Network Intervention Analysis

Unraveling the mechanisms of Cognitive Behavioural Therapy for personalised insomnia treatment

Over the past years we have introduced and expanded the use of network analysis to investigate treatment effects, collectively coined as Network Intervention Analysis (NIA). We have now applied NIA to (i) distinguish direct from indirect treatment effects, (ii) identify treatment differences, and (iii) identify treatment processes. 

We developed and applied these techniques all in the context of cognitive and behavioural therapy for insomnia (CBT-I). Its first application was inspired by a conversation I had with Jaap Lancee and Tanja van der Zweerde who presented their study on investigating the effect of insomnia treatment on the alleviation of depression symptoms at the European Insomnia Network meeting in 2016, later published in Psychological Medicine. They identified that, indeed, CBT-I also alleviated depression complaints, but were left wondering whether these effects on depression complaints are likely to occur because of alleviation in insomnia complaints, or whether CBT-I perhaps targets the depression complaints directly. Brainstorming at the EIN, we came up with the idea to explore this using network analysis. Crucially, in the network we would not include ‘depression’ and ‘insomnia’ as separate constructs, but rather look at their symptoms and their interrelations. We then came up with the idea to include a ‘treatment variable’ into the network, that would allow us to investigate which symptoms were predominantly affected by treatment, while taking the interrelationships with the other symptoms into account. That’s how we came up with and introduced Network Intervention Analysis (Blanken et al., 2019). 

Later, we continued our discussions on how we could apply a similar technique to shed light on different questions. We started out with contrasting the two active treatment components in CBT-I: cognitive versus behavioural therapy. Using the two RCTs out there (Harvey et al., 2014; Sunnhed et al., 2020) we used NIA to investigate which symptoms were differentially impacted by treatment (Blanken et al., 2021; Lancee et al., 2022). Interestingly, while both original RCTs did not identify differences between the two treatment components at an overall severity level, we consistently found different symptoms to be impacted by cognitive versus behavioural treatment. Possibly, by only looking at the overall severity level we miss more fine-grained effects at the symptom level. In our most recent paper we investigated differential treatment effects of behavioural therapy versus zolpidem medication and again identified treatment specific effects (Lancee et al., 2023). Accompanying my symposium presentation at World Sleep in 2022, I outlined the different ways in which we used NIA to better understand how CBT-I works in alleviating insomnia symptoms (Blanken et al., 2022).    

Collaborators: Jaap Lancee, Tanja van der Zweerde

related papers

Lancee J, Morin CM, […], Blanken TF. Symptom-specific effects of zolpidem and behavioral treatment for insomnia: A network intervention analysis. Sleep 2023; 46: zsad240. [PDF]

Blanken TF, T. Van der Zweerde, J. Lancee. Network Intervention Analysis: A novel way to unravel treatment targets and mechanisms of cognitive behavioral therapy for insomnia. Sleep Medicine 2022. 100: S131. [PDF]

Lancee J, Harvey AG, […], Blanken TF. Network Intervention Analyses of cognitive therapy and behavior therapy for insomnia: process measures and symptom specific effects. Behav Res Ther 2022; 153: 104100. [PDF]

Blanken TF, Jansson-Fröjmark M, Sunnhed R, Lancee J. Symptom-specific effects of cognitive therapy and behavior therapy for insomnia: A network intervention analysis. J Consult Clin Psychol 2021; 89: 364-370. [PDF]

Blanken TF*, Van der Zweerde T*, Van Straten A, et al. Treating co-occurring insomnia and depression symptoms: Introducing Network Intervention Analysis to reveal sequential, symptom-specific effects of online insomnia treatment. Psychother Psychosom 2019; 88: 52–54. [PDF]

* shared first/last authorship

related presentations

“Een uitleg en illustratie van Netwerk Interventie Analyse in het identificeren van sequentiële en symptoom-specifieke behandeleffecten”. Symposium presentation at Sleep2023, December 2023, Arnhem, The Netherlands. [PDF]

“Network Intervention Analysis: Using network analysis to explore sequential and symptom-specific treatment processes and effects”. Keynote presentation at the Experimental Psychopathology graduate school, December 2022, Driebergen, The Netherlands. [PDF]

“Network Intervention Analysis: Using network analysis to explore sequential and symptom-specific treatment processes and effects”. Symposium presentation at the International Conference for Urban Mental Health, October 2022, Amsterdam, The Netherlands. [PDF]

“Voorspellen, voorkomen en verlichten van depressie: nieuwe netwerk analyses wijzen op het primaire belang van insomnie.” Symposium presentation at Slaap2019, November 2019, Ermelo, The Netherlands. [PDF]

“Big data en netwerk interventie analyse onthullen subtypen van slapeloosheid en hoe behandeling op depressieve symptomen aangrijpt.” Symposium presentation at NVvP voorjaarscongres, April 2019, Maastricht, The Netherlands. [PDF]

“Introducing Network Intervention Analysis (NIA) to investigate sequential, symptom-specific treatment effects. A demonstration in co-occurring insomnia and depression.” European Insomnia Network Meeting, December 2018, Oxford, United Kingdom. [PDF]