# Patterns of ongoing thought in autistic individuals --- ## Introduction #### Autism Spectrum Disorder is expressed in a wide range of characteristics... - sensory processing (Leekam et al., 2007), - executive function (Craig et al., 2016), - attention (Moriuchi, Klin, & Jones, 2016), - social functioning (Baron-Cohen, 2000, Garfinkel et al., 2016, Tager-Flusberg et al., 2005) ---- ##### Lower level sensory characteristics (e.g. sensory sensitivity) alongside differences in higher-order processing (e.g. "theory-of-mind") ##### Although we have a relatively clear understanding of how autistic traits relate to performance on structured neurological tasks (Kenworthy et al., 2008), we know substantially less about how these traits influence patterns of ongoing thought. --- ## Turnbull (2020) ##### Previous study assessed how autistic traits relate to differences in ongoing thought in a neurotypical sample. ##### Found differences in modality of thought related to ASD traits and performance ##### However, investigation needs extending to clinical population. --- ## Aim ##### To investigate the differences in ongoing thought patterns between nonautistic and autistic individuals ##### To investigate how differences in ongoing thoughts relate to working memory task performance --- # Methods --- ## Participants ##### This study recruited a total of 228 participants (N females = 112, N males = 114, N NB = 9) with a mean age of 34 (SD = 13.9). ##### 149 participants (ASD=95, Controls=54, ADIE cohort) completed the 0-back/1-back mind wandering task --- ## Multidimensional Experience Sampling (MDES) ##### To examine the contents of participants thought we used well established technique referred to as multidimensional experience sampling ##### 13 experience sampling questions on 6 probe instances throughout the n-Back task. ----  ---- ## Principle Components Analysis ##### To identify common patterns of thought, we used a dimension reduction technique called PCA on the z-scored ESQ data  ###### _Turnbull et al (2020)_ --- ## Behavioural Task (n-back)  --- ## Task performance ##### Response Time / Accuracy = Inverse Efficiency Score (IES) ##### Averaged for each participant, for each condition (0-back, 1-back) --- ## Linear Mixed Modelling #### How does group membership & condition effect ongoing thought patterns? ###### Thought Pattern Scores ~ Group + Condition + Group:Condition + (1|subject) #### How do (average) ongoing thought patterns, group membership & condition effect task performance? ###### IES ~ Thought Pattern Scores + Group + Condition + Group:Condition + ... (1|subject) ###### Backward fitting fixed effects to find optimal model --- # Results --- ## ESQ Analysis (PCA) #### 1344 experience sampling observations were collected from 156 subjects Each probe featured 13 questions. ----  --- ## Comparing thoughts between groups and condition --- ## Task-Focus ###### Task-Focus ~ Group + Condition + Group:Condition + (1|subject)  ##### - Task-Focus significantly greater in 1-back condition ----  ###### [b = 0.34, 95% CI (0.22, 0.46), t(1173) = −5.51, P <0.001] --- ## Temporally Distant Thought ###### Temporally Distant Thought ~ Group + Condition + Group:Condition + (1|subject)  ##### - Temporally Distant thought significantly reduced in the 1-back condition ----  ###### [b = -0.25, 95% CI (-0.36, -0.14), t(1162.3) = −4.41, P <0.001] --- ## Modality ###### Modality ~ Group + Condition + Group:Condition + (1|subject)  ##### For controls, thinking in words greater in 1-back and reduced in 0-back ##### For ASD participants, no sig. difference in modality of thought between conditions. ----  ##### - Significant effect of Condition ###### [b = -0.15, 95% CI (-0.24, -0.06), t(1180.2) = −3.24, P = 0.0012] ##### - Significant Interaction of Condition:Group ###### [b = -0.20, 95% CI (-0.38, -0.01), t(1189.2) = −2.10, P = 0.036] --- ## Positive Engagement ###### Positive Engagement ~ Group + Condition + Group:Condition + (1|subject)  ##### - Positive engagement significantly increased during 1-back condition ##### - Positive engagement significantly reduced for ASD participants ---- ## Thought component 4 (Emotion)   ##### - Significant effect of Condition ###### [b = 0.23, 95% CI (0.14, 0.32), t(1174) = 4.86, P < .001] ##### - Significant effect of Group ###### [b = 0.45, 95% CI (0.15, 0.74), t(1181) = 3.00, P = 0.003] --- ## Task performance ##### After removing 18 missing values, and 25 outliers, there were 265 IES scores from 143 participants. ---- ### How do ongoing thoughts, condition and group membership relate to performance (IES)? ###### - Fit full model -> Backwards fit fixed effects -> Reduced model ---- ### How do ongoing thoughts, condition and group membership relate to performance (IES)? ##### Model found: ###### IES ~ Condition + Group ###### + Component 2 (Distant) + Component 3 (Modality) + Component 4 (Emotion) ###### + Group:Component 2 + Group:Component 4 ###### + (1|subject) --- ## Task Performance Results  ##### Task performance worse in 1-back ##### Task performance worse in ASD participants ---- ## Task Performance Results  ##### Task performance worse when thinking in images ---- ## Task Performance Results  ##### Task performance better when thinking in more temporally distant manner in ASD participants only ---- ## Task Performance Results  ##### Task performance better with increased positive engagement in ASD participants only --- # Discussion ---- ## Component Extraction #### Patterns of thought consistent with those extracted from a independent sample of neurotypical individuals (n=199, Turnbull et al 2020)   --- ## Differences in thought #### Ongoing thought patterns differed depending on condition AND group ---- ### Effect of condition ##### During 0-back, thoughts were 1) less task focused 2) more temporally distant 3) more in the form of images and 4) less positively engaged. ### Effect of group ##### ASD participants' thoughts were 1) less in the form of images during 0-back, and 2) less positively engaged during both conditions. ---- ### Difference in Modality  ##### Turnbull et al (2020) found autistic traits in neurotypical individuals associated with thinking more in words in a context unrelated manner ##### Our results extend findings of Turnbull et al (2020) to a sample of individuals with ASD diagnoses, demonstrate a broad similarity in results ----  ##### Inconsistent with views that autism is linked to excessive use of imagery (Kunda & Goel, 2011) ##### Consistent with finding that individuals with ASD use verbal strategies to mediate verbal short-term memory (Williams et al, 2012) ##### Generalizable? Result may be specific to use of verbal strategies in working memory tasks ----   ##### Overall, may reflect inflexibility in pattern of cognition to match task-context (Van Eylen et al, 2011) --- ## Differences in Performance ##### Task performance worse in 1-back ##### Task performance worse in ASD participants ---- ## Differences in Performance ##### Task performance worse when thinking in images (especially for ASD & 1-back) ----  ##### Turnbull (2020) also found task performance improved with greater word-based thinking (especially in 1-back). ##### Alterations to the modality of thought reflects a successful strategy for maintaining information working memory. ---- ## Differences in Performance ##### Task performance better when thinking in more temporally distant manner in ASD participants only - opposite for controls ##### Task performance better with increased positive engagement in ASD participants only --- ## In summary... ##### ASD linked to inflexibility in the modality of ongoing thought, as well as a negatively valenced thought pattern throughout the task. ##### Ongoing thoughts in ASD relate to task performance in ways that significantly differ from controls ---- ## Next steps... ##### Investigate patterns of brain activity and connectivity that may mediate these differences in ongoing thought --- # Questions?