University of Southern California Dissertations and Theses

Internet communication use, psychological functioning and social connectedness at older ages

(USC Thesis Other)

Internet communication use, psychological functioning and social connectedness at older ages

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

INTERNET COMMUNICATION USE, PSYCHOLOGICAL FUNCTIONING AND
SOCIAL CONNECTEDNESS AT OLDER AGES
by

Hyunju Shim

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements of the Degree
DOCTOR OF PHILOSOPHY
GERONTOLOGY

August 2020

ii

ACKNOWLEDGEMENTS
First of all, my sincere and deepest gratitude goes to Dr. Eileen Crimmins for guiding the
dissertation ideas from the beginning to the end of the program. I would like to thank her for her
tremendous support for making the progress toward my degree from developing my research
skills to supporting participation in various conferences, being inspiring in making me to think
and examine the research question from different perspectives, and being exemplary for not only
me but also other colleagues and professionals in terms of intellectual abilities and motherly
mentorship. Secondly, I would like to also thank Dr. Jennifer Ailshire for her consistent support
and encouragement. Her co-leadership with Dr. Crimmins has been instrumental in creating a
cooperative and supportive environment at the biodemography center, through which I learned
how to work with others on different projects with our unique abilities. I also thank another
committee member, Dr. Elizabeth Zelinski, for allowing me the opportunity to work with her
from my first year until the end of the program. Her sense of humor has also been a great source
of encouragement in the program and the one I hope to learn someday to pass on.
I also would like to express my gratitude to Dr. Jung Ki Kim for her mentorship and
guidance. Her patience, kindness, and faith will continue to guide me to aspire myself to become
a better researcher and a believer. Special thanks also go to my colleagues and friends at school:
Yeonjin Choi, Hyewon Kang, Yuan Zhang, Haena Lee, Stephen Frochen, Kylie Meyer, and
Deborah Hoe. I also would like to thank my mentors in Korea, Dr. Sungjae Choi and Dr.
Donghee Han, who have been a great example of gerontologists and the very reason I began to
pursue career in gerontology.
Lastly, my deepest gratitude to my family from the bottom of my heart, for allowing me
to go to on a graduate school ride and pursue the degree which no member has obtained before.
iii

None of us knew what it involved, and I am immensely thankful for my family for constantly
believing in me and encouraging me to finish the ride despite many ups and downs. Thank you
also to my husband for being so patient and supportive.
Thank you all sincerely for making this possible.

iv

TABLE OF CONTENTS

Acknowledgements....................................................................................................................................ii
List of Tables..............................................................................................................................................vi
List of Figures..........................................................................................................................................viii
Abstract.......................................................................................................................................................ix
Chapter One: Introduction.........................................................................................................................1
1.1. Internet Use as a Means of Social Connections......................................................................1
1.2. Theories on Well-being and Social Connections at Older Ages.......................................3
1.3. Internet Communication Use and Older People................................................................5
1.4. Three Indicators of Social Connectedness.........................................................................8
1.5. Objectives of the Present Dissertation.............................................................................11
Chapter Two: Living Alone, Electronic Communication and Psychological Functioning:
Findings from the National Health and Aging Trends Study........................................................13
2.1. Introduction.....................................................................................................................13
2.2. Methods...........................................................................................................................18
2.3. Results.............................................................................................................................23
2.4. Discussion.......................................................................................................................32
Chapter Three: Cross-National Examination of Differences in the Association between Internet
Use and Depressive Symptoms by Gender and Intergenerational Characteristics in the U.S. and
Korea.............................................................................................................................................38
3.1. Introduction.....................................................................................................................38
3.2. Methods...........................................................................................................................44
3.3. Results.............................................................................................................................48
3.4. Discussion.......................................................................................................................56
Chapter Four: Internet Use and Life Transitions in Marital and Work Status..............................60
4.1. Introduction.....................................................................................................................60
4.2. Methods...........................................................................................................................64
4.3. Results.............................................................................................................................69
v

4.4. Discussion........................................................................................................................83
Chapter Five: Discussion..............................................................................................................87
5.1. Discussion of Findings.....................................................................................................87
5.2. Limitations and Strengths................................................................................................88
5.3. Implications and Future Research....................................................................................91
References......................................................................................................................................94

vi

LIST OF TABLES

Table 2.1. Baseline Sample Characteristics..........................................................................................23
Table 2.2. Pearson’s Correlations among the Key Variables.............................................................24
Table 2.3. Growth Curve Models of Electronic Communication Use Predicting Well-Being......27
Table 2.4. Growth Curve Models of Electronic Communication Use Predicting Depressive
Symptoms................................................................................................................................28
Table 2.5. Growth Curve Models of Electronic Communication Use Predicting
Well-Being (2011-2014) ......................................................................................................29
Table 2.6. Growth Curve Models of Electronic Communication Use Predicting
Depressive Symptoms (2011-2014)...................................................................................30
Table 2.7. Growth Curve Models of Electronic Communication Use with Different Social
Participation Measures.........................................................................................................31
Table 3.1. Sample Characteristics in the U.S. and Korea...................................................................50
Table 3.2. Average frequency of contacts by geographic distance and internet use in the U.S....51
Table 3.3. Average frequency of contacts by geographic distance and internet use in Korea......51
Table 3.4. OLS results of internet use predicting depressive symptoms by gender in the U.S and
Korea.......................................................................................................................................52
Table 4.1. Baseline Sample Characteristics Among Those Who Remained Married and
Experienced Marital Transitions (N=5,323) (2010/2012) ...............................................70
Table 4.2. Baseline Sample Characteristics for the Employed and Those with Work Transitions
(N=2,498) (2010/2012) ........................................................................................................72
Table 4.3. Mean levels of outcomes by marital and work transitions in 2010/2012 and
2014/2016...............................................................................................................................74
Table 4.4. Results for Equations Estimating Associations between Internet Use, Marital
Transitions, and Depressive Symptoms..............................................................................76
Table 4.5. Results for Estimating Associations between Internet Use, Marital Transitions, and
Loneliness...............................................................................................................................77

vii

Table 4.6. Results for Estimating Associations between Internet Use, Work Transitions, and
Depressive Symptoms...........................................................................................................81
Table 4.7. Results for Estimating Associations between Internet Use, Work Transitions, and
Loneliness...............................................................................................................................82

viii

LIST OF FIGURES

Figure 3.1. Predicted standardized depressive symptoms by internet use and geographic distance
from the closest child in the U.S. .......................................................................................54
Figure 3.2. Predicted standardized depressive symptoms by internet use and geographic distance
from the closest child in Korea............................................................................................55
Figure 4.1. Predicted Levels of Depressive Symptoms and Loneliness by Internet Use and
Marital Transitions................................................................................................................79

ix

ABSTRACT
The use of the internet for older adults has emerged as an essential tool for enhancing
their quality of life and maintaining social connections in the recent years. Using the internet has
become an integral part of daily life of people in the 21st century, and advances in technology
offer innovative and various media of communications over different time zones and vast
distance to keep contact with family and friends, despite migration and declining household
sizes.
An increasing number of studies have documented evidence of positive effects of
technology use on psychological well-being and health of older adults. However, existing
literature primarily examined older people as one homogenous group, making it impossible to
examine how the differences in the individual and contextual factors affect the association
between internet use and mental health. Additionally, the role of internet use in intergenerational
relationship has not been explored despite its potential implications for maintaining family
support in the globalized world.
The purpose of the present dissertation is to address this gap by examining the effect of
internet use for older people situated in different individual and environmental contexts and
examine whether the effect of technology use shows heterogeneity depending on these contexts.
The first paper examines people with different environmental contexts: people who live alone
compared to people who live together with spouses and others. The National Health and Aging
Trends Study was analyzed to examine whether the use of electronic communication had
different impacts on psychological well-being and depressive symptoms for people who live
alone compared to people who live together with others. Growth curve models showed that the
effect of using electronic communication was significantly associated with higher psychological
x

well-being both at the initial level and over time, but this effect did not significantly differ by
living arrangements. For depressive symptoms, the effect of using electronic communication was
significantly associated with lower depressive symptoms at the initial level for both types of
living arrangements, but there was no significant effect over time for either group.
The second paper focuses on older people whose cultural and geographic contexts are
situated in the different contexts. Using the Health and Retirement Study in the U.S. and the
Living Profiles of Older People Survey in Korea, the paper examined whether the internet use
was differentially related to geographic proximity to adult children by gender in the two
countries where different gender norms and expectations about intergenerational relationship
exist. Ordinary least square regression models showed that internet use was significantly
associated with reduced depressive symptoms more for parents whose adult children were living
far away, with different effects by gender in the two countries.
The third paper investigates how significant transitions in marital and work status at older
ages are related to the association between internet use and psychological distress. Significant
transitions in marital status were defined as changes from being married to a divorce or
widowhood, while transitions in work status referred to changes from working to unemployment
or retirement. The Health and Retirement Study was used to examine whether internet use had
different effects for people who experienced marital and work transitions. Results showed that
internet use was significantly associated with reduced depressive symptoms and loneliness for
people who became divorced, but there was no significant effect for widowhood. For work
transitions, using the internet did not show statistically significant associations with either
depressive symptoms or loneliness.
xi

The present dissertation aims to extend the discussion of implications of internet use
among older adults to recognition of the heterogeneity in the effect of internet use and highlight
the need to develop various interventions specific to different environmental and individual
characteristics of this diverse aging population. The dissertation also uses three different sources
of longitudinal data to examine the effects of internet use comprehensively. The overall findings
contribute to growing literature on internet use and aging by expanding our understanding on
whether the implications of internet use for enhancing psychological well-being and social
connectedness are applicable across different groups and countries. The final discussion
addresses limitations and strengths of each study and presents future directions for further
research on examining the heterogeneity in the effect of internet use on diverse groups of older
adults to develop policies and programs that can effectively enhance psychological well-being
and promote social integration of older people.

1

Chapter One: Introduction

1.1. Internet Use as a Means of Social Connections
Social connection is a crucial determinant of one’s psychological well-being and health.
Throughout history, human beings have been recognized to be social animals. The centrality of
social connection in the quality of life is found in numerous studies documenting how a lack of
social interactions and meaningful relationships are linked to higher risks of mortality (Holt-
Lunstad, Smith & Layton, 2010; House, 2001), higher odds of developing comorbid conditions
(Carver, Beamish & Phillips, 2018), poorer immunological functioning (Ong & Allaire, 2005),
and steeper declines in physical and cognitive functioning (Burholt et al., 2017; Fingerman et al.,
2020). Decrease in social support was also found to be associated with a decrease in life
satisfaction and an increase in overall depression (Newsom & Schulz, 1996). Due to this
importance, recent epidemiological literature calls for establishing social connection as a priority
issue in public health (Holt-Lunstad, Robles & Sbarra, 2017).
Social connection is an encompassing term that includes the structural component (e.g.,
social isolation, living alone), the functional component (e.g., perceived social support,
loneliness), and the qualitative component (e.g., positive or negative aspects) of the social
relationships (Holt-Lunstad, 2018). Increasingly, empirical attention on how to enhance social
connection and decrease risks of social isolation has been growing in recent literature on aging
and health. A recent national estimate suggests that 24% of Americans aged 65 and older were
socially isolated, with increased risks among people in older ages, male, and those with low
education (Cudjoe et al., 2018). One of the areas that has significant implications for social
connection in the older population is how to optimize the use of existing technologies. Internet
2

use provides a means for social connection that is not constrained by geographic locations or
time, and physical functioning. Ang and Chen (2019) found that using online social networking
sites reduced the negative effect of pain on depressive symptoms while in-person social
interactions did not show a similar protective effect. The recent pandemic situation also presents
a particularly relevant context where the access to and the ability to use the internet becomes
crucial for maintaining social connections and health of older people. Internet communication
use can be an optimal way of preventing social isolation for an aging population under an
epidemiological crisis where physical isolation is mandated (Armitage & Nellums, 2020; Van
Orden et al., 2020). Indeed, a majority of nursing homes and senior living communities had to
temporarily replace visits from family and friends with online communication media, such as
emails or Skype, due to the pandemic situation (AARP, 2020). Yet, despite this increasing
demand in practice, the role of internet use in promoting social connections and communication
has received surprisingly little empirical attention.
The present dissertation aims to contribute to growing literature on aging, internet
communication, and social connection and address this gap in literature by examining how
internet communication use is related to the association between different aspects of social
connection and psychological functioning. Is internet communication use linked to reduced
depressive symptoms and higher psychological well-being among people situated in different
contexts of social connection? How does the effect of internet communication use vary by these
contextual differences? These questions have important implications for testing the potential
effectiveness of using the internet and other kinds of technologies in reducing risks of social
isolation, as primary existing literature does not examine the effectiveness of internet use for
multidimensional aspects of social connections. Thus, it is important to examine whether the
3

effect of internet use is heterogenous for different aspects of social connectedness to evaluate its
potential as a primary means of social connection and identify an effective strategy for different
groups of older population.
1.2. Theories on Well-being and Social Connections at Older Ages
Maintaining positive psychological functioning and adequate social engagement at older
ages are key components of successful aging (Rowe & Kahn, 1997). Although scholars differ in
how to define the concept of well-being, subjective well-being broadly has come to incorporate
two philosophical distinctions: hedonic well-being which indicates one’s affective state or
satisfaction of desires, and eudemonic well-being which indicates one’s evaluative sense of
purpose and personal growth in life (Poon & Cohen-Mansfield, 2011; Ryff, 1989). Early
scholarly literature on well-being focused mostly on hedonic definitions examining what factors
contribute to positive affect such as feelings of happiness in people’s life. A significant
breakthrough came with an identification of eudemonic aspect, which is closely linked to the
concept of self-actualization that has been proposed by many developmental psychologists
(Erikson, 1994; Maslow, 1973).
In order to provide a theoretically-based conceptualization, Ryff operationalized six
dimensions of well-being: self-acceptance, purpose in life, personal growth, environmental
mastery, autonomy, and positive relationships (Ryff, 1989). An evaluation of the model in a
national sample showed that the only dimension which scored higher among old-aged
participants compared to young-aged participants was positive relationships with others (Ryff,
1995). This indicated that positive relationship with others was the important component of older
people’s well-being that is resilient to age-associated changes in physical health and other
functioning. Social relationships can also offer a sense of meaning and purpose in life when the
4

societal roles of individuals diminish in their family and work with advances in age. The
importance of social relationships for enhancing well-being at older ages was underscored in
subsequent studies (Charles & Carstensen, 2010; Luong, Charles & Fingerman, 2011; Diener,
Oishi & Tay, 2018; Steptoe & Fancourt, 2019; Umberson & Karas Montez, 2010).
Therefore, dysfunctions in the structural aspect (e.g., social isolation) or the functional
aspect (e.g., loneliness) of social connection can lead to serious consequences on health and
psychological well-being. Berkman and Syme (1979)’s study in Alameda County laid a
foundational groundwork for highlighting the significance of social isolation on mortality and
health. Recent evidence suggests the effect of social isolation on health is comparable to
smoking, leading to elevated risks of mortality and morbidity (Holt-Lunstad et al., 2015; House,
2001; Laugesen et al., 2018; Pantell et al., 2013). The effects of loneliness on health were also
found to be of similar magnitude: being associated with cognitive decline (Cacioppo & Hawkley,
2009; Wilson et al., 2007); elevated blood pressure (Hawkley, Masi, Berry & Cacioppo, 2006); a
poor quality of sleep (Cacioppo et al., 2002); and increased risks of mortality (Manemann et al.,
2018).
Social gerontologists have long battled to characterize how social relationships change
with age, especially at older ages. Disengagement theory depicts old age as the time of
withdrawal and discontinuity from social roles and relationships, arguing that the size of social
network and social interactions inevitably decrease with age (Cumming & Henry, 1961).
Changes in family structure and kinship networks in the industrial societies have also been
indicated as major risk factors of social isolation in old age (Townsend, 1968). On the other
hand, activity theory posits that successful aging occurs when the individual tries to stay active
and engaged to maintain social roles and find substitutes for losses in work or close social
5

network (Havinghurst, 1968). Later theories on social network and connectedness in old age also
revealed that, while the objective size of social network may decrease, the relational quality
remains constant or even increases with age. For instance, the convoy model suggests that the
social roles create convoy membership through which people gain and give relational support for
going through various age-related events throughout the life course (Kahn & Antonucci, 1980).
The convoy model also identified different network characteristics, ranging from relationship
type, gender, contact frequency, and geographic distance as important measures for assessing
social connection in late life (Antonucci, Ajrouch & Birditt, 2014). Subsequent empirical studies
demonstrated that the overall size of social network decreased with age, but emotional closeness
with social partners steadily increased with age (Carstensen, Fung & Charles, 2003; Lang &
Carstensen, 1994). This reflects that a priority of social relationship shifts from expansion of a
size of networks for more opportunities of career and marital status to maintenance of close
networks for meaningful connections. Cornwell and his colleagues (2008) also examined how
age is associated with different indicators of social connectedness and found that while age is
negatively associated with the size of social network, positive associations existed between age
and frequency of social interactions and volunteering, indicating that how old age is linked to
social relationships is complicated and diverse by types of indicators.
1.3. Internet Communication Use and Older People
Internet has become an integral part of everyday life of older people. The use of products
that utilize the internet has risen sharply among older adults, ranging from computers, tablets,
and smart phones, all of which are critically related to daily life of many people. A recent
estimate suggests that 91% of Americans aged 50 and over own a computer, 75% have a
smartphone, 42% have a tablet, and 14% have wearables, with 94% stating that they used the
6

device to maintain social contacts (AARP Research, 2019). Internet use thus has become one of
the major influences on to quality of life of older adults through its various effects on health and
social networks (Charness & Boot, 2009; Cotten, 2017; Czaja, 2017).
Empirical research on the effect of internet use among older adults has increased
consequently. Internet use has been linked to higher psychological well-being and lower
depressive symptoms in numerous studies (Chopik, 2016; Cotten, Ford, Ford & Hale, 2014;
Elliot, Mooney, Douthit & Lynch, 2013), although the mechanisms through which this occur
remains unclear. Some evidence indicates that using the internet confers a more convenient
medium to contact people and stay in touch which reduces their loneliness (Cotten, Anderson &
McCullough, 2013), while others suggest that going online provides more opportunities for older
people to be socially engaged in their communities (Bixter, Blocker & Rogers, 2018; Czaja,
2017; Kim, Lee, Christensen & Merighi, 2017). There is also a growing body of research
suggesting that using the internet facilitates cognitive functioning (Calhoun & Lee, 2019;
Marmot, 2014; Xavier et al., 2014). Attention to the potential of internet usage for promoting
social connection of older people also goes beyond empirical research, and there has been an
increasing attention among organizations and policy makers on using the internet to enhance
social connectedness of older people (Elder & Retrum, 2012; National Science and Technology
Council, 2019). Overall, a majority of previous studies on the internet use and both physical
health and psychological well-being of older people have found the association to be primarily
positive.
However, not all research has concluded that internet use is positive. For instance,
Primack and colleagues reported that social media use was associated with greater odds of
having greater perceived social isolation based on the survey of adults aged 19-32 years
7

(Primack et al., 2017). Additionally, Kraut and his colleagues (1998) found that the average
hours spent using the internet per week were significantly associated with decreased family
communication, decreased size of local social network, and increased loneliness and depression.
Some evidence suggests the effect of internet use is likely to be different depending on the
purpose of usage. One recent longitudinal study examined the effect of internet use among a pre-
retirement group (aged 40-64 years) and a post-retirement group (aged 65-85 years) and found
that the general internet use was significantly associated with increased loneliness in both age
groups, while using the internet to contact friends and family was associated with decreased
loneliness (Hees, Tesch-Römer & Huxhold, 2019). Another study examining the effect of online
volunteer support for internet usage among community dwelling participants (aged 57 to 84
years) found that internet use significantly increased the number of social contacts and well-
being scores while decreasing loneliness, and the participants viewed the value of the internet
use for communication with family and friends more important than the value of using the
internet for finding information on health care (Jones, Ashurst, Atkey & Duffy, 2015).
These mixed findings in the existing literature regarding the effect of internet use call for
a more thorough examination of internet communication use among older adults. Most prior
studies treated older people as one aggregate group and did not test the internet as a means of
communication. While using internet communication may provide several benefits to older
people, it is important to examine whether and how the effect of use differs depending on the
individual and environmental factors, given the diversity of aging population as well as the
heterogeneity of aging experience. For instance, internet use for someone who live alone is likely
to be significantly different from someone who lives with a spouse or others as online
communication has more influence over the amount of interpersonal communication for the
8

former compared to the latter. Similarly, widowhood or bereavement commonly occurs to people
at older ages to whom the effect of using the internet for reducing depression can be questionable
compared to other individuals whose aging experience is less traumatic. Additionally, the effect
of internet use on social connection may differ by environmental and socio-historical contexts. In
a country where greater value is placed on intergenerational relationship, the internet use may
contribute more to offsetting feelings of isolation and depression. However, because examination
of internet use is a relatively nascent topic (Charness & Jastrzembski, 2009; Cotten, 2017),
studies recognizing the differential implications of internet communication use for different sub-
groups of older people have been lacking.
1.4. Three Indicators of Social Connectedness
Based on life course perspective-based theories on social connectedness discussed above,
I identify three main indicators of social connectedness at older ages as a focus in this
dissertation. I examine how internet communication use is related to these indicators to affect
social connectedness of older people: environmental factors that are related to individual’s
unique circumstances; social contacts and support from individual network; and significant life
events that cause relational loss in a close social network.
Environmental factors refer to environmental, social, and economic characteristics that
can cause people to be more susceptible to social isolation. One of the frequently studied factors
is living alone. Living alone was found to be an important determinant of social isolation among
adults aged 65 and over, indicating that while not every older adult living alone felt isolated,
nearly all adults who scored high on an isolation scale lived alone (Wenger, Davies,
Shahtahmasebi & Scott, 1996). The significance of this issue has also grown in recent decades,
as almost half (45%) of all women among the age group of 75 and over are living alone
9

