Attention-Deficit/Hyperactivity Disorder: Medication Treatment-dosing and Duration of Action

Supplements and Featured Publications, New Approaches to the Management and Treatment of Attention-Deficit/Hyperactivity Disorder, Volume 10, Issue 4 Suppl

Several trends have affected the current standardsfor medication dosing recommendations for the treatmentof attention-deficit/hyperactivity disorder(ADHD). Empirical evidence that treatment can effectimprovement across multiple domains of impairmenthas encouraged optimizing treatment throughout theday. For stimulants, the value of increasing the doseto improve effect has replaced former notions of minimizingexposure to medication by treating at thelowest dose demonstrating effect. Several new stimulantformulations control the rate of dose delivery tooptimize effect for longer intervals than immediate-releasemedication. This avoids morbidity associatedwith improper dose sculpting, inherent ups anddowns of effect, and poorer adherence to multipledaily doses of shorter-acting products. Dosing of thenew nonstimulant agent atomoxetine (Strattera) isbased primarily on weight. Atomoxetine appears tohave a longer duration of action in children and adolescents,but in adults it requires twice-daily dosing tomaintain effect across the day. Whereas its overalleffect size is smaller than stimulant agents, there is noway to predict an individual's response in advance,and for some atomoxetine may be the optimal medicationtreatment. Some head-to-head comparisonstudies between long-acting stimulant preparationsand atomoxetine, especially those focused on durationof effective action, are likely.

(Am J Manag Care. 2004;10:S99-S106)

Historical Overview

The use of medication for the treatmentof attention-deficit/hyperactivity disorder(ADHD) began in 1937 with Bradley'sserendipitous discovery of the behavioraland cognitive benefits of benzedrine. After asystematic study at the therapeutic schoolwhere he worked,1 Bradley continued totreat children with ADHD with benzedrine.Their improved academic performanceinspired the children to nickname the medication"math pills." This early linkage withschool performance continues to dominatemuch of the lay thinking regarding treatmentof ADHD, despite ever-increasing evidenceof beneficial effects on deficits insocial, family, and vocational domains.

Research on the relationship of dosestrength to effect started with Sprague andSleator.2 By contrasting 2 different dosinglevels (0.3 vs 1.0 mg/kg) in 20 patients, theydetermined that the higher dose hadimproved behavior control, but short-termvisual memory decreased. This study suggestedthat using higher doses of stimulantmedication may merely pacify children withbehavioral problems, rather than improvetheir condition. Individual dose amounts frequentlyreflected these fears, being substantiallylower than those prescribed today.Fortunately, this worrisome result has notbeen replicated in subsequent pharmacokinetic/pharmacodynamic studies.3

A focus limited to school performance,and a general mistrust and concern aboutany undiscovered risks of taking medicine,led to dosing patterns designed to minimizeoverall exposure to stimulants. Medicationwas typically prescribed for Monday throughFriday and only during the school year. Also,these short-acting medications were commonlygiven in the morning and atlunchtime, and a third dose, if given at all,was frequently "sculpted" downward instrength to avoid possible side effects and tominimize exposure to stimulants.4

Meanwhile, a large amount of evidenceabout the nonacademic morbidity of ADHD(and its response to stimulant treatment)was accumulating. Clear evidence about theimpact of ADHD on multiple domains,including social interactions, family relationships,and the development of subsequentcomorbidities, significantly influenced theexpert consensus medication treatmentparadigms of the landmark multimodal treatmentstudy of children with ADHD (MultimodalTreatment Study on ADHD [MTA]).5By design, these algorithms targeted optimizationof dosing (best response with minimalside effects) and set the standard forfull-day medication treatment.6

Concurrently, significant interest developedin pharmaceutical technologies thatpromised to deliver optimal medicationamounts across the whole day. Besides beingmore convenient than 3-times-daily dosing,long-acting medications can also provideoptimization of blood levels and significantreduction in morbidity associated with poordosing practices, missed doses, and rapidchanges in mental state from short-actingmedication. In addition, those with ADHD(especially adolescents and adults) havesymptoms, such as forgetfulness, inability tocomplete tasks, lack of follow-through, andeasy distractibility that made it difficult forthem to adhere to a 3- or 4-times-daily dosingschedule. Responding to the challengesthese symptoms cause, drug companieshave created products with controlled dosedelivery over an extended period that allowsfor optimized effect over a longer duration.

