INT’L. J. AGING AND HUMAN DEVELOPMENT, Vol.

78(3) 229-257, 2014

WHO ARE LONELY? LONELINESS IN DIFFERENT

AGE GROUPS (18-81 YEARS OLD), USING TWO
MEASURES OF LONELINESS*

MAGNHILD NICOLAISEN
Ageing and Health, Norwegian Centre for Research,
Education and Service Development
KIRSTEN THORSEN
Ageing and Health, Norwegian Centre for Research,
Education and Service Development;
University College of Buskerud and Vestfold; and
Norwegian Social Research (NOVA)

ABSTRACT
This study asks if the prevalence of loneliness in the population varies
depending on the measures used, with special focus on loneliness among
the elderly. The study compares loneliness in different age groups between
18 and 81 years old (N = 14,743) using two measures of loneliness: the
(indirect) six-item De Jong Gierveld Loneliness Scale and a single-item,
direct question about loneliness. Data are from the Norwegian LOGG
(Life Course, Generation, and Gender) study. We compare the findings on
loneliness according to age, gender, health, and partner status. Overall,
the two loneliness measures indicate a similar prevalence of loneliness,

*The NorLAG and LOGG surveys are financed by the Research Council of Norway
(grant no. 149564 and 168373), Ministry of Health and Care Services, Ministry of Labour,
Ministry of Children, Equality and Social Inclusion, Ministry of Local Government and
Regional Development, Norwegian Social Research (NOVA) and Statistics Norway. The
LOGG and NorLAG datasets are part of the ACCESS Life Course infrastructure project
funded by the National Financing Initiative for Research Infrastructure at the Research
Council of Norway (grant no. 195403) and NOVA.

229

Ó 2014, Baywood Publishing Co., Inc.

doi: http://dx.doi.org/10.2190/AG.78.3.b
http://baywood.com
230 / NICOLAISEN AND THORSEN

but attribute loneliness to somewhat different people. When using a

direct measure, loneliness is more prevalent among women; when using
the (indirect) De Jong Gierveld Scale, loneliness is more prevalent among
men. Also, the association between age and loneliness differed when
using the direct and the indirect measure.

“How, if at all, can we identify who is lonely and who is not?” (Victor, Bond,
& Scambler, 2009, p. 55). This question is still posed, even after decades of
research into the subject of loneliness. Loneliness is important to the quality
of life (Shiovitz-Ezra, Leitsch, Graber, & Karraker, 2009), and is strongly
associated with aspects of mental health, such as depression (Cacioppo,
Hawkley, & Thisted, 2010; Heikkinen & Kauppinen, 2011; Paul, Ayis, &
Ebrahim, 2006; Routasalo, Savikko, Tilvis, Strandberg, & Pitkälä, 2006;
Thorsen & Clausen, 2008). Loneliness is also associated with—but not
equivalent to—social isolation. Peplau and Perlman (1982, p. 3), in their
summary of the common elements of loneliness, still influential after more
than 30 years, state: “First, loneliness results from deficiencies in a person’s
social relationships. Second, loneliness is a subjective experience; it is not
synonymous with objective social isolation. People can be alone without
being lonely, or lonely in a crowd. Third, the experience of loneliness is
unpleasant and distressing.” Thus, in contrast to the objective state of social
isolation (being alone), loneliness is a subjective experience.
Important questions are: In the population, which people are lonely? Are
older people more lonely than younger people, or is it the other way around?
Are women more lonely than men? What factors are associated with loneli-
ness in the different age groups, indicating possible causes and different factors
influencing loneliness across age groups? At another level the question can
be asked: Are the answers to these questions and the picture we get of
loneliness in the population different depending on what measures are used?
More specifically: do they vary between “direct” and “indirect” measures?
Different measures have been used in loneliness research. However, few
studies have directly compared prevalences of loneliness yielded by these
measures. This article investigates rates of loneliness in men and women
across a range of age groups using two different measures:

1. a direct question about loneliness; and

2. the six-item De Jong Gierveld Loneliness Scale.

The study compares the empirical results in age groups 18-81 years old using
the two measures.
WHO ARE LONELY? / 231

WHAT IS LONELINESS?
THEORIES AND CONCEPTS OF LONELINESS

In most research on loneliness, being alone and being lonely are not
synonymous, but a lack of social contact and support are key antecedents
that can lead to loneliness. Weiss (1973), in his classical and still influential
study of loneliness, distinguished between two forms of loneliness: the experi-
ences of emotional and social isolation. The first form indicated missing
intimate social relations, the second the lacking of a wider social network.
The distinction between emotional and social loneliness may be particu-
larly relevant for studies involving older people because, due to stressful life
events such as the death of a spouse/partner and aging relatives or friends,
the probability of having or finding an intimate attachment figure decreases
with age (van Baarsen, Snijders, Smit, & van Duihn, 2001, p. 133). However,
studies indicate that even if the number of peripheral social contacts is greatly
reduced in very old age, there is little difference in the number of emotionally
close partners reported (Field & Minkler, 1988; Lang & Carstensen, 1994).
Across adulthood, an increasingly larger percentage of the network is emo-
tionally close social partners (Carstensen, Isaacowitz, & Charles, 1999).
In contrast to an explanation that focuses on losses in old age reducing the
size of the social network, socioemotional selectivity theory suggests that,
ideally, older people prefer to have relatively small networks, emphasizing
emotionally close social partners. The importance of emotional ties increases
in old age (Carstensen et al., 1999). Younger people, by contrast, have rela-
tively larger, more diverse networks that include a high proportion of relatively
new social partners. But, as people get older, they become increasingly aware
that time is “running out.” More social contacts feels superficial—in contrast
to deepening ties of close relationships and pursuing emotionally meaningful
goals (Carstensen et al., 1999).

DIFFERENT MEASURES OF LONELINESS

Survey researchers have used at least two different techniques in interviews

to measure feelings of loneliness: a direct self-rating question and indirect
scales. A common approach to the measurement of loneliness is a single
question with a rating response scale (Victor, Grenade, & Boldy, 2005). The
single question is known as the direct way of measuring loneliness because
the word “lonely” or “loneliness” is included in the question (Shiovitz-Ezra
& Ayalon, 2012). The direct self-report measure is frequently used in Britain,
continental Europe, and North America. It is simple to use and appears to
be highly acceptable to participants (Victor, Grenade, et al., 2005). This way of
232 / NICOLAISEN AND THORSEN

measuring loneliness is a phenomenological approach—letting people decide

for themselves what the concept of loneliness means to them. The meaning
of loneliness may thus vary among the respondents. When using a single-
item direct question, the researcher must accept and consider the individual’s
contextual aspects of loneliness (Jylhä, 2004). It has been suggested (Victor
et al., 2009; Victor, Grenade, et al., 2005) that the self-report measure is more
appropriate than composite scales when older people are the subjects of the
research. However, after looking at the studies referred to (Fees, Martin, &
Poon, 1999; Holmén & Furukawa, 2002; Russell, Peplau, & Cutrona, 1980),
we cannot see that the articles mentioned present strong evidence for this.
An important objection to the use of direct questions (single-question self-
rating scales, mentioning the term “loneliness”) to investigate loneliness is
that because of the social stigma linked to the phenomenon of loneliness,
using a direct question will result in under-reporting, especially among men
(de Jong Gierveld, van Tilburg, & Dykstra, 2006).
Because of the assumed problem of under-reporting, a number of scales
that approach the topic indirectly, without using the word loneliness, have been
developed (Victor, Scambler, Marston, Bond, & Bowling, 2005; Victor et al.,
2009). One widely used scale is the UCLA Loneliness Scale (Russell, 1996), a
unidimensional, 20-item Likert scale. Another widely used scale is the De Jong
Gierveld Loneliness Scale (de Jong Gierveld & Kamphuis, 1985; de Jong
Gierveld & van Tilburg, 1999). The De Jong Gierveld Loneliness Scale is
based on a multidimensional conceptualization of loneliness and influenced
by Weiss’s (1973) view that there are two basic forms of loneliness; emo-
tional and social. The 1985 version (de Jong Gierveld & Kamphuis, 1985)
consists of 11 items, 5 positively phrased, 6 negatively phrased. The De Jong
Gierveld Scale can be used as an unidimensional loneliness scale or as two
separate subscales (one for emotional and one for social loneliness) (de Jong
Gierveld et al., 2006). For use in large surveys the shorter six-item version
was constructed.
International comparative research has shown that both the 11-item scale
(van Tilburg, Havens, & de Jong Gierveld, 2004) and the short scale (de Jong
Gierveld & van Tilburg, 2010) are suitable for use in different locations
and countries. The (newer) short scale has been used in a number of studies
(e.g., Burke, Schnittger, O’Dea, Buckley, Wherton, & Lawlor, 2012; Fokkema
& Naderi, 2013; Han & Richardson, 2010; Leung, de Jong Gierveld, & Lam,
2008; McHugh, Casey, & Lawlor, 2011; Prieto-Flores, Forjaz, Fernandez-
Mayoralas, Rojo-Perez, & Martinez-Martin, 2011). Recently, some items
from the scale have also been used in a Norwegian study (Hansen, Slagsvold,
& Ingebretsen, 2012) based on LOGG-data (the same dataset as the one used
in this article).
WHO ARE LONELY? / 233

