Abstract
Objective:
To describe headache characteristics over time in patients with traumatic brain injury (TBI).
Setting:
Patients enrolled and followed between 2011–2020.
Participants:
There were 147 patients with TBI, with 74 mild TBI (mTBI), 49 moderate (modTBI), 24 severe (sTBI) and 20 individuals without brain injury (IWBI).
Design:
Regular surveys of headache characteristics in patients with TBI were conducted. Patients were enrolled as early as 30 days post-injury and followed up to 5 years, for 419 total visits and 80 patients with multiple return visits.
Main Measures:
Surveys of headache characteristics, including headache severity measured on a 0–10 Likert scale and headache frequency quantified as headaches per month. Patients with migraine-type headaches (n=39) were identified by a clinician-administered tool. Functional outcomes were measured using the Glasgow Outcome Scale-Extended (GOS-E) and quality of life by the Satisfaction with Life Scale (SWLS) and Short Form-36 (SF-36).
Results:
At their initial visit, patients with TBI had more severe and frequent headaches than IWBIs (median 5 v. 2.5, P<0.001; median 2 v. 0.2, P<0.001), as did patients with mTBI compared with modTBI/sevTBI (all P≤0.01). Migraines were associated with lower SWLS and SF-36 scores. Migraines and young age were associated with higher headache severity and frequency across timepoints. Longitudinally, time postinjury correlated with improvement in headache severity and frequency without differences by injury severity. However, time postinjury did not correlate with improvement in headache characteristics in a patient subgroup with moderate/severe headaches.
Conclusion:
Our findings suggest that patients with mild, moderate, or severe TBI see improvement in headaches over time. However, patients should be counseled that improvement is modest and seen more in patients with milder headache symptoms. Patients with migraine headaches in particular are at risk for worse headache characteristics with greater impact on quality of life.
Keywords: Traumatic brain injury, headache, migraine, natural history, quality of life
INTRODUCTION
Traumatic brain injury (TBI) is a common neurological condition with an estimated global incidence of 27 million cases.1 Headaches are among the most common symptoms of TBI,2 and they impose a significant burden of disease.3–5 Studies have estimated that as many as 90% of patients experience new or worse headaches within a year of TBI.6,7
Post-traumatic headache (PTH) is a secondary headache disorder defined by the International Classification of Headache Disorder third edition (ICHD-3) as new headaches beginning within seven days after trauma.8 Patients with PTH are more likely to be anxious or depressed, and less likely to return to work.9,10 Tension and migraine headaches are the most common subtypes of PTH.11,12 Migraine-type headaches in particular have been linked to higher headache severity and impact on daily life.12
Increasingly, headaches following TBI have been demonstrated to impact various functional and quality of life aspects of patients’ lives. Higher headache severity and frequency has been linked to poor sleep13,14 and has been shown to impact daily activities.12 Fortunately, PTHs have been found to generally improve over time, with the majority of patients (especially mild TBI [mTBI] patients) fully recovering within a few months of injury.9,15–19 Still, a significant number of patients continue to experience chronic headaches well beyond this timeframe, regardless of TBI severity.20–22
There is a scarcity of research about the natural history of headaches in a mixed TBI population beyond one year postinjury. Here, we follow 147 patients with TBI up to five years postinjury. The objectives of this study were to: 1) describe the characteristics of headaches in patients with TBI cross-sectionally and longitudinally; 2) identify differences in headache characteristics that are associated with injury severity; and 3) identify associations between headaches and quality of life or global outcomes.
METHODS
This analysis is part of an ongoing longitudinal, natural history study of TBI. All participants provided written informed consent to participate, and the study was approved by the NIH institutional review board. Inclusion criteria required patients i) be 18 years or older and ii) diagnosed with a non-penetrating TBI. Exclusion criteria were i) pregnancy, ii) contraindication to MRI, and iii) medical or psychological instability where they could not reasonably complete the study. There were 160 patients initially enrolled. Of those, 147 patients had complete headache data collected and were included in analyses. These 147 patients were followed up to five years postinjury, with visits at 30 (±10 days), 90 (±30 days), 180 (±30 days) and at one, two, three, four, and five years (±2 months). Of these 147 patients, 100 were enrolled for longitudinal follow-up between 30 days to one year postinjury, and 47 were enrolled cross-sectionally for a single visit between 30 days to 5 years postinjury. There were 419 completed visits. The same inclusion/exclusion criteria applied to individuals without brain injury (IWBIs) with the exception that they could not have history of head injury. All patients and IWBIs were volunteers who responded to advertisements. They all received extensive medical examinations to determine their overall health and relevant comorbidities.
