Robust symptom networks in recurrent major depression across different levels of genetic and environmental risk

Loo, Hanna M. van; Borkulo, Claudia D. van; Peterson, Roseann E.; Fried, Eiko I.; Aggen, Steven H.; Borsboom, Denny; Kendler, Kenneth S.

doi:10.1016/j.jad.2017.10.038

Cited by 43 publications

(72 citation statements)

References 59 publications

(89 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Second, the bootnet package in R (R Core Team, 2013) uses three bootstrapping routines to examine the stability of edge weight and node centrality estimates within networks. (1) A non-parametric bootstrap draws many (e.g., 1000; Fried et al, 2018) different subsamples with replacement from the observed data. The network edges estimated in these subsamples are pooled to generate a sampling distribution for each edge.…”

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

“…Finally, the NCT package (van Borkulo et al, under review) uses permutation testing to quantify differences between pairs of networks with, ideally, similar sample sizes. This test works by pooling the data from the two networks that are being compared, and then repeatedly randomly re-assigning the data into two groups to estimate many (e.g., 5000 ;Fried et al, 2018) new pairs of networks. This process results in a reference distribution of differences between the two networks, representing the null hypothesis that the networks are drawn from the same population.…”

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

“…As above, this is based on comparing the difference in the observed global strength values for each network to the reference distribution of differences from the permuted networks. In addition to these three tests, networks are often also compared using a "coefficient of similarity" (i.e., a Spearman rank correlation of the edges lists from a pair of networks; e.g., Borsboom et al, 2017;Fried et al, 2018;Rhemtulla et al, 2016). Borsboom et al (2017) described the methods in the bootnet and NCT packages as "powerful tools" for assessing the stability and consistency of PMRF networks (p. 990), and together with LASSO regularization they have become the default suite of methods for research on PMRF psychopathology networks.…”

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

“…Borsboom et al (2017) described the methods in the bootnet and NCT packages as "powerful tools" for assessing the stability and consistency of PMRF networks (p. 990), and together with LASSO regularization they have become the default suite of methods for research on PMRF psychopathology networks. Most of the latest studies use these methods to interpret cross-sectional networks, to examine change in networks over time, and to conduct replication studies (e.g., Beard et al, 2016;Fried et al, 2018;Fried et al, 2016;van Loo et al, 2018). These studies have largely concluded that network characteristics (i.e., estimated edges and symptom centrality estimates) are accurate, sufficiently stable to interpret, consistent over time, and largely generalizable.…”

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

“…Fried for comments on an earlier version of this manuscript. Thanks also to Fried and colleagues-who authored the study on the replicability of posttraumatic stress disorder symptom networks (Fried et al, 2018, Clin Psychol Sci)-for making their summary data, output, and code available, and for encouraging reanalysis of the data for further replicability research.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Quantifying the Reliability and Replicability of Psychopathology Network Characteristics

Forbes

Wright

Markon

et al. 2019

Multivariate Behavioral Research

View full text Add to dashboard Cite

Acknowledgements: Thanks to Nicholas R. Eaton and Eiko I. Fried for comments on an earlier version of this manuscript. Thanks also to Fried and colleagues-who authored the study on the replicability of posttraumatic stress disorder symptom networks (Fried et al., 2018, Clin Psychol Sci)-for making their summary data, output, and code available, and for encouraging reanalysis of the data for further replicability research. AbstractPairwise Markov random field networks-including Gaussian graphical models (GGMs) andIsing models-have become the "state-of-the-art" method for psychopathology network analyses. Recent research has focused on the reliability and replicability of these networks. In the present study, we compared the existing suite of methods for maximizing and quantifying the stability and consistency of PMRF networks (i.e., lasso regularisation, plus the bootnet and NetworkComparisonTest packages in R) with a set of metrics for directly comparing the detailed network characteristics interpreted in the literature (e.g., the presence, absence, sign, and strength of each individual edge). We compared GGMs of depression and anxiety symptoms in two waves of data from an observational study (n = 403) and reanalyzed four posttraumatic stress disorder GGMs from a recent study of network replicability. Taken on face value, the existing suite of methods indicated that overall the network edges were stable, interpretable, and consistent between networks, but the direct metrics of replication indicated that this was not the case (e.g., 39-49% of the edges in each network were unreplicated across the pairwise comparisons). We discuss reasons for these apparently contradictory results (e.g., relying on global summary statistics versus examining the detailed characteristics interpreted in the literature), and conclude that the limited reliability of the detailed characteristics of networks observed here is likely to be common in practice, but overlooked by current methods. Poor replicability underpins our concern surrounding the use of these methods, given that generalizable conclusions are fundamental to the utility of their results.Keywords: network analysis; network theory of mental disorders; psychopathology networks; conditional independence; replicability crisis Quantifying the reliability and replicability of psychopathology network characteristicsThe central tenet of the network theory of mental disorders is that psychopathology is due to dynamic causal interactions among symptoms (Borsboom, 2017). This theory encourages the quantification of the symptom-level structure of psychopathology using network analysis-a method with rapidly growing popularity in psychopathology research.

