CLINICAL SCIENCES Clinically Relevant

The effects of exercise on children with attention-deficit hyperactivity disorder

MARY TANTILLO, CHRISTINA M. KESICK, GEORGE W. HYND, and ROD K. DISHMAN Departments of Exercise Science and Special Education, The University of Georgia, Athens, GA; and U.S. Army Research Institute of Environmental Medicine, Natick, MA

ABSTRACT TANTILLO, M., C. M. KESICK, G. W. HYND, and R. K. DISHMAN. The effects of exercise on children with attention-deficit hyperactivity disorder. Med. Sci. Sports Exerc., Vol. 34, No. 2, pp. 203212, 2002. Purpose: The effects of exercise on children with attention-deficit hyperactivity disorder (ADHD) were evaluated by studying the rate of spontaneous eye blinks, the acoustic startle eye blink response (ASER), and motor impersistence among 8- to 12-yr-old children (10 boys and 8 girls) meeting DSM-III-R criteria for ADHD. Methods: Children ceased methylphenidate medication 24 h before and during each of three daily conditions separated by 24 48 h. After a maximal treadmill walking test to determine cardiorespiratory fitness (VO2peak), each child was randomly assigned to counterbalanced conditions of treadmill walking at an intensity of 6575% VO2peak or quiet rest. Responses were compared with a group of control participants (11 boys and 14 girls) equated with the ADHD group on several key variables. Results: Boys with ADHD had increased spontaneous blink rate, decreased ASER latency, and decreased motor impersistence after maximal exercise. Girls with ADHD had increased ASER amplitude and decreased ASER latency after submaximal exercise. Conclusions: The findings suggest an interaction between sex and exercise intensity that is not explained by physical fitness, activity history, or selected personality attributes. The clinical meaning of the eye blink results is not clear, as improvements in motor impersistence occurred only for boys after maximal exercise. Nonetheless, these preliminary findings are sufficiently positive to encourage additional study to determine whether a session of vigorous exercise has efficacy as a dopaminergic adjuvant in the management of behavioral features of ADHD. Key Words: ANXIETY, EYE BLINKS, METHYLPHENIDATE, MOTOR IMPERSISTENCE, PERSONALITY, SLEEP DISTURBANCE, STARTLE

bout 35% of school-age children have attentiondeficit hyperactivity disorder (ADHD), characterized by inattention and/or hyperactivity-impulsivity (1,2). Children with ADHD have a high rate of disciplinary problems in school, and they comprise 30 40% of the referrals to child guidance clinics (6). The most commonly prescribed drug for the treatment of ADHD, methylphenidate, is moderately effective (50), but not all children with ADHD respond favorably; about 20% experience side effects including high blood pressure, sleep problems, or mood disturbances (6). Thus, the potential benefits of other, behaviorally oriented, approaches to treatment merit study (44). Anecdotal reports from parents and teachers have suggested that vigorous exercise is beneficial for children with ADHD. If so, brief periods of exercise could practically be included in the school curriculum. We know of four published studies that reported mixed positive and null effects of physical
0195-9131/02/3402-0203/$3.00/0 MEDICINE & SCIENCE IN SPORTS & EXERCISE Copyright 2002 by the American College of Sports Medicine Submitted for publication November 2000. Accepted for publication May 2001.

activity on children with attentional and/or hyperactivity disorders. Those studies failed to either: 1) test a clinically diagnosed population (27,35), 2) quantify or standardize the exercise stimulus (15,27,35), or 3) implement valid behavioral measures (27,35,44). None used measures related to putative mechanisms of ADHD that might explain an effect of exercise, and all but one (27) were case reports of one or two children. The underlying neuroanatomical structures and neurophysiological mechanisms that explain ADHD are not fully known, but research suggests that brain dopaminergic systems are involved (6). Methylphenidate is a dopamine (DA) agonist (43) that has high binding affinity in the corpus striatum (53). Children diagnosed as ADHD have morphological asymmetry in the caudate nucleus of the striatum (10,17,23), which is mainly dopaminergic. Though a recent neuroimaging study indicated no effect of acute treadmill running on D2 receptor binding (an index of DA release) in the putamen of young men and women (54), studies of laboratory rats have found elevated DA levels (18,22) and turnover (21) in the striatum after acute treadmill running and increased extracellular DA in the nucleus accumbens of the ventral striatum after forced swimming (55). Moreover, c-fos expression (an index of neuronal activation) in the 203

striatum after acute treadmill running appears dependent on activation of D1 receptors (29). Hence, we reasoned that a test of the effects of exercise on children diagnosed with ADHD should include measures related to brain dopaminergic activity. Neuroimaging is costly and impractical in most research settings. Spontaneous eye blinks and the acoustic startle eye blink response (ASER) are technically simple, noninvasive measures that are sensitive to dopamine agonists (11,14,25) and have been used as dopaminergic probes among children diagnosed with ADHD (9,20). Spontaneous blink rates are correlated positively with dopamine levels in the caudate nucleus (49). The ASER is an obligatory, reflexive contraction of the orbicularis oculi muscle elicited by the presentation of a standardized acoustic stimulus that is processed from the auditory nerve by ventral cochlear root neurons projecting to the nucleus reticularis pontis caudalis and on to the facial nerve (11). Pharmacologic agonists of DA receptors augment startle amplitude in rat (12), and increased sensitivity of D1 receptors in the striatum has been implicated as a mediator of this effect (33), which appears partly dependent on a synergistic activation of D2 receptors (34). Because treadmill exercise training increases D2 receptor density in rat striatum (31), it is plausible that acute physical activity might lead to increased blink rate and an augmented ASER response. The diagnostic value of spontaneous blink rate and the ASER for ADHD has not yet been tested, so it is also important to determine whether exercise might alter behavioral measures that have clinical relevance for ADHD. Emerging evidence indicates that a hallmark feature of ADHD in children is motor impersistence (52), defined as the inability to sustain simple motor acts such as maintaining a conjugate gaze, keeping the mouth open, protruding the tongue, fixing the eyes centrally, or holding the eye lids shut. Voeller and Heilman (52) have provided evidence that the inability to inhibit perceptual-motor reflexes is related to the behavioral disinhibition commonly seen in children with ADHD. A reduction of motor impersistence after exercise, concomitantly with increases in spontaneous blink rate and the ASER, would suggest potential efficacy of exercise for helping to manage ADHD. We measured spontaneous eye blinks, the ASER, and motor impersistence before and after maximal and submaximal treadmill bouts of exercise in children diagnosed with ADHD and compared their responses with those after a quiet rest condition and with responses in a group of children without ADHD who served as controls. We hypothesized that acute exercise would lead to an increase in the rate of spontaneous eye blinks, an increase in amplitude and decrease in latency of the ASER, and a decrease in motor impersistence among children with ADHD that would be dependent on exercise intensity.

