`Efficacy of a Mixe d Amphetamin e Salts
`Compound in Adults With Attention-Deficit/
`Hyperact ivity Disorder
`
`Thomas Spencer, MD; Joseph Biederman, MD; Timothy Wile11S, MD; Stephen Faraone, PhD; Jefferson Prince, MD;
`Kristine Gerard, MD; Robert Doyle, MD; Asha Parekh, MD; Jake Kagan, BA; Sarah Kate Beannan, BA
`
`Backg-u•"• We report on a controll ed trial of a
`mixed amphetamine salts compound (Adderall , de."Xtro(cid:173)
`amphetamine sulfate, dextro -, levoamphetamine sul(cid:173)
`fate , dextroamphetamine aspartate, levoamphetamine
`aspartate, and dextroamphetamine
`saccharate) in th e
`treatment of adult attention-deficit/hyperactivity disor(cid:173)
`der (ADHD) .
`Meth•"•• This was a 7-week , randomiz ed, double (cid:173)
`blind , placebo-controlled , crossover study of Adderall
`in 27 well-characterized adults satisfying full DSM-IV
`criteria for ADHD of childhood onset and persistent
`symptoms into adulthood. Medication was titrated up
`to 30 mg twice a day. Outcome measures included the
`ADHD Rating Scale and the Clinical Global Impres(cid:173)
`sion Score . Comorbid psychiatric
`disord ers were
`assessed to test for potential effects on treatment out(cid:173)
`come.
`
`Results, Treatment with Adderall at an average oral dose
`of 54 mg (administ ered in 2 daily doses) was effective and
`well tolerated. Drug-specific improvement in ADHD symp(cid:173)
`toms was highly significant overall (42% decrease on the
`ADHD Rating Scale, P<. 001), and sufficiently robust to
`be detectable in a parallel groups comparison restricted
`to the first 3 weeks of the protocol (P<. 001). The percent(cid:173)
`age of subjects who improved (reduction in the ADHD rat(cid:173)
`ing scale of ~300,6) was significantly higher with Adderall
`treatment than with a placebo (70% vs 7%; P = .001).
`
`Co•clu slo••• Adderall was effective and well tolerated
`in the short-term treatment of adults with ADHD. More
`work is needed to evaluate the long-term effects of Adder(cid:173)
`all, or other amphetamine compounds , in the treatment
`of adults witl1 ADHD.
`
`Arch Gen Psychiatry. 2001;58:775-782
`
`I N CHILDREN with attention-deficit/
`
`hyperactivit y disorder (ADHD),
`the literature suggests that the per(cid:173)
`centage of responders is compa(cid:173)
`rable between the stimulants.'
`However, crossover studies of dextroam(cid:173)
`phetamine and methylphenidate (6 stud(cid:173)
`ies , 274 subjects) reveal differences in
`response on the individual level. Of re(cid:173)
`sponders , 52% responded equally well to
`both, 25% preferentially to amphetamine ,
`and 23% to meth ylphenidate. 2
`7 However,
`•
`these differences in response may be be(cid:173)
`cause of either efficacy or adverse effects.
`In adults with ADHD, controlled
`studies have reported an average re (cid:173)
`sponse of 54% of subjects both to meth(cid:173)
`ylphenidate (6 studies, 139 subjects) and
`pemohne (2 studies, 93 subjects) .8 To our
`knowledge, the only previous controlled
`trial of amphetamines in adults with ADHD
`was a recent short -tem1 study of dextro(cid:173)
`amphetamine in adults with broadly de(cid:173)
`fined ADHD indicating efficacy.9 In addi(cid:173)
`tion , there is a controlled study in normal
`men 10 as well as several case studiesu· 12 and
`•14 in adults with ADHD. For
`open series 13
`
`example, in a placebo-controlled , single(cid:173)
`dose crossover study of dextroamphet
`amine in nomlal men (N = 31) , Rapoport
`et al10 reported improved cognitive per(cid:173)
`formance . In an open , 6-week trial of dex(cid:173)
`troamphetamine in 18 adults with ADHD,
`dramatic changes were reported in behav(cid:173)
`ior, but not on cognitive measures. 13
`
`See also page 784
`
`Despite the well-documented effi(cid:173)
`cacy of stimulant drugs in the treatment of
`ADHD, their short duration of action com(cid:173)
`monly requires a 3-times-daily dosing
`schedule to obtain a daylong clinical ef(cid:173)
`fect. In children with ADHD the preva (cid:173)
`lence of after-school stinmlant use has in(cid:173)
`creased.15 Such after-school dosing has been
`recommended for AD HD-associated non(cid:173)
`academic adaptive dysfunctions in daily liv(cid:173)
`ing, communication , and socialization
`skills. 16 Similar adaptive dysfunctions are
`salient in adults with ADHD. Thus , a sim(cid:173)
`plified dosing regimen with a longer (cid:173)
`acting compound could be particularly im(cid:173)
`portant for adults with ADHD.
