Prevalence and Risk of Sleep Disturbances in Adolescents After Minor Head Injury Giora Pillar, MD, PhD, Eran Averbooch, MD, Neri Katz, MD, Nir Peled, MD, PhD, Yuval Kaufman, MD, and Eli Shahar, MD Sleep disturbances were reported in patients during the acute stage after minor head injury, and for some of these patients, the disturbances may become chronic. The purpose of the present study was to assess the prevalence and risk factors of the long-term sleep disturbances in adolescents after minor head injury. Unselected adolescents (98) who had experienced a minor head injury 0.5– 6 years before the institution of the study and 80 matched control subjects were interviewed and completed a detailed questionnaire. The prevalence of sleep disturbances was significantly larger among adolescents who experienced minor head injury compared with the control subjects (28% versus 11%, P < 0.05). Within the study group, those who developed long-term sleep disturbances manifested a greater body mass index (20.8 ⴞ 4.0 vs 18.4 ⴞ 2.8 kg/m, P ⴝ 0.005) and poorer parental education (fathers 11.0 ⴞ 4.0 vs 13.4 ⴞ 3.0 years, mothers 11.8 ⴞ 3.3 vs 13.2 ⴞ 2.9 years, P < 0.05 for both), compared with those who did not develop sleep disturbances. Our data indicate that subjective sleep disturbances may be evident in a fairly high percentage of adolescents after minor head injury, up to 28%, suggesting that minor head injury may not be as benign as previously estimated. Risk factors include heavier body mass and poorer parental education. © 2003 by Elsevier Inc. All rights reserved. Pillar G, Averbooch E, Katz N, Peled N, Kaufman Y, Shahar E. Prevalence and risk of sleep disturbances in adolescents after minor head injury. Pediatr Neurol 2003; 29:131-135. From the Sleep Laboratory, Rambam Medical Center and TechnionIIT, and Child Neurology Unit, Rambam Medical Center, Haifa, Israel. © 2003 by Elsevier Inc. All rights reserved. doi:10.1016/S0887-8994(03)00149-8 ● 0887-8994/03/$—see front matter Introduction Minor head injuries (MHIs) are considered frequent events among children and adolescents. In the United States alone, approximately 100,000 children are hospitalized every year after head injury [1]. An MHI is defined as a brain concussion commonly associated with transient loss or impairment of consciousness (for fewer than 30 minutes), vomiting, and a short anterograde or retrograde amnesia. The estimated Glasgow Coma Scale score in MHI is 13 or higher [2]. Results of a thorough investigation of patients with MHI, including computed tomography and magnetic resonance imaging scans, electroencephalogram, brainstem auditory-evoked responses, electronystagmograms, and audiograms, are commonly normal or may reveal minor transient abnormalities [2]. Most of these patients, if hospitalized, are discharged within 24 hours. However, although considered a benign event, MHI may be associated with long-term complications such as post-traumatic stress disorder (PTSD) or postconcussive syndrome [1,2], both of which are frequently associated with sleep disturbances. The postconcussive syndrome consists of any combination of headaches, memory and concentration impairments, anxiety, mood disorders, and sleep disorders [2]. It may overlap with PTSD, which is mainly a mental disorder that consists of flashbacks, nightmares, and inability to cope with daily life events. This syndrome that primarily affects adults include two types of sleep disturbances incorporated in the diagnostic criteria (based on the Diagnostic and Statistical Manual of Mental Disorders [DSM-IV R]): re-experiencing events (nightmares, criteria B) and a hyperarousal state (difficulty initiating and maintaining sleep, criteria D). Some researchers, on the other hand, believe that MHI may not induce a PTSD because the head injury is commonly associated with a loss of consciousness and short-term amnesia [3]. Communications should be addressed to: Dr. Pillar; Sleep Laboratory, Rambam Medical Center; Haifa 31096, Israel. Received September 24, 2002; accepted February 17, 2003. Pillar et al: Sleep Disturbances After Minor Head Injury 131 Adolescents and young adults who experienced MHI may manifest sleep disorders without necessarily exhibiting other symptoms of postconcussive syndrome or PTSD, although the incidence or