Association between Altered Cortisol Profiles and Neurobehavioral Impairment after Mild Traumatic Brain Injury in College Students

Abstract

Mild traumatic brain injury (mTBI) is the most common form of TBI, with more than 2.5 million TBI cases in the United States annually. Identification of easily obtainable biomarkers that track strongly with mTBI symptoms may improve our understanding of biological factors that contribute to mTBI symptom profiles and long-term outcomes. Notably, some individuals with mTBI exhibit circadian disruptions and elevated stress sensitivity, which in other clinical groups often correlate with disrupted secretion of cortisol, a glucocorticoid hormone that coordinates circadian and stress physiology. Here, we examined whether cortisol profiles could serve as a biomarker to complement the assessment of neurobehavioral sequelae after mTBI. We partnered with our on-campus health clinic to recruit college students seeking medical care after mTBI (n = 46) and compared this population to a well-matched non-injured student control group (n = 44). We collected data at an initial visit (shortly after injury in mTBI subjects) and one week later. At each visit, we evaluated neurobehavioral function using the Automated Neuropsychological Assessment Metric (ANAM). The subjects also provided cortisol samples through at-home saliva collection. We observed strong coherence between ANAM subjective and objective measures, indicating significant multi-dimensional impairment in subjects with mTBI. Further, female mTBI subjects exhibited diminished neurobehavioral function compared with males. Regardless of sex, decreased amplitude of diurnal cortisol and a blunted cortisol awakening response were associated with mTBI symptom severity and neurobehavioral impairment. Taken together, these findings suggest that salivary cortisol profiles may be a sensitive biomarker for studying underlying biological factors that impact mTBI symptoms and outcomes.

Introduction

Each year, more than 2.8 million Americans and 10 million individuals worldwide sustain traumatic brain injury (TBI).^1,2 The most common form is mild TBI (mTBI), representing approximately 80% of total TBI cases.³ Symptoms include physical impairments (e.g., headaches, hypersensitivity to sound and light) and disruptions in behavior, cognition, and emotion.⁴ Some symptoms are especially pronounced just after injury, but many tend to resolve within several months.⁵ Nevertheless, 30% of patients report persistent symptoms six months after injury, and 10-15% report symptoms after one year.^6,7

A major challenge for understanding the risk factors that contribute to long-term symptom persistence is the wide-ranging symptomology that varies between individuals during the first few weeks after injury. The development of assessment tools that can sensitively track individual neurobehavioral change after mTBI offers value as a means to study individual symptom progression and outcome.

Computerized neurocognitive tests are a good complement to standard clinical assessments.⁸ Representing one of these tests, the Automated Neuropsychological Assessment Metric (ANAM) is administered quickly (15–30 min) and yields extensive information regarding general health, emotional status, and cognitive capacity in a variety of populations, including those with mTBI.^9

–12

Physiological measures associated with mTBI may serve as additional complementary biomarkers providing novel mechanistic insights into mTBI symptom profiles and recovery.^13

–16 Cortisol is a hormone controlled by the neuroendocrine hypothalamic-pituitary-adrenal (HPA) axis. It can be conveniently measured in saliva, offering powerful research utility as a measure of an individual's ongoing physiological state without the confounding influence of the laboratory setting. Cortisol secretion exhibits a prominent circadian rhythm and a dynamic phasic response to stress.^17,18

In humans, normal cortisol diurnal profiles are characterized by very low trough levels around bedtime and a sharp rise that begins in the early morning and reaches a peak at the typical time of awakening. An additional transient increase occurs 30–60 min after awakening, and this increase is called the cortisol awakening response (CAR).^19,20 Cortisol levels then steadily decline until again reaching trough levels the following night.

Diurnal cortisol is controlled by the circadian system and provides a sensitive measure of systemwide circadian function.^21,22 The neural basis and functional roles of the CAR are less well established, but may be related to stress reactivity in addition to circadian control.^23
–25 Some physiological and psychological disorders involving stress sensitivity and circadian disruption are associated with decreased diurnal cortisol and blunted CAR.^26

–29 Notably, mTBI is associated with circadian dysregulation and increased stress sensitivity.^30,31 Therefore, we hypothesized that mTBI may be associated with disruptions in cortisol daily profiles.

Here, we investigate whether altered cortisol profiles are associated with neurobehavioral impairments after recent mTBI. We provide a detailed characterization of ANAM measures and corresponding salivary cortisol, including examination of sex differences, in college students with mTBI compared with non-injured controls. Although the ANAM has been widely used in study and evaluation of concussion symptoms in military personnel, there are few reports of studies that have characterized ANAM measures in other groups of young adults who experienced a recent mTBI. We administered two assessments, one week apart, finding that blunted CAR and decreased diurnal cortisol are significant predictors of neurobehavioral impairment after mTBI.

Methods

Subjects

All participants in the mTBI group were University of Colorado Boulder students seeking healthcare (2017–2019) at the on-campus Wardenburg Health Center in response to a recent brain injury. Wardenburg medical practitioners specializing in concussion assessment provided mTBI diagnosis based on a multi-factor assessment.³²

After mTBI diagnosis, physicians informed potential subjects about the ongoing study by handing out flyers with a brief study description and contact information for the research team. Control subjects were recruited through an introductory psychology course at the University of Colorado Boulder (Spring 2018).

Study design

All participants provided informed consent before participation. The University of Colorado Institutional Review Board approved all procedures (IRB#17-0069). Subjects in the mTBI group were tested in a quiet room at either Wardenburg or the Clinical Assessment of Injury, Recovery & Resilience (CAIRR) Neuroscience Laboratory. Control subjects were tested at the CAIRR Neuroscience Laboratory.

The study included an Initial Visit and 1-Wk-Follow-Up, each consisting of an in-person and at-home component. In-person sessions lasted ∼30 min, during which time subjects were given saliva collection tubes and instructions for at-home saliva self-sampling. At-home saliva sampling consisted of a bedtime and next morning awakening sample collection to be initiated the night of the initial visit and 1-Wk Follow-Up. Subjects with mTBI were compensated $20 for each completed session. Control subjects were compensated with class participation credits.

ANAM Neurobehavior

The ANAM Core Battery computerized neurocognitive assessment was used to collect demographic information, self-ratings of symptoms, mood, and sleep, and objective measures of various aspects of cognitive performance (see Supplementary Text).^12,33 For the eight tests of cognitive performance, simple reaction times are reported as mean reaction times (msec); code substitution, procedural reaction time, matching-to-sample, and mathematical processing performance are reported as throughput score (number of correct responses per minute of available response time); and Go/No-Go performance is reported as D-prime, the ability in standard deviation units to discriminate targets from distracters.

