In previous versions of the ICSD and DSM insomnia was often considered “secondary” to other conditions, for example, co-occurring psychiatric or medical conditions. In line with the increasing recognition of the variety of different, and complex, mechanisms by which sleep disturbance can be associated with other conditions (e.g., bidirectional relationships, shared risk) (see Gregory & Sadeh, 2016) such a distinction between “primary” and “secondary” insomnia is no longer made but rather, where present, all co-occurring conditions are coded separately, including insomnia.

The current ICSD-3 nosology for insomnia is simplified compared to earlier versions where a variety of different subtypes of insomnia were described, related to the supposed cause of the symptoms. These subtypes no longer form part of the formal diagnostic classification since it can often be clinically very difficult to determine the precise cause of insomnia (and there are frequently multiple factors which are contributing to both cause and maintenance of symptoms) and evidence to support all these subtypes is lacking. However, they are still described in the most recent ICSD-3 and perhaps serve as useful reminders of possible etiological/maintaining factors related to childhood insomnia problems. For this reason, they are briefly outlined below in Section 3.1.2. In children, behavioral insomnia of childhood (BIC) and anxiety-related insomnia are both common, and perhaps particularly pertinent for children with ADHD, and so will be outlined below.

In addition to BIC and sleep-related anxieties, as discussed above, other potential “causes” of insomnia which are highlighted in the ICSD are summarized below as they provide a useful framework for clinically relevant areas of enquiry at the assessment stage.

We lack high quality prevalence data on insomnia disorder in both typically developing young people and those with ADHD, making comparison between samples problematic. Reported rates for typically developing children range between 4% and 41% (Archbold, Pituch, Panahi, & Chervin, 2002; Owens & Mindell, 2011; Paavonen et al., 2000; Zhang et al., 2009) primarily reflecting the different definitions of insomnia which have been used and also the different inclusion criteria across studies. Further, most studies describe rates for symptoms of sleeplessness, only some of which may be due to insomnia. For example, in children and adolescents with ADHD bedtime resistance has been reported in 20%–31% and difficulty falling asleep in 11%–42% (Blader, Koplewicz, Abikoff, & Foley, 1997; Hvolby, Jørgensen, & Bilenberg, 2009; Stein, 1999) but without a diagnosis of insomnia, other potential “causes” of the difficulty falling asleep may not be ruled out (including, e.g., sleep disorders other than insomnia such as restless legs syndrome or circadian rhythm disturbances). Further, as indicated above, distinctions between different underlying causes of insomnia (such as behavioral factors or anxiety) are generally not made, although clearly have important implications, both for understanding any potential mechanisms underlying the link between the sleep disturbance and ADHD and also for approaches to management (see Chapter 9).

What is clear is that insomnia symptoms in children with ADHD are frequently reported. For example, a meta-analysis of 16 studies of unmedicated children and adolescents with ADHD suggested that parents reported frequent bedtime resistance, difficulties with sleep onset, night waking, and morning waking along with daytime sleepiness (Cortese, Faraone, Konofal, & Lecendreux, 2009). These results were supported by some of the objective measurements obtained via PSG and actigraphy. An instability and night to night variability in sleep has also been suggested by some studies (Gruber, Sadeh, & Raviv, 2000) although not all (Poirier & Corkum, 2015).

