Staying connected in old age: associations between bonding social capital, loneliness and well-being and the value of digital media

Abstract

Objectives

We examined associations between age, bonding social capital (BSC), loneliness and psychological, social and emotional well-being in old age. As in theory digital media can support both preservation of and access to someone’s social capital, we also explored associations between the use of WhatsApp, social network sites (SNS) and Internet applications and respectively BSC and loneliness.

Method

In this cross-sectional study a sample of 349 older adults (range = 70y–93y; Mean(SD)=74.8y(4.9y); 59.6% male) filled out a questionnaire including the bonding social capital subscale of the personal social capital scale for elderly (PSCSE), validated scales measuring social, emotional, and psychological well-being and loneliness, the SNS intensity scale and items measuring frequencies of Whatsapp use and use of various Internet functions. Also relevant demographic and other covariates were included.

Results

Regression analyses including relevant covariates confirmed our hypotheses, finding negative associations between BSC and both age and loneliness, and positive associations between BSC and psychological, social and emotional well-being. WhatsApp and Internet use were both found positively associated with BSC, whilst a negative association between WhatsApp use and loneliness was found. SNS use was not associated with BSC nor with loneliness.

Conclusion

Our findings indicate BSC as an important factor in positive aging and illustrate a widening gap between the need for socio-emotional resources versus their availability. With regard to digital media, we conclude that its value in old age should be sought in providing access to one’s bonding social capital rather than adding to it by expanding the number of social relations.

Keywords:

• Social capital
• well-being
• loneliness
• elderly
• digital media

Introduction

Throughout the lifespan, people build a social network containing their personal social capital. Social capital has been found positively related to quality of life (Nilsson et al., 2006; Nyqvist et al., 2013), well-being (Biddle, 2012; Forsman et al., 2013; Magson et al., 2014) and related positive concepts such as positive affect, optimism, life satisfaction, trust and hope (Blazer, 2002; Gallagher & Lopez, 2009, Nyqvist et al., 2013), whilst negative associations were found with loneliness (Nyqvist et al., 2016; Santini et al., 2016). The current study aims to further investigate the role of social capital in old age (70+ years) and to explore the potential value of digital media as a means to help support social capital of older adults, as it seems likely to decrease with age (e.g., Wrzus et al., 2013).

Social Capital in later life

Social capital has been described by Putnam (1995, p. 67) as “features of social organization such as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit,” which is considered the most used approach of this concept in behavioral research literature (Forsman et al., 2013; Nyqvist et al., 2013). One’s social capital will ideally contain the necessary socio-emotional and socio-economic resources to successfully walk one’s life path. These resources can be found in two related sub-dimensions of social capital: bridging social capital (Woolcock, 1998, Putnam, 2000), comprising mostly larger networks of relatively weak and more business-like connections (Bourdieu & Wacquant, 1992; Putnam, 2000; Zhang et al., 2011) and bonding social capital (Putnam, 2000), concerning the more intimate relationships with family and friends in a close social circle, providing socio-emotional resources (Adler & Kwon, 2002; Liu et al., 2016; Putnam, 2000).

The still limited research of social capital in later life stages indicates bonding social capital (BSC) in particular as an important ingredient for well-being and positive aging processes (Simons et al., 2020). These empirical results can be explained by the socioemotional selectivity theory (Carstensen, 1992), claiming that as the remaining life time shortens, people become more selective in seeking out activities or social interactions that provide positive experiences or emotions in the present. As a result, older people tend to prefer social relationships with socioemotionally important others that add to positive experiences (Carstensen et al., 2003; 2006). Also, as life events in older age more often involve experiences of loss—such as a decrease in work or other activities, physical abilities and mobility, but also the passing of loved ones (Forsman et al., 2013; Wrzus et al., 2013) —the need for social support, which can be found in one’s bonding social capital, may increase (Machielse & Duyndam, 2020). However some of these same events (e.g., losing a loved one, less participation in work or other activities) may very well negatively affect one’s available bonding social capital (Wrzus et al., 2013). The limited research of this association between age and one’s social network in later life points in this direction, suggesting that daily social interaction is negatively associated with age (McDonald & Mair, 2010) and socio-emotional networks seem to shrink in older age (Cornwell et al. 2008; Kalmijn, 2003; Wrzus et al., 2013). Apart from a decrease in the actual number of social relations, reduced physical abilities due to old age related conditions or moving to a nursing home, may also hinder access to one’s social capital (Chipps & Jarvis, 2016).

