Abstract
Loneliness is defined as a distressing feeling that arises from the discrepancy between one’s desired and perceived social relationships. Theorists have posited that loneliness involves affective, cognitive, and behavioral components that can be noticed by close family or friends. Little is known about whether social withdrawal, a behavioral marker of loneliness observed by knowledgeable others, shares similar associations with cognition as self-reported loneliness. The present study examined whether self-reported and informant-rated loneliness are related to cognitive function in older adulthood. Data come from Harmonized Cognitive Assessment Protocol subsample included in three national aging studies from the US (HRS; N = 2,821, Mage = 75.8), England (ELSA; N = 896, Mage = 73.6), and India (LACI-DAD; N = 2,994, Mage = 69.3). Respondents reported on 1-item loneliness and informants rated the respondent’s withdrawal behavior. Regression models were used to analyze the association between both measures of loneliness and cognition, controlling for respondent (age, gender, education, race/ethnicity) and informant (age, gender, education, spouse, known years, frequency of contact) covariates. Meta-analysis showed that both respondent-reported and informant-rated loneliness were independently associated with lower cognitive functioning (memory, speed–attention–executive, visuospatial ability, and fluency) and global cognition (MMSE). The associations between observed withdrawal and cognition did not vary by informant-level characteristics nor the respondent’s cognitive impairment status. The present study indicates that withdrawal behavior observed by informants is associated with cognitive function, supporting psychosocial observations provided by knowledgeable others can be utilized in detecting cognitive function.