Comparison of Fidgeting in Adolescents with Attention-Deficit/Hyperactivity Disorder Between Before and After Stimulant Medication Intake

Abstract

Objective:

Fidgeting is a common symptom in patients with attention-deficit hyperactivity disorder (ADHD). The current study investigated ADHD stimulant medication effects on fidgeting in adolescents with ADHD during a short research study session using wrist-worn accelerometers.

Method:

Adolescents with ADHD who had been taking stimulant medications (ADHD group) and adolescents without ADHD (control group) participated in the study. Accelerometer data were obtained from both wrists of each participant to track their hand movements during two hearing testing sessions. All subjects in the ADHD group abstained from their stimulant medications for at least 24 hours before their first session (off-med session). The second session (on-med session) was conducted about 60–90 minutes after medication intake. The control group participated in two sessions in a similar time frame.

Results:

The current study focuses on relationships between hand movements and stimulant medication in adolescents with ADHD. Both conditions were compared to evaluate the relationship of hand movements and stimulant medication. We hypothesized the ADHD group will exhibit less hand movements during the on-medication session in comparison to off-medication session.

Conclusion:

Wrist-worn accelerometer measures obtained during nonphysical tasks in a short duration may not provide hand movement differences between on-med and off-med conditions in adolescents with ADHD. ClinicalTrials.gov Identifier: NCT04577417.

Introduction

Attention-deficit/hyperactivity disorder and fidgeting

Attention-deficit/hyperactivity disorder (ADHD) is the most common neurodevelopmental disorder in children (American Psychiatric Association, 2013; Cortese and Coghill, 2018). Fidgeting, the generation of small movements through restlessness or nervousness, commonly occurs in individuals with ADHD (Cohen et al, 2018). Patients with ADHD tend to fidget more than typically developed children and adults. Both children and adults with ADHD report fidgeting as their most hyperactive symptom (Millstein et al, 1997). Previous studies suggest that fidgeting helps children with hyperactivity to maintain their attention (Hartanto et al, 2016; Sarver et al, 2015). Fidgeting is considered beneficial for individuals with ADHD because their motor activity maintains their attention while facilitating neurocognitive functioning (Cohen et al, 2018; Sarver et al, 2015).

Stimulant medication on fidgeting

Accelerometers have become increasingly popular to objectively measure motor movements in various clinical populations (Dane et al, 2000). The devices serve as a sensor-based measurement to record fine-grained activity (Gawrilow et al, 2014). Accelerometers have been used with ADHD patients to measure motor activity. Previous studies indicate that accelerometers have successfully measured physical activity levels in fingers (Hotham et al, 2018), nondominant wrists (Kam et al, 2011), dominant arms (Munoz-Organero et al, 2018), legs (Munoz-Organero et al, 2018; Wood et al, 2009), waist (Wood et al, 2009), and ankles (Hartanto et al, 2016; Munoz-Organero et al, 2018)^. These studies suggest that activity levels are higher in ADHD compared with non-ADHD peers and indicate the use of accelerometers for diagnostic measures. On the contrary, the effects of stimulant medication in relation to bodily movements have not been investigated in majority of the ADHD accelerometer studies. Stimulant medications such as methylphenidate and amphetamine are common to treat ADHD symptoms in children (Advokat and Scheithauer, 2013).

Stimulants are confirmed to alleviate the symptoms of ADHD. Few studies have examined stimulant medication in conjunction with fidgeting using accelerometers. Munoz-Organero and colleagues (Munoz-Organero et al, 2018) used an accelerometer to examine wrist and ankle movements in children with ADHD throughout a typical school day. The study found an increase in fine and gross motor movements in children with ADHD compared with typically developed children. Some children with ADHD in the study took stimulant medication, which eased wrist and ankle movement in children with ADHD according to the authors. Despite the previous finding on accelerometers, accelerometers are not widely used as medication effectiveness measures for a short nonphysical test. The current study investigated ADHD stimulant medication effects on wrist movements in adolescents with ADHD using accelerometers during a relatively short hearing testing.

