Knowing What We Don’t Know: Long-Term Psychiatric Outcomes following Adult Concussion in Sports

Abstract

Objective:

Amidst a growing concern regarding concussion in sports, there is an emerging link between sport concussion and mental health outcomes. This review summarizes the current literature addressing long-term psychiatric sequelae associated with sport concussion in adults.

Method:

Several databases were searched using a broad list of keywords for each of concussion, sports, and mental health, with a resultant 311 studies for initial review. After limiting studies based on duplication, appropriateness of data, and relevance, 21 studies remained pertaining to depression, anxiety, substance use, and behavioural changes, including those highlighting chronic traumatic encephalopathy (CTE).

Results:

Most studies identified suggested an increased prevalence of depressive symptoms related to concussion history. A conference abstract and qualitative study suggested increasing anxiety related to concussion history; however, a PhD dissertation found no relationship. In reviewing substance use, several studies mentioned use in athletes suspected of having concussion histories, although no link was established, while another noted undiagnosed concussion as leading to current substance misuse. Regarding behavioural changes, all studies identified occurrences of behaviour and/or cognitive changes in participants, with 2 studies suggesting a link with concussion history. With respect to CTE, concerns with mood, behaviour, cognition, and substance use were consistently highlighted, suggesting relations to previous sport concussion; however, the notion of different CTE subtypes and clear aetiology behind concussion severity or frequency was not consistently elucidated.

Conclusion:

There appears to be a growing body of evidence supporting the presence of long-term psychiatric and psychological sequelae following sport concussion in adults.

Clinical Implications

There is scientific evidence that long-term psychiatric complications following sport concussion occur in the adult population.

It is prudent for clinicians to screen adults who have sustained a sports concussion for depression, anxiety, substance use, or behavioural changes upon discontinuation from their athletic pursuits.

Health care providers must be aware that psychiatric complications may occur chronically or later in life secondary to sport concussion.

Limitations

Given the nature of retrospective studies, recall accuracy challenges the validity of self-reported concussion frequency and severity, as well as limits the accurate assessment of causal timelines.

The evolving definition of concussion, the lack of consistently implemented screening tools for concussion, and inconsistent use of psychiatric and psychological scales hinder the ability to thoroughly compare the current literature.

A paucity of literature on this specific topic ultimately limits the ability to establish definitive conclusions or relationships at present.

Introduction

Recently, there has been growing concern regarding long-term psychiatric consequences of sport concussion in adults. In the United States, the Centers for Disease Control and Prevention estimates that 1.6 to 3.8 million concussions occur in sports and recreational activities annually. ¹ According to the Canadian Institute for Health Information, sports and recreational activities were the third leading cause of traumatic brain injury (TBI) admissions to Canadian hospitals from 2003 to 2004. ²

Over the past decade, the popular sports press has questioned the link between chronic psychiatric difficulties and previous concussion in athletes; in turn, this has driven further scientific research in the field. Despite the recent interest to explore the relationship between chronic psychiatric changes and sport concussion in adults, there is a paucity of literature on the topic. In comparison, the pediatric literature on psychiatric outcomes in head injury appears to be more robust. ³ The objective of this review was to summarize the current literature addressing chronic (>3 months) psychiatric changes, including emotional and behavioural domains (depression, anxiety, substance abuse, and challenging behaviours) with sport concussion in adults.

Methods

Several strategies were used in searching the existing literature. Initially, MEDLINE, EMBASE, and PsycINFO were searched using OVID, followed by a PubMed search. Next, the resulting references from the 3 databases were combined using RefWorks, and 1 reviewer reviewed the titles and abstracts for relevance. Thereafter, the articles were subcategorized into topics: depression, challenging behaviours, substance use, and anxiety. Furthermore, additional studies were identified for review from the citations listed in the original compilation of references.

Given that the 4 databases varied in their listed keywords, the search terms slightly differed between databases. The first set of keywords focused on concussion, including terms such as concussion, brain concussion, traumatic brain injury, or head injury. The second set of keywords focused on the athletic nature of the intended participants and included the terms sports, athletics, and sports medicine, as well as specific sports such as hockey, football, lacrosse, rugby, baseball, soccer, basketball, martial arts, boxing, contact sport, and combat sport. The last set of keywords focused on the mental health aspect of this review, including terms such as mental health, depression, anxiety, obsession, traumatic stress, behaviour/al dysregulation, emotion/al dysregulation, anger, apathy, fear, psychiatry, and neuropsychiatry.

