While virtually all the antidepressant and mood stabilizing drugs found useful in the treatment of adults have been tested with children, systematic studies with results that can confidently be applied by child psychiatrists are rare. The familiar tricyclic antidepressants (Imipramine, desipramine, amitriptyline, nortriptyline), MAO inhibitors, and lithium have most frequently been reported in contrast to the newer agents such as the quadricyclics, monocyclics, tricyclic-like compounds, and those related to, but chemically distinct from the tricyclics.1,2
While a broad spectrum of childhood symptoms have been treated with antidepressants, serious methodological flaws limit the usefulness of most early studies. Improvements in diagnostic precision through the application of standardized assessment instruments, increased homogeneity of treatment groups, the use of more reliable outcome measures, and designs that adequately control for observer bias and individual patient differences make contemporary studies more useful. Drug trials in pediatric psychiatry have at times reflected the valid, but often naively optimistic, search for pharmacologically responsive syndromes in childhood. Much as the 1970s proved to be a decade in which stimulants were overused and nonpharmacologic treatments for hyperactive children neglected, the 1980s may be similar with regards to the use of antidepressant agents and the concept of depression in childhood.3 However, compared to the stimulant drugs, the risks associated with prescribing antidepressants are greater and their effects on cognition, growth/metabolism, and development are largely untested.
In order to understand the literature on antidepressant drug treatment in the pediatric age group, basic principles of pediatric pharmacology must be kept in mind. Children tend to exhibit developmental^ dependent differences in how they distribute, metabolize, and excrete drugs. These mechanisms are distinct from those seen in either neonates or adults.4 Children are able to tolerate higher mg/kg doses of pharmacologic agents than adults, and often require adult-like maintenance doses despite a lower absolute body weight.5
THE CYCLIC ANTIDEPRESSANTS
As discussed here, the term "cyclic antidepressants" is adopted as a pragmatic device to describe the addition of quadricyclics, monocyclics, and related non-tricyclic drugs to the more familiar category of tricyclic agents. Controlled trials with these drugs in the pediatric age group have generally been limited to Imipramine (IMI), desipramine (DMI), amitriptyline and nortriptyline (NOR). Consequently, unless specifically stated, IMI will be used as the prototype for this entire "drug class." Symptoms and syndromes in childhood subjected to treatment trials with these agents include depression, ADD with hyperactivity, behavior disorders, enuresis, and learning problems. Early papers6,7 reflect the overlap between ADD with hyperactivity and conduct disorder as subjects were described as being "hyperactive and/or aggressive." For the sake of discussion, these symptom clusters win be discussed together under the category of ADD with hyperactivity. Similarly, contusing reports and negative findings render the literature on antidepressant treatment of "learning problems"8 so vague as to be without clinical utility.
While the anticholinergic effects of most cyclic antidepressants may alter their uptake, the prototypical agent IMI tends to be rapidly absorbed from the gut, reaches peak plasma levels in .5 to 3 hours after a single dose under steady state conditions, and has a half-life that ranges from 6 to 15 hours in children.9 Individual half-life variation has been conservatively estimated to be over 14 fold.10 There is now evidence to suggest that certain genetically predisposed persons may be unable to degrade antidepressants and a host of other drugs through hydroxylation,11 thus risking exposure to toxic plasma levels at even low doses. Population estimates indicate that this autosomal recessive "poor metabolizing" trait may be present in 3% to 9% of Caucasians, 6% of Blacks, and a significantly higher percentage of Orientals and/or American Indians.12 Because of the reduced capacity for albumin binding and the smaller absolute adipose compartment in children compared to adults, higher circulating levels of free, unbound drug can be expected. When combined with the fact that hepatic biotransformation into active metabolites occurs more promptly in children because of their greater liver mass to total body mass ratio, there is reason to believe that, with IMI, the risk of exposure to cardiotoxic metabolites such as 2-OHimipramine may be greater in children than adults.
These drugs have not been evaluated with regard to their potentially positive or negative impact on cognition and academic performance.13 An initial stimulant-like effect tends to be present but is transient, and probably cannot be expected to significantly alter attention span, concentration, or freedom from distractibility in most children. Target symptoms in depressed children that may be responsive to treatment with the cyclic antidepressants have been reported to include dysphoria and suicidal ideation. Withdrawal and helplessness show a nonsignificant trend toward improvement.14 Work by Puig-Antich15 suggests that changes in anhedonia and depressed mood correlate closely with plasma drug levels in responders treated with Imipramine. Conversely, in some children, aggression, anger, and hostility may paradoxically worsen with treatment as lethargy, depression, and dejection improve.16 This finding has been supported in non-depressed children with behavior disorders or enuresis who occasionally become moody, tearful, Or irritable when treated with amitriptyline.17,18
The wide variety of side effects encountered with this drug class can best be understood by dividing them into four broad categories based on pharmacologic action: a) anticholinergic effects (central and peripheral), b) cardiovascular effects (quinidine-like, anticholinergic, and direct myocardial suppression), c) dose related phenomena encountered in poisoning, and d) miscellaneous effects.
Anticholinergic potency varies considerably from agent to agent, with nortriptyline and desipramine representing the least noxious and amitriptyline representing a prototype of the most noxious. Imipramine has intermediate effects. The newer tricyclic-like drugs range in potency from those with virtually no anticholinergic effects to those with effects comparable to the standard tricyclics. In general, central and peripheral anticholinergic side effects are more annoying than serious. They include sedation, blurred vision, dry mouth, drying of bronchial secretions, constipation, and urinary retention. The precipitation of narrow angle glaucoma, while uncommon, is serious and requires prompt attention.
