Association of socio-demographic, behavioral, and comorbidity-related factors with severity of periodontitis in Turkish patients

Abstract

Objective: To assess the prevalence of comorbidities and periodontal diseases severity in dental patients and to determine the socio-demographic, behavioral, and comorbidity-related predictors of periodontal diseases severity.

Material and methods: This retrospective study sample consisted of 2458 patients who referred to faculty dentistry clinic. Socio-demographic, behavioral, and comorbidity characteristics of study participants were collected using hospital database and self-reported questionnaire. Descriptive, bivariate, and multivariate analyses were used to analyze study data.

Results: Of these patients, 55.2% had mild-to-moderate periodontitis and 44.8% had severe periodontitis. The severity of periodontal disease was significantly associated with the presence of comorbidity, the number of comorbidities, age, gender, income level, smoking, and alcohol consumption. Patients with severe periodontitis were more likely being a current smoker, to report drinking alcohol sometimes or every day, to be a male, to have a pulmonary disease, to have an endocrinological and metabolic disorder, to have a cardiovascular disorder and to have a neurological disorder than those with mild/moderate periodontitis. Having a hematological disorder, having a muscle, skeletal and connective tissue disorder, to be a current smoker and lower age were found to be predictors of moderate periodontitis whereas being a female and lower number of comorbidities were predictors of mild periodontitis.

Conclusions: The severity of periodontitis was associated with socio-demographic, behavioral, and comorbidity characteristics of periodontal patients.

Keywords:

• Comorbidity
• health behavior
• periodontitis
• socio-demographic
• Turkey

Introduction

Periodontitis is common in developed and developing countries and affects approximately 20–50% of the world population [1]. A recent systematic analysis for the Global Burden of Disease Study 2016 indicated that severe periodontal disease, which may cause tooth loss, was the 11th most prevalent disease worldwide [2]. Systematic and narrative reviews on periodontal disease [3,4] and recent cohort study [5] provide epidemiologic and experimental evidence exists for the role of several modifiable and non-modifiable risk factors in the onset, progression, and severity of the periodontal disease. Besides the presence of dental biofilms, periodontitis shares similar inflammatory pathways, social determinants, and risk factors with the major non-communicable diseases such as heart disease, diabetes, cancer, and chronic respiratory disease [5,6].

According to the network model for the pathway of risk factors in the pathogenesis of periodontitis, several proximal and distal risk factors influence the onset and the severity of periodontal diseases [7]. Due to the multifactorial nature of periodontal diseases, periodontitis can be prevented, diagnosed early and effectively treated and controlled through the implementation of primary and secondary preventive prevention strategies at individual and population level [3,6,8]. Within a holistic approach, personalized medical and oral care should be planned considering the individual’s genetic trait, environmental condition, personal profile, and host susceptibility [5]. Previous studies conducted on dental patients showed that periodontal diseases found to be associated with various systemic disorders or diseases [9–15] as well as with smoking [9,15].

Recent nationwide studies showed that the prevalence of periodontal diseases is higher in Turkish adults [16,17]. In these studies, age, gender, living area, smoking habits, educational, and employment status were found to be important predictors of the severity of periodontal disease [16,17]. Only one study reported that periodontitis was a risk factor for systemic diseases in dental patients [14]. In Turkey, primary and secondary prevention strategies are urgently needed in controlling and treating periodontal disease considering the common risk factor approach. Oral disease prevention strategies are not integrated into chronic systemic disease preventive initiatives. Identifying predictors of periodontal disease severity may provide evidence about the importance of this issue for all health professionals as well as an opportunity for collaborative efforts aiming to control and prevention of periodontal diseases and comorbidities in both medical and dental practice.

This study aimed to assess the prevalence of comorbidities and periodontal disease severity in patients referred to periodontal treatment and to determine the socio-demographic, behavioral, and comorbidity-related predictors of periodontal disease severity. We hypothesized that significant differences would exist among periodontal patients with different severity levels regarding socio-demographic, behavioral, and comorbidities.

