Functional Connectivity Uniqueness and Variability? A Signature of Cognitive and Psychiatric Problems in Children

doi:10.21203/rs.3.rs-1514598/v1

Download PDF

Article

Functional Connectivity Uniqueness and Variability? A Signature of Cognitive and Psychiatric Problems in Children

https://doi.org/10.21203/rs.3.rs-1514598/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Brain functional connectivity (FC) derived from functional magnetic resonance imaging (fMRI) has been serving as a potential “fingerprint” for adults. However, intra-subject variation of FC can be substantial and carry biologically meaningful information, especially during adolescence, a time of intense brain changes. Here, for the first time, we performed a large-scale analysis on cross-scan FC stability and its association with a diverse range of health measures in children from the Adolescent Brain Cognitive Development (ABCD) data. Functional network connectivity (FNC), a network analog of FC, was extracted via an automated independent component analysis framework on 9071 subjects (age: 9~11 years) and compared across 4 scans. We found that the FNC profile can identify a given child from a large group with high accuracy (> 80%), though intra-subject variations exist cross-scan. The robustness of this finding is verified by replicating the child identification using the second-year scans and between longitudinal scans with a two-year interval. Moreover, cross-scan FNC stability was predictive of sleep condition, cognitive performance, and psychiatric problems in children, with higher stability correlated with better cognitive performance, longer sleep duration, and less psychotic expression. Meanwhile, parental psychopathology and prenatal exposure were associated with the FNC stability in their children. Overall, our findings show that a child’s connectivity profile is not only intrinsic but also exhibits reliable variability cross-scan, regardless of brain growth and development. The intra-subject connectivity stability may serve as a valuable biomarker to draw inferences on early cognitive and psychiatric behaviors in children.

children

cross-scan functional connectivity stability

cognition

psychopathology

sleep

Functional connectivity (FC) derived from functional magnetic resonance imaging (fMRI) data has shifted neuroimaging focus away from exploring localized brain activity toward characterizing co-activation patterns across distributed brain regions ^1,2. FC has been associated with cognition ³ and various neurological ⁴ and mental ^5,6 disorders. Although the population-based studies have revealed a great deal of knowledge on the FC organization profile, brain FC is assumed to be unique to individuals, regardless of how the brain is engaged during scanning ⁷. FC heterogeneity has long been appreciated in fMRI studies, even within the same population ^8–10. By using a multi-condition fMRI dataset from the Human Connectome Project, studies have shown that FC profile can distinguish adult subjects across scan sessions and even between distinct task conditions ^7,11,12, acting as a “fingerprint”. A recent study has captured five different functional clusters in the pulvinar with specific connectivity fingerprints, associating with distinct components of cognition ¹³. However, an individual’s FC is not constant but continuously changes with remarkable variations at different scales, adapting to internal and external demands ^14–16. That is to say, besides the intrinsic patterns, FC can also exhibit prominent intra-subject variability, which might underly important biological mechanisms ¹⁷. Many existing works have only examined the FC variation within a single scan ^18–20, and an individual’s cross-scan FC variability has received relatively little attention from the neuroimaging field. A comprehensive understanding of cross-scan FC is necessary for developing robust FC-based biomarkers.

Unlike adults, children show considerable growth and development in the brain ²¹ and might exhibit more temporal variability in brain FC ^22,23. The temporal variation in resting-state FC declines with aging, especially for those FC involved in frontal-parietal and occipital networks ²². Age is negatively associated with the variability of dynamic brain connectivity states captured by a clustering strategy on dynamic FC estimates ²³. The prominent variation in youths’ FC might be due to more neuroplasticity in the adolescent’s brain ^24,25 or be induced by larger motion artifacts ²⁶ of the adolescent subjects, though this is still far from understood. On the other hand, some other studies assume that FC is unique to an adolescent and show that FC profile can successfully identify an adolescent from a group of subjects ^25,27.

Despite such progress, we argue that the exploration of cross-scan FC in youths has been limited, as most studies have used relatively small numbers of participants (e.g., sample size < 1000). In addition, many prior studies failed to examine the robustness of FC results across multiple scans and between longitudinal sessions. More importantly, existing work has only focused on the fingerprint property of FC across scans, ignoring the intra-subject FC variation cross-scan and its potential relationship with individual behavior in children. The stability of FC across scans can carry meaningfully biological mechanisms, reflected by its relevance to adolescents’ neurocognitive development, adverse mental problem outcomes, and other healthy backgrounds. Therefore, there is the need for a reliable large-scale study as the present one that can include more than 9000 subjects with multiple scans collected from longitudinal sessions to comprehensively examine the relevance of the cross-scan FC stability to individual differences in behavior and mental conditions.

Here, we investigate the cross-scan FC stability in children using a multimodal database called Adolescent Brain Cognitive Development (ABCD). The ABCD database includes more than 11,800 subjects, with multiple scans collected from two longitudinal sessions. This dataset has collected a comprehensive range of measures related to mental problems, cognitions, and other health backgrounds ²⁸ that have been shown useful for the investigation of the relationship between adolescent behaviors and brain functions ²⁹. Our hypothesis is twofold. First, the FC profile of children shows intrinsic patterns across scans, and the intrinsic patterns can identify a given child from a large group of subjects. We also expected this identification would be successful between longitudinal sessions, regardless of the developments and changes in the children’s brains. Our second hypothesis is that children’s FC also shows intra-subject variations across scans, which will predict distinct behavioral phenotypes in children, such as cognitive performance, psychiatric problems, and sleep conditions.

Flowchart of the Cross-scan Functional Connectivity Analysis

Fig. 1 displays the flowchart of the cross-scan functional network connectivity (FNC) analysis. We first applied a Neurmark framework to extract robust intrinsic connectivity networks (ICNs) that are comparable across subjects, scans, and sessions. FNC was estimated using the time-courses (TCs) of ICNs from each scan. After obtaining the FNC matrix of each scan, cross-scan FNC similarity was measured by the correlation between FNC from different scans. Then individual identification was performed based on the cross-scan FNC similarity. Finally, we examined the associations between intra-subject FNC similarity and individuals' behaviors via a linear mixed-effect model (LMM).

