What this is
- This research examines the relationship between WeChat usage and among middle-aged and elderly individuals in China.
- It specifically investigates how mediates this relationship.
- Data were sourced from the China Health and Retirement Longitudinal Study (CHARLS) in 2018, focusing on depressive symptoms.
- Findings indicate that WeChat usage correlates with lower levels, with recreational activities being a significant mediating factor.
Essence
- WeChat usage is linked to lower levels among middle-aged and elderly individuals in China, with , particularly recreational activities, mediating this effect.
Key takeaways
- WeChat usage significantly correlates with a lower prevalence of , with an adjusted odds ratio of 0.701 (95% CI: 0.605-0.812).
- partially mediates the relationship between WeChat usage and , with recreational activities showing a significant mediating effect.
- The impact of WeChat usage on varies by age and gender, with stronger effects observed in the elderly compared to middle-aged individuals.
Caveats
- The study relies on cross-sectional data, limiting causal inferences between WeChat usage, , and .
- Only was examined as a mediating variable, leaving the potential influence of other factors unexplored.
- The classification of activities may lack objectivity due to the absence of a standardized classification.
Definitions
- Social participation: Engagement in activities that involve interactions with others in society or the community.
- Depression: A mental health condition characterized by persistent feelings of sadness and loss of interest, affecting daily functioning.
AI simplified
Background
Depression affects mood, quality of life, and physical health [1], and is an independent risk factor for increased mortality [2, 3]. The prevalence of depression among Chinese adults was 20% from 2011 to 2019 [4]. Depression is one of the most common mental health problems among the middle-aged and elderly and has become an important public health problem in China [5]. It not only poses a significant threat to physical and mental health and ability to live, but also brings a heavy burden to families and society. With the aging and the popularity of healthy aging, the prevention and treatment of depression in the middle-aged and elderly is becoming increasingly important.
With the high popularity of the Internet, social media has become a factor that affects mental health. The use of social media could reduce anxiety and loneliness and promote physical health [6], as well as reduce depression levels and increase life satisfaction [7]. Social media can provide many convenient services such as information retrieval and online communication for the middle-aged and elderly with mobility problems, so it is gradually winning their affection. Online socialization could overcome geographical barriers [8], and keep older adults in close contact with family and friends [9], therefore enhancing their social support [10, 11]. WeChat is a popular social media in China and has become an inseparable part of the work and life [12]. According to a report by the Chinese Academy of Social Sciences, WeChat has become the most commonly used online social tool among the elderly [13]. WeChat usage has some positive effects on both physical and mental health of middle-aged and older adults. WeChat usage could reduce the risk of depression in the elderly [14] and had a significant positive impact on subjective health status [15]. A study suggested that WeChat usage may also boost memory in older adults by reducing risk of depression [16]. However, some studies have noted that excessive use of social media, such as WeChat addiction, can have negative effects on the physical, psychological, and social health of users [17β19]. We attempted to clarify the correlation between WeChat usage and depression among Chinese middle-aged and older adults, and to explore whether there were mediating variables in the correlation.
Social participation is an important factor, among influences on depression. The definition of social participation is not entirely uniform. In general, social participation refers to activities in which individuals engage in interactions with others in society or the community [20]. In China, whatever way the elderly keep in touch with society is considered to be a form of social participation [21].
Studies showed that social participation was an important part of healthy aging [22], and great for reducing loneliness, relieving depressive symptoms [23, 24], increasing life satisfaction [25], and improving mental health [26]. The higher the frequency and levels of social participation, the lower the risk of depression in older adults [27]. Social participation had a positive effect on health by increasing their social capital and contact between friends and family [28]. Older adults who engaged in social participation tend to have better cognitive behaviors [29]. Many types of socially participating activities had effects on mental health, such as volunteering and healthy exercise [30], paid work [31], religious activity [32]. Different types of social participation could positively affect physical and mental health as well as life satisfaction of older adults through different mechanisms [33].
In the digital era, it has been suggested that social media usage could contribute to social participation among the middle-aged and elderly [34]. Does social participation play a mediating role between WeChat usage and depression in middle-aged and older populations? Do different types of social participation have different mediating effects? In this study, we aimed to 1) clarify the relationship between WeChat usage, social participation, and depression in the middle-aged and elderly in China; and 2) explore the mediating role of different types of social participation activities in the relationship.
