Copyright 2001 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2001
Piracetam is widely used as a purported means of improving cognitive function in children with Down syndrome. Its efficacy, however, has not been rigorously assessed.
To determine whether 4 months of piracetam therapy (80-100 mg/kg per day) enhances cognitive function in children with Down syndrome.
A randomized, double-blind, placebo-controlled crossover study.
Participants and Methods
Twenty-five children with Down syndrome (aged 6.5-13 years) and their caregivers participated. After undergoing a baseline cognitive assessment, children were randomly assigned to 1 of 2 treatment groups: piracetam-placebo or placebo-piracetam.
Main Outcome Measure
The difference in performance while taking piracetam vs while taking placebo on tests assessing a wide range of cognitive functions, including attention, learning, and memory.
Eighteen children completed the study, 4 withdrew, and 3 were excluded at baseline. Piracetam therapy did not significantly improve cognitive performance over placebo use but was associated with central nervous system stimulatory effects in 7 children: aggressiveness (n = 4), agitation or irritability (n = 2), sexual arousal (n = 2), poor sleep (n = 1), and decreased appetite (n = 1).
Piracetam therapy did not enhance cognition or behavior but was associated with adverse effects.
PIRACETAM is a member of the class of drugs known as nootropics, which are generally thought to enhance cognitive function in instances of brain dysfunction. Putative mechanisms of action differ depending on the disease process being modeled and include enhanced membrane fluidity,1 increased neurotransmitter release (eg, dopamine),2 protective effects on specific receptors (eg, glutamate),3 increased blood flow,4 enhanced corticosteroid function,5 and effects on calcium channel function.6 Piracetam has been administered to patients with diverse clinical conditions such as stroke,7 Alzheimer disease,8- 12 and developmental dyslexia.13- 15 However, research on the efficacy of piracetam and related compounds to ameliorate cognitive deficiencies in these populations, and in animal studies, has frequently produced small or inconsistent results.16
Down syndrome is associated with developmental delay, and affected children generally attain mental and cognitive capacity in the range of mild to moderate mental retardation. Interest in using piracetam to improve cognitive function in children with Down syndrome surged in North America after the television program Day One was nationally broadcast on January 19, 1995. In this program, claims were made that piracetam therapy improved cognitive function in a child with Down syndrome. This was followed by television reports on December 20, 1996 (Nightline), and on August 21, 1997 (48 Hours), and widespread dissemination of anecdotal evidence through Internet newsgroups. For example, parents of a 6-year-old girl indicated that " . . . her concentration and awareness have improved. Her speech has improved to the point that she is finally saying phrases and sentences; improvement is slow but she is finally making some" (May 1995). An 11-year-old boy became "healthier, happier, has more energy, pays attention better, is growing like crazy . . . absolutely full of energy" (January 1999). Partially in response to claims such as these, piracetam has gained widespread use among children with Down syndrome. Although thousands of children are believed to presently receive the drug, no well-designed studies testing the efficacy or assessing the adverse effects of piracetam therapy have been published in the peer-reviewed medical literature.
Unbiased assessment of the effects of piracetam therapy in Down syndrome, beneficial or harmful, is of great importance. Most of the media reports have indicated that cognitive functions such as learning, memory, and attention improve rapidly with the use of piracetam. To test these claims, we designed a randomized, double-blind, placebo-controlled crossover study.
This study used a double-blind, placebo-controlled design consisting of a baseline assessment and two 4-month treatment arms (Figure 1). To control for maturation effects, 2 treatment orders were used. Children were randomly assigned to receive piracetam in phase 1 and placebo in phase 2 (piracetam-placebo) or placebo in phase 1 and piracetam in phase 2 (placebo-piracetam). Cognitive evaluations were conducted at the end of each phase while the children were in school. Baseline assessments were conducted at the end of the 1996-1997 school year, phase 1 testing occurred at the end of the fall 1997 term, and phase 2 testing occurred near the end of the spring 1998 term. This study was approved by the research ethics board at The Hospital for Sick Children in Toronto, Ontario.
Flow chart of study design.
