Being cut off from social identity resources has shaped loneliness during the coronavirus pandemic: A longitudinal interview study with medically vulnerable older adults from the United Kingdom

Loneliness is a pernicious problem in older adulthood, associated with physical decline and isolation from valued social groups. However, the long-term evolving experiences of ageing, identity and loneliness have yet to be elucidated. We use a Qualitative Longitudinal Research interview approach with nine vulnerable older adults (Age mean = 79.4 years), in which five participants were interviewed twice between 2019 and 2020, and four participants were interviewed at three-time points from 2019 to 2021. This study aims to understand the unfolding experiences of ageing, social identity and loneliness during a prolonged period of social isolation during the Coro-navirus pandemic. A theoretically guided thematic analysis highlights that participants initially experience ‘ Categorisation as Vulnerable and Loss of Agency ’ and ‘ Shrinking Social Worlds ’ , leading to ‘ Undermining of Reciprocal Support ’ and ‘ Fears of Persistent Loneliness ’ . Findings suggest that interventions to ameliorate loneliness among older adults would benefit from addressing age-based stereotypes and emphasising the value of reciprocal contributions that older

five participants were interviewed twice between 2019 and 2020, and four participants were interviewed at three-time points from 2019 to 2021.This study aims to understand the unfolding experiences of ageing, social identity and loneliness during a prolonged period of social isolation during the Coronavirus pandemic.A theoretically guided thematic analysis highlights that participants initially experience 'Categorisation as Vulnerable and Loss of Agency' and 'Shrinking Social

| Loneliness experiences and cognitions
Loneliness is the distressing feeling experienced during perceived social isolation (Hawkley & Cacioppo, 2010).Loneliness is understood as distinct from social isolation, with emotional components such as the felt absence of desired personal relationships and cognitive components such as hyper-vigilance to social threats (Cacioppo et al., 2006).Furthermore, chronic loneliness can become a pernicious, self-perpetuating challenge; chronically lonely people report social anxiety and paranoia of others (Taube et al., 2016), which can then lead to poorer satisfaction with connections and further social withdrawal (Cacioppo & Hawkley, 2009).
On one hand, ageing does not inevitably lead to loneliness and older adults can show resilience to reduced social contact (Cohen-Mansfield, 2020).Indeed, evidence shows that older adults can value fewer social groups with greater emotional value (Carstensen, 2021) and can benefit from volunteering and increased neighbourhood activity after retirement, which may help to prevent loneliness (Bowe et al., 2021).On the other hand, one in four older adults in high-income countries experiences loneliness, with health and age-related declines as a key risk factor (Chawla et al., 2021;Martín-María et al., 2020).It is the chronic nature of ageing declines, leading to cyclical loneliness cognitions, which may further perpetuate chronic loneliness for this population.

| Current study: Ageing, social isolation, & loneliness
Given the complex challenge of loneliness for vulnerable older adults, this study employs a detailed exploration of how loneliness develops alongside ageing and social isolation.This study focuses on older adults with medical vulnerabilities during the Coronavirus pandemic.During the height of the pandemic in the UK, this population had a heightened risk of identity discontinuity as well as the risk of loneliness (Bowe et al., 2022;Polenick et al., 2021).Whilst many individuals appeared to weather the storm of the pandemic by relying on social, emotional and financial support of family (Stevenson et al., 2022), some older adults with medical vulnerabilities and limited social groups were particularly disadvantaged.The large-scale social disruption during pandemic created barriers to social identity resources and prolonged social isolation (Brooks et al., 2020).The social restrictions created by the UK Government advised those over 70 years of age and people with medical vulnerabilities to remain at home and distanced from other households (or 'in lockdown') in England from March-June 2020, a second lockdown from fifth Novembersecond December 2020, and a third lockdown from sixth Januaryeighth March 2021.Besides the national lockdowns, different regions in the UK variously entered 'tiered' systems of restrictions, in which restrictions on social activities were altered based on recorded levels of the coronavirus in the region, and in all stages of these tiered systems clinically and extremely vulnerable including people aged 70 years or older were advised to limit social contact, use of public spaces and to socially distance from others (Talbot & Briggs, 2021).
We explore the patterning of loneliness and social isolation for older adults during this 18-month period when social resources were scarce.We utilise a Qualitative Longitudinal Research (QLR) approach to capture the development and trajectories for a group of vulnerable older adults at up to three-time points.Finally, we use a theoretically driven thematic analysis to illuminate the interplay of theorised age-related identity changes and loneliness for vulnerable older adults as they unfold over the course of the pandemic.