Methods
Study design and study sample
We used data from the China Health and Retirement Longitudinal Study (CHARLS) in wave 4 of 2018 for cross-sectional analysis. CHARLS aims to collect high-quality data on households and individuals aged 45 and older in China to analyze aging and promote research on healthy aging. CHARLS surveyed participants for basic information, health status, health insurance and health care, and retirement. CHARLS was approved by the Ethics Review Committee of Peking University, and all participants signed an informed consent form before the investigation and voluntarily participated in the survey [35].
Data inclusion process
Depression
The dependent variable in this study was depression, which was set as a dichotomous variable based on the CHARLS questionnaire. The 10-item Center for Epidemiologic Studies Depression Scale (CES-D-10), which is a simplified version of The Center for Epidemiologic Studies Depression Scale, is included in the questionnaire of the CHARLS. Each question was scored on a scale of 0β3 (rarely or not at allβ=β0, not too muchβ=β1, sometimes or half the timeβ=β2, most of the timeβ=β3). The final score was calculated cumulatively, with a total score of 0β30. When the scores wereββ₯β10, the respondent was considered to have depressive symptoms, and below 10, the respondent was considered normal. The depression score also reflected the levels of depression in the sample; therefore, the depression was also examined as a continuous variable in this study. The higher the depression score, the higher the levels of depression (Cronbachβs alphaβ=β0.805).
WeChat usage
The independent variable of this study was WeChat usage status. In the questionnaire, participants were asked βDo you use WeChat?β and the answers included βyesβ and βnoβ. When participants answered βyesβ, they were considered to use WeChat, otherwise they were considered not to use WeChat.
Social participation
The mediating variable in this study was social participation. Participants were asked in the questionnaire whether they had engaged in 10 activities in the past month. Participants were considered to be socially engaged if they participated in any of the activities. We also measured the levels of social participation. Participants earn one point for engaging in an activity. Points were accumulated, and the total score ranged from 0 to 10 points. The higher the score, the higher the levels of social participation.
We also tried to explore which type or types of social participation have a mediating role. Based on the experience of previous papers [29], we categorized the 10 activities mentioned in the questionnaire into voluntary activities, recreational activities, cultural activities, and other activities. Voluntary activities included: providing assistance to family, friends or neighbors who do not live with participants; doing volunteer or charitable work; caring for a sick or disabled adult who does not live with participants. Recreational activities included: Interacting with friends; playing Mahjong, chess, cards, or going to community clubs; going to sports, social or other types of clubs. Cultural activities included: participating in a community-related organization; participating in an educational or training course. Other activities contained: stock investment; and other activities. Participants were considered to have taken part in the type if they engaged in any of the activities in that type. Participants earn one point for engaging in an activity. Voluntary activities, recreational activities, cultural activities and other activities were respectively scored cumulatively to measure the participation levels in these four types of activities.
Covariates
| Variable | Coding |
|---|---|
| Depression | <β10β=β0, β₯10β=β1 |
| Levels of depression | 0β~β30 |
| WeChat usage | Not using the WeChat =0, Using the WeChat =1 |
| Social participation | Noβ=β0, Yesβ=β1 |
| Levels of social participation | 0β~β10 |
| Voluntary activities | Noβ=β0, Yesβ=β1 |
| Levels of voluntary activities | Noβ=β0, One kindβ=β1, Two kindsβ=β2, Three kindsβ=β3 |
| Recreation | Noβ=β0, Yesβ=β1 |
| Levels of recreation | Noβ=β0, One kindβ=β1, Two kindsβ=β2, Three kindsβ=β3 |
| Cultural activities | Noβ=β0, Yesβ=β1 |
| Levels of cultural activities | Noβ=β0, One kindβ=β1, Two kindsβ=β2 |
| Other activities | Noβ=β0, Yesβ=β1 |
| Levels of other activities | Noβ=β0, One kindβ=β1, Two kindsβ=β2 |
| Age | β₯45 |
| Gender | Femaleβ=β0, Male =1 |
| Marital status | Unmarriedβ=β0, Marriedβ=β1 |
| Residence | Ruralβ=β1, Urbanβ=β2 |
| Education | No formal educationβ=β1, Elementary schoolβ=β2, Middle schoolβ=β3, High schoolβ=β4, College or aboveβ=β5 |
| Self-reported health | Very poorβ=β1, Poorβ=β2, Fairβ=β3, Goodβ=β4, Very goodβ=β5 |
| ADL | No impairedβ=β0, Impairedβ=β1 |
| Smoke status | Still haveβ=β1, Quitβ=β2, Noβ=β3 |
| Drink status | Noβ=β0, Yesβ=β1 |
| Sleep time | Take the log of sleep time |
| Employment | Noβ=β0, Yesβ=β1 |
| Pension insurance | Noβ=β0, Yesβ=β1 |
| Medical insurance | Noβ=β0, Yesβ=β1 |
| Chronic diseases | Noβ=β0, One kindβ=β1, Two kindsβ=β2, Three kindsβ=β3, Four kinds and moreβ=β4 |
Data analysis
We would use t-tests for continuous variables and chi-square tests for categorical variables to compare the sample characteristics between the WeChat users and non-WeChat users. Given that WeChat usage was not randomly distributed in the study, we used propensity score matching (PSM) to match the subjects. We used Psmatch2 to identify covariates and perform propensity score analysis. A 1:2 matching group was eventually constructed by nearest-neighbor matching, and the whole set of control variables was included in PSM.
