Comparative evaluation of knowledge levels and attitudes regarding periodontal health and periodontal disease among internal medicine physicians and dentists in Turkey: a cross-sectional survey study Authors: Beyza Bozmaoğlu, Emre Yaprak, Ahmet Doğukan Gündoğdu, Yasemin Turan Akansel, Tuncay Müge Alvur Published: 2026 ABSTRACT AbstractBackgroundPeriodontal diseases are closely related to various systemic conditions including diabetes mellitus (DM), cardiovascular diseases, adverse pregnancy outcomes, and rheumatoid arthritis (RA). Internal medicine physicians frequently encounter patients with these systemic conditions and are in a strategic position to facilitate early diagnosis and referral for periodontal problems. Despite international evidence suggesting significant knowledge gaps among medical practitioners regarding periodontal health, limited data are available from Turkey. The aim of this study was to evaluate the knowledge levels and attitudes of internal medicine specialists and residents regarding periodontal health and periodontal disease, and their relationship with systemic conditions, in comparison with dentists.MethodsThis cross-sectional descriptive survey study was conducted between April 2021 and December 2021 in Turkey. A content-validated, self-administered online questionnaire consisting of 7 sociodemographic questions and 29 questions related to periodontal health and periodontal disease knowledge and attitudes was distributed to internal medicine specialists/residents (IMS group, n = 107) and dentists (control group, n = 133) via social media platforms. Internal consistency was assessed using Cronbach’s alpha (α = 0.829). Normality of continuous variables was assessed using the Shapiro–Wilk test. Data were analyzed using chi-square tests, Kruskal–Wallis tests, and Mann–Whitney U tests, with statistical significance set at p < 0.05. Bonferroni correction was applied for multiple comparisons. Multiple linear and logistic regression models were used to adjust for potential confounders.ResultsA total of 240 participants met the inclusion criteria. Internal medicine specialists/residents demonstrated significantly lower knowledge levels compared to dentists across most periodontal knowledge domains (p < 0.05). Only 65.4% of the IMS group correctly defined periodontitis compared to 98.5% of dentists (p < 0.001). Knowledge gaps were particularly prominent regarding the relationship between periodontal disease and pregnancy complications (46.7% vs. 91%, p < 0.001), the effect of smoking on periodontal tissues (59.8% vs. 97.7%, p < 0.001), and the role of radiographic examination in periodontal diagnosis (37.4% vs. 88.8%, p < 0.001). After Bonferroni correction, 14 of 22 comparisons remained statistically significant. Notably, both groups showed high and comparable knowledge regarding the recognition of diabetes as a risk factor for periodontitis (IMS: 93.5% vs. dentists: 92.5%). Subgroup analysis revealed that specialists scored significantly higher than residents across all knowledge domains (total score: 16.94 ± 2.90 vs. 14.38 ± 3.73, p < 0.001). Multivariate analysis confirmed that group differences remained significant after adjusting for age, sex, marital status, and years since graduation (adjusted B = 4.65, p < 0.001). Self-assessment scores for periodontal knowledge significantly increased after completing the survey in the IMS group (4.02 to 5.34, p < 0.001).ConclusionsThe findings of this study suggest that internal medicine specialists and residents in Turkey may have lower knowledge levels regarding periodontal health and periodontal disease, and the periodontal-systemic health relationship compared to dentists. Specialists demonstrated significantly higher knowledge than residents across all domains. While these results should be interpreted within the context of the study’s methodological limitations, they point to a potential need for integration of oral health education into medical curricula and enhanced interprofessional collaboration between medicine and dentistry. Further research using representative sampling strategies and longitudinal designs is warranted to confirm these findings.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12909-026-09365-7.Keywords: Periodontal disease knowledge, Internal medicine physicians, Medical-dental collaboration, Oral-systemic health relationship, Periodontal awareness survey, Turkey, Cross-sectional study Background Oral health is closely related to general health, and periodontal diseases have been linked to numerous systemic conditions including diabetes mellitus (DM), coronary artery disease, adverse pregnancy outcomes, and rheumatoid arthritis (RA) [1–4]. Periodontitis, a chronic multifactorial inflammatory disease characterized by progressive destruction of tooth-supporting tissues, is the sixth most prevalent disease globally, with severe forms affecting approximately 11.2% of the adult population and milder forms affecting 45–50% of adults [5–7]. Beyond its oral manifestations, periodontitis has been proposed as a potential contributor to chronic systemic conditions through the dissemination of oral microorganisms, bacterial virulence factors, and pro-inflammatory molecules to distant body sites [8, 9]. The bidirectional relationship between periodontal disease and diabetes is particularly well-established. Diabetes is recognized as a major risk factor for periodontitis, while the presence of periodontal disease can impair glycemic control and increase insulin resistance in diabetic individuals [10]. Similarly, periodontal pathogens, notably Porphyromonas gingivalis, have been implicated in vascular and atherosclerotic changes contributing to cardiovascular disease etiology [9]. Furthermore, periodontal disease has been associated with increased risks of preeclampsia, preterm birth, and low birth weight in pregnant women [2]. Despite these well-documented associations, historical separation between medical and dental practice has resulted in a lack of synergistic interaction between