Abstract
Background
Evidence suggests a bidirectional relationship between oral health status and type 2 diabetes (T2D) in adults. Studies on associations between childhood oral health and T2D in adulthood are lacking.
Methods
This is a nationwide Danish registry-based cohort study of individuals born between 1963 and 1972, having at least one registration in the National Child Odontology Registry between 1972 and 1987 (n = 627,758). Follow-up lasted from 1995 to 2018. Main exposure variables were the highest achieved levels of dental caries and gingivitis between 1972 and 1987. The outcome was T2D diagnosis during follow-up. Data was analyzed using Cox-regression, stratified on sex, with age as the underlying timescale and highest achieved level of education between age 25–30 years as Cox-strata. Main analyses were conducted with and without age-restrictions (T2D diagnosis before/after age 40).
Results
Compared to lowest-level references, high levels of gingivitis associated with increased hazard ratios (HRs) of T2D in both males (HR [95% confidence interval]: 1.59 [1.47; 1.72]) and females (1.87 [1.68; 2.08]), as did severe dental caries (males: (1.15 [1.04; 1.27], in females: 1.19 [1.06; 1.35]). Below age 40, gingivitis associated with increased HRs in males (1.84 ([1.58; 2.15]) and females (1.94 [1.63; 2.30]). Above age 40, both exposures displayed higher HRs in males (high gingivitis: 1.52 [1.39; 1.66] vs. severe caries: 1.23 [1.09; 1.38]) and females (1.83 [1.59; 2.10] vs. 1.37 [1.17; 1.59]).
Conclusions
Data suggest an association between childhood dental caries and gingivitis with risk of receiving a T2D diagnosis in adulthood. However, results are affected by residual confounding warranting further studies.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00592-024-02437-4.
Keywords: Cohort studies, Dental caries, Diabetes mellitus, Type 2, Gingivitis
Introduction
Evidence points to a bidirectional relationship between oral health status and type 2 diabetes (T2D) in adults. Dental caries (DC), periodontitis, and gingivitis are the most common oral diseases, as well as amongst the most prevalent diseases worldwide [1]. DC is as a localized, chemical dissolution of a tooth surface caused by the metabolic activity of the tooth surface microbiota [2]. Periodontitis is an aggressive inflammation of the tooth-supporting tissues, which can eventually lead to tooth loss [3]. Gingivitis covers a wide range of disease states ranging from light bleeding by a single tooth to whole-mouth bleeding and inflammation, and can be seen as a precursor of periodontitis [4–6].
The association between periodontitis and T2D is the most well documented, suggesting that periodontitis treatment improves glycemic control in individuals with T2D [7, 8]. The association between gingivitis and T2D has not been as extensively explored. Nevertheless, studies of 2.144–93.647 individuals find higher levels of gingivitis in individuals with T2D compared to controls [9, 10]. Both gingivitis and periodontitis are inflammation-mediated diseases [5], and correspondingly systemic low-grade inflammation has been suggested as the link to systemic diseases like obesity and T2D [11]. Evidence for an association between DC and T2D is limited, but some studies indicate that individuals with poorly controlled T2D have higher rates of DC when compared to metabolically healthy individuals [12]. Evidence on whether DC and gingivitis can predispose to the development of diabetes is lacking, and the mechanisms potentially linking DC with systemic disease remain to be uncovered.
DC and gingivitis are highly prevalent in children and adolescents [1, 13]. Given the associations observed in adults, DC and gingivitis in childhood potentially have a lifelong impact on systemic health. A study of the association between childhood oral health and the adulthood risk of cardiovascular disease in 755 individuals found a higher prevalence of cardiovascular risk factors, risk of subclinical atherosclerosis and metabolic syndrome in adulthood, amongst those with oral infections or inflammation in childhood [14, 15]. However, there is a lack of studies on the association between childhood oral health and T2D in adulthood.
Data from The Danish National Child Odontology Register (SCOR) combined with data from the Register of Selected Chronic Diseases (RUKS) and the Civil Registration System (CRS) presents a unique opportunity to investigate the link between childhood oral disease and T2D in adulthood. In this paper we utilize this to first test the hypothesis that there is an association between childhood experience of DC and gingivitis with adulthood T2D. Secondarily, we hypothesize that the age at which the highest registered level of disease occurs, as well as the duration and severity of oral disease is associated with the incidence of T2D, which we examine in a subset of the study population.
Methods
This is a nationwide Danish registry-based cohort study of individuals born between 1963 and 1972, and with at least one registration in SCOR between 1972 and 1987. Individuals were followed through registry data from January 1st, 1995, or age 25 (whichever occurred last), and were registered with an event upon receiving a diagnosis of T2D or censored upon first instance of either a diagnosis with type 1 diabetes (T1D), death, disappearance, emigration, or by the end of follow-up on December 31st, 2018.