(Administration on Aging, 2018). Another important factor that has been often examined is
demographic and socioeconomic status. Marital status is often used to assess an individual’s
risks of social isolation. Additionally, having low education and income has been identified as
one the main risk factors of social isolation, although being a member of an ethnic minority was
associated with being less likely to be socially isolated compared to non-Hispanic Whites
(Cudjoe et al., 2018). The second chapter of this dissertation examines living arrangements as an
important contextual determinant of social isolation and tests whether using electronic
communication attenuates risks of poor psychological well-being and depression.
Social contacts and support indicate different types of social networks which can provide
instrumental and emotional support to older people. While definitions of social support varied in
prior research, numerous empirical studies have demonstrated that social support buffers the
effects of psychosocial distress and facilitates an individual’s coping skills for stressful life
events (Berkman & Syme, 1979; Pearson, 1986). These effects were found despite individual
variability in their capacity for intimacy (Lowenthal & Haven, 1968). Family members such as
spouses, adult children, friends, neighbors, and kin ties have been identified as the main sources
of support. Gender differences have also been found, with women more open to report negative
interactions with family or kin ties than men (Seeman, 2000).
However, the influence of each social network on the quality of life of older people has
been found to differ across different cultures, with findings on friends support as more beneficial
than family in Western countries whereas family support had a more significant role in non-
Western countries (Bélanger et al., 2016; Donnellan, Bennett & Soulsby, 2017; Poulin, Deng,
Ingersoll, Witt & Swain, 2012). With increases in migration among younger cohorts, many older
people live at a greater geographic distance from their adult children, which may cause
10

weakening of family ties. The third chapter of this dissertation aims to test whether the effect of
internet use moderates the relationship between depression and distance to adult children in two
different cultures – the U.S. and South Korea – and also examines this effect by gender.
Significant life events denote a wide range of life stressors that cause significant changes
in social network size, including but not limited to loss of spouse, loss of children or other family
members, loss of friends, loss of employment, and loss of physical and cognitive health. Life
events are primarily classified as proximal and distal events. Proximal events refer to the events
experienced in relatively recent time periods in adulthood such as a divorce, while distal events
refer to events that had happened with a longer time frame such as childhood abuse (Martin et
al., 2011). Wrzus and colleagues (2013) conducted a meta-analysis of 277 studies on social
network and life events of 177,635 participants from age 11 to 80 and over, and found that the
life-event-related change in social networks had more negative effects on social network sizes
than the age-related change in social networks, and divorce even reduced the size of the family
network substantially. Martial transitions have often been studied as one of the stressful life
events, and the effect of widowhood has been found to be the most potent out of all categories
(Ferraro & Barresi, 1982; Holmes & Rahe, 1967; Kitson, Babri, Roach & Placidi, 1989). Loss of
employment caused by unemployment or retirement also has been linked with causing
psychological distress (Mosca & Barrett, 2014; Segel-Karpas, Ayalon & Lachman, 2018),
especially among men (Oliffe et al., 2013; Shiba et al., 2017). Yet, social support has been found
to have a moderating role in this association (Van Solinge & Henkens, 2007). The fourth chapter
of this dissertation tests whether internet use can be a means of social support to moderate the
association between marital and work transitions and depression and loneliness of older people.
11

1.5. Objectives of the Present Dissertation
The objective of the present dissertation is to examine the role of internet use by focusing
on older adults situated in different contexts related to risk factors for deterioration in
psychological well-being and examine whether the use of internet communication is effective for
promoting positive psychological functioning for these people within these contexts. The specific
aims of each chapter are outlined below.
Chapter 2 examines the effect of using electronic communication on psychological well-
being and depressive symptoms of older people by living arrangements, with a particular focus
on people living alone compared to people living with spouses and others. In this paper, internet
communication use is measured based on the use of emails/texts usage. A rich set of social
network and engagement characteristics are accounted for in the model to better estimate the
effect of electronic communication use on predicting changes in psychological well-being and
depressive symptoms over eight years with controls for other factors affecting the outcomes.
This study tests whether internet communication use serves as an efficient means of enabling
communications and interactions to keep social contacts for older people living alone more so
than their counterparts living with spouses and others.
Chapter 3 compares the effect of internet use as a medium for social contacts on reducing
depression of older people by gender in two different countries, the U.S. and South Korea. The
main goal of this paper was to test the effect of the internet use on depressive symptoms in two
countries, and see if this effect is differentially related to distance from adult children between
men and women. Prior studies on internet use and depression did not focus on gender differences
although many studies found the prevalence of depression to be higher among women and have
been limited in its implications to mostly Western countries. This paper contributes to a growing
12

literature on the effect of internet use through cross-national research that also accounts for
gender differences in depressive symptoms.
Chapter 4 analyzes the effect of the internet use on significant life transitions and
psychological distress. Significant life transitions are measured with changes in marital status
(i.e., became married, became separated or divorced, and became widowed from being married)
and work status (i.e., became unemployed and became retired from working), and psychological
distress is assessed with depressive symptoms and loneliness. The overall aim of this paper is to
empirically test the effect of the internet use for people who had significant life events over two
time points.
The findings from the overall chapters will contribute to growing literature on internet
use and social connectedness by testing the potential of heterogeneity in the effect of internet use
on older people and advance our understanding on the effect of internet use on promoting social
connection of older population in different contexts. Three different datasets comprised of
nationally representative samples were used to address the questions raised in each chapter. The
implications of the overall results are also discussed at the end to inform the development of
programs and initiatives geared toward enhancing psychological well-being and reducing risks of
depression of older adults through the use of internet communication.

13

Chapter Two: Living Alone, Electronic Communication and Psychological Functioning:
Findings from the National Health and Aging Trends Study

2.1. Introduction
Internet use has become an influential factor on older adults’ health and well-being.
Approximately 63% of Americans aged 65 and over use the internet every day, for various
purposes ranging from shopping, online banking, renewing prescriptions, and staying in touch
with family and friends (National Telecommunications and Information Administration, 2018).
Many studies have found the internet usage to offer promising instruments for addressing
emerging issues and challenges associated with an aging population. Being able to access and
use the internet has been found to have implications for health care (Nam, Han & Gilligan, 2018;
Levine, Lipsitz & Linder, 2017; Levy, Janke & Langa, 2015); depressive symptoms (Elliot,
Mooney, Douthit & Lynch, 2013; Cotten, Ford, Ford & Hale, 2014), and psychological well-
being (Chopik, 2016; Sims, Reed & Carr, 2017).
One subgroup of older people who are at a greater risk of depression and poor
psychological well-being are those who live alone. Living alone has also been found to be an
important determinant of social isolation and loneliness among those aged 65 and over (Wenger,
Davies, Shahtahmasebi & Scott, 1996). While people who live alone might be more likely to
benefit from new means of communication, older people living alone have been found to be less
likely to have internet access at home than people living with others (van Deursen & Helsper,
2015). This subgroup also represents a fast-growing segment of the aging population in the U.S.
14

accounting for almost 30% (Ausubel, 2020). Among women aged 75 and over, the percentage is
close to a half (45%) (Administration on Aging, 2018).
The use of electronic communication for people living alone has the potential to be a tool
for addressing their physical risk of social isolation. Use of electronic communication has been
linked to higher social engagement and fewer depressive symptoms (Choi & DiNitto, 2013
a
;
Kim, Lee, Christensen & Merighi, 2016), but it remains unclear whether this association is
present for people living alone in the community. The psychological health of people who live
with others may be more influenced by interpersonal interactions than the use of electronic
communication. On the other hand, for people living alone, the use of electronic communication
may have greater benefits, as this can become a major source of social connections.
Using eight waves of a nationally representative and longitudinal data from the National
Health and Aging Trends Study, the current study examines how the use of electronic
communication measured by frequency of using e-mail/text usage is associated with the
trajectories of psychological well-being and depressive symptoms over eight years, and tests
whether this association differs for people who are living alone or residing with others. The
findings of the paper will be useful for determining whether it is appropriate to promote the use
of internet communication as a tool for improving psychological well-being and what the
implications are for people with different environmental contexts.

2.1.1. Background
Living arrangements are determined by a number of individual and societal factors.
Those entering old age recently are members of the baby-boom generation which has seen
15

marked changes in life circumstances including rising employment among women, divorce rates,
childlessness, and lower kin availability, changes which were unparalleled in previous
generations (Macunovich, Easterlin, Schaeffer & Crimmins, 1995). More economic resources
and improvements in health also enabled many people to live independently at older ages, and
intergenerational co-residence have become a choice ensuing from specific health declines
(Agree, 2018). Baby boomers are expected to be the loneliest generation in U.S. history with
about eight million people lacking a close kin, a spouse or living child (Adamy & Overberg,
2018); the proportion of Americans living alone has more than doubled from 13% in 1960 to
28% in 2019 (U.S. Census Bureau, 2019). This increasing prevalence of one-person households
calls for more empirical attention to how living alone is associated with the quality of life in later
adulthood.
Living arrangements also have a significant influence in meeting the needs and desires of
people at older ages. The person-environment (PE) fit perspective suggests that a mismatch
between salient needs of the individual and environmental properties leads to adverse
consequences on the well-being of older people (Kahana, 1982; Kahana, Liang & Felton, 1980;
Park, Han, Kim & Dunkle, 2017). With increasing age, people become more dependent on their
close network members, such as spouses or family members (Carstensen, Fung & Charles,
2003). Declines in health and functioning may also limit older people’s ability to engage in
social interactions outside the home. The physical living environment can thus also indicate the
social context for older people, where some have personal interactions while others do not. In
this regard, the use of electronic communication may enable interpersonal communication in the
absence of interactions with co-residing members in the household. For people who live alone,
16

their use of emails or texts to interact with family and other members in the community may be
relatively more important than for their counterparts who are living with others.
Living alone has been found to be linked to increased risks of depression (Djernes, 2006;
Stahl, Beach, Musa & Schulz, 2017); loneliness (Victor, Scambler, Bond, & Bowling, 2000);
more physical disabilities (Henning-Smith, Shippee & Capistrant, 2017); and higher mortality
risks (Klinenberg, 2016; Teguo et al., 2016). However, not all research findings indicate that
living alone leads to negative consequences on well-being. For instance, people living alone had
a lower risk of depression compared to people living with spouses and partners among women
aged 60 and over in a longitudinal analysis of the Nurses’ Health Study (Michael, Berkman,
Colditz & Kawachi, 2001). Henning-Smith (2016) found that older people living alone had
worse psychological well-being than those living with a spouse, but they did not have worse
well-being than those living with others without a spouse using the sample of adults aged 65 and
over in a cross-sectional analysis of the National Health Interview Survey. While the overall
evidence demonstrates the complexity of the implications of living arrangements on mental and
physical health of older people, this also indicates a gap in the existing literature on examining
the effect of living arrangements on psychological well-being of older people using a
longitudinal analysis of the nationally representative sample.
Using the internet specifically designed for communication may reduce potential risks of
social isolation for people living alone by enabling online social interactions. Challands and
colleagues (2017) found that online social connectedness, measured with how frequently the
participants used the internet for communicating with friends and family and meeting new
people online, significantly reduced depression for people aged 65 and over after their driving
cessation. Chopik (2016) also found that social technology use, including the use of e-mail,
17

social networking sites, and online chatting/instant messaging and others, was associated with
higher subjective well-being and fewer depressive symptoms through the mediating effect of
reduced loneliness. In this regard, the use of internet for communication may also be effective in
decreasing risks of depression and increasing psychological well-being for older adults living
alone in the community, a group that can benefit more from virtual social interactions than
people co-residing with their spouses or others.
The overall findings indicate the effect of electronic communication use on well-being
and depressive symptoms may be different for older people depending on their individual and
environmental characteristics. Understanding the effect of electronic communication use for
people living alone, a group that has been understudied despite their heightened vulnerability, is
crucial in evaluating the effectiveness of internet use as a potential tool for enhancing mental
health of older adults. Previous research suggests the use of electronic communication as one of
the potentially effective interventions for promoting psychosocial outcomes of older adults
(Forsman et al., 2018), but no prior study has examined how living alone is associated with the
relationship between internet use and psychological well-being and depressive symptoms among
people aged 65 and older. The present study aims to address this gap in literature by examining
how the use of electronic communication is associated with the trajectories of well-being and
depressive symptoms over the eight-year periods, and whether this association differs by living
arrangements using a nationally representative sample from the National Health and Aging
Trends Study.

18

2.1.2. Research Hypotheses
The present study examines whether the use of electronic communication affects the
trajectories of psychological well-being and depressive symptoms differently for people living
alone compared to people who live with others. Based on prior studies on internet use, well-
being, and depressive symptoms, it is expected that the use of electronic communication will be
associated with better psychological well-being (Hypothesis 1) and fewer depressive symptoms
(Hypothesis 2) among people who live alone compared to people who live with others.

2.2. Methods

2.2.1. Data
The current study used the 2011-2018 National Health and Aging Trends Study
(NHATS), a nationally representative sample of Medicare beneficiaries aged 65 and over. The
NHATS began in 2011 with a stratified three-stage sampling design that included oversampling
of Black and Hispanic older people as well as people aged 90 and older (Kasper, Judith &
Freedman, 2017). The NHATS is administered every year, and a replenishment sample was
added in 2015 to the original sample in order to account for sample attrition.
The sensitive data on the exact age and ethnicity of participants was also obtained and
merged for the final analysis. In 2011, a total of 8,245 respondents completed the survey, out of
which 7,075 participants were followed-up in Round 2. From Round 2 to 8, 230 participants who
had missing data on well-being, and 45 participants missing on social network sizes were
19

excluded. The final analytic sample consisted of 6,897 participants representing 2,304
participants living alone and 4,593 participants living with spouses or with spouses and others.

2.2.2. Measures
Psychological functioning
Psychological functioning was measured with two outcomes: subjective well-being and
depressive symptoms. Subjective well-being was measured with 7 questions that include both
hedonic and eudemonic aspects of well-being (Ryan & Deci, 2001). This scale has been
validated for its reliability and details are described in other studies (Freedman et al., 2014; Kim,
Lehning & Sacco, 2016). The summary score ranged from 0 to 22, with higher scores indicating
greater well-being (Cronbach’s α= .74). Depressive symptoms were measured with the Patient
Health Questionnaire 2 (Kroenke, Spitzer & Williams, 2003). Two screening questions on
depression were asked: “Over the last month, how often have you: (1) had little interest or
pleasure in doing things; (2) felt down, depressed, or hopeless?” The response was measured on
a 4-point Likert scale coded as not at all (0), several days (1), more than half the days (2), nearly
every day (3), and a summary score combining the responses to two questions was constructed as
an overall index measure (0-6) (Kroenke, Spitzer & Williams, 2003). The measure has been used
as an indicator of depressive symptoms in other NHATS studies (Elliot, Mooney, Douthit &
Lynch, 2013; Pohl, Cochrane, Schepp & Woods, 2018).

20

Use of Electronic Communication
The frequency of using electronic communication was assessed with a question on the
frequency of using emails or texts. For each wave, the NHATS asked the participants how often
they sent messages through emails or texts in the last month, and the response was categorized
into none (0), rarely/some (1), and most days (2). The frequency of using electronic
communication at baseline is used to indicate the use of electronic communication.

Covariates
The multivariate models also included a diverse range of covariates measured at baseline
indicating demographic, socioeconomic status, health, and social relationships that may be
simultaneously associated with psychological well-being, depressive symptoms, and use of
electronic communication. Sociodemographic covariates included age, gender, race/ethnicity,
and education. Health covariates included self-reported health, number of doctor-diagnosed
chronic conditions (count of arthritis, cancer, diabetes, heart disease, hypertension, lung disease,
stroke) and functional limitations in activities of daily living (ADL: a count of eating, bathing,
dressing, toileting, bed transfer, moving inside the house). Social relationship measures included
the number of close people in the participant’s social network (0-5), and the binary indicators of
whether the respondents had visits to family or friends, and whether they went out for social
events, participated in clubs, classes, or organized activities, and whether the participants
attended any religious services last year. Appendix A provides additional detail on how each
social relationship measure was coded.

21

2.2.3. Analytic Strategy
The use of electronic communication as well as demographic and socioeconomic
differences between people living alone and people living with others were compared using χ
2

tests and t-tests. Next, growth curve modeling was used to examine how the use of electronic
communication is associated with psychological well-being and depressive symptoms by living
arrangements over time. Multilevel growth curve models enable examination of how trajectories
of well-being and depressive symptoms changed over time within the same individual (intra-
individual differences) and how these patterns differ across individuals (inter-individual
differences) (Curran, Obeidat & Losardo, 2010; Raudenbush & Bryk, 2002). The mixed
command in STATA 14 was used to estimate the changes in psychological well-being and
depressive symptoms adjusting for repeated measurements within the same individual over time
(Level 1) and how this varies between individuals (Level 2). Equations for each level were as
below:
Level 1
𝑌 𝑖𝑡
= 𝜋 0𝑖 + 𝜋 1𝑖 𝑇𝑖𝑚𝑒 𝑖𝑡
+ 𝜀 𝑖𝑡

Where 𝑌 𝑖𝑡
is psychological well-being or depressive symptoms for individual i at Time t.
𝜋 0𝑖 is the intercept of the outcome for individual i, 𝜋 1𝑖 is the rate of change in the outcome for
individual i, 𝜀 𝑖𝑡
indicates random error in the outcome for individual i at Time t.

22

Level 2
𝜋 0𝑖 = 𝛾 0𝑝 + ∑ 𝛾 0𝑝 𝑍 𝑝𝑖
+ 𝑢 𝑝𝑖

Here 𝜋 0𝑖 represents how the initial level of psychological well-being or depressive
symptoms can vary across individuals depending on the vector of covariates, 𝑍 𝑝𝑖
, referring to
baseline age, gender, race/ethnicity, education, chronic conditions, ADL limitations, visits to
family/friends, going out for social events, attending religious services, participating in clubs or
organized activities, social network size, and the use of electronic communications. 𝑢 𝑝𝑖
shows
residual variance, referring to the deviation of an individual i from the overall intercept and
slope.
For specification of time, time in the survey was used because of a better model fit; a
quadratic term for time was tested but it was neither significant nor improved model fit, and thus
a linear specification of time was used in the models. Model 1 presents the unconditional model
showing the trend in well-being and depressive symptoms over time, Model 2 shows the model
adjusting for age, gender, race/ethnicity, and education, Model 3 presents the fully adjusted
model adjusting for all sociodemographic, health, and social network characteristics, and Model
4 shows the interaction between use of electronic communication and living arrangements.
Descriptive analyses used survey weights and stratum to account for the NHATS complex
sampling design (Kasper, Judith, Freedman, Vicki, 2017).

