Does Mild Traumatic Brain Injury Increase the Risk for Dementia? A Systematic Review and Meta-Analysis

Abstract

Background:

Mild traumatic brain injury (mTBI) is a putative risk factor for dementia; however, despite having apparent face validity, the evidence supporting this hypothesis remains inconclusive. Understanding the role of mTBI as a risk factor is becoming increasingly important given the high prevalence of mTBI, and the increasing societal burden of dementia.

Objective:

Our objective was to use the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) format to determine if an association exists between mTBI and dementia and related factors, and to quantify the degree of risk.

Methods:

In this format, two authors conducted independent database searches of PubMed, PsycInfo, and CINAHL using three search blocks to find relevant papers published between 2000 and 2020. Relevant studies were selected using pre-defined inclusion/exclusion criteria, and bias scoring was performed independently by the two authors before a subset of studies was selected for meta-analysis. Twenty-one studies met the inclusion criteria for this systematic review.

Results:

The meta-analysis yielded a pooled odds ratio of 1.96 (95% CI 1.698–2.263), meaning individuals were 1.96 times more likely to be diagnosed with dementia if they had a prior mTBI. Most studies examining neuropsychiatric and neuroimaging correlates of dementia found subtle, persistent changes after mTBI.

Conclusion:

These results indicate that mTBI is a risk factor for the development of dementia and causes subtle changes in performance on neuropsychiatric testing and brain structure in some patients.

Keywords

Alzheimer’s disease dementia mild traumatic brain injury

INTRODUCTION

Despite the benign nature of the term, mild traumatic brain injury (mTBI) is a significant risk factor for many serious conditions. An mTBI is defined as “an acute brain injury resulting from mechanical energy to the head from external physical forces, including loss of consciousness for 30 minutes or less, post-traumatic amnesia for less than 24 hours, and a Glasgow Coma Score of 13–15 after 30 minutes post-injury or upon first presentation for healthcare” [1]. Aside from the immediate and short-term effects of an mTBI, there is an increasing body of evidence linking mTBI to the development of long-term psychiatric and neurological conditions that include depression, bipolar disorder, and neurodegenerative diseases [2]. The constellation of neurodegenerative diseases linked to mTBI includes chronic traumatic encephalopathy (CTE), a tauopathy that has gained attention due to its prevalence in professional athletes from the National Football League (NFL) [3]. The incidence of mTBI is also high in the general population, with as many as 42 million people suffering an mTBI worldwide each year [4]. Thus, there is growing pressure to better understand the pathophysiology and long-term risks associated with mTBI to help direct treatment and prevention strategies to alleviate a growing health care burden.

Understanding mTBI in an aging population presents additional challenges. Cognitive abilities, including memory, processing speed, and reasoning, are thought to naturally decline with age [5]. The diagnosis of mild cognitive impairment (MCI) was introduced to identify individuals who demonstrate greater cognitive decline than seen in typical aging but have not yet reached dementia-level impairment [6]. Unlike MCI, a diagnosis of dementia requires impairments in daily-life functions in addition to cognitive difficulties [7]. Five main cognitive domains may be impaired during cognitive decline: learning and memory, language, visuospatial, executive, and psychomotor [7]. An MCI diagnosis typically requires one of these areas to be impaired, but several must be impaired for a diagnosis of dementia [7]. MCI is a risk factor for further neurodegeneration; for example, a 2009 meta-analysis found a 9.6% annual conversion rate of MCI to dementia in an adjusted clinical setting, and 4.9% in community studies [8]. Dementia is a term that encompasses several age-related neurodegenerative disorders that affect a large portion of the population: up to 7.2% of Canadians over the age of 65 and nearly 20% of Canadians over the age of 80 [9, 10]. Of all the dementias, Alzheimer’s disease (AD) is the most common, accounting for up to 80% of diagnoses [11]. While our understanding of dementia etiology remains limited, several dementia risk factors have been identified. The Alzheimer’s Society of Canada describes many risk factors for AD and other dementias including modifiable factors like high blood pressure and cholesterol, as well as non-modifiable factors such as sex and genetics [12]. Although risk factors are not causes of disease on their own, understanding and managing risk factors is an important strategy to minimize susceptibility to disease. This is especially important in the context of dementia, due to the lack of disease-modifying treatments.

Improving early diagnosis of dementia is important, as early detection has been found to have health, social, and societal benefits [13]. Currently, collecting a medical history and conducting a mental status examination is standard practice for the diagnosis of both MCI and dementia. Medical history is used to assess if the patient has impairment in daily functioning, while the mental status examination is used to assess the level of cognitive impairment [7]. Several cognitive assessments exist for diagnosing cognitive impairments, such as the Montreal Cognitive Assessment (MOCA), the Short Test of Mental Status (STMS), and the Mini-Mental State Examination (MMSE). These assessments also serve as objective measures of global cognitive functioning. There is a large interest in the early detection of MCI and dementia; protein biomarkers and advanced imaging techniques have shown promising results, but more research is needed [14, 15]. Since current treatments for AD and other dementias mainly focus on symptom control, strategies to reduce dementia risk involve controlling the known risk factors [16] and early detection [17]. Identification of these risk factors and early detection methods will allow future researchers to target interventions toward people at high risk of developing dementia.

Regarding traumatic brain injuries, it seems that a clear relationship exists between moderate and severe brain injury and the development of dementia [18 –21]. However, it is notable that the methodological rigor of studies that assess the association between TBI, dementia, and AD has recently been questioned [22], and further the association between TBI, AD, and other apolipoprotein E (APOE) specific neurodegeneration has been debated [21]. A 2013 review of neuroimaging studies suggests that moderate to severe TBIs cause reductions in total brain volume, dynamic degeneration of white matter, and loss of connectivity, all of which could predispose the brain to dementia later in life [23]. However, the relationship between mTBI and dementia is less understood and has become an extensive area of research largely in part to the finding of CTE in NFL players [24]. In 2004, the World Health Organization Collaborating Centre Task Force on mTBI found inconclusive evidence around mTBI and dementia after examining three studies carried out in the 1990s [25]. More recently, a 2014 meta-analysis did not demonstrate a relationship between mTBI and subsequent MCI/dementia [26], although, only one study was included in the analysis. While a 2016 meta-analysis suggested a positive correlation, only a small number of studies specifically examined mTBI and dementia, few of these used recent criteria for defining mTBI, and most were published before 2005 [2]. Due to the widespread prevalence of mTBI and the high clinical burden of MCI and dementia, identifying whether mTBI interacts with the development of these degenerative diseases is of high importance.

Given the inconsistencies in the body of evidence around mTBI and dementia [2 , 26], we aimed to provide an update of the literature in this area with this systematic review. Our objectives were to identify recent publications that examine remote mTBI (an injury sustained years in the past) and dementia, to quantify any associated risk, and to identify gaps and methodological issues in the literature to help direct future research. We also included MCI, and associated factors, in our search strategy because of the strong association between MCI and the subsequent development of dementia. Should strong evidence link mTBI to dementia, appropriate measures can be taken to identify at-risk populations and implement prevention measures. Conversely, strong evidence against mTBI as a risk factor for dementia may reduce the need for dementia risk management in favor of treatments that target other mTBI complications.

METHODS

The methodology for this review was based on the Preferred Reporting Items for Systematic Reviews and Meta-Analysis framework and is reported following the PRISMA 2009 checklist (Supplementary Table 1); the review protocol was not registered [27]. To identify primary, peer-reviewed articles published between 2000 and 2020 that characterized the risk of developing MCI and dementia after a remote history of mTBI, three blocks of search terms were developed. The first block of search terms identified mTBI studies and included the following terms: “concuss*”, “mild traumatic brain injur*”, “mTBI”, “mild head impact”, “mild head inj*. The second concerned the development of MCI/dementia and included the following terms: “Mild cognitive impairment”, “neurocognitive impairment”, “cognitive decline”, “cognitive impairment”, “mild neurocognitive disorder”, “major neurocognitive disorder”, “delirium”, “preclinical Alzheimer’s”, “neurodegenerative diseases”, “Alzheimer*”, “dementia*”. The third search block was meant to exclude review articles, and contained the terms “review”, meta-analysis”, “meta-review”, and “literature review”. The first block and second blocks were separated by an “AND” modifier, while the third block was separated by a “NOT” modifier.