The target sample size (N = 25) was chosen to allow detection of large performance differences (0.8 SD) between the piracetam and placebo phases with power of 80% and α = .05. A large effect size was chosen because most reports in the popular press have indicated immediate and substantive effects of piracetam treatment in children with Down syndrome.
Moderate- to high-functioning children with Down syndrome (aged 6.5-13 years) were recruited through pediatricians and Down syndrome support groups throughout southern Ontario. Initial interviews with the child's parent(s) and teacher determined the child's general health and cognitive abilities. Exclusion criteria were hearing, vision, language, or other physical or cognitive limitations that would interfere with their ability to complete the test battery; known problems in swallowing capsules; use of piracetam during the previous 6 months; and concurrent use of megavitamins.
Twenty-five children passed the initial telephone screening and were invited to The Hospital for Sick Children for baseline physical and cognitive assessment. All children were assessed with the Stanford-Binet Intelligence Scale, 4th Edition,17 to establish mental age equivalents, and the cognitive test battery was also administered. Children were examined by a physician, and a medical history was obtained from the parents. Testing was divided into two 3- to 4-hour sessions, with a break for lunch and additional breaks as needed.
The 14 tests selected for this study were culled from a variety of standardized tests and experimental paradigms and broadly covered the following functional domains: attention, learning and memory, perceptual abilities, executive function, and fine motor and visuomotor skills. Additional criteria were that the tests be suitable for children with Down syndrome and show minimal or no learning or practice effects with repeated administration. Brief descriptions of the tasks are listed in Table 1, and full descriptions can be obtained from the authors. The standardized tests in the battery are typically used to assess children in the 3- to 5 year-old age range, and the remaining tests have been used in 3- to 5-year-old children (eg, see Johnson18) or patients with Down syndrome (eg, see Dalton19). The child's effort on each task was monitored by the tester using a 5-point rating scale (5 = fully compliant and 1 = noncompliant).
Parents and teachers completed standardized questionnaires at each test phase (Table 2). Parent questionnaires provided 80 items that assessed activity levels, social behavior and well-being, stress, parenting and family issues, and the child's temperament. Teacher questionnaires provided 24 items that assessed activity levels, learning, and social behaviors in the school environment.
The cognitive battery was administered by the same tester (V.R.) at all visits to ensure the best possible rapport with the children. The 14 tests were divided into three 25-minute blocks. No more than one 15-minute task was in each block. To control for the order of test administration, 3 block orders were created and children were randomly assigned to 1 of the 6 possible combinations. Children received breaks between each block and as otherwise necessary. Parents sat in a nearby room and were brought into the test room only when poor compliance interfered with testing or the child's verbal responses were not readily understood.
Piracetam was donated in powder form by Medisca Pharmaceuticals (St-Laurent, Quebec) and was encapsulated in gelatin capsules at The Hospital for Sick Children. Children received piracetam, 80 to 100 mg/kg per day, in 3 doses 8 hours apart. Placebo was starch and was encapsulated and administered in the same manner. To monitor adverse effects and establish compliance, parents were interviewed on a weekly to biweekly basis during both treatment arms by a physician. Parents also were asked to note any substantial behavioral or cognitive changes (positive or negative) during each treatment arm in a diary provided to them. Psychologists involved in testing the children were masked to drug status; to ensure safety, a team of physicians who did not participate in the measurement of outcome was unmasked.
Across the 14 tests, 87 dependent variables were identified, including measures such as the number of correct responses and errors (eg, perseverations, omissions, and false alarms), reaction times, and the number of trials to the first correct response. The effectiveness of the randomization to treatment order was confirmed on the baseline data using between-group analysis of variance (SPSS version 10.0; SPSS Inc, Chicago, Ill; piracetam-placebo vs placebo-piracetam). The primary analyses were repeated-measures analyses of variance, with treatment phase (piracetam vs placebo) as the repeated measure. Cases with missing data were excluded on a test-by-test basis, which was necessary for only 5 tests (Go/No-Go, n = 15; Stroop Color/Shape, n = 16; Stroop Color/Word, n = 10; Delayed Match-to-Sample, n = 17; and Animal Pegs, n = 17). Huynh-Feldt–adjusted probabilities30 were used to evaluate the significance of the repeated-measures factors. The strength of all statistical results is indicated by η2, which indicates the proportion of total variability attributable to the factor.