Then, we used binary logistic regression and multiple linear regression on the matched groups to analyze the correlation between WeChat usage and depression. Since one of the dependent and mediating variables were both dichotomous, we used stepwise regression to examine the presence of a mediating effect. We also used the KHB method to check the mediating effect again and to analyze whether the mediating effect differed in different populations. The data were processed and analyzed using stata16.0.
Results
In Table 4, we replaced the dependent variable with the levels of depression, and WeChat usage was consistently and significantly associated with lower levels of depression. In Models 7 and 8, the presence and levels of social participation were significantly associated with lower levels of depression. The effect of WeChat on the levels of depression was also reduced with the inclusion of variables related to social participation, suggesting that social participation may also have a partially mediating role. Compared to Table 3, the significance of the association between education levels and depression levels increased, and the association of being married and employed with lower depression levels also showed significance.
Finally, we divided the population into two groups respectively according to age and gender: the middle-aged (45β59), the elderly (60β84), the female and the male. We tried to analyze whether the correlation between WeChat usage and depression (see Supplementary Material SβS) as well as the mediating role of social engagement (see Supplementary Material S, S, S, S, S, S, S, S, S, S, S) were heterogeneous because of age and gender. WeChat usage could reduce the levels and risk of depression in all groups and reduced the risk of depression better in the elderly than in the middle-aged. Apart from among women, social participation played the mediating role in almost all groups. However, the mediating effect of recreational activities was only present in the middle-aged. 1 2 3 4 5 6 7 8 9 10 11 12 13
| Variable | Before PSM (=β15,428)N | After PSM (=β4545)N | ||||
|---|---|---|---|---|---|---|
| WeChat user (=β2179)N | Non-WeChat user (=β13,249)N | valueP | WeChat user (=β1906)N | Non-WeChat user (=β2639)N | valueP | |
| Depressive symptom | <β0.001 | <β0.001 | ||||
| CES-Dβ<β10 | 1680 (77.1) | 8077 (61.0) | 1455 (76.3) | 1814 (68.7) | ||
| CES-Dβ>β=10 | 499 (22.9) | 5172 (39.0) | 451 (23.7) | 825 (31.3) | ||
| Levels of depression | 6.3βΒ±β0.1 | 8.8βΒ±β0.1 | <β0.001 | 6.4βΒ±β0.1 | 7.5βΒ±β0.1 | <β0.001 |
| Age | 55.3βΒ±β7.2 | 61.8βΒ±β9.3 | <β0.001 | 55.7βΒ±β7.4 | 56.2βΒ±β7.3 | 0.033 |
| Gender | <β0.001 | 0.558 | ||||
| Female | 985 (45.2) | 6887 (52.0) | 868 (45.5) | 1225 (46.4) | ||
| Male | 1194 (54.8) | 6362 (48.0) | 1038 (54.5) | 1414 (53.6) | ||
| Marital status | <β0.001 | 0.444 | ||||
| Unmarried | 158 (7.3) | 1754 (13.2) | 139 (7.3) | 177 (6.7) | ||
| Married | 2021 (92.7) | 11,495 (86.8) | 1767 (92.7) | 2462 (93.3) | ||