The literature search was completed on April 7, 2014. In total, 311 English-language publications were reviewed. Of those, 152 studies were discarded as they focused on other populations such as children or did not consider concussion as a distinct outcome. Of the remaining 159 studies, these were divided into the following subcategories with number of studies in parentheses: anxiety (19), behaviour (23), cognition (41), diagnosis/prevention/prognosis/return to play (72), general information/introduction (23), mood (54), pathology/imaging (29), substance abuse (8), and those studies whose abstracts or listings were unclear (22).

In reading those studies deemed relevant to mental health outcomes, the majority were excluded from this review, as they did not relate psychiatric outcomes specifically to concussion or were focused on short-term effects. Of the studies deemed appropriate for this review, 7 were focused on depression, 3 on anxiety, 4 on substance use concerns, and 4 on behavioural changes following sport concussion. In addition, 5 publications were included that studied or reviewed chronic traumatic encephalopathy (CTE) implications that focused specifically on psychiatric changes. Postconcussion syndrome publications that did not separate physical symptoms, such as headaches and dizziness, from emotional or behavioural symptoms were not included.

Results

Depression

Reviewing studies relating specifically to depressive symptoms, several were discarded from this section for the following reasons: 6 were reviews of CTE or postconcussion syndrome, 5 had study durations too short for our review, 4 discussed neuroimaging and neuropathological findings but did not relate clearly to concussion, and the remaining studies were irrelevant for other reasons. Ultimately, we uncovered 7 studies for review, including 2 that analyzed the same group of retired National Football League (NFL) players, ^4,5 with similar assessments separated by 9 years; a retrospective study of 30 retired NFL players ⁶ ; a qualitative study involving interviews of 5 NHL players ⁷ ; a dissertation evaluating anxiety and depression in athletes many years postconcussion ⁸ ; and 2 abstracts from conferences for which full studies could not be obtained. ^9,10

In 2007, Guskiewicz et al. ⁴ investigated the association between sport concussion and clinical depression amongst retired NFL players. A general health survey and the SF-36 were distributed to over 3500 athletes to assess current health status and function of activities of daily living. Of the 2552 completed surveys returned, 60.7% reported a history of concussion, divided into those suffering 1 to 2 concussions and those suffering greater than 3 concussions. Of the total respondents, 269 reported a previous diagnosis of depression. After controlling for age, career length, years since retirement, several comorbidities (including chronic pain), and the physical component of the SF-36 scores, the authors found a linear relation between depression and concussion history, suggesting that concussion frequency increases the risk depression later in life. Those players having suffered 1 to 2 concussions were at a 1.5 times risk of developing depression, while those having experienced greater than 3 concussions were at a 3 times risk for depression, compared to those retirees with no concussion history.

Kerr et al. ⁵ revisited this same population by resending the general health survey and SF-36 nine years later. This study determined the effects of recurrent concussions on the eventual diagnosis of depression over the 9-year period. ⁵ Of the 1316 respondents, 1044 remained for analysis after excluding those who had reported a history of depression on the first survey and those who provided incomplete data. Once again, respondents were stratified based on concussion history, in groups suffering 0, 1 to 2, 3 to 4, 5 to 9, and 10+ concussions. They found that 10.2% of respondents reported a new diagnosis of depression between survey periods, with a range of 3% reported in those who had not suffered a concussion, compared to 26.8% in those reporting over 10 concussive events. Interestingly, they found that those with new diagnoses of depression over the 9-year period between surveys tended to be younger, having been retired for fewer years, and reported lower SF-36 scores in the initial study; the association between concussion and depression was again independent of these features.