`
`From the Pedialric
`Psychophamwcology Unit ,
`Massachusetts General Hospital
`(Drs Spencer , Biederman ,
`Wilens, Faraone , Prince,
`Gerard , Doy le, and Parelili,
`Mr Kagan , and Ms Bearman)
`and the Department of
`Psycliiatry ; Harvard Medical
`School (Drs Spencer ,
`Biederman, Wilens, Faraone ,
`Prince, Gerard, Doyle ,
`and Parekh) , Boston .
`
`(REPRINTED) ARCH GEN PSYCHIATRY/VO L 58, AUG 2001
`775
`
`WWW .ARCHGENPSYCHIATRY.COM
`
`<02001 American Medical Associ
`Downloaded From: http ://archpsycj amanctwork .com/ by Brooke Raymond on 12/17/201
`
`Page 1
`
`SHIRE EX. 2020
`KVK v. SHIRE
`IPR2018-00290
`
`(cid:173)
`
`
`SUBJECTS AND METHODS
`
`SUBJECTS
`
`Subjects were 30 outpatient adults withADHD between 19
`and 60 years of age ascertained from clinical referrals. To
`be included , subjects had to satisfy full diagnostic criteria
`for DSM-IV ADHD based on clinical assessment con(cid:173)
`firmed by structured diagnostic interview. Attention-deficit/
`hyperactivity disorder diagnoses , including age of onset by
`7 years, were detenniued by self-report as well as school
`records and report by others as available. We excluded po (cid:173)
`tential subjects if they had any clinically significant chronic
`medical conditions, abnonnal baseline laboratory values,
`IQ less than 80 , delirium , dementia , or amnestic disor(cid:173)
`ders, any other clinicall y unstable psych iatric conditions
`(ie, bipolar disorder , psychosis ), drug or alcohol abuse or
`dependenc e within the 6 months preceding the snldy, pre(cid:173)
`vious adequate trial of Adderall, or current use of psycho(cid:173)
`tropics. We also excluded pregnant or nursing females. This
`study was approved by the institutional review board and
`all subjects completed a written informed consent before
`inclusion in the study.
`
`PROCEDURE
`
`This was a doubl e-blind , placebo-controlled, randomized,
`crossover trial, comparing Adderall with placebo. There were
`two 3-week treatment periods separated by 1 week of wash(cid:173)
`out to minimize carryover effects of medication . During
`washout, subjects received placebo pills to maintain the
`blind . The order of treatment (Add erall , placebo , or pla(cid:173)
`cebo, Adderall) was randomized by the resear ch phar(cid:173)
`macy . Weekly supplies of Adderall or placebo were dis(cid:173)
`pensed by the pharma cy in identical-appearing
`10-mg
`capsules. Study ph ysicians pre scribed medication under
`double-blind conditions in twice-a-day dosing (7:30 AM,
`2:30 PM). Compliance was monitored by pill counts a teach
`physician visit. Study medication was titrated up to 20 mgld
`(10 mg twice daily) by week 1, 40 mgld (20 mg twice daily)
`by week 2, and 60 mgld (30 mg twice daily) by week 3,
`unless adverse effects emerged. Although drug or placebo
`status was randomized , dose within each phase was not.
`Study treatment was always titrated from low to high dose
`
`to avoid exposure to high initial doses of active medica(cid:173)
`tion and to minimize adverse effects. Other psychoactive
`medications were not permitted during the protocol.
`
`ASSESSMENT
`
`Before inclusion in the study , patients underwent a com(cid:173)
`prehensive clinical assessment that included a psychiatric
`eva luation by a board-certified adult and child psychia(cid:173)
`, KG., R.D., and A.P.), a structured
`trist (T.S., T.W.,J.P.
`diagnostic interview , a medical history , and laboratory as(cid:173)
`sessments (liver function tests, complete blood counts, and
`electrocardiograms)
`. The structured diagnostic interview
`used was the Structured Clinical Interview for DSM-N, 211
`supplemented for childhood disorders by modules (DSM(cid:173)
`N ADHD and conduct disorder ) from the Kiddie Sched(cid:173)
`ule for Affective Disorders and Schizophrenia for School(cid:173)
`Age Children (Epidemiologic Version). 29 Diagnostic raters
`estimated a level of ADHD impairment (mild, moderate,
`or severe) by assessing the degree of dysfunction (social,
`familial, academic, and occupational) specifically attribut(cid:173)
`able to the ADHD symptoms.