Registries
Information was gathered through linkage of relevant registries using the unique Danish personal identification number (CPR-number): (1) SCOR data on childhood DC and gingivitis from 1972 to 1987, date of birth, and sex. (2) RUKS for data on diagnosis of T1D and T2D between January 1st, 1995, and December 31st, 2018. To be registered with T1D or T2D in RUKS either two relevant diagnoses (ICD-8/ICD-10: 249/E10 [T1D], 250/E11 [T2D]), or at least two registered purchases of antidiabetic prescription drugs are needed. A detailed description of the criteria for being registered with diabetes in RUKS can be found elsewhere [16]. For T1D, we identified diagnosed individuals to be able to censor them. 3) CRS provided data on status (alive, dead, disappeared, or emigrated from August 8th, 1967, to December 31st, 2018, and educational status from December 1st, 1981, to September 1st, 2016 (Fig. 1).
Fig. 1.
Data extraction process. RUKS Register of Selected Chronic Diseases
Exclusion criteria
Individuals were excluded from the study if they had been registered with codes denoting a faulty or invalid CPR-number, or if they had an event before beginning of their follow-up. Individuals without a registration of their level of education between ages 25 and 30 were also excluded from the study. For secondary analyses a subset of the study population with at least one registration at both ages 0–5, 6–11 and 12–18 were used (Fig. 2).
Fig. 2.
Selection of study participants. Flowchart of the selection of study participants for main and secondary (subset) analyses
Variables used in main analyses
Educational length was grouped into:
Short: Mandatory schooling between ages 6–16.
Medium: Secondary school and vocational education.
Long: Higher education of any duration.
Based on this classification we used the highest attained level of education observed during ages 25–30, which should be close to the individuals’ final adult education level.
Gingivitis was recorded using the Silness and Löe gingival index [4], and measured on four teeth (52, 55, 72 and 75 in the temporary dentition, 12, 16, 32 and 36 in the permanent dentition). An index-score ranging from 0 to 12 for the severity of gingivitis at each visit to the dentist was obtained by summing the index-score for each of the examined teeth [17]. The index-scores were 0 (healthy gingiva), 1 (slight changes in color and structure of gingiva but no bleeding), 2 (moderately inflamed gingiva, bleeding on probing), 3 (severely inflamed gingiva, bleeding, and ulceration). Based on odontological considerations this measure was categorized into gingivitis score (GS), using the highest registered sum index-score at any visit:
Low: Score of 0–4.
Moderate: Score of 5–8.
High: Score of 9–12.
The scoring of DC in SCOR has been described in detail previously [17]. The measure of DC used in this study is the number of destroyed, missing, and filled teeth in either the primary, permanent, or mixed dentition (dmft/DMFT). This was calculated for each visit as the sum of codes denoting cavity, filling, and extraction due to DC. Based on odontological considerations the dmft/DMFT measures were used as a categorical variable, dental caries score (DCS), using the highest registered dmft/DMFT in any dentition at any point in time:
Low: 0–4 affected teeth.
Moderate: 5–12 affected teeth.
High: 13–16 affected teeth.
Severe: ≥ 17 affected teeth.
Variables used in secondary analyses
For the secondary analyses we calculated the duration of DC and gingivitis by summing the number of periods in time (between ages 0–5, 6–11 and 12–18, roughly corresponding to the temporary, mixed, and permanent dentition), during which an individual had been registered as having DC (dmft/DMFT > 0) and gingivitis (gingivitis index-score > 0), resulting in a score from 0 to 3 for each disease. For the same subset of the study population, total disease severity over time was grouped based on the level of disease between the three examined timepoints, for gingivitis:
Low: gingivitis index score ≤ 4 at all 3 age ranges.
Moderate: gingivitis index score of 5–8 by at least one age range, no score ≥ 9.
High: gingivitis index score ≥ 9 by at least one age range.
For DC:
Low: dmft/DMFT ≤ 4 at all 3 age ranges.
Moderate: dmft/DMFT of 5–12 by at least one age range, no dmft/DMFT ≥ 13.
High: dmft/DMFT ≥ 13 by at least one age range.
Statistical methods
Baselines characteristics were described using frequencies and percentages for categorical variables and means and standard deviations for continuous variables.
Data analyses were done using Cox-regression modelling. Martingale residuals were plotted to assess the functional form and outliers for all continuous variables. To assess whether the Cox model assumptions were met Schoenfeld residuals, log–log curves, and cumulative hazards were plotted. The log–log curves for DCS and sex, respectively, were not parallel before age 40. Ideally this would have been solved by doing an interaction with time, however, due to limitations of the server we were working on, this was not possible. Instead, we examined whether there were any substantial time-dependent associations by running all analyses separately for males and females. Thus, primary analyses were run separately for time-at-risk below and above age 40 years. Further, the education-variable did not live up to the proportional-hazards assumption. The variable was used as strata in our models, allowing for differing baseline hazards between groups with differing education levels, thereby implicitly adjusting for the confounding effect of education in the models. Hazard ratios (HR) and 95% confidence intervals for the association between DCS and GS were produced using Cox-regression analysis with age as the underlying timevariable (see Fig. 3).
