Fatigue Following Moderate and Severe TBI

1.0. Fatigue following Traumatic Brain Injury (TBI) - definitions, measurement and prevalence.

Fatigue has been defined as "the awareness of a decreased capacity for physical and/or mental activity due to an imbalance in the availability, utilization and/or restoration of resources needed to perform an activity". TBI represents "an alteration in brain function, or other evidence of brain pathology, caused by an external force", and is among the most severe, disabling neurological disorders, with post-TBI fatigue (PTBIF) being one of the most common and debilitating chronic symptoms, regardless of injury severity. A challenge in studying PTBIF is related to difficulties in operationalization and assessment. Estimates of PTBIF rates thus vary from 21% to 73% due to heterogeneity of study populations, assessment strategies and study designs, e.g. time since injury, sampling of patients and fatigue measures employed. The numbers however typically by far exceed rates of fatigue in the general population, with one Norwegian study estimating 23%. Although numerous measures of fatigue exist, there is a lack of "gold standard" tools. Self-report instruments, VAS-scales and single item questionnaires are most common, and the Fatigue Severity Scale (FSS) has been documented to have good psychometric properties. There is however considerable overlap between measures of fatigue and other symptoms such as sleepiness, depression, and distress, resulting in less than optimal construct validity. Most studies of fatigue after TBI have furthermore used cross-sectional designs. Although very limited longitudinal data is available, there is some evidence to suggest that PTBIF levels may decline over the first 6 to 12 months post-injury, but remain steady or rise slightly thereafter. A recent study identified fatigue as the most common somatic complaint five years after TBI, affecting more than half of the patients. In summary, despite study variability, fatigue is a common and often chronic symptom after TBI. However, little longitudinal work has investigated individual variability in patterns of fatigue, rendering a lack of knowledge regarding sub-groups and clinical trajectories.

1.1. Contributing factors and theoretical accounts of PTBIF. Fatigue following TBI may be caused directly by the nervous system injury, i.e., primary fatigue, which is associated with deficits in cerebral networks mediating attention, arousal, and response speed (reticular activating system, limbic system, anterior cingulate, mid frontal, and basal ganglia). Primary fatigue may in turn be exacerbated by common symptoms after TBI, such as insomnia, pain, emotional distress, and reduced tolerance for physical and/or mental activity, i.e. secondary fatigue. Additional physical causes of PTBIF have also been noted, such as increased body mass index (BMI) leading to deconditioning, reduced physical activity and posttraumatic pituitary hormone deficiencies. Of notice, PTBIF does not seem to be related to the severity of injury as assessed by the Glasgow Coma Scale (GCS) or duration of post-traumatic amnesia (PTA). Ponsford et al. (2014) tested a model of the interrelationships between fatigue and daytime sleepiness, vigilance (i.e., sustained attention), anxiety, and depression after TBI. The findings indicated that fatigue after TBI is the precursor of anxiety, depression, and daytime sleepiness, and that fatigue exacerbates cognitive problems. However, this model needs to be replicated, as factors such as pre-and comorbid risk factors, premorbid fatigue, and personality were not included in the model, and the study used a cross-sectional design. The "coping hypothesis" put forward by van Zomeren and colleagues (1984) proposes that cognitive impairment, particularly reduced processing speed and attention, leads to a compensatory increase in brain activity and effort, which in turn results in fatigue. Some studies support this hypothesis, as increased brain (functional MRI) and autonomic reactivity (i.e., heart rate variability) was seen during cognitive effort in individuals with TBI compared to normal controls. Individuals with TBI have also been shown to experience greater fatigue while performing cognitive tasks requiring divided attention or other demanding cognitive processes. In their cost-benefit model, Boksem & Tops (2008) propose a theoretical framework emphasizing the role of automatic attributions of costs and benefits of mental and physical effort and rest in contributing to experienced fatigue. Mediated by dopaminergic structures, with medial frontal cortices playing a central role, it is proposed that heightened sensitivity to the short-term rewards of resting, in combination with a decreased sensitivity to the delayed rewards of long term goals, may lead to decreased intrinsic motivation to exert energy. Reward-related behavior is proposed to be associated with three factors; positive emotion towards the reward, a motivation to act in order to achieve it, and a capacity for learning stimulus-reward contingencies. An inverse relationship between self-reported reward sensitivity and fatigue has been reported in persons with Multiple Sclerosis, but no studies have explored the role of reward sensitivity PTBIF, despite the fact that both positive emotionality, motivated behavior and learning of reward contingencies are at risk of being affected by injury.

