Predictors of Loneliness among Youth in Middle Childhood and Adolescence during the COVID-19 Pandemic

• In Event: Impact of COVID-19 on Children and Youth

Abstract

Introduction: COVID-19 has negatively affected youth mental health (Meade, 2021; Sabato et al., 2021). As social isolation increased due to stay-at-home orders, the association between loneliness and mental health problems increased (Loades et al., 2020). Youth may be particularly susceptible to COVID-19 changes, as their social environments (e.g, school, family) are pivotal to social-emotional development (Carr, 2017; Oberle et al., 2014). Limited research exists on middle childhood (MC; aged 8-12) during the COVID-19 pandemic and even less focuses on loneliness, as most research on loneliness concentrates on adolescents (i.e., aged 13-17). Thus, we aimed to 1) examine the differences between MC and adolescent loneliness during COVID-19, and 2) better understand the impact of social environments and COVID-19 on loneliness.

Methods: Parents of youth (8-17 years) attending school in-person before the pandemic were recruited through Facebook May-July 2020 (T1) and November 2020-January 2021 (T2). At T1, parents reported on family demographic characteristics, whether their child was attending school (yes/no), and communicating with friends in-person (yes/no). Youth completed measures of their COVID-19 impact, loneliness, and perceived family functioning at T1 and T2 (Bevans et al., 2010; Kazak et al., 2021). Children (n=92, Mage=10.03) were mostly White and male. Adolescents (n=56, Mage=14.66) were mostly White and female (Table 1). Pearson’s correlations were used to examine relationships between focal variables. A hierarchical regression model examined T1 predictors of T2 loneliness. Step 1 included youth sex, race, ethnicity, and age. Step 2 added T1 youth-reported COVID-19 impact. Step 3 added T1 youth-reported family functioning. Step 4 added T1 parent-reported school attendance, and step 5 added T1 parent-reported in-person communication with friends.

Results: At T1, 18.5% (n=17) of children and 26.8% of adolescents (n=15) maintained in-person communication with friends. Mean loneliness differed significantly at T1 for children (µMC=54.30, SDMC=11.12) and for adolescents (µAdolescent=58.64, SDAdolescent=9.77; t(146)= -2.413, p=0.017). Both MC and adolescent loneliness were moderate at T2 (µMC=55.40, SDMC=11.34; µAdolescent=56.61, SDAdolescent=12.31). The difference was non-significant at T2. Mean COVID-19 impact score at T1 was 1.72 (SD=0.76) for children and 1.90 (SD=0.92) for adolescents, indicating low impact. The overall model predicting loneliness of those in MC was significant, R2=0.327, F(8, 83)= 5.044, p<.001 (Table 2). The adolescent model was also significant, R2=0.314, F(8, 46)= 2.629, p=0.018. Significant final predictors of greater loneliness at T2 for children were poorer family functioning, no school attendance, no in-person communication with friends, and being Hispanic. Poor family functioning was the only significant predictor of greater loneliness at T2 for adolescents.

Conclusion: Beyond COVID-19 impact, loneliness for both groups during this time may be influenced by family functioning. Additionally, for children, school and peer changes may play a vital role, indicating the importance of social context in this age group. Encouraging in-person interaction with friends and increasing school engagement may be protective factors against loneliness for children. For both groups, support for positive family functioning may help protect against loneliness within and outside of the context of COVID-19.

Authors

• Ashley Benhayoun, Nationwide Children's Hospital
  Presenting Author

• Anna Lorraine Olsavsky, Nationwide Children's Hospital

• Parishma Guttoo

• Valdeoso Patterson

• Terrah Foster Akard

• Cynthia A. Gerhardt, The Research Institute at Nationwide Children's Hospital

• Micah Skeens