Cardiovascular side effects are clearly the most serious and potentially treatment limiting pharmacologic actions encountered. Virtually all agents marketed in the United States exert a quinidine-like effect that slows conduction time and repolarization, thus prolonging EKG indices such as PR, QRS, QTs, and HV intervals. The overwhelming majority of children maintained on doses greater than or equal to the equivalent of 3.5 mg/kg/d of IMI exhibit a significant prolongation of the PR interval. Those children with the shortest pretreatment PR intervals tend to exhibit the greatest magnitude of change, although it is uncommon to exceed the 0.1 to 0.2 second normal range seen in 5 to 16 year olds unless unexpected complications such as unusually slow metabolism or hydroxylation defects are encountered. QRS complex widening, while relatively uncommon at doses under 5 mg/kg/d is likely to be encountered at higher doses. Except in the case of acute toxicity, arrhythmias tend to be suppressed rather than accentuated by these drugs.
Similarly, anticholinergic effects on AV node firing and refractory time are seen in most children, resulting in a mean increase in resting heart rate of 10 to 20 beats per minute.19 While blood pressure commonly drops in a dose dependent manner in adults, children may exhibit a paradoxical increase in blood pressure. They are probably also subject to the same idiosyncratic dose-independent orthostatic hypotensive effects seen occasionally in adults. Finally, animal studies suggest that the prototypical agent IMI and its major metabolite, DMI, may exert a direct negative effect on myocardial contractility that has not been examined in children.
Toxicity associated with overdose represents a serious danger that is not limited to intentional or accidental drug ingestion. Children with unusually short antidepressant drug half-lives (less than 7 hours) may fail to respond to doses many practitioners consider to be "standard" because their plasma levels are inadequate. In the absence of careful plasma level monitoring, higher doses may be prescribed and expose the child to greater risk. Similarly, those with unusually lengthy half-lives (greater than 12 hours) or those with genetically mediated defects in their ability to metabolize these drugs through hydroxylation are at risk and may unintentionally be made toxic on even low or modest doses. Fatalities after the accidental ingestion of IMI and amitriptyline have been described,20,21 with most deaths attributed to heart block and/or tachyarrhythmias. Mild symptoms may include drowsiness, restlessness, vomiting, midriasis, hallucinations, and tachycardia, while more serious symptoms range from ataxia, convulsions, and systolic hypertension, to coma, hypotension, tachyarrhythmias, and cardio-respiratory arrest. Severe symptoms may be seen in children who have ingested up to 10 mg/kg of amitriptyline, and are almost always present when higher doses have been ingested.22
Miscellaneous side effects may be encountered and can be problematic. Rapid withdrawal from maintenance doses greater than or equal to 3.5 mg/ kg/d of the prototypical agent IMI is likely to be accompanied by uncomfortable flu-like symptoms such as nausea, vomiting, abdominal pain, drowsiness, diminished appetite, tearfulness, apathy, and headache. Cholinergic rebound has been implicated as the most likely cause for this syndrome in children,23 and has received support in the adult literature.24-26 Because these symptoms are rarely encountered when medication is discontinued slowly, it is recommended that children be withdrawn gradually over a one to two week period. All cyclic antidepressants should be used cautiously in children with CNS damage, autism or autistic-like syndromes, and/or preexisting seizure disorders due to an alleged lowering of the seizure threshold,27 although it is unlikely that such an effect would be significant in otherwise healthy children. Finally, while the scope of this article does not permit discussion of the controversy surrounding the concept of prepubertal manic-depressive illness, there is evidence in adults that tricyclics may indeed precipitate mania in vulnerable or predisposed individuals.28 The applicability of these findings to children is unclear. However, in a group of rigorously diagnosed depressed adolescents, the most statistically significant predictor of early onset bipolar illness was the precipitation of mania by tricyclics.29
Guidelines for Use in Specific Syndromes
Depression. Limitations imposed by past debates over phenomenology and diagnosis of depression in children are further complicated by the paucity of adequately controlled drug studies in this age group. Recent advances in the standardization of diagnostic criteria, however, strongly support the validity of depression as a syndrome in childhood carrying a significant enough morbidity to warrant treatment with medication.30-32
Approximately 75% of all children treated with tricyclics in open, uncontrolled trials prior to 1975 were said to have responded positively. Adequately controlled, blind treatment trials performed more recently support this finding. Puig-Antich and his group continue to provide updates on their study of the relationship between IMI and DMI plasma levels and treatment response in prepubertal children meeting Research Diagnostic Criteria for a diagnosis of major depressive disorder,15·33 and have elaborated a set of guidelines for treatment and treatment monitoring that are useful in clinical practice.
Target symptoms that have been shown to improve most consistently in positive drug responders include dysphoria, anhedonia, and suicidal ideation, while irritability, anger, aggression, and crying may paradoxically worsen. An adequate trial of the prototypical agent IMI is probably a duration of five weeks with plasma levels of IMI plus its chief metabolite, DMl, maintained above 146 ng/ml (preferably around 200 ng/ml). Depressed children with psychotic features respond quite differently than those without and require significantly higher plasma levels with optimal clinical response at or above 400 ng/ml.19 In general, plasma IMI plus DMI levels, and plasma DMI levels alone predict clinical response while variables such as dose, IMI plasma levels alone, and endogenous/nonendogenous classification of symptoms do not. It is not clear whether a ceiling plasma level above which drug effectiveness diminishes exists with IMI, although this has been suggested.34 Preliminary studies with children35 support work done with adult and geriatric patients in which nortriptyline (NOR) seems to be far safer than IMI due to its lower incidence of side effects, lower maintenance dose requirement in order to achieve therapeutic plasma levels, and more precise "therapeutic window" of 50 to 150 ng/ml. While no final data are available, children may not respond to NOR until they reach plasma levels of 75 ng/ml and should probably be maintained within the mid range of the "therapeutic window." Amitriptyline has a limited role in the treatment of depressed children. It has no obvious advantages over agents such as IMI and NOR, is highly anticholinergic and, as yet, exhibits no clear relationship between plasma level and clinical response even in adults. Some of the newer cyclic agents promise to be as effective and perhaps safer, but remain untested in children.