Using these search blocks, the co-first authors independently searched the databases PubMed, CINAHL, and PsycInfo and exported results found between the years 2000 and 2020 to the Mendeley© citation manager. Following the PRISMA flow diagram protocol, duplicates were removed and all papers were screened using the inclusion/exclusion criteria (Table 1). The authors also conducted forward and reverse searches of previous review articles to ensure all relevant papers were included in this review. Papers selected in the screening process were read in full and assessed for eligibility as defined by the search criteria. The remaining papers were compared between authors and any discrepancies were discussed and resolved. A total of 21 papers were included in this review (Fig. 1). Studies were included if they were primary research papers available in full text and published in English after 2000, included human subjects, included a measure of assessing cognitive impairment, and included a clinical diagnosis of mTBI. Conversely, studies were excluded if they violated any of the inclusion criteria, only examined treatments, or examined only the acute effects of mTBI (rather than the long-term effects of a remote injury). A list of relevant full-text articles was generated and read in full to create a final list of papers for inclusion in this review. After the generation of the final list, relevant information from the selected studies was extracted by the co-first authors. Relevant information included outcome measures, findings related to dementia risk, and population characteristics.

Table 1

Inclusion/exclusion criteria

Inclusion Criteria	Exclusion Criteria
1. Primary, peer-reviewed studies	1. Studies that only examined treatments
2. English	2. Non-primary research papers (reviews, posters, abstracts)
3. Human	3. Published in a language other than English
4. 2000-present	4. Non-human studies
5. Full text available	5. No full-text available
6. Used a screening tool used to diagnose MCI or Dementia, such as the MMSE or MOCA or a clinical diagnosis	6. Examines the acute effects of the injury exclusively and not long-term neurodegenerative changes
7. Used a clinical diagnosis of mTBI according to the WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury and the Centers for Disease Control and Prevention or a similar definition	7. No use of a screening tool such as the MMSE, MoCA or a clinical diagnosis of MCI/Dementia
8. No clinical diagnosis of mTBI

Fig. 1

Graphical depiction of the search process. After duplicate removal, articles identified through database searching were screened using their title and abstract for inclusion, generating a final list of 57 full text articles to screen for inclusion in the systematic review.

Bias scoring was performed independently by two authors using the Scottish Intercollegiate Guidelines [28]. The authors individually answered a checklist of questions regarding the methodology of each study, and from that studies were classified as high quality, acceptable, or unacceptable (Supplementary Table 2). Discrepancies between authors were resolved by consensus.

A subset of studies was selected for inclusion in a meta-analysis. To be included in the meta-analysis, studies had to include the risk of dementia after sustaining an mTBI as an outcome measure and quantify this risk using an odds ratio. Also, the authors had to provide the raw data needed to perform the meta-analysis (see Table 2 for narrow inclusion/exclusion criteria). Data were pulled directly from the publications where possible; for four studies, the authors were contacted directly to obtain data. Pooled odds ratios (ORs), 95% confidence intervals (CIs), and tests for heterogeneity were performed using MedCalc Statistical Software [29]. A random effects model was chosen to account for any potential variations in effect size caused by potential covariates in the selected participant populations. OR weights for each study were assigned based on the inverse of their respective variances. The meta-analysis forest plot was made using the ggplot2 package for Rstudio [30].

Table 2

Narrow inclusion/exclusion criteria for inclusion in the meta-analysis

Inclusion Criteria	Exclusion Criteria
Provided an objective measure of dementia risk using an odds ratio or equivalent measure	Data was not available and authors did not provide data after being contacted

RESULTS

Study selection

The co-first authors individually performed a literature search of the PubMed, CINAHL, and PsycInfo databases to find a total of 1,011 studies for initial screening. After removing duplicates, 699 studies remained and the abstracts were screened for relevance. Following the screening, 57 full-text articles were assessed for eligibility, while 642 did not meet the inclusion/exclusion criteria. From the full-text articles, 36 were excluded: six had the wrong severity of injury, twelve had no confirmation of mTBI, two used animal models, while 16 did not use a screening tool for MCI/dementia. A total of 21 studies were included in this review, six of which were included in the quantitative analysis (see Fig. 1 for a stepwise breakdown of the search process).

Objective risk of dementia after mTBI

The outcome measures of twelve studies concerned identifying the risk of cognitive impairment (i.e. MCI, dementia, AD) following mTBI [31 –42]. Specifics about each study, including population characteristics, definitions of mTBI and dementia used, outcome measures, time between injury and diagnosis, main findings and bias scores can be found within Table 3.

Table 3

Objective risk of dementia following mTBI

Paper	Population Characteristics	mTBI Definition	Outcome Measures	Time between injury and dementia diagnosis	Findings	Bias Score	Main effect of remote mTBI
Barnes et al., 2018 [31]	Veterans who had received a TBI diagnosis (n = 178,779), and a propensity matched sample (n = 178,779) recruited from the Veterans Health Administration	TBI evaluations performed by trained professional, recorded using ICD-9 codes	A diagnosis of dementia as defined by ICD-9 codes	3.6 y (all TBI severities)	After adjustment for age, medical comorbidities, and psychiatric disorders, the adjusted HR for dementia diagnosis was 2.36 (95% CI, 2.10, 2.66) for mild TBI without LOC; 2.51 (95% CI, 2.29–2.76) for mild TBI with LOC; 3.19 (95% CI, 3.05–3.33) for mTBI with LOC unknown	HQ	Increased risk of dementia
Helmes et al., 2011 [32]	A subset of Canadian adults recruited from the CHSA-I and CHSA-II (n = 2,431) who scored below cut-off on a global cognitive screening tool. Reassessment occurred 5 y after initial intake.	Self-report confirmed by informant, TBI with LOC 0–30 min	A diagnosis of dementia (DSM-III-R)	27.8 y (SD = 22.6) (all TBI severities)	Having a history of mTBI did not improve the predictive ability of dementia diagnosis. Age remained the most related factor.	A	No change in risk of dementia
Lee et al., 2013 [35]	A general population of Taiwanese adults who had sustained an mTBI (n = 28,551), and controls who had not sustained an mTBI (692,382). Data was obtained from the NHI database	ICD-9 codes were used to identify mTBI	Presence of dementia or AD	1.0 y (95% CI 0.77–1.23)	In adults aged 65+, the adjusted HR was 3.27 (95% CI, 2.67–4.00).	HQ	Increased risk of dementia
Li et al., 2016 [36]	Adults aged 55–90 who had sustained a TBI (N = 79; mTBI = 71), and a non-TBI control group (n = 1197). Data was obtained from ADNI database	Self-reported concussion history corroborated by informants at ADNI clinical site	Age of Onset of MCI or AD	Age at injury = 35.4 (SD = 24.8)	Adults with mTBI history have earlier ages of onset of MCI than controls by about 4 y (p = 0.016)	A	Earlier age of onset of MCI
Liao et al., 2020 [37]	A general population of Taiwanese adults aged 65 + with a history of LOAD (n = 4,600) and a matched control group (n = 4,600). Data was obtained from both the NHIRD or CIC databases.	ICD-9 codes were used to identify mTBI	Identification of factors associated with LOAD	Not reported	Incidence of concussion up to 4 y prior to LOAD diagnosis has significant positive effects on incidence (total effects; 0.050)	A	Increased risk of LOAD
McMillan et al., 2014 [38]	A general population of patients from a Glasgow hospital identified as having mTBI (n = 2,428), a matched non-head related injury control group (n = 2,428), and a matched control group (n = 2,428).	GCS score between 13–15	Identification of death rates and comorbidities 15 y post injury	Not reported	In adults aged under 65 at baseline, 15 y later dementia rates in the mTBI group were 0.37/1000/y, while the control group had a rate of 0.14/1,000/y	U	Increased risk of dementia* *No statistical tests conducted
Nordström &Nordström, 2018 [39]	A general population of Swedish adults with data available in the Swedish National Patient Register. The first cohort was prospectively followed after mTBI (n = 491,252), the second cohort consisted of siblings with discordant TBI status (n = 93,940) and the third cohort was a retrospective case-control cohort (n = 404,877).	ICD-10 codes were used to identify mTBI	Diagnosis of dementia using ICD 8–10 codes	15.3 y for total cohort	Risk of dementia after mTBI has a dose dependent relationship, with mild TBIs having less associated risk than severe or multiple TBIs. The adjusted OR for mTBI and dementia was 1.63 (95% CI 1.57–1.70), 1.49 (95% CI 1.23–1.80) and 1.59 (95% CI 1.54–1.65) in cohorts 1, 2 and 3, respectively	HQ	Increased risk of dementia
Plassman et al., 2000 [40]	World War II Veterans who has sustained a TBI (n = 1,422), pneumonia (n = 1,740), or a non-head related injury (n = 2,282). Participants were recruited from the pension database for Veteran Affairs.	mTBI categorized using a scale described by Frankowski et al., 1985 [78]	Cognitive assessments, neuropsychological tests, APOE genotype, diagnosis of AD	Not reported	The mTBI group showed no significant risk of dementia or AD	A	No change in risk of AD
Redelmeier et al., 2019 [41]	General population of adults aged 66 + with a diagnosis of concussion, and who were prescribed a statin (n = 7,058), and who were not prescribed a statin (n = 21,757), and a non-head related injury control group prescribed a statin (n = 77,898), and not prescribed a statin (n = 229,992). Data was obtained from the OHIP database.	ICD-9 codes were used to identify mTBI	Presence of dementia according to ICD-9 codes from OHIP database, with diagnosis on 2 separate dates to avoid false positives	3.9 y for all patients	Patients not receiving a statin had a dementia incidence of 43 cases per 1,000 patients annually, which is greater than 2x the population norm	HQ	Increased risk of development of dementia
Sercy et al., 2020 [42]	General population of adults who sought treatment for mTBI (n = 964) or another injury (n = 5,567), recruited from six level 1 trauma centers across the US. Medical information was obtained from the USCDC NDI database.	ICD-9 codes were used to identify mTBI	Mortality, causes of death and comorbidities after mTBI	5 y	OR of patients with a history of mTBI developing a neurodegenerative disease was found to be 1.62 (95% CI 0.88–2.98)	A	Increased risk of developing a neurogenerative disease
Sundström et al., 2007 [33]	Dementia patients (n = 181) and controls (n = 362) data obtained from Betula prospective longitudinal study of memory, health and aging.	Self-reports corroborated by medical record in some patients; ability of patient to perform cognitive tests taken as indicative of mild injury	Risk of dementia and APOEɛ4 genotyping	More than 5 y before diagnosis	APOE positive status increased the risk of developing dementia with head injury	U	Increases risk only in APOE4 + patients
Yang et al., 2019 [34]	General Taiwanese population with craniofacial trauma and mTBI (n = 399,057), facial bone fracture (n = 23,890), or moderate to severe TBI (n = 149,475). Data was obtained from the NHIRD database.	ICD-9 codes were used to identify mTBI	Incidence of dementia in patients with various types of head injuries	Not reported	A standardized incidence ratio for mTBI and dementia was found to be 1.78 (95% CI; 1.73–184)	HQ	Increased risk of development of dementia following mTBI