and currently taking methylphenidate (10 20 mg, 2 4 times per day, with weekend use as needed) and 25 children without ADHD (11 boys and 14 girls) aged 8 12 yr were recruited from the community by using flyers and newspaper advertisements. All ADHD children were judged to be positive responders to stimulant medication treatment. None wore contact lenses. One boy in the ADHD group was a nonresponder to the acoustic stimulus, and another boy in that group did not participate in the submaximal exercise condition. ASER data on the day of the VO2peak condition were not obtained for one girl in the control group because of equipment failure. Children and their parents affirmed that caffeinated beverages had not been consumed within 4 h of testing. Children ceased methylphenidate medication 24 h before and during each of three daily conditions separated by 24 48 h. Methylphenidate has a duration of action of 1 4 h and a pharmacokinetic half-life of 2 4 h (26,42). Within the dosages used by our participants, brain clearance is about 90 min (53). Experimental Design Upon entering the laboratory, a parent of each participant completed and signed a medical history questionnaire, a consent form approved by the Institutional Review Board, and the Pittsburgh Sleep Quality Index (PSQI) (8) for the child. The child also signed a consent form and completed the Trait version of the State-Trait Anxiety Inventory for Children (STAI-C) (48) and the Junior Eysenck Personality Questionnaire (JEPQ) (16). Participants also completed a daily physical activity record (5), had their height measured, and were weighed on balance scales. The sum of triceps brachii and calf skinfolds measured by a Lange caliper was used to estimate percent body fat (46). Both the ADHD and control participants underwent two bouts of exercise and one quiet rest condition, each on consecutive days. The time of day when testing occurred was standardized for each child. The exercise intensities were chosen to be consistent with intensities that have evoked a brain dopaminergic response in rat (21) and to determine whether responses to exercise were dose dependent. The first bout of exercise was a peak oxygen uptake (VO2peak) test on a motor-driven treadmill for the purposes of determining physical fitness and providing a condition of maximal exercise intensity. The second bout of exercise was conducted at a submaximal intensity ~6575% VO2peak. The durations of the submaximal exercise and rest conditions were matched for each child with the time required to complete the maximal exercise test. The second and third days of testing consisted of either a submaximal treadmill exercise bout or quiet rest in a sound attenuated, thermoneutral (22 1C) chamber administered randomly, resulting in a balanced order among groups. State anxiety, assessed by the STAI-C, rate of spontaneous eye blinks, ASER amplitude (ASERAMP) and latency (ASERLAT), and motor impersistence assessed by the Motor Impersistence Battery (MIB) (52) were administered before and after each condition. The JEPQ was administered on
http://www.acsm-msse.org

METHODS
Participants Eighteen children (10 boys and 8 girls) clinically diagnosed according to the DSM III-R guidelines for ADHD (1) 204
Official Journal of the American College of Sports Medicine

two of the three testing days immediately after administration of the STAI-C. The PSQI was administered on the first day of testing. During the rest condition, participants watched a silent cartoon (The New Adventures of Winnie the Pooh: Alls Well That Ends Well) (The Walt Disney Company, Burbank, CA) designed to standardize visual and auditory cues that affect reflexive eye blinks (45). Participants rated their interest in the video and how pleasant they thought the video was on separate 10-point scales. Verbal anchors on the two separate scales at the number 1 were very interesting or very pleasant, and 10 was not interesting or not pleasant. Ratings of pleasantness and interest in the video were similarly neutral (ranged from 4 to 5) across diagnosis and sex.

MIB The MIB required the participants to complete, for as long as possible up to 1 min, the following motor tasks: eye closure, tongue protrusion, right and left lateral gaze fixation (point specified), central fixation without distraction, right and left lateral gaze fixation (point unspecified), central fixation with distraction, and head turning to the right and left. Participants were given two trials per task. The mean was used for analysis. Spontaneous Eye Blinks and the Acoustic Startle Eye Blink Response After mild abrasion of the skin with alcohol swabs, two Beckman silver/silver-chloride miniature surface electrodes were filled with electrode cream and affixed to the participant by using double-stick adhesive collars. One electrode was placed 0.25 0.50 cm inferior to the exocanthus of the left eye and the other electrode was placed suborbitally 1 cm medial and slightly inferior to the first electrode. A third pregelled, silver/silver-chloride ground electrode (Eaton Electrode Co., Ann Arbor, MI) was placed on the forehead. Electrode impedance was measured with a Grass Instruments (Astro-Med, Inc., Warwick, RI) model EZM 4, elecat each trode impedance meter and maintained at 2 4 electrode site. The mean impedance at each electrode site was the same for each group and did not change from preto post-treatment in either group (P 0.40). The ASER was elicited by the binaural delivery of a burst of white noise via Sony model MDR-200 stereo circumaural headphones. The acoustic stimulus was a 50-ms burst at 95 dB(A) with a rise and fall time of 10 ms. Ten trials of the stimulus occurred 8 15 min before and after each condition with an intertrial period of 2535 s. Spontaneous eye blinks were recorded for a period of 3 min before elicitation of the ASER. The broadband noise was produced by a GrasonStradler 455C (West Concord, MA) noise generator gated through a Massey Dickinson (Massey Dickinson, Saxonville, MA) amplifier. The amplifier was used to generate an appropriate stimulus intensity and set the rise and fall time of the stimulus. The noise stimulus was calibrated by a General Radio 1551-C (General Radio Co., Concord, MA) sound level meter. The ASER and spontaneous eye blinks were measured by electromyography (EMG). The surface muscle biopotential was amplified 1000 2000 by using a Grass Instruments model P511 amplifier with frequencies below 3 Hz and above 1000 Hz filtered. The raw EMG signal was integrated and rectified by a MED Associates Inc. (Georgia, VT) dual-following integrator by using a nearly instantaneous time constant. A Kenwood CS-8010 digital storage oscilloscope (Kenwood, Tokyo, Japan) was used to store and display the output waveform from the integrator. Latency and amplitude of the stored waveform were measured digitally by using cursors displayed on the oscilloscope. In the few cases when the amplitude onset was not dramatic, the measurement of latency was defined by a deflection from the prestimulation baseline level to 10% of the total amplitude of the ASER waveform response to the
Medicine & Science in Sports & Exercise