Bonding social Capital, well-being and loneliness

Following these notions, a widening gap can be assumed between the need for socio-emotional resources or bonding social capital (BSC) versus its availability in old age. Considering the earlier mentioned research literature regarding the association between social capital and well-being related concepts as well as the importance of BSC in later life stages, this gap may negatively affect well-being of older adults and induce feelings of loneliness.

Well-being has been conceptualized in the research literature as consisting of three subdimensions (Keyes et al., 2008; Lamers et al., 2011). Emotional well-being concerns the presence of positive affect and one’s life satisfaction (Keyes, 2009). Social well-being can be described as someone’s functioning in their social community and society in general (Keyes, 1998). Effective functioning and self-realization are the key elements of psychological well-being (Ryff & Singer, 1998). It can be assumed that BSC is positively associated with each subdimension of well-being as it provides positive socioemotional relationships that help experience positive emotions and contains resources for autonomy and self-realization. It also facilitates feelings of belonging and a sense of coherence (Putnam, 2000).

An increase of feelings of loneliness, conceptualized as a perceived lack of both quantity and quality of relationships (De Jong Gierveld & Van Tilburg, 2006), also seems a relevant consequence of the earlier mentioned widening gap, as BSC is someone’s main resource of socio-emotional relationships and contacts. As loneliness has earlier been found negatively associated with well-being in older age (Dong et al., 2012; Golden et al., 2009), and might even have a mediating role in the association between bonding social capital and well-being (Simons et al., 2020), it seems vital for the well-being of older adults to maintain a certain degree of BSC and to support access to it.

Digital media use

The literature suggests that digital media offer opportunities for social interaction when mobility is limited (e.g., Chen and Schulz, 2016; Page-Tan, 2021; Simons et al., 2021) and in that capacity may serve as a means to provide access to one’s social capital or even help preserve it (e.g., Chen & Li, 2017; Ryan et al., 2017). Empirical research on such benefits of digital media use for older adults is still limited though and not yet conclusive. A review study by Chen and Schulz (2016) supports positive associations between digital media use and social support, social connectedness and social isolation of elderly, whilst no clear results were found for loneliness. A study by Hajek and König (2019) found that daily users of online social networks scored lower on social isolation than less or non- frequent users, whilst other studies indicate that online social networks do not necessarily help diminish feelings of loneliness (Aarts et al., 2015; Van Ingen et al., 2017).

The COVID-19 pandemic has recently shown us however, that digital media can indeed play a crucial role in the continuing of functioning of our society in unusual times when close physical proximity is severely restricted. It is likely to assume that during the COVID-19 pandemic, the use of digital media by older adults also increased, following a trend that started about a decade ago (CBS, 2019; Eurostat, 2019; König et al., 2018). In the Netherlands for instance, Internet use in general increased in the age group 75+ from 5.2% in 2012 to 39.8% in 2019 (CBS, 2019). Even though older generations are still considerably less inclined to use the available digital media (Embarak et al., 2021), regarding the possibilities they have to offer to support communication with others and overcome physical distance (Bell et al., 2013; Page-Tan, 2021), ; it is worthwhile to explore their potential value in old age.

The current study

The current study aims to further explore the importance of BSC in old age (70+ years) by examining associations between BSC, the three subdimensions of well-being and loneliness. Additionally, the potential value of several applications of digital media is explored, as in theory digital media offer possibilities to maintain access to one’s social capital when personal circumstances change and may also help form new relationships (Bell et al., 2013; Page-Tan, 2021; Simons et al., 2021).

Following research findings as earlier discussed we assume that bonding social capital (BSC) decreases with age (Hypothesis 1). Furthermore we assume a negative association between BSC and loneliness (Hypothesis 2) and positive associations between BSC and, respectively, social, emotional and psychological well-being (Hypothesis 3).

To explore the potential value of digital media, different applications were included supporting different online interactions: WhatsApp text messaging for one-to-one and one-to many interaction (WhatsApp Inc., 2020); Social Network Sites (SNS) for more (semi-)public interaction within an online community; and Internet services supporting interaction (email) and different regular activities such as shopping, making reservations, online banking and email correspondence. These applications are actually used by adults over 70 (CBS, 2019; Eurostat, 2019), allowing to explore their current potential for this age group. Associations between the use of these applications and both BSC and loneliness were explored.

