Typologies of loneliness, living alone and …...Typologies of loneliness, living alone and socialisolation,andtheirassociationswith physical and mental health KIMBERLEY J. SMITH*

Typologies of loneliness, living alone andsocial isolation, and their associations withphysical and mental health

KIMBERLEY J. SMITH* and CHRISTINA VICTOR†‡

AbstractThe relationship between living alone, loneliness and social isolation, and how theyare associated with health remain contentious. We sought to explore typologiesbased on shared experiences of loneliness, social isolation and living alone usingLatent Class Analysis and determine how these groups may differ in terms of theirphysical and mental health. We used Wave of the English Longitudinal Study ofAgeing (N = ,; mean age = .) and responses to the University of California,Los Angeles (UCLA) loneliness scale, household composition, participation insocial/societal activities plus frequency of contact with friends, family and relativesfor the Latent Class Analysis. The optimal number of groups was identified usingmodel-fit criteria. The socio-demographic characteristics of groups and health out-comes were explored using descriptive statistics and logistic regression. We iden-tified a six-cluster typology: Group , no loneliness or isolation; Group ,moderate loneliness; Group , living alone; Group , moderate isolation; Group, moderate loneliness, living alone; and Group , high loneliness, moderate isola-tion (with high likelihood of living alone). Groups experiencing loneliness and/orisolation were more likely to report poorer physical and mental health even afteradjusting for socio-demographic confounders, this was particularly notable forGroup . Our results indicate that different typologies of living alone, lonelinessand isolation can be identified using data-driven techniques, and can be differen-tiated by the number and severity of issues they experience.

KEY WORDS – isolation, loneliness, living alone, typologies, health, older adults.

* Department of Psychological Sciences, School of Psychology, University of Surrey,Guildford, UK.

† Institute of Health, Environment and Societies, Brunel University London,Uxbridge, UK.

‡ Department of Clinical Sciences, Brunel University London, Uxbridge, UK.

Ageing & Society , , –. © Cambridge University Press This is an Open Access article, distributed under the terms of the Creative Commons Attributionlicence (http://creativecommons.org/licenses/by/./), which permits unrestricted re-use, dis-tribution, and reproduction in any medium, provided the original work is properly cited.doi:./SX

terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0144686X18000132Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 24 Aug 2020 at 01:22:12, subject to the Cambridge Core

https://www.cambridge.org/core/terms

https://doi.org/10.1017/S0144686X18000132

https://www.cambridge.org/core

Introduction

Are loneliness, social isolation and living alone the same or different?Researchers and theorists have long hypothesised that loneliness, social iso-lation and living alone are separate constructs despite overlapping features(Victor et al. ). The identification of typologies of loneliness and socialisolation has long been of interest to researchers (Townsend ; Tunstall; Weiss ). Identifying typologies tells us how these concepts arerelated and can help us to identify better those people who may be athigh risk of the adverse consequences of loneliness and social isolationsuch as poor health and wellbeing (Heinrich and Gullone ;Luanaigh and Lawlor ; Valtorta et al. b). Andersson () pro-posed a four-fold typology to summarise the relationship between lonelinessand social isolation: neither lonely nor isolated, isolated but not lonely,lonely but not isolated, and both lonely and isolated. However, littleresearch has explicitly examined whether these typologies may reflect thelived experience of older adults nor how different typologies might linkwith physical and mental health.

The concepts of loneliness, social isolation and living alone

Loneliness has been conceptualised in a number of different ways: as a dis-crepancy between a persons desired and perceived quality and quantity ofsocial relationships (Walton et al. ), a perceived deprivation in socialcontact (Townsend ), perceived social isolation (Hawkley andCacioppo ) or a lack of people with whom to share emotional andsocial experiences (Rook ). On the other hand, social isolation hasbeen conceptualised as the measure of a person’s integration and meaning-ful communication with their community, family and friends (Victor et al.), the objective experience of being alone (Hawkley and Cacioppo), a lack of meaningful social ties (Lubben and Gironda ) or alack of integration with social networks (Rook ). Whilst these defini-tions suggest conceptual overlap, research shows there is only a moderateassociation between loneliness and social isolation (Golden et al. ).Social isolation and loneliness can also be theoretically differentiated

from living alone, a simple enumeration of household size (Victor et al.). The evidence for the relationship of loneliness and social isolationwith living alone is mixed, with studies variously reporting no associationbetween loneliness (Zebhauser et al. ) or social isolation (Larson,Zuzanek and Mannell ) and living alone, high levels of loneliness/iso-lation amongst those living alone (Tunstall ) or that living alone is astrong risk factor for loneliness (Sundström et al. ). Despite possible

Kimberley J. Smith and Christina Victor



https://doi.org/10.1017/S0144686X18000132


theoretical differences, living alone has been included in measures of socialisolation (Shankar et al. , ; Steptoe et al. b; Victor et al. )rather than being evaluated as a potentially separate category.

Loneliness, social isolation, living alone and health

Existing theory and evidence suggest that a more nuanced examination ofthe relationships between these three concepts could be useful, especially indeveloping our understanding of the association of these issues with healthand wellbeing outcomes. Previous qualitative research provides us withrich and meaningful data on the lived experiences of loneliness, social iso-lation and living alone. They indicate that loneliness and social isolation arecomplex individual experiences which have different perceived causes andconsequences (Cloutier-Fisher, Kobayashi and Smith ; Dahlberg). However, quantitative work has typically defined loneliness andsocial isolation in predetermined, binary and stringent ways (Shankaret al. , ; Steptoe et al. b) that may not accurately encapsulatethe complexity of these experiences.Previous work argues that researchers, policy makers and practitioners

need to appreciate how we define and measure loneliness, isolation andliving alone so we can better identify how these experiences might belinked with important health and wellbeing outcomes (Perissinotto andCovinsky ; Valtorta and Hanratty ; Valtorta et al. a).Loneliness, living alone and social isolation are all independently asso-

