The present study aimed at: (1) exploring rate and clinical features of superimposed bipolar disorder (BD) in Italian children with attention deficit hyperactivity disorder (ADHD), compared with a community sample, matched for age and gender; (2) exploring predictors of BD in ADHD children, by comparing ADHD children with or without superimposed BD. We studied 173 consecutive drug-naïve outpatients with ADHD (156 males and 17 females, mean age of 9.2 ± 2.3 years, age range 6–17.5 years), diagnosed with a clinical interview (Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL)); the control group consisted of a community-based sample of 100 healthy children. The rate of children with a diagnosis of BD was higher in the ADHD group (29/173, 16.7%) compared with controls (1/100, 1%), (P < 0.001). Among the 29 children with ADHD + BD, 16 (55.2%) had a Bipolar Disorder-Not Otherwise Specified (BD-NOS), and 11 (37.9%) showed ultrarapid cycling. Compared with children with ADHD without BD, they showed a higher rate of combined sub-type (21/29, 72.4%), a higher score at ADHD-Rating Scale (total score and hyperactivity subscale), higher rates of major depression, oppositional defiant disorder and conduct disorder. In summary, children with ADHD present a higher risk for developing a superimposed BD. The identification of clinical features with an increased risk of BD can improve diagnosis, prognosis and treatments.
A major feature of juvenile bipolar disorder (BD) is its frequent ‘atypicality’, compared with adult ‘standards’, in terms of presentation, course and pattern of co-morbidity (Wozniak et al., 1995, Findling et al., 2001 and Masi et al., 2006a). Classical, adult-like symptomatology of manic-depressive illness, with clear episodes of euphoric mood and inflated self-esteem, alternated with depressive episodes is not frequent in children with BD (Geller and Luby, 1997). Another marker of the earliest forms of BD is the high co-morbidity with disruptive behaviour disorders, namely Attention Deficit Hyperactivity Disorder (ADHD) (Faraone et al., 1997, Kim and Miklowitz, 2002 and Galanter and Leibenluft, 2008). The rate of this co-morbidity is still debated in the literature, ranging from 30% to 90% (Kim and Miklowitz, 2002). Furthermore, even in adult patients, ADHD and BD co-morbidity is well established (47% of ADHD and 21% of bipolar adult patients, according to a meta-analysis), even though still understudied (Wingo and Gaemi, 2007). Differential diagnosis among early-onset mania, severe ADHD, and the co-occurrence of both the disorders may be very difficult, given the overlap of some symptoms (hyperactivity, impulsivity/aggressiveness, distractibility and emotional lability) (Kim and Miklowitz, 2002 and Galanter and Leibenluft, 2008).
The relationship between ADHD and BD is far from clear. Three models have been suggested to explain the overlap between paediatric BD and ADHD: (a) ADHD symptoms be a prodrome to paediatric BD in some cases, (b) BD and ADHD are distinct disorders but share an association with emotional difficulty in childhood, and (c) BD may be a severe variant of personality traits in which an underlying dysfunction in affective and cognitive circuitry associated with emotion regulation causes both temperamental difficulties and clinical symptoms (West et al., 2008). Recent evidence supports the notion that ADHD–BD co-morbidity has a strong and specific biological basis (Biederman et al., 2008, Pavuluri et al., 2009 and Lopez-Larson et al., 2009).
The issue is not merely nosological, because ADHD co-morbidity affects several clinical features of BD, and it can be a meaningful predictor of prognosis and treatment (Biederman et al., 1996, Faraone et al., 1997, Masi et al., 2004a and Masi et al., 2006b). In our previous studies, we explored the effect of ADHD co-occurrence on BD phenomenology and treatment by analysing a sample of children and adolescents with BD, and stratifying them according to the ADHD co-morbidity. According to these findings, ADHD is associated with an earlier onset of BD (Masi et al., 2007), a chronic course and an irritable mood (Masi et al., 2006b and Masi et al., 2006c), a greater resistance to treatments (Masi et al., 2004a), a more frequent diagnosis of Bipolar Disorder-Not Otherwise Specified (BD-NOS) (Masi et al., 2007) and a lesser efficacy of lithium (Masi et al., 2010). Most of these features are consistent with the definition of severe mood dysregulation, according to Leibenluft and co-workers (Leibenluft et al., 2003).
These previous studies did not explore the occurrence of BD in a sample of children with ADHD. According to a review article by Spencer et al. (2001), children with ADHD are 10 times as likely to develop BD than age-matched and gender-matched controls. However, ethnic and sociocultural differences are meaningful in the emergence of early onset BD (Kennedy et al., 2004).
The present study was aimed at: (1) assessing the rate and the clinical features of BD in a large, consecutive sample of Italian children with co-morbid ADHD, compared with a control, epidemiological sample of children attending elementary and junior high schools, matched for age and gender; (2) exploring possible elements associated with an increased risk of BD, comparing ADHD children with or without superimposed BD. Our hypotheses were that BD is more frequent in a sample of ADHD, consistently with data from the literature; that BD occurrence is associated with a greater severity of ADHD; and that the ADHD–BD co-occurrence is associated with a heavier co-morbidity.