STAI-C Forms C-1 and C-2 each contain 20 items that measure state, i.e., temporary, and trait, i.e., general, anxiety, respectively. An exaggerated startle response is a feature of generalized anxiety disorder (2), and comorbidity of ADHD with anxiety disorders is common (47). Thus, comparisons between the ADHD and control groups on the dependent variables after exercise might be confounded by group differences in anxiety. Also, anxiety could confound sex comparisons, as girls score higher on the STAI-C than boys (48).

JEPQ This 81-item questionnaire assesses three primary personality traits: introversion-extraversion, emotional stability-neuroticism, and psychoticism, as well as conformity, which is an index of motivated distortion of scale responses or lying. Eysenckian personality theory contends that extraverts tend to seek out stimulation and are generally impulsive, whereas introverts are quiet, reserved, and not impulsive (16). Neuroticism and psychoticism influence neurological and behavioral responses. Hence, the effects of exercise might be confounded with personality if children with ADHD were more extraverted, neurotic, or psychotic than the control children. The interday reliability for the JEPQ exceeded 0.90, so the mean of the two scores was used for analysis.

PSQI The PSQI has seven components: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, daytime dysfunction, and a global sleep score. The seven components reflect sleep habits during the past month. Circadian/ sleep patterns could possibly affect the outcome of the ASER. To account for potential circadian influences, participants were tested at the same time of day for each of the three conditions 1.5 h. The global sleep score was used for analysis.
EXERCISE AND ADHD

205

startle stimulus EMG. The EMG signals were calibrated with a series of input voltages (5 mV to 200 mV) generated from an external calibration signal generator for each gain setting (10 mV to 5 mV) employed in the study. Regression equations were derived from the calibration signals and used to convert EMG signals to voltages. Heart rate, systolic blood pressure, diastolic blood pressure, and mean arterial blood pressure were recorded once each minute during measurement of spontaneous eye blinks and the ASER by using a Finapres model 2300 blood pressure monitor (Ohmeda, Englewood, CO). Metabolic and Cardiorespiratory Measures During the exercise test, metabolic measures were recorded using a computer automated open-spirometry system. The volume of air inspired was measured using a mechanical flow meter REF (model 9200, Rayfield Equipment, Inc., Waitsfield, VT). Oxygen and carbon dioxide concentrations were measured using Ametek electronic gas analyzers calibrated using the micro-Scholander technique. The gas analyzers and flow meters were interfaced with an IBM personal computer (using VISTA Corporation data acquisition software, Rayfield Equipment Inc.), which com puted a 1-min running average of VO2, VE, VE/VO2, and RER at 15-s intervals. Heart rate was sampled during the last 15 s of each minute by using a Polar recorder (Polar Oy, Kempele, Finland). Treadmill Protocol VO2peak exercise test. The participant walked on the treadmill at 3.5 mph for 3 min at a 10% grade as a warm-up. After a short break, the participant walked at 7.5% grade for 3 min. The grade was then increased to 10% for the ensuing 3 min while the speed remained at 3.5 mph. The grade of the treadmill was increased 2.5% every 3 min thereafter while remaining at 3.5 mph until the participant was not able to continue. The time to reach voluntary maximum ranged from 5 to 25 min. Heart rate was monitored and recorded every minute of the VO2peak bout. Ninety percent of the children, balanced among groups, attained the criteria for VO2peak were an RER 1.0 and heart rate within 10% of age-predicted maximum or 190 beatsmin 1 (41). After
TABLE 1. Characteristics of children with ADHD and children without ADHD (controls). ADHD N Age (yr) Height (cm) Weight (kg) Percent fat VO2peak (mL kg min 1) Heart rate (beats min 1)a Diastolic BP (mm Hg)a Systolic BP (mm Hg)a MAP (mm Hg)a Physical activity (min day 1)b Physical activity (MET day 1)c
a b c

oral instructions to anchor the scale end points, Borgs category scale (6 20) was used to obtain ratings of perceived exertion during the final 15 s of each minute of exercise. Participants communicated RPE to the researcher by using hand signals. Submaximal exercise test (~6575% VO2peak). The aforementioned measurement procedures were also used during the submaximal treadmill exercise bout. Partic ipants walked at ~6575% of their VO2peak at 3.5 mph for the same duration as their maximal exercise test (i.e., 525 min). Participants began walking at the grade in which they reached ~60% VO2peak (mLkg 1min 1) during their VO2peak tread mill test. Once 60% VO2peak (mLkg 1min 1) was reached, the treadmill grade was increased or decreased to maintain ~60% VO2peak for the duration of the exercise bout. Responses during the bout, reported as the means of the last minute of percent VO2peak (mLkg 1min 1), did not differ across sex or ADHD diagnosis status (F(56,65) 0.32; P 0.58). Design and Statistical Analysis A mixed model factorial design was employed with between-group factors for ADHD diagnosis and sex and within-subject factors for condition and time. The sample size provided a statistical power at alpha 0.05 of 0.80 for main effects and 0.60 for the condition time effect within each group (39). To increase power for tests of the group condition time effects, a trend analysis for linear and quadratic functions (ADHD vs matched controls by sex) with repeated measures on conditions (VO2peak, 6575% O2peak, and quiet rest) and time (pre and post conditions) V was used for analyses of responses of spontaneous blink rate, ASER amplitude and latency, motor impersistence, and state anxiety. The analysis was repeated using the variables of VO2peak, sleep quality, state anxiety, and psychoticism as covariates. The trend analysis also was repeated for ASER amplitude and latency after natural log transformations to adjust for any nonnormality of the data (38). Trend analysis was again completed for spontaneous eye blinks and ASER amplitude and latency. The intertrial and interday reliabilities of pretest responses, as well as the intratrial reliability of the MIB subscales, were estimated by the Cronbach coef-

