Influence of obesity on the outcome of non-surgical periodontal therapy - a systematic review

Background Obesity and periodontitis are important chronic health problems. Obesity is associated with an increased prevalence of periodontitis. Whether obesity also affects the outcome of non-surgical periodontal therapy is to date still unclear. Methods A systematic review of studies referenced in SCOPUS, MEDLINE, PubMed, Cochrane, CINAHL, Biosis and Web of Science was performed. Titles, abstracts and finally full texts were scrutinized for possible inclusion by two independent investigators. Quality and heterogeneity of the studies were assessed and the study designs were examined. Probing pocket depth reduction was analyzed as primary surrogate parameter for therapeutic success after non-surgical periodontal therapy. Results One-hundred-and-fifty-nine potentially qualifying studies were screened. Eight studies fulfilled the inclusion criteria and were analyzed. Three of eight studies failed to show an influence of obesity on pocket depth reduction after non-surgical therapy. The remaining five studies documented a clear negative effect on the outcome of non-surgical periodontal therapy. The finally included studies did not correspond to the highest level of quality (RCTs). Due to the heterogeneity of the data a meta-analysis was not possible. Conclusion The literature on the effect of obesity on the treatment outcome of non-surgical periodontal therapy remains controversial. The data, however, support that obesity is not only a factor associated with poorer periodontal health but might also result in inferior response to non-surgical treatment of periodontitis.


Background
The prevalence of obesity is increasing worldwide and is becoming one of the most important health hazards [1], as obesity is highly associated with increased overall morbidity and mortality [2].
Obesity is defined with a body mass index (BMI; body weight in kilogram divided by the square of the height in meters (kg/m 2 )) of at least 30.0 kg/m 2 [3], whereas overweight is defined with a BMI of 25-29.9 kg/m 2 . Normal weight is characterized by a BMI ranging between 19 to 24.9 kg/m 2 [4]. In this context, BMI seems a valuable parameter to predict obesity-related disease risks in a wide range of populations [2]. There are, however, some limitations: Firstly, risk assessment by BMI is less applicable in persons over 65 years of age because they generally have a higher body fat content for the same BMI. Secondly, the abdominal (central, visceral, android) type of obesity, which is more often seen in men, is associated with higher morbidity than the rather female type of gluteofemoral (peripheral, gynoid) obesity and, thirdly, the BMI cut-off points for overweight and obesity are too high for Asian people [2]. In addition, current large studies have indicated that measurement of waist circumference (WC) or waist-to-hip-ratio (WHR) may be a better disease risk predictor than BMI [5,6]. There is, however, currently intensive research and debate as to whether BMI, WC, WHR, or all of them should be used to assess disease risk [2].
For the purpose of this systematic review, however, only BMI is the most frequently reported data of obesity in a large number of studies.
Adipose tissue contains usually 5-10 % macrophages, but the adipose tissue of obese patients shows up to 60 % macrophage infiltration [4]. Adipocytes secrete bioactive molecules called adipokines, that can modify or trigger inflammation and fat metabolism locally or systemically as signaling molecules to liver, muscle and endothelium [4]. Therefore, the adipose tissue can be considered as an important metabolically active endocrine organ [4].
This explains how obesity acts as a risk factor for several chronic diseases: Hypertension, type 2 diabetes, dyslipidemia, and coronary heart disease are so closely related to obesity that obesity itself is often considered to be a systemic disease. This disease also affects dental health [7]. Accordingly obese persons require attention of physicians and dentists [2].
Among dental pathologies, periodontitis is a very common, primarily bacterial inflammatory disease, which destroys teeth surrounding soft tissues and bone. It leads to pocket formation and ultimately to loss of teeth if no effective treatment is applied [8]. Periodontitis is no longer considered only an oral health issue but also a public health problem, as it constitutes a risk factor for cardiovascular conditions, poor glycemic control in diabetics and adverse outcome of pregnancy [4,8]. These correlations coincide with obesity and general health.
Recently, it has been suggested that obesity is a possible risk factor for periodontitis [8]. One study identified obesity even as the second strongest risk factor for periodontitis preceded only by smoking [9]. The first report on the relationship between obesity and periodontal disease appeared in 1977. Perlstein and co-workers [10] found greater alveolar bone resorption in obese than in non-obese rats. Under healthy oral conditions, obesity itself did not promote periodontal damage, but in the presence of bacterial plaque accumulation periodontal inflammation was more severe in obese than in nonobese animals. With concomitant arterial hypertension, plaque accumulation caused even more pronounced periodontal destruction than with obesity alone. These results suggest that a combination of risk factors, such as the one defined by the metabolic syndrome, elicit a more severe periodontal effect [10,11]. Chaffee et al. [12] found in their meta-analysis an increased prevalence odds ratio for obesity among subjects with periodontal disease of approximately one-third, a greater mean clinical attachment loss (CAL) among obese individuals, a higher BMI among subjects with periodontal disease, and a trend for linear increase in the odds of periodontal disease with increasing BMI [4,12]. Finally the association reported between obesity and periodontitis was less strong than that reported between periodontal disease and adverse pregnancy outcomes [12,13] or cardiovascular events [12,14]. There seems, however, to be a stronger obesity-periodontitis association in women, non-smokers and younger individuals than in the general adult populations [12]. In addition, smoking remains another well-studied predisposing factor for periodontitis [12,15,16]. Thus, BMI and smoking share a complex relationship [17]. This relationship can be inverse in certain populations [12,18,19].
The biological mechanism by which obesity predisposes to periodontitis is not fully understood [8]. Compared to individuals with normal weight individuals with obesity have higher levels of circulating tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6), which are also secreted from adipose tissue and are involved in the pathophysiology of both obesity and periodontitis. Not surprisingly, serum levels of these cytokines decrease with loss of weight [20].
The objective of this systematic review was to study the hypothesis whether the clinical outcome, in terms of pocket depth reduction, after non-surgical periodontal therapy in non-obese is better than in obese individuals. To verify this hypothesis, we systematically reviewed all retrievable, qualitatively adequate clinical investigations, which focused on this topic.

