Transitions in loneliness in later life: the role of social comparisons and coping strategies

Abstract This study explored the coping strategies and social comparisons used by older adults on different loneliness trajectories (decreased loneliness, stable loneliness and degenerating loneliness). The adaptive consequences of social comparison in later life are recognised as an important strategy for preserving life satisfaction regardless of age-related losses. Coping strategies are also important in managing loneliness. Narrative interviews were conducted with lonely older adults (N = 11) who had participated in Wave One of the Maintaining Function and Well-being in Later Life Study Wales (CFAS Wales). The study found key differences in the coping strategies employed by older adults on different loneliness trajectories. Differences in coping styles between those who reported decreased loneliness and those who were chronically lonely stemmed from perceptions as to whether loneliness was modifiable or not. Different types of social comparison were also found to modulate the loneliness experience. The findings indicate that higher-order strategies (problem, emotional and meaning focused) are not distinct entities but are part of a dynamic process. The management of loneliness in later life may be dependent on several factors, including older adults’ interpretations of the cause of loneliness. These interpretations will have implications for interventions aimed at alleviating chronic loneliness, where the focus may have to be on changing older adult's perceptions of unmodifiable loneliness.


Introduction
This article explores the role of coping strategies and social comparisons in shaping the loneliness trajectory. We draw on qualitative data to look at the coping strategies utilised by older adults on three different loneliness trajectories: decreased loneliness, stable loneliness and degenerating loneliness. We explore how these strategies shape and influence the loneliness trajectory. We will argue that how an individual understands the event(s) that trigger loneliness is pivotal in the strategies they choose to adopt to address or manage loneliness. This choice of strategies has implications for the individual's loneliness trajectory: whether they recover or go on to become chronically lonely.
Loneliness is a universal experience; one that is inherent within the human condition. Indeed, it has been argued that loneliness is an evolutionary mechanism signifying our need for social connection (Cacioppo and Patrick, 2008). This is important as older adults are not passive actors, and on recognising loneliness, they will adopt strategies to manage or alleviate it. Yet, we know very little about how older adults themselves manage loneliness, whether the strategies employed differ for individuals on different loneliness trajectories or if the chosen coping strategies influence the loneliness trajectory of that individual. This article addresses this deficit by exploring, from the perspective of lonely older adults themselves, how older adults on different loneliness trajectories have utilised coping strategies and social comparisons and how this has shaped their loneliness trajectory.
In this article, loneliness is defined as a situation in which there is a discrepancy between the desired and achieved number, frequency and/or quality of social relationships (De Jong Gierveld, 1998). Social comparison refers to the ways in which individuals evaluate their situation by comparing themselves to others (Festinger, 1954). Coping strategies refer to individuals' cognitive and behavioural efforts, to manage external and internal stressors which are perceived to exceed personal resources (Lazarus and Folkman, 1984).

Transitions in loneliness
Loneliness is not a linear experience, levels may change over time, with loneliness decreasing, remaining stable or worsening. Research has identified five distinct loneliness trajectories: loneliness as a constant feature of the lifecourse, as a new experience, loneliness that is perceived to be increasing (degenerating), loneliness perceived to be decreasing and those who never experience loneliness (Victor et al., 2009).
The literature on transitions or trajectories in loneliness to date has focused on the structural, functional and socio-demographic features of older people's lives and how these are associated with changes in levels of loneliness over time (Hawkley and Kocherginsky, 2018). For example, Newall et al. (2014) used longitudinal data to explore patterns of stability and change in loneliness across four loneliness trajectories: became lonely, overcame loneliness, persistently lonely and persistently not lonely. They identified several factors, including changes in health, income, environment and perceptions of control as key factors in loneliness transitions. However, we have found no research to date that has drawn on qualitative data to explore the role of coping strategies or social comparisons in shaping the loneliness trajectory from the perspective of older adults themselves.

