A 12-week, double-blind, randomized, placebo-controlled trial of oral clonidine in three fixed doses (4, 6, and 8 mcg/kg/day) using a crossover design was conducted with 10 children who had hyperkinetic disorder (mean age 7.6 years ±.54). All had comorbid mental retardation. Both parents' ratings on the Parent Symptom Questionnaire and clinicians' ratings on the Hillside Behaviour Rating Scale showed a marked dose-related response to clonidine in hyperactivity, impulsivity, and inattention. Drowsiness was a common side effect of clonidine. It wore off by the 2nd to 4th week in most cases. Thus, clonidine is a safe and effective medication in young hyperkinetic children with comorbid mental retardation.
Hyperkinetic disorders (World Health Organization, 1993) often occur in children with mental retardation (Scott, 1994). Hyperkinetic disorder is comparable with attention deficit hyperactivity disorder (ADHD) of combined type as noted in the Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition—DSM IV (American Psychiatric Association, 1994). Children with hyperkinetic disorders have excessive overactivity and inattention. They exhibit impulsivity, social disinhibition, and behavior problems, which make learning difficult for them. If they already have cognitive impairments due to mental retardation, learning is made even more difficult. Comorbid hyperkinetic disorders and mental retardation in children pose great problems in management for both their parents and teachers. Behavior therapy or pharmacotherapy or both are required to manage children with hyperkinetic disorders. Pharmacotherapy, which reduces hyperactivity and inattention, is needed for children with moderate to severe hyperkinetic disorders. The response of a drug on these symptoms, when part of another disorder, may be qualitatively and quantitatively different. For example, in hyperactive children with autism, central nervous system (CNS) stimulants may cause a withdrawn or “frozen state” and a worsening of stereotypies, so that any benefit due to reduction in hyperactivity is offset by adverse effects (Campbell et al., 1972). It is, therefore, important that comorbidity of symptoms or disorders be taken into account when speaking of effectiveness of a drug on one or more symptoms of hyperkinetic disorders.
Thus, a drug may have a different response on hyperkinetic disorders symptoms when comorbid mental retardation is present. Indeed, ADHD children with comorbid mental retardation show a poor response to and more side effects with CNS stimulants (Green, 1995). Aman, Marks, Turbott, Wilsher, and Merry (1991), in a randomized controlled trial of methylphenidate (CNS stimulant) and thioridazine in children with subaverage lQs and attention deficit disorder (ADD), found that mental age and IQ may be important determinants of drug response. Below IQ 45, there was a greatly reduced likelihood of attention deficit responding positively to methylphenidate both in home or school situations. The CNS stimulants (e.g., Ritalin, Adderal) are the “gold standard” in pharmacotherapy of hyperkinetic disorders. Stimulants have several dose-limiting short-term adverse effects (e.g., such as anorexia, insomnia), and long-term adverse effects on growth. These may exacerbate tics (Taylor, 1994).The alternatives, such as tricyclics and neuroleptics, are also limited by their serious side effects, such as cardiac toxicity and tardive dyskinesia, respectively.
Clonidine may be an alternative drug for hyperkinetic children. Studies of clonidine with placebo (Hunt, Minderaa, & Cohen, 1985) or methylphenidate (Hunt, 1987) as controls showed that clonidine was as effective as methylphenidate in children with ADHD. Clonidine was also reported to be safe and effective in ADHD symptoms comorbid with tic disorder (Leckman et al., 1991; Singer et al., 1995; Steingard, Biederman, Spencer, Wilens, & Gonzales, 1994), autistic disorder (Fankhauser, Karumanchi, German, Yates, & Karumanchi, 1992; Jaselskis, Cook, Fletcher, & Leventhal, 1992), and conduct disorder (Connor, Barkley, & Davis, 2000; Schvehla, Mandoki, & Sumner, 1994). A recent meta-analysis of clonidine for the treatment of children with ADHD reported moderate treatment effects for the common symptoms of ADHD in children and adolescents in doses of .1 mg to .3m per day. Common side effects reported were sedation, irritability, drop in blood pressure, dryness of mouth, and dizziness. They supported the use of clonidine as a second-tier treatment for ADHD (Connor, Fletcher, & Swanson, 1999). To our knowledge, effectiveness of clonidine has not been demonstrated in children with hyperkinetic disorders and comorbid mental retardation. The present study was, therefore, designed to determine the effectiveness of cIonidine in hyperkinetic children with comorbid mental retardation in a randomized, double-blind, placebo-controlled trial using a crossover design. Our aim was to determine dose-related, short-term (6-week) response in children with hyperkinetic disorders and comorbid mental retardation.
