Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder

Abstract

Objective: To characterize head movements in children with ADHD using an ex-Gaussian distribution and examine associations with out-of-scanner sustained attention. Method: Fifty-six children with ADHD and 61 controls aged 9 to 11 years completed the Sustained Attention to Response Task (SART) and resting-state functional magnetic resonance imaging (fMRI). In-scanner head motion was calculated using ex-Gaussian estimates for mu, sigma, and tau in delta variation signal and framewise displacement. Sustained attention was evaluated through omission errors and tau in response time on the SART. Results: Mediation analysis revealed that out-of-scanner attention lapses (omissions during the SART) mediated the relationship between ADHD diagnosis and in-scanner head motion (tau in delta variation signal), indirect effect: B = 1.29, 95% confidence interval (CI) = [0.07, 3.15], accounting for 29% of the association. Conclusion: Findings suggest a critical link between trait-level sustained attention and infrequent large head movements during scanning (tau in head motion) and highlight fundamental challenges in measuring the neural basis of sustained attention.

Keywords

ADHD resting-state fMRI ex-Gaussian distribution neuroimaging movement

Introduction

Head motion is generally considered a major confounder to be removed from magnetic resonance imaging (MRI) data (Power et al., 2012). If not appropriately managed, head motion can distort MRI analyses, leading to data loss and hence costs of wasted participant time, research time, and MRI resources. In-scanner motion may be particularly problematic for neuroimaging studies in psychiatric research, which generally seek to compare data between clinical and neurotypical groups. Greater head motion has been demonstrated across several conditions, including ADHD, autism spectrum disorder, and schizophrenia when compared with controls (Pardoe et al., 2016).

Head movement may impact MRI results at multiple stages. A typical first step in standard MRI analysis is to remove participants with extreme head motion. This means that participants in clinical groups will be more likely to be excluded from analysis, particularly those with more severe symptoms. Head motion is then typically regressed out in the remaining participants (Friston et al., 1996), meaning patients will experience more data loss when regressing head motion out of the brain signal, potentially leading to unreliable or spurious results (Andrews-Hanna et al., 2007; Power et al., 2012; Satterthwaite et al., 2012).

Particular concern is building around head motion distortions in imaging research of ADHD, a pervasive neurodevelopmental disorder with core symptoms of inattention and/or hyperactivity and impulsivity (Polanczyk et al., 2007). Pardoe et al. (2016) recently considered children from the ADHD-200 consortium (mean age = 11.8 years; Milham et al., 2012) and found that those diagnosed with ADHD (n = 332) had significantly greater head motion than controls (n = 549). Furthermore, Couvy-Duchesne et al. (2016) investigated pairs of monozygotic (N = 95) and dizygotic (N = 144) adult twins (mean age = 22 years) and found that head motion was moderately correlated with maternally reported inattentive and hyperactive-impulsive symptoms. This correlation was partly due to shared genetic factors between head motion and ADHD symptoms. An investigation by Kong et al. (2014) further found that differences in scores on an impulsivity scale partially accounted for differences in head motion during resting-state functional MRI (fMRI) and diffusion tensor imaging (DTI) in both children (n = 245, mean age = 11.7 years) and adults (n = 581, mean age = 20.5 years), and accounted for differences in head motion between children with ADHD (n = 102, mean age = 12.1 years) and without ADHD (n = 143, mean age = 11.4 years) during resting-state fMRI. These studies not only provide initial evidence that head motion may be a common feature of ADHD, but also suggest a potential link between head motion and core cognitive deficits in ADHD (e.g., impulsivity), although it is worth noting studies such as Engelhardt et al. (2017) and Costa Dias et al. (2013) that did not observe this relationship between head motion and ADHD symptomology.

Whereas Kong et al. (2014) considered impulsivity, many other cognitive deficits are implicated in ADHD. Sustained attention is a key functional deficit in ADHD, frequently measured using go/no-go tasks, such as the Sustained Attention to Response Task (SART; Robertson et al., 1997) and Continuous Performance Test (CPT; Riccio et al., 2002). Studies examining performance on sustained attention tasks in ADHD have generally found evidence of worse performance on tasks such as the SART in ADHD samples. This was indicated by a larger standard deviation of response time (SD of RT), a measure of response time variability (RTV; Adamo, Di Martino, et al., 2014; Braet et al., 2011; Johnson, Barry, et al., 2008, although see Johnson, Kelly, et al., 2007) and more omission errors (failure to respond to a target; Bellgrove et al., 2005, 2006; Johnson, Robertson, et al., 2007; O’Connell et al., 2004, 2006; although see Adamo, Huo, et al., 2014; Johnson, Kelly, et al., 2007).

The specific deconstruction of response times (RTs) using an ex-Gaussian distribution has been proposed as a more sensitive method to investigate RT differences in clinical populations. This approach separates the RT distribution into Gaussian and exponential components (Heathcote, 1996; Ratcliff & Murdock, 1976). Analysis of RTs using an ex-Gaussian distribution gives rise to mu (mean RT of the Gaussian component), sigma (SD of RT of the Gaussian component), and tau (mean or SD of RT of the exponential component, indicating the degree of infrequent long RTs; Leth-Steensen et al., 2000). This provides researchers with more sensitive measures to distinguish different types of RTV, such as distinguishing individuals producing consistent RTs most of the time (small sigma) but with several extremely long RTs (large tau), from those who have few extremely long RTs (small tau) but consistent fluctuations around the average RT (large sigma). Large tau has since been consistently observed in ADHD samples (Hervey et al., 2006; Kofler et al., 2013; Leth-Steensen et al., 2000; Tye et al., 2016; Vaurio et al., 2009), taking the place of SD of RT as the most appropriate measure of RTV in ADHD studies (Leth-Steensen et al., 2000). A larger tau in RT is proposed to occur due to lapses in sustained attention, leading to intermittent long responses during a task (Hervey et al., 2006). In a similar way, head motion during resting-state fMRI may arise due to lapses in sustained attention during scanning. If large head movements and long RTs are both influenced by underlying occurrences of lapses in attention, it is important to understand whether the distribution of head motion measures may mirror RT distributions in being positively skewed. This would mean that previous studies using a Gaussian distribution may not have sufficiently captured potential individual or group differences in head motion (explaining inconsistencies in previous ADHD research, for example, Engelhardt et al., 2017; Pardoe et al., 2016) and an ex-Gaussian distribution may be more representative.