Using Veteran Affairs/Department of Defense guidelines for TBI classification,23 our sample consisted of 74 mild TBI (mTBI), 49 moderate (modTBI), and 24 severe (sTBI). Self-reported headache severity (0–10 Likert scale with 0 being no pain) and frequency (headaches per month) were collected at each time point. IWBIs (n=20) had data collected at one time point. Patients and controls reported headache characteristics using an in-house survey instrument completed by clinician interview. Patients were then retrospectively categorized by ICHD-3 (ICHD-3 1.1 or 1.2) criteria for migraine with or without aura. Headache diagnoses prior to injury, medical diagnoses including history of depression and anxiety, and current medications were collected from patients by self-report and medical record review at their initial visit. Overall outcomes were assessed using the Glasgow Outcome Scale Extended (GOS-E), a clinician-rated scale assessing overall function after head injury on a 1–8 scale, with 1 indicating death and 8 upper good recovery.24,25 Patient quality of life was measured using the Satisfaction with Life Scale (SWLS), a measure that assess global life satisfaction on a scale from 5–35, with 35 indicating very high life satisfaction.26,27 Health related quality of life was measured using the mental and physical health component summary scores of the Short Form-36 (SF-36).28,29 These scores are transformed to a standardized scale from 0–100, with 50 and 10 as the US population mean and standard deviation, respectively.28
Statistical Analysis
The Wilcoxon rank sum test by and Student’s t test were used for comparisons between non-normally and normally distributed continuous variables, respectively. Chi-Square test was used for categorical variables. P-values were corrected for multiple comparisons by the Benjamini-Hochberg procedure. Moderate/severe headache was defined as pain intensity of 4 or greater in agreement with previous literature.30 Linear mixed-effect models with random intercepts for each patient were generated to model repeat (longitudinal) measures of headache severity and frequency. These models were covaried for age, sex, migraine status, TBI severity, and time since injury. Headache frequency was log transformed to meet model assumptions. Linear mixed-effect models using the same covariates were developed to investigate recovery over time by each injury severity, with and without interaction terms. We investigated whether there was an interaction effect between injury severity and recovery over time by comparing individual patient’s beta estimates for mTBI, modTBI, and sTBI patients. The same analysis was conducted for interactions between headache disorder diagnosis prior to head injury and recovery over time. A sensitivity analysis to determine the impact of patient loss to follow-up was conducted using only patients who returned to at least 3 time points (n=67). Demographic differences were compared between patients who returned to multiple time points and those who did not. Separate linear mixed models were developed investigating change in GOS-E, SWLS, or SF-36 summary components. These models used random intercepts for each patient and were covaried for age, sex, TBI severity, and history of depression. One model included migraine-type headaches as a covariate and another model included headache severity and headache frequency as covariates. All statistical analyses were performed in R (v. 3.6.3, The R Foundation for Statistical Computing).
RESULTS
Demographics and clinical characteristics
Patient demographics are summarized in Table 1. Briefly, the average age of patients with TBI was 45.2 (SD 16.5), and 66% were male. The majority of patients were white (75%) and 14% were black. There were no statistically significant demographic or clinical differences between patients with TBI and IWBIs. The number of patients with TBI presenting at each time point was as follows: 30 day, (n=26); 90 day, (n=33); 180 day, (n=47); 1 year, (n=89); 2 year, (n=74); 3 year (n=53); 4 year, (n=57); 5 year, (n=40), for 419 total visits. The number of patients presenting for initial assessment and included in cross-sectional analysis at each timepoint was as follows: 30 day, (n=26); 90 day, (n=26); 180 day, (n=21); 1 year, (n=42); 2 year, (n=12), 3 year, (n=4), 4 year, (n=13), 5 year, (n=3). Of the 100 patients enrolled longitudinally, 80 (80%) had multiple visits with complete headache severity and frequency data, constituting 352 visits. All patients with multiple visits were enrolled within one year of injury. The number of patients with multiple visits presenting for initial assessment at each timepoint was: 30 day, (n=20), 90 day, (n=22), 180 day, (n=12), 1 year, (n=26). There were 30 patients with mTBI, 34 modTBI, and 16 sTBI with multiple visits. The loss to follow-up of patients enrolled longitudinally was 20 total patients (20%), of which there were 10 mTBI patients (25%), 8 modTBI patients (19%), and 2 sTBI patient (11%). The number of patients of different headache severities presenting for their initial visit were similar at each timepoint (Supplemental Digital Content, Table 1).
Table 1.
