The COVID-19 epidemic brought into prominence a different epidemic, that of loneliness, which has recently been defined as a modern behavioral phenomenon, similar to suicides and opioid use (Jeste et al., Reference Jeste, Lee and Cacioppo2020). The measures taken to stop the spread of the COVID-19 virus involved reducing the social sphere and avoiding in-person social interactions, which increased social isolation and loneliness. But loneliness was evident even before the COVID-19 pandemic and was attributed to social anomie resulting from the “rapid growth of technology, social media, globalization, and polarization of societies” (Jeste et al., Reference Jeste, Lee and Cacioppo2020, p.553). “The epidemic of loneliness” was confirmed in a recent systematic review and meta-analysis based on studies conducted between 2000 and 2019 in 113 countries and territories. Loneliness at a problematic level was found to be a common experience globally, especially in older people (Surkalim et al., Reference Surkalim2022). Losses experienced more often in old age because of the death of a spouse, siblings, and friends together with negative age-biased stereotypical attitudes, beliefs, and discriminatory behaviors toward older people increase loneliness in old age (Shiovitz-Ezra et al., Reference Shiovitz-Ezra, Erlich and Ayalon2023).
Unlike social isolation, loneliness is defined as a subjective feeling of lack of emotionally close and meaningful social ties (Perlman and Peplau, Reference Perlman and Peplau1998). Loneliness is a consequence of a perceived gap between the social resources available to individuals and those they would like to have. The gap refers to the number and frequency of social interactions and to their quality (Perlman and Peplau, Reference Perlman and Peplau1998). According to the evolutionary approach to loneliness, this gap creates significant discomfort and distress that motivates people to reengage and reconnect socially. Reconnecting with others is essential for survival and continuity (Spithoven et al., Reference Spithoven, Cacioppo, Goossens and Cacioppo2019).
The growing interest in loneliness is motivated by its prevalence as well as by the cumulative evidence of its detrimental effects on various aspects of health. Loneliness is associated with increased depression (Cacioppo et al., Reference Cacioppo, Hawkley and Thisted2010) and death wishes (Ayalon and Shiovitz-Ezra, Reference Ayalon and Shiovitz-Ezra2011). It has also been consistently found to be associated with compromised cognitive function (Rafnsson et al., Reference Rafnsson, Orrell, d’Orsi, Hogervorst and Steptoe2020) and to serve as a risk factor for heart disease (Thurston and Kubzansky, Reference Thurston and Kubzansky2009), inflammation, and metabolic deregulation (Shiovitz-Ezra and Parag, Reference Shiovitz-Ezra and Parag2019). Prospective studies have shown loneliness to be the cause of the increased risk of mortality. These negative health effects of loneliness were evident also during the COVID-19 pandemic, which cause psychological distress and low responsiveness to adopting preventive health behaviors such as wearing a mask, maintaining hand hygiene, and social distancing (Stickley et al., Reference Stickley, Matsubayashi and Ueda2020).
Although research has often treated loneliness as a unidimensional phenomenon measured at one time, some studies have suggested that its nature is more complex. Early theories referred to loneliness as consisting of two components. Based on the social need theory, Weiss (Reference Weiss1973) distinguished between emotional and social loneliness. The two types of loneliness result from an absence of a particular social provision, albeit a different one. Weiss identified six social provisions: attachment, social integration, reliability, alliance, guidance, reassurance of worth, and opportunity for nurturance, arguing that underlying emotional loneliness was the social provision of attachment, whereas underlying social loneliness was the provision of social integration.
Therefore, emotional loneliness refers to distress attributed to a lack of emotional figures and close or intimate relationships, and social loneliness to the lack of a broader social network of friends and colleagues. Weiss also proposed that the different types of loneliness produced different symptoms, such as anxiety, oversensitivity to minimal cues, feeling of abandonment, vigilance to threat, nameless fear, and constant appraisal for emotional loneliness, and boredom, depression, aimlessness, and marginality for social loneliness. Weiss also suggested that both types shared core experiences of poor concentration, distress, tension, restless depression, and amorphous and unfocused dissatisfaction.
