Cogniphobia in Mild Traumatic Brain Injury

Introduction

Fear and avoidance of movement, or “kinesophobia,” is an important perpetuating factor in chronic pain conditions. It is thought to contribute to the chronicity of pain (e.g., through muscle disuse) and mediate the relationship between pain and disability. ^{1 –3} Similarly, patients with chronic headache tend to avoid a range of headache triggers to manage their condition. ⁴ The extent of avoidance appears more related to fearful appraisal of headache pain than to pain severity. ^5,6 Although avoiding precipitating factors can be adaptive to some degree, excessive avoidance can lead to marked lifestyle changes, psychological comorbidity, as well as sensitize patients to headache triggers such that headache is elicited more readily when triggers cannot be avoided. ^4,5 “Cogniphobia” refers to fearful avoidance of a specific headache trigger, mental exertion. ^7,8 Cogniphobia may contribute to diminished cognitive effort and performance, as well as self-imposed restrictions for cognitively demanding activities, leading to increased disability. ^8,9 In one study, greater self-reported avoidance of mental exertion correlated with worse performance on a challenging serial arithmetic task in young adults with chronic headaches. ⁹

Headache is among the most common symptoms after mild traumatic brain injury (mTBI). ^10,11 One in three patients continue to experience headaches 12 months after an mTBI. ¹⁰ Cogniphobia may help explain why some patients with persistent headaches perform poorly on neuropsychological testing and in daily life long after an mTBI; ^8,12 however, empirical evidence for this link is lacking. The present study investigates cogniphobia in a treatment-seeking mTBI sample. We hypothesized that cogniphobia will be associated with 1) greater headache severity, 2) lower cognitive performance, and 3) greater behavioral avoidance of other kinds. Evidence supporting these hypotheses would add further detail to an evolving biopsychosocial conceptualization of mTBI outcome and potentially help inform the design of novel interventions.

Methods

Participants were recruited from consecutive referrals to four outpatient concussion clinics in the greater Vancouver area from March 2015 to August 2016. The eligibility criteria were: 1) being 18–65 year of age, 2) having sustained an mTBI by the World Health Organization Neurotrauma Task Force operational definition ¹³ within the past 6 months, 3) being fluent in English, 4) having been employed prior to injury, because the current study was embedded in a larger research program investigating return to work following mTBI, and 5) reporting having experienced at least one headache during the week prior to assessment. The present study received approval from the University of British Columbia Behavioural Research Ethics Board, the Vancouver Coastal Health Research Institute, and the Fraser Health Research Institute. Participants completed a battery of tests in a single session at the time of their first clinic visit that included the following measures.

Results

Of 200 consecutively screened referrals, 80 met the eligibility criteria and consented. Figure 1 displays reasons for non-enrollment in a flow diagram. The demographic and clinical characteristics of the final sample are shown in Table 1.

OPEN IN VIEWER

FIG. 1.

Flow diagram. MSVT, Medical Symptom Validity Test.

Cogniphobia-Avoidance and Cogniphobia-Dangerousness were correlated (r[60] = 0.649, p < 0.001). Neither Cogniphobia-Avoidance or Cogniphobia-Dangerousness were associated with age (r[58] = 0.004, p = 974; r[58] = −0.116, p = 0.386) or time since injury (r[60] = −0.081, p = 0.538; r[60] = −0.119, p = 0.366). Men and women reported similar levels of Cogniphobia-Avoidance (mean = 33.48, SD = 4.83 vs. mean = 35.27, SD = 5.39) and Cogniphobia-Dangerousness (mean = 18.00, SD = 3.22 vs. mean = 18.12, SD = 3.89).

The main analyses are presented in Table 2. Cogniphobia-Avoidance, and to a lesser extent, Cogniphobia-Dangerousness, correlated significantly with headache severity and other measures of avoidance (of physical movement and traumatic stress). Despite a severely restricted range on the MSVT (i.e., 61.7%, 60%, and 51.7% of sample achieved the ceiling score on the Immediate Recognition, Delayed Recognition, and Consistency subtests, respectively), there were significant correlations between Cogniphobia-Avoidance and the MSVT subtests. Cogniphobia-Avoidance correlated with only one NIHTB-CB fluid cognition test at p < 0.01, the Picture Sequence Memory test.

Low scores on the MSVT may reflect avoidance of mental effort caused by headache fear and/or unmeasured confounds, such as intentional below-capacity performance for secondary gain (malingering). To investigate whether the relationship between Cogniphobia-Avoidance and the NIHTB-CB fluid cognition tests held in participants without performance validity concerns, we reran the correlations after excluding participants who failed the MSVT (performed below established cutoffs for credible performance; n = 20). These results are also reported in Table 2. The correlation between Cogniphobia-Avoidance and Picture Sequence Memory was very modestly attenuated.

