Loneliness has been defined as a feeling resulting from a deficiency in a person’s social relations, acknowledged though a subjective evaluation of one’s achieved and desired levels of social relations (Perlman and Peplau, Reference Perlman, Peplau, Duck and Gilmour1981). The discrepancy between the relations a person has and the relations that s/he wishes to have may concern quantitative aspects, such as how many relations a person has or how often s/he meets or has contacts with family and friends, or it may concern qualitative aspects, such as trust or intimacy (de Jong Gierveld, Reference de Jong Gierveld1998). In the 1970s, Weiss (Reference Weiss1973) identified two dimensions of loneliness: the social and the emotional. While social loneliness refers to the lack of a broader group of contacts to which one belongs or having an engaging social network, emotional loneliness refers to the absence of intimate and emotionally close relations. Some instruments developed to measure loneliness identify different dimensions of loneliness, while other instruments and all single items are unidimensional.

In addition to the social and emotional dimensions of loneliness, there is literature on existential loneliness. Compared to social and emotional loneliness, existential loneliness has less to do with social relations and is instead related to a lack of meaning in life (van Tilburg, Reference van Tilburg2021, p. e336), an awareness of one’s mortality and fundamental separateness from other people and the universe (Bolmsjö et al., Reference Bolmsjö, Tengland and Rämgård2019). It has been argued that existential loneliness is mostly experienced in life-threatening situations and research on existential loneliness has often been undertaken in care contexts including end-of-life care (Ettema et al., Reference Ettema, Derksen and van Leeuwen2010). There is less research on the prevalence, risk factors for, and consequences of this type of loneliness compared to social and emotional loneliness, and the literature discussed in this commentary does not focus on existential loneliness.

A meta-analytic study provided a pooled estimate of 28% of people aged over 65 years living in high-income countries experiencing loneliness (Chawla et al., Reference Chawla, Kunonga, Stow, Barker, Craig and Hanratty2021). Loneliness can vary both in intensity and frequency, and a minority of 8% of the total sample experienced severe loneliness. There are substantial variations in the prevalence of loneliness across countries, with the lowest rates found in Northern Europe and higher rates in Southern and Eastern Europe (Chawla et al., Reference Chawla, Kunonga, Stow, Barker, Craig and Hanratty2021). Media often report an increase in loneliness and the co-chair of the British Loneliness Commission argued that loneliness “has escalated from personal misfortune into a social epidemic” (Policy Exchange, 2017). However, research shows that prevalence rates are stable. For example, a Swedish study found no significant changes in loneliness over two decades, from 1992 until 2014 (Dahlberg et al., Reference Dahlberg, Agahi and Lennartsson2018). In England, no increase was found in loneliness when comparing results from a survey of older adults in 1999 with surveys undertaken between 1945 and 1960 (Victor et al., Reference Victor2002). The COVID-19 pandemic has, though, included a number of restrictions on social life that may potentially increase loneliness in the population, particularly among the oldest old, and several research studies point in this direction (for a review, see Buecker and Horstmann, Reference Buecker and Horstmann2021; cf. Dahlberg, Reference Dahlberg2021).

Loneliness in older adults increases the risk of poor well-being, physical and mental health problems, and even mortality (Courtin and Knapp, Reference Courtin and Knapp2017; Rico-Uribe et al., Reference Rico-Uribe, Caballero, Martin-Maria, Cabello, Ayuso-Mateos and Miret2018; Solmi et al., Reference Solmi2020). It has been argued that the aging process can pose challenges, including health problems, impaired functional ability, cognitive impairment, and life events such as loss of close relationships, that may contribute to both depression and loneliness, for example, by causing difficulties in the maintenance of relationships (Van As et al., Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022; van Zutphen et al., Reference van Zutphen, Kok, Rijnhart, Rhebergen, Huisman and Beekman2021). A broad review on health and well-being consequences of loneliness in old age found that depression was the most commonly studied outcome (Courtin and Knapp, Reference Courtin and Knapp2017), and several reviews and meta-analyses have shown that loneliness is associated with an increased risk of depression (e.g., Cohen-Mansfield et al., Reference Cohen-Mansfield, Hazan, Lerman and Shalom2016; Erzen and Cikrikci, Reference Erzen and Cikrikci2018; Solmi et al., Reference Solmi2020). However, previous reviews have pointed out that the majority of studies on the association between loneliness and depression are cross-sectional (Cohen-Mansfield et al., Reference Cohen-Mansfield, Hazan, Lerman and Shalom2016; Courtin and Knapp, Reference Courtin and Knapp2017). This limits the potential to determine what is cause and effect, and in a review of longitudinal risk factors of loneliness in older adults depression was identified as one of the few risk factors for which there was strong evidence of an association (Dahlberg et al., Reference Dahlberg, McKee, Frank and Naseer2022).

