Association between attention-deficit/hyperactivity disorder (ADHD) and dental caries risk: a retrospective cross-sectional study in children and adolescents

1. Introduction

Attention-deficit/hyperactivity disorder (ADHD) is a commonly diagnosed neurodevelopmental disorder in childhood, characterized by a persistent pattern of inattention, hyperactivity, and impulsivity that interferes with functioning across settings [1]. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) recognizes three presentations: predominantly inattentive, predominantly hyperactive-impulsive, and combined [1]. In the United States, the prevalence of ADHD among children and adolescents has increased steadily, with recent estimates ranging from 9 to 11 percent [2, 3, 4].

Children with ADHD often face challenges in maintaining regular health habits, and less frequent toothbrushing, poorer oral hygiene, and more frequent daily eating and drinking have been reported in this population [5, 6, 7]. Additionally, reduced salivary flow has been reported as a side effect of the stimulant medications commonly prescribed for ADHD, although a review found that this reduction did not appear to explain the higher caries prevalence in children with ADHD [8].

Prior studies examining the relationship between ADHD and dental caries have yielded mixed findings. Some reported significantly higher caries rates in children with ADHD [9, 10], and in a national survey of children with ADHD symptoms the association with caries persisted after adjustment for socioeconomic level, sugar intake, toothbrushing, and dental visits [11], whereas others found no significant difference [6, 12] or only slightly higher caries prevalence [13]. Many of those studies were small, with sample sizes ranging from 34 to 851, and carried a high risk of bias from unaddressed confounding, limiting the generalizability of their conclusions [7, 14]. Behavioral traits inherent to ADHD, such as impulsivity, difficulty with routine, poor compliance, and increased snacking, are potential mediators in caries etiology. Several studies reported that children with ADHD demonstrated less consistent oral hygiene routines and more frequent daily eating and drinking compared to neurotypical peers [6, 12]. Moreover, reduced supervision and executive dysfunction may impair follow-through on recommended preventive dental behaviors.

To address this gap, the present study used a large hospital-based dataset to investigate whether pediatric patients with ADHD were more likely to carry a diagnosis of dental caries than their peers without ADHD. Using retrospective diagnostic data from over 258,000 pediatric records, we aimed to quantify this association and assess its significance in a real-world clinical population. We hypothesized that children diagnosed with ADHD would show a significantly higher prevalence of dental caries than the general pediatric hospital population.

2. Materials and methods

This retrospective cross-sectional study was conducted using the i2b2 clinical data platform to examine anonymized electronic health records from pediatric encounters at the University of Florida Health Shands Hospitals. The study window extended from June 2011 through April 2025. All data were obtained through the i2b2 search engine, which provides a totally de-identified data set. The University of Florida Health Center Institutional Review Board has determined that research conducted solely using this IRB-approved de-identified data set does not meet the definition of human research and therefore does not require IRB approval (determination letter dated 06 April 2021; R. Peter Iafrate, Pharm.D., Chairman, Health Center IRB).

Throughout this manuscript, the term ADHD is used in accordance with current DSM-5 (2013) and ICD-11 nomenclature [1]. Although the historical term attention-deficit disorder (ADD) appeared in earlier diagnostic systems, the ICD-10 F90.x codes used for data extraction in this study encompass all recognized presentations of the disorder, including the predominantly inattentive type (F90.0), the predominantly hyperactive type (F90.1), and the combined type (F90.2). These subtypes were defined for data extraction purposes but were not analyzed separately.

Diagnoses were defined according to standardized clinical coding systems used in administrative health data. ADHD was identified using ICD-10 code F90, which represents diagnostic categories for attention-deficit hyperactivity disorders (World Health Organization, International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, ICD-10). Dental caries was identified using ICD-10 code K02.9, which denotes unspecified dental caries, encompassing decay in enamel, dentin, or cementum without further specification. ICD-10 is not a standardized clinical tool for measuring dental caries severity; its use here reflects administrative coding practices rather than clinical caries assessment systems such as decayed, missing, and filled teeth (DMFT) or International Caries Detection and Assessment System (ICDAS). All diagnostic codes reflected formal definitions recognized by the World Health Organization and the Centers for Medicare & Medicaid Services.