Methods

Participants

Two groups of adolescents, 13–19 years of age, were recruited from the tri-state area of Delaware, Pennsylvania, and Maryland. The ADHD group consisted of 18 adolescents with a documented ADHD diagnosis based on the fifth edition of Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2013). Participants in the ADHD group were under treatment of the same drug and dosage of ADHD stimulant medication for at least 3 months before the study visit. Their ADHD stimulant medications included either amphetamine-based or methylphenidate-based drugs prescribed by their primary care provider. Dosage of ADHD stimulant medication varied among the participants with ADHD. The control group consisted of 23 adolescents with typical development history. All the participants in the control group did not take any prescribed medications and did not have any neuropsychiatric or neurological diagnosis.

All participants passed otoscopic inspection of their ears, and exhibited normal hearing based on tympanometry, audiometry, and distortion products of otoacoustic emissions testing. Participant characteristics are summarized (Table 1). All participants received monetary compensation for their participation.

Table 1.

Characteristics of Participants

	ADHD group (n = 18)	Control group (n = 23)
Age	M = 15.1	M = 14.9
Sex (M:F)	13:5	11:12
ADHD presentations
Predominantly inattentive	8	NA
Predominantly Hyperactive/impulsive	2	NA
Combined	8	NA
Stimulants subtypes
Amphetamine-based drug^a	6	NA
Methylphenidate-based drug^b	12	NA
Hand dominance
Left	2	9
Right	16	14

Amphetamine-based stimulants include Adderall and Vyvanse.

Methylphenidate-based stimulants include Concerta, Dexmethylphenidate, Focalin, Metadate CD.

ADHD, attention-deficit/hyperactivity disorder; NA, not applicable.

Procedure

All procedures were approved by the Nemours Institutional Review Board (local no.: 1600804). Test procedures were thoroughly explained to participants and their parents during the consent process. Participants and their parents provided their consent to participate in the study by signing the online consent documents before their visit for testing. Participants or their parents completed an online questionnaire pertaining to demographic information and medical history, including their medications and dominant hand. ADHD participants were asked to come to the laboratory in the morning without taking their stimulant medication on their visit day. The first session was the off-medication condition and the second session was the on-medication condition for all participants (Fig. 1). During each session, participants completed a series of auditory and speech perception tasks. The ADHD participants were asked to take stimulant medication after completing the off-medication session. Depending on the participant's stimulant medication, a different wait time (ranging 60–90 minutes) was selected before starting their on-medication session.

FIG. 1.

Data collection process.

Additional tests, such as pure-tone audiometry, NIH Toolbox Cognition Battery (www.NIHToolbox.org), and Conners 3 Self-Report Short form (Conners, 2009), were conducted during participant wait times between sessions. The participants in both groups completed two sessions within the same day. During their visit, subjects wore a wrist-worn GENEActiv accelerometer on both wrists.

Materials

Wrist-worn accelerometers (GENEActiv Original; Activinsights, Ltd, UK) were used to collect hand movement data. The accelerometers were set to record hand movement with the sampling frequency of 100 Hz on a triaxial platform (X, Y, Z). Each participant wore two wrist-worn devices on dominant and nondominant hand for the duration of their visit (3–4 hours). We extracted the accelerometer data during the off-medication and on-medication sessions to compare the differences in hand movement data between the two sessions. Each session took about 40 minutes to complete. Nondominant hand data have missing data files of two subjects (one control and one ADHD) and the dominant hand data have missing files from one control subject due to procedural errors. GENEActiv PC Software (version 3.3) was used to extract the accelerometer data from the start and end times for each session and saved in raw format as binary files. Then each binary file was converted to csv files using the 10-second time window (10-second epoch). The devices track the three axials of movement, body temperature, ambient light, and recording button status. In this study, we analyzed the triaxial data of dominant and nondominant hand movements.