Controls N 8 8 8 8 8 8 8 8 8 8 8 Girls Mean 10.24 145.70 45.90 23.28 31.85 76.89 63.40 104.69 80.17 19.55 140.18 SD 1.77 15.84 18.69 7.41 5.23 14.54 13.12 17.98 13.12 15.85 82.96 N 11 11 11 11 11 11 11 11 11 11 11 Boys Mean 10.27 143.00 38.41 24.59 40.61 87.58 74.54 112.18 89.06 26.49 294.55 SD 1.56 11.30 9.21 7.43 10.62 12.74 12.79 15.99 12.60 8.26 220.70 N 14 14 14 14 14 13 13 13 13 14 14 Girls Mean 9.71 141.39 36.19 25.58 43.32 80.02 65.26 107.74 88.59 24.18 159.59 SD 1.55 7.68 10.20 9.94 13.54 9.92 9.43 13.17 15.32 14.35 186.42

Boys Mean 9.8 146.73 38.31 17.86 41.10 80.02 67.76 111.51 88.59 26.23 254.93

SD 1.55 5.03 4.42 5.10 7.46 10.70 9.16 14.48 15.32 11.83 255.59

10 10 10 10 10 10 10 10 10 10 10

Mean of the first 3 min during the counting of spontaneous eye blinks before any treatment condition. Measured using the Daily Activity Record (4). Parental report of childs physical activity on the Medical History Questionnaire.

206

Official Journal of the American College of Sports Medicine

http://www.acsm-msse.org

TABLE 2. Personality,a sleep,b trait anxiety,c and video ratingsd in children with ADHD and children without ADHD (controls). ADHD N Extraversiona Psychoticisma Neuroticisma Conformitya Sleepb Trait anxietyc Video interestd Video pleasantnessd
a b

Controls N 8 8 8 8 8 8 8 8 Girls Mean 16.38 2.94 9.31 10.87 6.00 29.27 5.21 5.17 SD 20.4 1.02 3.93 5.32 4.24 7.89 2.00 0.94 N 11 11 11 11 11 11 11 11 Boys Mean 17.36 3.59 8.14 9.05 2.27 30.42 5.42 5.09 SD 3.16 3.01 5.45 2.77 1.62 6.77 0.82 0.37 N 14 14 14 14 14 14 14 14 Girls Mean 17.57 2.68 9.75 11.89 2.00 31.89 5.14 5.00 SD 4.09 2.32 5.58 6.72 2.51 8.72 0.34 0.54

Boys Mean 18.85 5.43 10.05 9.55 3.60 32.92 4.30 4.17

SD 4.04 3.48 3.22 4.68 3.37 7.58 1.67 1.59

10 10 10 10 10 10 10 10

Measured using the Junior Eysenck Personality Questionnaire (16). Measured using the Pittsburgh Sleep Quality Index (8). c Measured using the State-Trait Anxiety Inventory for Children (48). d Measured using a 10-point rating of video interest or pleasantness.

ficient . The mean of 10 trials within each session was used for ASER amplitude and latency. Means SD are used in the text to describe participant characteristics and responses to the experimental conditions. Diagnosis sex ANOVA was used to test group differences in participants characteristics. Parental reports in the medical history questionnaire of their childs physical activity were expressed as METsd 1 by using standard metabolic equivalents. Self-report by the child on the Daily Activity Record (5) was expressed as mind 1. Resting heart rate and systolic, diastolic, and mean arterial blood pressures were determined by calculating the mean of values assessed during measurement of spontaneous eye blinks before each condition. Diagnosis sex condition time ANOVA with repeated measures on condition and time were conducted to determine any group differences at rest and during exercise.

RESULTS
Participant characteristics. The children diagnosed with or without ADHD were equated on attributes of age, height, weight, percent body fat, resting heart rate, resting systolic diastolic, and mean arterial blood pressure, VO2peak, physical activity, trait anxiety, and personality measures of extraversion, neuroticism, psychoticism, and the conformity scales (P 0.05). The ADHD group had worse global sleep scores (PSQI) compared with the control group (F(56,65) 8.49; P 0.006) (see Tables 1 and 2).

There were sex differences among the ADHD children on two covariates: boys had a higher VVO2peak (41.10 7.46 mLkg 1min 1) than girls (31.85 5.23 mLkg 1min 1) (F(56,65) 5.93; P 0.02) and a higher psychoticism score (5.43 3.48) than girls (2.94 1.02) (F(56,65) 4.27; P 0.046). Responses at rest and during exercise. No significant group time linear or quadratic trends across conditions occurred for heart rate, systolic blood pressure, diastolic blood pressure, or mean arterial blood pressure at rest. No effects were found for the physiological responses during exercise (see Tables 3 and 4). Spontaneous eye blinks. A diagnosis sex time linear trend across conditions occurred for spontaneous eye blinks (F(56,36) 5.06; P 0.03). Figure 1 illustrates that boys with ADHD had a faster blink rate per minute after maximal exercise (23.1 14.4 vs 13.4 6.9), with little change after quiet rest or submaximal exercise. The girls with ADHD and control participants had little change in spontaneous eye blinks across time. Pretest latencies were consistent across conditions. No other interactions were obtained, and results were unchanged by covariate analysis. Acoustic startle eye blink response amplitude (ASERAMP). A diagnosis sex time linear trend across conditions effect was obtained for ASERAMP (F(56,21) 3.70; P 0.06). Figure 2 illustrates that the effect was explained by an increase in ASERAMP after the submaximal exercise condition in girls with ADHD (0.63 0.46 mV vs 0.43 0.12 mV). Also, control boys had an increase, but control girls had a decrease, in ASERAMP after quiet rest. No

