Abstract
Background: The long-term impacts of traumatic brain injury (TBI) in older adults are not well known. Our objective was to describe the association between late-life TBI, incident dementia, and health care needs.
Methods: We conducted a retrospective cohort study using linked health administrative data in Ontario, Canada, and included community-dwelling individuals older than 65 years with a new TBI between Apr. 1, 2004, and Mar. 1, 2020, and up to 17 years of follow-up. People with and without TBI were 1:1 matched on age, sex, and propensity score. We compared rates of incident dementia (≤ 5 yr and > 5 yr), use of publicly funded home care, and admission to a long-term care home, using cause-specific hazard models. We used conditional inference trees to identify subgroups at high risk for outcomes based on associations with individual and intersecting social determinants of health.
Results: We included 132 113 matched pairs. Late-life TBI was associated with an increased rate of incident dementia (≤ 5 yr: hazard ratio [HR] 1.69, 95% confidence interval [CI] 1.66 to 1.72; > 5 yr: HR 1.56, 95% CI 1.53 to 1.59), use of publicly funded home care (HR 1.30, 95% CI 1.29 to 1.31), and admission to a long-term care home (HR 1.45, 95% CI 1.42 to 1.47). Dementia was more common among older females from low-income neighbourhoods than among male peers (29.0% v. 24.7%). Residents of smaller communities received less home care than those in larger communities (60.1% v. 64.6%) and had a higher probability of long-term care home admission (26.3% v. 21.7%).
Interpretation: Late-life TBI in community-dwelling older adults is associated with higher rates of incident dementia, use of publicly funded home care, and admission to a long-term care home, with older females and residents of smaller communities experiencing poorer outcomes. Clinical awareness and policy planning that reflects these needs and population heterogeneity are important.
Protocol registration: Open Science Framework identifier 10.17605/OSF.IO/KZT3F
More than 1 million people in Canada aged 65 years or older will sustain a traumatic brain injury (TBI) in their lifetime.1,2 Fall-related TBIs have increased over time and account for 50% of all TBIs across age groups.3,4 Traumatic brain injury is clinically diagnosed when an individual experiences a direct impact on the head or a biomechanically plausible indirect force (e.g., whiplash injury), along with 1 of the following: loss of consciousness, posttraumatic amnesia, disorientation, or neurologic signs such as slurred speech, muscle weakness, or vision changes.5 The clinical severity of TBI is commonly classified according to the total Glasgow Coma Scale score as mild (13 to 15), moderate (9 to 12), or severe (3 to 8).6 Traumatic brain injury is associated with substantial direct and indirect health care costs, from readmission to hospital, home care use, and caregiving, with higher costs in older than younger adults.7,8
Older adults may face unique challenges such as cognitive and functional decline following a TBI,9,10 leading to increased care needs.11 Traumatic brain injury is recognized as a midlife risk factor for dementia.12,13 However, population-level estimates of dementia incidence following TBI in late life (i.e., > 65 yr), including the middle- and oldest-old, are unknown.14 Further, older adults with TBI are a heterogeneous group; therefore, their health outcomes should be studied through an equity lens.15,16 Factors such as education, socioeconomic status, or identifying as racialized minorities may confer a higher risk of dementia and lower access to health care services, but there remains an important research gap in how individual and intersecting social determinants of health affect older adults with TBI.17–19
To help clinicians advise older adults and their families on prognosis and strategies to mitigate cognitive and functional decline while also informing equitable population-based care solutions,20 we had 2 study aims. We sought first to describe the magnitude of the association between late-life TBI and rate of incident dementia and care needs, including use and intensity of of publicly funded home care and admission to a long-term care facility; and second to identify social determinants of health associated with a greater risk of dementia and increased care needs in older adults following TBI.