Methods
The review was conducted according to the PRISMA criteria [21]. The research question was explored using the PICO method [22). The focused question addressed was: Does non-surgical periodontal therapy (I) have a different outcome in obese chronic periodontitis patients (P), than in non-obese chronic periodontitis patients (C), regarding periodontal pocket depth reduction as the main clinical periodontal parameter (O).

Search strategy and review process
An electronic search of SCOPUS, MEDLINE, PubMed, Cochrane, CINAHL, Biosis and Web of Science was carried out considering articles published up to January 2016 in English or German language. The search was performed in two steps. The first electronic search started at 20.11.14 and an update has been done at 5.1. 16. This is shown in Table 1. For the database search, a combination of subject headings (MeSH terms and CINAHL headings) and free text search was used. An example of a detailed strategy (Medline/OvidSP) is shown in Table 2.
The same search protocol was applied to all databases.
Two of the authors (FAG, PRS) screened the titles for potential eligibility according to the inclusion criteria. Based on the abstract screening, 18 studies were selected for full text review. Scores were independently allocated by both authors to each publication according to their suitability for the present review (see inclusion criteria). Any discrepancies were resolved by consensus.

Inclusion criteria
To be included studies had to be clinical interventional studies regarding the outcome of non-surgical periodontal therapy in obese or non-obese patients. The studies had to display the diagnosis of chronic periodontitis. Key parameters to be reported were data for pocket probing depth (PPD) and BMI.

Exclusion criteria
Studies were excluded for the following reasons: animal studies, case reports, commentaries, unsuitable exposure or outcome measures, confounding medical diagnoses (e.g. pregnancy or any systemic disease, such as diabetes, in addition to metabolic syndrome), confounding systemic medical treatments such as immunosuppressive treatments, cortisone or antibiotic treatment as well as confounding local treatments such as treatment of perimucositis, gingival overgrowth or surgical periodontal treatment. Studies including either the diagnosis of aggressive periodontitis or of peri-implantitis were excluded as well.

Outcome measures
The primary outcome measure is PPD after non-surgical periodontal treatment.