The De Jong Gierveld Loneliness Scale has also met with some criticism.
One strand points at the lack of precision in the items in the scale, which may
result in aspects of mental health and well-being other than loneliness being
measured (Shaver & Brennan, 1991). Kirova (2003) suggests that multiple
question scales focus more on measuring individuals’ perceived social deficits.
Items like “I often feel rejected” and “I experience a general sense of empti-
ness” may refer to other mental states—such as depression, for instance.
Further, asking about numbers of close relationships (many, plenty, enough)
may trigger a negative feeling of being socially unsuccessful and make people
think that they “should” have more contacts. Thus, also items included in the
multi-question loneliness scales may elicit “socially acceptable” responses.
Given the decision to measure loneliness indirectly, nevertheless most
measures (like the De Jong Gierveld Scale) are validated by showing that
they correlate with direct self-reports of loneliness (Shaver & Brennan,
1991, p. 248).

PREVIOUS RESEARCH ON LONELINESS ASSOCIATED

WITH AGE, GENDER, HEALTH, MARITAL STATUS,
AND LIVING ARRANGEMENTS
In our society, social structural characteristics influence the opportunities
that people have to create and maintain an optimal network of social relation-
ships (de Jong Gierveld, 1987). Factors such as age, gender, partner status,
and health are main variables used in loneliness studies (e.g., Aartsen
& Jylhä, 2011; Fokkema, de Jong Gierveld, & Dykstra, 2012; Sundström,
Fransson, Malmberg, & Davey, 2009; Victor & Bowling, 2012; Victor &
Yang, 2012). These factors have been shown to have an impact on loneliness,
and are therefore applied in this study.
The relationship between age and loneliness seems inconclusive. Reviews
of cross-sectional studies show that loneliness is more prevalent among
older adults, particularly those of advanced age (Dykstra, 2009; Dykstra,
van Tilburg, & de Jong Gierveld, 2005). Multivariate analyses show that their
loneliness is mainly the product of factors such as the loss of partners and
friends due to death, which contribute to a reduced social network. The most
significant factor is the loss of one’s partner. Other studies, however, find that
younger and older people have the highest risk of experiencing loneliness
and that the curve for the prevalence of loneliness is U-shaped (Victor & Yang,
2012). In the younger population, personality characteristics have proved
to be more important than among the elderly (de Jong Gierveld, 1998).
Also, the association between gender and loneliness is inconclusive.
Using the direct measure of loneliness, there is considerable agreement that
234 / NICOLAISEN AND THORSEN

women are consistently more likely to report being lonely than men (Jylhä,
2004; Pinquart & Sörensen, 2001; Savikko, Routasalo, Tilvis, Strandberg,
& Pitkälä, 2005; Victor & Yang, 2012). The finding that men are less lonely
appears to be related to men’s greater reluctance to disclose socially
undesirable feelings (Borys & Perlman, 1985). Some studies find that
female gender is independently associated with loneliness (Savikko et al.,
2005), others find no gender difference once other factors like marital
status, age, health, and living arrangements are controlled for (Aartsen &
Jylhä, 2011; Jylhä, 2004; Nicolaisen & Thorsen, 2012; Victor, Scambler, et al.,
2005). When measures that avoid the word loneliness are used, such as
the De Jong Gierveld Loneliness Scale or the UCLA Loneliness Scale, no
gender differences are consistently found (Pinquart & Sörensen, 2003).
Some studies, using separate scales for social and emotional loneliness,
with the aid of the De Jong Gierveld subscales of loneliness, find that
men appear to be less emotionally lonely but more socially lonely than
women (de Jong Gierveld & van Tilburg, 2010; Dykstra & de Jong Gierveld,
2004). Steed, Boldy, Grenade, and Iredell (2007) found that gender differences
regarding social networks and loneliness emerged only when loneliness was
measured by the De Jong Loneliness Scale.
One challenge in measuring the impact of social networks on loneliness—
applying the 11-item De Jong Gierveld Scale—relates to the fact that contact
with friends and confidants is included in this measure of loneliness.
The six-item scale does not have questions which ask specifically about contact
with friends but explores friendly social relations by asking—for example—
about people one trusts completely, people one can lean on when one has
problems, and people one feels close to (see the Methods section for the
wording of the six-scale items), with answers indicating satisfaction with
the number of people.
Regardless of the measure of loneliness that is used, marital status and
living arrangements are consistently found to be correlated with loneliness.
Those who are married or living with someone are less likely to be lonely
than those who live alone (Nicolaisen & Thorsen, 2012; Pinquart & Sörensen,
2003; Sundström et al., 2009; Victor & Yang, 2012).
Being in poor health (by various measures) is also associated with loneli-
ness. The association between health and loneliness seems to go both ways:
being in poor health predicts loneliness (Cohen-Mansfield, Shmotkin, &
Goldberg, 2009; Tijuis, de Jong Gierveld, Feskens, & Kromhout, 1999),
and loneliness predicts poor health (Luo, Hawkley, Waite, & Cacioppo, 2012;
Nummela, Seppänen, & Uutela, 2011). Subjective health is a commonly
used measure in loneliness studies. It seems to be a summary of one’s status
in many health-related domains (Deeg & Bath, 2003; Moum, 1992).
WHO ARE LONELY? / 235