Acronyms are separated by column. Population Characteristics: CHSA, Canadian Study of Health and Aging; mTBI, mild traumatic brain injury; NHI, National Health Index; TBI, traumatic brain injury; ADNI, Alzheimer’s Disease Neuroimaging Initiative; NHIRD, National Health Insurance Research Database; CIC, Catastrophic Illness Certificate; OHIP, Ontario Health Insurance Program; USCDC, United States Centre for Disease Control; NDI, National Death Index. mTBI Definition: ICD-9, International Classification of Disease - 9th edition; LOC, loss of consciousness; GCS, Glasgow Coma Scale. Outcome Measures: DSM, Diagnostic and Statistical Manual for Mental Disorders –Third Edition; MCI, mild cognitive impairment; AD, Alzheimer’s disease, LOAD, late onset Alzheimer’s disease; APOE, Apolipoprotein E. Time Between Injury and Dementia Diagnosis: y, year; SD, standard deviation; CI, confidence interval. Findings: HR, hazard ratio; OR, odds ratio. Bias Score: HQ, high quality, A, acceptable, U, unacceptable.

Three retrospective studies considered the effects of mTBI on the development and/or onset of later-life dementias (including AD) [31 , 34]. Two studies found an increased proportion of people diagnosed with dementia had a history of at least one mTBI [31, 34], while the third found no change in the number of people diagnosed with dementia [32]. Barnes et al. (2018) [31] examined the relationship between TBI severity, loss of consciousness (LOC), and dementia diagnosis in 357,558 Veterans. The researchers found an adjusted hazard ratio for dementia diagnosis of 2.36 (95% CI, 2.10, 2.66) for mTBI without LOC; 2.51 (95% CI, 2.29–2.76) for mTBI with LOC; and 3.19 (95% CI, 3.05–3.33) for mTBI with LOC unknown [31]. Yang et al. (2019) [34] assessed dementia risk following craniofacial trauma in 501,889 patients. Researchers found that patients with past mTBI were more likely to be diagnosed as having dementia later in life than matched controls (2.75% versus 1.28%; p < 0.01), further, the same team found a standardized incidence ratio of 1.78 (1.73–1.84) for mTBI and dementia. Helmes et al. (2001) [32] investigated the development of dementia in adults with a history of mTBI using samples from two separate waves of the community-based Canadian Study of Health and Aging. Participants who identified as having a TBI before the first wave were followed up within the second wave to assess cognitive status five years later. The researchers found that having a history of mTBI was not associated with a dementia diagnosis.

Six prospective studies considered the effects of mTBI on the development of later-life dementias (including AD) [33 , 40–42]. Five of the six papers found increased rates of dementia diagnosis in at least one subset of individuals [33 , 42]; however, one of these papers did not use statistical assessments to verify this conclusion [38]. The objective of Lee et al. (2013) [35] was to identify the incidence of dementia in Taiwanese patients aged 18+ with and without histories of concussion. A total of 720,933 adults were followed for 8 years post-injury. In all participants, researchers found the adjusted hazard ratio to be 3.26 (95% CI, 2.69–3.94), and in adults 65+ the hazard ratio was 3.27 (95% CI, 2.67–4.00) for the diagnosis of dementia following remote mTBI. McMillan et al. (2014) [38] explored mortality rates and associated comorbidities 15 years after mTBI in a sample of 7,284 adults in Glasgow. At admission, after mTBI the mean age was 39 (range 14–98), and 15 years later the dementia rate in those with mTBI was 0.37/1000/year versus case controls (0.14/1000/year). However, no statistical assessments were done due to low power. Redelmeier et al. (2019) [41] tested whether statin use is associated with a change in risk of dementia in older adults after a concussion; for this review, we are only reporting on the group that did not receive a statin. It was found that mTBI control patients accounted for 3,677 dementia cases over 85,339 patient-years (mean, 3.9 years), equal to an incidence of 43/1,000 patients annually, greater than twice the normal population. Sercy et al. (2020) [42] explored the mortality rate and causes of death in mTBI patients five-years post-injury, and found an odds ratio of 1.62 (95% CI, 0.88–2.98) for the development of a neurodegenerative disease, although this was not specific to dementia. Sundström et al. (2007) [33] examined 181 subjects with dementia and 362 controls and assessed the prevalence of mTBI using a case definition provided by the authors. They found that mTBI conferred an increased risk of dementia in only carriers of APOE ɛ4 genotype. In contrast, Plassman et al. (2000) [40] investigated the association between mTBI and dementia in a military population. The researchers found no effect of mTBI on the development of later-life dementia (hazard ratio: 2.56, 95% CI: 0.68 to 9.67), no difference in age of onset of dementia, and no interaction effect between mTBI and APOE ɛ4 genotype on the risk of dementia.

Two studies retrospectively examined if a history of mTBI influences the age of onset in adults with a clinical diagnosis of MCI and/or dementias [36, 37]. Liao et al. (2020) [37] examined how prior conditions, including mTBI, are associated with the development of late-onset AD in a general Taiwanese population. In this sample of 9,200 adults aged 65+, mTBI within 9 years was positively associated with late onset AD diagnosis. Li et al. (2016) [36] assessed if a history of TBI and severity are associated with the age of onset of MCI and dementia. Researchers found that mTBI is related to an earlier age of onset than adults with no history of mTBI (68.5±1.1 years; 95% CI 66.3–70.7 versus 70.9±0.2 years; 95% CI 70.5–71.4).

One study included both prospective and retrospective analyses [39]. Nordstrom & Nordstrom (2018) assessed all inhabitants of Sweden aged 50 years and over. In the first cohort, 164,334 individuals with TBI were age and sex-matched to two individuals between 1964 and 2012, and the National Patient Register (NPR) was prospectively searched for diagnosis of dementia. In patients with mTBI, the adjusted odds ratio of developing dementia in the 48 years follow up was 1.63 (95% CI, 1.57–1.70). The second cohort involved siblings in the NPR who had different diagnoses of TBI (i.e., no TBI versus TBI), to control for familial factors. Researchers found a slight increase in dementia rates in the second cohort with an adjusted odds ratio of 1.49 (95% CI, 1.23–1.80). The third cohort was a prospective analysis of individuals diagnosed with dementia in the NPR. Researchers found a small increase in having a history of mTBI in adults diagnosed with dementia (aOR, 1.59; 95% CI, 1.54–1.65). In all cohorts, mTBI had the weakest association with dementia when compared with severe TBI and repeated TBIs.