Because cardiovascular effects can be expected with all the commonly used tricyclic antidepressants, a baseline EKG and regular EKG monitoring of PR and QRS intervals should always be undertaken. While probably too conservative, the safety standards proposed by Puig-Antich32 are prudent to follow. It is recommended that resting heart rate be maintained at or under \30 beats per minute, that PR intervals be less than or equal to 0.21 seconds, that QRS interval be maintained at less than a 30%-38% increase over baseline values, and that blood pressure be kept at less than or equal to 145 over 95 in order for drug maintenance to be considered safe. Attention should be paid to adequately preparing children and their parents for the likelihood that annoying, minor side effects will emerge. Parents should always understand the meaning of more serious side effects as well as the dangers associated with accidental ingestion of larger doses. In clinical practice, prescribing small quantities of these medications and meeting frequently with the patient and their family may minimize the risk of dangerous overdosage.
Gradual titration to adequate plasma levels should be performed since absolute dose and mg/ kg/d calculations do not correspond with improvement in drug responders. Imipramine is most safely given three times per day while nortriptyline, the agent which is likely to emerge as the pharmacologic treatment of choice in depressed children, may be given twice a day because of its longer half-life (Figure). Single daily doses should always be avoided as they result in pharmacokineticprofiles more closely resembling those of a parenteral than of an oral dosing, and expose children to unnecessarily high "bolus" like levels of potentially toxic metabolites. Predictive kinetic models have been developed in populations of depressed adult and geriatric patients in which maintenance dose can be estimated on the basis of drug half-life determination following the administration of a single, low test dose of either IMI or NOR. While nomograms generated from this model are only now being established for children, it is likely that titration to therapeutic plasma levels may soon be facilitated by using a standardized "kinetic study" and that slow, rapid, and compromised metabolizers can thus be identified in advance.
Finally, caution is recommended in treating depressed children with medication for several reasons. Because none of the cyclic antidepressants are FDA approved for use in depressed children, clinicians may risk liability should serious or unexpected cardiovascular side effects occur. Patients with genetically mediated hydroxylation defects and those not evaluated with plasma levels and EKGs are at highest risk for adverse effects. The reliability of commercially available tricyclic antidepressant plasma levels is questionable.36,37 It is suggested that the clinician be absolutely certain that the laboratory on which he/she relies utilizes satisfactory quality controls and that occasional "split samples" be sent as a test.
Attention Deficit Disorder with Hyperactivity
The stimulant drugs are most commonly used in practice for the treatment of ADD (methylphenidate, dexedrine, pemoline) and appear to be effective over 75% of the time in children over the age of 6 years. They are infrequently associated with serious side effects.
Early studies designedsto evaluate the efficacy of tricyclic antidepressants as an alternative drug treatment of hyperactivity are limited to IMI and amitriptyline.6'8,3840 While suffering from serious méthodologie limitations, the evidence from this work suggests that hyperactive children may respond to this drug class, although side effects and the possibility that symptoms become refractory to treatment after weeks or months are often cited as reasons not to use tricyclics. In a one year follow-up study of hyperactive children previously treated with either methylphenidate/plaeebo or IMI/ placebo in a double-blind, crossover design,41,42 both methylphenidate and IMI were found to be superior to placebo. While it was originally reported that methylphenidate was more effective than IMI on the basis of teacher ratings and other variables related to side effects, follow-up with children still maintained on medication failed to distinguish between the two drug groups. Both groups seemed to have maintained clinical gains, and both were associated with anorexia and weight loss. Surprisingly, the high dose methylphenidate group suffered the most side effects. This may not accurately reflect differences between the two groups, due to more treatment dropouts during the year in the IMI group than in the methylphenidate group, presumably on the basis of side effects.
Figure. Guidelines for prescribing common tricyclic agents for prepubertal depression.
Several contemporary studies support the hypothesis that tricyclic antidepressants may be an effective treatment alternative for some hyperactive children. Werry43 tested methylphenidate against IMI in a double-blind, placebo controlled, crossover study and concluded that IMI may actually be superior to both methylphenidate and placebo. A stimulant-like effect on cognition and motor performance as well as a weak elevation of mood were observed in the IMI group. Based on dose-response characteristics nearly as flat as those for methylphenidate, higher IMI doses (2 mg/kg/d) were found to be associated with relatively more side effects than additional clinical benefits when compared to lower doses (1 mg/kg/d). However, the high dose was superior on the basis of global parent and psychiatrist ratings, as well as on the basis of the degree to which it reduced seat movement. A recently published study by Garfinkel44 compared methylphenidate, desipramine (DMI), and clomipramine (CMI) under rigorously controlled, double-blind conditions. While methylphenidate resulted in greater behavioral and cognitive improvement than either of the antidepressants, all were superior to placebo. Both classes of drugs raised systolic blood pressure equally but, because of its association with sleep disturbance, methylphenidate was consistently rated as having more side eïects than either of the tricyclics. Tricyclics were found to have two potential advantages over stimulants. First, due to their longer halflife and sustained action, they control "after school" behavior better and tend not to be associated with evening symptom "rebound" or escape. Secondly, as described in several earlier studies, the tricyclics appear to exert a weakly positive effect on mood and self-perception, although this finding awaits further substantiation.
Tricyclics, particularly the less anticholinergic agents such as IMI and DMI, seem to be nearly as effective in reducing ADD target symptoms as the stimulants. Anecdotal reports regarding reduced tricyclic effectiveness over time are not entirely justified, although children maintained on tricyclics may be more likely to be taken off medication by their parents because of the side effects than those who are maintained on stimulants.