prevalence of both syndromes after MHI has not been fully clarified. Subjective evaluation of individuals after MHI revealed complaints of difficulties in initiating and maintaining sleep, earlymorning awakenings, decreased ability to function, and a generally decreased sleep quality [4,5]. When these patients complained of nightmares as well, the diagnosis of PTSD-related sleep disorder was established, although in some children PTSD may exist even without nightmares [6]. Only few studies have attempted to objectively measure sleep-behavior patterns after MHI. For the acute phase after MHI, some polysomnographic findings have been reported [7-10]. However, long-term effects of minor head injury on sleep patterns in adolescents have not been consistently studied. Adolescents who complained of long-term sleep disturbances after MHI, both at their home and in the laboratory, were studied using objective methods [11]. Findings indicate objective evidence of insomnia in these adolescents. However, the prevalence of such a phenomenon has not been fully clarified. Thus, the purpose of the present study was to assess the prevalence of long-term sleep disturbances after MHI in adolescents and to identify risk factors for their occurrence. Table 1. General comparison between study and control groups Study group Control group n Age (Range), Years Boys (%) BMI (Range), kg/m2 98 80 13.5 ⫾ 2.3 (8-18) 12.4 ⫾ 2.5 (8-18) 68 57 19.0 ⫾ 3.2 (14.0-28.8) 18.8 ⫾ 2.8 (10.6-27.2) None of the comparisons revealed significant differences between the groups. Abbreviations: BMI ⫽ Body mass index n ⫽ Number of participants and parental age, education, and socioeconomic status), general health information, complaints associated with the head injury (for the study group only), and evaluation of sleep quality. All quantitative questions instructed the participant to grade answers from 0 (never) to 4 (always). Prespecified criteria to diagnose sleep disorder were determined (for this study) and considered whenever the summary score of the following eight questions exceeded 12: 1. 2. 3. 4. 5. 6. 7. 8. Difficulties falling asleep Frequent awakenings from sleep Frequent movements in sleep Early-morning awakening Anxiety/fear on arousals from sleep Nonrestorative sleep Bad dreams/nightmares Daytime somnolence Patients and Methods Study Protocol and Data Analysis Patients The patients were selected from the archives of the Rambam Medical Center in Haifa, a tertiary hospital, the largest in the north of Israel. A list was formed of all patients aged 7 to 15 years who were admitted with MHI between 1993 and 1999 (ICD-10 code S06.0), having manifested an estimated Glasgow Coma Scale of 13 or above on admission to the emergency room. Three hundred candidates were randomly selected, half of them as the study group (every other name from the list). The number of participants was further reduced because of either address change or language obstacles. The study group was therefore composed of 98 patients. Healthy control subjects at the same ages were randomly selected at local elementary school classes and consisted of a group of 80 participants with no known history of MHI. This group was selected randomly without any sampling bias or any knowledge regarding their sleep characteristics and therefore was representative of the prevalence of sleep disorders in the general population at this age range in Israel. All participants (study and control groups) filled out a specifically designed questionnaire. To ensure compliance, one of the investigators met each of the participants and supervised him or her during the process of filling out the questionnaire. The controls filled out the questionnaire at school, again with one of the investigators present in the class for assistance. The age of the participants at the time of the questionnaire completion ranged between 8 and 18 years. The study was approved by the Human Subject Committee of Rambam Medical Center. Questionnaire The questionnaire consisted of 60 questions detailing demographic personal data (age, sex, race, religion, height, weight, etc.), family data (including marital status of the parents; number of brothers and sisters, 132 PEDIATRIC NEUROLOGY Vol. 29 No. 2 All patients were investigated by a detailed questionnaire in an attended