Functional Networks

53 ICNs were extracted by the Neurromark framework, with activation peaks falling on the majority of cortical and subcortical gray matter areas across the whole brain. The ICNs were assigned to seven different functional domains according to their anatomical locations and functional information ¹⁴, including subcortical (SC), auditory (AUD), visual (VS), sensorimotor (SM), cognitive-control (CC), default-mode (DM), and cerebellar (CB) domains. Details of the spatial maps and coordinates of ICNs are provided in the supplementary materials.

Intra-subject FNC Shows High Similarity across Scans

Fig. 2a displays the FNC of subjects with maximum and minimum intra-subject FNC similarities between scans. Children show different levels of cross-scan FNC similarity. For subject INVX2TWFHMP, the FNC of scan 1 and the FNC of scan 2 share the highest intra-subject similarity (r = 0.9448). In contrast, for subject INVZF82Z2N1, the FNC show more variability between scan 1 and scan 2, where the intra-subject FNC similarity is only r = 0.1914. Fig. 2b displays the percentage of children with intra-subject FNC similarity larger than a given percentage of inter-subject FNC similarity, from 60% to 99%. Our results show that intra-subject FNC similarity is larger than a majority of inter-subject FNC similarity, though intra-subject FNC variability exists. The FNC shows the highest intra-subject similarity between scan 1 and scan 2. More than 90% of subjects have intra-subject FNC similarity larger than 60% of inter-subject FNC similarity and more than 65% of subjects have intra-subject FNC similarity larger than 99% of inter-subject FNC similarity. The intra-subject FNC shows the lowest similarity between scan 1 and scan 4. Still, more than 80% of subjects have intra-subject FNC similarity larger than 60% of inter-subject FNC similarity, and about 40% of subjects have intra-subject FNC similarity larger than 99% of inter-subject FNC similarity.

The results are replicated by examining the scans from the second-year session. Similarly, subjects can have different levels of cross-scan FNC similarity. FNC shows the highest intra-subject similarity between scan 1 and scan 2 and the lowest intra-subject similarity between scan 1 and scan 4. FNC also shows intra-subject similarities between longitudinal scans. Although a two-year time interval between scans incurred a significant decrease in intra-subject similarity, the intra-subject similarity is still larger than the majority of inter-subject FNC similarity, especially when the FNC was averaged within the session.

Individual Identification using Whole-brain FNC

Fig. 3 shows the identification results of each pair of identification. At the baseline session, the identification accuracy was 93.99%, 84.78%, 81.87%, and 93.10% based on the database-target of scan 1-scan 2, the target-database of scan 1-scan 3, the target-database of scan 1-scan 4, and the target-database of scan 1-scan mean respectively. The identification was replicated by using the FNC of the second-year scans. Similar to the results from the baseline, the highest identification accuracy 95.16% was achieved based on the database-target of scan 1-scan 2, while the lowest identification accuracy 82.80% was achieved based on the database-target of scan 1-scan 4.

The individual identification was further performed using the FNC of between longitudinal scans. Scans from the baseline session were the database and scans from the second-year session were the target. Although more intra-subject FNC variations were introduced, the FNC of a child from the baseline session can still be used to identify his/her FNC from the second-year follow-up session. The highest accuracy was 91.43%, which was achieved by averaging the FNC across all four scans within each session before identification.

The nonparametric permutation testing shows that the average identification accuracy was 50% if the identity was shuffled for each scan. The real identification accuracy was significantly higher than the accuracy obtained by the permutation tests.

FNC Stability Predicts Cognitive Performance

The cognitive measures were positively correlated with the intra-subject FNC stability (False discovery rate [FDR] corrected, q < 0.05). Specifically, 10 out of 10 of the cognitive summary scores were positively correlated with FNC stability, with correlation r values ranging from 0.0376 to 0.1070. The Total Composite Score was the score most significantly positively correlated with the FNC stability (r = 0.1070, Cohen’s d = 0.2152, p = 4.82×10^-24). For the neurocognitive battery in the subdomain, TPVT was the score most significantly positively correlated with the FNC stability (r = 0.0841, Cohen’s d = 0.1688, p = 1.54×10^-15) while TFT was the score least significantly positively correlated with intra-subject FNC stability (r = 0.0376, Cohen’s d = 0.0753, p = 3.68×10^-4). To better visualize the associations, we divided the children into four groups from low cognitive performance to high cognitive performance according to each cognitive score (group 1: 0%~25%, group 2: 25%~50%; group 3: 50%~75%, and group 4: 75%~100%) and the averaged cross-scan FNC stability within each group is displayed using bar plots in Fig. 4. Clear increasing trends can be observed along group 1 to group 4, indicating that children with good cognitive performance tended to have higher FNC stability.

FNC Stability Predicts Psychiatric Problems

The psychopathological measures of children were negatively correlated with the intra-subject FNC stability. 12 out of 20 psychiatric problem scores show significantly negative correlations with FNC stability, with r values ranging from -0.0257 to -0.0496 (FDR corrected, q < 0.05). The social problem score was the score most significantly negatively correlated with the FNC stability (r = -0.0496, Cohen’s d = -0.0992, p = 2.38×10-6). Similarly, we divided the children into four groups according to each psychopathological measure. The mean and the standard error of the mean for the cross-scan FNC stability of each group were displayed in Fig. 4. The FNC stability show decreasing trends along group 1 to group 4, indicating that children with high psychiatric problem scores tended to have lower FNC stability.

FNC Stability Correlates with Sleep Conditions and Screen Usage

We further found significant associations between FNC stability and the sleep conditions of children. The cross-scan FNC stability was negatively correlated with the sleep duration score (r = -0.0752, Cohen’s d = -0.1508, p = 7.74×10^-13). In the ABCD measurement system, high sleep duration score indicates short sleep duration (1 = 9-11 hours; 2 = 8-9 hours; 3 = 7-8 hours; 4 = 5-7 hours; 5 = Less than 5 hours). The FNC stability was also negatively correlated with the score that evaluates how long an adolescent falls asleep (sleepdisturb2_p). A higher score in sleepdisturb2_p indicates a longer time to fall asleep. The FNC stability was negatively correlated with other sleep behaviors of adolescents, such as sleepdisturb24_p (evaluates a child feels unable to move when waking up in the morning) and sleepdisturb26_p (evaluates a child falls asleep suddenly in inappropriate situations). Higher scores in these measurements indicate more frequently that the event happens (1 = Never; 2 = Occasionally (once or twice per month or less); 3 = Sometimes (once or twice per week); 4 = Often (3 or 5 times per week); 5 = Always). The overall results indicate that children with worse sleep conditions (e.g., shorter sleep duration or longer time to fall asleep) tended to have lower FNC stability.