these two healthcare disciplines [11]. Most patients visit medical physicians more frequently than dentists, positioning medical doctors, particularly internal medicine specialists, as key gatekeepers for early detection and referral of oral health problems [12]. Consequently, adequate knowledge of periodontal diseases and their systemic implications among medical practitioners is essential for comprehensive patient care. International studies conducted in Nigeria [13], China [14], India [15], the United States [16], Israel [17], Croatia [18], and France [19] have assessed the periodontal awareness and knowledge levels of medical practitioners with varying results, generally indicating significant knowledge gaps. A recent cross-sectional study from Croatia demonstrated that both physicians and dentists in primary healthcare had limited oral-systemic health awareness, emphasizing the need for curriculum modification and interprofessional training [18]. However, limited data are available from Turkey, and no study has specifically compared the periodontal knowledge of internal medicine specialists and residents with that of dentists in the Turkish context. The primary objective of this study was to evaluate the knowledge levels and attitudes of internal medicine specialists and residents in Turkey regarding periodontal health and periodontal disease, and their relationships with systemic conditions, using dentists as a comparison group. Methods Study design and setting This was a cross-sectional descriptive epidemiological survey study conducted between April 2021 and December 2021 across Turkey. The study was approved by Kocaeli University Non-Interventional Clinical Research Ethics Committee (Decision No: GOKAEK-2021/6.03, Date: 18.03.2021). All methods were performed in accordance with the Declaration of Helsinki and relevant guidelines. Informed consent was obtained from all participants; completion of the questionnaire following the information page was accepted as implied consent. Participants and sampling The target population comprised internal medicine specialists, internal medicine residents, and dentists practicing in Turkey. The online questionnaire, created using Google Forms, was distributed via social media platforms (WhatsApp, Facebook) to reach potential participants across the country (Fig. 1). A total of 245 individuals responded. Five undergraduate dental students were excluded, as they had not yet completed their core clinical education and were not practicing clinicians; their inclusion would have introduced a fundamentally different level of training, making group comparisons inappropriate. By contrast, internal medicine residents are postgraduate physicians actively engaged in clinical practice and were therefore considered appropriate for inclusion. This yielded a final sample of 240 participants: 47 internal medicine specialists (19.6%), 60 internal medicine residents (25%), and 133 dentists (55.4%). Internal medicine specialists and residents were combined into a single group (IMS group, n = 107) for the primary comparative analyses, as both groups undergo the same core internal medicine training program and practice within similar clinical environments. Subgroup comparisons between specialists and residents were also performed to explore potential differences within the IMS group. STROBE flow diagram illustrating participant recruitment, inclusion, and exclusion. A total of 245 responses were received; 5 undergraduate dental students were excluded, yielding a final analytic sample of 240 participants (IMS group: n = 107 [specialists: n = 47, residents: n = 60]; Dentist group: n = 133). IMS = Internal Medicine Specialists/Residents; STROBE = Strengthening the Reporting of Observational Studies in Epidemiology Survey instrument The survey instrument was developed de novo specifically for this study, building upon a comprehensive review of established instruments used in comparable international research [14, 16, 18–20]. Specifically, items assessing periodontal disease definition, etiology, and clinical features were adapted from Lin et al. [14] and Owens et al. [16]; items related to periodontal-systemic disease relationships were informed by Cohen et al. [19] and Al-Habashneh et al. [20]; and items on clinical attitudes and referral practices were developed based on themes identified in Gambhir et al. [21]. Several items related to medication-related osteonecrosis of the jaw (MRONJ), organ transplantation, and implant-related periodontal disease were newly developed for this study to reflect clinical scenarios relevant to the Turkish internal medicine context. After identifying key thematic domains and adapting them to the Turkish context, content validity was established through an expert review conducted by two periodontologists and one internal medicine specialist. These experts independently evaluated the relevance, clarity, and appropriateness of the items, leading to systematic revisions that ensured conceptual clarity. Prior to its implementation in the main study, the revised questionnaire was piloted with a sample of 20 participants to assess comprehensibility, feasibility, and item wording. Final adjustments were subsequently made based on the feedback obtained during this pilot phase. The internal consistency of the 22 knowledge items was assessed using Cronbach’s alpha. The instrument demonstrated good overall reliability (α = 0.829), with domain-specific values as follows: Domain 1 (definition, etiology, and clinical features; 8 items): α = 0.605; Domain 2 (periodontal-systemic relationships; 4 items): α = 0.615; Domain 3 (pregnancy-related knowledge; 4 items): α = 0.659; Domain 4 (medications, smoking, and clinical knowledge; 6 items): α = 0.622. Group-specific reliability was also acceptable (IMS: α = 0.733; dentists: α = 0.737). Corrected item-total correlations ranged from 0.246 to 0.578, confirming that all items contributed meaningfully to the overall construct. Formal test-retest reliability and factor analysis were