PREVIOUS COMPARISONS OF LONELINESS

BY DIFFERENT MEASURES

Previous studies using and comparing different measures of loneliness

directly, that is, using different measures of loneliness in the same survey and
explicitly comparing the results, are few. In an Australian study, Victor,
Grenade, et al. (2005) compared the results of two approaches to measuring
loneliness, a self-report (direct, global measure) and an aggregate measure
(the De Jong Gierveld Loneliness Scale). Victor, Grenade, et al. (2005)
focused on a comparison of prevalence estimates yielded by the two loneliness
measures for key demographic characteristics and discuss possible reasons
for differences. However, their sample is restricted in number (N = 353) and
includes only older people (65 and over). Victor, Grenade, et al. (2005) found
that the prevalence of loneliness was remarkably similar in the De Jong
Gierveld Scale and the self-report questionnaire, especially at the severe
end of the distribution. However, the De Jong Gierveld Scale generated a
higher percentage in the intermediate loneliness category and a subsequent
lower percentage in the “never” lonely group, compared to the direct single-
item measure.
Using the same Australian sample as Victor, Grenade, et al. (2005), Steed
et al. (2007) explored the prevalence of loneliness, employing three measures:
a direct question of loneliness, the De Jong Gierveld Loneliness Scale, and
the UCLA Loneliness Scale. In addition to exploring prevalence, Steed et al.
(2007) aimed to identify the demographic factors associated with loneliness.
They found no gender differences when using the UCLA Loneliness Scale.
When measuring loneliness directly, women were more lonely than men.
When using the De Jong Gierveld Scale, no significant gender difference
in loneliness was found. Steed et al. found no evidence of a relationship
between age and loneliness (using either measure).
A third relevant study (Shiovitz-Ezra & Ayalon, 2012) compared and
explored the degree of concordance between two measurement approaches—
a single direct question and a multi-item scale (the shortened version of the
UCLA Loneliness Scale)—and investigated whether the same people were
classified as lonely by each approach. Shiovitz-Ezra and Ayalon (2012) studied
people aged 55 and over, using a large U.S. sample of approximately 2,000
observations. Many of the respondents (57%) who reported being lonely on
the direct item were classified as not lonely on the indirect scale.
The samples in these three studies are restricted in age, all of them study
people in the second half of life, and have a rather restricted number of
respondents. How the results concerning loneliness among the elderly
compare with those from younger age groups, and the factors related to
236 / NICOLAISEN AND THORSEN

loneliness, will deepen our understanding of loneliness among people in the

different stages of life. But the question arises: will the results vary with
the measures applied?

THE AIM OF THE STUDY

The primary aim of the study is to analyze loneliness and related factors
in different age groups in a large Norwegian sample with an age range of
18-81 and 14,743 respondents. The large sample makes it possible to include
all adult age groups, young, middle-aged, and older adults, and compare
loneliness rates, with special focus on the loneliness of the elderly. Our study
has a much larger sample than previous studies. Loneliness studies have
used different measures, direct and indirect. The study compares prevalence
rates of loneliness in different subgroups, using one direct and one indirect
measure of loneliness: a global single-item direct question on loneliness
and the six-item De Jong Gierveld Loneliness Scale. The analyses compare
rates of loneliness in different age groups, measured by the two approaches,
to explore whether the results differ according to the measures used. We
investigate associations between loneliness and age, gender, health, and
partner status. Do the two measures yield similar or different findings on
loneliness in different age groups of men and women? And is one of the
dimensions (emotional or social loneliness) more dominant in some groups
than in others?

METHOD

Data
The study uses data from the cross-sectional LOGG (Life Course, Gener-
ation and Gender) study conducted in 2007-2008 with a sample of people
18–81 years old at the time of the interview. LOGG merges two studies: the
second wave of the Norwegian Life Course, Ageing and Generation
study (NorLAG) and the first wave of the Generations and Gender Survey
(GGS-Norway). The LOGG study uses a combination of telephone interviews
(CATI) and postal questionnaires. The response rate to the telephone inter-
view was 60.0%, and 72.5% of those interviewed by telephone subsequently
answered the postal questionnaire. Response rates were lower in the youngest
(18-29) and the oldest (70-79) groups and among persons with lower education
(Bjørshol, Høstmark, & Lagerstrøm, 2010). The sample does not include
people living in institutions. This study uses data from those who answered
all the loneliness questions (N = 14,743). The data on loneliness and health are
WHO ARE LONELY? / 237

from the telephone interview. Age and marital status/partnership are data
taken from public records (Statistics Norway), added with the respondents’
informed consent.
The data analyses are performed on four age groups, 18-29 years old, 30-49
years old, 50-64 years old, and 65-81 years old. In general, many of the
youngest group (18-29 years old) are students, do not have children, and have
not yet “settled down.” Most people 30-49 years old are in the middle of
their working careers. In this age group, most people still have children
living in the household. In the age group 50-64 years old, most households
are smaller, as children have grown up and moved out. Some people have left
the work force, most of them because of health problems. In the oldest group,
65-81 years old, employment rates are low because most people have left
the work force. As people approach advanced age, deteriorating health and
the loss of a partner become more prevalent.

Measuring Instruments
Measures of Loneliness

The study used two measures of loneliness: a direct single question and
a multiple-question scale, the six-item De Jong Gierveld Loneliness Scale
(de Jong Gierveld & van Tilburg, 2006). Whereas the De Jong Gierveld
measure assesses severity of loneliness, the single-item measure probes fre-
quency of loneliness (Iecovich, 2013). The direct single question assesses
loneliness by the global question: “Do you feel lonely?” Response categories
are 1 = “often,” 2 = “sometimes,” 3 = “seldom,” or 4 = “never.” The values
were recoded so that higher scores indicate more loneliness.
The De Jong Gierveld Loneliness Scale (as used in the GGS) encompasses
two components of loneliness: emotional and social. None of the items refers
directly to loneliness and the word “loneliness” was not used in the set of items.
The items of the scale related to emotional loneliness are three negatively
formulated items: “I experience a general sense of emptiness,” “I miss having
people around,” and “Often, I feel rejected.” The items of social loneliness
are three positively formulated items: “There are plenty of people that I can
lean on in case of trouble,” “There are many people that I can count on
completely,” and “There are enough people that I feel close to.” Answer
categories for the De Jong Gierveld scale varies according to how the data
are collected. The response categories are “yes!,” “yes,” “more or less,”
“no,” and “no!,” or “yes,” “more or less,” and “no,” and differs for face-to-face
interviews, telephone interviews, and mail questionnaires. In the LOGG
study, the items had five similar response categories: “strongly agree,” “tend to
agree,” “neither agree nor disagree,” “tend to disagree,” “strongly disagree.”
238 / NICOLAISEN AND THORSEN

Processing the scale data entailed counting neutral and positive answers
(“neither agree nor disagree,”1 “strongly agree,” and “tend to agree”) on the
negatively formulated items (measuring emotional loneliness) and counting
neutral and negative answers (“neither agree nor disagree,” “strongly dis-
agree,” and “tend to disagree”) on the positively formulated items (measuring
social loneliness).2 The resulting loneliness scale scores range from 0
(not lonely) to 6 (intensely lonely).
The reliability coefficient (Cronbach’s alpha) is 0.67 for the total
sample. The alphas of the four age groups in our study, 18-29, 30-49,
50-64, and 65-81 are, respectively: 0.67, 0.68, 0.67, and 0.63. That is, for
most groups, the reliability scores are about 0.7 and are thus considered to
be quite good.3
Using the De Jong Gierveld Scale, those who score 2 or more experience
loneliness (to various degrees). This dichotomization is in accordance with
previous studies (Fokkema & Naderi, 2013). Similarly, using the direct
one-item measure of loneliness, those who experience loneliness often or
sometimes are considered as “lonely” (to various degrees) (i.e., Holmén
& Furukawa, 2002; Thorsen, 1990; Tiikkainen & Heikkinen, 2005). In the
literature, we have not found further cut-points on the short De Jong Scale,
comparable to the responses to the direct one-item measure of loneliness.
Hence, we have used the dichotomized versions when comparing the fre-
quencies of the two measures directly.

Independent Variables

Respondent’s age is defined as the number of complete years lived at the

time of the interview. Partner status is measured by a dichotomy coded 1 =

1 The “neither agree nor disagree” answers (equivalent to “more or less”) are not
considered to be neutral answers but indicators of loneliness (de Jong Gierveld &
Kamphuis, 1985), given the social stigma associated with loneliness (Dykstra et al., 2005).
2 de Jong Gierveld and van Tilburg (1999) note that in view of the available computer
programs, they had to dichotomize the item scores. New releases of the computer programs
allow multi-categorical item scores. For the time being, de Jong Gierveld and van Tilburg
(1999) prefer the scale score based on dichotomous scores, which facilitate comparison
of results with earlier studies. However, some recent studies (using the 11-item scale)
have used scale scores based on multi-categorical item scores (Iecovich, 2013).
3 When using the original five-item scores, the reliability/internal validity is better:
For the unidimensional, overall loneliness scale, reliability of internal consistency
showed a Cronbach’s a of 0.75 calculated with the original multi-categorical (five
response) item scores (strongly agree–strongly disagree), compared to an a of 0.67 with
the dichotomized item scores.
WHO ARE LONELY? / 239

married/cohabitant, and 0 = not married/not cohabitant (including those who

are divorced/separated or widowed). Subjective health is assessed by the
question: “How would you best describe your current health?” Answer cate-
gories range from 1 = excellent to 5 = poor.