Children’s screen usage is also negatively correlated with cross-scan FNC stability. 14 out of 14 youth screen time utilization scores, including the use of television, internet, cell phone, and video games, show negative correlations with individuals’ FNC stability (FDR corrected, q<0.05). Children with more screen usage tended to have lower FNC stability. Details of the correlations statistics can be found in the supplementary materials.

FNC Stability Correlates with Parent Psychopathology and Prenatal Exposure

Moreover, parental dimensional psychopathology showed significant correlations with their children’s FNC stability (Fig. 5). Specifically, the positive questions in the parents’ psychopathology assessment, including asr_q15_p (I am pretty honest), asr_q73_p (I meet my responsibilities to my family), asr_q88_p (I enjoy being with people), asr_q98_p (I like to help others), asr_q106_p (I try to be fair to others), asr_q123_p (I am a happy person), were positively correlated with the FNC stability of children with r values ranging from 0.0315 to 0.0583 (FDR corrected, q < 0.05). In contrast, the negative questions in the parents’ psychopathology assessment were negatively correlated with the FNC stability of children with r values ranging from -0.0287 to -0.0482 (FDR corrected, q < 0.05). These results indicate that parents with positive behaviors will result in higher FNC stability in children while parents with negative behaviors will result in lower FNC stability in children.

Our analysis also showed that prenatal exposure before and during pregnancy was associated with FNC stability in children. Parents with prenatal exposure to tobacco and marijuana will result in lower FNC stability in children (FDR corrected, q < 0.05). Also, a planned pregnancy will result in higher FNC stability in children. The age of the parents during the pregnancy showed significant correlations with FNC stability as well. While older mothers will result in higher FNC stability in children, fathers’ ages between 30~40 years old (when the child was born) result in the highest FNC stability in adolescents.

The Fingerprint Property of Adolescent’s FNC

Brain FC and its network analog, FNC, are believed to provide a window into brain function and intrinsic brain organization ^2,40–42. Neuroimaging studies have successfully established that adults’ FC profile shows substantial inter-subject variability, and such variability can distinguish individuals from another scan ^7,12. Unlike adults, children might show more intra-subject variability in FC due to the developments in the brain ⁴³. More heterogeneous brain states and confounding effects in youths (e.g., head motions) ^44,45 might also influence the individual identifications. Recent research has found that the identification of youths is not significantly different from that of adults ²⁵. However, this study relied on adolescents with a large age range (12~18 years). Brain developments at different ages can be different, potentially introducing unpredictable variability in the identification. Also, the results based on such a small sample size (N = 140) can be biased by the sampling variability, further resulting in replication failures ⁴⁶. Our study used large data from the ABCD project with almost the same age to show that the cross-scan FNC similarity can be robustly observed in children. We noted that the identification accuracy decreased and the intra-subject FNC variability increased as the time interval of scans increased. These results indicate that the assumption of the FNC uniqueness might oversimplify the interrelationships between brain regions. Accumulated evidence has shown that brain connectivity is highly dynamic with dramatic variations across scans, tasks, and time ^47–49. An individual’s FC will show both similarity and variability across scans, depending on the brain state in which the FC is measured ¹⁷. Our result provides the first evidence that single-subject FC will exhibit cross-scan variability that might be indicative of brain state during imaging. In other words, as a subject stayed in the scanner for a longer time, their brain state might be far away from the initial “resting-state”, resulting in a larger difference between the brain FNC.

We successfully performed the individual identification between longitudinal scans with a two-year interval, although with reduced overall accuracy. This finding is in line with a previous result based on a relatively small sample size, which suggested that a larger time interval can incur a significant decrease in identification accuracy ²⁵. We further found that averaging FNC across scans within each session can increase the accuracy. Growth and development inside the children’s brain will introduce FC variation intra-subject, which is associated with children’s neurodevelopment and behaviors ^50,51. Averaging FNC within the same session can mitigate the heterogeneity induced by transient brain states, but not the variability induced by brain development. Our result suggests that the decreased identification accuracy between longitudinal scans can be due to both brain development and the difference in the temporal brain conditions. The successfully longitudinal identification further supports that the FC profile contains fundamental properties that are unique to each child, regardless of the FC developments during adolescence ²⁵.

Individual FNC Stability Predict Adolescent’s Behaviors

Besides the intra-subject similarity, children’s FNC also showed intra-subject FNC variability across scans. The balance between cross-scan FNC similarity and variability does not appear to be driven by random noise but captures physiological and psychological information. The cross-scan FNC stability was positively correlated with cognitive performance, including reading recognition, pattern comparison, memory, etc. Previously, neuroimaging studies have been typically focusing on the FC strength and suggesting that its inter-subject variability is relevant to individual differences in behaviors ^3,7,52. However, brain FC is not static, but with considerable variation between tasks and rest, across scans, and even within a single scan ^17,49,53. Spontaneous FC variations can predict the performance of different cognitive tasks ^54,55. Literature also showed that individuals with stable FC across-time show advanced cognitive performance, reflected by the increased accuracy and the more stable response time ^56,57. Our finding has extended the investigation of FC variability within a single scan to the investigation of FC stability across scans and showed robust relationships between cross-scan FC stability and cognitions. A possible explanation of this finding is that the resting-state is in a “relaxed” brain condition that ameliorates adaptive reconfiguration of brain networks in the context of cognitive tasks. A stable FC during the execution of cognitive tasks is associated with successful cognition and more difficult task conditions require increased stability of FC ^56,57. The stable FC during the resting-state might facilitate the brain switching from a relaxed condition to a task demand condition that purportedly requires sustained cognition, consequently resulting in better cognitive performance ⁵⁸.