Guidelines for the use of tricyclics in ADD have not been established and stimulants remain the first drug class of choice in this disorder. Tricyclics seem to represent a rational treatment alternative in the event that stimulants are either ineffective or contraindicated, or when nonpharmacologic treatments that are often equally effective or superior to medication cannot properly be applied.4547 Other candidates for a trial of tricyclics may include those in whom sleep disturbance, severe late afternoon symptom escape, or affective symptoms (irritability, dysphoria) cannot be managed by stimulant dose reduction, the addition of a late afternoon dose, or changing to a different stimulant. One infrequently encounters children, such as those with Tourette's syndrome or chronic motor tics, in whom the use of dopaminergic drugs such as the stimulants are contraindicated. In these cases, tricyclics may be the first drug of choice although other agents, such as Clonidine, are being evaluated for such purposes.
At this time, the use of low doses of the prototypical agent IMI offers the best margin of risk to benefit, with 2 mg/kg/d representing a reasonable ceiling dose. Because plasma tricyclic concentration varies from child to child, doses up to 3.5 mg/kg/d may be necessary despite the fact that plasma level guidelines for use in this disorder have not yet been established. Symptomatic improvement in positive responders is likely to be prompt; in some cases benefits will be obvious within a day or two. As in childhood depression, divided doses, adherence to EKG safety criteria, and the careful preparation of parents and children are recommended, the clinician should also avoid using doses higher than 3.5 mg/kg/d whenever possible.
Enuresis is a common childhood problem estimated to occur in nearly 20% of 5-year-olds, and to decline progressively by age 12 to 14 to less than 2%. Etiologic explanations are diverse but unproven, and range from those with little empirical support (psychoanalytic conflict or symptom substitution theories, sleep arousal disorder theories) to those with acceptable data to support them (decreased bladder capacity due to insufficient cortical inhibition or as a result of deficient bladder stretching, habit deficiency resulting from inadequate learning experiences and/or inappropriate reinforcement contingencies).48
Antidepressants, particularly IMI, have commonly been used to treat enuretic children. Again, while many hypotheses have been forwarded to explain the mechanism through which these drugs exert their effect, few are empirically supported. Mood elevation, REM onset suppression, interference with deep sleep stages, and anticholinergic effects on detrusor muscle tone appear to be unsatisfactory explanations. Enhanced involuntary control over urethral sphincter tone via tricyclic mediated alpha-adrenergic receptor sensitization to sympathetic motor innervation has been more critically examined.49 Reviews of clinical trials of tricyclics in enuretic children13 suggest that this drug class is superior to placebo 60% to 80% of the time. Total remission is infrequent, and up to 95% of children who respond positively relapse when medication is discontinued. Reports of doses used range from 25 to 125 mg of the prototypical agent IMI with 75 mg representing an implicitly established "ceiling." Plasma levels corresponding to positive response have not been reported.
Alternative nonpharmacologic treatments for enuresis, particularly nocturnal enuresis, have been shown to be highly effective when properly implemented.48 Retention control with bladder expansion training appears to be a promising treatment for diurnal enuresis while modifications of the "urine alarm" (bell and pad) technique, dry bed training, or combinations of these approaches are usually sufficient to treat nocturnal enuresis.
While enuresis is the only childhood "behavior syndrome" in which the use of tricyclics has been approved by the FDA, it may very well be the least logical and least compelling reason to choose this medication. Nonpharmacologic approaches are more time consuming but are safer and probably more effective in the long run. Pharmacotherapy with IMI or a similar agent may be warranted on rare occasions, and then only for brief periods of time. The older child being sent off to camp or on vacation with nonfamily members who is significantly embarrassed or distressed by his/her symptom may represent one such instance. Because enuresis is an innocuous, usually self-limited disorder, subjecting a child to even marginal risk associated with low tricyclic doses should be carefully examined. Should the clinician decide to use this drug class as a treatment for enuresis, doses should be kept as low as possible (1 to 2 mg/kg/d), risks and side effects should be carefully explained to the parent and child, and safety criteria must be rigorously monitored.
Separation-Anxiety Disorder, or school refusal as it has most commonly been studied in the pharmacologic literature, is a relatively frequent syndrome in childhood that may represent the intersection of etiologically related psychosocial and biological variables. Contemporary work identifies this syndrome as a heterogeneous cluster of symptoms rather than a single disorder with a clear etiology.
Drug studies done with young rhesus monkeys50 imply that IMI may be valuable in altering maladaptive social behavior associated with "separationanxiety." In children, while the rate of placebo response may approach 50%, 70% of those treated with a combination of IMI and nonpharmacologic therapies improved and reported significantly greater degrees of subjective comfort, anxiety reduction, and reduction in somatic symptoms than did those treated with a combination of placebo and nonpharmacologic therapies.5153 Doses used in these studies range from 50 to 200 milligrams of IMI per day but were not reported in terms of mg/kg/d. In a placebo controlled, doubleblind study designed to evaluate the efficacy of clomipramine as a treatment for school refusers divided into three discrete age groups, Berney et al54 concluded that drug treatment offers no demonstrable long-term benefits, although shortterm improvement in depressive symptom ratings and global symptom severity ratings were reported. The authors point out that, while they used doses comparable to those recommended for use in depressed adults (up to 75 mg/d), they were probably lower than the equivalent doses of IMl found so effective in earlier studies.
Since school refusal is usually responsive to nonpharmacologic treatment and the spontaneous recovery rate in young children is high, the routine use of tricyclics in this syndrome cannot be recommended. Evidence from controlled drug studies suggests that IMI may be a useful adjunctive treatment for children refractory to nonpharmacologic interventions or for those in whom significant anxiety and/or dysphoria persist after they are able to return to school.55 Nonpharmacologic treatments such as parent management training, behavior therapy and/or family therapy should always be utilized along with pharmacotherapy as the efficacy of medication alone has not yet been tested. When tricyclics are used, it may be best to titrate to maintenance dose and monitor safety criteria in accordance with standards set for depressed children. Finally, révaluation of the child in a drug free state after three to five weeks of adequate maintenance therapy appears advisable.
