Description
Purpose: Atypical neural connectivity is thought to underlie the complex cognitive and behavioral profile observed in Autism Spectrum Disorders (ASDs). Literature has reported findings of both under- and overconnectivity in individuals with ASDs. A recent meta-analysis spurred a developmental hypothesis, which predicts functional overconnectivity in pre-pubertal children with ASDs, but underconnectivity in adolescents and young adults with ASDs, compared to matched typically developing (TD) individuals. However, relevant empirical findings remain mixed, warranting further study. Design/Methodology/Approach: Independent components analysis and dual regression were applied to 518 participants from the Autism Brain Imaging Database Exchange and in-house data, with ASD and TD groups matched on motion, age, PIQ, sex, and handedness. Correlational analyses of component BOLD signals were used to examine the effects of diagnostic group and age on between- and within network connectivity patterns. Diagnostic group differences were tested using median split age categories. Next, differences were tested for three age categories. Finally, differences were tested with age as a continuous variable. Results: Main effects for diagnostic group or for the interaction between diagnostic group and age were not observed. With age as 3 factors, main effects of age were detected for all but one network connectivities (p(FDR) < 0.05). For age as continuous variable, main effects of age were seen for 19 of 27 network connections supporting that on average large- scale functional networks have non-linear developmental trajectories. Functional connectivity measures indicated that when tracking development from age 6 to 18 years, connectivity trajectories took an inverted-U shape. Conclusion: Trajectories of functional connectivity in resting-state networks of participants between the ages of 6 and 18 years show a non-linear developmental pattern. Adolescents around puberty (10-14 years old) showed increased connectivity compared to participants in early childhood and late adolescence. The current findings are not in agreement with the developmental hypothesis. Overall, when considering differences in neural network connectivity between ASD and TD samples, it is imperative to distinguish the developmental stage and age of the sample, as these parameters may affect network connectivity levels in both ASD and TD individuals.