fashion, and the collected data represented the status of the participants at the time of questionnaire completion. Only a few questions tried to retrospectively assess the impact of head injury. Two-tailed unpaired t tests for independent samples were used to compare questionnaire data between the study and control groups. Chi-square comparison for proportions was used to compare prevalence of sleep disturbances among the study and control groups. Thereafter, the study group was divided into two subgroups: those with sleep disturbances (summary score exceeding 12 to the eight questions presented above) and those without sleep disturbances. Two-tailed unpaired t tests for independent samples were used to assess potential risk factors for developing sleep disturbances after MHI. Correlation analyses were performed to assess the relationships between selected sleep complaints and potential etiologic factors. P ⬍ 0.05 was considered statistically significant in each comparison. Results The age, sex, and body habitus distribution were similar in the MHI (study) group and the control group (Table 1). In addition, no statistical differences were evident regarding race or religion, number of family members, socioeconomic status, or parental marital or educational status. The average Glasgow Coma Scale for the study group on admission was 14.7 ⫾ 0.6. According to the prespecified cutoff for the definition of sleep disturbances (see Methods), the study group consisted of a significantly greater percentage of post-MHI adolescents who complained of Table 2. Discussion Sleep schedule regulation of the two groups Nocturnal Sleep Nocturnal Sleep Duration Weekdays Duration Weekends Habitual (Range), Hours (Range), Hours Napping Study group Control group P 8.2 ⫾ 1.2 (4-11) 8.2 ⫾ 0.9 (7-11) 0.72 9.7 ⫾ 1.8 (5-13) 10.3 ⫾ 1.4 (7-13) 0.013 38% 31% 0.43 sleep disturbances (28% vs 11% for controls, P ⬍ 0.05). The average score of sleep complaints was significantly higher in the study group than the control group (10.1 ⫾ 4.8 vs 8.2 ⫾ 3.2, P ⬍ 0.05). In addition, the study group demonstrated significantly more complaints of sleep bruxism (10% vs 4%, P ⬍ 0.05). There were no differences between the groups with respect to other potential sleep disturbances, such as sleepwalking or nocturnal enuresis. Also, no differences were evident between the groups regarding bedtimes, wakeup times, or sleep times in weekdays (8.2 hours per night for both groups), although the study-group patients reported significantly shorter sleep time during weekends (9.7 ⫾ 1.8 hours vs 10.3 ⫾ 1.4 hours, P ⬍ 0.05, Table 2). Within the study group, the general score of sleep disorders correlated with headaches (R ⫽ 0.45, P ⬍ 0.05) and with general anxiety score (R ⫽ 0.56, P ⬍ 0.05). Moreover, most patients reported that the headaches, anxiety, and sleep problems occurred after head injury (85%). The study group was further divided into two subgroups: those with sleep disturbances (n ⫽ 27, 28%), and those without (n ⫽ 71, 72%). Comparison measures between both subgroups are summarized in Table 3. Body weight was significantly higher in patients with sleep disturbances, and their parents had significantly poorer education, exemplified in fewer years at school. No differences in other dimensions were apparent between both groups (including age, sex, religion, medical diseases, socioeconomic status, number of siblings, or the number of people sleeping in the same room). Table 3. Comparisons between children with and without sleep disturbances, within the children who underwent minor head injury Age (years) Sex (% boys) Body mass index, kg/m2 Paternal education (years) Maternal education (years) Number of brothers/sisters Study Patients With Sleep Disturbances Study Patients Without Sleep Disturbances P 14.2 ⫾ 2.1 70 20.8 ⫾ 4 11.0 ⫾ 4 11.8 ⫾ 3.3 3.3 ⫾ 1.2 13.2 ⫾ 2.3 67 18.4 ⫾ 2.8 13.4 ⫾ 3 13.2 ⫾ 2.9 2.9 ⫾ 1.4 NS NS 0.005 0.002 0.049 0.176 Children who developed sleep disturbances were heavier, and their parents had fewer years of education. Abbreviation: NS ⫽ Not significant The present study reveals that 28% of adolescents who experienced MHI exhibited long-term sleep disturbances (6 months to 6 years). The prevalence of similar sleep complaints in matched control subjects was only 11%, which is statistically