Another theoretical division is based on the temporal patterns of loneliness, pointing to potential changes over time and challenging the perception of loneliness as a stable phenomenon. Thus, loneliness may be a short-term transient condition that occurs in response to relatively minor changes in the social arena that are limited in time. Alternatively, it may be experienced for a longer duration in response to stressful life events, which become more prevalent in old age, such as bereavement and retirement. It has been argued that in both cases, after a relatively short time of psychological distress caused by transient or situational factors, many individuals manage to cope and recover. For others, however, loneliness continues, and they experience it as chronic or persistent. It has been argued that stable states of loneliness were the result of long-term difficulties in developing meaningful social relationships (de Jong-Gierveld and Raadschelders, Reference De Jong-Gierveld, Raadschelders, Peplau and Perlman1982). Some individuals may be caught in a loneliness loop that prevents them from recovering. According to the loneliness model, loneliness is followed by feeling unsafe, which sets off implicit hypervigilance for social threats that produce cognitive biases causing the social world to be perceived as a more threatening place. Lonely people tend to remember more negative social information and have worse expectations from social interactions. Negative social expectations encourage behaviors by others that confirm the lonely persons’ expectations, setting in motion a self-fulfilling prophecy (Hawkley and Cacioppo, Reference Hawkley and Cacioppo2010).
Awad et al. (Reference Awad, Shamay-Tsoory and Palgi2023) addressed the complexity of the phenomenon of loneliness by adopting Weiss’ typology of examining social and emotional loneliness separately and tracking the fluctuations of loneliness for six weeks using a weekly diary. This method is used infrequently, despite its advantage in overcoming memory biases that may affect the way feelings and experiences are reported using a broader time frame. The study was conducted on a convenience sample of 55 participants aged 65 and over, who were mostly in good to very good health and enjoyed above-average financial conditions; therefore, the ability to generalize the findings is limited. Variability in health is important because of the accumulating evidence of associations between loneliness and health. Awad and colleagues tested how changes in interactions with family, close individuals, neighbors, and friends, as well as participation in social activities, affected week-to-week variations in loneliness. They found fluctuations in loneliness, with participants reporting severe loneliness at baseline and later reporting no loneliness, or vice versa. The instability of loneliness was evident also in both its social and emotional components. Furthermore, it was found that social loneliness was less sensitive to changes in objective social life. Weekly interactions with friends were related to emotional loneliness, contrary to the hypothesis according to which they correlated with fluctuations in social loneliness, which is defined as the absence of a broad social network of friends and others in the external social circle.
The study by Awad et al. expands the findings of recent works that address correlates of loneliness by exploring its complex multidimensional nature. Spitzer et al. (Reference Spitzer, Segel-Karpas and Palgi2022) found that close relationships negatively affected loneliness and that subjective age moderated this association for the benefit of those who felt subjectively younger. This study assessed loneliness at one time point and did not explore its dimensions. A recent systematic review (Van As et al., Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2021) that explored the longitudinal associations between loneliness and depression in the older population also failed to address fluctuations in loneliness. The authors emphasized that one of the eligibility criteria was that loneliness had to be measured at baseline and depression at follow-up. Therefore, the effect of change in loneliness on depression could not be assessed. Similarly, Creese et al. (Reference Creese2021), who explored trajectories of depression and anxiety between 2015 and 2020 concerning COVID-19 loneliness, measured mental health outcomes at several time points, but loneliness only in 2020. This study also failed to explore changes in loneliness and their consequences and conceptualized loneliness as a unidimensional construct, contrary to Awad et al., who tested the multidimensional and unstable nature of loneliness.
