Abstract |
Background The fact that the contemporary world has been ageing rapidly is one of
the greatest challenges health care and social security systems have to deal with.
Social and gerontological research that has been implemented up to date, mainly at
the country level, has extensively addressed the health of people of third and fourth
age and has thus indicated several determinants associated with their social
environment. In this context, well-being has been denoted to predict longer survival
among older people and higher utilization of health care services. Therefore, wellbeing
has become an important objective of ageing-related public health and social
policies. However, the study of older adults’ well-being and the comprehensive
evaluation of the factors related to its configuration remain relatively unexplored,
whereas the respective existing findings of nationally comparative research have been
limited.
Aim The current study aimed at assessing the prevalence of six different positive
well-being outcomes and their accumulated presence (4+) as well, among adults aged
65 years of age and older of the eleven European countries who took part in the first
wave of the SHARE survey (Survey of Health, Ageing and Retirement in Europe),
according to: (i) their socio-demographic characteristics, (ii) social engagement (1st
Chapter), social isolation (2nd Chapter) and loneliness (3rd chapter). Furthermore,
frequency of feelings of loneliness was examined in relation to adverse health
conditions, stressful life events and social isolation (4rth Chapter). In addition, the
utilization of preventive health services was measured according to social isolation
indicators and their multiple clustering (5th Chapter). Lastly, we were interested to
study the potential differences in well-being, social engagement, social isolation,
loneliness and preventive health services utilization among the different countries and
geographical regions under scrutiny.
Subjects and Methods The data of the present study pertains to a subsample of
adults aged 65 and older which was retrieved from the first wave of the cross-national
longitudinal SHARE survey which was conducted between 2004 and 2005 in eleven
European countries (Austria, Belgium, France, Germany, Denmark, Switzerland,
Greece, Spain, Italy, the Netherlands and Sweden). This survey was organized and
coordinated centrally at the Mannheim Research Institute for the Economics of
Ageing (MEA, Germany) under the collaborative effort of multidisciplinary national
teams of more than 150 researchers worldwide and more than 60 working groups,
including the research team of the Department of Social Medicine, in the Faculty of
Medicine of University of Crete. The target population of the study concerned households with at least one member
aged 50 and over, including their probably younger partners or spouses, and it was
selected according to the complex multistage stratification design that was
implemented so that this population to be representative of the European population
aged over 50 years. At the individual level, the average weighted response rate which
was achieved ranged from 73.7% in Spain to 93.3% in Germany, whereas at the
household level the lowest response rate was reached in Switzerland (38.8%) and the
highest in France (81.0%). For the purposes of the current investigation, the analyses
comprised individuals aged 65 years and older within the SHARE sample, which
yielded a study population of 7,025, 5,129, 5,074 and 6,971 respondents.
Well-being was gauged as the clustering of six indicators: quality of life, depressive
symptomatology, self-perceived health, life satisfaction, chronic conditions and Body
Mass Index (BMI). High well-being was equated with reporting high quality of life,
exhibiting absence of depressive symptomatology, perceiving health status as very
good, being very satisfied with life, suffering from one or none chronic health
condition and having normal BMI. The clustering of more than four well-being
indicators (4+) was considered to be indicative of higher well-being and referred to as
multiple presence of positive well-being outcomes. Additionally, we assessed the
socio-demographic characteristics of the participants, their social engagement, social
isolation, loneliness, adverse health conditions, stressful life events and preventive
health services utilization.
Data were analyzed using the IBM-SPSS v21.0. Weights were applied according to
the complex sampling design of the study, reflecting non-responses and stratification
design. The prevalence of well-being indicators and the respective 95% confidence
intervals (95% CIs) were estimated according to the complex sampling design.
Furthermore, analysis of covariance and multivariate regression analysis were applied
in order to search for potential differences and associations between well-being and
social engagement, social isolation and loneliness. Furthermore, we estimated
frequency of feelings of loneliness with the respective p-values and we examined
frequency of loneliness according to adverse health conditions, stressful life events
and social isolation. Multivariate regression models were calculated with the
respective Odds Ratios (ORs) so as to study the effect of the above factors on
frequency of loneliness. Lastly, preventive health services utilization was measured
according to a composite score of twelve different items (12-item composite score)
and the distribution of this score was investigated according to social isolation. In
addition, the utilization of the distinct components of preventive care under study was
also examined according to different indicators of social isolation through multiple
logistic regression analysis Moreover, in order to detect possible national variations, we estimated the weighted
prevalence and the corresponding confidence intervals of the frequency of lacking
indicators of well-being and the occurrence of social isolation and loneliness across
the eleven SHARE European countries. Lastly, country-specific differences in the
association between the frequency of activity participation and well-being clustering
were also addressed by means of simple linear regression analysis.