Delayed circadian rhythms are reported in individuals with DSWPD (e.g., more than 2 hour delay in the timing of core body temperature nadir and dim light melatonin onset (DLMO) time) (Chang, Reid, Gourineni, & Zee, 2009). A number of factors are proposed to increase the risk for this, such as a reduction in the homeostatic sleep drive (so that readiness to sleep will build more slowly, as seen during adolescence (Crowley, Acebo, & Carskadon, 2007)), having a longer than average circadian period (Micic et al., 2013) (which itself is associated with having an evening chronotype, e.g., Brown et al., 2008), and having an altered relationship between circadian rhythm timing and the sleep period (Okawa & Uchiyama, 2007) so that any phase-shifting effect of daylight is less potent and changes in light sensitivity may also play a role (i.e., with increased sensitivity to evening light delaying the circadian clock or reduced sensitivity to morning light resulting in a failure to advance the circadian clock), as suggested by the results of Aoki, Ozeki and Yamada’s study (2001), where, compared to controls, DSWPD patients showed greater phase-delay effects from light. Genetic factors influence chronotype (Taillard, Philip, & Bioulac, 1999) and also play a role in DSWPD, with higher rates of eveningness in the first- and second-degree relatives of patients with DSPWD than for those of the general population (Ancoli-Israel, Schnierow, Kelsoe, & Fink, 2001). Specific variations of the clock genes PER3 (Archer et al., 2003) and CRY1 (Patke et al., 2017) have also been described in association with DSWPD.

Whilst the exact prevalence of DSWPD in children from the general population is unknown there are estimates of between 0.13% and 3% in adolescents (see Wolfson & O’Malley, 2012). The number of adolescents showing some features of DSWPD is likely much greater, this being a time when a variety of biological parameters serve to increase risk (see Wolfson & O’Malley, 2012). Of note, a large study of 9338 Norwegian adolescents found that adolescents with DSWPD had significantly more symptoms of inattention and hyperactivity/impulsivity than adolescents without DSWPD and that ADHD symptoms were predictive of DSWPD (Sivertsen, Harvey, Pallesen, & Hysing, 2015).

Again, exact prevalence rates of DSPWD in children with ADHD are not available but research, with both adults and children with ADHD, suggests that this is a disorder to which they might be particularly vulnerable. For example, studies of adults with ADHD have suggested a link between ADHD and delayed sleep–wake rhythms, including increased “eveningness” preference, delayed DLMO and CLOCK gene expression and consistent resulting sleep patterns (i.e., longer sleep latencies, later waking up time) (see Baird, Coogan, Siddiqui, Donev, & Thome, 2012; Snitselaar, Smits, van der Heijden, & Spijker, 2017). Studies with children are more limited but raise interesting questions. Similar to studies with adults, a delay of melatonin secretion in children with ADHD compared to controls has been described (van der Heijden, Smits, Van Someren, & Gunning, 2005; van der Heijden, Smits, Van Someren, Ridderinkhof, &, Gunning, 2007) along with increased melatonin levels both day and night in a study that included children and adolescents (Büber et al., 2016). Associations between increased eveningness preference and a delay in sleep onset (with sleep assessed by both subjective and objective measures) has been repeatedly documented in children with ADHD (Gruber et al., 2012; van der Heijden, Stoffelsen, Popma, & Swaab, 2018) but there have been inconsistent findings concerning whether there are actually chronotype differences between children with ADHD and controls with some studies suggesting increased “eveningness” preference in children with ADHD compared to controls (Durmuş, Arman, & Ayaz, 2017; Gruber et al., 2012) and others finding no group differences (van der Heijden et al., 2018). Given that differences in chronotype between children with and without ADHD are not consistently seen, it is important to consider the possibility that any delayed sleep onset in children with ADHD may be due to reasons other than biological rhythm abnormalities (e.g., limit setting problems, bedtime resistance, distractibility, etc.).

Disorders of arousal occurring in childhood are frequently considered as developmentally normal in the absence of any serious consequences, and most resolve spontaneously by adolescence (American Academy of Sleep Medicine, 2014). There is often a strong family history for arousal disorders and a genetic predisposition has been identified (Hublin & Kaprio, 2003; Nguyen et al., 2008). Whilst the precise mechanism underlying these disorders is not known, it seems that both genetic and environmental factors play a role since various common factors can increase the risk for disorders of arousal, in predisposed individuals: these include sleep deprivation (including as a result of co-occurring sleep disorders), which may alter the intensity and quantity of deep NREM sleep, and anything which can increase the likelihood of arousals (including stress, illness, environmental triggers such as sleeping in an unfamiliar or noisy/light environment and, commonly for children, having a full bladder) (see Mahowald & Bornemann, 2011). This highlights the importance of ensuring good sleep and good sleep hygiene and, highly relevant for children with ADHD, treating any other co-occurring sleep disorders in children who are prone to disorders of arousal as this in itself might resolve disorders of arousal (Guilleminault, Palombini, Pelayo, & Chervin, 2003).