Some limitations in Awad et al. must be noted. The social correlates of loneliness were all in the objective sphere and related to the frequency of meeting with members of the social network and to social participation. No quality measures were taken into account, although it has been shown that conceptually loneliness was related to the quality of social ties. For example, recently, McComb et al. (Reference McComb, Goldberg, Flett and Rose2020) tested the concept of mattering to others, which is the feeling of being important to others in relation to trait (more stable) or state (changeable) loneliness. The results demonstrate that feelings such as being insignificant and unimportant to others were related to increased state and trait loneliness. In a previous study, perceived family strain, measured by experiences of too many demands and criticism by family members, was positively associated with feelings of loneliness in subsamples of both married and unmarried participants, but showed a stronger association in unmarried participants. Perceived support from the family, expressed by being open and reliant on family members, was associated with decreased loneliness but only in the unmarried participants (Shiovitz-Ezra and Leitch, Reference Shiovitz-Ezra and Leitsch2010). Moreover, marital status has been repeatedly shown to serve as a protective condition against loneliness; therefore, testing for differences in marital status in the associations between the frequency of social relationships and fluctuations in loneliness is important. However, in the Awad et al. study, only gender differences were explored.
Awad et al. measured loneliness weekly, whereas other studies examined changes in loneliness several times a day, referring to it as “momentary loneliness”. Other studies have tested profiles of loneliness over months or years and reported deleterious effects of loneliness experienced persistently and chronically. Persistent loneliness was found to be associated with significant negative health consequences (e.g. Tao et al., Reference Tao2022).
These studies describe the complexity of the phenomenon of loneliness and call for longitudinal research designs that measure loneliness starting from several times a day to periods of weeks, months, and years. They recommend developing profiles that include individuals who have recovered from loneliness and those who present an unstable sequence of recovery and relapse. Testing of correlates and consequences of the diverse loneliness profiles may deepen our understanding of loneliness as a complex concept. Clinical interventions for preventing loneliness can incorporate the insights derived from such studies into the complex reality in which they operate.
You have
Access
The COVID-19 epidemic brought into prominence a different epidemic, that of loneliness, which has recently been defined as a modern behavioral phenomenon, similar to suicides and opioid use (Jeste et al., Reference Jeste, Lee and Cacioppo2020). The measures taken to stop the spread of the COVID-19 virus involved reducing the social sphere and avoiding in-person social interactions, which increased social isolation and loneliness. But loneliness was evident even before the COVID-19 pandemic and was attributed to social anomie resulting from the “rapid growth of technology, social media, globalization, and polarization of societies” (Jeste et al., Reference Jeste, Lee and Cacioppo2020, p.553). “The epidemic of loneliness” was confirmed in a recent systematic review and meta-analysis based on studies conducted between 2000 and 2019 in 113 countries and territories. Loneliness at a problematic level was found to be a common experience globally, especially in older people (Surkalim et al., Reference Surkalim2022). Losses experienced more often in old age because of the death of a spouse, siblings, and friends together with negative age-biased stereotypical attitudes, beliefs, and discriminatory behaviors toward older people increase loneliness in old age (Shiovitz-Ezra et al., Reference Shiovitz-Ezra, Erlich and Ayalon2023).
Unlike social isolation, loneliness is defined as a subjective feeling of lack of emotionally close and meaningful social ties (Perlman and Peplau, Reference Perlman and Peplau1998). Loneliness is a consequence of a perceived gap between the social resources available to individuals and those they would like to have. The gap refers to the number and frequency of social interactions and to their quality (Perlman and Peplau, Reference Perlman and Peplau1998). According to the evolutionary approach to loneliness, this gap creates significant discomfort and distress that motivates people to reengage and reconnect socially. Reconnecting with others is essential for survival and continuity (Spithoven et al., Reference Spithoven, Cacioppo, Goossens and Cacioppo2019).