Methods
We followed the Sex and Gender Equity in Research Guidelines and the Strengthening the Integration of Intersectionality Theory in Quantitative Health Equity Analysis checklist in conducting this study.21,22 We used the Strengthening the Reporting of Observational Studies in Epidemiology and Reporting of Studies Conducted using Observational Routinely-Collected Data reporting statements (Appendix 1, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content) to guide reporting.23,24
Setting and data sources
We examined health administrative databases in Ontario, Canada, to assemble our cohort. These databases were linked using unique encoded identifiers and analyzed at ICES (Appendices 2 and 3, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content).25
The Ontario Health Insurance Plan Claims Database captured physician visits and diagnoses, the Canadian Institute for Health Information Discharge Abstract Database captured acute care hospital admissions, and the National Ambulatory Care Reporting System captured emergency department visits and other ambulatory encounters. We obtained patient sociodemographic characteristics and vital status from the Registered Persons Database. Additional data sources included the ICES Physician Database for provider specialty, the Home Care Database for home care services, the Continuing Care Reporting System for continuing care and long-term care use, the Ontario Drug Benefit claims database for outpatient prescription drugs, and the National Rehabilitation Reporting System for inpatient rehabilitation. We also used the Ontario Mental Health Reporting System; the Ontario Diabetes Database; the Ontario Marginalization Index; the Immigration, Refugees and Citizenship Canada Permanent Residents’ database; and the ICES-derived dementia cohort to ascertain comorbid conditions.
Study design
For this retrospective matched cohort study, we selected community-dwelling older adults (age > 65 yr) who sustained a new TBI between Apr. 1, 2004, and Mar. 1, 2020 (Appendix 4, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content). The exposure was late-life TBI diagnosis, identified through an emergency department visit or hospital admission, using International Statistical Classification of Diseases and Related Health Problems, 10th Revision codes (Appendix 5A, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content). This definition of TBI has a positive predictive value of 91.5%.26 We designated the day of TBI diagnosis as the cohort entry date, and for matched individuals, we randomly assigned index dates to match the distribution of index dates in the exposed group. We classified TBI severity as noncomplex (≤ 3 d) or complex (> 3 d) based on length of hospital stay (including in the emergency department), using a pragmatic, expert-consensus cut-off reflecting clinical complexity.27
We excluded individuals with prior dementia or TBI, and those residing in long-term care homes, receiving palliative care within 2 years before cohort entry, or older than 105 years at baseline. To ensure comparable look-back and follow-up periods, a random index date was assigned to individuals without TBI. Individuals were followed from the TBI diagnosis until incident dementia, end of follow-up (Mar. 1, 2021), or death (Figure 1). This observation period allowed sufficient time for dementia emergence. 12 Our primary outcome was incident physician-diagnosed dementia, defined using a validated algorithm with a sensitivity of 79.3% and specificity of 99.1% (Appendix 5B).28 Our secondary outcomes were rate of publicly funded home care use (including nursing, in-home personal support, case management, occupational therapy, and physiotherapy), publicly funded home care intensity for those receiving publicly funded home care (number of days per person-year), and rate of admission to a long-term care home.
Cohort study time frame of community-dwelling older adults who sustained a new traumatic brain injury (TBI) matched to those without TBI in Ontario, Canada. Accrual window = period during which individuals are being enrolled (recruited) into study; follow-up period = period after an individual’s entry to cohort until the end of study; index date = date of entry into cohort; look-back window = the period used to establish the baseline study population and to capture comorbidities and prior health care use, including outpatient physician and emergency department visits; observation window = period during which to look for outcomes.
We matched individuals based on age, sex, and a clinically derived propensity score of baseline covariates, and used greedy nearest neighbour 1:1 matching without replacement with a caliper width of 0.1.29,30 These covariates are shown in Appendix 6 (available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content). They were selected based on the existing literature and clinical judgment (Appendices 1B, 6, and 7A, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content).