Participants and Measures
Participants were all consecutive 6- to 15-year-old children with hyperkinetic disorder who attended the Child and Adolescent Psychiatric Clinic of King George's Medical College, Lucknow. The parents of all children with hyperkinetic disorders and mental retardation, who fulfilled the selection criteria were approached for the study. The parent of 1 child did not give consent. The remaining 10 children (8 boys, 2 girls, mean age 7.6 years (standard error [SE] = .54 months)) were started on the trial. All 10 children completed the study. Four children were studying in primary school, and 6 were not in school. This is understandable because there are no special schools for children with mental retardation in the part of India in which the study was conducted. Children with mild mental retardation go to regular schools as long as they can be mainstreamed.
Mean duration of the hyperkinetic disorder was 4.6 years ± .6. All the children belonged to families of middle socioeconomic status. None of them had a family history of hyperkinetic disorders.
All children had comorbid mental retardation (4, mild; 5, moderate; and 1, severe). Their mean IQ was 48.2 (SE = 4.1, range = 30 to 69). Two children had seizure disorder. One of them was maintained on 500 mg per day of sodium valproate for the last 3 years; the other one was on 600 mg per day of carbamazepine and 400 mg per day of sodium valproate for the preceding 5 years. One girl had congenital hypothyroidism but was euthyriod on 125 mcg per day of thyroxin for the preceding 2 years. Mental retardation was due to congenital hypothyroidism and birth anoxia in one child each. The causes of mental retardation were not known in the remaining 8 children. Five children had comorbid conduct problems as reported on the Parent Symptom Questionnaire and Kiddie Schedule for Affective Disorder and Schizophrenia—Parent Version but not serious enough to merit an International Classification Diseases—ICD-10 (World Health Organization, 1993) diagnosis of conduct disorder. Two children in the past had shown a poor response to neuroleptics. A complete history was obtained for each child from the parent (mother or father). All children received physical examinations. The Hillside Behavior Rating Scale (Gittelman & Klein, 1985) was used as a clinical measure of the symptoms of hyperkinetic disorder. The Kiddie Schedule for Affective Disorder and Schizophrenia–Parent Version (KSADS-P), a semistructured standardized psychiatric interview schedule (Puig-Antich, Chambers, & Tambrizi, 1983) was administered at baseline to identify comorbid psychopathology. The fourth author psychometrically assessed intelligence using the Wechsler Intelligence Scale for Children (WISC) Indian Adaptation (Malin, 1968), Vineland Social Maturity Scale (Vineland– Indian Adaptation, Malin, 1972), and Seguin Form Board Test (Bharath Raj, 1971). Baseline pulse rate, blood pressure, and weight were also recorded.
Each child was seen by one of the first three authors, who are all physicians. Using information from the procedure just described, we made a consensus diagnosis using the ICD-10 classification of mental and behavioral disorders. All the children thus diagnosed as having hyperkinetic disorders and mental retardation were included in the study provided they were not taking any drugs for hyperkinetic disorders for the preceding month and their parents gave informed consent for participation. The exclusion criteria were (a) comorbid psychosis, affective disorder, or pervasive developmental disorder; (b) uncontrolled epilepsy or epilepsy controlled on phenobarbital (c) concomitant use of any other drug known to have an interaction with clonidine, and (d) and contraindication to the use of clonidine (e.g., heart disease).