ciated with similar health outcomes such as depression (Alpass andNeville ; Dean et al. ; Hawthorne ), poorer cognitive func-tioning (Bassuk, Glass and Berkman ; Cacioppo and Hawkley ;Zunzunegui et al. ), cardiovascular disease (Orth-Gomér, Rosengrenand Wilhelmsen ; Valtorta et al. b), poorer self-rated health(Kharicha et al. ; Lee et al. ; Nummela, Seppänen and Uutela), worsened physical functioning (Iliffe et al. ; Kharicha et al.; Perissinotto, Cenzer and Covinsky ) and lifestyle (Alpass andNeville ; Davis et al. ).When researchers examine loneliness and social isolation with health sim-

ultaneously results are often mixed, with studies reporting loneliness andsocial isolation being associated with different health outcomes (Cornwelland Waite ; Elovainio et al. ; Holwerda et al. ; McHugh et al.; Shankar et al. ; Steptoe et al. b). However, there is a lackof work that examines how the experiences of living alone, loneliness andsocial isolation combine, and how these groups might be linked with health.One way to determine whether meaningful typologies based on shared

characteristics of social isolation, loneliness and living alone can be

Loneliness, isolation, living alone and health



https://doi.org/10.1017/S0144686X18000132


uncovered quantitatively is to use data-driven methods such as Latent ClassAnalysis (LCA) as a method to identify distinct sub-groups based on socialrelationships in older populations (Burholt, Dobbs and Victor ;Ellwardt, Aartsen and van Tilburg ). The first aim of our study was toidentify groups of older people based on shared characteristics of loneli-ness, social isolation and living alone using LCA. The second aim of thisstudy was to determine how each of these groups differed in terms ofhealth and socio-demographic characteristics.

Methods

Data-set and participants

This study utilised data from Wave of the English Longitudinal Study ofAgeing (ELSA). ELSA is a longitudinal prospective cohort study of adultsaged or older that has been ongoing since (Wave ). Baseline par-ticipants were sampled from eligible participants who had taken part in theHealth Survey for England (HSE) in either , or (Steptoeet al. a). Core data collection for the ELSA study takes place everytwo years. As well as following-up baseline participants, refreshment sam-pling (i.e. sampling new participants from HSE) for ELSA has taken placeat Waves , , and . Our analysis utilises Wave , the most recently pub-lished wave of ELSA (data collection –). A total of , peopletook part in this wave. We excluded people aged less than (N = ),those who had a proxy answer on their behalf (N = ) or had incompletedata for indicators of loneliness, social isolation and living alone (N = ),giving an analytic sample of , (mean age = .; % female).

Indicators of loneliness and social isolation

The indicators of loneliness and social isolation used for our analysis werebased upon previous work from ELSA exploring these topics. Our indicatorsof loneliness were the three questions from the three-item University ofCalifornia, Los Angeles (UCLA) Loneliness Scale (Hughes et al. ):How often do you feel you lack companionship? How often do you feelleft out? How often do you feel isolated? Responses were: hardly ever/never, sometimes or always.We used the social isolation index developed for use with ELSA (Shankar

et al. ) based on the five items of:

. Household composition (living alone, living in multi-personhousehold).




https://doi.org/10.1017/S0144686X18000132


. Participation in societal/social activities (not involved/involved). Thisincluded participation in any social/recreational activities (e.g. eveningclasses, social clubs and sports clubs), societal (e.g. political parties ortenants groups) or voluntary activities.

. Communication with family (more than once a month/less than once amonth/no family).

. Communication with relatives (more than once a month/less than oncea month/no family).

. Communication with friends (more than once a month/less than once amonth/no family).

Socio-demographic characteristics

We examined the following socio-demographic characteristics: age, sex(male/female), ethnicity (white/non-white), marital status (married/notmarried/divorced or separated/widowed), employment status (employed/unemployed/retired), education (no qualifications/high school or college-level qualifications/higher education qualifications), net wealth minuspension (quintiles and /quintiles –) – an indicator of wealth used insimilar studies (Shankar et al. ).

Health-related outcomes (physical and mental health)

Functioning was assessed by calculating the number of issues with the follow-ing activities of daily living (ADLs): getting out of bed, dressing, walking,bathing, eating and/or using the toilet. Groups were categorised accordingto whether they experienced no issues or one or more issues with ADLs inline with previous work (Gale, Cooper and Aihie Sayer ).Number of chronic conditions involved summing the number of chronic

conditions people reported having from a list of included in ELSA (for alist of conditions, see Dhalwani et al. ). Conditions were categorised asnone, one, or two or more (multi-morbidity) (Dhalwani et al. ).Self-rated health was assessed using a single item where people rated their

own health on a five-point scale that ran from excellent to poor. Responseswere dichotomised as excellent/very good/good or fair/poor in line withprevious work (Badawi et al. ).Depressive symptoms were measured using the eight-item Center for

Epidemiological Studies Depression Scale (CES-D ). This item measuresthe past week experience of eight indicators of depressive symptomatology(responses yes/no). For the sake of this analysis, the item on loneliness(‘I have felt lonely’) was removed in line with previous work that has exam-ined loneliness and depressive symptoms using ELSA (Shankar et al. ).




https://doi.org/10.1017/S0144686X18000132


Responses were then categorised as high (⩾) versus low (<) depressivesymptoms in line with previous work (Au et al. ; Hamer et al. ).However, as this cut-point was validated for the full eight-item scale wealso ran sensitivity analyses with a cut-point of (cut-point identified forthe % with the highest depressive symptom score) to check our results.

Analysis

The indicators of loneliness, social isolation and living alone were groupedinto meaningful groups using LCA (Latent Gold .). LCA uses mathe-matical modelling to group categorical and/or continuous variables intoclasses (in this paper referred to as groups) based on likelihood estimates(Hagenaars and McCutcheon ; McCutcheon ; Vermunt andMagidson ). These classes represent groupings of data where peopleresponded similarly to our indicators of loneliness, social isolation andliving alone. To determine the optimal number of groups to derive we com-pared model fit from one to ten groups (run with , iterations). Wedetermined the best model fit by examining the Bayesian InformationCriteria (BIC), log-likelihood ratio (LLR), p-value and number of para-meters. The model that demonstrated parsimony between the lowestnumber of parameters, highest p-value, and lowest BIC and LLR was chosen.Once the optimal number of groups was identified, we examined the

latent (previously unobserved) shared loneliness, social isolation and livingalone characteristics that defined each group. Groups were then comparedon socio-demographic characteristics and health-related outcomes using chi-squared analysis for categorical variables and one-way analysis of variance forcontinuous variables. We then ran logistic regression analyses to examine therelationship between group membership and each health and wellbeingoutcome (issues with ADLs, self-rated health, multi-morbidity and depressivesymptoms) by comparing all groups who reported loneliness, isolation and/or living alone to a reference group that was neither lonely nor isolated.Logistic regression analyses were first run unadjusted and then adjustedfor socio-demographic characteristics (age, sex, marital status, wealth,employment and education). All analyses were conducted with SPSS ..