Fig. 3.
Plots of the cumulative hazard of T2D based on childhood DCS and GS, stratified by sex. Panel A The cumulative hazard of T2D in males based on the highest registered childhood DCS. Panel B The cumulative hazard of T2D in females based on the highest registered childhood DCS. Panel C The cumulative hazard of T2D in males based on the highest registered childhood GS. Panel D The cumulative hazard of T2D in females based on the highest registered childhood GS. DCS dental caries score, GS gingivitis score
All models were stratified by sex as we theorized that sex may moderate any associations between oral health and T2D due the protective effect of estrogen in adult females [18].
Main analyses:
Model (1) DCS + GS + strata (highest achieved level of education).
Model (2) DCS + GS + DCS * GS + strata (highest achieved level of education).
Secondary analyses:
Model (1) Age at highest registered DCS + duration of DC + severity over time of DC + strata(highest achieved level of education).
Model (2) Age at highest registered GS + duration of GS + severity over time of GS + strata(highest achieved level of education).
All analyses were conducted using the packages prodlim [19], survival [20], timereg [21] and mets [22] in R v.4.3.1 [23] with Rstudio (IDE version 2023.09.1 + 494). Reporting followed STROBE guidelines.
Results
Study size and participants
After removing individuals with missing information on educational level (n = 7,273), a total of 627,758 individuals were included in the study population (Fig. 1). The subset study population used for secondary analyses of the effect of oral health at different ages and disease severity over time, included 64,660 individuals with registrations within each age-range, 0–5, 6–11 and 12–18 years.
In total, 12,252 (3.8%) males and 7,990 (2.6%) females were diagnosed with incident T2D (Table 1). Most individuals regardless of sex and T2D diagnosis had a moderate DCS and GS as their highest registered level of disease (all groups: 46.1–67.7%). The percentage of individuals with a high to severe level of DCS was greater amongst individuals diagnosed with T2D (approximately 23% in those with T2D to 18.5% in the study population) but did not differ greatly between sexes. More females with T2D had a high GS (4.9%) compared to those with no T2D (2.5%). In males 7.2% of those with T2D had a high GS compared to 4.7% in the study population (for visualizations see supplementary Figs. 2–3). Amongst those diagnosed with T2D 36.8% had a low level of education, in the study population this was 21.2%.
Table 1.
Descriptive characteristics of the study population stratified by sex (see next page)
| Males | Females | |||
|---|---|---|---|---|
| Study population | T2D | Study population | T2D | |
| Number of individuals (%) | 321,933 (100) | 12,252 (3.8) | 305,825 (100) | 7,990 (2.6) |
| Follow-up age, mean (SD) years | 49.4 (4.4) | 43.8 (5.6) | 49.7 (4.1) | 41.7 (6.6) |
| Follow-up time, mean (SD) years | 22.5 (3.8) | 16.4 (5.2) | 22.8 (3.3) | 14.4 (6.1) |
| Education, N (%) | ||||
| Low | 72,255 (22.4) | 4,440 (36.2) | 63,972 (20.9) | 3,015 (37.5) |
| Medium | 166,194 (51.6) | 6,314 (51.5) | 141,623 (46.3) | 3,704(46.1) |
| High | 83,484 (25.9) | 1,498 (12.2) | 100,230(32.8) | 1,312 (16.3), |
| Any DC, N (%) | 314,437 (97.7) | 12,075 (98.4) | 299,984 (98.1) | 7,890 (98.8) |
| Highest DC, mean (SD) | 9.13 (4.2) | 9.72 (4.2) | 9.27 (4.1) | 9.95 (4.1) |
| DCS, N (%) | ||||
| Low | 48,442 (15.1) | 1,382 (11.3) | 41,552 (13.6) | 821 (10.3) |
| Moderate | 214,039 (66.5) | 8,128 (66.3) | 207,179 (67.7) | 5,306 (66.4) |