The children's ages fell naturally into 2 nonoverlapping age groups: younger (n = 8; mean [SE] age, 8.0 [0.3] years; range, 7.0-8.9 years) and older (n = 10; mean [SE] age, 11.4 [0.3] years; range, 10.0-13.0 years). To ensure that no treatment effects were being masked by differences in age at baseline, additional analyses were conducted separately for younger and older children.
Based on telephone interviews with parents, we identified a subset of children (n = 11) for whom the parents indicated some cognitive improvement during piracetam treatment (eg, "sharper," "attentive," "brighter," or "more focused"). (Two parents indicated cognitive improvement during placebo treatment, one of whom indicated adverse cognitive effects during piracetam treatment. No parent indicated decreased cognitive performance with placebo use.) The "improvement" group was analyzed separately to determine whether parents' anecdotal comments corresponded to any measurable improvement due to piracetam use. (The diaries provided to the parents were not completed consistently and were not analyzed.) Finally, performance of individual children was examined to determine whether any child demonstrated beneficial effects while taking piracetam.
Twenty-five children were invited to participate based on telephone screening, and 18 were included in the final sample. Three children could not complete the baseline assessment and were not enrolled; their verbal reasoning age equivalents (2.5-3.0 years) were significantly lower than those of the study children. Four enrolled children withdrew during the course of the study: 3 had difficulty swallowing the capsules and withdrew during phase 1 and the fourth required heart surgery in phase 2 (placebo condition). Baseline demographics and performance on the Stanford-Binet Intelligence Scale for the 22 enrolled children are shown in Table 3.
Of the 87 items selected for analysis, 22 reflected easier components of the tasks. These items had no variability because all children had perfect performance, and they were not analyzed further. Raw scores while taking piracetam and placebo for the 65 remaining variables were analyzed. To aid visualization of the magnitude of the differences between the piracetam and placebo arms across tests, all raw scores were converted to z-score differences from the grand average. All data are reported as mean ± SE.
Children were generally compliant during testing, as seen in a mean effort rating of 4.2 ± 0.2 across all tasks (range, 3.4-4.6). Performance on the standardized tests concurred with mental age equivalents determined from the Stanford-Binet Intelligence Scale. For example, on the McCarthy tests, the mean raw scores for Verbal Fluency (9.9 ± 1.3), Verbal Memory (10.6 ± 1.1), and Tapping (2.4 ± 0.2) are typical for children aged 4.5, 3.5, and 4.0 years, respectively. There were no differences at baseline attributable to the randomization to treatment order (P>.09 for all).
The children also worked well at the tasks in the 2 test arms of the study. The mean effort ratings were identical on the piracetam and placebo arms (4.6 ± 0.2; range, 4.2-4.9).
The differences between the piracetam and placebo arms on the primary outcome battery are shown in Figure 2, with reaction times and test scores shown in panel A and errors shown in panel B. Only one measure, an error measure on the Stroop attention task, reached statistical significance (F1,16 = 5.66, P<.03, η2 = 0.26). On average, children made 2.1 ± 0.8 errors while taking piracetam and 3.3 ± 1.2 errors while taking placebo. As shown in Figure 2, the differences between the piracetam and placebo arms were small, with a maximal difference of 0.55 SD.
Mean ± SEM differences (and 95% confidence intervals of the differences [horizontal lines]) between the piracetam and placebo arms for all primary outcome measures (A and B) and significant differences from secondary outcome measures (C). A, Measures such as reaction times, total scores, and number correct. B, Measures that reflect task errors such as false alarms, omissions, naming color instead of shape, and perseverations. C, The 6 measures from the parent and teacher questionnaires that approached or reached significance. Positive numbers indicate better performance while taking piracetam and negative numbers indicate better performance while taking placebo. Scores are expressed as z-score deviations from the grand average of piracetam and placebo scores. The letter in parentheses after each measure name indicates whether the measure came from a test designated as primarily reflecting attention (A), memory (M), perceptual abilities (P), executive function (E), or fine motor skills (FM) and are listed as in Table 1. Strp indicates Stroop; RT, reaction time; Diff, difference; AudCPT, auditory continuous performance task; VisCPT, visual CPT; FA, false alarms; VLearn, Verbal Learning; Max, maximum recalled; E, error; B1, B2, blocks 1 and 2; T1-T4, trials 1-4; DMTS, Delayed Match-to-Sample; CBCL, Child Behavior Check List; CBQ, Children's Behavior Questionnaire; TRF, Teacher's Report Form; and asterisk, significant differences.
