ADHD symptom presentation and trajectory in adults with borderline and mild intellectual disability

  • Published on

  • View

  • Download

Embed Size (px)


  • ADHD symptom presentation and trajectory in adultswith borderline and mild intellectual disabilityjir_1270 668..677

    K. Xenitidis,1 E. Paliokosta,2 E. Rose,3 S. Maltezos1 & J. Bramham4

    1 South London and The Department of Maudsley Foundation Trust, Adult ADHD Service, London, UK2 Institute of Psychiatry, Child and Adolescent Psychiatry, London, UK3 University of Hertfordshire, Hatfield, UK4 University College Dublin, UCD School of Psychology, Dublin, Ireland


    Background This study examined symptoms andlifetime course of Attention Deficit HyperactivityDisorder (ADHD) in adults with borderline andmild Intellectual Disability (ID).Method A total of 48 adults with ID and ADHDwere compared with 221 adults with ADHDwithout ID using the informant Barkley scale forchildhood and adulthood symptoms.Results The ADHD/ID group presented withgreater severity of (adult and childhood) symptomscompared with the non-ID group. For the ADHD/non-ID group, most symptoms improved signifi-cantly from childhood to adulthood, whereas onlytwo symptoms changed significantly for the IDgroup. Principal component analysis revealed scat-tered loading of different items into five compo-nents for the ADHD/ID group that were notconsistent with the classic clusters of inattentive,hyperactive and impulsive symptoms. A negativecorrelation was found between severity of symptomsand IQ.

    Conclusions ADHD in adults with ID may have amore severe presentation and an uneven and lessfavourable pattern of improvement across thelifespan in comparison with adults without ID.

    Keywords adults, Attention Deficit HyperactivityDisorder, ID, symptoms profile


    Attention Deficit Hyperactivity Disorder (ADHD)is a common neurodevelopmental disorder causingsignificant distress and impairment. (Faraone et al.2000) DSM-IV diagnostic criteria require at leastsix symptoms in each domain of inattention andhyperactivity/impulsivity for the combined type tobe diagnosed or symptoms in one domain for theinattentive or hyperactive type, respectively. ICD-10classification of hyperkinetic disorder correspondsto the DSM-IV criteria of the combined type.ADHD diagnostic validity for adults has been onlyrecently established. (Toone 2004; Zwi &York2004) Comorbidity is the norm rather than theexception (Kutcher et al. 2004), and this often com-plicates the diagnostic process and affects treat-ment. More specifically, the coexistence of

    Correspondence: Dr Elena Paliokosta, Institute of Psychiatry,Department of Child and Adolescent Psychiatry, London, UK(e-mail:

    Journal of Intellectual Disability Research doi: 10.1111/j.1365-2788.2010.01270.x

    volume 54 part 7 pp 668677 july 2010668

    2010The Authors. Journal Compilation 2010 Blackwell Publishing Ltd

  • hyperactivity and ID has long been noted but thevalidity of the diagnosis of ADHD in people withID has recently attracted attention (Antshel et al.2006). However, studies have shown a higherprevalence (Dekker & Koot 2003) and a worseprospective outcome of ADHD in children andadults with ID (Lambert et al. 1987; Peterson et al.2001).Diagnostic overshadowing may lead to the attri-

    bution of hyperactive, impulsive and inattentivesymptoms to the ID itself and thus obscure theidentification of ADHD in this population (Reiss &Szyszko 1983) as clinical presentation of ADHD inthis group is not well established. (Sevin et al. 2003)Additionally, clinicians feel more confident aboutmaking a diagnosis of ADHD in patients withoutID (Jopp & Keys 2001; Gillberg et al. 2004) ratherthan in those with ID.This could be also due to adifficulty in establishing whether activity and atten-tion levels are consistent or not with the develop-mental stage of the individual with ID as DSM-IVcriteria indicate.The phenomenology of ADHD and ID has been

    studied much more in children and adolescents.Although the presentation of mental disorders inpeople with ID is often atypical, there is evidencethat children with ID and ADHD exhibit classicalsymptoms of ADHD such as poor selective atten-tion (Pearson et al. 1996), off-task behaviour andfidgeting compared with their ID peers who dontpresent with ADHD symptoms. (Handen et al.1998) Ishii et al. (Ishii et al. 2003) found no statisti-cally significant difference in the presentation ofDSM-IV symptoms of ADHD in children with IDcompared with children of normal intelligence.Using observational measures, Fee et al. (Fee et al.1994) also concluded that children with ID andADHD show a pattern of ADHD symptoms similarto that of normal intelligence children diagnosedwith ADHD. Simonoff et al. (Simonoff et al. 2007)found a negative linear relationship between ADHDsymptoms and IQ (beta = -0.087, P < 0.001) inadolescents with mild ID. Neither the profiles ofADHD symptoms nor the comorbidity withemotional/behavioural problems differed accordingto the presence of ID.These findings suggest thatthe clinical presentation of ADHD is similar acrossindividuals with different intellectual levels inchildhood.