TABLE 3. Responses to maximal exercise (VO2peak) in children with ADHD and children without ADHD (controls). ADHD N VO2peak (mL kg min 1) Peak RER Peak VE/VO2 (L min 1) Peak HR (beats min 1) Peak RPEb Recovery HR (beats min 1)a Recovery MAP (mm Hg)a Recovery SBP (mm Hg)a Recovery DBP (mm Hg)a Boys Mean SD N Girls Mean SD N Boys Mean SD Controls N 14 14 14 14 14 14 14 14 14 Girls Mean 43.32 1.04 37.03 201.6 19.00 101.3 79.31 108.4 65.36 SD 10.20 0.07 4.53 7.57 1.80 7.00 14.82 25.13 13.69

10 41.10 7.46 8 31.85 5.23 11 40.61 10.62 10 1.04 0.06 8 1.04 0.07 11 1.06 0.06 10 34.89 4.00 8 34.9 4.00 11 34.98 4.45 10 192.7 13.14 8 193.6 9.75 11 195.7 8.45 10 18.90 1.97 8 18.00 2.73 11 19.09 1.64 10 97.03 13.34 8 92.33 11.55 11 85.45 9.75 10 85.90 15.73 8 76.75 15.53 11 89.00 16.49 10 108.4 14.11 8 101.5 13.65 11 114.8 19.19 10 70.43 15.89 8 60.75 12.99 11 74.55 15.99 a Measurement obtained 6 9 min post VO2peak exercise during the 3-min period when spontaneous eye blinks were counted. The mean for the 3 b Ratings of Perceived Exertion (RPE) according to Borgs 6 20 category rating scale.

min is reported.

EXERCISE AND ADHD

Medicine & Science in Sports & Exercise

207

 TABLE 4. Responses to submaximal (6575% VO2peak)a treadmill walking in children with ADHD and children without ADHD (controls). ADHD N VO2 (mL kg min 1) %VO2peak RER VEVO2a (L min 1) RPEb HR (beats min 1) Recovery HR (beats min 1)a Recovery MAP (mm Hg)a Recovery SBP (mm Hg)a Recovery DBP (mm Hg)a
a b

Controls N 8 8 8 8 8 8 8 8 8 8 Girls Mean 20.24 65.27 0.97 32.18 12.11 153.0 89.58 74.25 101.5 60.04 SD 3.51 7.01 0.03 2.49 5.71 10.24 13.18 18.69 13.65 13.54 N 11 11 11 11 11 11 11 11 11 11 Boys Mean 27.67 62.75 0.95 32.22 14.73 156.9 76.12 80.24 102.2 67.61 SD 6.73 11.77 0.06 3.6 3.41 13.85 10.13 18.50 21.52 16.80 N 14 14 14 14 14 14 14 14 14 14 Girls Mean 24.65 73.91 0.92 30.83 12.71 148.9 90.24 78.81 106.4 65.24 SD 4.43 11.21 0.06 3.58 3.43 16.45 7.18 15.61 23.02 14.87

Boys Mean 31.32 74.07 0.90 32.36 12.6 166.2 90.41 77.48 101.2 63.26

SD 3.97 10.79 0.07 5.08 4.16 25.88 9.68 16.16 15.89 13.47

9 9 9 9 9 9 9 9 9 9

Measured 6 9 min post submaximal exercise during the 3 min counting of spontaneous eye blinks. The mean for the 3 min is reported. Ratings of Perceived Exertion according to Borgs 6 20 category rating scale.

other interactions were obtained and groups otherwise did not change. Other than the low value for the ADHD girls before the submaximal exercise bout, pretest amplitudes

were similar across groups and conditions. Results were unchanged by covariate analysis. Also, diagnosis sex time linear trend analysis after log transformation yielded similar results (F(56,21) 3.12; P 0.08).

FIGURE 1Spontaneous eye blink rate before and after conditions of quiet rest, submaximal exercise (6575% VO2peak), or maximal exer cise (VO2peak) among boys and girls with attention-deficit hyperactivity disorder (ADHD) or without ADHD.

FIGURE 2Acoustic startle eye blink response amplitude (ASERAMP) before and after conditions of quiet rest, submaximal exercise (6575% VO2peak), or maximal exercise (VO2peak) among boys and girls with attention-deficit hyperactivity disorder (ADHD) or without ADHD. http://www.acsm-msse.org

208

Official Journal of the American College of Sports Medicine

FIGURE 3Acoustic startle eye blink response latency (ASERLAT) before and after conditions of quiet rest, submaximal exercise (6575% VO2peak), or maximal exercise (VO2peak) among boys and girls with attention-deficit hyperactivity disorder (ADHD) or without ADHD.

Acoustic startle eye blink response latency (ASERLAT). A diagnosis sex time quadratic trend across conditions was obtained for ASER latency (F(56,65) 5.03; P 0.03). Figure 3 illustrates that boys with ADHD had a decreased latency after the VO2peak condition (41.5 3.6 ms vs 46.8 5.5 ms) with little change after the resting or submaximal exercise conditions. The girls with ADHD had decreased ASER latency after the submaximal exercise (39.8 2.7 ms vs 43.7 3.6 ms) with little change after the resting or maximal VO2peak conditions. Control participants had no change in latency across conditions or time. Pretest latencies were consistent across conditions. No other interactions were obtained, and results were unchanged by covariate analysis. Also, diagnosis sex time by quadratic trend analysis after natural log transformation yielded similar results (F(56,65) 4.58; P 0.04). Motor impersistence and state anxiety. Effects for sex time linear trend across conditions (F(56,36) 7.09; P 0.01) and diagnosis sex time linear trend
EXERCISE AND ADHD