At baseline, only one of the questionnaire items differed as a function of assignment to treatment order. Parents of children in the placebo-piracetam group reported greater feelings of isolation (14.1 ± 0.9) than did parents in the piracetam-placebo group (9.8 ± 0.9; F1,16 = 11.61, P<.01, η2 = 0.42). In the comparison of piracetam and placebo scores, 6 items from the parent and teacher questionnaires reached or approached significance (Figure 2C). For parents, these indicated improvements in leadership (FACES III: F1,13 = 4.45, P<.055, η2 = 0.26), fewer thought problems (Child Behavior Check List: F1,15 = 4.93, P<.04, η2 = 0.25), and poorer attention while taking piracetam (Children's Behavior Questionnaire: F1,15 = 5.43, P<.03, η2 = 0.27). Teachers indicated that the children seemed happier (Teacher's Report Form: F1,11 = 8.04, P<.02, η2 = 0.42), had fewer internalizing problems (Teacher's Report Form: F1,13 = 8.65, P<.01, η2 = 0.40), and had fewer total problems while taking piracetam (F1,13 = 5.01, P<.04, η2 = 0.28). We note, however, that all of these differences, although statistically significant, were small from a clinical perspective. For example, on the parent questionnaire, thought problem scores below 67 are not considered clinically significant. Changes of the magnitude found here would not be interpreted by health care professionals as indicative of either improvement or decrease on the factor. Scores while taking piracetam and placebo were within the reference range, at 56.3 ± 1.5 and 60.1 ± 1.6, respectively. Similarly, for teachers, the total problem scores were in the clinically relevant range (>60) at 62.1 ± 2.7 for the piracetam arm and 64.4 ± 2.6 for the placebo arm.
To ensure that age-related beneficial effects of piracetam therapy were not being masked in the primary analysis, we examined the data for the younger and older children separately. Significant treatment × age group interactions were found for only 3 measures. Older children had better performance while taking placebo on 2 measures (Tapping and perseverations in Verbal Learning, P<.05), whereas younger children showed no differences while taking piracetam vs placebo. On the first trial to correct reversal in the Go/No-Go task, young children were better while taking placebo (4.3 trials) and older children were better while taking piracetam (3.9 trials; P<.05 for both).
Analysis of the remarks made during telephone interviews indicated that 11 parents made some comment of improved cognition or attention during the piracetam arm. The data for these children were analyzed separately (mean ± SD age, 10.3 ± 0.6 years; range, 7.0-12.4 years). None of the cognitive measures reached or approached significance in this subsample.
Inspection of performance on a case-by-case basis indicated that occasionally individual children showed meaningful changes in performance (improved or worsened relative to baseline) on 1 or 2 measures while taking piracetam. However, these effects were not consistent across children of similar ages or across tasks within a child.
Adverse effects during the piracetam phase reported by caregivers were mostly associated with central nervous system stimulatory effects and were seen in 7 children: aggressiveness or violent behavior (n = 4), agitation or irritability (n = 2), sexual arousal (n = 2), poor sleep (n = 1), and decreased appetite (n = 1). In 1 boy (aged 10.6 years at baseline), previously noted inappropriate sexual and aggressive behaviors increased during piracetam use and were especially disruptive at home and at school. No such effects were reported in the placebo arm. To ensure that the adverse effects were not overshadowing possible beneficial effects of piracetam treatment, performance was analyzed separately for the 11 children not showing behavioral problems. No beneficial effects of piracetam were identified, and the 2 groups did not differ from each other on any measure.
