Introduction

Up to 95% of traumatic brain injuries (TBIs) are classified as being mild in severity.[1] Whilst the impact on cognitive, emotional and behavioural functioning of moderate and severe TBI have been well documented,[2] much less is known about the impact of mild-TBI, especially in the longer-term. Given 40% of mild-TBIs occur in early to mid-adulthood,[1] the societal and personal burden of mild-TBI may be higher than anticipated. For example, whilst the costs of moderate and severe injury are higher on a per person basis, as mild-TBI occurs more frequently than moderate to severe TBI, the overall costs of mild-TBI are three times higher.[3]

Whilst there is considerable variation in how TBI affects a person and their level of functioning,[4] people commonly experience a range of impairments referred to as post-concussion symptoms. These include cognitive (e.g., difficulty concentrating and taking longer to think), somatic (e.g., headaches, fatigue or dizziness) and emotional symptoms (e.g., feeling irritable, frustrated or restless).[5] Cognitive symptoms have been linked to reduced productivity at work,[6, 7] higher health service use, and engagement in antisocial behaviour.[8] Whilst many recover naturally from mild-TBI within weeks or months of the injury,[9, 10] evidence has now revealed that symptoms can persist for between 24% and 48% of those experiencing a mild-TBI one year post-injury.[9, 11–14] However, longitudinal, controlled, follow up studies within a mild-TBI cohort have not been conducted to determine whether symptoms persist in the longer-term and the impact of mild-TBI on community participation remains unclear.

One of the difficulties in studying mild-TBI is that some people do not attend hospital post-injury and seek assistance in the community from a General Practitioner or allied health professional such as a physiotherapist or choose not to seek medical attention at all. Additionally, mild-TBIs can often be overshadowed in the context of other injuries that are sustained concurrently which require more urgent medical attention. Consequently, studies that only explore mild-TBI cases that attend hospital services may not be representative of the mild-TBI spectrum. A further difficulty is that symptoms experienced following a mild-TBI can be caused by other medical conditions or illness.[14] Therefore, to determine the isolated effects of mild-TBI, it is important to compare the symptom experience to controls without history of mild-TBI. Additionally, whilst it is important to understand whether symptoms persist in the longer term, it is also vital to determine the impact that symptoms have on people’s ability to participate in everyday activities. In this study we compare the experience of post-concussion symptoms and community participation in a mild-TBI cohort, four-years post-injury, with age and gender matched controls to determine if there are long-term impacts following mild-TBI.

Materials and methods

Ethical approval was obtained from the Northern Y Health and Disability Ethics Committee of New Zealand (NTY/09/09/095) and the Auckland University of Technology Ethics Committee (09/265).

Participants

Participants who had experienced a mild-TBI were recruited through an TBI incidence and outcome study (Brain Injury Incidence and Outcomes New Zealand (NZ) in the Community, BIONIC) undertaken between 2010 and 2011.[15] The BIONIC study identified all TBI cases that occurred in the Hamilton and Waikato districts of NZ over a 12-month period. TBI was defined using the World Health Organisation criteria as an “acute brain injury resulting from mechanical energy to the head from external physical forces….the manifestations of TBI should not be due to drugs, alcohol, medications or caused by other injuries or medical conditions” (p115).[16] Participants were considered eligible if they had described an incident meeting likely to result in a TBI (as defined above) and met at least one of the following criteria; dazed or confused after the accident, loss of consciousness for 30 minutes or less, and/or not being able to remember what happened during the accident. As many people do not seek medical attention following injury, for self-reported cases details of the accident were obtained and reviewed by a diagnostic team of neurologists, clinicians and neuropsychologists to determine eligibility for the study. Severity of TBI was operationalised as a record of post-traumatic amnesia (PTA) for less than 24 hours and/or Glasgow Coma Scale (GCS) score of 13–15. Cases where there was no GCS or PTA recorded in the medical record were classified as being a mild-TBI.