All participants were randomly placed into clonidine- or placebo-first treatment groups. Clonidine-first children were given clonidine (sugar-mixed suspension with a strength of 10 mcg/ml) in doses of 4-, 6-, and 8-mcg /kg/day (hereafter referred to, as 4-, 6-, or 8-mcg dose) in two or three divided doses for 2 weeks each for a total period of 6 weeks. This was followed by administration of placebo (sugar-mixed suspension of calcium carbonate) identical in look, smell, and taste dispensed in a similar manner for the next 6 weeks. Placebo-first children were given placebo for the first 6 weeks, followed by clonidine in the same schedule for the next 6 weeks.
Each child was assessed on Parent Symptom Questionnaire (Conners, 1970) and Clinical Global Impression Scale (National Institute of Mental Health, 1985) at the baseline. Information on the Parent Symptom Questionnaire was obtained from the parent, and the Clinical Global Impression Scale was rated by the first author. After starting, treatment, children were given weekly assessments on the Parent Symptom Questionnaire, Hillside Behavior Rating Scale, Clinical Global Impression Scale, and Dosage Record Treatment Emergent Symptom Scale (Campbell & Palij, 1985) by the first author throughout the study period of 12 weeks. Blood pressure, pulse rate, and weight were also recorded weekly. The treatment was manipulated by the second author, who was the only author who knew which child received clonidine or placebo in what dose and when and who supplied the treatment suspension weekly to the parents after all assessments were done. The parent, the children, and the clinic rater (the first author) were blind to the content of treatment suspension received by the children. The treatment was started in the clonidine phase of study with the 4-mcg dose and increased to 6- and 8-mcg dose at the end of the 2nd and 4th week, respectively, if improvement did not occur. The criteria for improvement were (a) a rating of 1 on the gross motor activity item from the Hillside Behavior Rating Scale and (b) a rating of not at all on the Parent Symptom Questionnaire items constantly fidgeting, always climbing, unable to stop a repetitive activity, and acts as if driven by a motor.
Comparison between baseline, placebo, and each of the 4-, 6-, and 8-mcg doses of clonidine were analyzed using paired t test (two-tailed). Because there was no significant difference between the baseline and placebo treatment at the end of 2nd, 4th, and 6th weeks, we only report the results of comparison between the baseline and the clonidine treatment.
Parent ratings on Parent Symptom Questionnaire (Table 1) showed significant dose-related improvement on the overall behavior rating and on the Impulsive–Hyperactive factor. One girl achieved a normal activity level on 6-mcg clonidine dose. Thus, there were only 9 children in the 8-mcg clonidine dose group. The level of improvement was more with 6- and 8-mcg doses, ps < .01, than with 4-mcg dose, p < .05. The children on the 8-mcg dose showed significant improvement, p < .02, as compared to 6-mcg dose on Impulsive–Hyperactive factor but not on overall behavior rating. The Conduct factor improved significantly only with 8-mcg dose, p < .05. No significant improvement was seen on other Parent Symptom Questionnaire factors.
Ratings on Hillside Behavior Rating Scale (Table 2) showed significant improvement on gross-motor activity with all the three doses of clonidine. Here also, one girl (the same one as discussed earlier) achieved average activity level on the 6-mcg clonidine dose and, therefore, was not given the 8-mcg clonidine dose. Thus, the number of children in 8-mcg clonidine dose group was only 9. This dose showed maximum effect, p < .01, as compared to 4- and 6-mcg doses. Distractibility and concentration improved more with 6- and 8-mcg doses, ps < .01, than with the 4-mcg dose, p < .05. No significant difference was found between 6- and 8-mcg doses in this regard. Frustration tolerance showed equally significant improvement only with 6- and 8-mcg doses, p < .05. Cooperation improved more with 6- and 8-mcg doses, ps < .02, than with the 4-mcg dose, p < .05. The 8-mcg dose did not improve cooperation further as compared to the 6-mcg dose. Interest in task showed maximum improvement with the 8-mcg dose as compared to 6- and 4-mcg, ps < .05 and .01, respectively. Impulsivity also improved most with the 8-mcg dose, p <. 01.