In the general population, sleepwalking and sleep terrors tend to be more frequent at their onset (usually early childhood) and to decrease over time with many problems resolving by adolescence/early adulthood. Disorders of arousal are relatively common in typically developing preadolescent children; rates of 17% for sleepwalking in children, peaking about 8–12 years and 1%–6% for sleep terrors, typically resolving before adolescence have been reported (Ohayon, Guilleminault, & Priest, 1999). More precise rates were reported by Partinen and Hublin (2000) in relation to different frequencies of occurrence (e.g., sleepwalking “always or often” in 3% and “now and then” in 3.5%–15% and sleep terrors “always or often” in 1%–3% and “now and then” in 6%–29%). Part of the difficulty with obtaining accurate prevalence rates relates to the different assessment methods and criteria used so that the frequency of episodes are not always accounted for.

There is less information about the prevalence of disorders of arousal in children with ADHD compared to some other sleep disorders, perhaps because many parent report questionnaires frequently provide subscale scores for “parasomnias” (grouping together different types of parasomnias). Some studies have reported a general association between “parasomnias” and ADHD symptoms in young people aged 6–15 (e.g., Gau, 2006), but there are also contradictory findings where rates of parasomnias have been compared between children with and without ADHD, for example, Owens, Maxim, Nobile, McGuinn, and Msall (2000) reported significantly higher rates of parasomnias in children with ADHD compared to controls (aged 5–10 years) whilst Gruber et al. (2012) found no difference in their sample of children aged 7–11 years. Importantly, Lycett, Mensah, Hiscock, and Sciberras (2014) noted that in their sample aged 5–13 years, parasomnias are one of the range of sleep disorders which can present and persist over time, emphasizing the need for broad screening of sleep disturbances in children with ADHD.

Where discrete parasomnias have been investigated research suggests that children with ADHD might have increased risk for both sleepwalking and sleep terrors. For example, in children aged 5–11 years, parent report of frequent sleepwalking was 11.1% for children with ADHD compared to only 1.9% for controls and similarly sleep terrors in 13.3% of the children with ADHD and in only 1.4% of the controls (Hvolby et al., 2009). Silvestri et al. (2007) reported high rates of reported sleepwalking (47.6%) and sleep terrors (38%) in 42 unmedicated children with ADHD (mean age 8.9 years). Although, as might be expected, follow-up video PSG captured lower rates (sleepwalking 2.3% and sleep terrors 4.7%), the study highlights that such specific problems are commonly reported. Results of this study are also noteworthy because interictal epileptiform discharges were also recorded in 53.1% of the ADHD sample, highlighting the need for careful differential diagnosis for episodes occurring in association with sleep (see Chapter 8). Further, an increased percentage of deep NREM sleep was reported in the children in this study. Taken together with the results of Miano et al. (2006) who noted greater instability of sleep in children with ADHD aged 6–13 years compared to controls, these data suggest possible mechanisms whereby children with ADHD might be at increased risk for disorders of arousal.

Interestingly, Gau and Chiang (2009) explored sleep problems in 281 young people aged 10–17 who had been diagnosed with ADHD when they were younger (mean aged 6.7 years), comparing those who still met the criteria for a diagnosis (n=145) with those who now had a subclinical ADHD (n=136) and also with unaffected controls (n=185). They found that adolescents with a childhood diagnosis of ADHD, regardless of whether or not they still met the criteria for an ADHD diagnosis, were more likely to have current and lifetime sleep terrors. The pattern was slightly different for sleepwalking where adolescents with persistent ADHD had a higher likelihood than the controls of having lifetime sleepwalking disorder and adolescents with subthreshold ADHD were more likely than the control subjects to have current sleepwalking disorder. Overall, findings support the idea of increased risk for disorders or arousal in children with ADHD and also suggest that the presence of disorders of arousals in children with ADHD may not be wholly determined by the severity of current ADHD symptoms.