Coping with loneliness
Previous research has identified the range of coping strategies used by lonely older adults. Strategies identified include increasing social interaction and providing a distraction from loneliness, e.g. by engaging in solitary pursuits or adopting unhealthy behaviours such as excessive alcohol use (Pettigrew and Roberts, 2008;Hauge and Kirkevold, 2010). Coping strategies have been categorised into two higher orders: problem-focused and emotion-focused (Baker and Berenbaum, 2007).
Problem-focused coping strategies are defined as efforts to modify loneliness by increasing or improving the quality of existing relationships or by developing new social contacts. In contrast, emotion-focused coping involves managing or reducing the emotional distress associated with loneliness by distracting from loneliness or engaging in solitary pursuits. These emotion-focused strategies include attempts to diminish the emotional impact of loneliness, also referred to as regulative coping (Peplau and Perlman, 1982) or non-social behaviours (Pettigrew and Roberts, 2008).
A third dimension of coping, meaning focused, has also been identified by Folkman (2008). Meaning focused, coping is an appraisal-based strategy, which is less situation-specific than either problem or emotion coping mechanisms; instead, it draws on beliefs and value systems to motivate and sustain coping activities and maintain wellbeing (Folkman, 2008). It is argued that this strategy can act as a catalyst helping a person re-appraise their priorities and validate their sense of self in light of their loneliness. Which, in turn, can help them to make positive changes or to find a new focus in their life (Hutchinson et al., 2008).
Existing research on the effectiveness of coping strategies in reducing or alleviating loneliness is contradictory. Rook and Peplau (1982) suggested that both problem-and emotion-focused strategies may be effective in reducing loneliness. However, this is inconsistent with the findings of Schoenmakers et al. (2015) who found that older adults who utilised emotion-focused strategies, such as lowering expectations for social interaction, were more likely to be chronically lonely. On the other hand, older adults who recovered from loneliness were found to have utilised more problem-focused coping strategies (Schoenmakers et al., 2015). However, a significant limitation in the study was that it captured the coping options that non-lonely participants selected for other lonely people, rather than the coping strategies they had chosen. This article addresses this deficit by exploring the coping strategies employed by lonely older adults.

Social comparisons
To date, the role of social comparisons as a coping mechanism for loneliness has received little attention in the academic literature. However, social comparisons are an important element in the experience of loneliness . Humans have an innate desire to understand self and to construct meaningful accounts of their personal experiences . One way of achieving this is through self-evaluation. Through self-evaluation, individuals make a subjective assessment of their social relationships, comparing their current situation to their past experiences, and other people's relationships (Revenson, 1981;. When their relationships are perceived to fall short, it is argued they will experience loneliness. This discrepancy between achieved and desire social relationships is a central tenet of cognitive discrepancy theory on which this article draws (Perlman and Peplau, 1981;Peplau and Perlman, 1982;. The prevailing social norms may influence an individual's expectations about their social relationships, and these may change with age (Perlman and Peplau, 1981). For example, a person's desired level of social relations may be tempered by their expectations about the sorts of relationships that are possible in a given situation, such as within a particular geographic location combined with a set of personal health and material resources (Perlman and Peplau, 1981: 40). On the other hand, they may continue to hold the same expectations concerning desired social relations that they held earlier in the lifecourse (Dykstra et al., 2005). Social norms and desirable personal attributes are identified during socialisation and behaviours are adapted accordingly. Socialisation is defined as a lifelong process whereby an individual learns the norms, values, behaviours and social skills that are acceptable in society. Thus, a deficit in social relationships may be interpreted as loneliness and result in the adoption of strategies to ameliorate the situation.
We argue that social comparison can be considered as a regulative coping strategy that moderates feelings of loneliness. Comparisons to people who are perceived to be more advantaged (upward) are likely to result in a negative self-evaluation. In contrast, comparisons with people who are perceived to be more disadvantaged (downward) are likely to result in positive self-evaluations. While research has demonstrated that favourable social comparisons help preserve mental wellbeing in later life (Wills, 1981;Beaumont and Kenealy, 2004), there is limited evidence of the role of social comparisons as a coping strategy to manage loneliness. This article will build on the existing literature by exploring the coping strategies employed by older adults on three different loneliness trajectories.
The article addresses the following questions: (1) What are the coping strategies used by older people who report recovering from loneliness, stable loneliness and degenerating loneliness? (2) What types of social comparisons are made by older people who report recovering from loneliness, stable loneliness and degenerating loneliness? (3) Why do some older people adopt specific coping strategies or types of social comparison?

Sample
Participants for this study were recruited from the Wave One sample of the Maintaining Function and Well-being in Later Life Study (CFAS Wales). CFAS Wales is a longitudinal cohort study of health and ageing, conducted between 2011 and 2014 in Wales, United Kingdom (UK). Quantitative sample design details are reported by Burholt and Sardani (2018). A total of 3,593 interviews were conducted with participants aged 65 and older at Wave One. The qualitative sample was drawn from Wave One participants. The exclusion and inclusion criteria were sent to the CFAS Wales Research Team at Cambridge University who selected the sample (N = 20) using a purposive sampling frame. As the purpose of the study was to explore stability and change in loneliness, it was important to recruit individuals with lived experience of loneliness. Therefore, the purposive sampling frame selected individuals who had stated they were sometimes or always lonely in response to a single item on the questionnaire which asked: 'Do you feel lonely?' Using data from the Wave One sample, SPSS was used to select randomly ten cases in North Wales and ten cases in South Wales. The rationale for the sample size was to enable a detailed case-by-case analysis to identify the meanings behind loneliness for each participant.
Potential participants were contacted by letter inviting them to take part, along with an information sheet outlining the study and what participation involved; this was followed up by telephone a week later. Of the 20 people contacted, 11 participants aged 67-84 agreed to an interview (six women, five men). Nine people were unwilling or unable to participate, could not be located, had died or were too confused to be interviewed.
Ethical approval and permission to conduct the study within the Betsi Cadwaladr and Abertawe Bro Morgannwg University Health Boards were granted by the North Wales Research Ethics Committee and the All Wales Primary Care Research Management and Governance Office.