The current study will therefore apply an ex-Gaussian distribution to head motion data, yielding measures for mu (indicating mean head motion most of the time), sigma (indicating SD of head motion most of the time), and tau (indicating frequency or magnitude of abnormally large head movements). Behavioral findings suggest that tau in RT distinguishes ADHD from control individuals rather than mu or sigma in RT, an effect thought to be driven by more lapses in sustained attention in those with ADHD. We hypothesize that a similar effect may be evident with head motion, and hence that larger tau in head motion may distinguish children with ADHD from control children, whereas mu and sigma in head motion are comparable between the two groups. If this hypothesis is confirmed, then a mediation analysis could be used to investigate whether the relationship between greater head motion and ADHD might be mediated by an individual’s sustained attention capacity (as measured by omission errors and tau in RT on the SART). Head motion during task-based fMRI has been generally attributed to task-related movement artifacts in previous work; however, if head motion during resting-state fMRI is linked to performance outside of the scanner, then head motion may be better understood as an individual-level trait and form part of the phenotypic presentation of ADHD. To our knowledge, this approach has not been used to simultaneously investigate head motion, sustained attention, and ADHD, and differs from previous work that has centered primarily on blood-oxygen level dependent (BOLD) signal interpretation.

It is predicted that omission errors and tau in RT, respectively, on the SART will mediate the relationship between ADHD diagnosis and head motion in the scanner. Furthermore, the mediating effect of omission errors or tau in RT will be specific to the relationship between ADHD status and tau in head motion, not mu or sigma in head motion, reflecting a specific relationship between infrequent large head movements during MRI scanning and poorer sustained attention.

Method

Participants

Data were collected as part of the Neuroimaging of the Children’s Attention Project (NICAP), a longitudinal community-based study of development and outcomes of children with and without ADHD, approved by the Royal Children’s Hospital Human Research Ethics Committee in Melbourne. Details of the recruitment procedures and study protocol have been previously described (Sciberras et al., 2013; Silk et al., 2016). Briefly, participants were recruited at 6 to 8 years of age, using a two-stage screening and case-confirmation procedure using the Conners 3 ADHD Index (Conners, 2008) and Diagnostic Interview Schedule for Children (DISC-IV; Shaffer et al., 2000). Three years later, 179 participants with written informed consent by the parent completed a cognitive assessment and MRI scan at the Murdoch Children’s Research Institute, Melbourne. Resting-state fMRI for participants with complete behavioral data meeting inclusion criteria (discussed below) were available for a final sample of 56 ADHD individuals and 61 non-ADHD control children, aged 9.7 to 11.9 years (see Table 1 for demographic details).

Table 1.

Demographic Characteristics of Participants.

Variable	ADHD	Control	Test statistic	p
Number	56	61
Males (n, %)	39 (69)	33 (54)	χ² = 2.98	.084
Age (M, SD)	10.5 (0.5)	10.5 (0.4)	t(115) = 0.13	.897
IQ (M, SD)	96.3 (13.7)	103.5 (14.0)	t(115) = 2.83	.006
SES (M, SD)	1,017 (41)	1,018 (47)	t(115) = 0.02	.988
Hyperactive-impulsive symptoms criterion (median, IQR)	4.0 (4.0)	0.0 (1.0)	U = 502	<.001
Inattentive symptoms criterion (median, IQR)	7.0 (3.8)	0.0 (2.0)	U = 258	<.001

Note. ADHD group had a significantly lower intelligence quotient (IQ; p = .006), more hyperactive-impulsive (p < .001), and more inattentive symptoms (p < .001) than the control group. SES = socioeconomic status; IQR = interquartile range.

A subset of 12 children with ADHD were taking medication on the day of the assessment and MRI scan. These children had significantly greater hyperactive-impulsive, U = 857, p < .001, and inattentive, U = 727, p = .032, symptoms than children with ADHD not taking medication; however, there were no significant differences between medicated and nonmedicated ADHD children in terms of any SART performance measures: omission errors, U = 307, p = .395, and tau in RT, t(54) = 1.06, p = .295. Hence, both medicated and nonmedicated children were retained in the ADHD sample.

Procedure

Children and their families attended a 3.5-hr assessment session at the Murdoch Children’s Research Institute, Royal Children’s Hospital, Melbourne, Australia. Full details of the assessment protocol are provided in Silk et al. (2016). A diagnostic interview with the parent was completed to confirm current ADHD status using the DISC-IV (Shaffer et al., 2000). The direct assessment with the child included a cognitive battery (during which the SART was completed), mock MRI, and MRI scan. Researchers conducting child assessment and MRI scan were blinded to diagnostic group. The Wechsler Abbreviated Scale of Intelligence was administered upon recruitment 3 years prior, and the FSIQ-2 (Full Scale Intelligence Quotient) was computed using the vocabulary and matrix reasoning subtests (Wechsler, 1999).

Sustained attention

The SART fixed version (Manly et al., 2003), with evidence of high test–retest reliability and ecological validity (Johnson, Barry, et al., 2008; Smilek et al., 2010), was run through E-prime to measure sustained attention (see Figure 1). The outcome measures for sustained attention were total omission errors, the lack of response to digits other than 3 (Wickens, 2002), and the ex-Gaussian distribution measure tau in RT.

Figure 1.

Procedure of the Sustained Attention to Response Task (SART).