To better appreciate the clinical significance of the Picture Sequence Memory finding, we dichotomized the outcome as impaired (Picture Sequence Memory Test age-adjusted standard score <85) versus not, and calculated the risk associated with scoring above the cutoff on the Cogniphobia-Avoidance (raw score >36) that optimally discriminated impaired versus normal Picture Sequence Memory Test performance in this sample. In the full sample, the risk ratio for low memory test performance associated with high avoidance of mental exertion was 3.71 (95% confidence interval = 1.76–7.80). In the sample who passed the MSVT, the risk ratio was 3.39 (95% confidence interval = 1.16–9.91).

Migraine headaches are classically worsened by activity ^31,32 and can cause reversible cognitive dysfunction. ³³ To explore the possibility that the link among cogniphobia, headache pain, and reduced cognitive performance was restricted to migraine-type headache, we performed a supplementary analysis limited to only selected participants (n = 32) who probably did not have migraine-type headaches based on their reporting no more than mild nausea and sensitivity to noise on the BC-PSI (both item scores = 0–2; note that photobobia is not measured by the BC-PSI). As in the primary analyses, Cogniphobia-Avoidance correlated with MSVT Delayed Recognition (r[32] = −0.529, p = 0.002), MSVT Consistency (r[32] = −0.461, p = 0.007), and Picture Sequence Memory Test (r[32] = −0.540, p = 0.001] in the subgroup without migraine features. The correlation with Dimensional Change Card Sort was r(32) = −0.283, p = 0.117. Cogniphobia-Dangerousness did not significantly correlate with any NIHTB-CB fluid cognition test (p > 0.05). After removing participants who failed the MSVT from this subgroup, the correlation with the Picture Sequence Memory Test held (r[28] = −0.386, p = 0.042) and a correlation between Cogniphobia-Avoidance and the Dimensional Change Card Sort test emerged (r[24] = −0.386, p = 0.042).

Further supplementary analyses aimed to determine if the relationship between avoidance of mental exertion and performance on the Picture Sequence Memory Test would remain after adjusting for headache severity. In linear regression modeling with the sample who passed the MSVT, Cogniphobia-Avoidance significantly predicted Picture Sequence Memory Test performance when it was the only predictor (β = −0.335, t = −2.632, p = 0.011). When headache intensity during the assessment (β = −0.241, t = −1.571, p = 0.122) and headache severity (frequency × intensity) over the past week (β = 0.023, t = 0.151, p = 0.880) were added to the model as covariates, the coefficient for Cogniphobia-Avoidance remained significant (β = −0.384, t = −2.13, p = 0.043).

Finally, to appreciate the relative contribution of cogniphobia to cognitive test performance (Picture Sequence Memory Test), we ran a linear regression that included other candidate predictors of MTBI outcome. ^34,35 In this model, Cogniphobia-Avoidance (β = −0.322, t = −2.821, p = 0.006) and performance validity (MSVT) failure (β = −.384, t = −3.281, p = 0.002) independently predicted Picture Sequence Memory performance, but a history of mental health treatment, history of prior mTBI(s), presence of loss of consciousness with the index mTBI, time since index mTBI, and current post-concussion symptom severity (BC-PSI Total Score) did not significantly contribute to the prediction model.

Discussion

A number of factors other than brain injury may influence neuropsychological test performance and daily functioning in post-acute mTBI. Intentional underperformance for secondary gain is one well-established example. ³⁶ Other examples may include performance anxiety induced by creating an expectation of low performance by drawing attention to the subjects' mTBI diagnosis (“diagnosis threat”), ^37,38 or exerting extra effort by deploying top-down control over automatic skills, as when a golfer “chokes” on a high-stakes short putt. ^12,39 The present study suggests that cogniphobia may also be relevant to understanding chronic problems after mTBI. Given that fear-avoidance is common in various pain and headache disorders, ^1,9,32 it can also be expected in people with post-traumatic headaches following mTBI. Theoretically, phobic anticipation of headache pain may motivate some patients to withhold mental effort and restrict their participation in mentally taxing activities, contributing not only to reduced neuropsychological test performance but potentially to functional disability, above and beyond the headache pain itself.