Results

LCA: identifying groups based on shared loneliness and isolationcharacteristics

The LCA showed that the lowest BIC value was demonstrated by the seven-group model (see Table ), although the value for the six-group model was




https://doi.org/10.1017/S0144686X18000132


comparable. As the number of groups increased, the LLR indicated thatmodel fit improved (see Table ), however, the number of parameters indi-cated that fewer groups was preferable whilst the p-value indicated anymodel fit from three groups onwards had a good fit. We used the six-group model as it demonstrated the greatest parsimony between the fourmodel-fit indicators (for model-fit statistics, see Table ).The six groups identified differed in terms of their likelihood to exhibit

different loneliness and living alone characteristics (see Table ) which wecharacterised as:

. Group : no loneliness or isolation (.% of sample).

. Group : moderate loneliness (.% of sample).

. Group : living alone (.% of sample).

. Group : moderate isolation (.% of sample).

. Group : moderate loneliness, living alone (.% of sample).

. Group : high loneliness, moderate isolation (with high likelihood ofliving alone) (.% of sample).

When differentiating groups based on loneliness, those groups with lowloneliness were identified by a high likelihood of responding ‘hardly everor never’ to loneliness questions. Those groups with moderate lonelinesswere identified by a high likelihood of responding ‘sometimes’ to lonelinessquestions. Finally, those groups with high loneliness were identified by ahigh likelihood of responding ‘always’ to loneliness questions. When differ-entiating groups based on isolation, low isolation was indicated by a high

T A B L E . Model fit based on the number of groups to which the data arefitted

Number of groups BIC p Number of parameters LLR

,. <. −. ,. <. −. ,. . −. ,. . −. ,. . −. ,. . −. ,. . −. ,. . −. ,. . −. ,. . −.

Notes: BIC: Bayesian Information Criteria. LLR: log-likelihood ratio. Lower numbers for indica-tors of model fit (BIC and LLR) indicate better model fit. Higher p-values also indicate bettermodel fit. The best match between model fit and p-values that results in the lowest number ofparameters represents the optimum number of groups.




https://doi.org/10.1017/S0144686X18000132


likelihood of frequent contact with friends, children and family plus havingsocial/societal involvement. Groups with moderate isolation were identifiedby having a higher likelihood of reporting positively to one or two of thesocial isolation indicators such as indicating less-frequent contact with

T A B L E . Characteristics of the six groups (classes) uncovered with LatentClass Analysis

Group Group Group Group Group Group

% . . . . . .Lacks companionship:Hardly ever or never . . . . . .Sometimes . . . . . .Always <. . . <. . .

Feels left out:Hardly ever or never . . . . . <.Sometimes . . . . . .Always <. . <. <. . .

Feels isolated:Hardly ever or never . . . . . .Sometimes . . . . . .Always <. . <. <. . .

Lives alone:No . . . . . .Yes . . . . . .

Contact with child:Frequent contact . . . . . .Infrequent contact . . . . . .No children . . . . . .

Contact with relatives:Frequent contact . . . . . .Infrequent contact . . . . . .No relatives . . . . . .

Contact with friends:Frequent contact . . . . . .Infrequent contact . . . . . .No friends . . . . . .

Social/societal involvement:Not involved . . . . . .Involved . . . . . .

Notes: Group : no/low loneliness and isolation. Group : moderate loneliness. Group : livesalone. Group : moderate isolation. Group : moderate loneliness, lives alone. Group : highloneliness and moderate isolation. Results shown are likelihood estimates of members withineach group responding in a particular way to each question (e.g. . indicates a .% like-lihood of giving that response/meeting that criteria). Results are rounded up to three decimalplaces and so not all results may add up to .




https://doi.org/10.1017/S0144686X18000132


friends, relatives or children plus reporting no social/societal involvement.High social isolation would be indicated by a higher likelihood of respond-ing positively to all indicators of social isolation.The first group, almost half of the sample, had a high likelihood of not

being lonely, socially isolated or living alone. Groups – demonstrated ahigh probability of experiencing one of our three factors. Group hadthe highest likelihood of reporting moderate feelings of loneliness(responding sometimes to feelings of loneliness). Group had thehighest likelihood of being isolated as measured by contact with family,friends and children when compared with other groups. Group had ahigh likelihood (%) of living alone but a low likelihood of exhibitingloneliness or other isolation characteristics. Groups and demonstratedcombinations of our three indicators. The fifth group showed a high like-lihood of reporting moderate loneliness and living alone but a high like-lihood of frequent contact with people and social/societal involvementand the highest likelihood of having no children (%) or relatives(%). The sixth group was most likely to report always being lonely,and also had a relatively high likelihood (%) of living alone and notbeing involved in social/societal activities (%). While not as high asGroup , this group also demonstrated a greater likelihood of havingless frequent contact with relatives (%) and a relatively high likelihoodof having no children (%) and the highest likelihood of having nofriends (%).

Comparing groups on socio-demographic and health-related outcomes

We firstly compared groups on socio-demographic characteristics (seeTable ), and there were significant differences for age, gender, maritalstatus, net wealth, employment and education between the six groups. Interms of age, Groups and , both characterised by a high probability ofliving alone, were the oldest, the most likely to be retired and had thehighest proportions of widows. Two of the three groups that had thehighest proportion of married and employed participants had been charac-terised by a low likelihood of being lonely (Groups and ). However, thegroup that was characterised by moderate loneliness also had a high propor-tion of married and employed respondents (Group ). Furthermore, thegroup characterised by experiencing no issues with loneliness or social iso-lation (Group ) were the best educated. Those groups with the highest pro-portions of females were the groups most likely to live alone (Groups , and ) whilst the group with the highest proportion of males was thegroup that reported moderate isolation (Group ). The group that hadthe highest proportion of respondents in the two lowest wealth quintiles




https://doi.org/10.1017/S0144686X18000132


T A B L E . Associations of group membership with socio-demographics, physical health and mental health

Group Group Group Group Group Group

% . . . . . .Mean age (SD)*** . (.) . (.) . (.) . (.) . (.) . (.)