To evaluate the impact of social determinants of health on outcomes, we applied Crenshaw’s theory of intersectionality as a guiding framework.31 Crenshaw’s work recognizes how experiences are shaped by multiple axes of marginalization and the social processes that create inequities.32 Examining health outcomes at intersections of identity, social position, oppression, privilege, policies, or institutional practices can identify patient subgroups who have unmet needs or experience care disparities. Social determinants of health variables were selected from existing literature and included age at TBI diagnosis; sex; immigration status (Canadian-born/long-term resident or immigrants born outside Canada);33–35 neighbourhood income quintile;36 community size; metropolitan influenced zone (available for rural communities and based on the proportion of workforce commuting to a metropolitan area);37 and neighbourhood-level ethnic diversity quintile, which reflects the proportion of residents who are recent immigrants or self-identify as visible minorities (Appendix 7B).38
Statistical analyses
We evaluated baseline imbalances between groups using standardized differences and considered them negligible if they were 0.1 or less (Appendices 6 and 8, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content).39 We constructed Cox proportional hazards models to derive cause-specific hazard ratios (HRs) and 95% confidence intervals (CIs) of incident dementia, publicly funded home care use, and admission to a long-term care home.40 We accounted for the competing risk of death and used robust standard errors to account for within-pair correlation induced by matching, ensuring accurate variance estimation despite nonindependence of observations.41,42 We used cumulative incidence function curves to estimate probability of outcomes over time.40 The 5-year time axis division was selected based on evaluation of the cumulative incidence function curve (≤ 5 yr and > 5 yr), addressing the initial proportional hazards assumption violation (Appendix 9, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content).43 To examine intensity of publicly funded home care use among individuals receiving it, we used modified Poisson regression to calculate an incidence rate ratio (IRR) and its 95% CI comparing the number of days per person-year. We completed E-value sensitivity analysis to determine the minimum strength of association that an unmeasured confounder would need to have to explain away an exposure–outcome association.44 We conducted post hoc sensitivity analyses to test the robustness of our results, including subgroup analyses by TBI diagnosis setting (emergency department v. hospital admission) and exclusion of dementia diagnoses within 3 months following TBI (Appendix 10, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content).45
Equity-informed analysis of social determinants of health
Using a theory-based framework of intersectional identities and positions, we predefined intersections of social determinants of health using a descriptive table46,47 and stratified by sex based on prior work indicating a higher risk of dementia among community-living females.12,48 We applied conditional inference trees as an exploratory analysis to evaluate how social determinants of health intersect (e.g., income and community size) with each other to affect risks of outcomes (Appendix 11, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content).49 Trees were fitted on the set of patients with new TBI, and outcomes of incident dementia, publicly funded home care use, and long-term care home admission, to identify the most impactful social determinants of health and their intersections.50 We set the minimum terminal node size to 5% of the cohort to ensure stable estimates. We used a significance threshold of p < 0.05 to iteratively select splitting variables.51,52 We conducted analyses using SAS v9.4 and Rstudio v3.6.3.52,53
Ethics approval
The use of data in this project is authorized under section 45 of Ontario’s Personal Health Information Protection Act and does not require review by a research ethics board.
Results
We included 132 113 matched pairs of individuals with and without TBI. The mean age of the TBI cohort was 77.2 (standard deviation 7.7) years on the date of cohort entry; 58.2% were female (Table 1 and Figure 2). Complex TBI patients made up 9.8% of the cohort. We achieved adequate matching based on our prespecified threshold (Table 1).
Characteristics of community-dwelling older adults who sustained a new traumatic brain injury (TBI) matched to those without TBI in Ontario, Canada
Cohort creation before and after matching of community-dwelling older adults who sustained a new traumatic brain injury (TBI) matched to those without TBI in Ontario, Canada. Controls were randomly assigned an index date based on their matched case, after which the remaining exclusion criteria were applied relative to that date. Note: LTCH = long-term care home, OHIP = Ontario Health Insurance Plan. See Related Content tab for accessible version.
Association between new TBI and incident dementia
Individuals with new TBI had a higher rate of incident dementia than matched individuals without TBI (Table 2). The median time to incident dementia diagnosis was 7.2 (interquartile range 2.6 to 14.4) years in the TBI and 9.2 (interquartile range 4.1 to 16.4) years in the non-TBI group. After we accounted for the competing risk of death, TBI was associated with a higher rate of incident dementia in the first 5 years (HR 1.69, 95% CI 1.66 to 1.72) and beyond 5 years (HR 1.56, 95% CI 1.53 to 1.59).