23

2.3. Results
Table 1 presents the weighted sample characteristics by living arrangements on the
selected indicators of demographic, socioeconomic, health, and social network characteristics.
On average, the use of overall electronic communication was lower among people living alone
compared to people living with others. The proportion of non-users of emails/texts was much
higher among people living alone compared to people living with others (59.67% vs. 48.52%, p
<.001). The overall sociodemographic and health profile of people living alone also was poorer
than their counterparts who were living with others. The mean age of the sample was
approximately 74, with the range of 65 to 105. People living alone were significantly older
(76.37 [SD=7.82] vs. 73.24 [SD=6.02], p <.001); more likely to be women (73.88% vs.49.35%, p
<.001); more likely to have less than high school education (19.37% vs.16.97%, p <.001).
There were statistically significant differences between people living alone and people
living with others also in race/ethnicity, a count of chronic conditions, and functional limitations.
In general, the social interactions of the two groups did not differ much. A slightly higher
proportion of people visited family or friends among people living together (90.53%) than people
living alone (88.99%), and a similar trend was observed for going out for social events (85.28%
vs. 76.31%, p <.001). No significant difference was found for social network size, attending
religious services, and whether the participants engaged in clubs, classes, or organized activities.
Table 2.1. Baseline Sample Characteristics
Variables of interest Total (N=6,897)
Living Alone
(n = 2,304)
Living with Others
(n = 4,593) P
Frequency of electronic communication use

None 51.84 59.67 48.52 <.001
Some days 21.80 18.52 23.18

Most days 26.36 21.81 28.30

Age 74.17 (6.72) 76.37 (7.82) 73.24 (6.02) <.001
Female (%) 56.66 73.88 49.35 <.001
Ethnicity

24

NH White 84.30 84.53 84.19 <.05
Black 7.41 8.99 6.74

Hispanic 5.63 4.60 6.08

Others 2.66 1.88 2.99

Education

Less than high school 17.69 19.37 16.97 <.001
High school diploma 26.11 29.05 24.86

College and above 56.20 51.58 58.17

Chronic conditions 2.04 (1.30) 2.13 (1.38) 2.00 (1.27) <.01
(0-7)

ADL limitations (0-6) 0.50 (1.05) 0.56 (1.15) 0.47 (1.00) <.05
Social Network (0-5) 2.04 (1.33) 2.02 (1.40) 2.05 (1.29) 0.546
Visit family/friends 90.07 88.99 90.53 <.01
Going out for enjoying social
events
82.61 76.31 85.28 <.001
Attend religious services 60.35 59.23 60.83 0.298
Participate in clubs, classes, or
organized activities
43.53 44.18 43.25 0.561
Note. Mean and SD in parentheses and weighted percentages.

The bivariate analyses showed significant correlations across the key variables of interest.
Age and gender were negatively associated with a frequency of use of emails/texts while
education, social network size, visiting family/friends showed positive associations (Table 2).

Table 2.2. Pearson’s Correlations among the Key Variables

1. Frequency
of E-
mail/Text use
2. Age 3. Gender 4. Education
5. Visit
family /
friends
6. Social
network
size
1. Frequency of
E-mail/Text use 1

2. Age -0.28*** 1

3. Gender -0.08*** 0.09*** 1.00

4. Education 0.43*** -0.12*** -0.05*** 1.00

5. Visit
family/friends
0.14*** -0.09*** 0.05* 0.13*** 1.00

6. Social network
size
0.16*** -0.05*** 0.18*** 0.15*** 0.12*** 1.00

25

Table 3 shows the results on the association between the use of electronic
communication and psychological well-being from growth curve models. Model 1 shows that the
average score at baseline was 17.06, and psychological well-being decreased over time (β= -
0.134; SE = 0.007, p < .001). When age, gender, race/ethnicity, education are included in Model
2, the use of electronic communication is positively associated with psychological well-being
while living alone is negatively associated. Model 3 presents the fully adjusted model with
additional inclusion of health and social network characteristics, and the use of electronic
communication independently showed positive associations for the initial level of psychological
well-being, with a significant increase over time for the category use of communication on most
days (β= 0.051; SE = 0.021, p < .05). Living alone also remains negatively associated with
psychological well-being (β= -0.320; SE = 0.078, p < .001), but this association is not significant
over time. The interaction between living alone and the use of electronic communication in
Model 4 was not significant and the main effect of using electronic communication and living
alone remained significant similar to Model 3.
In Table 4 which presents the models for predicting depressive symptoms, Model 1
indicates that the average score on depressive symptoms was 0.93, with a significant increase
over time (β= 0.016; SE = 0.003, p < .001). Model 2 shows that the use of electronic
communication is negatively associated with depressive symptoms while living alone is
positively associated. When a full set of covariates are included in Model 3, the use of electronic
communication remains negatively associated with depressive symptoms but there is no
significant effect over time. On the other hand, whereas living alone at baseline was associated
with increased initial levels of depressive symptoms (β= 0.096; SE = 0.027, p < .001), this effect
decreases over time (β= -0.014; SE = 0.007, p < .05). Similar to psychological well-being, there
26

was no significant interaction between living alone and the use of electronic communication in
Model 4 while the main effect remained the same as in Model 3.

Sensitivity Analyses
Because the replenishment sample was added in 2015 and there is a concern for sample
attrition over the years in modeling the changes in psychological well-being and depressive
symptoms, the multi-level models were re-estimated using the data from 2011 to 2014 to test the
robustness of the findings. The results are presented in Supplementary Table 5-6. When the
time frame is limited to 2011 to 2014, there was a significant interaction between living alone
and the use of electronic communication over time (β= 0.191; SE = 0.086, p < .05) and changes
in psychological well-being. However, the result remained the same for depressive symptoms,
with only the intercept for the main effect significant whereas no interaction coefficient was
significant. This showed that the result examining trajectories of depressive symptoms was more
robust and less influenced by sample attrition than psychological well-being.
Additionally, due to variance in the proportion of people visiting family or friends and
going out for social events, further sensitivity analyses were conducted, which are shown in
Supplementary Table 7. The first two models were estimated with the social engagement
variable summarizing all four social participation measures, and the second two models were run
without the two variables. The main results for the interaction of psychological well-being and
depressive symptoms with living arrangements remained identical to Table 3-4.

27

28

29

30

31

32

2.4. Discussion
The proportion of older people who are living alone has been rapidly growing, but no
prior study has focused on how the effect of using electronic communication is associated with
psychological well-being and depressive symptoms of people living alone compared to people
co-residing with other people. This paper aimed to examine how the use of electronic
communication was associated with psychological well-being and depressive symptoms among
people aged 65 and over by living arrangements. The results showed that the prevalence of using
electronic communication was substantially lower among people who live alone compared to
people living with others. However, the effect of using emails/texts did not differ by living
arrangements. Moreover, using emails/texts was only significantly associated with the initial
levels of depressive symptoms but not change in depressive symptoms over time for both people
living alone and people living with others. The overall findings suggest that the use of electronic
communication did not provide a means to supporting mental health and reducing risks of social
isolation for older people living alone in the community.
This is the first study to examine the effect of using electronic communication by living
arrangements. The study provides the national estimate of electronic communication use and its
effect for people living alone compared to people living with others among the age group of 65
and over. The first hypothesis was supported in that a very low prevalence of electronic
communication use was found among people living alone compared to people living with others.
The result is consistent with prior studies highlighting “digital divide” in certain sub-groups of
older population. Prior research has linked lower levels of education and economic resources
such as income (Choi & DiNitto, 2013
b
; Yu, Ellison, McCammon & Langa, 2016) as the main
reason for digital divide. The current study found that the average years of education among
33

people living alone was significantly lower compared to people living with others. Additionally,
the lower prevalence of electronic communication use among people living alone may be in part
due to affordability of accessing the internet at home (van Deursen & Helsper, 2015). Choi and
DiNitto (2013
b
) noted that people who were living alone showed higher scores on their
confidence and interest in using the internet, and low-income adults who did not use the internet
did not show reluctance to learning to use it if they could. This indicates that more focused
efforts should be taken to expand access to the internet for older adults living alone in the
community through public policies and programs.
The findings provide evidence that the effect of electronic communication did not differ
by living arrangements for enhancing psychological well-being, which did not support
Hypothesis II. The finding on psychological well-being suggests that while the use of electronic
communication appeared to have significant effects on higher psychological well-being both at
the initial level and change over time, this effect did not significantly differ by living
arrangements. Positive relationships between psychological well-being and internet use have
been well documented in other studies (Heo, Chun, Lee, Lee, & Kim, 2015; Zambianchi &
Carelli, 2018). For instance, Sims and colleagues (2017) found that a higher use of information
and communication technology was significantly associated with higher psychological well-
being across all domains in a cross-sectional analysis (i.e., life satisfaction, goal attainment, and
loneliness). In another study using two waves from the English Longitudinal Study of Aging, a
significant association between internet use and psychological well-being was found, but only in
the eudemonic measure (Quintana, Cervantes, Sáez & Isasi, 2018). The present study used the
summative measure of well-being assessing both hedonic and eudemonic components derived
from the confirmatory factor analyses (Kim, Lehning & Sacco, 2016), and adds to this growing
34

body of literature that using electronic communication does not lead to higher psychological
well-being among people living alone in the community. Further studies are necessary to clarify
the association between electronic communication use and psychological well-being for various
sub-groups of older adults and how this is related to different types of living arrangements.
The hypothesis on depressive symptoms was also not supported in that no difference was
found in the effect of electronic communication use for reducing depressive symptoms by living
arrangements over time. Most prior studies underscore the effect of internet use in reducing
depressive symptoms among older adults. Elliot and colleagues (2013) found a significantly
negative association between the frequency of emails/texts and depressive symptoms among
community-dwelling older people in their sub-analyses using the 2011 NHATS, and Lee and
colleagues (2018) reported the similar results among cancer survivors using the same data.
Cotton and colleagues (2014) examined the effect of using the internet from the Health and
Retirement Study over six years and found that using the internet was significantly associated
with a reduced probability of depression. The current paper extends this prior research to the
longitudinal analysis of the effect of using electronic communication on depressive symptoms,
and suggests the heterogeneity in the effect of electronic communication use in this association.
Although the use of electronic communication was significantly associated with the reduced
initial levels of depressive symptoms, the association was not significant for change over time
and this was consistent for both people living alone and people living together with others. The
present study raises possibility that the observed effect of using the electronic communication
may be different at intraindividual level. Without taking into account changes within individual
and across contextual levels, promoting electronic communication use as a tool to reduce social
isolation of older people may not be as effective.
35

There are several limitations that need to be noted and addressed in future studies. First,
the participants’ perception and attitudes toward the internet were not assessed due to limitations
of the survey questionnaire. As people’s perception of usefulness, confidence, and attitudes have
been found to be associated with use of technology (Venkatesh, Morris, Davis & Davis, 2003),
including these traits in the model may have given a more in-depth interpretation of the results.
Including this assessment will be useful to directly identify the cause of the low usage rate
among people living alone aged 65 and over in future analyses. Second, the response category
regarding the frequency of e-mails did not indicate the specific time frame regarding the
frequency of usage. Future studies can use a mixed method design to ask a more specific
question to assess the frequency of electronic communication use that is effective for reducing
depression for older people living alone in the community. Third, examination of loneliness was
not possible due to absence of the psychological measure. Since this trait has recently emerged as
a key factor in depression of people who live alone (Jacob, Haro & Koyanagi, 2019; Stahl,
Beach, Musa & Schulz, 2017), the pathways through which electronic communication use
influences well-being and depression could have been better analyzed with loneliness.
In conclusion, the overall findings contribute to delineate the effect of electronic
communication use and how the usage is differentially associated with psychological well-being
and depressive symptoms for people who live alone compared to people living together with
others and offer two policy implications. The first is to design internet-based interventions that
recognize differential implications of electronic communication use. The effect of electronic
communication use may differ as older population is composed of heterogenous groups with
different individual contexts. The second is to identify an effective strategy to reduce barriers to
accessing the internet. Digital divide has been raised as a concern for people who are ethnic
36

minority (Mitchell, Chebli, Ruggiero & Muramatsu, 2018) and low-educated (Yu et al., 2016).
The present study adds people who live alone as another category of people experiencing this
barrier among diverse groups of aging population in the United States. Overall, the findings call
for more policies and programs to expand access and use of electronic communication and more
tailored interventions to reduce potential risks of social isolation and depression for older people
living alone in the community.

37

Appendix. The Social Participation Measures from the National Health and Aging Trends Study
No. Variable Questions in the Survey Range / Response Categories
1 Visits to family
or friends
In the last month, did {you/SP} ever visit
in person with friends or family not living
with {you/him/her}, either at
{your/his/her} home or theirs?
1. Yes
2. No
3. Refused
4. Don’t know

2 Go out for
social events
In the last month, {besides participating in
club or group activities,} did {you/SP}
ever go out for enjoyment? This includes
things like going out to dinner, a movie, to
gamble, or to hear music or see a play.
1. Yes
2. No
3. Refused
4. Don’t know

3 Participate in
clubs, classes,
or organized
activities
In the last month, {besides religious
services,} did {you/SP} ever participate in
clubs, classes, or other organized
activities?
1. Yes
2. No
3. Refused
4. Don’t know

4 Attend any
religious
services
In the last month, did {you/SP} ever attend
religious services?
1. Yes
2. No
3. Refused
4. Don’t know

38

Chapter Three: Cross-National Examination of Differences in the Association between
Internet Use and Depressive Symptoms by Gender and Intergenerational Characteristics in
the U.S. and Korea

3.1. Introduction
Depression among older persons presents a global public health challenge. The World
Health Organization (WHO) estimates that mental health disorders are a leading cause of the
global burden of disease for older populations, as the prevalence of depression increases for both
men and women over age 55 (WHO, 2017). Social interactions with family and friends have
been found to play an important role in decreasing risks of depression by encouraging people to
stop ruminating over stress inducing factors and seek meaningful activities (Fiske, Wetherell &
Gatz, 2009). Internet use offers a potential medium that can reduce depression among older
people by enabling social interactions through online communication. Prior studies have
generally found that internet use was associated with reduced depression among older adults
(Cotten, Ford, Ford & Hale, 2014; Elliot, Mooney, Douthit & Lynch, 2014).
One of the primary reasons for older adults using the internet is to maintain contacts and
relationships with family members including adult children and grandchildren (Gubernskaya &
Treas, 2016; Harwood, 2000). While both Western and Asian countries have experienced rapid
changes in family structure, intergenerational changes occurred in an earlier period and at a more
gradual pace in Western countries than in Asian countries. Expectations about intergenerational
relationships and gender roles in the family context are also widely different in these two
regions, with greater expectations for frequent intergenerational contacts in Asian countries than
Western countries. Thus, the use of the internet for maintaining contacts between parents and
39

adult children may have different implications for older adults in Asian countries compared to
Western countries. For instance, Teo and his colleagues (2015) found that a higher frequency of
in-person contact (i.e., visits in person) with children, but not non face-to-face contact (i.e.,
talking over the phone), was associated with fewer depressive symptoms in the U.S. In contrast,
Roh and his colleagues reported that the effects of non face-to-face contact and face-to-face
contact with children were similar in reducing risks of depression in Korea (Roh et al., 2015).
Gender is another factor that may significantly affect the association of the use of the
internet and depressive symptoms. Cross-country research has consistently found the prevalence
of depression to be higher among women than men (Crimmins, Kim & Solé-Auró, 2010;
Kuehner, 2017). In addition, the patterns of social interactions with other family members differ
between men and women, with women generally expected to be the kin-keepers who maintain
relationship (Kim, Birditt, Zarit & Fingerman, 2019). There is also gender inequality in internet
use across countries. Ono & Zavodny (2007), in a comparison study of diffusion of computers
and internet use among 15 to 59 year-olds in five countries (i.e., U.S., Sweden, Japan, Korea,
Singapore), found that women used computers and the internet less than men in Japan and Korea
but there was no gender difference in the U.S. and Sweden. This suggests that a gender-specific
perspective may be useful in evaluation of the association between internet use and depressive
symptoms in cross-national research.
In this context, a comparative study on the internet use and depressive symptoms in the
U.S. and Korea may provide an important insight into understanding the role of internet use as a
medium for intergenerational contacts for reducing risks of depression. Two countries have a
widespread use of the internet, and the role of family is pivotal in addressing different needs of
older populations such as caregiving (Jang, Avendano & Kawachi, 2012; Solé-Auró &
40

Crimmins, 2014). At the same time, they also have different cultural values and expectations for
gender roles and intergenerational relationships (Ha, Yoon, Lim & Heo, 2016; Sung, 1990).
Korea presents a dynamic context of population aging and use of internet communication. The
country is projected to see one of the fastest increases in the older population among OECD
countries by 2050 (Yoon, 2013) and has one of the most widespread uses of the internet in the
Asian region (World Bank, 2017). Internet use can provide a medium to enable intergenerational
support over distance and time in the 21
st
century when many adult children increasingly live
apart from their older parents. Examination of the role of internet use for intergenerational
relationship on depressive symptoms of older adults in a cross-national context can elucidate
whether intergenerational contacts can present one mechanism linking internet use and
depressive symptoms for older people over differences in gender norms and cultural values. The
current paper used nationally representative data on older persons from Korea and the U.S. to
examine whether internet use is associated with depressive symptoms differentially between men
and women, and test whether geographic distance to adult children is linked to the relationship
between internet use and depressive symptoms accounting for intergenerational exchanges of
support and other demographic and social characteristics. The goal of this analysis is to
empirically determine whether using the internet provides a means of communication to reduce
depressive symptoms when older parents are living far from their children in countries with
different sociohistorical contexts.

41

3.1.1. Background
Globally, an estimated 322 million people suffer from depression with a higher
prevalence among women and peaking at ages 55-74 (WHO, 2017). Depression is linked to
many other health conditions in later life, including dementia (Byers & Yaffe, 2011; Mirza et al.,
2016; Wilson et al., 2002) and stroke (Goodwin, 2006). Depression is also linked to suicide
among adults aged 65 and over (Morgan et al., 2018), with a particularly high rate in South
Korea (Naghavi, 2019). With increases in age, people may become more vulnerable to
experiencing negative life events, including a loss of employment, a loss of health, and a loss of
members in their family or other types of social networks. These events, juxtaposed with changes
in physical functioning and capacity, can encourage the development of negative beliefs about
the self as well as the future (Fiske, Wetherell & Gatz, 2009).
Social support is an important resource for buffering the effect of depression. Positive
perceptions about availability of social support, frequency of contacts, and social network sizes
have been consistently associated with a lower risk of depression (Nam & Lee, 2019; Oxman &
Hull, 2001; Robins & Block, 1989). In particular, social support from children has been found to
play a crucial role in reducing risks of mental health problems among older parents (Bengtson &
Roberts, 1991; Djundeva, Mills, Wittek & Steverink, 2015). Internet use can play a key role in
maintaining contacts and social relationships over distance and time, and many studies have
linked using the internet to reduced depressive symptoms among older adults (Cotten, Ford, Ford
& Hale, 2014; Harerimana, Forchuk & O'Regan, 2019; Shaw & Gant, 2004). Examining the role
of social contacts in the association between the internet use and depressive symptoms may
elucidate under what circumstances internet use is related to reduced depressive symptoms.
Additionally, whether and how this association differs by gender within the context of
42

intergenerational relationships will add an international insight into evaluating the effect of
internet use.
Adult children can be important sources of emotional, financial, and physical support to
older adults. They may be more important in countries with less established pension systems
which do not provide adequate financial support in old age for most of the populations. In 2015,
Korea began to expand the Basic Livelihood Security Program (BLSP) to offer pension benefits
to people living below the minimum poverty threshold among those over 65, but the program is
still in its infancy. The average income of Koreans aged 65 and over is less than 70% of the
national average income (OECD, 2017). Due to this factor, nearly 95% of older Koreans report
private financial transfers, such as financial support from adult children, as the main source of
income. A recent national estimate shows that approximately 73% received irregular financial
support from adult children while 57% received regular financial support (Jung et al., 2017).
Additionally, a recent study suggests that Korean parents aged 65 and over who received
financial support from their adult children reported higher levels of happiness in both low and
high income groups, indicating that the presence of financial support plays a key role in older
parents’ well-being regardless of their socioeconomic status (Chun & Kim, 2016). On the other
hand, in the United States where a long and well-established social security system exists, only
about 11-12% of adult children aged 35 and over reported giving any financial support to older
parents (Friedman, Park & Wiemers, 2015). Older parents’ dependence on financial support
from their adult children also varies by race and ethnicity, with Black mothers more likely to
receive financial support from adult children than Whites (Park, 2017).
Children’s residential proximity also signifies an important source of care and support for
older parents. Older people living close to their adult children were less likely to relocate to a
43

care institution than people whose children were living far away (van der Pers, Kibele & Mulder,
2015). Acute health events, such as the onset of cardiovascular diseases, have also been found to
be associated with a higher likelihood of adult children’s closer geographic proximity to parents
(Choi, Schoeni, Langa & Heisler, 2015). Older parents who were living far away from their adult
children had higher depressive symptoms compared to parents who were living together with
their children (Liang & Zhang, 2017). The internet offers a tool to communicate and exchange
support over geographic distance and different time zones and may thus facilitate reducing
depressive symptoms and promoting well-being by enabling social contacts for older parents
whose children live far away.

3.1.2. Research Hypotheses
The current study contributes to existing literature on the use of the internet and
depressive symptoms by highlighting the roles of intergenerational interaction and exchange of
support in this association. Given the previous findings on using the internet and depression
(Cotten, Ford, Ford & Hale, 2014; Elliot, Mooney, Douthit & Lynch, 2014), it is predicted that
using the internet will lead to reduced depressive symptoms in the U.S. and Korea (Hypothesis
1). It is also hypothesized that the geographic proximity to children will be related to this
association but this association will differ between men and women, with women showing more
significant effect than men in using the internet for maintaining intergenerational contacts
(Hypothesis 2). Given a wider span of geographic distance in the U.S. compared to Korea, the
effect of internet use on geographic distance is expected to be more pronounced in the U.S. than
Korea.