Salivary cortisol

Subjects were instructed to collect saliva samples at home by spitting through a short plastic straw into six pre-labeled 1.5 mL microcentrifuge tubes. Saliva samples were provided at bedtime, immediately on waking, and 15, 30, 45, and 60 min after waking. Subjects kept a log in which they recorded their bedtime, time of awakening, and the time of collection for each saliva sample. Subjects were instructed to refrain from caffeine, food intake, and strenuous exercise during the collection period. They were also instructed to store saliva samples in their home freezer until the samples were returned to the research team. Returned samples were stored in the laboratory at -20°C.

Salivary cortisol concentrations were quantified using the Salimetrics Cortisol Enzyme Immunoassay Kit (Product#1-3002-5). Each sample was run in duplicate. All samples from one participant were run on one multi-titer plate to minimize the influence of between plate variability. Similarly, mTBI and control samples were counterbalanced across plates. The mean within-plate coefficient of variation was 7.6% for high-concentration standards and 12.1% for low-concentration standards, and the between-plate coefficient of variation was 10.4% for high-concentration standards and 14.1% for low-concentration standards.

We characterized cortisol profiles using two parameters. “Diurnal Amplitude” represents the difference between the average of the five morning cortisol samples and the bedtime sample (average morning cortisol−bedtime cortisol), illustrating the underlying circadian rhythm of cortisol secretion. “CAR Magnitude” (CARi) represents the area under the curve of the five morning cortisol samples in reference to the first waking sample (cumulative CAR area−waking cortisol), illustrating the extent of an increase in cortisol after waking.²⁷

Statistical analyses

Analyses were performed using R Studio (Version 1.1.423) and SPSS (IBM, Version 25.0.0.0). Group differences in demographic information collected at the Initial Visit were assessed through either two-tailed t-test comparisons of means or Fisher exact test of group differences in frequency. The ANAM and cortisol measures were all assessed separately via two-way analysis of variance (ANOVA), with injury condition and sex as separate factors. Mixed model ANOVAs were used to evaluate salivary cortisol profiles, with injury condition and sex as between-group factors and time after awakening as a within-group factor. Follow-up within-group repeated-measure ANOVAs were used to assess the effect of time for the five morning cortisol samples.

Pearson correlation coefficients with nominal probability levels were used to assess correlations between dependent measures. Linear multiple regression analyses were used to assess whether cortisol parameters (Diurnal Amplitude or CAR Magnitude) predicted ANAM neurobehavioral measures. For multiple regression analyses, a four-factor linear model included the cortisol parameter, injury condition, sex, and injury-condition-by-cortisol-parameter interaction. For cortisol profiles, there were some missing data because of inadequate saliva volume or implausible outlier values.³⁴ Missing data usually involved a single sample and were estimated using the mean value of the temporally flanking samples. Graphs are presented as means ± standard error of the mean (SEM). A significance level of alpha <0.05 was used for inferential statistics.

Results

Demographic and Injury Details

Control and mTBI subjects were well matched by age (young adults), sex (balanced male-female proportion), and ethnicity (mostly white) (Table 1). Subjects with mTBI reported significantly more previous concussions (78.3%) than control subjects (22.7%), consistent with previous observations of college students in this setting.³² Most subjects with mTBI received a diagnosis and began participation within one week after injury (mean = 6.5 days, median = 4.5 days) (Table 2). The majority of injuries resulted from falls or sports-related accidents (67% of cases).

Table 1.

Demographic Information for Mild Traumatic Brain Injury and Control Subjects

Demographics	mTBI	Control	Statistical comparison of group differences
Number of subjects	46	44
Mean age (SEM)	20.5 (0.39) years	20.6 (9.071) years	p = 0.90 (t test)
Female	52.2% (24/46)	40.9% (18/44)	p = 0.30 (Fisher exact test)
Ethnicity (non-white)	15.2% (7/46)	22.7% (10/44)	p = 0.43 (Fisher exact test)
Prior history of TBI	78.3% (36/46)	22.7% (10/44)	p < 0.0001 (Fisher exact test)
Subject drop-out after Initial Visit	28.3% (13/46)	6.8% (3/44)	p = 0.01 (Fisher exact test)

mTBI, mild traumatic brain injury; SEM, standard error of the mean.

Bold type indicates significant effect.

Table 2.

Mild Traumatic Brain Injury Details

mTBI Details	Males	Females	Overall
Time between injury and assessment	Median = 5.0 days, Mean (SEM) = 7.86 (1.99) days	Median = 3.5 days, Mean (SEM) = 5.29 (0.78) days	Median = 4.5 days, Mean (SEM) = 6.5 (1.02) days
Cause of injury	Fall (10)Sports (6)Vehicle accident (3) Struck by object (0) Other (3)	Fall (9)Sports (6)Vehicle accident (2)Struck by object (2) Other (5)	Fall (19)Sports (12)Vehicle accident (5) Struck by object (2) Other (8)

mTBI, mild traumatic brain injury; SEM, standard error of the mean.

ANAM Neurobehavior-related Measures

Symptoms

At the Initial Visit, subjects with mTBI endorsed a significantly greater number of symptoms (mean = 13.8, SEM = 0.71) than controls (mean = 7.2, SEM = 0.72). In addition, subjects with mTBI rated those symptoms as more severe than controls, resulting in a significantly higher mean symptom severity rating (Fig. 1, Table 3). There was a marked sex-by-injury-condition interaction for mean symptom severity ratings (Fig. 1B, Table 3). Females with mTBI rated mean symptom severity significantly higher than males with mTBI (p < 0.05, Bonferroni post hoc test). Symptom severity ratings did not differ between control males and females.

FIG. 1.

Symptom ratings for mild traumatic brain injury (mTBI) and control subjects. (A) Average ratings of mTBI and control subjects for 21 separate symptoms (5 point severity scale) during Initial Visit and 1-Wk Follow-Up. *Indicates group differences (Student t test, p < 0.05). (B) Mean symptom severity ratings of individual female and male mTBI and control subjects are superimposed on group means (bar height) and standard error of the mean (error bars). Ratings of subjects that dropped out after the Initial Visit are also denoted (see Figure Key). Italicized letters under the Initial Visit panel indicates significant results of two-way analysis of variance (a – main effect of injury condition, b – main effect of sex, c – group X sex interaction). (C) Mean symptom severity ratings of mTBI and control groups based on whether subjects were present for both session visits or only the Initial Visit. *Significant difference at Initial Visit between mTBI and control subjects that were present for both visits; ‡significant difference at Initial Visit between mTBI subjects that were present for only first visit compared with subjects with mTBI present for both visits (Tukey post hoc test, p = 0.006).

Table 3.