The ex-Gaussian distribution is the convolution of two independent distributions: a Gaussian distribution with mean mu and standard deviation sigma and an exponential distribution with rate (mean and standard deviation) tau. Tau in RT is the ex-Gaussian measure of primary interest to sustained attention research, indicating the degree of positive skew in the RT distribution. RTs were modeled using the ex-Gaussian distribution and parameter estimates obtained using an iterative maximum likelihood estimation in MATLAB R2017a (The MathWorks Inc., Natick, MA) following the procedure of Lacouture and Cousineau (2008).

Resting-state fMRI and head motion

Images were acquired using a 3-Tesla Siemens TIM Trio MRI scanner (Erlangen, Germany). Resting-state functional images were acquired with 250 whole brain volumes over 6 min 33 s (repetition time = 1,500 ms; echo time = 33 ms; flip angle = 85°; field of view = 255 mm; number of slices = 60; in-plane resolution = 2.45 mm × 2.45 mm, multiband = 3). Two additional BOLD sequences were acquired with standard and reverse phase encoding directions for distortion correction lasting 24 s (repetition time = 3,980 ms, echo time = 33 ms, flip angle = 85°, phase encoding direction = reversed). Participants were instructed to stare at a white fixation cross on a black screen and not think of anything in particular.

The FSL Motion Outliers script of the FMRIB Software Library (FSL, Version 5.0; https://fsl.fmrib.ox.ac.uk/fsl) was used to obtain the established head motion metrics delta variation signal (DVARS) and framewise displacement (FD) from the resting-state fMRI images. Taking the BOLD signal time series for each brain volume captured, DVARS indicates the root mean square (RMS) of the change between the current and previous volume, where unnaturally large spikes in BOLD signal are indicative of head movement. In contrast, FD reflects the average change in head motion (rotation and translation parameters) from previous to current volume. Following checks that DVARS and FD values followed an ex-Gaussian distribution, head motion metrics obtained for each volume and person were analyzed using an ex-Gaussian analysis (as described for tau in RT above), yielding estimates for head motion outcome measures mu, sigma, and tau in DVARS and FD.

Statistical testing

Statistical testing was run using Statistical Package for the Social Sciences (SPSS, Version 23) software. Following assessment of variable normality, group differences in head motion and sustained attention performance were tested with independent samples t tests for normally distributed variables and Mann–Whitney U tests for nonnormally distributed variables. Following this, mediation analysis was conducted using PROCESS, a mediation and moderation modeling tool for SPSS described in Hayes (2017). The model selected (Model 4—basic mediation) tested whether sustained attention performance (mediator variable M) mediated the effect of ADHD diagnosis (predictor variable X) in predicting head motion (outcome variable Y). Age and sex were included as covariates. Following recommendations from Dennis et al. (2000), IQ was not included as a covariate. Effects were considered significant if 95% confidence intervals (CIs) based on 5,000 bootstrap samples did not include zero.

Results

Group Differences in Head Motion

Initial Mann–Whitney U tests examined differences between ADHD and control groups in in-scanner head motion (Table 2). Tau in DVARS was significantly larger in the ADHD (Mdn = 7.69, interquartile range [IQR] = 10.73) than control (Mdn = 5.21, IQR = 6.59) group, U = 1,276, p = .019. Similarly, tau in FD was significantly larger in the ADHD (Mdn = 0.156, IQR = 0.215) than control (Median = 0.093, IQR = 0.070) group, U = 1,062, p < .001, demonstrating that the ADHD group produced more and/or greater abnormally large head movements (see Figure 2 for distribution plots of tau in DVARS and tau in FD). There were no significant differences between the ADHD and control group in mu or sigma in head motion (see Supplementary Figure 1 for relevant distribution plots).

Table 2.

ADHD and Control Group Differences in in-Scanner Head Motion.

Motion variable	Median (IQR)		Test statistic	p
Motion variable	ADHD	Control	Test statistic	p
Mu in DVARS	29.60 (9.63)	30.79 (7.85)	1,990	.125
Sigma in DVARS	2.38 (1.59)	2.44 (1.70)	1,830	.507
Tau in DVARS	7.69 (10.73)	5.21 (6.59)	1,276	.019
Mu in FD	0.063 (0.046)	0.060 (0.038)	1,563	.430
Sigma in FD	0.022 (0.024)	0.023 (0.020)	1,783	.684
Tau in FD	0.156 (0.215)	0.093 (0.070)	1,062	<.001

Note. Mann–Whitney U tests conducted for nonnormally distributed head motion variables. IQR = interquartile range; DVARS = delta variation signal; FD = framewise displacement.

Figure 2.

Violin plots depicting the distribution of head motion measures tau in delta variation signal (DVARS) (left) and tau in framewise displacement (FD) (right).

Group Differences in Sustained Attention Performance

A Mann–Whitney U test revealed greater omission errors on the SART in the ADHD (Mdn = 8.0, IQR = 10.0) than control (Mdn = 3.0, IQR = 7.5) group, U = 1,227, p = .008. Despite a strong trend toward greater tau in RT in the ADHD group, an independent samples t test indicated no significant differences in tau in RT between the ADHD (M = 155, SD = 82) and control (M = 127, SD = 77) group, t(114) = −1.85, p = .066. See Figure 3 for distribution plots of omission errors and tau in RT.

Figure 3.

Violin plots depicting the distribution of sustained attention performance measures omission errors (left) and tau in response time (RT) (right).

Mediation Analysis Predicting In-Scanner Head Motion

Given that omissions and large tau in RT are proposed to indicate lapses in attention, a mediation analysis was conducted to investigate the potential role of sustained attention performance (omission errors) in mediating the head motion behavior seen in children with ADHD during MRI scanning. Figure 4 depicts the model testing the mediating role of omission errors in the demonstrated relationship between ADHD diagnosis and tau in DVARS. Omission errors mediated the relationship between ADHD diagnosis and tau in DVARS, indirect effect: B = 1.29, 95% CI = [0.07, 3.15], accounting for 29% of the association between ADHD diagnosis and tau in DVARS. ADHD diagnosis significantly predicted omission errors, B = 2.67, SE = 1.21, and omission errors predicted tau in FD, B = 0.006, SE = 0.002; however, omission errors did not appear to mediate the relationship between ADHD diagnosis and tau in FD, indirect effect: B = 0.02, 95% CI = [0.00, 0.05]. Results held when age and sex were included as covariates (see Supplementary Tables 1 and 2).