44

3.2. Methods

3.2.1. Data
The paper used two nationally representative data sets to examine the association
between the use of internet and depressive symptoms by gender within the context of
intergenerational exchange. For the United States, the 2016 Health and Retirement Study
(HRS), a biennial survey of nationally representative Americans aged 50 and over, was used.
In 2016, a total of 20,918 participants were interviewed, and the sample was restricted to
people who were born in 1952 or before for age comparability with the Korean survey as age
was differently calculated between the two countries. Of the 10,421 participants who were age
eligible, we excluded 2,575 respondents missing information on the frequency of contact with
children; 726 participants missing data on depression; 1,187 respondents missing information
on intergenerational financial transfers; 798 respondents on distance from the closest child;,
and 15 respondents on living arrangements. The final analytic sample consisted of 5,120 (1,690
men and 3,366 women) persons with complete information.
For Korea, the 2017 Living Profiles of Older People Survey (LPOPS), a nationally
representative survey of South Koreans aged 65 and over conducted through coordination
between the Ministry of Health and Welfare and the Korea Institute for Health and Social
Affairs (KIHASA), was used. In 2017, a total of 10,299 participants were interviewed. After
excluding 216 proxy respondents, 223 respondents who did not have any children, and 252
respondents who had missing information on the frequency of contact with children, the final
analytic sample included 9,608 respondents (3,847 men and 5,761 women). The participants
without children were excluded from both countries.

45

3.2.2. Measures
Depressive symptoms
Depressive symptoms were measured with an 8-item version of the Center for
Epidemiologic Studies (CES-D) scale in the HRS, a short version of the 20-item CES-D scale
(Radloff, 1977). A summary score consisted of the number of yes responses to 8 questions and
had a range of 0 to 8. The questions asked whether respondents experienced the following
symptoms all or most of time: felt depressed, everything was an effort, sleep was restless, felt
alone, felt sad, felt happy (reverse coded), enjoyed life (reverse coded), and could not get going.
The measure has been examined for its validity and specificity, and have been used in many
prior studies as an indicator of depressive symptoms (Cotten, Ford, Ford & Hale, 2014; Steffick,
2000; Kwon, Kim, Lee & Park, 2017; Leggett, Sonnega & Lohman, 2018).
In the LPOPS, depressive symptoms were measured with a 15-item Korean version of the
Geriatric Depression Scale (SGDS-K), which was originally developed by Sheik and Yesavage
(Sheik & Yesavage, 1986) and translated into Korean (Bae & Cho, 2004). The scale measured 1-
week prevalence of the respondents’ reported symptoms using yes/no responses to 15 questions.
A summary score was calculated using the response to all 15 items with a range of 0 to 15,
following the method used in prior studies (Lee, Oh & Hong, 2018).
Because the two scales had different ranges, depressive symptom scores were
standardized by calculating the z-standardized depressive symptom scores using the sample
means and standard deviations in each country to minimize differences. Therefore, the regression
estimates indicate differences in depressive symptoms in terms of standard deviations in the
respective country.
46

Use of the Internet
In the HRS, the use of the internet was assessed with the question, “Do you regularly use
the Internet (or the World Wide Web) for sending and receiving e-mail or for any other purpose,
such as making purchases, searching for information, or making travel reservations?” In the
LPOPS, the participants were asked to indicate whether they were able to do the following
activities using the internet, computer, mobile phone, and tablet PC: receiving text messages,
sending text messages, searching information, reading news, taking photos/videos, listening to
online music, watching videos online, playing online games, using social media, doing online
shopping, or other activities. Participants were coded as the internet users if they were able to do
any of the listed activities except for texting in order to maintain comparability and reduce
correlations with contact measures.

Covariates
Sociodemographic measures indicating age, education, and wealth as potential
confounders were included in the analyses. Age was measured in years; education was
categorized into three groups (0-11 years [reference], 12 years, and more than 12 years); and
wealth was categorized into three groups (low [reference], middle, high).
Health may also affect depression so the covariates included a count of chronic
conditions (arthritis, cancer, diabetes, heart disease, hypertension, lung disease, stroke: range 0 to
7) and whether the participants had any limitations with activities of daily living.
In addition to demographic and social correlates, the indicators of intergenerational
exchange of support and contact between older parents and adult children were included. The
47

three indicators were: (1) financial exchange (none [reference], downward only, upward only,
and both), (2) distance from the closest child (co-residence [reference], living within 10 miles,
and living more than 10 miles), and (3) frequency of contacts with children (none [reference],
once or twice a year, every few months, once or twice a month, once or twice a week, three or
more times a week). Contact measures were composed of meeting up, speaking, writing in the
HRS and meeting up, contact via phone, letter, or texting in the Korean survey. A composite
measure averaging the frequency of the modes of contact is used for the analysis.

3.2.3. Analytic Strategy
First, the characteristics of participants in the U.S. and Korea stratified by gender were
compared to see if there are any significant differences in their demographic and social
characteristics, internet use, and depressive symptoms. Next, ordinary least squares regression
(OLS) analyses were performed to examine the effect of internet use on depressive symptoms
separately for two countries. OLS regressions were used as the dependent variables were
standardized in terms of deviations from the mean in similar ways in the two countries, and the
regression parameter estimates thus indicated the standardized differences in the two depression
scales with different ranges. In the first set of analyses, how use of the internet, gender, and
intergenerational factors are related to depressive symptoms in the two countries was tested and
whether the effect differed by gender in the two countries was investigated. In the second set of
analyses, the effect of using the internet on depressive symptoms was estimated in separate
regressions stratified by gender for each country to investigate how the effect differed between
men and women. The Model 1 tested the association between internet use and depressive
48

symptoms adjusting for a full set of covariates on demographic, social, physical health, and
intergenerational characteristics. The Model 2 included an interaction term to see if internet use
reduces depressive symptoms more for older parents living at a greater distance from their
children by adding an interaction term between the internet use and distance from the closest
child.

3.3. Results
Table 1 presents the weighted sample characteristics by gender in the U.S. and Korea. In
the U.S., approximately 57% of men and 56% of women were internet users, while in Korea, the
proportion was 47% for men and 30% for women. On the other hand, women showed higher
levels of depressive symptoms than men in both countries. Women were slightly older than men
and less likely to have high levels of education and wealth. Women also were significantly more
likely to live alone, had a higher number of chronic conditions and limitations with activities of
daily living.
The characteristics of intergenerational interaction and support differed significantly
between men and women in the two countries. For financial support, in the U.S. more than half
of men (59%) and women (62%) did not exchange any financial transfers with adult children,
and the most frequent category of financial transfer was parents giving to their adult children
(38% for men and 31% for women). In Korea, the most frequent category of financial transfer
was from adult children to parents (63% for men and 71% for women), and a bidirectional
exchange was also much more frequent (33% for men and 26% for women). An average
frequency of contacts with children was slightly higher in Korea (3.83 [SD=0.85] for men and
49

3.88 [SD=0.85] for women) compared to the U.S (2.80 [SD=1.09] for men and 3.13 [SD=1.09]
for women), although men had less frequent contact with children than women in both countries.
Although co-residence with adult children was higher in Korea, the proportion of parents living
far from their closest adult child was also higher in Korea. While 55% of parents reported living
within 10 miles of their closest child in the U.S., the percentage was only 31% for Korean
parents for the same category, and almost half reported living more than 10 miles from their
closet child.
50

51

The average frequency of contacts by geographic distance to the closest child and
internet use are shown in Table 2 for the U.S. and Table 3 for Korea. In both countries, internet
users had more frequent contact with adult child than non-users for each category of geographic
distance. Except for men in the U.S., frequency of contacts was highest among those who lived
farthest from their closest child and who used the internet. Similarly, significant associations
between geographic distance and frequency of contacts were observed for women in the U.S. and
for women and men in Korea, but not for men in the U.S.

Table 3.2. Average frequency of contacts by geographic distance and internet use in the U.S.
Men

Women

Mean SD P Mean SD P
Co-residing and non-users 2.45 (1.01) 0.79 2.43 (1.20) <.001
Live within 10 miles and non-users 2.75 (1.02)

2.65 (1.11)

Live more than 10 miles and non-users 2.23 (1.11)

2.96 (0.99)

Co-residing and users 3.13 (1.19)

2.95 (1.21)

Live within 10 miles and users 3.16 (0.97)

3.10 (1.00)

Live more than 10 miles and users 2.77 (1.08) 3.65 (0.97)

Table 3.3. Average frequency of contacts by geographic distance and internet use in Korea
Men Women

Mean SD P Mean SD P
Co-residing and non-users 3.64 (0.95) <.001 3.65 (0.94) <.001
Live within 10 miles and non-users 3.56 (0.85)

3.68 (0.82)

Live more than 10 miles and non-users 4.23 (0.67)

4.31 (0.68)

Co-residing and users 3.78 (0.83)

3.85 (0.92)

Live within 10 miles and users 3.60 (0.81)

3.76 (0.76)

Live more than 10 miles and users 4.32 (0.65) 4.33 (0.64)
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001
52

53

Table 4 shows the OLS regression results predicting depressive symptoms in the U.S.
and Korea accounting for demographic, economic, social, and intergenerational characteristics.
In Model 1 presenting the fully adjusted model, the main effect of using the internet was
inversely associated with depressive symptoms in both countries except for men in the U.S.
When the interaction term between internet use and distance from the closest child was included
in Model 2, the results were significant for women who were living more than 10 miles from the
closest child (b=-0.257 [SE=0.123], p < .05) in the U.S, but not for men (b=0.151 [SE=0.166], p
= .36). For Korea, the results were opposite, with men showing lower depressive symptoms with
internet use for more than 10 miles (b=-0.226 [SE=0.080], p < .01), and marginally associated
for living within 10 miles (b=-0.153 [SE=0.083], p = .06). No significant interaction was found
between internet use and distance from the closest child for women in Korea. The overall results
showed that our hypothesis was partially supported in that we found significant differences
between men and women in the interaction effect between the distance from a child and internet
use, but only one gender was found to have a significant effect in both countries.
Figure 1 graphs the interaction from Model 2 in Table 4 for the U.S. For the U.S., the
graded levels of depressive symptoms by internet use and the distance from the closest child was
more pronounced among women compared to men. For the interaction from Model 2 for Korea
depicted in Figure 2, men showed the graded levels of depressive symptoms by internet use and
geographic proximity to the closest child with the users living farthest from the closest child
showing the lowest levels. For women in Korea, those who lived within 10 miles of the closest
child who used the internet showed the lowest levels of depressive symptom.

54

Figure 3.1. Predicted standardized depressive symptoms by internet use and geographic distance
from the closest child in the U.S.

-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0.50
Co-residence Within 10 Miles More than 10
Miles
Predicted Depressive Symptoms
Geographic Distance from the Adult Child
U.S - Men
Non-Users
Users
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0.50
Co-residence Within 10 Miles More than 10
Miles
Predicted Depressive Symptoms
Geographic Distance from the Adult Child
U.S - Women
Non-Users
Users
*
55

Figure 3.2. Predicted standardized depressive symptoms by internet use and geographic distance
from the closest child in Korea

-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0.50
Co-residence Within 10 Miles More than 10
Miles
Predicted Depressive Symptoms
Geographic Distance from the Adult Child
Korea - Men
Non-Users
Users
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0.50
Co-residence Within 10 Miles More than 10
Miles
Predicted Depressive Symptoms
Geographic Distance from the Adult Child
Korea - Women
Non-Users
Users
**
56

3.4. Discussion
The overall findings highlight the importance of internet use for communication over
geographic distance for older parents in cross-national research. The study found that a
proportion of internet users was significantly higher in the U.S., with a more balanced use
between men and women. For Korea, the usage was higher among men compared to women,
indicating that the gender gap still exists in internet use in this age group. Depressive symptoms
were higher among women than men in both countries, which was consistent with prior research
on gender differences in the prevalence of depression (Kuehner, 2017). Internet use was
associated with lower depressive symptoms for older women in the U.S. and older men in Korea
who were living far away from their closest child while no significant effect was found for the
other gender in both countries.
While prior literature has reported the effect of using the internet in attenuating
depressive symptoms (Cotten, Ford, Ford & Hale, 2014; Elliot, Mooney, Douthit & Lynch,
2014; Jun & Kim, 2016), a test including social contacts with a gender-specific examination
from the present study adds nuance to this finding. The current study found that depressive
symptoms were higher among women compared to men in both countries, which was consistent
with prior research on depression in the respective countries (Cho, Jin & Kang, 2018; Crimmins,
Kim & Solé-Auró, 2010). Additionally, the effect of using the internet on depression was
different by gender within the context of intergenerational exchanges of support and
communication. For women in the U.S., using the internet was significantly associated with
lower depressive symptoms when they resided far from their closest child. This association was
not statistically significant for men. Two factors can potentially explain this finding. First, the
expression of intergenerational support may exhibit differently between men and women. For
57

men, providing financial transfers was a more prevalent mode of support, while for women
giving emotional support through social contacts was more common. This pattern was also
observed in Korea where men received less and gave more financial support while women had
more frequent contacts with children, which may indicate different perceptions regarding the
value of different modes of intergenerational support between men and women. Second, the
result may indicate that men and women may have different purposes for using the internet in the
U.S. Kim and her colleagues (2017) found that using information and communication technology
(ICT) was positively associated with all types of women’s social engagement, but for men, only
certain types of social engagement such as in-person visits showed positive associations, through
a study of a nationally representative sample of Medicare recipients aged 65 and over. This may
indicate that the internet use has stronger associations with social interactions for older women
while older men may use the internet for different purposes in the U.S.
In Korea, however, using the internet was significantly associated with lower depressive
symptoms for men when they resided more than 10 miles away from their closest child. This
may stem mostly from the gender gap in internet use in this sample. Gender disparity in internet
use highlights a persistent gap in levels of education among older women compared to older men
in the country (Park, 2014), a primary influencer of access to using the internet among this age
group (Wong, Law, Fung & Lee, 2010; Yu et al., 2016). Given that the use of the internet as well
as levels of education are becoming more balanced between men and women among younger age
groups, this result is likely to change and future research can investigate generational differences
in the effect of using the internet in Korea. Also, a recent report from the National Statistical
Office reported that older parents aged 65 and over in Korea showed higher life satisfaction
when they lived independently from their children than people who were co-residing with their
58

children (Shim, 2018). Therefore, the result from the current study may indicate that the co-
residence status can negatively affect psychological well-being of older men in Korea.
Contrary to the hypothesis on country differences, the effect of internet use on geographic
distance to children was similar in the U.S. and Korea. The present study suggests that
intergenerational social contacts may be one mechanism behind the association between internet
use and reduced depressive symptoms. This signifies the importance of intergenerational context
in the use of the internet among older adults. Internet communication use can be a tool to connect
and strengthen intergenerational relationships. Using the internet with an intergenerational
exchange of support and resources can be a medium to connect different generations and
promote well-being of both older and younger adults (Han & Braun, 2013). Promoting internet
communication use as a means to strengthen intergenerational relationship thus can have
significant implications and prevent social isolation of people within the community through
mutual social interactions.
Several limitations need to be noted for the interpretation of these findings. First, the
current study is cross-sectional and no causal inference can be made from the results. For
instance, the observed relationship can be caused by the possibility that depressed older people
contacted their adult children less. Future studies can investigate this association in a longitudinal
design. Second, several measures were not completely comparable between the two countries.
For instance, depression measure in the HRS utilized the Center for Epidemiologic Studies
(CES-D) scale while the LPOPS used the Korean version of the Geriatric Depression Scale
(SGDS-K). To address the differences in the range of these scales, the present study used the
standardized scores and found the similar effect of internet use in reducing depressive symptoms
regarding geographic distance from adult children. Third, the present study did not examine
59

different types of family contexts and how this can influence the outcome. In the HRS, more
detailed measures on step families and a history of marriage were available, which was not the
case for the LPOPS survey due to its cross-sectional design. The result on the mediating role of
social contacts and other intergenerational factors may appear different depending on marital
history and other family contexts, and future research can examine whether these factors lead to
variability in the effect of internet use. Finally, future studies can also expand a pool of countries
as the present findings are limited to examination of two countries.
Despite these limitations, the overall findings highlight the importance of
intergenerational factors in the effect of using the internet on depressive symptoms among older
men and women in the U.S. and Korea. The current study indicates that social contacts have a
mediating role in this association in both countries, providing a medium to communicate across
vast distance and different time zones. As global migration and advance in technology increases,
this finding may have important implications for promoting intergenerational relationship and
exchange of support in the future.

60

Chapter Four: Internet Use and Life Transitions in Marital and Work Status

4.1. Introduction
The significance of psychological distress, such as depression and loneliness, in later
adulthood have emerged as a key challenge in the recent years. A global estimate suggests that
the U.S. ranks third among countries in terms of the burden of disease for mental disorders
(WHO, 2018). More than 6.5 million Americans aged 65 and over are affected by depression,
with the prevalence expected to increase more among those aged 65 and over compared to other
age groups by 2050 (National Alliance on Mental Illness, 2009; Heo et al., 2008). Additionally,
one in three Americans aged 45 and over reported feeling lonely and the prevalence increased to
seven in ten when assessing lifetime exposures to feeling lonely (AARP, 2018
a
). Psychological
distress in later life can have serious consequences. Late life depression has been linked to
suicides, and it is reported to be the cause for over 80% of late life suicides (Bickford, Morin,
Nelson, & Mackin, 2019; Van Orden & Conwell, 2011). Poor psychological functioning has also
been associated with other adverse health outcomes (Gleason, Pierce, Walker & Warnock, 2013;
Heser et al., 2018), and thus has important implications for public health.
Psychological distress can be triggered by a number of factors. Significant life transitions
in old age including a loss of a spouse, employment, and other life events have been found to
affect people’s social network and lead to higher depression (Wrzus, Hänel, Wagner & Neyer,
2013). In particular, widowhood is recognized as one of the most stressful life events that can
cause people to withdraw from social contacts (Ferraro & Barresi, 1982). Divorce rates have also
been increasing rapidly among those aged 50 and over (Stepler, 2017). Additionally, work
transitions such as retirement have been identified as a mutually reinforcing factor of major
61

depressive symptoms (Calvo, Sarkisian & Tamborini, 2012; Doshi, Cen & Polsky, 2008; Segel-
Karpas, Ayalon & Lachman, 2018). The overarching evidence suggests that transitions in marital
and work status in later life can have significant effects on mental health and psychological well-
being, and warrant more empirical attention in identifying an effective strategy to reduce risks of
psychological distress caused by these events.
Maintaining good psychological health provides a key resource to adapting to different
transitions in later life. Internet use has been found in prior studies to reduce depression and
enhance psychological functioning of older people, but evaluation of this effect on people
undergoing different types of life transitions has not been tested. Examination of whether internet
use has any effects on people who experience significant life events on psychological distress
can provide useful information on clarifying the effect of internet use on enhancing social
connection of older people, a call that has been increasingly initiated and implemented (AARP,
2018
b
; Elder & Retrum, 2012; O’Rourke, Collins & Sidani, 2018). People who experience
transitions in marital and work status at older ages are more likely to be exposed to increased
risks of depressive symptoms and loneliness, and the examination of whether the internet can be
used to strengthen their social support going through these transitions can enhance our
understanding on the effect of internet use on social connectedness of older people. Thus, the
current study examines how internet use is associated with change in depressive symptoms and
loneliness for people who experience transitions in marital and work status using the Health and
Retirement Study over the 4-year period.