In 2013, Didehbani et al. ⁶ investigated the relationship between remote sport concussions and depression in retired professional athletes by comparing Beck Depression Inventory (BDI-II) scores between study subjects and controls. Thirty retired NFL players with neuropsychological evaluations were recruited for this study. The average age for the subjects was 58.6 years, with careers averaging 9.17 years and an average history of 3.97 concussions. They were compared against 29 age- and IQ-matched controls with similar educational backgrounds and an absence of concussions. The authors found the retired athletes scored significantly higher on all aspects of the BDI-II than the control group, with the number of concussions correlating with increased BDI-II results. On further analysis, increased concussion frequency correlated with symptoms such as sadness, pessimism, guilt, self-criticalness, worthlessness, thoughts of past failure, and suicidal ideation. This study found that 40% of the athletes endorsed mild to moderate depression, a figure higher than the national average.

A qualitative study by Caron et al. ⁷ examined the experiences of 5 retired NHL players who had suffered from concussions, ultimately ending their careers. Following an interpretative phenomenological analysis for qualitative data, the authors reported that all players had some degree of anxiety, depression, and suicidal ideation; however, these psychiatric features could not definitively be linked to the subjects’ concussion histories or to the adjustment following retirement.

Two conference abstracts by Hong et al. ⁹ and Canto et al. ¹⁰ offered differing results regarding the link between concussion and depressive symptoms. Hong et al. ⁹ studied the prevalence of depression among college athletes who had suffered injury or concussion, using the Center for Epidemiological Studies Depression Scale. Although they found a higher prevalence of depression among athletes than previous studies, they did not identify a relation between depression and injury or concussion. Finally, Canto et al. ¹⁰ presented an abstract that aimed to discuss neuropsychological and cognitive changes seen in retired Filipino boxers, using the Symptoms Suggestive of Cognitive Impairment, Hospital Anxiety and Depression Scale (HADS), and the Montreal Cognitive Assessment. This study found a relationship between increasing frequency of knockouts (each knockout thought to reflect a sport concussion) and worsening neuropsychological/neuropsychiatric scores on all 3 scales. ¹⁰

Marchese ⁸ wrote a dissertation looking at the number of concussions in action-sport athletes, including but not limited to BMX bikers, rollerbladers, and skateboarders, as it related to depressive symptoms using the BDI-II. Amongst 19 participants averaging 10.63 concussions, on average 24.74 months postinjury, she found no correlation between the number of concussions and results on the BDI-II.

Anxiety

Most publications reviewed for the anxiety section were omitted for reasons described in the depression section. In addition, several studies that include anxiety, written by Stern et al., ^11,12 Yi et al., ¹³ and Gavett et al., ¹⁴ are reviewed in the section dealing with CTE. There was limited evidence of anxiety-specific concerns following concussion, with only Canto et al. ¹⁰ and Caron et al., ⁷ as well as a PhD dissertation by Marchese, ⁸ offering mentions.

Canto et al. ¹⁰ found a positive correlation between the number of knockdowns (each knockout thought to reflect a sport concussion) and the score on the HADS in retired Filipino boxers. Furthermore, they found a positive relationship between the length of time since retirement and higher HADS scores.

A qualitative study by Caron et al. ⁷ described the experience of 5 National Hockey League (NHL) players who retired due to sport concussion. In the article, 2 players described detailed experiences of anxiety that persisted many years postconcussion and postretirement. One retired NHL player reported experiencing anxiety “absolutely,” feeling as though “everything is going too fast.”

From her dissertation, Marchese ⁸ found there was no relationship between the number of concussions and anxiety (Beck Anxiety Inventory) in 19 action-sport athletes (average number of concussions 10.63 per athlete and 24.74 months postinjury).

Substance Abuse

Chronic substance abuse is another potential related concern in athletes who sustained a sport concussion, although again limited evidence is available. For this section, 4 studies relating substance abuse and sport concussion were reviewed. Caron et al., ⁷ in a quantitative study of 5 NHL athletes forced to terminate their careers prematurely due to symptoms related to concussion, noted substance use as a response to involuntary retirement, although no link was highlighted amongst study participants. A case study by Nowak et al. ¹⁵ noted moderate alcohol use in a former world boxing champion with progressive cognitive decline; however, there was no direct link made between concussions and substance use. Two other studies analyzed the link between substance abuse and sport concussion in adults. A 2010 study by Tait et al. ¹⁶ attempted to determine the incidence of marijuana and alcohol abuse in patients with a history of self-reported concussion. Of the patients surveyed, 101 people reported sustaining a concussion (40% sport-related) over a 4-year period. The study found individuals with a previous brain injury were at an increased risk of developing marijuana abuse but not alcohol abuse compared to those without a history of brain injury.