Figure 4.

Mediation model depicting full mediation of omission errors on the relationship between ADHD diagnosis and tau in delta variation signal (DVARS).

Discussion

This study investigated the role of lapses in sustained attention in the association between ADHD and in-scanner head motion. Many previous studies have investigated sustained attention performance in individuals with ADHD (Tamm et al., 2012). Researchers have also become increasingly interested in the neural basis of sustained attention and have naturally focused on children with and without ADHD (Christakou et al., 2013; Rubia, 2011), however, results have been variable (Cubillo & Rubia, 2010). Recent evidence suggests that inconsistency in MRI research could be the result of greater head motion in those with ADHD than controls distorting the MRI data and leading to data loss (Pardoe et al., 2016), yet no studies have considered the nature of this head motion in more detail or how it might specifically relate to sustained attention rather than general ADHD diagnosis or symptoms. The current study modeled head motion as an ex-Gaussian distribution, with the hypothesis that tau in head motion (infrequent large head movements) would be predicted by sustained attention performance over and above ADHD diagnosis. Furthermore, this relationship was predicted to be specific to tau in head motion, whereby mu or sigma in head motion would not be related to attention problems.

Initial group comparisons revealed greater tau in DVARS and tau in FD in the ADHD group compared with the control group, suggesting that those with ADHD have more frequent or extreme large head movements during scanning. These findings accord with previous research demonstrating greater head motion in the ADHD than control group (Kong et al., 2014; Pardoe et al., 2016). However, the current study further found that there were no group differences in mu or sigma in head motion, indicating that average head motion and smaller fluctuations in head motion were comparable between the ADHD and control groups. Using this ex-Gaussian decomposition of in-scanner head motion, the current results provide new insights into the nature of head movements in the scanner. Infrequent large head movements in the scanner specifically, rather than broadly greater head motion, may be characteristic of ADHD. Lapses in sustained attention are common in ADHD (Johnson, Robertson, et al., 2007; Leth-Steensen et al., 2000) and could explain such occasional extreme head movements.

Previous research has suggested that omission errors and tau in RT in the SART may be behavioral indicators of lapses in sustained attention. The current study found evidence of greater omission errors in the ADHD than control group, which is consistent with previous research (see Kuntsi et al., 2010), and suggests that this group may have more frequent lapses in sustained attention during the SART (Johnson, Robertson, et al., 2007). Interestingly, there were no significant differences in tau in RT between the ADHD and control group in the current study, contrasting with previous research finding greater tau in RT in those with ADHD (Karalunas et al., 2014; Tye et al., 2016). Larger tau in RT has been proposed to occur due to lapses in sustained attention (Leth-Steensen et al., 2000); hence, a link between infrequent long RTs and infrequent large head movements was expected. This result needs to be replicated in future studies to be confirmed.

Targets in the fixed SART are predictable and frequent; hence, missing a target (represented by an omission error) is proposed to be caused by a lapse in sustained attention to the task (Johnson, Robertson, et al., 2007). In a similar way, infrequent large head movements in the scanner may be attributed to lapses in sustained attention. This hypothesis was tested in the current study using a mediation analysis to investigate whether infrequent large head movements evident in the ADHD group may be better predicted by sustained attention performance (omission errors on the SART) than general ADHD group membership. When omission errors were included as a mediator, they mediated the relationship between the ADHD group and tau in DVARS, explaining 29% of the association between ADHD group and tau in DVARS. The results suggest that children with ADHD have more infrequent large head movements during fMRI in general; however, participants with ADHD demonstrating particular deficits in sustained attention outside of the scanner may be more at risk of committing these large head movements during scanning. This aligns with behavioral research suggesting that within those with ADHD there may be subgroups of individuals possessing greater deficits in sustained attention (Johnson, Kelly, et al., 2007; Johnson, Kelly, et al., 2008). Furthermore, although previous work has found moderate correlations between head motion and inattentive symptoms (Couvy-Duchesne et al., 2016), this study is the first to consider the potential link between out-of-scanner sustained attention performance and in-scanner head motion. It is interesting to note that there may be differences in head motion measures depending on an individual’s specific distribution of ADHD symptoms (e.g., greater tau in head motion in individuals with more inattentive symptoms and greater sigma in those displaying more hyperactive-impulsive symptoms). Taking a dimensional approach, it will be important for future research to examine the potential association between the different ADHD symptom domains and head motion measures, and how this changes over development.

Omission errors did not mediate the relationship between the ADHD group and tau in FD (lower limit of the indirect effect was zero). This is surprising, given that attention lapses did mediate the relationship between ADHD group and tau in head motion when head motion was measured using DVARS. FD represents the sum of the absolute value of the change in brain position (displacement values), from the previous to current volume, whereas DVARS calculates the RMS change in the temporal derivative of the BOLD time series from previous to current volume (Sylvester et al., 2013). Despite differences in the calculation, and hence focus of each measure, it is currently unknown whether DVARS or FD represents motion data more accurately and studies could continue the work of Power et al. (2014) to determine which metric best reflects physical head motion.

Establishing evidence of a link between lapses in attention outside of the scanner and tau in in-scanner head motion is an important conceptual point for imaging studies in psychiatry, which often regard head motion as nuisance signal alone, and highlights a fundamental barrier for successful neuroimaging studies of sustained attention in ADHD. Recent work has shown that removal of participants with excessive head motion leads to a sampling bias toward those with greater cognitive ability (Wylie et al., 2014), and necessary regression procedures for head motion may lead to increases in false-negative rates by creating a much lower signal-to-noise ratio in clinical than control groups (Couvy-Duchesne et al., 2016; Kong et al., 2014). This could explain why previous MRI findings have been unable to establish a clear neural basis for sustained attention impairment in ADHD as this two-step motion correction process may result in studies comparing the least severe (or highest functioning) children with ADHD to controls.