different, suggesting that the outcome after minor head injury may not be as benign as previously assumed. Also, risk factors for prolonged persistence of sleep disturbances after MHI included a higher body mass index as well as significantly shorter duration of maternal and paternal education. Other symptoms commonly associated with emotional stress, such as anxiety, headaches, and bruxism, are also more prevalent in adolescents who experienced MHI compared with control subjects. The primary aim of this study was to assess the prevalence of long-term sleep disturbances in adolescents after MHI. Almost one third of adolescents complained of sleep disturbances after such an event, compared with only 11% of age-, sex-, religion-, and socioeconomic statusmatched controls. The 11% prevalence of complaints about sleep in the general population of adolescents is not surprising and is in concert with previous reports in Israel [12]. However, a 2.5-fold greater prevalence after a minor head injury that occurred 0.5 to 6 years before the present study is indeed surprising and challenging. Several explanations may account for the high prevalence of sleep disturbances in adolescents with MHI. Many adolescents reportedly develop sleep disturbances within the acute stage of MHI [13,14]. With time, when sleep disturbances commonly improve [13-15], these adolescents may have already developed secondary emotionally derived insomnia [15]. In this disorder, the acute stress that leads to insomnia, instead of resolving with time, leads to substantial concern and distress now focusing on the sleep itself. Ultimately, a vicious cycle of poor sleep leading to further anxiety regarding insomnia is created. This explanation is also supported by the increased prevalence of bruxism in the study group, which may result from stress, and result in sleep fragmentation and further nonrestorative sleep. Another explanation may be a premorbid personality or character, which consequently may result in sleep disturbances after a relatively minor trauma. There were no relevant data regarding either sleep patterns or personality characteristics of the patients before the MHI, and therefore this explanation cannot be assessed. A third possible explanation is a substantial discrepancy between subjective complaints and objective findings (i.e., that their sleep may be objectively normal). Such an occurrence is still worrisome because insomnia is defined as the perception of inadequate quantity or quality of sleep, with associated daytime disturbances. The definition of insomnia relies on the perspective of the patient, and therefore diagnostic testing is not generally required because sleep history alone is sufficient to establish the diagnosis. Subjective vs objective complaints in adolescents with sleep disturbances after MHI were recently compared [11] and, Pillar et al: Sleep Disturbances After Minor Head Injury 133 although tending to be exaggerated, the subjective complaints were largely confirmed by objective sleep measures in the patients’ homes and in the laboratory. However, other studies focusing on the functional consequences of trauma have detected various discrepancies between subjective complaints and objective sleep disorders, especially in insomniacs with PTSD [15-17]. The evolution of an accompanying PTSD may also explain the high prevalence of sleep disturbances in these adolescents, although a full-blown diagnosis of PTSD could be established on clinical grounds in this study’s patients. Despite the sleep complaints and reported disrupted sleep patterns, most of the patients managed to properly function in everyday life, although a few complained of serious functional difficulties because of increased daytime somnolence, which prevented them from attending school and was associated with learning and concentration difficulties, increasing anxiety, and headaches and therefore required medical attention. Regardless of the explanation for the increased incidence of sleep disturbances in the examined adolescents, these results indicate a substantial and non-negligible impact of the MHI associated with long-term sleep disturbances. Although the findings reveal significant insomnia, the patients had severe additional sleep complaints, including high frequency of movements