Another study by Cottler et al. ¹⁷ focused on prescription opioid use and misuse in retired NFL players, in the context of injury and pain histories. In a phone survey of 644 individuals, 52% reported opioid use during their career, with 71% reporting misuse, while 15% of these continued to use opioids in a manner deviant of its prescription. In analyzing common features amongst varying use patterns, Cottler et al. ¹⁷ found that increased pain, heavy drinking, and undiagnosed concussion were more likely linked to current misuse. Overall, 98% of misusers reported undiagnosed concussion histories, with this feature identified as the strongest predictor of misuse.

Challenging Behaviours

We reviewed challenging behaviours, with specific interest paid to aggressive behaviours, as related to sport concussion in adults. It was clear that several studies overlapped with those assessed for the depression section. Ultimately, 3 studies were identified for review. Of these, 2 studies used assessment tools looking at behavioural changes.

Seichepine et al. ¹⁸ compared behavioural changes in college and professional football players to healthy controls using the Behaviour Rating Inventory of Executive Functioning. Sixty-four players were subdivided by age, younger or older than 40 years, as well as self-reported concussion frequency. There were no significant differences between college and professional football players across the 9 domains of executive functioning. However, compared to healthy controls, the players demonstrated impairment in 7 domains, including inhibit, shift, emotional control, initiate, working memory, plan/organize, and task monitor. Older players demonstrated worse emotional control and initiate scores compared to younger players. Self-reported concussion frequency correlated with worse scores on emotional control and initiation. Younger players (<40 years of age) scored worse on inhibition.

The second study by Kaste et al. ¹⁹ examined 14 amateur and professional boxers who underwent neurological and psychological testing. Although there was no clear link established between concussion frequency and behavioural changes, 2 participants were noted to have had episodes of “embarrassing inappropriate behaviour,” as well as 12 participants taking longer on the trail-making test compared to established norms.

A qualitative study by Caron et al. ⁷ explored the experiences of 5 professional hockey players who had retired due to complications related to concussions. In addition to their mood concerns, several participants alluded to changes in their behaviour and executive functioning. One individual reported that concussions altered his demeanour, causing angry outbursts and becoming “a different person.”

Chronic Traumatic Encephalopathy

CTE is a neurodegenerative disease that occurs years following concussion or repetitive head trauma. It is only definitively diagnosed postmortem by its characteristic distribution of tau-immunoreactive neocortical neurofibrillary tangles, neuropil threads, and glial tangles. Recent literature suggests CTE may involve several characteristic clinical features. ^{11 –14,20} Cognitively, individuals tend to demonstrate difficulties with memory and executive functioning, while behaviourally, they may present as having poor impulse control and aggressive behaviours. With respect to mood, patients have been known to demonstrate concerns with emotional lability ranging from apathy to depression and anxiety, as well as reporting suicidal thoughts and self-harm behaviours. We reviewed 2 retrospective studies comparing neuropathologically confirmed CTE with next-of-kin report of the subjects’ clinical features, as well as 3 review articles. ^{11 –14,20} We evaluated the individual manuscripts that constituted the data for the 3 cited review articles and came to the same conclusions that were reached in the published reviews. Therefore, to avoid redundancy, we chose to not repeat the detailed annotation of each study. For the purpose of this review, only those sections of the studies relating to a psychiatric focus are summarized.

In 2011, Stern et al. ¹¹ wrote an article on long-term consequences of repetitive brain trauma (RBT). He discussed the clinical presentation of 50 confirmed cases of CTE in subjects who had played contact sports or were enlisted in the military and exposed to blast injuries. Next of kin were interviewed by a neuropsychologist blinded to the pathological findings of CTE, following the death of the subject. Stern et al. ¹¹ described CTE as a slow, gradually progressing phenomenon, typical presenting in midlife, several years removed from RBT, and often presenting earlier than dementias such as Alzheimer disease and frontotemporal dementia. The authors noted that early symptoms of CTE may involve cognitive, behavioural, or affective changes, with the cognitive changes comparable to those seen in other neurodegenerative diseases. The behavioural and affect changes included depression, emotional lability, impulse control, “short-fuse” outbursts, and suicidal ideation. Furthermore, Stern et al. ¹¹ also identified substance abuse and suicide completion in these subjects.