Analyses
The data were analyzed using SPSS version 19.0. In the analyses we measure
loneliness by a direct measure of loneliness and by the six-item De Jong
Gierveld Loneliness Scale and explore and compare the associations between
loneliness by the respective two measures and key correlates including
gender, age, subjective health, and partner status. Blockwise multiple regres-
sion analysis is applied to assess the relative significance of the independent
variables for loneliness by the two measures. Age is introduced at step 1.
At step 2, gender is entered. At step 3, health is entered. At step 4, partner
status is entered into the model. Each model is estimated separately for the
four age groups under study (18-29, 30-49, 50-64, and 65-81). An alpha
level of 0.05 is used for all statistical tests.
Both parametric and non-parametric tests were performed because loneli-
ness score distributions were skewed. That is, most respondents do not feel
lonely and/or have low scores on the De Jong Gierveld Scale. Since these
tests revealed similar results, only parametric test results are presented.

RESULTS
The mean age of the sample is 46.41 (SD 16.01). The sample consists
of slightly more women than men (50.7% vs. 49.3%), and the majority
(55.5%) are in excellent or very good health. Most respondents are married
or cohabitant (67.4%) (see Table 1).
Table 2 displays the frequency of loneliness, measured by the six-item
De Jong Gierveld Loneliness Scale, and by the the direct single question.
According to the De Jong Gierveld Scale, 24% experience loneliness (to
various degrees, score 2-6). According to the direct measure, 21% experience
loneliness often or sometimes.
Overall, the prevalence of loneliness seems quite similar when using the
De Jong Gierveld Scale and the direct self-rating measure. Descriptive statistics
of the scores in the different age groups under study (Table 3) reinforce this
impression of similarity in the results.
However, three out of four age groups seem to be slightly lonelier when
using the De Jong Gierveld Scale than when using the direct measure
of loneliness. The only exception is the youngest group, aged 18-29. But
240 / NICOLAISEN AND THORSEN

Table 1. Characteristics of the Sample

Variables N %

Total sample 14,743 100

Gender
Men 7,268 49.3
Women 7,475 50.7

Age
18-29 2,550 17.3
30-49 5,946 40.3
50-64 3,948 26.8
65-81 2,299 15.6

Subjective health
Excellent 3,542 24.0
Very good 4,638 31.5
Good 3,489 23.7
Fair 2,330 15.8
Poor 729 4.9

Marital status
Married/cohabitant 9,944 67.4
Not married/cohabitant 4,799 32.6

age patterns of prevalence vary somewhat: age correlates significantly with

loneliness using both the De Jong Gierveld Scale (p < 0.001) and the direct
measure of loneliness (p < 0.001), however in different ways. When using the
direct measure, the oldest group is most lonely, followed by the youngest
group, and then the two groups in the middle (50-64 years old and 30-49
years old) (Table 3). When using the De Jong Gierveld Scale, people are
lonelier the older they are. That is, there is a significant linear trend where
loneliness is most severe in the oldest group.
The correlation between the scores on the direct question and the short
De Jong Gierveld Loneliness Scale was positive and significant, r = 0.47,
p < 0.001, but rather modest. A comparison of the two measures’ classification
of respondents as lonely and not lonely show that 12% (1,727 respondents)
are characterized as lonely by both measures, while 67% (9,899 respondents)
of our sample are characterized as not lonely by both measures. The remainder
WHO ARE LONELY? / 241

Table 2. Loneliness Measured by the Short De Jong Gierveld

Loneliness Scale and by a Direct Single Question: “Do You Feel Lonely?”

De Jong Gierveld

Score Frequency Percent

No loneliness 8,105 55.0

1 3,128 21.3
2 1,731 11.7
3 901 6.1
4 474 3.2
5 261 1.8
Severe loneliness 133 0.9
Total 14,743 100.0

Direct single question

Response Frequency Percent

Never 6,141 41.7

Seldom 5,531 37.5
Sometimes 2,678 18.2
Often 393 2.7
Total 14,743 100.0

(21%) is characterized as lonely by one measure but not lonely by the other
measure (table not shown).
Also, the association between gender and loneliness differs depending on
which measure we use. If we measure loneliness with the direct single question,
women are significantly more often lonely than men. Using the De Jong
Gierveld Scale, men are significantly lonelier than women in the younger
groups (18-29 years old and 30-49 years old), but not in the older groups (50-64
years old and 65-81 years old) (table not shown).
We examined whether there was a gender difference in emotional loneli-
ness and social loneliness respectively in our material, using cross-tabulations
and chi-square significance tests. The results from these additional analyses
showed that men are more socially lonely than women in all age groups, and
242 / NICOLAISEN AND THORSEN

Table 3. Loneliness in Different Age Groups:

Descriptive Statistics of the Short De Jong Gierveld Loneliness Scale
and of the Direct Single Question: “Do You Feel Lonely?”

De Jong Gierveld

Total
Variable sample 18-29 30-49 50-64 65-81

% with feelings of 23.7 21.2 22.3 23.8 30.2

loneliness (2-6)

% with maximum score 0.9 0.7 1.0 0.9 0.8

on loneliness scale (6)

Direct single question

Total
Variable sample 18-29 30-49 50-64 65-81

% with feelings of 20.9 22.9 18.7 20.0 25.6

loneliness (sometimes
or often)

% with maximum score 2.7 2.1 2.2 3.2 3.7

on loneliness (often)

especially among those aged 30-49 (Chi square = 71.00, p < .001) and 50–64
(Chi square = 32.12, p < .001). Women are somewhat more emotionally lonely
than men, but only in the two oldest age groups, among those aged 50-64
(Chi square = 8.66, p = .034) and 65-81 (Chi square = 29.61, p < .001).

Multivariate Analysis

Next, we study the associations between the two measures of loneliness

and the factors included in our study (age, gender, partner status, and health).
Which of the factors included account for the most of the explained variance
in loneliness? Blockwise multiple regression was used to assess the relative
significance of the independent variables for loneliness measured, respec-
tively, by the six-item De Jong Gierveld Loneliness Scale (Table 4) and by
the direct single question (Table 5).
WHO ARE LONELY? / 243

In the total sample, when measuring loneliness by the De Jong Gierveld

Scale, health is the number 1 predictor, accounting for the most of the explained
variance in loneliness for the total sample (Table 4). Partner status is number 2,
age is number 3, and gender is number 4. Regarding the age-groups: when
health, partner status, and gender are controlled for (step 4), the group aged
30-49 is the loneliest, followed by the group aged 65–81, and then the group
aged 50–64. All three age groups are significantly more lonely than those
aged 18–29 (used as the reference group).
When feelings of loneliness are measured by the direct single-item question,
partner status is the most important predictor of loneliness (Table 5). Health
is number 2, gender number 3, and age is number 4. When health, partner
status, and gender are controlled for, the oldest age group (65-81) is sig-
nificantly less lonely than the youngest group (18-29). This is not the case
for the age groups 30-49 and 50-64.
Which of the factors included in our study (age, gender, partner status, and
health) accounts for more variance in loneliness in the different age groups?
Using either measure, in most age groups, partner status and health are the
two factors most strongly related to loneliness. In the groups aged 18-29, 30-49,
and 50-64, poor health is the most important predictor measured by the De Jong
Gierveld Scale (Table 4). Measured by the direct single item, partner status
is the number 1 predictor (Table 5). By both measures, we find that in the
oldest age group (65-81), not having a spouse or cohabitant is the most
important predictor of loneliness.
Concerning gender, we performed an ancillary analysis (tables not shown),
investigating which variables were the most important predictors of loneliness
for men and women in different age groups when using the two loneliness
measures. The results corresponded for men and women, and were similar to
the results for the whole sample, with two exceptions: when using the De Jong
Gierveld Scale, only poor health was significantly associated with loneliness
for young women (18-29 years old); for the men in the youngest group, both
health (number 1) and partner status (number 2) were significantly related
to loneliness.