A number of different MRI processing methods have been developed, which aim to reduce the effect of head motion on imaging measures such as functional connectivity (Caballero-Gaudes & Reynolds, 2017). A recent study by Satterthwaite et al. (2019) suggests that the most successful technique for minimizing the effect of head motion on fMRI data may be to apply a combination of different regression strategies (e.g., global signal regression and regression of time series from white matter, cerebrospinal fluid, and realignment parameters). The optimal denoising approach still requires further consideration, however, with the most effective denoising techniques such as global signal regression (particularly when considering groups with differing levels of in-scanner head motion) also appearing to amplify the distant-dependent relationship between head motion and functional connectivity (Satterthwaite et al., 2013). A longer term solution may be to consider alternative approaches to resting-state fMRI data collection. Research has found that functional connectivity observed during resting-state fMRI may be stable within individuals over short time periods (e.g., over several minutes; Laumann et al., 2016). One approach to avoid contamination of MRI data due to head motion could therefore be to continuously measure motion during scanning and extend the MRI scan time until a minimum number of clean or motion-free frames has been acquired (Dosenbach et al., 2017). This could reduce biases in clinical and control group comparisons by ensuring that both ADHD and control samples, for example, have the same number of clean data frames available for analysis. A brief sustained attention task completed before scanning could be one feasible method of identifying children with a high risk of extreme head motion, allowing MRI scan times to be adjusted in advance for participants who are likely to require a longer scan time to achieve the desired level of data quality.

The current findings therefore have critical implications for appropriate collection and treatment of MRI data in hyperkinetic and inattentive populations such as those with ADHD (Posner et al., 2014). Nevertheless, there are key points that must be considered in interpreting the current findings and firmly establishing the relationship between head motion and cognitive performance. Being the first study to employ an ex-Gaussian distribution to systematically decompose head motion data, replication of these findings using independent cohorts will be important to ensure consistency of results across multiple testing sites and samples. Furthermore, although the current study primarily considered sustained attention, previous ADHD work has also found a relationship between scores on an impulsivity scale and head motion (Kong et al., 2014). Considering the ex-Gaussian procedure implemented in the current study, tasks such as CPT yielding a measure of impulsivity (Riccio et al., 2002) could provide a deeper understanding of the relationship between head motion behavior and cognitive performance, potentially also revealing a link between greater head motion most of the time (larger mu) or greater fluctuations around average head motion (larger sigma) and impulsivity performance. Research by Van Dijk et al. (2012) and Engelhardt et al. (2017) additionally suggests that while the degree of head motion is stable within participants over a short period, head motion is also related to age. Therefore, applying the current method to longitudinal data could establish whether this relationship between head motion and sustained attention performance is stable within participants over time or alters through development.

Considering the implications of the current results more broadly, Bright and Murphy (2015) found that variance removed when regressing out head motion possessed a recognizable network structure resembling those networks observed in functional connectivity studies. Future work could extend the current findings to establish the neural networks associated with head motion and their potential correspondence with sustained attention networks. There is also initial evidence linking specific alleles to poorer sustained attention in ADHD (Johnson, Kelly, et al., 2008), as well as evidence of shared genetic factors between head motion and inattentive symptoms (Couvy-Duchesne et al., 2016). In light of this, further work could take advantage of genetic data to explore whether sustained attention performance and head motion might arise from common genetic influences. In addition, while the current study focused on ADHD, there are several other disorders, such as autism spectrum disorder and multiple sclerosis, which similarly display impaired cognitive ability and greater head motion in the scanner (Christakou et al., 2013; Pardoe et al., 2016; Wylie et al., 2014). Differences in cognitive ability and head motion are also present even in neurotypical samples (Robertson et al., 1997). Applying an ex-Gaussian distribution analysis to better understand head motion behavior in these populations could be invaluable to promote more effective management of head motion across both clinical and nonclinical samples.

In summary, initial MRI research has revealed a relationship between ADHD and increased in-scanner head motion; however, no studies to date had examined the nature of this head motion and its behavioral correlates beyond basic symptom levels. This study employed a novel application of an ex-Gaussian distribution analysis to head motion data and demonstrated a specific link between poor sustained attention outside of the scanner and infrequent large head movements during resting-state fMRI. These results challenge whether current standard MRI scanning and analysis protocols are adequate to validly investigate the neural basis of sustained attention. Findings also add to evidence of subgroups within the broad ADHD umbrella possessing distinct cognitive and head motion profiles (Fair et al., 2012). Recommendations for future work include investigating the relationship between head motion and brain activity, other cognitive impairments including impulsivity, and genetic factors. More careful collection of MRI data, such as through real-time monitoring of in-scanner head motion, and careful treatment of head motion during analysis would maximize the validity of future studies. This could allow for previously unseen clarity in the ADHD neuroimaging literature, ultimately leading to increased progress toward understanding the neural basis of sustained attention impairments in ADHD and hence targets for interventions to promote effective brain development.