Gender:***Male . . . . . .Female . . . . . .

Marital status:***Married . . . . . .Divorced/separated . . . . . .Single . . . . . .Widowed . . . . . .

Net wealth (minus pension):***Quintiles – (wealthiest) . . . . . .Quintiles and (poorest) . . . . . .

Educational qualifications:***Degree or higher education . . . . . .High school/college qualifications . . . . . .No qualifications . . . . . .

Employment:***Employed (full/part/self) . . . . . .Unemployed . . . . . .Retired . . . . . .

Functioning:***No issues with ADLs . . . . . .One or more issues with ADLs . . . . . .

Kimberley

J.Smith

andChristina

Victor

terms of use, available at https://w

ww

.cambridge.org/core/term

s. https://doi.org/10.1017/S0144686X18000132D

ownloaded from

https://ww

w.cam

bridge.org/core. IP address: 54.39.106.173, on 24 Aug 2020 at 01:22:12, subject to the Cambridge Core


https://doi.org/10.1017/S0144686X18000132


Self-rated health:***Excellent/very good/good . . . . . .Fair/poor . . . . . .

Number of chronic conditions:***None . . . . . .One . . . . . .Two or more . . . . . .

Depressive symptoms:***Low . . . . . .High . . . . . .

Notes: SD: standard deviation. ADL: activities of daily living. Group : no/low loneliness and isolation. Group : moderate loneliness. Group : lives alone.Group : moderate isolation. Group : moderate loneliness, lives alone. Group : high loneliness and moderate isolation.Significance levels: *p < ., **p < ., ***p < ..

Loneliness,isolation,living

aloneand

health


ww



ownloaded from

https://ww

w.cam



https://doi.org/10.1017/S0144686X18000132


and the highest proportion of unemployment was Group (high loneliness,moderate isolation, high likelihood of living alone).There were also significant differences between the six groups for all

four health and wellbeing outcomes: functioning (issues with ADLs),self-rated health, number of chronic conditions and depressive symptoms(see Table ). The groups who reported isolation and/or loneliness(Groups , , and ) had the highest proportion of two or more issueswith ADLs). This association was strongest in the group with both highloneliness and isolation (Group ). All of the groups who experiencedloneliness (Groups , and ) had a high likelihood of reporting fair/poor self-rated health (see Table ) and this was most pronounced in thegroup characterised by high loneliness and social isolation. The threegroups who had the highest proportion of multi-morbidity (two or morechronic conditions) were the group characterised by living alone(Group ) and the two characterised by shared loneliness and social isola-tion (Groups and ). All groups experiencing isolation and/or loneli-ness were more likely to exhibit high depressive symptoms (Groups , , and ). This association was strongest in the group with high lonelinessand isolation (Group ).We then ran a series of logistic regression analyses to determine the asso-

ciation between group membership and health and wellbeing, adjusting forsocio-demographic characteristics. Crude and adjusted estimates are pre-sented in Table . Group , no loneliness or isolation, was our reference cat-egory for both our crude and adjusted analyses. For all crude analyses,Groups – inclusive reported poorer outcomes on all four physical andmental health outcomes. All associations were attenuated after adjustmentfor Group (lives alone). Furthermore, the associations between Group

with multi-morbidity was also attenuated after adjustment. The most pro-nounced associations between group and health was observed for highdepressive symptoms, where all groups (except Group ) were more likelythan Group to report high depressive symptoms, and this was particularlypronounced in the group with high loneliness and isolation (Group ).

Sensitivity analysis

As the cut-point we used for our depressive symptom analysis (⩾) was basedon a cut-off validated for the eight-item scale, we re-ran our analysis with acut-point of ⩾ to see what effect this had on our results. For our cross-tabu-lation analysis, the association between group membership and depressivesymptoms remained highly significant (χ(, N = ,) = ., p <.). For the proportions of high depressive symptoms, see Table A inthe online supplementary material.




https://doi.org/10.1017/S0144686X18000132


T A B L E . Logistic regression analyses of association between group membership and physical and mental health

Group

One or more issues with ADLs Fair/poor self-rated health Two or more chronic conditions High depressive symptoms

UnadjustedOR (% CI)

AdjustedOR (% CI)

UnadjustedOR (% CI)

AdjustedOR (% CI)

UnadjustedOR (% CI)

AdjustedOR (% CI)

UnadjustedOR (% CI)

AdjustedOR (% CI)

.

(.–.)***.

(.–.)***.

(.–.)***.

(.–.)***.

(.–.)***.

(.–.)***.(.–.)***

.(.–.)***

.(.–.)**

.(.–.)

.(.–.)***

.(.–.)

.(.–.)***

.(.–.)

.(.–.)***

.(.–.)

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)**

.(.–.)**

.(.–.)

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

.(.–.)***

Notes: ADL: activities of daily living. OR: odds ratio. CI: confidence interval. Group : no/low loneliness and isolation. Group : moderate loneliness.Group : lives alone. Group : moderate isolation. Group : moderate loneliness, lives alone. Group : high loneliness and moderate isolation. Alladjusted models are adjusted for age, sex (male/female), marital status (married or partner/single/divorced or separated), employment (employed/unemployed/retired), qualifications (university or higher education/high school or college/no qualifications) and net non-pension wealth (topthree quintiles/bottom two quintiles). . Reference group: no issues with ADLs (N in fully adjusted model = ,). . Reference group: excellent/very good/good self-rated health (N in fully adjusted model = ,). . Reference group: no chronic conditions (N in fully adjusted model = ,).. Reference group: low/no depressive symptoms as indicated by a score of < on the Center for Epidemiological Studies Depression Scale (N infully adjusted model = ,).Significance levels: *p < ., **p < ., ***p < ..