    However, it is not clear whether these observa-tions also apply to adults with ADHD and ID andwhether the symptoms improve from childhood toadulthood as it is the case for non-ID individualswith ADHD. Adults with ADHD and mild or bor-derline ID assessed by our team at the AdultADHD Service at the Maudsley Hospital werefound to be significantly more impaired in terms ofselective attention and response inhibition com-pared with adults with ADHD and normal cogni-tive function during neuropsychological assessment(Rose et al. 2009). The present study aims todescribe the clinical presentation and course ofADHD in adults with ID and compare ADHDsymptomatology in adults with and without ID.Thefollowing research questions were addressed:1 What is the ADHD presentation of adults withID? Based on the neuropsychological findings men-tioned above we hypothesised that adults with IDwould present with increased severity of their symp-toms compared with non-ID adults.2 Is the symptom trajectory from childhood intoadulthood different for the two groups?3 Do the symptoms in the ID group co-occur inthe same pattern as in the non-ID group?4 Finally, how do symptoms correlate to IQ?



    This study was a retrospective study based on thecopies of Barkley Scales (informant version)(Conners & Barkley 1985) for adulthood and child-hood symptoms found within the medical caserecords of adults who received clinical assessmentsfor ADHD in a specialist outpatient clinic.


    The study was conducted at the Adult ADHDService at the Maudsley Hospital, London.This is aSpecialist Service established in 1994 providingdiagnostic assessments and treatment for adultswith ADHD from throughout the UK.


    The sampling frame of the present study consistedof all adults who attended the clinic between 2001

    669Journal of Intellectual Disability Research volume 54 part 7 july 2010

    K. Xenitidis et al. ADHD in adults with intellectual disability

    2010The Authors. Journal Compilation 2010 Blackwell Publishing Ltd

  • and 2006. All patients diagnosed with ADHD(n = 269) who had completed neuropsychologicalassessment and provided informant rating scaleswere included in the sample. Of those, 48 patientswere found to have IQ equal to or below 80and were defined as the ADHD/ID group, while221 patients with IQ above 80 were the ADHD/non-ID group.We decided to include people in theborderline (IQ between 70 and 80) range of intel-lectual functioning in the ADHD/ID groupbecause of the similarities in the relevant manage-ment issues. Additionally, because the number ofthe patients in the ID group was relatively small itwas not possible to conduct separate analysis forthe borderline intelligence group. All patients wereattending the clinic for assessment of their currentADHD symptoms. 52% of the ID group and 42%of the non-ID group had a previous diagnosis andhad received some treatment in the past for the dis-order; however, the exact treatment regimens foreach patient were not available.

    Measures of ADHD symptoms

    Given that ADHD is a neurodevelopmental disor-der, it is necessary to establish that symptoms werepresent in childhood. Pre-assessment questionnaireswere posted to each patient referred, in order toobtain preliminary data about ADHD symptoms inchildhood and adulthood.These include theBarkley Scales (patient and informant version)(Barkley & Murphy 1998; Conners & Barkley 1985)for adulthood and childhood symptoms.This scalecontains the 18-symptom items for ADHD fromDSM-IV-TR that are scored from 0 (never orrarely) to 3 (very often) and a total score derivedfrom adding symptom scores. All data analyses wereconducted based on informant-rating scales. Thereason for choosing informant vs. self-report scaleswas the fact that the latter were missing or incom-plete in a significant proportion of patients with IDbut also because diagnosis during childhood isbased traditionally on parents/informant reportsabout hyperactivity.A comprehensive neuropsychological evaluation

    was conducted at the time of assessment includingtheWechsler Adult Intelligence Scale-III (Wechsler1997) or theWechsler Abbreviated Intelligence

    Scale. (Wechsler 1999) A DSM-IV diagnosis ofADHD was assigned by experienced clinicians.