| High | 48,993 (15.2) | 2,182 (17.8) | 46,677 (15.3) | 1,449 (18.1) |
| Severe | 10,459 (3.3) | 560 (4.6) | 10,417 (3.4) | 414 (5.2) |
| Any gingivitis, N (%) | 312,276 (97.0) | 11,937 (97.4) | 293,199 (98.9) | 7,696 (96.3) |
| Highest gingivitis, mean (SD) | 5.75 (2.3) | 6.17 (2.3) | 5.13 (2.2) | 5.62 (2.3) |
| GS, N (%) | ||||
| Low | 103,511 (32.2) | 3,207 (26.2) | 130,786 (42.8) | 2,803 (35.1) |
| Medium | 203,402 (63.2) | 8,165 (66.6) | 167,312 (54.7) | 4,799 (60.1) |
| High | 15,020 (4.7) | 880 (7.2) | 7,727 (2.5) | 388 (4.9) |
| Number of individuals in subset analyses (%) | 32,703 (100) | 867 (2.7) | 31,993 (100) | 661 (2.1) |
| Age at highest recorded DCS, N (%) | ||||
| < 12 years | 27,125 (83.0) | 721 (83.2) | 26,705 (83.5) | 554 (82.3) |
| ≥ 12 years | 5,542 (16.9) | 146 (16.8) | 5,288 (16.5) | 117 (17.7) |
| DC duration, N (%) | ||||
| 0–2 periods | 10,248 (31.4) | 215 (24.8) | 9,733 (30.4) | 182 (27.5) |
| 3 periods | 22,419 (68.6) | 652 (75.2) | 22,260 (69.6) | 479 (72.5) |
| DCS severity over time, N (%) | ||||
| Low | 8,039 (24.6) | 147 (16.9) | 7,463 (23.3) | 142 (21.5) |
| Moderate | 17,374 (53.2) | 483 (55.7) | 17,647 (55.2) | 355 (53.7) |
| High—Severe | 7,254 (22.2) | 237 (27.3) | 6,883 (21.5) | 164 (24.8) |
| Age at highest GS, N (%) | ||||
| < 12 years | 25,087 (76.8) | 630 (72.7) | 26,837 (83.9) | 517 (78.2) |
| ≥ 12 years | 7,580 (23.2) | 237 (27.3) | 5,156 (16.1) | 144 (21.8) |
| Gingivitis duration, N (%) | ||||
| 0–2 periods | 10,079 (30.9) | 213 (24.6) | 11,120 (34.8) | 208 (31.5) |
| 3 periods | 22,588 (69.2) | 654 (75.3) | 20,873 (65.2) | 453 (68.5) |
| GS severity over time, N (%) | ||||
| Low | 9,088 (27.8) | 179 (20.7) | 12,155 (37.9) | 211 (31.9) |
| Moderate | 22,147 (67.8) | 625 (72.1) | 19,050 (59.5) | 425 (64.3) |
| High | 1,432 (4.4) | 63 (7.3) | 788 (2.5) | 25 (3.8) |
DC dental caries, DCS dental caries score, GS gingivitis score, N number, SD standard deviations
In the subset study population 55.7%-75.3% had DC and gingivitis for the duration of their time in SCOR regardless of T2D diagnosis, but with more than 72% having their highest registered DCS and GS before age 12.
The association between T2D, dental caries and gingivitis
Analysis with no age-restriction showed that having severe DCS compared to a low DCS in childhood was associated with the incidence of T2D in adulthood (HR = 1.19, CI: 1.06; 1.35 in females and HR = 1.15, CI:1.04; 1.27 in males) (Table 2, unadjusted estimates in Supplementary Table 1). Likewise, a high GS was associated with a higher incidence of T2D compared to a low GS (HR = 1.87, CI: 1.68; 2.08 in females and HR = 1.59, CI: 1.47; 1.72 in males). Amongst those diagnosed with T2D before age 40, the HR of T2D in both sexes was higher with increasing GS, while the incidence of T2D across levels of DCS were similar. Amongst those diagnosed after age 40 the incidence of T2D was higher in those with severe compared to low DCS (HR = 1.37, CI: 1.17; 1.59 in females and HR = 1.23, CI: 1.09; 1.38 in males), and a high compared low GS (HR = 1.83, CI: 1.59; 2.10 in females and HR = 1.52, CI: 1.39; 1.66 in males). Analyses using non-discretized DC and gingivitis variables overall supported these findings, though DC was not statistically significant in males after adjustment for education (Supplementary Table 2). Only one interaction was significant (Supplementary Table 3); in females diagnosed with T2D before age 40 having moderate DCS with a moderate to high GS is associated with a lower interaction HR for T2D (HR = 0.79, CI: 0.63; 0.98).
Table 2.
