Complaints of feeling cold, without actual alteration in body temperature, are occasionally seen in patients who have suffered brain injury. This feeling can be distressing to those who experience it. Patients may wear excessive amounts of clothing and adjust the thermostat so that other members of the family are uncomfortable. Although this is not a commonly reported symptom of traumatic brain injury (TBI), Hibbard et al.1 have found that in a sample of 331 individuals with TBI, 27.9% complained of changes in body temperature and 13% persistently felt cold. Eames,2 while conducting a study of the cognitive effects of vasopressin nasal spray in patients with TBI, reported incidentally that 13 patients had the persistent feeling of coldness, despite normal sublingual temperature. All were treated with vasopressin nasal spray for 1 month. Eleven of these patients stopped complaining of feeling cold after 1 month of treatment, and 1 other patient had improvement in the symptom, but without complete relief. We describe below a series of 6 patients with brain injury whose subsequent complaints of feeling cold were treated with DDAVP (1-d-amino-8-d-arginine-vasopressin; intranasal vasopressin or desmopressin acetate).
Patients who had suffered brain injury were selected for treatment on the basis of their subjective complaints of being cold. These patients attended brain injury rehabilitation programs and were referred for evaluation for pharmacological treatment of emotional, cognitive, or behavioral problems subsequent to brain injury. No patients were referred for treatment of the complaint of feeling cold. All patients spontaneously noticed feeling cold. Patients were treated twice a day for 1 month with intranasal DDAVP. Response was assessed at that time and 1 month after discontinuation of DDAVP. A positive response was based on the patient's reporting no longer feeling cold or on the caregiver's reporting improvement in behavior. Patients were followed for up to 6 months after discontinuation of DDAVP.
Five patients showed improvement after treatment with DDAVP. Three of the 6 patients reported complete improvement, one reported 80% improvement, and one reported 50% improvement. One patient initially reported a response to treatment, but on further questioning it did not appear that the target symptom had improved.
Patient 1. This 18-year-old African American woman sustained a closed head injury in a motor vehicle accident (MVA). She reported that she was in a coma for approximately 2 weeks. CT scan at the time of the injury revealed petechial hemorrhages at the gray—white matter junction of the left frontal and posterior right frontal lobes, mild left temporal lobe edema, and evidence of air-fluid level in the right sphenoid sinuses. She was referred for treatment of periods of depressed mood. She also noted that since the injury she felt cold all the time. Past psychiatric history was significant for prior hospitalizations for depression, as well as periods of hypomania of up to 6 months' duration.
Mood problems responded to treatment with lamotrigine and gabapentin. Two weeks after initiating treatment with DDAVP nasal spray, she noted a 50% decrease in coldness. At the time of her last evaluation, 2 months after discontinuing DDAVP, this degree of improvement persisted.
Patient 2. This 26-year-old African American woman sustained a brain injury after an MVA where she was an unbelted passenger and was thrown from the vehicle. She had a loss of consciousness (duration not known) and required intubation at that time. Two 0.5×0.5 cm left temporal lobe contusions were demonstrated on CT imaging. Past history was significant for daily marijuana use.
She was referred for treatment of irritability, decreased interest, decreased appetite, and increased sleep. These symptoms responded to sertraline. In addition, she noted that she felt cold all the time since the accident. After 1 month of treatment with DDAVP nasal spray, she noted an 80% decrease in coldness, which has persisted for 6 weeks since discontinuing the DDAVP.
Patient 3. This 20-year-old Caucasian woman sustained a brain injury in an MVA. She was in coma for 30 to 60 minutes. She sustained a left frontal parietal subdural hematoma demonstrated on MRI. Electroencephalogram (EEG) was abnormal, with bilateral (predominately frontal) cerebral dysfunction. An EEG 5 months later revealed normal background and scattered posterior sharp waves.
The patient was referred for treatment of depression, anxiety, and irritability. These problems responded to venlafaxine, sodium valproate, and alprazolam. Methylphenidate improved problems with attention and concentration. She also noted that since the injury she felt cold all the time. Thyroid function tests were normal. Past psychiatric history was significant for hospitalization for anorexia nervosa. She denied that these feelings of coldness occurred during exacerbation of her eating disorder.
The patient noted that the feelings of coldness remitted 3 days after starting DDAVP. She reported that her hands and feet felt warmer. This response has persisted during the 8 months since treatment.