62

4.1.1 Conceptual Background
I draw upon theories on depression and life course development as the main theoretical
framework for the present study. Beck’s depression model posits that depression occurs from
interactions between predisposing factors, such as one’s beliefs and attitudes about oneself, and
precipitating factors, referring to stressful life situations that activate these predispositions (Beck,
1967). This model highlights the role of life situations as the main agent triggering self-blame
and negative expectations about life, which in turn cause depressive symptoms. The later
evaluation of the model, along with other theoretical models on depression, also calls for greater
attention on the examination of the role of stressful life events in assessing one’s vulnerability to
depression (Robins & Block, 1989). Based on this theoretical framework, empirical examination
of late-life depression calls for more attention to the role of external life events such as
widowhood and unemployment.
Additionally, life course development theories on individuals’ social networks and
relationships demonstrate how risks of social isolation can be amplified in undergoing significant
life transitions. Socioemotional selectivity theory suggests that people select their close network
of people to invest in more meaningful relationships with a limited time horizon, finding that the
frequency of social interactions decreased with age, but emotional closeness with social partners,
such as spouses and siblings, steadily increased with age (Carstensen, Fung, & Charles, 2003).
The absence of a spouse from a divorce or widowhood may thus have a more profound effect on
people in late adulthood than young adulthood. Another theory that focuses on how social
relationship changes with the life course perspective is the convoy model (Kahn & Antonucci,
1980). This perspective suggests that the social roles (e.g., in family and work) create convoy
membership through which people gain and give support, ultimately affecting their well-being
63

and health. When the individual loses employment through a layoff, the size of the convoy is
likely to shrink. Retirement, on the other hand, can lead to positive effects including increases in
life satisfaction (Hansson et al., 2017), despite a similar reduction in the convoy membership.
The methods of how people connect with others became more diversified with advances
in communication technology in the 21
st
century. Telephones have now been replaced with
mobile phones that also provide access to the internet and social media throughout the day. Most
prior studies reported positive associations between internet use and social connectedness among
older adults (Shapira, Barak & Gal, 2007). Chopik (2016) constructed a social technology use
measure which included older people’s use of e-mail, social networking sites, smartphones,
online video/phone calls and instant messaging, and found that social technology use was
significantly associated with fewer depressive symptoms and higher subjective well-being
through mediation by reduced loneliness. Heo and his colleagues also found that using the
internet was positively related to social support, which in turn contributed to reduced loneliness
and higher psychological well-being (Heo, J., Chun, Lee, Lee & Kim, 2015). On the other hand,
Cotten and colleagues (2013) found that the frequency of using the internet was not associated
with lower levels of perceived social isolation, indicating the possibility that perceived social
isolation may be more related to face-to-face contact than online contact. However, these studies
were mostly cross-sectional and did not examine multiple dimensions of social relationships in
examining the effect of internet use. If using the Internet leads to stronger social support, it may
be possible that the internet use offers protective effects to attenuating risks of depression and
isolation among people who experience significant transitions in marital and work status at older
ages. Using the internet may provide a means to compensate for the loss of offline relationships
with the online connection or expand an individual’s social network to find people with similar
64

experiences and share knowledge about coping mechanisms. The current study thus examines
whether the internet use buffers the effect of significant transitions in marital and work status on
depression and loneliness accounting for both qualitative and quantitative aspects of social
relationships of older adults using data from the Health and Retirement Study.

4.1.2 Research Hypotheses
The present study examines how internet use affects the relationship between life
transitions and psychological functioning in later adulthood by using nationally- representative,
longitudinal data from the Health and Retirement Study. Because most prior studies found
internet use to be associated with decreased depression and loneliness (Chopik, 2016; Cotten,
Ford, Ford & Hale, 2014), I hypothesize that marital transitions, measured by a divorce and
widowhood, will have less depressive effects among internet users compared to non-users. I also
hypothesize that internet use will reduce the adverse influence of marital transitions on loneliness
(Hypothesis 1). I also expect a similar association to be observed between internet use and the
relationship between work transitions and these psychological outcomes (Hypothesis 2).

4.2. Methods
4.2.1. Data
The current study used two interviews (2010/2012 and 2014/2016) of respondents to the
Health and Retirement Study (HRS), a nationally representative sample of Americans aged 50
and over. The HRS began in 1992 with a multi-stage area probability design that included
oversampling of Black and Hispanic older people and collects information from face-to-face
65

interviews and telephone surveys. The HRS is administered every two years (Sonnega et al.,
2014).
Since 2006, the HRS has been collecting psychosocial information of the participants
through the “Leave-Behind” (LB) questionnaire, which is randomly distributed to a half sub-
sample off those participating in the main survey at each wave. The response rates ranged from
73% to 88%, and more details for each psychosocial survey year are specified in the HRS
guidelines (Smith, Ryan, Fisher, Sonnega & Weir, 2017). The psychosocial measures were asked
more consistently after 2010, and thus the 2010/2012 and 2014/2016 intervals were chosen for
the current study. Transitions were coded as occurring in two intervals to match the outcome
variables.
In order to focus on the effect of transitions, I defined different samples for analysis of
marital and work status. For marital status, I restricted the baseline sample to only sample
members who were married at the baseline and excluded those who were never married or
divorced or widowed. In 2010/2012, a total of 10,123 respondents who were married in 2010
completed the LB questionnaire and a total of 8,429 respondents completed and returned the
questionnaire in 2014/2016. A total of 6,763 respondents had the complete LB questionnaires for
both time points, and those who had missing data on loneliness (n=1,094), depression (n=89),
and perceived support from children and other family members (n=257) were excluded. The final
analytic sample was 5,323 participants for marital transitions. For work transition, I focused the
baseline sample on those who were working and excluded those who were unemployed or retired
at baseline, which resulted in more loss of participants because a majority of participants were
retired (61% of those aged over 50). A total of 4,865 respondents who were working in 2010
returned the LB questionnaire for 2010/2012, and a total of 4,480 respondents completed the LB
66

questionnaire in 2014/2016. The respondents who had the complete LB questionnaires at both
time points were 3,629, and those who had missing data on loneliness (n=760) and other social
relationship characteristics (n=371) were excluded. The final analytic sample was 2,498
participants for work transitions.

4.2.2. Measures
Psychological Functioning
Depression was assessed with an 8-item version of the Center for Epidemiologic Studies
(CES-D) scale, a short version of the 20-item CES-D scale (Radloff, 1977). A summary score of
yes/no responses to 8 questions was used as a depression measure, with a range of 0 to 8. This
measure has been examined as an indicator of depressive symptoms in older Americans in other
HRS studies (Recksiedler & Stawski, 2019).
Loneliness was measured by the UCLA loneliness scale (Hughes, Waite, Hawkley &
Cacioppo, 2004) that was included in the LB questionnaire. The scale included 3 items which
asked participants how much of the time they felt (1) lacked companionship, (2) left out, (3)
isolated from others, with a three-point scale response ranging from 1 (hardly or never) to 3
(often). All responses to items were reverse coded and averaged to create a loneliness scale with
a resulting Cronbach’s alpha which matched the scale reliability indicated in the HRS guideline
for each survey year (Smith, Ryan, Fisher, Sonnega & Weir, 2017).

67

Use of the Internet
The HRS had a question that asked “Do you regularly use the Internet (or the World
Wide Web) for sending and receiving e-mail or for any other purpose, such as making purchases,
searching for information, or making travel reservations?,” with a binary response (No=0; Yes=1)
and the response at the baseline was included in the model.

Life Transitions
Two life transitions were selected as potential risk factors for depression and loneliness:
Changes in marital status and work status. For marital status, participants were asked in each
wave to indicate their marital status (i.e., married, divorced, separated, never married, widowed).
Divorced and separated were combined into a single category based on distribution and practical
infeasibility with completely isolating the effect of one from the other across the waves. Changes
in marital status were categorized as consistently married, became separated or divorced, and
became widowed based on their changes during the two intervals. Individuals who were never
married, who were consistently divorced or widowed, and who were re-married were all
excluded in the study. Employment status was classified as working (working part-time, full-
time), retired, and unemployed (unemployed, not in the labor force, disabled) in each wave.
Change in work status was categorized into consistently working, became unemployed, and
became retired. Like marital transitions, the participants who were constantly unemployed or
retired, and who became employed were excluded from the final sample.

68

Covariates
Demographic characteristics at baseline included age, gender, race/ethnicity (White
[reference], Black, Hispanic, and other race), education (less than high school [reference], high
school, some or more than college), and total household wealth (low [reference], middle, high).
Health measures at baseline included self-reported health (1 poor to 5 excellent), a count of
chronic conditions (hypertension, cancer, diabetes, heart disease, stroke, arthritis, lung disease)
and whether the participant had any limitations in activities of daily living (walking across a
room, eating, dressing, bathing, toileting, and getting into bed). Social support measures included
a number of indicators to assess multiple components of social relationships. Social network size
was constructed as a quantitative indicator with a summary index of the respondents’
composition of social networks (0 to 4). Perceived social support was included as a qualitative
indicator with an average score of responses to three questions on measuring the positive
perceived support in three relationships (i.e., children, other family, friends), and the average
scores on perceived support from each relationship were included separately (0 to 3).

4.2.3. Analytic Strategy
Descriptive characteristics were examined first to indicate sample characteristics in
2010/2012. The negative binomial regressions were estimated to examine the differences in the
effect of internet use and each transition on depressive symptoms and loneliness accounting for
the initial level of each outcome. This approach allows the examination of between-wave
residual change in depressive symptoms and loneliness from baseline to follow-up.

69

The equation was as below:
𝑌 𝑖 2
= 𝛽 0
+ 𝛽 1
𝑋 𝑖 1
+ 𝛽 2
𝑋 𝑖 2
… 𝛽 𝑛 𝑋 𝑖𝑛
+ 𝛽 𝑛 𝑌 𝑖 1
+ 𝜀 𝑖

Where the outcome of interest for individual i are indicated as 𝑌 𝑖 1
at wave 1 and
𝑌 𝑖 2
at wave 2, 𝑋 𝑖𝑛
is the values of independent variables measured at waves 1, 𝛽 0
is a constant,
and 𝜀 𝑖
refers to the error term.
First, the model with baseline internet use was estimated to examine the main effect of
internet use on depressive symptoms and loneliness at follow-up. Second, we added each
transition variable, and demographic and socioeconomic controls. The third model included
health and social support characteristics to examine how internet use and significant life
transitions in marital and work status were independently associated with depression and
loneliness at follow-up, adjusting for all baseline covariates in demographic, socioeconomic,
health, and social support characteristics. Finally, we examined whether the effect of significant
life transitions differed by internet use within each category of martial and work transitions by
inclusion of an interaction term. Sampling weights were used in all analyses to account for
differential probability of sampling and complex survey design of the HRS. Statistical analyses
were conducted in STATA 14.1 (StataCorp, TX).

4.3. Results
Table 4.1. presents the weighted baseline sample characteristics stratified by marital
transition status. From 2010/2012 to 2014/2016, about 91% (n=4,855) remained married, while
2% (n=119) became divorced and 7% (n=349) became widowed. The average age at baseline
70

was 63.01 years (SD= 8.60, range 50-92), and approximately a half of participants were women
(52%). A majority of the sample was non-Hispanic White (87%) and had more than college
education (60%). The participants who became widowed were more likely to be older and
women, non-Hispanic White, and had lower levels of education than those who became
separated or divorced or remained married. Overall, 57% of the sample used the internet, and the
usage was highest among the participants who remained married and lowest among the
participants who became widowed. The participants who became separated or divorced had the
lowest self-reported health, but the participants who became widowed had the highest number of
chronic conditions and the highest proportion of people reporting any limitations with activities
of daily living. For social relationship, the participants who became separated or divorced had
the smallest size social network (M=3.62, [SD=0.60]) compared to those who remained married
(M=3.89, [SD=0.39]) or became widowed (M=3.72, [SD=0.55]). The perceived support from
children and friends were also lowest among the participants who became separated or divorced,
although this difference was not statistically significant.

Table 4.1. Baseline Sample Characteristics Among Those Who Remained Married and
Experienced Marital Transitions (N=5,323) (2010/2012)

Total
(N=5,323)
Married
(n=4,855)
Separated/
Divorced
(n=119)
Widowed
(n=349)

Variables
Mean (SD)/
(%)
Mean (SD)/
(%)
Mean (SD)/
(%)
Mean (SD)/
(%) P
Age 63.01 (8.60) 62.77 (8.35) 57.30 (7.15) 70.38 (9.66) <.001
Female 52.34 51.19 48.93 76.03 <.001
Race/ethnicity

White 86.68 86.70 80.98 89.06 <.05
Black 4.93 4.76 9.64 5.99

Hispanic 5.62 5.72 4.07 4.47

Others 2.77 2.82 5.31 0.48

Education

71

< High School 9.52 9.03 10.17 18.51 <.001
High School 31.05 30.79 27.83 37.54

>=College 59.43 60.18 62.00 43.95

Wealth

Low 14.84 14.24 29.45 19.41 <.001
Middle 35.44 35.32 40.22 35.41

High 49.72 50.44 30.33 45.18

Use Internet 57.16 66.33 63.80 42.70 <.001
Self-reported health
(1: Poor - 5: Excellent)
3.53 (0.97) 3.55 (0.96) 3.27 (1.20) 3.34 (0.98) <.001
Count of chronic
conditions
1.54 (1.24) 1.53 (1.24) 1.22 (1.22) 2.03 (1.20) <.001
Have ADL limitations 8.95 8.58 12.22 14.41 <.001
Social network size
(0-4)
3.88 (0.41) 3.89 (0.39) 3.62 (0.60) 3.72 (0.55) <.001
Perceived support
from children (0-3)
2.20 (0.70) 2.19 (0.70) 1.99 (0.81) 2.33 (0.67) 0.09
Perceived support
from family (0-3)
1.85 (0.84) 1.84 (0.85) 1.86 (0.84) 1.82 (0.88) 0.76
Perceived support
from friends (0-3)
2.04 (0.73) 2.04 (0.72) 1.99 (0.76) 2.12 (0.76) 0.26
Note. Numbers represent weighted means, with standard errors in parentheses, and weighted percentages.
* p < 0.05, ** p < 0.01, *** p < 0.001
For work transitions, the average age of participants at baseline was younger with 57.93
(SD= 5.51, range 50-92) years as shown in Table 4.2. More variation was found for the work
transitions, as approximately 64% (n=1,595) of the participants who were working at baseline
remained employed, while 6% became unemployed (n=140) and 30% (n=763) became retired.
About 84% of the sample was non-Hispanic White and 68% had college education or above. The
participants who retired were more likely to be older and non-Hispanic White, while the
participants who became unemployed were more likely to be women and have lower levels of
education. The participants who became retired had poorer self-reported health, more number of
chronic conditions and ADL limitations than the participants in other categories of work
transitions. On the other hand, the participants who became unemployed had the smallest size of
social network (M=3.59, [SD=0.59]), compared to those who remained employed (M=3.71,
72

[SD=0.50]) or became retired (M=3.68, [SD=0.55]), although this difference was not statistically
significant. There was also no statistically significant difference in terms of perceived support
from children, family, and friends by work transition status.
Table 4.2. Baseline Sample Characteristics for the Employed and Those with Work Transitions
(N=2,498) (2010/2012)

Total
(N=2,498)
Employed
(n=1,595)
Unemployed
(n=140)
Retired
(n=763)
Variables
Mean (SD)/
(%)
Mean (SD)/
(%)
Mean (SD)/
(%)
Mean (SD)/
(%) P
Age 57.93 (5.51) 56.61(4.80) 56.31 (4.85) 61.45 (5.67) <.001
Female 54.55 53.56 60.21 55.89 0.40
Race/ethnicity

White 84.17 84.21 81.48 84.56 0.47
Black 7.05 6.65 6.41 8.15

Hispanic 5.89 6.14 7.92 4.90

Others 2.89 3.00 4.19 2.39

Education

< High School 5.95 5.41 9.38 6.67 <.05
High School 26.96 26.04 19.84 30.50

>=College 67.09 68.55 70.78 62.83

Wealth

Low 21.93 22.62 28.28 19.09 0.13
Middle 37.30 38.10 33.93 35.95

High 40.77 39.28 37.79 44.96

Use Internet 73.62 76.36 68.57 68.81 <.01
Self-reported health
(1: Poor - 5:
Excellent)
3.70 (0.92) 3.73 (0.90) 3.73 (1.09) 3.62 (0.94)
0.08
Count of chronic
conditions
1.12 (1.07) 1.00 (0.98) 0.84 (1.04) 1.46 (1.20)
<.001
Have ADL limitations 3.96 2.32 2.95 6.72 <.001
Social network size
(0-4)
3.70 (0.52) 3.71 (0.50) 3.59 (0.59) 3.68 (0.55)
0.16
Perceived support
from children (0-3)
2.14 (0.73) 2.14 (0.71) 2.10 (0.87) 2.12 (0.75)
0.58
Perceived support
from family (0-3)
1.90 (0.83) 1.90 (0.82) 2.10 (0.72) 1.88 (0.85)
0.80
Perceived support
from friends (0-3)
2.10 (0.73) 2.12 (0.72) 2.14 (0.69) 2.05 (0.74) 0.10
Numbers represent weighted means, with standard errors in parentheses, and weighted percentages.
* p < 0.05, ** p < 0.01, *** p < 0.001

73

Table 4.3. shows mean values of depressive symptoms and loneliness at baseline in
2010/2012 and follow-up in 2014/2016 by marital and work transition status. The levels of each
outcome measure is shown at baseline for each marital and work transition status group and then
at follow-up after the participants experienced transitions. A graded increase in both depressive
symptoms and loneliness was observed by marital transition status, with the participants
becoming separated or divorced showing the highest values followed by people who became
widowed, and people who remained married showing the lowest values. The mean levels of
depressive symptoms and loneliness in all marital transition status increased slightly from
2010/2012 to 2014/2016. The differences in the mean levels of depressive symptoms and
loneliness by each martial and work transition status were statistically significant at both time
points.
On the other hand, the mean levels of depressive symptoms by work status increased over
time, whereas the mean levels of loneliness did not show much change. The participants who
become unemployed exhibited the highest values of depressive symptoms and loneliness
followed by people who became retired, and the participants who remained working showed the
lowest values. Although the mean levels of depressive symptoms were significantly different by
work transition status, the mean levels of loneliness did not show statistically significant
difference at either time points.

74

Table 4.3. Mean levels of outcomes by marital and work transitions in 2010/2012 and 2014/2016
Mean Depression Scores (0 - 8)

2010/2012

2014/2016

M SD P M SD P
Martial Transitions

Remained married 0.84 (1.48) <.001 0.93 (1.57) <.001
Became sep/divorced 1.78 (2.09)

2.08 (2.45)

Became widowed 1.24 (1.75)

1.98 (2.23)

Mean Loneliness Scores (1 - 3)

2010/2012 2014/2016

M SD P M SD P
Martial Transitions

Remained married 1.34 (0.46) <.001 1.37 (0.48) <.001
Became sep/divorced 1.74 (0.51)

1.78 (0.60)

Became widowed 1.47 (0.53) 1.64 (0.52)
Mean Depression Scores (0 - 8)

2010/2012

2014/2016

M SD P M SD P
Work Transitions

Remained working 0.75 (1.37) <.05 0.79 (1.43) <.001
Became unemployed 1.28 (1.82)

1.55 (2.07)

Became retired 0.89 (1.61) 1.07 (1.75)

Mean Loneliness Scores (1 - 3)

2010/2012

2014/2016

M SD P M SD P
Work Transitions

Remained working 1.41 (0.50) 0.86 1.41 (0.50) 0.83
Became unemployed 1.58 (0.57)

1.59 (0.54)

Became retired 1.41 (0.51) 1.41 (0.51)
+
p < 0.10,
*
p < 0.05,
**
p < 0.01,
***
p < 0.001

The negative binomial regression models assessing the effect of marital transitions and
internet use predicting change in the levels of depressive symptoms are presented in Table 4.4.
The effect of baseline internet use was significantly associated with decreased depressive
symptoms after 4-year periods (β= -0.445; SE = 0.053, p < .001) (Model 1). With inclusions of
demographic and socioeconomic factors in Model 2, the baseline internet use remained
significantly associated with reduced depressive symptoms (β= -0.177; SE = 0.057, p < .01). For
marital transition status, both becoming separated or divorced and becoming widowed were
75

associated with increased depressive symptoms, although the magnitude of effect was larger for
widowhood (Model 2 to Model 4). When the interaction between internet use and marital
transitions was included in the Model 4, using the internet and becoming separated or divorced
was associated with reduced depressive symptoms for people who used the internet at baseline
(β= -0.710; SE = 0.329, p < .05). There was no significant interactive effect for becoming
widowed.
Table 4.5 shows the models of internet use and marital transitions for loneliness. The
effect of internet use was significant for loneliness after the 4-year period (β= -0.035; SE =
0.013, p < .01), although the magnitude of the effect was smaller than depressive symptoms
(Model 1). Using the internet at baseline remained significantly associated with lower loneliness
(β= -0.029; SE = 0.012, p < .05) when adjusting for demographic and socioeconomic
characteristics (Model 2). The association became attenuated to marginal significance in Model 3
with inclusion of health and social network characteristics as controls, and the magnitude of the
effect of internet use decreased by 24%. In Model 4 when the interaction between marital
transitions and internet use was added, there was a significant interaction for using the internet
for people who became separated or divorced (β= -0.198; SE = 0.083, p <.05), but not for
widowhood.

76

77

78

Figure 1 depicts the predicted levels of depressive symptoms from Model 4 in Table 4 and
loneliness from Model 4 in Table 5. The predicted level of depressive symptoms was highest among the
participants who became separated and divorced and did not use the internet and lowest the participants
people who were married and used the internet. The gap by internet use for the predicted levels of
depressive symptoms was wider among the participants who became separated or divorced compared to
the participants in other categories. The similar effect is found for loneliness, where the participants who
exhibited the highest predicted level of loneliness were the participants who experienced separation or a
divorce, but the predicted level was lower when they used the internet.