We employed Conners 3 Self-Report Short form to estimate severity of two ADHD core symptoms (inattention and hyperactivity/impulsivity). The inattention score and hyperactivity/impulsivity raw scores were converted to the T-score (ranging 40–90) by the appropriate age and sex norm values. We also employed the NIH Toolbox Flanker Inhibitory Control and Attention (Flanker) Test to measure a participant's visual attention and inhibitory control. The Flanker test was conducted by an iPad. The age-corrected standardized scores were computed automatically at the end of each test based on a participant's accuracy and response time.

Data analysis

We computed mean of standard deviation data and mean of absolute mean values for each participant and each session using the 10-second epoch data to see differences between sessions and between groups. Two-way repeated measures of analysis of variance (ANOVA) was conducted for each axial measure, using group as the between-subjects factor and session as the within-subjects factor. The Pearson partial correlation coefficient was calculated between each hand movement measure and ADHD measures (Conners 3 scaled inattention scores and scaled hyperactivity/impulsivity scores, and NIH Toolbox Flanker Inhibitory Control and Attention Test age-corrected scores) using age as a control variable. The statistical tests were performed using the IBM SPSS Statistics version 27. A p < 0.05 was considered statistically significant in this study.

Results

We assumed that ADHD symptoms of impulsivity or restlessness will be reduced by the stimulant medications, resulting in less fidgeting within each wrist. We expected less hand movements in the ADHD group during the on-medication session compared with the off-medication session.

For the control group, we do not expect to see a change in hand movements between both sessions. Unintentional hand movements are expected to remain relatively the same for both sessions in both wrists.

Nondominant hand movement

The mean of absolute mean for each axial and group was analyzed for nondominant hand movement (Fig. 2). The results of Mauchly's test of sphericity indicated violation of the sphericity assumption. The Greenhouse and Geisser correction was employed.

FIG. 2.

Mean of absolute mean values (in g) of each axial data from nondominant hand in the session 1 (solid bars) and the session 2 (striped bars) in the control and ADHD groups (blue and red bars, respectively). For the ADHD group, session 1 and session 2 correspond to the off-medication and on-medication sessions, respectively. ADHD, attention-deficit/hyperactivity disorder.

The two-way repeated measures of ANOVA results for absolute mean measures and standard deviation measures of the nondominant hand data indicated that both main group effect and the interaction effect were not significant for all axial mean data (Table 2). The mean of absolute mean values was significantly larger in the on-med session than the off-med session for y- axial data although the difference was very small (M = 0.01 g for both groups). The z-axial mean value was larger in the session 1 (off-medication session for the ADHD) than in the session 2 (on-medication session for the ADHD group). No significant difference was found between sessions for the x-axial data, which involves side-to-side movements. The results indicated moderate-to-large effect sizes on the session comparisons. Because each subject was sitting in a chair during the sessions, the results suggest similar size movement in lateral rotation (x) of hand but smaller size movement to raise their nondominant hand in the session 2 (on-medication).

Table 2.

Results of the Two-Way Analysis of Variance on Nondominant Hand Data

Measures	Axis	Factor	F	p	η²
Absolute mean	X	Session	3.891	0.056	0.095
		Group	0.046	0.832	0.001
		Group × session	0.010	0.922	<0.001
	Y	Session	4.410	0.043	0.106
		Group	<0.001	0.997	<0.001
		Group × session	0.086	0.771	0.002
	Z	Session	11.978	0.001	0.245
		Group	0.024	0.878	0.001
		Group × session	0.018	0.893	<0.001
Standard deviation	X	Session	22.235	<0.001	0.375
		Group	0.288	0.594	0.008
		Group × session	0.448	0.508	0.012
	Y	Session	14.928	<0.001	0.287
		Group	0.090	0.766	0.002
		Group × session	0.143	0.707	0.004
	Z	Session	21.587	<0.001	0.368
		Group	0.001	0.972	<0.001
		Group × session	1.589	0.215	0.041

Greenhouse–Geisser correction was used due to violation of the sphericity assumption.

