Neuroendocrine dysfunction following traumatic brain injury is emerging as an important area of scientific inquiry. Hormonal deficits after traumatic brain injury (TBI) were initially described in the early 20th century, but remained largely unaddressed in both research and clinical contexts until recently.¹,² Sex steroid and growth hormone deficiencies are among the most commonly reported posttraumatic neuroendocrine disturbances. Single or multiple pituitary-target hormone disruption has been reported in 25% to 69% of persons with TBI in cross-sectional studies.³—¹⁰ Whereas recovery from such disturbances occurs in some individuals,³,⁵,⁹,¹¹,¹² the extent of their resolution is highly variable. In a series of patients studied at 3 months and 1 year after TBI, no patients with complete panhypopituitarism recovered normal function at 1 year, although many with isolated or multiple deficits showed improvement.¹³ The precise time course of the decline and eventual recovery in specific pituitary hormones is unknown.

Screening for neuroendocrine deficits is not performed routinely in the acute postinjury period, and these abnormalities often go undiagnosed and untreated in persons with TBI.¹³,¹⁴ In light of the potentially adverse effects of hormonal deficits on the rate and extent of both physical and cognitive recovery, evaluation of pituitary-target gland function and the theoretical beneficial effects of hormonal replacement strategies require further investigation. Whereas the literature generally supports an association between neuroendocrine changes and the severity of brain and systemic trauma, secondary brain damage such as hypoxia, and medical complications such as infection, the complexity of these multiple and often interacting variables complicates the analysis. In this article, we focus primarily on the neuroendocrine dysfunction that can be attributed specifically to TBI. The paucity of publications and lack of consistency of design preclude a meta-analytic approach to this review. Accordingly, this review is largely descriptive, reviewing the data on hormonal abnormalities after TBI. We also discuss the potential benefits that hormonal replacement, particularly growth hormone and sex steroid replacement, might have on the rate of recovery after TBI.

TBI occurs in 1.4 million people each year in the United States.¹⁵ Children, older adolescents, and adults older than age 75 are at highest risk for TBI. The frequency is higher among men in almost every age group, although gender differences in injury occurrence are most striking among children, adolescents, and young adults. Approximately 50,000 persons die in the United States each year as a result of their brain injury, and an additional 235,000 persons with TBI require hospitalization. In addition, approximately 1.1 million persons are treated and released every year from emergency departments with mild TBI (or "concussion"). An uncertain number of individuals experience TBI of relatively mild severity that may not be evaluated at the time of injury and may only become known if recovery is incomplete and produces neurological or neurobehavioral problems that require clinical attention.

Approximately 80,000 to 90,000 persons/year will experience some permanent disability as a result of TBI, although the extent and severity of that disability are highly variable.¹⁵ Among the most common sequelae of TBI are cognitive impairments, disturbances of emotion and behavior, and somatic symptoms, such as headache, dizziness, visual disturbance, and mobility deficits.¹⁶ Secondary complications of TBI in the acute postinjury are frequent and often include deconditioning, altered nutritional status, weight loss, and infection; the complications are particularly common after TBI of moderate or greater severity, and adversely affect recovery from TBI.¹⁶

In light of the critical role of endocrine axes on the response to injury/stress, the effects of neuroactive hormones on neuronal repair, and the implications of neuroendocrine dysfunction on cognitive, emotional, behavioral, and physical function, the integrity of these systems is essential for optimal neurological and neurobehavioral recovery after TBI.

Posttraumatic neuroendocrine dysfunction may arise as a result of direct mechanical injury, cytotoxic processes, or both, to the CNS components of the hypothalamic-pituitary-target organ axes.¹⁷,¹⁸ Acceleration/deceleration forces may strain and/or shear white matter projections between these structures, particularly when those forces affect deep and medial structures, including the medial temporal lobes (i.e., entorhinal-hippocampal complex, amygdala), diencephalon, and brain stem.¹⁹—²¹ Vascular insults (i.e., hypoxia and/or ischemia), edema, and necrosis may complicate biomechanical and cytotoxic injury.²² The role of hormonal excess or deficiency in the cascade of events that follow TBI is incompletely understood, but several lines of investigation suggest that these processes are important to consider (Table 1).