Loneliness,isolation,living

aloneand

health


ww



ownloaded from

https://ww

w.cam



https://doi.org/10.1017/S0144686X18000132


When we re-ran the logistic regression analysis with the cut-off of , allresults remained significant at the same level as when the cut-off was ,even for unadjusted analyses (see Table B in the online supplementarymaterial). Thus, using a cut-point of produced comparable results to acut-point of .

Discussion

Using clustering methodologies on a large data-set, we have been able touncover meaningful sub-groups of older people based on shared character-istics of loneliness, living alone and social isolation in a community popula-tion of English older adults. Our results indicate that existing typologies thattake into account the potential overlap or independence of social isolationand loneliness, such as the groupings suggested by Andersson (),reflect the experience of older adults. However, our work develops this typ-ology further by demonstrating the presence of groups characterised by ahigh prevalence of living alone. Our work also suggests that we need to dif-ferentiate between those experiencing loneliness or isolation, or in combin-ation, and that the severity of those issues may also be important in definingdifferent typologies, as suggested by Victor et al. (). Therefore, wesuggest considering both the number of issues that people have (i.e.none, one issue independently, two or more issues in combination) andthe severity of these issues (e.g. is loneliness moderate or severe in its inten-sity) will be important in defining different typologies in future work.The largest single group identified within our sample did not experience

social isolation, loneliness and did not live alone (Group ). This group wasmostly younger, married, wealthier, better educated and had fewer issueswith health and wellbeing. This observation provides support to the ideathat health and wellbeing are important correlates of loneliness andsocial isolation in older adults (Heinrich and Gullone ; Luanaighand Lawlor ). It is possible that increased community and socialcontact could have helped this group to maintain their health, however, itcould also be the case that this group do not experience loneliness andsocial isolation because they do not have any health issues (Berkman andGlass ; Ferlander ). Future research could examine the directionof causality between maintenance of health and lack of loneliness/social iso-lation. There is evidence suggesting that loneliness and social isolation arerisk factors for the development of serious illness (Valtorta et al. b) andmortality (Elovainio et al. ; Holt-Lunstad et al. ; Steptoe et al.b). However, evidence also suggests that poor health is also a riskfactor for the development of loneliness (Victor et al. ). Thus,




https://doi.org/10.1017/S0144686X18000132


examining longitudinal relationships between health and loneliness, isola-tion and living alone could be important.When creating typologies, we have suggested it could be important to dif-

ferentiate between groups who experience an issue independently or incombination with another issue. We identified two groups who livedalone. One of these groups only experienced living alone (Group ),whereas the other group experienced living alone with moderate loneliness(Group ). Both groups that lived alone were older than all other groupsand were more likely to be widowed. However, Group also reported lone-liness despite a similar demographic profile to Group . When compared tothe reference group, Group had a higher likelihood of reporting depres-sive symptoms and poorer health whereas for Group , after adjustment forconfounders, there was no association between group membership andhealth. Depressive symptoms (Barg et al. ; Cacioppo et al. ) andpoorer health (Victor et al. ) have previously been pointed to as import-ant issues strongly associated with loneliness. Thus, it is possible the poorerhealth andmental wellbeing of Group could be why this group also experi-ences loneliness. Future longitudinal work could further unpick the causal-ity of this association.The identification of Group was also of theoretical and policy/practice

interest. Previous theoretical conceptualisations of loneliness and isolationhave differentiated these concepts from living alone, which is simply ameasure of the type of household in which a person lives (Victor et al.). The social isolation index from which our indicators of social isola-tion used in this analysis were derived includes living alone (Shankar et al.). However, the identification of Group supports the presuppositionof Victor et al. () that living alone can be differentiated from lonelinessand other indicators of social isolation. Further investigation of Group

could also be particularly interesting to determine why, in the face ofmany of the socio-demographic risk factors for experiencing social isolationand loneliness such as living alone, widowhood, being female and older age(Pinquart and Sorensen a; Wenger et al. ), they are less likely toreport being lonely or socially isolated.We also identified two groups who experienced moderate loneliness

(Group ) and some isolation (Group ) independently. Group werebroader similar to Group in socio-demographic terms but had worsehealth outcomes. Previous work has indicated that loneliness is stronglyassociated with health status (Luanaigh and Lawlor ; Victor et al.), thus it is possible that poorer psychological and physical healthexplained the moderate loneliness that characterised Group . Group

was particularly notable as it had a higher proportion of males than allother groups. This could be because men are less likely to report




https://doi.org/10.1017/S0144686X18000132


(Pinquart and Sörensen b) or recognise (Borys and Perlman )feelings of loneliness compared with women, or that they are more likelyto experience social rather than emotional loneliness (Dahlberg andMcKee ). Our work adds to pre-existing literature by indicating thatgender may be an important factor to examine when separating outgroups based on loneliness and social isolation.We have suggested that alongside taking the broad threefold typologies

into consideration, the severity of these experiences could be importantin differentiating between groups. This is particularly evident for Group. This group had the most severe profile of loneliness and social isolation,and also had a high likelihood of living alone. This group had the most pro-nounced association with all indicators of physical and mental health.Previous work has indicated that loneliness and depression share a strongrelationship (Barg et al. ; Cacioppo et al. ). Our work adds tothis by showing, when compared with a group with low/no loneliness or iso-lation, that loneliness, living alone and social isolation are linked withdepressive symptoms. Our work also suggests that as loneliness severityincreases, the association with depressive symptoms increases, as hypothe-sised by Victor et al. ().However, it is important to note that most groups who experienced either

social isolation and/or loneliness were all more likely to have high depres-sive symptoms when compared with the group with no loneliness or isola-tion, even after we adjusted for socio-demographic variables. Thisindicates that mental wellbeing should be considered in all people whomay be socially isolated or lonely. Past work indicates that loneliness andsocial isolation share different relationships with health outcomes(Cornwell and Waite ; Holwerda et al. ; Shankar et al. ;Steptoe et al. a). Our work adds to the research in this field byshowing that when compared with a group with no loneliness or isolation,any group who experiences loneliness and/or social isolation are morelikely to report poor health. This work also indicates identifying groupswith differing characteristics of loneliness and social isolation using data-driven methodologies could represent an interesting avenue to tell usmore about how loneliness, social isolation and living alone are linkedwith health.