In addition to the correlations with cognitions, we found negative correlations between FNC stability and the dimensional psychopathology in children. Children with less cross-scan FNC stability will have more severe dimensional psychopathological problems and more frequent mania episodes. One interpretation of this finding is that decreased FC stability underlies the dysregulated brain rhythms that characterize psychiatric problems. Of note, increased rumination is associated with higher medial prefrontal cortex to insula FC variability, suggesting that the intra-subject FC heterogeneity might trigger rumination by enhancing sensitivity to self-referential information ⁵⁹. It is also suggested that the unstable FC may be associated with deficits in executive functioning and reflect weaknesses in brain circuits responsible for cognitive control ⁶⁰. A growing body of literature has linked dynamic FC patterns to psychiatric problems. Individuals with autism spectrum disorder have larger FC variability in time, associated with the increase depending on autism symptom severity ⁶¹. By using magnetoencephalography, researchers have shown that schizophrenia patients exhibit more trials-to-trials network topology variability during a 2-back working memory task ⁶². Increased FC variability has also been observed in both patients with depressed bipolar disorder and major depressive disorder, who shared overlapping symptoms that typically confound the diagnosis ⁶³. Our results complement the prior work by showing that an individual’s FNC exhibits reliable variability across scans, which might signify underlying biological mechanisms in mental health. The stability of cross-scan FNC can add information to the connectivity strength and will be a potential brain biomarker that predicts early psychiatric problems in children.

Interestingly, our analysis further showed that the FNC stability of children is also relevant to the parental conditions. The development history of children can be an important indicator of later mental and psychical behaviors in youths ²⁸. The increase of drug use among pregnant mothers has become a social problem in the US ⁶⁴. Prenatal cannabis exposure is associated with a greater risk for psychopathology in adolescents ⁶⁵. Here in this study, we found that prenatal tobacco and marijuana exposure is associated with lower cross-scan FNC stability during middle childhood. Considering the associations between FNC stability and cognition and dimensional psychopathology, this result underscores the potential to use FNC stability to advance our understanding of the relationships among prenatal drug usage and cognitive developments and mental health among offspring. Another interesting finding of our present study is that the parents’ psychopathology was correlated with the FNC stability of children, where the positive behaviors in parents were associated with higher FNC stability in their children while the negative behaviors in parents were associated with lower FNC stability in their children. We speculate that the family environment might influence the stability of FC in children. This speculation is supported by a further analysis showing a positive correlation between children’s FNC stability and neighborhood safety, an important living environmental factor. The inherited characteristic can be another cause of these relationships. Analysis including genetic data is needed in future studies for validating this hypothesis.

In sum, our study revealed that children’s FNC shows certain similarity and substantial variability across scans from the same day and longitudinal sessions. The intrinsic patterns of FNC successfully identify each child from a large cohort of 9071 subjects. Besides, cross-scan FNC stability carries important psychological and physiological information underlying distinct behavioral phenotypes in children. To the best of our knowledge, this is the first attempt to investigate the cross-scan FNC on children at a large-scale. Given this foundation, future neuroimaging studies should move beyond the focus of the connectivity uniqueness to an individual, but to investigate the cross-scan connectivity variability, probing how intra-subject stability and variability of an individual’s functional brain organization provide a window into neuropsychological mechanisms.

Participants and behavioral assessments

The present study used a longitudinal dataset shared by the ABCD, the largest long-term study of brain development and child health in the United States (https://abcdstudy.org/). We used the release 2.01 of the ABCD dataset, containing over 11,800 children aged 9-11 years, with two imaging sessions (baseline and the second-year) and multiple scans within each session. The ABCD study incorporated a comprehensive range of measures, including neurocognitive battery, physical and mental health assessments, socioeconomic, ethnic, and other health backgrounds, to assess predictors and outcomes related to different domains ^28,29. The parent’s full written informed consent and the child’s assent were obtained under protocols approved by the Institutional Review Board. We performed data quality control (QC) on the preprocessed fMRI data via the Neuromark framework ³⁰. The Neuromark QC yield 9071 subjects for the baseline analysis, 2918 subjects for the second-year analysis, and 2290 subjects for the cross-sessional analysis; each analysis has at least four good resting-state fMRI scans within each session. The basic demographics can be found in Table I.

Table I. Basic Demographics of Subjects

Basic Demographics	Baseline	Second-year
Total Subject	9071	2918
Age (month)	119.06 ± 7.52	142.86 ± 7.51
Gender (F/M)	4365/4706	1333/1585
Height (inch)	55.28 ± 3.35	60.04 ± 3.53
Weight (lbs)	82.97 ± 23.52	107.57 ± 31.99
Race (W/B/H/A/O)	4771/1325/1863/181/929	1631/317/629/57
Cognition (nihtbx_totalcomp)	86.44 ± 8.99	90.94 ± 7.11
Psychiatric problem (cbcl_scr_syn_totprob)	45.73 ± 11.35	44.78 ± 11.27
Sleep disturbance (sleepdisturb1_p)	1.72 ± 0.81	1.99 ± 0.86

Neuromark Framework

To capture reliable intrinsic connectivity networks (ICNs) and their corresponding TCs for each subject and each scan, a robust independent component analysis (ICA)-based framework called Neuromark ³⁰ was applied to the ABCD data. Unlike atlas-based methods that typically assume fixed brain regions across subjects, the Neuromark can identify brain networks comparable across subjects while adapting single-subject variability with the networks. The effectiveness of Neuromark has been demonstrated in previous work, with a wide range of brain markers and abnormalities identified in different populations ^31–36. More details of the Neuromark framework are provided in ³⁰ and the supplementary materials.

Functional Network Connectivity

Children’s data can be noisy with more confounding effects, such as larger head motions. We therefore performed four additional post-processing steps to carefully regress out the remaining noise in the TCs of ICNs: 1) detrending linear, quadratic, and cubic trends, 2) removal of detected outliers, 3) multiple regression of the head motions parameters (3 rotations and 3 translations) and their derivatives, 4) band-pass filtering with a cutoff frequency of 0.01 Hz-0.15 Hz. Pearson correlation coefficients between post-processed TCs were calculated to measure the FNC for each scan.

FNC Similarity and Subject Identification

We calculated the correlation between whole-brain FNC from different scans to measure the FNC similarity. Specifically, the correlation between FNC of scans from the same subject was defined as the intra-subject FNC similarity while the correlation between FNC of scans from different subjects was defined as the inter-subject FNC similarity. For each subject, there are one intra-subject similarity and 9070 inter-subject similarities. Then we calculated the percentage of children who have intra-subject FNC similarity larger than a given percentage of inter-subject FNC similarities (60%, 70%, 80%, 90%, 95%, and 99%).