`To have been given a full diagnosis of adult ADHD, the
`subject must have (1) met full DSM-N criteria (at least 6 of
`9 symptoms) for inattentive or hyperactive/impulsive sub(cid:173)
`types30 by the age of 7 years as well as currently ( within the
`past month) ; (2) described a chronic course of ADHD symp(cid:173)
`toms from childhood to adulthood; and (3) endorsed a mod(cid:173)
`erate or severe level of impairment attribu ted to the ADHD
`symptoms. Diagnostic reliability of the structured
`inter(cid:173)
`views was established by having 3 experie nced , board(cid:173)
`certified child and adult psychiatrists diagnose the condi(cid:173)
`tions of 35 subjects from audio taped interviews made by the
`assessment staff. The mean K was 0.91. A K of 1.0 was ob(cid:173)
`tained for ADHD with a 95% confidence interval of0.8 to 1.0.
`To assess intellectual functioning , we administered sub(cid:173)
`tests of the Wechsler Adult Intelligence Scale, Revised3l and
`the Wide Range Achievement Test , Revised. 32 1.earning d.is(cid:173)
`abilities33 were defined by the procedure recommended by
`Reynolds l< that provides a statistical method for operation(cid:173)
`alizing the differ ence between achievement and intelli(cid:173)
`gence scores . Famil y history was determined by question(cid:173)
`ing the subject about the presen ce of psychiatric disorders
`in first- or second-degree relatives . Socioeconomic status
`was measured by the Hollingshead Four-Factor Index of
`
`There are sustained-release preparations available
`for both methylphenidate
`and dextroamphetamine
`(spansules ). While some reports have indicated an equal
`response to methylphenidate immediate-release and sus(cid:173)
`tained-release,'·17·18 others have not. 19-22 Dextroamphet(cid:173)
`ami.ne spansules may be more consistently effective\ how(cid:173)
`ever, few studies have examined effectiveness beyond 4
`hours ."·23 While relatively long-acting , pemoline has been
`relegated to second-line status because of concerns about
`hepatotoxicity . 2
`"
`An additional longer-acting amphetamine product
`is the mixed amphetamine salts compound, Adderall.
`Adderall consists of 25% levoamphetamine and 75% de,'{(cid:173)
`troamphetamine in 4 salts. Recent controlled studies in
`children have reported that Adderall is as effective as meth(cid:173)
`ylphenidate immediate-release in tl1e in1provement ofbe-
`
`havior in classroom and recreational settings and in in(cid:173)
`creased academic performance, and that the time course
`of the response is longer , as shown in detailed pharma(cid:173)
`26 While there has been no direct
`codynarnic studies .25
`•
`comparison between Adderall and dextroamphetamine ,
`there are theoretic reasons for potential differences . Pre(cid:173)
`vious reports comparing levoamphetamine with dextro(cid:173)
`amphetamine have suggested that some children re(cid:173)
`spond preferentially to each isomer. 27
`We now report results of a randomized, placebo(cid:173)
`controlled clinical trial of an amphetamine (Adderall) in
`the treaonent of adults with AD HD. We hypothesized that
`a twice-daily dosing regimen of Adderall at clinically rel(cid:173)
`evant doses will be effective in the treatment of adults
`with ADHD and provide adequate daylo.ng coverage of
`their symptoms.
`
`(REPRINTED) ARCH GEN PSYCHIATRY/VOL 58, AUG 2001
`776
`
`WWW .ARCHGENPSYCHIA TRY.COM
`
`C2001 American Medical Associ
`Downloaded From: http: //archpsycJamanetwork.co1n/ by Brook e Raymond on 11/17/2015
`
`Page 2
`
`
`
`Social Status /' with low values indicating high socioeco(cid:173)
`nomic status.
`To assess change during treatment, board-eligible or
`board-certified psychiatrists used the following scales. Over(cid:173)
`all severity and change in severity of ADHD was assessed
`with the Clinical Global Impression Scale.36 The Clinical
`Global Impression Scale includes global severity (1, not ill,
`to 7, extremely ill ) and global improv ement (1, very much
`improved, to 7, very much worse) scales. The ADHD Rat(cid:173)
`ing Scale,37 shown to be sensitive to drug effects in pedi(cid:173)
`
`atric38 and adult39-" 2 populations , assesses each of the 18
`individual criteria symptoms of ADHD in DSM-Non a se(cid:173)
`verity grid (0, not present ; 3, severe; overall minimum score ,
`O; maximum score, 54) . Five raters independentl y re(cid:173)
`viewed audiotaped interviews of 5 subjects. An intraclass
`correlation of 0. 99 was obtained for interrater reliability of
`the ADHD symptom checklist. For depression , we used the
`17-item Hamilton Depression Scale (HAM-D) (minimum ,
`O; maximinn , 52)' 3 and the Beck Depression Inventory
`(minimum , O; maximum, 63) ..... For anxiety, we used the
`Hamilton Anxiety Scale (HAM-A) (minimum , O; maxi (cid:173)
`mum , 56). ~' The presenc e of adverse ex periences was elic(cid:173)
`ited by open-ended questions at each visit. We adminis(cid:173)
`tered the HAM-D, HAM-A, and Beck Depressio n Inventory
`before and after each arm of the study . All other symptom
`rating scales were administered weekly. Raters were blind
`to treatment assignment.