Patient 4. This 33-year-old female Caucasian police officer sustained a brain injury when her patrol car was struck by another vehicle at high speed. She was in a coma for approximately 11 days. CT scan revealed diffuse cerebral edema, perisylvian contusions, and a lesion in the right frontal convexity. She was referred for evaluation because of apathy, lack of motivation, and depression. She was on sertraline, to which methylphenidate was added with significant benefit.
She also noted feeling cold all the time. After treatment with DDAVP nasal spray, she noted that these feelings were significantly improved. This improvement persisted 1 month after DDAVP was discontinued.
Patient 5. This 64-year-old Caucasian woman had a pineal tumor that was treated initially with radiation therapy. She later had a ventriculoperitoneal shunt inserted. Several weeks later, she was placed on warfarin for treatment of right lower extremity thrombosis. She had surgery for a hernia, and during the surgery sustained an intratumoral bleed with extension into the right ventricle. She was in a coma for 1 week. CT later revealed a hemorrhage into the midbrain with extension of the hemorrhage into the third and lateral ventricles. A mild to moderate degree of hydrocephalus was noted. She was referred for evaluation of problems with depression, lethargy, decreased arousal, and sleep problems. These problems never adequately responded to trials of stimulants and antidepressants. She also constantly complained of feeling cold, and her husband noted that her skin felt cold. Thyroid function tests were normal.
Within 2 weeks of treatment with DDAVP nasal spray, she no longer felt cold and her husband noted that her skin was warmer. Improvement has persisted for the past 6 months.
Patient 6. This 46-year-old Caucasian woman, while on a plane ride of several hours, sustained bilateral anterior thalamic infarcts. MRI revealed areas of increased signal in the anterior portions of both thalami. In addition, there were punctate areas of hypointensity within the left thalamus.
She was referred for treatment of depression, tearfulness, and decreased motivation. After the injury, she also believed that her husband was an imposter. This symptom gradually improved, although problems with mood and motivation did not significantly respond to trials of sertraline, bupropion, selegiline, methylphenidate, dextroamphetamine, or sodium valproate. She also complained of persistently feeling cold.
She initially claimed that these feelings improved 1 day after starting DDAVP. Several months later, she stated that the cold feelings were still present, but not as severe as prior to DDAVP. It later was not clear if there had been any significant improvement with DDAVP. In addition, her husband noted that she had always felt cold, although it became more severe after the injury.
Physiology of Vasopressin
Vasopressin, which is also known as antidiuretic hormone, is produced by the paraventricular and supraoptic nuclei of the anterior hypothalamus as a preprohormone, is changed by Golgi apparatus to a prohormone, and is then transported in vesicles to axon terminals. Most of the axons carrying vasopressin are within the supraopticohypophyseal tract and terminate in the posterior pituitary, where vasopressin can be released into the general circulation. Other axons terminate in the median eminence of the basal hypothalamus, where vasopressin regulates the secretion of adrenocorticotropic hormone, and in the cerebrospinal fluid of the third ventricle.3
Vasopressin has several physiological effects, including conservation of water by concentrating urine (through its effect on the kidney) and promoting vasoconstriction in high concentrations. Because of these effects, the most common clinical uses are in the treatment of diabetes insipidus, nocturnal enuresis (with vasopressin nasal spray), and bleeding esophageal varices (with intravenous preparations). Vasopressin may also have a role in facilitating learning and memory.
Release of vasopressin is generally controlled by osmoreceptors in the hypothalamus. A decrease in plasma volume also causes release of vasopressin. Several neurotransmitters have effects on vasopressin levels. Acetylcholine and angiotensin II stimulate release of vasopressin. Norepinephrine, atrial natriuretic peptide, and gamma-aminobutyric acid are inhibitory.3
Vasopressin and Temperature Regulation
Temperature regulation is controlled by the hypothalamus.4 Heat receptors are found in the anterior hypothalamus. The preoptic area of the anterior hypothalamus responds to increased temperature by releasing vasopressin to prepare for an increase in sweating. This area stimulates sweating and inhibits heat production from shivering and chemical thermogenesis. Cold receptors are located in the posterior hypothalamus. In response to decreased temperature, the posterior hypothalamus stimulates the sympathetic nervous system and results in shivering, vasoconstriction, and increased generation of heat.