Data extraction
A list with exclusion reasons for each paper was generated. Total number of patients, demographic data, origin of study, outcome measurements 2, 3, 6 and 12 months after therapy and the impact of obesity on the treatmentoutcome were extracted. In addition, the exact definition of chronic periodontitis, the assessment of the periodontal disease and the number of smokers included in the studies were summarized. Data on the individual definition of obesity and the systemic examinations were also collected. Nonsurgical periodontal treatment measures, treatment time, periodontal maintenance and adverse events were also recorded for each study separately (Tables 3, 4, 5, 6 and 7).
The quality of the included studies was assessed through the Newcastle-Ottawa Quality Assessment Scale (Table 8).

Selection of studies
Initially, 159 studies were identified by electronic search by the two reviewers (FAG, PRS). Full text analysis of the 18 potentially qualified reports led to exclusion of 10 other studies. Additional five titles [23][24][25][26][27] were identified by hand search but after full text analysis, all these articles had to be excluded based on the inclusion and exclusion       (Table 3). Therefore, eight publications [8,[28][29][30][31][32][33][34] from the electronic and hand search fulfilled the criteria. However, it was not possible to compare the raw data so that we had to reduce our analysis to a qualitative analysis. This process is summarized in a flow-chart ( Fig. 1).

Summary of studies: characteristics (PICO)
Total number of patients, demographic data, origin of study, outcome measurements 2, 3, 6 and 12 months after therapy and the impact of obesity on the treatmentoutcome are summarized in Table 4. Definition of chronic periodontitis, periodontal assessment and the amount of smokers included are depicted in Table 5.
The definition of obesity and systemic examination were summarized in Table 6. Non-surgical periodontal treatment, treatment time, periodontal maintenance and limitations are shown in Table 7.

Population
Of the eight finally analyzed studies, clinical trials comprised 516 participants. One study [8] enrolled only women. The prevalence of smoking among male patients affected with periodontitis was so high that Al-Zahrani and co-workers [8] were unable to correct for smoking in the male population. Accordingly, only women were included in their study. Overall the studies comprised between 26 up to 260 subjects ( Table 4).
Two papers (8,34] reassessed the PPD 2 months after therapy. Two studies [29,30] reassessed their patients after 3 months. Another two papers [28,31] reevaluated the patients after 3 and 6 months and one study [32] reassessed the PPD three, six and 12 months after therapy. Only Lakkis et al. [33] measured the periodontal pocket depths already 4 to 6 weeks after nonsurgical therapy (Table 3). Oral hygiene was instructed additionally to the non-surgical periodontal therapy in all studies (Table 7).
Due to the heterogeneity of the study designs with respect to outcome measures and treatment protocol, as well as variation in study population, sample size, and/or    CI confidence interval, CPNW normal weight patients with periodontitis, CPO obese patients with periodontitis, kg kilogram, m meter, mm millimeter, Nb Δ PD number of improving PD, n.r. not reported, PD probing depth, PHNW normal weight periodontally healthy control subjects, PPD periodontal pocket depth, p probability, SD standard deviation, (n.) sig. (not) significant, stat. statistically, WHO world health organisation statistical methods, a statistical synthesis of the results of the included studies was not possible. So the authors decided to analyze the papers on a qualitative way [22]. A meta-analysis was not possible.