Cause-specific hazard models for rate of incident dementia, home care use, and long-term care home admission among community-dwelling older adults who sustained a new traumatic brain injury (TBI) matched to those without TBI in Ontario, Canada
Association between new TBI and support services use
Late-life TBI was associated with an increased rate of publicly funded home care use (HR 1.30, 95% CI 1.29 to 1.31) and admission to a long-term care home (HR 1.45, 95% CI 1.42 to 1.47). Patients with complex TBI had a significantly higher rate of all outcomes than those with noncomplex TBI (Appendix 12, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content); cumulative incidence function curves are shown in Figure 3. Among individuals who received publicly funded home care (n = 140 406, 53.1% of sample), those with new TBI had 87 days per person-year of home care, compared with 84 days among those without, with an IRR of 1.02 (95% CI 1.02 to 1.02). The E-value showed that an unmeasured confounder would need to have an association of at least 2.45-fold with new TBI and incident dementia to fully account for the observed association.44 Subgroup analyses by setting of TBI diagnosis (emergency department v. hospital admission) and excluding early dementia diagnoses yielded results consistent with the primary analyses (Appendix 10).
Cumulative incidence function for (A) incident dementia, (B) publicly funded home care use, and (C) long-term care home admission in community-dwelling older adults who sustained a new traumatic brain injury (TBI) matched to those without TBI in Ontario, Canada.
Impact of individual and intersecting social factors on outcomes in individuals with new TBI
Females with new TBI were older than males (Table 3). Age was the most important predictor of all outcomes (Figure 4). Individuals aged 85 years and older had the highest probability of incident dementia, at 31.3%. Among individuals aged 66 to 74 years, those who were from low-income neighbourhoods had a higher probability of incident dementia than those from high-income neighbourhoods (Figure 4A). Individuals aged 75 years and older and from larger communities had a greater probability of receiving publicly funded home care than those from smaller communities (Figure 4B). Individuals living in smaller, low-income, and less ethnically diverse communities were more likely to be admitted to a long-term care home than those from larger, high-income, and more ethnically diverse neighbourhoods (Figure 4C).
Baseline social determinants of health stratified by sex (female v. male) among older adults who sustained a new traumatic brain injury in Ontario, Canada
Conditional inference trees showing associations between (A) age, sex, and neighbourhood income quintile and probability of incident dementia; (B) age and community size and probability of home care use; and (C) age, community size, and neighbourhood income quintile and probability of long-term care home (LTCH) admission among community-dwelling older adults who sustained a new traumatic brain injury (TBI) in Ontario, Canada. Note: Q = quintile. Conditional inference trees illustrate associations between prespecified social determinants of health and the probability of incident dementia, publicly funded home care use, and long-term care home admission among community-dwelling older adults with a new TBI. Six candidate variables (age, sex, neighbourhood income quintile, neighbourhood ethnicity quintile, immigrant status, and community size) were considered for each outcome; variables that do not appear in each tree were not associated with that outcome at the significance threshold (p < 0.05). Within each tree, variables may appear in some age groups and not others because there were no significant associations in those subgroups. See Related Content tab for accessible version.
Interpretation
Using a population-based matched cohort study of more than 260 000 older adults with and without TBI, we found that new TBI was associated with an increased rate of incident dementia, publicly funded home care use and intensity, and admission to a long-term care home. The hazard of incident dementia was highest in the first 5 years following TBI. We also identified that age and key social determinants of health were associated with outcomes: older age (≥ 75 yr), female sex, and low area-level income were associated with higher dementia risk; older age and living in larger communities were associated with higher probability of publicly funded home care use; and older age, residence in smaller communities, low area-level income, and low ethnic diversity were associated with higher probability of long-term care home admission.