Deficits in each of the hypothalamic-pituitary-target organ axes have been reported after TBI.¹ The location of the specific cell types, with somatotrophs (making growth hormone) and the gonadotrophs (making the luteinizing hormone and follicle stimulating hormone) being the most lateral and injury-prone may, in part, explain the frequencies of growth hormone and gonadotropin deficiency after TBI.¹⁴,²³ Data supporting specific pituitary-target hormone defects are reviewed below.

Central hypogonadism, with low levels of luteinizing hormone (LH), follicle stimulating hormone (FSH), and testosterone in men and estradiol in women, has been reported within hours of brain insult, with eventual recovery of the hypothalamic-pituitary-gonadal (HPG) axis.²⁴ In a cross-sectional study of 50 men with moderate to severe TBI studied 7 to 20 days postTBI, 79% (N=30) of the subjects had low serum testosterone, which correlated with their Glascow Coma Scores (GCS).²⁵ This study included subjects with elevated prolactin levels, which may have contributed to low testosterone levels. Another report showed that 14 of 21 subjects (67%) studied 1 week after transfer to rehabilitation had abnormally low testosterone levels.¹¹ In a cross-sectional study of 102 TBI survivors studied at 6 to 36 months postevent, 11.8% had gonadotropin deficiency and hypogonadism.⁹ Seven of 50 (14%) male and female subjects studied 12 to 64 months after TBI were found to have low testosterone or estradiol levels.⁷ Although the precise timeline of recovery of the hypothalamic-pituitary-gonadal (HPG) axis has not been established prospectively, hypogonadism was only detected in one of 46 men more than 4 years (49 [SD=8] months) after their initial injury.⁵ Another series of subjects evaluated at 6 and 12 months after injury also showed that the majority of them (34/40=85%) regained normal gonadal function by 1 year.²⁶

The mechanisms of suppression of the HPG axis after TBI are multifactorial. Acute, severe illness, and stress are known to adversely affect normal HPG axis function,²⁷ as are brain damage and dysfunction.⁶,²⁸ TBI-associated hypogonadism may involve decreased LH pulse amplitude, but not pulse frequency, as a result of injury to the pituitary.¹² This is in contrast to a direct injury to the hypothalamic gonadotropin releasing hormone (GnRH) neurons or effects of stress, which would result in a decreased LH pulse frequency. Medications commonly used in the management of TBI and its sequelae, in particular opiates, also can suppress GnRH-induced LH secretion. The inflammatory cascade induced by TBI may also play a role in HPG axis dysfunction at the level of the gonad, as cytokines can suppress Leydig cell function, and thereby the normal production of testosterone in the testis.²⁷

In other models of severe illness, testosterone levels decrease acutely and dramatically into the prepubertal range. Testosterone levels were shown to correlate with APACHE (Acute Physiology and Chronic Healthy Evaluation) scores in a study of 59 men in an ICU.²⁹ The more severely ill men (APACHE scores of >15) had an average level of 8.2 nmol/liter on admission (healthy age matched comparison subjects ranged from 9.7 to 33.7 nmol/l). Levels fell to 3.7 nmol/liter by day 3 and reached a nadir of 1.2 nmol/liter. Although less dramatic, men with an APACHE score of <10 average level reached a nadir of 7.2 nmol/liter. Another study of ICU admissions found 29 of 30 (96%) of men had testosterone levels below the lower limit of normal for age.³⁰

Recent studies demonstrate that patients with severe burns also have profound hypogonadism. Six men studied approximately 2 weeks after severe burns had an average level of 36.6 ng/dl (normal range=262 to 1593 ng/dl).³¹ The authors of this study advocated physiological replacement of testosterone to improve the catabolic state of these subjects. Although hypogonadism can contribute to the catabolic state seen in the critically ill, it is not known if this response is adaptive or detrimental to the recovery process.²⁷ The effects of hypogonadism on neurological and neurobehavioral function and recovery after TBI also require further investigation.

Identification and treatment of androgen deficiency may be particularly relevant to the treatment of men with TBI. Basic research studies suggest that aromatization of testosterone to estradiol is critical to the neuroprotective effects of testosterone on astroglia after acute brain injury.³²,³³ At the cellular level, estradiol has been shown to have neurotrophic effects.³⁴ In embryonic hippocampal cells, pretreatment with estrogen before injury improved cell survival.³² Rats given an aromatase inhibitor by infusion in the cerebral ventricle experienced neuronal loss in the hippocampus to a greater degree than control rats, and additional studies looking at an aromatase knockout mouse confirmed these neuroprotective effects.³³ Others have shown that estrogen blocks secretion of inflammatory mediators, such as inducible NO and prostaglandin E2 and matrix metalloproteinase-9 and complement C3 receptor after liposaccaride-induced neuronal injury.³⁵ If similar neuroprotection can be extended to persons with TBI, this intervention could afford an opportunity to mitigate the effects of injury and/or facilitate recovery following injury.

