To Lose or Not to Lose: Developmental Pathways to Regression in Rett Syndrome
Objectives: The objectives of this exploratory study are: (i) to depict atypical neurofunctions in the first two years of life and define potential early behavioural biomarkers pinpointing RTT before the onset of regression; (ii) to reveal developmental trajectories including the acquisition and loss of functions in the following developmental domains: motor development, speech-language and socio-communicative development; and (iii) to develop a vocalisation-based automated detection approach.
Methods: Utilizing a novel pattern recognition approach, we applied a signal-analytical paradigm to assess video footage of the prodromal development of individuals later diagnosed with RTT (intelligent audio analysis on retrospective data; Marschik et al., 2017). Moreover, early motor development was analysed with the standardised General Movement Assessment (GMA; e.g., Einspieler et al., 2017; Novak et al., 2017). Verbal utterances were linguistically captured following established schemes of vocalisation categorisation (e.g., Nathani et al., 2006) and socio-communicative functions are evaluated referring to classifications of the Inventory of Potential Communicative Acts (IPCA; Sigafoos et al., 2006). Data are extracted from our large-scale database GUARDIAN (Pokorny et al., 2016), comprising – besides prospectively collected data – home-video material of children with various neurodevelopmental and genetic disorders (N=118; n=40 with RTT or the preserved speech variant of RTT: PSV; n=52 with autism spectrum disorder: ASD) and typically developing children (TD; total n=33).
Results: Adding to recent findings, we observed atypical neurofunctions in the above-mentioned domains in the pre-regression development of individuals with RTT or PSV: abnormal general movement patterns, a marker for the integrity of the developing nervous system, were observed in all individuals. Atypical pre-linguistic vocalisations were present in 100% of the analysed sample already during the first year of life. As a predecessor of breathing irregularities, atypical inspiratory vocalisations while cooing or babbling were found in 70% of the individuals. A limited repertoire of communicative forms and functions as well as a reduced volubility and complexity of pre-linguistic vocalisations were present in all participants. Intelligent audio analyses automatically differentiating individuals with RTT or ASD versus TD individuals revealed recognition accuracies of 87.9% for RTT and 75% for ASD. Spectral vocalisation parameters, such as Mel-frequency cepstral coefficients, spectral flux, or the Hammarberg index, yielded the highest differentiation effects. Decline of communicative forms and functions and loss of purposeful hand use coincided with an increase of prototypical hand stereotypies. The spontaneous motor repertoire and the productive and receptive verbal abilities in concert point to different developmental disease pathways: (i) developmental milestones are achieved and functions deteriorate or are lost (regression); and (ii) an increase in atypical behaviours and the lack of achievement of expected functions mimic developmental ‘regression’ (pseudo-regression).
Conclusions: Our results are promising towards defining reliable neurofunctional markers for an early (automated) detection of RTT, unravelling within-syndrome traits, and distinguishing developmental trajectories. Our exploratory findings question the prevailing understanding of regression in RTT.