A cross-sectional study of 72 T1D and 86 HC subjects was conducted at the Diabetes and Dental Units of the Institute for maternal and child health IRCCS “Burlo Garofolo” (Trieste, Italy) between May 2022 and December 2023 during routine outpatient visits.

Inclusion criteria for T1D participants were: diagnosis for at least 1 year, age between 6 and 21 years, and absence of other types of diabetes mellitus (i.e., type 2, monogenic diabetes, cystic fibrosis-related diabetes). For HC selection, subjects similar to T1D subjects in terms of sex and age were included. Then, subjects with HbA1c > 6% (> 42 mmol/mol), measured with finger pricks using portable instrumentation (QuikRead go, A. De Mori S.p.A., Milan, Italy), were excluded. Furthermore, subjects diagnosed with any other form of diabetes, obesity and other metabolic and autoimmune disorders, and family history of diabetes, were excluded.

Additionally, all subjects (both T1D and HC) with immunodeficiencies and neoplasms and who had taken antibiotics or anti-inflammatory drugs within 14 days prior to enrolment were excluded.

Each subject who met the eligibility criteria and presented to the clinic during the study period was invited to participate in the study.

At the time of enrolment, demographic data, including age, sex, and anthropometric measurements were collected for all participants. BMI standard deviation scores (BMI-SDS) were calculated using WHO reference charts [25] via the Growth Calculator 4 software (http://www.weboriented.it/gh4/).

For T1D subjects, clinical data at onset, including age at onset and the presence of ketoacidosis (DKA) (calculated as a venous pH of < 7.3 and/or a bicarbonate (HCO3) level of < 18 mmol/L) [26], type of diabetes therapy, insulin requirements and HbA1c from capillary blood measurements over the past year, were collected by medical records. HbA1c and insulin requirements at recruitment were used to calculate IDAA1c (Insulin-Dose Adjusted A1c) using the formula: HbA1c (%) + 4 × insulin dose (units/kilogram/day) [27]. T1D subjects were categorized into two groups: poor glycaemic control (PGC) and good glycaemic control (GGC) using as cut-off the median HbA1c value over the past year (HbA1c ≥ 7% or < 7%, ≥ or < 53 mmol/mol, respectively) [28].

Finally, data from the most recent visit were retrieved, including: fasting lipid profile (total cholesterol [TC]; low-density lipoprotein cholesterol [LDL-C]; high-density lipoprotein cholesterol [HDL-C]; triglycerides [TG]), serum and urine creatinine, albumin-to-creatinine ratio (ACR). Estimated glomerular filtration rate (eGFR) were calculated using the Schwartz equation: 0.413 × height (cm)/serum creatinine mg/dL [29].

The research project was approved by ethic committee (CEUR-2018-Em-323-Burlo). Before the enrolment all participants or their parents/guardians (for participants aged < 18 years) provided written informed consent.

At the time of enrolment, each participant completed a questionnaire on oral hygiene habits, which included questions about the number of daily brushings, the methods used for oral cleaning (manual toothbrush, electric toothbrush, dental floss, etc.), the frequency of dental visits, and the reasons for attending such visits.

The oral clinical examination was conducted using a mirror and a CP12 Periodontal Probe. The oral-dental assessment included the definition of the state of the dentition and dental formula, the state of caries and periodontal health, and salivary pH. Validated standard assessment tools were used. Dental health was assessed by recording the percentage of caries in the dentition (permanent, mixed, deciduous), excluding third molars (range 0–28) out of the total number of teeth in the arch. Unerupted, congenitally lost, or supernumerary teeth, teeth extracted for orthodontic reasons, teeth filled for reasons other than caries, and non-exfoliated deciduous teeth in the permanent dentition were not taken into consideration. Where clinically indicated, intraoral x-rays were performed, or recently performed panoramic x-rays were requested.

Periodontal health was assessed by calculating the bleeding on probing index (BoP), recording, for each tooth, the bleeding sites after gingival probing at six sites (mesial, mid, and distal on both the buccal and lingual surfaces). This number was divided by the total number of sites available in the mouth and multiplied by 100. The BoP index is expressed as a percentage.

The O’Leary Plaque Index (PI) was used to assess the oral hygiene status of individuals. After applying a plaque disclosing agent (Mira-2-Ton®, Hager & Werken GmbH & Co., Duisburg, Germany), the number of stained surfaces was recorded for each tooth. Four surfaces were evaluated for each tooth (distal, mesial, buccal, and lingual). PI was calculated by dividing the total number of surfaces by the surfaces with plaque. Third molars were excluded from this count.

Moreover, BoP (%), PI (%), and Periodontal Pocket Depth (PPD, measured in mm) parameters were assessed, and subjects were classified into three categories: Health (BoP ≤ 10%, PI ≤ 25%, PPD ≤ 3 mm); Plaque-induced gingivitis (BoP > 10%, PI > 25%, PPD ≤ 3 mm); Non-plaque-induced gingivitis (BoP > 10%, PI < 25%, PPD ≤ 3 mm) (Adapted from Chapple ILC, et al., 2018) [30]. Given the extremely small number of subjects with non-plaque-induced gingivitis, these individuals were excluded from the study.

For each participant 2 mL of saliva were collected and immediately stored at -80 °C until further processing. A panel of cytokines, chemokines and growth factors was assessed in all the saliva samples using magnetic bead-based multiplex immunoassays (Bio-Plex Pro™ human cytokine; Bio-Rad Laboratories, Milan, Italy). Samples were centrifuged al 10,000 g for 10 min at room temperature (RT) prior to analysis. Assays were performed according to the manufacturer’s instructions.

The concentrations of the immune soluble factors were determined using the Bio-Plex-200 system (Bio-Rad Corp., Hercules, CA, USA) and analysed with Bio-Plex Manager software (v.6; Bio-Rad). Results were reported as concentration (pg/mL).

Contingency tables were used to summarize demographic and anthropometric data, oral hygiene habits, dental visit information, and clinical and inflammatory parameters.

Categorical variables were expressed as percentages (%), while continuous variables were assessed for normality using the Shapiro–Wilk test and reported as mean ± standard deviation (sd) or median and interquartile range (IQR), as appropriate. Cytokines and chemokines that were not normally distributed were log-transformed.

Differences between HC and all T1D subjects, as well as between T1D participants stratified by GCC and PGC, were evaluated using the χ2 test or Fisher’s test for categorical variables, and the t test or non-parametric Mann–Whitney U test for continuous variables. The same statistical approach was applied to compare clinical outcomes and salivary cytokine levels in T1D subjects with and without oral diseases.

Causal mediation analyses were conducted to assess whether the associations between salivary cytokine levels and oral pathologies (caries, gingivitis) were mediated by glycaemic control (defined by HbA1c values) or diabetes status. For mediation analysis, two linear regression models were specified: one modelling the mediator as a function of the exposure, and one modelling the outcome as a function of both the exposure and the mediator. Mediation effects were estimated using the mediation R package, with non-parametric bootstrapping (1.000 simulations) to compute 95% confidence intervals (CI). The Average Causal Mediation Effect (ACME), Average Direct Effect (ADE), Total Effect, and Proportion Mediated were reported.

Statistical significance was defined as p-value ≤ 0.05. All statistical analysis were performed using R software v.2023.12 (www.r-project.org).