Materials and methods

This study was carried out on a series of consecutive patients seeking treatment for periodontitis in the Periodontology Department, Faculty of Dentistry, Bolu Abant Izzet Baysal between the years 2014 and 2016. A total of 2458 patients (1111 male and 1347 female) were selected from 8750 consecutive patients who referred to the Department of Periodontology. The inclusion criteria were the presence of periodontitis as classified according to the criteria of the American Academy of Periodontology [18]. Patients were excluded if they: (1) had gingivitis; (2) were under the age of 18 years; (3) were pregnant; (4) who could not read, write, or understand Turkish, and (5) were non-responders or declined to participate. This study was approved by the ethical board of the Bolu Abant Izzet Baysal University, Bolu, Turkey (No: 2018/228).

Data were obtained using our hospital database and a self-reported questionnaire including medical history, socio-demographic characteristics (gender, age, educational level, and family monthly income), health behaviors (smoking and alcohol use), and clinical data.

During study time, the International Classification of Diseases, 10th revision (ICD-10) codes were not being used in our hospital. Therefore, the comorbidities list was prepared by a research assistant through a review of medical records. Then, the patients were categorized as cardiovascular disorders and interventions, lung diseases, endocrinological and metabolic disorders, dermatological diseases, gastrointestinal diseases, genitourinary disorders, infectious diseases, allergic disorders, liver and biliary disorders, hematological disorders, muscle, skeletal and connective tissue disorders, neurological disorders, psychiatric disorders, cancers, and ear, nose, and throat disorders according to the Merck Manual [19].

For lifestyle variables, patients were categorized into three groups based on self- reported smoking data: (i) infrequent smokers/non-smokers: those who had quit smoking after smoking for less than 10 years and those with no history of smoking; (ii) former heavy smokers: those who had smoked daily for 10 or more years but had stopped smoking (iii) current smokers: those who are smoking currently [20,21]. Taking into account the suggestion of a systematic review conducted by Fiorini et al. [22] that a possibility of periodontitis risk reversal within 10 years after smoking cessation, we classified the patients who had quit smoking after smoking for less than 10 years and those with no history of smoking into same group.

Patients were classified into two groups according to alcohol consumption: none versus sometimes or every day [16].

Taking into account the 1997 Eight-Year Compulsory Education Law, the education level of the patient was divided into two groups according to the completed education years and the highest education level attained: ≤8 years of schooling versus > 8 years of schooling [23]. Monthly household income was classified into two categories according to the basic wage at the time of the study: minimum wage and less versus above minimum wage [20].

A diagnosis of periodontitis is based on the combination of periodontal and gingival inflammation, determined as the presence of sub-gingival bleeding on probing, the number and depth of periodontal pockets, the amount of clinical attachment loss, and presence of radiologically evaluated alveolar bone loss [18]. Clinical parameters were assessed by probing with a Williams periodontal probe (Hu-Friedy, Chicago, IL, USA). Patients fulfilled the criteria of periodontitis proposed by Tonetti and Claffey as well as the criteria for severity of periodontitis of the American Academy of Periodontology [18,21].

Statistical analyses

Descriptive statistics (mean, standard deviation, frequencies, percentages) were used to summarize the data. Differences between patient groups were examined with Pearson chi-square test for categorical variables. To compare the socio-demographic, comorbidity status and behavioral characteristics of patient groups, continuous data were analyzed with the Mann–Whitney U and Kruskal–Wallis tests followed by Dunn’s post hoc multiple comparisons. Finally, three separate binary logistic regression analyses with stepwise backward elimination (likelihood ratio) were applied to identify the significant associations between the independent variables (gender, age, income, educational level, the presence of comorbidity, the number of comorbidities, comorbidities, smoking status, and alcohol) and the dependent variables (the severity of periodontal disease designed as mild, moderate and severe). Sample size for logistic regression was calculated using the following formula: n = 100 + 50(i), where i refers to number of independent variables in the final model. Based on the above formula, the minimum required sample size for 23 independent variables was found to be 1250 [24]. Estimates of model fit (Omnibus test), the unstandardized beta (B), the standard error (SE), and odds ratios (ORs) with their corresponding 95% confidence intervals (CIs) were computed. Data were analyzed using IBM SPSS Statistics Version 21 for Windows (SPSS Inc., Chicago, IL, USA).