CONCLUSIONS AND DISCUSSION

Loneliness among older people is a theme with many aspects. It is related
to mental and physical health, quality of life and well-being, and will influence
both the needs for and effect of health services and care, and the need for
social support among the elderly.
People in different age groups—from young to old age—may experience
loneliness differently, and the factors associated with loneliness may vary, and
Table 4. Predictors of Loneliness Measured by the De Jong Gierveld Loneliness Scale among
People in Different Age Groups
De Jong Gierveld Loneliness Scale
Total Sample

Model 1 Model 2 Model 3 Model 4

Variables B SD Beta p B SD Beta p B SD Beta p B SD Beta p

Age 30-49a .06 .03 .02 .050 .06 .03 .02 .050 .01 .03 .00 .738 .24 .03 .09 <.001

Age 50-64a .11 .03 .04 .001 .11 .03 .04 .001 –.07 .03 –.02 .045 .16 .03 .05 <.001
244 / NICOLAISEN AND THORSEN

Age 65-81a .31 .04 .09 <.001 .31 .04 .09 <.001 .07 .04 .02 .085 .23 .04 .07 <.001

Female –.09 .02 –.04 <.001 –.12 .02 –.05 <.001 –.14 .02 –.06 <.001

Health .26 .01 .23 <.001 .25 .01 .23 <.001

Partner –.52 .02 –.19 <.001

R square .005 .007 .056 .087

R square .005 .001 .049 .031

change

F change 26.38 19.23 769.40 498.05

(p-value) (<.001) (<.001) (<.001) (<.001)

18-29 years
Model 1 Model 2 Model 3 Model 4
Variables B SD Beta p B SD Beta p B SD Beta p B SD Beta p

Age .00 .01 .01 .548 .01 .01 .01 .519 –.00 .01 –.00 .856 .01 .01 .03 .127
Female –.12 .05 –.05 .019 –.15 .05 –.06 .002 –.11 .05 –.05 .022
Health .28 .02 .23 <.001 .28 .02 .22 <.001
Partner –.23 .06 –.09 <.001
R square .000 .002 .053 .059
R square .000 .002 .050 .006
change
F change .361 5.56 135.38 16.36
(p-value) (.548) (.019) (<.001) (<.001)
30-49 years
Age .00 .00 .01 .549 .00 .00 .01 .521 –.00 .00 –.01 .441 .00 .00 –.00 .897
Female –.17 .03 –.07 <.001 –.21 .03 –.08 <.001 –.21 .03 –.08 <.001
Health .28 .02 .24 <.001 .27 .02 .23 <.001
Partner –.55 .04 –.18 <.001
R square .000 .004 .059 .090
R square .000 .004 .055 .031
change
WHO ARE LONELY?

F change .36 25.59 343.93 202.42

(p-value) (.549) (<.001) (<.001) (<.001)
/ 245
Table 4. (Cont’d.)
50-64 years
Model 1 Model 2 Model 3 Model 4
Variables B SD Beta p B SD Beta p B SD Beta p B SD Beta p
Age .01 .01 .04 .018 .01 .01 .04 .020 .00 .01 .01 .459 .00 .01 .01 .482
Female –.07 .04 –.03 .079 –.10 .04 –.04 .018 –.16 .04 –.06 <.001
Health .24 .02 .22 <.001 .23 .02 .21 <.001
Partner –.65 .05 –.21 <.001
R square .001 .002 .050 .093
R square .001 .001 .047 .043
246 / NICOLAISEN AND THORSEN

change
F change 5.61 3.08 196.42 188.14
(p-value) (.018) (.079) (<.001) (<.001)
65-81 years
Age .02 .01 .08 <.001 .02 .01 .08 <.001 .02 .01 .06 .002 .01 .01 .04 .072
Female .09 .06 .03 .124 .07 .06 .03 .202 –.09 .06 –.03 .108
Health .23 .02 .20 <.001 .22 .02 .19 <.001
Partner –.65 .06 –.23 <.001
R square .007 .008 .047 .094
R square .007 .001 .039 .048
change
F change 15.01 2.36 93.66 120.94
(p-value) (<.001) (.124) (<.001) (<.001)
aReference category: age 18-29.
WHO ARE LONELY? / 247

also depend on what loneliness measures are used. Studies exploring the
impact of different measures have focused on elderly people (Shiovitz-Ezra
& Ayalon, 2012; Victor, Grenade, et al., 2005). This study has empirically
used two frequently used measures of loneliness, a single direct question of
loneliness and the (indirect) six-item De Jong Gierveld Loneliness Scale,
comparing loneliness among men and women in different age groups (18-29,
30-49, 50-64, 65-81). Overall, the results from using the direct and indirect
measures indicated a similar prevalence of loneliness. However, we find that
the two different measures seem to provide rather different pictures of loneli-
ness in different age groups, ranging from 18 to 81 years. This age-range is
larger than the ones used in previous studies measuring loneliness among
elderly people both by a direct and by an indirect measure of loneliness
(Shiovitz-Ezra & Ayalon, 2012; Victor, Grenade, et al., 2005).
When applying the direct measure of loneliness in the present study, the
youngest (18-29 years old) and the oldest (65-81 years old) groups reported
loneliness most often. Applying the De Jong Gierveld Scale, we found a
significant and positive association between age and loneliness; the older
the more lonely.
As people reach older age, various changes and events—for example,
becoming retired, the deterioration of health and functional capacity, the
deaths of one’s partner and/or peers—become more likely and more prevalent,
resulting in more limited networks. Although de Jong Gierveld states that
loneliness is a subjective experience and, as such, is not directly related to
situational factors (de Jong Gierveld & van Tilburg, 1999), several scale items
(the social loneliness items) emphasize and relate to the size of the respondent’s
network in different aspects of social relations. Thus, the scale may indicate
feelings of social deficiency, feelings of insufficient social relationships (or
feelings of not living up to social expectations) not always experienced as
loneliness. The same may apply to a sense of emptiness or a feeling of rejection.
Furthermore, our multivariate analyses yielded different rankings of the
(key) factors associated with loneliness by the two measures, both for the whole
sample and for the different age groups under study. Overall, partner status
showed the strongest association when loneliness was measured by the single
direct question. Health showed a stronger association with loneliness than
did partner status when loneliness was measured indirectly by the De Jong
Gierveld Loneliness Scale.
Social losses in old age make people more vulnerable to loneliness. But
elderly people have coping strategies and reactions, making adaptations to
changed social realities (Baltes & Carstensen, 1996). They may lower their
aspirations and expectations concerning social network and social activities.
Older people with smaller networks and fewer social activities may not feel
Table 5. Predictors of Loneliness Measured by the (Single Item) Direct Question about Loneliness among
People in Different Age Groups
Direct Measure of Loneliness
Total Sample

Model 1 Model 2 Model 3 Model 4

Variables B SD Beta p B SD Beta p B SD Beta p B SD Beta p

Age 30-49a –.14 .02 –.08 <.001 –.14 .02 –.09 <.001 –.17 .02 –.10 <.001 .03 .02 .02 .096

Age 50-64a –.13 .02 –.07 <.001 –.13 .02 –.07 <.001 –.22 .02 –.12 <.001 –.01 .02 –.01 .528
248 / NICOLAISEN AND THORSEN

Age 65-81a –.09 .02 –.04 <.001 –.09 .02 –.04 <.001 –.21 .02 –.09 <.001 –.06 .02 –.03 .013

