خواب و رفتارهای خودآسیبی ذهنی و عینی در کودکان: یک مطالعه کلی جمعیت

Abstract Significant association between sleep disturbances and suicidal ideation and/or attempts is reported in adults and adolescents. However, there is paucity of studies exploring the association between sleep and self-harm behaviors (SHB) in young children and are limited to only subjective sleep measures. We examined the association between SHB and both subjective and objective sleep in a population-based sample of 5–12 yr old. Parents of every student in 3 local school (K-5) districts (n=7312) was sent a screening questionnaire. Randomly selected children from this sample underwent a comprehensive history, physical examination, a 9-h overnight polysomnogram and completed several questionnaires. Among the final sample (n=693), 27 children had SHB with adjusted prevalence of 3%. There was no difference in age, gender, obesity, or socioeconomic status in subjects with or without SHB. Significantly more children with SHB had subjective sleep difficulty and depression. Difficulty maintaining sleep and frequent nightmares were associated with SHB independent of depression or demographics. Polysomnographic %REM-sleep was significantly higher in the SHB group after adjusting for demographics and depression. These data indicate that parent reported sleep disturbances are independently associated with SHB. It is possible that higher REM-sleep is a non-invasive biomarker for risk of self-harm behaviors in young children.

. Introduction Suicide is the tenth leading cause of death in the United States (Kochanek et al., 2011). According to the 2009 National vital statistics report, there were 36,547 deaths from suicide with a death rate of 11.9 for 100,000. The number of deaths from suicide in the 5–14 age group were 266 in the year 2009 in USA, with a suicide death rate of 0.7 for 100,000 (Kochanek et al., 2011). For every completed suicide, several non-lethal suicide attempts occur (Pfeffer, 1988 and Maris, 2002). Suicidal thoughts and attempts in children, collectively can not only lead to mortality but also can cause significant morbidity as a result of non-lethal injury (Pfeffer, 1997 and Doshi et al., 2005). Several factors are known to increase the risk of suicidal ideation and/or attempts including sleep disturbances (Ialongo et al., 2004 and Bernert and Joiner, 2007). Numerous studies in adults suggest strong association between sleep disturbances and suicidal ideation, suicidal attempts and completed suicide. (Fawcett et al., 1990, Turvey et al., 2002, Fujino et al., 2005, McCall et al., 2010 and Bjørnaard et al., 2011; Bjørngaard et al., 2011). Similarly, several studies in adolescents found an association between suicide and sleep disturbances (Tishler et al., 1981, Choquet and Menke, 1989, Choquet and Kovess, 1993, Vignau et al., 1997, Roberts et al., 2001, Bailly et al., 2004, Liu, 2004 and Barbe et al., 2005; Bernert and Joiner, 2007, Fitzgerald et al., 2011 and Lee et al., 2012). However, most of these studies in children were in subjects aged 13 years or more except for the study by Roberts et al. (2001) and Barbe et al. (2005). Roberts et al. (2001) found that in children (age 10 to 17 yr) insomnia or hypersomnia increased risk of suicidal ideation. Barbe et al. (2005) examined depressed children of 7 to 17 years of age and found that depressed suicidal children presented more frequently with insomnia. Both of these studies had a sample of children that were relatively younger; however, to our knowledge none have examined the association of sleep and self-harm behaviors in general population sample of children younger than 10 years old. Additionally, the sleep disturbances were assessed by only subjective report in these two studies with relatively younger children (Roberts et al., 2001 and Barbe et al., 2005). However, the subjective report of sleep is limited in comparison to sleep assessment as done by a comprehensive polysomnogram. An objective polysomnogram provides data on physiological sleep measures such as rapid eye movement (REM) sleep and non-rapid eye movement sleep (NREM) along with objective measures of sleep latency, REM-latency, sleep efficiency, etc. Thus, it is important to assess both subjective and objective polysomnographic sleep in children with self-harm behaviors. Sleep disturbances are common in children with a prevalence of 20–30% (Stores, 1996, Anders and Eiben, 1997, Liu et al., 2000, Owens et al., 2000, Sadeh et al., 2000 and Singareddy et al., 2009). In this study we examined the relationship between subjective and objective polysomnographic sleep and self-harm behaviors in young children aged 5–12 yr. We hypothesize that young children with self-harm behaviors will have increased subjective and objective sleep disturbances.