THE MONOAMINE OXIDASE INHIBITORS
The monoamine oxidase (MAO) inhibitors were the first drugs with specific antidepressant effects to be marketed and formally tested in treatment trials with adults. Although ongoing work with adults suggests that these agents are effective as a second drug of choice in depressed, tricyclic nonresponder and perhaps the drug of choice in certain atypical mood disorders or phobic-anxiety disorders, complementary data in pediatric populations are unavailable. Advances in determining the mechanisms through which certain food-drug and drug-drug interactions mediate dangerous hypertensive crises have significantly increased the margin of safety of these drugs in clinical practice with adults. Yet, dietary restrictions are difficult to enforce Ui the pediatric age group.
Studies performed with children have been anecdotal, poorly controlled, and by and large limited to trials on diagnostically heterogeneous or predominantly phobic-obsessional groups. Rapoport and Mikkelsen13 reviewed several studies in which MAO inhibitors were used to treat phobic children with results that, at best, are merely suggestive. No recent or methodologically acceptable trials have been performed with children who satisfy diagnostic criteria for major depressive disorder. There appear to be no legitimate uses for the MAO inhibitors in the treatment of children at this time, although controlled experimental use may yield guidelines in the future. It is suggested that MAO inhibitors not be prescribed for children unless new findings with adequate empirical support emerge.
Although lithium has not been shown to be a particularly specific treatment for acute depression, patients with bipolar type depression and/or those refractory to other treatments may respond positively, and its efficacy in the acute treatment and prophylaxis of manic-depressive illness in adults is well documented.56
In existing studies of the use of lithium in children drug conditions and outcome measures have been poorly controlled. While these studies often claim high rates of positive drug response and may, in fact, accurately reflect useful clinical data upon which more satisfactory work can be based, they must be interpreted with caution. Rapoport, Mikkelsen, and Werry57 carefully reviewed six controlled trials performed with diagnostically heterogeneous groups of children and concluded that, although lithium has no clear indications in the treatment of prepubertal children, it may be effective in early adolescent onset bipolar illness, as well as in certain behavior disorders characterized by severe aggression. Hyperactivity, although proposed by some authors to overlap with mania in children with positive family histories of bipolar illness and cyclic variations in mood,58 does not seem to be an acceptable indication for treatment with lithium. While manic-depressive illness in prepubertal children undoubtedly exists, the identification of those at risk or those exhibiting clear cut symptoms remains controversial.
Pharmacologic and Clinical Effects
The lithium ion is rapidly absorbed in the gut, reaching peak serum levels in roughly two to four hours with an effective half-life of 24 hours. Because it is not protein bound, it is distributed throughout the entire intra- and extra-cellular fluid compartment, and is subject to the same pharmacokinetic factors peculiar to children discussed earlier. Similarly, since it is not metabolized in any way, it is cleared entirely by the kidney where excretion/reabsorption rates are limited by the rate of sodium exchange in the proximal tubules. While blood levels are influenced by numerous factors, steady-state concentrations of lithium can be expected to occur within five days after multiple dosings.
The effect of lithium on cognition and academic performance in children has not been evaluated. In adults, even therapeutic doses sometimes lead to subjective complaints of mild confusion and decreased concentration,59,60 although the literature yields conflicting reports about these problems.61,62 Aggression, impulsivity, irritability, and affective lability may improve in some children treated with lithium.57,63
Mild symptoms, often likely to occur in association with peaks in serum lithium concentration, are common and include tremor, diarrhea, drowsiness, and blurred vision. Polyuria and polydipsia are commonly seen, and nausea following the ingestion of lithium on an empty stomach can be eradicated by taking the drug with meals. Probably because of lithium's effect on intracellular electrolyte balance, cell membrane permeability, and action potential threshold in the heart, sinus node depression can occur, resulting in bradycardia64,65 that is reversible after discontinuing the drug. Flattening of T waves may be common but is of little concern, and at toxic blood levels, the QT interval can be prolonged.
Metabolic and endocrine side effects of lithium pose more serious problems for use with children. A reduction in circulating thyroid hormone with compensatory hypersecretion of TSH as well as clinical hypothyroidism are commonly observed but tend to be reversible once the drug is withdrawn. Adverse effects on carbohydrate metabolism and a transient increase in Cortisol levels have been reported to occur in adults, but have not been evaluated in children. Perhaps of even more concern in children who are still growing are reports of hypercalcemia and bone demineralization said to occur in some adults maintained on lithium for long periods of time.66-68 This effect on calcium metabolism has been hypothesized to be mediated at the renal level by elevated levels of circulating parathyroid hormone.
Neurotoxicity can be seen in overdoses or accidental poisoning with large doses of lithium. Initial symptoms are exaggerations of the more commonly encountered mild side effects and include lethargy, nausea, vomiting, diarrhea, and polydipsia. Progression to confusion, ataxia, muscle fasciculation, marked hypoactivity, and diffuse EEG slowing can lead to coma and sometimes death. Severe toxicity may be difficult to manage clinically. Because lithium excretion is regulated by sodium exchange rates in the proximal kidney tubules, any factors that result in sodium wasting (concurrent use of diuretics, low salt diets, fluid loss through perspiration during vigorous exercise) tend to simultaneously increase serum lithium concentration and predispose the individual to potentially toxic side effects. Fatalities have been reported.67 Miscellaneous side effects include leukocytosis,69 skin rashes, and reversible nephrogenic diabetes insipidus. Another controversy beyond the scope of this article is the potentially damaging effect of longterm lithium maintenance on the microanatomy of the kidney.70
Guidelines for Use in Specific Syndromes
The topic of prepubertal manic-depressive disorder remains controversial and children only infrequently satisfy adult criteria for this diagnosis. Developmentally mediated differences in symptom expression between children and adults have not been critically examined but may generate useful information regarding childhood precursors of adult mania. At least 20% of adult patients satisfying DMS-III criteria for mania, exhibit symptoms before the age of 20 years.71 These findings suggest that further prospective work may yield clinically relevant data, particularly if and when accurate ratings of symptom periodicity72 and/or biological markers can be applied.