We performed individual identification using the FNC similarity. For each subject, we compared his/her intra-subject FNC similarity with a randomly picked inter-subject FNC similarity. The predicted identity was that with the larger correlation value. We performed this step for every subject to obtain an identification vector, which can be used to calculate the overall identification accuracy. The whole procedure was repeated 1000 times to estimate the distribution of the identification accuracy. The identification was also performed using domain-based FNC, and results are provided in the supplementary materials.

We further performed nonparametric permutation testing to assess the statistical significance of identification accuracy. We permuted subject identity for the FNC of scans to shuffle the intra-subject and inter-subject FNC similarity. The same identification was performed 1000 times on the permuted data to have the identification accuracy for the permuted data.

Test-retest Reliability and Longitudinal Identification

We further calculated the FNC similarity between other scans (between scan 1 and scan 3, between scan 1 and scan 4, and between scan 1 and scan mean [mean FNC across scans 2~4]) from the baseline session. We also calculated the FNC similarity between scans from the second-year session. The same identification was performed using the FNC similarity between scans. To investigate whether the FNC profile can identify an individual from a longitudinal scan while there are developmental changes in the brain, we measured the FNC similarity between a scan from the baseline session and a scan from the second-year session and then performed the identification based on the FNC similarity between longitudinal scans.

Cognitive Measures

The cross-scan FNC stability was measured by the intra-subject similarity between FNC of scan 1 and mean FNC across scans 2~4. To show the test-retest reliability of the associations, we also replicated the results using different FNC stability measures and using the second-year data. Details of the replication can be found in the supplementary materials.

We first investigate the associations between FNC stability and cognitive assessments. The cognitive performance of each adolescent was measured via the NIH Cognition Battery Toolbox (abcd_tbss01) ²⁹. Higher scores indicate better cognitive performance. The NIH neurocognitive battery contains 7 distributional characteristics, including the Toolbox Picture Vocabulary Task (TPVT), the Toolbox Oral Reading Recognition Task (TORRT), the Toolbox Pattern Comparison Processing Speed Test (TPCPST), the Toolbox List Sorting Working Memory Test (TLSWMT), the Toolbox Picture Sequence Memory Test (TPSMT), the Toolbox Flanker Task (TFT), and the Toolbox Dimensional Change Card Sort Task (TDCCS). There are also 3 composite scores, including a Crystalized Intelligence Composite and a Fluid Intelligence Composite, and a Total Score Composite. In total 10 cognitive scores were used in the analysis. The detailed information of each score can be found in ²⁹.

Mental Problem Measures

The associations between cross-scan FNC stability and children’s mental health conditions were also investigated. The mental health conditions of children were measured by the Parent Child Behavior Checklist Scores (abcd_cbcls01). These checklist scores contain 11 syndrome scales related to psychiatric problems and 1 total Syndrome Scale, 6 DSM-Oriented scales, and 3 CBCL Scale2007 Scales. In total 20 scores from the Parent Child Behavior Checklist Scores were used for the investigation. The ABCD Parent General Behavior Inventory-Mania (abcd_pgbi01) was also used to assess the subsyndromal mania. It contains 10 scores that evaluate the children’s behaviors of mania. Higher scores of the mental health measures indicate dimensional psychopathology.

Sleep Conditions and Screen Usage

The assessments of sleep conditions and sleep disorders of children were measured by ABCD Parent Sleep Disturbance Scale for Children (abcd_sds01). It includes 26 questionnaires to evaluate the sleep disturbance of each child. For example, question 1 is “How many hours of sleep does your child get on most nights?” and question 2 is “How long after going to bed does your child usually fall asleep?”. The scores will be between 1 to 5, with a higher score indicating a worse sleep condition (e.g., fewer hours of sleep and longer time to fall asleep). The screen time utilization of youth, which is measured by the ABCD Youth Screen Time Survey (abcd_stq01) was also used to investigate its relationships with cross-scan FNC stability. It contains 14 scores that evaluate the screen usage of a child during the weekdays and weekends, with higher scores indicating longer screen usage.

Parental Behaviors and Prenatal Exposure

We were also interested in the potential relationships between parental factors and the children’s FNC stability. The prenatal exposure before and during pregnancy, measured by ABCD Developmental History Questionnaire (dhx01) was used for the investigation. We focused on prenatal exposure to tobacco, alcohol, and marijuana, and the parents’ age when the child was born. The parent dimensional psychopathology, measured by ABCD Parent Adult Self Report Raw Scores Aseba (pasr01), was also used in the analysis. These scores evaluate parent dimensional psychopathology from either positive question (e.g., question q15: I am pretty honest) or negative question (e.g., question q12: I feel lonely). The higher scores in the positive question indicate better condition of parents while the higher scores in the negative question indicate worse dimensional psychopathological condition.

Association between Cross-scan FNC Stability and Behaviors

An LMM was adopted to investigate the associations between intra-subject cross-scan FNC stability and the behavioral assessments. The LMM was also used to examine the associations between the children’s FNC stability and their parents’ conditions and neighborhood safety. The ABCD data contain related data at sites and within families due to twins and siblings. The LMM can model families nested within the site to take account of this effect. It has been successfully applied in previous ABCD studies and identified meaningful brain-wide associations with a wide range of individual behaviors ^37,38. In this work, cross-scan FNC stability was modeled as the dependent variable, while each score/behavior was modeled as a fixed effect. Age, gender, race, height, and weight that were considered as confounding effects were modeled as other fixed effects. The family structures and sites were modeled as random effects ³⁷. The correlation r-value, t-statistic, and effect size Cohen’s d was obtained for each association analysis to reflect the relationship between FNC stability and a behavioral score. The results are corrected by false discovery rate (FDR) correction ³⁹.