Patients with more severe post-traumatic headaches reported greater avoidance of mental exertion so as to not elicit or worsen a headache, confirming that the phenomenon of cogniphobia applies to mTBI in a clinic sample. The origin of beliefs regarding the dangerousness of mental exertion after MTBI may come from early experiences with activity-related symptom exacerbations, ^{1,40 –42} inappropriate application of the “rest until asymptomatic” maxim by health professionals, ⁴³ or media misinformation. Although likely adaptive in the acute phase of recovery, avoiding mental exertion is not necessary to achieve a good clinical outcome from mTBI. ^40,44 It may be reinforced by relief from pain and/or other symptoms, and come to unduly restrict a patient's return to pre-injury activities and exacerbate their emotional distress. Also, avoiding activity and other headache triggers can over time make those triggers more potent. ⁴⁵

Our primary hypothesis was that cogniphobia would be associated with lower cognitive performance. We found that avoidance of mental exertion was associated with objectively assessed cognitive effort (performance on the “easy” subtests of the MSVT), despite a severely restricted range on these variables. It is possible that withholding effort motivated by cogniphobia accounted for reduced effort test performance, but it is also possible that other factors (e.g., compensation seeking) explained both reduced effort test performance and high scores across questionnaires, including the cogniphobia scale. Of the five NIHTB-CB fluid cognition tests, avoidance of mental exertion was significantly related to one (the Picture Sequence Memory Test), even after removing participants who failed the MSVT and adjusting for multiple testing. A supplementary analysis also found that this relationship held after controlling for headache severity (current and over the past week). Participants who reported high avoidance of mental exertion were two to four times more likely to have a low score (< 1 SD) on the Picture Sequence Memory Test than those reporting low avoidance. It is unclear why avoidance of mental exertion correlated less strongly (and nonsignificantly) with other measures of fluid cognition. The Picture Sequence Memory test requires examinees to reproduce the sequential order of 15–18 pictures. Because a high information load (intended to exceed the capacity of working memory) is presented on the first trial, this task may be perceived as especially taxing and therefore sensitive to cogniphobia. Future research demonstrating that examinees rate their perceived exertion as higher during the Picture Sequence Memory test relative to the other NIHTB-CB tests would support the theory that subjective effort mediates the effects of fear-avoidance beliefs on cognitive performance. ⁹

The present study demonstrated that avoidance of mental exertion was associated with other forms of avoidance in mTBI, including avoidance of physical activity and traumatic stress. Cogniphobia may be a specific manifestation of avoidant coping, a broader psychological construct that negatively influences behavioral engagement in daily activities. Exposure-based treatments have been shown to successfully target avoidance in other health conditions, ^1,2 and even to decrease sensitivity to headache triggers, ⁴⁶ and, therefore, should be considered for chronic mTBI. It is also shown in Table 1 that participants with high scores on the Cogniphobia-Avoidance scale were somewhat less likely to have a college degree. Low education and health literacy might also contribute to cogniphobia.

The cross-sectional design limits us from drawing conclusions about the causal nature of the relationship between cogniphobia and neuropsychological performance. Patients may perform poorly on cognitive testing because they are avoiding mental exertion and/or they may avoid mental exertion because they accurately perceive having limited cognitive capacity (or other reasons). Further research is required to better understand the clinical implications of our findings. Patients with greater cogniphobia were more likely to produce an impaired score on one test in a five test battery. This effect may not be sufficient to significantly alter treatment recommendations. The present study was also limited by not having medical records available to document the mTBI in many cases. We relied on a structured interview of the injury event, signs of altered mental status, and potential confounds of altered mental status to determine whether participants met mTBI diagnostic criteria. The optimal cutoff score for the Cogniphobia-Avoidance scale used in the relative risk analyses and for stratification in Table 1 may be specific to the present sample and might not replicate in future studies. It should be cross-validated before it is applied clinically. Fear of headache may motivate other maladaptive behaviors not measured in the present study, such as analgesic overuse ³ and restricted head movement. ³² Another limitation of the present study is that participants were not medically evaluated for specific headache disorders common after mTBI, such as migraine versus tension type. We did, however, find an effect of cogniphobia in a subgroup that probably did not have migraine-type headache based on their reporting relatively mild or absent nausea and phonophobia, suggesting that cogniphobia is not restricted to the migraine phenotype of post-traumatic headache. Finally, our findings should generalize to patients who are seeking treatment for persistent problems related to mTBI, but not to those seen in other settings.

The risk for false negative and false positive findings should be considered. Our sample may have been too small to detect true differences in cognitive domains other than memory. With effect size estimates derived from the present work, future studies on cogniphobia can prospectively calculate sample size requirements necessary to achieve adequate power. We conducted a number of supplementary (post-hoc) analyses to explore the robustness of the cogniphobia–memory performance relationship. Given the purpose of these supplementary analyses and the modest sample size, we did not further adjust for multiple testing, which could have resulted in spurious findings. These novel findings require replication.

Conclusion

In conclusion, patients who reported avoiding mental exertion because of concerns about headache performed worse on memory testing, but not on other cognitive tests. They also reported worse headaches and greater avoidance of physical activity and psychological trauma. The results partially support our hypothesis that cogniphobia contributes to cognitive difficulties in some patients with chronic symptoms after mTBI. Future research is warranted on cogniphobia and its potential role in a biopsychosocial model of poor outcome from mTBI.