Elaboration of diagnostic criteria for prepubertal manic-depressive disorder date to 1960 when Anthony and Scott73 concluded that mania does occur in childhood but that it is exceedingly rare. Others have proposed criteria74,75 that are probably overly inclusive. Faulty logic, arguing that lithium responsive syndromes in children are, by definition, affective disorders regardless of presenting symptoms further confuses the issue. Adolescent onset bipolar illness is a less controversial topic and lithium has regularly been used to treat this population. No acceptable studies to date have empirically tested the efficacy of lithium in children meeting strict diagnostic criteria for major depressive disorder, and its application to atypical childhood affective disorders has been explored in only a preliminary way.
Similarly, lithium as a treatment for hyperaggressive children refractory to other interventions, particularly those with affective lability and cyclic fluctuations in mood, has only been tested in small, diagnostically heterogeneous samples with inadequate control measures. Lithium does appear to exert a nonspecific antiaggressive effect in humans that has been substantiated in a number of controlled and uncontrolled drug trials.76 Together, these findings suggest that lithium may be useful in certain aggressive children in whom the risk of maintenance therapy can be justified.
Given the absence of acceptable, rational guidelines for the use of lithium in childhood affective and behavior disorders, adherence to standards developed for adults seems prudent. Titration to maintenance doses should be based on clinical response with serum levels used to monitor the potential for toxicity. Children may require higher mg/kg/d doses to achieve serum levels equivalent to those in adults because their renal clearance of the drug is higher and more efficient. However, there is no evidence to suggest that they are any less sensitive to side effects, and the effect of lithium on cognition in children, even at low serum levels, remains unknown. Aggressive children with treatment refractory disorders of impulse control may respond promptly at doses lower than those required to exert an antimanic effect77 although the data are by no means conclusive. These topics have been reviewed in greater detail elsewhere.78
Baseline measures of renal, thyroid, cognitive, and metabolic (particularly calcium and carbohydrate metabolism) functioning should always be obtained prior to initiating lithium therapy, and regular monitoring is advised. Full therapeutic effects may be seen in children after 7 to 10 days of maintenance when levels are kept within the 0.6 to 1.5 mEq/L norms established for adults.57 Measurements of salivary lithium concentration hold some promise as an alternative to frequent venipuncture in children79 but have not, as yet, been standardized satisfactorily. Other less invasive procedures such as measuring lithium concentration in microsamples via finger sticks or ear lobe punctures with equipment routinely used in monitoring diabetic children are available.80
Given the paucity of controlled studies in which children are treated with lithium, other pharmacologic interventions should be considered before lithium is selected. The potential for serious long term and largely unexplored side effects dictates that the clinician carefully weigh the risks and potential benefits, adequately prepare children and their parents, and select only those in whom symptoms are quite serious before deciding to use lithium. Treatment trials should be brief and drug free révaluations undertaken. Further, children on diuretics, salt restrictions, or those suffering from diabetes are not likely candidates for safe treatment with lithium.
1. Feighner JP: The new generation of antidepressants. / Clin Psychiatry 1983; 44:49-55.
2. Kelwala S, Stanley M. Gershon S: History of antidepressants: Successes and failures. J Clin Psychiatry 1983; 44:40-58.
3. Werry [S: Pharmacotherapy, in Lahey B. Kazdin A (eds): Advances in Clinical Child Psychology. New York, Plenum Press. 1982. vol 5.
4. Rane A, Wilson J: Clinical pharmacokinetics in infants and children. Clin Pharmacokinei 1976; 1:2-24.
5. Bnant RH: An introduction to clinical pharmacology, in Werry JS (ed): Pediatric Psychopharmacology: The Use of Behavior Modifying Drugs in Children. New York, Brunner/Mazel. 1978.
6. Huessy H. Wright A: The use of Imipramine in children's behavior disorders. Acta Paedopsychiatr 1970; 37:194-199.
7. Yepes L, Balka E, Winsbcrg B. cl al: Amitriptyline and methylphenidate treatment of behaviorally disordered children. / Child Psychol Psychiatry 1977; 18:39-52.
8. Rapoport |: Childhood behavior and learning problems treated with Imipramine. International loumal of Neuropsychiatry 1965; 1:635-642.
9. Perei JM. Stiller RL. Glàssman AH: Studies on plasma level/effect relationships in Imipramine therapy. Communications in Psychopharmacology 1978; 2:429-439.
10. Morselli PL, Cuche H. Zarifian E: Pharmacokinetics of psychotropic drugs in the pediatric patient. Advances in Biological Psychiatry 1978; 2:70-85.
11. Eichelbaum M. Bertilsson L. Sâwe J. et al: Polymorphic oxidation of sparteine and debrisoquine: Related pharmacogenetic entities. Clin Pharmacol Ther 1 982 ; 31:184-186.
12. Kalow W: The metabolism of xenobiotics in different populations. Can I Phvsiol Pharmacol 1982: 60:1-12.
13. Rapoport JL, Mikkelscm EJ: Antidepressants, in Werry IS (cd): Pediatric Psychopharmacology: The Use of Behavior Modifying Drugs in Children. New York. Brunner/Mazel, 1978.
14. Peed TA. Law W: Imipramine treatment of depressed children: A double-blind pilol study. Clin Psychophennacol 982; 2:107-110.