Interviews
Interviews were conducted between February and April 2013 in participants' homes. Interviews were audio-recorded and transcribed verbatim. Interviews took, on average, between 40 and 75 minutes. Two interviews were cut short, one due to an unexpected visitor and another due to the participant's distress.
Data were collected using narrative interviews. Narrative interviews centre on the stories participants tell (Kvale and Brinkmann, 2009) so were appropriate for a study aiming to explore the lived experience of loneliness, stability and change. Narrative interviews enable the participant to engage in an evolving conversation in which the participant and the interviewer co-produce the meaning of the events and experiences (Mishler, 1995). They begin with a broad opening question; in this study, participants were invited to talk about changes that had occurred in their social relationships as they aged, in particular how they had experienced loneliness and/or social isolation, and how they had coped with loneliness and/or isolation. Participants were allowed to talk with minimal interrruptions, other than points of clarification. It was anticipated that since the initial interview, some individuals might have experienced a change in levels of loneliness. However, at the time of the qualitative interview, there was no information available on exactly how much time had passed between interviews. Therefore, a decision was taken to focus on changes experienced in the previous year. This information was used to group individuals according to their loneliness trajectory in the group analysis. A topic guide using open-ended questions, developed from the research literature, was used to guide as an aide-mémoire.
Participants were asked to reflect on their experiences and understandings of loneliness and social isolation. Handwritten notes were taken during the interview to record observations of the participant's mood, as well as questions or issues raised by the participant and areas which needed further exploration. After each interview, interview notes were expanded before verbatim transcripts were produced for each audio recording.

Data analysis
NVivo 10 (QSR International, Melbourne) was used to assist the analysis and manage the data. The data were analysed using Interpretative Phenomenological Analysis (IPA) (Smith et al., 2009). IPA was chosen as the study was concerned with exploring and understanding participants' subjective experiences of loneliness, the meanings they ascribe to those experiences, and how they thought and acted in response to their experience. Adopting a case-by-case approach, each transcript was analysed in detail before moving on to the next case. Using the process outlined by Smith et al. (2009), the transcripts were coded using a mixture of theory-driven deductive codes and data-driven inductive codes, as well as in vivo codes. The deductive codes drew on themes from the existing literature, including those on coping strategies (problem, meaning and emotion focused), and the data-driven inductive codes included upward and downward comparisons. Loneliness trajectories were also grounded in the data, based on participants' response to the question: 'Are you more or less lonely than a year ago?' Those responding that they were less lonely than a year ago were categorised as decreasing (N = 5), those who responded they were lonelier than a year ago were categorised as degenerating (N = 2), while those participants who stated their level of loneliness was unchanged were categorised as stable (N = 4). At this stage, questions were asked of the data to identify conceptual similarities between cases (i.e. similarities and differences in coping strategies between those on different loneliness trajectories) (see Table 1).
Themes with a weaker evidential base were put to one side and later re-examined and used to highlight the diversity of the experience of loneliness and social isolation. Themes with a stronger evidential base were regrouped and reordered into themes and sub-themes. Each theme was given a short description which linked it to relevant segments of text (Table 2). A second independent researcher then checked the codes.

Results
The analysis of the narrative interviews found that older adults utilised a range of coping strategies to manage or ameliorate loneliness, many of which were consistent with those identified in the literature. The results are framed around the loneliness trajectory and the coping strategies used by individuals on that trajectory. Four broad coping strategies were identified: (a) problem-focused coping, (b) meaningfocused coping, (c) emotion-focused coping, and (d) social comparison. Differences in the use of strategies for individuals on different loneliness trajectories are highlighted, and the implications for the loneliness trajectory are discussed.

Decreasing loneliness
Five participants reported a decreased sense of loneliness over the preceding year. The group-level analysis showed that individuals on a decreased loneliness trajectory had utilised a wider range of problem-, emotion-and meaning-focused coping strategies than their peers on different loneliness trajectories. Indeed, their narratives indicated a determination to overcome loneliness: Talking to strangers • Side-effect of being alonecan't stop talking • Talk to people in the shops • I found, mind, you'll probably discover it now, unless my daughter, [daughter] calls here, I don't see anybody in the week, but when I meet somebody, I can't shut up. • I try then to say hello to everybody, you know, when you go down the shop. • I find myself sometimes actually starting a conversation with someone just to have someone to talk to you.