Outcome
Generally obese patients were found to have deeper periodontal pockets at baseline in all studies.
Three [8,29,30] of the eight papers [8,[28][29][30][31][32][33][34] reported no major negative impact of obesity on response to periodontal therapy in terms of PPD reduction (mm). Al-Zahrani et al. [8] assessed the reduction of PPD (mm) comparing obese with normal-weighted women. There was no statistically significant effect of obesity on treatment outcome. Duzagac et al. [29] assessed the clinical response to non-surgical periodontal treatment, according to the severity of periodontitis based on probing depth < 4 mm vs. ≥ 4 mm. Patients with and without obesity showed similar clinical healing in terms of percentage and number of sites with probing depth < 4 mm and ≥ 4 mm. So they failed to show an effect of obesity on the treatment outcome dependent on the severity of the disease. Eldin et al. [30] also found no effect when comparing an overweight group with an obese group.
The difference between the groups in reduction of PPD was not significant (Table 4). Five [28,[31][32][33][34] of the eight papers showed a negative effect of obesity on the healing after non-surgical periodontal therapy. Bouaziz et al. [28] revealed that normal-weight patients had a better response to periodontal treatment than obese patients. This effect was specially observed for moderate-to-deep pockets. This fact suggests that the more severe the periodontitis the more pronounced is the negative effect of obesity on periodontal treatment outcome. They showed in the multivariate analysis that obesity was significantly associated with percentage changes of PD > 5 mm and numbers of improving sites (p ≤ 0.05). In the univariate analysis all periodontal parameters improve more in patients suffering from more severe periodontitis at baseline. Other patient characteristics, like age, sex, obesity, and WHR, were not associated with periodontal parameter changes. Gonçalves et al. [31] showed that patients with obesity and chronic periodontitis had a lower PDD reduction than patients without obesity. The measurement of the  reduction in PPD (mm) at full-mouth sites showed after 3 months a not statistically significant difference (p = 0.08) between the obese group compared to the group without obesity. However after 6 months there was a statistically significant difference (p = 0.04). At this time point, especially deep sites (PPD ≥ 7 mm) showed a significantly better result in the group without obesity (p = 0.04). Another study of Gonçalves et al. [32] reported that patients with obesity had a significantly greater mean PD (6 months p-value = 0.04, 12 months p value = 0.03) than patients without obesity at six and 12 months post-therapy. The data of a study by Suvan et al. [34] corroborated these findings and showed that obesity was an independent impact value of poorer periodontal treatment outcome 2 months after therapy. The extent of the association between poorer periodontal treatment and obesity was similar to that of smoking (p = 0.02). They worked with 260 patients. This secondary analysis consisted of individuals participating in five clinical studies of non-surgical periodontal therapy over a 7year period. Lakkis et al. [33] selected 30 patients who were obese; 15 of them had previously undergone bariatric surgery, whereas the other half (n = 15) did not loose any weight and served as a control group. The bariatric surgery group reached a statistically significant greater mean PPD reduction (0.45 mm versus 0.28 mm) compared with the control (no surgery) group (p = 0.007; Table 4). No individuals in the sample had missing data for any of the covariates listed in this study.
CHX chlorhexidine, min minute, mm millimeter, n.r. not reported, PD pocket depth, SRP scaling and root planing

Limitations
Some patients could not be reviewed or discontinued the study for personal reasons [28,29,31]. Weight loss and pregnancy were additional reasons to be excluded in the study of Duzagac et al. [29] ( Table 7). The quality of the included studies was evaluated through the Newcastle-Ottawa Quality Assessment Scale (Table 8).