Meta-analysis of quantitative studies

Six studies met the narrow criteria (Table 2) and were included in the meta-analysis [31 , 39–41]. Using a random-effects OR model, prior mTBI was found to be associated with the later development of dementia (OR 1.96, 95% CI 1.698–2.263, p < 0.001) (Fig. 2, Table 4). The studies exhibited a high degree of heterogeneity (I² = 97.30%).

Fig. 2

Forest plot of individual and pooled ORs for studies included in the meta-analysis.

Table 4

Pooled Odds Ratio for developing dementia after remote mTBI

Study	Intervention	Controls	Odds Ratio	95% CI	Weight (%) Random
Barnes et al., 2018 [31]	3,972/91,888	4,698/174,081	1.629	1.560 to 1.701	21.01
Lee et al., 2013 [35]	127/28,424	944/691,438	3.283	2.727 to 3.952	15.87
Nordström &Nordström, 2018 [39]	5,846/108,463	7,314/216,077	1.626	1.570 to 1.684	21.14
Redelmeier et al., 2019 [41]	3,677/21,757	19,717/229,992	2.169	2.087 to 2.254	21.09
Yang et al., 2019 [34]	4,451/394,607	2,497/394,607	1.791	1.705 to 1.882	20.89
Total (random effects)	18,073/645,139	35,170/170,6195	1.972	1.705 to 2.281	100

Structural and neuropsychiatric changes after mTBI

We identified nine studies that examined structural and/or neuropsychiatric outcomes associated with MCI/dementia after remote mTBI (Table 5). For inclusion of this section, studies had to use a global cognitive screening tool, such as the MMSE or MOCA, to delineate cognitive function. Specifics about each study, including population characteristics, the definition of mTBI and measures of global cognitive functions used, outcome measures, time between injury and diagnosis, main findings and bias scores can be found within Table 5.

Table 5

Structural and neuropsychiatric changes after mTBI

Paper	Population Characteristics	mTBI Definition	Outcome Measures	Findings	Bias Score	Main effect of remote mTBI
Esopenko et al., 2017 [48]	Former pro hockey players recruited through NHL Alumni Association (n = 33) with age matched controls from the community (n = 18)	A blow to the head followed by clinical symptoms	Performance on neuropsych tests, cognitive tests, and APOE4 status	Alumni tested significantly worse on executive function tests but not cognitive tests. No association between APOE4 status and cognitive test performance. Two alumni were excluded due to low MOCA score.	A	Decreased neuropsych performance
Hart et al., 2013 [44]	Retired NFL players with mTBI history (n = 32), and without (n = 2) recruited from NFL Players Association meeting and healthy, matched controls (n = 85)	AAN Practice Parameter Guidelines for grading concussion (1997).	Performance on neuropsych tests, neurological assessment, neuroimaging, dementia or MCI diagnosis	No correlation between neuropsychiatric measured and concussions. Total white matter lesion volumes significantly different in mTBI with cognitive impairment group. Lower incidence of dementia in retired players, higher incidence of MCI.	A	Increased risk of MCI, white matter reduction
Montenigro et al., 2017 [49]	n=93 former football players, High school n = 17, College n = 76, Participants were subset of ongoing longitudinal study (LEGEND)	CDC definition of concussion	Performance on cognitive neuropsych tests	Cumulative Head Impact Index predicts risk of cognitive impairment later in life	A	Increased risk of later life cognitive impairment
Ozen et al., 2015 [51]	Elderly patients with remote TBI history; mTBI n = 5, Control n = 15, recruited from Waterloo Research in Aging Pool	American Congress of Rehabilitation Medicine definition (1993)	Performance on neuropsych tests, self-report scales, open-ended interviews	No difference between mTBi, mod/severe TBI on any measure, but no statistics were performed to assess effects of TBI severity on performance	U	No effect was detected
Strain et al., 2015 [43]	Former NFL athletes with no memory impairment (n = 20), NFL athletes with MCI (n = 8), Cognitively healthy controls (n = 21), controls with MCI (n = 6) recruited from NFL Players Association.	AAN Practice Parameter Guidelines for grading concussion (1997)	Performance on neuropsych tests, hippocampal volume assessment	Former NFL athletes without cognitive impairment performed worse than healthy controls. Athletes with concussion had lower hippocampal volumes at all age ranges.	A	Decreased neuropsych. performance and hippocampal volume
Tremblay et al., 2013 [46]	mTBI group (n = 15), control group (n = 15), all participants were university athletes between age of 51 and 75 recruited with the help of university athletics organization	2009 Consensus Statement on Concussion in Sports	Performance on neuropsych tests, APOEɛ4 genotyping, neuroimaging	Reduced performance on most neuropsych tests and diminished cortical thickness and enlarged ventricle size in mTBI group, which are considered a marker for MCI development	HQ	Decreased neuropsych performance, and imaging changes analogous to MCI
Tremblay et al. 2019 [45]	Remote mTBI (n = 15) patients recruited from university athletics associations with matched controls (n = 15) recruited from community, recent mTBI (n = 19) recruited from ER with matched controls (n = 25).	2009 Consensus Statement on Concussion in Sports	MMSE and neuroimaging	Total brain volume did not differ between mTBI to control, but diffusivity was increased in remote mTBI group only in the corpus callosum and frontal white matter; no difference on MMSE	A	White matter reduction
Wammes et al., 2017 [50]	Young healthy (n = 22) and young mTBI (n = 17) recruited from university, older healthy (n = 22) and older mTBI (n = 21) recruited from Waterloo Research in Aging Pool	Any closed head injury that resulted from the head being hit, the head striking an object, or any acceleration/deceleration force that resulted in LOC (1 min to 6 h) or PTA (1 min to 7 days)	Performance on neuropsych tests and Autobiographical interview	Remote mTBI had little effect on neuropsych testing, semantic priming, or autobiographical interview content	A	Little to no effect of remote mTBI
Yang et al., 2015 [47]	mTBI+MCI/D (n = 6), mTBI without MCI/D (n = 21), healthy controls (n = 10) all recruited from the Taiwan National Health Insurance database	CDC definition of concussion	Performance on cognitive tests, APOEɛ4 genotyping, Neuroimaging	No statistics performed on cognitive tests, but loss of consciousness was identified as a risk factor for dementia. Allele frequency of APOE4 was significantly higher in mTBI+D group than controls; Amyloid accumulation was higher in mTBI+D but no difference between mTBI - D and HC group	A	Increased dementia risk with LOC, no changes in amyloid levels

NHL, National Hockey League, NFL, National Football League; AAN, American Academy of Neurology; MCI, mild cognitive impairment; LEGEND, Longitudinal Examination to Gather Evidence of Neurodegenerative Disease; CDC, Center for Disease Control and Prevention; TBI, traumatic brain injury; MMSE, Mini-Mental Status Examination; LOC, loss of consciousness; PTA, post-traumatic amnesia; ADHD, attention deficit hyperactivity disorder.

The following four studies all used MRI imaging to assess brain volume changes in addition to neuropsychiatric testing. Strain et al. (2015) assessed 28 retired NFL athletes in Texas with an mTBI history by using tests of verbal episodic memory and neuroimaging [43]. They found that verbal episodic memory was reduced in athletes compared to controls. Furthermore, hippocampal volume was significantly diminished compared to healthy controls at all age groups in athletes with at least 1 mTBI, but not in athletes without an mTBI history; this difference appeared more pronounced at increasing ages. In another study examining professional athletes, Hart et al. (2013) [44] assessed retired NFL athletes who had an mTBI history using neurocognitive assessments and neuroimaging. They found significant group differences in tests of naming, word-finding and visual and verbal episodic memory in cognitively impaired NFL players compared to controls, but no correlations between the number of mTBIs and years played in the NFL [44]. Neuroimaging studies using diffusion tensor imaging revealed increased diffusivity in regions of white matter in impaired NFL players, particularly in the frontal, parietal, and temporal regions and corpus callosum, and changes in blood flow between impaired players and healthy controls. These findings were similar to those of Tremblay et al. (2019) [45], who examined mTBI patients with different ages at mTBI onset using diffusion-weighted imaging. Remote mTBI patients had increased white matter abnormalities in the frontal region and anterior corpus callosum compared to controls, whereas recent mTBI patients had no significant differences in these parameters compared to controls [45]. Finally, Tremblay et al. (2013) [46] recruited former university-level athletes with and without an mTBI history and evaluated them using neuropsychiatric testing and MRI of anomalies in neurometabolites and structure. The mTBI group had reduced episodic and semantic memory, abnormal neurometabolites in the medial temporal region and prefrontal cortex, increased ventricle size, and reduced cortical thickness relative to controls [46].