3. Results Among the final sample of 693 children, 27 had self-harm behaviors (SHB) and 666 children did not have self-harm behaviors (No-SHB). The adjusted prevalence of self-harm behaviors in this general population sample was 3%. Children in the two groups (SHB versus No-SHB) did not differ in age, gender, obesity, or socioeconomic status (Table 1). Children with self-harm behaviors had significantly higher T scores on depression (P<0.001)( Table 1). Table 1. Characteristics of study sample. SHB (n=27) No-SHB (n=666) P Females, % 54.2% 52% 0.82 Age 8.82 (1.9) 8.77 (1.7) 0.91 ⁎SES, % 38.9% 45.5% 0.59 Obesity, % 25.3% 14.1% 0.16 Dep (T scores) 59.45 (7.9) 53.76 (5.8) <0.001 SHB=self-harm behaviors; SES=socio economic status; Dep=depression T scores. All data are adjusted for sampling weight. ⁎ We had data on only 518 subjects on SES. Table options Subjective sleep disturbances in children with or without self-harm behaviors are presented in Table 2. Significantly more children with self-harm behaviors had difficulty initiating sleep (P<0.001), difficulty maintaining sleep (P<0.001), excessive daytime sleepiness (P<0.001), and frequent nightmares (P<0.001). Multivariable logistic regression model showed that difficulty maintaining sleep and frequent nightmares were associated with self-harm behaviors even after controlling for depression (see Table 3). The association between difficulty maintaining sleep and frequent nightmares with self-harm behaviors remained strong and significant even after further controlling for age, gender, obesity or socioeconomic status. Difficulty initiating sleep and excessive daytime sleepiness were not significantly associated with SHB after adjusting for depression and any of the above mentioned demographic factors. Table 2. Subjective sleep characteristics of study sample. SHB (n=27) (%) No-SHB(n=666) (%) P Subjective sleep DIS 42.2 13.3 <0.001 DMS 68.5 21.3 <0.001 EDS 11.8 2.4 0.010 Nightmares 26.3 5.5 <0.001 Any sleep comp 83.1 27.1 <0.001 SHB=self-harm behaviors; DIS=difficulty initiating sleep; DMS=difficulty maintaining sleep; EDS=excessive daytime sleepiness; Any sleep comp.=presence of any of the sleep complaints (DIS, DMS, EDS or frequent nightmares). All data are adjusted for sampling weight. Table options Table 3. Multivariable logistic regression models to examine the association between subjective sleep disturbances with self-harm behaviors. Model 1 Model 2 Model 3 Model 4 P OR P OR P OR P OR DIS 0.07 2.79 0.08 2.75 0.06 2.84 0.12 2.50 DMS 0.01 6.92⁎ 0.001 7.40⁎ <0.001 7.59⁎ 0.001 7.01⁎ EDS 0.20 3.20 0.25 2.88 0.26 2.73 0.43 2.25 Nightmares 0.004 5.58⁎ 0.003 5.95⁎ 0.003 6.04⁎ 0.012 4.90⁎ All data are adjusted for sampling weight. DIS=difficulty initiating sleep, DMS=difficulty maintaining sleep, EDS=excessive daytime sleepiness; Model 1: DIS, DMS, EDS, or nightmares adjusted for age and depression; Model 2: DIS, DMS, EDS, or nightmares adjusted for gender and depression; Model 3: DIS, DMS, EDS, or nightmares adjusted for obesity and depression; Model 4: DIS, DMS, EDS, or nightmares adjusted for socioeconomic status and depression. Table options Adjusted polysomnographic measures in subjects with and without self-harm behaviors are presented in Table 4. Percent of REM sleep was significantly higher in those with SHB (P=0.045), even after adjusting for age, gender, obesity, socioeconomic status, and depression. Surprisingly, there was a trend towards increase of total sleep time in children with self-harm behaviors. Sleep latency, REM latency, and other polysomnographic measures including apnea hypopnea index were not significantly different between the groups. Table 4. Objective sleep characteristics in subjects with and without self-harm behaviors (SHB) adjusted for age, gender, obesity, socioeconomic status and depression. SHB (n=27) No-SHB (n=666) P Objective sleep Sleep latency 24.79 (29.4) 28.84 (24.3) 0.535 REM latency 147.31(70.4) 157.51 (85.3) 0.590 Total sleep time 478.61 (57.5) 455.64 (47.5) 0.072 Sleep efficiency 89.13 (10.0) 85.56 (8.2) 0.106 % Stage 1 2.04 (4.2) 3.62 (3.4) 0.088 % Stage 2 46.48 (13.5) 46.00 (11.2) 0.871 % SWS 28.54 (13.0) 30.50 (10.7) 0.496 % REM 22.95 (6.7) 19.99 (5.5) 0.045⁎ WASO 33.77 (40.9) 48.07 (33.7) 0.115 AHI 0.85 (2.1) 0.80 (1.7) 0.905 All data are adjusted for age, gender, obesity, socioeconomic status, depression, and sampling weight. The standard deviation is presented in parenthesis (S.D.). AHI=apnea hypopnea index.