79

Figure 4.1. Predicted Levels of Depressive Symptoms and Loneliness by Internet Use and Marital
Transitions

80

Turning to work transitions, internet use showed much more attenuated effects compared to the
results for marital transitions. For depressive symptoms shown in Table 4.6, the main effect of baseline
internet use was only significant without any covariates (β= -0.424; SE = 0.089, p <.001) (Model 1). The
significance of the effect disappeared with inclusion of demographic and socioeconomic characteristics
(Model 2). Both becoming unemployed and retired were significantly associated with increased
depressive symptoms in Model 2 and Model 3. When the interaction between work transition status and
internet use was added in Model 4, there was a marginally significant interactive effects for the
participants who became unemployed (β= +0.525; SE = 0.311, p =.091), but no significant effect was
found for the participants who became retired.
For loneliness presented in Table 4.7, the main effect of internet use was not significant for
loneliness after 4-year periods (β= -0.033; SE = 0.022, p =.127) (Model 1). Both becoming unemployed
and becoming retired did not show any significant associations with change in loneliness. An inclusion of
interaction term in Model 4 showed the similar association between becoming unemployed and internet
use observed in the models for depressive symptoms. The baseline internet use showed marginally
significant associations for the participants who became unemployed (β= 0.125; SE = 0.074, p =.092). On
the other hand, the perceived support from friends (β= -0.025; SE = 0.011, p <.05) was significantly
associated with reduced loneliness and the perceived support from children also showed marginally
significant associations (β= -0.018; SE = 0.011, p =.088).

81

82

83

4.4. Discussion
The present study aimed to estimate the effect of internet use for people aged 50 and over
who experienced marital and employment transitions on their depression and loneliness over four
years. While using the internet was linked to reduced depression and other psychological
outcomes in prior studies (Chopik, 2016; Cotten, Ford, Ford & Hale, 2014; Elliot, Mooney,
Douthit & Lynch, 2013; Sum, Mathews, Hughes & Campbell, 2008), this is the first study to test
whether and how exposures to marital and work transitions add variability to this association.
The results on marital transitions showed that using the internet buffered the negative effect of
becoming separated or divorced on depressive symptoms. However, no significant effect was
found for widowhood. For work transitions, internet use did not offer protective effects on
depressive symptoms or loneliness.
The results on marital transitions reveal heterogeneity in the effect of internet use on
different categories of transitions on depression and loneliness. The result from the present study
indicates that internet use may buffer the effect of a divorce on depression for older adults, but
not widowhood. For loneliness, the similar effect was found for people who became separated or
divorced, although the magnitude of effect was much smaller. Internet use can offer a medium
for seeking alternative partners through online dating (Davis & Fingerman, 2015) or to find a
social network with whom they can share their feelings which may not be understood even by
their family for people who became divorced (Wrzus, Hänel, Wagner & Neyer, 2013). This may
help reduce feelings of depression or loneliness among people who experience transitions in
marital status at older ages. In the current study, no effect of the internet use was found for
widowhood on depressive symptoms or loneliness. Widowhood has been found to have the most
poignant effect on people among all marital transitions (Holmes & Rahe, 1967; Kitson, Babri,
84

Roach & Placidi, 1989). Prior studies on marital disruption and widowhood reported that people
who experienced a divorce expressed more anger or avoidance while those who experienced
widowhood reported more grief (Farnsworth, Pett & Lund, 1989). Thus, for people who became
widowed, their needs may be more related to face-to-face interactions or social engagement to
address their feelings of depression than internet use (Isherwood, King & Luszcz, 2012; Min, Li,
Xu & Chi, 2018; Utz, Carr, Nesse & Wortman, 2002). However, further research is necessary to
clarify the association between internet use and marital transitions on depressive symptoms and
loneliness in the present study.
In this study, the effect of internet use on reducing depressive symptoms and loneliness
either for people who became unemployed or people who became retired did not reach statistical
significance. Previous examination of the effect of internet use on depression and loneliness
treated older adults as one homogeneous group (Chopik, 2016; Elliot, Mooney, Douthit &
Lynch, 2013; Sum, Mathews, Hughes & Campbell, 2008), and called for more investigation of
the effect of using the internet for people with different aging experiences. The findings indicate
using the internet may not confer protective effects for adverse psychological distress following
unemployment. Further studies are necessary to validate the findings in the present paper. The
reasons for the marginal effect of using the internet for work transitions may be interpreted in the
context of both micro- and macro-level factors influencing depression and loneliness caused by
work transitions. At the individual level, older people with abilities to use the internet tend to be
more highly educated (Hong & Cho, 2016; Yu, Ellison, McCammon & Langa, 2016), and to
have had professions with more stability and less unemployment risks (Mincer, 1991).
Therefore, their unemployed status may be more likely to be prompted by circumstances that
limit their abilities to work (Chu et al., 2016), and being unemployed may cause them to feel
85

isolated from their peers who are similarly educated and still in the labor force. At the macro-
level, the dynamic changes of economic situations from the boom in the late 1990s to the
recession in 2008 also made retirement in later life more of an involuntary withdrawal among the
younger cohorts approaching retirement (Cahill, Giandrea & Quinn, 2013). In this regard, using
the internet may not offer an effective means of reducing risks of depression and other adverse
psychological outcomes associated with a loss of employment.
Several limitations need to be noted. First, the long-term effect of these transitions and
internet use are not estimated. Previous literature suggests differences in the extent of
psychological distress and social behaviors between people who recently became widowed and
those who had been widowed for longer periods (Ferraro, Mutran & Barresi, 1984; Isherwood,
King & Luszcz, 2012). For instance, people who became recently widowed have more acute
slopes of depressive symptoms, but this increase may decrease over time. By the same token, the
patterns of social interaction between people who became recently widowed and people who
have been widowed for longer time spans may be different, all of which can differently be
related to the effect of internet use. Additionally, internet use was only assessed with the general
usage question than the usage question assessing specific purposes. The current study used the
analytic strategy to account for the effect of timing of transition and internet use, but is limited in
scope for two time points. Future research can examine how duration of each marital and work
transition is associated with more detailed measures of internet use. Additionally, the
mechanisms behind reduced depression and increased loneliness are not estimated in the present
study. The use of mediated analyses may provide more information on how internet use is related
to marital and work transitions to affect different psychological outcomes accounting for various
measures of social support and social relationships.
86

Despite these limitations, the overall findings demonstrate the complexity of the effect of
internet use on different life transitions for older adults’ depression and loneliness. The present
study examined the effect of internet use on psychological distress of people in the context of
marital and work transitions accounting for both quantitative and qualitative indicators of social
connectedness. The results highlight the importance of different contextual factors in examining
the effect of internet use for older adults on their social relationships. Given the heterogeneity of
aging experience, the effect of internet use is likely to differ for individuals with different
contexts, and it is crucial to identify the effective intervention strategy accounting for these
differences in reducing risks of depression or loneliness. Without such effort, internet
interventions may underestimate the complementary needs of human contacts in enhancing
psychological well-being for helping individuals cope with significant life transitions in marital
and work status at older ages.

87

Chapter Five: Discussion

5.1. Discussion of Findings
The present dissertation sought to understand how individual’s environmental, cultural,
and situational factors contribute to differences in the effect of internet communication use on
psychological well-being and social connectedness of older people. Across three chapters, the
role of the internet as a main source of social connection was examined in various contexts,
including physical environments, sociohistorical contexts, and significant life transitions. By
investigating these different contexts, the present dissertation aimed to highlight heterogeneity of
aging processes in examination of the effect of internet use, which has received less attention in
exploring the potential of internet use for older population in existing research. The findings
from each chapter inform us of how differences in environmental, sociohistorical, and individual
contexts contribute to variability in the effect of internet use. The present chapter summarizes the
main findings in each chapter and discusses their limitations and implications for future research.
The first chapter focused on how electronic communication use is associated with older
adults’ psychological well-being and depressive symptoms by types of living arrangements using
the person-environment perspective. Using electronic communication was associated with higher
psychological well-being and lower depressive symptoms, but the effect did not differ
significantly for people living alone compared to people living with spouses and others.
The second chapter compared the association between internet use and depressive
symptoms in Korea and the U.S. within the context of intergenerational relationship and transfers
of financial support. The results showed that using the internet was associated with lower
depressive symptoms for people living far from their closest child, and the association differed
88

by gender with the effect being significant for women in the U.S. and men in Korea.
The third chapter tested how significant life transitions characterized by changes in
marital and work status are associated with depressive symptoms and loneliness, and whether the
internet use differently affected this association using life-course development theories on social
relationship and depression models. For marital status, using the internet was associated with
lower depressive symptoms and loneliness for people who became separated or divorced. For
work status, the overall effect of internet use was not significant in reducing depressive
symptoms or loneliness for work transitions.
The overall findings contribute to a growing literature on technology and aging by
demonstrating that the effect of internet use is heterogenous depending on the contextual factors
in which an individual is situated. Research on internet use and older people has been gaining
attention in the recent years, yet few studies explored whether and how the effect varies by
differences in environmental, sociohistorical, and individual contexts. The present dissertation
calls for a more multi-dimensional and theoretically-based approach in examining the values and
effectiveness of internet use in enhancing psychological well-being and addressing social
isolation of older people.

5.2. Limitations and Strengths
Three papers also have several limitations that need to be noted for the interpretation of
findings. For the first chapter, one of the notable limitations is sample attrition across different
waves. Sample attrition across waves requires a particular attention for researchers studying the
National Health and Aging Trend Study, as the participants lost to attrition tend to show
significantly different sociodemographic profiles than the participants with complete data
89

(Xiang, Chen & Kim, 2019). To address this concern, sensitivity analyses were performed to
examine whether the effect of electronic communication use differed in a shorter time frame, and
the results only showed minor differences for psychological well-being and were consistent for
depressive symptoms. However, the use of advanced statistical techniques such as multiple
imputation could have provided a more holistic picture of the observed effect of electronic
communication use by living arrangements.
The second chapter is limited to correlational examination as the survey in Korea is
cross-sectional. The Living Profiles of Older People’s Survey in Korea is a nationally
representative survey administered to Koreans aged 65 and over via a collaboration between
Ministry of Health and Social Welfare and the Korea Institute for Health and Social Affairs. Yet,
the survey is cross-sectional despite having a wide range of data on demographic, social,
environmental, health, and internet use. A sister study of the Health and Retirement Study, the
Korea Longitudinal Study of Aging, does not assess the use of internet from survey participants
and only has one question asking whether the respondents used the internet when searching for a
job. Considering technological advancement in the country and the Korea Longitudinal Study of
Aging being one of the few publicly available longitudinal surveys of older Koreans (Yoon,
2013), incorporating the question on internet use seems particularly important for future research
in the country.
The limitation of the final chapter is related to a small sample size for each marital
transition and work transition category. The number of participants who became separated or
divorced and who became unemployed was particularly smaller compared to other groups. Due
to this reason, other studies examining the marital and work transitions excluded the respondents
who became separated or divorced or who became unemployed, and only compared married to
90

widowed and working to retired as time-varying predictors (Lee, Chi & Ailshire, 2018).
However, these groups were included in the chapter as the research question concerned
examining the effect of internet use on all categories of marital and work transitions compared to
not having any transition and how this is related to changes in depressive symptoms and
loneliness over time.
Despite these limitations, the current dissertation also has several strengths. First, the
present study has a longitudinal examination of two nationally representative studies in the U.S.
By doing so, the current study contributes to expanding the knowledge on the effect of internet
communication use among older Americans. Second, the dissertation extends the applicability of
the effect of using the internet to the non-Western context by doing a cross-country comparison
with Korea. Third, the overall findings indicate the complexity of the effect of internet use
among older adults with different environmental and individual contexts. Although a recent
study begins to raise a question on the effectiveness of internet use on mental health of older
people (Hees, Tesch-Römer & Huxhold, 2019), a prevalent view depicts the effect as mostly
positive (Chopik, 2016; Cotten, Ford, Ford & Hale, 2014; Elliot, Mooney, Douthit & Lynch,
2013) and does not fully reveal the complexity of the effect of internet use. With increasing
availability of different devices and their widespread use, the role of the internet in the lives of
older adults is likely to expand, and more empirical attention will be called for to promote
development of programs and products to make online communication more age-friendly and
convenient for older people to use and to help them age in place.
The present dissertation has a number of limitations but does offer the initial evidence of
presenting a complex effect of internet communication use among older population. More
research from both national surveys and experimental studies will better clarify the role of
91

internet use in maintaining their social network and support as well as improving the quality of
lives for older adults.

5.3. Implications and Future Research
The findings from the current dissertation provide several implications in regards to
theory and future research. From the theoretical perspective, a set of three papers highlight the
importance for development of theoretical framework in the field of aging and technology to
describe complex dynamics between older people and the internet. Due to a lack of a
comprehensive theoretical framework linking internet use to people in terms of their physical
environment and individual characteristics, the current dissertation derived the main hypotheses
using available sociological and gerontological theories, using a person-environmental theory,
socioemotional selectivity theory, convoy theory, and theory of depression. While these models
were well-validated, development of a theoretical model that encompasses an ecological,
sociological, and environmental aspects of internet use will be necessary to empirically test
various implications of internet use among older population, and how it affects their behaviors,
physical health, and psychological well-being.
For research and practice, the present dissertation indicates variability in the effect of
internet use among heterogenous groups of older adults depending on their unique
environmental, cultural, and individual contexts. Empirical attention to the effect of internet use
among older people is expected to exponentially increase on various topics including health care,
caregiving, social isolation, psychological well-being, and cognitive functioning in the
foreseeable future. The role of a researcher for engaging in technology-related research therefore
has become increasingly important. Additionally, an incorporation of a detailed internet use
92

measure in nationally representative surveys will promote more research on internet use among
older population to identify effective intervention strategies and programs to meet the growing
interests of both public and private organizations on the differential implications of internet use
for older people. Most nationally representative surveys that are publicly available do not have a
detailed or validated measure of internet use. The survey questions are usually vague, without a
defined concept of the purpose of usage or the frequency of usage. The development of a
validated internet use measure will be useful in more rigorously testing the effect of internet use
and isolating concerns for endogeneity.
Lastly, the overall findings suggest that internet communication use may be a
complement rather than a substitute for older people in maintaining their social contacts and
relationships. The development of policies and programs aimed at addressing social isolation of
older people using the internet need to be aware of the implications of the present findings and
consider heterogeneity of aging population. For instance, Waycott and colleagues (2016) found
that some older participants avoided participating in technology-based interventions for social
isolation. Future research should pay more attention on how to design technological
interventions that does not become futile effort by better examining the diverse needs of older
participants. Additionally, a lack of comparable technological interventions makes it challenging
to evaluate the effectiveness of internet communication use in addressing social isolation of older
people (Bartlett, Warburton, Lui, Peach & Carroll, 2013; Khosravi, Rezvani & Wiewiora, 2016).
Given the initial findings of the present dissertation and its limited scope, future research using a
mixed method design that incorporates both qualitative interviews and quantitative assessment of
technology use will offer a more comprehensive evaluation of effectiveness of internet
communication use in addressing social isolation and enhancing psychological well-being of
93

older people.
Technology is rapidly advancing and its use is spreading widely among older people in
different countries, and its significance in affecting the quality of life of people will become
more pronounced with the upcoming generations of older adults who are expected to be more
familiar with different types of technological devices and programs. The findings from the
current dissertation can be used as an initial step toward building a theoretical framework for
deepening understanding on the heterogeneity of internet use on diverse aging population. More
direct incorporation of older adults’ voice in designing and developing products and policies
related to technology use will contribute to making technology useful and age-friendly for
supporting successful aging and promoting social integration.

94

References
Administration on Aging. (2018). 2017 Profile of older Americans. Retrieved from
https://acl.gov/sites/default/files/Aging%20and%20Disability%20in%20America/2017OlderAme
ricansProfile.pdf
Adamy, J., & Overberg, P. (2018). The loneliest generation: Americans, more than ever, are aging
alone. The Wall Street Journal.
Agree, E. M. (2018). Demography of Aging and the Family. In Future directions for the
demography of aging: Proceedings of a workshop (pp. 159-186). Washington, DC: National
Academies Press.
American Association of Retired Persons. (2020). Families Anxious Over Loved Ones in
Nursing Homes, Assisted Living. AARP News. Retrieved from
https://www.aarp.org/caregiving/health/info-2020/preventing-coronavirus-in-nursing-homes.html
(Accessed on May 19, 2020)
American Association of Retired Persons. (2018a). A National Survey of Adults 45 and Older:
Loneliness and Social Connections. AARP Foundation. Retrieved from
https://www.aarp.org/content/dam/aarp/research/surveys_statistics/life-leisure/2018/loneliness-
social-connections-2018.doi.10.26419-2Fres.00246.001.pdf (Accessed on December 12, 2019)
American Association of Retired Persons. (2018b, July 24). Tech Training Builds Connections
and Confidence for Older Adults. AARP. Retrieved from https://www.aarp.org/home-
family/personal-technology/info-2018/technology-training-for-older-adults.html?intcmp=AE-HF-
BB (Accessed on July 26, 2018)
American Association of Retired Persons Research. (2019). 2019 Tech Trends and the
50+.Retrieved from https://www.aarp.org/content/dam/aarp/research/surveys_statistics
/technology/2019/2019-technology-trends.doi.10.26419-2Fres.00269.001.pdf (Accessed
on April 16, 2020)
95

Ang, S., & Chen, T. Y. (2019). Going Online to Stay Connected: Online Social Participation
Buffers the Relationship Between Pain and Depression. The Journals of Gerontology: Series
B, 74(6), 1020-1031.
Antonucci, T. C., Ajrouch, K. J., & Birditt, K. S. (2014). The convoy model: Explaining social
relations from a multidisciplinary perspective. The Gerontologist, 54(1), 82-92.
Armitage, R., & Nellums, L. B. (2020). COVID-19 and the consequences of isolating the
elderly. The Lancet Public Health.
Ausubel, J. (2020). Older people are more likely to live alone in the U.S. than elsewhere in the
world. Pew Research Center, Available from https://www.pewresearch.org/fact-
tank/2020/03/10/older-people-are-more-likely-to-live-alone-in-the-u-s-than-elsewhere-in-the-
world/
Bae, J. N., & Cho, M. J. (2004). Development of the Korean version of the Geriatric Depression Scale
and its short form among elderly psychiatric patients. Journal of psychosomatic research, 57(3),
297-305.
Bartlett, H., Warburton, J., Lui, C. W., Peach, L., & Carroll, M. (2013). Preventing social
isolation in later life: findings and insights from a pilot Queensland intervention study. Ageing &
Society, 33(7), 1167-1189.
Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. University of
Pennsylvania Press.
Bélanger, E., Ahmed, T., Vafaei, A., Curcio, C. L., Phillips, S. P., & Zunzunegui, M. V.
(2016). Sources of social support associated with health and quality of life: a cross-sectional
study among Canadian and Latin American older adults. BMJ open, 6(6), e011503.
Bengtson, V. L., & Roberts, R. E. (1991). Intergenerational solidarity in aging families: An
example of formal theory construction. Journal of Marriage and the Family, 856-870.
Berkman, L. F., & Syme, S. L. (1979). Social networks, host resistance, and mortality: a
96

nine-year follow-up study of Alameda County residents. American journal of
Epidemiology, 109(2), 186-204.
Bickford, D., Morin, R. T., Nelson, J. C., & Mackin, R. S. (2019). Determinants of suicide-
related ideation in late life depression: associations with perceived stress. Clinical gerontologist,
1-9.
Bixter, M. T., Blocker, K. A., & Rogers, W. A. (2018). Enhancing social engagement of
older adults through technology. In Aging, Technology and Health (pp. 179-214).
Academic Press.
Burholt, V., Windle, G., Morgan, D. J., & CFAS Wales team. (2017). A social model of
loneliness: The roles of disability, social resources, and cognitive impairment. The
Gerontologist, 57(6), 1020-1030.
Byers, A. L., & Yaffe, K. (2011). Depression and risk of developing dementia. Nature Reviews
Neurology, 7(6), 323.
Cacioppo, J. T., & Hawkley, L. C. (2009). Perceived social isolation and cognition. Trends in
cognitive sciences, 13(10), 447-454.
Cacioppo, J. T., Hawkley, L. C., Berntson, G. G., Ernst, J. M., Gibbs, A. C., Stickgold, R., &
Hobson, J. A. (2002). Do lonely days invade the nights? Potential social modulation of sleep
efficiency. Psychological Science, 13(4), 384-387.
Cahill, K. E., Giandrea, M. D., & Quinn, J. F. (2013). Retirement patterns and the
macroeconomy, 1992–2010: The prevalence and determinants of bridge jobs, phased retirement,
and reentry among three recent cohorts of older Americans. The Gerontologist, 55(3), 384-403.
Calhoun, D., & Lee, S. B. (2019). Computer usage and cognitive capability of older adults:
Analysis of data from the Health and Retirement Study. Educational Gerontology, 45(1),
22-33.
Calvo, E., Sarkisian, N., & Tamborini, C. R. (2012). Causal effects of retirement timing on
97

subjective physical and emotional health. Journals of Gerontology Series B: Psychological
Sciences and Social Sciences, 68(1), 73-84.
Carstensen, L. L., Fung, H. H., & Charles, S. T. (2003). Socioemotional selectivity theory
and the regulation of emotion in the second half of life. Motivation and emotion, 27(2), 103-123.
Carver, L. F., Beamish, R., & Phillips, S. P. (2018). Successful Aging: Illness and Social
Connections. Geriatrics, 3(1), 3.
Challands, K. G., Lacherez, P., & Obst, P. L. (2017). Does online social connectedness buffer
risk of depression following driving cessation? An analysis of older drivers and ex-
drivers. Cyberpsychology, behavior, and social networking, 20(4), 232-237.
Charles, S. T., & Carstensen, L. L. (2010). Social and emotional aging. Annual review of
psychology, 61, 383-409.
Charness, N., & Boot, W. R. (2009). Aging and information technology use: Potential and
barriers. Current Directions in Psychological Science, 18(5), 253-258.
Charness, N., & Jastrzembski, T. S. (2009). Gerontechnology. In Future interaction design
II (pp. 1-29). Springer, London.
Cho, J., Jin, Y., & Kang, H. (2018). Weight status, physical activity, and depression in Korean
older adults. Journal of epidemiology, JE20170083.
Choi, N. G., & DiNitto, D. M. (2013
a
). Internet use among older adults: association with health
needs, psychological capital, and social capital. Journal of Medical Internet Research, 15(5).
Choi, N. G., & DiNitto, D. M. (2013
b
). The digital divide among low-income homebound older
adults: Internet use patterns, eHealth literacy, and attitudes toward computer/Internet
use. Journal of medical Internet research, 15(5), e93.
Choi, H., Schoeni, R. F., Langa, K. M., & Heisler, M. M. (2015). Older adults’ residential
proximity to their children: Changes after cardiovascular events. Journals of Gerontology Series
B: Psychological Sciences and Social Sciences, 70(6), 995-1004.
98