Demographics, clinical, and headache characteristics
| IWBI (n=20) | All TBI (n=147) | Sig | Mild TBI (n=74) | Moderate TBI (n=49) | Severe TBI (n=24) | |
|---|---|---|---|---|---|---|
| Age, yrs, mean (SD) | 44.6 (11.9) | 45.2 (16.5) | ns | 45.4 (15.1) | 48.9 (17.5) | 37.4 (16.3) |
|
Sex, no. (%) Male |
12 (60) | 97 (66) | ns | 44 (60) | 41 (84) | 12 (50) |
|
Race, no. (%) White |
11 (55) | 110 (75) | ns | 53 (72) | 38 (78) | 19 (79) |
| African American |
6 (30) | 20 (14) | ns | 13 (18) | 5 (10) | 2 (8) |
| Asian | 1 (5) | 3 (2) | ns | 1 (1) | 1 (2) | 1 (4) |
| Multiple Races | 2 (10) | 12 (8) | ns | 5 (7) | 5 (10) | 2 (8) |
| Native American |
0 (0) | 1 (1) | ns | 1 (1) | 0 (0) | 0 (0) |
| Education, yrs, mean (SD) | 17 (4) | 15.7 (2.8) | ns | 16.1 (3.1) | 15.5 (2.1) | 14.3 (2.4) |
| Past Military Service, no. (%) | 2 (10) | 28 (19) | ns | 20 (27) | 5 (10) | 3 (12) |
| Headache diagnosis prior to injury, no. (%) | na | 20 (14) | na | 11 (15) | 4 (8) | 5 (21) |
|
Current headache medication, no. (%) Any |
na | 67 (46) | na | 37 (50) | 22 (45) | 8 (33) |
| Acetaminophen | na | 24 (16) | na | 11 (15) | 10 (20) | 3 (13) |
| NSAID | na | 24 (16) | na | 12 (16) | 8 (16) | 4 (17) |
| Opioid | na | 5 (3) | na | 2 (3) | 2 (4) | 1 (4) |
| Triptan | na | 7 (5) | na | 6 (8) | 1 (2) | 0 |
| Sedative | na | 8 (5) | na | 4 (5) | 3 (6) | 1 (4) |
| Anti-migraine prophylactic | na | 12 (8) | na | 9 (12) | 2 (4) | 1 (4) |
| Depression, no (%) | 0 (0) | 28 (19) | ns | 15 (20) | 8 (16) | 4 (17) |
| Anxiety, no (%) | 1 (5) | 18 (12) | ns | 10 (14) | 5 (10) | 3 (13) |
| BMI, mean, (SD) | 26.2 (4.0) | 27.3 (4.2) | ns | 28.1 (4.2) | 27.4 (3.7) | 24.7 (4.1) |
| Diabetes, no. (%) | 0 (0) | 8 (5) | ns | 3 (4) | 0 (0) | 0 (0) |
| Hypertension, no. (%) | 1 (5) | 27 (18) | ns | 17 (23) | 10 (20) | 0 (0) |
| HA Severity, median (IQR) | 2.5 (0–5) | 5 (1–9) | *** | 6 (2–10) | 4 (0–8) | 4.5 (0–10.5) |
| HA Frequency/mo, median (IQR) | 0.2 (0–1) | 2 (0–17.7) | *** | 8 (0–34) | 2 (0–9.8) | 2 (0–6.9) |
| HA Duration, hrs, median (IQR) | 1 (0–7) | 2.8 (0–10.3) | ns | 4 (0–27) | 2 (0–6) | 1.3 (0–6) |
| Migraine criteria, n (%) | 3 (15) | 39 (27) | ns | 30 (41) | 4 (8) | 5 (21) |
| Moderate/severe headache, n (%) | 9 (45) | 88 (60) | ns | 58 (78) | 18 (37) | 12 (50) |
| SWLS, median (IQR) | 30.5 (38–33) | 24 (12–36) | *** | 24 (11–37) | 26 (16–36) | 22 (10.5–33.5) |
| SF-36 physical, median (IQR) | 59.7 (55–64.4) | 46.7 (33–60.4) | *** | 45.5 (30.7–60.3) | 48(33.3–62.6) | 46.7 (36.2–57.2) |
| SF-36 mental, median (IQR) | 56.9 (53.3–60.4) | 48.1 (32.6–63.5) | *** | 45.5 (27.3–63.7) | 49.5 (35.2–63.8) | 50.2 (38.8–61.6) |
| GOS-E, median (IQR) | na | 6 (4–8) | na | 6 (4–8) | 6 (4–8) | 5 (3–7) |
Abbreviations: IWBI = individual without brain injury; TBI = traumatic brain injury; NSAID = nonsteroidal anti-inflammatory drug; BMI = body mass index; SWLS= Satisfaction with Life Scale; SF-36=36-item short form survey; GOS-E= Glasgow Outcome Scale-Extended
P<0.05,
P<0.01,
P<0.001
There were 20 patients (14%) with diagnosed headache conditions prior to injury, 121 without prior headache diagnoses (82%), and six (4%) for whom prior headache diagnoses could not be determined. Of those patients with prior headache diagnoses, 16 had migraines, one had tension-type headaches, one had cluster headaches, and two had unspecified headache disorders. There were 67 patients (46%) taking medications for headaches at their initial visit after their head injury, and 53 of those had no headache diagnosis prior to injury. A similar proportion took medications for headaches across TBI severity groups: 37 mTBI (50%), 22 modTBI (45%), and 8 sTBI (33%). Headache medications commonly used can be seen in Table 1.
Demographics and clinical characteristics were largely similar between patients with TBI in the longitudinally or cross-sectionally enrolled groups (Supplemental Digital Content, Table 2). Longitudinally enrolled patients were more likely to be older (mean 45.3 v 39.9 years, P<0.01) and less likely to have experienced a mTBI (40% v. 72%, P<0.001). Demographics were also similar between patients with TBI who presented to multiple time points compared to those who presented to only one visit. Patients who presented to multiple time points were also older (P<0.001) and less likely to have experienced an mTBI (P<0.01) (Supplemental Digital Content, Table 3).
Cross-sectional analysis of headache characteristics and quality of life
Initial visits occurring median 0.8, IQR 0.2–1.3 years postinjury were used for cross-sectional analysis. There were 118 patients with TBI (80%) with headaches compared to 11/20 IWBIs (55%, P=0.025). There was no difference in the prevalence of headaches between male and female patients (90% v. 82%, P=0.333). Patients with TBI had more severe headaches compared to IWBIs (Table 1, median 5, IQR 1–9 vs. 2.5, IQR 0–5, P<0.001) and they had higher frequency of headaches (median 2/month, IQR 0–17.7 vs. 0.2/month, IQR 0–1; P<0.001). Patients with diagnosed headache conditions prior to injury had more severe headaches (median 7.5, IQR 4.5–10.5; median 5, IQR 0–10, respectively, P=0.003) but not more frequent compared to patients without prior headache diagnoses (median 8, IQR 0–19; median 2, IQR 0–18, respectively, P=0.225).