Another relevant trend has been the realizationthat ADHD is a developmental disorder.As such, some symptoms presentdifferently or may change in severity withage. For many years it was assumed thatwhen overt hyperactivity abated, typicallyduring adolescence, the disorder hadresolved.7 Further and more careful studyrevealed that rather than abating, this symptomchanged into a subjective feeling of restlessness.8 Because many of the symptomsrequired for the diagnosis are defined for thechildhood presentation (eg, "climbs excessively"),only about half of adults will retainthe full disorder on formal diagnosis.9However, with different criteria, such asthe retention of enough ADHD symptomsto cause significant impairment in everydaylife, persistence estimates are greaterthan 80%.10

It was thought for some time that adultswith ADHD did not have as robust aresponse to stimulant medication as children.Four early studies of methylphenidate(MPH) in adults with ADHD yieldedresponse rates of only 50%,11 but the estimatedmean daily weight-corrected dose ofMPH was 0.6 mg/kg/day, which was muchlower than the 1 mg/kg/day commonly prescribedto treat children with the disorder.12When he increased doses to 1 mg/kg/day forMPH or to an average of 54 mg/day of mixedamphetamine salts (MAS) in adults,Spencer demonstrated response rates similarto those in children (MPH 78%13; MAS70%14).

The advantages of full-day, optimized dosingfor maximum treatment have been furthersupported by the shift in mental healthtreatment focus from symptom control tooverall outcome effects, early intervention,prevention, and quality-of-life improvement.This shift in treatment has spawnedresearch on quality-of-life measures (eg, theimpact of ADHD and stimulant treatment ondriving),15 comorbidity development (eg, theprotective effects stimulant treatment hason preventing future substance use disorder),16 and the increase in patient andparental life satisfaction with atomoxetinetreatment.4,17,18

New Stimulant Preparations

Methylphenidate-based Products.

Theconceptual foundation for all of the newstimulant-based products was pioneered inthe University of California, Irvine (UCI)Lab School. In addition, most of the newstimulant preparations, once manufactured,have been tested for efficacy and durationof action with the UCI Lab School Protocol,often for the US Food and Drug Administration(FDA) pivotal trial submissions.19Essentially, the UCI Lab School Protocolinvolves comparing subjects against themselveson successive Saturdays, with differentdosing conditions and probing multipletimes across the day for ADHD symptoms ina school-like setting. Primary outcomemeasures typically include changes in theSwanson, Kotkin, Agler, M Flynn, andPelham (SKAMP) rating scale, or in thenumber of problems attempted or correctlycompleted in a math test (with difficulty calibratedfor each individual to be mildly boring,so as to test attention not mathematicsability). The SKAMP is a highly operationalizedclassroom observation system,providing unobtrusive counts of specific,observable "off-task" behaviors that correlateto specific ADHD symptoms, and arecaptured in provocative classroom situations,such as structured student-teacherinteraction. For example, the class may beasked to participate in a game of namingcapitals. When given a state name, somechildren will raise their hands while otherswill "blurt out" cities despite repeatedinstruction to raise their hands beforeanswering.

The original landmark work involved carefulelucidation of the relationship of MPHblood levels to their effects on ADHD symptoms.20 This research determined that ashort-term, rapid tolerance to the effects ofMPH accrues, necessitating an ascendingblood level across the day to optimize effects.A subsequent study determined the averageoptimal blood level profile for a targeted 12-hour duration of action.21 Redesigning thecapsule shape allowed for the osmotic releaseoral system (OROS) technology to be used toprovide a dose delivery system that wouldrecreate the optimal blood level profile inOROS MPH (Concerta).22

By design, OROS MPH delivers 22% of itsnominal MPH dose immediately in an overcoat,whereas the remaining 69% is deliveredacross 10 hours (the MPH in OROS MPH is91% bioavailable)23 at a gradually increasingrate by an osmotic pump. This provides 12hours of even effect independent of foodintake.24

All other long-acting stimulant systemsuse beaded, 2-bolus delivery systems withvarying ratios of immediate- to delayed-releasebeads of MPH hydrochloride (HCl)(Ritalin LA). MPH HCl is a 50%:50% ratioproviding a relatively larger initial effect,sacrificing later effects to acute tolerance,with a total duration of action of 8 hours.MPH HCl USP extended-release capsules(Metadate CD) contain a 30%:70% ratio providingincreased blood levels later in the dayto overcome acute tolerance and demonstrateeven effects across 8 hours.