Main effects for the association between T2D and the highest registered DCS and GS at any point in time in SCOR
| Males | Females | |||
|---|---|---|---|---|
| HR (CI) | P | HR (CI) | P | |
| No age restriction | ||||
| DCS: Low | 1.00 (ref) | - | 1.00 (ref) | - |
| Moderate | 1.08 (1.02; 1.14) | 0.01 | 1.04 (0.97; 1.12) | 0.31 |
| High | 1.09 (1.02; 1.17) | 0.01 | 1.07 (0.98; 1.17) | 0.11 |
| Severe | 1.15 (1.04; 1.27) | 0.005 | 1.19 (1.06; 1.35) | 0.003 |
| GS: Low | 1.00 (ref) | - | 1.00 (ref) | - |
| Moderate | 1.23 (1.18; 1.28) | < 0.001 | 1.24 (1.18; 1.29) | < 0.001 |
| High | 1.59 (1.47; 1.72) | < 0.001 | 1.87 (1.68; 2.08) | < 0.001 |
| Age < 40 | ||||
| DCS: Low | 1.00 (ref) | - | 1.00 (ref) | - |
| Moderate | 1.02 (0.91; 1.14) | 0.77 | 0.94 (0.84; 1.05) | 0.25 |
| High | 0.94 (0.82; 1.08) | 0.36 | 0.88 (0.77; 1.01) | 0.07 |
| Severe | 0.93 (0.75; 1.15) | 0.48 | 0.98 (0.79; 1.19) | 0.81 |
| GS: Low | - | - | - | - |
| Moderate | 1.28 (1.18; 1.39) | < 0.001 | 1.17 (1.08; 1.27) | < 0.001 |
| High | 1.84 (1.58; 2.15) | < 0.001 | 1.94 (1.63; 2.30) | < 0.001 |
| Age ≥ 40 | ||||
| DCS: Low | 1.00 (ref) | - | 1.00 (ref) | - |
| Moderate | 1.09 (1.03; 1.17) | 0.007 | 1.13 (1.02; 1.25) | 0.01 |
| High | 1.14 (1.06; 1.24) | < 0.001 | 1.23 (1.09; 1.38) | 0.005 |
| Severe | 1.23 (1.09; 1.38) | < 0.001 | 1.37 (1.17; 1.59) | < 0.001 |
| GS: Low | 1.00 (ref) | - | 1.00 (ref) | - |
| Moderate | 1.21 (1.15; 1.27) | < 0.001 | 1.28 (1.21; 1.36) | < 0.001 |
| High | 1.52 (1.39; 1.66) | < 0.001 | 1.83 (1.59; 2.10) | < 0.001 |
Highest achieved level of education (observed between 25–30 years) used as strata
CI confidence interval, DCS dental caries score, GS gingivitis score, HR hazard ratio, P p-value, ref reference group
Testing at a 95% confidence level
The association of T2D with the age at the highest registered levels of oral disease, duration, and severity over time of dental caries and gingivitis
In females, neither age at disease, severity over time, nor duration of DC were related to the incidence of T2D (Table 3, unadjusted estimates in Supplementary Table 4). In males, having a moderate severity DCS for the duration of the time in SCOR was associated with a higher incidence of T2D (HR = 1.32, CI: 1.07; 1.64), while timing and duration of DC were not. Having the highest GS registered at or above age 12 was associated with a higher incidence of T2D (HR = 1.37, CI: 1.13; 1.65) in females, and in males (HR = 1.25, CI: 1.08; 1.45). In males high severity GS throughout the time in SCOR was associated with an HR of 1.61 (CI: 1.19; 2.16) compared to a low GS, and having gingivitis throughout the time in SCOR, e.g. within all three defined age-ranges and regardless of the severity, was associated with an HR of 1.24 (CI: 1.06; 1.45) compared to having gingivitis at two or less periods in time. Similar patterns appeared in the effect of gingivitis on the HRs of T2D for females, though these did not reach statistical significance.
Table 3.
Secondary (subset) analyses of the hazard of T2D by age at highest registered DCS and GS, duration, and severity over time of DC and gingivitis
| Males | Females | |||
|---|---|---|---|---|
| HR (CI) | P | HR (CI) | P | |
| DC | ||||
| Age at highest registered DCS | ||||
| Below age 12 | 1.00 (ref) | - | 1.00 (ref) | - |
| Age 12 or above | 1.01 (0.84; 1.22) | 0.91 | 1.10 (0.89; 1.35) | 0.37 |
| DCS over time | ||||
| Low | 1.00 (ref) | – | 1.00 (ref) | – |
| Moderate | 1.32 (1.07; 1.64) | 0.01 | 0.91 (0.72; 1.14) | 0.39 |
| High—Severe | 1.28 (0.95; 1.71) | 0.10 | 0.89 (0.64; 1.24) | 0.49 |
| DC duration | ||||
| Up to 2 periods in time | 1.00 (ref) | – | 1.00 (ref) | – |
| 3 periods in time | 1.07 (0.89; 1.29) | 0.47 | 1.09 (0.89; 1.35) | 0.40 |
| Gingivitis | ||||
| Age at highest registered GS | ||||
| Below age 12 | 1.00 (ref) | – | 1.00 (ref) | – |
| Age 12 or above | 1.25 (1.08; 1.45) | 0.003 | 1.37 (1.13; 1.65) | < 0.001 |
| GS over time | ||||
| Low | 1.00 (ref) | – | 1.00 (ref) | – |
| Moderate | 1.19 (1.004; 1.42) | 0.04 | 1.11 (0.94; 1.32) | 0.23 |
| High | 1.61 (1.19; 2.16) | 0.002 | 1.41 (0.93; 2.15) | 0.10 |
| Gingivitis duration | ||||
| Up to 2 periods in time | 1.00 (ref) | – | 1.00 (ref) | – |
| 3 periods in time | 1.24 (1.06; 1.45) | 0.008 | 1.06 (0.89; 1.26) | 0.47 |
Highest achieved level of education (observed between 25 and 30 years) used as strata. Testing at a 95% confidence level
CI confidence interval, DCS dental caries score, GS gingivitis score, HR hazard ratio, P p-value
Discussion
In this registry-based study we found a higher incidence of T2D among adult Danes that had DC and gingivitis in childhood, thereby confirming our main hypothesis for this study. In both sexes, having one’s highest registered level of gingivitis after the age of 12 was associated with a higher incidence of T2D in adulthood. The results should be interpreted with caution due to residual confounding.