Limitations and considerations for future work

There are methodological caveats to our findings. When interpreting theseresults it is important to note that LCA is an exploratory technique, and thatmodel fit is not absolute. Both the number of groups examined and thecharacteristics of the groups were determined using likelihood estimates.




https://doi.org/10.1017/S0144686X18000132


It should be noted that there was considerable within-group heterogeneityfor many of the indicators we examined, so while groups were identified ashaving a high likelihood of exhibiting certain characteristics, this is not tosay every single person within this group definitively exhibited these charac-teristics (just that they were the group they were most likely to belong tobased on their profile). Future work could use LCA methodologies withindifferent populations to determine whether the proposed broad typologiesare generalisable to different populations and samples. A related consider-ation for interpretation is that these groups were uncovered in a specificcohort, and may not generalise to other populations or cultures.Measurement issues should also be considered. Both the loneliness and

social isolation indicators examined were short, and likely would notcapture the full complexity of experiences of social isolation or loneliness.There is also increasing evidence suggesting that the chronicity and chan-ging nature of loneliness and social isolation are important issues to con-sider (Dahlberg et al. ; Dykstra, Van Tilburg and de Jong Gierveld; Victor and Bowling ), but we could not address them. Futureresearch could address these limitations by re-running this analysis in astudy with more comprehensive loneliness and isolation measures thattake chronicity into account.The indicators of social isolation used in this study have also been used by

other researchers to operationalise differing theoretical constructs such associal connectedness (Cornwell and Waite ) and social detachment(Jivraj, Nazroo and Barnes ; Tomaszewski and Barnes ). Thus,while we have used these items as indicators of social isolation, they couldbe indicating social connectedness, social detachment or other related con-cepts such as social exclusion, which future research could try to disentan-gle. Furthermore, it is important to determine whether isolation was anactive or passive process. Qualitative work suggests that one can differentiatebetween voluntary and involuntary loneliness (Dahlberg ), and it isplausible to assume these different types of loneliness/isolation will have dif-ferent associations with health and wellbeing. The cross-sectional nature ofthis analysis means we cannot infer causality. It is plausible that there is a bi-directional association of loneliness, living alone and social isolation withhealth and wellbeing, and we will explore this in our future work.

Conclusions and implications

Using LCA we have been able to uncover different groups based on sharedexperiences of loneliness, social isolation and living alone, showing the dif-ferent experiences older adults can have. This has important implications




https://doi.org/10.1017/S0144686X18000132


for the creation of typologies based on loneliness, social isolation and livingalone. Our work indicates considering the number of issues that peoplehave (i.e. none, one, two or more) and then the severity of these is importantin defining different typologies. However, in order to make use of these typ-ologies it is important to replicate these results in different samples. We haveshown that different groups share different associations with health andwellbeing. This work indicates that the lived experience of loneliness,social isolation and living alone in older adults is complex, and thattaking the number of issues and severity of issues into account will beimportant for researchers and clinicians working with groups of olderadults who may be experiencing these issues. It also demonstrates thatliving alone is conceptually separate from loneliness and isolation, andhas limited utility as a measure of these complex concepts.

Supplementary material

To view supplementary material for this article, please visit https://doi.org/./SX.

References

Alpass, F. M. and Neville, S. . Loneliness, health and depression in older males.Aging & Mental Health, , , –.

Andersson, L. . A model of estrangement—including a theoretical understand-ing of loneliness. Psychological Reports, , , –.

Au, B., Smith, K. J., Gariépy, G. and Schmitz, N. . The longitudinal associationsbetween C‐reactive protein and depressive symptoms: evidence from the EnglishLongitudinal Study of Ageing (ELSA). International Journal of Geriatric Psychiatry,, , –.

Badawi, G., Pagé, V., Smith, K. J., Gariépy, G., Malla, A., Wang, J., Boyer, R.,Strychar, I., Lesage, A. and Schmitz, N. . Self-rated health: a predictor forthe three year incidence of major depression in individuals with Type II diabetes.Journal of Affective Disorders, , , –.

Barg, F. K., Huss-Ashmore, R., Wittink, M. N., Murray, G. F., Bogner, H. R. andGallo, J. J. . A mixed-methods approach to understanding loneliness anddepression in older adults. Journals of Gerontology: Psychological Sciences and SocialSciences, B, , S–.

Bassuk, S. S., Glass, T. A. and Berkman, L. F. . Social disengagement and inci-dent cognitive decline in community-dwelling elderly persons. Annals of InternalMedicine, , , –.

Berkman, L. F. and Glass, T. . Social integration, social networks, social support,and health. In Berkman, L. F. and Kawachi, I. (eds), Social Epidemiology. OxfordUniversity Press, Oxford, –.

Borys, S. and Perlman, D. . Gender differences in loneliness. Personality andSocial Psychology Bulletin, , , –.



https://doi.org/10.1017/S0144686X18000132

https://doi.org/10.1017/S0144686X18000132

https://doi.org/10.1017/S0144686X18000132


https://doi.org/10.1017/S0144686X18000132


Burholt, V., Dobbs, C. and Victor, C. . Social support networks of older migrantsin England andWales: the role of collectivist culture. Ageing & Society, , , –.

Cacioppo, J. T. and Hawkley, L. C. . Perceived social isolation and cognition.Trends in Cognitive Sciences, , , –.

Cacioppo, J. T., Hughes, M. E., Waite, L. J., Hawkley, L. C. and Thisted, R. A. .Loneliness as a specific risk factor for depressive symptoms: cross-sectional andlongitudinal analyses. Psychology and Aging, , , –.

Cloutier-Fisher, D., Kobayashi, K. and Smith, A. . The subjective dimension ofsocial isolation: a qualitative investigation of older adults’ experiences in smallsocial support networks. Journal of Aging Studies, , , –.

Cornwell, E. Y. and Waite, L. J. . Social disconnectedness, perceived isolation,and health among older adults. Journal of Health and Social Behavior, , , –.

