits and granulation tissue there by reducing the occurrence of the signs of gingivitis and periodontitis(18). This is in line with our ndings that noted a decline in number of teeth with pockets by ninety-one percent and a decrease in proportion of participants with periodontitis from ninety-seven-point six percent to thirty-nine-point one percent.
Periodontal treatment whether surgical or non-surgical has been demonstrated to decrease the median relative proportions and frequencies of periodonto-pathogenic bacteria (19). These bacteria have been shown to decrease at 2 months and 4 months following periodontal treatment (20). Our study noted a decrease in the frequencies of all the tested bacteria except Fusobacterium nucleatum. This is in contrast to a study that reported a signi cant decrease in all the tested bacteria following periodontal treatment except for Aggregatibacter actinomycetemcomitans (21). The reasons for this discrepancy are not well explained. One study identi ed smoking as an important factor for differences in changes in subgingival biodiversity following periodontal treatment (22). However, from our study, only one participant was a smoker and this small number cannot be statistically used to explain this variation. Despite these discrepancies, it is basic knowledge that periodontal treatment involves removal of plaque and calculus which are the reservoirs for periodonto-pathogenic bacteria (23). It is thus effective in reducing the proportions of these bacteria thereby improving the morbidity of patients with periodontitis.
Periodontal treatment whether associated with or without antibiotics can improve periodontal status of patients with diabetes mellitus (24) (27). Our study demonstrated an improvement in Hba1c levels in 39% of the participants. The factors associated with this improvement were young age, shorter duration of diabetes mellitus, BMI less than 27.7 and high HBA1c levels at base line.This is partly in agreement with Kaur et al, 2015 who reported that although non-surgical periodontal therapy improved glycemic control and periodontal health in patients with type two diabetes, patients with poor baseline glycemic control had less clinical improvement than those with good glycemic control (28). Actually, a study conducted among veterans with poorly controlled diabetes mellitus reported no effect on glycemic control 4 months following periodontal treatment (29). Our study also identi ed short duration of diabetes mellitus and BMI as one of the factors associated with improvement in glycemic control. This is in agreement with Fiseha et al, 2018 who reported long duration of diabetes and Chetouia et al, 2020 who reported longer duration of diabetes and BMI to be associated with poor glycemic control (30,31). It thus imperative to say that periodontal treatment can improve metabolic control in a number of diabetic patients. However, other factors need to be investigated and adequately addressed so as to improve morbidity of patients living with diabetes mellitus.

Conclusion
Non-surgical periodontal treatment can be effective in improving periodontal status and metabolic control in a signi cant number of patients with diabetes mellitus and periodontitis. However, further studies need to be conducted to ascertain why some patients do not experience an improvement in their metabolic control following periodontal treatment.