Further clarification of dose responsehas been demonstrated for OROS MPH aswell. Stein et al, in a weekly forced-dose(placebo, 18, 36, and 54 mg), randomized,counterbalanced design, demonstrated aclear dose-response relationship for thehyperactive-impulsive and combined subtypes.25 Although present, the doseresponserelationship was less dramatic forthe inattentive type. Also, doses for OROSMPH up to 72 mg have been explored systematicallyin adolescents with ADHD. A30% reduction in the ADHD rating scale wasnot achieved by 37% of patients until theyhad reached 72 mg in a systematic weeklytitration schedule that increases dosage.26

Comparisons Between MPH Preparations

There have been few studies directlycomparing short-acting MPH with longer-actingpreparations. Pelham et al comparedboth OROS MPH and MPH 3 times daily withplacebo in a laboratory school setting. Thenumber of participants (68) in this studywas large enough to distinguish between theeffects of active treatment and placebo, butnot large enough to detect many differencesbetween the long- and short-acting agents.4Despite this small number of patients, parentratings of ADHD symptoms were statisticallysignificant in difference between OROSMPH and MPH 3 times daily. In a separatedouble-dummy, double-blind, placebo-controlledstudy, parents blindly selectedOROS MPH 2:1 over MPH 3 times daily.4Finally, in a small pilot study intenselyexamining driving ability in a simulator,Cox et al reported significantly better drivingperformance at 8:00 PM in subjects receivingOROS MPH compared with MPH 3 timesdaily.15 In each of these studies, an argumentcan be made that the relative nominal dosingwas not ideal, because there was a conversionfrom the MPH dose to the nominalOROS MPH dose equivalent. However, it isalso likely that these studies, underpoweredas they are, are detecting the contrastbetween the smooth action of optimizeddelivery and the multiple onset and offsetaction of bolus dosing.

Comparison studies between the intermediate(8 hour) and long-acting (12 hour) MPHpreparations consistently show that higherand increasing blood levels yield strongereffects. For instance, Lopez et al 27 demonstratedthat MPH HCl 20 mg had a strongereffect than OROS MPH 36 mg during thefirst 4 hours of behavioral observations. BecauseMPH HCl 20 mg provides an initialbolus of 10 mg MPH and OROS MPH 36 mgprovides an initial bolus of 8 mg, any initialgreater effect would be correlated to the largerinitial dose. It is unclear why no comparisonto OROS MPH 54 mg was undertaken,which presumably would have demonstratedsuperiority in the first 4 hours to MPH HCl20 mg (initial dose of 12 vs 10 mg) or inferiorityto MPH HCl 40 mg (initial dose of 20 vs12 mg).

r

This independence of the initial dose-responserelationship to preparation wasmade even more evident in the Comparisonof Once-daily Extended-release MethylphenidateFormulations in Children withAttention-Deficit/Hyperactivity Disorder inthe Laboratory School (COMAC Study)where MPH HCl USP and OROS MPH werecompared.28 Comparison doses were basedon similarity of nominal total strength(OROS MPH 18 mg vs MPH HCl USP 20mg; OROS MPH 36 mg vs MPH HCl USP40 mg, etc). When sampled over the 8-hour duration of action of MPH HCl USP,larger effects were seen presumablybecause of the inherently larger dose (20mg during 8 hours of effect vs 18 mg during12 hours of effect). When the effectsize during the initial 4 hours wasregressed on the amount of initial bolus,the correlation coefficient average was =0.9 across several measures (eg, SKAMPdeportment 0.932, SKAMP attention 0.865,PERMP [PERMAnent Product, ie, numberof problems correct or attempted on the10-question math test] correct 0.912).These results strongly suggest differencesin the first 4 hours are not caused by preparationbut by the dose amount in the initialcompartment.