Two-Way Analysis of Variance Results: Neurobehavioral Differences Assessing Injury Condition Group, Sex, and Their Interaction within Initial Visit and within 1-Week Follow-Up Sessions

ANAM Task	Group	Sex	Group x Sex
Initial Visit
Mean symptom severity	F(1,86) = 38.29, p<0.001	F(1,86) = 8.95, p = 0.004	F(1,86) = 4.50, p = 0.037
Simple reaction time	F(1,86) = 24.17, p<0.001	F(1,86) = 4.74, p = 0.032	F(1,86) = 8.23, p = 0.005
Simple reaction time repeated	F(1,85) = 5.48, p = 0.022	F(1,85) = 11.75, p = 0.001	F(1,85) = 5.278, p = 0.024
Code substitution	F(1,86) = 1.90, p = 0.172	F(1,86) = 2.51, p = 0.117	F(1,86) = 2.18, p = 0.143
Code substitution delayed	F(1,85) = 1.44, p = 0.233	F(1,85) = 1.32, p = 0.253	F(1,85) = 1.40, p = 0.24
Go-No/Go	F(1,85) = 0.26, p = 0.614	F(1,85) = 3.623, p = 0.06	F(1,85) = 0.05, p = 0.817
Matching-to-sample	F(1,85) = 0.80, p = 0.374	F(1,85) = 5.27, p = 0.024	F(1,85) = 0.12, p = 0.729
Mathematical processing	F(1,86) = 2.81, p = 0.097	F(1,86) = 8.88, p = 0.004	F(1,86) = .012, p = 0.912
Procedural Reaction Time	F(1,86) = 3.92, p = 0.051	F(1,86) = 3.76, p = 0.056	F(1,88) = 0.07, p = 0.792
Restlessness	F(1,86) = 9.84, p = 0.002	F(1,86) = 0.49, p = 0.485	F(1,86) = 0.15, p = 0.70
Anger	F(1,86) = 11.42 , p = 0.001	F(1,86) = 0.99, p = 0.324	F(1,86) = 0.43, p = 0.514
Anxiety	F(1,86) = 5.52, p = 0.021	F(1,86) = 1.73, p = 0.193	F(1,86) = 0.45, p = 0.504
Depression	F(1,86) = 10.97, p = 0.001	F(1,86) = 0.24, p = 0.627	F(1,86) = 0.089, p = 0.767
Fatigue	F(1,86) = 16.55, p<0.001	F(1,86) = 1.15, p = 0.286	F(1,86) = 1.75, p = 0.19
Happiness	F(1,86) = 30.74, p<0.001	F(1,86) = 1.55, p = 0.217	F(1,86) = 0.56, p = 0.456
Vigor	F(1,86) = 20.35, p<0.001	F(1,86) = 4.35, p<0.001	F(1,86) = 0.88, p = 0.352
Sleepiness	F(1,86) = 14.29, p<0.001	F(1,86) = 3.68, p = 0.058	F(1,86) = 1.96, p = 0.165
One-Week Follow-Up
Mean symptom severity	F(1,70) = 3.47, p = 0.067	F(1,70) = .59, p = 0.445	F(1,70) = 1.64, p = 0.204
Simple reaction time	F(1,70) = 6.97, p = 0.01	F(1,70) = 2.73, p = 0.103	F(1,70) = 7.90, p = 0.006
Simple reaction time repeated	F(1,70) = 3.54, p = 0.064	F(1,70) = 6.02, p = 0.017	F(1,70) = 9.70, p = 0.003
Code substitution	F(1,70) = 0.01, p = 0.916	F(1,70) = 10.17, p = 0.002	F(1,70) = 0.00, p = 0.964
Code substitution delayed	F(1,70) = 0.54, p = 0.466	F(1,70) = 7.21, p = 0.009	F(1,70) = 0.48, p = 0.491
Go-No/Go	F(1,70) = 2.85, p = 0.096	F(1,70) = 2.43, p = 0.124	F(1,70) = 3.59, p = 0.062
Matching-to-sample	F(1,70) = 0.01, p = 0.906	F(1,70) = 9.88, p = 0.002	F(1,70) = 0.04, p = 0.835
Mathematical processing	F(1,70) = 0.04, p = 0.835	F(1,70) = 9.04, p = 0.004	F(1,70) = 0.91, p = 0.344
Procedural reaction time	F(1,70) = 0.41, p = 0.526	F(1,70) = 11.06, p = 0.001	F(1,70) = 0.15, p = 0.70
Restlessness	F(1,70) = 0.14, p = 0.712	F(1,70) = 0.20, p = 0.654	F(1,70) = 0.05, p = 0.826
Anger	F(1,70) = 0.83, p = 0.366	F(1,70) = 0.00, p = 0.986	F(1,70) = 0.27, p = 0.607
Anxiety	F(1,70) = 0.52, p = 0.475	F(1,70) = 0.09, p = 0.769	F(1,70) = 0.01, p = 0.923
Depression	F(1,70) = 0.33, p = 0.567	F(1,70) = 0.67, p = 0.417	F(1,70) = 1.47, p = 0.23
Fatigue	F(1,70) = .059, p = 0.443	F(1,70) = 0.06, p = 0.814	F(1,70) = 2.09, p = 0.153
Happiness	F(1,70) = 1.00, p = 0.321	F(1,70) = 1.11, p = 0.295	F(1,70) = 2.46, p = 0.122
Vigor	F(1,70) = 1.37, p = 0.246	F(1,70) = 3.49, p = 0.066	F(1,70) = 5.09, p = 0.027
Sleepiness	F(1,70) = 0.15, p = 0.699	F(1,70) = 0.03, p = 0.861	F(1,70) = 0.15, p = 0.699

ANAM, Automated Neuropsychological Assessment Metric.

Bold type indicates significant effect.

At the 1-Wk Follow-Up, overall mean symptom severity ratings no longer differed between mTBI and control subjects (Fig. 1, Table 3). Subjects with mTBI, however, still reported greater symptom severity for some symptoms that are generally considered physical in nature: headache, sensitivity to light, sensitivity to noise, balance disturbances, blurred vision, and feeling slowed down (Fig. 1A).

Thirteen subjects with mTBI did not return for the 1-Wk Follow-Up, whereas only three controls did not return (Table 1). At the Initial Visit, the mean symptom severity ratings of the 13 subjects with mTBI who did not return were significantly greater than those of the subjects with mTBI who did return (Tukey post hoc test, p = 0.006; Fig. 1C). In fact, the five most extreme mean symptom severity ratings at the Initial Visit were from subjects with mTBI who did not return (Fig. 1B). There was no difference in the mean symptom severity ratings of controls between the Initial Visit and 1-Wk Follow-Up.