Results

The study sample consisted of 2458 patients (54.8% of females and 45.2% males) with a mean age of 52.75 ± 12.26 years. From 2458 patients, 22.5% of participants were diagnosed with mild periodontitis, 32.7% with moderate periodontitis and 44.8% were classified as having severe periodontitis. In 925 (37.6%) patients, no comorbidities were found. Six hundred 24 patients presented with comorbidities. 38.4% reported one disorder, 19.3% reported two, 3.5% reported three, 0.8% reported four, 0.2% reported five, and 0.1% reported six. Of these, 17.2% (423) declared to be current smokers, 13.2% (342) to be former heavy smokers and 68.9% to be infrequent smokers/non-smokers. Concerning alcohol consumption, 83.6% reported that they did not drink alcohol at all, 16.4% said they drank alcohol sometimes or every day. When the distribution of comorbidities was examined, it was found that 37.9% were cardiovascular disorders, 26.9% were endocrine and metabolic disorders, 6.33% were pulmonary diseases, 4.5% were liver disorders, 3.1% were muscle, skeletal, and connective tissue disorders, 2.4% were neurological disorders, 2.6% were psychiatric disorders, 2.4% were allergic disorders, 1.6% were gastrointestinal diseases, 1.5% were cancers, 1.3% were genitourinary disorders, 0.8% were hematological disorders, 0.3% were infectious diseases, 0.3% were ear, nose, and throat diseases, 0.3% were biliary diseases and 0.2% were dermatological disorders (Data not shown).

As seen in Table 1, significant differences in the severity of periodontal disease were found related to the presence of comorbidity (p < 0.001), the number of comorbidities (p < 0.001), age (p = 0.001), gender (p < 0.001), income level (p = 0.001), smoking status (p < 0.001), and alcohol consumption (p < 0.001). Post hoc comparisons using Dunn’s test revealed the following significant differences in the number of comorbidities among all patients’ groups (p < 0.001). The significant difference was found between patients with moderate periodontitis and patients with severe periodontitis in terms of age (p = 0.001).

Table 1. Socio-demographic, behavioral, and comorbidity characteristics of study participants according to the severity of the periodontal disease.

Characteristics	Patients with mild periodontitis	Patients with moderate periodontitis	Patients with severe periodontitis	p Value
	n = 553	n = 803	n = 1102
Age (years), Mean (SD)^a	52.31 ± 11.94	51.87 ± 11.98	53.61 ± 12.57	0.001**
Gender (n, %)^b
Male (n = 1111)	198 (17.8)	366 (32.9)	547 (49.2)	<0.001
Female (n = 1347)	355 (26.4)	437 (32.4)	555 (41.2)
Income level (n, %)^b
Minimum wage and less (n = 1114)	219 (19.7)	353 (31.7)	542 (48.7)	0.001
Above minimum wage (n = 1344)	334 (24.9)	450 (33.5)	560 (41.7)
Educational level (n, %)^b
≤8 years (n = 1339)	293 (21.9)	429 (32)	617 (46.1)	0.392
>8 years (n = 1119)	260 (23.2)	374 (33.4)	485 (43.3)
Alcohol consumption (n, %)^b
None (n = 2056)	498 (24.2)	687 (33.4)	871 (42.4)	<0.001
Sometimes or every day (n = 402)	55 (13.7)	116 (28.9)	231 (57.5)
Smoking status (n, %)^b
Infrequent smokers/non-smokers (n = 1693)	425 (25.1)	584 (34.5)	684 (40.4)	<0.001
Former heavy smokers (n = 342)	57 (16.7)	107 (31.3)	178 (52)
Current smokers (n = 423)	71 (16.8)	112 (36.5)	240 (56.7)
Comorbidiety (n, %)^b
No (n = 925)	296 (32)	324 (35)	305 (33)	<0.001
Yes (n = 1533)	257 (16.8)	479 (31.2)	797 (52)
Number of comorbidities, Mean (SD)^a	0.65 ± 0.79	0.92 ± 0.95	1.06 ± 0.91	<0.001*