Results More than four indicators of well-being were observed for 10.2% of the
respondents, whereas for 14.4% of the total sample no positive well-being outcomes
were rendered. The majority of the sample was found with one or two well-being
indicators (28.9 and 27.9%, respectively). The presence of 4+ wellbeing indicators
was significantly more common among participants aged 65–74 years than among
those of age 75–84 years, whereas the prevalence of accumulated well-being
indicators among the oldest-old participants of the study did not differ significantly
from the their younger counterparts aged 75-84 years old.
The prevalence of 4+ well-being indicators was shown to be more than twice as high
(23.2%; 95% CI 20.5–26.1) in Northern countries compared to Central countries
(11.2%, 95% CI 9.7–12.8) and more than three times as high compared to Southern
ones (7.2%; 95% CI 5.8–9.0). A significantly lower proportion of the participants
who had not participated in any productive or/and social activities were found with
high quality of life, in relation to productively and socially active ones. Likewise, the
proportion of adults who were attested with low depression score, rated their health as
very good, were very satisfied with life and displayed less than two chronic diseases
was significantly higher among those with frequent productive or/and social activity
participation over the course of the previous month.
This pattern was consistent for most well-being indicators and remained after their
clustering, with 4+ indicators of well-being being significantly more prevalent among
frequent participants in productive or/and social activities, than infrequent ones
(15.0%, 95% CI 12.9–17.4 vs. 7.2%; 95% CI 6.1–8.5). Clustering of well-being
indicators was found to correlate at a significant level with frequent participation in
productive (ORs=1.35, p=0.007) and social activities (ORs=1.57, p<0.001).
Accordingly, a higher score of well-being indicators was evident among older adults
participating frequently in productive or/and social activities, in relation to those who
had not participated in any activities over the course of the previous month (2.1 vs.
1.7, respectively, p<0.05). The correlation between frequent productive or/and social
activity participation and multiple presence of well-being indicators was 0.050
(p=0.045). Additionally, well-being was found to be significantly associated with specific
indicators of social isolation. More particularly, individuals contacting their offspring
daily or at least once a month displayed a significantly higher mean well-being score
(1.80), in relation to those reporting less frequent or no parent-child contact (1.40)
(p=0.028). The above pattern was also ascertained regarding participants with at least
one social activity, in relation to their socially inactive peers (1.93 vs. 1.70, p=0.001).
A higher mean well-being score was thus observed among older adults living in
partnered households (1.90), compared to their unpartnered counterparts (1.69)
(p=0.007). Older individuals with an accumulation of social isolation indicators
indicated a lower mean well-being score (1.69), in comparison to the least isolated
individuals (1.94). However the above difference did not reach statistical significance
(p=0.200).
Regarding the prevalence of well-being outcomes according to loneliness, it was
found that individuals who declared feeling lonely most of the time over the course of
the previous week had a significantly lower mean score of indicators of well-being
(1.07), relative to their non lonely counterparts (1.36) (p-trend=0.002). Moreover, the
proportion of the respondents being very satisfied with their life was significantly
higher among those with no feelings of loneliness (40.5%; 95% CI 38.1-42.9), as
compared to adults with very frequent endorsement of loneliness feelings (10.5%;
95% CI 7.1-15.3). Likewise, multiple clustering of well-being indicators was
significantly more prevalent among non lonely individuals (15.5%; 95% CI 13.8-
17.2), in relation to their lonely seniors (6.9%; 95% CI 3.7-12.4).
As far as the assessment of loneliness is concerned, persistent feelings of loneliness,
endured most of the time, were mostly reported by females (11.2%), compared to
males (7.2%), whereas frequent feelings of loneliness were reported by 30.8% of
males and 47.9% of females (p<0.001). In addition, loneliness was indicated to be
unequally distributed among different age groups and individuals of different
educational and income status. Specifically, the proportion of individuals declaring to
feel lonely most of the time was significantly higher among the oldest-old participants
(85+) (12.4%), as compared to their younger counterparts, aged 65-74 (7.7%) and 75-
84 years old (11.9%) (p<0.001). Moreover, older adults who had obtained more years
of schooling were found to suffer from significantly less frequent feelings of
loneliness, in comparison to adults with the least years of education (p<0.001).
Likewise, 13.6% of the respondents with the lowest household income declared to
feel lonely, in relation to 7.3% of those belonging to the highest income quartile.
As regards the association of loneliness with adverse health conditions, stressful life
events and social isolation significant differences were shown to exist. More
particularly, individuals with one or more chronic conditions reported more frequent feelings of loneliness, in comparison to adults suffering from less than two chronic
diseases (p=0.015). This pattern also held true for older people with more than one
limitations in activities of daily living ((I)ADL) (p<0.001) or more than one disease
symptoms (p=0.002) and more than four depressive symptoms (p<0.001).