The growing interest in loneliness is motivated by its prevalence as well as by the cumulative evidence of its detrimental effects on various aspects of health. Loneliness is associated with increased depression (Cacioppo et al., Reference Cacioppo, Hawkley and Thisted2010) and death wishes (Ayalon and Shiovitz-Ezra, Reference Ayalon and Shiovitz-Ezra2011). It has also been consistently found to be associated with compromised cognitive function (Rafnsson et al., Reference Rafnsson, Orrell, d’Orsi, Hogervorst and Steptoe2020) and to serve as a risk factor for heart disease (Thurston and Kubzansky, Reference Thurston and Kubzansky2009), inflammation, and metabolic deregulation (Shiovitz-Ezra and Parag, Reference Shiovitz-Ezra and Parag2019). Prospective studies have shown loneliness to be the cause of the increased risk of mortality. These negative health effects of loneliness were evident also during the COVID-19 pandemic, which cause psychological distress and low responsiveness to adopting preventive health behaviors such as wearing a mask, maintaining hand hygiene, and social distancing (Stickley et al., Reference Stickley, Matsubayashi and Ueda2020).
Although research has often treated loneliness as a unidimensional phenomenon measured at one time, some studies have suggested that its nature is more complex. Early theories referred to loneliness as consisting of two components. Based on the social need theory, Weiss (Reference Weiss1973) distinguished between emotional and social loneliness. The two types of loneliness result from an absence of a particular social provision, albeit a different one. Weiss identified six social provisions: attachment, social integration, reliability, alliance, guidance, reassurance of worth, and opportunity for nurturance, arguing that underlying emotional loneliness was the social provision of attachment, whereas underlying social loneliness was the provision of social integration.
Therefore, emotional loneliness refers to distress attributed to a lack of emotional figures and close or intimate relationships, and social loneliness to the lack of a broader social network of friends and colleagues. Weiss also proposed that the different types of loneliness produced different symptoms, such as anxiety, oversensitivity to minimal cues, feeling of abandonment, vigilance to threat, nameless fear, and constant appraisal for emotional loneliness, and boredom, depression, aimlessness, and marginality for social loneliness. Weiss also suggested that both types shared core experiences of poor concentration, distress, tension, restless depression, and amorphous and unfocused dissatisfaction.
Another theoretical division is based on the temporal patterns of loneliness, pointing to potential changes over time and challenging the perception of loneliness as a stable phenomenon. Thus, loneliness may be a short-term transient condition that occurs in response to relatively minor changes in the social arena that are limited in time. Alternatively, it may be experienced for a longer duration in response to stressful life events, which become more prevalent in old age, such as bereavement and retirement. It has been argued that in both cases, after a relatively short time of psychological distress caused by transient or situational factors, many individuals manage to cope and recover. For others, however, loneliness continues, and they experience it as chronic or persistent. It has been argued that stable states of loneliness were the result of long-term difficulties in developing meaningful social relationships (de Jong-Gierveld and Raadschelders, Reference De Jong-Gierveld, Raadschelders, Peplau and Perlman1982). Some individuals may be caught in a loneliness loop that prevents them from recovering. According to the loneliness model, loneliness is followed by feeling unsafe, which sets off implicit hypervigilance for social threats that produce cognitive biases causing the social world to be perceived as a more threatening place. Lonely people tend to remember more negative social information and have worse expectations from social interactions. Negative social expectations encourage behaviors by others that confirm the lonely persons’ expectations, setting in motion a self-fulfilling prophecy (Hawkley and Cacioppo, Reference Hawkley and Cacioppo2010).