| Design
Qualitative Longitudinal Research (QLR) is a distinct methodological paradigm, informed by the 'temporal turn' in the social sciences (Thomson & McLeod, 2015).In addition to sensitising researchers to the temporal orientations of their design, it highlights how chronic engagements with research contexts afford an appreciation of the diachronic evolution of social and psychological processes.Repeat interviews in particular offer the opportunity to capture intervening events and return to previous discussions and deepen understanding of participants' experiences (Vincent, 2013).In this study, we availed of the opportunities afforded by repeat interviews before and during the pandemic to track how a crisis served to disrupt the pre-existing social relationships of the participants and their experiences of loneliness as a consequence.Therefore, we were able to chart the unfolding identity processes occurring as a result of the pandemic.
The longitudinal research design provided an outline for data collection with Braun and Clarke's reflexive thematic analysis (Braun & Clarke, 2006;Braun & Clarke, 2019).This analytical approach provided a rigorous method for processing the qualitative longitudinal data (Braun & Clarke, 2022).In addition, Braun and Clarke (2019) encourage analysts to be specific about the theoretical frameworks that drive their interpretations.In this case, epistemologically, this study was a critical realist with an expectation that the longitudinal interviews would allow us to explore the relationships between experiences and self-concept over time, between participant's interviews (Archer et al., 2013).Theoretically, the analysis explored identity change processes from the perspective of the social identity tradition (Haslam et al., 2018) and social cognitive conceptualisation of loneliness (Hawkley & Cacioppo, 2010).
Longitudinal semi-structured interviews spanning 14 months elicited in-depth loneliness experiences of nine older adults who were identified as medically vulnerable.Participants were interviewed prior to the pandemic in late 2019-2020 and interviewed again during two pandemic lockdowns in 2020 and late 2020-2021 in England.

| Participants
Participants were recruited for a study prior to the pandemic in November 2019.The original study explored social connections in older adulthood, interviewing participants over the age of 50 years who had co-morbidities.Participants were existing clients of a partner organisation (AgeUK Exeter), who had sought practical or social support (e.g., befriending, help with applying for government benefits).The sampling strategy was purposive, selecting individuals who could discuss experiences of being socially and medically vulnerable older adults.Following nationwide pandemic measures, participants agreed to continue to be re-interviewed to understand the loneliness and wellbeing impacts on older people during the pandemic.
At time-point one and two, participants were nine older people (seven female, two male), ranging from 60 to 88 years old (M = 79.4years).All were white British, living in the South West of England and all but one participant lived alone.At time-point three we re-interviewed four of these participants (three female, M = 83.5 years).
Attrition at time three can be attributed to the nature of the vulnerable population, that is, due to death (n = 1), declining health or cognition (n = 2), or loss of contact (n = 2).Table 1 details participant demographic information.

| Interview procedure
The semi-structured interview schedule contained four topics; exploring social lives during the pandemic, resilience, support and the future.Questions are outlined in Data S1.Interviewer, one completed interviews at time-point one and two, was an employee of the partner organisation who supported recruitment and was known to all participants.Interviewer two completed time-point three interviews and was a researcher who had met participants prior to time-point one through rapport-building activities at the onset of the study.Neither interviewers were older adults.A timeline indicating the interview time points and the context of the English coronavirus restrictions is provided in Figure 1.

| Ethical considerations
This study obtained ethical approval from the Nottingham Trent University BLSS (Business, Law and Social Sciences) Ethics Committee before commencing data collection and updated ethical approval following the onset of the pandemic.Participants were fully informed and consented to each interview.
T A B L E 1 Participant demographic information and interview time-point participation.F I G U R E 1 A timeline of the messaging and national social restrictions for older adults during the study period.

| Data analysis
All interviews were audio-recorded and transcribed verbatim.Thematic analysis was guided by Braun and Clarke's (2006) six-step reflexive thematic analysis and driven by a consideration of (a) psychological conceptualisations of social identity change and loneliness and (b) temporal changes in experiences expressed across the participant sample.One main coder undertook analysis, creating approximately 1,500 initial codes using QSR NVIVO and two additional coders conducted regular critical consistency checks.
Coding involved (1) familiarisation with the transcripts (2) initial codes generated deductively utilising the full transcripts.Each coder explored a participant's data across time points to identify changes in connections and loneliness.At step (3) coders jointly developed themes, using social identity-informed theoretical reasoning to develop themes from the initial codes.Nine preliminary themes were developed, and in analysis step (4) themes were reviewed for fit with the coded data, theory, and the data emerging from time-point three interviews.In step (5), theme names were re-defined as fitting with four main themes and (6) the report was produced.