Mood and Sleepiness

At the Initial Visit, mTBI and control subjects differed significantly on all mood subdomains and sleepiness (Fig. 2A, Table 3). There was a main effect of sex for the item vigor, with females overall reporting lower vigor than males. At the 1-Wk Follow-Up, there were no differences between mTBI and control subjects in mood or sleepiness. There was, however, a significant injury-condition-by-sex interaction, with female subjects with mTBI reporting less vigor than any other group. At both visits, for mTBI and control subjects alike, all mood states and sleepiness were strongly correlated with mean symptom severity ratings (Table 4).

FIG. 2.

Automated Neuropsychological Assessment Metric (ANAM) mood ratings and cognitive task performance for mild traumatic brain injury (mTBI) and control subjects. The ANAM-Mood ratings (A) and ANAM-Cognitive Task performance (B) for female and male control and mTBI subjects at the Initial Visit and 1-Wk Follow-Up. Italicized letters under each bar graph indicate significant results of two-way analyses of variance (a – main effect of injury condition group, b – main effect of sex, c – group X sex interaction). *Significant difference from same sex control group; ‡significant difference from opposite sex of same injury condition (Fisher least significant difference, p < 0.05).

Table 4.

Correlation of Neurobehavioral Measures with Mean Symptom Severity Ratings

	Initial Visit				1-Wk Follow-Up
	mTBI		Control		mTBI		Control
	r	P	r	P	r	P	r	P
Simple reaction Time (msec)	0.260	0.084	-0.161	0.296	0.266	0.135	0.254	0.109
Simple reaction time repeated (msec)	0.254	0.093	0.074	0.632	0.241	0.177	0.133	0.406
Code substitution throughput	-0.094	0.540	0.076	0.626	-0.338	0.055	-0.028	0.863
Code substitution delayed throughput	-0.156	0.305	-0.009	0.955	-0.276	0.120	-0.129	0.422
Go/No Go D-prime	-0.157	0.302	0.115	0.463	-0.291	0.100	0.017	0.915
Matching-to-sample throughput	-0.442	0.002	-0.012	0.939	-0.268	0.131	0.099	0.537
Mathematical processing throughput	-0.367	0.013	-0.007	0.963	-0.463	0.007	0.230	0.148
Procedural reaction time throughput	-0.300	0.045	0.109	0.481	-0.457	0.008	0.137	0.392
Restlessness	0.535	<0.001	0.537	<0.001	0.641	0.000	0.620	<0.001
Anger	0.621	<0.001	0.408	0.006	0.781	0.000	0.609	<0.001
Anxiety	0.517	<0.001	0.615	<0.001	0.542	0.001	0.672	<0.001
Depression	0.501	<0.001	0.590	<0.001	0.699	0.000	0.519	0.001
Fatigue	0.743	<0.001	0.695	<0.001	0.876	0.000	0.657	<0.001
Happiness	-0.369	0.013	-0.394	0.008	-0.406	0.019	-0.484	0.001
Vigor	-0.401	0.006	-0.409	0.006	-0.416	0.016	-0.246	0.122
Sleepiness	0.675	<0.001	0.400	0.007	0.660	0.000	0.392	0.011

mTBI, mild traumatic brain injury; r, Pearson correlation coefficient; P, nominal probability.

Bold type indicates nominal significance.

Cognition

At the Initial Visit, females with mTBI exhibited longer simple reaction times than male mTBI or control subjects (Fig. 2B, Table 3). There were also significant overall sex differences for matching-to-sample and mathematical processing, with females exhibiting lower throughput scores than males. For subjects with mTBI, there was also a significant negative correlation between mean symptom severity rating and throughput performance on the matching-to-sample, mathematical processing, and procedural reaction time tasks (Table 4).

Impaired throughput performance of subjects with mTBI was not simply a result of slower reaction times, but also a result of poorer accuracy. For example, mean symptom severity and percent correct responses were negatively correlated in subjects with mTBI for matching-to-sample (r = -0.39, p = 0.009) and procedural reaction time (r = -0.39, p = 0.008). Cognitive performance and mean symptom severity ratings were not correlated in controls.

At the 1-Wk Follow-up, mTBI females continued to exhibit significantly longer simple reaction times than mTBI males or controls (Fig. 2B, Table 3). Similar to the Initial Visit, for subjects with mTBI, poorer throughput on mathematical processing and procedural reaction time was associated with greater mean symptom severity ratings (Table 4). Throughput scores and mean symptom severity ratings were not correlated in controls.

Salivary cortisol

Sample return

Of the 46 subjects with mTBI who underwent neurobehavioral assessment at the Initial Visit, 33 (72%) returned saliva samples. The 13 subjects who did not return saliva samples also did not return for neurobehavioral assessment at the 1-Wk Follow-up. Of the 33 subjects with mTBI who returned for neurobehavioral assessment at the 1-Wk Follow-Up, 29 (88%) returned saliva samples.

Diurnal cortisol

Bedtime cortisol was uniformly low for all subjects. For most subjects, the cortisol sample provided on awakening was substantially elevated (∼7-fold mean increase). These data illustrate a robust diurnal rhythm of cortisol secretion. There were no injury condition or sex differences in bedtime or initial waking cortisol levels (Supplementary Fig. S1).

CAR

At both the Initial Visit and 1-Wk Follow-Up, the mean responses of control subjects exhibited increases in average morning cortisol over the first 30–45 min after waking that then declined by 60 min (main effect of within-session time; repeated-measure ANOVA p < 0.001; Fig 3). In contrast, the mean responses of subjects with mTBI exhibited a relatively flat CAR at both visits (no significant effect of within-session time; repeated-measure ANOVA, p > 0.05). There were no main effects or interactions for injury condition and sex at either session.

FIG. 3.

Morning cortisol awakening response (CAR) for mild traumatic brain injury (mTBI) and control subjects. Self-collected morning salivary cortisol levels across the first 60 min after in-home waking at the Initial Visit or 1-Wk Follow-Up. Top panel (A) shows mean ± standard error of the mean for all subjects; middle (B) and bottom panel (C) show data separated by males and females, respectively. Note the generally blunted CAR for mTBI males and females at both the Initial Visit and 1-Wk Follow-Up. Asterisk next to an injury condition group in the figure key for each graph indicates a significant within-subject effect of time after awakening for that group (within-subjects analysis of variance). Note that most control groups, but not mTBI groups, had a significant within-subject effect of time after awakening.

Examination of individual subject CAR profiles for the 20% of subjects with mTBI with either the greatest or least mean symptom severity ratings compared with a random sample of 20% of control subjects illustrates that most subjects in the severely symptomatic mTBI subject group lacked a positive CAR response in comparison with the other two groups (Supplementary Fig. S2).