A genetic predisposition to nightmares has been reported (Hublin, Kaprio, Partinen, & Koskenvuo, 1999). In adults and adolescents, nightmares have repeatedly been seen to be associated with various personality traits and psychopathologies (American Academy of Sleep Medicine, 2014; Blagrove, Farmer, & Williams, 2004; Zadra & Donderi, 2000) but these associations may also be mediated by other factors such as distress associated with the nightmare, coping style and nightmare chronicity (see Nielsen & Zadra, 2005). Whether the same is true for children is not as extensively explored; some studies have suggested that elevated general anxiety rates are related to nightmare frequency and severity in childhood and emphasized the role of parenting practices which promote dependent sleeping as a risk factor (Simard, Nielsen, Tremblay, Boivin, & Montplaisir, 2008). Sleep disturbances such as sleep deprivation, sleep fragmentation, or irregular sleep–wake schedules (which can alter the quantity or intensity of REM sleep) may also serve to increase the risk for the occurrence of nightmare (American Academy of Sleep Medicine, 2014). The impact of the content of TV shows/films has been reportedly mixed although parents and children themselves have certainly noted this to play a role (see Moore & Mindell, 2012).

Occasional nightmares are very common in children from the general population but frequent nightmares far less so (1%–5% of preadolescent children), with rates tending to peak about 9 years (Fisher & Wilson, 1987) and decline in nightmare frequency after age 12 (Nielsen & Zadra, 2005). As noted above, a tendency for parasomnias (which includes nightmares along with sleep terrors, sleepwalking, and other parasomnias) to be associated with ADHD symptoms and to be more commonly reported in children with ADHD compared to controls has been documented (Gau, 2006; Lycett et al., 2014; Owens et al., 2000). Limited prevalence data specifically addressing nightmares also indicate that nightmares are more common in children with ADHD than in controls. For example, Grünwald and Schlarb (2017), reported 20.5% fulfilled the ICSD criteria for regular nightmares (2.3% severe) in their sample of 46 unmedicated children with ADHD aged 6–13 years. They also found nightmares to be more common in those children who also had co-occurring insomnia. In a study where parent report of children (aged 4–12 years) with and without ADHD were compared, Hvolby et al.’s (2009) results suggested that the children with ADHD had increased rates (13.3%) compared with controls (1.4%). Interestingly, the percentage of children with ADHD complaining (to parents) of frightening dreams was even higher (37.8%), and much more common than for the control group of children (7.5%) (Hvolby et al., 2009).

There has also been a suggestion that, compared to control children, the dreams of children with ADHD are more negatively toned and include more misfortunes/threats, negative endings, and physical aggression towards the dreamer although other features (e.g., dream recall frequency, dream length, dream bizarreness) were not reported to differ between groups (Schredl & Sartorius, 2010).

Because normal sleep changes across development and some sleep disorders may be more likely to be present (and be noticed by parents) at certain ages, it is important that sleep is considered in a developmental context. Whilst some sleep problems in young children (notably some parasomnias, enuresis, and some sleep-related movement disorders) can typically be expected to be outgrown it is vital to emphasize that this is not the case for all children or all sleep disorders, which fits with the findings of Fisher et al. (2014) who found self-reported sleep problems (mainly difficulty with initiating and maintaining sleep and poor quality sleep) to be present in children, adolescents, and adults with ADHD, but also that the presence of any sleep disturbance, the number of co-occurring sleep problems and complaints of daytime fatigue actually tended to increase from childhood to adolescence and adulthood.