| RESULTS AND DISCUSSION
Four main themes were generated from the analysis (see Figure 2).Themes are ordered in terms of their temporal occurrence in participants' accounts of their experiences: beginning with their social vulnerabilities and progressing to capture how their social contact and loneliness experiences evolved over time.

| Theme 1: Categorisation as vulnerable and loss of agency
Ageing-related health changes typically undermined the participants' ability to engage with the groups and group activities which were valuable to their identities.While declining physical health need not entail poorer access to group resources (e.g., prior to the pandemic, one participant joined a support group which helped him to adapt to his illness) our medically vulnerable participants often experienced frustration as their illnesses impacted their usual ways of connecting with others.Notably, these changes were not usually within participants' control and therefore, when participants discussed social contact with groups, they discussed this in terms of a loss of agency and limitations to their social contact.As we see below, Christine faced driving restrictions which prevented her from visiting friends and activity groups: Christine, 78 years, T1.
I'd like to go out more in the evenings because there are things on in the evenings but I can't do because I am not allowed to drive at night you see.So, I just don't get there and I have to be in, even Cinderella's better off than me she has to be in by midnight I've got to be back home at four o'clock in the afternoon.
The participants suffered from a range of ailments and as a result, often were categorised by others as an older adult.This was particularly evident at time points two and three when their age-related vulnerability to COVID-19 often became an issue.As Ann highlighted below, the people in the participants' social networks often treated them in accordance with negative stereotypes of older age, for instance with expectations that they had reduced intelligence and capabilities.The participants then positioned other people as lacking in understanding, and this could act as a barrier to connecting with new groups.
Ann, 88 years, T3.Moving through to time points two and three, participants experienced more restricted contact with others during the pandemic.For prolonged periods, participants were unable to leave their homes.In order to cope they constructed new routines around hobbies or pre-arranged telephone contact with others.However, telephone contact was often a poor replacement for physical group meetings.Moreover, by time-point two participants began to express fears that younger, more able members of their social groups were moving on and leaving them behind.
Indeed, participants were often let down by those who had arranged to call, which reinforced their sense of aloneness, as Jean reports: Jean, 80 years, T2.
It's getting harder for me because I am not seeing people you know…it can be a whole week…I am waiting for this call 'cos I thought they'd ring and if they can't get me they ring again but they obviously didn't because I waited there and nobody rang me so I had nothing that week either…I could have stayed by the phone all day.
Participants often refocused on undertaking independent activities such as card making, gardening, or making items to donate to charity.Such examples had a dual purpose of filling time while alone and allowing participants to keep salient a valued group identity for instance as a gardener (e.g., gardening was important for Christine who at time-point one noted 'My garden is everything to me').However, by time-point three these replacement activities had become inadequate in satisfying social needs.In the face of enduring social isolation, participants spoke of needing personal resilience through 'digging deep', or 'not allowing the miseries to last'.For John, below, such individual activities now ironically emphasised his feelings of loneliness and poorer well-being.
it's quite lonely because I like being with people obviously not just doing things on my own.It's the practical things at home I can, I'm not.A lot of things that have kept me busy but they not necessarily make you feel better.
This theme illustrated that ageing and ill health had restricted participant's access to social groups, and left them vulnerable to negative stereotyping which, in turn, reduced participants' desire to connect with others.These restrictions were compounded by the pandemic.Participants were left dependent on others to make contact with them, hence losing agency in their social interactions, culminating in low mood and boredom.