Association of salivary cortisol and ANAM neurobehavior

At the Initial Visit, blunted CAR Magnitude strongly predicted mean symptom severity—i.e., lower CAR Magnitude was associated with higher symptom severity (Table 5). Blunted CAR Magnitude also predicted anger, depression, and fatigue. For mean symptom severity and anger, there were significant interactions between CAR Magnitude and injury condition (Table 5). These effects were driven by significant negative correlations in subjects with mTBI between CAR Magnitude and mean symptom severity (r = -0.407, p = 0.019; Fig. 4A) and CAR Magnitude and anger (r = -0.465, p = 0.006). These variables were not correlated when assessing control subjects alone (p > 0.05). At the Initial Visit, Diurnal Amplitude was not a significant predictor for any of the neurobehavioral measures.

FIG. 4.

Association of cortisol awakening response (CAR) Magnitude and Diurnal Amplitude with mean symptom severity. (A) CAR Magnitude of mTBI, but not control subjects, was significantly correlated with mean symptom severity at the Initial Visit. (B) Diurnal Amplitude of mTBI, but not control subjects, was significantly correlated with mean symptom severity at the 1-Wk Follow-Up. These results support the significant predictor effect of CAR Magnitude by Injury Condition and Diurnal Amplitude by injury condition on mean symptom severity (multiple regression, Table 5). r, Pearson correlation coefficient; p indicates nominal probability level.

Table 5.

Significant Linear Multiple Regression Results for Cortisol Parameters as Predictor of Neurobehavioral Measures

	beta	SE	t	p
Initial Visit
CAR Magnitude predictor of:
Mean symptom severity	-1.9	0.57	-3.31	0.001
Anger	-1.54	0.44	-3.51	0.001
Depression	-2.21	0.68	-3.27	0.002
Fatigue	-2.28	1.02	-2.24	0.028
CAR Magnitude by injury condition predictor of:
Mean symptom severity	-1.23	0.58	-2.14	0.036
Anger	-1.21	0.44	-2.76	0.007
1-Wk Follow-Up
CAR Magnitude predictor of:
Simple reaction time	-116.22	49.09	-2.37	0.021
Simple reaction time repeated	-211.92	81.32	-2.61	0.011
Procedural reaction time throughput	28.32	11.22	2.52	0.014
Diurnal Amplitude predictor of:
Mean symptom severity	-1.52	0.47	-3.21	0.002
Simple reaction time (msec)	-117.47	41.72	-2.82	0.007
Simple reaction time repeated (msec)	-176.01	70.08	-2.51	0.015
Procedural reaction time throughput	27.22	9.25	2.94	0.005
Go/No-Go D-Prime	2.22	0.82	2.72	0.008
Math processing throughput	10.39	4.94	2.10	0.039
Anger	-1.46	0.46	-3.19	0.002
Diurnal Amplitude by injury condition predictor of:
Go/No-Go D-Prime	-1.78	0.806	-2.21	0.031

SE, standard error; CAR, cortisol awakening response.

Each linear model includes cortisol parameter (CAR Magnitude or Diurnal Amplitude), injury condition group, sex, and cortisol parameter by injury condition group as predictors.

Interestingly, although at the 1-Wk Follow-Up mean symptom severity ratings did not differ between mTBI and control subjects, CAR Magnitude and Diurnal Amplitude significantly predicted some of the neurobehavioral measures (Table 5). Blunted CAR Magnitude significantly predicted longer simple reaction time, longer simple reaction time repeated, and poorer procedural reaction time throughput. Reduced Diurnal Amplitude significantly predicted increased mean symptom severity (Fig. 4B), slower simple reaction time, slower simple reaction time repeated, and increased anger.

Reduced Diurnal Amplitude also significantly predicted poorer performance scores for procedural reaction time, Go/No-Go, and mathematical processing. Regarding the Go/No-Go task, there was an interaction between Diurnal Amplitude and injury condition, driven by a stronger correlation between Diurnal Amplitude and Go/No-Go D-prime scores for subjects with mTBI (r = 0.505, p = 0.001) than control subjects (r = 0.283, p = 0.022).

Discussion

This study is the first to examine associations between neurobehavior and salivary cortisol profiles shortly after mTBI. Research subjects were young adult non-varsity athlete college students who sought medical care in a university-based health clinic after recent mTBI and a well-matched non-injured student control group. These subjects may better represent mTBI symptomology in the general population of young adults than the more widely studied military personnel and athletes with mTBI.

As expected, the subjects with mTBI reported more symptoms and greater symptom severity than controls at the Initial Visit. The controls, however, also endorsed a number of physical, emotional, and cognitive symptoms, perhaps reflecting the stressful aspects of being a college student. This illustrates the importance of including well-matched control subjects for study of mTBI symptoms and symptom severity.

At the Initial Visit, subjects with mTBI reported pronounced alteration of all mood states and increased sleepiness compared with controls. They also had overall slower reaction times, and this difference was especially pronounced in females with mTBI. These data are consistent with other work showing that slower reaction time is a sensitive indicator of cognitive impairment.^33,35

We also found significant correlations between some of the cognitive tests and symptom severity for subjects with mTBI. At the 1-Wk Follow-up, subjects with mTBI continued to rate some symptoms as having greater severity than controls, and these symptoms were primarily physical in nature. As was the case for the Initial Visit, females with mTBI continued to exhibit slower simple reaction times. An association between symptom severity and cognitive performance on select tasks also persisted for subjects with mTBI at the 1-Wk Follow-Up.

It is noteworthy that females with mTBI reported greater symptom severity than males. Several other studies of sports-related concussion in high school and college students have reported increased symptom severity ratings of females within the first several weeks after concussion.^36,37 Our study indicates that this sex difference extends to non-varsity athlete college age individuals.

This sex difference may reflect psychosocial factors that contribute to differences in self-report symptom severity tendencies. For example, there may be a greater tendency for males than females to downplay symptom severity.³⁸ However, there was no mTBI sex difference for each of the separate mood ratings. In addition, there was a sex difference in the objective measure of simple reaction time, with females with mTBI overall showing pronounced longer reaction times than males with mTBI. We also note that there were no sex differences for the symptom severity ratings among the control subjects.

Taken together, these data suggest that mTBI females overall experienced greater symptom severity and that this was appropriately reflected in their subjective ratings. Other studies have found that females are more likely to have worse long-term mTBI clinical outcomes than males.^39

–43 This study shows that sex differences are apparent soon after mTBI and persist to the 1-Wk Follow-Up. Early detection of females with greater post-mTBI impairment may be helpful in developing treatment strategies that reduce risk of poor long-term outcomes for these individuals.

The cortisol profiles of the control subjects align with well-established adult salivary cortisol profiles, including a robust CAR.³⁴ Although bedtime and waking cortisol profiles in subjects with mTBI were similar to those of control subjects, subjects with mTBI exhibited a blunted CAR at both the Initial Visit and 1-Wk Follow-Up.