These diagnoses were charted into the electronic health records by qualified healthcare providers and were approved by attending physicians, dentists, and surgeons. For dental caries, this could involve direct oral examinations by dental professionals, referrals from primary care, or notations from multidisciplinary team assessments conducted by qualified dental and medical providers; however, as this is retrospective administrative data, specific diagnostic criteria (e.g., clinical visual inspection, radiographic confirmation, or parent-reported history) were not standardized across records, and no inter-rater calibration was applied.

All patients aged 0 to 17 years with at least one recorded hospital encounter during the study period were eligible for inclusion. Patients with a diagnostic code for ADHD (F90.x) were assigned to the exposed group. All others comprised the comparison group (i.e., patients without any recorded F90 code). No exclusions were applied based on comorbid conditions, number of visits, or visit type to preserve the representativeness of the hospital population.

Demographic variables included age, sex, and race when available. Race was recorded at intake and was categorized as White, Black, or Other, based on standard reporting practices in Florida hospital systems. Because the i2b2 platform stratifies patients by decade of life, the full 0-to-17-year age range was retained for analysis without further subdivision.

All data queries were executed within the secure i2b2 interface. Individual-level charts were not accessed, and no free-text records, laboratory values, or radiographic data were reviewed. The study relied entirely on structured, coded data. Because no clinical abstraction or interpretation was performed, calibration or adjudication procedures were not required.

Descriptive statistics were calculated to summarize the demographic composition of the population and the overall prevalence of each diagnosis. Condition frequencies were also examined by racial subgroup. The association between ADHD and dental caries was assessed using odds ratios with 95 percent confidence intervals. Statistical analysis was performed using the MedCalc Odds Ratio Calculator (MedCalc Software Ltd., Ostend, Belgium), which calculates the odds ratio and its 95% confidence interval as described by Altman [15] and the p-value as described by Sheskin [16]. Statistical significance was determined using a two-tailed test with a significance level of 0.05. Multivariable logistic regression was not feasible due to constraints in the i2b2 platform, which limits access to individual-level data and does not support adjustments for confounders such as age, sex, socioeconomic status, or detailed demographics.

3. Results

A total of 258,914 pediatric patients aged 0 to 17 years were identified in the hospital dataset during the study period. Among this cohort, 8125 patients (3.14%) carried a diagnosis of dental caries, and 18,444 patients (7.12%) were diagnosed with ADHD (ICD-10 F90) (Table 1).

Of the 18,444 patients with ADHD, 1662 also carried a diagnosis of dental caries. This corresponds to a caries prevalence of 9.01% within the ADHD group, compared to a prevalence of 2.69% among patients without ADHD. Patients with ADHD were therefore over three times more likely to carry a caries diagnosis (Table 2).

The calculated odds ratio for the association between ADHD and dental caries was 3.59 (95% CI: 3.39 to 3.79; p < 0.001) (Table 2).

Race-stratified analysis showed variation in diagnostic prevalence across demographic groups (Table 1). Among children with a diagnosis of dental caries, 43.5% were White, 34.5% were Black, and 22.0% were classified as Other Race. In contrast, among patients with ADHD, 63.1% were White, 20.7% were Black, and 16.2% were classified as Other Race. While White children composed approximately 55% of the total hospital population, they accounted for a larger share of ADHD diagnoses. Black children were proportionally more likely to carry a caries diagnosis, while children categorized as Other Race showed lower prevalence of both conditions. These race-based comparisons are exploratory and were not hypothesis-driven.

All findings were derived from structured diagnostic coding within the i2b2 interface. Individual charts, radiographs, and free-text notes were not reviewed.

4. Discussion

This study examined whether children and adolescents with a diagnosis of ADHD carry a higher burden of dental caries than the general pediatric population. Using a large hospital-based dataset, we found a notable difference in caries prevalence among patients with ADHD compared to their peers. While 3.14 percent of the overall population carried a caries diagnosis, the rate was 9.01 percent among patients with ADHD, nearly a threefold increase. The association was statistically significant, with an odds ratio of 3.59 (95% CI: 3.39 to 3.79; p < 0.001). These findings suggest an association between ADHD and an increased prevalence of dental caries diagnoses, though further research is needed to clarify the mechanisms underlying this relationship.

