Dynamic symptom associations in posttraumatic stress disorder: a network approach

ABSTRACT

Background and objective: The current study aimed to investigate the within-day symptom dynamics in PTSD patients, specifically focusing on symptoms that most predict changes in other symptoms. The study included a baseline diagnostic assessment, followed by an assessment using the experience sampling method (ESM) via a smartphone.

Method: Participants answered questions related to their PTSD symptoms four times per day for 15 consecutive days (compliance rate 75%). The clinical sample consisted of 48 treatment-seeking individuals: 44 with PTSD as a primary diagnosis, and four patients with subsyndromal PTSD, all of whom had not yet begun trauma-focused treatment. The ESM assessment included the 20 items from the PTSD Checklist for DSM-5, five items from the International Trauma Questionnaire (ITQ) assessing disturbances in relationships and functional impairment, and two items from the Clinician-Administered PTSD Scale for DSM-5 assessing symptoms of depersonalization and derealization.

Results: Temporal networks (prospective associations between symptoms) showed that changes in hypervigilance predicted changes in the greatest number of symptoms at the next time point. Furthermore, hypervigilance showed temporal connections with at least one additional symptom from each of the DSM-5 PTSD symptom clusters.

Conclusions: Results show that the contemporaneous network (representing the relationship between given symptoms within the same assessment occasion) and the temporal network (representing prospective associations between symptoms) differ and that it is important to estimate both. Some findings from earlier research are replicated, but heterogeneity across studies remains. Future studies should include potential moderators.

HIGHLIGHTS

• We investigated within-day symptom dynamics in PTSD patients using experience sampling technology.

• Temporal and contemporaneous symptom networks differed; thus, it is important to estimate both.

• Changes in hypervigilance were an important predictor of symptoms at the next time point.

Antecedentes y objetivo: El presente estudio buscó investigar las dinámicas de los síntomas dentro de un día en pacientes con TEPT, centrándose específicamente en los síntomas que más predicen los cambios en otros síntomas. El estudio incluyo una evaluación diagnostica inicial, seguida de una evaluación mediante el método de muestreo de experiencias (ESM en su sigla en inglés) a través de un teléfono inteligente.

Método: Los participantes contestaron preguntas relacionadas con sus síntomas del TEPT cuatro veces al día por 15 días consecutivos (tasa de cumplimiento 75%). La muestra clínica consistió en 48 individuos que buscaban tratamiento: 44 con TEPT como diagnóstico principal, y cuatro pacientes con TEPT subclínico, todos los cuales aún no habían empezado un tratamiento centrado en el trauma. La evaluación del ESM incluyó los 20 ítems de la Lista de Chequeo de TEPT para el DSM-5, cinco ítems del Cuestionario Internacional de Trauma (ITQ en sus siglas en ingles) que evalúa las alteraciones en las relaciones y el deterioro funcional, y dos ítems de la Escala de TEPT administrada por el clínico según el DSM-5 que evalúa los síntomas de despersonalización y desrealización.

Resultados: Las redes temporales (asociaciones prospectivas entre síntomas) mostraron que los cambios en hipervigilancia predijeron los cambios en la mayor cantidad de síntomas en el siguiente momento. Además, la hipervigilancia mostró conexiones temporales con al menos un síntoma adicional de cada grupo de síntomas de TEPT del DSM-5.

Conclusiones: Los resultados muestran que las redes contemporáneas (que representa la relación entre los síntomas dados dentro de la misma ocasión de evaluación) y la red temporal (que representa las asociaciones prospectivas entre síntomas) difirieren y que es importante estimar ambas. Se replican algunos hallazgos de investigaciones previas, pero persiste la heterogeneidad entre los estudios. Futuras investigaciones deberían incluir potenciales moderadores.

KEYWORDS:

• Posttraumatic stress disorder
• ESM
• network analysis
• multilevel VAR
• contemporaneous
• temporal

PALABRAS CLAVE:

• Trastorno de estrés postraumático
• ESM
• análisis de redes
• VAR de multiniveles
• contemporáneo
• Temporal

Acknowledgements

Experience sampling data was collected with the Tellmi App, which was developed with support from a grant from the German Research Foundation to Felix Schönbrodt (SCHO 1334/5-1).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study and the R Code are accessible for reviewers in OSF at https://osf.io/upaxk/, and will be made openly accessible via OSF upon publication.

Notes

1 These four participants attended one of the specialized treatment centers for their PTSD symptomatology but did not meet full DSM-5 criteria in the structured interviews.

2 When data collection for this study started, SCID-5-CV was not yet available in all outpatient centers. With the intention of simplifying the process for the patients, if a diagnostic was already provided with the SCID-IV and CAPS, we did not repeat the diagnostic part, as all patients were diagnosed according to DSM-5 criteria.

3 As a VAR model assumes the stationarity for each time series, we confirmed that there was no significant time trend on the PCL, ITQ, or CAPS scores.

4 Compliance rate was computed as the number of valid ESM responses divided by total number of beeps. Then, the compliance rate was correlated with the PCL score (both specified at the person level), which was not statistically significant.

Additional information

Funding

MS was supported by scholarship from BAYHOST – Bavarian Academic Center for Central, Eastern and Southeastern Europe.