At the Initial Visit, the blunted CAR Magnitude in subjects with mTBI was a significant predictor of symptom severity, even when controlling for injury condition and sex. Decreased CAR Magnitude also significantly predicted increased anger, depression, and fatigue. Although at the 1-Wk Follow-Up there were no group differences in overall mean symptom severity, both blunted CAR and decreased diurnal cortisol in subjects with mTBI were significant predictors of symptom severity and/or decreased cognitive performance.

Some subjects with mTBI dropped out after their Initial Visit (28%). These individuals reported greater symptom severity than subjects with mTBI who returned for the 1-Wk Follow-Up. Consequently, although we saw no significant difference in symptom severity ratings between mTBI and control subjects at the 1-Wk Follow-Up, the dataset does not capture individuals with the most severe initial symptoms. Of further note, those who dropped out of the study also did not collect and return saliva samples for the Initial Visit. Thus, the strong association of the cortisol measures with neurobehavior at the Initial Visit and 1-Wk Follow-Up is particularly striking given that the relationship was evident despite the absence of some of the initially most symptomatic subjects.

Taken together, these data indicate that reduced amplitude of diurnal cortisol and a blunted CAR are associated with increased neurobehavioral symptoms in individuals with recent mTBI, providing strong support for cortisol as a biomarker of neurobehavioral impairment after mTBI. However, the underlying mechanisms of altered cortisol profiles after acute mTBI are not yet clear. While cortisol profiles could reflect a generalized stress state that is positively correlated with symptom severity, subjects with mTBI did not exhibit overall elevated cortisol levels compared with control subjects, as would be expected if the differences in cortisol profiles were primarily driven by differential exposure to acute stress. Instead, mTBI symptom severity was strongly associated with decreased CAR Magnitude.

A blunted CAR has been associated with depression, chronic fatigue syndrome, posttraumatic stress disorder (PTSD), sleep disorders, chronic pain, and other aberrant physiological and psychiatric conditions,^{25,27
–29,44} suggesting some underlying dysregulation of HPA axis function under these circumstances. The extent to which a generalized stress state contributes to altered cortisol profiles surrounding mTBI would be best assessed in future studies that include an acute injured control group without mTBI.

Differences in daily behavioral patterns of subjects recovering from recent mTBI compared with control subjects may also impact daily cortisol profiles. For example, daily activity levels, sleep patterns, and light exposure can modulate cortisol profiles.^24,25 Future work will need to characterize daily activity, sleep, and light exposure profiles after mTBI to account for such differences.⁴⁵

An additional consideration for use of salivary cortisol measures as a biomarker relevant to mTBI research is that dysregulation of cortisol profiles is associated with stress-related disorders such as major depression and PTSD.^24,27,46 Both depression and PTSD have relatively high comorbidity with mTBI.^47
–49 Thus, altered cortisol profiles present in subjects with mTBI with these comorbid disorders may reflect more complex underlying processes.

Regardless of the underlying mechanisms, we suggest that these alterations in cortisol signaling might contribute to neurobehavioral impairment through circadian disruption. Diurnal rhythms in the secretion of glucocorticoid hormones (cortisol in humans and corticosterone in rats and mice) serve as a coordinator of circadian biological clocks located throughout the brain and body.^21,50,51 Blunted or mistimed cortisol diurnal peaks may compromise peripheral and brain physiology.^21,52 For example, optimal function of the rat prefrontal cortex circadian clock requires a daily surge of corticosterone at the appropriate time each day.⁵³ Further, disruption of diurnal corticosterone impairs a type of learning important for emotional regulation.⁵⁴

A larger CAR amplitude in humans may help to augment the diurnal cortisol amplitude, resulting in stronger cortisol signals for circadian clocks. Thus, mTBI-induced disruptions in cortisol profiles—namely, a reduced amplitude of diurnal cortisol and a blunted CAR—may disrupt glucocorticoid signaling to circadian clocks in the brain, contributing to neurobehavioral impairment after acute mTBI. In turn, clinical assessment of cortisol profiles may shed light on pathology after injury and aid in the assessment of underlying neurobiological responses to various therapeutic interventions.

Conclusions

We found that a blunted CAR and decreased diurnal cortisol were associated with greater symptom severity, increased negative mood states, decreased positive mood states, and impaired cognitive performance in subjects with mTBI. These disruptions in cortisol point to alterations in the central nervous system that may contribute to dysregulated circadian function after injury. In addition, dysregulated cortisol may exacerbate some of the physiological and neurobehavioral consequences of mTBI. Our results support further investigation of cortisol and its association with mTBI recovery and outcome. Moreover, they indicate the potential for behavioral and pharmacological interventions that normalize cortisol profiles as novel therapeutics in mTBI recovery.^21,55

Footnotes

Acknowledgments

We are grateful for the extensive advice and assistance with conducting this study within a busy student health center provided by John Breck, Tracy Casault, Melissa Batey, Sarah Jirkovsky, Mary McQueen, Matt McQueen, and the rest of the Medical Services team at the University of Colorado Boulder. Undergraduate students Piper Doering and Mason Valdez provided valuable technical assistance.

Authors' Contributions

Conception and design of the study: Theresa D. Hernández (TDH), Robert L. Spencer (RLS), Eduardo Villegas (EV), Matthew J. Hartsock (MJH). Data collection and cortisol assays: EV, Bo L.L.G. Aben (BLLGA), MJH. Data analysis, Figures, and Tables: EV, RLS, Kristen N. Lenahan (KNL). Drafting of the manuscript: EV, RLS, MJH. Interpretation of the findings and final edits: EV, TDH, RLS, MJH, KNL, BLLGA.

Funding Information

Funding was provided by a Research and Innovation SEED Grant from the Research and Innovation Office at the University of Colorado Boulder and by the PAC-12 Student Athlete Health and Well-Being Initiative.

Author Disclosure Statement

No competing financial interests exist.

References

Taylor

C.A.

, Bell

J.M.

, Breiding

M.J.

, and Xu

(2017). Traumatic brain injury–related emergency department visits, hospitalizations, and deaths United States, 2007 and 2013. MMWR. Surveillance Summaries, 66, 1–16.

Choe

M.C.

(2016). The pathophysiology of concussion. Curr. Pain Headache Rep. 20, 42.

Laker

S.R.

(2011). Epidemiology of concussion and mild traumatic brain injury. PM R 3, Suppl. 2, S354–S358. 4..

Kay

(1993). Neuropsychological treatment of mild traumatic brain injury. J. Head Trauma Rehabil. 8, 74–85.

Sharp

D.J.

, and Jenkins

P.O.

(2015). Concussion is confusing us all. Pract. Neurol. 15, 172–186.