SD: standard deviation.

^a Statistical evaluation by the Kruskal–Wallis test with post hoc Dunn test.

^b Statistical evaluation by the Pearson’s Chi-square test.

*Post hoc comparisons using Dunn’s test revealed following significant differences among all patients’ groups at p < 0.001 significance level; **Post hoc comparisons using Dunn’s test revealed following significant differences between patients with moderate and severe periodontitis at p = 0.001 significance level. During study period, the monthly minimum wage in Turkey ranged from TRY 1134 (520 US Dolar in 2014) to TRY 2558 (850 US Dolar in 2016).

In this study, statistically significant differences were found between patients without comorbidity and with comorbidity regarding gender (p = 0.001), periodontal disease severity (p < 0.001), and income level (p < 0.001; Table 2).

Table 2. Socio-demographic, behavioral, and periodontal disease characteristics of study patients according to the presence of comorbidity.

Characteristics	Patients without comorbidity n = 925	Patients with comorbidity n = 1533	p Value
Age (years), Mean (SD)^a	52.20 ± 12.18	53.08 ± 12.29	0.062
Gender (n, %)^b
Male (n = 1111)	457 (41.1)	654 (58.9)	0.001
Female (n = 1347)	468 (34.7)	879 (65.3)
Income level (n, %)^b
Minimum wage and less (n = 1114)	341 (30.6)	773 (69.4)	<0.001
Above minimum wage (n = 1344)	584 (43.5)	760 (56.5)
Educational level (n, %)^b
≤8 years (n = 1339)	481 (35.9)	858 (64.1)	0.056
>8 years (n = 1119)	444 (39.7)	675 (60.3)
Alcohol consumption (n, %)^b
None (n = 2056)	763 (37.1)	1293 (62.9)	0.228
Sometimes or every day (n = 402)	162 (40.3)	240 (59.7)
Smoking status (n, %)^b
Infrequent smokers/non-smokers (n = 1693)	626 (37)	1067 (63)	0.491
Former heavy smokers (n = 342)	138 (40.4)	204 (59.6)
Current smokers (n = 423)	161 (38.1)	262 (61.9)
The severity of periodontal diseases^b
Mild (n = 553)	296 (53.5)	257 (46.5)	<0.001
Moderate (n = 803)	324 (40.3)	479 (59.7)
Severe (n = 1102)	305 (27.7)	797 (72.3)

SD: standard deviation.

^a Statistical evaluation by the Mann–Whitney U test.

^b Statistical evaluation by the Pearson’s Chi-square test. During study period, the monthly minimum wage in Turkey ranged from TRY 1134 (520 US Dolar in 2014) to TRY 2558 (850 US Dolar in 2016).