On the Clinical Global Impression Scale, Severity of Illness (Table 3) showed a linear dose-related improvement, maximizing with the 8-mcg dose (baseline 4.6 [SE =.3] to 2.7 [SE = .5], p < .01). Global improvement (Table 4) showed that 5 children improved minimally, whereas 1 showed much improvement on the 4-mcg dose. On 6 mcg, 2 showed minimal improvement; 5 much improvement, and 1 very much improvement. On the 8-mcg dose, 5 showed much improvement and 2, very much improvement. Two children were nonresponders to clonidine. The 8-mcg dose had maximum Efficacy Index (4 mcg =1.3 [SE =.2]; 6mcg, 1.7 [SE =.2]; 8 mcg, 2.6 [SE .5]).
Five children developed drowsiness, 4 on the 4-mcg dose and 1 on the 8-mcg dose. Out of 4 children who initially developed mild drowsiness on 4 mcg, 2 continued to have mild drowsiness also on 6 mcg. For 1 child it increased to moderate severity, whereas the other developed tolerance to it. Of the 3 children who continued to have drowsiness on 8 mcg, 2 developed tolerance and 1 with moderate drowsiness did not. One child developed mild drowsiness only on the 8-mcg dose. Thus, by the end of 6 weeks of clonidine treatment, only I child each had mild or moderate drowsiness. Drowsiness occurred about 30 minutes to 1 hour after the clonidine dose and lasted for about less than 2 hours. Only 1 girl developed mild dryness of mouth and anorexia on a 6-mcg dose that persisted till the end of 6 weeks. She did not require an 8-mcg dose and, therefore, we are unable to determine whether her dryness of mouth and anorexia would have increased on 8 mcg. Seven children had 3% mean drop in systolic blood pressure on the 4-mcg dose, and all children had 6.9% and 8.9% mean drop on 6- and 8-mcg doses, respectively. The drop in blood pressure did not produce any clinical adverse effect.
In our study clonidine had a dose-related effect on all the core symptoms (hyperactivity, inattention, and impulsivity) of hyperkinetic disorders. Additional symptoms such as those of conduct disorder and frustration tolerance, cooperation, and interest in task were also improved. Clonidine had a differential effect on hyperactivity, impulsivity, and inattention. Hyperactivity and impulsivity improved with increasing doses from 4- to 8-mcg, whereas improvement in inattention occurred only at the 4- and 6-mcg doses. The 8-mcg dose did not further improve inattention.
Findings of this study are similar to those reported by Hunt et al. (1985). They reported a significant reduction in hyperactivity, p = .004, and inattention, p = .02, on 5-mcg/kg/day clonidine dose given over a period of 8 weeks. In an open study, Hunt (1987) found that clonidine increases calmness and frustration tolerance but does not diminish distractibility. In our study distractibility did improve significantly, p <.01, on the 6-mcg dose.
As compared to earlier studies (Connor et al., 1999), children in our study were younger (7.6 years, SD = .05) and had comorbid mental retardation. Although none of the children in our study had diagnosable conduct disorder, its symptoms were present in 5 children. These children improved significantly on clonidine. This has also been observed by Schvehla et al. (1994) and Connor et al. (2000).
Hunt et al. (1985) reported that in clonidine-treated ADHD children, the symptoms return within 2 to 4 days when they were put on placebo. In this study only 4 children received clonidine first and placebo later on. Of these, 2 were nonresponders. The remaining 2 responders showed gradual deterioration in their hyperkinetic disorders symptoms in the 2nd and 3rd weeks of placebo treatment. They deteriorated from much improved to minimally improved category and relapsed completely in the 4th week. A definite comment on how long the effect of clonidine lasts cannot be made because the study design was not appropriate to measure this effect.
Side effects of clonidine in our study were minimal. Drowsiness was the main side effect in 5 children, who developed tolerance in 2 to 4 weeks, with the exception of 1 child. Drowsiness was not dose related because it occurred only in 2 out of 9 children on the 8-mcg dose. It is possible that in these 2 children also, tolerance would have developed had the trial continued. Qualitatively, the parents could well differentiate between drowsiness and beneficial therapeutic effect of clonidine upon the symptoms. Maximum drop in blood pressure was 8.9%, which was not clinically significant.