15. Puig-Antich J. Perel JM, Lupatkin W. et at: Plaamalevelsolin,ipramine (IMI) and desmethylimipramine (DMI) and clinical response in prepubersal major depreasive disorder. J. Am Acad Child Psychiatry 1979: 18616-627.
16. Pallmeyer TP, Petel TA: Effects of imipramine on agwession and dejection in depressed children. Am J Psychiatry 1979: 11:1472.1473.
17. Puissant AF. Ditman KS, Greenfield R: Amitriptyline in children enassetis. Clin Psychophannacol Ther 1966: 7:21-25.
18. Kraft IA. Ardali C, Duffy, et at: Use of amitriptyline in childhood behavior disturbances. International Journal of Neisropsychiatry 1966: 2:611-624.
19. Puig-Antich J. Pen lM. Lupaikin W. et al: Imipramise in prepubertal major depressive disorders. Arch Gen Psychiatry, to be published.
20. Cool KM. Shanks RA: Amitriptyline and imipramine poisoning In children. Br Med J 1974; 1:261.263.
21. Saraf K. Klein D, Gittel,nan-lclein R, et al: Insipramine and aide effects In children. Psychopharmacologia 1974: 37:265-275.
22. Steel CM. O'Duffy Brown 55: Clinical effectsand treatmene ofimipramine and asnittiptyline poisoning in children, Ar Med J 1967: 3:663.667.
23. Law W. Petti TA, Kasdin AE: Withdrawal symptoms after graduated cessstion of imipramine in children. Am J Psychiatry 1981: 138:647-650.
24. Shatan C: Withdrawal symptoms after abrupt tennination of lmipraminr. Can I Psychiatry 1966: tltstsppt): 150-158.
25. Kramer IC. Klein DE, Fink M: Withdrawal symptoms following discontinuation of imiprsnsine therapy. Am,! Ps-ychiatsy 1961: 118:549-550.
26. Dlltaver SC. Kronfol Z. Sackelisres IC, ci alt Antidepressant withdrawal tyndromes: Evidence supporting the cholinergic overdrivt hypothesis. IC/in Psychopharsnacol 1983: 3:157-164.
27. Petti TA. Campbell M: tmipramine and seisurea. Am I Psychiatry t975: 5; 538-540.
28. van Scheyen ~D, van Kammen UP: Comiprsmine-induced mania in unipolar depression. Arch Gen Psychiatry 1979: 36:560.
29. Strober M. Carlson 0: Bipolar illness in adolescents with major depression. Arch Gas Psychiatry 1982: 39:549-555.
30. Puig-Antich I. Gittelman R: Depression in childhood and adolescence, in Paykel ES (cdl: Handbook of Affective Disosders. New `York. Guilford, 1982
31. Kasdin AE. gancurello MD. Unit AS: Childhood depression, in Burrows GD. Wet-nt IS feds): Advances in Human Psychophar,naco!ogy. Greenwich. CN, Al Preat, vol 4, to be publiahed.
32. Puig-Anticb I: Antidepressant treatment in children: Current state of the ri/idence, in Friedman E. Gershon S. Mann I teds): Depression and Antidepressanta: Implications for Cause and Treatment, New York, Raven Press. 1981.
33. Ptaig.Antich J: Affective disorders in childhood: A review and perspective, in Blinder BI ted) The Psychiatric Clinics of Nrsrrh America. Philadelphia, WE Saunders Co, 1980. vol 3.
34. Preslcorn SI-I. Weller ER. Wcller RA: Depression in chitdren: Relationship between plasma imipramine levels and response. I din Psychiatry 1982: 43:450-453,
35. Geller B. Percl IM. Knitter EF. ct a!: Nortriptyline in major depressive disorder in children: Reaponse. steady-state plasma levels, predictive kinetics, and phartnacokineties. Psychaphannacol Bull 1983: 19:62-65.
36. Amsterdam ID. Brunswick Dl. Mendela I: Reliability of commercially available trlcyclic antidepressant levels, J Clin Psychiatry 1980: 4t:206-207.
37. Baumann P. Breyer-Pfsfl U, Kuas HI, et al: Ouality control of amlu-ipiyline and nortriptytine plasma level assessments: A multicenter etudy. Pharmacopsychiatry 1982: 15:1 56-160.
38. Krskowski A: Amiteiptyline in treatment of hyperkinceic children, a double-blind study. Psychosomatics 1965: 6:355-360.
39. Waiter I, l-loffmsn 5, Polizos Pet at: Outpatient treatment ofhypeeaciive achool children with ltnipramine, Am I Psychiatry 1974, 131:587-591.
40. Gittelman-Klein R: Pilot clinical trial of insipramine in hyperlcinetic children, in Conners CK ted): Clinical Use of Stimulant Drugs in Children. Amsterdam, Excerpts Mediea. 1974.
41. Rapoporl IL. Quinn PD, Bradbard G. et al: lmiprsmine and methylphensdate treatments of hyperactive boys. Arch Gen Psychiatry 1974: 30:789-793.
42. Quinn P. Rapoport I: One-year follow-op of hyperactive boys treated with imipramine or methytptsenidate. Am! Psychiatry 1975: 132:241-245.
43. Worry IS. Asnan MG, Diamond E: Imiprsmine and n'aeihylphenidate in hyperactive ehildrett/ Child Psycho! Psychiatry 1980: 21:27-35.
44. Garfinkel SD, Wender P14, Stoman L. et al: Tricyclic antidepressant and mothylphenidate treatment of attention deticia disorder in children, I Ant Acad Child Psychiatry 1983: 22:343-348,
45. cantwell UP: CNS activating drugs in the treatment of the hyperactive chtld. m Brady JP. Brodie UK (edal: Controversy in Psychiatry. Philadelphia. Saonders, 1978.