The highest HRs of T2D were found with high levels of GS. This could be explained by the low-grade inflammation inferred by gingivitis, as it may cause a long-lasting reprogramming of granulopoiesis resulting in a persisting neutrophil hyper-responsiveness, in a similar way as has been suggested for periodontitis [24]. In adipose tissues, neutrophils are then activated and release a range of inflammatory factors, contributing to the development of insulin resistance. In line with this theory of inflammation, one may speculate if the greater effect of gingivitis in teenagers compared to children in this study, is due to a stronger immune response to low-grade inflammation of the more mature teenage immune-system [25].
While research on the genetics of gingivitis and T2D is lacking, there is evidence that periodontitis and T2D are genetically linked [26, 27]. Given the similarities between the two oral diseases [5], these genes could also play a role in the association to T2D observed in this study. Evidence suggests that DC has a high heritability and that DC correlates with several traits of metabolic disease on a genetic level [28, 29]. Thus, the associations of DC and T2D in our study may have a considerable genetic component.
In Denmark, there has historically been a great focus on reducing the DC incidence and prevalence in children and adolescents. Meanwhile, gingivitis has received little attention. Today registration of gingivitis in SCOR has long ceased to be a mandatory [17]. Our study suggests that the impact of gingivitis may be as relevant as the impact of DC for the subsequent incidence of T2D. One interpretation of this could be that efficient prevention of gingivitis in childhood, including targeted education and oral hygiene instructions in high-risk individuals, may have a long-term protective effect on T2D development later in life.
A significant limitation of our study is that the registry data lack information about potential confounders. We do not have direct measurements of lifestyle factors e.g. smoking and eating habits. Infrequent tooth-brushing in the teenage years has been shown to correlate with higher rates of smoking, an obesogenic lifestyle, and a lower level of education later in life [30]. Accordingly, using the highest achieved level of education as strata considerably lowered the HRs in this study. While hopefully having captured some of the confounding effects of lifestyle-factors, our estimates are bound to be affected by residual confounding. Thus, we cannot ultimately conclude that DCS and GS are causes of T2D. Studies with individual level information on lifestyle factors are warranted.
We were not able to account for individuals travelling in and out of Denmark. We may assume that individuals who leave the country, develop diabetes, and then return, will still be registered as having T2D, as anyone prescribed medication via the Danish healthcare system will be registered in RUKS. The prevalence of T2D in our study population corresponds well with the nationwide disease prevalence in the period [31]. Thus, most T2D-cases have likely been registered, though there may be a delay in the date of diagnosis for some individuals, somewhat artificially increasing survival times.
The subset of the study population used for the analyses presented in Table 3 excludes the oldest individuals in our dataset. Due to being born between 1963 and 1966 they do not have registrations in SCOR between ages 0–5, as SCOR didn’t exist at the time. As an effect of their current age, the excluded individuals are the ones at the highest risk of T2D in the available time-at-risk. This may increase the probability of finding no associations between oral health and T2D in the subset analyses and produce more uncertain effect estimates.
Strengths of our study include the large sample size, with nationwide registrations of both exposures and outcome, the length of follow-up, and the detailed registrations of oral health.
In conclusion, we report evidence for the association between childhood experience of DC and gingivitis with receiving a T2D diagnosis in adulthood. Our findings suggest the need for a greater awareness of gingivitis in childhood dental care. However, studies adjusting for confounding variables to a greater extent than possible here are needed.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We wish to extend our most sincere thanks to Professor Claus Henrik Nielsen for sharing with us his knowledge on the immune system and his thoughts on how it may have impacted our findings.
Author contributions
Nikoline Nygaard: Contributed to conception, design, data acquisition, analysis and interpretation, drafted and critically revised the manuscript and gave final approval. Merete Markvart: Contributed to conception and design, data acquisition and interpretation, critically revised the manuscript and gave final approval. Daniel Belstrøm: Contributed to conception and design, data acquisition and interpretation, critically revised the manuscript and gave final approval. Torben Hansen: Contributed to conception and design, data acquisition and interpretation, critically revised the manuscript and gave final approval. Evelina Stankevic: Contributed to conception and design, data acquisition, analysis and interpretation, critically revised the manuscript and gave final approval. Lars Ängquist: Contributed to conception and design, data acquisition, analysis and interpretation, critically revised the manuscript and gave final approval. Anja Olsen: Contributed to conception and design, data acquisition and interpretation, critically revised the manuscript and gave final approval. Anne Kirstine Eriksen: Contributed to conception and design, data acquisition, analysis and interpretation, critically revised the manuscript and gave final approval. All authors gave their final approval and agree to be accountable for all aspects of the work.