Awad et al. (Reference Awad, Shamay-Tsoory and Palgi2023) addressed the complexity of the phenomenon of loneliness by adopting Weiss’ typology of examining social and emotional loneliness separately and tracking the fluctuations of loneliness for six weeks using a weekly diary. This method is used infrequently, despite its advantage in overcoming memory biases that may affect the way feelings and experiences are reported using a broader time frame. The study was conducted on a convenience sample of 55 participants aged 65 and over, who were mostly in good to very good health and enjoyed above-average financial conditions; therefore, the ability to generalize the findings is limited. Variability in health is important because of the accumulating evidence of associations between loneliness and health. Awad and colleagues tested how changes in interactions with family, close individuals, neighbors, and friends, as well as participation in social activities, affected week-to-week variations in loneliness. They found fluctuations in loneliness, with participants reporting severe loneliness at baseline and later reporting no loneliness, or vice versa. The instability of loneliness was evident also in both its social and emotional components. Furthermore, it was found that social loneliness was less sensitive to changes in objective social life. Weekly interactions with friends were related to emotional loneliness, contrary to the hypothesis according to which they correlated with fluctuations in social loneliness, which is defined as the absence of a broad social network of friends and others in the external social circle.
The study by Awad et al. expands the findings of recent works that address correlates of loneliness by exploring its complex multidimensional nature. Spitzer et al. (Reference Spitzer, Segel-Karpas and Palgi2022) found that close relationships negatively affected loneliness and that subjective age moderated this association for the benefit of those who felt subjectively younger. This study assessed loneliness at one time point and did not explore its dimensions. A recent systematic review (Van As et al., Reference Van As, Imbimbo, Franceschi, Menesini and Nocentini2021) that explored the longitudinal associations between loneliness and depression in the older population also failed to address fluctuations in loneliness. The authors emphasized that one of the eligibility criteria was that loneliness had to be measured at baseline and depression at follow-up. Therefore, the effect of change in loneliness on depression could not be assessed. Similarly, Creese et al. (Reference Creese2021), who explored trajectories of depression and anxiety between 2015 and 2020 concerning COVID-19 loneliness, measured mental health outcomes at several time points, but loneliness only in 2020. This study also failed to explore changes in loneliness and their consequences and conceptualized loneliness as a unidimensional construct, contrary to Awad et al., who tested the multidimensional and unstable nature of loneliness.
Some limitations in Awad et al. must be noted. The social correlates of loneliness were all in the objective sphere and related to the frequency of meeting with members of the social network and to social participation. No quality measures were taken into account, although it has been shown that conceptually loneliness was related to the quality of social ties. For example, recently, McComb et al. (Reference McComb, Goldberg, Flett and Rose2020) tested the concept of mattering to others, which is the feeling of being important to others in relation to trait (more stable) or state (changeable) loneliness. The results demonstrate that feelings such as being insignificant and unimportant to others were related to increased state and trait loneliness. In a previous study, perceived family strain, measured by experiences of too many demands and criticism by family members, was positively associated with feelings of loneliness in subsamples of both married and unmarried participants, but showed a stronger association in unmarried participants. Perceived support from the family, expressed by being open and reliant on family members, was associated with decreased loneliness but only in the unmarried participants (Shiovitz-Ezra and Leitch, Reference Shiovitz-Ezra and Leitsch2010). Moreover, marital status has been repeatedly shown to serve as a protective condition against loneliness; therefore, testing for differences in marital status in the associations between the frequency of social relationships and fluctuations in loneliness is important. However, in the Awad et al. study, only gender differences were explored.
Awad et al. measured loneliness weekly, whereas other studies examined changes in loneliness several times a day, referring to it as “momentary loneliness”. Other studies have tested profiles of loneliness over months or years and reported deleterious effects of loneliness experienced persistently and chronically. Persistent loneliness was found to be associated with significant negative health consequences (e.g. Tao et al., Reference Tao2022).
These studies describe the complexity of the phenomenon of loneliness and call for longitudinal research designs that measure loneliness starting from several times a day to periods of weeks, months, and years. They recommend developing profiles that include individuals who have recovered from loneliness and those who present an unstable sequence of recovery and relapse. Testing of correlates and consequences of the diverse loneliness profiles may deepen our understanding of loneliness as a complex concept. Clinical interventions for preventing loneliness can incorporate the insights derived from such studies into the complex reality in which they operate.
Conflict of interest
None.