Female .17 .01 .11 <.001 .16 .01 .10 <.001 .14 .01 .09 <.001

Health .13 .01 .18 <.001 .12 .01 .17 <.001

Partner –.47 .01 –.27 <.001

R square .004 .015 .045 .108

R square .004 .011 .030 .063

change

F change 18.66*** 167.45 467.81 1037.66

(p-value) (<.001) (<.001) (<.001) (<.001)

18-29 years
Model 1 Model 2 Model 3 Model 4
Variables B SD Beta p B SD Beta p B SD Beta p B SD Beta p

Age .00 .01 .01 .794 .00 .01 .00 .862 –.00 .01 –.01 .685 .02 .01 .07 .001
Female .15 .03 .09 (<.001) .13 .03 .08 <.001 .18 .03 .12 <.001
Health .13 .02 .16 <.001 .12 .02 .16 <.001
Partner –.33 .04 –.20 <.001
R square .000 .009 .033 .064
R square .000 .009 .025 .031
change
F change .07 22.09 64.63 84.97
(p-value) (.794) (<.001) (<.001) (<.001)
30-49 years
Age –.00 .00 –.03 .027 –.00 .00 –.03 .022 –.01 .00 –.04 .001 –.01 .00 –.03 .008
Female .14 .02 .09 <.001 .12 .02 .08 <.001 –.12 .02 .08 <.001
Health .14 .01 .19 <.001 .13 .01 .19 <.001
Partner –.49 .02 –.26 <.001
R square .001 .009 .045 .112
R square .001 .008 .036 .066
change
WHO ARE LONELY?

F change 4.92 47.33 226.85 442.99

(p-value) (.027) (<.001) (<.001) (<.001)
/ 249
Table 5. (Cont’d.)
50-64 years
Model 1 Model 2 Model 3 Model 4
Variables B SD Beta p B SD Beta p B SD Beta p B SD Beta p
Age .00 .00 .02 .032 .00 .00 .02 .251 –.00 .00 –.00 .867 –.00 .00 –.00 .806
Female .17 .03 .11 <.001 .16 .03 .10 <.001 .11 .03 .07 <.001
Health .12 .01 .18 <.001 .11 .01 .17 <.001
Partner –.52 .03 –.27 <.001
R square .000 .011 .043 .113
R square .000 .011 .032 .070
250 / NICOLAISEN AND THORSEN

change
F change .99 43.85 129.89 312.81
(p-value) (.319) (<.001) (<.001) (<.001)
65-81 years
Age .01 .00 .05 .011 .01 .00 .05 .023 .01 .00 .03 .092 .00 .00 –.00 .937
Female .29 .04 .16 <.001 .28 .04 .16 <.001 .14 .04 .08 <.001
Health .12 .02 .15 <.001 .11 .02 .14 <.001
Partner –.57 .04 –.31 <.001
R square .003 .028 .052 .137
R square .003 .026 .023 .085
change
F change 6.53 60.45 56.22 226.79
(p-value) (.011) (<.001) (<.001) (<.001)
aReference category: age 18-29.
WHO ARE LONELY? / 251

lonely, or the feelings of loneliness may be diminished (Baltes & Baltes, 1990).
The age difference in loneliness is linear, and larger when using the De Jong
Gierveld Scale than when using the direct measure of loneliness. This may
indicate some adjustment in aspiration, concerning the number of social rela-
tionships. However, due to the loss of a partner, close friends, or family
members through death, the probability of lacking an intimate attachment
figure increases with age. Thus, when people grow older, emotional loneli-
ness (Weiss, 1973) may become more widespread, and the loss of the most
important close relationships may increasingly have more impact on loneliness
than the number of social contacts.
Young people, by contrast, may have great aspirations for social success,
including many friends, being popular, living in a context of constant com-
parisons with other young people. They are preparing for the future, and
seeking diverse social networks (Carstensen et al., 1999), and hoping for
intimate social relations—a partner or spouse and a family life.
In accordance with previous studies (Jylhä, 2004; Nicolaisen & Thorsen,
2012; Pinquart & Sörensen, 2001; Savikko et al., 2005; Victor & Yang, 2012),
women were more often lonely than men in all age groups, using the direct
measure of loneliness. Using the De Jong Gierveld Scale, men were lonelier
than women. This gender difference was found in the two youngest age
groups, but not in the two oldest groups (50-64 and 65-81 years old). When
other variables were controlled for, the gender difference was found in all
age groups, except in the oldest group (65-81).
The different findings using the two measures may be related to men’s
greater reluctance to “admit” to being lonely (Borys & Perlman, 1985) when
asked directly. An additional separate analysis of our data, splitting the De Jong
Gierveld items into the two sub-groups of emotional and social loneliness,
indicated that the gender difference was especially pronounced regarding
social loneliness, which accords with previous studies (de Jong Gierveld &
van Tilburg, 2010; Dykstra & de Jong Gierveld, 2004). This was found in all
age groups. Men were more socially lonely than women, or put differently:
it was easier for men to admit to a lack of social contacts than to emotions
of missing contact. We found that women were more emotionally lonely
than men in the two oldest age groups (50-64 and 65-81) with this measure,
which would be expected since women generally lose their partner earlier
than men and live alone for a longer time.
Returning to the definition of loneliness by Peplau and Perlman (1982),
which summarizes the common elements of the loneliness concept in
studies: loneliness is the negative feeling of a gap between the contact that is
desired and the contact that is experienced. Lonely people want more social
contact—qualitatively and quantitatively—than they currently feel they have.
252 / NICOLAISEN AND THORSEN

This definition leaves open what kind of contact is desired—whether it is

social, that is, based on the size of the network (as in “social loneliness”), or
emotional, based on a lack of intimate relations (as in “emotional loneliness”).
Most significantly, a central aspect of Peplau and Perlman’s (1982) definition is
that people can change their ambitions for contact and adjust them to their
prospects. Lowered expectations can protect against feelings of loneliness
in old age.
The complex interrelationship between aspirations for contact and actual
contact will depend on social ambitions in different stages of life, whether the
person is a youth, an adult, or in later life. It is probably also influenced by
cultural ideals and norms regarding the number and content of social rela-
tionships with family members and friends in different periods of life.
Aspirations for contact in different age groups may be an important factor
influencing the inconclusive gender and age patterns found in studies.
Older people report as many social contacts as younger people (Field &
Minkler, 1988; Lang & Carstensen, 1994). However, loss of one’s partner in
old age is a central factor leading to loneliness. A main measure directed
at alleviating loneliness among the elderly is to include them in social life
where they can find acquaintances as well as close friends. They may also
need support to stay in contact with family and friends.
There are some limitations of the study. The cross-sectional design does not
allow us to draw conclusions about causations. For example, health problems
may lead to loneliness and isolation, and loneliness to health problems—
especially concerning mental health. The loss of respondents is more pro-
nounced in the youngest and the oldest age groups, and among people in
poor health. The selected loss of respondents may be a limitation regarding
generalization of the findings. However, this kind of sample selection is usual
in surveys. Strengths of the study include the large sample size and the broad
age range, permitting an extensive account of the prevalence of loneliness
among men and women across age groups.
The study opens for questions about the meanings of social relations, and
aspirations for different forms and qualities of social contact during the life
course. The meaning of having many or few social contacts may be quite
different among the young, middle aged, and old adult people. People in older
age groups have a remarkable capacity to adapt to impending—also negative—
events in life, adjusting aspirations and coping with adversities, and thus be
more able to upheld subjective well-being.
The meaning of social contact, friendship, partnership, and intimacy would
seem to be changing at present. The internet is providing new opportunities
for instant (“faceless”) intimacy over worldwide distances, via Skype and
Facebook, for example. Future research should study social relationships,
WHO ARE LONELY? / 253

friendship, and social support in the (evolving) technological reality of today,

especially among the elderly. More qualitative studies of loneliness among
men and women in different age groups are warranted to get rich descriptions
of experiences of loneliness. Future work on loneliness could also aspire
to develop new instruments, tapping the complexities of the experience of
loneliness in different age groups today. The aim should be to develop more
refined theories about the qualities of loneliness among people in different
phases of life.