Method

Procedure and sample

This quantitative cross-sectional study was carried out as part of a larger research program on positive aging and has been approved by the local research ethics committee (cETO, 2012; approval date December 2017). Inclusion criteria for participation in the overall research program were age ≥ 50 years, and sufficient command of the Dutch language to understand instructions and informed consent. As the current study focused on old age, only respondents of 70 and older (n = 349) of the larger sample (n = 913) were included. Demographics of our sample are further described in the Results section.

Data collection

Participants were invited to fill out a questionnaire that started with a consent page that, in accordance to the American Psychological Association Ethical principles of psychologists and code of conduct (American Psychological Association, 2010), explained that (1) participation in this study was voluntary and could be stopped at any time without reason or adverse consequences; (2) submitting the filled out questionnaire included the consent to the careful and secure anonymous use of the data for this study, in compliance with privacy rights.

Measurements

Demographic variables and covariates

The questionnaire started with several items concerning demographic variables and relevant covariates. In addition to gender and age, we included education (0 = high school/vocational education or less, 1= undergraduate degree or higher), living situation (0 = independently, 1 = in a nursing home/residing with someone) and financial resources to provide for oneself and (future) needed care (1= definitely enough, 2= probably enough, 3 = probably not enough, 4 = definitely not enough). Education and financial resources are associated with social capital (Han et al., 2014) and someone’s living situation can be of influence on social environment (Chipps & Jarvis, 2016). Furthermore we included covariates relationship (0 = no partner; 1= with partner) and perception of physical health (1 = very poor − 5 = excellent), as these were found positively associated with mental well-being (e.g., Cho et al., 2011; Hooghe & Vanhoutte, 2011), and are likely to undergo changes in older age.

Well-being

Well-being was measured with the Dutch Continuum Mental Health Short Form (MHC-SF, Lamers et al., 2011), consisting of three items measuring emotional well-being (e.g., “In the last month how often did you feel satisfied?”); six items measuring psychological well-being (e.g., “In the last month how often did you feel that you have experiences that challenge you to grow and become a better person?”); and five items measuring social well-being (e.g., In the last month how often did you feel that our society is becoming a better place for people"). These items were scored on a 6-point Likert scale, (1 = never; 2 = once or twice; 3 = about once a week; 4 = 2 or 3 times a week; 5 = almost every day; and 6 = every day). The mean score for each subscale was computed.

Bonding social capital (BSC)

BSC was measured with the subscale bonding social capital of the Personal Social Capital Scale for the Elderly (PSCSE, Simons et al., 2020), consisting of five categories of statements. The first four categories each include a statement—respectively (1) “I have ……. family members/relatives”; (2) “I keep a routine contact with ……. family members/relatives”; (3) “I have ……. family members/relatives that I can trust”; and (4) “I can ask ……. family members/relatives for help” —that was repeatedly scored for four different social groups (family members/relatives; close friends; acquaintances; others) on a 5-point Likert scale (5 = many; 4 = reasonably many; 3 = some; 2 = few; 1 = none). The fifth category consists of five statements addressing someone’s access to certain resources via personal social networks (e.g., “I know ……. people with certain political or other influential power” or “I know ……. people with broad social connections”), scored on the same 5-point Likert scale. BSC was then calculated by adding the mean scores of each category.

Loneliness

The 6-item Loneliness scale of De Jong Gierveld and Van Tilburg (2006) was used to measure someone’s experienced degree of loneliness (e.g., “I miss having people around me” and “There are many people I can trust completely”—reverse coded). Respondents could answer to these statements with either (1) “no,” (2) “more or less,” or (3) “yes.” Loneliness was then computed by adding the scores on each item.

Digital media use

SNS use was measured by the Facebook intensity scale (Ellison et al., 2007) that we applied to any social network site the respondents might use. This scale consists of six items measuring emotional connectedness (i.e., “SNS have become part of my daily routine”, and “I feel part of a social media community”) scored on a 5-point Likert scale (1 strongly disagree—5 strongly agree) and two items respectively measuring the duration of use (less than 10 min per day, 10–30 min per day, 31–60 min per day, 1–2 h per day, 2–3 h per day, more than 3 h per day) and the quantity of friends or connections in their most used SNS (1–10, 11–50, 51–100, 101–150, 151–200, 201–300, 301–400, 401–600, 601–800, 801–1000, 1001+). The overall score for SNS use was then computed by adding standardized scores of all eight items. Additionally we asked respondents to indicate how frequently (1 = daily, 2 = weekly, 3 = monthly, or 4 = less) they looked at SNS, posted themselves or reacted to posts of others.