Rutherford

W.H.

, Merrett

J.D.

, and McDonald

J.R.

(1979). Symptoms at one year following concussion from minor head injuries. Injury, 10, 225–230.

Spinos

, Sakellaropoulos

, Georgiopoulos

, Stavridi

, Apostolopoulou

, Ellul

, and Constantoyannis

(2010). Postconcussion syndrome after mild traumatic brain injury in Western Greece. J. Trauma, 69, 789–794.

Arrieux

J.P.

, Cole

W.R.

, and Ahrens

A.P.

(2017). A review of the validity of computerized neurocognitive assessment tools in mild traumatic brain injury assessment. Concussion 2, CNC31.

Reeves

D.L.

, Winter

K.P.

, Bleiberg

, and Kane

R.L.

(2007). ANAM^® Genogram: historical perspectives, description, and current endeavors. Arch Clin. Neuropsychol. 22, Suppl. 1, S15–S37.

10.

J.K.

, Guskiewicz

K.M.

, Mihalik

J.P.

, Schmidt

J.D.

, Kerr

Z.Y.

, and McCrea

M.A.

(2013). Reliable change, sensitivity, and specificity of a multidimensional concussion assessment battery: implications for caution in clinical practice. J. Head Trauma Rehabil. 28, 274–283.

11.

Nelson

L.D.

, LaRoche

A.A.

, Pfaller

A.Y.

, Lerner

E.B.

, Hammeke

T.A.

, Randolph

, Barr

W.B.

, Guskiewicz

, and McCrea

M.A.

(2016). Prospective, head-to-head study of three computerized Neurocognitive Assessment Tools (CNTs): reliability and validity for the assessment of sport-related concussion. J. Int. Neuropsychol. Soc. 22, 24–37.

12.

Kane

R.L.

, Roebuck-Spencer

, Short

, Kabat

, and Wilken

(2007). Identifying and monitoring cognitive deficits in clinical populations using Automated Neuropsychological Assessment Metrics (ANAM) tests. Arch. Clin. Neuropsychol. 22, Suppl. 1, S115–S126.

13.

Levin

H.S.

, and Diaz-Arrastia

R.R.

(2015). Diagnosis, prognosis, and clinical management of mild traumatic brain injury. Lancet Neurol. 14, 506–517.

14.

Mayer

, Quinn

, and Master CL (2017) The spectrum of mild traumatic brain injury: a review. Neurology 89, 623–632.

15.

Dadas

, Washington

, Diaz-Arrastia

, and Janigro

(2018) Biomarkers in traumatic brain injury (TBI): a review. Neuropsychiatric disease and treatment 14, 2989–3000.

16.

Kim

H.J.

, Tsao

J.W.

, and Stanfill

A.G.

(2018). The current state of biomarkers of mild traumatic brain injury. JCI insight, 3, e97105.

17.

Spencer

R.L.

, and Deak

(2017). A users guide to HPA axis research. Physiology and, Behavior, 178, 43–65.

18.

Spencer

R.L.

, Chun

L.E.

, Hartsock

M.J.

, and Woodruff

E.R.

(2020). Neurobiology of stress-health relationships, in: Cohen, L.M., ed. The Encyclopedia of Health Psychology,. L.M. Cohen (ed). Wiley-Blackwell: Hoboken, NJ, pp. 21–35.

19.

Clow

, Thorn

, Evans

, Hucklebridge

(2004). The awakening cortisol response: methodological issues and significance. Stress, 7, 29–37.

20.

Pruessner

J.C.

, Wolf

O.T.

, Hellhammer

D.H.

, Buske-Kirschbaum

, Von Auer

, Jobst

, Kaspers

, and Kirschbaum

(1997). Free cortisol levels after awakening: a reliable biological marker for the assessment of adrenocortical activity. Life Sci. 61, 2539–2549.

21.

Spencer

R. L.

, Chun

L. E.

, Hartsock

M. J.

, Woodruff

E. R.

(2017). Glucocorticoid hormones are both a major circadian signal and major stress signal: how this shared signal contributes to a dynamic relationship between the circadian and stress systems. Front. Neuroendocrinol. 49, 52–71.

22.

Buijs

R.M.

, and Escobar

(2007). Corticosterone and activity: the long arms of the clock talk back. Endocrinology, 148, 5162–5164.

23.

Wilhelm

, Born

, Kudielka

B.M.

, Schlotz

, and Wust

(2007). Is the cortisol awakening rise a response to awakening?. Psychoneuroendocrinology, 32, 358–366.

24.

Fries

, Dettenborn

, and Kirschbaum

(2009). The cortisol awakening response (CAR): facts and future directions. Int. J. Psychophysiol. 72, 67–73.

25.

Clow

, Hucklebridge

, Stalder

, Evans

, and Thorn

(2010). The cortisol awakening response: more than a measure of HPA axis function. Neurosci. Biobehav. Rev. 35, 97–103.

26.

Adam

E.K.

, Quinn

M.E.

, Tavernier

, McQuillan

M.T.

, Dahlke

K.A.

, and Gilbert

K.E.

(2017). Diurnal cortisol slopes and mental and physical health outcomes: a systematic review and meta-analysis. Psychoneuroendocrinology, 83, 25–41.

27.

Chida

, and Steptoe

(2009). Cortisol awakening response and psychosocial factors: a systematic review and meta-analysis. Biol. Psychol. 80, 265–278.

28.

Kudielka

, and Kirschbaum

(2003). Awakening cortisol responses are influence by health status and awakening time but not by menstrual cycle phase. Psychoneuroendocrinology, 28, 35–47.

29.

Carol

E.E.

, and Mittal

V.A.

(2015) Resting cortisol level, self-concept, and putative familial environment in adolescents at ultra high-risk for psychotic disorders. Psychoneuroendocrinology 57, 26–36.

30.

Wickwire

E.M.

, Schcnyer

D.M.

, Germain

, Smith

M.T.

, Williams

S.G.

, Lettieri

C.J.

, McKeon

A.B.

, Scharf

S.M.

, Stocker

, Albrecht

, Badjatia

, Markowitz

A.J.

, and Manley

G.T.

(2018). Sleep, sleep disorders, and circadian health following mild traumatic brain injury in adults: review and research agenda. J. Neurotrauma, 35, 2615–2631.

31.

Hutchison

M.G.

, Mainwaring

, Senthinathan

, Churchill

, Thomas

, and Richards

(2017). Psychological and physiological markers of stress in concussed athletes across recovery milestones. J. Head Trauma Rehabil. 32, E38–E48.

32.

Breck

, Bohr

, Poddar

, McQueen

M.B.

, and Casault

(2019). Characteristics and incidence of concussion among a US collegiate undergraduate population. JAMA Netw. Open, 2, e1917626.