FIGURE 4 Motor Impersistence Battery (MIB) scores before and after conditions of quiet rest, submaximal exercise (6575% VO2peak), or maximal exercise (VO2peak) among boys and girls with attentiondeficit hyperactivity disorder (ADHD) or without ADHD. High scores indicate low impersistence.

across conditions (F(56,36) 3.37; P 0.07) were obtained for MIB. Figure 4 shows that boys with ADHD had improved motor impersistence scores after maximal exercise (992.7 83.7 s vs 790.5 145.8 s) but no change after rest or submaximal exercise. Control participants had no change in motor impersistence across conditions or time. Results were unchanged by covariate analysis, and there were no effects for state anxiety. Reliabilities. Intertrial reliability for spontaneous eye blinks was 0.89 0.98 for all participants, ranging from 0.66 to 0.98 for the ADHD group and from 0.82 to 0.98 for the control group. Intertrial reliability was 0.93 to 0.98 for ASERAMP and 0.78 to 0.96 for ASERLAT and similar between the ADHD and control groups. The intratrial reliability for the MIB was ~0.75 (range 0.63 0.79) for all trials, excepting a reliability of 0.55 for the first test administration after submaximal exercise.
Medicine & Science in Sports & Exercise

209

DISCUSSION
The main findings were increased spontaneous blink rate and decreased latency of the ASER in boys with ADHD after maximal exercise and decreased latency and increased amplitude of ASER after submaximal exercise in girls with ADHD. Neither spontaneous blink rate nor the ASER was affected by exercise among control children, consistent with previous findings in young adults asymptomatic for ADHD (37,51). Boys with ADHD had improved performance on motor impersistence after maximal exercise and performed worse than controls on the first day of testing, so they may present a unique ADHD symptomatology compared with girls (19,28,36). The sex by exercise intensity interaction effect observed for the ASER is not readily explainable, but it plausibly could be accounted for by greater striatal D2 receptor density in boys (3). The size of the significant effects (i.e., posttest pretest/pooled SD) was statistically large, exceeding 1 SD for blink rate and ASER latencies and 1.75 SD for motor impersistence. However, the clinical importance of the results for the management of behavioral features of ADHD is unknown, and the methods we used do not permit the conclusion that exercise had a dopaminergic effect. The decreased ASER latency in the children with ADHD might be partly explained by their longer latencies before each condition compared with the control children, permitting more room for change. However, latencies among children with ADHD did not change after quiet rest and responded differently for boys and girls, depending upon exercise intensity. Though elevated baseline values before the maximal exercise test for spontaneous blink rate, ASER latency, and motor impersistence among boys with ADHD might be explained by a testing effect, a similar pattern was not observed for the girls with ADHD. It is unclear why there was not a concurrent change in ASERAMP as large as was observed for ASERLAT. Nonetheless, girls with ADHD had a moderately large increase in ASERAMP of 0.70 standard deviation after submaximal exercise. The different latencies observed between the ADHD and control groups before each condition are explainable by differences in neurophysiology that might affect excitatory postsynaptic potentials which could shorten neural latency. Pharmacologic and neuroimaging studies have indicated that dysfunctional brain dopaminergic pathways contribute to behavioral and neuropsychological features of ADHD in children (24,30). These pathways may also be involved in the regulation of spontaneous eye blinks (9). The spontaneous blink rate is modulated by dopaminergic neural activity in the caudate nucleus (49), where morphological variability REFERENCES
1. AMERICAN PSYCHIATRIC ASSOCIATION. Diagnostic and Statistical Manual of Mental Disorders, 3rd Ed. Rev. Washington, DC: American Psychiatric Association, 1987, pp. 50 53. 2. AMERICAN PSYCHIATRIC ASSOCIATION. Diagnostic, and Statistical Manual of Mental Disorders, 4th Ed., Text Revision. Washington, DC: American Psychiatric Association, 2000, pp. 1943.

in the brains of children with ADHD has been reported (10,17,23). The findings are not explained by level of cardiorespira tory fitness. The groups were matched on VO2peak and activity history, and the submaximal exercise bout was standardized relative to each participants VO2peak. Nonetheless, an influence of absolute exercise intensity on the different responses between the ADHD boys and girls can not be discounted, as the boys had a higher VO2peak and exercised at a higher absolute submaximal intensity. The fitness levels of the children were in the low normal range (32), and the sum of skinfolds was higher than average among the control boys. Heart rate and blood pressure measurements taken at rest, during, and after maximal and submaximal exercise were consistent with prior reports on children of this age with (4) or without (7,40) ADHD. As expected (6), the children with ADHD in this study reported more sleep problems than the controls, but covariate analyses revealed that sleep problems did not influence the results. The DSM III-R does not differentiate between inattention and hyperactivity-impulsivity, which constrains effective treatment because the two types appear to differ in terms of sex ratios, comorbidity, and pharmacologic intervention (6,28,36). To reduce the heterogeneity associated with a DSM III-R diagnosis, we delimited the study to children who were taking methylphenidate and judged by their parents and physicians to be favorable responders (13). To our knowledge, this is the first controlled, experimental study of the effects of acute exercise on behavioral and dopaminergic-like responses in children diagnosed as having ADHD. Future studies like this one might include: 1) neuroimaging of striatal dopamine release after exercise, 2) a dopamine antagonist to directly examine the effect of exercise on dopaminergic mechanisms, 3) neuropsychological assessments to further clarify diagnosis, and 4) maintenance of methylphenidate dosage during testing. Though the preliminary findings we report suggest that exercise has potential efficacy as a dopaminergic adjuvant for treating ADHD behavior, the lack of uniform changes in eye blinks and motor impersistence after exercise makes such a conclusion premature. Nonetheless, the findings are sufficiently positive to encourage additional study aimed at determining their clinical meaning and underlying neuropharmacological mechanisms.
Address for correspondence: Rod K. Dishman, Ph.D., Department of Exercise Science, Ramsey Student Center, 300 River Road, Athens, GA 30602-6554; E-mail: rdishman@coe.uga.edu.