One strength of identifying cases through the incidence study, was that both cases presenting to hospital as well as cases receiving medical attention through community health care services, concussion clinics, sports clubs and self-referrals were identified. Participants were also identified through searches of claims made through NZ’s national accident compensation provider. This approach aimed to prevent potentially skewing any findings of longer-term impact through only including those who sought hospital based medical treatment following the injury. Whilst the incidence study included mild-TBI cases of all ages and severities, for the purposes of this study only those cases aged >12 years at the time of injury (thereby meeting the criteria of being aged ≥16 years at the four year follow up assessment to provide informed consent) and identified as having experienced a mild-TBI were extracted. From the incidence cohort of 341 adult participants who experienced a mild-TBI and agreed to complete a further follow up assessment 12-months post-injury, 232 (68.0%) completed a follow-up assessment at fouryears post-injury (see Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Participant flow diagram.

https://doi.org/10.1371/journal.pone.0191655.g001

Controls were initially recruited through general advertising in the community (email circulations, advertisements on electronic community noticeboards and posters in public spaces). Further participants were then targeted by age and sex through the Resarchnow registry (https://www.researchnow.com) to ensure each mild-TBI case could be age and sex matched to a control participant in NZ. As our mild-TBI cohort included people who sought medical treatment in the community or did not seek medical treatment at all post-injury, a community based rather than a hospital or medically based control group was deemed to be more comparable.

Results

As shown in Table 1, TBI participants were identified from a range of sources. Of the 232 identified participants, 210 (90.5%) had sought medical attention post-injury, with 186 (88.6%) of those seeking medical attention within 24 hours. The majority of cases seeking medical attention attended hospital (70.2%), although the sample also included people seeking medical attention in the community (27.5%) as well as self-referrals who did not seek medical treatment (7.3%). There were no differences between the mild-TBI cohort and controls on sociodemographic and contextual variables, except for employment status (Table 1). There were significantly higher rates of post-injury unemployment in the mild-TBI group.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Participant characteristics of the mild-TBI and controls samples.

https://doi.org/10.1371/journal.pone.0191655.t001

Of the 210 participants who sought medical attention after injury, 159 (75.7%) were discharged home, 9 (4.2%) received an onward referral, 6 (2.9%) left medical services before being seen by a clinician due to long wait times and 36 (17.1%) were hospitalised.

Comparing the mean scores on the outcome measures between the two groups, controlling for prior TBI and levels of depression (see Table 2), the mild-TBI group experienced significantly more self-reported cognitive symptoms four-years post-injury than controls. There were also differences between the groups in terms of participation four-years post-injury, with the mild-TBI group reporting significantly poorer participation across all domains compared to the controls. Effect sizes were most notable for productivity and being out and about, although the effects were small. A post-hoc power calculation revealed that the study had 90% power to detect an effect size of 0.2 (alpha = 0.05).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Mean scores on the outcome measures between the two samples.

https://doi.org/10.1371/journal.pone.0191655.t002

As shown in Table 3, the regression models all predicted more than 40% of the variance in long-term symptoms for mild-TBI cases. Baseline symptom scores were predictive of longer term cognitive and somatic symptoms with depression, sleep quality and prior TBI revealed as independent predictors of long-term symptoms in the models. Regression models explained less variance (3–27%) in participation outcomes, with only age and education independently predicting rate of productivity, and age and depression independently linked to getting out and about four-years post-injury. Only highest education level was correlated with social relations four years post-injury, and the overall regression model was not significant.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Regression models of four year outcomes.

https://doi.org/10.1371/journal.pone.0191655.t003

Discussion

This study has revealed initial evidence to suggest that people who have sustained a mild-TBI have significantly increased self-reported cognitive symptoms (including forgetfulness, poor concentration and taking longer to think) and reduced community participation four-years post-injury compared to controls. This difference remained significant even after controlling for prior TBI and depression, although the effect sizes were small. The persistence of cognitive symptoms four years post-injury, suggests that cognitive symptoms which fail to resolve in the acute phase post-injury are likely to become chronic, and impact participation without intervention. People experiencing acute cognitive symptoms, who have a history of TBI, sought acute medical attention, experience poor sleep quality and with high levels of anxiety or depression are more at risk of persistent longer term cognitive symptoms.

This is the first study to identify that adults experience significantly greater community participation difficulties than controls four years after their injury. These findings compliment previous research which has highlighted the impact of mild-TBI on work productivity four years after mild-TBI.[6, 25] Indeed, a changing sense of self, needing to manage limitations caused by the TBI symptoms and needing support to cope with the unpredictability of the future have been important factors affecting employment following mild-TBI. The current findings highlight that the impact of mild-TBI extends beyond paid employment and also affects wider community participation. Current interventions to increase participation have revealed limited impact and new approaches to increase participation after mild-TBI are needed.[26]

One of the strengths of this study was the inclusion of participants who attended hospital post-injury in addition to people who received medical attention in the community or whom did not seek medical treatment. Mild-TBI is a heterogenous group and the finding that presentation for medical advice post-injury was associated with increased somatic and cognitive symptoms is likely to reflect the injury severity, rather than indicating that medical advice is detrimental. The findings suggest that those seeking medical attention are those at greatest risk of poorer long term outcome and highlights this interaction as an important point of contact where interventions could help to improve long term outcome.