Androgen replacement also might improve posttraumatic cognitive impairments, and particularly memory disturbances, by direct action on CNS androgen receptors or via action on the estrogen receptor after aromatization to estradiol.³⁶—⁴⁰ Testosterone (from the testis or adipose tissue) and the weaker adrenal precursors dehydroepiandrosterone (DHEA) and its sulfate (DHEA-S) are converted into either more potent androgens by 5-α reductase to dihydrotestosterone, (DHT) or into estrogens via aromatization. In epidemiological cross-sectional studies, estradiol and testosterone levels are correlated with cognitive performance in older men and women.³⁶—⁴⁰ Free testosterone levels have been shown to predict memory performance and cognitive status in elderly men, and cognitive decline in women has been correlated with free estradiol concentration.³⁸,³⁹ In addition to effects on cognition, normal levels of sex steroids may be required for the maintenance of other neurobehavioral functions, and particularly motivation. For example, in chronic schizophrenics, low estradiol levels correlated with lower levels of cognitive performance and increased apathy and anhedonia.⁴¹ If sex steroids afford similar benefits among persons with TBI, their use in this context would be a novel and potentially productive contribution to their postinjury rehabilitation.

Androgens also exert anabolic effects, thereby improving muscle mass and lean body mass.⁴² In seven hypogonadal men, treatment with testosterone injections for 10 weeks led to a significant increase in fat-free mass from 56.0 to 60.9 kg. These men also had increases in cross-sectional areas of the triceps and quadriceps muscle.⁴² Testosterone has also been used in other ill populations, such as patients with HIV, to increase muscle mass and lean body mass.⁴³,⁴⁴ Treatment of men with TBI using physiological testosterone administration might improve their ability to participate in and benefit from physical therapy, and thereby improve functional outcome after injury.

Growth hormone deficiency is another common sequelae of TBI and may become a persistent problem for some of these individuals. Growth hormone is produced and released by the somatotroph cells of the anterior pituitary. The lateral location of these cells and their predisposition to vascular insufficiency, stalk injury, or anoxia may in part explain the relatively high frequency of posttraumatic growth hormone deficiency,⁷ which ranges from 9% to 28%.⁵,⁶,⁸ This wide range of estimates of growth hormone deficiency reflects controversy regarding the best methods of diagnosing growth hormone deficiency in adults.

The production and release of growth hormone is stimulated by growth hormone releasing hormone (GHRH) and suppressed by growth hormone release-inhibiting factor (also known as somatostatin) and by glucose. Growth hormone activates the liver protein, insulin-like growth factor 1 (IGF-1), to act on target tissues to decrease protein catabolism, mobilize fat, decrease carbohydrate utilization, and increase insulin resistance.⁴⁵ growth hormone has a short half-life in plasma, and its episodic secretion varies throughout the day, making single measurements of growth hormone difficult to interpret. IGF-1 has a long half-life and, although it is used to screen for growth hormone excess, is not a useful screen alone for adult growth hormone deficiency. Although insulin tolerance testing (ITT) to induce hypoglycemia and growth hormone secretion has been regarded as the "gold standard" for growth hormone reserve, the risk of hypoglycemia makes it unsuitable for testing of persons with recent TBI.

Reports describing the prevalence of posttraumatic growth hormone deficiency vary considerably with respect to their methods of testing, the criteria used to define deficiency, and the time postinjury at which the assessment is performed. Accordingly, estimates of the prevalence of growth hormone deficiency are wide ranging. In a series of 50 subjects with TBI, growth hormone, IGF-1, and dynamic testing with glucagon were performed "acutely" (i.e., 9 to 20 days postTBI). Nine of the 50 subjects (18%) had growth hormone levels of <5 ng/ml after stimulation testing.³ However, when these growth hormone deficient subjects were compared with those with normal growth hormone function, there was no difference in age, GCS, IGF-1 level, or relation to other pituitary deficiencies such as ACTH or the gonadotropins, LH and FSH. When 34 TBI subjects were assessed with growth hormone releasing hormone (GHRH) provocative testing, partial growth hormone deficiency (defined as a peak growth hormone between 3 and 5mcg/liter; normal >5mcg/liter) was observed in three (9%). IGF-1 levels were similar between the groups, confirming that this measure is not a good screening test for growth hormone deficiency postTBI.⁶ When 50 subjects were tested at a longer interval (12 to 64 months) after their injury, low growth hormone levels after GHRH and arginine stimulation test were observed in 28% of subjects.⁷ In this group, only seven subjects also had low IGF-1 levels, and one subject with low IGF-1 level had normal provocative testing. Thus, IGF-1 was insensitive as a screen for growth hormone deficiency in the context of TBI.