Bivariate analyses revealed that gender (p < 0.001), alcohol consumption (p < 0.001), smoking status (p < 0.001), income level (p = 0.002), the number of comorbidities (p < 0.001), lower prevalence of pulmonary disease (p = 0.004), endocrine and metabolic disorder (p < 0.001), cardiovascular disorder (p < 0.001), muscle, skeletal, and connective tissue disorder (p = 0.027), and neurological disorder (p = 0.022) were associated with mild periodontitis; smoking status (p = 0.006), age (p = 0.001), higher prevalence of hematological disorder (p = 0.016), muscle, skeletal, and connective tissue disorder (p = 0.018) were associated with moderate periodontitis; and age (p < 0.001), gender (p < 0.001), income level (p = 0.001), alcohol consumption (p < 0.001), smoking status (p < 0.001), the number of comorbidities (p < 0.001), higher prevalence of pulmonary disease (p < 0.001), endocrine and metabolic disorder (p = 0.001), cardiovascular disorder (p < 0.001), neurological disorder (p = 0.011) were associated with severe periodontitis (data not shown).

As seen in Table 3, the results of separate binary logistic regression analyses showed patients with severe periodontitis were more likely to be current smokers (OR = 1.185, 95% CI = 1.008 to 1.394), to report drinking alcohol sometimes or every day (OR = 1.187, 95% CI = 1.045 to 1.342), to be male (OR = 0.912, 95% CI = 0.833 to 0.999), to have a pulmonary diseases (OR = 1.413, 95% CI = 1.190 to 1.677), to have a endocrine and metabolic disorders (OR = 1.144, 95% CI = 1.042 to 1.256), to have a cardiovascular disorders (OR = 1.254, 95% CI = 1.152 to 1.366), and to have a neurological disorders (OR = 1.403, 95% CI = 1.069 to 1.842) than those with mild/moderate periodontitis. Having a hematological disorder (OR = 1.564, 95% CI = 1.010 to 2.423), having a muscle, skeletal, and connective tissue disorder (OR = 1.306, 95% CI = 1.035 to 1.648), to be current smoker (OR = 1.230, 95% CI = 1.045 to 1.449), and lower age (OR = 0.991, 95% CI = 0.984 to 0.999) were found to be predictors of moderate periodontitis whereas being female (OR = 1.236, 95% CI = 1.109 to 1.378) and lower number of comorbidities (OR = 0.598, 95% CI = 0.529 to 0.677) were important predictors of mild periodontitis.

Table 3. Separate binary logistic regression analyses for associations between each level of periodontal disease severity and study variables.

	B	SE	p Value	OR	95% CI for OR
					Lower	Upper
Significant predictors for severe periodontitis
Alcohol consumption
None	Reference
Sometimes or every day	0.172	0.065	0.008	1.187	1.045	1.342
Smoking status
Infrequent smokers/non-smokers	Reference
Former heavy smokers	0.074	0.083	0.375	1.077	0.915	1.267
Current smokers	0.17	0.083	0.04	1.185	1.008	1.394
Gender
Male	Reference
Female	−0.092	0.047	0.049	0.912	0.833	0.999
Pulmonary diseases
No	Reference
Yes	0.346	0.087	<0.001	1.413	1.19	1.677
Endocrinological and metabolic disorders
No	Reference
Yes	0.134	0.048	0.005	1.144	1.042	1.256
Cardiovascular disorders and intervention
No	Reference
Yes	0.227	0.044	<0.001	1.254	1.152	1.366
Neurological disorders
No	Reference
Yes	0.339	0.139	0.15	1.403	1.069	1.842
Significant predictors for moderate periodontitis
Hematological disorders
No	Reference
Yes	0.448	0.223	0.045	1.564	1.01	2.423
Muscle, skeletal, and connective tissue disorders
No	Reference
Yes	0.267	0.119	0.024	1.306	1.035	1.648
Smoking status
Infrequent smokers/non-smokers	Reference
Former heavy smokers	0.022	0.088	0.804	1.022	0.86	1.214
Current smokers	0.207	0.083	0.013	1.23	1.045	1.449
Age^a	−0.009	0.004	0.024	0.991	0.984	0.999
Significant predictors for mild periodontitis
Gender
Male	Reference
Female	0.212	0.055	<0.001	1.236	1.109	1.378
Number of comorbidities^a	−0.514	0.063	<0.001	0.598	0.529	0.677

B: the unstandardized coefficients; SE: standard error; OR: odds ratio; CI: confidence interval.