not performed, which is acknowledged as a limitation. The questionnaire consists of two sections. The first section contains 7 sociodemographic questions including age, sex, marital status, education status, graduation year, work experience, and subspecialty. The second section comprises 29 questions assessing knowledge and attitudes regarding periodontal health and periodontal disease, with 26 multiple-choice and 3 multi-select questions. Questions covered the following domains: periodontal disease definition, etiology and clinical features (questions 2–9); periodontal-systemic disease relationships including diabetes (questions 10–13); periodontal disease and pregnancy (questions 14–17); medication effects on periodontal tissues, organ transplantation, and implant-related periodontal disease (questions 18–23); and clinical attitudes and referral practices (questions 24–27). Self-assessment of periodontal knowledge was evaluated before and after completing the survey using a 0–10 numerical rating scale (questions 1 and 28). The source of participants’ periodontal knowledge was assessed in question 29. Statistical analysis Data were transferred from Google Forms to Microsoft Excel and IBM SPSS Statistics Version 20.0 for analysis. Descriptive statistics included frequencies, percentages, means, standard deviations, medians, and interquartile ranges. The normality of all continuous variables was assessed using the Shapiro–Wilk test. As all continuous variables showed significant departures from normality (p < 0.05; see Results), non-parametric tests were used for continuous data comparisons. Categorical variables were analyzed using the chi-square test (or Fisher’s exact test when expected cell counts were < 5). The Mann–Whitney U test was used for comparisons of continuous and ordinal variables between two groups, and the Kruskal–Wallis test was used for comparisons across three or more groups. Paired pre-post self-assessment scores were compared using the Wilcoxon signed-rank test. The significance level was set at p < 0.05. Given the 22 simultaneous comparisons across individual knowledge items, the risk of type I error inflation was addressed by performing a sensitivity analysis with Bonferroni correction (adjusted threshold: p < 0.0023). Post-hoc power analysis was conducted for all primary comparisons. Additionally, multiple linear regression was performed to examine the independent effect of group (IMS vs. dentist) on total knowledge scores after adjusting for potential confounders (age, sex, marital status, and years since graduation). Binary logistic regression was used to calculate adjusted odds ratios (OR) with 95% confidence intervals for key knowledge items. Within the IMS group, a separate linear regression model was constructed to evaluate whether the specialist-resident difference in knowledge was independent of age, sex, and marital status. Variance inflation factors (VIF) were computed to assess multicollinearity; all values were below the threshold of 3, indicating acceptable collinearity levels. Results Demographic characteristics Demographic characteristics stratified by group are presented in Table 1. Of the 240 participants, 167 (69.6%) were female and 73 (30.4%) were male. The IMS group was significantly older than the dentist group (median: 30 vs. 26 years, p < 0.001) and had a higher proportion of married participants (51.4% vs. 30.8%, p = 0.002). Sex distribution did not differ significantly between groups (p = 0.093). The IMS group had earlier graduation years (median: 2017 vs. 2019, p = 0.001). In the IMS group (n = 107), 47 were specialists and 60 were residents. Among the specialists, 30 had subspecialties including endocrinology and metabolism (n = 6), gastroenterology (n = 3), hematology (n = 8), nephrology (n = 1), rheumatology (n = 7), and medical oncology (n = 5). Table 1Demographic characteristics of participants by groupCharacteristicIMS Group (n = 107)Dentist Group (n = 133)p-valueAge (years) Mean ± SD31.8 ± 7.229.4 ± 9.2 Median (IQR)30 (27–34)26 (24–30) < 0.001ᵃ Sex, n (%)0.093ᵇ Female68 (63.6%)99 (74.4%) Male39 (36.4%)34 (25.6%)Marital status, n (%) 0.002ᵇ Married55 (51.4%)41 (30.8%) Single52 (48.6%)92 (69.2%)Graduation year Median (IQR)2017 (2013–2019)2019 (2016–2020) 0.001ᵃ Graduated ≥ 2011, n (%)94 (87.9%)112 (84.2%)0.437ᵇIMS composition, n (%)N/A Specialists47 (43.9%) Residents60 (56.1%)U testBold p-values indicate statistical significance (p < 0.05)ᵃ Mann–Whitneyᵇ Chi-square test Demographic characteristics of participants by group U testBold p-values indicate statistical significance (p < 0.05) ᵃ Mann–Whitney ᵇ Chi-square test Periodontal disease knowledge Table 2 presents the comparative analysis of periodontal disease knowledge between the IMS group and dentists. Across all basic periodontal knowledge questions, IMS participants demonstrated significantly lower correct response rates compared to dentists. The most pronounced knowledge gap was observed in the role of radiographic examination in periodontal diagnosis, where only 37.4% of IMS participants answered correctly compared to 88.8% of dentists (p < 0.001). The definition of periodontitis was correctly identified by 65.4% of IMS versus 98.5% of dentists (p < 0.001). After Bonferroni correction, 6 of 8 items in this domain retained statistical significance. Table 2Periodontal disease knowledge assessment: correct response ratesQuestion (Correct answer)Dentists n (%)IMS n (%)p-valueDefinition of periodontitis (T)131 (98.5)70 (65.4) < 0.001 Clinical signs: bleeding, mobility, recession (T)114 (85.7)71 (66.4) < 0.001 Periodontitis can cause total tooth loss (T)123 (92.5)65 (60.7) < 0.001 Gingival bleeding is an important sign (T)126 (94.7)92 (86.0) < 0.001 Asymptomatic progression is possible (T)110 (82.7)86 (80.4) 0.023 Radiographic examination is important (T)117 (88.8)40 (37.4) < 0.001 Oral hygiene and plaque as etiological factors (T)131 (98.5)98 (91.6) 0.024 Oral complications of periodontitis (T)130 (97.7)98 (91.6) 0.011 Bold p-values indicate statistical