These presumed preparation differencesin the 0- to 4-hour period could simply beequalized clinically by changing the nominaldose of the preparation (eg, increasing theOROS MPH to 27 mg), thereby providing thesame initial dose of 6 mg; 22% of 27 mg forOROS MPH, 30% of 20 mg for MPH HCl USP.In the 8- to 12-hour period, OROS MPH wassuperior, presumably because of the extendeddelivery and increased amount of MPHdosed during that time period. In that timerange, the shorter-acting MPH preparations(MPH HCl USP and MPH HCl) are unable todeliver additional medication regardless ofthe nominal dose, and would be unable toeffect change clinically unless a twice-dailydosing strategy was adopted.

Amphetamine-based Products.

MAScomprised of 75% dextroamphetamine and25% levoamphetamines were evaluatedusing the UCI Lab School protocol.29 As withMPH, higher doses led to greater effects. Inaddition, the duration of action was extendedby increasing the dose. These phenomenawere demonstrated in the extended-releasebeaded system form (MAS XR [Adderall]) aswell.30 In the case of MAS XR, the 10-mgstrength effects differed from placebo for10.5 hours; for the 20- and 30-mg strengths,12-hour differences from placebo wereestablished. Also, comparing subjects againstthemselves for each higher dosage strength,the effect on all measures was greater than orequal to the lower dosage strengths.

Nonstimulant Products

While several antidepressants in the pasthave demonstrated some efficacy in thetreatment of ADHD, it was not until the FDAapproval of atomoxetine that a well-studiednonstimulant became available for this indication.The development of atomoxetine washeavily influenced by safety concerns drawnfrom reported cases of sudden death withthe use of desipramine in children.31 Thisinitially compelled relatively slow dosagetitration with electrocardiograms performedat each dosage adjustment. As these safetyconcerns were mitigated, the titrationschedule became more rapid, eventuallyleading to a recommendation of 3 days takinghalf of the target dose followed by thetarget dose. When atomoxetine was deliveredon this schedule, early reports demonstratedinitial response within 24 hours.32However, subsequent clinical experiencehas led to the common belief that effectsaccumulate over time, necessitating a longerrange expectation and an indirect pharmacodynamiceffect.

Pharmacokinetic data have demonstratedidentical blood level profiles when the doseis adjusted for weight. This has suggestedweight-based dosing. In the FDA pivotal parallel-group dose-ranging study comparingplacebo, 0.5, 1.2, and 1.8 mg/kg, there wasno increased response rate between the 1.2-and 1.8-mg/kg groups.17 This suggests thatthere is little compelling reason to increasethe dose beyond 1.2 mg/kg in all but rarecases. Of interest, the response rate in adolescentswas similar for the 0.5- and 1.2-mg/kg doses, hinting that perhaps there maybe a therapeutic window with greater benefitat intermediate doses.

Duration of action was initially expectedto be relatively short because the half-life ofatomoxetine is 5 hours. In fact, all the pivotalFDA submission trials administeredtwice-daily dosing with the exception of onestudy. Encouraged by the apparent smoothnessof action, investigators explored theonce-daily dosing model with a novel duration-of-action testing paradigm. Parentswere given hand-held electronic organizersthat probed responses to the Daily ParentRating of Evening and Morning Behavior(DPREMB) scale in real time. This is anunnormed rating scale with high face validitythat asks parents to rate the applicabilityof 8 items (eg, how much difficulty didyour child have with falling asleep?) on aLikert scale (0 = not at all, 3 = severe). In a6-week study of once-daily-dosed atomoxetineversus placebo, statistically significantbut small afternoon/evening effectswere seen. Effects at 24 hours (eg, thenext morning prior to dosing) using the 4-item morning version of the DPREMBscale were also assessed but proved not tobe statistically significant in their differencefrom placebo.32

Because of failure to detect full-dayeffects, the twice-daily dosing requirementfor adults has remained. It is unclear whythe duration of action for atomoxetineappears to extend beyond expected for ashort half-life drug in children and adolescents.One postulation is that there may besome enduring receptor presence or effectthat extends beyond the presence of atomoxetinein the blood.32 Another explanationis that the perceived effect is a halo effect,shining from the benefits of treatment earlierin the day and unperturbed by a dramaticrebound or end-of-day effect that canoccur with stimulants.