Similar to the results of the absolute mean data, the two-way repeated measures of ANOVA on standard deviation measures indicated that the main session effect was significant for all triaxial data, but the main group effect and the interaction effect were not significant for all triaxial data (Table 2). The mean of standard deviation of hand movement (g) was significantly higher for the session 2 in both groups for all axial data (F = 0.33, 0.04, and 0.70 for X, Y, and Z, respectively, ps < 0.05), suggesting less stable movement in the session 2 (on-medication session for the ADHD) than the session 1 (off-medication session for the ADHD) in both groups (Fig. 3).

FIG. 3.

Mean of standard deviation of mean values (in g) of each axial data from nondominant hand in the session 1 (solid bars) and the session 2 (striped bars) in the control and ADHD groups (blue and red bars, respectively). For the ADHD group, session 1 and session 2 correspond to the off-medication and on-medication sessions, respectively. ADHD, attention-deficit/hyperactivity disorder.

Dominant hand movement

Two-way repeated measures of ANOVA on dominant hand measures indicated very similar results as we found for nondominant hand measures (Table 3). The mean of absolute mean was significantly larger in session 1 than in the session 2 for the x-axial data (Fig. 4). Although the mean of absolute mean was larger in session 1 than in the session 2 for the y-axial data, there was no significant difference between the sessions for the dominant hand data. As for the z-axial data, similar to the nondominant hand results, we found a significantly smaller absolute mean in the session 2 than in the session 1. The standard deviations in the session 2 were larger than the standard deviations in the session 1 in all three-axial data (Fig. 5).

FIG. 4.

Mean of absolute mean values (in g) of each axial data from dominant hand in the session 1 (solid bars) and the session 2 (striped bars) in the control and ADHD groups (blue and red bars, respectively). For the ADHD group, session 1 and session 2 correspond to the off-medication and on-medication sessions, respectively. ADHD, attention-deficit/hyperactivity disorder.

FIG. 5.

Mean of standard deviation of mean values (in g) of each axial data from dominant hand in the session 1 (solid bars) and the session 2 (striped bars) in the control and ADHD groups (blue and red bars, respectively). For the ADHD group, session 1 and session 2 correspond to the off-medication and on-medication sessions, respectively. ADHD, attention-deficit/hyperactivity disorder.

Table 3.

Results of the Two-Way Analysis of Variance on Dominant Hand Data

Measures	Axis	Factor	F	p	η²
Absolute mean	X	Session	1.700	0.020	0.043
		Group	0.018	0.895	<0.001
		Group × session	0.087	0.770	0.002
	Y	Session	1.356	0.252	0.034
		Group	0.195	0.662	0.005
		Group × session	2.031	0.162	0.051
	Z	Session	5.572	0.023	0.128
		Group	0.050	0.824	0.001
		Group × session	0.736	0.396	0.019
Standard deviation	X	Session	24.548	<0.001	0.392
		Group	0.140	0.710	0.004
		Group × session	0.496	0.485	0.013
	Y	Session	18.356	<0.001	0.326
		Group	0.026	0.873	0.001
		Group × session	0.210	0.649	0.006
	Z	Session	21.367	<0.001	0.360
		Group	0.018	0.8942	<0.001
		Group × session	0.728	0.399	0.019

Greenhouse–Geisser correction was used due to violation of the sphericity assumption.

Relationships between hand movement data and ADHD measures

No significant correlation was found between each hand movement measure and Conners 3 inattention scaled scores, between each hand movement measure and Conners 3 hyperactivity and impulsivity scaled scores, and between each hand movement measure and age-corrected Flanker test scores for both dominant hand (rs = −0.14 to 0.08, ps = 0.25–0.94) and nondominant hand data (rs = −0.11 to 0.12, ps = 0.48–0.89).