significance (p < 0.05)T True statement, IMS Internal Medicine Specialists/Residents Periodontal disease knowledge assessment: correct response rates Bold p-values indicate statistical significance (p < 0.05) T True statement, IMS Internal Medicine Specialists/Residents Periodontal-systemic disease relationships Knowledge regarding the relationship between periodontal disease and systemic conditions showed a more nuanced pattern (Table 3). Regarding the recognition of diabetes as a risk factor for periodontitis, both groups demonstrated high correct response rates (IMS: 93.5% vs. dentists: 92.5%), with IMS participants showing a marginally higher rate. Although this difference reached statistical significance (p = 0.034), the small absolute difference (1%) and loss of significance after Bonferroni correction suggest limited clinical relevance. This was the only item on which IMS participants showed a higher correct response rate than dentists. No significant differences were found regarding periodontal effects on glycemic control (p = 0.914) or the beneficial effect of periodontal treatment on diabetes prognosis (p = 0.148). However, IMS participants demonstrated significantly lower awareness of associations between periodontal disease and cardiovascular diseases, cerebrovascular diseases, peripheral arterial diseases, and rheumatoid arthritis (68.2% vs. 87.2%, p < 0.001; significant after Bonferroni correction). Table 3Knowledge of periodontal-systemic disease relationshipsQuestion (Correct answer)Dentists n (%)IMS n (%)p-valueDM is a risk factor for periodontitis (T)123 (92.5)100 (93.5) 0.034 Periodontitis impairs glycemic control (T)108 (81.2)86 (80.4)0.914Periodontal treatment benefits DM prognosis (T)102 (76.7)71 (66.4)0.148Association with CVD, cerebrovascular, PAD, RA (T)116 (87.2)73 (68.2) < 0.001 T True statement, DM Diabetes mellitus, CVD Cardiovascular disease, PAD Peripheral arterial disease, RA Rheumatoid arthritis Knowledge of periodontal-systemic disease relationships T True statement, DM Diabetes mellitus, CVD Cardiovascular disease, PAD Peripheral arterial disease, RA Rheumatoid arthritis Pregnancy-related periodontal knowledge Significant knowledge gaps were observed regarding the relationship between periodontal disease and pregnancy (Table 4). Less than half of IMS participants (46.7%) recognized the increased risk of adverse pregnancy outcomes in women with periodontal disease, compared to 91% of dentists (p < 0.001). Only 34.6% of IMS correctly identified that tooth loss during pregnancy is not caused by mineral transfer from the mother’s teeth to the fetus, versus 90.2% of dentists (p < 0.001). Similarly, only 41.1% of IMS knew that the second trimester is the optimal period for dental procedures during pregnancy, compared to 91.7% of dentists (p < 0.001). All four items in this domain remained significant after Bonferroni correction. Table 4Knowledge of periodontal disease and pregnancyQuestion (Correct answer)Dentists n (%)IMS n (%)p-valueIncreased pregnancy complication risk (T)121 (91.0)50 (46.7) < 0.001 Gingival enlargement during pregnancy (T)131 (98.5)80 (74.8) < 0.001 Tooth loss due to mineral transfer to fetus (F)*120 (90.2)37 (34.6) < 0.001 Best trimester for dental procedures: 2nd (T)122 (91.7)44 (41.1) < 0.001 T True statement, F False statement (correct answer: selecting “Wrong”)* Percentages reflect those who correctly identified the statement as false Knowledge of periodontal disease and pregnancy T True statement, F False statement (correct answer: selecting “Wrong”) * Percentages reflect those who correctly identified the statement as false Medication effects, smoking, and clinical attitudes Knowledge regarding the effects of medications and smoking on periodontal tissues was significantly lower in the IMS group compared to dentists (Table 5). Only 59.8% of IMS participants knew that smoking accelerates periodontal destruction while suppressing gingival bleeding, compared to 97.7% of dentists (p < 0.001). Awareness of MRONJ was also lower among IMS participants (88.8%) compared to dentists (98.5%, p = 0.003), though this item did not retain significance after Bonferroni correction (adjusted p = 0.079). Regarding clinical attitudes, only 23.4% of IMS participants reported referring patients with suspected dental infections directly to a dentist without prescribing antibiotics or antiseptic mouthwash. The majority (76.6%) prescribed antibiotics and/or antiseptic mouthwash before referral. This referral pattern question was assessed within the IMS group only, as it specifically addresses medical practitioners’ management of dental complaints; therefore, a direct comparison with dentists was not performed for this variable. Only 25.2% of IMS participants frequently recommended oral hygiene practices to their patients, versus 86.5% of dentists (p < 0.001). Table 5Knowledge of medication effects, smoking, organ transplantation, and implant-related periodontal diseaseQuestion (Correct answer)Dentists n (%)IMS n (%)p-valueDrug-induced gingival overgrowth (T)129 (97.0)90 (84.1) < 0.001 MRONJ awareness (T)131 (98.5)95 (88.8) 0.003 Pre-MRONJ dental clearance needed (T)131 (98.5)92 (86.0) < 0.001 Pre-transplant dental clearance needed (T)131 (98.5)92 (86.0) < 0.001 Smoking accelerates destruction and suppresses bleeding (T)130 (97.7)64 (59.8) < 0.001 Periodontal diseases can affect implants (T)124 (93.2)65 (60.7) < 0.001 T True statement, IMS Internal Medicine Specialists/Residents, MRONJ Medication-Related Osteonecrosis of the JawBold p-values indicate statistical significance (p < 0.05) Knowledge of medication effects, smoking, organ transplantation, and implant-related periodontal disease T True statement, IMS Internal Medicine Specialists/Residents, MRONJ Medication-Related Osteonecrosis of the Jaw Bold p-values indicate statistical significance (p < 0.05) Self-assessment and knowledge sources Mean self-assessment scores for periodontal knowledge before the survey were 4.02 ± 2.09 (out of 10) for IMS and 8.11 ± 1.69 for dentists (p < 0.001). After completing the survey, IMS participants revised their self-assessed knowledge score significantly upward to 5.34 ± 2.10 (Wilcoxon signed-rank test, p < 0.001), while dentists also showed a modest but significant increase to 8.41 ± 1.60 (p = 0.007). This unexpected increase among IMS participants, occurring despite substantial objective knowledge gaps, is discussed further in the Discussion section. Among IMS participants, the primary sources of periodontal knowledge were medical education (66.4%), information exchange with dentists (36.4%), media (16.8%), and scientific publications (15%). Notably, 15% reported having no prior knowledge on the subject. Subgroup analysis: Specialists versus residents Exploratory subgroup analyses comparing internal medicine specialists (n = 47) and residents (n = 60) revealed significant differences across all knowledge domains (Table 6). Specialists were significantly older than residents (37.4 ± 7.5 vs. 27.4 ± 1.9 years, p < 0.001), while sex distribution was comparable (female: 61.7% vs. 65.0%, p = 0.881). Table 6Knowledge domain scores: Specialists vs. ResidentsDomain (score range)Specialists (n = 47)Residents (n = 60)p-valueᵃDomain 1: Definition/Etiology (0–8)6.36 ± 1.295.35 ± 1.68 0.001 Domain 2: Systemic (0–4)3.40 ± 0.992.83 ± 1.21 0.004 Domain 3: Pregnancy (0–4)2.23 ± 1.151.77 ± 1.00 0.024 Domain 4: Meds/Smoking (0–6)4.94 ± 1.114.43 ± 1.32 0.031 Total Score (0–22) 16.94 ± 2.90 14.38 ± 3.73 < 0.001 Values are presented as mean ± SDBold p-values indicate statistical significance (p < 0.05)ᵃ Mann–Whitney U test Knowledge domain scores: Specialists vs. Residents Values are presented as mean ± SD Bold p-values indicate statistical significance (p < 0.05) ᵃ Mann–Whitney U test Specialists demonstrated significantly higher total knowledge scores than residents (16.94 ± 2.90 vs. 14.38 ± 3.73, p < 0.001). Domain-specific analyses showed that specialists outperformed residents in Domain 1 (definition and etiology: 6.36 ± 1.29 vs. 5.35 ± 1.68, p = 0.001), Domain 2 (systemic relationships: 3.40 ± 0.99 vs. 2.83 ± 1.21, p = 0.004), Domain 3 (pregnancy: 2.23 ± 1.15 vs. 1.77 ± 1.00, p = 0.024), and Domain 4 (medications and smoking: 4.94 ± 1.11 vs. 4.43 ± 1.32, p = 0.031). At the item level, the most pronounced differences were observed for clinical signs of periodontitis (83.0% vs. 53.3%, p = 0.003), the effect of periodontitis on glycemic control (93.6% vs. 70.0%, p = 0.005), CVD/PAD/RA associations (80.9% vs. 58.3%, p = 0.023), and identification of the optimal trimester for dental procedures during pregnancy (59.6% vs. 26.7%, p = 0.001). In contrast, several items showed comparable knowledge between specialists and residents, including the definition of periodontitis (p = 0.124), radiographic examination (p = 1.000), diabetes as a risk factor (p = 1.000), and smoking effects (p = 0.878). Pre-survey self-assessment scores were significantly higher among specialists than residents (4.83 ± 2.07 vs. 3.38 ± 1.89, p < 0.001), and this difference persisted post-survey (6.28 ± 1.88 vs. 4.60 ± 1.98, p < 0.001). Multivariate analysis Multiple linear regression confirmed that the group effect on total knowledge scores was independent of demographic confounders (Table 7). After adjusting for age, sex, marital status, and years since graduation, being a dentist was associated with 4.65 points higher knowledge (B = 4.647, 95% CI: 3.86–5.44, p < 0.001), an estimate virtually unchanged from the unadjusted model (B = 4.804). Years since graduation was independently associated with higher scores (B = 0.186, p = 0.006), while sex and marital status were not significant predictors. The model explained 42.9% of the variance in knowledge scores (R² = 0.429). Table 7Multiple linear regression models for total knowledge score (0–22)VariableB95% CI p B95% CI p Model 1 (unadjusted)Model 2 (adjusted)Group (Dentist vs. IMS) 4.855 [4.09, 5.62] < 0.001 4.647 [3.86, 5.44] < 0.001 Age (years)———-0.166[-0.29, -0.04]0.011Sex (Male vs. Female)———-0.038[-0.86, 0.78]0.928Marital status (Married)———0.629[-0.23, 1.49]0.152Years since graduation———0.186[0.06, 0.32] 0.006 R² 0.401 0.429 Model 1: unadjusted. Model 2: adjusted for age, sex, marital status, and years since graduationAll VIF values < 3, indicating acceptable multicollinearity Multiple linear regression models for total knowledge score (0–22) Model 1: unadjusted. Model 2: adjusted for age, sex, marital status, and years since graduation All VIF values < 3, indicating acceptable multicollinearity Binary logistic regression for key knowledge items confirmed that group differences remained highly significant after adjustment (Table 8). Dentists had dramatically higher adjusted odds of correct responses for periodontitis definition (adjusted OR = 56.17, p < 0.001), radiographic examination (OR = 17.99, p < 0.001), pregnancy complications (OR = 11.29, p < 0.001), mineral transfer misconception (OR = 17.26, p < 0.001), best trimester identification (OR = 19.44, p < 0.001), and smoking effects (OR = 30.20, p < 0.001). Notably, diabetes-related items showed no significant group difference after adjustment (DM as risk factor: OR = 0.70, p = 0.513; glycemic control: OR = 1.00, p = 0.997), confirming that comparable diabetes knowledge is a genuine finding independent of demographic confounders. Table 8Adjusted odds ratios for correct responses: Dentists vs. IMS (reference)Knowledge itemUnadj ORAdj ORᵃ95% CIp-valuePeriodontitis definition34.62 56.17 [12.34–255.60] < 0.001 Radiographic examination12.25 17.99 [8.50–38.04] < 0.001 DM as risk factor0.860.70[0.23–2.07]0.513Glycemic control1.051.00[0.50–1.99]0.997Pregnancy complications11.50 11.29 [5.39–23.67] < 0.001 Mineral transfer (False)17.46 17.26 [8.20–36.33] < 0.001 Best trimester15.88 19.44 [8.67–43.61] < 0.001 Smoking effects29.11 30.20 [8.70–104.88] < 0.001 OR > 1 indicates higher odds of correct response for dentists compared to IMS (reference group)ᵃ Adjusted for age, sex, marital status, and years since graduation Adjusted odds ratios for correct responses: Dentists vs. IMS (reference) OR > 1 indicates higher odds of correct response for dentists compared to IMS (reference group) ᵃ Adjusted for age, sex, marital status, and years since graduation Within the IMS group, specialist status was independently associated with 2.65 points higher knowledge after adjusting for age, sex, and marital status (B = 2.654, 95% CI: 0.75–4.56, p = 0.007; Table 9). Notably, age was not a significant predictor (p = 0.885) when specialist status was included in the model, indicating that the specialist-resident knowledge gap reflects differences in training level and clinical experience rather than age per se. Table 9IMS-only linear regression: Predictors of total knowledge score (n = 107)VariableB95% CIp-valueSpecialist (vs. Resident) 2.654 [0.75, 4.56] 0.007 SignificantAge (years)-0.010[-0.14, 0.12]0.885Not significantSex (Male vs. Female)-0.373[-1.77, 1.02]0.596Not significantMarital (Married)0.021[-1.50, 1.54]0.979Not significantR² = 0.127, Adj R² = 0.093 IMS-only linear regression: Predictors of total knowledge score (n = 107) Discussion This study examined the knowledge levels and attitudes of internal medicine specialists and residents in Turkey regarding periodontal health and periodontal disease, using dentists as a reference group. While the comparison with dentists provides a useful benchmark, the more clinically meaningful insight lies in characterizing the nature and significance of the knowledge gaps within the IMS group and identifying domains where knowledge is relatively preserved. A hierarchical pattern of knowledge deficits emerged among IMS participants. Knowledge was relatively preserved in domains with direct clinical relevance to internal medicine practice, particularly the diabetes-periodontitis relationship, where over 93% correctly identified diabetes as a risk factor. In contrast, knowledge was markedly deficient in domains less directly encountered in internal medicine training: pregnancy-periodontitis associations (46.7% correct), smoking’s dual effect on periodontal tissues (59.8%), and the role of radiographic examination (37.4%). This pattern suggests that IMS participants’ periodontal knowledge is shaped by clinical exposure to specific comorbidities rather than by systematic oral health education, with important implications for curriculum development. Regarding the basic definition of periodontitis, 65.4% of IMS participants answered correctly, which is comparable to findings from Nigeria (68.8%) [11] but higher than those reported from India (43.3–51.7%) [15, 22]. Interestingly, IMS participants demonstrated relatively strong knowledge about gingival bleeding as an important periodontal sign (86%), consistent with the understanding that bleeding is a cardinal symptom of periodontal disease, while similar studies in Nigeria reported a much lower rate (11%) [13]. The relatively high awareness of the diabetes-periodontitis relationship among IMS participants deserves particular attention. Over 93% correctly identified diabetes as a risk factor for periodontitis, and approximately 80% recognized the adverse effect of periodontitis on glycemic control. These results are comparable to findings from Hong Kong (90% and 75.7%, respectively) [23] and North Carolina (78% and 82%, respectively) [14], and substantially higher than those reported from Nigeria (13.7%) [13], Kuwait (44%) [24], and Jordan (32–50%) [20]. This relatively favorable result likely reflects the direct clinical relevance of the diabetes-periodontitis relationship to internal medicine practice and the growing emphasis on this association in medical education. Notably, the diabetes risk factor item was the only instance in which IMS participants showed a marginally higher correct response rate than dentists (93.5% vs. 92.5%, p = 0.034), though this finding did not survive Bonferroni correction and the absolute difference was minimal. In contrast, IMS participants showed marked knowledge deficits in several other domains. The pregnancy-periodontitis relationship was correctly identified by less than half of IMS participants (46.7%), compared to 91% of dentists. This finding is concerning given the established associations between periodontal disease and adverse pregnancy outcomes including preeclampsia, preterm birth, and low birth weight [2, 25]. International comparisons show similar patterns, with correct response rates ranging from 11% in Nigeria [13] to 83.6% in Brazil [26]. The widespread misconception that pregnancy-related tooth loss results from mineral transfer to the fetus (believed by 29% of IMS participants) further underscores the need for targeted education. The particularly low awareness in this domain may reflect the fact that pregnancy management is not a core focus of internal medicine training, resulting in limited exposure to obstetric-periodontal interactions. The low awareness of smoking’s dual effect on periodontal tissues—accelerating destruction while masking gingival bleeding—among IMS participants (59.8%) is clinically significant. Smoking is one of the most important modifiable risk factors for periodontitis, and its vasoconstrictive effect on gingival tissues can mask clinical signs of inflammation, potentially delaying diagnosis [27]. This finding is particularly noteworthy because smoking cessation counseling is a routine component of internal medicine practice, yet the periodontal implications of smoking appear to be largely overlooked in medical training. The finding that only 23.4% of IMS participants would refer patients with suspected dental infections directly to a dentist without prescribing antibiotics or mouthwash reveals a concerning self-reported pattern of practice. The majority (76.6%) reported prescribing antibiotics and/or antiseptic mouthwash before referral, potentially contributing to antibiotic overuse and delaying definitive dental treatment. This finding highlights the communication gap between medical and dental practitioners and the need for clearer referral pathways. However, it should