In a following study, Stern et al. ¹² retrospectively studied the clinical presentation of athletes with neuropathologically confirmed CTE by interviewing next of kin. Thirty-six of the 81 subjects from Boston’s Center for the Study of Traumatic Encephalopathy were studied after excluding those suffering head injury from nonathletic causes, without next of kin, or with comorbid neurological disease. Of the 36 subjects, all were male, with the majority having played football. Interviews with family were conducted by a neuropsychologist who was blinded to pathology. Eleven subjects were noted to have initial cognitive changes, such as memory and executive function impairments, prior to or without developing mood and behaviour symptoms. Thirteen subjects were found to have behaviour concerns, such as explosivity, impulsivity, and violence, prior to the appearance of mood and cognitive symptoms. Nine of the remaining 12 subjects experienced depression and hopelessness as their index concern, and 3 were asymptomatic. Interestingly, patients presenting initially with mood and behaviour disturbances typically experienced their presentation at a younger age, with similar age of death and similar neuropathological changes. Overall, Stern et al. ¹² suggested that CTE comprises 2 differing presentations, one of earlier mood and behaviour concerns, which evolves to include cognitive concerns, and the other form initially presenting with cognitive concerns and proportionally fewer individuals later developing mood/behaviour issues.

Three recent review articles ^13,14,20 discussed the phenotypic presentation of athletes diagnosed postmortem with CTE. All came to similar conclusions: a history of multiple sport concussions in athletes can lead to a marked onset of impaired cognition, mood difficulties, and challenging behaviours initially presenting years following retirement from sport. Symptoms such as memory and executive function impairments, depression, apathy, impulsivity, emotional lability, aggression, substance abuse, and suicidal ideation were all described as potential long-term sequelae of multiple previous concussions in athletes. Furthermore, McCrory ²⁰ noted a correlation between these changes and the number of head injuries, with identifiable risk factors for developing CTE such as beginning careers in teen years, careers spanning 10 years or longer, a less technical “slugger” style of boxing, and boxing match totals greater than 20. Gavett et al. ¹⁴ and Yi et al. ¹³ supported the idea that CTE tended to manifest in either of 2 presentations: one initially presenting with cognitive impairments, the other involving behavioural and affective disturbances at index presentation.

Conclusion

Having scoured the available literature, the paucity of relevant research into long-term psychiatric consequences following sport concussion in adults is initially surprising, especially given the increased attention paid by the media and sports organizations. However, considering that the link between concussion and mental health outcomes is a rather new area of interest, it is reasonable to acknowledge the limited opportunity to obtain representative data. The quality of studies reviewed varied widely and included abstracts, a dissertation, and a small qualitative study. Due to the diversity of literature reviewed, we initially tried to apply a grading system to assess the risk of bias and methodological quality. Using the methods described by Li and Liu, ³ Cappa et al., ²¹ and Satz et al., ²² we aimed to select studies with acceptable low-risk bias and high-quality methodology. However, had we used this approach, only 2 studies would have met the criteria; therefore, we were not able to use this more stringent methodology. That said, there appears to be a growing area of literature suggesting that long-term sequelae such as depression, anxiety, substance abuse, and challenging behaviours may occur following sport concussion, especially in athletes with repetitive head injuries in contact sports and longstanding careers. Worth noting, and always a challenge in retrospective studies, are limitations of self-report, recall accuracy, the varying definitions and criteria for concussion over an individual’s playing career, and the various scales used to diagnose depression, anxiety, and challenging behaviours.

Given the limited literature we have, more research is necessary to determine at-risk populations, develop appropriate diagnostic tools, prognosticate psychiatric outcomes, and develop appropriate treatments. We recommend future studies with design strengths such as a prospective longitudinal follow-up of a cohort of athletes; a control group of peers matched for age, gender, and socioeconomic status; standardized psychiatric interviews; input from a significant other; assessors blind to group; a clear definition of severity of injury; an objective documentation of actual injuries; a baseline assessment; and a neuroimaging assessment.