How is linear regression being used in this study? Why linear regression?

Background:It is well-documented that the mental health of pregnant and postpartum women is essential for maternal, child, and family well-being. Of major public health concern is the perinatal mental health impacts that may occur during the ongoing coronavirus disease 2019 (COVID-19) pandemic. It is essential to explore the symptom experience and predictors of mental health status, including the relationship between media use and mental health.

Materials and Methods:The purpose of this study is to evaluate the experiences of pregnant and postpartum women (n= 524) in the United States in the early phase of the COVID-19 pandemic. This cross-sectional online observational study collected psychosocial quantitative and qualitative survey data in adult pregnant and postpartum (up to 6 months postdelivery) women in April-June 2020.

Results:Multivariable linear regression models were used to evaluate predictors of depressive symptoms, anxiety, and post-traumatic stress disorder. The most common predictors were job insecurity, family concerns, eating comfort foods, resilience/adaptability score, sleep, and use of social and news media. Qualitative themes centered on pervasive uncertainty and anxiety; grief about losses; gratitude for shifting priorities; and use of self-care methods including changing media use.

Conclusions:This study provides information to identify risk for anxiety, depression, and PTSD symptoms in perinatal women during acute public health situations. Women with family and job concerns and low resilience/adaptability scores seem to be at high risk of psychological sequelae. Although use of social media is thought to improve social connectedness, our results indicate that increased media consumption is related to increased anxiety symptoms.