| Residence | <β0.001 | <β0.001 | ||||
| Rural | 937 (43.0) | 9965 (75.2) | 906 (47.5) | 1450 (54.9) | ||
| Urban | 1242 (57.0) | 3284 (24.8) | 1000 (52.5) | 1189 (45.1) | ||
| Education | <β0.001 | <β0.001 | ||||
| No formal education | 33 (1.5) | 2772 (20.9) | 33 (1.7) | 72 (2.7) | ||
| Elementary school | 506 (23.2) | 6277 (47.4) | 499 (26.2) | 847 (32.1) | ||
| Middle school | 775 (35.6) | 2926 (22.1) | 736 (38.6) | 997 (37.8) | ||
| High school | 657 (30.2) | 1138 (8.6) | 516 (27.1) | 623 (23.6) | ||
| College or above | 208 (9.5) | 136 (1.0) | 122 (6.4) | 100 (3.8) | ||
| Sleep time | 1.8βΒ±β0.3 | 1.8βΒ±β0.4 | <β0.001 | 1.8βΒ±β0.3 | 1.8βΒ±β0.3 | 0.678 |
| Self-reported health | <β0.001 | 0.049 | ||||
| Very poor | 53 (2.4) | 813 (5.8) | 51 (2.7) | 91 (3.4) | ||
| Poor | 206 (9.5) | 2688 (20.3) | 200 (10.5) | 328 (12.4) | ||
| Fair | 1103 (50.6) | 6580 (49.7) | 986 (51.7) | 1347 (51.0) | ||
| Good | 446 (20.5) | 1564 (11.8) | 357 (18.7) | 431 (16.3) | ||
| Very good | 371 (17.0) | 1604 (12.1) | 312 (16.4) | 442 (16.7) | ||
| ADL | <β0.001 | 0.844 | ||||
| No impaired | 2143 (98.3) | 12,395 (93.6) | 1870 (98.1) | 2587 (98.0) | ||
| Impaired | 36 (1.7) | 854 (6.4) | 36 (1.9) | 52 (2.0) | ||
| Smoke status | 0.235 | 0.501 | ||||
| Still have | 643 (29.5) | 3676 (27.7) | 563 (29.5) | 792 (30.0) | ||
| Quit | 309 (14.2) | 1917 (14.5) | 273 (14.3) | 346 (13.1) | ||
| No | 1227 (56.3) | 7656 (57.8) | 1070 (56.1) | 1501 (56.9) | ||
| Drink status | <β0.001 | 0.042 | ||||
| No | 1043 (47.9) | 8845 (66.8) | 966 (50.7) | 1418 (53.7) | ||
| Yes | 1136 (52.1) | 4404 (33.2) | 940 (49.3) | 1221 (46.3) | ||
| Employment | <β0.001 | 0.418 | ||||
| No | 629 (28.9) | 4431 (33.4) | 570 (29.9) | 760 (28.8) | ||
| Yes | 1550 (71.1) | 8818 (66.6) | 1336 (70.1) | 1879 (71.2) | ||
| Pension insurance | <β0.001 | 0.602 | ||||
| No | 164 (7.5) | 1397 (10.5) | 150 (7.9) | 219 (8.3) | ||
| Yes | 2015 (92.5) | 11,852 (89.5) | 1756 (92.1) | 2420 (91.7) | ||
| Medical insurance | 0.033 | 0.739 | ||||
| No | 41 (1.9) | 352 (2.6) | 41 (2.2) | 53 (2.0) | ||
| Yes | 2138 (98.1) | 12,897 (97.3) | 1865 (97.8) | 2586 (98.0) | ||
| Chronic diseases | <β0.001 | 0.348 | ||||
| No | 597 (27.4) | 2645 (20.0) | 499 (26.2) | 702 (26.6) | ||
| One | 566 (26.0) | 3131 (23.6) | 492 (25.8) | 664 (25.2) | ||
| Two | 405 (18.6) | 2722 (20.5) | 357 (18.7) | 516 (19.6) | ||
| Three | 292 (13.4) | 1889 (14.3) | 261 (13.7) | 314 (11.9) | ||
| Four and more | 319 (14.6) | 2862 (21.6) | 297 (15.6) | 443 (16.8) | ||
| Variable | Model 1 | Model 2 | Model 3 | Model 4 |
|---|---|---|---|---|
| OR (95%CI) | OR (95%CI) | OR (95%CI) | OR (95%CI) | |
| WeChat usage | 0.711*** | 0.701*** | 0.729*** | 0.731*** |
| (0.620β0.816) | (0.605β0.812) | (0.628β0.846) | (0.629β0.850) | |
| Social participation (Yes) | 0.792** | |||
| (0.683β0.919) | ||||
| Levels of Social participation | 0.922* | |||
| (0.866β0.982) | ||||
| Age | 0.987** | 0.971*** | 0.97*** | 0.971*** |
| (0.978β0.997) | (0.960β0.983) | (0.959β0.982) | (0.959β0.982) | |