`Since ADHD has been associated with cognitive impair(cid:173)
`ments ,46 we included neuropsy chological measures to test for
`potential drng effects on cognition . Based on our review of
`the literanu e and our previous neuropsyc hologic studies with
`ADHD children and adults, 46 we selected neuropsychologi(cid:173)
`cal tests that measure sustained attention, response inhibi(cid:173)
`tion , set shifting and categorization, selective attention and
`visual scanning, and organization and recall of visual con(cid:173)
`structions. The test battery included an auditory version of
`the Continuous Performance Test: 7
`·~ the Stroop test : 9 and
`the Rey-Osterrieth Complex Figure. ' 0 This neuropsychologi(cid:173)
`cal battery was administered 3 times, at baseline and after each
`arm of the study.
`
`SUBJECT CHARACTERISTICS
`
`thr ee prospective participants applied for
`One hundred
`entry into the study . Of these, 30 were enrolled. Of the
`
`73 who were not enrolled , 17 did not complete the ini(cid:173)
`tial evaluation; 14 were excluded because of current sub(cid:173)
`stance or alcohol abuse; 12 did not meet full DSM-N cri(cid:173)
`teria for ADHD; 11 met entry criteria but were unable to
`commit to the demands of a co ntroll ed study; 12 were
`excluded for medical conditions and/or current use of
`concomitant medications
`(4 seizures, 3 sensory motor
`impairment, 2 hypertension , and 3 other); and 7 were
`exduded
`for unstable psychiatric conditions or current
`use of psychotropics
`(3 psychosis, 1 bipolar , and 3
`depression ).
`in the study, 27 (90%) com(cid:173)
`Of the 30subjectsenrolled
`pleted it . Three subjects did not complete the first treat(cid:173)
`ment arm : 1 after the first week and 2 after the second week,
`and were not included in the final analyses . These 3 pa(cid:173)
`tients were receiving placebo and never received Adderall.
`Thus , tl1e final sample consisted of 15 men and 12 women
`(age: mean± SD, 38± 9 .3 years). Seventy-eight percent (2 1/
`27) met criteria for ADHD combined type in childhood ( 44%
`[12127] currently) and 22% (6127) met criteria for ADHD
`predominantly
`inattentive
`type in childhood
`(5 6% [15/
`27] currently). No one met criteria for ADHD predo(cid:173)
`minantly hyp eractive-imp ulsi ve type. Ten (37%) of the 27
`subjects had been diagnosed as having ADHD previously
`and had received other medications (8 other stimulants;
`2 desipramine ).
`
`STATISTICAL ANALYSIS
`
`The primary outcome measures were the ADHD Rating Scale
`and the Clinical Global lmpression Scale. Improvement was
`defined either as a 30% reduction in the ADHD Rating Scale
`or "much " or "very much improved " on the Clinical Global
`Impression Scale. For statistical tests of change between 2
`points in time, we used the McNemar test (for binary data ),
`the paired t test (for continuous data ), or the Wilcoxon
`signed rank test (for ordinal data ). For analyses that used
`all of the time points in our data set, we used random ef(cid:173)
`fects , cross-sectional time-series models using the method
`of generalized estimating equations ( GEE) as described by
`Liang and Zeger' 1 and Zeger et al52 These models esti(cid:173)
`mated main effects of drug (Adderall vs placebo), time ( week
`in study), and order (Adderall first vs placebo first), as well
`as interactions among these effects. Significance was set at
`the .05 level and all tests were 2-tai led .
`
`RI '>l I I'>
`
`As depicted in Taltl• I , 93% (N = 25) of ADHD subjects
`had at least 1 lifetime comorbid psychiatric disorder.