As published in the current issue of this journal, Van As et al. (Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022) have undertaken a review to examine the other direction of that relationship, that is, whether loneliness is a risk factor of depression in older adults, limited to healthy community-living persons aged 60 years or older. It should be noted that the studies in the review used validated self-report measures and not clinically diagnosed depression as the outcome. There is variation in how depression is labeled and measured in the literature considered below, but for concision I will use the term depression in this commentary.

Unlike most previous reviews, Van As et al. (Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022) only included studies with longitudinal design, thus avoiding some of the interpretive restrictions imposed by cross-sectional designs. They found that all included articles confirmed the association between loneliness and depression. Another novel feature of their review is its second outcome: the clinical course of depression. They concluded that loneliness increases the probability of worsening depression and hinders remission from depression. Such findings have been confirmed in a later study showing that during the COVID-19 pandemic the risk of increase in depression in persons aged 50 years or older was higher among those experiencing loneliness than among non-lonely people (Creese et al., Reference Creese2021).

A total of 10 studies were identified and included in the review, of which five studies considered the first outcome (depression at follow-up), two studies considered the latter outcome (the clinical course of depression such as new onset, remission, or maintenance), and three studies considered both outcomes (Van As et al., Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022). Taken together, the number of research studies focusing on older adults in this field is thus limited. There is evidence that loneliness is more common in certain subgroups of older adults, such as women, people of higher ages, and people living alone (Dahlberg et al., Reference Dahlberg, McKee, Frank and Naseer2022). With only a total of 10 relevant studies included in the review, Van As et al. (Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022) were not able to examine whether the association between loneliness and depression varied across different subgroups of older adults. Another limitation of this review is that no risk-of-bias assessment was made of the included studies, which means that the reliability of the synthesis is unclear. Future reviewers should conduct such assessments and, where possible, also do meta-analyses of the findings from included studies.

Still, there is evidence that depression is both a risk factor for and an outcome of loneliness in older adults. There may also be a bidirectional association between loneliness and depression. This has rarely been examined in samples of older adults, and the findings are mixed: while some studies confirmed reciprocal effects between depression and loneliness (e.g., van Zutphen et al., Reference van Zutphen, Kok, Rijnhart, Rhebergen, Huisman and Beekman2021), other studies found that loneliness more robustly predicts depression at the subsequent wave of data collection than vice versa (e.g., Hsueh et al., Reference Hsueh, Chen, Hsiao and Lin2019). Thus, further research is needed in order to fully disentangle the relationship between loneliness and depression. As part of this, it is essential to explore potential mechanisms influencing this relationship. In previous reviews, several mechanisms through which loneliness influences health have been suggested, including adverse health behaviors, impaired sleep, neurobiological mechanisms (e.g., age-related changes in stress responses, elevated blood pressure, and diminished immunity), and negative social cognition (Hawkley and Cacioppo, Reference Hawkley and Cacioppo2010; Ong et al., Reference Ong, Uchino and Wethington2016). Mechanisms specific for the pathway from loneliness to depression are less discussed.

Van As et al. (Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022) call for future research on factors moderating the association between depression and loneliness, especially factors that could decrease the harmful effects of loneliness on depression. Such factors could include individual resources, such as social relations and the functional ability to engage in social activities, or meso and macro-level resources such as supportive neighborhoods and welfare systems. When considering the influence of different socio-cultural contexts, it would also be of importance to examine whether the pathway between loneliness and depression manifests in different ways under extraordinary circumstances such as pandemics.

There is limited research differentiating between the different dimensions of loneliness. Three of the 10 studies included in the review by Van As et al. (Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022) used the de Jong Gierveld Loneliness Scale, which includes subscales on social and emotional loneliness. Only one of these studies considered these dimensions of loneliness separately in the analyses, and found that both social and emotional loneliness contributes to depression severity (Holvast et al., Reference Holvast, Burger, de Waal, van Marwijk, Comijs and Verhaak2015). None of the studies considered existential loneliness. Thus, it can be concluded that more research is needed on the different dimensions of loneliness as their associations with depression could vary in strength and as modifying factors may differ.

Van As et al. (Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2022) have contributed to the research field by synthesising research evidence on the longitudinal association between loneliness and depression in healthy community-living older adults. In line with previous reviews, they conclude that loneliness increases the risk of depression. In addition, they find that loneliness increases the risk of worsening depression and hinders remission from depression. Given the small number of studies identified in the review, there is an evident need for further longitudinal research on the association between loneliness and depression in older adults. The review specifically underlines the need for explorations of moderating factors of this association. Then, it would be of interest to distinguish between the effects of different dimensions of loneliness. When seen in the light of reviews on depression as a risk factor for loneliness, it is apparent that further research is also needed on the complex and potentially reciprocal relationship between loneliness and depression: to what extent do lonely people become sad and/or sad people lonely?