One might assume that in our study the presence of drowsiness indicated that clonidine was being used, and because of this the parents were not blind to the medication employed: this possibility had to be analyzed. The parents were given treatment suspension (clonidine or placebo) weekly. Ten children were given clonidine or placebo for 6 weeks each. The parents did not know whether the dose or the contents of the treatment suspension would change weekly or not. Thus, if a child developed drowsiness one week but not another, the parents could not know whether the contents of or the dose in the treatment suspension changed or not. Also, of 8 responders on the 4-mcg dose, only 3 (37.7%) developed mild drowsiness and 5 did not. In the latter 5 children, the parents should have expected a nonresponse due to absence of drowsiness, and yet a beneficial response was reported. Of the 2 nonresponders on the 4-mcg dose, 1 child did develop drowsiness, yet his parents reported a nonresponse. The presence of drowsiness does not predict a beneficial therapeutic response perceived by the parents. This is even clearer on the 8-mcg dose, where only 2 out of 7 responders (28.6%) had drowsiness. Finally, statistical analysis of the parental and clinical ratings of the responders who did not develop drowsiness also showed a significant improvement in responders. As an example, the baseline and 8-mcg clonidine dose results were compared in responders who did not develop drowsiness. On the Parent Symptom Questionnaire Impulsive–Hyperactive factor, the children significantly improved on clonidine, p = .001. Also, on the clinical measure Hillside Behavior Rating Scale gross-motor activity, children again significantly improved, p = .002. These results confirm that the therapeutic effects of clonidine were not due to drowsiness and any undoing of blinding was not responsible for therapeutic effects of clonidine in this study.
Only 2 children did not show any response. These children did not have any specific characteristic except that both were boys and had a history of previous nonresponse to neuroleptics given for hyperactivity. One had moderate mental retardation with seizure disorder. The other one had mild mental retardation. Because methylphenidate or other CNS stimulants were (at the time this study was conducted) not available in India, it cannot be said whether these children would have responded to those medications.
Aman et al. (1991) found that methylphenidate is not effective in children with lQs less than 45. In our study, the responders had severe to mild mental retardation. Hunt et al. (1985) reported that girls did not have a favorable response to clonidine. In our study both the girls were responders. Neuroleptics are helpful at reducing hyperactivity in children with mental retardation (Burke & Menolascino, 1968), but in higher, sedating doses, these medications impair learning and cognition. Also, in long-term use, these produce troublesome tardive dyskinesia in children with mental retardation (Paulson, Rizvi, & Crane, 1975). Clonidine in this regard is better. In our study, it produced drowsiness only in 28.6% responding children, even on 8-mcg dose, and maximum drop in blood pressure was 8.9%, which was not clinically significant. However, pulse and blood pressure should be monitored regularly, especially in young children and those with mental retardation because they may not be able to report the cardiovascular side effects. Caution must be exercised while combining clonidine with stimulants or other psychotropic medications. Also clonidine should be taken regularly as per dose schedule, and use as an as-needed medication for sedation should be avoided because abruptly decreasing plasma levels of clonidine, fluctuating stimulant and clonidine plasma levels, or unknown clonidine stimulant interaction may lead to noradrenergic overarousal leading to cardiovascular instability (Cantwell, Swanson, & Connor, 1997).
The results of the present study should be interpreted in view of the following limitations. One, ratings of the parents might have been biased because they might have guessed whether or not the child was on medication because of the improvement in symptoms due to the single crossover design. Second, ratings of the clinical observer might have been biased because he might have guessed medication status of the child by the measurement of the side effects.
The results of our study show that clonidine is an effective drug in treatment of hyperkinetic disorders or ADHD of combined type with comorbid mental retardation. In this group, it may be preferable over the stimulants and neuroleptics because it has fewer side effects. Long-term studies of clonidine treatment on a larger sample are required to confirm the findings.
Authors:Vivek Agarwal, MD, Junior Resident in Psychiatry; Prabhat Sitholey, MD, Professor in Psychiatry; Sudhir Kumar, MD, Associate Professor in Psychiatry; Mata Prasad, Professor in Clinical Psychology, Department of Psychiatry, King George's Medical College, Lucknow (U.P.), India. Reprint requests should be sent to Vivek Agarwal, B-1, 10/69 sector ‘K,’ Aliganj, Lucknow-226024, India.