be noted that this represents self-reported behavior, which may not accurately reflect actual clinical practice patterns. Furthermore, given that data collection occurred in 2021, it is possible that antimicrobial prescribing practices among internal medicine physicians may have evolved in the interim period, particularly in light of growing global and national emphasis on antimicrobial stewardship programs. Future studies should reassess these prescribing patterns to determine whether the increased awareness of antimicrobial resistance has influenced referral and prescribing behaviors in the context of dental infections. The subgroup analysis comparing specialists and residents yielded important findings. Specialists scored significantly higher than residents across all four knowledge domains (total score: 16.94 vs. 14.38, p < 0.001), suggesting that clinical experience and advanced training contribute to the accumulation of periodontal knowledge. The most pronounced differences were observed for items related to clinical manifestations (clinical signs: 83.0% vs. 53.3%, p = 0.003) and systemic interactions (glycemic control: 93.6% vs. 70.0%, p = 0.005), domains where direct clinical exposure is most likely to reinforce learning. However, both subgroups showed similarly low knowledge regarding radiographic examination (38.3% vs. 36.7%, p = 1.000) and smoking effects (61.7% vs. 58.3%, p = 0.878), indicating that these deficits are structural and unlikely to be resolved by clinical experience alone. These findings support the need for formal educational interventions rather than reliance on experiential learning. The self-assessment findings provide additional insight. Pre-survey self-assessment scores were low (4.02/10), indicating that IMS participants were aware of their limited periodontal knowledge. After completing the survey, IMS participants revised their self-assessed knowledge significantly upward to 5.34 (p < 0.001). This is an unexpected finding, particularly given the concurrent demonstration of substantial knowledge gaps across multiple domains. The increase in self-assessment scores is difficult to reconcile with the objective knowledge deficits and cannot be adequately explained within a cross-sectional design. Possible explanations include a calibration effect following exposure to detailed survey questions or an anchoring bias; however, these remain speculative, and this finding warrants further investigation through longitudinal study designs. Specialists reported higher self-assessment scores than residents both before (4.83 vs. 3.38, p < 0.001) and after (6.28 vs. 4.60, p < 0.001) the survey. Among specialists, no significant differences in knowledge scores were found across subspecialties (Kruskal–Wallis H = 6.969, p = 0.324), though the small subgroup sizes limit interpretation. Descriptively, endocrinologists (18.40 ± 2.30) and medical oncologists (18.60 ± 2.19) showed the highest scores, possibly reflecting greater clinical exposure to periodontal-systemic interactions in their practice, while gastroenterologists scored lower (15.33 ± 3.79). These exploratory findings warrant confirmation in larger studies. Recent advances in artificial intelligence (AI) and machine learning offer promising avenues for periodontal disease detection that may help bridge interdisciplinary knowledge gaps. Wu et al. [28] demonstrated that machine learning approaches can achieve accurate periodontal disease classification from medical images, while Jafer et al. [29] provided a comprehensive review of AI applications in periodontal diagnosis and prediction. These emerging technologies have the potential to serve as clinical decision support tools for non-dental healthcare providers, enabling more accurate identification of patients requiring periodontal referral. The multivariate analyses provide important methodological reassurance. The near-identical group effect in unadjusted (B = 4.80) and fully adjusted (B = 4.65) linear regression models confirms that the observed IMS-dentist knowledge gap is a genuine training-related difference rather than an artifact of demographic imbalance. The logistic regression results further demonstrate that the adjusted odds of correct responses for most items were 11–56 times higher among dentists, while diabetes-related items showed no group difference—a pattern consistent with domain-specific, clinically mediated knowledge acquisition. The IMS-only regression revealed that specialist status, not age, was the primary driver of within-group knowledge differences (specialist B = 2.654, p = 0.007; age B = -0.010, p = 0.885). This finding suggests that clinical experience and advanced training genuinely contribute to periodontal knowledge accumulation beyond what would be expected from aging alone, further supporting the value of targeted educational interventions at earlier stages of medical training. This study has several limitations that warrant careful consideration. First, although the instrument demonstrated good overall internal consistency (Cronbach’s α = 0.829), formal test-retest reliability and more comprehensive validity assessments (construct validity, criterion validity) were not performed, limiting the ability to fully evaluate the measurement properties of the instrument. Second, the use of convenience sampling via social media platforms limits the representativeness of the findings. The response rate could not be precisely calculated, and the self-selection of participants may have introduced selection bias, as those with greater interest in periodontal health may have been more likely to participate. Third, the professional credentials of participants were not formally verified. While detailed sociodemographic questions helped establish respondent profiles, the absence of direct credential verification should be acknowledged. Fourth, the cross-sectional design precludes establishing causal relationships. Fifth, self-reported knowledge and referral patterns may not accurately reflect actual clinical