Discussion

In this study, we investigated wrist movements before and after stimulant medication intake. Wrist movements in adolescents with ADHD were compared with age-matched typically developed peers. We expected that the control group would exhibit about the same amount of hand movement in both sessions. We also expected that the ADHD group would exhibit more hand movements after stimulant medication intake than before medication intake. Contrary to our predictions, we did not find any significant group difference in any of our hand movement measures. Results of standard deviation measures show more unstable hand movement in both ADHD and control group during the session 2 (on-medication session for the ADHD) compared with the session 1 (off-medication session for the ADHD) for both dominant and nondominant hands. Larger standard deviations of movement measures could have been exerted due to possible boredom during the session 2. The on-med session was always the second session in this study. The study was designed to obtain data in one visit to avoid missing data for the second visit. Due to this design constraint, we did not counterbalance the order of on-med and off-med conditions.

It is possible that novelty of tasks in the first off-med session might fade during the second on-med session, which could explain that both ADHD and control groups exhibited the same finding of the standard deviation results.

As for the absolute size of movement difference between the two sessions, we found large movement size in x- and z-axial data for both dominant and nondominant hands, whereas there were generally small y-axial movement in both dominant and nondominant hand data although y-axial nondominant hand data were significantly different between the two sessions. Because our participants were in a sitting position during the data collection, their hand movements are mostly due to raising their hands during the task to respond to our auditory stimuli for hearing tests. It seems that overall different magnitudes of the movement measures in both hands were attributed to a task-related hand movement, and not reflecting fidgeting. It is also possible that we did not capture finger movements (e.g., finger tapping) by wrist worn sensors.

The small sample size of participants with varied ADHD stimulant drugs and dosages in this study may have affected our findings. We were not able to analyze our data by stimulant drugs due to our small sample size. Another related limitation of this study is that we did not have any baseline measures before individuals with ADHD had started their medication treatment because all ADHD participants in this study were recruited after they had started their medication treatment for ADHD and had been on their medication for 3 months before the study participation. In addition to issues related with different drugs and dosages, our participants with ADHD vary their ADHD presentations. Similar to the drug-related analyses, we were not able to conduct further analyses based on ADHD presentations because the numbers of each subgroup are not large enough. However, hyperactive symptoms measured by gross motor activity can be found across the three ADHD subtypes in previous studies (Dane et al, 2000).

Indeed, the correlation results in this study showed no significant relationships between self-reported ADHD symptom measures (inattention and hyperactivity/impulsivity scores) and the movement measures as well as between the Flanker test scores and the movement measures. However, it is possible to find a clear difference in movement measures between the two conditions or two groups if the participants in the ADHD group shared the same type of ADHD presentation.

In addition, the 40-minute testing session may not have been sufficient to capture fidgeting on the devices. Participants were asked to sit still in a chair and focus on listening tasks. Although participants were old enough to retain task attentiveness, less movement could have been naturally exerted due to the manner of testing. The listening tasks during each session were relatively simple tasks and did not require a high cognitive demand, which may affect activity levels of patients with ADHD across subtypes (Kofler et al, 2016).

Conclusions

The relationship between ADHD stimulant medication and the common symptom of fidgeting has not been well studied. Learning more about the relationship between fidgeting and stimulant medication could provide objective measures of medication effects of hyperactivity symptoms in ADHD. However, careful consideration should be taken for selection of tasks during activity movement measurement. Tasks with low cognitive demands in a laboratory setting may not be suitable to exert fidgeting.

Clinical Significance

Within the ADHD population, accelerometers may be able to provide sufficient data on hand or leg movement to indicate stimulant medication effects. Consideration should be taken for cognitive demands of tasks.

Footnotes

Acknowledgments

The authors thank the participants and their family members for their contribution to this project. The authors express their gratitude to Josephine Elia, MD, Hal Byck, MD, Jessica Rayfield, and Thierry Morlet, PhD at Nemours for their recruitment and mentorship.

