Abstract
Major Depressive Disorder (MDD) is a mental illness that greatly impacts patients as well as the people around them. It is estimated to become the biggest contributor to the global burden of disease in the upcoming decades. MDD is hard to understand because it is a highly heterogeneous illness: patients with the same MDD diagnosis may display vastly different symptom profiles, illness course, and responses to treatment. Furthermore, reliable biological markers for MDD have yet to be found.
MDD and psychiatric disorders, in general, have traditionally been analyzed as if they are physical illnesses; i.e. as a latent cause that leads to the manifestation of depression symptoms. Importantly, these symptoms are assumed to be mutually independent: their co-occurrence is completely explained by the assumed underlying disorder. The use of the latent variable model to represent MDD is problematic because both the existence of a single latent cause and the independence of symptoms are hard to justify. The current dissertation aims to investigate the use of an alternative disease model of MDD: the network model. Recently, network models have been increasingly applied in psychiatric research, forming an interesting alternative to latent variable models, because they do not assume a single latent cause for MDD and they can model (in)direct relationships among symptoms.
This thesis considers the application of network models in mental health research on cross-sectional and longitudinal study data. Moreover, we discuss the strengths, weaknesses, possibilities, and challenges of the network approach in psychiatric research.
MDD and psychiatric disorders, in general, have traditionally been analyzed as if they are physical illnesses; i.e. as a latent cause that leads to the manifestation of depression symptoms. Importantly, these symptoms are assumed to be mutually independent: their co-occurrence is completely explained by the assumed underlying disorder. The use of the latent variable model to represent MDD is problematic because both the existence of a single latent cause and the independence of symptoms are hard to justify. The current dissertation aims to investigate the use of an alternative disease model of MDD: the network model. Recently, network models have been increasingly applied in psychiatric research, forming an interesting alternative to latent variable models, because they do not assume a single latent cause for MDD and they can model (in)direct relationships among symptoms.
This thesis considers the application of network models in mental health research on cross-sectional and longitudinal study data. Moreover, we discuss the strengths, weaknesses, possibilities, and challenges of the network approach in psychiatric research.
Original language | English |
---|---|
Qualification | Doctor of Philosophy |
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 3-Feb-2020 |
Place of Publication | [Groningen] |
Publisher | |
Print ISBNs | 978-94-034-2369-2 |
Electronic ISBNs | 978-94-034-2368-5 |
DOIs | |
Publication status | Published - 2020 |