behavior. Sixth, specialists and residents were combined for the primary analysis, though significant differences between these subgroups were identified. Stratification data were not collected for the dentist group. Seventh, multiple comparisons increase the risk of type I error; a Bonferroni correction was applied, and 14 of 22 comparisons retained significance. Eighth, data collection occurred in 2021; although the core periodontal-systemic disease relationships remain current, and no major curricular reform has been implemented in Turkey since then, the time elapsed should be considered. Given that this is a 2026 submission based on 2021 data, physicians’ attitudes and clinical practices may have evolved during this interval, particularly with respect to antimicrobial stewardship awareness and the increasing integration of oral-systemic health concepts into continuing medical education. Readers should therefore interpret these findings as reflecting the state of knowledge and practice at the time of data collection. Ninth, although multivariate analyses were performed to control for measured confounders (age, sex, marital status, years since graduation), other unmeasured confounders (e.g., institutional training environment, prior interprofessional exposure, personal interest in oral health) could not be accounted for and may have influenced the results. Finally, the study was conducted in Turkey, and generalizability to other countries should be interpreted cautiously. Several directions for future research are recommended: (a) multi-center studies with stratified random sampling; (b) rigorously validated instruments with comprehensive psychometric assessment including test-retest reliability, construct validity, and criterion validity; (c) subgroup analyses by subspecialty with adequately powered samples; (d) exploration of additional confounding variables through more comprehensive multivariate models; (e) longitudinal studies assessing knowledge changes following educational interventions; and (f) investigation of AI-based clinical decision support tools for periodontal screening in medical settings [28, 29]. Despite these limitations, the study provides valuable insights into the state of periodontal knowledge among internal medicine physicians in Turkey. The finding that 66.4% of IMS participants obtained their periodontal knowledge from medical education, yet demonstrated significant knowledge gaps, suggests that the current medical curriculum may not adequately address oral-systemic health relationships. This is consistent with international observations that medical curricula generally lack sufficient oral health content [18, 30, 31]. Conclusions The findings of this cross-sectional survey suggest that internal medicine specialists and residents in this Turkish sample may have insufficient knowledge regarding periodontal health and periodontal disease, and the relationship between periodontal and systemic conditions compared to dentists. A hierarchical pattern of knowledge gaps emerged, with relatively preserved knowledge in domains directly relevant to internal medicine practice (particularly diabetes-periodontitis relationships), but marked deficits in pregnancy-related periodontal issues, smoking’s effects on periodontal tissues, and diagnostic and referral practices. Specialists demonstrated significantly higher knowledge than residents across all domains, suggesting that clinical experience contributes to knowledge accumulation, though structural deficits in radiographic and smoking-related knowledge persisted regardless of training level. These findings point toward the potential value of integrating oral health education into medical curricula and fostering interprofessional collaboration. However, given the study’s limitations—including convenience sampling, the absence of test-retest reliability data, and the cross-sectional design—these recommendations should be interpreted cautiously and confirmed through larger, methodologically rigorous studies. Supplementary Information Supplementary Material 1. Acknowledgements The data presented in this manuscript were derived from the dental specialty thesis of Beyza Bozmaoğlu. The authors thank all the participants who voluntarily completed the questionnaire. Artificial intelligence use During the preparation of this work, the authors used Gemini (Google) exclusively for English language polishing (grammar, syntax, and phrasing). Gemini was not used for generating scientific content, interpreting data, formulating hypotheses, or drawing conclusions. All AI-suggested changes were individually reviewed, verified for accuracy, and approved by all authors. The authors take full responsibility for the content of the publication. Abbreviations Artificial intelligence Cardiovascular disease Diabetes mellitus Internal medicine specialists/residents Interquartile range Medication-related osteonecrosis of the jaw Peripheral arterial disease Rheumatoid arthritis Strengthening the Reporting of Observational Studies in Epidemiology Authors' contributions BB collected data, performed statistical analysis, interpreted results and drafted the manuscript. EY supervised the study, contributed to study design, data interpretation and critically revised the manuscript. ADG conducted data interpretation. YTA contributed to data interpretation and improved the manuscript. TMA contributed to study design, provided methodological guidance and critically revised the manuscript. All authors read and approved the final manuscript. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Data availability The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate This study was approved by Kocaeli University Non-Interventional Clinical Research Ethics Committee (Decision No: GOKAEK-2021/6.03, Date: 18.03.2021). All methods were performed in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants; completion of the questionnaire following the information page was accepted as implied consent. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.