Our results align with some earlier investigations that reported similar trends, though findings in the literature are mixed. Blomqvist et al. [6] found that 13-year-olds with ADHD were far less likely than controls to brush their teeth every evening (48 percent versus 82 percent), although differences in caries indices at that age did not reach statistical significance. Grooms et al. [10] reported higher caries rates in children with ADHD in both primary and permanent dentitions, and Melwani-Sadhwani et al. [9] similarly reported significantly more caries and restorative treatment in children with ADHD. A recent systematic review reported significantly higher caries in four of fourteen studies and less frequent daily brushing among children with ADHD [7], and Chau et al. [12] found no difference in caries prevalence but significantly poorer oral hygiene and oral health behaviors in adolescents with ADHD. These results have often been attributed to behavioral features characteristic of the condition. Children with ADHD may struggle with routine hygiene, exhibit impulsivity around snacking, or face challenges adhering to dietary advice [5, 6, 12]. A recent meta-analysis found that children with ADHD had significantly greater caries experience than their peers, with a pooled odds ratio of 3.31 [14]. Other studies raised questions about the strength of this association. Mielnik-Błaszczak et al. [13], for instance, found that caries prevalence and decayed, missing, and filled (DMF) scores were higher in children with ADHD but that the differences did not reach statistical significance. Drumond et al. [14] similarly noted that the included studies rarely addressed confounding, and social determinants such as household income and parental education are known to shape caries outcomes [17, 18]. Our findings, while unadjusted for these variables, offer insight into real-world diagnostic prevalence patterns in a large and diverse clinical population.

In addition to behavioral contributors, pharmacologic treatment for ADHD may present further risk. Stimulant medications such as methylphenidate and amphetamine-based formulations are widely prescribed in this population, and their effect on saliva is not uniform across studies: one case-control study found dry mouth commonly reported but no significant reduction in stimulated salivary flow [19], two studies documented reduced salivary flow or buffering capacity during methylphenidate therapy [20, 21], and a review concluded that medication-related salivary reduction did not appear to explain the higher caries prevalence observed in children with ADHD [8]. Where it occurs, medication-induced xerostomia may impair saliva’s protective functions, including acid neutralization, mechanical cleansing, and remineralization of enamel surfaces [22, 23]. Salivary hypofunction is typically recognized only after damage to the oral tissues has occurred, leaving clinicians to treat new and recurrent carious lesions [24]. Additionally, many pediatric medications, including some used for ADHD management, contain sugar to improve palatability, which may independently contribute to caries risk.

Despite its strengths, this study has several limitations. A key limitation is that the data were derived from a hospital-based population rather than from community or population-based samples. Children who present to a hospital setting may differ systematically from the general pediatric population, for example in terms of underlying medical complexity, socioeconomic background, or health-seeking behavior. In addition, dental caries in hospital records may be underdiagnosed or inconsistently recorded, as oral health is not always the primary focus of care. These factors may limit the external validity and generalizability of our findings to broader community settings.

No multivariable adjustment was performed for key confounders including age, sex, race, or socioeconomic status. Due to constraints of the i2b2 platform, which limits access to individual-level data, multivariable logistic regression could not be conducted. The reported odds ratios are therefore unadjusted and may be subject to residual confounding. This limitation restricts the ability to isolate the independent contribution of ADHD to caries risk and represents an important direction for future work employing multivariable regression models.

The comparison group consisted of the total hospital population, which may include patients with undiagnosed or subclinical ADHD who did not receive an F90 code during the study period. The inclusion of such patients in the non-ADHD group could attenuate the observed association.

Although ADHD subtype codes (F90.0, F90.1, F90.2) were defined in the methods for data extraction purposes, subtypes were not analyzed separately. It is possible that different ADHD presentations carry different levels of caries risk, and this represents a limitation of the current analysis.