^a Continous variables.

The binary logistic regression models indicated a good fit (Ombinus test: chi-square = 117.422, p < 0.001 for mild periodontitis; Ombinus test: chi-square = 25.391, p < 0.01 for moderate periodontitis; and Ombinus test: chi-square = 117.768, p < 0.001).

Discussion

In Turkey, there is a scarcity of oral health literacy about the link between periodontal diseases and comorbidities among medical and dental professionals as well as the public [25]. Patients who have comorbidities generally do not have any information about links between their general health and oral health and referred late by medical doctors to dentists for a dental examination, which leads to reducing the effective management of periodontal diseases among these patients.

In this study, we compared the prevalence of comorbidities among patients with periodontitis and tried to clarify the relationship between periodontal disease severity and systemic diseases. We found that 62.4% of the patients had one or more comorbidities. When the distribution of comorbidities was examined, the highest rates were seen in cardiovascular disorders and interventions (%37.9) and endocrine and metabolic disorders (%26.9), followed by respiratory diseases (%6.33). These findings are consistent with previous studies conducted among periodontal patients [8,13]. Higher prevalence of cardiovascular disorders and endocrine and metabolic diseases may be explained by the fact that the pathophysiology of these diseases is similar [26].

Confirming the study hypothesis, bivariate analyses showed that the severity of periodontitis was associated with age, gender, income level, alcohol consumption, smoking status, having comorbidity, and number of comorbidities among Turkish periodontal patients. Multivariate analyses revealed that patients with severe periodontitis were more likely to be current smokers, to report drinking alcohol sometimes or every day, to be male, to have a pulmonary disease, endocrine and metabolic disorders, cardiovascular disorders, and neurological disorders than those with mild/moderate periodontitis. In line with this study, previous studies conducted on dental patients showed that periodontal diseases found to be associated with various systemic disorders or diseases [9–15]. The number of systemic conditions [10] and a combined manifestation of systemic diseases [9] was associated with severe periodontal disease.

In contrast to our study, some studies suggested that chronic periodontitis severity was not significantly associated with systemic conditions [11–13]. These differences may be related to the study design and population characteristics. We found statistically significant differences between patients without comorbidity and with comorbidity regarding gender, periodontal disease severity, and income level. Similar to our study, Zhao et al. [5] found that participants who were diagnosed with more comorbidities exhibited greater mean bone loss/age and the less mean number of remaining teeth as severe periodontitis. In contrast to our study, Sperr et al. [27] and Boutros et al. [11] reported that there were no significant differences between the severity of periodontal disease and co-existence systematic disease.

Bivariate analysis showed that patients with moderate periodontitis were younger than those with severe periodontitis. This may be explained by the fact that the risk of periodontal disease increases with the advancing age [16,17,28,29]. Our findings are in line with recent national studies using the Community Periodontal Index and loss of periodontal attachment [16,17] that found that the prevalence of >3 mm of attachment loss in age group 35 to 44 years was lower than the 65 + year-old age group. Another study using the Community Periodontal Index of Treatment Needs index reported that the prevalence of periodontal disease of adult patients in South Eastern Turkey was found to be higher in patients aged 57–69 years [30].

Lower income was related to both the severity of periodontal diseases and comorbidity status among Turkish patients which is consistent with previous studies [28,31]. This can be explained by the fact that patients with lower income levels face some problems in accessing oral health and general health services as well as in the utilization of these services.

Being female and a lower number of comorbidities were important predictors of mild periodontitis. This result is not surprising because women may have a higher risk of developing systemic diseases than men due to physiological factors and metabolic changes [31,32]. Being male was found to be an important predictor of severe periodontitis. Similarly, Ghosh-Dastidar et al. [29] reported that men were more likely than women to have a periodontal disease. This may be related to their lifestyle-related behaviors and health service attendance pattern, because men are mainly problem-oriented dental attenders and they were less likely than women to perform healthy behaviors [33].