ABCD = Adolescent Brain Cognitive Development; AUD = auditory domain; CB = cerebellum domain; CC = cognitive control domain; DM = default-mode domain; FC = functional connectivity; FDR = False discovery rate; FNC = functional network connectivity; fMRI = functional magnetic resonance imaging; ICA = independent component analysis; ICN = intrinsic connectivity network; LMM = linear mixed-effect model; QC = quality control; SC = subcortical domain; SM = sensorimotor domain; TC = time-course; TDCCS = Toolbox Dimensional Change Card Sort Task; TFT = Toolbox Flanker Task; TLSWMT = Toolbox List Sorting Working Memory Test; TORRT = Toolbox Oral Reading Recognition Task; TPCPST = Toolbox Pattern Comparison Processing Speed Test; TPSMT = Toolbox Picture Sequence Memory Test; TPVT = Toolbox Picture Vocabulary Task; VS = visual domain

Data Availability

The code of Neuromark framework and the Neuromark template have been released and integrated in the group ICA Toolbox (GIFT, https://trendscenter.org/software/gift/), which can be downloaded and used directly by users worldwide. The ABCD data used in the present study can be accessed upon application from NDA. Other MATLAB codes of this study can be obtained from the corresponding author with reasonable request.

Acknowledgements

This work was supported by National Institutes of Health (No. R01DA049238 and R01MH118695, R01MH117107), National Science Foundation (No. 2112455). Use of the ABCD (https://abcdstudy.org/) dataset is acknowledged. A full list of supporters of ABCD project is available at https://abcdstudy.org/nih-collaborators.

Author Contributions

Zening Fu and Vince Calhoun designed the study; Zening Fu performed the data analysis; Zening Fu, Jingyu Liu, Jing Sui, and Vince Calhoun wrote the paper. Mustafa Salman helped with data preprocessing. All authors contributed to the results interpretation and discussion.

Competing Interests

The authors declare no competing interests.

1. Biswal B, Zerrin Yetkin F, Haughton VM, Hyde JS. Functional connectivity in the motor cortex of resting human brain using echo‐planar mri. Magn Reson Med. 1995;34(4):537-541. doi:10.1002/mrm.1910340409

2. Biswal BB, Mennes M, Zuo XN, et al. Toward discovery science of human brain function. Proc Natl Acad Sci U S A. 2010;107(10):4734-4739. doi:10.1073/pnas.0911855107

3. Park HJ, Friston K. Structural and functional brain networks: From connections to cognition. Science (80- ). 2013;342(6158). doi:10.1126/science.1238411

4. Greicius M. Resting-State Functional Connectivity in Neuropsychiatric Disorders. Vol 21.; 2008. doi:10.1001/jamapsychiatry.2015.0484

5. Arbabshirani MR, Plis S, Sui J, Calhoun VD. Single subject prediction of brain disorders in neuroimaging: Promises and pitfalls. Neuroimage. 2017;145:137-165. doi:10.1016/j.neuroimage.2016.02.079

6. Arbabshirani MR, Kiehl KA, Pearlson GD, Calhoun VD. Classification of schizophrenia patients based on resting-state functional network connectivity. Front Neurosci. 2013;(7 JUL). doi:10.3389/fnins.2013.00133

7. Finn ES, Shen X, Scheinost D, et al. Functional connectome fingerprinting: Identifying individuals using patterns of brain connectivity. Nat Neurosci. 2015;18(11):1664-1671. doi:10.1038/nn.4135

8. Cribben I, Wager TD, Lindquist MA. Detecting functional connectivity change points for single-subject fMRI data. Front Comput Neurosci. 2013;(OCT). doi:10.3389/fncom.2013.00143

9. Gordon EM, Nelson SM. Three types of individual variation in brain networks revealed by single-subject functional connectivity analyses. Curr Opin Behav Sci. 2021;40:79-86. doi:10.1016/j.cobeha.2021.02.014

10. Anderson JS, Ferguson MA, Lopez-Larson M, Yurgelun-Todd D. Reproducibility of single-subject functional connectivity measurements. Am J Neuroradiol. 2011;32(3):548-555. doi:10.3174/ajnr.A2330

11. Li K, Wisner K, Atluri G. Feature selection framework for functional connectome fingerprinting. Hum Brain Mapp. 2021;42(12):3717-3732. doi:10.1002/hbm.25379

12. Sarar G, Rao B, Liu T. Functional connectome fingerprinting using shallow feedforward neural networks. Proc Natl Acad Sci U S A. 2021;118(15):2021. doi:10.1073/pnas.2021852118

13. Guedj C, Vuilleumier P. Functional connectivity fingerprints of the human pulvinar: Decoding its role in cognition. Neuroimage. 2020;221:117162. doi:10.1016/j.neuroimage.2020.117162

14. Allen EA, Damaraju E, Plis SM, Erhardt EB, Eichele T, Calhoun VD. Tracking whole-brain connectivity dynamics in the resting state. Cereb Cortex. 2014;24(3):663-676. doi:10.1093/cercor/bhs352

15. Buzsáki G. Rhythms of the Brain.; 2009. doi:10.1093/acprof:oso/9780195301069.001.0001

16. van den Heuvel MP, Hulshoff Pol HE. Exploring the brain network: A review on resting-state fMRI functional connectivity. Eur Neuropsychopharmacol. 2010;20(8):519-534. doi:10.1016/j.euroneuro.2010.03.008

17. Finn ES, Scheinost D, Finn DM, Shen X, Papademetris X, Constable RT. Can brain state be manipulated to emphasize individual differences in functional connectivity? Neuroimage. 2017;160:140-151. doi:10.1016/j.neuroimage.2017.03.064

18. Chang C, Glover GH. Time-frequency dynamics of resting-state brain connectivity measured with fMRI. Neuroimage. 2010;50(1):81-98. doi:10.1016/j.neuroimage.2009.12.011

19. Zalesky A, Fornito A, Cocchi L, Gollo LL, Breakspear M. Time-resolved resting-state brain networks. Proc Natl Acad Sci U S A. 2014;111(28):10341-10346. doi:10.1073/pnas.1400181111

20. Fu Z, Du Y, Calhoun VD. The Dynamic Functional Network Connectivity Analysis Framework. Engineering. 2019;5(2):190-193. doi:10.1016/j.eng.2018.10.001

21. Johnson SB, Blum RW, Giedd JN. Adolescent Maturity and the Brain: The Promise and Pitfalls of Neuroscience Research in Adolescent Health Policy. J Adolesc Heal. 2009;45(3):216-221. doi:10.1016/j.jadohealth.2009.05.016