during sleep, bruxing teeth during sleep, and daytime sleepiness. An important question is whether these characteristics are the consequences of a direct trauma to the head or may generally develop after any trauma, resulting from the emotional distress. One way to try to delineate the question is by comparing the neurobehavioral sequelae of head-injured and otherwise bodily-injured patients who sustained minor injury. Such a study was performed by Farmer et al. [18], but sleep abnormalities were not specifically addressed. Other studies referring to adolescents and young adults who had undergone trauma other than MHI, including road traffic accidents without head injury [19], child abuse [20], or burns [21], have also reported sleep disturbances to some extent. Thus, although the origin of the sleep problems in MHI is difficult to determine precisely, it may likely be related to overall emotional stress rather than direct traumatic impact of cerebral cortex. This supposition is also supported by findings that the insomniac complaints were also associated with anxiety, headaches, and bruxism. Another surprising finding in the present study is a significantly shorter sleep time during weekends (9.7 ⫾ 1.8 hours vs 10.3 ⫾ 1.4 hours, P ⬍ 0.05, Table 2) reported by the study group patients. If indeed their sleep is disrupted during weekdays, their time in bed during weekends would be expected to increase. Previous studies on sleep in adolescents have demonstrated that a shorter lag between sleep during weekdays and weekends is associated with better school performance [22]. Also, students with lower grades reported greater delays of their sleep schedules over weekends [23]. Thus, generally 134 PEDIATRIC NEUROLOGY Vol. 29 No. 2 keeping a similar schedule and total sleep time in weekends may be advantageous. However, in the case of this study, adolescents from the study group apparently did not elect to sleep less during weekends but rather manifested the inability to extend their sleep during weekends despite a desire to do so (i.e., spontaneous early-morning awakening). Whether this affected school performance remains speculative because it was not specifically assessed in the present study. Within the MHI group, sleep complaints were observed in heavier adolescents and in those whose parents had fewer years of education. The reason why heavier adolescents are at increased risk for the development of longterm sleep complaints after MHI has apparently not been previously reported and remains unknown. Heavier adolescents may experience lower self-esteem and therefore be prone to develop sleep disturbances after MHI [24,25]. Regarding parental education, several studies have outlined the association between poorer parental education and decreased sleep quality of their children [12]. Perhaps parents with poorer education are more prone to develop emotional distress and anxiety and hence unable to implement techniques of addressing the overwhelming emotional stress. This study has several limitations. First, the response rate for participation was approximately two thirds (98 participants of 150 candidates). Theoretically, one may assume that those who manifest sleep disturbances will more likely agree to participate than those who remain asymptomatic, which may artifactually increase the prevalence of sleep disturbances in this population. However, the distortion factor is unimportant because most individuals quit the study mainly because of technical reasons. In fact, the few individuals who intentionally chose not to participate in the study indicated that they actually did experience sleep disturbances but refused to participate because they were in a process of litigation and insurance claims. Second, although this study was prospective with respect to the protocol and the questionnaires introduced to the study group and control group, it was retrospective with respect to analyzing the clinical aspects of the MHI. As such, there are no data regarding the participants’ sleep patterns before the MHI and there is no full proof that the complaints are the direct consequence of MHI. However, the latter is probably the case because a relatively wellmatched control group was chosen, and the same methods were applied in both groups. Thus, with the MHI being the