In contrast to previous studies [16,30], we did not find any significant relationship between educational level and periodontal disease severity. Compared with other Turkish studies, this discrepancy may be due to the educational variations between regions in Turkey and the exclusion of patients who could not read, write, or understand Turkish. More educational programs are needed to improve the level of knowledge and awareness regarding the relationship between systemic and periodontal diseases and the risk factors for periodontal disease for all patients and the general population [16,17,30].

Existing systematic reviews and meta-analyses of the effect of smoking and alcohol consumption on periodontitis reported that alcohol consumption and smoking were associated with an increased risk of periodontitis and these life-style related factors should be evaluated together with other risk factors for periodontitis because of having a deleterious effect on the incidence and progression of periodontitis [34,35]. In this study, being current smokers was found to be an important life-style related predictor of severe and moderate periodontitis, but alcohol consumption was only predictor of severe periodontitis. Similar findings regarding smoking were obtained from previous studies conducted on patients’ group [9,15,16,28]. Additionally, findings obtained from this study are consistent with recent systematic reviews and meta-analyses showing that current smokers have higher risk of periodontitis and tooth loss than quitters and never smokers [34,36]. As stated in method section, the patients who had quit smoking after smoking for less than 10 years and those with no history of smoking were included into same group because evidences showed that a possibility of periodontitis risk reversal within 10 years after smoking cessation [22] and the risk for periodontitis incidence and progression could be reversed following smoking cessation to the same level as that of never smokers [36].

Similar to our study, earlier studies also demonstrated the effect of alcohol consumption on periodontal disease [16,37]. We found significant associations among alcohol consumption, smoking, and severe periodontitis, which are consistent with previous studies reporting the interactive effects of smoking and drinking on the presence of periodontitis [38,39]. Tezal et al. [40] found that only alcohol consumption was associated with moderately increased severity of periodontitis but there was no significant interaction between smoking and alcohol consumption for periodontitis severity. Further studies to confirm this interaction and to test possible underlying mechanisms are needed.

According to OECD Health at a Glance 2019 report, daily smoking rates among Turkish adults is higher than the OECD average despite tobacco control policies but alcohol consumption level is lower [41]. In Turkey, there are policies for alcohol control including the increase in alcohol sales taxes, advertising bans, and comprehensive restrictions. Previous studies reported that lower alcohol consumption rate in Turkish adults could mainly be due to the socio-cultural, and religious norms besides these control policies [16,42]. Oral health literacy and education programs should be planned for patients and the general public to raise oral health awareness regarding the hazardous effects of smoking and alcohol consumption on oral health [16,39,43].