22. Chen Y, Wang W, Zhao X, et al. Age-related decline in the variation of dynamic functional connectivity: A resting state analysis. Front Aging Neurosci. 2017;9(JUN):203. doi:10.3389/fnagi.2017.00203

23. Xia Y, Chen Q, Shi L, et al. Tracking the dynamic functional connectivity structure of the human brain across the adult lifespan. Hum Brain Mapp. 2019;40(3):717-728. doi:10.1002/hbm.24385

24. Jalbrzikowski M, Larsen B, Hallquist MN, Foran W, Calabro F, Luna B. Development of White Matter Microstructure and Intrinsic Functional Connectivity Between the Amygdala and Ventromedial Prefrontal Cortex: Associations With Anxiety and Depression. Biol Psychiatry. 2017;82(7):511-521. doi:10.1016/j.biopsych.2017.01.008

25. Jalbrzikowski M, Liu F, Foran W, et al. Functional connectome fingerprinting accuracy in youths and adults is similar when examined on the same day and 1.5-years apart. Hum Brain Mapp. 2020;41(15):4187-4199. doi:10.1002/hbm.25118

26. Power JD, Cohen AL, Nelson SM, et al. Functional Network Organization of the Human Brain. Neuron. 2011;72(4):665-678. doi:10.1016/j.neuron.2011.09.006

27. Demeter D V., Engelhardt LE, Mallett R, et al. Functional Connectivity Fingerprints at Rest Are Similar across Youths and Adults and Vary with Genetic Similarity. iScience. 2020;23(1):100801. doi:10.1016/j.isci.2019.100801

28. Barch DM, Albaugh MD, Avenevoli S, et al. Demographic, physical and mental health assessments in the adolescent brain and cognitive development study: Rationale and description. Dev Cogn Neurosci. 2018;32:55-66. doi:10.1016/j.dcn.2017.10.010

29. Luciana M, Bjork JM, Nagel BJ, et al. Adolescent neurocognitive development and impacts of substance use: Overview of the adolescent brain cognitive development (ABCD) baseline neurocognition battery. Dev Cogn Neurosci. 2018;32:67-79. doi:10.1016/j.dcn.2018.02.006

30. Du Y, Fu Z, Sui J, et al. NeuroMark: an automated and adaptive ICA based pipeline to identify reproducible fMRI markers of brain disorders. NeuroImage Clin. Published online August 11, 2020:102375. doi:https://doi.org/10.1016/j.nicl.2020.102375

31. Fu Z, Tu Y, Calhoun VD, et al. Dynamic functional network connectivity associated with post-traumatic stress symptoms in COVID-19 survivors. Neurobiol Stress. 2021;15. doi:10.1016/j.ynstr.2021.100377

32. Tu Y, Fu Z, Mao C, et al. Distinct thalamocortical network dynamics are associated with the pathophysiology of chronic low back pain. Nat Commun. 2020;11(1). doi:10.1038/s41467-020-17788-z

33. Fu Z, Iraji A, Turner JA, et al. Dynamic state with covarying brain activity-connectivity: On the pathophysiology of schizophrenia. Neuroimage. 2021;224:117385. doi:10.1016/j.neuroimage.2020.117385

34. LI kaicheng, Fu Z, Luo X, et al. The influence of cerebral small vessel disease on static and dynamic functional network connectivity in subjects along Alzheimer’s disease continuum. Brain Connect. Published online November 16, 2020. doi:10.1089/brain.2020.0819

35. Fu Z, Sui J, Espinoza R, et al. Whole-brain Functional Connectivity Dynamics associated with Electroconvulsive Therapy Treatment Response. Biol Psychiatry Cogn Neurosci Neuroimaging. Published online July 23, 2021. doi:10.1016/j.bpsc.2021.07.004

36. Fu Z, Sui J, Turner JA, et al. Dynamic functional network reconfiguration underlying the pathophysiology of schizophrenia and autism spectrum disorder. Hum Brain Mapp. Published online September 23, 2020:hbm.25205. doi:10.1002/hbm.25205

37. Cheng W, Rolls E, Gong W, et al. Sleep duration, brain structure, and psychiatric and cognitive problems in children. Mol Psychiatry. 2021;26(8):3992-4003. doi:10.1038/s41380-020-0663-2

38. Dick AS, Garcia NL, Pruden SM, et al. No evidence for a bilingual executive function advantage in the nationally representative ABCD study. Nat Hum Behav. 2019;3(7):692-701. doi:10.1038/s41562-019-0609-3

39. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B. 1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.x

40. Biswal B, Zerrin Yetkin F, Haughton VM, Hyde JS. Functional connectivity in the motor cortex of resting human brain using echo‐planar mri. Magn Reson Med. 1995;34(4):537-541. doi:10.1002/mrm.1910340409

41. Londei A, D’Ausilio A, Basso D, Belardinelli MO. A new method for detecting causality in fMRI data of cognitive processing. Cogn Process. 2006;7(1):42-52. doi:10.1007/s10339-005-0019-5

42. Jafri MJ, Pearlson GD, Stevens M, Calhoun VD. A method for functional network connectivity among spatially independent resting-state components in schizophrenia. Neuroimage. 2008;39(4):1666-1681. doi:10.1016/j.neuroimage.2007.11.001

43. Váša F, Romero-Garcia R, Kitzbichler MG, et al. Conservative and disruptive modes of adolescent change in human brain functional connectivity. Proc Natl Acad Sci U S A. 2020;117(6):3248-3253. doi:10.1073/pnas.1906144117

44. Satterthwaite TD, Wolf DH, Loughead J, et al. Impact of in-scanner head motion on multiple measures of functional connectivity: Relevance for studies of neurodevelopment in youth. Neuroimage. 2012;60(1):623-632. doi:10.1016/j.neuroimage.2011.12.063

45. Satterthwaite TD, Wolf DH, Ruparel K, et al. Heterogeneous impact of motion on fundamental patterns of developmental changes in functional connectivity during youth. Neuroimage. 2013;83:45-57. doi:10.1016/j.neuroimage.2013.06.045