`The mean±SD number of comorbid diagnoses was
`2. 9 ± 2.5 per subject. Baseline ratings of depression
`(HAM-D, 4.3, and Beck Depression Inventory , 6.2) and
`anxiety (HAM-A, 6.0) were low. Using standard cutoff
`points for moderate severity on ratings of depression
`(HAM-D, > 16; Beck Depression Inventory , >19) and
`amdety (HAM-A, >2 1), only 11% (N=3) of subjects had
`baseline scores of depression or arnd.ety that were mod(cid:173)
`erately severe or worse. Sixty-seven percent of ADHD
`adults had 1 or more first- or second-degree relatives
`with ADHD. Despite average to above-average intelli(cid:173)
`gence (mean±SD, 108±11), 37% of the subjects
`
`required tutoring in school and 19% had repeated at
`least 1 grade.
`
`EFFICACY
`
`Averaged across both periods, at week 1 the average daily
`doses of Adderall and placebo were both 20 mg; by week
`2, 38.5 mg and 40 mg; by week 3, 53.7 mg and 59.3 mg,
`respectively. Examining the first and second periods sepa(cid:173)
`
`rately, inspection of the ,1.,... shows some evidence of a
`
`carryover effect in that the mean value of the ADHD
`Rating Scale of the medication-first group at week 4
`(placebo-washout) did not fully return to the baseline
`(Figure, B). However, the order effect (medication first vs
`medication second) was not significant (random effects:
`z = 0. 99, P = .3 2). Despite the weak order effect, we found a
`
`(REPRINTED ) ARCH GEN PSYCHIATRY/VOL 58, AUG 2001
`777
`
`WWW.ARCHGENPSYCHIATRY.COM
`
`Downloaded From : http://archpsycJamanetwork.com/
`
`«:>2001 American Medical Assoc
`by Brooke Raymond on 12/17/201
`
`
`
`Page 3
`
`
`
`Table 1. Cllnlcal and Demographic Characteristics
`of Sample (N = 27)*
`
`Oancgr~cs
`Male, No. (%)
`'Mite, No. (%)
`Jlge, mea, (ID) , y
`Soc:ioea:nrnic staus, mea, (s:>)t
`
`Psychiaric disordErs, No. (%)
`M~or ~ession with severe irfl)airma,t
`M~or ~ession with a lest
`moderate irfl)airma,t
`Multiple .;nxiety disocdErs (2:2)
`Pl. lest 1 m>dety dsorda'
`9.Jbstanoe ~ IOB
`pjcohol dependEnoe
`1¥rtisocial personality dsocda'
`Oni.ld disorda'
`kiy comorbid disorda'
`Pa;t GAF, mea, (SO)
`OJrrent GAF, mea, (S))
`
`Rmly hlstocy of dsocda's, No. (%)
`J>Dl-0
`Depression
`kixiety
`kitisocial personality
`9.Jbstanoe abuse dependEnoe
`
`Qig,itivetesting , mea, (SO)
`W3dlsiE!' PdiJt lntallga,ce S;;ales
`R'eedom fran dstr~ibility
`IQ
`RJll-scale IQ
`
`.Adllevement srores, mea, (SO)
`INRAT subscale percentiles
`Pl'itrmetic
`Remng
`
`J',aamc U1da'ariietEmS1t , No. (%):j:
`Pl'ithmetic
`Rea:ing
`School faih.l'e, No. (%)
`Repalled gra:Je
`Ainment in spEriai da;;s
`Tutc:xing
`
`15(56)
`26(96)
`38.8 (9.27)
`2.0(0 .73)
`
`Past
`
`Current
`
`2(7)
`12(44)
`
`1 (4)
`3(11)
`
`1 (4)
`7 (26)
`
`5(19)
`14 (52)
`4(15)
`7(26)
`0(0)
`6(22)
`0(0)
`6(22)
`6 (22)
`25 (93)
`53(5 .C6)
`61 (4.58)
`
`18(67)
`15(56)
`10(37)
`3(11)
`11 (41)
`
`102(12.05)
`100(11.32)
`
`47 (23 .85)
`00(22 .64)
`
`7(26)
`0(0)
`
`5(19)
`2(7)
`10(37)
`
`, ient of RJndioning; !'D-0 ,
`*GAF i ndicaes Globa PSSE!SSI.
`atentimdetidthlyper~ivity disorda'; INRAT. 'Mde Paige .AdlletEmS1t
`Test; and alipses, not appliatiie
`tS:Jdoa:onomic status W<S meE6U'ed by the Hollingsheaj Four-Fa::.tc:x
`Incle< cA Social staus ,36 with lc,,y values indicaing hig, socioeo:nonic
`sta:us.
`:j:leer'nng dsabilities33 were defined by the procejll'e ra:armended by
`Reynolds,"' 'A>tiich prO\lides a staistical method fc:x opera:ionali2ing the
`dffE!'ence between a::hievement end intaliga,oe sco-es.