To the best of our knowledge, this was the first study to determine the socio-demographic, behavioral and comorbidity-related predictors of periodontal disease severity among adult periodontal patients. However, this study had some limitations. First, we aimed to investigate the associations between the periodontal disease severity and coexistence of systemic diseases; therefore, we did not include non-periodontitis/gingivitis group. New population and hospital-based case-control studies may provide an additional information regarding the relationship between periodontal disease and systemic conditions by comparing groups with different diseases and healthy. Second, this cross-sectional study was conducted with a sample of adult patients who referred to the department of periodontology at our faculty, limiting the generalizability of the results of the Turkish patient population. This retrospective study could not explain causation and changes over time in the severity of periodontitis and comorbidities status as well as other studied variables including socio-demographic and behavioral. Future longitudinal population-based and multicenter clinical studies assessing the association between the severity of periodontitis and comorbidities status may provide valuable information for health and oral health professionals in understanding possible link between these diseases as well as in developing evidence-based guidelines for healthcare professionals and individuals. The cross-sectional design of this study did not permit us to establish the temporal relationship between cause (smoking and alcohol consumption) and effect (periodontitis). We acknowledge that the self-reported smoking and alcohol use status used in this study may lead to underreporting these risky behaviors because of social desirability. Future longitudinal studies using both the objective measures and the self-reported data on duration and quantity of current smoking and alcohol consumption are needed to investigate the dose–response relationship between smoking and alcohol consumption and periodontitis [34,38,40]. We did not use a structural equation modeling (SEM) approach to explaining the causal pathways for any association because of the limited number of socio-demographic and behavioral factors. After this pilot study, we aim to plan a new study for developing theoretical models of the pathways that link all possible factors to periodontal disease severity using the SEM approach. In this study, increasing age was only found to be a significant predictor of moderate periodontitis in the multivariate analyses. It should be noted that we could not determine an appropriate age threshold for the severity of the periodontitis in this study used a clinical-based sample because age was treated as a continuous variable. Future clinical and population-based studies aiming to assess an appropriate age threshold for moderate and severe periodontitis among Turkish adults may provide valuable insight into developing guideline for the appropriate age threshold at which to recommend the evaluation of moderate and severe periodontitis for the general population as well as for individuals with systemic disease [44]. In addition, some studies reported that the differences between smokers and never-smokers in the severity of periodontal disease increased with age confirming a smoke exposure related response [45,46]. Future studies are needed to determine the age-related differences in the dose-response relationship of smoking in patients with periodontitis. We found that severe periodontitis was associated with having a pulmonary disease, endocrine and metabolic disorders, cardiovascular disorders, and neurological disorders while having a hematological disorder and a muscle, skeletal and connective tissue disorder was related to moderate periodontitis. This may be explained by the fact that periodontal diseases and these diseases are highly prevalent among Turkish adults [16,17,42]. In line to our study results [15,28,46], some researchers found that systemic diseases are associated with periodontitis severity and influenced by periodontal treatment. However, further prospective cohort studies which take a broad range of confounders into account are needed to confirm and clarify the link between severity of periodontal diseases and systemic disease.

It should be noted that clinical, socio-demographic, and behavioral data was obtained from the hospital database. Thus, we could not give any intra-interrelated reliability for examiners. Patients were categorized according to the Merck Manual [19] because ICD-10 codes were not being used in our hospital during study time. To date, there are no epidemiological studies using the periodontitis case definitions. To the best of our knowledge, this is the first study using the American Academy of Periodontology 1999 classification of periodontitis [18] to assess a wide range of risk factors for periodontal disease. New studies are needed using the comorbidity index or ICD-10 codes for the classification of comorbidities and a new classification scheme for periodontal diseases [47], which may allow a comprehensive understanding of the periodontal status in patients and the general Turkish population and the assessment of risk factors as well as to allow future international comparability. Considering the proposed causal network model for periodontitis [7], various socio-demographic, behavioral [48,49], general, and oral health-related risk factors should be examined in future studies, because work-related qualification, living area, and dental attendance were found to be associated with the likelihood of periodontal disease [5,26,28]. Despite these limitations, this study provides valuable information that is helpful to the clinicians in identifying the highly susceptible individuals as well as in developing personalized oral health care plan and evidence-based guideline [5,44].

Conclusion

The results of this study indicated that the severity of periodontitis was associated with socio-demographic, behavioral and comorbidity characteristics of adult periodontal patients. Detecting patients at risk for periodontal diseases may provide for developing effective primary and secondary preventive strategies via collaborative efforts of medical and dental providers as well as for controlling of negative effects of periodontal diseases and comorbidities considering a bidirectional relationship between these diseases.

Author contributions

TP, GU, and KP conceived the ideas; TP and GU collected the data; TP, GU, and KP analyzed the data; and TP, GU, and KP led the writing.

Acknowledgment

The authors thank Dr. Ömer UYSAL who works at Bezmialem Vakıf University for conducting statistical analyses.

Disclosure statement

No potential conflict of interest was reported by the author(s).