Mental Health and (Online) Behaviors during the COVID-19 Pandemic in Spain: A Network-Based Approach

Abstract

The COVID-19 pandemic has contributed to the increase in mortality and morbidity rates globally, but it has also led to a generalized worsening of mental health and risk behaviors in different population groups regardless of the measures adopted by different governments. In this paper, using data from a Spanish survey of emotional well-being, we aim to explore through mixed graphical models the complex structure of relationships between the mental health of populations, their lifestyles, and forms of cultural and leisure consumption during the pandemic. The results bring to light some interesting findings, such as the association between teleworking and greater rest or greater stress with the use of social media, a variable that enables the connection with other mental health problems of greater severity. Increased physical activity and the consumption of streaming content at home, as well as increased care for family, friends, and neighbors, are some of the variables that show relevant associations. These findings highlight the usefulness and versatility of this network approach for the study of health behaviors and health outcomes, which offer the researcher a holistic and organic view of the relational structure of complex data characterized by high dimensionality and variables with different levels of measurement.

Keywords: health behaviors; lifestyles; mental health; mixed graphical models; social network analysis.

Figure 1

Distribution of responses in the mental health variables (%). The response categories are (1) All or most of the time; (2) A good part of the time; (3) Some of the time; and (4) At no time or almost no time.

Figure 2

Distribution of responses in behavioral variables (%).

Figure 3

Mixed graphical model using the full dataset (k  =  2).

Figure 4

Centrality measures for variables in the model.