46. Barletey RA: Hyperactive Children: A Handbook for Diagnosis and Treatment. New York, The Goilford Press. 1981.
47. Rosa DM, Rosa SA: Hyperactivity: Current Issues. Research. and Theory. ed 2. New York. lohn Wiley & Sons, 1982.
48. Doleya DM: Enureals and encopreaia, in Ollendick Th, Iersen M (edt): Handbook on Child Psychopathology. New York. Plenum Press. 983.
49. Mahoney DI. Laferte go, Mahoney JE: Observationa on sphincter-augmenting effect of imtpramine in children with urtnary incontinence. Urology 1973; 1:317.323.
50. Suomi SI, Seaman SF. Lewis 1K. et a!: Effects of imipramine treatment of separation-induced sexist diaordcra in rhesus monkeys. Arch Get, Psychiatry 1978; 35:321-325.
51. Rsbmer Cl, Klein DF; Imipramine treatment of school phobia. Compr Psychiatry 1969; 10:387-390."
52. Gittelman-KIein R, Klein DF, Oaks G: Controlled imipramine treatment of school phobia. Arch Gen Psychiatry 1971; 25:204-207.
53. Gittelman-KIein DF: School phobia: Diagnostic considerations m the light of imipramine effects. J Nerv Ment Dis 1973; 156:199-215.
54. Bemey T, Kolvm I. Bhate SR, et al: School phobia: A therapeutic trial with clomipramine and short-term outcome. Br I Psychiatry 1981; 138: 110-118.
55. Gittelman-Klein R. Klein D: Separation anxiety in school refusal and its treatment with drugs, in Hersov L. Berg I (eds) : Out of School. New York, John Wiley & Sons. Ltd, 1980.
56. Coppen A. Metcalfe M. Wood K: lithium, in Paykel ES (ed): Handbook of Affective Disorders. New York, Guilford. 1982.
57. Rapoport (L. Mikkelsen Eh Werry JS; Antiroanic, antianxiety, hallucinogenic and miscellaneous drugs, in Werry IS (eds); Pediatric Psychopharmacology: The Use of Behavior Modifying Drugs in Children. New York. Brunner/Mazel, 1978.
58. Dyson L. Bareai A: Treatment of children of lilhium-respondlngparents. Current Therapeutic Research 1970; 12:286-290.
59. Iudd LL. Hubbard B, Janowsky DS, et al: The effect of lithium carbonate on affect, mood, and personality of normal subjects, Arch Gen Psychiatry 1977; 34:346-351.
60. Judd LL, Hubbard B, lanowsky DS, et al: The effect of lithium carbonate on the cognitive functions of normal subjects. Arch Gen Psychiatry 1977; 34:355-357.
61. Squire LR, Iudd LL, lanowsky DS, et al: Effects of lithium carbonate on memory and other cognitive functions, Am I Psychiatry 1980; 1 37: 1042-1046.
62. Friedman M), Culver CM, Ferrell RB: On the safety of long-term treaiment with lithium. Am J Psychiatry 1977; 134: 123-1 126.
63. Lena B: Lithium in child and adolescent psychiatry. Arch Gen Psychiatry 1979; 36:854-855.
64. Roose SP, Nürnberger G?, Dunner DL. et al: Cardiac sinus node dysfunction during lithium treatment. Am I Psychiatry 1979; 136:804-805.
65. Wilson JR. Kraus ES. Builas MM, et al: Reversible sinus-node abnormalities due to lithium carbonate therapy. N Engl / Med 1 976; 294: 1 223-1224.
66. Garfinkel PE, Ezrin C, Stancer HC: Hypothyroidism and hyperparathyroidism associated with lithium. Lancet 1973; 2:331-332.
67. Jefferson JW, Griest IH: Primer of Lithium Therapy. Baltimore, Williams & Wiikins. 1977.
68. Franks RD, Dubovsky SL, Lifshitz M. et ah Long-term lithium carbonate therapy causes hyperparathyroidism. Arch Gen Psychiatry 1982; 39:1074-1077.
69. Baidessarmt RJ. Upinski JF: Lithium salts: J970-1975. Ann intern Med 1975; 83:527-533.
70. Strayhom |M: Foundations of Clinical Psychiatry. Chicago, Year Book Medical Publishers, 1982.
71. Loranger AW. Levine PM: Age at onset of bipolar affective illness. Arch Gen Psychiatry 1978: 35:1345-1348.
72. Rancurello MD, Costello AJ, Unis AS: Cyclic fluctuations in behavior: A strategy for exploring the phenomenology of bipolar affective disorder in children. Presented at the 29th Annual Meeting of the American Academy of Child Psychiatry. Washington. DC, October 1982.
73. Anthony EJ, Scott P: Manic-depressive psychosis in childhood. I Child Psychol Psychiatry 1960: 1:53-72.
74. Lena B: Lithium treatment of children and adolescents, in Johnson FN (ed); Handbock of Lithium Therapy. Lancaster. MTP Press, 1980.
75. Davis RE: Manic-depressive variant syndrome of childhood: A preliminary report, Am I Psychiatry 1979: 136:702-706.
76. Marini JL. Sheard MH: Antiaggressive effect of lithium in man. <4cta Psychiatr Scand 1977; 55:269-286.
77. Lena B: Lithium therapy in hyperaggressive behavior in adolescence, in Sandler M (ed): Psychopharmacology of Aggression. New York. Raven Press. 1979.
78. Jefferson JW: The use of lithium in childhood and adolescence: An overview. / CKn Psychiatry 1982: 43:174-177.
79. Lena B, Bastaste MD: The reliability of salivary lithium estimations in children. International Research Communication System iournal of Medical Science 1978; 6:208.
80. Cooper TB: Monitoring lithium dose levels: Estimation of lithium in blood, in Johnson FN (ed>: Handbook of Lithium Therapy. Lancaster, MTP Press, 1980.
Figure. Guidelines for prescribing common tricyclic agents for prepubertal depression.