Technology
• Technology as a way of alleviating loneliness • My grandson talked me into an iPad for Christmasso I bought one of those, so I'm on the internet and I can get in touch with anyone. This determination to overcome loneliness was evident in the range of coping strategies utilised by older adults on a decreased loneliness trajectory. Strategies employed included a range of problem-focused coping strategies such as returning to social groups that they had attended in the past: One older woman living, in a rural setting, gathered a group of women who were also widowed or divorced with whom to socialise: I've even got a club together, a lot of us go out for a meal every week, or every fortnight, you know, down to local restaurants or to try a restaurant, you know? They've got no husbands a lot of them, it's all women, yeah. Just one of us, I think, has got a husband. I think we're all either separated or divorced. (Female,73,widowed) For this woman, the formation of new social networks comprising people in similar circumstances was a successful coping strategy as it helped to normalise her new situation as a widow in a way that socialising with couples may not have. In contrast, others joined new groups or took up new activities: I think I'm less lonely because of the contacts, and the course I'm doing, I think that's made a difference; … I look forward to it so much, just making contact with other people. (Male,78,separated) This male participant had experienced a shrinking social network due to retirement, relocation and the end of his marriage. The lack of an existing network meant that he joined several interest groups and started a university course as a way of facilitating social contact and forming a new social network. However, not all of his problem-focused coping strategies were successfully employed. For example, another approach he had employed was to instigate conversations while out shopping. Indeed, he was explicit in noting that his sole motivation in starting conversations with strangers was the hope of establishing some social contact: Although this approach instigated conversations, it did not lead to any meaningful connections. Yet this man's narrative suggested that he also perceived this strategy as an insurance against future loneliness: Some people they just give up and I don't intend doing that … You've got to keep trying to make contact with people. (Male,78,separated) Revisiting social groups or joining shared interest groups were successful problem-focused strategies that increased the likelihood that participants would find others with similar interests, increasing the possibility that they would alleviate loneliness. In contrast, although chatting to strangers in shops initiates some social contact, it is not a successful problem-focused strategy as it creates a fleeting connection rather than a meaningful one. As will be discussed in relation to the other loneliness trajectories, some other problem-focused strategies were also unsuccessful. Another coping strategy used by individuals on a decreased loneliness trajectory was meaning-focused coping, which drew on the individual's beliefs and value systems (Folkman, 2008). For a small number of participant's faith was identified as an important strategy for managing and alleviating loneliness. Narratives revealed that church congregations often provided an essential source of comfort and support in a time of crisis: If I didn't have my religion, I don't know what I would have done, to be honest with you … saved me, I think I'd been a lost soul. (Male B,72,divorced) Faith also provided an important source of social contact and a sense of continuity: I don't have to wait for anybody to find out who I am or anything, you know, we are always with one another, like you know … We do look after one another like, you know … They'll ring me every day, whoever, all my friends. I've got family down here, all [Jehovah] Witnesses, down the road, and we just see each other all the time, like, you know, and we socialise a lot. (Female A,67,widowed) The support offered by faith groups helped individuals cope with loss and loneliness, while the social interaction provided by church social events provided an opportunity to meet up with others who were 'on the same wavelength'. For these individuals, a shared belief system not only connected them to a wider religious community, but it also provided a sense of belonging. This sense of belonging was instrumental in helping them cope with loneliness.
Participants on a decreased loneliness trajectory were also found to have used social comparisons, in particular downward social comparisons. Downward social comparisons are a regulative coping strategy that can alleviate emotional distress resulting from loneliness (Ferring and Hoffmann, 2007). This was supported in the narrative of a recently widowed woman. The woman compared her ability to do the things she enjoyedsuch as helping at a lunch club or dinner with friendswith that of her sister, who was housebound due to deteriorating health. She also made comparisons, comparing her situation and that of her mother, who had cared for her father (who had dementia), noting that her mother was 'not a widow … not a wife'. The frequent comparisons she made between her life and those she perceived to be worse off appeared to remind her that she had been fortunate with her relationship with her husband, noting: We had a good innings together, so you've got to look at that as well as a plus haven't you. (Female,73,widowed) For this woman, downward comparisons acted as a defence mechanism that enabled her to adapt to life without her husband. Although she reported a decrease in loneliness, she was the most prolific user of downward social comparisons. This may indicate that this was a strategy the woman has used previously to promote wellbeing, or it may mean that despite reporting decreased loneliness, she was still adapting to the loss of her husband and her new situation.
Only one participant on the decreased loneliness trajectory used both downward and upward social comparisons. When speaking about his first wife, this male participant made a downward comparison, comparing her actions and experience of loneliness to his intended actions: She [first wife] was in this place on her own, and she wouldn't go out shopping, you know, because she had a heart condition and she was a bit worried … So that is my experience of someone else being very lonely, you know, I mean, she apparently she used to just not even bother to go to bed in the end, she'd sit in the chair all night long, and watch television. That's pathetic, isn't it really? When you think about it, but then that's how some people are, they just give up, and I don't intend doing that. (Male,78,separated) However, he also referenced one upward comparison when he compared his loneliness to that of a lonely widowed neighbour: I said to this woman that's lost her husband here because he died very suddenly of a heart attack, I said 'I don't like saying this but to be honest it's worse when your partner leaves you because they're still around', 'oh where's your husband?' well he's not going to come back. (Male,78,separated) The breakdown of this man's second marriage had a negative impact on his self-esteem and confidence. His use of social comparison appears to reflect his continued emotional loneliness and suggests that he was adapting to his newly single status. Although this man used a combination of upward and downward comparisons, he reported decreasing loneliness, suggesting that his use of downward comparison was being used as an adaptive mechanism to alleviate loneliness.
This was supported to some extent in the narrative of two females who had both been widowed for several years and made only one reference to a downward social comparison. The first woman compared her situation, having a large family around her, to people who were socially isolated: I feel sorry for those people who are socially isolated, I know people who are on their own, and that's it, you know, but no it's the opposite, it's the opposite for me. (Female A,68,widowed) The second older woman, when reflecting on how she copes with loneliness, noted: Participant: We've got lovely memories, and that's what you live on. When you're lonely, that's what you think about. Interviewer: And that doesn't make the feelings of loneliness worse? Participant: No, no, to think about the good times that we had, no, it just makes you feel good that you had them, I mean some people never get memories like that, I mean I'm lucky. (Female,84,widowed) This quote suggests that as the individual adapts over time, their use of downward social comparisons declines. Downward social comparisons appeared to remind participants that despite being lonely, they were more fortunate than most people. This appeared to have a positive impact, helping them adapt to the changes imposed on them. Indeed, based on the evidence presented, it could be argued that downward social comparisons have an adaptive purpose, allowing older people to recover from loneliness.