The developmental sleep patterns of children with ADHD have been seen to differ from children without ADHD and differences in sleep between children with and without ADHD have been found to be apparent early in life, before the diagnosis of ADHD had been made. For example in a large birth cohort study of over 8000 children in the United Kingdom, Scott et al. (2013) reported the sleep characteristics of children, as reported by parents at eight time points between when the children were 6 and 140 months old. Compared to the rest of the cohort, the children who were diagnosed with ADHD at age 7 showed reduced night time sleep duration which was most marked during infancy and primary age and increased night waking, particularly after about age 5. Although the absolute group differences in sleep/wake minutes were not large, the tendency for sleep duration to decrease by more than 1 standard deviation between adjacent assessment points was predictive of the presence of ADHD for all but the first two assessments (6–18 months), highlighting that individual developmental trajectories (rather than absolute levels) might be an important way to conceptualize some of the sleep disturbances associated with ADHD.

There are few longitudinal studies examining the impact of age, however, those that there are suggest that age is not a significant risk factor during childhood. Hansen and colleagues (2013) found that baseline levels of sleep problems, and not age, were related to the presence of sleep problems 18 months later (present in 70.6% of the children with ADHD) although their overall “clinical” sample of 7–13 year olds included children with ADHD and anxiety.

In a larger group of children, all with ADHD, Lycett et al. (2014) examined behavioral sleep problem trajectories in children with ADHD aged 5–13 years over a 12-month period, with assessments at baseline, 6 and 12 months. They too found age not to be a significant risk factor associated with the sleep problem trajectories but their results highlighted that a range of sleep problems (including insomnia symptoms, daytime sleepiness, and parasomnias) can be persistent; whilst 41% did not have any significant sleep problems at any time point, 49% presented with transient sleep problems (i.e., a sleep problem at one or two assessment points), and 10% showed sleep problems at all three assessments.

The inherent sex differences in the prevalence of ADHD itself contribute to the difficulty of investigating sex differences in the sleep of children with ADHD, since the number of females is typically considerably less than the number of males in most samples. For example, the prospective cohort study by Lycett et al. (2014) described above, found that sex was not a significant risk factor for sleep problem trajectories, however, over 87% of their sample of 195 were boys. Others too have also found little indication of sex-related sleep differences in children with ADHD (Becker, Pfiffner, Stein, Burns, & McBurnett, 2017; Fisher et al., 2014).

Other work does suggest some sex differences in sleep problems; Becker, Cusick, Sidol, Epstein, and Tamm (2018) examined parent-reported sleep problems in a relatively large sample of primary-aged children with ADHD (n=181, 31% girls). Compared to boys, girls were reported to have poorer sleep across a range of sleep problems (bedtime resistance (but not extended sleep onset latency), anxiety, sleep duration, night wakings, parasomnias, daytime sleepiness, and total sleep problems) and 75% of girls, compared to 53% of boys met the cut-off criteria for a clinically significant sleep problem. However, they also investigated the extent to which sex moderated any associations between comorbid psychiatric symptoms and sleep and found that despite the sex differences in the rates of sleep problems, sex did not have a moderating effect on the associations between comorbid symptoms and sleep.

Sex differences in the profile of sleep problems in adolescence are less striking. Gau and Chiang (2009) carefully assessed a range of sleep disorders pertinent to this chapter (insomnia, hypersomnia, circadian rhythm sleep disorder, sleep terror disorder, sleepwalking disorder, and nightmare disorder) according to DSM-IV (APA, 2000) criteria. No sex difference in the rates of any sleep problems were reported in their large sample of adolescents, with the exception of nightmares which were more common in girls (both current and lifetime rates).

Firstly, results are mixed both in terms of whether ADHD subtype differences in sleep are detected or, if they are, which subtype and which type of sleep disturbance are associated. For example, multiple studies have found no subtype-related differences in parent reported “total sleep disturbance” (e.g., Grünwald & Schlarb, 2017) or for specific types of sleep disturbances such as dyssomnias (Corkum, Moldofsky, Hogg-Johnson, Humphries, & Tannock, 1999; Grünwald & Schlarb, 2017) and parasomnias (although this included “night waking” too) (Corkum et al., 1999).