Supplemental Material

FigS1 – Supplemental material for Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder

Supplemental material, FigS1 for Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder by Phoebe Thomson, Katherine A. Johnson, Charles B. Malpas, Daryl Efron, Emma Sciberras and Timothy J. Silk in Journal of Attention Disorders

Supplemental Material

Supplementary_Table_1 – Supplemental material for Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder

Supplemental material, Supplementary_Table_1 for Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder by Phoebe Thomson, Katherine A. Johnson, Charles B. Malpas, Daryl Efron, Emma Sciberras and Timothy J. Silk in Journal of Attention Disorders

Supplemental Material

Supplementary_Table_2 – Supplemental material for Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder

Supplemental material, Supplementary_Table_2 for Head Motion During MRI Predicted by out-of-Scanner Sustained Attention Performance in Attention-Deficit/Hyperactivity Disorder by Phoebe Thomson, Katherine A. Johnson, Charles B. Malpas, Daryl Efron, Emma Sciberras and Timothy J. Silk in Journal of Attention Disorders

Footnotes

Authors’ Note

This study was presented as an abstract at the Australasian Cognitive Neuroscience Society Conference, Melbourne, Australia, November 22–25, 2018.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by the National Health and Medical Research Council of Australia (NHMRC; 1065895, 1008522) and the Collier Foundation, Melbourne, Australia. E.S. was supported by NHMRC Early Career (1037159) and Career Development (1110688) Fellowships. T.J.S. was supported by an NHMRC Career Development Award (1004637). D.E. is supported by a Clinical Scientist Fellowship from the Murdoch Children’s Research Institute. The research was also supported by The Royal Children’s Hospital, The Royal Children’s Hospital Foundation, Department of Pediatrics at the University of Melbourne, Australian Commonwealth Government, and the Victorian Government’s Operational Infrastructure Support Program.

ORCID iDs

Phoebe Thomson

Emma Sciberras

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Phoebe Thomson is a doctoral student in the Department of Paediatrics at the University of Melbourne and Murdoch Children’s Research Institute, Australia. Her research focuses on the developmental trajectories of neural and cognitive function in children and adolescents with ADHD.

Katherine A. Johnson is an associate professor in Developmental Psychology at the Melbourne School of Psychological Sciences, University of Melbourne. Her research is in the area of developmental cognitive neuroscience. She is interested in executive control of behaviour, with a special focus on attention control in infants, children, and young adults.

Charles B. Malpas is a senior clinical neuropsychologist in the Department of Neurology at the Alfred Hospital and a senior research fellow in the Clinical Outcomes Research Unit, Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Australia. Charles also holds honorary appointments at the Murdoch Children’s Research Institute and the Department of Neuroscience, Monash University.

Daryl Efron is a developmental-behavioral pediatrician at the Royal Children’s Hospital, Melbourne, Australia; a senior research fellow at the Murdoch Children’s Research Institute, Melbourne, Australia; and associate professor in the Department of Paediatrics, The University of Melbourne.

Emma Sciberras is an associate professor and clinical psychologist in the School of Psychology at Deakin University, Australia. She is also an honorary research fellow at the Murdoch Children’s Research Institute and the Department of Paediatrics at the University of Melbourne, Australia.

Timothy J. Silk is an associate professor in the School of Psychology, Deakin University, Australia, and honorary fellow in the Developmental Imaging group at the Murdoch Children’s Research Institute, Melbourne, Australia.

References

Adamo

Di Martino

Esu

Petkova

Johnson

Kelly

. . . Zuddas

(2014). Increased response-time variability across different cognitive tasks in children with ADHD. Journal of Attention Disorders, 18(5), 434–446.

Adamo

Huo

Adelsberg

Petkova

Castellanos

F. X.

Di Martino

(2014). Response time intra-subject variability: Commonalities between children with autism spectrum disorders and children with ADHD. European Child and Adolescent Psychiatry, 23(2), 69–79.

Andrews-Hanna

J. R.

Snyder

A. Z.

Vincent

J. L.

Lustig

Head

Raichle

M. E.

Buckner

R. L.

(2007). Disruption of large-scale brain systems in advanced aging. Neuron, 56(5), 924–935.

Bellgrove

M. A.

Hawi

Gill

Robertson

I. H.

(2006). The cognitive genetics of attention deficit hyperactivity disorder (ADHD): Sustained attention as a candidate phenotype. Cortex, 42(6), 838–845.

Bellgrove

M. A.

Hawi

Kirley

Gill

Robertson

I. H.

(2005). Dissecting the attention deficit hyperactivity disorder (ADHD) phenotype: Sustained attention, response variability and spatial attentional asymmetries in relation to dopamine transporter (DAT1) genotype. Neuropsychologia, 43(13), 1847–1857.

Braet

Johnson

K. A.

Tobin

C. T.

Acheson

McDonnell

Hawi

. . . Bellgrove

M. A.

(2011). fMRI activation during response inhibition and error processing: The role of the DAT1 gene in typically developing adolescents and those diagnosed with ADHD. Neuropsychologia, 49(7), 1641–1650.

Bright

M. G.

Murphy

(2015). Is fMRI “noise” really noise? Resting state nuisance regressors remove variance with network structure. NeuroImage, 114, 158–169.

Caballero-Gaudes

Reynolds

R. C.

(2017). Methods for cleaning the BOLD fMRI signal. NeuroImage, 154, 128–149.

Christakou

Murphy

Chantiluke

Cubillo

Smith

Giampietro

. . . Murphy

(2013). Disorder-specific functional abnormalities during sustained attention in youth with attention deficit hyperactivity disorder (ADHD) and with autism. Molecular Psychiatry, 18(2), 236–244.

10.

Conners

C. K.

(2008). Conners 3rd edition: Manual. Multi-Health Systems.

11.

Costa Dias

T. G.

Wilson

V. B.

Bathula

D. R.

Iyer

S. P.

Mills

K. L.

Thurlow

B. L.

. . . Grayson

D. S.

(2013). Reward circuit connectivity relates to delay discounting in children with attention-deficit/hyperactivity disorder. European Neuropsychopharmacology, 23(1), 33–45.

12.

Couvy-Duchesne

Ebejer

J. L.

Gillespie

N. A.

Duffy

D. L.

Hickie

I. B.

Thompson

P. M.

. . . Medland

S. E.

(2016). Head motion and inattention/hyperactivity share common genetic influences: Implications for fMRI studies of ADHD. PLOS ONE, 11(1), Article e0146271.

13.

Cubillo

Rubia

(2010). Structural and functional brain imaging in adult attention-deficit/hyperactivity disorder. Expert Review of Neurotherapeutics, 10(4), 603–620.

14.

Dennis

Lockyer

Lazenby

A. L.

(2000). How high-functioning children with autism understand real and deceptive emotion. Autism, 4(4), 370–381.