Stable loneliness
Four individuals reported that their loneliness had remained stable or unchanged in the preceding year. The coping strategies employed by individuals on the stable loneliness trajectory included emotion-focused and, to a lesser extent, problem-focused coping strategies. Although to some extent, participants on all three loneliness trajectories utilised emotion-focused coping strategies to varying degrees, for participants reporting stable loneliness, emotion-focused coping strategies were a primary strategy. These fell into two key dimensions: lowering of expectations and distracting from loneliness.
The lowering of expectations was a strategy used by individuals to explain why contact with their family networks was less frequent than they may have desired: Participants using this strategy all had dispersed family networks, and each had developed an explanatory framework in which they could make sense of infrequent contact with their family. Some participants justified infrequent contact by citing the geographical distance from their family, while others referred to the busy lives of relatives.
Distracting from loneliness was another coping strategy used by all participants regardless of their loneliness trajectory. However, for individuals on the stable and degenerative isolation trajectories, it was a key strategy adopted to manage the negative feelings of loneliness. Participant's narratives revealed a range of activities were employed to distract from loneliness and isolation, many of which involved engaging in solitary pursuits such as gardening: Other participants read or completed puzzles to distract them from feeling lonely, although as indicated in the narrative below, it was not always successful: I do crosswords, magazines, keep myself busy. It passes the time. And I go out, I go for rides in the car, again, it's lonely. (Female B,68,widowed) Some individual's used shopping as a temporary distraction from loneliness, with one woman describing it as 'things, to fill your loneliness' (Female A, 68, widowed).
However, for many individuals television was the primary form of distraction, as illustrated below: And I watch the television a lot more now than I used to, put it on first thing in the morning and not saying I watch it all day but I'm back and fore to it all the time, to have something to do. (Male,84,widowed) Some participants turned to alcohol to distract themselves and to regulate negative feelings. This point is illustrated by a divorced male participant who, when asked how he managed loneliness, noted that: Well, [I] go and have a drink … I don't drink in the house, though, never, I'm not a big drinker in that sense, but I do … I do have a few pints when I go out. (Male A,72,divorced) Ultimately, although distraction can temporarily reduce the influence of loneliness, it does little to alleviate it in the long term unless some form of social activity is incorporated into the distraction activity. This point is illustrated by one male participant who believed that developing a new skill at adult education classes would help distract from his feelings of loneliness: In the beginning, it was for something to do, for the computers, but I got so used to it after then it's taken on a different meaning to me like, you know. I go for the company now; now we have a talk, it's good, you know. (Male B,72,divorced) As previously noted, distracting from loneliness is an emotion-focused coping strategy; however, for one man, this strategy had unintended benefits in the form of widening his social network. His relationship with other class members motivated his continued attendance at the classes, and this could also be considered a problem-focused strategy. It was evident from his narrative that he derived great pleasure from the computer classes, and they had taken on new meaning. The man noted that 'everybody tells each other their troubles' and this sharing of problems suggested that he felt he belonged to the group and was a valued group member. The fact that the classes had taken on new meaning suggests that what began as an emotion-focused strategy to distract from loneliness changed over time to a problem-focused strategy and eventually to a meaning-focused strategy; which, in turn, had a positive impact on his level of loneliness. The sense of belonging to the group implied acceptance and gave him a level of security that his social needs would be met (Cobb, 1976), which in turn contributed to his decreasing sense of loneliness.
Individuals on the stable loneliness trajectory used few problem-focused strategies. The analysis showed that strategies that were successful for some individuals on a decreased loneliness trajectory were less effective for others on a stable loneliness trajectory. This was illustrated by one childless widow reporting stable loneliness, who had returned to an old social group. For this woman, her motivation in returning to the group was to alleviate her loneliness. Yet, the approach proved ineffective as the woman noted a difference between the quality of friendships that she had with past group members compared to those she had with the current social group: Although I've got a lot of friends, my dear, dearest friends have all gone. Perhaps they were older than me, and they've died in their sixties or seventies sort of thing, but the friends I've made now in church and that are a different type of people, you know? You know, they're not my friends that I like to go out with, but they're friends there. Different people in different places, you know. (Female,78,widowed) These findings highlight the importance of the quality of social relationships and also that relationships need to be meaningful for the older adult if they are to alleviate loneliness.
The analysis also showed that only one participant on the stable loneliness trajectory used social comparisons, making an upward comparison, to compare her situation to that of a recently widowed friend: A friend of [husband] died, and we went to the funeral, and she was so, you know, and I said, I said to my sister 'I hope it doesn't hit her like it hit me', although they've got a family, so that makes a difference, she's got grandchildren. (Female,78,widowed) This participant had described how she had gone into a 'very dark place' in the aftermath of her husband's death. She appeared to be using upward comparison to justify and explain her own continued grief, suggesting that it would have been easier if she had children. In this instance, the use of an upward comparison may reflect a struggle to adapt to her loss. However, it could be argued that upward comparisons were contributing to the stability of loneliness, as the social situation was perceived to be unchangeable.