`
`significant effect of drug for both the first period ( week 0-3:
`Adderall , 15; placebo , 12; z =5.5 , P<. 001) and the second
`period (week 4-7: Adderall, 12; placebo, 15;z =5.7, P<. 001).
`In addition the average change scores (ADHD Rating Scale)
`were similar in each period of the study (35% vs 51 % de(cid:173)
`crease while receivingAdderall and 5% vs 7% increase while
`receiving placebo; first vs second period, respectively). While
`none of the subjects worsened while receiving Adderall ,
`55% (15/27) worsened while receiving placebo.
`In addition , we analyzed the results after combin(cid:173)
`ing the first and second period . Response to Adderall at(cid:173)
`tained significance by the first week of treaunent (z = 3 .1,
`
`P= .002 ), with further improvement by week 2 (z=5 .6,
`P< .001) and week3 (z =6.3, P< .001) . Overall , there was
`a very significant drug by ti.me interaction
`for ADHD
`symptoms (z = 6.4 , P< .001 ) without significant main ef(cid:173)
`fects of drug (Adderall or placebo ) or time (baseline and
`weeks 1, 2, and 3) .
`To further evaluate the absolute rate of improve (cid:173)
`ment , we analyzed end-of-treaune nt results (averaged
`across both periods ) using a preestabli..shed definition of
`improvement of more than a 30% reduction on the ADHD
`Rating Scale (see the "Subjects and Metl10ds" section). Us(cid:173)
`ing this definition , 70.4% (19/27) of patients showed im(cid:173)
`provement of ADHD symptoms while receiving Adderall
`compared with only 7.4% (2/27) who were receiving pla(cid:173)
`cebo (x 21=13.8 , P<. 001). Similarly, when improvement
`was defined as much or very much improved on the Clini(cid:173)
`cal Global Improvement Scale, 66 . 7% (18/27) of patients
`receiving Adderall were rated as improved compared with
`only 3.7% (1127) receiving placebo (x 2
`, = 14.2, P<. 001).
`In addition, Adderall treatment significantly reduced the
`Global Severity Scale ratings ofADHD (4.7 ±0.7 to 3.4± 1.0;
`z=4.3, P<. 001). In contrast, placebo did not (4.6±0.7 to
`4.4±0 .9; z =0 .8, P= .45).
`Adderall treaunent (averaged across both periods )
`was associated with clinica lly an d statistically signifi(cid:173)
`cant improvement of all but 2 of the 18 individua l ADHD
`symptoms, with the notable effects obseived for symp(cid:173)
`toms in both subclusters of hyperactivity/i..mpulsivity
`and inattention (T•ltle 2 ). However, fewer of the indi(cid:173)
`vidual hyperacti ve/i..mpulsi vity items would achieve
`significance when corrected for multip le comparisons .
`In contrast , the effect of placebo on individua l ADHD
`symptoms was negligible.
`Other than race and socioeconomic status, our sample
`represented a group of adults with diverse clinical charac(cid:173)
`teristics (Table 1). Therefore , we examined each of these
`characteristics as potential confounders . While we did not
`have sufficient power to fully examine this issue, a detaile d
`analysis revealed no effects of ADHD subtype (combined
`vs predominately inattentive ), sex , age, hi.story of comor(cid:173)
`bid disorders , lifetime hi.story of treaonent , current comor (cid:173)
`bid disorders , or positive family hi.story of psychiatric dis(cid:173)
`order on rates of improvement while receivi..ngAdderall or
`placebo . As mentioned earlier, baseline ratings of depres(cid:173)
`sion (HAM-D, BDI) and anxiety (HAM-A) were very low
`and were not affected by treaonent with Adderall.
`At baseline , adults with ADHD in this sntdy per(cid:173)
`formed comparably with non-AD HD adults on some cog(cid:173)
`nitive tests (Rey-Osterrieth Complex Figure, Stroop [In(cid:173)
`terference condition]). On other tests (Stroop [Word and
`Color], and Continuous Performance Test ) adults with
`ADHD were found to have mild difficulty at baseline with
`equal improvement while receiving medication and pla(cid:173)
`cebo (T•ltl• 3 ) .