Furthermore, significant were the differences noted in the distribution of loneliness
frequency between individuals living in social isolation, as indicated by solitary
living, social disengagement and childlessness (p<0.001). In addition, for 12.2% of
widowed older people frequent feelings of loneliness were endorsed, relative to 7.8%
of those living in partnered households. In a similar vein, the likelihood of persistent
endorsements of feelings of loneliness was twice as high among older adults whose
offspring had recently moved out from parental nest, in relation to those whose child
still shared the same house with them, in both models of multiple logistic regression
analysis (ORs=2.08; 95% CI 1.24-3.48 and ORs=1.75; 95% CI 1.03-2.96,
respectively). Lastly, 27.8% of Italian and 26.1% of Greek older individuals were
categorized as severely lonely, which applied to 6.0% of the elderly in Denmark and
5.0% in the Netherlands.
With reference to the utilization of preventive care according to social isolation, the
mean score of preventive health services was documented to be significantly higher
among adults residing with a partner or spouse (p=0.001), being married (p=0.004),
having at least one child (p=0.046) and being involved in any kind of productive
or/and social activity (p=0.023). Additionally, respondents presenting multiple
presence of social isolation indicators (4+) were found to have a significantly lower
mean score of preventive health services utilization, as compared to their non isolated
partners (37.6 vs. 41.8, p=0.046). Further, older adults living unpartnered, as well as
those being socially disengaged, indicated a significantly lower likelihood to have
visited a dentist (ORs=0.69; 95% CI 0.52-0.91 and ORs=0.70; 95% CI 0.54-0.89,
respectively). Similarly, individuals with no activity participation had lower odds of
having ever undertaken sigmoidoscopy or colonoscopy (ORs=0.74; 95% CI 0.57–
0.96). Significant differences were discernible between the eleven European countries
under investigation as regards the distribution of health services utilization score
among socially isolated older people. Specifically, the mean score of preventive
health services utilization among adults with 4+ social isolation indicators was 49.6 in
France and 26.0 in Greece.
Furthermore, the rate of the multiple presence of social isolation indicators was
approximately 9.0–22.0% in southern Europe, relative to 13.0–25.0% among older
people in northern and central Europe. Moreover, the proportion of adults being
identified with more than 4 indicators of social isolation was the highest in Sweden
(25.2%) and the lowest in Greece (8.8%). Conclusions According to the afore-mentioned cross-national empirical findings the
social factors under study were found to be significantly associated with specific
positive well-being outcomes and their multiple clustering, as well. In addition,
preventive health services utilization was significantly related to different indicators
of social isolation. The main conclusions that could be drawn are as follows: (i) wellbeing
outcomes are socially distributed, with individuals with the least years of
education and the lowest household income level being ascertained with the lowest
likelihood of presenting multiple clustering of well-being indicators, (ii) well-being
dimensions are differently distributed among the eleven European countries and the
three geographical regions under investigation, which is in accordance with the wellestablished
well-being “north-south pattern”, with significantly better outcomes being
consistently attested among Northern Europeans, (iii) frequent productive and social
activity participation is significantly related to well-being, (iv) specific elements of
people’s social environment which pertain to social isolation are significantly related
to well-being outcomes, (v) preventive health services utilization is associated with
social living conditions and social isolation, (vi) non lonely older adults present better
well-being outcomes and have a greater well-being mean score, whereas multiple
clustering of well-being indicators seems to be less common among individuals
declaring to feel lonely most of the time, (vii) frequency of reporting feelings of
loneliness is significantly related to specific adverse health conditions, stressful life
events and social isolation indicators and, (viii) recent departure of the last offspring
from parental nest was the most significant independent predictor of loneliness.
It becomes evident from the above findings that there seem to be specific factors of
older people’s social and family context which are significantly associated with their
level of well-being and can possibly hold beneficial implications for multiple wellbeing
aspects. Several of these factors could be subjected to modification through to
suitable interventions which could lead to the amelioration of individuals’ well-being
as they age. Particularly, the enhancement of chances for active engagement in social
activities and the encouragement of an active lifestyle could be to the benefit of older
people’s well-being and respective strategies should therefore be oriented towards
satisfying their needs for social integration and meaningful social connections. In
addition, social and public health policies aiming at the improvement of later-life
well-being ought to prioritize the mitigation of psycho-social distress and human pain
which is reflected on the occurrence of loneliness as a key risk factor for health and
well-being. Furthermore, the current results afford important empirical evidence on
identifying factors which possibly bear upon social engagement, social isolation and
loneliness and could thus extend current knowledge and understanding on well-being
promoting factors and risk factors for poor well-being outcomes in old age.
|