Associations between attention‐deficit hyperactivity disorder (ADHD) symptom remission and white matter microstructure: A longitudinal analysis

Abstract

Background Attention‐deficit hyperactivity disorder (ADHD) is associated with white matter (WM) microstructure. Our objective was to investigate how WM microstructure is longitudinally related to symptom remission in adolescents and young adults with ADHD. Methods We obtained diffusion‐weighted imaging (DWI) data from 99 participants at two time‐points (mean age baseline: 16.91 years, mean age follow‐up: 20.57 years). We used voxel‐wise Tract‐Based Spatial Statistics (TBSS) with permutation‐based inference to investigate associations of inattention (IA) and hyperactivity‐impulsivity (HI) symptom change with fractional anisotropy (FA) at baseline, follow‐up, and change between time‐points. Results Remission of combined HI and IA symptoms was significantly associated with reduced FA at follow‐up in the left superior longitudinal fasciculus and the left corticospinal tract (CST; PFWE = 0.038 and PFWE = 0.044, respectively), mainly driven by an association between HI remission and follow‐up CST FA (PFWE = 0.049). There was no significant association of combined symptom decrease with FA at baseline or with changes in FA between the two assessments. Conclusions In this longitudinal DWI study of ADHD using dimensional symptom scores, we show that greater symptom decrease is associated with lower follow‐up FA in specific WM tracts. Altered FA thus may appear to follow, rather than precede, changes in symptom remission. Our findings indicate divergent WM developmental trajectories between individuals with persistent and remittent ADHD, and support the role of prefrontal and sensorimotor tracts in the remission of ADHD.

Longitudinal changes of ADHD symptoms in association with white matter microstructure: a tract-specific fixel-based analysis

“…Because our smaller sample size at an older age range is not suitable for a data-driven search to discover any new relevant regions or tracts, the present analyses were intended to further understand the nature of our previous results. To compare FBA metrics to our previous FA findings, our first follow-up analysis used the same exact sample as our most recent longitudinal DTI study [17]. We hypothesized that, like our earlier findings, ADHD symptom improvement would be associated with lower follow-up WM microstructure in the lSLF and lCST.…”

“…For continuous measures of symptom dimension severity and in accordance with our previous report, we used raw total Conners' Parent Rating Scale (CPRS) scores from W1 and W2, and Conners' Adult ADHD Rating Scale (CAARS) scores from W3 for hyperactivity-impulsivity (HI) and inattention (IA) [17,32,33]. Here, we define symptom change (Δ) as the Conners' score difference: Δscore = scorefollow-up -scorebaseline Baseline and follow-up scores were always positively correlated with each other (Figure S1).…”

“…At W2, single-shell DWI data were acquired with a 1.5-Tesla AVANTO scanner (Siemens, Erlangen, Germany) equipped with an 8-channel receive-only phased-array head coil using the following parameters: echo time/repetition time (TE/TR) = 97/8500 ms; GRAPPA-acceleration factor 2; voxel size = 2 × 2 × 2.2 mm; b-values = 0 (5 volumes, interleaved) and 1000 (60 directions) s/mm 2 ; twice refocused pulsed-gradient spin-echo EPI; no partial Fourier. More details of this MRI data acquisition have been described previously [16,17]. Because our models only included follow-up neuroimaging data as a further investigation of the aforementioned analyses, we did not include W1 DWI.…”

“…A longitudinal design reveals temporal dynamics of underlying neurobiological processes and increases statistical power by reducing inter-subject variability. The NeuroIMAGE study and its latest follow-up, DELTA, is a longitudinal cohort of ADHD-affected probands, their siblings, and unaffected controls from childhood to adulthood [10,[14][15][16][17]. We previously demonstrated that, at two different time-points and in two partly overlapping NeuroIMAGE samples, more improvement in combined and hyperactivity-impulsivity symptom scores over time was associated with lower FA at follow-up in an area where the left corticospinal tract (lCST) crosses the left superior longitudinal fasciculus (lSLF) [10,17].…”

“…The NeuroIMAGE study and its latest follow-up, DELTA, is a longitudinal cohort of ADHD-affected probands, their siblings, and unaffected controls from childhood to adulthood [10,[14][15][16][17]. We previously demonstrated that, at two different time-points and in two partly overlapping NeuroIMAGE samples, more improvement in combined and hyperactivity-impulsivity symptom scores over time was associated with lower FA at follow-up in an area where the left corticospinal tract (lCST) crosses the left superior longitudinal fasciculus (lSLF) [10,17]. In that same report, we also systematically demonstrated that symptom change is associated with neither baseline FA nor change in FA from baseline to follow-up.…”