Chopik, W. J. (2016). The benefits of social technology use among older adults are
mediated by reduced loneliness. Cyberpsychology, Behavior, and Social
Networking, 19(9), 551-556.
Chu, W. M., Liao, W. C., Li, C. R., Lee, S. H., Tang, Y. J., Ho, H. E., & Lee, M. C. (2016).
Late-career unemployment and all-cause mortality, functional disability and depression among
the older adults in Taiwan: A 12-year population-based cohort study. Archives of gerontology and
geriatrics, 65, 192-198.
Chun, M. A. & Kim, J. H. (2016). The effects of receiving financial support from/to their
offsprings on happiness among Korean older persons, Korean Journal of Gerontological Social
Welfare, 71(2), 249-268.
Cornwell, B., Laumann, E. O., & Schumm, L. P. (2008). The social connectedness of older
adults: A national profile. American sociological review, 73(2), 185-203.
Cotten, S. R. (2017). Examining the roles of technology in aging and quality of
life. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 72(5), 823-
826.
Cotten, S. R., Anderson, W. A., & McCullough, B. M. (2013). Impact of internet use on
loneliness and contact with others among older adults: cross-sectional analysis. Journal
of medical Internet research, 15(2), e39.
Cotten, S. R., Ford, G., Ford, S., & Hale, T. M. (2014). Internet use and depression
among retired older adults in the United States: A longitudinal analysis. Journals of Gerontology
Series B: Psychological Sciences and Social Sciences, 69(5), 763-771.
Crimmins, E. M., Kim, J. K., & Solé-Auró, A. (2010). Gender differences in health: results from
SHARE, ELSA and HRS. European journal of public health, 21(1), 81-91.
Cudjoe, T. K., Roth, D. L., Szanton, S. L., Wolff, J. L., Boyd, C. M., & Thorpe, R. J.
99

(2018). The epidemiology of social isolation: National health and aging trends study. The
Journals of Gerontology: Series B.
Cumming, E., & Henry, W. E. (1961). Growing old, the process of disengagement. Basic books.
Curran, P. J., Obeidat, K., & Losardo, D. (2010). Twelve frequently asked questions about
growth curve modeling. Journal of cognition and development, 11(2), 121-136.
Czaja, S. J. (2017). The role of technology in supporting social engagement among older
adults. Public Policy & Aging Report, 27(4), 145-148.
Davis, E. M., & Fingerman, K. L. (2015). Digital dating: Online profile content of older and
younger adults. Journals of Gerontology Series B: Psychological Sciences and Social
Sciences, 71(6), 959-967.
Diener, E., Oishi, S., & Tay, L. (2018). Advances in subjective well-being research. Nature
Human Behavior, 2(4), 253.
Djernes, J. K. (2006). Prevalence and predictors of depression in populations of elderly: a
review. Acta Psychiatrica Scandinavica, 113(5), 372-387.
Djundeva, M., Mills, M., Wittek, R., & Steverink, N. (2015). Receiving instrumental support in
late parent–child relationships and parental depression. The Journals of Gerontology: Series
B, 70(6), 981-994.
Donnellan, W. J., Bennett, K. M., & Soulsby, L. K. (2017). Family close but friends
closer: exploring social support and resilience in older spousal dementia carers. Aging & mental
health, 21(11), 1222-1228.
100

Doshi, J. A., Cen, L., & Polsky, D. (2008). Depression and retirement in late middle‐aged US
workers. Health services research, 43(2), 693-713.
Elder, K., & Retrum, J. (2012). Framework for isolation in adults over 50: AARP
Foundation/isolation Framework Project. San Diego CA: Research Works.
Elliot, A. J., Mooney, C. J., Douthit, K. Z., & Lynch, M. F. (2013). Predictors of older
adults’ technology use and its relationship to depressive symptoms and well-being. Journals of
Gerontology Series B: Psychological Sciences and Social Sciences, 69(5), 667-677.
Erikson, E. H. (1994). Identity and the life cycle. WW Norton & Company.
Farnsworth, J., Pett, M. A., & Lund, D. A. (1989). Predictors of loss management and well-being
in later life widowhood and divorce. Journal of Family Issues, 10(1), 102-121.
Ferraro, K. F., & Barresi, C. M. (1982). The impact of widowhood on the social relations
of older persons. Research on Aging, 4(2), 227-247.
Ferraro, K. F., Mutran, E., & Barresi, C. M. (1984). Widowhood, health, and friendship support
in later life. Journal of Health and Social Behavior, 246-259.
Fingerman, K. L., Huo, M., Charles, S. T., & Umberson, D. J. (2020). Variety is the spice of late
life: Social integration and daily activity. The Journals of Gerontology: Series B, 75(2), 377-388.
Fiske, A., Wetherell, J. L., & Gatz, M. (2009). Depression in older adults. Annual review of
clinical psychology, 5, 363-389.
Forsman, A. K., Nordmyr, J., Matosevic, T., Park, A. L., Wahlbeck, K., & McDaid, D. (2018).
101

Promoting mental wellbeing among older people: technology-based interventions. Health
promotion international, 33(6), 1042-1054.
Freedman, V. A., Kasper, J. D., Spillman, B. C., Agree, E. M., Mor, V., Wallace, R. B., & Wolf,
D. A. (2014). Behavioral adaptation and late-life disability: a new spectrum for assessing public
health impacts. American Journal of Public Health, 104(2), e88-e94.
Friedman, E. M., Park, S. S., & Wiemers, E. E. (2015). New estimates of the sandwich
generation in the 2013 panel study of income dynamics. The Gerontologist, 57(2), 191-196.
Gleason, O. C., Pierce, A. M., Walker, A. E., & Warnock, J. K. (2013). The two-way
relationship between medical illness and late-life depression. Psychiatric Clinics, 36(4), 533-544.
Goodwin, G. M. (2006). Depression and associated physical diseases and symptoms. Dialogues
in clinical neuroscience, 8(2), 259.
Gubernskaya, Z., & Treas, J. (2016). Call home? Mobile phones and contacts with mother in 24
countries. Journal of Marriage and Family, 78(5), 1237-1249.
Ha, J. H., Yoon, H., Lim, Y. O., & Heo, S. Y. (2016). The effect of widowhood on parent–child
relationships in Korea: Do parents’ filial expectations and geographic proximity to children
matter?. Journal of cross-cultural gerontology, 31(1), 73-88.
Han, D., & Braun, K. L. (2013). Promoting active ageing through technology training in Korea.
In Digital Literacy: Concepts, Methodologies, Tools, and Applications (pp. 572-589). IGI Global.
Hansson, I., Buratti, S., Thorvaldsson, V., Johansson, B., & Berg, A. I. (2017). Changes
102

in life satisfaction in the retirement transition: Interaction effects of transition type and individual
resources. Work, Aging and Retirement, 4(4), 352-366.
Harerimana, B., Forchuk, C., & O'Regan, T. (2019). The use of technology for mental healthcare
delivery among older adults with depressive symptoms: A systematic literature
review. International journal of mental health nursing, 28(3), 657-670.
Harwood, J. (2000). Communication media use in the grandparent-grandchild
relationship. Journal of Communication, 50(4), 56-78.
Havighurst, R. J. (1968). Personality and patterns of aging. The gerontologist, 8(1_Part_2), 20-
23.
Hawkley, L. C., Masi, C. M., Berry, J. D., & Cacioppo, J. T. (2006). Loneliness is a unique
predictor of age-related differences in systolic blood pressure. Psychology and aging, 21(1), 152.
Hees, E., Tesch-Römer, C., & Huxhold, O. (2019). Internet use and loneliness: Cure or
cause? Longitudinal Analysis of older adults’ Internet use. Innovation in Aging, 3(Suppl 1), S14.
Henning-Smith, C. (2016). Quality of life and psychological distress among older adults: The
role of living arrangements. Journal of Applied Gerontology, 35(1), 39-61.
Henning-Smith, C., Shippee, T., & Capistrant, B. (2018). Later-life disability in environmental
context: Why living arrangements matter. The Gerontologist, 58(5), 853-862.
Heo, J., Chun, S., Lee, S., Lee, K. H., & Kim, J. (2015). Internet use and well-being in older
adults. Cyberpsychology, Behavior, and Social Networking, 18(5), 268-272.
Heo, M., Murphy, C. F., Fontaine, K. R., Bruce, M. L., & Alexopoulos, G. S. (2008). Population
103

projection of US adults with lifetime experience of depressive disorder by age and sex from year
2005 to 2050. International Journal of Geriatric Psychiatry: A journal of the psychiatry of late
life and allied sciences, 23(12), 1266-1270.
Holmes, T. H., & Rahe, R. H. (1967). The social readjustment rating scale. Journal of
psychosomatic research.
Holt-Lunstad, J. (2018). Why social relationships are important for physical health: A systems
approach to understanding and modifying risk and protection. Annual review of psychology, 69,
437-458.
Holt-Lunstad, J., Robles, T. F., & Sbarra, D. A. (2017). Advancing social connection as a public
health priority in the United States. American Psychologist, 72(6), 517.
Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and
social isolation as risk factors for mortality: a meta-analytic review. Perspectives on
psychological science, 10(2), 227-237.
Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: a
meta-analytic review. PLoS med, 7(7), e1000316.
Hong, Y. A., & Cho, J. (2016). Has the digital health divide widened? Trends of health-related
internet use among older adults from 2003 to 2011. Journals of Gerontology Series B:
Psychological Sciences and Social Sciences, 72(5), 856-863.
House, J. S. (2001). Social isolation kills, but how and why?. Psychosomatic medicine, 63(2),
273-274.
Hughes, M. E., Waite, L. J., Hawkley, L. C., & Cacioppo, J. T. (2004). A short scale for
104

measuring loneliness in large surveys: Results from two population-based studies. Research on
aging, 26(6), 655-672.
Isherwood, L. M., King, D. S., & Luszcz, M. A. (2012). A longitudinal analysis of social
engagement in late-life widowhood. The International Journal of Aging and Human
Development, 74(3), 211-229.
Jacob, L., Haro, J.M., Koyanagi, A. (2019). Relationship between living alone and common
mental disorders in the 1993, 2000 and 2007 National Psychiatric Morbidity Surveys.
PLoS ONE 14(5): e0215182. https://doi.org/10.1371/journal.pone.0215182
Jang, S. N., Avendano, M., & Kawachi, I. (2012). Informal caregiving patterns in Korea and
European countries: a cross-national comparison. Asian Nursing Research, 6(1), 19-26.
Jones, R. B., Ashurst, E. J., Atkey, J., & Duffy, B. (2015). Older people going online: its value
and before-after evaluation of volunteer support. Journal of medical Internet research, 17(5),
e122
Jun, H. J., & Kim, M. Y. (2016). What accounts for the relationship between internet use and
suicidal ideation of Korean older adults? A mediation analysis. Journals of Gerontology Series B:
Psychological Sciences and Social Sciences, 72(5), 846-855.
Jung, K. Oh, Y., Lee, Y., Oh, M., Kang, E., Kim, K., Hwang, N., Kim, S., Lee, S., Lee, S. &
Hong, S. (2017). Report on the 2017 Living Profiles of Older People Survey (LPOPS). Ministry
of Health and Welfare. 1-736.
Kahana, E. (1982). A congruence model of person-environment interaction. Aging and the
environment: Theoretical approaches, 97-121.
105

Kahana, E., Liang, J., & Felton, B. J. (1980). Alternative models of person-environment fit:
Prediction of morale in three homes for the aged. Journal of Gerontology, 35(4), 584-595.
Kahn, R. L., & Antonucci, T. C. (1980). Convoys over the life course: Attachment, roles
and social support. In P. B. Baltes & O. G. Brim Jr. (Eds.), Life-span development and behavior
(Vol. 3, pp. 253–286). New York, NY: Academic Press.
Kasper, Judith D. and Freedman, Vicki A. (2017). National Health and Aging Trends Study User
Guide: Rounds 1-6 Final Release [internet]. Baltimore: Johns Hopkins University School of
Public Health. Available at www.NHATS.org.
Khosravi, P., Rezvani, A., & Wiewiora, A. (2016). The impact of technology on older
adults’ social isolation. Computers in Human Behavior, 63, 594-603.
Kim, J., Lee, H. Y., Christensen, M. C., & Merighi, J. R. (2017). Technology access and use, and
their associations with social engagement among older adults: Do women and men
differ?. Journals of Gerontology Series B: Psychological Sciences and Social
Sciences, 72(5), 836-845.
Kim, K., Birditt, K. S., Zarit, S. H., & Fingerman, K. L. (2019). Typology of Parent–Child Ties
Within Families: Associations With Psychological Well-Being. Journal of family psychology:
JFP: journal of the Division of Family Psychology of the American Psychological Association
(Division 43).
Kim, K., Lehning, A. J., & Sacco, P. (2016). Assessing the factor structure of well-being in older
adults: findings from the National Health and Aging Trends Study. Aging & mental health, 20(8),
814-822.
106

Kitson, G. C., Babri, K. B., Roach, M. J., & Placidi, K. S. (1989). Adjustment to
widowhood and divorce: A review. Journal of Family Issues, 10(1), 5-32.
Klinenberg, E. (2016). Social isolation, loneliness, and living alone: Identifying the risks for
public health. American journal of public health, 106(5), 786.
Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukophadhyay, T., & Scherlis, W. (1998).
Internet paradox: A social technology that reduces social involvement and psychological well-
being?. American psychologist, 53(9), 1017.
Kroenke, K., Spitzer, R. L., & Williams, J. B. (2003). The Patient Health Questionnaire-2:
validity of a two-item depression screener. Medical care, 1284-1292.
Kuehner, C. (2017). Why is depression more common among women than among men?. The
Lancet Psychiatry, 4(2), 146-158.
Kwon, E., Kim, B., Lee, H., & Park, S. (2018). Heterogeneous trajectories of depressive
symptoms in late middle age: critical period, accumulation, and social mobility life course
perspectives. Journal of aging and health, 30(7), 1011-1041.
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.
Laugesen, K., Baggesen, L. M., Schmidt, S. A. J., Glymour, M. M., Lasgaard, M.,
Milstein, A., ... & Ehrenstein, V. (2018). Social isolation and all-cause mortality: a population-
based cohort study in Denmark. Scientific reports, 8(1), 4731.
Lee, H. Y., Kim, J., & Sharratt, M. (2018). Technology use and its association with health and
depressive symptoms in older cancer survivors. Quality of Life Research, 27(2), 467-477.
107

Lee, S., Oh, E., & Hong, G. R. S. (2018). Comparison of factors associated with fear of falling
between older adults with and without a fall history. International journal of environmental
research and public health, 15(5), 982.
Lee, Y., Chi, I., & Ailshire, J. A. (2018). Life transitions and leisure activity engagement
among older Americans: findings from a national longitudinal study. Ageing & Society, 1-28.
Leggett, A. N., Sonnega, A. J., & Lohman, M. C. (2018). The association of insomnia and
depressive symptoms with all‐cause mortality among middle‐aged and old adults. International
journal of geriatric psychiatry, 33(9), 1265-1270.
Levine, D. M., Lipsitz, S. R., & Linder, J. A. (2017). Changes in everyday and digital
health technology use among seniors in declining health. The Journals of Gerontology: Series A.
Levy, H., Janke, A. T., & Langa, K. M. (2015). Health literacy and the digital divide
among older Americans. Journal of general internal medicine, 30(3), 284-289.
Liang, D., & Zhang, D. (2017). Children’s Geographic Proximity and Older Parents’ Depressive
Symptoms in China. Gerontology and Geriatric Medicine, 3, 2333721417729501.
Luong, G., Charles, S. T., & Fingerman, K. L. (2011). Better with age: Social relationships
across adulthood. Journal of social and personal relationships, 28(1), 9-23.
Macunovich, D. J., Easterlin, R. A., Schaeffer, C. M., & Crimmins, E. M. (1995). Echoes of the
baby boom and bust: Recent and prospective changes in living alone among elderly widows in
the United States. Demography, 32(1), 17-28.
Manemann, S. M., Chamberlain, A. M., Roger, V. L., Griffin, J. M., Boyd, C. M., Cudjoe, T.
108

K., ... & Finney Rutten, L. J. (2018). Perceived social isolation and outcomes in patients with
heart failure. Journal of the American Heart Association, 7(11), e008069.
Marmot, M. G. (2014). English Longitudinal Study of Aging: can internet/e-mail use reduce
cognitive decline?. Journals of Gerontology Series A: Biomedical Sciences and Medical
Sciences, 69(9), 1117-1121.
Martin, P., Deshpande-Kamat, N., Poon, L. W., & Johnson, M. A. (2011). The model of
developmental adaptation: Implications for understanding well-being in old-old
age. In Poon, L. W. & Cohen-Mansfield, J. (Eds.), Understanding Well-Being in
the Oldest Old (p. 65–78). New York, NY: Cambridge.
Maslow, A. H. (1973). Dominance, self-esteem, self-actualization: germinal papers of AH
Maslow. Thomson Brooks/Cole.
Michael, Y. L., Berkman, L. F., Colditz, G. A., & Kawachi, I. (2001). Living arrangements,
social integration, and change in functional health status. American journal of
epidemiology, 153(2), 123-131.
Min, J., Li, Y., Xu, L., & Chi, I. (2018). Psychological vulnerability of widowhood: financial
strain, social engagement and worry about having no care-giver as mediators and
moderators. Ageing & Society, 38(11), 2356-2375.
Mincer, J. (1991). Education and unemployment (No. w3838). National Bureau of Economic
Research.
Mirza, S. S., Wolters, F. J., Swanson, S. A., Koudstaal, P. J., Hofman, A., Tiemeier, H., & Ikram,
109

M. A. (2016). 10-year trajectories of depressive symptoms and risk of dementia: a population-
based study. The Lancet Psychiatry, 3(7), 628-635.
Mitchell, U. A., Chebli, P. G., Ruggiero, L., & Muramatsu, N. (2018). The digital divide in
health-related technology use: The significance of race/ethnicity. The
Gerontologist, 59(1), 6-14.
Morgan, C., Webb, R. T., Carr, M. J., Kontopantelis, E., Chew-Graham, C. A., Kapur, N., &
Ashcroft, D. M. (2018). Self-harm in a primary care cohort of older people: incidence, clinical
management, and risk of suicide and other causes of death. The Lancet Psychiatry, 5(11), 905-
912.
Mosca, I., & Barrett, A. (2014). The impact of voluntary and involuntary retirement on
mental health: evidence from older Irish adults. IZA Discussion Papers, No. 8723.
Naghavi, M. (2019). Global, regional, and national burden of suicide mortality 1990 to 2016:
systematic analysis for the Global Burden of Disease Study 2016. BMJ, 364, l94.
Nam, E. J., & Lee, J. E. (2019). Mediating effects of social support on depression and
suicidal ideation in older Korean adults with hypertension who live alone. Journal of Nursing
Research, 27(3), e20.
Nam, S., Han, S. H., & Gilligan, M. (2018). Internet Use and Preventive Health Behaviors
Among Couples in Later Life: Evidence from the Health and Retirement Study. The
Gerontologist.
National Alliance on Mental Illness. (2009). Depression in older persons: Fact sheet. Retrieved
from https://www.ncoa.org/wp-content/uploads/Depression_Older_Persons_FactSheet
110