Conversely, low IGF-1 levels do not necessarily indicate growth hormone deficiency. In a study of 170 subjects with TBI, dynamic testing was done on a subset of 44 subjects with low IGF-1 levels (less than 200 μg [SD=2]) or any other documented pituitary deficiency. When growth hormone testing was performed with a combination of GHRH and growth hormone releasing peptide (GHRP-6), only six of 44 subjects were clearly growth hormone deficient (peak growth hormone level of <10 μg/liter). When the results of ITT testing, glucagon testing, and IGF-1 levels were taken together, four additional subjects whose GHRH-GHRP-6 results had been indeterminate were also classified as growth hormone deficient.⁴ Overall, the prevalence was much lower (5.8%) than in prior studies that did not use these rigorous definitions. Importantly, in a follow-up study of 50 subjects initially tested for growth hormone deficiency in the acute period, only 5 of 48 (10%) were still deficient or had become deficient by the end of 1 year after brain injury.²⁶

We may be able to obtain more consistent estimates of posttraumatic growth hormone deficiency by using the recently developed consensus-panel recommendations, which suggest that the most sensitive and specific approach to the diagnosis of growth hormone deficiency in adults is provocative testing with GHRH in combination with arginine or GHRP-6.⁴⁶ After more clearly defining the true scope of this problem, its influence on posttraumatic neurological and neurobehavioral function and recovery may be more readily apparent.

It is unclear whether growth hormone replacement would facilitate recovery following TBI. Animal models suggest that growth hormone supplementation after a hypoxic injury can reduce neuronal loss.⁴⁷ To the extent that an individual’s TBI is complicated by hypoxic injury, growth hormone replacement may mitigate the contribution of this factor to injury and its sequelae. In a double-blind, placebo-controlled trial of 24 adults, 6 months of growth hormone therapy led to increases in lean body mass of 5.5 kg and decreases in fat mass of 5.7 kg.⁴⁸ Treatment has also been shown to increase exercise capacity.⁴⁹,⁵⁰ In addition, growth hormone replacement may improve cognition, psychological well being, and quality of life. If these benefits were seen in persons with TBI, it could increase the likelihood of effective participation in and benefit from neurorehabilitation.⁵¹

Among critically ill patients, the use of growth hormone is associated with improvements in nitrogen balance, strength, and weaning from mechanical ventilation.⁵²,⁵³ However, two prospective, double-blind, randomized, placebo-controlled trials in intensive care unit (ICU) patients showed an increased morbidity and mortality in those receiving high dose growth hormone.⁵⁴ The authors questioned whether hyperglycemia induced by the supraphysiological doses of growth hormone, effects of high dose growth hormone on other endocrine axes, or stimulation of excess lipolysis played a role in the adverse outcomes observed. These preliminary studies in a medically diverse patient population without documented growth hormone deficiency suggest a need for prospective investigation of the effects of physiologic growth hormone replacement among persons with posttraumatic growth hormone deficiency after TBI.

Both excessive activation and deficiency in the hypothalamic-pituitary-adrenal (HPA) axis have been reported after TBI. Corticotropin releasing hormone (CRH) from hypothalamic neurons activates ACTH in the pituitary, which then acts on the adrenal glands to induce the release of mineralcorticoids, glucocorticoids (cortisol), and adrenal precursor hormone, dehydroepiandrostenedione (DHEA). Adrenal insufficiency can result from central or peripheral defects, but persons with TBI would be expected to have central ACTH deficiency. Normally, cortisol production should increase in times of stress or illness to maintain blood pressure and fight infection; thus, cortisol deficiency due to disruption anywhere along the hypothalamic-pituitary-adrenal axis may be life threatening.

Cortisol levels vary according to the severity of the brain injury and duration of time since the event. Subjects studied immediately postinjury showed an early decrease in cortisol followed by an increase 5 days after the injury.⁵⁵ A recent study of persons with brain injuries in an ICU showed normal cortisol levels, but elevated free cortisol levels.⁵⁶ These patients must be observed closely: after the stress of acute illness, deficits in the HPA axis resulting from TBI may become apparent. In subjects tested 6 and 12 months after injury, 8 of 50 were diagnosed with HPA deficits in the acute setting. Five new cases and four recoveries were reported at 6 months, and no new cases or recoveries were reported at 1 year.²⁶

Other studies of acute and chronic posttraumatic HPA axis function are mixed with respect to methodology, but relatively consistent with respect to their findings of disturbances in this axis. In the acute period after TBI, 8 of 50 and 8 of 34 (16% to 24%) subjects were reported as ACTH deficient by stimulation with glucagon or corticotropin releasing hormone (CRH).³,⁶ A study of 40 persons with TBI in an intensive care unit similarly showed 6 of 40 (15%) to have adrenal insufficiency as diagnosed by a low dose (1mcg) cosytropin stimulation test.⁵⁷ ACTH stimulation testing of a group of subjects 49 months postinjury confirmed that although that 32 of 70 (45%) had a low baseline cortisol, only five (7.1%) had an abnormal cosytropin stimulated cortisol level (levels less than 18 μg/dl after 250 mcg of cosyntropin). In a series of 170 individuals screened with the insulin tolerance test at least 1 year after the acute event, only 6.4% had abnormal results.⁴ Similar to results seen with random screening of growth hormone or IGF-1, random cortisol levels did not predict ACTH deficiency, and stimulation testing is required for appropriate diagnosis and treatment. Collectively, these studies suggest that acute disturbances of HPA axis function are not uncommon, and that the majority—but not all—of persons with such disturbances recover normal HPA axis function by 1 year postinjury.

Similar to the challenges associated with assessment of gonadal and growth hormone levels, consensus is lacking regarding the optimal methods with which to assess HPA axis function. There is, however, general agreement that a random cortisol level is not a sensitive or specific test unless it is very low or very high; accordingly, HPA axis function assessments usually employ stimulation testing using cosyntropin, CRH, insulin, or glucagon. The most commonly used test is the cosyntropin stimulation test: 250 mcg of cosyntropin (ACTH) is given and cortisol levels are checked at 0, 30, and 60 minutes. More recently, the low dose test using 1 mcg of cosyntropin has been employed, which is more sensitive for diagnosing central ACTH deficiency, which would be the expected origin of defective HPA axis function after TBI.⁵⁸,⁵⁹ However, even the 1 mcg test may not suffice to diagnose ACTH deficiency early in its course, before adrenal tissue has had time to deplete its stores of cortisol. The previously used gold standard insulin tolerance test induces hypoglycemia and must be used with caution in this population, given the risk of inducing hypoglycemic seizures and metabolic stress in an already compromised brain.¹

Another challenge in the assessment of HPA axis function after TBI is alteration in cortisol binding proteins produced by acute illness, which affect tissue levels of free cortisol and thereby make the diagnosis of adrenal insufficiency difficult. Recent studies have debated the best way to diagnose and treat adrenal insufficiency in medical and surgical ICUs.⁶⁰,⁶¹ At present, it appears that measurement of serum free cortisol, which avoids the confound of illness-related alterations in cortisol binding proteins, may be better suited to this task than measurement of standard cortisol levels.⁶²

In summary, there is no current consensus on the optimal approach to assess the HPA axis in persons with TBI, particularly in the acute period. However, the literature would support the use of serum-free cortisol measurements after ACTH stimulation at multiple time-points. Since there is evidence of both increases and decreases in cortisol in persons with TBI, and because cortisol deficiency may be life threatening, it is imperative to assess HPA axis function in this population and provide at least short-term treatment when significant deficits are discovered.

The effects of TBI on the pituitary-thyroid axis may be particularly difficult to separate from the consequences of acute illness. Additionally, injury to pituitary thyrotropes may be less common than to other areas of the pituitary, due to their central location. Thyrotropin releasing hormone (TRH) from the hypothalamus activates thyroid stimulating hormone (TSH) secretion from the pituitary to control the release of T4 and trithyroiodine (T3) from the thyroid gland. The thyroid hormone exerts a variety of effects on peripheral tissues, including maintenance of normal oxygen consumption and metabolism, protein turnover, and sympathetic nervous system activity.