Comparison Between Stimulants andNonstimulants

No formal studies comparing atomoxetinewith stimulant agents have been published.However, at least one has been presented asa poster. This 3-week, practice-based, comparativestudy of OROS MPH and atomoxetinein 651 children showed a significantlygreater reduction in total symptom ratingscores for those in the OROS MPH group.Remarkably, this difference, which was firstsignificant at week 1, steadily increased overthe course of the next 2 weeks. Using alternativeoutcome criteria, the drug responserate (25% reduction in rating scale considereda response) was also greater in theOROS MPH study arm. Safety and tolerabilitywas comparable in both groups.33

Another possible source of comparisondata comes from one of the large FDA pivotaltrials for atomoxetine, which included asmall parallel group taking MPH immediaterelease (IR) as an active control.34 Thisdesign is meant to determine the nature ofthe study should atomoxetine fail to separatefrom placebo (eg, if the test drug doesnot separate from placebo, then either thepositive control does, suggesting the testdrug to be ineffective, or the positive controldoes not, implying the study is ineffective).Some have argued that this design might beused as a weak comparison study, demonstratingrough equivalence between theeffect of MPH IR and atomoxetine. However,examination of the dosing of MPH IR revealsa modal dosing paradigm of twice daily anda suboptimal mean dose strength of 0.85mg/kg/day. If this were a comparison study,it would suggest that optimally dosed atomoxetineis equivalent to suboptimal dosingof MPH IR.

Although few formal studies that comparedrugs across classes and groups have yet tobe published, many are under way. Mostwill vary in methodology and ability tocompare proper dosing and duration ofaction measurement. In the meantime, thebest available methodology to make com-parisons for groups of patients is to evaluateeffect size (ES) (Mean Effect Rating onTrial Medication—Mean Effect Rating onPlacebo/Pooled Experimental Standard Deviation).Although such a statistical techniqueinforms comparisons across differentstudies, it is no substitute for direct comparisonin a single study.

ES evaluation consistently demonstratesa stronger ES for stimulants (average 1.0)35than atomoxetine (average 0.695 in childrenand adolescents),32 despite the broad rangeof ESs reported for stimulants. One explanationfor the wide range of ESs for stimulanttreatment may be daily dosing shifts (MTAdosing determination paradigm, ES 0.6),36addition of behavioral treatment (OROSMPH = 2.0),4 or dosing strength (ES 0.7when dosing < 0.5 mg/kg vs 0.9 when dosing> 5 mg/kg).37 Thus, the differences betweenatomoxetine and stimulant ES may begreater if care is taken to provide optimaldosing across time for both agents.

Discussion

Several things are clear in the relationshipbetween dose and symptom controlwith stimulant medication. First, increasedeffect is obtained by increasing a stimulant'snominal dose within the nontoxic range.Second, increased and smoother effect isobtained by optimizing the delivery of thatdose. Third, increased nominal dose lengthensthe duration of effect for MAS and possiblyMPH. Fourth, it is impossible to predictan individual's optimal nominal dose, whichmust be determined with a clinical titrationacross a broad range of dosage strengths(higher dosing yielding improved responsebut limited by dose-related side effects).

With atomoxetine the dose-response relationshipappears to be weight based, suggestingminimal need for optimizing dose bytitration. While duration of some effect intothe evening has been established, differentmeasures from the morning were used andthe effects were small. It is not clear whenchildren require twice-daily dosing to maintaineffectiveness throughout the day, but itis established that adults do.

As previously mentioned, some studieshave been published to demonstrate differencesin short-term effects on ADHD symptomsamong stimulants; others have beenundertaken to compare atomoxetine withstimulants. It is unlikely that further studieswill be undertaken comparing active stimulanttreatment, either short acting to longacting or between long-acting preparations,because the closeness of ES with standardoutcome measures would require a substantialnumber of participants (and thereforeexpense) to be powerful enough to detecta difference. The exceptions are studiesdesigned (as we have reviewed) to selectivelyhighlight times where one drug does nothave duration of action, or where one drug isspecifically underdosed, in which case a differencein effect could be significant with amuch smaller number of participants.