Our study provides a population-level and long-term assessment quantifying the association between incident dementia and new TBI in older adulthood. Although TBI has been studied as a risk factor for dementia in adulthood,13,54,55 our findings emphasize its significant association with an increased rate of incident dementia, even when it is sustained in late life, and how this risk changes over time. This critical information can assist clinicians in guiding older patients and their families to better understand long-term risks. Raj and colleagues56 and Zhang and colleagues57 reported associations between TBI and increased risk of dementia among individuals sustaining a TBI before ages 64 and 69 years, respectively, leaving a gap in our understanding of dementia risk among the oldest age groups.20 Our study addresses this gap with a median age of 77 years and a follow-up period of up to 17 years. Despite differences in study populations, settings, and follow-up duration, our effect estimates align with those from these earlier studies (Appendix 13, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.250361/tab-related-content). Another cohort study including subgroups of older adults had a shorter follow-up of less than 7 years.58 We found that the association between new TBI and incident dementia was strongest in the first 5 years of follow-up, likely owing to reduced latency to dementia in individuals with TBI and the increased incident rate of dementia with age, attenuating the association in older age groups.59,60 One of the most common reasons for TBI in older adulthood is sustaining a fall, which is often preventable.4 By targeting fall-related TBIs, we can potentially reduce TBI-associated dementia in this population.61
Our exploratory analyses found that in older adults with new TBI, those who were female, aged 75 years and older, and from low-income areas had a higher probability of developing dementia than those who were aged 66 to 74 years, male, and from high-income areas. Regardless of genetic risk, socioeconomic deprivation independently increases dementia risk.62 Socioeconomic disadvantage may exacerbate TBI-related vulnerability through factors such as limited access to rehabilitative care, comorbidities, and reduced cognitive reserve.20,63 Interventions to counteract downstream effects of low socioeconomic status may include improving access to primary care, healthy diet, and spaces for physical activity.62,64,65 Moreover, we noted that patterns of support services utilization diverged markedly: individuals living in larger communities were more likely to receive home care, whereas those from smaller, low-income, and less ethnically diverse communities were more likely to be admitted to a long-term care home. These patterns may reflect limited service availability and structural barriers that disproportionately affect these populations.66,67 Residents of smaller communities may encounter challenges including limited community specialized care and potential transportation challenges, underscoring the need for targeted policies to improve service accessibility (e.g., day programs, care partner respite, and meal delivery services) and promote equity.66–68 Our findings suggest that, to better align limited health resources with population needs, specialized programs such as community-based dementia prevention programs, and support services should be prioritized for female older adults (≥ 75 yr) living in smaller communties and low-income and low-diversity areas.
Limitations
We could not describe the potential impacts of some social determinants of health, including gender identity, education, and social capital.33,69 The Canadian algorithm for identifying dementia has lower sensitivity for mild disease owing to diagnostic uncertainty in the early stages and does not distinguish between dementia subtypes (e.g., Alzheimer, vascular, or frontotemporal dementia).28,70 We also could not assess the cyclical link between cognitive impairment and TBI, where sustaining a TBI increases dementia risk, and dementia, in turn, raises the likelihood of falls and subsequent TBIs. This complex interplay was beyond our study’s scope. Additionally, our study identified individuals based on newly recorded TBIs for which the individual sought hospital assessment and does not account for recurrent injuries. Home care use does not include privately paid services. Our findings may not be generalizable to individuals managed solely in outpatient care and with repeated TBIs. Future research should explore the impact of recurrent injuries, particularly in older, community-dwelling adults.
Conclusion
New late-life TBI was associated with higher rates of incident dementia, use of publicly funded home care, and admission to a long-term care home in a matched population-based cohort study with up to 17 years of follow-up among community-dwelling older adults. Among individuals with new TBI, older females from low-income neighbourhoods were more likely to develop incident dementia than their male counterparts. Residents of smaller communities were less likely to access home care and had higher probability of admission to a long-term care home if they were from low-income and less-diverse areas. Reducing injury burden and addressing disparities in outcomes will require upstream preventive efforts, such as fall prevention, and tailored access to post-TBI rehabilitation and support services.