| Gender (Male) | 0.625*** | 0.694** | 0.682** | 0.683** |
| (0.546β0.717) | (0.552β0.873) | (0.542β0.857) | (0.543β0.859) | |
| Marital status (Married) | 0.666** | 0.784 | 0.782 | 0.784 |
| (0.518β0.855) | (0.597β1.030) | (0.595β1.028) | (0.597β1.031) | |
| Residence (Urban) | 0.77*** | 0.762*** | 0.759*** | 0.763*** |
| (0.668β0.887) | (0.649β0.894) | (0.646β0.890) | (0.650β0.896) | |
| Education | ||||
| Elementary school | 1.013 | 1.046 | 1.07 | 1.069 |
| (0.672β1.527) | (0.666β1.642) | (0.681β1.681) | (0.681β1.680) | |
| Middle school | 0.658* | 0.724 | 0.752 | 0.75 |
| (0.436β0.993) | (0.461β1.138) | (0.478β1.183) | (0.477β1.180) | |
| High school | 0.524** | 0.595* | 0.622* | 0.623* |
| (0.342β0.803) | (0.372β0.949) | (0.389β0.994) | (0.389β0.996) | |
| College or above | 0.585* | 0.705 | 0.752 | 0.764 |
| (0.346β0.992) | (0.398β1.250) | (0.423β1.336) | (0.429β1.360) | |
| Self-reported health | ||||
| Poor | 0.648* | 0.641* | 0.649* | |
| (0.429β0.979) | (0.424β0.968) | (0.430β0.981) | ||
| Fair | 0.311*** | 0.31*** | 0.315*** | |
| (0.211β0.459) | (0.210β0.457) | (0.213β0.464) | ||
| Good | 0.184*** | 0.186*** | 0.189*** | |
| (0.119β0.283) | (0.121β0.287) | (0.122β0.291) | ||
| Very good | 0.12*** | 0.121*** | 0.123*** | |
| (0.077β0.188) | (0.077β0.189) | (0.078β0.193) | ||
| ADL (Impaired) | 2.693*** | 2.69*** | 2.66*** | |
| (1.620β4.478) | (1.614β4.482) | (1.599β4.424) | ||
| Smoke status | ||||
| Quit | 1.004 | 0.998 | 0.996 | |
| (0.785β1.284) | (0.780β1.276) | (0.779β1.274) | ||
| No | 1.058 | 1.048 | 1.047 | |
| (0.841β1.331) | (0.832β1.318) | (0.832β1.317) | ||
| Drink status (Yes) | 0.928 | 0.945 | 0.943 | |
| (0.786β1.095) | (0.800β1.117) | (0.798β1.114) | ||
| Sleep time | 0.297*** | 0.3*** | 0.3*** | |
| (0.230β0.383) | (0.233β0.387) | (0.233β0.387) | ||
| Employment (Yes) | 1.167 | 1.157 | 1.164 | |
| (0.963β1.414) | (0.955β1.402) | (0.960β1.410) | ||
| Pension insurance (Yes) | 0.991 | 0.999 | 0.999 | |
| (0.762β1.289) | (0.768β1.300) | (0.768β1.299) | ||
| Medical insurance (Yes) | 0.581* | 0.588* | 0.588* | |
| (0.355β0.949) | (0.360β0.961) | (0.360β0.960) | ||
| Chronic diseases | ||||
| One | 1.128 | 1.135 | 1.136 | |
| (0.911β1.395) | (0.917β1.405) | (0.918β1.406) | ||
| Two | 1.299* | 1.322* | 1.32* | |
| (1.035β1.630) | (1.053β1.659) | (1.052β1.657) | ||
| Three | 1.646*** | 1.689*** | 1.68*** | |
| (1.276β2.123) | (1.308β2.181) | (1.302β2.169) | ||
| Four and more | 2*** | 2.062*** | 2.074*** | |
| (1.563β2.561) | (1.609β2.643) | (1.617β2.659) | ||
| Constant | 2.661** | 172.044*** | 187.642*** | 173.886*** |
| (1.296β5.464) | (54.026β547.865) | (58.778β599.020) | (54.568β554.111) | |
| N | 4545 | 4545 | 4545 | 4545 |
| Variable | Model 5 | Model 6 | Model 7 | Model 8 |
|---|---|---|---|---|
| WeChat usage | β0.873*** | β0.787*** | β0.716*** | β0.692*** |
| (0.170) | (0.153) | (0.155) | (0.157) | |
| Social participation (Yes) | β0.418** | |||
| (0.158) | ||||
| Levels of Social participation | β0.172** | |||
| (0.063) | ||||
| Age | β0.036** | ββ0.071*** | ββ0.072*** | ββ0.072*** |
| (0.012) | (0.012) | (0.012) | (0.012) | |
| Gender (Male) | β1.520*** | β1.175*** | β1.208*** | β1.217*** |
| (0.173) | (0.237) | (0.237) | (0.238) | |
| Marital status (Married) | β1.898*** | ββ1.220*** | ββ1.221*** | ββ1.212*** |
| (0.332) | (0.299) | (0.299) | (0.299) | |
| Residence (Urban) | β0.775*** | β0.689*** | ββ0.695*** | ββ0.682*** |
| (0.178) | (0.168) | (0.168) | (0.168) | |
| Education | ||||
| Elementary school | β0.240 | ββ0.165 | ββ0.125 | ββ0.116 |
| (0.573) | (0.513) | (0.513) | (0.513) | |
| Middle school | β1.495** | β1.029* | β0.962 | β0.950 |
| (0.571) | (0.513) | (0.513) | (0.513) | |
| High school | β2.270*** | β1.701** | β1.621** | ββ1.596** |
| (0.584) | (0.525) | (0.525) | (0.526) | |
| College or above | β2.499*** | β1.820** | β1.712** | ββ1.646** |
| (0.683) | (0.614) | (0.615) | (0.617) | |
| Self-reported health | ||||
| Poor | β2.662*** | ββ2.682*** | ββ2.662*** | |
| (0.483) | (0.482) | (0.482) | ||
| Fair | β5.230*** | β5.234*** | β5.206*** | |
| (0.454) | (0.454) | (0.454) | ||
| Good | β6.485*** | β6.458*** | β6.426*** | |
| (0.487) | (0.487) | (0.487) | ||
| Very good | β7.633*** | β7.616*** | β7.576*** | |
| (0.490) | (0.490) | (0.491) | ||
| ADL (Impaired) | 2.472*** | 2.457*** | 2.438*** | |
| (0.565) | (0.565) | (0.565) | ||
| Smoke status | ||||
| Quit | β0.405 | β0.417 | β0.421 | |
| (0.249) | (0.249) | (0.249) | ||
| No | β0.141 | β0.157 | ββ0.160 | |
| (0.233) | (0.233) | (0.233) | ||
| Drink status (Yes) | β0.201 | ββ0.165 | ββ0.160 | |
| (0.174) | (0.174) | (0.174) | ||
| Sleep time | β3.814*** | β3.797*** | β3.796*** | |
| (0.272) | (0.272) | (0.272) | ||
| Employment (Yes) | 0.428* | 0.411* | 0.423* | |
| (0.202) | (0.201) | (0.201) | ||
| Pension insurance (Yes) | 0.073 | 0.085 | 0.088 | |
| (0.285) | (0.285) | (0.285) | ||
| Medical insurance (Yes) | β1.512** | β1.486** | β1.479** | |
| (0.542) | (0.542) | (0.542) | ||
| Chronic diseases | ||||
| One | 0.248 | 0.265 | 0.269 | |
| (0.211) | (0.211) | (0.211) | ||
| Two | 0.578* | 0.608** | 0.611** | |
| (0.234) | (0.234) | (0.234) | ||
| Three | 1.066*** | 1.112*** | 1.112*** | |
| (0.272) | (0.272) | (0.272) | ||
| Four and more | 1.866*** | 1.918*** | 1.947*** | |
| (0.268) | (0.269) | (0.270) | ||
| Constant | 13.698*** | 27.387*** | 27.528*** | 27.382*** |
| (0.930) | (1.233) | (1.233) | (1.232) | |
| N | 4545 | 4545 | 4545 | 4545 |
| Variable | Model 9 | Model 10 | Model 11 | Model 12 | Model 13 | Model 14 | Model 15 | Model 16 |
|---|---|---|---|---|---|---|---|---|
| WeChat usage | 0.703*** | 0.726*** | 0.711*** | 0.706*** | 0.701*** | 0.734*** | 0.711*** | 0.705*** |
| (0.606β0.816) | (0.625β0.843) | (0.613β0.824) | (0.609β0.818) | (0.604β0.813) | (0.631β0.852) | (0.614β0.825) | (0.608β0.817) | |
| Voluntary activities (Yes) | 0.975 | |||||||
| (0.820β1.159) | ||||||||
| Recreation (Yes) | 0.812** | |||||||
| (0.701β0.940) | ||||||||
| Cultural activities (Yes) | 0.718 | |||||||
| (0.506β1.017) | ||||||||
| Other activities (Yes) | 0.792 | |||||||
| (0.504β1.244) | ||||||||
| Levels of voluntary activities | ||||||||
| One | 0.941 | |||||||
| (0.782β1.133) | ||||||||
| Two | 1.134 | |||||||
| (0.762β1.688) | ||||||||
| Three | 1.4 | |||||||
| (0.555β3.535) | ||||||||
| Levels of recreation | ||||||||
| One | 0.859 | |||||||
| (0.733β1.007) | ||||||||
| Two | 0.697** | |||||||
| (0.557β0.873) | ||||||||
| Three | 0.788 | |||||||
| (0.483β1.284) | ||||||||
| Levels of Cultural activities | ||||||||
| One | 0.746 | |||||||
| (0.522β1.067) | ||||||||
| Two | 0.395 | |||||||
| (0.088β1.765) | ||||||||
| Levels of other activities | ||||||||
| One | 0.764 | |||||||
| (0.483β1.208) | ||||||||
| Two | 6.116 | |||||||
| (0.310β120.763) | ||||||||
| Constant | 172.835*** | 180.163*** | 165.791*** | 169.944*** | 174.239*** | 179.866*** | 166.274*** | 170.102*** |
| (54.249β550.642) | (56.502β574.465) | (52.058β528.000) | (53.340β541.447) | (54.668β555.334) | (56.295β574.685) | (52.210β529.536) | (53.380β542.052) | |
| N | 4545 | 4545 | 4545 | 4545 | 4545 | 4545 | 4545 | 4545 |
| Variable | Model 17 | Model 18 | Model 19 | Model 20 | Model 21 | Model 22 | Model 23 | Model 24 |
|---|---|---|---|---|---|---|---|---|
| WeChat usage | ββ0.757*** | ββ0.733*** | β0.758*** | ββ0.769*** | ββ0.760*** | β0.709*** | β0.758*** | β0.770*** |
| (0.155) | (0.155) | (0.154) | (0.154) | (0.155) | (0.156) | (0.154) | (0.154) | |
| Voluntary activities (Yes) | β0.245 | |||||||
| (0.180) | ||||||||
| Recreation (Yes) | β0.312* | |||||||
| (0.155) | ||||||||
| Cultural activities (Yes) | ββ0.613 | |||||||
| (0.325) | ||||||||
| Other activities (Yes) | β0.535 | |||||||
| (0.420) | ||||||||
| Levels of voluntary activities | ||||||||
| One | β0.304 | |||||||
| (0.192) | ||||||||
| Two | 0.078 | |||||||
| (0.413) | ||||||||
| Three | β0.110 | |||||||
| (1.018) | ||||||||
| Levels of recreation | ||||||||
| One | β0.181 | |||||||
| (0.169) | ||||||||
| Two | β0.616** | |||||||
| (0.227) | ||||||||
| Three | β0.525 | |||||||
| (0.483) | ||||||||
| Levels of Cultural activities | ||||||||
| One | β0.602 | |||||||
| (0.339) | ||||||||
| Two | β0.719 | |||||||
| (1.000) | ||||||||
| Levels of other activities | ||||||||
| One | β0.558 | |||||||
| (0.422) | ||||||||
| Two | 1.294 | |||||||
| (3.575) | ||||||||
| Constant | 27.404*** | 27.458*** | 27.309*** | 27.349*** | 27.406*** | 27.414*** | 27.31*** | 27.35*** |
| (1.233) | (1.233) | (1.233) | (1.233) | (1.233) | (1.233) | (1.233) | (1.233) | |
| N | 4545 | 4545 | 4545 | 4545 | 4545 | 4545 | 4545 | 4545 |
| Variable | Model 25 | Model 26 | Model 27 | Model 28 |
|---|---|---|---|---|
| Social participationOR (95%CI) | Level of social participationCoefficients(Standard error) | RecreationOR (95%CI) | Level of recreationOR (95%CI) | |
| WeChat usage | 2.137*** | 0.552*** | 2.078*** | 2.103*** |
| (1.880β2.429) | (0.036) | (1.836β2.353) | (1.877β2.356) | |
