The U.S. Food and Drug Administration has only recently required additional safety data in medications that are likely to be used in children, in addition to drugs specifically approved for children. Already, there are a surprising number of medications known to cause psychiatric adverse events in some children.
For example, aggression, excitement, irritability or hallucinations are known to occur with alcohol(Drug information on alcohol), antihistamines, barbiturates, tricyclic antidepressants, acetazolamide(Drug information on acetazolamide) (Dazamide, Diamox), benzodiazepines, steroids, bromides, primidone(Drug information on primidone) (Mysoline), sulthiamine and stimulants (Rie and Rie, 1980). Regrettably, there has been no systematic reporting of pre- or postmarketing behavioral adverse events in children who have been given nonapproved medications, so frequency and severity are unknown.
Finally, there are really two issues embedded in the question of behavioral toxicity to medications, which I will address. The first is whether activation events occur under different circumstances in young people versus adults, and the second is whether there is any future diagnostic predictive validity to the activation if it occurs. I will specifically cover activation-like events predominantly in stimulants (stimulant rebound, stimulant response in children/teens with bipolar spectrum conditions, stimulant psychosis) and in antidepressants, including SSRI activation.
Rebound (increased hyperactivity, irritability, dysphoria occurring as stimulant medication wears off) is often described by parents and can be mistaken for mania, but is rarely examined in stimulant trials (Sarampote et al., 2002). When it is specifically examined, it is either difficult to document (e.g., Nolan et al., 1999), or the irritability that is usually measured generally improves with treatment (e.g., Greenhill et al., 2001). In one of the few systematic studies of rebound, Carlson and Kelly (2003) reported that some children (11%) behaved worse at night even without medication, while others (21%) had transient rebound. Only 9% of their sample of hospitalized children had clear, persistent, stimulation-related dysphoria, i.e., rebound severe enough to stop the medication. Within this sample, there were no indications that this response occurred more often in children with manic symptoms.
Intravenous and abused stimulants can produce euphoria, mania, hallucinations and paranoia in adults. For this reason, there is some concern that stimulants, even at prescribed doses, will induce/perpetuate mania or psychosis in children. There is surprisingly little systematic data, let alone controlled studies, on this subject in adults or children. In a post hoc analysis of the Multimodal Treatment of Attention Deficit Hyperactivity Disorder study data, stimulant response was compared in 579 children with attention-deficit/hyperactivity disorder and with and without a mania proxy based upon a Child Behavior Checklist (CBCL) (Galanter et al., 2003). Stimulants worked equally well in both groups and no one became manic. Similarly, in an inpatient study of children with ADHD, those who also had manic symptoms did no worse on stimulants than children without manic symptoms (Carlson and Kelly, 1998). Thus, direct and systematic studies of children with parent-described manic symptoms do not suggest inevitable worsening on stimulants, though, like rebound, some children could not tolerate the medication. Important to note, however, is that these children did not meet criteria for classic mania as defined by DSM-IV and Leibenluft et al. (2003).
How common and which adults become psychotic on stimulants and why remains unclear. Nontherapeutic use is usually studied. Apparently, people with mild psychotic symptoms or schizotypal traits and/or those who begin methamphetamine abuse at an early age are more vulnerable to psychosis; also, a large enough dose will produce at least a transient psychosis in almost anyone (Chen et al., 2003; Curran et al., 2004). Stimulant-induced psychosis on ostensibly therapeutic doses of medication is rarely reported in children (Cherland and Fitzpatrick, 1999; Young, 1981), so little is known of the risk factors.
I have had child patients develop paranoia on what was a previously therapeutic dose of stimulant taken after a drug holiday. (The paranoia subsequently resolved at a lower dose.) I have also seen children develop hallucinations at a younger age but not when retried on stimulants a few years later. Interestingly, in two small case series, children with psychosis or schizophrenia with ADHD symptoms were successfully treated with stimulants, provided they took antipsychotic medication concurrently (Pine et al., 1993; Tossell et al., 2004).