Patients with mTBI had more severe and frequent headaches compared to modTBI or sTBI patients (Table 1, P=0.002, P=0.005; and P=0.011, P=0.038, respectively). More patients with mTBI had headaches than sTBI (91% v. 71%, P=0.039) but not modTBI (84%, P=0.390).
Patients with TBI had lower scores on the SWLS, SF-36 physical component summary, and SF-36 mental component summary compared to IWBIs (Table 1, all P<0.001). Patients with TBI with migraine-type headaches had significantly lower scores on the SWLS scale and SF-36 mental component compared to those without migraine-type headaches (Supplemental Digital Content Table 4). Patients who returned for one visit had no differences in their headache characteristics, functional outcomes, or quality-of-life outcomes at their initial visit compared to those who returned for multiple visits (Supplemental Digital Content Table 3).
Longitudinal changes in headaches
There was a longitudinal decrease in headache severity among patients with TBI (β=−0.203; P=0.003, Table 2). Older age was associated with lower headache severity across all timepoints (β=−0.047; P=0.003), while presence of migraine-type headaches was associated with higher severity (β=2.284; P=<0.001). A history of moderate or severe TBI was associated with lower headache severity (β=−1.390 and −2.003, P=0.001 and <0.001, respectively). There was no association between headache severity and gender. Similar associations were found with headache frequency: a longitudinal decrease in frequency (β=−0.155; P<0.001, Table 2), an increase in those with migraine-type headaches (β=0.630; P=0.003), and decreases in older individuals (β=−0.023.; P=0.008) as well as mod/sTBI patients (β=−0.950; P=0.003; β=−1.045; P=0.010). These findings were robust to sensitivity analysis using only patients who returned to 3 or more visits (Supplemental Digital Content Table 5). There was no significant difference in recovery over time across injury severities (Figure 1). There was also no interaction of headache diagnosis prior to injury on improvement over time in headache severity or frequency (Supplemental Digital Content Figure 1). Mixed-effect models for each TBI injury severity group could not be developed due to violations of modeling assumptions.
Table 2.
Longitudinal changes in headache severity and frequency in patients with TBI
| Predictors | Headache Severity | Headache Frequency* | ||
|---|---|---|---|---|
| β | p-value | β | p-value | |
| Migraine | 2.284 | 1.68e-12 | 0.630 | 0.003 |
| Time since injury, year | −0.203 | 0.003 | −0.155 | 7.01e-4 |
| Moderate TBI | −1.390 | 0.001 | −0.950 | 0.003 |
| Severe TBI | −2.003 | 2.29e-4 | −1.045 | 0.010 |
| Age | −0.047 | 5.87e-4 | −0.023 | 0.008 |
| Sex, male | −0.655 | 0.106 | −0.079 | 0.795 |
Abbreviations: TBI=traumatic brain injury
β and p-values from linear mixed-effect models with random intercept for each subject.
Log-transformed dependent variable.
Figure 1. Association between brain injury severity and change in severity/frequency of post-traumatic headache.
(A) β-coefficients of change over time in headache severity vs. injury severity and (B) β-coefficients of change over time in headache frequency vs. injury severity. Beta coefficients were extracted for each patient from the mixed-effect analysis. Headache frequency values were log transformed before model generation.
Subgroup analysis: cross-sectional headache characteristics in patients with migraine-type and moderate/severe headache
At their cross-sectional visit 39 (27%) patients with TBI had migraine-type headaches compared to three (15%) IWBIs. Of those 39 patients, 30 (77%) did not have a migraine diagnosis prior to injury. There were 88 (60%) patients with TBI who had moderate/severe headaches compared to nine (45%) IWBIs (P=0.31). The moderate/severe headache group included all but one TBI patient with migraine-type headaches. The median headache severity and frequency for the migraine-type group and the moderate/severe headache group was severity 8 (IQR 4.5–11.5), 6 (IQR 3–9), P<0.001, and frequency 8 (IQR 0–34), 8 (IQR 0–34), P=0.164, respectively, higher than the general cohort of patients with TBI (all P<0.001).
Patients with mTBI were more likely to have migraine-type headaches or moderate/severe headaches compared to mod/sTBI (39% vs 14%, P<0.001; 80% v. 55%, P=0.002). The percentage with migraine-type headaches in the overall cohort varied little over time, remaining between 21% and 19% from year 1 to year 5 while the percentage with moderate/severe headaches decreased non-significantly from 60% to 48% (P=0.279, Figure 2).
Figure 2: Proportion of patients with TBI with migraines or moderate/severe headaches over time.
A) Proportion of patients with TBI with migraine-type headaches at each visit interval B) Proportion of patients with TBI with moderate/severe headaches at each time interval. The error bar corresponds to 95% CI.