TSH initially falls during illness, and then rises during recovery (i.e., the "sick euthyroid syndrome"). In the ICU setting, severe suppression of T3 levels correlates with ICU mortality.⁶³ Replacement with the thyroid hormone has not been shown to benefit this population.⁶⁴,⁶⁵ Free T4 is a measure of the nonprotein-bound T4 that is biologically active and is generally a more accurate assessment of thyroid function; however, it can also become low during acute illness. As a result, it is often difficult to assess deficiencies of the thyroid axis during the acute phases of and early recovery from TBI, similar to any acute illness.

Between 2% and 15% of persons with TBI have low plasma levels of free T4 with normal or low TSH, indicating central hypothyroidism.³,⁶,⁷ However, when TSH levels in response to TRH stimulation were obtained, only one in 22 subjects (4.5%) had an abnormal response consistent with central thyroid dysfunction.¹⁰ TRH stimulation testing, once considered the gold standard for diagnosis for central hypothyroidism, however, is no longer feasible in the United States due to the lack of manufacture of TRH; accordingly, the identification of subtle deficits in this neuroendocrine axis is at present extremely difficult.

In summary, the substantial effects of nonthyroidal illness on thyroid hormone levels and the low incidence of central hypothyroidism in the TBI population suggest that measuring thyroid hormone levels to diagnose central hypothyroidism in the acute period following TBI is neither easily done nor particularly useful. Since T4 levels will start to rise during illness recovery, a free T4 measurement later during recovery/rehabilitation would be indicated if symptoms consistent with hypothyroidism are present. If hypothyroidism were diagnosed during the recovery period, treatment should be initiated with physiologic replacement with T4.

Alterations in prolactin levels due to TBI have been described in several studies.³,⁵,⁷ Prolactin is synthesized and released by lactotroph cells in the anterior pituitary. Its main function is to stimulate lactation in postpartum women; however, it is increased by pituitary stalk compression and by many medications. In contrast to other pituitary hormones, prolactin is under tonic inhibitory control from the hypothalamic dopamine neurons.⁶⁶ In addition to stalk injury and medications, stress, suckling, and the thyroid stimulating hormone (TSH) modulate prolactin secretion. Sertonergic pathways activate prolactin release, explaining why some selective serotonin reuptake inhibitors (SSRIs) may lead to hyperprolactinemia. Elevated prolactin, in turn, modulates other endocrine systems. Hyperprolactinemia is associated with inhibition of GnRH-induced LH secretion and suppression of the gonadal axis. Use of dopamine agonists to suppress hyperprolactinemia may counter such effects, but the use of such agents must be considered in light of their potential beneficial and/or untoward effects on posttraumatic cognitive, emotional, and behavior function, as well as their potential interaction with other medications.⁶⁶

Hyperprolactinemia has been described in 26 of 50 (52%) individuals during the acute period post-TBI with a negative correlation between prolactin level and GCS score.³ Many of these individuals recovered at 6 months and only six of 48 (13%) studied at 12 months remained hyperprolactinemic at 1 year;²⁶ other studies suggest that few experience persistent prolactin elevations.⁵,⁷,⁹ The early incidence of hyperprolactinemia may be due to primary injury effects on the pituitary stalk, posttraumatic deficits of central dopamine resulting in a reduction in tonic suppression of prolactin release, medications given to the patient that result in hyperprolactinemia or some combination of these and other factors.

TBI appears to be associated with dysfunction across all of the major domains of neuroendocrine function, although the severity and clinical implications of such dysfunction are highly variable. Despite the frequency of such problems, information is lacking on clear criteria for the diagnosis of deficiencies, their relationship to neurological and neurobehavioral function, expectations regarding time to recovery of neuroendocrine function, and the effects of neuroendocrine interventions on short-term and long-term recovery after TBI. Future efforts, particularly in the area of sex steroid and growth hormone replacement interventions, are needed in view of both basic science and clinical research demonstrating beneficial effects of physiological hormone replacement on various aspects of cognition, mood, and physical functioning.⁴²,⁵¹,⁶⁷—⁷¹ Despite the inherent difficulties in performing clinical research in this population, such research is needed to test whether these hormonal replacement strategies may improve the rate of and ultimate level of recovery among persons with TBI.