15.

Dosenbach

N. U.

Koller

J. M.

Earl

E. A.

Miranda-Dominguez

Klein

R. L.

Van

A. N.

. . . Nguyen

A. L.

(2017). Real-time motion analytics during brain MRI improve data quality and reduce costs. NeuroImage, 161, 80–93.

16.

Engelhardt

L. E.

Roe

M. A.

Juranek

DeMaster

Harden

K. P.

Tucker-Drob

E. M.

Church

J. A.

(2017). Children’s head motion during fMRI tasks is heritable and stable over time. Developmental Cognitive Neuroscience, 25, 58–68.

17.

Fair

D. A.

Bathula

Nikolas

M. A.

Nigg

J. T.

(2012). Distinct neuropsychological subgroups in typically developing youth inform heterogeneity in children with ADHD. Proceedings of the National Academy of Sciences, 109(17), 6769–6774.

18.

Friston

K. J.

Williams

Howard

Frackowiak

R. S.

Turner

(1996). Movement-related effects in fMRI time-series. Magnetic Resonance in Medicine, 35(3), 346–355.

19.

Hayes

A. F.

(2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press.

20.

Heathcote

(1996). RTSYS: A DOS application for the analysis of reaction time data. Behavior Research Methods, Instruments, & Computers, 28(3), 427–445.

21.

Hervey

A. S.

Epstein

J. N.

Curry

J. F.

Tonev

Eugene Arnold

Keith Conners

. . . Hechtman

(2006). Reaction time distribution analysis of neuropsychological performance in an ADHD sample. Child Neuropsychology, 12(2), 125–140.

22.

Johnson

K. A.

Barry

Bellgrove

M. A.

Cox

Kelly

S. P.

Dáibhis

. . . Fitzgerald

(2008). Dissociation in response to methylphenidate on response variability in a group of medication naive children with ADHD. Neuropsychologia, 46(5), 1532–1541.

23.

Johnson

K. A.

Kelly

S. P.

Bellgrove

M. A.

Barry

Cox

Gill

Robertson

I. H.

(2007). Response variability in attention deficit hyperactivity disorder: Evidence for neuropsychological heterogeneity. Neuropsychologia, 45(4), 630–638.

24.

Johnson

K. A.

Kelly

S. P.

Robertson

I. H.

Barry

Mulligan

Daly

. . . Hawi

(2008). Absence of the 7-repeat variant of the DRD4 VNTR is associated with drifting sustained attention in children with ADHD but not in controls. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 147(6), 927–937.

25.

Johnson

K. A.

Robertson

I. H.

Kelly

S. P.

Silk

T. J.

Barry

Dáibhis

. . . Gallagher

(2007). Dissociation in performance of children with ADHD and high-functioning autism on a task of sustained attention. Neuropsychologia, 45(10), 2234–2245.

26.

Karalunas

S. L.

Geurts

H. M.

Konrad

Bender

Nigg

J. T.

(2014). Annual research review: Reaction time variability in ADHD and autism spectrum disorders: Measurement and mechanisms of a proposed trans-diagnostic phenotype. Journal of Child Psychology and Psychiatry, 55(6), 685–710.

27.

Kofler

M. J.

Rapport

M. D.

Sarver

D. E.

Raiker

J. S.

Orban

S. A.

Friedman

L. M.

Kolomeyer

E. G.

(2013). Reaction time variability in ADHD: A meta-analytic review of 319 studies. Clinical Psychology Review, 33(6), 795–811.

28.

Kong

X. Z.

Zhen

H. H.

Wang

Liu

. . . Liu

(2014). Individual differences in impulsivity predict head motion during magnetic resonance imaging. PLOS ONE, 9(8), Article e104989.

29.

Kuntsi

Wood

A. C.

Johnson

K. A.

Andreou

Arias-Vasquez

Buitelaar

J. K.

. . . Uebel

(2010). Separation of cognitive impairments in attention-deficit/hyperactivity disorder into 2 familial factors. Archives of General Psychiatry, 67(11), 1159–1166.

30.

Lacouture

Cousineau

(2008). How to use MATLAB to fit the ex-Gaussian and other probability functions to a distribution of response times. Tutorials in Quantitative Methods for Psychology, 4(1), 35–45.

31.

Laumann

T. O.

Snyder

A. Z.

Mitra

Gordon

E. M.

Gratton

Adeyemo

. . . Greene

D. J.

(2016). On the stability of BOLD fMRI correlations. Cerebral Cortex, 27(10), 4719–4732.

32.

Leth-Steensen

Elbaz

Z. K.

Douglas

V. I.

(2000). Mean response times, variability, and skew in the responding of ADHD children: A response time distributional approach. Acta Psychologica, 104(2), 167–190.

33.

Manly

Owen

A. M.

McAvinue

Datta

Lewis

G. H.

Scott

S. K.

. . . Robertson

I. H.

(2003). Enhancing the sensitivity of a sustained attention task to frontal damage: Convergent clinical and functional imaging evidence. Neurocase, 9(4), 340–349.

34.

Milham

M. P.

Fair

Mennes

Mostofsky

S. H.

(2012). The ADHD-200 consortium: A model to advance the translational potential of neuroimaging in clinical neuroscience. Frontiers in Systems Neuroscience, 6, Article 62.

35.

O’Connell

R. G.

Bellgrove

M. A.

Dockree

P. M.

Robertson

I. H.

(2004). Reduced electrodermal response to errors predicts poor sustained attention performance in attention deficit hyperactivity disorder. Neuroreport, 15(16), 2535–2538.

36.

O’Connell

R. G.

Bellgrove

M. A.

Dockree

P. M.

Robertson

I. H.

(2006). Cognitive remediation in ADHD: Effects of periodic non-contingent alerts on sustained attention to response. Neuropsychological Rehabilitation, 16(6), 653–665.

37.