The results of the linear mixed-effect models for the moderate/severe headache subgroup are shown in Table 3. Headache severity score was higher in those with migraine-type headaches and lower in mod/sTBI, and headache frequency was lower in those who sustained a modTBI. However, there was not a significant association between headache severity or frequency and time since injury for the moderate/severe headache subgroup.
Table 3.
Longitudinal changes in headache severity and frequency in patients with TBI with moderate/severe headaches
| Predictors | Headache Severity | Headache Frequency* | ||
|---|---|---|---|---|
| β | p-value | β | p-value | |
| Migraine | 0.724 | 0.005 | 0.136 | 0.545 |
| Time since injury, year | −0.022 | 0.754 | −0.082 | 0.173 |
| Moderate TBI | −0.824 | 0.017 | −1.05 | 0.002 |
| Severe TBI | −1.183 | 0.006 | −0.206 | .613 |
| Age | −0.005 | 0.581 | 0.001 | 0.954 |
| Male Sex | −.234 | 0.440 | 0.097 | 0.741 |
Abbreviations: TBI=traumatic brain injury
β and p-values from linear mixed-effect models with random intercept for each subject.
Log-transformed dependent variable.
Association between headache characteristics and longitudinal changes in global outcomes
Mixed-effect models could be generated for the GOS-E scale but not the SWLS scale or the SF-36 scale due to violations of modeling assumptions. Time since injury was associated with improved GOS-E score in both mixed-effect models (β=0.196 and 0.180; P<0.001; Supplemental Digital Content Table 6, Supplemental Digital Content Table 7). Headache frequency was inversely associated with GOS-E (β=−0.020; P<0.001) (Supplemental Digital Content Table 7). Severe TBI was associated with lower GOS-E (β=−0.776 and −0.939; P=0.010 and P=0.002). Headache severity was negatively correlated with GOS-E, although the relationship did not reach significance (β=−0.035; P=0.066). Migraine status was not a significant predictor of GOS-E (Supplemental Digital Content Table 6).
DISCUSSION
The key findings of this investigation were: 1) patients with TBI had more frequent and severe headaches than IWBIs, with patients with mTBI having the most frequent and severe headaches compared to mod/sTBI; 2) various factors were associated with increased headache severity and frequency in this TBI cohort, including migraine-type headaches, young age, mTBI, and recent injury; 3) in the overall TBI cohort headache severity and frequency improved over time regardless of TBI injury severity or history of headaches prior to injury, while in a smaller subgroup of patients with more severe headaches there was no improvement in headache characteristics over time; and 4) more frequent headaches were associated with worse functional outcome while migraine-type headaches were associated with worse quality of life.
It is widely known that patients with TBI have a higher headache burden than healthy controls.3,31 However, the literature is divided on whether mTBI, modTBI, or sTBI patients have the most severe headaches after injury. Some studies of veterans have found that mod/sTBI patients have higher symptom burden,19,32 while studies in the civilian population have generally found that the burden is similar across injury severities2,6,16,20 and some studies have also found that symptoms are greater in mTBI patients.33,34 Here, we found that patients with mTBI have both more severe and more frequent headaches by about a factor of two (e.g., in mTBI compared to modTBI the median severity is six v. three). The discrepancy between this finding and previous literature may be explained by the longer time course of this study and by our unique patient population. Most previous studies followed patients for one year postinjury, during which time many patients with mTBI fully recover.15,17,22 Since we followed patients with mTBI up to five years after injury, and many of our patients enrolled months postinjury and complained of headaches, our finding may best be interpreted that those patients with mTBI who continue to have headaches past one year of injury have relatively severe headaches. The lower headache severity and frequency in mod/sTBI might also be partially explained by a greater likelihood to receive treatment for headaches, although we found similar rates of headache medication treatment between each severity group. Interestingly, despite the higher headache severity and frequency among mTBI patients, we did not find a difference in the rate of recovery by injury severity. This lack of difference suggests that headache recovery after head injury is relatively consistent regardless of injury type, although it should be interpreted with caution due to the small sample sizes in each injury severity group.
Previous literature agrees with our finding that migraine-type headaches and young age are associated with headache severity and frequency in TBI.9,35 Our finding of the high percentage of patients with mTBI with migraine-type headaches (41%) also agrees with a previous study that found a prevalence of 49%.7 We were unable to further subclassify headache types phenotypes because the clinician administered form in this study was designed only to classify headaches as migraine-type per ICHD-3 criteria. We did not find an association between gender and more prevalent, severe, or frequent headaches, which conflicts with previous literature.6,9,20,36 This may be explained by the broad definition of headache used in this study, which included anyone who experienced headaches in the previous year. It is noteworthy that there was no interaction between headache condition prior to injury and recovery over time, suggesting that patients experience recovery regardless of prior headache history. This study’s major contribution is the long-time course over which we followed patients, and in particular the finding that headache severity and frequency may continue to improve up to five years post TBI. However, the improvement over time is modest. For example, from year one to year five, headache severity is predicted to decrease by about one point on a 0–10 scale and headache frequency by about 2 per month.