Funding
Open access funding provided by Copenhagen University
The Danish Dental Association (grant number 100) and the Novo Nordisk Foundation (grant number NNF18CC0034900 and NNF23SA0084103) provided funding for the costs related to obtaining data for this study.
Availability of data and materials
The full data from SCOR underlying the present article are available from The Danish Health Data Authority [32], but restrictions apply to the availability of these data. The R-script created for cleaning data and generating the analyses presented in the present paper can be found on GitHub [33].
Declarations
Conflict of interest
The authors have no conflicts of interest to declare.
Ethical approval
Research projects using only registry data do not need informed consent from individuals in the study cohort, nor ethical approval as per the Danish National Committee on Health Research Ethics. The use of registry data in this study was approved by the Danish Data Protection Agency via the Office of Research and Innovation, University of Copenhagen (case number 514–0496/20–300). Data for the present study was used under license from the Danish Health Data Authority. All data was anonymized prior to our receiving it from the Danish Health Data Authority. All methods were carried out in accordance with relevant guidelines and regulations.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.World Health Organization, “Oral Health,” World Health Organisation. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/oral-health. Accessed 27 Nov 2024
- 2.Kumar S, Tadakamadla J, Johnson NW (2016) Effect of toothbrushing frequency on incidence and increment of dental caries: a systematic review and meta-analysis. J Dent Res 95(11):1230–1236. 10.1177/0022034516655315 [DOI] [PubMed] [Google Scholar]
- 3.Slots J (2020) Primer on etiology and treatment of progressive/severe periodontitis: a systemic health perspective. Periodontol 2000 83(1):272–276. 10.1111/prd.12325 [DOI] [PubMed] [Google Scholar]
- 4.Löe H (1967) The gingival index, the plaque index and the retention index systems. J Periodontol 38(6):610–616. 10.1902/jop.1967.38.6.610 [DOI] [PubMed] [Google Scholar]
- 5.Van Dyke TE, Bartold PM, Reynolds EC (2020) The nexus between periodontal inflammation and Dysbiosis. Front Immunol 11:511. 10.3389/fimmu.2020.00511 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lundtorp-Olsen C et al (2024) Supragingival microbiota, cytokines, and proteins in individuals with different trajectories in experimental gingivitis. J Oral Microbiol 16(1):2372861. 10.1080/20002297.2024.2372861 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Simpson TC et al (2022) Treatment of periodontitis for glycaemic control in people with diabetes mellitus. Cochrane Database Syst Rev. 10.1002/14651858.CD004714.pub4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wu C et al (2020) Epidemiologic relationship between periodontitis and type 2 diabetes mellitus. BMC Oral Health 20(1):204. 10.1186/s12903-020-01180-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang Q, Li ZX, Yang Y, Wang CX, Wang LM, Wang LH (2017) Correlation between self-reported gingival bleeding and type 2 diabetes mellitus in aged ≥18 years adults in China. Zhonghua Liu Xing Bing Xue Za Zhi Zhonghua Liuxingbingxue Zazhi 38(3):346–349. 10.3760/cma.j.issn.0254-6450.2017.03.014 [DOI] [PubMed] [Google Scholar]
- 10.Mattout C, Bourgeois D, Bouchard P (2006) Type 2 diabetes and periodontal indicators: epidemiology in France 2002–2003. J Periodontal Res 41(4):253–258. 10.1111/j.1600-0765.2006.00862.x [DOI] [PubMed] [Google Scholar]
- 11.Holmstrup P et al (2017) Comorbidity of periodontal disease: two sides of the same coin? An introduction for the clinician. J Oral Microbiol 9(1):1332710. 10.1080/20002297.2017.1332710 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.De Lima AKA, Amorim Dos Santos J, Stefani CM, Almeida De Lima AD, Damé-Teixeira N (2020) Diabetes mellitus and poor glycemic control increase the occurrence of coronal and root caries: a systematic review and meta-analysis. Clin Oral Investig 24(11):3801–3812. 10.1007/s00784-020-03531-x [DOI] [PubMed] [Google Scholar]
- 13.Albandar JM, Tinoco EMB (2002) Global epidemiology of periodontal diseases in children and young persons. Periodontol 2000 29(1):153–176. 10.1034/j.1600-0757.2002.290108.x [DOI] [PubMed] [Google Scholar]