Degenerating loneliness
Participants reporting degenerating loneliness were found to have used the fewest number of coping strategies, focusing primarily on distracting from loneliness as described above. As with their peers on stable loneliness trajectories, those reporting a worsening of loneliness also engaged in solitary activities to distract themselves from loneliness: I feel lonely in the evenings but yes I overcome them, I sort of do something isn't it, I go and do some wardrobes or whatever, clean some wardrobes, put them all back in though, erm, yeah I just got to go and do something haven't I, to ease it. (Female,67,widowed) Even though all participants utilised emotion-focused coping strategies to some extent, for those for whom emotion-focused strategies was a primary coping strategy, this had implications for their loneliness trajectory. Although emotion-focused coping strategies, such as lowering expectations, regulated the negative emotions related to loneliness, when used in isolation it did little to change the quantity or quality of social contact. Indeed, there is some evidence that this strategy was adopted by older people who considered loneliness and the circumstances surrounding it to be unmodifiable. For example, the participants expressed a sense of powerlessness to change their circumstances: I've got used to it by now [loneliness]. Nothing has changed anyhow. I think you learn to live with that sort of thing, don't you? (Male,72,divorced) [It's] mainly realising that life's not the same, life has changed, you know, and I can't live it the way I was living it before. (Female,67,widowed) What was evident from the narratives of the two participants reporting a worsening of loneliness was that they were both struggling to adapt to their changed circumstances. For these participants, loneliness was insurmountable and their narratives were interspersed with references to the factors that were responsible for not only maintaining their loneliness but mediating the deterioration. To some extent both believed that their disabilities, physical and sensory, were contributing to the chronic nature of their loneliness. While the female participant also internalised her loneliness, attributing it to her perceived limitations in social situations: I can still talk to people, you know, and mix with people, socialise with people yeah, all right, I've never been good at that, that was my husband, he was, he was well known to mix and I was just sitting there but everybody knew me then, didn't they, through him. (Female,67,widowed) This inability to change the circumstances surrounding loneliness was reflected in the absence of other coping strategies, instead these participants focused their attention on distracting from their negative emotions; or, as in the case of the male participant, putting barriers in the way of solutions: I don't even try very hard, that's another thing, you want to try something to, you know, do it. But I'm looking at all the obstacles before I get there. (Male,72,divorced) There was also some evidence to suggest that emotion-focused strategies, when used alone, led to a pessimistic view of the future social world and loneliness into the foreseeable future: They say time will tell, but to me, time has gone. It's [the loneliness has] gone worse as time has gone on … Those feelings are going to be worse, the more disabled I get. (Female,67,widowed) This partially supports the argument made by Schoenmakers et al. (2015) that suggests that emotion-focused coping strategies have limited use in alleviating loneliness.
The analysis found no evidence of social comparisons in the narratives of the two individuals on a degenerating loneliness trajectory. Similarily, there were no examples of either participant utilising problem-focused coping strategies aimed at increasing the amount of social contact they had. The male participant did, however, speak about his use of technology, as as a means of social contact. Yet, he noted that he was using it solely to maintain contact with dispersed kin rather than as a way of connecting to a variety of social network members: My daughter in [town in England], er, she comes on Skype, she's got it on her phone so, … so they come on here pretty regular. (Male,74,divorced) This man's use of technology was consistent with that of a female participant reporting stable loneliness: My grandson talked me into an iPad for Christmasso I bought one of those, so I'm on the internet, and I can get in touch with anyone. (Female B,68,widowed) Although she noted she 'can get in touch with anyone', there was no evidence that she was using technology to develop new social contacts. Technological solutions are often posited as a potential solution to loneliness. However, using technology in such a passive way as a substitute for other forms of communication proved to be an unsuccessful strategy and, in the case of the male participant, may have served to enhance his loneliness.
Based on the evidence presented, it could be argued that that cognitive processes may be central to how an individual will respond and manage loneliness, with the nature of the coping response being influenced and shaped by the underlying perceptions of whether loneliness is modifiable or not.