However, since the ES differencesbetween atomoxetine and stimulants aremore sizable (0.35-0.7 vs 0.5-2.0), a studywith fewer participants could detect a differencewith an investment small enough toprovide a return on investment for a manufacturer.It is likely that such studies willalso be designed to highlight differences withstrategic decisions regarding dosing (bothdosing paradigms and dosing strength) aswell as duration of action (eg, time andspecificity of measurement of effects as wellas dosing paradigms). For example, long-actingstimulants will likely be compared withatomoxetine in a study that emphasizesafternoon and evening effects.

Dosing stimulants to optimize symptomcontrol (rather than a "minimal effectivedose" model) is supported by the idea thatimproved symptom control will lead to betteroutcomes. Although the formal evaluationof the relationship between symptomcontrol and quality of life (QOL) has notbeen extensively studied, the dramaticimpact of treatment on symptoms is obvious.However, the link between symptomcontrol and QOL is taken for granted exceptin occasional cases (eg, the protective role ofearly stimulant treatment to later substanceuse disorders), making the subtleties of thisrelationship less likely to be studied or fundedby governmental agencies, such as theNational Institute of Mental Health.

Other reasons for excluding QOL studiesinclude the expense, both financially and inhuman costs. Pharmaceutical companiesneed FDA approval to market a medication.This approval process requires a demonstrationof effect compared with placebo, wheredifferences are large (eg, for stimulant medicationthe ES difference is an average of1.0) and require a relatively small number ofparticipants to detect. The measurement ofgeneral QOL effects would be diluted bymany confounds and would require a muchlonger period of study (eg, a medication thatquickly decreases impulsivity would stillrequire some time to effect a change in achild's number of friends). To undertakesuch a study would be significantly moreexpensive than demonstrating symptomcontrol effect versus placebo and would providemuch less return on investment.

The human expense of these studieswould be large as well. Any study requiring along-term placebo phase would be unlikelyto pass an institutional review board review.Even if passed, such a study would be virtuallyimpossible to implement because of difficultyrecruiting and retaining participants.Any study of, or even retrospectively reviewing,long-term QOL outcome differencesbetween short-acting and long-acting stimulantpreparations would be biased by selection,because patients of means and/oreducation are extremely reluctant to settlefor the difficulties and morbidity associatedwith short-acting stimulant treatment.

Summary

Careful examination of the dose-responserelationship has led to numerous stimulantformulations that attempt to optimize medicationdelivery across the day. In addition,clear dose-effect relationships can be manipulatedby clinically increasing the nominaldose of a stimulant preparation. Optimizingdose delivery with careful attention to maximizingsymptom improvement across thewhole day is likely to improve overall outcomes.

For the treatment of the individual ADHDpatient, the American Academy of Pediatricsrecommendation is to optimize effect acrossthe full day. With stimulants, the long-actingpreparation OROS MPH is designed tosmoothly deliver MPH in an increasing dosethat yields consistent effects for 12 hours. Inthe amphetamine class, MAS XR delivers a10.5-hour effect in lower doses and a 12-hour effect in the higher doses. Treatmenteffects are dose related and patients shouldbe titrated to the optimal dose, deliveringthe most effect with minimal dose-relatedside effects. Atomoxetine in adults should bedosed twice daily. In children and adolescents,careful clinical measurement of afternoonand evening effects may also supporttwice-daily dosing of atomoxetine to optimizeeffect. Dosing amount is primarilydetermined by weight, and increasing thedose over the recommended maximum isunlikely to yield beneficial results.

Am J Psychiatry

1. Bradley C. The behavior of children receiving benzedrine. . 1937;94:577-585.

Science

2. Sprague RL, Sleator EK. Methylphenidate in hyperkineticchildren: differences in dose effects on learningand social behavior. . 1977;198:1274-1276.