- 14.Pussinen PJ et al (2019) Association of childhood oral infections with cardiovascular risk factors and subclinical atherosclerosis in adulthood. JAMA Netw Open 2(4):e192523. 10.1001/jamanetworkopen.2019.2523 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Pussinen PJ et al (2020) Childhood oral infections associate with adulthood metabolic syndrome: a longitudinal cohort study. J Dent Res 99(10):1165–1173. 10.1177/0022034520929271 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.The Danish Health Data Authority Algorithms for Selected Chronic Diseases and Severe Mental Disorders. The Danish Health Data Authority, Nov. 2022. [Online]. Available: https://www.esundhed.dk/Emner/Operationer-og-diagnoser/Udvalgte-kroniske-sygdomme-og-svaere-psykiske-lidelser#tabpanel988928C7E4B84712940C30D63EC7544C. Accessed 20 Apr 2023
- 17.Nygaard N et al (2023) The national child odontology registry (SCOR): a valuable resource for odontological and public health research. BMC Oral Health 23(1):608. 10.1186/s12903-023-03199-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ruiz-Romero GA, Álvarez-Delgado C (2024) Effects of estrogens in mitochondria: an approach to type 2 diabetes. AIMS Mol Sci 11(1):72–98. 10.3934/molsci.2024006 [Google Scholar]
- 19.Gerds TA Product-Limit Estimation for Censored Event History Analysis. (2023). [Online]. Accessed 12 Nov 2023. Available: https://cran.r-project.org/web/packages/prodlim/prodlim.pdf
- 20.Thernau TM A Package for Survival Analysis in R. (2023). [Online]. Accessed 12 Nov 2023. Available: https://cran.r-project.org/web/packages/survival/index.html
- 21.Scheike TH, Zhang M-J (2011) Analyzing competing risk data using the R timereg package. J Stat Softw. 10.18637/jss.v038.i02 [PMC free article] [PubMed] [Google Scholar]
- 22.Holst KK, Scheike TH, Hjelmborg JB (2016) The liability threshold model for censored twin data. Comput Stat Data Anal 93:324–335. 10.1016/j.csda.2015.01.014 [Google Scholar]
- 23.R Core Team, R: A language and environment for statistical computing. (2021). R Foundation for Statistical Computing, Vienna, Austria. [Online]. Accessed 12 Nov 2023. Available: https://ropensci.org/blog/2021/11/16/how-to-cite-r-and-r-packages/
- 24.Irwandi RA, Chiesa ST, Hajishengallis G, Papayannopoulos V, Deanfield JE, D’Aiuto F (2022) The roles of neutrophils linking periodontitis and atherosclerotic cardiovascular diseases. Front Immunol 13:915081. 10.3389/fimmu.2022.915081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Simon AK, Hollander GA, McMichael A (2015) Evolution of the immune system in humans from infancy to old age. Proc R Soc B Biol Sci 282(1821):20143085. 10.1098/rspb.2014.3085 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wang Y et al (2022) Mendelian randomization supports the causal role of fasting glucose on periodontitis. Front Endocrinol 13:860274. 10.3389/fendo.2022.860274 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yu Y-H, Steffensen B, Ridker PM, Buring JE, Chasman DI (2023) Candidate loci shared among periodontal disease, diabetes and bone density. Front Endocrinol 13:1016373. 10.3389/fendo.2022.1016373 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Haworth S et al (2020) Heritability of caries scores, trajectories, and disease subtypes. J Dent Res 99(3):264–270. 10.1177/0022034519897910 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Shungin D et al (2019) Genome-wide analysis of dental caries and periodontitis combining clinical and self-reported data. Nat Commun 10(1):2773. 10.1038/s41467-019-10630-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Koivusilta L, Honkala S, Honkala E, Rimpelä A (2003) Toothbrushing as part of the adolescent lifestyle predicts education level. J Dent Res 82(5):361–366. 10.1177/154405910308200507 [DOI] [PubMed] [Google Scholar]
- 31.Carstensen B, Rønn PF, Jørgensen ME (2020) Prevalence, incidence and mortality of type 1 and type 2 diabetes in Denmark 1996–2016. BMJ Open Diabetes Res Care 8(1):e001071. 10.1136/bmjdrc-2019-001071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.The Danish Health Data Authority, “Apply for SCOR data at hte Danish Health Data Authority’s Research Services,” The Danish Health Data Authority. [Online]. Accessed 20 Feb 2024. Available: https://sundhedsdatastyrelsen.dk/da/forskerservice/om-forskerservice/nyheder_forskerservice/ansoeg-om-scor_data111219
- 33.Nygaard N Niknyg-Overview. GitHub. [Online]. Available: https://github.com/Niknyg. Accessed 21 Mar 2024
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The full data from SCOR underlying the present article are available from The Danish Health Data Authority [32], but restrictions apply to the availability of these data. The R-script created for cleaning data and generating the analyses presented in the present paper can be found on GitHub [33].