Dahlberg, K. . The enigmatic phenomenon of loneliness. International Journal ofQualitative Studies on Health and Well-being, , , –.

Dahlberg, L., Andersson, L., McKee, K. J. and Lennartsson, C. . Predictors ofloneliness among older women and men in Sweden: a national longitudinalstudy. Aging & Mental Health, , , –.

Dahlberg, L. and McKee, K. J. . Correlates of social and emotional loneliness inolder people: evidence from an English community study. Aging & Mental Health,, , –.

Davis, M., Murphy, S., Neuhaus, J. and Lein, D. . Living arrangements anddietary quality of older US adults. Journal of the American Dietetic Association, ,, –.

Dean, A., Kolody, B., Wood, P. and Matt, G. E. . The influence of living alone ondepression in elderly persons. Journal of Aging and Health, , , –.

Dhalwani, N. N., O’Donovan, G., Zaccardi, F., Hamer, M., Yates, T., Davies, M. andKhunti, K. . Long terms trends of multimorbidity and association with phys-ical activity in older English population. International Journal of Behavioral Nutritionand Physical Activity, , , .

Dhalwani, N. N., Zaccardi, F., O’Donovan, G., Carter, P., Hamer, M., Yates, T.,Davies, M. and Khunti, K. . Association between lifestyle factors and the inci-dence of multimorbidity in an older English population. Journals of Gerontology,A, , –.

Dykstra, P. A., Van Tilburg, T. G. and de Jong Gierveld, J. . Changes in olderadult loneliness: results from a seven-year longitudinal study. Research on Aging,, , –.

Ellwardt, L., Aartsen, M. and van Tilburg, T. . Types of non-kin networks andtheir association with survival in late adulthood: a Latent Class Approach.Journals of Gerontology: Psychological Sciences and Social Sciences, B, , –.

Elovainio, M., Hakulinen, C., Pulkki-Råback, L., Virtanen, M., Josefsson, K.,Jokela, M., Vahtera, J. and Kivimäki, M. . Contribution of risk factors toexcess mortality in isolated and lonely individuals: an analysis of data from theUK Biobank cohort study. The Lancet Public Health, , , e–.

Ferlander, S. . The importance of different forms of social capital for health.Acta Sociologica, , , –.

Gale, C. R., Cooper, C. and Aihie Sayer, A. . Prevalence of frailty and disability:findings from the English Longitudinal Study of Ageing. Age and Ageing, , ,–.

Golden, J., Conroy, R. M., Bruce, I., Denihan, A., Greene, E., Kirby, M. andLawlor, B. A. . Loneliness, social support networks, mood and wellbeing in




https://doi.org/10.1017/S0144686X18000132


community-dwelling elderly. International Journal of Geriatric Psychiatry, , ,–.

Hagenaars, J. A. and McCutcheon, A. L. . Applied Latent Class Analysis.Cambridge University Press, Cambridge.

Hamer, M., Molloy, G. J., de Oliveira, C. and Demakakos, P. . Leisure time phys-ical activity, risk of depressive symptoms, and inflammatory mediators: the EnglishLongitudinal Study of Ageing. Psychoneuroendocrinology, , , –.

Hawkley, L. C. and Cacioppo, J. T. . Loneliness matters: a theoretical and empir-ical review of consequences and mechanisms. Annals of Behavioral Medicine, , ,–.

Hawthorne, G. . Perceived social isolation in a community sample: its preva-lence and correlates with aspects of peoples’ lives. Social Psychiatry and PsychiatricEpidemiology, , , –.

Heinrich, L. M. and Gullone, E. . The clinical significance of loneliness: a litera-ture review. Clinical Psychology Review, , , –.

Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T. and Stephenson, D. .Loneliness and social isolation as risk factors for mortality: a meta-analyticreview. Perspectives on Psychological Science, , , –.

Holwerda, T. J., Deeg, D. J., Beekman, A. T., van Tilburg, T. G., Stek, M. L., Jonker, C.and Schoevers, R. A. . Feelings of loneliness, but not social isolation, predictdementia onset: results from the Amsterdam Study of the Elderly (AMSTEL).Journal of Neurology, Neurosurgery, and Psychiatry, , –.

Hughes, M. E., Waite, L. J., Hawkley, L. C. and Cacioppo, J. T. . A short scale formeasuring loneliness in large surveys: results from two population-based studies.Research on Aging, , , –.

Iliffe, S., Kharicha, K., Harari, D., Swift, C., Gillmann, G. and Stuck, A. E. .Health risk appraisal in older people : the implications for clinicians and commis-sioners of social isolation risk in older people. British Journal of General Practice, ,, –.

Jivraj, S., Nazroo, J. and Barnes, M. . Change in social detachment in older agein England. In Banks, J., Nazroo, J. and Steptoe, A. (eds), The Dynamics of Ageing:Evidence from the English Longitudinal Study of Ageing – (Wave ). Institute forFiscal Studies, London, –.

Kharicha, K., Iliffe, S., Harari, D., Swift, C., Gillmann, G. and Stuck, A. E. .Health risk appraisal in older people : are older people living alone an ‘at-risk’group? British Journal of General Practice, , , –.

Larson, R., Zuzanek, J. and Mannell, R. . Being alone versus being with people:disengagement in the daily experience of older adults. Journal of Gerontology, , ,–.

Lee, H. Y., Jang, S.-N., Lee, S., Cho, S.-I. and Park, E.-O. . The relationshipbetween social participation and self-rated health by sex and age: a cross-sectionalsurvey. International Journal of Nursing Studies, , , –.

Luanaigh, C. Ó. and Lawlor, B. A. . Loneliness and the health of older people.International Journal of Geriatric Psychiatry, , , –.

Lubben, J. and Gironda, M. . Centrality of social ties to the health and well-being of older adults. In Berkman, B. and Harootyan, L. K. (eds), Social Workand Health Care in an Aging Society: Education, Policy, Practice and Research.Springer, New York, –.

