Oscillatory motor patterning is impaired in neurofibromatosis type 1: a behavioural, EEG and fMRI study

Abstract

Background: Neurofibromatosis type1 (NF1) is associated with a broad range of behavioural deficits, and an imbalance between excitatory and inhibitory neurotransmission has been postulated in this disorder. Inhibition is involved in the control of frequency and stability of motor rhythms. Therefore, we aimed to explore the link between behavioural motor control, brain rhythms and brain activity, as assessed by EEG and fMRI in NF1.

Methods: We studied a cohort of 21 participants with NF1 and 20 age- and gender-matched healthy controls, with a finger-tapping task requiring pacing at distinct frequencies during EEG and fMRI scans.

Results: We found that task performance was significantly different between NF1 and controls, the latter showing higher tapping time precision. The time-frequency patterns at the beta sub-band (20-26 Hz) mirrored the behavioural modulations, with similar cyclic synchronization/desynchronization patterns for both groups. fMRI results showed a higher recruitment of the extrapyramidal motor system (putamen, cerebellum and red nucleus) in the control group during the fastest pacing condition.

Conclusions: The present study demonstrated impaired precision in rhythmic pacing behaviour in NF1 as compared with controls. We found a decreased recruitment of the cerebellum, a structure where inhibitory interneurons are essential regulators of rhythmic synchronization, and in deep brain regions pivotally involved in motor pacing. Our findings shed light into the neural underpinnings of motor timing deficits in NF1.

Keywords: EEG; Inhibition; Motor coordination; Neurofibromatosis type 1; fMRI.

Fig. 1

Experimental design of the behavioural task during the EEG task (concerning the fMRI design, see text). The participants were instructed to make an audio-paced tapping using both index fingers, synchronously or alternately, as indicated by the visual cue. The tapping frequency was set at 1, 3 or 5 Hz, and it was set by a beep

Fig. 2

a Tapping time histograms (relative tapping frequency) for the synchronous condition. b Power at the ideal (cues) tapping frequency for controls and NF1 for the synchronous (S) and alternating (A) conditions. Healthy controls (green) performed better than participants with NF1 (red) at all the conditions, except the 5 Hz condition. The horizontal lines indicate the mean and standard deviation

Fig. 3

Sensitivity and specificity analysis of the power at the expected frequency of tapping. ROC curves were computed for both synchronous and alternating conditions at every frequency of finger tapping (1, 3 and 5 Hz). The best results were found for alternated tapping at 3 Hz (**), which ROC curve showed a sensitivity of 84% and a specificity of 74%, and for the synchronous tapping at 1 Hz (*) showing a sensitivity of 74% and a specificity of 47% to discriminate patients with NF1 from healthy controls

Fig. 4

Time-frequency plots of the control and NF1 groups during synchronous finger tapping at 1 Hz. A similarly strong periodical variation in the beta band is conspicuous in both groups and at the ideal motor tapping frequency in the beta sub-band of 20–26 Hz, centred in the desynchronization beta peak (23 ± 3 Hz) and with a modulation matching behaviour. Note that the colour peaks just reflect maxima and minima positions, and it is the difference that needs to be considered for statistical analysis

Fig. 5

Significant differences between the control group and the NF1 group (t(76) > 3.26, p < 0.05, FDR corrected for multiple comparisons, minimum cluster size of 20 voxels) during performance matched conditions (5 Hz synchronous and alternate)