46. Marek S, Tervo-Clemmens B, Calabro FJ, et al. Towards Reproducible Brain-Wide Association Studies Affiliations. Johnny Uriarte. 2020;11:15-18. doi:10.1101/2020.08.21.257758

47. Shine JM, Koyejo O, Poldrack RA. Temporal metastates are associated with differential patterns of time-resolved connectivity, network topology, and attention. Proc Natl Acad Sci U S A. 2016;113(35):9888-9891. doi:10.1073/pnas.1604898113

48. Cole MW, Reynolds JR, Power JD, Repovs G, Anticevic A, Braver TS. Multi-task connectivity reveals flexible hubs for adaptive task control. Nat Neurosci. 2013;16(9):1348-1355. doi:10.1038/nn.3470

49. Hutchison RM, Womelsdorf T, Allen EA, et al. Dynamic functional connectivity: Promise, issues, and interpretations. Neuroimage. 2013;80:360-378. doi:10.1016/j.neuroimage.2013.05.079

50. Wang H, Fan L, Song M, et al. Functional Connectivity Predicts Individual Development of Inhibitory Control during Adolescence. Cereb Cortex. 2021;31(5):2686-2700. doi:10.1093/cercor/bhaa383

51. Ernst M, Torrisi S, Balderston N, Grillon C, Hale EA. FMRI functional connectivity applied to adolescent neurodevelopment. Annu Rev Clin Psychol. 2015;11:361-377. doi:10.1146/annurev-clinpsy-032814-112753

52. Ferreira LK, Busatto GF. Resting-state functional connectivity in normal brain aging. Neurosci Biobehav Rev. 2013;37(3):384-400. doi:10.1016/j.neubiorev.2013.01.017

53. Allen EA, Damaraju E, Eichele T, Wu L, Calhoun VD. EEG Signatures of Dynamic Functional Network Connectivity States. Brain Topogr. 2018;31(1):101-116. doi:10.1007/s10548-017-0546-2

54. Chen T, Cai W, Ryali S, Supekar K, Menon V. Distinct Global Brain Dynamics and Spatiotemporal Organization of the Salience Network. PLoS Biol. 2016;14(6). doi:10.1371/journal.pbio.1002469

55. Douw L, Wakeman DG, Tanaka N, Liu H, Stufflebeam SM. State-dependent variability of dynamic functional connectivity between frontoparietal and default networks relates to cognitive flexibility. Neuroscience. 2016;339:12-21. doi:10.1016/j.neuroscience.2016.09.034

56. Elton A, Gao W. Task-related modulation of functional connectivity variability and its behavioral correlations. Hum Brain Mapp. 2015;36(8):3260-3272. doi:10.1002/hbm.22847

57. Hutchison RM, Morton JB. Tracking the brain’s functional coupling dynamics over development. J Neurosci. 2015;35(17):6849-6859. doi:10.1523/JNEUROSCI.4638-14.2015

58. Cohen JR. The behavioral and cognitive relevance of time-varying, dynamic changes in functional connectivity. Neuroimage. 2018;180:515-525. doi:10.1016/j.neuroimage.2017.09.036

59. Kaiser RH, Whitfield-Gabrieli S, Dillon DG, et al. Dynamic Resting-State Functional Connectivity in Major Depression. Neuropsychopharmacology. 2016;41(7):1822-1830. doi:10.1038/npp.2015.352

60. Banich MT, Milham MP, Atchley RA, et al. Prefrontal regions play a predominant role in imposing an attentional “set”: Evidence from fMRI. Cogn Brain Res. 2000;10(1-2):1-9. doi:10.1016/S0926-6410(00)00015-X

61. Li Y, Zhu Y, Nguchu BA, et al. Dynamic Functional Connectivity Reveals Abnormal Variability and Hyper-connected Pattern in Autism Spectrum Disorder. Autism Res. 2020;13(2):230-243. doi:10.1002/aur.2212

62. Siebenhühner F, Weiss SA, Coppola R, Weinberger DR, Bassett DS. Intra- and Inter-Frequency Brain Network Structure in Health and Schizophrenia. PLoS One. 2013;8(8). doi:10.1371/journal.pone.0072351

63. Chen G, Chen P, Gong JY, et al. Shared and specific patterns of dynamic functional connectivity variability of striato-cortical circuitry in unmedicated bipolar and major depressive disorders. Psychol Med. Published online 2020:1-10. doi:10.1017/S0033291720002378

64. Young-Wolff KC, Sarovar V, Tucker LY, et al. Self-reported Daily, Weekly, and Monthly Cannabis Use among Women before and during Pregnancy. JAMA Netw Open. Published online 2019. doi:10.1001/jamanetworkopen.2019.6471

65. Paul SE, Hatoum AS, Fine JD, et al. Associations between Prenatal Cannabis Exposure and Childhood Outcomes: Results from the ABCD Study. JAMA Psychiatry. 2021;78(1):64-76. doi:10.1001/jamapsychiatry.2020.2902

There is NO Competing Interest.

Supplementarymaterialfinal.doc
Supplemental material

Download PDF

Version 1

posted

You are reading this latest preprint version

Functional Connectivity Uniqueness and Variability? A Signature of Cognitive and Psychiatric Problems in Children

Status:

Version 1

Abstract

Figures

Introduction

Results

Flowchart of the Cross-scan Functional Connectivity Analysis

Functional Networks

Intra-subject FNC Shows High Similarity across Scans

Individual Identification using Whole-brain FNC

FNC Stability Predicts Cognitive Performance

FNC Stability Predicts Psychiatric Problems

FNC Stability Correlates with Sleep Conditions and Screen Usage

FNC Stability Correlates with Parent Psychopathology and Prenatal Exposure

Discussion

The Fingerprint Property of Adolescent’s FNC

Individual FNC Stability Predict Adolescent’s Behaviors

Materials and methods

Participants and behavioral assessments

Neuromark Framework

Functional Network Connectivity

FNC Similarity and Subject Identification

Test-retest Reliability and Longitudinal Identification

Cognitive Measures

Mental Problem Measures

Sleep Conditions and Screen Usage

Parental Behaviors and Prenatal Exposure

Association between Cross-scan FNC Stability and Behaviors

Abbreviations

Declarations

Data Availability

Acknowledgements

Author Contributions

Competing Interests

References

Additional Declarations

Supplementary Files

Status:

Version 1