As opposed to the overall TBI cohort, we found that patients with more severe headaches did not experience recovery over time. This analysis was in part constrained by a small sample size, but the low absolute value of the model’s coefficients (−0.02 change in severity per year) suggests that lack of power does not fully explain the finding. The lack of improvement in these patients with more severe headaches may be due to a unique neurodegenerative or psychological process within the subgroup. The nature of such a process is unknown and warrants further study. While we did not collect data on headache treatment, it is important to emphasize that early education on and screening for headache following brain injury, as well as early multimodal treatment approaches, are necessary to prevent and improve impacts of headaches on function and quality of life after TBI.37,38
Other studies have found that headaches are associated with decreased quality of life and disability.3,39,40 Our findings that frequent headaches are associated with worse functional outcome and that migraine-type headaches are associated with worse quality of life agrees with this prior work.41 Unfortunately, we were unable to generate more advanced models of these relationships due to violations of modeling assumptions. Further study is needed.
This study had limitations. First, the study cohort was a volunteer sample of patients who enrolled at least 30 days after injury. This may have biased the selection of patients towards those with more severe symptom burden. Second, 20 patients (20%) were lost to follow-up, which is an inherent issue in longitudinal studies. These patients lost to follow up were more likely to have mild TBIs (50%, compared to 38% of patients who returned for multiple visits). This may have biased the longitudinal assessment of recovery rate in mild TBIs, particularly if patients experiencing better recovery were less likely to return for subsequent visits. Third, there was variability in the time points for longitudinal follow-up; however, the mixed -effect model helped account for this variability. Fourth, pre-injury history was collected from medical records and patient self-report rather than a standardized survey, so it is possible that previous headache disorders were undercounted in the analysis. Lastly, as this study was designed to assess the subacute and chronic effects of TBI over several years, we did not have headache data within seven days of injury, so we were unable to classify headaches as posttraumatic as defined by ICHD-3.
CONCLUSION
TBI patients had more frequent and severe headaches than IWBIs. Migraine-type headaches, younger age, and mild TBI were associated with more severe and frequent headaches. Migraine-type headaches were also particularly associated with lower quality of life. Patients experienced improvement over time in headache severity and frequency regardless of their TBI severity; however, in a subgroup of patients with moderate/severe headaches time since injury did not have a significant association with improvement in headaches. Confirmatory longitudinal studies with larger sample size and less variability in follow-up are needed, as well as further investigation into the mechanism sustaining these potentially more persistent, severe headaches.
Supplementary Material
ACKNOWLEDGEMENTS
We thank the patients and volunteers who participated in this study. We would also like to thank Larry Tang, PhD, for providing independent review of the statistical methods for this paper.
CONFLICTS OF INTEREST AND SOURCE OF FUNDING
Footnotes
TRANSPARENCY AND OPENNESS
The deidentified data and code is available by request of a qualified researcher
Contributor Information
Spencer Brent Flynn, National Institutes of Health Clinical Center.
Brian Moore, Center for Neuroscience and Regenerative Medicine (CNRM), Bethesda, Maryland. The Henry M. Jackson Foundation for the Advancement of Military Medicine (HJF), Bethesda, Maryland, USA.
Andre van der Merwe, Center for Neuroscience and Regenerative Medicine (CNRM), Bethesda, Maryland. The Henry M. Jackson Foundation for the Advancement of Military Medicine (HJF), Bethesda, Maryland, USA.
Anita Moses, Center for Neuroscience and Regenerative Medicine (CNRM), Bethesda, Maryland. The Henry M. Jackson Foundation for the Advancement of Military Medicine (HJF), Bethesda, Maryland, USA.
Jessica Lo, Rehabilitation Medicine Department, National Institutes of Health, Bethesda, Maryland.
Leighton Chan, Rehabilitation Medicine Department, National Institutes of Health, Bethesda, Maryland. Center for Neuroscience and Regenerative Medicine (CNRM), Bethesda, Maryland.
REFERENCES