| Theme 2: Shrinking social worlds
Ageing processes were reported to trigger a shrinking of participant's social worlds.Participants felt increasingly distant from their social groups, and this led to feelings of social and emotional loneliness.These processes were present from the beginning of the research period, prior to COVID-19.For instance, participants such as John experienced 'othering' from friends due to his worsening health; for John, his wheelchair became a symbol of the difference between him and his friends: John, 60 years, T1.
A few weeks' ago, I went to his [friend's] surprise birthday party and I wasn't very well so I was in my (wheel)chair and a few people that are friends…they talked to everyone else but they don't talk to me and I don't know what their problem is and I haven't heard or seen them since.
By time point two, pandemic lockdown measures enforced a further loss of social contact.Seven of the participants relied on a network of support personnel who provided regular practical support such as shopping and gardening.The halt of support services raised participants' awareness of the vulnerable nature of their social connections, throwing into relief how much they had relied on these service-based relationships.It was noted that while the contact was characterised as a keen loss, actual contact did not entirely cease with others.Indeed, participants received phone calls from others or offers of shopping from neighbours.However, as social contact was dramatically reduced for in-person visits, including a reduction in support visitors and a reliance on the telephone and digital contact, social worlds were experienced a smaller and more vulnerable to loss.For instance, Margaret was distressed when she realised how impoverished her networks were as she did not have a family to turn to when employed personnel were not available: Margaret, 81 years, T2.
I found that [the pandemic restrictions] absolutely devastating because I lost the whole care package.I lost Age UK, I lost my cleaner, I lost my gardener,…that was a point at which I realised that I was isolated…I'd no idea at that point how, what I was going to do for that having no family.
Some participants also ruminated on the decline of contact from friends or younger family during the course of the pandemic.This explicitly evoked loneliness for Mary, who noted that she received fewer phone calls as the pandemic progressed and attributed this to other people adapting 'to how it's going to be'.This was interpreted as a sign to Mary that other people had a lesser need for a phone call as they had 'adapted' to the circumstances, leaving behind Mary who still felt the need to receive phone calls.As a result, she experienced this as lonely: Mary, 85 years, T2: Interviewer: Did you find that people then phoned you more during the sort of first few days, first week or so and then as time's gone on less?Mary: Yes.That is exactly what I found.As time has gone on it's got less and less people have adapted to how it's going to be which isn't really very nice…it is just very lonely really.
Loss of contact was permanent for participants whose friends had passed away.Given the participants' advancing age, death among their peers could be a natural, though lamented, shrinking of social worlds.However, pandemic restrictions on funeral attendance removed the opportunity to commemorate losses to access to collective mourning.The loss of friends also meant that participants could struggle to maintain a positive sense of shared identity and this could also undermine their perceived value to their community.For instance, Patricia described feelings of worthlessness following the loss of her friend: Patricia, 70 years, T2.
With this lockdown was when my friend passed away and I couldn't get to her funeral that really I went through a bad time then…that was very hard, very, very hard indeed you know but um…I've been here reminiscing about my past and like I told you the other day that really, I feel that I've wasted my life.
This theme displays how processes of ageing reduced participant's social worlds which exacerbated their perceived isolation during the pandemic.The pandemic then further impoverished their networks, emphasising the distance between the participants and their valued groups and eliciting loneliness.

| Theme 3: Undermining reciprocity
The participants emphasised that both ageing and the pandemic undermined social support within valued groups by reducing opportunities for reciprocal exchange.At time-point one, several participants described their homes as a place where friends or family could drop in unannounced.In effect they served as an informal gathering point for their groups where they were available for others in need and, in turn, believed others would be available for them in a crisis: Janet, 86 years, T1.
I am aware that if I needed somebody you know there are various people I would phone and if they were doing something or some crisis or something you know if I sit here long enough, I know somebody will drop in.
At time point two, support from informal connections was even more important for the participants who by then faced COVID-related restrictions on their activity.However, at this stage, these offers of help were often characterised as unilateral and non-reciprocal, positioning the recipient, the older vulnerable adults, as vulnerable and needy.Consequently, participants could feel burdensome in these relationships.Participants also found that the support they received from others could be insensitive to their needs, for example, participants received unwanted food deliveries or were invited to participate in Zoom calls they did not enjoy.This resulted in a perceived lack of belonging, recognition, or empathy from others.As in the extract below, David found himself uncomfortable with family Zoom calls because they did not suit his conversational style and, as a result, he felt unable to participate with his family: David, 87 years, T2.
Interviewer: What do you think of that [zoom] then?
David:I hate it.I really hate it…I haven't got anything to say I just sit there and they chat away and I haven't got anything to say really.
At time-point three, most participants had made efforts to establish some form of independence from the help of others.This tendency was particularly evident in relationships that were relatively superficial and lacked a common bond of identity.For instance, in the extract below Christine described a neighbour offering to go shopping for her.At time point two, Christine had accepted this neighbour's help, but at the third time point revealed that she had moderated how much she allowed them to buy her.One interpretation of this is it illustrates how being helped was experienced by older adults as a delicate balance between feeling supported and being burdensome: Christine, 78 years, T3.
[neighbour] wanted to know if I wanted shopping which was lovely….and I didn't like to bother them too much, you know…they did pick up a couple of lots of shopping but then you see I'd got the hang of the online shopping and when I was asking them to buy me a couple of bits I really needed more, but didn't really like to ask them to carry me a whole load of shopping you know.But, it's chicken and egg really.
Christine's account of not accepting neighbourly help may have indicated a developing resilience to the pandemic, as she was able to assert her independence through online shopping.However, Christine's account did not focus on her skills and self-reliance, instead, she referred to a 'chicken and egg' relationship, a cyclical, interpersonal process in which she had chosen not to burden her neighbours with carrying her full shopping load, and therefore, the help offered was inadequate.In turn, this made her less comfortable asking for additional help.The independence here is characterised as a loss of a 'lovely' offer, in favour of a more solitary solution.Thus, our interpretation was that over time, non-reciprocal support gave way to independent and solitary interactions, leading Christine to the conditions of loneliness.The lived experience of loneliness was explored in the final theme.