Discussion
This review focused on obesity and the outcome after non-surgical periodontal therapy and has shown that currently there is no really robust scientific evidence to reach solid conclusions and recommendations.
Three papers [8,29,30] were found, which did not find any statistically significant negative impact of obesity on the response to non-surgical periodontal therapy, whereas five papers [28,[31][32][33][34] showed the opposite, i.e. a clearly negative influence of obesity on the treatment outcomes.
Al-Zahrani et al. [8] included only women. Gonçalves and co-workers [31] did not consistently apply the accepted definitions of overweight and obesity, but rather included the waist-to-hip-ratio (WHR) for their definitions, probably leading to inclusion of patients with a BMI inferior to 30 kg/m 2 into the obesity group. Thus Table 8 Newcastle -Ottawa Quality Assessment Scale Stars represent "awards" for each quality item and serve as a quick visual assessment. A maximum of one 'star' for each item within the 'Selection'/'Exposure' categories and a maximum of two 'stars' for 'Comparability' can be awarded the results are difficult to interpret and compare with the other studies and are therefore not applicable for patients with a BMI ≥40 kg/m 2 [31]. Nonetheless, these studies still show a clearly better tendency with regard to the treatment response for patients without obesity as defined in their study. Lakkis et al. [33] chose another interesting way to find an impact of obesity on the outcome of non-surgical periodontal treatment. They compared obese people who had undergone bariatric surgery (BS) with obese who did not. After weight loss in the BS group, a reduction in total adipocytes might have resulted in a decrease in adipokines and pro-inflammatory mediators released by those adipose cells. This systemic inflammatory reduction might have played a role in reducing the insulin resistance resulting in a better outcome after periodontal therapy as suggested by the authors [33]. Some limitations in the specific profile of the obese patients (nondiabetic, non-smoker) in the paper of Bouaziz et al. (28) may restrict the extrapolation of the results to the whole obese population. Furthermore, the small sample size may also limit the power of this study. Duzagac et al. [29] failed to include a control group of periodontally healthy controls with obesity. Additionally the mean periodontitis parameters were within the limits of "moderate" periodontitis, and the WHR and BMI values of these obese patients were predominantly below those characterizing morbid obesity. So, the results of this study may not be extrapolated to those with severe periodontitis or morbid obesity. The second included study of Gonçalves and co-workers [32] assumes that the high interpatient variance in adipokine levels may reduce the statistical power to detect treatment effects, as previously reported. The results presented by Suvan et al. [34] may have been influenced by study limitations linked with unequal numbers in BMI categories and sample size. In addition, there may have been limitations with regard to the interpretation associated with the post hoc secondary analysis experimental design, although variation in clinical assessment and treatment was minimized by examiner and treatment clinician stability. This study did not constitute a higher level of evidence in the context of evidencebased health care levels of scientific evidence [34].
Overall, obesity is an obvious, visible stigma so that the studies cannot be considered blinded. This may be another possible bias in each of the studies.
Five studies [8,28,29,31,32] excluded smokers. Since smoking influences periodontal health these studies are biased [9] and may not be fully representative for the typical overall population. Overall it appears, however, that a positive effect of normal weight is present in nonsmokers [28,31,32] and in smokers [34].
As mentioned before, the included studies differed in statistical methods, populations, sample sizes, definition of chronic periodontitis, definition of obesity, time of outcome measurement, smoking status, periodontal assessment and non-surgical periodontal therapy. Therefore, it was not only impossible to perform a meta-analysis but also draw clear conclusions.
Nonetheless there is a consensus in the studies that obesity is associated with different baseline PPD levels. The large cohort in the study of Suvan et al. [34] and the long term results of Gonçalves et al. [31,32] may lead to the conclusion that obesity is an important negative factor which influences non-surgical periodontal therapy.
In summary, all studies [8,28,29,[31][32][33][34] included in this review validated the efficacy of non-surgical periodontal therapy, except the study of Eldin and co-workers [30] who did not report any efficacy of the therapy (Table 3).
Clinically, it appears obvious that a therapy is necessary to reach periodontal health independent of the patient's body mass index.
Because this systematic review provided only moderate evidence that obesity is an important factor for nonsurgical periodontal therapy, future prospective cohort studies are needed to confirm these findings [36]. Such trials should be of high methodological quality. They should control important confounding factors such as smoking status, severity of chronic periodontitis, severity of obesity. Every patient should get the same periodontal treatment and periodontal maintenance. Overall, there is a possibility to solve the research question even though blinding of the examiners to obesity or non-obesity status is not practically possible.
Clinicians should know that obesity may have some influence on periodontal status and are likely to have a negative impact on the clinical outcome of conservative treatment, even if this systematic review found only five [28,[31][32][33][34] out of eight papers [8,[28][29][30][31][32][33][34] corroborating the influence of obesity on the clinical periodontal outcome focusing on PPD as surrogate parameter for periodontal healing.
Clinicians might consider a weight reduction diet as an additional treatment for periodontal health with a positive effect expected after 6 and 12 months [31,32]. Also it should not be neglected that weight control has substantial other beneficial health effects which on their own justify such a recommendation.

Conclusion
This systematic review indicates a possible negative relationship between obesity and poorer treatment outcome in obese patients after non-surgical therapy based on the results of five out of eight studies. Three of these studies denied an impact of obesity on the treatment. The potentially inferior healing response could be based on pathophysiological inflammatory models.
Baseline levels showed also a poorer periodontal health in patients with obesity compared with non-obese patients.
No study found any better dental health parameters in obese than in non-obese individuals, and although dental health may not be the most important target for arriving at a near normal body weight, a person who can keep his or her body weight near normal might, in addition to all other established health benefits, count on having better periodontal health than if they are obese.