1.2. Risk factors for fatigue in the general population. In the general population, associations with fatigue have been found with demographic variables such as gender and educational level. Personality factors, such as neuroticism, have been linked to fatigue. Being high on neuroticism and perfectionism, and low on extraversion, constitute risk factors for experiencing exertion and fatigue, while extraverted people seem to be protected against high fatigue. Neuroticism might also moderate the role of catastrophizing cognitions and emotional distress on fatigue severity. Recent studies have also demonstrated that optimal levels of conscientiousness are related to a broad range of favorable health related behaviors and health outcomes, even including longevity. To our knowledge, conscientiousness, and personality in general, has not been much investigated in the context of PTBIF, nor in twin studies of fatigue. Emotional distress, anxiety and depression are commonly associated with fatigue, and may serve as both predisposing and perpetuating factors. Similarly, musculoskeletal pain, mood and sleep disturbances are associated with experienced fatigue in community dwelling adults.

In summary, there is reason to believe that individuals with TBI will be predisposed or vulnerable to develop PTBIF to variable degrees, as pre- and comorbid characteristics may act as both risk- and protective factors. Surprisingly, individual and general risk factors that may potentially contribute or even account for PTBIF, have not been extensively studied in populations with brain injury. Personality traits have been reported to be significant in mediating outcome following TBI, but the relationship between personality traits and persistent PTBIF has not previously been investigated.

2.0. Research Aims and Hypotheses. The primary aim of this study is to establish the frequency and severity of PTBIF in a representative sample of patients with moderate to severe TBI during the first year and a half post-injury, including investigation of potential clinical subgroups and trajectories of fatigue. We expect to find that fatigue will be a substantial and persisting problem for a large proportion of TBI patients. We furthermore hypothesize that there will be distinct clinical subgroups where the role of intra-individual factors such as premorbid fatigue, personality, somatic complaints, resilience, emotional distress, reward sensitivity and injury-related cognitive deficit will contribute differentially to persistent fatigue.

A secondary aim of the study is to explore the replicability of Ponsford´s (2014) model, wherein cognitive problems predicted fatigue, which in turn resulted in emotional distress. As we include a broader array of relevant individual factors, we hypothesize that the results will show multiple pathways to persistent fatigue, and expect to establish a more complex multidirectional model between the included factors.

3.0, Methodology. This longitudinal study takes on a broad clinical and theoretical perspective in exploring factors that might be predictive of persistent fatigue in a representative sample of patients with moderate-to-severe TBI over the first year and a half after injury, including comparison with a non-clinical twin-sample. A prospective longitudinal design will be applied, where a representative sample of patients with moderate and severe TBI will be assessed at 6 and 12 months post injury.

3.1. Statistical analyses and power. Analyses will be performed using the Statistical Package for the Social Sciences Version 22, with p<0.05 as level of significance. Change over time will be assessed using mixed models analysis at 3 time points: acute phase (T1), 6 months post-injury (T2), and 12 months post-injury (T3). Predictors of fatigue at 6 and 12 months post injury will be explored with linear regression analysis. The fatigue model presented by Ponsford et al. (2014) will be tested using path analysis, within a structural equation modelling framework. Based on previous reports, FSS scores ≥ 5 (on a scale ranging from 1-7) are interpreted as indicative of severe fatigue. According to the incident number of hospital-admitted adults with TBI in Eastern region of Norway, it is estimated that 130 patients with moderate-to-severe TBI, aged 18 - 65 years, will be eligible for study inclusion within one year. With an expected dropout rate of 20 - 30 %, we assume that 91-104 patients will be able to complete follow-ups. The statistical strength calculated with G*power 3.1.7 shows that a regression analysis with a correction for 10-13 predictors will yield a power of 81 % (effect size of 0.20, alpha of 0.05). One-point difference is a conservative indicator of an important change in individual FSS scores. Taking into account the total sample size, analysis of mean difference from T2 to T3 would permit a critical t of 1.98-1.99 and effect size of 0.28-0.30.

4.0. Plan for Activities, Visibility and Dissemination. Study results will be shared with the scientific community through scientific articles to be published in peer-reviewed international journals in the fields of neurology and rehabilitation, and through participation in scientific conferences. The PhD candidate will turn in his thesis containing three scientific articles addressing the study aims to the Department of psychology at the University of Oslo. Senior researchers in the group will publish additionally with him on the data set obtained in this study. Given the strong collaborative effort, the comprehensive data set, and the strong theoretical basis for this project, it has potential to result in high impact publications. Results will also be communicated through an ongoing dialogue with user organizations and central clinicians at Sunnaas Rehabilitation Hospital and Oslo University Hospital. A designated group of researchers and patient representatives within the project group will be responsible for planning of popularized presentation of results in order to communicate with society at large and to relevant patient groups. The research group will seek funding for hosting at least one large-scale scientific conference addressing mechanisms of fatigue in different clinical populations.