Internet use was measured by the sum of functions of Internet (making reservations, ordering food/dinner, ordering groceries; online shopping; online banking; contact with health care workers; playing games; email; and making new contacts) that respondents indicated to use (no = 0, yes = 1).

WhatsApp use was measured by adding up the scores on two items measuring the frequency of WhatsApp use for social contact: (1) “How often do you use WhatsApp socially/for fun, for example to keep in touch or to inform how someone is doing?” and (2) “How often do you use WhatsApp for practical matters, for instance to organize a gathering, or arrange something?” (0= (almost) never; 1 = once/a few times every three months; 2 = once/a few times every month; 3 = once a week; 4= a few times a week; 5 = every day; 6 = a few times a day; 7 = five or more times a day).

Analyses

Statistical analyses were performed with SPSS (V24, IBM Corp.NY). Reliability of scales (Cronbach’s alpha) was examined, (Pearson’s) correlations between main study variables were computed, and frequencies of use of media applications were described.

Addressing Hypothesis 1 associations between age (independent variable) and BSC (dependent variable) were examined with multiple regression analysis. Hypotheses 2 and 3 were tested by multiple regression analyses, examining the association between BSC (independent variable) and respectively loneliness and the three subscales of well-being (dependent variables).

To explore the association between the digital media applications (independent variables) and respectively BSC and loneliness (dependent variables), separate multiple regression analyses were performed for each application as they were used by a different subsample. In addition, all three applications were explored in one regression model (for both BSC and loneliness) among respondents using all three applications.

All regression analyses used method Enter, stepwise and included the a priori defined covariates gender, age, physical health, relationship, education level and financial resources.

Results

One respondent had to be removed because of missing data on the well-being and loneliness scales. Another respondent failed to fill out the loneliness scale and was given the mean score of loneliness of the total sample. With regard to missing demographic data, listwise deletion was used. This resulted in deletion of four cases in the regression analyses (missing data for education, see Table 1).

Table 1. Demographic variables.

Age M(SD)		74.76(4.88), range 70-93
Gender	Female	40.4% (n = 141)
	Male	59.6% (n = 208)
Relationship	No	28.1% (n = 98)
	Yes	71.9% (n = 251)
Residential situation	Independent in own home	97.1% (n = 339)
	Independent with care in own home	2% (n = 7)
	Living in with relative or in nursing home	.6% (n = 2)
	Missing	.3% (n = 1)
Education	Low	52.4% (n = 183)
	High	46.4% (n = 162)
	Missing	1.1% (n = 4)
Health (self-report)	Poor	.9% (n = 3)
	Moderate	11.5% (n = 40)
	Good	43% (n = 150)
	Very good	29.8% (n = 104)
	Excellent	14.9% (n = 52)
Financial resources	Maybe/maybe not enough	41.5% (n = 145)
	Definitely enough	58.5% (n = 204)

Note. n = 349.

Descriptives

Most respondents were in a relationship (n = 251; 71.9%), living independently in their own home (n = 339; 97.1%), and experiencing a good to excellent (n = 306; 87.7%) physical health. There were more male (n = 208; 59.6%) than female respondents, whilst level of education was almost evenly distributed. A small majority (n = 204; 58.5%) was not experiencing any financial distress. Age was skewed with an average of almost 75 years.

Table 2 presents reliability scores, other relevant statistics and correlations of core variables. As expected social, psychological and emotional well-being intercorrelate as subscales of the overall well-being scale MHC-SF. Positive correlations were found between BSC and the well-being subscales and negative correlations between BSC and respectively age and loneliness. Internet use and WhatsApp use correlate positively with BSC and negatively with both age and loneliness. Use of SNS correlates negatively with emotional well-being.

Table 2. Descriptives, reliability and correlations of main study variables.

