Great review article: Robinaugh2020 - The network approach to psychopathology, a review of the literature 2008-2018 and an agenda for future research

Overview of the Dutch project New Science of Mental Disorders: Roefs2022 - A new science of mental disorders, Using personalised, transdiagnostic, dynamical systems to understand, model, diagnose and treat psychopathology

What network theory is

Mental disorders as systems of interacting components, rather than mere passive reflections of an underlying common cause (McNally, 2020)

The current dominating model relies on the Germ theory of disease. Also referred to as “common cause” model, Borsboom2018 - Brain disorders Not really Why network structures and “latent variable” model.

McNally2016 - Can network analysis transform psychopathology : Instead of focusing on “hallmark symptoms”, we should put weight on symptoms that are non-specific and appear in many disorders since they may be particularly important (centrality hypothesis and comorbidity hypothesis). Recommends more frequent measurement through apps (rating mood several times per day).

van den Hout2014 - Psychiatric Symptoms as Pathogens : If we say that symptoms are caused by diagnosis it becomes a tautology and circle reasoning. Underlying biological mechanisms or even markers have yet to be seen.

Jones2021 - So What, Implications of Network Theory for Research and Practice

Network approach to psychopathology

Systems-based approaches may be the best way to cumulatively build and improve theoretical models of mental disorders [@borsboom2022] - Borsboom2022 - Systems-based approaches to mental disorders are the only game in town

Terms and analysis

Network nodes

Network edges

Axiom of local independence

Tipping point

Bistable system : A system in which there are two equillibrium states. Depressed vs non-depressed. Both states are relatively stable once established.

Example of critical slowing down and tipping points in MDD - Lempuut2014 - Critical slowing down as early warning for the onset and termination of depression

Symptom networks can have contagion effects where activation spreads through the network - connectivity hypothesis

When causal relations among symptoms are strong, the onset of one symptom will lead to the onset of others - causality hypothesis

Symptoms occuring in multiple disorders can bridge activation between the networks - comorbidity hypothesis

Central symptoms have greater potential to spread activation throughout the network - centrality hypothesis



Bootstrapping and Regularization in network analysis - Epskamp2018 - Estimating psychological networks and their accuracy, A tutorial paper

Node predictability

Different types of networks

Association networks : only strength of association, undirected

Partial correlation networks : adjusting for the influence of all other nodes

Relative importance networks : adding direction

Bayesian networks

Tipping point and Critical slowing down as early indicators of rapid change in systems Scheffer2012 - Anticipating Critical Transitions

Network structure analyses

Olatunji2019 - What is at the core of OCD, A network analysis of selected obsessive-compulsive symptoms and beliefs : Negative appraisals of intrusive thoughts were the most central symptoms, but their predictive power on OCD symptom severity was weak.

Meier2020 - Obsessive-compulsive symptoms in eating disorders, a network investigation : ED and OCD clusters distinct from one another. Interference due to obsessions most central OCD symptom.

Forrest2022 - Change in Eating-Disorder Psychopathology Network Structure in Patients With Binge-Eating Disorder, Findings From Treatment Trial With 12-Month Follow-Up : Strength centrality estimates pre-trt, post-trt and 12-month follow-up

McNally2017 - Co-morbid obsessive-compulsive disorder and depression, a Bayesian network approach : Distress from obsessions as a bridge symptom -> sadness -> rest of MDD cluster

Hellberg2022 - Co-occurring depression and obsessive-compulsive disorder, A dimensional network approach: Obsessions and negative affect were central to symptom networks in patients with co-occurring OCD and depression

Litvin2022 - A Network examination of Symptoms of Obsessive-compulsive disorder and sleep disturbances: In patients with primary OCD and no MDD, there were weak associations between symptoms of insomnia and OCD.

Cervin2020 - The Centrality of Doubting and Checking in the Network Structure of Obsessive-Compulsive Symptom Dimensions in Youth : Doubting/checking most central symptom

Ong2021 - Obsessive-compulsive Symptoms and Other Symptoms of the At-risk Mental State for Psychosis, A Network Perspective : Depression most central symptom, anxiety important bridge between OC and other symptoms (general psychopathology questionnaire)

Kuckertz2022 - Does the network structure of obsessive-compulsive symptoms at treatment admission identify patients at risk for non-response : Baseline network differences (Y-BOCS items) between responders and non-responders.

Berlim2021 - The network structure of core depressive symptom-domains in major depressive disorder following antidepressant treatment, a randomized clinical trial : Networks more connected at post-treatment, most central symptoms were fatigue (pre) and depressed mood (post).

Blanken2019 - Introducing Network Intervention Analysis to Investigate Sequential, Symptom-Specific Treatment Effects, A Demonstration in Co-Occurring Insomnia and Depression : Weekly assessments of symptom networks. First effects on sleep issues, later effects on sleep satisfaction. Effects on depressive symptoms likely indirect via sleep improvements.

Blanken2021 - Symptom-specific effects of cognitive therapy and behavior therapy for insomnia, A network intervention analysis : Specific CT and BT effects, but not 100% overlap with the 2022 study below. No process measures.

Lancee2022 - Network Intervention Analyses of cognitive therapy and behavior therapy for insomnia, Symptom specific effects and process measures : Specific effects of CT and BT at mid- and post-treatment, in line with theory of respective trt.

Cervin2020 - Symptom-specific effects of cognitive-behavioral therapy, sertraline, and their combination in a large randomized controlled trial of pediatric anxiety disorders : One study investigated the network effects of CBT and sertraline in treating pediatric anxiety, and found that the treatments were effective through reductions in the same symptoms [@cervin2020b]

Bernstein2021 - Mechanisms of cognitive-behavioral therapy effects on symptoms of body dysmorphic disorder, a network intervention analysis : similar to Blanken, repeated analyses. CBT more effective in targeting BDD thoughts and behavior change, SPT more effective in improving insight. Changes align with therapy modules in CBT

Moment-to-moment mood connections are stronger in individuals with a mental disorder, suggesting that those individuals are susceptible to downward spirals of negative mental states

Remaining questions

Recent developments include methods for aggregating results in meta-analyses, confirmatory network analysis and longitudinal network analysis [@epskamp2022] - Epskamp2022 - New trends in network modeling of psychopathology

Formal theories

Transclude of block next-steps

McNally2021 - Network Analysis of Psychopathology, Controversies and Challenges : Varying coefficients between studies does not mean that network analyses per se are non-replicable.

From Roefs2022 - A new science of mental disorders, Using personalised, transdiagnostic, dynamical systems to understand, model, diagnose and treat psychopathology:

Transclude of block targeting

Clinical utility

  • Can we use central nodes as treatment targets?
    • “Positive cascade” or low hanging fruit
  • Do more central symptoms predict OCD severity and other clinical outcomes better than non-central symptoms?
    • We might also see that common factors are more important than the CBT specific targets
  • Experimental manipulation of central symptoms -> wide effects on other parts of the network? Here we have multiple useful examples from the experimental literature, reviewed in van den Hout2014 - Psychiatric Symptoms as Pathogens. High responsibility -> Compulsions, and Perseveration of compulsions -> Uncertainty
  • More frequent data collection, similar to mood ratings in depression, interesting to explore?
  • Confirmatory network analyses, meta-analyses, not just exploratory results
  • Can network structures help predict onset of mental illness (See Boschloo 2016)

How is therapy effective from a network point of view?

  • Lowering or removing one central symptom
  • Remove an external stressor (too much to do at work, bullying in school)
  • Change the relationship between two symptoms