Pardoe

H. R.

Hiess

R. K.

Kuzniecky

(2016). Motion and morphometry in clinical and nonclinical populations. NeuroImage, 135, 177–185.

38.

Polanczyk

De Lima

M. S.

Horta

B. L.

Biederman

Rohde

L. A.

(2007). The worldwide prevalence of ADHD: A systematic review and metaregression analysis. American Journal of Psychiatry, 164(6), 942–948.

39.

Posner

Park

Wang

(2014). Connecting the dots: A review of resting connectivity MRI studies in attention-deficit/hyperactivity disorder. Neuropsychology Review, 24(1), 3–15.

40.

Power

J. D.

Barnes

K. A.

Snyder

A. Z.

Schlaggar

B. L.

Petersen

S. E.

(2012). Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. NeuroImage, 59(3), 2142–2154.

41.

Power

J. D.

Mitra

Laumann

T. O.

Snyder

A. Z.

Schlaggar

B. L.

Petersen

S. E.

(2014). Methods to detect, characterize, and remove motion artifact in resting state fMRI. NeuroImage, 84, 320–341.

42.

Ratcliff

Murdock

B. B.

(1976). Retrieval processes in recognition memory. Psychological Review, 83(3), 190–214.

43.

Riccio

C. A.

Reynolds

C. R.

Lowe

Moore

J. J.

(2002). The continuous performance test: A window on the neural substrates for attention? Archives of Clinical Neuropsychology, 17(3), 235–272.

44.

Robertson

I. H.

Manly

Andrade

Baddeley

B. T.

Yiend

(1997). “Oops!”: Performance correlates of everyday attentional failures in traumatic brain injured and normal subjects. Neuropsychologia, 35(6), 747–758.

45.

Rubia

(2011). “Cool” inferior frontostriatal dysfunction in attention-deficit/hyperactivity disorder versus “hot” ventromedial orbitofrontal-limbic dysfunction in conduct disorder: A review. Biological Psychiatry, 69(12), e69–e87.

46.

Satterthwaite

T. D.

Ciric

Roalf

D. R.

Davatzikos

Bassett

D. S.

Wolf

D. H.

(2019). Motion artifact in studies of functional connectivity: Characteristics and mitigation strategies. Human Brain Mapping, 40(7), 2033–2051.

47.

Satterthwaite

T. D.

Elliott

M. A.

Gerraty

R. T.

Ruparel

Loughead

Calkins

M. E.

. . . Gur

R. E.

(2013). An improved framework for confound regression and filtering for control of motion artifact in the preprocessing of resting-state functional connectivity data. NeuroImage, 64, 240–256.

48.

Satterthwaite

T. D.

Wolf

D. H.

Loughead

Ruparel

Elliott

M. A.

Hakonarson

. . . Gur

R. E.

(2012). Impact of in-scanner head motion on multiple measures of functional connectivity: Relevance for studies of neurodevelopment in youth. NeuroImage, 60(1), 623–632.

49.

Sciberras

Efron

Schilpzand

E. J.

Anderson

Jongeling

Hazell

. . . Nicholson

J. M.

(2013). The Children’s Attention Project: A community-based longitudinal study of children with ADHD and non-ADHD controls. BMC Psychiatry, 13(1), Article 18.

50.

Shaffer

Fisher

Lucas

C. P.

Dulcan

M. K.

Schwab-Stone

M. E.

(2000). NIMH Diagnostic Interview Schedule for Children Version IV (NIMH DISC-IV): Description, differences from previous versions, and reliability of some common diagnoses. Journal of the American Academy of Child and Adolescent Psychiatry, 39(1), 28–38.

51.

Silk

T. J.

Genc

Anderson

Efron

Hazell

Nicholson

J. M.

. . . Sciberras

(2016). Developmental brain trajectories in children with ADHD and controls: A longitudinal neuroimaging study. BMC Psychiatry, 16(1), Article 59.

52.

Smilek

Carriere

J. S.

Cheyne

J. A.

(2010). Failures of sustained attention in life, lab, and brain: Ecological validity of the SART. Neuropsychologia, 48(9), 2564–2570.

53.

Sylvester

C. M.

Barch

D. M.

Corbetta

Power

J. D.

Schlaggar

B. L.

Luby

J. L.

(2013). Resting state functional connectivity of the ventral attention network in children with a history of depression or anxiety. Journal of the American Academy of Child and Adolescent Psychiatry, 52(12), 1326–1336.e5.

54.

Tamm

Narad

M. E.

Antonini

T. N.

O’Brien

K. M.

Hawk

L. W.

Epstein

J. N.

(2012). Reaction time variability in ADHD: A review. Neurotherapeutics, 9(3), 500–508.

55.

Tye

Johnson

K. A.

Kelly

S. P.

Asherson

Kuntsi

Ashwood

K. L.

. . . McLoughlin

(2016). Response time variability under slow and fast-incentive conditions in children with ASD, ADHD and ASD+ ADHD. Journal of Child Psychology and Psychiatry, 57(12), 1414–1423.

56.

Van Dijk

K. R.

Sabuncu

M. R.

Buckner

R. L.

(2012). The influence of head motion on intrinsic functional connectivity MRI. NeuroImage, 59(1), 431–438.

57.

Vaurio

R. G.

Simmonds

D. J.

Mostofsky

S. H.

(2009). Increased intra-individual reaction time variability in attention-deficit/hyperactivity disorder across response inhibition tasks with different cognitive demands. Neuropsychologia, 47(12), 2389–2396.

58.

Wechsler

(1999). Manual for the Wechsler abbreviated intelligence scale (WASI). The Psychological Corporation.

59.

Wickens

T. D.

(2002). Elementary signal detection theory. Oxford University Press.

60.

Wylie

G. R.

Genova

DeLuca

Chiaravalloti

Sumowski

J. F.

(2014). Functional magnetic resonance imaging movers and shakers: Does subject-movement cause sampling bias? Human Brain Mapping, 35(1), 1–13.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.15 MB

0.02 MB