One study primarily examined amyloid accumulation using positron emission tomography. Yang et al. (2015) [47] identified remote mTBI patients and controls and found that cognitive performance and mean amyloid accumulation scores were not significantly different between normal mTBI and control groups, but were different in mTBI patients with dementia. Four studies used neuropsychiatric testing as an outcome measure, but not neuroimaging or risk of dementia [47]. Esopenko et al. (2017) [48] assessed National Hockey League (NHL) alumni with a concussion history compared to healthy controls, and found that the NHL group performed significantly lower on tests of executive and intellectual function. They also found that executive and intellectual function was associated with the number of concussions within the NHL group. However, the NHL group performed similarly to controls on objective tests of cognition, and APOE ɛ4 status was not associated with neuropsychological impairment in this study [48]. Montenigro et al. (2017) [49] assessed repetitive head impacts using a cumulative head impact index (CHII) in 93 former amateur football players and found that the risk of mood, behavioral and cognitive impairment increased with CHII score in a dose-dependent manner. Those with the highest CHII exposure were 9-times more likely to develop objective cognitive impairment, as assessed by the Brief Test of Adult Cognition by Telephone, later in life [49]. Wammes et al. (2017) [50] assessed young and old participants with and without remote mTBI to assess the effect of remote mTBI on lingering memory issues. They found significant effects of mTBI on word recall in elderly remote mTBI patients, but most neuropsychological testing revealed no significant differences compared to age-matched controls [50]. Finally, Ozen et al. (2015) [51] assessed older adults with and without a TBI history, but only five adults in the TBI group had a mild injury. Though neuropsychological test performance appeared reduced in the mTBI group compared to non-TBI controls, the study included only a small number of participants with mTBI, and no statistical analysis was performed on the mTBI data [51].

DISCUSSION

This review and meta-analyses of studies published between January 2000 and March 2020 revealed an increase in the risk of dementia following a history of diagnosed mTBI, with a pooled odds ratio of 1.96 (95% CI 1.698–2.263, p < 0.001) and subtle evidence of persistent changes in neuropsychiatric test performance and imaging. Previous reviews on this topic have yielded mixed conclusions; Godbolt et al. (2014) [26] reported that mTBI does not increase the risk dementia development, while Perry et al. (2016) [2] reported that mTBI does increase risk of dementia. While we believe that this review provides a compelling argument that a history of diagnosed mTBI increases the risk of dementia, we also acknowledge that a risk factor is simply an increased chance of developing a disease, rather than a definitive cause. This review has been split into two main components: identifying the risk of dementia after mTBI and identifying dementia-related factors after mTBI, such as structural and cognitive abnormalities. This differentiation was made to create a comprehensive review that provides both quantitative and qualitative evidence about the risk of dementia after mTBI. It is important to identify risk factors for dementia, as they serve as a starting place for clinicians and researchers to identify interventions for at-risk individuals.

Objective risk of dementia after mTBI

We included twelve studies that aimed to identify the risk of dementia years after mTBI [31 –42]. Of the twelve studies included, eight found an increased risk of dementia after mTBI in at least one subset of the examined population [31–35 , 38–42], and two studies found that a history of mTBI affects the onset of dementia [36, 37]. Overall it appears that prior mTBI increases the risk of developing later dementia. This is in contrast to the review conducted by Godbolt et al. (2014) [26]; however, the current review was able to capture information from eleven studies deemed “High Quality” or “Acceptable”, while Godbolt et al. (2014) only captured one [26]. These discrepancies may stem from differences between inclusion criteria for the respective review; for example, in the current review, included studies must have used physician-diagnosed concussions, have had the head injury confirmed by an informant, and/or read a strictly defined definition of concussion to participants to assist in their assessment of injuries. We made this a criterion to exclude studies where participants who are potentially at risk of dementia and may have a history of mTBI are asked to remember mTBI events throughout their life. Due to the memory problems associated with mTBI and dementia, this raises the potential of study bias; this was not an inclusion criterion in the other reviews. In their review, Godbolt et al. (2014) [26] recommended that future researches ensure they produce studies with appropriate methodology and enough power to draw conclusions from. Eight studies included in this review, all of which were deemed to have appropriate power and low risk of bias, have been published since the Godbolt et al. (2014) [26] review, and so we applaud the researchers for acknowledging the paucity of research in the area and designing studies to address these gaps.

It is worth noting that a majority of the studies in this section searched health databases to investigate any associations between mTBI and dementia. While this is an excellent method to ensure accurate diagnoses, it opens a discussion regarding universal health coverage. Data collected in this manner may be most generalizable when coming from countries that provide universal health coverage, and comparisons between countries with different types of healthcare systems should be taken with caution [52]. Notably, studies based in Taiwan used data from the National Health Insurance program which provides universal health insurance and enrolls up to 99% of the population [34 , 37]. In countries where universal health care is unavailable, people may be less likely to see a physician for a multitude of reasons, including financial barriers and limited resources, which may affect the results of database searches. Another challenge exists in retrospectively searching databases, regardless of universal healthcare, in that historically, fewer people may have seen a physician for diagnosis of mTBI due to the previous misconception of concussions and mTBI as an insignificant event. A Canadian study found that concussion rates were stable between 1994/1995–2005/2006 [53]. However, by 2014 an increase in reporting by up to 250% was observed without any indication of a plateau. This trend was observed in both sexes, young and older populations, and sport and non-sport related injuries, and likely represents increased awareness of concussion and mTBI [53].

Meta-analysis of quantitative studies

This meta-analysis supports an association be-tween remote mTBI and a later diagnosis of de-mentia, including AD. For this meta-analysis, only studies that explicitly reported mTBI and dementia rates were included. If this information was not provided in the paper, we contacted the authors to gain this information. In one case, we were not able to contact the authors, and so the data was not included in the present analysis. We also only included studies that were rated as high quality or acceptable to reduce bias in our results. Because we only included studies that assessed the occurrence of mTBI prior to one-year before dementia diagnosis, we feel confident in the directionality of this association.

TBI is the strongest environmental risk factor for the later development of dementia [54]. It seems that the severity of impact plays a role in the risk of dementia such that more severe impacts have more associated risk [2, 39], but the role of repetitive injuries is still under investigation [2]. To compare the current finding to non-brain injury-related risk factors, vascular risk factors (e.g., diabetes, hypertension, and obesity) have ORs ranging from 2.0–2.3 [55]. Further, an OR between 1.80–9.05 has been observed between the APOE ɛ4 allele and the development of AD [56]. While these risk factors may not be directly comparable, it is noteworthy that we have identified an odds ratio similar to these classically associated risk factors.

The current meta-analysis only includes data from six studies. While it provides evidence that there is an association between remote mTBI and later development of dementia, more research is needed in this area to better determine the strength of association.

Neuropsychiatric changes after mTBI

We included four studies that used neuropsychiatric test performance as the primary outcome measure. Overall, mTBI seems to result in mildly reduced performance on several neuropsychiatric tests, though some studies did not detect an effect. Most studies that examined subjective measures of cognitive function after mTBI found persistent deficits in domains such as executive function and visual, verbal and episodic memory, and one study found a dose-dependent increase in risk for a variety of cognitive impairments after mTBI. However, three papers found little changes in neuropsychiatric testing between control and mTBI groups. Taken together, these results are consistent with other systematic reviews conducted in the field that report subtle changes in cognition in some patients following mTBI, especially after repeated injuries [57, 58]. This is well illustrated by a systematic review of eleven meta-analyses, which reported that despite an overall good prognosis for the vast majority of mTBI patients, a subset of patients suffered from persistent cognitive deficits [58]. Furthermore, patients with repeated head injury were at a larger risk for reduced cognitive performance [59]. Despite these findings, it is important to emphasize that some reviews report essentially complete recovery after mTBI and that the majority of patients with an mTBI recover [60]. Overall though, the large body of evidence suggests that a small subset of mTBI patients, especially those with repeated injuries, have persistent, subtle cognitive impairment. This impairment could be associated with dementia through persistent damage and a reduction in cognitive reserve [61], through association with a long-term neurodegenerative process such as tau deposition or amyloid aggregation, or could be an independent deficit unrelated to dementia.

Neuroimaging changes after mTBI

We included five studies that used neuroimaging changes as the primary outcome measure. Neuro-imaging mainly revealed changes in white matter volume, particularly in the medial temporal and frontal regions. The vulnerability of these regions to white matter abnormalities after TBI has been reported elsewhere, and they are considered susceptible to impact due to their mobility and proximity to bony structures [62].