The significant differences observed between the mild-TBI cohort four-years post-injury and controls, whilst small, were notable given that it has been previously perceived that most cases of mild-TBI resolve in the days to weeks following injury.[27] This finding extends evidence of the identified effects of mild-TBI within a year post-injury.[4, 9, 14, 28, 29] by providing evidence of persistent effects four-years post-injury. It was interesting that there was less than a 0.1 change in the total RPQ mean symptom score identified in this study compared to the one year outcomes previously reported for this sample,[9] suggesting chronicity of symptoms. The findings in our community based sample are consistent with the findings of a recent systematic review of the impact of sport-related concussion, which revealed that some former athletes are at risk of cognitive impairment for many decades post-injury.[30] Exactly how long these symptoms persist for following mild-TBI remains unknown. Only six participants within this mild-TBI sample received active treatment following their injury and consequently it is hoped that the trajectory of symptoms could be improved with intervention and is an important area of exploration for future research. Indeed evidence of the effectiveness of early intervention and information provision has been encouraging, although the quality of evidence remains low and further trials of interventions specifically for mild-TBI are needed.[31]

The recovery journey following TBI has been found to be complex and influenced by a myriad of personal, social, and environmental factors.[32] It was interesting that of the factors previously found to be predictive of symptoms one year post-injury for this sample (age, gender, living situation, comorbidities, prior TBI, ethnicity and gender)[9] only prior TBI remained a significant predictor four-years post-injury. As additional factors such as mood and sleep quality were associated with longer cognitive symptoms this suggests that psychosocial factors play more of a role in long-term cognitive symptoms as found in other studies in Europe.[33] Identifying those with longer term risk factors and use of interventions to help with sleep quality and mood may help to improve longer-term symptom presentation following mild-TBI. The finding that baseline symptoms and injury characteristics were not significantly associated with longer-term participation outcomes suggests that other personal and contextual factors not assessed in this study may be involved and require further investigation.

Given that one third of people have been found to experience at least one TBI by the age of 25 years,[34] it is challenging and unrepresentative to recruit adult controls who have been TBI free throughout their lifetime. There is evidence that there may be cumulative effects from sustaining more than one TBI.[35] To account for the potential impact of prior TBIs on the outcome assessments, for the purposes of this study controls were included if they had experienced a TBI prior to the study timeframe. There were no significant differences between the two groups for prior TBI to control for this potential confounder. However, the timing and number of prior TBIs was based on participant self-report and is likely to be influenced by recall and diagnostic bias. Longitudinal birth cohort studies where TBIs sustained over the lifetime are more accurately recorded and verified may help to control for the impact of prior TBI on longer-term outcome.

There were higher rates of unemployment in the mild-TBI group in comparison to controls which could indicate some imbalance between the groups. The lower employment rates in the mild-TBI group are consistent with previous analyses highlighting higher unemployment rates following a mild-TBI in comparison to the NZ population.[6] Previous research has highlighted how the cognitive symptoms experienced post mild-TBI can directly impact work performance.[6, 36] Our findings support and extend this finding through mirroring the impact of cognitive symptoms on broader productivity (including both paid work as well as unpaid).

Acknowledgments

We thank the research team for their dedication and performance, the staff at the Coroner’s office in Hamilton; the staff of the NZ Accident Compensation Corporation and Health Information Service; Waikato District Health Board; Waikato University Staff; the many doctors, nurses and rehabilitation professionals and service providers such as ABI Management and administrative staff within and outside Hamilton; and the BIONIC participants and their families and friends.

Bionic Research Group Members included: Valery Feigin, Alice Theadom, Suzanne Barker-Collo, Kelly Jones, Kathryn McPherson, Amy Jones, Braden Te Ao, Paul Brown, Peggy Fairbairn-Dunlop (Auckland University of Technology); Rob Kydd, P Alan Barber, Varsha Parag, Shanthi Ameratunga (University of Auckland), Nicola Starkey (University of Waikato), Tony Dowell (University of Otago), Michael Kahan, Grant Christey (Waikato Distract Health Board), Natalie Hardaker (Accident Compensation Corporation).