`
`ADVERSE EFFECTS
`
`Adverse effects were reported on the entire sample , at any
`time during treatment and tabulated in T•ltle 4 . Adder(cid:173)
`all was well tolerated and no serious adverse effects were
`obseived . Of individual adverse effects reported , only Adder(cid:173)
`all-associated appetite supp ression and agitation reach ed
`
`(REPRINTED} ARCH GEN PSYCHIATRY/VO L 58, AUG 2001
`778
`
`WWW.ARCHGENPSYC HIA TRY.COM
`
`©2001 American Medical Assoc
`Downloaded From : http://archp sycjaman etwork.com/ by Brooke Raymond on 12/17/201
`
`Page 4
`
`
`
`
`
`Table 3. Neuropsychological Functioning at Baseline and
`at the End of Treatment of the Placebo and
`Adderall Conditions*
`
`Baseline
`Placebo
`Stroop Test
`Wo:d T-Soore
`44.3 (9. 7)
`48.3ft (9.6)
`COio: T-Scx:lre
`41.0 (8.7)
`43.6t11 (9.4)
`COio: 'M:lrd T-Scx:lre 43.4 (10.4)
`49.511[ (11.1)
`lnta'fen,·ireT-Scx:lre 49.7(6 .6)
`53.1111(7.6)
`
`Add era II
`
`49.91'§ (8. 7)
`46.ot§ (8.2)
`48.6t11 (11.8)
`50.1 (7.8)
`
`Q:ipy O'ga,ization
`Q:ipy kt:,Jacy
`Delay O'ga,izatioo
`Delay/!a:uracy
`
`Rey-Osterrieth Complex Figure
`10.4 (3.4)
`10.5 (3.3)
`63.6 ( 4.3)
`63.5 (3.5)
`7.1 (3.9)
`8.6 (3. 7)
`41.5(10 .0)
`51 (10.4)
`
`Hts
`Onissioos
`L.ae
`
`Continuous Perlonnance Test, No.
`76.7
`84.5
`20.6
`14.4
`2.78
`1.04
`
`9.6 (3.4)
`64.0 (2.1)
`8.2(4 .1)
`50.3(7 .8)
`
`85.7111
`12. 7t11
`1.39
`
`* Wues are meen (SO) unless othe'wise indicated.
`t'J.; baleline by 'Mlcoxon signed rri
`test.
`tf' < .05.
`§P < .0001.
`IIP<. 01.
`,JP<. 001.
`
`Table 4. Adverse Events While Receiving Adderall
`and Placebo
`
`Adverse Event
`
`lnsormia
`L.c:ss a cqieute
`Depressioo
`Pnxiety
`1-teBkhe
`Dy rrouth
`,Agitaion
`F!iig..ie
`lndlgestioo
`l.t'ira-y Ired infedioo
`Gistrcintestinal pan
`Pailc atta::k
`"8.Jsea
`Sirus problems
`B'onchitis
`ClxJgl
`Cbnfusia,
`Light -heje:j
`
`Drug,
`No.(%)
`
`10 (37)
`8(29 .6)
`1 (3.7)
`7(25 .9)
`3(11.1)
`4(14 .8)
`6 (22.2)
`1 (3.7)
`1 (3.7)
`1 (3.7)
`1 (3.7)
`1 (3.7)
`1 (3.7)
`0(0 .0)
`0(0 .0)
`0(0 .0)
`1 (3.7)
`1 (3.7)
`
`Placebo,
`No. (o/o)
`
`4(14 .8)
`3(11 .1)
`2(7.4)
`4(14 .8)
`2 (7.41)
`3(11.1)
`2(7 .4)
`1 (3.7)
`2(7 .4)
`0(0 .0)
`0(0 .0)
`0(0 .0)
`0(0 .0)
`1 (3.7)
`1 (3.7)
`1 (3.7)
`0(0 .0)
`0(0 .0)
`
`x'*
`3.6
`5.0
`1.0
`1.8
`0.33
`1.0
`4.0
`0.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`1.0
`
`p
`
`.06
`.03
`.32
`.18
`.56
`.32
`.05
`>. 99
`.32
`.32
`.32
`.32
`.32
`.32
`.32
`.32
`.32
`.32
`
`*Lsing McNerra' exa:t test.
`
`at an average oral daily dose of 54 mg , was well toler(cid:173)
`ated and effective . Although this was a crossover de(cid:173)
`sign , reduction
`in ADHD symptoms was sufficiently
`robust to be detectable in a parallel group comparison
`during the first 3 weeks of the protocol (P< .001). These
`results confirm the study hypothesis and suggest that
`Adderall is a well-tolerated and effective treatment for
`adults with ADHD.
`Although the duration of drug action was not mea(cid:173)
`sured directly , our results suggest that twice-daily dos(cid:173)
`ing of Adderall may be comparable with meth ylpheni(cid:173)
`date using a thrice-daily dosing .39 Subjects in our study
`indicated that twice-daily dosing was sufficient to cover
`the entire day and there was no subjective sense of medi(cid:173)
`cation "wear-off " in between doses .
