Loneliness Among Adults Dependent on Age and Multiple Psychological and Environmental Factors

Looking at the patterns of association between age and loneliness, they often vary. This study examines age-related differences in risk and protective factors for loneliness.

Researchers from University of California San Diego identified from a survey study a complex age-dependent relationship between loneliness and contributing environmental and psychological factors among Americans. These findings were published in the Journal of Clinical Psychiatry.

Adults (N=2843) were recruited from the Amazon Mechanical Turk (AMT) online crowdsourcing marketplace between April and May 2019. The survey of 90 items measured sociodemographic characteristics including age, race/ethnicity, education, marital status, income, and living situation (number of people in household) and also included the University of California Loneliness Scale (UCLA-4).

Participants were aged mean 42.9 (standard deviation [SD], 12.7) years. Stratified by their decade of age, respondents differed significantly for all measured sociodemographic characteristics (P <.001), symptoms of physical health (P <.001), symptoms of mental and cognitive health (P <.001), negative psychological features (P <.001), social interactions (P £.021), and positive psychological features (P £.013; except for insight [P =.276] and social advising [P =.464] subscores).

Average loneliness score differed by decade of age (F[4,2887], 11.5; P <.001). Loneliness was highest among those in their twenties (UCLA-4, 9.2; SD, 2.5), followed by during their forties (UCLA-4, 8.9; SD, 2.6), thirties (UCLA-4, 8.8; SD, 2.7), and fifties (UCLA-4, 8.5; SD, 2.6). Participants in their sixties were the least lonely (UCLA-4, 8.1; SD, 2.6).

This pattern of loneliness, with a second peak among those in their forties, was made up of 2 independent quadratic functions (Wald statistic, 5.50; P =.019), with 1 between 20 and 44 years and the other between 45 and 69 years of age.

A model with discrete age accounted for 52.1% of the variance of loneliness (Wald statistic, 5.48; P =.019) in which the nonlinear curve peaked at 47.7 years of age. Individuals who reported greater loneliness were more likely to not have a spouse or partner, had poorer sleep, lower prosocial behaviors, a smaller social network, lower self-efficacy, and more anxiety.

In a second model which included the interaction between decade of age and cofactors, 52.3% of the variance of loneliness was explained with marital status (P <.001), prosocial behaviors (P <.001), social network (P <.001), and sleep quality (P =.02) being significant contributors to loneliness. Among respondents aged £59 years, poor social self-efficacy (P <.001) and high anxiety (P <.005) were associated with greater loneliness.

All bivariate correlations were significant across age groups but were more strongly related among younger individuals (c2, 34.7; P <.001).

This study was possibly biased by the self-reported design which may have caused some under or over reporting of categories.

These data suggested that many aspects of life contributed to loneliness among adults, and that unique combinations of factors contributed to individuals differently depending on their decade of life.


Nguyen TT, Lee EE, Daly RE, et al. Predictors of loneliness by age decade: Study of psychological and environmental factors in 2,843 community-dwelling Americans aged 20-69 years. J Clin Psychiatry. 2020;81(6):20m13378. doi: 10.4088/JCP.20m13378