Gender differences in loneliness across the lifespan: A meta-analysis

Loneliness is a universal phenomenon that can be experienced throughout life (Qualter et al., 2015). It is defined as the unpleasant feeling that occurs when people perceive their network of social relationships to be deficient in a quantitative or qualitative way (Perlman & Peplau, 1981). Throughout the lifespan, sources of loneliness differ and may include deficient relations with parents, friends, or a romantic partner. However, the experience of dissatisfaction with one’s network of social relationships (i.e., loneliness) is likely the same across ontogeny. Gender differences in loneliness have been frequently examined, but theoretical notions are scarce and conflicting, and findings have been largely inconsistent. The present meta-analysis aimed to present an overview of theoretical ideas on gender differences in loneliness across the lifespan and to synthesize previous findings in a quantitative way. In addition to merely synthesizing previous work into one overall effect size, we further aimed to examine whether gender differences in loneliness vary in size or direction according to study characteristics, such as the loneliness measure used, and sample characteristics, such as age and nationality. For the large majority of studies, it was not clear from the title or the abstract whether gender differences in loneliness were examined. A meta-analysis, however, should be systematic and comprehensive. Therefore, in our literature search, we aimed to capture all studies that have used a questionnaire to measure loneliness. In total, 2,078 such publications from 1978 to 2013, written in English, Dutch, German, or French, were scrutinized for relevant information on gender differences. We were able to compute an effect size for 662 gender comparisons using a loneliness scale and those data were obtained in 37 different countries (some studies included multiple samples and a multidimensional measure of loneliness). A multi-level meta-analytic approach (Van den Noortgate & Onghena, 2003) was used to account for statistical dependency (i.e., multiple effect sizes within one study) and for the fact that different measures of loneliness were included. Preliminary analyses showed a significant difference suggesting higher loneliness scores for males than females. However, this effect size was very small (d = 0.05), which actually seems to suggest that, on average, there are no differences in loneliness. Preliminary moderator analyses suggested that effect sizes did not differ across countries, but did differ across age with males being slightly more lonely than females in college and university students only. Implications for current research on loneliness are discussed.status: accepte