| Theme 4: Fears of persistent loneliness
When participants considered their futures, they often conveyed fears of being forgotten, having lost confidence, or being left behind by others and often contemplating futures alone.For instance, Mary (85 years) ruminated on the possibility of suffering the fate of her neighbour whose body was not found until days after his death.At time points two and three, in particular, many participants conceptualised the longer-term future as irreversibly altered with permanent barriers to connectedness.Moreover, participants justified these worries in terms of their age and mortality.
For example, Ann suggested that people might never hug again within her lifetime.Further, many participants considered the likelihood that their health may deteriorate, they may leave their homes for institutional care, or they may not live to see-through activities.All contributed to a sense of social fatalism and loneliness.
One thing I feel personally is, I haven't got many more years to go…I'm going to be another year older and getting decrepit…at eighty-eight, you just don't know what's going to be around the corner for you, healthwise.I'm an optimist, but I think you have to be realistic…I've missed so much time in fun.
Conversely, by time-point three (a year after the onset of the pandemic) two dissenting voices expressed positive expectations of the future, showing greater faith in their activities being restored.In one instance, David envisaged an eventual return to normality.David had experienced falls by time-point three but had purchased a mobility scooter, maintaining an active lifestyle and continuing to visit shops.Margaret at time-point three had restoration of her health after a period of hospital treatment.The restoration of usual activities evidently allowed both participants to access and enact positive social identities that had been discontinued in lockdown.Margaret in particular was hopeful for the eventual continuity of the social world she had known before the pandemic.
Well, I'm fairly optimistic about the future.I mean I've always been pretty resilient about things.I must admit that the last few months have been very difficult, but we'll get over this…because the first lockdown was so frightening for me because of the additional health problems, this time around, it doesn't feel as bad.
belonging in their communities (Bowe et al., 2022).To learn from this, a key site for intervention for older adult loneliness may be to re-establish and encourage reciprocal support exchange within older adults' social groups.
Our sample, though a suitable size for affording unique qualitative longitudinal insights into the unfolding experiences of loneliness, is too small to afford generalisations for all older people.Our design also suffers from attrition and the more vulnerable among the participants are less represented in the final stages.Moreover, all participants are white British living with acquired medical vulnerabilities which limits our ability to generalise findings to other socially vulnerable groups.During the height of the pandemic, the UK experienced some of the worst infection and death rates in the European Union and these findings may uniquely reflect the interplay between cultural categorisations of older adults in an environment in which older people faced particularly lengthy social restrictions (Søraa et al., 2020) and the development of loneliness (Polenick et al., 2021).
With the legacy of COVID-19 social isolation policies, loneliness in older people is a public health issue.Whilst the social isolation protection policies have been warranted, they may have instigated long-term poorer well-being through loneliness in older adults, alongside processes of negative ageing categorisation, reduced social worlds and loss of reciprocal support.The loneliness-related cognitions are likely to have longer-term consequences for the fracturing of social relationships.As we move through and beyond the pandemic, we emphasise the social needs of the most vulnerable members of society and advocate for positive ageing identities and reciprocally supportive groups in order to help older adults to overcome loneliness-related distress.
Worlds', leading to 'Undermining of Reciprocal Support' and 'Fears of Persistent Loneliness'.Findings suggest that interventions to ameliorate loneliness among older adults would benefit from addressing age-based stereotypes and emphasising the value of reciprocal contributions that older Time point 1: Face-to-face interviews took place in November 2019-January 2020 before the pandemic.The interviews, (mean length = 46.11minutes, range 23-62), took place either at an Age UK centre or in the participant's home.Time point 2: Interviews took place in May-June 2020 shortly after the first UK national lockdown.The interviews (mean length = 41.82 minutes, range 23-60) took place on the phone (n = 5) or socially distanced outside participant's homes (n = 4).Time point 3: Telephone interviews (mean length = 39.9, range 15-49) took place in January 2021 during the third UK national lockdown.
Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/casp.2724 by The Open University, Wiley Online Library on [01/08/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Schematic map of themes and subthemes.
I'll tell you something that happens when you get old, some people start talking to you as though you're a bit backward and you don't know what you're talking about.They think if you're old, you're losing your marbles!My family don't, because they know that I'm definitely on top of anything they can say to me.But umm, sometimes strangers will talk to you as if you don't know what you're on about.