Authors' Contributions

S.S.: writing—original draft (lead); formal analysis (equal); and review and editing (equal). A.M.: writing—original draft (supporting); and writing—review and editing (equal). K.N.: conceptualization (lead); writing—original draft (supporting); methodology (lead); formal analysis (equal); and writing—review and editing (equal).

Disclosures

All the authors contributed to the original article and had no conflicts of interest with any other party. S.S., A.M., and K.N. declare that they have no conflicts of interest.

References

Advokat

, Scheithauer

. Attention-deficit hyperactivity disorder (ADHD) stimulant medications as cognitive enhancers. Front Neurosci, 2013; 7; doi: 10.3389/fnins.2013.00082

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 5th ed. American Psychiatric Publishing: Arlington, VA, USA; 2013.

Cohen

, Bravi

, Minciacchi

. The effect of fidget spinners on fine motor control. Sci Rep, 2018; 8(1):3144; doi: 10.1038/s41598-018-21529-0

Conners

CK.

Conners 3rd Edition Manual. Multi-Health Systems: Toronto, Canada; 2009.

Cortese

, Coghill

. Twenty years of research on attention-deficit/hyperactivity disorder (ADHD): Looking back, looking forward. Evid Based Ment Health, 2018; 21(4):173–176; doi: 10.1136/ebmental-2018-300050

Dane

, Schachar

, Tannock

. Does actigraphy differentiate ADHD subtypes in a clinical research setting?. J Am Acad Child Adolesc Psychiatry, 2000; 39(6):752–760; doi: 10.1097/00004583-200006000-00014

Gawrilow

, Kühnhausen

, Schmid

, et al. Hyperactivity and motoric activity in ADHD: Characterization, assessment, and intervention. Front Psychiatry, 2014; 5; doi: 10.3389/fpsyt.2014.00171

Hartanto

, Krafft

, Iosif

, et al. A trial-by-trial analysis reveals more intense physical activity is associated with better cognitive control performance in attention-deficit/hyperactivity disorder. Child Neuropsychol, 2016; 22(5):618–626; doi: 10.1080/09297049.2015.1044511

Hotham

, Haberfield

, Hillier

, et al. Upper limb function in children with attention-deficit/hyperactivity disorder (ADHD). J Neural Transm Vienna Austria 1996, 2018; 125(4):713–726; doi: 10.1007/s00702-017-1822-8

10.

Kam

, Lee

, Cho

S-M

, et al. High-resolution actigraphic analysis of ADHD: A wide range of movement variability observation in three school courses—A pilot study. Healthc Inform Res, 2011; 17(1):29–37; doi: 10.4258/hir.2011.17.1.29

11.

Kofler

, Raiker

, Sarver

, et al. Is hyperactivity ubiquitous in ADHD or dependent on environmental demands? Evidence from meta-analysis. Clin Psychol Rev, 2016; 46:12–24; doi: 10.1016/j.cpr.2016.04.004

12.

Millstein

, Wilens

, Biederman

, et al. Presenting ADHD symptoms and subtypes in clinically referred adults with ADHD. J Atten Disord, 1997; 2(3):159–166; doi: 10.1177/108705479700200302

13.

Munoz-Organero

, Powell

, Heller

, et al. Automatic extraction and detection of characteristic movement patterns in children with ADHD based on a convolutional neural network (CNN) and acceleration images. Sens Basel, 2018; 18(11):3924–3942; doi: 10.3390/s18113924

14.

Sarver

, Rapport

, Kofler

, et al. Hyperactivity in attention-deficit/hyperactivity disorder (ADHD): Impairing deficit or compensatory behavior?. J Abnorm Child Psychol, 2015; 43(7):1219–1232; doi: 10.1007/s10802-015-0011-1

15.

Wood

, Asherson

, Rijsdijk

, et al. Is overactivity a core feature in ADHD? Familial and receiver operating characteristic curve analysis of mechanically assessed activity level. J Am Acad Child Adolesc Psychiatry, 2009; 48(10):1023–1030; doi: 10.1097/CHI.0b013e3181b54612