`These results extend to amphetamines other find(cid:173)
`ings documenting a highly similar pattern of drug re(cid:173)
`sponsivity between children and adults with ADHD to
`anti-ADHD medications including methylphenidate, 39
`pemoline ,42 desipramine / 0 tomoxetine ,11 and bupro(cid:173)
`pion. 53 The similarities in drug response across the life
`span provide further support for the informativeness of
`trials of adults with ADHD in drug development pro (cid:173)
`grams for ADHD.
`Traditional analyses in clinical trials examine out(cid:173)
`come using a cutoff score . We used a 30% cutoff of
`ADHD symptoms
`(ADHD Rating Scale ) to define
`improvement.
`In another report , we have addressed
`the issue of cutoffs by use of a novel analytic
`tech(cid:173)
`nique, the drug-placebo response curve, that examines
`the entire range of symptom change scores .54 In addi (cid:173)
`tion, in this study the negligible overall response
`to
`placebo was composed of individuals who in1proved
`and individuals who worsened. Another pharmaco(cid:173)
`therapy study of adults with ADHD also revealed
`worsening while receiving placebo .5
`"
`
`Although we evaluated a range of neuropsychologi(cid:173)
`cal outcomes using a battery of tests that measures e.'(ecu(cid:173)
`tive functions , most subjects performed well on this bat(cid:173)
`tery. The relative good function at baseline is consistent
`with other studies showing adult ADHD to be associated
`with mild neuropsychological deficits.16 This created a "ceil(cid:173)
`ing" effect that did not allow the detection of medication(cid:173)
`associated cognitive improvements . The development of
`tests that are more sensitive to neuropsychological dys(cid:173)
`function in ADHD adults may be required to assess fully
`the effect of phannacological treatments .
`The absence of meaningful associations between
`Adderall treatment and ratings of anxiety and depression
`indicate that Adderall-associated ADHD improvement was
`unlikel y to be secondary to improvement in co morbid de(cid:173)
`pression or amdety in our sample . They also indicate that
`Adderall treatment was not associated with worsening of
`an.·dety or depression in this sample that had a frequent
`history of comorbidity with these disorders . While we did
`not have sufficient statistical power to fully e.'(amine the
`effects of potential confounding factors, the absence of sex
`and comorbidity effects in the treatment response of adults
`with ADHD is consistent with prior studies with other
`medications, and does not support the practice of exclud(cid:173)
`ing comorbid cases in clinical trials of adults with ADHD.
`Although none of our subjects suffered from pre(cid:173)
`existing hypertension , patients with poorly controlled hy(cid:173)
`pertension may not be eligible for stimulant treatment
`until their blood pressure is well controlled. Special moni(cid:173)
`toring may be required in patients with borderline hy(cid:173)
`pertension receiving Adderall or other stimulant drugs .
`Until more is known about long-term treatment in adults ,
`periodic assessment of blood pressure may be war(cid:173)
`ranted in patients e:1,,-posed to stimulants.
`The results of this smdy should be viewed in light
`of methodological
`limitations. These include the rela-
`
`(REPRINTED ) ARCH GEN PSYCHIATRY/VO L 58, AU G 2001
`780
`
`WWW .ARCHGENPSYCHlATRY .COM
`
`Downloaded From: http ://archpsycjamanctwork.com/
`
`<t>lOOl American Medical Assoc
`by Brooke Raymond on 12/17/201
`
`Page 6
`
`
`
`tively small subject size, use of a crossover design, and a
`relatively short eh'Posure to medication . While studies
`in children suggest a rapid response to stimulants , in clini(cid:173)
`cal practice a more gradual dose escalation is the rule .
`In our study , the dose was increased weekly; thus we can(cid:173)
`not disentangle dose and time effects. It is possible that
`continued exposure would lead to increased effective(cid:173)
`ness of long-tem1 Adderall treatment . While the use of a
`crossover design provides increased statistical power , the
`evidence of a minor carryover effect would suggest that
`in future studies , a longer washout period or a parallel
`design may be more optimal. Nevertheless, reduction
`in ADHD symptoms was robust enough to be detectable
`in a parallel group comparison . Despite their robust(cid:173)
`ness , our results could not address the impact of Adder(cid:173)
`all on functioning and quality of life. Such information
`is critical to further inform the risk vs benefit analysis of
`treatment with Adderall. Longer studies with appropri (cid:173)
`ate instrumentation
`assessing these domains will be
`needed to address these important issues.
`Despite these linntations , this study has shown that
`Adderall significantly improved ADHD symptoms and was
`well tolerated. These promising initial results provide sup(cid:173)
`port for further studies of Adderall or other amphet(cid:173)
`amine compounds in the treatment of adult ADHD us (cid:173)
`ing a wide range of doses over an extended period of
`treatment and with more detailed assessment of func(cid:173)
`tioning and quality of life.
`
`Accepted for publication October 31, 2000.
`Supported in part by