McCutcheon, A. L. . Latent Class Analysis. Sage, Newbury Park, California.McHugh, J., Kenny, R., Lawlor, B., Steptoe, A. and Kee, F. . The discrepancy

between social isolation and loneliness as a clinically meaningful metric:




https://doi.org/10.1017/S0144686X18000132


findings from the Irish and English longitudinal studies of ageing (TILDA andELSA). International Journal of Geriatric Psychiatry, , , –.

Nummela, O., Seppänen, M. and Uutela, A. . The effect of loneliness andchange in loneliness on self-rated health (SRH): a longitudinal study amongaging people. Archives of Gerontology and Geriatrics, , , –.

Orth-Gomér, K., Rosengren, A. and Wilhelmsen, L. . Lack of social support andincidence of coronary heart disease in middle-aged Swedish men. PsychosomaticMedicine, , , –.

Perissinotto, C. M., Cenzer, I. S. and Covinsky, K. E. . Loneliness in olderpersons: a predictor of functional decline and death. Archives of InternalMedicine, , , –.

Perissinotto, C. M. and Covinsky, K. E. . Living alone, socially isolated or lonely –what are we measuring? Journal of General Internal Medicine, , , –.

Pinquart, M. and Sorensen, S. a. Influences on loneliness in older adults: ameta-analysis. Basic and Applied Social Psychology, , , –.

Pinquart, M. and Sörensen, S. b. Gender differences in self-concept and psycho-logical well-being in old age: a meta-analysis. Journals of Gerontology: PsychologicalSciences and Social Sciences, B, , –.

Rook, K. S. . The negative side of social interaction: impact on psychologicalwell-being. Journal of Personality and Social Psychology, , , –.

Shankar, A., Hamer, M., McMunn, A. and Steptoe, A. . Social isolation and lone-liness: relationships with cognitive function during years of follow-up in theEnglish Longitudinal Study of Ageing. Psychosomatic Medicine, , , –.

Shankar, A., McMunn, A., Banks, J. and Steptoe, A. . Loneliness, social isolation,and behavioral and biological health indicators in older adults. Health Psychology,, , –.

Steptoe, A., Breeze, E., Banks, J. and Nazroo, J. a. Cohort profile: the EnglishLongitudinal Study of Ageing. International Journal of Epidemiology, , , –.

Steptoe, A., Shankar, A., Demakakos, P. and Wardle, J. b. Social isolation, lone-liness, and all-cause mortality in older men and women. Proceedings of the NationalAcademy of Sciences of the United States of America, , , –.

Sundström, G., Fransson, E., Malmberg, B. and Davey, A. . Loneliness amongolder Europeans. European Journal of Ageing, , , –.

Tomaszewski, W. and Barnes, M. . Investigating the dynamics of social detach-ment in older age. In Banks, J., Breeze, E., Lessof, C. and Nazroo, J. (eds), Living inthe st Century: Older People in England. The Institute of Fiscal Studies, London,–.

Townsend, P. . Isolation, desolation and loneliness. In Shanas, E., Townsend, P.,Wedderburn, D., Friis, H., Milhoj, P. and Stehouwer, J. (eds), Old People in ThreeIndustrial Societies. Routledge & Kegan Paul, London, –.

Tunstall, J. . Old and Alone: A Sociological Study of Old People. Routledge & KeganPaul, London.

Valtorta, N. and Hanratty, B. . Loneliness, isolation and the health of olderadults: do we need a new research agenda? Journal of the Royal Society of Medicine,, , –.

Valtorta, N. K., Kanaan, M., Gilbody, S. andHanratty, B. a. Loneliness, social iso-lation and social relationships: what are we measuring? A novel framework for clas-sifying and comparing tools. BMJ Open, , , e.

Valtorta, N. K., Kanaan, M., Gilbody, S., Ronzi, S. and Hanratty, B. b. Lonelinessand social isolation as risk factors for coronary heart disease and stroke: sys-tematic review and meta-analysis of longitudinal observational studies. Heart,, , –.




https://doi.org/10.1017/S0144686X18000132


Vermunt, J. K. and Magidson, J. . Latent class cluster analysis. In Hagenaars, J.and McCutcheon, A. (eds), Applied Latent Class Analysis. Cambridge UniversityPress, Cambridge, –.

Victor, C. R. and Bowling, A. . A longitudinal analysis of loneliness among olderpeople in Great Britain. Journal of Psychology, , , –.

Victor, C., Scambler, S., Bond, J. and Bowling, A. . Being alone in later life:loneliness, social isolation and living alone. Reviews in Clinical Gerontology, , ,–.

Victor, C. R., Scambler, S. J., Bowling, A. and Bond, J. . The prevalence of, andrisk factors for, loneliness in later life: a survey of older people in Great Britain.Ageing & Society, , , –.

Walton, C. G., Shultz, C. M., Beck, C. M. and Walls, R. C. . Psychological corre-lates of loneliness in the older adult. Archives of Psychiatric Nursing, , , –.

Weiss, R. S. . Loneliness: The Experience of Emotional and Social Isolation. MIT Press,Cambridge, Massachusetts.

Wenger, G. C., Davies, R., Shahtahmasebi, S. and Scott, A. . Social isolationand loneliness in old age: review and model refinement. Ageing & Society, , ,–.

Zebhauser, A., Hofmann-Xu, L., Baumert, J., Häfner, S., Lacruz, M., Emeny, R.,Döring, A., Grill, E., Huber, D. and Peters, A. . How much does it hurt tobe lonely? Mental and physical differences between older men and women inthe KORA‐Age Study. International Journal of Geriatric Psychiatry, , , –.

Zunzunegui, M.-V., Alvarado, B. E., Del Ser, T. and Otero, A. . Social networks,social integration, and social engagement determine cognitive decline in commu-nity-dwelling Spanish older adults. Journals of Gerontology: Psychological Sciences andSocial Sciences, B, , S–.

Accepted January ; first published online April

Address for correspondence :Kimberley J. Smith,University of Surrey,Stag Hill Campus,Guildford GU XH, UK

E-mail: [email protected]



mailto:[email protected]


https://doi.org/10.1017/S0144686X18000132


Documents

Typologies of loneliness, living alone and …...Typologies of loneliness, living alone and socialisolation,andtheirassociationswith physical and mental health KIMBERLEY J. SMITH*