All the neuroimaging studies included in this review found at least some significant differences in patients with a history of mTBI, most of whom were professional athletes. These included alterations in white matter tracts and brain volume in areas associated with memory, such as the hippocampus. One study found similar changes in white matter tracts in patients with post-concussive syndrome [63], and similar trends were identified in a recent systematic review [64]. While residual, long-term structural changes similar to those found in post-concussive syndrome are concerning enough, some evidence exists that white matter abnormalities seen in mTBI have similarities with those seen in AD [65]. These abnormalities include changes in the corpus callosum and temporal regions (such as the parahippocampal gyrus), both of which were affected by mTBI [66, 67].

In summary, the neuroimaging and neuropsychiatric studies presented provide potential explanations for the increased risk of dementia seen in mTBI patients. Firstly, it is possible that following mTBI, persistent injury to the hippocampus, white matter tracts and other regions causes affected individuals to be more vulnerable to the age-related brain changes associated with dementia that affect similar regions. Alternatively, mTBI could trigger a long-term neurodegenerative process that becomes clinically apparent only with age, analogous to the tau-related changes seen in patients with CTE [68]. These changes may be associated with altered levels of neurometabolites, which has already been demonstrated in symptomatic CTE patients retired from the NFL [69].

Risk of bias and limitations

The intent of rating the bias within studies is to better identify the validity of results concerning the current review, and is not done to discredit any research conducted in this area. Out of 21 studies, six were rated as high quality, twelve as acceptable, and three as unacceptable using the Scottish Intercollegiate Guidelines [20]. High quality studies had to meet most items on the bias scoring checklist, have a high enough sample size, and have no major methodological issues. Acceptable studies also met most of the guidelines but had some methodological issues. The most common issues involved a combination of the following: low statistical power, nonstandard definitions of mTBI, poor controls, selective reporting of data, and conducting neuropsychiatric tests without a trained professional. A 2007 study by Sundström et al. [33] was rated as unacceptable because of their definition of mTBI. Although the authors used structured interviews and medical records to examine head impact history, their justification of mild severity was inappropriate. The authors stated that participants had only sustained mild injuries, and this statement was based on an assumption that any participant who had sustained a TBI and was able to carry out a test battery had a mild injury. Given that most patients who survive moderate to severe TBI can function without assistance [70], we felt that this definition of mTBI was unable to accurately distinguish TBI severity. Due to the high probability of participants in the mTBI group having sustained a more severe injury, we were unable to draw strong conclusions from this paper. Although the studies by Ozen et al. (2015) [51] and McMillan et al. (2014) [38] were well done in terms of methodology, for our purposes, they were considered unacceptable, as no statistical analyses were performed on the specific results of interest (i.e., the association between remote mTBI and later development of dementia or related indicators). Because no statistical analyses were performed, results due to chance cannot be ruled out, and firm conclusions cannot be drawn.

One limitation of our review is that we did not analyze the effect of the number of mTBIs; some of the differences between studies are likely mediated by different exposure levels. Most of the studies included in this review did not include information on the number of injuries, so we were not able to examine this important variable. While evidence suggests that a single TBI may cause amyloid and tau pathology, the effect of a single mTBI is unclear [71]. Conversely, repetitive injuries have been studied as a risk factor for neurodegenerative disorders such as CTE [72], and pathological changes such as alterations in cerebral blood flow [73]. Presumably, patients with more mTBI exposure are more susceptible to developing dementia, and those with fewer injuries may have a much lower risk. Another potential limitation is that we only included studies that used a clinical diagnostic tool such as the MMSE/MoCA as a screening measure or outcome, and that a clinical definition of mTBI had to be met. Although these criteria were intended to help capture only relevant, high quality studies, it is possible that they led us to exclude relevant papers. Furthermore, we were unable to include one study in the present meta-analysis because we were unable to acquire the relevant numbers after contacting the authors. The exclusion of this study may have resulted in a slight overestimate of the risk of dementia using our pooled OR. Finally, we did not measure the risk of bias across studies, so our conclusions may have been affected by factors such as publication bias.

Future directions

This review highlights a putative association between dementia and mTBI, and discusses long-term neuroimaging and neuropsychiatric changes after mTBI. Future epidemiological studies should compare the effect of multiple versus single mTBIs to determine what level of mTBI exposure causes a significant increase in dementia risk. To better understand the mechanism of this risk, further longitudinal studies investigating tau and amyloid accumulation in mTBI patients should be conducted. In addition, this review provides a starting point for clinicians and researchers to develop prevention strategies for individuals at risk of developing dementia due to a history of mTBI. From this review, it is apparent that people with a history of mTBI show alterations in tracts, reduced brain volume in areas associated with memory, and slight decreases in cognitive performance, all of which may be risk factors for later development of dementia. While there is currently no cure for dementia, much research has focused on preventative mechanisms including exercise-based, cognitive training, and mindfulness techniques. Erickson et al. (2011) demonstrated that aerobic exercise training leads to increased hippocampal volume and associated improvements in spatial memory in healthy older adults [74]. Furthermore, Musteata et al. (2019) demonstrated improvements in cognitive performance in healthy older adults, while Spaner et al. (2019) demonstrated cognitive improvements in adults with subjective memory complaints following cognitive training programs [75, 76]. Another intervention under investigation is mindfulness training, a type of meditation. Hölzel et al. (2011) found that 8-weeks of mindfulness training increased grey matter volume in region specific areas, including the left hippocampus [77]. Pharmacological treatments, such as cholesterol management via statins, are a further area of interest in reducing dementia risk post-concussion. One study included in this review also included a patient group prescribed statins post-injury and found that patients who took statins following mTBI had a 13% decreased risk of dementia development [41]. Future researchers should investigate if these types of intervention programs have benefits to individuals at risk of dementia due to remote mTBI.

Conclusions

Remote mild traumatic brain injury confers an increased risk of dementia, with a pooled odds ratio of 1.96 (95% CI 1.698–2.263). In addition, neuropsychiatric and neuroimaging tests reveal long-term deficits and changes in brain structure after mTBI, that are associated with MCI and dementia. Future research should continue to examine mechanistic links between mTBI and dementia, and aim to identify risk management strategies in mTBI patients susceptible to dementia.

DISCLOSURE STATEMENT

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/20-0662r2).

Footnotes

The supplementary material is available in the electronic version of this article: .

References

Kristman

, Borg

, Godbolt

, Salmi

, Cancelliere

, Carroll

, Holm

, Nygren-De Boussard

, Hartvigsen

, Abara

, Donovan

, Cassidy

(2014) Methodological issues and research recommendations for prognosis after mild traumatic brain injury: Results of the international collaboration on mild traumatic brain injury prognosis. Arch Phys Med Rehabil 95 (3 Suppl), S265–277.

Perry

, Sturm

, Peterson

, Pieper

, Bullock

, Boeve

, Miller

, Guskiewicz

, Berger

, Kramer

, Welsh-Bohmer

(2016) Association of traumatic brain injury with subsequent neurological and psychiatric disease: A meta-analysis. J Neurosurg 124, 511–526.

Mckee

, Cantu

, Nowinski

, Hedley-Whyte

, Gavett

, Budson

, Santini

, Lee

H-S

, Kubilus

, Stern

(2009) Chronic traumatic encephalography in athletes: Progressive tauopathy after repetitive head injury. J Neuropathol Exp Neurol 68, 709–735.

Gardner

, Yaffe

(2015) Epidemiology of mild traumatic brain injury and neurodegenerative disease. Mol Cell Neurosci 66, 75–80.

Harada

, Natelson Love

, Triebel

(2013) Normal cognitive aging. Clin Geriatr Med 29, 737–752.

Petersen

(2009) Early diagnosis of Alzheimer’s disease: Is MCI too late? Curr Alzheimer Res 6, 324–330.

Knopman

, Petersen

(2014) Mild cognitive impairment and mild dementia: A clinical perspective. Mayo Clin Proc 89, 1452–1459.

Mitchell

, Shiri-Feshki

(2009) Rate of progression of mild cognitive impairment to dementia - Meta-analysis of 41 robust inception cohort studies. Acta Psychiatr Scand 119, 252–265.

Drummond

, Birtwhistle

, Williamson

, Khan

, Garies

, Molnar

(2016) Prevalence and management of dementia in primary care practices with electronic medical records: A report from the Canadian Primary Care Sentinel Surveillance Network. CMAJ Open 4, E177–E184.

10.