33.

Vincent

A.S.

, Roebuck-Spencer

T.M.

, Cox-Fuenzalida

L.E.

, Block

, Scott

J.G.

, and Kane

(2018). Validation of ANAM for cognitive screening in a mixed clinical sample. Appl. Neuropsychol. Adult, 25, 366–375.

34.

Aardal

, and Holm

A.C.

(1995). Cortisol in saliva—reference ranges and relation to cortisol in serum. Eur. J. Chem. Clin. Biochem. 33, 927–932.

35.

Bleiberg

, Kane

R.L.

, Reeves

D.L.

, Garmoe

W.S.

, and Halpern

(2000). Factor analysis of computerized and traditional tests used in mild brain injury research. Clin. Neuropsychol. 14, 287–294.

36.

Covassin

, Elbin

R.J.

, Harris

, Parker

. and Kontos

(2012). The role of age and sex in symptoms, neurocognitive performance, and postural stability in athletes after concussion. Am. J. Sports Med. 40, 1303–1312.

37.

Alsalaheen

, Almeida

, Eckner

, Freeman

, Ichesco

, Popovich

, Streicher

, and Lorincz

(2021). Do male and female adolescents report symptoms differently after concussion?. Brain Inj. 35, 698–704.

38.

Kroshus

, Baugh

C.M.

, Stein

C.J.

, Austin

S.B.

, and Calzo

J.P.

(2017). Concussion reporting, sex, and conformity to traditional gender norms in young adults. J. Adolesc. 54, 110–119.

39.

Bazarian

J.J.

, Blyth

, Mookerjee

, He

, and McDermott

M.P.

(2010). Sex differences in outcome after mild traumatic brain injury. J. Neurotrauma, 27, 527–539.

40.

Styrke

, Sojka

, Björnstig

, Bylund

P.O.

, and Stålnacke

B.M.

(2013). Sex differences in symptoms, disability, and life satisfaction three years after mild traumatic brain injury: a population-based cohort study. J. Rehabil. Med. 45, 749–757.

41.

Hsu

H.L.

, Chen

D.Y. T.

, Tseng

Y.C.

, Kuo

Y.S.

, Huang

Y.L.

, Chiu

W.T.

, Yan

F.X.

, Wang

W.S.

, and Chen

C.J.

(2015). Sex differences in working memory after mild traumatic brain injury: a functional MR imaging study. Radiology, 276, 828–835.

42.

Gupte

R.P.

, Brooks

W.M.

, Vukas

R.R.

, Pierce

J.D.

, and Harris

J.L.

(2019). Sex differences in traumatic brain injury: what we know and what we should know. J. Neurotrauma, 36, 3063–3091.

43.

Levin

H.S.

, Temkin

N.R.

, Barber

, Nelson

A.D.

, Robertson

, Brennan

, Stein

M.B.

, Yue

J.K.

, Giacino

J.T.

, McCrea

M.A.

, Diaz-Arrastia

, Mukherjee

, Okonkwo

D.O.

, Boase

, Markowitz

AJ.

, Bodien,Y., Taylor

, Vassar

M.J.

, Manley

G.T.

, and the TRACK-BI

Investigators.

(2021). Association of sex and age with mild traumatic brain injury–related symptoms: a TRACK-TBI study. JAMA Netw. Open, 4, e213046.

44.

Rauch

S.A.M.

, King

, Myra Kim

, Powell

, Rajaram

, Venners

, Simon

N. M.

, Hanmer

, and Liberzon

(2020). Cortisol awakening response in PTSD treatment: Predictor or mechanism of change. Psychoneuroendocrinology, 118, 104714.

45.

Ancoli-Israel

, Cole

, Alessi

, Chambers

, Moocroft

, and Pollak

(2003). The role of actigraphy in the study of sleep and circadian rhythms. Sleep, 26, 342–392.

46.

Daskalakis

N.P.

, Lehrner

, and Yehuda

(2013). Endocrine aspects of post-traumatic stress disorder and implications for diagnosis and treatment. Endocrinol. Metab. Clin. North Am. 42, 503–513.

47.

Bay

, Sikorskii

, and Gao

(2009). Functional status, chronic stress, and cortisol response after mild-to-moderate traumatic brain injury. Biol. Res. Nurs. 10, 213–225.

48.

Hoge

C.W.

, McGurk

, Thomas

J.T.

, Cox

A.L.

, Engel

C.C.

, and Castro

C.A.

(2008). Mild traumatic brain injury in U.S. Soldiers returning from Iraq. N. Engl. J Med. 358, 453–463.

49.

Stein

M.B.

, Kessler

R.C.

, Heeringa

S.G.

, Jain

, Campbell-Sills

, Colpe

L.J.

, Fullerton

C.S.

, Nock

M.K.

, Sampson

N.A.

, Schoenbaum

, Sun

, Thomas

M.L.

, Ursano

R.J.

, Army

STARRS collaborators.

(2015). Prospective longitudinal evaluation of the effect of deployment-acquired traumatic brain injury on posttraumatic stress and related disorders: results from the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS). Am. J. Psychiatry, 172, 1101–1111.

50.

Balsalobre

, Brown

S.A.

, Marcacci

, Tronche

, Kellendonk

, Reichardt

H.M.

, Schutz

, and Schibler

(2000). Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science, 289, 2344–2347.

51.

Dickmeis

(2009). Glucocorticoids and the circadian clock. J. Endocrinol. 200, 3–22.

52.

Oster

, Challet

, Ott

, Arvat

, Ronald de Kloet

, Dijk

D. J.

, Lightman

, Vgontzas

, and Van Cauter

(2017). The functional and clinical significance of the 24-hour rhythm of circulating glucocorticoids. Endocr. Rev. 38, 3–45.

53.

Woodruff

E.R.

, Chun

L.E.

, Hinds

L.R.

, and Spencer

R.L.

(2016). Diurnal corticosterone presence and phase modulate clock gene expression in the male rat prefrontal cortex. Endocrinology, 157, 1522–1534.

54.

Woodruff

E.R.

, Chun

L.E.

, Hinds

L.R.

, Varra

N.M.

, Tirado

, Morton

S.J.

, McClung

C.A.

, and Spencer

R.L.

(2018). Coordination between prefrontal cortex clock gene expression and corticosterone contributes to enhanced conditioned fear extinction recall. eNeuro, 5, e0455–18.2018.

55.

Hernández

T.D.

, Brenner

L.A.

, Walter

K.H.

, Bormann

J.E.

, and Johansson

(2016). Complementary and alternative medicine (CAM) following traumatic brain injury (TBI): opportunities and challenges. Brain Res. 1640, 139–151.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

1.10 MB

3.92 MB