Background: There is a paucity of past research concerning longitudinal changes in loneliness in old age and in the factors that correlate with loneliness in older persons. This study examines (a) whether loneliness increases with age, (b) characteristics of the lonely older population, and (c) predictors of becoming lonely in old age.
Methods: Data are from the Cross-Sectional and Longitudinal Aging Study, a national longitudinal study on the health, mental, and social status of people aged 75 and older living in Israel. Loneliness was measured as the frequency at which participants had felt lonely during the last month (range: 0–3). The analysis included two waves of data, with an average gap of 3.5 years between them, with 1147 participants at wave 1 and 588 participants at wave 2.
Results: Data showed an increase in the mean level of loneliness over time, rising from 0.62 to 0.80, p<0.001. The strongest correlate of being lonely at wave 1 was unmarried status. Different variables predicted loneliness in the married and the unmarried. Among the married, lonelier persons tended to have experienced more traumatic events, had more doctor's visits, and manifested less cognitive vitality. In contrast, among the unmarried, loneliness was higher in those with financial difficulties, poor health, and less social network support. The longitudinal transition to becoming lonely was more likely in women, those with insufficient financial resources, and people with poorer health.
Conclusions: Results highlight the following implications for the development of preventive programs: (a) helping older persons maintain and develop social networks for preventing increased loneliness, (b) taking demographic factors, health factors, and previous trauma into account when designing interventions, and (c) targeting unmarried women for interventions.