Manuel

, Garner

, Finès

, Bancej

, Flanagan

, Tu

, Reimer

, Chambers

, Bernier

(2016) Alzheimer’s and other dementias in Canada, 2011 to 2031: A microsimulation Population Health Modeling (POHEM) study of projected prevalence, health burden, health services, and caregiving use. Popul Health Metr 14, 37.

11.

Crous-Bou

, Minguillón

, Gramunt

, Molinuevo

(2017) Alzheimer’s disease prevention: From risk factors to early intervention. Alzheimers Res Ther 9, 1–9.

12.

Alzheimer Society of Canada, https://alzheimer.ca/en/Home.

13.

Prince

, Bryce

, Ferri

(2011) World Alzheimer Report 2011. The benefits of early diagnosis and intervention. Alzheimer’s Disease International, London. 90

14.

Gyebnár

, Szabó

, Sirály

, Fodor

, Sákovics

, Salacz

, Hidasi

, Csibri

, Rudas

, Kozák

, Csukly

(2018) What can DTI tell about early cognitive impairment? - Differentiation between MCI subtypes and healthy controls by diffusion tensor imaging. Psychiatry Res Neuroimaging 272, 46–57.

15.

Sheinerman

, Tsivinsky

, Crawford

, Mullan

, Abdullah

, Umansky

(2012) Plasma microRNA biomarkers for detection of mild cognitive impairment. Aging (Albany. NY) 4, 590–605.

16.

Weller

, Budson

(2018) Current understanding of Alzheimer’s disease diagnosis and treatment. F1000Res 7, F1000 Faculty Rev-1161.

17.

Weimer

, Sager

(2009) Early identification and treatment of Alzheimer’s disease: Social and fiscal outcomes. Alzheimers Dement 5, 215–226.

18.

Barnes

, Kaup

, Kirby

, Byers

, Diaz-Arrastia

, Yaffe

(2014) Traumatic brain injury and risk of dementia in older veterans. Neurology 83, 312–319.

19.

Gardner

, Rubenstein

, Wang

KKW

, Korley

, Yue

, Yuh

, Mukherje

, Valadka

, Okonkwo

, Diaz-Arrastia

, Manley

(2018) Age-related differences in diagnostic accuracy of plasma glial fibrillary acidic protein and tau for identifying acute intracranial trauma on computed tomography: A TRACK-TBI study. J Neurotrauma 35, 2341–2350.

20.

Wang

, Lin

, Sung

, Wu

, Hung

, Wang

, Huang

, Lu

, Chen

, Tsai

(2012) Population based study on patients with traumatic brain injury suggests increased risk of dementia. J Neurol Neurosurg Psychiatry 83, 1080–1085.

21.

Huang

, Lin

, Lee

, Huang

, Tai

, Wang

, Yang

, Sun

, Wang

(2018) Is traumatic brain injury a risk factor for neurodegeneration? A meta-analysis of population-based studies. BMC Neurol 18, 1–8.

22.

Hicks

, James

, Spitz

, Ponsford

(2019) Traumatic brain injury as a risk factor for dementia and Alzheimer disease: Critical review of study methodologies. J Neurotrauma 36, 3191–3219.

23.

Bigler

(2013) Traumatic brain injury, neuroimaging, and neurodegeneration. Front Hum Neurosci 7, 395.

24.

Omalu

, DeKosky

, Minster

, Kamboh

, Hamilton

, Wecht

(2005) Chronic traumatic encephalopathy in a National Football League player. Neurosurgery 57, 128–133.

25.

Carroll

, Cassidy

, Peloso

, Borg

, von Holst

, Holm

, Paniak

, Pépin

; WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury (2004) Prognosis for mild traumatic brain injury: Results of the WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury. J Rehabil Med (43 Suppl), 84-105.

26.

Godbolt

, Cancelliere

, Hincapié

, Marras

, Boyle

, Kristman

, Coronado

, Cassidy

(2014) Systematic review of the risk of dementia and chronic cognitive impairment after mild traumatic brain injury: Results of the international collaboration on mild traumatic brain injury prognosis. Arch Phys Med Rehabil 95, S245–S256.

27.

Moher

, Liberati

, Tetzlaff

, Altman

, Antes

, Atkins

, Barbour

, Barrowman

, Berlin

, Clark

, Clarke

, Cook

, D’Amico

, Deeks

, Devereaux

, Dickersin

, Egger

, Ernst

, Gøtzsche

, Grimshaw

, Guyatt

, Higgins

, Ioannidis

JPA

, Kleijnen

, Lang

, Magrini

, McNamee

, Moja

, Mulrow

, Napoli

, Oxman

, Pham

, Rennie

, Sampson

, Schulz

, Shekelle

, Tovey

, Tugwell

(2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Ann Intern Med 151, 264–269.

28.

Harbour

, Miller

(2001) A new system for grading recommendations in evidence based guidelines. Br Med J 323, 334–336.

29.

MedCalc Statistical Software version 16.4.3. MedCalc Software bv, Ostend, Belgium. https://www.medcalc.org.

30.

Wickham

(2016) ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag, New York.

31.

Barnes

, Byers

, Gardner

, Seal

, Boscardin

, Yaffe

(2018) Association of mild traumatic brain injury with and without loss of consciousness with dementia in US Military Veterans. JAMA Neurol 75, 1055–1061.

32.

Helmes

, Østbye

, Steenhuis

(2011) Incremental contribution of reported previous head injury to the prediction of diagnosis and cognitive functioning in older adults. Brain Inj 25, 338–347.

33.

Sundström

, Nilsson

L-G

, Cruts

, Adolfsson

, Van Broeckhoven

, Nyberg

(2007) Increased risk of dementia following mild head injury for carriers but not for non-carriers of the APOE epsilon4 allele. Int Psychogeriatr 19, 159–165.

34.

Yang

J-R

, Kuo

C-F

, Chung

T-T

, Liao

H-T

(2019) Increased risk of dementia in patients with craniofacial trauma: A nationwide population-based cohort study. World Neurosurg 125, e563–e574.

35.

Lee

Y-K

, Hou

S-W

, Lee

C-C

, Hsu

C-Y

, Huang

Y-S

, Su

Y-C

(2013) Increased risk of dementia in patients with mild traumatic brain injury: A nationwide cohort study. PLoS One 8, e62422.

36.

, Risacher

, McAllister

, Saykin

(2016) Traumatic brain injury and age at onset of cognitive impairment in older adults. J Neurol 263, 1280–1285.

37.

Liao

J-Y

, Lee

CT-C

, Lin

T-Y

, Liu

C-M

(2020) Exploring prior diseases associated with incident late-onset Alzheimer’s disease dementia. PLoS One 15, e0228172.

38.

McMillan

, Weir

, Wainman-Lefley

(2014) Mortality and morbidity 15 years after hospital admission with mild head injury: A prospective case-controlled population study. J Neurol Neurosurg Psychiatry 85, 1214–1220.

39.

Nordström

, Nordström

(2018) Traumatic brain injury and the risk of dementia diagnosis: A nationwide cohort study. PLoS Med 15, e1002496.

40.

Plassman

, Havlik

, Steffens

, Helms

, Newman

, Drosdick

, Phillips

, Gau

, Welsh-Bohmer

, Burke

, Guralnik

, Breitner

(2000) Documented head injury in early adulthood and risk of Alzheimer’s disease and other dementias. Neurology 55, 1158–1166.

41.

Redelmeier

, Manzoor

, Thiruchelvam

(2019) Association between statin use and risk of dementia after a concussion. JAMA Neurol 76, 887–896.

42.

Sercy

, Orlando

, Carrick

, Lieser

, Madayag

, Vasquez

, Tanner

2nd , Rubin

, Bar-Or

(2020) Long-term mortality and causes of death among patients with mild traumatic brain injury: A 5-year multicenter study. Brain Inj 34, 556–566.

43.

Strain

, Womack

, Didehbani

, Spence

, Conover

, Hart

, Kraut

, Cullum

(2015) Imaging correlates of memory and concussion history in retired National Football League athletes. JAMA Neurol 72, 773–780.

44.

Hart

, Kraut

, Womack

, Strain

, Didehbani

, Bartz

, Conover

, Mansinghani

, Lu

, Cullum

(2013) Neuroimaging of cognitive dysfunction and depression in aging retired national football league players: A cross-sectional study. JAMA Neurol 70, 326–335.

45.

Tremblay

, Desjardins

, Bermudez

, Iturria-Medina

, Evans

, Jolicoeur

, De Beaumont

(2019) Mild traumatic brain injury: The effect of age at trauma onset on brain structure integrity. Neuroimage Clin 23, 101907.