ACKNOWLEDGMENTS

The authors would like to thank the IJAHD reviewers and the editor for
their valuable suggestions and comments.

REFERENCES

Aartsen, M., & Jylhä, M. (2011). Onset of loneliness in older adults: Results of a 28
year prospective study. European Journal of Ageing, 8(1), 31-38. doi: 10.1007
/s10433-011-0175-7
Baltes, P. B., & Baltes, M. M. (1990). Psychological perspectives on successful
aging: The model of selective optimalization with compensation. In P. B. Baltes &
M. M. Baltes (Eds.), Successful aging. Perspectives from the behavioral sciences.
Cambridge, UK: Cambridge University Press.
Baltes, M. M., & Carstensen, L. (1996). The process of successful ageing. Ageing
and Society, 16(4), 397-422. doi: 10/1017/50144686X00003603
Bjørshol, E., Høstmark, M., & Lagerstrøm, B. O. (2010). Livsløp, generasjon og kjønn.
LOGG 2007. Dokumentasjonsrapport [Life course, generation, and gender.
LOGG 2007. Documentation report]. Oslo, Norway: Statistisk sentralbyrå. Notater
19/2010.
Borys, S., & Perlman, D. (1985). Gender differences in loneliness. Personality and
Social Psychology Bulletin, 11(1), 63-74. doi: 10.1177/0146167285111006
Burke, K. E., Schnittger, R., O’Dea, B., Buckley, V., Wherton, J. P., & Lawlor, B. A.
(2012). Factors associated with perceived health in older adult Irish population.
Aging & Mental Health, 16(3), 288-295. doi: 10.1080/13607863.2011.628976
Cacioppo, J. T., Hawkley, L. C., & Thisted, R. A. (2010). Perceived social isolation
makes me sad: 5-year cross-lagged analyses of loneliness and depressive symp-
tomatology in the Chicago Health, Aging, and Social Relation Study. Psychology
and Aging, 25(2), 453-463. doi: 10.1007/s10433-011-0175-7
Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously.
A theory of socioemotional selectivity. American Psychologist, 54(3), 165-181.
doi: 10.1037/0003-066X.54.3.165
254 / NICOLAISEN AND THORSEN

Cohen-Mansfield, J., Shmotkin, D., & Goldberg, S. (2009). Loneliness in old age:
Longitudinal changes and their determinants in an Israeli sample. International
Psychogeriatrics, 21(6), 1160-1170. doi: 10.1017/S1041610209990974
Deeg, D. J., & Bath, P. A. (2003). Self-rated health, gender and mortality in older
persons: Introduction to special section. The Gerontologist, 43(3), 369-371. doi:
10.1093/geront/43.3.369
de Jong Gierveld, J. (1987). Developing and testing a model of loneliness. Journal of
Personality and Social Psychology, 53(1), 119-128. doi: 10.1037/0022-3514.53.1.119
de Jong Gierveld, J. (1998). A review of loneliness: Concepts and definitions, deter-
minants and concequences. Reviews in Clinical Gerontology, 8, 73-80.
de Jong Gierveld, J., & Kamphuis, F. (1985). The development of a Rasch-type
loneliness scale. Applied Psychological Measurement, 9(3), 289-299.
de Jong Gierveld, J., & van Tilburg, T. (1999). Manual of the loneliness scale. Retrieved
from http://home.fsw.vu.nl/TG.van.Tilburg/manual_loneliness_scale_1999.html
de Jong Gierveld, J., & van Tilburg, T. (2006). A 6-item scale for overall, emotional,
and social loneliness: Confirmatory tests on survey data. Research on Aging, 28(5),
582-598. doi: 10.1177/0164027506289723
de Jong Gierveld, J., & van Tilburg, T. (2010). The De Jong Gierveld short scales
for emotional and social loneliness: Tested on data from 7 countries in the
UN generations and gender surveys. European Journal of Ageing, 7(2), 121-130.
doi: 10.1007/s10433-010-0144-6
de Jong Gierveld, J., van Tilburg, T., & Dykstra, P. A. (2006). Loneliness and social
isolation. In A. Vangelisti & D. Perlman (Eds.), Cambridge handbook of personal
relationships (pp. 485-500). Cambridge, MA: Cambridge University Press.
Dykstra, P. A. (2009). Older adult loneliness: Myths and realities. European Journal
of Ageing, 6(2), 91-100. doi: 10.1007/s10433-009-0110-3
Dykstra, P. A., & de Jong Gierveld, J. (2004). Gender and marital-history differences
in emotional and social loneliness among Dutch older adults. Canadian Journal
on Aging, 23(2), 141-155. doi: http://dx.doi.org/10.1353/cja.2004.0018
Dykstra, P. A., van Tilburg, T., & de Jong Gierveld, J. (2005). Changes in older adults
loneliness: Results from a seven-year longitudinal study. Research on Aging, 27(6),
725-747. doi: 10.1177/0164027505279712
Fees, B. S., Martin, P., & Poon, L. W. (1999). A model of loneliness in older adults.
Journal of Gerontology: Psychological Sciences, 54B(4), P231-P239. doi: 10.1093/
geronb/54B.4.P231
Field, D., & Minkler, M. (1988). Continuity and change in social support between
young-old and old-old or very-old age. Journal of Gerontology: Psychological
Sciences, 43(4), P100-106. doi: 10.1093/geronj/43.4.P100
Fokkema, T., de Jong Gierveld, J., & Dykstra, P. A. (2012). Cross-national differences
in older adult loneliness. The Journal of Psychology, 146(1-2), 201-228. doi:
10.1080/00223980.2011.631612
Fokkema, T., & Naderi, R. (2013). Differences in late-life loneliness: A comparison
between Turkish and native-born older adults in Germany. European Journal of
Ageing. Published online: March 6, 2013. doi: 10.1007/s10433-013-0267-7
WHO ARE LONELY? / 255

Han, J., & Richardson, V. E. (2010). The relationship between depression and lone-
liness among homebound older persons: Does spirituality moderate this rela-
tionship? Journal of Religion & Spirituality in Social Work, 29, 218-236.
doi: 10.1080/15426432.2010.495610
Hansen, T., Slagsvold, B., & Ingebretsen, R. (2012). The strain and gains of caregiving:
An examination of the effects of providing personal care to a parent on a range of
psychological outcomes. Social Indicators Research, Published online: September
11, 2012. doi: 10.1007/s11205-012-0148-z
Heikkinen, R.-L., & Kauppinen, M. (2011). Mental well-being: A 16-year follow-up
among older residents in Jyväskylä. Archives of Gerontology and Geriatrics, 52(1),
33-39. doi: http://dx.doi.org/10.1016/j.archger.2010.01.017
Holmén, K., & Furukawa, H. (2002). Loneliness, health and social network among
elderly people—A follow-up study. Archives of Gerontology and Geriatrics, 35(3),
261-274.
Iecovich, E. (2013). Psychometric properties of the Hebrew version of the de Jong
Gierveld Loneliness Scale. Educational Gerontology, 39, 12-27. doi: 10.1080/
03601277.2012.660860
Jylhä, M. (2004). Old age and loneliness: Cross-sectional and longitudinal analyses
in the Tampere Longitudinal Study on Aging. Canadian Journal on Aging, 23(2),
157-168. doi: 10.1353/cja.2004.0023
Kirova, A. (2003). Assessing children’s experiences of loneliness through conver-
sations. Field Methods, 15(1), 3-24. doi: 10.1177/1525822X02239572
Lang, F. R., & Carstensen, L. L. (1994). Close emotional relationships in late life:
Further support for proactive aging in the social domain. Psychology and Aging,
9(2), 315-324. doi: 10.1037/0882-7974.92.315
Leung, G. T. Y., de Jong Gierveld, J., & Lam, L. C. W. (2008). Validation of
the Chinese translation of the 6-item De Jong Gierveld Loneliness Scale in
elderly Chinese. International Psychogeriatrics, 20(6), 1262-1272. doi: 10.1017/
S1041610208007552
Luo, Y., Hawkley, L. C., Waite, L. J., & Cacioppo, J. T. (2012). Loneliness, health, and
mortality in old age: A national longitudinal study. Social Science & Medicine,
74(6), 907-914. doi: 10.1016/j.socscimed.2011.11.028
McHugh, J. E., Casey, A.-M., & Lawlor, B. A. (2011). Psychosocial correlates
of aspects of sleep quality in community-dwelling Irish older adults. Aging &
Mental Health, 15(6), 749-755. doi: 10.1080/13607863.2011.562180
Moum, T. (1992). Self-assessed health among Norwegian adults. Social Science &
Medicine, 35(7), 935-947. doi: 10.1016/0277-9536(92)90108-3
Nicolaisen, M., & Thorsen, K. (2012). Impairments, mastery, and loneliness. A
prospective study of loneliness among older adults. Norwegian Journal of
Epidemiology, 22(2), 143-150.
Nummela, O., Seppänen, M., & Uutela, A. (2011). The effects of loneliness and
change in loneliness on self-rated health (SRH): A longitudinal study among ageing
people. Archives of Gerontology and Geriatrics, 53(2), 163-167. doi: 10.1016/
j.archger.2010.10.023
256 / NICOLAISEN AND THORSEN