	α	Mean	SD	Min	Max	Pearson’s correlation
n = 349						1	2	3	4	5	6	7	8	9
1. BSC	.88	15.54	2.86	7.65	24	1
2. Age		74.76	4.88	70	93	−.16**	1
3. SWB	.75	3.53	1.04	1.00	6.00	.41***	−.13*	1
4. PWB	.84	4.28	1.02	1.00	6.00	.34***	−.03	.70***	1
5. EWB	.83	4.88	.89	1.67	6.00	.26***	−.14*	.46***	.56***	1
6. Loneliness	.76	1.33	.39	1.00	2.67	−.51***	.02	−.33***	−.37***	−.37***	1
7. SNS^1		.00	5.87	−10.37	13.07	.05	−.02	.00	.01	−.17*	.12	1
8. Internet^2		6.01	1.71	0	9	.13*	−.17**	.09	.09	.10	−.13*	.02	1
9. WhatsApp^3		.34	1.36	−2.94	4.16	.25***	−.18**	.03	.04	.10	−.15**	.03	.26***	1

Note. SWB = social well-being; PWB = psychological well-being; EWB = emotional well-being; ¹n = 165, computed by adding standardized scores of SNS use statements (α=.88), time using SNS and contacts on SNS; ²n = 343; ³n= 273; *p<.05, **p<.01, ^***p<.001.

Those who used Internet (n = 343) mostly used it for sending/receiving emails (n = 340; 99.1%), online banking (n = 323; 94.2%), online shopping (n = 295; 86.0%) and making reservations (dinner, theater etc.) (n = 291; 84.8%). WhatsApp (n = 273) was used for both practical and social purposes comparably frequent (resp. 30.1% (n = 84) and 32.7% (n = 89) indicated to use WhatsApp only once a week or less; 32.7% (n = 89) and 35.2% (n = 96) several times per week; and 33.3% (n = 91) and 28.4% (n = 78) daily). Table 3 presents the reported frequencies of different actions on SNS, indicating that the majority of our sample rather looks at or reacts to other posts, than posting themselves.

Table 3. Frequency of actions in SNS use.

	Daily	Weekly	Monthly	less	missing
Look	61.2% (101)	19.4% (32)	6.1% (10)	6.1% (10)	0.6% (1)
Post	7.9% (13)	23% (38)	16.4% (27)	45.5% (75)	0.6% (1)
React to posts of friends	27.9% (46)	32.7% (54)	12.7% (21)	19.4% (32)	0.6% (1)
React to posts of others	5.5% (9)	10.9% (18)	6.1% (10)	70.3% (116)	0.6% (1)

Note. SNS users (n = 165).

BSC, loneliness and well-being

Regression analysis showed a negative association between age and BSC (dependent variable), (F=(6,338) = 4.379, p<.001; Adj R²=.06, ΔR²=.01, β=-.12, p=.04). This is in line with Hypothesis 1, indicating that older adults in our sample possess less BSC. Of the included covariates, health and financial resources were both positively associated with BSC (resp. β=.13, p=.04;; β=.14, p=.01).

In line with Hypothesis 2, a negative association was found between BSC and loneliness (Hypothesis 2) (dependent variable), (F=(7,337) = 21.59, p<.001; Adj R²=.30, ΔR²=.25, β=-.52, p<.001), indicating that older adults with a higher score on BSC experienced less feelings of loneliness. In this model health and age were also negatively associated with loneliness (resp. β=-.11, p=.02; β=-.11, p=.02), although earlier no correlation was found between age and loneliness (see Table 2).

Table 4 presents the results of regression analysis to test associations between BSC and the subscales of well-being (Hypothesis 3). Positive associations were found with respectively emotional (β=.24, p<.001), psychological (β=.34, p<.001), and social wellbeing (β=.42, p<.001), showing that in our sample adults with higher scores on BSC also had higher scores on each subdimension of well-being. Health was positively associated with emotional and psychological wellbeing; gender and having a partner were negatively associated with psychological and social wellbeing, indicating that the male respondents and respondents without a partner experience higher levels of these sub-dimensions of wellbeing.

Table 4. Results of regression analysis of association between BSC and subscales of well-being.

Dependent variable	Emotional well-being				Psychological well-being			Social well-being
Predictor	B(SE)	95% CI		B(SE)		95% CI		B(SE)	95% CI
		LL	UL			LL	UL		LL	UL
Age	−.01(.01)	−.03	.01	,01(.01)		−.01	.03	−.01(.01)	−.03	.01
Gender	.22(.11)	−.01	.44	.39***(.11)		.17	.60	.33**(.11)	.12	.54
Partner	−.11(.12)	−.35	.13	−.36**(.12)		−.59	.13	−.29*(.12)	−.51	−.06
Health	.23***(.06)	.11	.35	.14*(.06)		.02	.25	.06(.06)	−.06	.17
Education	−.15(.10)	−.36	.05	−.07(.10)		−.27	.13	.11(.10)	−.08	.31
financial resources	−.02(.11)	−.23	.19	−.10(.10)		−.30	.11	−.19(.10)	−.39	.01
BSC	.24***(.05)	.14	.35	.35***(.05)		.24	.45	.42***(.05)	.32	.52
	Adj. R^2 = .12 F(7,337)=7.771, p<.001				Adj. R^2 = .16 F(7,337)=10.20, p<.001			Adj. R^2 = 20. F(7,337)=13.34, p<.001

Note. Model: Enter, *p<.05, **p<.01. Gender (0 = female, 1 = male); partner (0 = no, 1 = yes); education (0 = low; 1 = high); financial resources (0 = maybe/maybe not enough; 1 = definitely enough); Standardized scores were included for emotional well-being, psychological well-being, social well-being, health, age and BSC.

Exploring associations with digital media

Associations between the digital media applications and respectively BSC and loneliness were explored in separate regression models (Internet, n = 343; WhatsApp, n = 273; SNS, n = 165). Table 5 presents the results of the significant models WhatsApp use was positively associated with BSC (β=.25, p<.001). Both WhatsApp use (β=-.19, p=.002) and Internet use (β=-.13, p=.020) were negatively associated with loneliness. SNS use was not associated with BSC, nor with loneliness. These results indicate that older adults in our sample who used more Internet services experienced less feelings of loneliness. More frequent Whatsapp users not only experienced less feelings of loneliness, but also scored higher on BSC.

Table 5. Results of regression analysis of association between digital media use and respectively BSC and loneliness.

Dependent variable	BSC				Loneliness			Loneliness
Predictor	B(SE)	95%CI		B(SE)		95%CI		B(SE)	95%CI
		LL	UL			LL	UL		LL	UL
Age	−.01(.01)	−.04	.02	−.02(.02)		−.05	.01	−.02(.01)	−.04	.01
Gender	−.12(.12)	−.29	.25	−.14(.13)		−.40	.12	−.05(.12)	−.28	.18
Partner	−.02(.14)	−.29	.25	−.09(.15)		−.38	.19	−.17(.13)	−.42	.08
Health	.11(.07)	−.02	.25	−.20**(.07)		−.35	−.06	−.19**(.06)	−.31	−.07
Education	.04(.11)	−.18	.26	.08(.12)		−.16	.31	.10(.11)	−.12	.31
Financial resources	.29*(.11)	.06	.51	−.17(.12)		−.41	.07	−.15(.11)	−.37	.06
WhatsApp use	.35***(.08)	.19	.52	−.28**(.09)		−.45	−.10
Internet use								−.13*(.05)	−.23	−.02
	Adj. R^2 =.09 F(7,262)=4.95, p<.001				Adj. R^2 = .07 F(7,262)=3.86, p=.001			Adj. R^2 = .04 F(7,331)=3.13, p=.002

Note. Model: Enter, ^*p<.05, ^**p<.01, ^***p<.001; Gender (0 = female, 1 = male); partner (0 = no, 1 = yes); education (0 = low; 1 = high); financial resources (0 = maybe/maybe not enough; 1 = definitely enough); Standardized scores were included for emotional well-being, psychological well-being, social well-being, health, age and BSC.

Associations between use of each application and BSC and loneliness were also explored in one model, including respondents that used all three applications (n = 138, Mean(SD)_age =73,60(4.23), range 70–90). The regression model with BSC as dependent variable (F(9,127) =2.66, p=.007; Adj R²=.10) found a positive association with WhatsApp use (β=.26, p=.003) and Internet use (β=.20, p=.020) . In the regression model with loneliness as dependent variable (F(9,127) =3.04, p=.003; Adj R²=.12) both WhatsApp and Internet use were negatively associated with loneliness (resp. β=-.19, p=.04; β=-.19, p=.03). In both models financial resources were found to be a significant covariate (resp. β=−.17, p=.044, β=−.19, p=.025). SNS use was not found a significant factor in either model. These results are largely in line with the results from the separate regression models, indicating that in this subsample more frequent WhatsApp use and use of more Internet services are associated with less feelings of loneliness and higher scores on BSC, while SNS use does not appear to be associated with either.

Discussion

We examined the associations between BSC and respectively loneliness and mental well-being in old age and explored the potential value of digital media in this stage of life. We assumed that BSC decreases in old age (Wrzus et al., 2013), whilst the need for socio-emotional resources increases (Machielse & Duyndam, 2020), resulting in a gap possibly affecting wellbeing in later life. Indeed we found a negative association between age and BSC, supporting our first hypothesis, although it must be acknowledged that the variance in BSC explained by age was relatively small. The amount of overall social capital (both bonding and bridging) built during the lifespan—which is among other things associated with level of education and financial resources (Han et al., 2014; Zhang et al., 2011) and someone’s history of social participation or employment (Engbers et al., 2017) —can be assumed to explain part of the variance in BSC in later life stages as well.

Furthermore, a negative association between BSC and loneliness was found (Hypothesis 2) as well as positive associations between BSC and all three subdimensions of well-being, indicating that the availability of socioemotional resources is associated with respectively greater positive affect and life satisfaction, more effective functioning and self-realization, and better social functioning (Hypothesis 3). These results confirm earlier research findings regarding the relationship between BSC and respectively well-being (Biddle, 2012; Forsman et al., 2013; Magson et al., 2014) and loneliness (Nyqvist et al., 2016; Santini et al., 2016), and emphasise the relevance of our exploration of the potential value of digital media as a means to preserve BSC and maintain access to it.

Our exploratory analyses of the association between digital media and BSC resulted in positive associations with respectively WhatsApp and Internet use. We also found a negative association between WhatsApp use and loneliness. No significant associations between BSC or loneliness and SNS use were found in our sample.

To help understand and interpret these associations with respect to the potential value of digital media for BSC in old age, we looked at the purposes for which the included digital media applications were used by our sample. WhatsApp was measured as the frequency of use for either social contact and fun or practical matters. For both types of use comparable frequencies were reported. As WhatsApp is an instant messaging application, mostly used for communication with known others (WhatsApp Inc., 2020), it can be assumed that the potential value of this use can be found in facilitating access to BSC rather than preserving or elaborating it. With regard to Internet applications, the vast majority of our sample reported email use, which also supports social interaction with current relations rather than initiating new contacts. Other applications that were used had a more practical character, such as online banking and shopping.

SNS, which use was not found to be a significant factor, are actually a suitable means for making new contacts, as they allow for interaction and sharing information with both known and unknown others, as well as access to what (un)known others either publicly or privately share. Our sample however, was rather passive and reactive in their use of SNS, as they mostly looked or reacted to posts of friends. Still, evidence of positive associations between SNS use and social capital is growing (Ryan et al., 2017) and recent research findings indicate positive associations between SNS use and social capital in the second half of life (Hajek & König, 2019) and in an age group of a wider age range (>50) than our sample (Simons et al., 2021). Explained variances in the latter study were small though, and a comparable passive and reactive use of SNS was reported, which may be characteristic for current older age groups (Waycott et al., 2013).

In summary, our results indicate a negative association between age and BSC, supporting the theoretical assumption of a widening gap between need and availability of BSC in old age. Considering the found associations between BSC and respectively loneliness and well-being in the current study as well as earlier research (e.g., Biddle, 2012; Magson et al., 2014; Nyqvist et al., 2016; Santini et al., 2016), this gap can eventually affect well-being in this age group. Digital media help to bridge physical distance and have proven to enable online social interaction when physical social interaction becomes less obvious. Our results show significant associations between digital media applications that support interaction within one’s current social network, and both BSC and loneliness. This indicates that in our sample, the potential value of digital media should be sought in providing access to one’s BSC, rather than adding to it by expanding the number of social relations.

Limitations

As we conducted a cross-sectional study, no conclusions about causality can be drawn. It is therefore quite possible that the found associations have a reverse direction or are at least reciprocal. Having a larger social network for instance will probably involve more social interaction and may consequently involve more use of digital media as well. However, this does not necessarily make our results less interesting as these developed digital skills may still help to maintain access to one’s social capital when circumstances change and physical interaction becomes more difficult.

Although no correlation was found between age and loneliness, the regression model examining the association between BSC and loneliness indicated age as a significant (negative) variable. In a post-hoc regression analysis examining this association, however, no significant association between age and loneliness was found. Apparently, the decline in BSC with age in our sample does not manifest itself in a positive association between age and loneliness. This might be explained by the earlier mentioned socioemotional selectivity theory (Carstensen, 1992; Carstensen et al., 2006), which claims that in later life people become more selective in their social contacts, focusing on those adding to positive experiences. This may result into less social relations, but will not necessarily induce feelings of loneliness (English & Carstensen, 2014; Nicolaisen & Thorsen, 2017). Given physical health was reported as good to excellent by a large majority of our sample, who also largely live independently in their own homes, our respondents may not yet have been confronted much with the discussed life events that can affect their social capital or access to it. This might indicate that in our sample the decline in BSC is rather related to a selective attitude toward social contacts, than to changing conditions with regard to for instance physical health and housing. This is possibly due to the skewed distribution of age. While this fits the distribution of age of the general population of 70+ (CBS, 2020), the current study would have benefited from a sample including more old-old respondents.

One covariate, having a life partner, acted differently from what we expected and was found negatively associated with psychological and social well-being. This might be explained by the assumption that in this age group a life partner is more likely to suffer from illness or impaired abilities due to aging, which can have an impact on a person’s quality of life (Rees et al., 2001). However, our data do not provide information to further explore this assumption.

We used a validated scale to measure SNS use, whose reliability and validity have been shown to be satisfactory in several studies (e.g., Jenkins-Guarnieri et al., 2013; Orosz et al., 2016), but was also criticized for lack of formal systematic validation (Sigerson & Cheng, 2018). Moreover, as the three included applications have different functionalities and purposes, the measurements of use also differ. The scale for SNS use for instance included emotional connectedness, while WhatsApp use and Internet were measured by respectively frequency of use and the number of used services. This complicates assessing or interpreting the results of the three applications in relation to each other. In addition, as there was no widely used scale available of Internet and WhatsApp (text messaging) usage for older adults, we created a fairly simple measure that was both easy to understand for our sample and sufficiently informative for our exploratory research question. However it must be taken account of that these measures therefor give a still quite simplified presentation and must be interpreted with caution in our analysis.

Finally, we only included respondents that indeed use digital media and our data was, even though a paper questionnaire was available, mostly collected via an online questionnaire. This approach may have led to the earlier mentioned rather still capable, healthy and independently living group of respondents, and did not allow a comparison between users and non-users, which could have supplemented our search for the benefits of using digital media for older adults.

Practical and theoretical implications

By our search for means to help older adults preserve their BSC and consequently their well-being, we have contributed to the growing body of research of positive aging (Hill, 2011) and the limited research literature on the value of information and communication technologies for supporting social capital in old age (Barbosa Neves et al., 2018). As noted earlier in the Discussion section, our results confirm other research findings concerning the association between BSC and respectively loneliness and well-being in old age. The association between BSC and loneliness in later life stages however, may manifest itself differently or less strongly, if a decline in BSC is mainly the result of a more selective attitude toward social contacts. This notion is important to consider in further research of BSC in relation to positive aging.

Although it was explained that in theory digital media are able to meet several challenges characteristic for old age, which may threaten social capital in later life (Chen & Li, 2017; Utz & Muscanell, 2015), practice is more unruly and older generations can be reluctant to use new technologies for various reasons. From a lifespan perspective, it can be argued that since these technologies were introduced only during the second half of their lives, they were not incorporated in their daily routines as is the case for younger generations, growing up in an already digitalized society. This may be illustrated by the passive and reactive use of for instance SNS, and our finding that digital media were merely used to interact with known others and not to initiate new relationships. However considering the earlier mentioned notion of older adults being more selective toward their social contacts, this type of use can still be of value.

Older adults may also experience difficulties in learning how to use new technologies (Embarak et al., 2021). To find a better match between theory and practice and benefit from the support digital media have to offer, educational programs for older adults to become more familiar with the use of digital media (Embarak et al., 2021) can be helpful. Considering the results of the current study, the focus of these programs should be on the use of digital media for sufficiently “rich” social interaction with important others, when physical contact is limited. This can help older adults to maintain access to their BSC and the socioemotional resources it contains. In addition to educating older adults, adapting the media applications, especially the interfaces, more to their specific needs, can also be effective (e.g., Goumopoulos et al., 2017).

Conclusion

Old age comes with new challenges that involve the loss of certain skills and abilities as well as (contact with) friends and family members. This can negatively affect one’s bonding social capital and consequently mental well-being. In theory, digital media and online societies can address some of these challenges and be helpful in bridging physical distance and accessing important others. By educating current and future older generations how to benefit from these technologies in their daily activities and social interactions and developing user friendly interfaces that meet their needs, this theory can become practice that can actually add to or help preserve one’s bonding social capital and well-being in old age.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Funding

The author(s) reported there is no funding associated with the work featured in this article.