46.

Tremblay

, De Beaumont

, Henry

, Boulanger

, Evans

, Bourgouin

, Poirier

, Théoret

, Lassonde

(2013) Sports concussions and aging: A neuroimaging investigation. Cereb Cortex 23, 1159–1166.

47.

Yang

S-T

, Hsiao

I-T

, Hsieh

C-J

, Chiang

Y-H

, Yen

T-C

, Chiu

W-T

, Lin

K-J

, Hu

C-J

(2015) Accumulation of amyloid in cognitive impairment after mild traumatic brain injury. J Neurol Sci 349, 99–104.

48.

Esopenko

, Chow

, Tartaglia

, Bacopulos

, Kumar

, Binns

, Kennedy

, Müller

, Levine

(2017) Cognitive and psychosocial function in retired professional hockey players. J Neurol Neurosurg Psychiatry 88, 512–519.

49.

Montenigro

, Alosco

, Martin

, Daneshvar

, Mez

, Chaisson

, Nowinski

, Au

, McKee

, Cantu

, McClean

, Stern

, Tripodis

(2017) Cumulative head impact exposure predicts later-life depression, apathy, executive dysfunction, and cognitive impairment in former high school and college football players. J Neurotrauma 34, 328–340.

50.

Wammes

, Good

, Fernandes

(2017) Autobiographical and episodic memory deficits in mild traumatic brain injury. Brain Cogn 111, 112–126.

51.

Ozen

, Fernandes

, Clark

, Roy

(2015) Evidence of cognitive decline in older adults after remote traumatic brain injury: An exploratory study. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 22, 517–533.

52.

Ridic

, Gleason

, Ridic

(2012) Comparisons of health care systems in the United States, Germany and Canada. Mater Sociomed 24, 112–120.

53.

Gordon

, Kuhle

(2018) ‘Reported concussion’ time trends within two national health surveys over two decades. Brain Inj 32, 843–849.

54.

Smith

, Johnson

, Stewart

(2013) Chronic neuropathologies of single and repetitive TBI: Substrates of dementia? Nat Rev Neurol 9, 211–221.

55.

Kloppenborg

, van den Berg

, Kappelle

, Biessels

(2008) Diabetes and other vascular risk factors for dementia: Which factor matters most? A systematic review. Eur J Pharmacol 585, 97–108.

56.

Jun

, Naj

, Beecham

, Wang

, Buros

, Gallins

, Buxbaum

, Ertekin-Taner

, Fallin

, Friedland

, Inzelberg

, Kramer

, Rogaeva

, St George-Hyslop

, Cantwell

, Dombroski

, Saykin

, Reiman

, Bennett

, Morris

, Lunetta

, Martin

, Montine

, Goate

, Blacker

, Tsuang

, Beekly

, Cupples

, Hakonarson

, Kukull

, Foroud

, Haines

, Mayeux

, Farrer

, Pericak-Vance

, Schellenberg

(2010) Meta-analysis confirms CR1, CLU, and PICALM as Alzheimer disease risk loci and reveals interactions with APOE genotypes. Arch Neurol 67, 1473–1484.

57.

Frencham

KAR

, Fox

, Maybery

(2005) Neuropsychological studies of mild traumatic brain injury: A meta-analytic review of research since 1995. J Clin Exp Neuropsychol 3, 334–51.

58.

Karr

, Areshenkoff

, Garcia-Barrera

(2014) The neuropsychological outcomes of concussion: A systematic review of meta-analyses on the cognitive sequelae of mild traumatic brain injury. Neuropsychology 28, 321–336.

59.

Belanger

, Vanderploeg

(2005) The neuropsychological impact of sports-related concussion: A meta-analysis. J Int Neuropsychol Soc 11, 345–357.

60.

Schretlen

, Shapiro

(2003) A quantitative review of the effects of traumatic brain injury on cognitive functioning. Int Rev Psychiatry 15, 341–349.

61.

Bigler

, Stern

(2015) Traumatic brain injury and reserve. In Handbook of Clinical Neurology Elsevier B.V., pp. 691-710.

62.

McAllister

(2011) Neurobiological consequences of traumatic brain injury. Dialogues Clin Neurosci 13, 287–300.

63.

Messé

, Caplain

, Paradot

, Garrigue

, Mineo

, Soto Ares

, Ducreux

, Vignaud

, Rozec

, Desal

, Pélégrini-Issac

, Montreuil

, Benali

, Lehéricy

(2011) Diffusion tensor imaging and white matter lesions at the subacute stage in mild traumatic brain injury with persistent neurobehavioral impairment. Hum Brain Mapp 32, 999–1011.

64.

Biagianti

, Stocchetti

, Brambilla

, Van

Vleet T

(2020) Brain dysfunction underlying prolonged post-concussive syndrome: A systematic review. J Affect Disord 262, 71–76.

65.

Fakhran

, Yaeger

, Alhilali

(2013) Symptomatic white matter changes in mild traumatic brain injury resemble pathologic features of early Alzheimer dementia. Radiology 269, 249–257.

66.

Yoshiura

, Mihara

, Koga

, Ohyagi

, Noguchi

, Togao

, Ogomori

, Miyoshi

, Yamasaki

, Kaneko

, Ichimiya

, Kanba

, Honda

(2006) Mapping of subcortical white matter abnormality in Alzheimer’s disease using diffusion-weighted magnetic resonance imaging. Acad Radiol 13, 1460–1464.

67.

Van Hoesen

, Augustinack

, Dierking

, Redman

, Thangavel

(2006) The parahippocampal gyrus in Alzheimer’s disease: Clinical and preclinical neuroanatomical correlates. Ann N Y Acad Sci 911, 254–274.

68.

Katsumoto

, Takeuchi

, Tanaka

(2019) Tau pathology in chronic traumatic encephalopathy and Alzheimer’s disease: Similarities and differences. Front Neurol 10, 980.

69.

Lin

, Ramadan

, Stern

, Box

, Nowinski

, Ross

, Mountford

(2015) Changes in the neurochemistry of athletes with repetitive brain trauma: Preliminary results using localized correlated spectroscopy. Alzheimers Res Ther 7, 13.

70.

Jochems

, van Wessem

KJP

, Houwert

, Brouwers

, Dankbaar

, van Es

, Geurts

, Slooter

AJC

, Leenen

LPH

(2018) Outcome in patients with isolated moderate to severe traumatic brain injury. Crit Care Res Pract 2018, 3769418.

71.

Washington

, Villapol

, Burns

(2016) Polypathology and dementia after brain trauma: Does brain injury trigger distinct neurodegenerative diseases, or should they be classified together as traumatic encephalopathy? Exp Neurol 275 Pt 3, 381–388.

72.

Maroon

, Winkelman

, Bost

, Amos

, Mathyssek

, Miele

(2015) Chronic traumatic encephalopathy in contact sports: A systematic review of all reported pathological cases. PLoS One 10, e0117338.

73.

Clark

, Weigand

, Bangen

, Merritt

, Bondi

, Delano-Wood

(2020) Repetitive mTBI is associated with age-related reductions in cerebral blood flow but not cortical thickness. J Cereb Blood Flow Metab, doi: 10.1177/0271678X19897443

74.

Erickson

, Voss

, Prakash

, Basak

, Szabo

, Chaddock

, Kim

, Heo

, Alves

, White

, Wojcicki

, Mailey

, Vieira

, Martin

, Pence

, Woods

, McAuley

, Kramer

(2011) Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A 108, 3017–3022.

75.

Musteata

, Yoshida

, Baranzini

, Spaner

, Taneja

, Abutalebi

, Christie

, Author

(2019) Perceptual-cognitive training can improve cognition in older adults with subjective cognitive decline. Ageing Sci Ment Health Stud 3(6), 1–15.

76.

Spaner

, Musteata

, Kenny

, Gawryluk

, Hofer

, Christie

(2019) 3-Dimensional multiple object tracking training can enhance selective attention, psychomotor speed, and cognitive flexibility in healthy older adults. Ageing Sci Ment Health Stud 3(4), 1–12.

77.

Hölzel

, Carmody

, Vangel

, Congleton

, Yerramsetti

, Gard

, Lazar

(2011) Mindfulness practice leads to increases in regional brain gray matter density. Psychiatry Res 191, 36–43.

78.

Frankowski

, Annegers

, Whitman

(1985) Epidemiological and descriptive studies. Part I: The descriptive epidemiology of head trauma in the United States. In Central Nervous System Trauma Status Report, Becker DP, Povlishock JT, eds. National Institute of Neurological and Communicative Disorders and Stroke, Bethesda, pp. 33-43.