Piracetam has received significant attention in the media for its purported beneficial effects on cognition in children with Down syndrome. Because enhancement has been reported on several cognitive functions, we explicitly designed the present study to capture a wide range of cognitive functions and behaviors. Thus, our primary outcome measure incorporated tests of sustained and short-term attention and memory; tests that assess processing in the visual and auditory domains; tests that examine verbal abilities, such as word fluency and learning and remembering new words; and basic fine motor and perceptual motor skills. The study is strengthened by the addition of a set of standardized parent and teacher questionnaires to evaluate behavior, temperament, and social and academic performance. The 2 test sessions of primary importance (piracetam and placebo) were conducted within a single school year, and well within each school term, providing a degree of stability to our measurements. It also strengthens the teacher reports, as the same teacher conducted both assessments. Because of the inherent difficulties in assessing changes in cognitive function in cognitively impaired individuals, we were careful to include tasks that children with Down syndrome could complete. In addition, our screening measures ensured that the participating children would have sufficient capacity to understand and perform all tasks. These goals were met for most measurements.
The results of this study indicate no consistent or pervasive beneficial effects of piracetam therapy over placebo use in these children. Across 3 analyses of the formal test battery, 2 measures indicated better performance while taking piracetam and 2 indicated better performance while taking placebo. For the parent and teacher reports, all detected changes were either within the reference range or remained in a clinically relevant range. Given the small number of significant effects and the mixed nature of the findings, we conclude that there is no strong evidence to indicate that piracetam use improved cognition in these children, either in the formal test battery or from parent and teacher reports.
As is typical in clinical trials, some children were lost to the study. Because this reduced our sample size below that needed to detect large effects of the drug, we were careful to examine multiple aspects of the data, including relaxing the threshold for statistical significance, to ensure that positive effects of piracetam treatment were not being masked. We did not identify even a single case that would suggest the possibility that piracetam therapy generally improved cognition. Most telling, perhaps, is that although 11 parents noted that cognitive function seemed to be better when children were taking piracetam, this did not translate into measurable beneficial effects of piracetam over placebo.
The dose of piracetam used by us was within the range used in adults and in children with dyslexia.13 Despite this, potentially serious adverse effects were experienced by 7 of the 18 children while taking piracetam. This finding excludes the option of dose escalation.
The results of this study are strikingly different from the anecdotal testimonials presented in the popular press. One possibility that might partially account for the perceived beneficial effects of piracetam therapy might lie in the stimulatory effects of this medication. This type of behavioral stimulation could easily be confused with actual cognitive improvement in the absence of objective measures. Especially relevant to parent observations of alertness and better focus, none of our attention tasks revealed enhanced performance while taking piracetam over placebo on any measure. In conclusion, the results of this study indicate that piracetam treatment is associated with adverse effects and does not substantially enhance cognition or behavior.
Accepted for publication December 1, 2000.
This research was supported by the Motherisk Research Fund and by a grant from The Scottish Rite, Toronto, Ontario.
Presented in part at the Pediatric Academic Societies/Society for Pediatric Research annual meeting, San Francisco, Calif, May 3, 1999.
The Down Syndrome Association of Toronto, provided valuable help in recruiting participants for this study. Piracetam was donated by Medisca Pharmaceuticals. Arthur J. Dalton, PhD, graciously allowed us to use the Dalton/McMurray Visual Memory Test. We thank Zina Levichek, MD, and Dionne Laslo, MA, for their assistance in data analysis; Orna Citrin, MD, for help with evaluating adverse effects of medication use; and especially the parents and children who participated in the study.
Corresponding author: Nancy J. Lobaugh, PhD, Imaging Research and Cognitive Neurology Unit, Sunnybrook and Women's College Health Sciences Centre, 2075 Bayview Ave, Room S604, Toronto, Ontario, Canada M4N 3M5 (e-mail: firstname.lastname@example.org).
Lobaugh NJ, Karaskov V, Rombough V, Rovet J, Bryson S, Greenbaum R, Haslam RH, Koren G. Piracetam Therapy Does Not Enhance Cognitive Functioning in Children With Down Syndrome. Arch Pediatr Adolesc Med. 2001;155(4):442-448. doi:10.1001/archpedi.155.4.442