- 1.James SL, Bannick MS, Montjoy-Venning WC, et al. Global, regional, and national burden of traumatic brain injury and spinal cord injury, 1990–2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2019;18(1):56–87. doi: 10.1016/S1474-4422(18)30415-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dikmen S, Machamer J, Fann JR, Temkin NR. Rates of symptom reporting following traumatic brain injury. J Int Neuropsychol Soc. 2010;16(3):401–411. doi: 10.1017/S1355617710000196 [DOI] [PubMed] [Google Scholar]
- 3.Marcus DA. Disability and chronic posttraumatic headache. Headache. 2003;43(2):117–121. doi: 10.1046/j.1526-4610.2003.03028.x [DOI] [PubMed] [Google Scholar]
- 4.Abu Bakar N, Tanprawate S, Lambru G, Torkamani M, Jahanshahi M, Matharu MS. Quality of life in primary headache disorders: A review. Cephalalgia. 2016;36(1):67–91. doi: 10.1177/0333102415580099 [DOI] [PubMed] [Google Scholar]
- 5.Lo J, Chan L, Flynn S. A Systematic Review of the Incidence, Prevalence, Costs, and Activity and Work Limitations of Amputation, Osteoarthritis, Rheumatoid Arthritis, Back Pain, Multiple Sclerosis, Spinal Cord Injury, Stroke, and Traumatic Brain Injury in the United States: A 2019 Update. Arch Phys Med Rehabil. 2021;102(1):115–131. doi: 10.1016/j.apmr.2020.04.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hoffman JM, Lucas S, Dikmen S, et al. Natural history of headache after traumatic brain injury. J Neurotrauma. 2011;28(9):1719–1725. doi: 10.1089/neu.2011.1914 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lucas S, Hoffman JM, Bell KR, Dikmen S. A prospective study of prevalence and characterization of headache following mild traumatic brain injury. Cephalalgia. 2014;34(2):93–102. doi: 10.1177/0333102413499645 [DOI] [PubMed] [Google Scholar]
- 8.Olesen J. Headache Classification Committee of the International Headache Society (IHS) The International Classification of Headache Disorders, 3rd edition. Cephalalgia. 2018;38(1):1–211. doi: 10.1177/0333102417738202 [DOI] [PubMed] [Google Scholar]
- 9.Yilmaz T, Roks G, De Koning M, et al. Risk factors and outcomes associated with post-traumatic headache after mild traumatic brain injury. Emerg Med J. 2017;34(12):800–805. doi: 10.1136/emermed-2015-205429 [DOI] [PubMed] [Google Scholar]
- 10.Minen MT, Boubour A, Walia H, Barr W. Post-Concussive Syndrome: a Focus on Post-Traumatic Headache and Related Cognitive, Psychiatric, and Sleep Issues. Curr Neurol Neurosci Rep. 2016;16(11). doi: 10.1007/s11910-016-0697-7 [DOI] [PubMed] [Google Scholar]
- 11.Lew HL, Lin PH, Fuh JL, Wang SJ, Clark DJ, Walker WC. Characteristics and treatment of headache after traumatic brain injury: A focused review. Am J Phys Med Rehabil. 2006;85(7):619–627. doi: 10.1097/01.phm.0000223235.09931.c0 [DOI] [PubMed] [Google Scholar]
- 12.Hoffman JM, Lucas S, Dikmen S, Temkin N. Clinical Perspectives on Headache After Traumatic Brain Injury. PM R. 2020;12(10). doi: 10.1002/pmrj.12338 [DOI] [PubMed] [Google Scholar]
- 13.Shekleton JA, Parcell DL, Redman JR, Phipps-Nelson J, Ponsford JL, Rajaratnam SMW. Sleep disturbance and melatonin levels following traumatic brain injury. Neurology. 2010;74(21):1732–1738. doi: 10.1212/WNL.0b013e3181e0438b [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Castriotta R, Wilde M, Lai J, Atanasov S, Masel B, Kuna S. Prevalence and consequences of sleep disorders in traumatic brain injury - PubMed. J Clin Sleep Med. 2007;15(3):349–356. https://pubmed.ncbi.nlm.nih.gov/17694722/. Accessed October 22, 2020. [PMC free article] [PubMed] [Google Scholar]
- 15.Levin HS, Mattis S, Ruff RM, et al. Neurobehavioral outcome following minor head injury: A three-center study. J Neurosurg. 1987;66(2):234–243. doi: 10.3171/jns.1987.66.2.0234 [DOI] [PubMed] [Google Scholar]
- 16.Rohling ML, Binder LM, Demakis GJ, Larrabee GJ, Ploetz DM, Langhinrichsen-Rohling J. A meta-analysis of neuropsychological outcome after mild traumatic brain injury: Re-analyses and reconsiderations of Binder et al. (1997), Frencham et al. (2005), and Pertab et al. (2009). Clin Neuropsychol. 2011;25(4):608–623. doi: 10.1080/13854046.2011.565076 [DOI] [PubMed] [Google Scholar]
- 17.Nordhaug LH, Linde M, Follestad T, et al. Change in Headache Suffering and Predictors of Headache after Mild Traumatic Brain Injury: A Population-Based, Controlled, Longitudinal Study with Twelve-Month Follow-Up. J Neurotrauma. 2019;36(23):3244–3252. doi: 10.1089/neu.2018.6328 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Walker WC, Seel RT, Curtiss G, Warden DL. Headache after Moderate and Severe Traumatic Brain Injury: A Longitudinal Analysis. Arch Phys Med Rehabil. 2005;86(9):1793–1800. doi: 10.1016/j.apmr.2004.12.042 [DOI] [PubMed] [Google Scholar]
- 19.Suri P, Stolzmann K, Iverson KM, et al. Associations Between Traumatic Brain Injury History and Future Headache Severity in Veterans: A Longitudinal Study. Arch Phys Med Rehabil. 2017;98(11):2118–2125.e1. doi: 10.1016/j.apmr.2017.04.008 [DOI] [PubMed] [Google Scholar]