Discussion
In this article, we have examined the role of coping strategies and social comparisons in shaping the loneliness trajectories of older adults reporting decreased, stable and degenerating loneliness. The analysis of data from interviews with 11 older adults has highlighted a complex picture of the strategies employed by older adults to manage loneliness. Those individuals reporting a decreased sense of loneliness utilised a wide range of coping strategies. In contrast, those on a degenerating loneliness trajectory used the fewest types of coping strategies.
Our analysis highlighted the key differences in the coping strategies and social comparisons employed by older adults on different loneliness trajectories. Research by Schoenmakers et al. (2015) has shown that differences in coping styles between those who transition out of loneliness and those who are chronically lonely stem from whether the individual perceived loneliness to be modifiable.
This finding is consistent with previous research which has acknowledged that perceived control is an important aspect of coping (Skinner and Zimmer-Gembeck, 2011). Similarly, Newall et al. (2014) found that perceived control was an important factor in persistent loneliness. Where individuals perceive loneliness events to be controllable, they are more likely to expect the coping strategies they employ to be effective in addressing the situation (Skinner and Zimmer-Gembeck, 2011). Locus of control relates to the extent to which an individual believes they have control over the outcome of their lives (Lefcourt, 1992;Rotter, 1966). Individuals with a high internal locus of control are more likely to perceive loneliness as controllable and therefore less of a threat. This assertion is supported by our findings which found that only those reporting stable or degenerating loneliness (chronically lonely) attributed their loneliness to unmodifiable personal circumstances (personality characteristics or physical/ sensory impairments). We argue that individuals who are confident that they can modify their circumstances to alleviate loneliness (those with a high locus of control) will employ a wider range of strategies, directing their efforts on problem-focused coping strategies or downward social comparisons to alleviate the problem. When loneliness is perceived to be unmodifiable, older adults believing they have no control over their circumstances (low locus of control) focus their efforts on emotion-focused coping strategies rather than problem-focused strategies. This has implications for their loneliness trajectory, as emotion-focused coping strategies are relatively ineffective in alleviating loneliness (Thoits, 1995;Schoenmakers et al., 2015). Indeed, we would argue that emotion-focused strategies, when used in isolation, may contribute to the stability of loneliness.
Similarly, downward social comparisons appear to provide an adaptive mechanism which can improve wellbeing in light of loneliness and loss. Social comparison theory argues that individuals who are experiencing distress can improve their subjective wellbeing by comparing themselves to others whom they perceive to be worse off than themselves (Wills, 1981). Indeed, Suls et al. (2002) argue that strategic downward comparisons can help older adults cope with a threatening situation. In this context, downward social comparisons could be considered to be a form of cognitive adaptation (Taylor, 1983;Gibbons and Gerrard, 1991), with participants using downward comparison to adjust to their situation.
The analysis found that downward social comparisons were only used by participants reporting a decreased sense of loneliness. There was nothing in the participants' narratives to suggest that downward comparison had been adopted previously to address difficult periods or situations in their lives. However, four participants spontaneously referred to downward social comparisons, suggesting that it is an adaptive technique used by participants to help them adapt to new situations. For example, participants who were recently bereaved appeared to make more frequent downward social comparisons than participants who had been lonely for a longer period of time. This finding is consistent with that of Beaumont and Kenealy (2004), who found that downward social comparisons were both adaptive and functional, enhancing quality of life among older adults.
Making downward comparisons to those perceived to be worse off can precipitate a transition out of loneliness, as the individual reappraises their circumstances in a more positive light (Mussweiler and Strack, 2000). This is important, as the way an individual structures experience is related to the way they think and act (Beck, 1989). Thus, a positive appraisal of the potential for changing social relationships will influence and shape the coping strategies employed.
Although this finding has the potential to be incorporated into loneliness interventions, more research is needed to assess whether downward social comparisons are a lifelong approach to appraising situations, or whether social comparisons can be taught or changed in later life. Also, we would argue that further research is required to examine whether specific characteristics are associated with downward comparison (sense of control, personality characteristics, resilience and competitiveness) and whether meso/macro factors influence social comparisons.
Our analysis indicates that the 'higher-order strategies' (problem, emotional and meaning focused) identified in previous research may not be as clear-cut as initially suggested. We found evidence that emotion-focused coping regulated negative emotions, thus enabling the individual to focus on problem-focused strategies to address loneliness (Folkman, 2010). This strategy, in turn, became meaning focused as the values and goals of the older adult were met (Carver and Connor-Smith, 2010;Folkman, 2010). This suggests that higher-order strategies are not distinct entities but are part of a dynamic process of coping strategies employed by older adults. This finding was consistent with those identified in previous research studies (Peplau and Perlman, 1982;Rokach and Brock, 1998;Pettigrew and Roberts, 2008;Schoenmakers et al., 2015).
The article uses data from Wales, however, the findings would likely be applicable for older adults in other countries. Coping is process orientated and contextual: different strategies can be adopted in different contexts or at different time, and strategies may also change over time (Schoenmakers et al., 2015). For older adults living in other countries, the choice of coping strategies may be affected and shaped by the individual's cultural background and the availability of support. Regardless of this, it is likely that the complexity of loneliness trajectories may mean that individuals need to draw on multiple strategies to address loneliness. For older adults, the management of loneliness may be dependent on the interpretation of the cause and the coping strategies adopted in the past to manage difficult periods during the lifecourse.