Paul, C., Ayis, S., & Ebrahim, S. (2006). Psychological distress, loneliness and dis-
ability in old age. Psychology, Health & Medicine, 11(2), 221-232. doi: 10.1080/
13548500500262945
Peplau, L. A., & Perlman, D. (Eds.). (1982). Loneliness. A sourcebook of current
theory, research and therapy. New York, NY: John Wiley and Sons.
Pinquart, M., & Sörensen, S. (2001). Influences on loneliness in older adults: A
meta-analysis. Basic and Applied Social Psychology, 23(4), 245-266. doi: 10.1207/
153248301753225702
Pinquart, M., & Sörensen, S. (2003). Risk factors for loneliness in adulthood and
old age—A meta-analysis. In S. Shohov (Ed.), Advances in psychology research
(Vol. 19, pp. 111-143). Hauppauge, NY: Nova Science.
Prieto-Flores, M.-E., Forjaz, M. J., Fernandez-Mayoralas, G., Rojo-Perez, F., & Martinez-
Martin, P. (2011). Factors associated with loneliness of noninstitutionalized
and institutionalized older adults. Journal of Aging and Health, 23(1), 177-194.
doi: 10.1177/0898264310382658
Routasalo, P. E., Savikko, N., Tilvis, R. S., Strandberg, Y. E., & Pitkälä, K. H.
(2006). Social contacts and their relationship to loneliness among aged people—
A population-based study. Gerontology, 52(3), 181-187.
Russell, D. (1996). UCLA loneliness scale (Version 3): Reliability, validity, and factor
structure. Journal of Personality Assessments, 66(1), 20-40.
Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA Loneliness
Scale: Concurrent and discriminant validity evidence. Journal of Personality and
Social Psychology, 39(3), 472-480.
Savikko, N., Routasalo, P., Tilvis, R. S., Strandberg, T. E., & Pitkälä, K. H. (2005).
Predictors and subjective causes of loneliness in an aged population. Archives of
Gerontology and Geriatrics, 41(3), 223-233. doi: 10.1016/j.archger.2005.03.002
Shaver, P. R., & Brennan, K. A. (1991). Measures of depression and loneliness.
In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of personality
and social psychological attitudes (pp. 195-289). San Diego, CA: Academic Press.
Shiovitz-Ezra, S., & Ayalon, L. (2012). Use of direct versus indirect approaches to
measure loneliness in later life. Research on Aging, 34(5), 572-591. doi: 10.1177/
0164027511423258
Shiovitz-Ezra, S., Leitsch, S., Graber, J., & Karraker, A. (2009). Quality of life and
psychological health indicators in the national social life, health, and aging project.
Journal of Gerontology: Social Sciences, 64B(S1), i30-i37. doi: 10.1093/geronb/
gbn020
Steed, L., Boldy, D., Grenade, L., & Iredell, H. (2007). The demographics of loneliness
among older people in Perth, Western Australia. Australasian Journal on Ageing,
26(2), 81-86.
Sundström, G., Fransson, E., Malmberg, B., & Davey, A. (2009). Loneliness among
older Europeans. European Journal of Ageing, 6(4), 667-675. doi: 10.1007/
s10433-009-0134-8
Thorsen, K. (1990). Ensom og alene, sammen og lykkelig? Ensomhet i uliki alders-
grupper [Lonely and alone, together and happy? Loneliness in different age
groups]. Oslo, Norway: Norsk Gerontologisk Institutt. N6i-rapport 2-1990.
WHO ARE LONELY? / 257

Thorsen, K., & Clausen, S.-E. (2008). Funksjonshemning, ensomhet og depresjon:

Hva betyr ensomhet for om personer med funksjonshemning opplever depresjon?
[Impairments, loneliness and depression. The impact of loneliness on depression
among people with impairments]. Tidsskrift for Norsk Psykologforening, 45, 19-27.
Tiikkainen, P., & Heikkinen, R.-L. (2005). Associations between loneliness, depressive
symptoms, and perceived togetherness. Aging & Mental Health, 9(6), 526-534.
doi: 10.1080/13607860500193138
Tijhuis, M. A. R., De Jong Gierveld, J., Feskens, E. J. M., & Kromhout, D. (1999).
Changes in and factors related to loneliness in older men. The Zutphen elderly
study. Age and Ageing, 28, 491-495. doi: 10.1093/ageing/28.5.491
van Baarsen, B., Snijders, T. A. B., Smit, J. H., & van Duijn, M. A. J. (2001). Lonely
but not alone: Emotional isolation and social isolation as two distinct dimen-
sions of loneliness in older people. Educational and Psychological Measurement,
61(1), 119-135. doi: 10.1177/00131640121971103
van Tilburg, T., Havens, B., & de Jong Gierveld, J. (2004). Loneliness among older
adults in the Netherlands, Italy, and Canada: A multifaceted comparison. Canadian
Journal on Aging, 23(2), 169-180. doi: http://dx.doi.org/10.1353/cja.2004.0019
Victor, C. R., Bond, J., & Scambler, S. J. (2009). The social world of older people.
Maidenhead, UK: Open University Press.
Victor, C. R., & Bowling, A. (2012). A longitudinal analysis of loneliness among
older people in Great Britain. The Journal of Psychology, 146(3), 313-331.
doi: 10.1080/00223980.2011.609572
Victor, C. R., Grenade, L., & Boldy, D. (2005). Measuring loneliness in later life:
A comparison of different measures. Reviews in Clinical Gerontology, 15, 63-70.
doi: 10.1017/S0959259805001723
Victor, C. R., Scambler, S. J., Marston, L., Bond, J., & Bowling, A. (2005). Older
people’s experiences of loneliness in the UK: Does gender matter? Social Policy
and Society, 5(1), 27-38. doi: 10.1017/S1474746405002733
Victor, C. R., & Yang, K. (2012). The prevalence of loneliness among adults: A case
study of the United Kingdom. The Journal of Psychology, 146(1-2), 85-104.
Weiss, R. S. (1973). Loneliness: The experience of emotional and social isolation.
Cambridge, MA: MIT Press.

Direct reprint requests to:

Magnhild Nicolaisen
Ageing and Health
Norwegian Centre for Research, Education and Service Development
Oslo University Hospital, Ullevaal, Bld. 37A
NO-0407 Oslo, Norway
e-mail: [email protected]