single factor distinguishing between the groups, the sleep complaints can indeed be associated with the MHI. However, a better prospective study, to begin with patient follow-up at the onset of MHI, is needed to confirm the results. In summary, despite these limitations, this study is important and the results are valid. To date, health professionals usually perceive MHI as a minor and negligible transient event with no long-term consequences. Neurologic examination and imaging results are usually normal, and most patients are discharged from the emergency wards and receive no further medical attention. The abnormal sleep patterns detected in the present study suggest that MHI is not necessarily as benign as considered previously. Therefore, awareness of possible sleep disturbances should be taken into account in any adolescent who has undergone MHI. Treatment should be given before learning disability and impaired development occur. References [1] Haslam RHA. Head injuries. In: Nelson WE, Behrman RE, Kliegman RM, Arvin AM, eds. Nelson textbook of pediatrics, 15th ed. Philadelphia: W.B. Saunders Company, 1996:1719-23. [2] Kaufman DM. Head trauma. In: Kaufman DM, ed. Clinical neurology for psychiatrists, 4th ed. Philadelphia: W.B. Saunders Company, 1995:545-8. [3] McMillan TM. Minor head injury. Curr Opin Neurol 1997;10: 479-83. [4] Parsons LC, Ver Beek D. Sleep-awake patterns following cerebral concussion. Nurs Res 1981;31:260-4. [5] Perlis ML, Artiola L, Giles DE. Sleep complaints in chronic postconcussion syndrome. Percept Mot Skills 1997;84:595-9. [6] Terr LC. Psychic trauma in children and adolescents. Psychiatr Clin North Am 1985;8:815-35. [7] Lenard HG, Pennigstroff H. Alterations in the sleep patterns of infants and young children following acute head injuries. Acta Pediatr Scand 1970;59:565-71. [8] Harada M, Minani R, Hattori E, et al. Sleep in brain damaged patients: An all-night study of 105 cases. Kumamoto Med J 1976;29: 110-27. [9] Enomoto T, Ono Y, Nose T, et al. Electroencephalography in minor head injury in children. Childs Nerv Syst 1986;2:72-9. [10] Parsons LC, Crosby LJ, Perlis M, et al. Longitudinal sleep EEG power spectral analysis studies in adolescents with minor head injury. J Neurotrauma 1997;14:549-59. [11] Kaufman Y, Tzichinski O, Epstein R, et al. Sleep disorders in children in the long term after minimal head injury. Pediatr Neurol 2001;24:129-34. [12] Sadeh A, Raviv A, Gruber R. Sleep patterns and sleep disruptions in school age children. Dev Psychol 2000;36:291-301. [13] Hefez A, Metz L, Lavie P. Long-term effects of extreme situational stress on sleep and dreaming. Am J Psychiatry 1987;144: 344-7. [14] Casey R, Ludwig S, McCormick MC. Morbidity following minor head trauma in children. Pediatrics 1986;78:497-502. [15] Dagan Y, Bleich A, Lavie P. Elevated awakening thresholds in sleep stage 3– 4 in war-related PTSD patients. Biol Psychiatry 1991;30: 618-22. [16] Horton AM. Posttraumatic stress disorder and mild head trauma: Follow-up of a case study. Percept Mot Skills 1993;76:243-6. [17] Pillar G, Malhotra A, Lavie P. PTSD and sleep: What a nightmare. Sleep Med Rev 2000;4:183-200. [18] Farmer MY, Singer HS, Mellits ED, et al. Neurobehavioral sequelae of minor head injuries in children. Pediatr Neurosci 1987;13: 304-8. [19] Kramer M, Kinney L. Sleep patterns in trauma victims with disturbed dreaming. Psychiatr J Univ Ottawa 1988;13:12-16. [20] Ellis A, Stores G, Mayou R. Psychological consequences of road traffic accidents in children. Eur Child Adolesc Psychiatry 1998;7: 61-8. [21] Gold CA, Teicher MH, Hartman CR, et al. Increased nocturnal activity and impaired sleep maintenance in abused children. J Am Acad Child Adolesc Psychiatry 1997;36:1236-43. [22] Hofman WF, Steenhof L. Sleep characteristics of Dutch adolescents are related to school performance: Sleep-wake research in the Netherlands. 1997;8:51-5. [23] Wolfson AR, Carskadon MA. Sleep schedules and daytime functioning in adolescents. Child Dev 1998;69:875-87. [24] Johnson EO, Breslau N. Sleep problems and substance use in adolescence. Drug Alcohol Depend 2001;64:1-7. [25] Taub JM, Hawkins DR. Aspects of personality associated with irregular sleep habits in young adults. J Clin Psychol 1979;35:296-304. Pillar et al: Sleep Disturbances After Minor Head Injury 135