The main effect of vasopressin on this system is to inhibit thermoregulatory heat production.5 Increasing hypothalamic temperature stimulates vasopressin release, whereas decreasing hypothalamic temperature inhibits its release.6 Vasopressin decreases the brain temperature of febrile more than nonfebrile rats, and it may act on the febrile set point (an antipyretic effect), but it does not decrease the normal temperature set point.7 Exposure of normal human subjects to acute cold stress resulted in inhibition of vasopressin release.8
Temperature Regulation After TBI
Fever is the most commonly observed temperature disturbance after brain injury, especially if brainstem structures are involved. Many fevers are the result of infection related to the injury, since the blood—brain barrier has been disrupted. However, there have been a few reports of chronic thermoregulatory difficulties following TBI. Whyte et al.9 report on a patient with central fevers immediately following TBI who later developed occasional core body temperature elevations in a warm environment. This patient apparently had difficulties with dissipating heat, or possibly with sensing ambient temperature elevation.
The supraoptic nucleus (in the anterior hypothalamus) appears to be the most vulnerable area of the hypothalamus to traumatic injury.10 Autopsy studies have shown that a high number of patients who die immediately following TBI have hypothalamic and pituitary damage.11 In evaluating 106 cases of fatal brain injury, Crompton12 found 45 hypothalamic lesions; there were microhemorrhages in 31 cases and ischemic necrosis in 26 cases. Traumatic and experimental lesions of the anterior hypothalamus result in hyperthermia.4,13
Hypothesis of Mechanism of Action of Vasopressin
Ischemic areas of the brain may have higher temperatures than core body temperatures.14 It is possible that if this occurs in the anterior hypothalamic area, the body receives the incorrect perception of fever. However, because of damage to the anterior hypothalamus, there is insufficient vasopressin to regulate the appropriate response, with a resultant feeling of persistent coldness. In addition, the general sensation of feeling cold may inhibit vasopressin release.
The administration of exogenous vasopressin may act in two ways to reset the hypothalamic thermostatic mechanism. Vasopressin may decrease local brain temperature, enabling the hypothalamus to adjust to an appropriate set point. In addition, a brief exposure to vasopressin may reset the hypothalamic feedback mechanisms where vasopressin release has been inhibited by the sensation of coldness.
Although vasopressin has been used previously for treatment of cognitive deficits after TBI, with variable success,15 this is only the second reported trial of vasopressin for treatment of subjective feeling of coldness after brain injury.2 No patients in our series reported side effects due to DDAVP.
High doses of intranasal DDAVP have been reported to cause transient headaches and nausea. Other side effects occasionally noted are nasal congestion, rhinitis, flushing, and abdominal cramps. In susceptible individuals who are at risk for water intoxication and hyponatremia, fluid intake must be adjusted. Some medications, such as the selective serotonin reuptake inhibitors, may increase the risk of hyponatremia in these individuals.
Several characteristics of this group of patients deserve comment. First, all patients reported in this paper are women. We do not know whether this phenomenon is more common in women, but 9 of the 13 individuals in the report by Eames were male.2 Thus, there may be some selection bias in our sample. Second, temperature regulation is complex, and it is mediated by many neurotransmitters, including dopamine.16 Several patients were on dopaminergic medications during administration of DDAVP (bupropion, selegiline, dextroamphetamine), and this may have affected response. Third, the single patient in this series who did not respond had thalamic lesions. The absence of observable hypothalamic involvement, or the nontraumatic nature of this injury, may have prevented response to DDAVP.
It is striking that the beneficial effects of DDAVP persisted after the treatment period ended. We speculate that DDAVP may reverse physiologic effects of a relative deficit in vasopressin in the hypothalamus that was caused by injury to the vasopressin precursor—producing cells in the anterior hypothalamus. DDAVP thus may correct an internal temperature set point disrupted by the brain injury.
We found vasopressin to be an effective treatment for this distressing symptom in this series of patients. Future work should include placebo-controlled trials of the effects of vasopressin that incorporate clinical measurements of body temperature and peripheral vasopressin effects such as urine volume, serum osmolality, and serum sodium.
This research was presented at the ninth annual meeting of the American Neuropsychiatric Association, Honolulu, HI, February 1—3, 1998.