Ritalin: Theory and Practice

3. Greenhill L, Osman B.. 2nd ed. Larchmont, NY: Mary Ann Liebert; 2000.

Pediatrics

4. Pelham WE, Gnagy EM, Burrows-Maclean L, et al.Once-a-day Concerta methylphenidate versus three times- daily methylphenidate in laboratory and naturalsettings. . 2001;107:e105.

Arch Gen Psychiatry

5. A 14-month randomized clinical trial of treatment strategies for attention-deficit/hyperactivity disorder. TheMTA Cooperative Group. Multimodal Treatment Studyof Children with ADHD. . 1999;56:1073-1086.

Arch Gen Psychiatry

6. Arnold LE, Abikoff HB, Cantwell DP, et al. NationalInstitute of Mental Health Collaborative Multimodal Treatment Study of Children with ADHD (the MTA):design challenges and choices. . 1997;54:865-870.

Am J Psychiatry

7. Hill JC, Schoener EP. Age-dependent decline of attentiondeficit hyperactivity disorder. . 1996;153:1143-1146.

J Clin Psychiatry

8. Spencer T, Biederman J, Wilens TE, Faraone SV. Adults with attention-deficit/hyperactivity disorder: acontroversial diagnosis. . 1998;59(suppl)7:59-68.

Arch Gen Psychiatry

9. Mannuzza S, Klein RG, Bessler A, Malloy P, LaPadula M. Adult outcome of hyperactive boys: educationalachievement, occupational rank and psychiatricstatus. . 1993;50:565-576.

Annu Rev Med

10. Wilens TE, Biederman J, Spencer TJ. Attentiondeficit/hyperactivity disorder across the lifespan. . 2002;53:113-131.

J Am Acad Child Adolesc Psychiatry

11. Spencer T, Biederman J, Wilens T, Harding M, O'Donnell D, Griffin S. Pharmacotherapy of attention-deficithyperactivity disorder across the life cycle. . 1996;35:409-432.

Psychiatr Clin North Am

12. Wilens TE, Biederman J. The stimulants. . 1992;15:191-222.

Arch Gen Psychiatry

13. Spencer T, Wilens T, Biederman J, Faraone SV, Ablon JS, Lapey K. A double-blind, crossover comparisonof methylphenidate and placebo in adults withchildhood-onset attention-deficit hyperactivity disorder. . 1995;52:434-443.

Arch Gen Psychiatry

14. Spencer T, Biederman J, Wilens T, et al. Efficacy of a mixed amphetamine salts compound in adults withattention-deficit/hyperactivity disorder. . 2001;58:775-782.

J Am Acad Child Adolesc Psychiatry

15. Cox DJ, Merkel RL, Penberthy JK, Kovatchev B, Hankin CS. Impact of methylphenidate delivery profileson driving performance of adolescents with attention-deficit/hyperactivity disorder: a pilot study. . 2004;43:269-275.

Pediatrics

16. Wilens TE, Faraone SV, Biederman J, Gunawardene S. Does stimulant therapy of attention-deficit/hyperactivitydisorder beget later substance abuse? A meta-analytic review of the literature. . 2003;111:179-185.

Pediatrics

17. Michelson D, Faries D, Wernicke J, et al. Atomoxetine in the treatment of children and adolescentswith attention-deficit/hyperactivity disorder: a randomized,placebo-controlled, dose-response study. . 2001;108:e83.

18. Wigal S. Concerta: preference in children. Poster presented at: The American Psychological Associationannual meeting; August 22-25, 2002; Chicago, Ill.

J Atten Disorder

19. Swanson JM, Lerner M, Wigal T, et al. The use of a laboratory school protocol to evaluate concepts aboutefficacy and side effects of new formulations of stimulant medications. . 2002;6(suppl 1):S73-S88.

Clin Pharmacol Ther

20. Swanson J, Gupta S, Guinta D, et al. Acute tolerance to methylphenidate in the treatment of attentiondeficit hyperactivity disorder in children. . 1999;66:295-305.

Psychopharmacol Bull

21. Swanson JM, Wigal SB, Udrea D, et al. Evaluation of individual subjects in the analog classroom setting: I.Examples of graphical and statistical procedures for within-subject ranking of responses to different deliverypatterns of methylphenidate. . 1998;34:825-832.