In other work, Mayes et al. (2009), found that compared to ADHD-I (and controls), children with ADHD-C were reported by parents to have more problems with all areas of sleep (including dyssomnias, nightmares, and sleepwalking). The ADHD-HI subtype was not included in this study by Mayes et al, but all three subtypes were compared in the work by Grünwald and Schlarb (2017). They similarly made comparisons across a range of sleep disorders, assessed by parent report but in relatively small, although well-defined samples. They found only parasomnias to differ between the subtypes; in this study the parents of the ADHD-HI group reported significantly more parasomnias than the parents of the ADHD-I group (with the ADHD-C group somewhere in the middle). Wagner and Schlarb (2012) also found parents of the children with ADHD-HI reported the highest overall sleep problem scores compared to the other subtypes and also showed highest rates of specific sleep difficulties including bedtime resistance and daytime sleepiness. In contrast, in adolescents, Chiang et al. (2010) found the only difference between the subtypes was that the ADHD-HI group was less likely to have insomnia than both the ADHD-C and ADHD-I groups.

An alternative aspect of sleep, persistence of problems over a 12-month period, was examined in relation to subtype by Lycett et al. (2014). Results suggested that the ADHD-C appeared to be more likely to have persistent sleep problems over a 12-month period, compared to both children with ADHD-I and ADHD-HI.

Secondly, whilst the profiles of sleep disturbance associated with different ADHD subtypes remains unclear a number of studies have reported daytime sleepiness to be prominent in children with ADHD-I (Le Bourgeois, Avis, Mixon, Olmi, & Harsh, 2004; Lecendreux, Konofal, Bouvard, Falissard, & Mouren-Simeoni, 2000; Mayes et al., 2009), although, as mentioned above, these results too are not without contradiction (e.g., Wagner & Schlarb, 2012). It is not known whether subtype-specific abnormalities of sleep underlie this sleepiness as parent-reported sleep of this subtype was not particularly notable in the relatively large and well-defined sample of children with ADHD-I which was studied by Becker et al. (2017); some children with ADHD-I (14%) had reduced (parent reported) sleep duration (compared to reference values) and 31.3% were reported to have sleep onset latency greater than 20 minutes (although this appeared to be related to children taking stimulant medication) but night wakings and early wakings were not prominent.

The challenge of examining subtype differences is illustrated in the review and meta-analysis of Díaz-Román, Hita-Yáñez, and Buela-Casal (2016) who investigated objective sleep differences in children with ADHD, excluding children with diagnosed sleep disorders. Whilst they found few objective differences between the sleep of children with and without ADHD they noted that the number of children with the ADHD-HI subtype whose data were included in the review were very small (n=16) compared to the children with ADHD-I (n=63) and ADHD-C (n=112). If subtype differences in sleep are prominent, such disparities in subtype sample sizes could affect the outcome of meta-analyses where subtypes are not examined separately, with sleep problems for children with ADHD potentially being severely over or under estimated.

When considering factors which might influence the risk of sleep disturbance in children with ADHD, an important area to consider is co-occurring psychiatric disorders. Children with ADHD have increased risk for anxiety, depression, and oppositional behavior (August, Realmuto, MacDonald, Nugent, & Crosby, 1996; Wilens et al., 2002), with about one-third of children having one psychiatric comorbidity and another third having more than one (see Accardo et al., 2012). This is of significance since these problems are themselves associated with increased risk for sleep disturbance (Gregory & Sadeh, 2016).