Footnotes
-
Competing interests: Manav Vyas is an associate editor for the Canadian Journal of Neurological Sciences. Peter Hoang has received honoraria from the Canadian Coalition for Seniors’ Mental Health. Mark Bayley reports grants from the Canadian Institutes of Health Research, Brain Canada Foundation, Heart and Stroke Foundation, UHN Foundation, National Institutes of Health Research, Praxis Spinal Cord Institute, and Ministry of Health of Ontario, and meeting support from the Government of Mauritius and Brain Injury Australia. Dr. Bayley is chair of Rehabilitation Care Alliance of Ontario and receives a stipend as medical director of the University Health Network Rehabilitation Program. No other competing interests were declared.
-
This article has been peer reviewed.
-
Contributors: Yu Qing Huang, Jennifer Watt, Sharon Straus, Manav Vyas, Corita Vincent, and Susan Bronskill contributed to the conception and design of the study. Yu Qing Huang formulated the research question and developed the study proposal and the data creation plan, including cohort definitions and covariates. Jennifer Watt, Sharon Straus, Manav Vyas, and Susan Bronskill provided substantial revisions to the study protocol and data creation plan. Following data curation by Zhiyin Li and Jun Guan to ensure patient privacy and data integrity, Yu Qing Huang conducted the statistical analyses. Methodological decisions and interpretation of results were reviewed in collaboration with Jennifer Watt, Sharon Straus, Manav Vyas, Peter Hoang, Corita Vincent, Alan Tam, Mark Bayley, and Susan Bronskill. Yu Qing Huang drafted the manuscript. All authors contributed meaningfully to critical revisions, approved the final version to be published, and agreed to be accountable for all aspects of the work.
-
Funding: This study was funded by the Knowledge Translation Program of St. Michael’s Hospital and Physicians Services Incorporated Foundation Resident Research Grant. This study is a chapter of Yu Qing Huang’s doctoral thesis work in Clinical Epidemiology and Health Care Research at Institute of Health Policy, Management and Evaluation, University of Toronto. Yu Qing Huang is supported by the Alzheimer Society Research Program doctoral award. Yu Qing Huang and Peter Hoang are supported by the University of Toronto Department of Medicine Eliot Phillipson Clinician–Scientist Training Program. Peter Hoang is supported by the Vanier Canada Graduate Scholarship. Dr. Manav Vyas holds a New Investigator Award from Heart and Stroke Foundation of Canada.
-
Data sharing: The data set from this study is held securely in coded form at ICES. Although data-sharing agreements prohibit ICES from making the data set publicly available, access may be granted to those who meet prespecified criteria for confidential access, available at The full data creation plan and underlying analytic code are available from the authors on request, with the understanding that the code may include ICES-specific macros and templates that are not publicly accessible or may not be fully transferable outside the ICES environment. This study used the Johns Hopkins ACG System Version 10 Aggregated Diagnosis Groups.
-
Disclaimers: Sharon Straus is a member of the CMAJ Governance Council and was not involved in the editorial decision-making process for this article. This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC). The funders had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. This study used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under licence from Canada Post Corporation and Statistics Canada. Parts of this material are based on data and/or information compiled and provided by the Canadian Institute for Health Information (CIHI), Immigration, Refugees and Citizenship Canada (IRCC) current to May 15, 2024, and the MOH. Some data in our study were adapted from Statistics Canada’s Ontario Census Area Profiles or CENSUS library, which was last updated on Dec. 22, 2023. However, the analyses, conclusions, opinions and statements expressed in the material are those of the author(s), and not necessarily those of CIHI, IRCC, MOH, and Statistics Canada. The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred. The authors thank IQVIA Solutions Canada Inc. for use of its Drug Information File and Toronto Community Health Profiles Partnership for providing access to the Ontario Marginalization Index.
This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY-NC-ND 4.0) licence, which permits use, distribution and reproduction in any medium, provided that the original publication is properly cited, the use is noncommercial (i.e., research or educational use), and no modifications or adaptations are made. See:
link