Limitations
The research outlined in this article was conducted in Wales, UK, using a small sample of older adults who had self-identified as lonely in response to a single survey item. The stigma associated with loneliness may mean that some people may have been reluctant to identify as lonely and therefore did not answer the loneliness question truthfully. However, the study also used the De Jong Gierveld loneliness measure to overcome this limitation, as it indirectly measures loneliness. Four people refused to participate in the qualitative research. We acknowledge that this may have resulted in selection bias, as those agreeing to participate may differ from those who refused. In this study, loneliness was a relatively new experience for all participants. However, there were three variations in terms of the loneliness trajectory (stable, increasing or decreasing). The lack of coverage of the five loneliness trajectories identified by Victor et al. (2009) was partly due to a sampling decision to only include individuals who had reported being lonely in Wave One of the CFAS Wales survey. As the qualitative sample did not capture any individuals for whom loneliness was a constant feature of their life, it is likely that the coping strategies discussed may not reflect those for whom loneliness has been a constant feature of the lifecourse.

Implications
The findings of the study suggest that the management of loneliness in later life may be dependent on several factors, including older adults' interpretations of the cause of loneliness. The interpretation of the causes of loneliness, in turn, shapes and influences the coping strategies employed. Interventions aimed at alleviating loneliness in chronically lonely older adults may need to focus on changing their perceptions of unmodifiable loneliness through the use of motivational interviewing or cognitive behavioural therapy, or by offering support in engaging in new social situations.
We conclude that for coping to be effective, a wide range of coping strategies need to be employed, including problem-, meaning-and emotion-focused strategies. Strategies used in isolation will have little impact on stable or degenerating loneliness. In particular, emotion-focused coping strategies, when used in isolation, are largely ineffective. Furthermore, older adults themselves must define both the problem and the most appropriate solution to address loneliness.
Financial support. This work was supported by the Economic and Social Research Council (grant number ES/I900993/1). The CFAS Wales study was supported by the Economic and Social Research Council (grant number RES-060-25-0060) and Higher Education Funding Council Wales as 'Maintaining Function and Well-being in Later Life: A Longitudinal Cohort Study'.
Ethical standards. Ethical approval and permission to conduct the study within the Betsi Cadwaladr and Abertawe Bro Morgannwg University Health Boards were granted by the North Wales Research Ethics Committee and the All Wales Primary Care Research Management and Governance Office.