3. ANDERSEN S. L., and M. H. TEICHER. Sex differences in dopamine receptors and their relevance to ADHD. Neurosci. Biobehav. Rev. 24:137141, 2000. 4. BALLARD, J. E., R. A. BOILEAU, E. K. SLEATOR, B. H. MASSEY, and R. L. SPRAGUE. Cardiovascular responses of hyperactive children to methylphenidate. JAMA 236:2870 2874, 1976.

210

Official Journal of the American College of Sports Medicine

http://www.acsm-msse.org

5. BARANOWSKI, T., R. J. DWORKIN, C. J. CIESLIK, et al. Reliability and validity of self-report of aerobic activity: family health project. Res. Q. Exerc. Sport 17:309 317, 1984. 6. BARKLEY, R. A. Attention Deficit Hyperactivity Disorder: A Handbook for Diagnosis and Treatment, 2nd Ed. New York: Guilford Press, 1998, pp. 1 628. 7. BOILEAU, R. A., A. BONEN, V. H. HEYWARD, and B. H. MASSEY. Maximal aerobic capacity on the treadmill and bicycle ergometer of boys 1114 years of age. J. Sports Med. 17:153162, 1977. 8. BUYSSE, D. J., C. F. REYNOLDS, T. H. MONK, S. R. BERMAN, and D. J. KUPFER. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 28: 193213, 1988. 9. CAPLAN, R., D. GUTHRIE, and S. KOMO. Blink rate in children with attention-deficit hyperactivity disorder. Biol. Psychiatry 39:1032 1038, 1996. 10. CASTELLANOS, F. X., J. N. GIEDD, W. L. MARSH, et al. Quantitative brain magnetic resonance imaging in attention-deficit hyperactivity disorder. Arch. Gen. Psychiatry 53:607 616, 1996. 11. DAVIS, M. The neurophysiological basis of acoustic startle modulation: research on fear motivation and sensory gating. In: Attention and Orienting: Sensory and Motivational Processes, P. J. Lang, R. F. Simons, and M. Balaban (Eds.). Mahwah, NJ: Lawrence Erlbaum, 1997, pp. 69 96. 12. DAVIS, M., and G. K. AGHANJANIAN. Effects of apomorphine and haloperidol on the acoustic response. Psychopharmacology 47: 217223, 1976. 13. DENNY, C. B., and M. D. RAPPORT. Predicting methylphenidate response in children with ADHD: theoretical, empirical, and conceptual models. J. Am. Acad. Child Adolesc. Psychiatry 38:393 401, 1999. 14. ELSWORTH, J. D., M. S. LAWRENCE, R. H. ROTH, et al. D1 and D2 dopamine-receptors independently regulate spontaneous blink rate in the vervet monkey. J. Pharmacol. Exp. Ther. 259:595 606, 1991. 15. ETSCHEIDT, M. A., and T. AYLLON. Contingent exercise to decrease hyperactivity. J. Child Adolesc. Psychother. 4:192198, 1987. 16. EYSENCK, H. J., and S. B. G. EYSENCK. Manual for the Eysenck Personality Questionnaire (Junior & Adult). San Diego: Educational and Industrial Testing Service, 1975, pp. 119. 17. FILIPEK, P. A., M. SEMRUD-CLIKEMAN, R. J. STEINGARD, P. F. RENSHAW, D. N. KENNEDY, and J. BIEDERMAN. Volumetric MRI analysis comparing subjects having attention-deficit hyperactivity disorder with normal controls. Neurology 48:589 601, 1997. 18. FREED, C. R., and B. K. YAMAMOTO. Regional brain dopamine metabolism: a marker for the speed, direction, and posture of moving animals. Science 229:62 65, 1985. 19. GAUB, M., and C. L. CARLSON Gender differences in ADHD: a meta-analysis and critical review. J. Am. Acad. Child Adolesc. Psychiatry 36:1036 1045, 1997. 20. GOLDSTEIN, D. J., and T. D. BLUMENTHAL. Startle eyeblink elicitation in attention-deficit disordered children using low-intensity acoustic stimuli. Percept. Motor Skills 80:227231, 1995. 21. HATTORI, S., M. NAOI, and H. NISHINO. Striatal dopamine turnover during treadmill running in the rat: relation to the speed of running. Brain Res. Bull. 35:41 49, 1994. 22. HEYES, M. P., E. S. GARNETT, and G. COATES. Nigrostriatal dopaminergic activity is increased during exhaustive exercise stress in rats. Life Sci. 42:15371542, 1988. 23. HYND, G. W., K. L. HERN, E. S. NOVEY, et al. Attention deficithyperactivity disorder and the asymmetry of the caudate nucleus. J. Child. Neurol 8:339 347, 1993. 24. HYND, G. W., M. SEMRUD-CLIKEMAN, A. R. LORYS, E. S. NOVEY, and D. ELIOPULOS. Brain morphology in developmental dyslexia and attention deficit disorder/hyperactivity. Arch. Neurol. 47:919 926, 1990. 25. KARSON, C. N. Spontaneous eyeblink rates and dopaminergic systems. Brain 106:643 653, 1983. 26. KIMKO, H. C., J. T. CROSS, and D. R. ABERNETHY. Pharmacokinetics and clinical effectiveness of methylphenidate. Clin. Pharmacokinet. 37:457 470, 1999.
EXERCISE AND ADHD