A number of studies have suggested that internalizing disorders confer an increased risk for sleep disturbance in children with ADHD. For example, Accardo et al. (2012) compared children and adolescents (6–19 years) with ADHD who had comorbid anxiety, comorbid depression, and no comorbidities, finding that increased rates of sleep overall sleep difficulties (and specifically bedtime resistance, increased sleep onset latency and more night wakings) were reported for those with comorbid anxiety compared to those without comorbidities. Whilst the depressed young people with ADHD did not have significantly elevated scores for overall sleep difficulties, they did have increased sleep onset latency and also more problems with sleep duration (which, in the scale used in this study, indicated problems with the length of sleep, either too short or too long, and irregularity of sleep duration). Similar problems with difficulty falling asleep, restless sleep, waking too early, sleeping less than normal, and elevated “overall problems” scores in association with comorbid anxiety or depression were described by Mayes et al. (2009) in their sample of the same broad age range. Whilst Accardo et al. found that parasomnias did not differ between young people with ADHD with and without these comorbidities, Mayes et al. did find that both nightmares and sleepwalking were more common in youth with ADHD who had comorbid anxiety and depression, compared to those who were comorbidity free.

In a younger sample of Norwegian children (7–13 years), Hansen, Skirbekk, Oerbeck, Richter, and Kristensen (2011) also described anxiety (with and without ADHD) to be associated with increased night waking (compared to those with ADHD only and controls) but the children with comorbid anxiety and ADHD had more bedtime resistance and sleep anxiety compared to children with only ADHD, suggesting ADHD and anxiety may confer an additive risk for some sleep problems.

It has been suggested (Corkum et al., 1999; Mick, Biederman, Jetton, & Faraone, 2000) that the presence of comorbid oppositional defiant disorder (ODD) could help explain bedtime and settling problems in children with ADHD from a behavioral perspective. However, data do not clearly support such a suggestion and the link between sleep disturbance and externalizing disorders is perhaps less clear than for internalizing disorders as results have been very mixed. For example, Hvolby et al. (2009) found that only fear of the dark was significantly more common in a group with ADHD and comorbid ODD (compared to those with only ADHD) and that there was no connection between comorbid ODD and problems related to bedtime and falling asleep. Other studies have also found limited relationships between externalizing disorders and sleep as assessed by parent report (Virring, Lambek, Jennum, Møller, & Thomsen, 2017) and objectively by actigraphy (Hvolby, Jørgensen, & Bilenberg, 2008).

In contrast, some studies have found comorbid externalizing symptoms to be associated with increased rates of different, specific, parent-reported sleep problems of night waking and sleepiness (not bedtime difficulties) as well as total sleep disturbance scores (Choi, Yoon, Kim, Chung, & Yoo, 2010) but only the work of Gruber et al. (2012) supported the idea of a link between externalizing symptoms and some “settling” issues. They investigated the contribution of psychiatric/behavioral factors (internalizing and externalizing symptoms) and circadian factors (morningness/eveningness preference) to various aspects of disturbed sleep, assessed by both parent report and PSG in both children with ADHD and controls. Results indicated that in both groups of children, externalizing problems made significant independent contributions to the explained variance in parental reports of bedtime resistance, whereas an evening circadian tendency contributed both to parental reports of sleep onset delay and to PSG-measured sleep-onset latency. These results emphasize the need for careful assessment of precise types of “settling difficulties” as each may have different risk factors.

What does seem clear is that both internalizing and externalizing problems have been seen to correlate with some parent-reported sleep problems in children with ADHD (Cortese, Konofal, Yateman, Mouren, & Lecendreux, 2006;Willoughby, Angold, & Egger, 2008). Further, co-occurring internalizing and externalizing comorbidities have been significantly associated with higher odds of having sleep problems, both transiently and persistently, over a 12-month period (alone, only externalizing problems (and medication use) were associated with higher odds of persistent sleep problems over this time period) (Lycett et al., 2014). Lycett and colleagues (2015) also demonstrated that co-occurring internalizing and externalizing disorders seem to be associated with the greatest risk of sleep problems in children with ADHD.

It has been suggested that the presence of one psychiatric comorbidity alone increases the risk for both insomnia and nightmares in children with ADHD (Gau & Chiang, 2009). Whilst the details provided above suggest that there might be different risks associated with different types of comorbidity (which have yet to be fully delineated), together these findings emphasize the clinical need for particular attention to sleep in children with ADHD who have comorbid mental health problems.