Arch Gen Psychiatry

22. Swanson J, Gupta S, Lam A, et al. Development of a new once-a-day formulation of methylphenidate forthe treatment of attention-deficit/hyperactivity disorder:proof-of-concept and proof-of-product studies. . 2003;60:204-211.

J Clin Pharmacol

23. Modi NB, Lindemulder B, Gupta SK. Single- and multiple-dose pharmacokinetics of an oral once-a-dayosmotic controlled-release OROS (methylphenidate HCl) formulation. . 2000;40:379-388.

Biopharm Drug Dispos

24. Modi NB, Wang B, Hu WT, Gupta SK. Effect of food on the pharmacokinetics of osmotic controlled-releasemethylphenidate HCl in healthy subjects. . 2000;21:23-31.

Pediatrics

25. Stein MA, Sarampote CS, Waldman ID, et al. A dose-response study of OROS methylphenidate in childrenwith attention-deficit/hyperactivity disorder. . 2003;112:e404.

26. Spencer T, Greenhill L. Treatment of adolescents with ADHD with Concerta. Poster presented at: 50thAnnual Meeting of the American Academy of Child and Adolescent Psychiatry; October 14-19, 2003; MiamiBeach, Fla.

Paediatr Drugs

Paediatr Drugs

Paediatr Drugs

27. Lopez F, Silva R, Pestreich L, Muniz R. Comparative efficacy of two once daily methylphenidate formulations(Ritalin LA and Concerta) and placebo in children with attention deficit hyperactivity disorder across theschool day. . 2003;5:545-555. Erratum in: . 2003;5:661. . 2003;5:832.

Pediatrics

28. Swanson JM, Wigal SB, Wigal T, et al. COMACS Study Group. A comparison of once-daily extended-releasemethylphenidate formulations in children with attention-deficit/hyperactivity disorder in the laboratoryschool (the COMACS Study). . 2004;113:e206-e216.

J Am Acad Child Adolesc Psychiatry

29. Swanson JM, Wigal S, Greenhill LL, et al. Analog classroom assessment of Adderall in children withADHD. . 1998;37:519-526.

J Am Acad Child Adolesc Psychiatry

30. McCracken JT, Biederman J, Greenhill LL, et al. Analog classroom assessment of a once-daily mixedamphetamine formulation, SLI381 (Adderall XR), in childrenwith ADHD. . 2003;42:673-683.

Drug Saf

31. Wernicke JF, Faries D, Girod D, et al. Cardiovascular effects of atomoxetine in children, adolescents,and adults. . 2003;26:729-740.

Am J Psychiatry

32. Michelson D, Allen AJ, Busner J, et al. Once-daily atomoxetine treatment for children and adolescents withattention deficit hyperactivity disorder: a randomized,placebo-controlled study. . 2002;159:1896-1901.

33. Kemner JE, Starr HL, Ciccone PE, Lynch JM. OROS MPH provides greater ADHD symptom improvementthan atomoxetine. Poster presented at: 157th AnnualMeeting of the American Psychiatric Association; May 1-6, 2004; New York, NY.

J Am Acad Child Adolesc Psychiatry

34. Kratochvil CJ, Heiligenstein JH, Dittmann R, et al. Atomoxetine and methylphenidate treatment in childrenwith ADHD: a prospective, randomized, open-labeltrial. . 2002;41:776-784.

35. Faraone S, Biederman J, Spencer T, Aleardi M, Pagano C. Using meta-analysis to draw conclusionsabout ADHD medication effects. Poster presented at: 156th Annual Meeting of the American PsychiatricAssociation; May 17-22, 2003; San Francisco, Calif.

J Am Acad Child Adolesc Psychiatry

36. Greenhill LL, Swanson JM, Vitiello B, et al. Impairment and deportment responses to differentmethylphenidate doses in children with ADHD: theMTA titration trial. . 2001;40:180-187.

J Atten Disorder

37. Conners CK. Forty years of methylphenidate treatmentin attention-deficit/hyperactivity disorder. . 2002;6(suppl 1):S17-S30.