27. KLEIN, S. A., and J. L. DEFFENBACHER. Relaxation and exercise for hyperactive impulsive children. Percept. Motor Skills. 45:1159 1162, 1977. 28. LAHEY, B. B., C. L. CARLSON, and P. J. FRICK. Attention deficit disorder without hyperactivity: a review of research relevant to DSM-IV. In: DSM IV Sourcebook, Vol. 1, T. A. Widiger, A. J. Frances, W. Davis, and M. First (Eds.). Washington, DC: American Psychiatric Press, 1996, pp. 550 580. 29. LISTE, I., M. J. GUERRA, H. J. CARUNCHO, and J. L. LABANDEIRAGARCIA. Treadmill running induces striatal Fos expression via NMDA glutamate and dopamine receptors. Exp. Brain. Res. 115: 458 68, 1997. 30. LOU, H. C., L. HENRIKSEN, P. BRUHN, H. BORNER, and J. B. P. NEILSEN Striatal dysfunction in attention deficit and hyperkinetic disorder. Arch. Neurol. 46:48 52, 1989. 31. MACRAE, P. G., W. W. SPIRDUSO, G. D. CARTEE, R. P. FARRAR, and R. E. WILCOX. Endurance training effects on striatal D2 dopamine receptor binding and striatal dopamine metabolite levels. Neurosci. Lett. 79:138 144, 1987. 32. MCCORMACK, W. P., K. J. CURETON, A. T. BULLOCK, and P. G. WEYAND. Metabolic determinants of 1-mile run/walk performance in children. Med. Sci. Sports Exerc. 23:611 617, 1991. 33. MELONI, E. G., and M. DAVIS. Enhancement of the acoustic startle response by dopamine agonists after 6-hydroxydopamine lesions of the substantia nigra pars compacta: corresponding changes in c-Fos expression in the caudate-putamen. Brain Res. 879:93104, 2000. 34. MELONI, E, G., and M. DAVIS. Enhancement of the acoustic startle response in rats by the dopamine D1 receptor agonist SKF 82958. Psychopharmacology 144:373380, 1999. 35. MOLLOY, G. N. Chemicals, exercise and hyperactivity: a short report. Int. J. Disabil. Dev. Educ. 36:57 61, 1989. 36. MORGAN, A. E., G. W. HYND, C. A. RICCIO, and J. HALL. Validity of DSM-IV ADHD predominantly inattentive and combined types: relationship to previous DSM diagnoses/subtype differences. J. Am. Acad. Child Adolesc. Psychiatry 35:325333, 1996. 37. OCONNOR, P. J., and S. J. PETRUZZELLO. Spontaneous eyeblinks and state anxiety following exercise. Int. J. Neurosci. 62:57 63, 1992. 38. ORNITZ, E. M., R. T. RUSSELL, H. YUAN, and M. LIU. Autonomic electroencephalographic and myogenic activity accompanying startle and its habituation during mid-childhood. Psychophysiology 33:507513, 1996. 39. POTVIN, J., and R. W. SCHUTZ. Statistical power in repeated measures ANOVA. Behav Res. Methods Instrum.Comput. 32:347 356, 2000. 40. RIOPEL, D. A., A. B. TAYLOR, and R. A. HOHEN. Blood pressure, heart rate, pressure-rate product and electrocardiographic changes in healthy children during treadmill exercise. Am. J. Cardiol. 4:697704, 1979. 41. ROWLAND, T. W. Aerobic exercise testing protocols. In: Pediatric Laboratory Exercise Testing Clinical Guidelines, T. W. Rowland (Ed.). Champaign, IL: Human Kinetics Publishers, 1993, pp. 19 41. 42. SHADER, R. I., J. S. HARMATZ, J. R. OESTERHELD, D. X. PARMELEE, F. R. SALLEE, and D. J. GREENBLATT. Population pharmacokinetics of methylphenidate in children with attention-deficit hyperactivity disorder. J. Clin. Pharmacol. 39:775785, 1999. 43. SILVER, L. B. Attention-Deficit Hyperactivity Disorder: A Clinical Guide to Diagnosis and Treatment. Washington, DC: American Psychiatric Press, 1992, pp. 1164. 44. SILVERSTEIN, J. M., and D. B. ALLISON. The comparative efficacy of antecedent exercise and methylphenidate: a single-case randomized trial. Childcare Health Dev. 20:47 60, 1994. 45. SIMONS, R. F., and M. F. ZELSON. Engaging visual stimuli and reflex blink modification. Psychophysiology 22:44 49, 1985. 46. SLAUGHTER, M. H., T. G. LOHMAN, R. A. BOILEAU, et al. Skinfold equations for estimation of body fatness in children and youth. Hum. Biol. 60:709 723, 1988. 47. SPENCER, T., J. BIEDERMAN, and T. WILENS. Attention-deficit/hyperactivity disorder and comorbidity. Pediatr. Clin. North Am. 46:915927, 1999.
Medicine & Science in Sports & Exercise

211

48. SPIELBERGER, C. D., C. D. EDWARDS, J. MONTUORI, R. LUSHENE, and D. PLATZEK. STAIC Preliminary Manual for the State-Trait Anxiety Inventory for Children (Forms C-1 and C-2). Palo Alto CA: Consulting Psychologist Press, 1973, pp. 111. 49. TAYLOR, J. R., J. D. ELSWORTH, M. S. LAWRENCE, J. R. SLADEK, R. H. ROTH, and D. E. REDMOND Spontaneous blink rates correlate with dopamine levels in the caudate nucleus of MPTP-treated monkeys. Exp. Neurol. 158:214 220, 1999. 50. THURBER, S., and C. E. WALKER. Medication and hyperactivity: a meta-analysis. J. Gen. Psychol. 108:79 86, 1983. 51. TIEMAN, J. G., L. J. PEACOCK, K. J. CURETON, and R. K. DISHMAN. Acoustic startle eye blink response after acute exercise. Int. J. Neurosci. 106:2133, 2001.

52. VOELLER, K. S., and K. M. HEILMAN. Motor impersistence in children with attention deficit hyperactivity disorder: evidence for right hemisphere dysfunction. Ann. Neurol. 24:323, 1998. 53. VOLKOW, N. D., Y. S. DING, J. S. FOWLER, et al. Is methylphenidate like cocaine? Studies on their pharmacokinetics and distribution in the human brain. Arch. Gen. Psychiatry 52:456 463, 1995. 54. WANG, G. J., N. D. VOLKOW, J. S. FOWLER, et al. PET studies of the effects of aerobic exercise on human striatal dopamine release. J. Nucl. Med. 41:13521356, 2000. 55. YADID, G., D. H. OVERSTREET, and A. ZANGEN. Limbic dopaminergic adaptation to a stressful stimulus in a rat model of depression. Brain Res. 896:43 47, 2001.

212

Official Journal of the American College of Sports Medicine

http://www.acsm-msse.org