show abstract

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

Section: Current Psychopathology Network Methodsmentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Quantifying the Reliability and Replicability of Psychopathology Network Characteristics

Forbes

Wright

Markon

et al. 2019

Multivariate Behavioral Research

View full text Add to dashboard Cite

show abstract

Mindfulness as practice: A network analysis of FMI data

Smith

Kempton

Williams

et al. 2021

Couns and Psychother Res

View full text Add to dashboard Cite

By committing to latent variable models, mindfulness research has aimed to transform observable practices into an identifiable real 'mindfulness' experience which is claimed to exist beyond what is directly observed. Recently, an alternative methodology has been developed which allows mindfulness to be modelled as a complex system or network at the level of self-report. This study hypothesised that a more densely connected network of observable practices is indicative of a greater level of development of mindfulness. MethodsMindfulness networks were estimated for practitioners and non-practitioners using the Friedberg Mindfulness Inventory (FMI). A total of 371 regular mindfulness practitioners, 224 non-practitioners and 59 irregular practitioners were recruited online from the Amazon Mechanical Turk database. ResultsComparisons of practitioners' and non-practitioners' networks indicated that network density did not significantly differ, whereas evidence was found in support of a significant difference in network structure. An exploratory analysis revealed that the FMI item representing the mindfulness practice of Acceptance was substantially more central in the Practitioners FMI network, relative to its position in the Non-practitioners FMI network. FMI items representing the mindfulness practices of Self-kindness and Returning to the Present were substantially more peripheral to the practitioners FMI network relative to their position in the non-practitioners FMI network. Conclusions. MINDFULNESS AS PRACTICE 3The study provides proof-of-principle support for investigating mindfulness as a complex network at the level of self-report. However, the lack of difference in network density indicates that future research is needed to examine network dynamics in the context of regular mindfulness practice.

show abstract

Network structure of depression symptomology in participants with and without depressive disorder: the population-based Health 2000–2011 study

Hakulinen

Fried

Pulkki-Råbäck

et al. 2020

Soc Psychiatry Psychiatr Epidemiol

View full text Add to dashboard Cite

Purpose Putative causal relations among depressive symptoms in forms of network structures have been of recent interest, with prior studies suggesting that high connectivity of the symptom network may drive the disease process. We examined in detail the network structure of depressive symptoms among participants with and without depressive disorders (DD; consisting of major depressive disorder (MDD) and dysthymia) at two time points. Methods Participants were from the nationally representative Health 2000 and Health 2011 surveys. In 2000 and 2011, there were 5998 healthy participants (DD−) and 595 participants with DD diagnosis (DD+). Depressive symptoms were measured using the 13-item version of the Beck Depression Inventory (BDI). Fused Graphical Lasso was used to estimate network structures, and mixed graphical models were used to assess network connectivity and symptom centrality. Network community structure was examined using the walktrap-algorithm and minimum spanning trees (MST). Symptom centrality was evaluated with expected influence and participation coefficients. Results Overall connectivity did not differ between networks from participants with and without DD, but more simple community structure was observed among thosewith DD compared to those without DD. Exploratory analyses revealed small differences between the samples in the order of one centrality estimate participation coefficient. Conclusions Community structure, but not overall connectivity of the symptom network, may be different for people with DD compared to people without DD. This difference may be of importance when estimating the overall connectivity differences between groups with and without mental disorders.

show abstract

Robust symptom networks in recurrent major depression across different levels of genetic and environmental risk

Abstract: Similar depressive symptom networks for MD patients with a higher or lower genetic or environmental risk suggest that differences in these etiological influences may produce similar symptom networks downstream for severely depressed women.

Cited by 43 publications

References 59 publications

Quantifying the Reliability and Replicability of Psychopathology Network Characteristics

Quantifying the Reliability and Replicability of Psychopathology Network Characteristics

Mindfulness as practice: A network analysis of FMI data

Network structure of depression symptomology in participants with and without depressive disorder: the population-based Health 2000–2011 study

Contact Info

Product

Resources

About