_2009.pdf
National Science and Technology Council. (2019). Emerging Technologies to Support an Aging
Population. Washington, DC: Author. Retrieved from https://www.whitehouse.gov/wp-
content/uploads/2019/03/Emerging-Tech-to-Support-Aging-2019.pdf
National Telecommunications and Information Administration. (2018). Digital Nation Data
Explorer, Available from https://www.ntia.doc.gov/data/digital-nation-data-
explorer#sel=internetUser&demo=age&pc=prop&disp=chart
Organization for Economic Co-operation and Development (OECD). (2017). Pensions at a
Glance 2017: OECD and G20 indicators, OECD Publishing, Paris. Available from:
https://www.oecd-ilibrary.org/social-issues-migration-health/pensions-at-a-glance-2017/old-age-
dependency-ratio_pension_glance-2017-22-en. Accessed October 29, 2019
Oliffe, J. L., Rasmussen, B., Bottorff, J. L., Kelly, M. T., Galdas, P. M., Phinney, A., &
Ogrodniczuk, J. S. (2013). Masculinities, work, and retirement among older men who experience
depression. Qualitative health research, 23(12), 1626-1637.
Ong, A. D., & Allaire, J. C. (2005). Cardiovascular intraindividual variability in later life: the
influence of social connectedness and positive emotions. Psychology and aging, 20(3), 476.
Ono, H., & Zavodny, M. (2007). Digital inequality: A five country comparison using
microdata. Social Science Research, 36(3), 1135-1155.
O’Rourke, H. M., Collins, L., & Sidani, S. (2018). Interventions to address social connectedness
and loneliness for older adults: a scoping review. BMC geriatrics, 18(1), 214.
111

Oxman, T. E., & Hull, J. G. (2001). Social support and treatment response in older depressed
primary care patients. The Journals of Gerontology Series B: Psychological Sciences and Social
Sciences, 56(1), P35-P45.
Pantell, M., Rehkopf, D., Jutte, D., Syme, S. L., Balmes, J., & Adler, N. (2013). Social isolation:
a predictor of mortality comparable to traditional clinical risk factors. American journal
of public health, 103(11), 2056-2062.
Park, H. (2014). Inequality of educational opportunity in Korea by gender, socio-economic
background, and family structure. In Equality in Asia-Pacific (pp. 182-200). Routledge.
Park, S. S. (2017). Life events and black–white differences in adult children’s financial
assistance to mothers. The Gerontologist, 58(5), 883-893.
Park, S., Han, Y., Kim, B., & Dunkle, R. E. (2017). Aging in place of vulnerable older adults:
Person–environment fit perspective. Journal of Applied Gerontology, 36(11), 1327-1350.
Pearson, J. E. (1986). The definition and measurement of social support. Journal of
Counseling & Development.
Pohl, J. S., Cochrane, B. B., Schepp, K. G., & Woods, N. F. (2018). Falls and social isolation of
older adults in the National Health and Aging Trends Study. Research in gerontological
nursing, 11(2), 61-70.
Poon, L. W., & Cohen-Mansfield, J. (Eds.). (2011). Understanding well-being in the oldest old.
Cambridge University Press.
Poulin, J., Deng, R., Ingersoll, T. S., Witt, H., & Swain, M. (2012). Perceived family and friend
112

support and the psychological well-being of American and Chinese elderly persons. Journal of
cross-cultural gerontology, 27(4), 305-317.
Primack, B. A., Shensa, A., Sidani, J. E., Whaite, E. O., yi Lin, L., Rosen, D., ... &
Miller, E. (2017). Social media use and perceived social isolation among young adults in the
US. American journal of preventive medicine, 53(1), 1-8.
Quintana, D., Cervantes, A., Sáez, Y., & Isasi, P. (2018). Internet use and psychological well-
being at advanced age: Evidence from the English longitudinal study of
aging. International journal of environmental research and public health, 15(3), 480.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general
population. Applied psychological measurement, 1(3), 385-401.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data
analysis methods (2
nd
Ed.). Thousand Oaks, CA: Sage.
Recksiedler, C., & Stawski, R. S. (2019). Marital transitions and depressive symptoms among
older adults: examining educational differences. Gerontology, 65(4), 407-418.
Robins, C. J., & Block, P. (1989). Cognitive theories of depression viewed from a diathesis-
stress perspective: Evaluations of the models of Beck and of Abramson, Seligman, and
Teasdale. Cognitive Therapy and Research, 13(4), 297-313.
Roh, H. W., Lee, Y., Lee, K. S., Chang, K. J., Kim, J., Lee, S. J., ... & Son, S. J. (2015).
Frequency of contact with non-cohabitating adult children and risk of depression in elderly: A
community-based three-year longitudinal study in Korea. Archives of gerontology and
geriatrics, 60(1), 183-189.
Rowe, J. W., & Kahn, R. L. (1997). Successful aging. The gerontologist, 37(4), 433-440.
113

Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of research on
hedonic and eudaimonic well-being. Annual review of psychology, 52(1), 141-166.
Ryff, C. D. (1995). Psychological well-being in adult life. Current directions in psychological
science, 4(4), 99-104.
Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the meaning of
psychological well-being. Journal of personality and social psychology, 57(6), 1069.
Seeman, T. E. (2000). Health promoting effects of friends and family on health outcomes
in older adults. American journal of Health promotion, 14(6), 362-370.
Segel-Karpas, D., Ayalon, L., & Lachman, M. E. (2018). Retirement and depressive
symptoms: A 10-year cross-lagged analysis. Psychiatry research, 269, 565-570.
Shapira, N., Barak, A., & Gal, I. (2007). Promoting older adults’ well-being through Internet
training and use. Aging & Mental Health. 11(5):477-84.
Shaw, L. H., & Gant, L. M. (2004). In defense of the Internet: The relationship between Internet
communication and depression, loneliness, self-esteem, and perceived social support. Internet
Research, 28(3).
Sheikh, J. I., & Yesavage, J. A. (1986). Geriatric Depression Scale (GDS): Recent evidence and
development of a shorter version. Clinical Gerontologist: The Journal of Aging and Mental
Health, 5(1-2), 165–173. https://doi.org/10.1300/J018v05n01_09
Shiba, K., Kondo, N., Kondo, K., & Kawachi, I. (2017). Retirement and mental health:
114

does social participation mitigate the association? A fixed-effects longitudinal analysis. BMC
public health, 17(1), 526.
Shim, S. (2018). Analysis of Factors of Life Satisfaction of Older Koreans. KOSTAT Statistical
Report, Statistics Korea, Available from http://kostat.go.kr/sri/srikor/srikor_pbl/4/index.
board?bSeq=&pageNo=3&rowNum=10&amSeq=&sTarget=&sTxt=(accessed on
November, 7, 2019)
Sims, T., Reed, A. E., & Carr, D. C. (2017). Information and communication technology
use is related to higher well-being among the oldest-old. The Journals of Gerontology: Series
B, 72(5), 761-770.
Smith, J., Ryan, L., Sonnega, A., & Weir, D. (2017). Psychosocial and lifestyle questionnaire
2006–2016. Ann Arbor, MI: Survey Research Center, Institute for Social Research.
Solé-Auró, A., & Crimmins, E. M. (2014). Who cares? A comparison of informal and formal
care provision in Spain, England and the USA. Ageing & Society, 34(3), 495-517.
Sonnega, A., Faul, J. D., Ofstedal, M. B., Langa, K. M., Phillips, J. W., & Weir, D. R. (2014).
Cohort profile: The health and retirement study (HRS). International journal of
epidemiology, 43(2), 576-585.
Stahl, S. T., Beach, S. R., Musa, D., & Schulz, R. (2017). Living alone and depression: the
modifying role of the perceived neighborhood environment. Aging & mental health, 21(10),
1065-1071.
Steffick, D. E. (2000). Documentation of affective functioning measures in the Health and Retirement
Study. Ann Arbor, MI: University of Michigan.
115

Stepler, R. (2017). Led by Baby Boomers, divorce rates climb for America’s 50+
population. Pew Research Center. Retrieved from https://www.pewresearch.org/fact-
tank/2017/03/09/led-by-baby-boomers-divorce-rates-climb-for-americas-50-population/
(Accessed on July 26, 2018)
Steptoe, A., & Fancourt, D. (2019). Leading a meaningful life at older ages and its relationship
with social engagement, prosperity, health, biology, and time use. Proceedings of the
National Academy of Sciences, 116(4), 1207-1212.
Sum, S., Mathews, R. M., Hughes, I., & Campbell, A. (2008). Internet use and loneliness in
older adults. CyberPsychology & Behavior, 11(2), 208-211.
Sung, K. T. (1990). A new look at filial piety: Ideals and practices of family-centered parent care
in Korea. The Gerontologist, 30(5), 610-617.
Teguo, M. T., Simo-Tabue, N., Stoykova, R., Meillon, C., Cogne, M., Amiéva, H., & Dartigues,
J. F. (2016). Feelings of loneliness and living alone as predictors of mortality in the elderly: the
PAQUID study. Psychosomatic medicine, 78(8), 904-909.
Teo, A. R., Choi, H., Andrea, S. B., Valenstein, M., Newsom, J. T., Dobscha, S. K., & Zivin, K.
(2015). Does mode of contact with different types of social relationships predict
depression in older adults? Evidence from a nationally representative survey. Journal of
the American Geriatrics Society, 63(10), 2014-2022.
Townsend, P. (1968). The structure of the family. In E. Shanas, P. Towsend, D. Wedderburn, H.
116

K. Friis, P. Milhot, & J. Stehouwer (Eds.), Old people in three industrial societies (pp. 132-176).
Abingdon and New York: Routledge.
U.S. Census Bureau (2019). Current Population Survey, Historical Households Tables.
[November 2019]. Available from https://www.census.gov/data/tables/time-
series/demo/families/households.html.
Umberson, D., & Karas Montez, J. (2010). Social relationships and health: A flashpoint for
health policy. Journal of health and social behavior, 51(1_suppl), S54-S66.
Utz, R. L., Carr, D., Nesse, R., & Wortman, C. B. (2002). The effect of widowhood on older
adults' social participation: An evaluation of activity, disengagement, and continuity theories. The
Gerontologist, 42(4), 522-533.
van der Pers, M., Kibele, E. U., & Mulder, C. H. (2015). Intergenerational proximity and the
residential relocation of older people to care institutions and elsewhere. Ageing & Society, 35(7),
1429-1456.
van Deursen, A. J., & Helsper, E. J. (2015). A nuanced understanding of Internet use and non-
use among the elderly. European Journal of Communication, 30(2), 171-187.
Van Orden, K. A., Bower, E., Lutz, J., Silva, C., Gallegos, A. M., Podgorski, C. A., Santos, E. J.,
& Conwell, Y. (2020). Strategies to Promote Social Connections Among Older Adults During
‘Social Distancing’ Restrictions. The American Journal of Geriatric Psychiatry, Advance online
publication. https://doi.org/10.1016/j.jagp.2020.05.004
Van Orden, K., & Conwell, Y. (2011). Suicides in late life. Current psychiatry reports, 13(3),
234-241.
117

Van Solinge, H., & Henkens, K. (2007). Involuntary retirement: The role of restrictive
circumstances, timing, and social embeddedness. The Journals of Gerontology Series B:
Psychological Sciences and Social Sciences, 62(5), S295-S303.
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of
information technology: Toward a unified view. MIS quarterly, 425-478.
Victor, C., Scambler, S. J., Bond, J., & Bowling, A. (2000). Being alone in later life: Loneliness,
social isolation, and living alone. Reviews in Clinical Gerontology, 10, 407-417.
Waycott, J., Vetere, F., Pedell, S., Morgans, A., Ozanne, E., & Kulik, L. (2016, May).
Not for me: Older adults choosing not to participate in a social isolation intervention.
In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 745-
757). ACM.
Wenger, G. C., Davies, R., Shahtahmasebi, S., & Scott, A. (1996). Social isolation and
loneliness in old age: review and model refinement. Ageing and Society, 16(03), 333-358.
Wilson, R. S., Barnes, L. L., De Leon, C. M., Aggarwal, N. T., Schneider, J. S., Bach, J., ... &
Bennett, D. A. (2002). Depressive symptoms, cognitive decline, and risk of AD in older
persons. Neurology, 59(3), 364-370.
Wilson, R. S., Krueger, K. R., Arnold, S. E., Schneider, J. A., Kelly, J. F., Barnes, L. L., ... &
Bennett, D. A. (2007). Loneliness and risk of Alzheimer disease. Archives of general
psychiatry, 64(2), 234-240.
Wong, Y. C., Law, C. K., Fung, J. Y. C., & Lee, V. W. P. (2010). Digital divide and social
118

inclusion: policy challenge for social development in Hong Kong and South Korea. Journal of
Asian Public Policy, 3(1), 37-52.
World Bank. (2017). Individuals using the Internet (% of population). International
Telecommunication Union, World Telecommunication/ICT
Development Report and database, and World Bank estimates. Available from
https://data.worldbank.org/indicator/IT.NET.USER.ZS (accessed on October, 22, 2019)
World Health Organization. (2018).Global Health Estimates 2016: Disease burden by Cause,
Age, Sex, by Country and by Region, 2000-2016. Geneva, Switzerland. Retrieved from
https://www.who.int/healthinfo/global_burden_disease/estimates/en/index1.html (accessed on
October, 24, 2019)
World Health Organization. (2017). Depression and Other Common Mental Disorders.
Available from https://apps.who.int/iris/bitstream/handle/10665/254610/WHO-MSD-MER-
2017.2-eng.pdf (accessed on September, 20, 2019)
Wrzus, C., Hänel, M., Wagner, J., & Neyer, F. J. (2013). Social network changes and life
events across the life span: a meta-analysis. Psychological bulletin, 139(1), 53.
Xavier, A. J., d’Orsi, E., de Oliveira, C. M., Orrell, M., Demakakos, P., Biddulph, J. P., &
Marmot, M. G. (2014). English Longitudinal Study of Aging: can internet/e-mail use reduce
cognitive decline?. Journals of Gerontology Series A: Biomedical Sciences and Medical
Sciences, 69(9), 1117-1121.
Xiang, X., Chen, J., & Kim, M. (2019). Trajectories of Homebound Status in Medicare
Beneficiaries Aged 65 and Older. The Gerontologist.
119

Yoon, H. S. (2013). Korea: Balancing economic growth and social protection for older
adults. The Gerontologist, 53(3), 361-368.
Yu, R. P., Ellison, N. B., McCammon, R. J., & Langa, K. M. (2016). Mapping the two levels of
digital divide: Internet access and social network site adoption among older adults in the
USA. Information, Communication & Society, 19(10), 1445-1464.
Zambianchi, M., & Carelli, M. G. (2018). Positive Attitudes towards Technologies and facets of
well-being in Older Adults. Journal of Applied Gerontology, 37(3), 371-388.

Abstract (if available)

Abstract The use of the internet for older adults has emerged as an essential tool for enhancing their quality of life and maintaining social connections in the recent years. Using the internet has become an integral part of daily life of people in the 21st century, and advances in technology offer innovative and various media of communications over different time zones and vast distance to keep contact with family and friends, despite migration and declining household sizes. ❧ An increasing number of studies have documented evidence of positive effects of technology use on psychological well-being and health of older adults. However, existing literature primarily examined older people as one homogenous group, making it impossible to examine how the differences in the individual and contextual factors affect the association between internet use and mental health. Additionally, the role of internet use in intergenerational relationship has not been explored despite its potential implications for maintaining family support in the globalized world. ❧ The purpose of the present dissertation is to address this gap by examining the effect of internet use for older people situated in different individual and environmental contexts and examine whether the effect of technology use shows heterogeneity depending on these contexts. The first paper examines people with different environmental contexts: people who live alone compared to people who live together with spouses and others. The National Health and Aging Trends Study was analyzed to examine whether the use of electronic communication had different impacts on psychological well-being and depressive symptoms for people who live alone compared to people who live together with others. Growth curve models showed that the effect of using electronic communication was significantly associated with higher psychological well-being both at the initial level and over time, but this effect did not significantly differ by living arrangements. For depressive symptoms, the effect of using electronic communication was significantly associated with lower depressive symptoms at the initial level for both types of living arrangements, but there was no significant effect over time for either group. ❧ The second paper focuses on older people whose cultural and geographic contexts are situated in the different contexts. Using the Health and Retirement Study in the U.S. and the Living Profiles of Older People Survey in Korea, the paper examined whether the internet use was differentially related to geographic proximity to adult children by gender in the two countries where different gender norms and expectations about intergenerational relationship exist. Ordinary least square regression models showed that internet use was significantly associated with reduced depressive symptoms more for parents whose adult children were living far away, with different effects by gender in the two countries. ❧ The third paper investigates how significant transitions in marital and work status at older ages are related to the association between internet use and psychological distress. Significant transitions in marital status were defined as changes from being married to a divorce or widowhood, while transitions in work status referred to changes from working to unemployment or retirement. The Health and Retirement Study was used to examine whether internet use had different effects for people who experienced marital and work transitions. Results showed that internet use was significantly associated with reduced depressive symptoms and loneliness for people who became divorced, but there was no significant effect for widowhood. For work transitions, using the internet did not show statistically significant associations with either depressive symptoms or loneliness. ❧ The present dissertation aims to extend the discussion of implications of internet use among older adults to recognition of the heterogeneity in the effect of internet use and highlight the need to develop various interventions specific to different environmental and individual characteristics of this diverse aging population. The dissertation also uses three different sources of longitudinal data to examine the effects of internet use comprehensively. The overall findings contribute to growing literature on internet use and aging by expanding our understanding on whether the implications of internet use for enhancing psychological well-being and social connectedness are applicable across different groups and countries. The final discussion addresses limitations and strengths of each study and presents future directions for further research on examining the heterogeneity in the effect of internet use on diverse groups of older adults to develop policies and programs that can effectively enhance psychological well-being and promote social integration of older people.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Age integration in late life: sociodemographic & psychosocial correlates of intergenerational-only, peer-only, and age-integrated social networks

PDF

Is stress exposure enough? Race/ethnic differences in the exposure and appraisal of chronic stressors among older adults

PDF

Social determinants of physiological health and mortality in China

PDF

Self-perceptions of Aging in the Context of Neighborhood and Their Interplay in Late-life Cognitive Health

PDF

Biomarkers of age-related health changes: associations with health outcomes and disparities

PDF

Examination of the long-term psychosocial and functioning effects of a healthy living, life management behavior intervention for older adults

PDF

A biodemographic approach to understanding sociodemographic disparities in kidney functioning on three dimensions: individual, population, and cross-national

PDF

International sex and age differences in physical function and disability

PDF

Life transitions, leisure activity engagement, and cognition among older adults

PDF

Estimating survival in the face of pain: evidence from the health and retirement study

PDF

Mechanisms of stress effects on learning and decision making in younger and older adults

PDF

Statistical algorithms for examining gene and environmental influences on human aging

PDF

Psychological and physiological pathways from social support exchanges to health: a lifespan perspective

PDF

Three essays on modifiable determinants of shingles: risk factors for shingles incidence and factors affecting timing of vaccine uptake

PDF

Children 's migration and the financial, social, and psychological well-being of older adults in rural China

PDF

Palliative care: breaking the knowledge barrier among community-based older adults

PDF

The Internet activities, gratifications, and health consequences related to compulsive Internet use

PDF

Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults

PDF

Family caregiver well-being: generational differences and coping strategies

PDF

Translating research into practice: a community-based medication management intervention

Asset Metadata

Creator Shim, Hyunju (author)

Core Title Internet communication use, psychological functioning and social connectedness at older ages

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Gerontology

Publication Date 07/25/2020

Defense Date 05/12/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag intergenerational characteristics,internet communication,life transitions,living arrangements,OAI-PMH Harvest,social support

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Crimmins, Eileen M. (committee chair), Ailshire, Jennifer A. (committee member), Zelinski, Elizabeth M. (committee member)

Creator Email hyunjush@usc.edu,simhyunju@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-345567

Unique identifier UC11665403

Identifier etd-ShimHyunju-8745.pdf (filename),usctheses-c89-345567 (legacy record id)

Legacy Identifier etd-ShimHyunju-8745.pdf

Dmrecord 345567

Document Type Dissertation

Rights Shim, Hyunju

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA