Skip Navigation
Skip to contents

JPMPH : Journal of Preventive Medicine and Public Health

OPEN ACCESS
SEARCH
Search

Articles

Page Path
HOME > J Prev Med Public Health > Volume 59(5); 2026 > Article
Original Article
Social Capital in Digitally Mediated Communities: Conditional Associations With Depressive Mood in Korea
Kyungmin Baekorcid
Journal of Preventive Medicine and Public Health 2026;59(5):489-497.
DOI: https://doi.org/10.3961/jpmph.26.305
Published online: July 13, 2026
  • 1,020 Views
  • 37 Download

Information Sociology, Soongsil University, Seoul, Korea

Corresponding author: Kyungmin Baek, Information Sociology, Soongsil University, 369 Sangdo-ro, Dongjak-gu, Seoul 06978, Korea, E-mail: kbaek37@ssu.ac.kr
• Received: April 2, 2026   • Revised: May 1, 2026   • Accepted: June 5, 2026

Copyright © 2026 The Korean Society for Preventive Medicine

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

prev next
  • Objectives
    This study examined whether engagement with different online community environments moderates the association between social capital and depressive mood.
  • Methods
    Using the 2023 Korea Social Integration Survey (n=8221), a nationally representative survey conducted by the Korea Institute of Public Administration, social capital was measured as an index of participation in 9 types of voluntary and community organizations, and depressive mood was assessed using a single-item measure of day-to-day emotional distress. Ordinary least squares regression models with interaction terms were used to determine whether the association between social capital and depressive mood varied across 3 types of online community environments: messenger services, social networking services (SNS), and membership-based online communities.
  • Results
    Higher levels of social capital were associated with higher levels of depressive mood (b=0.327), contrary to the conventional assumption that social participation is uniformly beneficial. Interaction analyses indicated that this association varied across online community environments. Frequent messenger service use was associated with a weaker positive association between social capital and depressive mood (b=−0.084), whereas SNS use showed a smaller and more conditional moderating effect (b=−0.042). Use of membership-based online communities was not a statistically significant moderator.
  • Conclusions
    The mental health implications of social participation depend on how social engagement is organized across offline and online contexts. Relationally oriented digital environments, particularly messenger services, may attenuate the emotional burden associated with extensive social participation. These findings underscore the need to consider digitally mediated community environments in public mental health research and interventions.
In public health and social epidemiology, social capital has long been studied as a determinant of mental health operating through psychosocial pathways that link social structure to individual well-being [1,2]. A substantial body of research has documented a robust association between social capital and mental health, generally emphasizing its protective role through social support, trust, and a sense of belonging [1–4]. Social capital has typically been conceptualized as a collective community resource that enables individuals to access emotional, informational, and normative support through participation in social networks [5] and buffers the effects of stressors on mental health [6,7]. From this perspective, higher levels of social capital are widely assumed to promote psychological well-being and protect against emotional distress.
However, this assumption rests on the premise that social participation primarily generates supportive exchanges rather than social obligations. Less attention has been paid in public health to the parallel literature on the negative dimensions of social capital, which Portes [8] termed the “downsides” of social embeddedness. Participation in multiple social groups often entails expectations of availability, responsiveness, and conformity, which can impose psychological strain and emotional burden. Classic sociological research has long suggested that overlapping social networks expose individuals to competing role expectations across intersecting relationships, potentially increasing rather than alleviating stress [9–11]. From a social-epidemiological perspective, sustained obligations and overlapping role demands impose chronic psychosocial demands that require repeated mobilization of emotional and cognitive resources without adequate replenishment, leading to resource depletion and sustained stress exposure—a recognized pathway to depressive symptomatology [2,12,13]. Accordingly, the mental health consequences of social capital are not uniform but depend on how social participation is organized and experienced.
In Korea, social capital research has generally reported protective associations of interpersonal trust and informal social participation with depression [14,15], whereas evidence on formal organizational participation has been less consistent. Lee et al. [16], using community health survey data from a single province, found that informal participation and trust were associated with lower depression, whereas formal organizational participation (e.g., political parties, religious organizations, and civic associations) showed no significant protective association. The present study extends this line of work by using 2023 nationally representative data and measuring formal participation across 9 organizational types on a 5-point intensity scale, an operationalization better suited to capturing the cumulative role demands of sustained engagement across multiple settings.
Recent social changes further complicate this relationship. Social participation that traditionally occurred in offline community settings is increasingly mediated through online environments [17–19]. Messaging services (e.g., WhatsApp and KakaoTalk), social networking platforms (e.g., Facebook and Instagram), and membership-based online communities (e.g., Reddit and NAVER Café) have become integral contexts in which social relationships are maintained, coordinated, and experienced. These online environments differ in relational structure, interaction intensity, and expectations of engagement [19–21]. Rather than simply extending offline communities, online community environments may reorganize how social participation is enacted and how its emotional consequences are experienced. For example, obligations that once required face-to-face coordination and immediate responsiveness may be redistributed through asynchronous messaging or platform-based interaction, altering how social demands are perceived and managed.
This perspective raises an underexplored question about whether online community environments alter the mental health implications of social participation. Although prior studies have examined either the mental health effects of social capital or the psychological implications of digital media use, less attention has been paid to how online community environments condition the relationship between social capital and mental health. In particular, it remains unclear whether engagement with online community environments amplifies, mitigates, or otherwise reshapes the emotional consequences of social participation.
To address this gap, this study examines whether and how the association between social capital and depressive mood varies across different online community environments. Using nationally representative survey data from Korea, the study investigates the moderating roles of messenger services, social networking services (SNS), and membership-based online communities. It thereby contributes to a more nuanced understanding of social participation as a conditional social determinant of mental health in digitally mediated societies.
Theoretical Background
Prior research has generally viewed social capital as beneficial for mental health because it facilitates social support, trust, and a sense of belonging [1,3]. However, several studies caution that social participation also entails obligations and expectations that may generate psychological strain, suggesting that the mental health effects of social capital are not uniform but context-dependent [10,22].
Classic sociological studies further clarify this mechanism by emphasizing overlapping social networks. When individuals occupy multiple, intersecting social roles, they are exposed to competing expectations that may intensify role-related strain rather than alleviate stress [9,11,23]. Research on network multiplexity similarly shows that dense and overlapping ties can increase normative pressure and emotional demands, thereby transforming social embeddedness into a potential source of psychological burden [24]. Taken together, these perspectives suggest that the mental health effects of social capital depend not only on its volume but also on the organizational contexts in which social relationships are embedded.
Recent social changes have added a new dimension to this relationship. Social participation that was once primarily rooted in offline community settings is now increasingly organized and experienced through online environments. Messaging services, social networking platforms, and membership-based online communities have become key infrastructures through which social relationships are maintained, coordinated, and sustained. These online community environments differ markedly in their relational structures.
Messenger services and membership-based online communities tend to involve repeated interaction, bounded membership, and clear expectations of responsiveness, closely mirroring the relational intensity of offline groups. In contrast, SNS are characterized by more diffuse ties, lower interactional obligation, and less clearly defined social boundaries. These structural differences suggest that online environments may shape how individuals manage the demands and obligations associated with social capital, thereby conditioning whether social participation amplifies or attenuates emotional distress [20,25].
In relationally intensive environments such as messenger services and membership-based online communities, social participation is often embedded in stable, bounded interactional settings. These platforms facilitate ongoing coordination, routine communication, and predictable patterns of engagement, which may distribute social demands over time and reduce the immediacy of role-related pressures. As a result, the emotional burden associated with extensive social participation may become more manageable, attenuating the extent to which social capital translates into day-to-day emotional distress.
By contrast, SNS typically involve more diffuse and loosely bounded networks, in which social ties are less clearly defined and expectations of engagement are more ambiguous. In such contexts, individuals with lower levels of social capital may use SNS as a compensatory resource for social connection and visibility, potentially buffering emotional distress. However, as social capital increases, the relative contribution of SNS to managing social demands may diminish, offering limited capacity to reorganize the obligations and expectations associated with extensive social participation. Consequently, SNS may have a weaker, more context-dependent moderating influence on the relationship between social capital and mental health [17,20,26].
Building on this perspective, the present study examines whether the association between social capital and depressive mood varies across online community environments, with particular attention to whether relationally intensive and diffuse platforms differentially condition the mental health consequences of social participation. Based on this discussion, the study proposes the following hypotheses.

Hypothesis 1. Higher levels of social capital are associated with higher levels of depressive mood

Because organized social activities may entail greater social obligations and role-related strain, greater involvement in these activities may coincide with higher levels of day-to-day emotional distress.

Hypothesis 2. The positive association between social capital and depressive mood is weaker among individuals who more frequently use online community environments

Online environments that facilitate continuous and direct interaction may reorganize how social participation is experienced, thereby attenuating the emotional burden associated with social capital.
Data
The hypotheses were tested using data from the Korea Social Integration Survey. These data were collected by the Korea Institute of Public Administration, a government-funded policy research organization. The survey aimed to identify changes in the public’s perceptions of and attitudes toward social integration, with the goal of reducing conflict in Korean society and contributing to social integration. It included items assessing socioeconomic status, political orientation, social media use—including messenger apps, video platforms, and SNS—and various dimensions of subjective well-being. This study used data collected in 2023 from 8221 respondents.
Measures

Depressive mood

The outcome variable was self-reported depressive mood, measured using a single item that asked respondents how depressed they felt on the previous day. Responses were coded so that higher values indicated higher levels of depressive mood. This measure captures day-to-day emotional distress rather than clinically diagnosed depression and is commonly used in population-based mental health research [27,28].

Social capital

Social capital was measured as an index of participation in multiple voluntary and community organizations [3,29]. Respondents reported their level of involvement across 9 types of organized social activities: political parties; labor unions, business, or professional associations; religious organizations; hobby or leisure clubs; civic organizations; local community or neighborhood groups; alumni or hometown associations; volunteer or charitable organizations; and social economy organizations, such as social enterprises and cooperatives. For each activity, responses were measured on a 5-point Likert scale ranging from no membership to active participation, with higher scores indicating greater involvement. The social capital index was constructed as the mean of these items. This measure captures the structural dimension of social capital by reflecting individuals’ embeddedness in organized social networks (Cronbach’s alpha=0.77).

Online community context

Online community contexts were assessed according to the frequency of use of 3 types of digitally mediated social platforms: messenger services, SNS, and membership-based online communities (e.g., NAVER Café). For each platform, higher values indicated more frequent use.

Covariates

All models controlled for socio-demographic and socioeconomic factors known to be associated with mental health. Sex was coded 1 for female and 0 for male. Age was available only as a 5-category ordinal variable (1=19–29, 2=30–39, 3=40–49, 4=50–59, 5=60 years or older). Educational attainment was recoded from the original 8 response categories into a 4-category ordinal variable (1=elementary school or less, 2=middle school, 3=high school, 4=college or higher, including graduate education). Monthly pre-tax household income was recoded from the original 12 categories into a 7-category ordinal variable (million Korean won [KRW]: 1=less than 1, including no income; 2=1 to less than 2; 3=2 to less than 3; 4=3 to less than 4; 5=4 to less than 5; 6=5 to less than 6; 7=6 or more). Self-rated health was rated on a 5-point scale (1=very poor to 5=very good), and general trust was rated on a 4-point scale (1=cannot be trusted at all to 4=can be trusted very much). Perceived social status and life satisfaction were each rated on an 11-point scale from 0 to 10, with higher values indicating higher perceived status and greater satisfaction, respectively. Religious affiliation (1=has a religion), spousal status (1=has a spouse), and employment status (1=employed) were coded as binary indicators. Age, education, and household income were entered into the models as continuous covariates, so their coefficients represent the average change in depressive mood associated with moving to the next-higher category. Descriptive statistics for all variables included in the analysis are presented in Table 1.
Statistical Analysis
Ordinary least squares (OLS) regression was used to examine the association between social capital and depressive mood and to test whether this association was moderated by engagement with online community environments. Four models were estimated. Model 1 included social capital and all covariates. Models 2 to 4 added the 3 online community context variables and, in turn, an interaction term between social capital and messenger service use (model 2), social networking service use (model 3), and membership-based online community use (model 4).
Ethics Statement
This study used secondary analysis of publicly available data collected by the Korea Institute of Public Administration. Institutional review board approval was not required because the study used publicly available data without identifiable personal information.
OLS regression analyses were conducted to examine the association between social capital and depressive mood and to assess whether this association varied across digital media exposure contexts. Table 2 presents the results of 4 models.
Model 1 reports the baseline OLS regression model without interaction terms. In this model, higher levels of social capital were significantly associated with higher levels of depressive mood (b=0.327) after adjustment for sex, age, educational attainment, self-rated health, perceived social status, household income, religious affiliation, spousal status, general trust, life satisfaction, and employment status. This positive association suggests that greater involvement in organized social activities may be accompanied by increased social demands, obligations, or role-related strain, reflected in higher levels of day-to-day emotional distress [22,30]. This finding supports hypothesis 1.
Models 2–4 introduced interaction terms between social capital and different digital media exposure contexts. Model 2 examined the moderating role of messenger service use. The interaction between social capital and messenger service use was negative and statistically significant (b=−0.084), indicating that the positive association between social capital and depressive mood was weaker among individuals with more frequent messenger use. As illustrated in Figure 1A, depressive mood increased with higher levels of social capital across all groups; however, the slope of this increase differed markedly by messenger use frequency. Among individuals who never used messenger services, depressive mood rose steeply as social capital increased, whereas this increase was more moderate among those who used messenger services 1–2 day/wk. In contrast, among daily messenger users, the association between social capital and depressive mood was substantially attenuated, resulting in relatively small differences in depressive mood across levels of social capital.
Model 3 examined SNS use as a moderating factor. The interaction between social capital and SNS use was also negative and statistically significant (b=−0.042), although its magnitude was smaller than that observed for messenger services. As shown in Figure 1B, depressive mood increased with higher levels of social capital across all SNS use groups, but differences between groups were less pronounced. Among individuals with very low social capital, those who used SNS more frequently reported slightly higher levels of depressive mood than non-users. As social capital increased, predicted depressive mood converged across SNS use groups, resulting in relatively similar levels at moderate levels of social capital. This pattern suggests that SNS use may function as a compensatory social resource among individuals with lower levels of offline social capital but may offer limited additional mental health benefits as social capital increases.
Model 4 focused on participation in membership-based online communities. The interaction between social capital and participation in membership-based online communities was negative but did not reach conventional levels of statistical significance. As noted in Table 2, depressive mood increased more sharply with higher levels of social capital among individuals who did not participate in such communities, whereas a more moderate increase was observed among those who participated 1–2 day/wk. Among frequent participants, the association between social capital and depressive mood appeared weaker; however, confidence intervals overlapped across groups, indicating that these differences should be interpreted with caution. Collectively, these results provide partial support for hypothesis 2.
To facilitate interpretation, the marginal effects of social capital on depressive mood were plotted across levels of digital media use. Overall, the results indicate that although higher social capital was, on average, associated with higher depressive mood, this association was substantially conditioned by exposure to online community environments. In particular, more relational and interaction-oriented digital environments —such as messenger services and, to a lesser extent, SNS—were associated with a weaker association between social capital and depressive mood.
This study examined how engagement with online community environments conditions the association between social capital and depressive mood. Contrary to the dominant assumption that social capital uniformly promotes mental health, the baseline analysis showed that higher levels of social capital were associated with higher levels of depressive mood. This finding aligns with sociological perspectives emphasizing the burdens of social participation, including role strain and accumulated social obligations [9–11], and is consistent with social-epidemiological accounts in which sustained obligations and overlapping role demands lead to chronic mobilization of psychosocial resources, resource depletion, and sustained stress exposure—a recognized pathway to depressive symptomatology [2,12,13].
This association was not uniform across social contexts. Engagement with online community environments substantially modified the relationship between social capital and depressive mood. Frequent messenger service use was associated with marked attenuation of the positive association between social capital and depressive mood. A similar, though weaker, moderating pattern was observed for SNS use. These findings suggest that relationally oriented online environments may reorganize how social participation is experienced, potentially reducing the emotional burden associated with extensive social involvement. The weaker moderating role of SNS compared with messenger services may reflect differences in relational character. Messenger services typically involve more dyadic and reciprocal interaction, whereas SNS more often involve broadcast-oriented communication and diffuse exposure. In addition, SNS affordances may foreground performative self-presentation and upward social comparison [31,32], attenuating the psychological relief that interactive communication might otherwise provide. Participation in membership-based online communities showed a similar pattern in the predicted values, although the interaction effect did not reach conventional levels of statistical significance. This suggests that although structured online communities may offer contexts in which social participation is redistributed or managed, their moderating role may be more contingent or heterogeneous. Together, these results indicate that online community environments do not automatically translate social capital into mental health benefits but can condition its emotional consequences.
These findings suggest that public mental health research and practice should distinguish between the quantity of social participation and the conditions under which participation occurs. The social capital measure used in this study captures the cumulative quantity of formal organizational participation but does not directly assess its qualitative conditions, such as perceived voluntariness, role compatibility, or emotional reciprocity. Policies and interventions that promote social participation as a pathway to improved mental well-being often implicitly assume that greater involvement is uniformly beneficial. The present results indicate that this assumption may not always hold: cumulative social participation may, under certain conditions, be associated with elevated rather than reduced depressive mood. Policies that seek to expand social participation as a population-level mental health strategy should therefore attend to the qualitative conditions under which participation is sustained, rather than treating the volume of engagement as a sufficient indicator of beneficial social capital. In this respect, relationally oriented digital tools such as messenger services may be considered supplementary resources that redistribute the demands of offline participation rather than substitutes for in-person engagement, and they could be integrated as complements within community mental health programs.
Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference, and reverse causality cannot be ruled out. Second, depressive mood was measured using a single-item indicator of day-to-day emotional distress rather than clinical depression. Third, the analysis focused on how online environments structure the experience of social participation rather than on the composition of specific networks. Fourth, the role-strain account adopted here is one of several plausible mechanisms through which cumulative social participation may be associated with depressive mood. Other mechanisms—including participation rewards that fall short of expectations, intensified social comparison, and declining relational quality despite quantitative engagement—may also contribute to this pattern. For instance, extensive but unrewarding participation may increase depressive mood, whereas messenger and SNS use may partially compensate for disappointment when offline participation falls short of expected returns. The present data do not allow adjudication among these accounts, and future research using longitudinal designs and direct measures of perceived participation rewards, social comparison, and relational quality would help clarify which mechanisms are most consequential.
Despite these limitations, this study extends the Korean public health literature on social capital and mental health in 2 respects. First, using nationally representative data and a measure of social participation that captures cumulative engagement across multiple organizational types, the study documents that social capital is not uniformly protective against depressive mood and may, beyond a certain level of accumulation, be positively associated with emotional distress. Second, it shows that this association is meaningfully conditioned by the digital environments through which contemporary social participation is increasingly sustained, an issue largely unexamined in prior research linking social capital to mental health outcomes. As social participation increasingly unfolds across digital spaces, longitudinal and mechanism-oriented research will be needed to clarify the precise pathways linking cumulative social participation to depressive mood and to identify which features of digital relational contexts most effectively buffer this association.
The Korea Social Integration Survey dataset analyzed in this study is available from the Korea Institute of Public Administration.

Conflict of Interest

The author has no conflicts of interest associated with the material presented in this paper.

Funding

None.

Acknowledgements

None.

Author Contributions

All work was done by Baek K.

Figure 1
Interactions between social capital and online community environment. (A) Predicted depressive mood by social capital at 3 levels of messenger service use, based on model 2. (B) Predicted depressive mood by social capital at 3 levels of social networking services (SNS) use, based on model 3. Predicted values were estimated with all covariates held at their means; shaded areas indicate 95% confidence intervals. On the horizontal axis, “very low,” “middle,” and “very high” correspond to social capital index scores of 1, 3, and 5, respectively. “Never,” “1–2 day/wk,” and “every day” correspond to the lowest, intermediate, and highest response categories of the 7-point frequency-of-use scale.
jpmph-26-305f1.jpg
jpmph-26-305f2.jpg
Table 1
Descriptive statistics of study variables
Variables Observations (n) Mean±SD Min Max
Depressive mood 8221 2.82±1.95 0 10
Sex (1=female) 8221 0.51±0.50 0 1
Age 8221 3.43±1.44 1 5
Educational attainment 8221 3.20±0.91 1 4
Self-rated health 8221 3.63±0.75 1 5
Perceived social status 8221 5.52±1.46 0 10
Household income 8221 4.63±1.86 1 7
Religious affiliation (1=has a religion) 8221 0.31±0.46 0 1
Spousal status (1=has a spouse) 8221 0.66±0.47 0 1
General trust 8221 2.53±0.57 1 4
Life satisfaction 8221 6.43±1.46 0 10
Employment status (1=employed) 8221 0.74±0.44 0 1
Social capital 8221 1.61±0.64 1 5
Messenger service use 8221 5.72±1.98 1 7
Social networking service use 8221 3.59±2.42 1 7
Membership-based online community use 8221 4.58±2.30 1 7

SD, standard deviation; Min, minimum; Max, maximum.

Table 2
Ordinary least squares regression models predicting depressive mood
Variables Model 1 Model 2 Model 3 Model 4
Sex (1=female) 0.100 (0.040)* 0.097 (0.040)* 0.100 (0.040)* 0.100 (0.040)*
Age −0.050 (0.019)* −0.034 (0.020) −0.033 (0.020) −0.034 (0.020)
Educational attainment −0.028 (0.030) 0.004 (0.031) 0.013 (0.031) 0.012 (0.031)
Self-rated health −0.330 (0.032)*** −0.307 (0.032)*** −0.304 (0.032)*** −0.307 (0.032)***
Perceived social status 0.068 (0.015)*** 0.058 (0.015)*** 0.060 (0.015)*** 0.059 (0.015)***
Household income −0.018 (0.012) −0.003 (0.013) −0.005 (0.013) −0.005 (0.013)
Religious affiliation (1=has a religion) −0.028 (0.044) −0.025 (0.044) −0.029 (0.044) −0.023 (0.044)
Spousal status (1=has a spouse) −0.071 (0.049) −0.064 (0.049) −0.059 (0.049) −0.059 (0.049)
General trust −0.123 (0.035)*** −0.121 (0.035)*** −0.123 (0.035)*** −0.123 (0.035)***
Life satisfaction −0.466 (0.015)*** −0.466 (0.015)*** −0.469 (0.015)*** −0.466 (0.015)***
Employment status (1=employed) 0.099 (0.049)* 0.101 (0.049)* 0.107 (0.049)* 0.109 (0.049)*
Social capital 0.327 (0.032)*** 0.823 (0.110)*** 0.498 (0.061)*** 0.454 (0.075)***
Messenger service use - 0.081 (0.029)** −0.047 (0.012)*** −0.046 (0.012)***
Social networking service use - 0.066 (0.011)*** 0.134 (0.024)*** 0.065 (0.011)***
Membership-based online community use - −0.065 (0.012)*** −0.066 (0.012)*** −0.022 (0.025)
Social capital×Messenger service use - −0.084 (0.017)*** - -
Social capital×Social networking service use - - −0.042 (0.013)** -
Social capital×Membership-based online community use - - - −0.026 (0.014)
Constant 6.695 (0.189)*** 6.015 (0.244)*** 6.460 (0.208)*** 6.529 (0.217)***
Observations 8,221 8,221 8,221 8,221
R2 0.161 0.171 0.170 0.169

Values are presented as b (standard error).

Model 1 includes social capital and all covariates. Models 2–4 add the 3 online community context variables and, respectively, an interaction term between social capital and messenger service use (model 2), social networking service use (model 3), and membership-based online community use (model 4).

* p<0.05,

** p<0.01,

*** p<0.001.

Figure & Data

References

    Citations

    Citations to this article as recorded by  

      Figure
      • 0
      • 1
      Social Capital in Digitally Mediated Communities: Conditional Associations With Depressive Mood in Korea
      Image Image
      Figure 1 Interactions between social capital and online community environment. (A) Predicted depressive mood by social capital at 3 levels of messenger service use, based on model 2. (B) Predicted depressive mood by social capital at 3 levels of social networking services (SNS) use, based on model 3. Predicted values were estimated with all covariates held at their means; shaded areas indicate 95% confidence intervals. On the horizontal axis, “very low,” “middle,” and “very high” correspond to social capital index scores of 1, 3, and 5, respectively. “Never,” “1–2 day/wk,” and “every day” correspond to the lowest, intermediate, and highest response categories of the 7-point frequency-of-use scale.
      Graphical abstract
      Social Capital in Digitally Mediated Communities: Conditional Associations With Depressive Mood in Korea
      Variables Observations (n) Mean±SD Min Max
      Depressive mood 8221 2.82±1.95 0 10
      Sex (1=female) 8221 0.51±0.50 0 1
      Age 8221 3.43±1.44 1 5
      Educational attainment 8221 3.20±0.91 1 4
      Self-rated health 8221 3.63±0.75 1 5
      Perceived social status 8221 5.52±1.46 0 10
      Household income 8221 4.63±1.86 1 7
      Religious affiliation (1=has a religion) 8221 0.31±0.46 0 1
      Spousal status (1=has a spouse) 8221 0.66±0.47 0 1
      General trust 8221 2.53±0.57 1 4
      Life satisfaction 8221 6.43±1.46 0 10
      Employment status (1=employed) 8221 0.74±0.44 0 1
      Social capital 8221 1.61±0.64 1 5
      Messenger service use 8221 5.72±1.98 1 7
      Social networking service use 8221 3.59±2.42 1 7
      Membership-based online community use 8221 4.58±2.30 1 7
      Variables Model 1 Model 2 Model 3 Model 4
      Sex (1=female) 0.100 (0.040)* 0.097 (0.040)* 0.100 (0.040)* 0.100 (0.040)*
      Age −0.050 (0.019)* −0.034 (0.020) −0.033 (0.020) −0.034 (0.020)
      Educational attainment −0.028 (0.030) 0.004 (0.031) 0.013 (0.031) 0.012 (0.031)
      Self-rated health −0.330 (0.032)*** −0.307 (0.032)*** −0.304 (0.032)*** −0.307 (0.032)***
      Perceived social status 0.068 (0.015)*** 0.058 (0.015)*** 0.060 (0.015)*** 0.059 (0.015)***
      Household income −0.018 (0.012) −0.003 (0.013) −0.005 (0.013) −0.005 (0.013)
      Religious affiliation (1=has a religion) −0.028 (0.044) −0.025 (0.044) −0.029 (0.044) −0.023 (0.044)
      Spousal status (1=has a spouse) −0.071 (0.049) −0.064 (0.049) −0.059 (0.049) −0.059 (0.049)
      General trust −0.123 (0.035)*** −0.121 (0.035)*** −0.123 (0.035)*** −0.123 (0.035)***
      Life satisfaction −0.466 (0.015)*** −0.466 (0.015)*** −0.469 (0.015)*** −0.466 (0.015)***
      Employment status (1=employed) 0.099 (0.049)* 0.101 (0.049)* 0.107 (0.049)* 0.109 (0.049)*
      Social capital 0.327 (0.032)*** 0.823 (0.110)*** 0.498 (0.061)*** 0.454 (0.075)***
      Messenger service use - 0.081 (0.029)** −0.047 (0.012)*** −0.046 (0.012)***
      Social networking service use - 0.066 (0.011)*** 0.134 (0.024)*** 0.065 (0.011)***
      Membership-based online community use - −0.065 (0.012)*** −0.066 (0.012)*** −0.022 (0.025)
      Social capital×Messenger service use - −0.084 (0.017)*** - -
      Social capital×Social networking service use - - −0.042 (0.013)** -
      Social capital×Membership-based online community use - - - −0.026 (0.014)
      Constant 6.695 (0.189)*** 6.015 (0.244)*** 6.460 (0.208)*** 6.529 (0.217)***
      Observations 8,221 8,221 8,221 8,221
      R2 0.161 0.171 0.170 0.169
      Table 1 Descriptive statistics of study variables

      SD, standard deviation; Min, minimum; Max, maximum.

      Table 2 Ordinary least squares regression models predicting depressive mood

      Values are presented as b (standard error).

      Model 1 includes social capital and all covariates. Models 2–4 add the 3 online community context variables and, respectively, an interaction term between social capital and messenger service use (model 2), social networking service use (model 3), and membership-based online community use (model 4).

      p<0.05,

      p<0.01,

      p<0.001.


      JPMPH : Journal of Preventive Medicine and Public Health
      TOP