The inclusion of children aged 0 to 4 years, where ADHD diagnosis is uncommon [25], may bias the comparison group. Children in this age group who receive a caries diagnosis are unlikely to also carry an ADHD diagnosis at that stage. As a result, the inclusion of younger children with caries but without ADHD may lower the observed odds ratio, since these children increase the caries count in the non-ADHD group but are unlikely to carry an F90 diagnosis. Ideally, analysis would be restricted to ages 5 through 17 or stratified by age groups (e.g., 0 to 9 years and 10 to 17 years), but due to the i2b2 platform’s decade-level age stratification and lack of support for finer age breakdowns without individual chart access, such refinement was not possible. This constraint may contribute to a modest underestimation of the true association and highlights a key limitation of using aggregated administrative data platforms for detailed subgroup analyses.

Comorbid conditions such as autism spectrum disorder, which may independently increase caries risk through shared behavioral, sensory, and oral hygiene challenges, were not excluded from the analysis. While preserving the representativeness of the hospital population, the inclusion of patients with such comorbidities may limit the specificity of the observed association between ADHD and caries.

Diagnostic codes in administrative data do not account for clinical severity, medication adherence, or timing of diagnosis. Caries diagnoses may be under-reported in hospital-based data, as routine dental visits often occur outside the primary health system. Likewise, the i2b2 platform does not capture socioeconomic status, household environment, or oral health behaviors, each of which may mediate or confound the relationship between ADHD and caries. There may also be a selection bias, as children with ADHD could be more likely to receive dental care for caries in a hospital-based setting rather than in private community practices. Furthermore, the prevalence of dental caries in this study (9.01 percent in the ADHD group and 2.69 percent in the non-ADHD group) is lower than that reported in other ADHD-caries studies, which often use full clinical dental examinations rather than hospital-based ICD codes. This difference may result from ICD codes capturing only more severe or symptomatic cases, potentially underestimating the true prevalence.

Nonetheless, the present findings offer population-level evidence of an association between ADHD and dental caries in children. The large sample size, standardized coding, and consistent effect size across racial strata suggest that the relationship warrants further investigation. Future studies would benefit from multivariable adjustment for age, sex, socioeconomic status, and comorbid conditions, as well as the addition of pharmacy data, dental chart abstraction, age-stratified analyses, and behavioral assessments to explore this association in greater depth.

5. Conclusions

This retrospective analysis found a statistically significant association between ADHD and increased prevalence of dental caries in a large pediatric hospital population. Children with ADHD were more than three times as likely to carry a diagnosis of caries compared to those without ADHD. While administrative coding data do not capture important contextual factors such as dietary patterns, socioeconomic status, or treatment adherence, the strength of the observed association supports the need for heightened preventive dental care in children with ADHD. Future research integrating multivariable adjustment, caries severity indices, and behavioral assessments will be important to further clarify this relationship.

Availability of data and materials

The data supporting the findings of this study were obtained from the i2b2 clinical data platform at the University of Florida, which provides a totally de-identified data set. Due to institutional data use agreements and the platform’s restrictions on data export, the underlying datasets are not publicly available. Aggregate results are reported in full within the manuscript. Inquiries regarding data access may be directed to the corresponding author.

Author contributions

DKL—conceptualized the study, performed data collection and analysis, and wrote the initial draft. MG—contributed to study design, methodology review, and manuscript revision. JK—provided expertise in oral diagnostics, data aggregation, supervised the project, and contributed to manuscript revision. All authors read and approved the final manuscript.

Ethics approval and consent to participate

This study utilized a totally de-identified data set provided by the i2b2 clinical data platform at the University of Florida. The University of Florida Health Center Institutional Review Board has determined that research conducted solely using this IRB-approved de-identified data set does not meet the definition of human research and therefore does not require IRB approval (determination letter dated 06 April 2021; R. Peter Iafrate, Pharm.D., Chairman, Health Center IRB). Consent to participate was not applicable.

Acknowledgment

The authors would like to thank the University of Florida Health Shands Hospital system for their support. An LLM was used exclusively for grammar, punctuation, and syntax refinement. No content generation, data analysis, or interpretation of findings was performed using an LLM.

Funding

This research was conducted without external funding. No grants, financial support, or sponsorships were received for the completion of this study.

Conflict of interest

The authors declare no conflict of interest. Marcio Guelmann is serving as one of the Editorial Board members of this journal. We declare that Marcio Guelmann had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to AS.