- 20.Walker WC, Marwitz JH, Wilk AR, et al. Prediction of headache severity (density and functional impact) after traumatic brain injury: A longitudinal multicenter study. Cephalalgia. 2013;33(12):998–1008. doi: 10.1177/0333102413482197 [DOI] [PubMed] [Google Scholar]
- 21.Voormolen DC, Haagsma JA, Polinder S, et al. Post-Concussion Symptoms in Complicated vs. Uncomplicated Mild Traumatic Brain Injury Patients at Three and Six Months Post-Injury: Results from the CENTER-TBI Study. J Clin Med. 2019;8(11):1921. doi: 10.3390/jcm8111921 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Faux S, Sheedy J. A prospective controlled study in the prevalence of posttraumatic headache following mild traumatic brain injury. Pain Med. 2008;9(8):1001–1011. doi: 10.1111/j.1526-4637.2007.00404.x [DOI] [PubMed] [Google Scholar]
- 23.Traumatic Brain Injury: Updated Definition and Reporting [press release]. Assistant Secretary of Defense. [Google Scholar]
- 24.Wilson JTL, Pettigrew LEL, Teasdale GM. Structured interviews for the glasgow outcome scale and the extended glasgow outcome scale: Guidelines for their use. J Neurotrauma. 1998;15(8):573–580. doi: 10.1089/neu.1998.15.573 [DOI] [PubMed] [Google Scholar]
- 25.Jennett B, Bond M. Assessment of Outcome after Severe Brain Damage. Lancet. 1975;305(7905):480–484. doi: 10.1016/S0140-6736(75)92830-5 [DOI] [PubMed] [Google Scholar]
- 26.Pavot W, Diener E, Colvin CR, Sandvik E. Further Validation of the Satisfaction With Life Scale; Evidence for the Cross-Method Convergence of Well-Being Measures. J Pers Assess. 1991;57(1):149–161. doi: 10.1207/s15327752jpa5701_17 [DOI] [PubMed] [Google Scholar]
- 27.Diener E, Emmons RA, Larsen RJ, Griffin S. The Satisfaction With Life Scale. J Pers Assess. 1985;49(1):71–75. doi: 10.1207/s15327752jpa4901_13 [DOI] [PubMed] [Google Scholar]
- 28.Anderson C, Laubscher S, Burns R. Validation of the Short Form 36 (SF-36) health survey questionnaire among stroke patients. Stroke. 1996;27(10):1812–1816. doi: 10.1161/01.STR.27.10.1812 [DOI] [PubMed] [Google Scholar]
- 29.Stewart A, Ware J. Measuring Functioning and Well-Being: The Medical Outcomes Study Approach. Duke University Press; 1992. doi: 10.7249/cb361 [DOI] [Google Scholar]
- 30.Hong CK, Joo JY, Shim YS, et al. The course of headache in patients with moderate-to-severe headache due to mild traumatic brain injury: a retrospective cross-sectional study. J Headache Pain. 2017;18(1). doi: 10.1186/s10194-017-0755-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rutherford WH, Merrett JD, McDonald JR. Symptoms at one year following concussion from minor head injuries. Injury. 1979;10(3):225–230. doi: 10.1016/0020-1383(79)90015-9 [DOI] [PubMed] [Google Scholar]
- 32.Beswick-Escanlar V, Lee T, Hu Z, Clark L. Increasing severity of traumatic brain injury is associated with an increased risk of subsequent headache or migraine: a retrospective cohort study of U.S. active duty service members, 2006–2015 - PubMed. MSMR. 2016;23(7):2–8. https://pubmed.ncbi.nlm.nih.gov/27501937/. Accessed October 22, 2020. [PubMed] [Google Scholar]
- 33.Haas D. Chronic Post-Traumatic Headaches Classified and Compared with Natural Headaches: 10.1046/j1468-298219961607486.x. 2016;16(7):486–493. doi: [DOI] [PubMed] [Google Scholar]
- 34.JR C CB Chronic daily headache in the posttrauma syndrome: relation to extent of head injury. Headache. 2001;41(6):559–564. doi: 10.1046/J.1526-4610.2001.041006559.X [DOI] [PubMed] [Google Scholar]
- 35.Metti A, Schwab K, Finkel A, et al. Posttraumatic vs nontraumatic headaches: A phenotypic analysis in a military population. Neurology. 2020;94(11):e1137–e1146. doi: 10.1212/WNL.0000000000008935 [DOI] [PubMed] [Google Scholar]
- 36.Dwyer B. Posttraumatic Headache. Semin Neurol. 2018;38(6):619–626. doi: 10.1055/s-0038-1673692 [DOI] [PubMed] [Google Scholar]
- 37.Schwedt TJ. Post-traumatic headache due to mild traumatic brain injury: Current knowledge and future directions: 10.1177/0333102420970188. 2020;41(4):464–471. doi: 10.1177/0333102420970188 [DOI] [PubMed] [Google Scholar]
- 38.Iaccarino MA, Bhatnagar S, Zafonte R. Rehabilitation after traumatic brain injury. Handb Clin Neurol. 2015;127:411–422. doi: 10.1016/B978-0-444-52892-6.00026-X [DOI] [PubMed] [Google Scholar]
- 39.Guitera V, Muñoz P, Castillo J, Pascual J. Quality of life in chronic daily headache: A study in a general population. Neurology. 2002;58(7):1062–1065. doi: 10.1212/WNL.58.7.1062 [DOI] [PubMed] [Google Scholar]
- 40.Abu Bakar N, Tanprawate S, Lambru G, Torkamani M, Jahanshahi M, Matharu MS. Quality of life in primary headache disorders: A review. Cephalalgia. 2016;36(1):67–91. doi: 10.1177/0333102415580099 [DOI] [PubMed] [Google Scholar]
- 41.Martins HA de L, Martins BBM, Ribas VR, et al. Life quality, depression and anxiety symptoms in chronic post-traumatic headache after mild brain injury. Dement e Neuropsychol. 2012;6(1):53–58. doi: 10.1590/S1980-57642012DN06010009 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.