    Statistical analysis

    We conducted KolmogorovSmirnov test to checkour measures regarding normal distribution. Asresults were consistent with non-normal distributionfor item scores although distribution was normal fortotal scores, non-parametric tests, MannWhitney,were conducted to compare total Barkley score aswell as each item score for the ID and non-IDgroup. Subsequently, we conductedWilconxonsigned ranks test, a non-parametric test, withingroups to compare adulthood vs. childhood symp-toms scores in order to examine different course inthe symptomatology between the ID and the non-IDgroup.We further conducted a principal componentanalysis of Barkley scale to investigate different pat-terns of loading of items. A rotated componentmatrix was produced usingVarimax method withKaiser Normalization. Finally, we correlated itemsscores with IQ score. Significance level for all testswas set at P < 0.05. All analyses were conductedusing spss statistical package. (Field 2005)


    The ADHD/ID group consisted of 34 men and 14women, while the ADHD/non-ID group of 140men and 81 women.The ADHD/ID group (n = 48)had a mean age of 24.63 years (range 1643,SD = 7.394) and mean IQ of 70.27 (range 5280,SD = 8.115). The ADHD/non-ID group (n = 221)had a mean age of 29.31 years (range 1757,SD = 8.908) and mean IQ 106.27 (range 81145,SD = 14.295). There was no significant differencefor age and sex distribution between the twogroups. Cronbachs alpha for our data is 0.878,which shows high internal reliability. KolmogorovSmirnov z-test was significant for all symptomsitems except the Current Barkley total score sonon-parametric tests were conducted.

    Symptom severity: comparison of ADHD/ID groupwith ADHD/non-ID group

    Adulthood symptoms

    Total Barkley score in adulthood was used as anindicator of symptom severity for ADHD/ID group

    670Journal of Intellectual Disability Research volume 54 part 7 july 2010

    K. Xenitidis et al. ADHD in adults with intellectual disability

    2010The Authors. Journal Compilation 2010 Blackwell Publishing Ltd

  • compared with ADHD/non-ID group.There was astatistically significant difference with ADHD/IDgroup scoring higher compared with ADHD/non-ID group. Results are presented in Table 1.ADHD/ID group had also higher scores in allBarkley items. In five out of nine inattentive itemsand in four out of nine hyperactivity/impulsivityitems these differences were statistically significant.

    Childhood symptoms

    Similarly, ADHD/ID group presented a statisticallysignificant higher total Barkley score in childhood.As indicated in Table 1, this group differed signifi-cantly from the non-ID group in eight out of nineinattentive and six out of nine hyperactivity/impulsivity symptoms.

    Symptoms within trajectory: group comparisonbetween childhood and adulthood scores

    Within each group, paired samples statistics wereconducted for each item comparing childhood andadulthood scores in order to assess the lifetimecourse of the disorder. For the ADHD/non-IDgroup there was significant decrease in ratingsindicative of improvement over time.This was thecase for the total score of Barkley scale as well asfor almost all symptom items in both domains ofinattention and hyperactivity/impulsivity (Table 1).This is a biopsychological pattern consistent withthe general notion that ADHD symptoms tend toimprove with (brain and psychosocial) maturation.For the ADHD/ID group, although there was anoverall decrease in scores, there were significant dif-ferences only for two inattentive and two hyperac-tive items and the total score.

    Symptoms pattern: clustering of symptoms in theADHD/ID and the ADHD/non-ID group

    Symptoms principal component analysis for adult-hood symptoms revealed that three components forthe ADHD/non-ID group accounted for 64.26% ofthe variance for ratings. All the inattentive symp-toms loaded for the first component and all threeimpulsivity items (blurting out answers, difficultywaiting turn and interrupting) and one hyperactivity

    item (talking excessively) loaded for the secondcomponent. Hyperactivity symptoms loaded for thethird component.For the ADHD/ID group, five components were

    identified with scattered loading of different itemsto account for 73.98% of the variance for ratings.The first component included the three impulsivitysymptoms (blurting out answers, difficulty waitingturn and interrupting) and one hyperactivity item(talking excessively) suggesting that these symptomscould represent a discrete cluster of difficulties forthe adult ADHD/ID patients. However, the othercomponents did not follow any particular patternwith only four inattentive symptoms (organisationproblems, avoiding mental effort, losing things andforgetting) loading with one hyperactive symptomin component 2.

    Correlation of symptoms withintellectual functioning

    Pearsons product moment correlations between IQand Barkleys total score in adulthood were con-ducted and showed a negative correlation betweenscores and IQ (-4.15, P < 0.001). A negative corre-lation was also found for each Barkleys item sepa-rately. This means that the lower the IQ the higherthe Barkley scores and this was the case when eachgroup was analysed separately and when the wholesample was analysed altogether.


    Our results indicate that the clinical presentation ofADHD in adults with ID has certain differenceswhen compared with non-ID adults with ADHD.Mean total score on Barkley scale was significantlyhigher in adulthood for the ADHD/ID group indi-cating greater severity. This difference was evenmore pronounced for childhood total score. Addi-tionally, both inattentive and hyperactivity/impulsiveitems on Barkley scale were higher rated for theADHD/ID group.It could be argued that higher scores for ADHD

    symptoms in chil...


View more >