Psychosocial factors associated with loneliness change during COVID‐19 pandemic in Italian older‐old individuals: Findings from the InveCe.Ab cohort

Abstract

AbstractBackgroundLoneliness is an issue for many older adults, significantly affecting physical, cognitive and mental health. The aim of the present study was to evaluate if loneliness worsened in older‐old individuals during the COVID‐19 nationwide lockdown in Italy, compared to their pre‐pandemic status. Moreover, we explored which pre‐pandemic psychosocial factors predict loneliness change.MethodThe present investigation was conducted on 409 older adults without dementia and aged between 79 and 84 years, participating in the fourth evaluation wave of the InveCe.Ab study in 2018 (NCT01345110) and then performing a telephone survey between March and May 2020, during the first Italian nationwide lockdown. Loneliness was evaluated with the UCLA 3‐item loneliness scale both in 2018 and 2020. The following pre‐pandemic measures, collected as part of the in‐person multidimensional assessment in 2018, were considered: age, sex, education, social contact, leisure and physical activity, digital technology usage (yes/no), perceived support from family network (yes/no), personality traits (psychoticism, extraversion, neuroticism), depressive symptoms (Geriatric Depression Scale, GDS), global cognition (Mini Mental State Examination, MMSE) and baseline loneliness. To compare pre‐post pandemic loneliness, we performed the Wilcoxson signed‐rank test. Kendal tau rank correlations were calculated to preliminarily explore the associations between each psychosocial factors and loneliness change (delta). Finally, stepwise multiple linear regression based on the p‐value criteria was used to test which variables independently predicted loneliness change.ResultLoneliness showed a significant increase during the lockdown (p<0.001). Loneliness change was associated with lower GDS (τ = ‐0.221, p<0.001), psychoticism (τ = ‐0.086, p = 0.024), neuroticism (τ = ‐0.116, p = 0.002) and higher extraversion (τ = 0.079, p = 0.035), MMSE (τ = 0.106, p = 0.006), social contacts (τ = 0.091, p = 0.015) leisure activities (τ = 0.103, p = 0.006). Multiple regression showed that only 2 psychosocial factors explained 27% of the variance (R2 = 0.27, F = 74,7, p<0.001). Specifically, baseline loneliness (β = ‐0.70, p<0.001) and perceived support from family network (β = ‐0.50, p = 0.007) predicted loneliness change.ConclusionLoneliness increased during the first nationwide lockdown among older‐old individuals in Italy, compared to their pre‐pandemic level. This change was most prominent among those with lower pre‐pandemic loneliness levels and it was mitigated by perceived support by family network.