| Age | 0.991 | β0.005 | 0.993 | 0.995 |
| (0.981β1.001) | (0.003) | (0.984β1.003) | (0.986β1.004) | |
| Gender (Male) | 0.698*** | β0.246*** | 0.601*** | 0.589*** |
| (0.574β0.850) | (0.056) | (0.496β0.729) | (0.493β0.704) | |
| Marital status (Married) | 0.981 | 0.043 | 1.051 | 1.109 |
| (0.766β1.257) | (0.071) | (0.825β1.339) | (0.888β1.385) | |
| Residence (Urban) | 0.939 | 0.041 | 0.907 | 0.966 |
| (0.818β1.078) | (0.040) | (0.792β1.038) | (0.853β1.094) | |
| Education | ||||
| Elementary school | 1.497 | 0.286* | 1.692* | 1.695* |
| (0.991β2.262) | (0.122) | (1.107β2.584) | (1.124β2.555) | |
| Middle school | 1.988** | 0.463*** | 2.246*** | 2.137*** |
| (1.316β3.003) | (0.121) | (1.471β3.430) | (1.419β3.218) | |
| High school | 2.306*** | 0.613*** | 2.453*** | 2.458*** |
| (1.510β3.523) | (0.124) | (1.591β3.783) | (1.618β3.734) | |
| College or above | 3.233*** | 1.01*** | 3.343*** | 3.488*** |
| (1.930β5.415) | (0.145) | (2.003β5.579) | (2.160β5.632) | |
| Self-reported health | ||||
| Poor | 0.81 | 0.002 | 0.845 | 0.957 |
| (0.546β1.200) | (0.114) | (0.574β1.243) | (0.670β1.368) | |
| Fair | 0.964 | 0.141 | 0.994 | 1.105 |
| (0.665β1.397) | (0.107) | (0.691β1.431) | (0.792β1.543) | |
| Good | 1.351 | 0.341** | 1.328 | 1.464* |
| (0.905β2.017) | (0.115) | (0.897β1.965) | (1.024β2.094) | |
| Very good | 1.202 | 0.333** | 1.284 | 1.442* |
| (0.805β1.796) | (0.116) | (0.866β1.903) | (1.005β2.068) | |
| ADL (Impaired) | 0.846 | ββ0.203 | 0.833 | 0.746 |
| (0.535β1.337) | (0.134) | (0.529β1.312) | (0.489β1.138) | |
| Smoke status | ||||
| Quit | 0.879 | β0.093 | 0.91 | 0.906 |
| (0.716β1.079) | (0.059) | (0.744β1.112) | (0.754β1.089) | |
| No | 0.838 | β0.114* | 0.773** | 0.742*** |
| 0.691β1.015) | (0.055) | (0.641β0.933) | (0.624β0.881) | |
| Drink status (Yes) | 1.493*** | 0.244*** | 1.508*** | 1.464*** |
| (1.293β1.726) | (0.041) | (1.309β1.737) | (1.286β1.666) | |
| Sleep time | 1.193 | 0.103 | 1.313* | 1.291* |
| (0.955β1.490) | (0.064) | (1.054β1.637) | (1.051β1.585) | |
| Employment (Yes) | 0.829* | ββ0.029 | 0.815* | 0.77*** |
| (0.702β0.980) | (0.048) | (0.692β0.959) | (0.662β0.894) | |
| Pension insurance (Yes) | 1.146 | 0.092 | 1.092 | 1.072 |
| (0.911β1.442) | (0.067) | (0.869β1.370) | (0.867β1.327) | |
| Medical insurance (Yes) | 1.308 | 0.189 | 1.243 | 1.292 |
| (0.847β2.019) | (0.128) | (0.805β1.919) | (0.860β1.942) | |
| Chronic diseases | ||||
| One | 1.197* | 0.126* | 1.146 | 1.178* |
| (1.008β1.422) | (0.050) | (0.967β1.357) | (1.007β1.377) | |
| Two | 1.377** | 0.193*** | 1.302** | 1.294** |
| (1.136β1.669) | (0.055) | (1.079β1.573) | (1.090β1.537) | |
| Three | 1.636*** | 0.267*** | 1.495*** | 1.504*** |
| (1.304β2.053) | (0.064) | (1.199β1.866) | (1.230β1.838) | |
| Four and more | 1.742*** | 0.474*** | 1.633*** | 1.745*** |
| (1.395β2.176) | (0.064) | (1.314β2.029) | (1.431β2.130) | |
| Constant | 0.509 | β0.028 | 0.315* | |
| (0.186β1.390) | (0.292) | (0.116β0.856) | ||
| N | 4545 | 4545 | 4545 | 4545 |
| Effect | Ξ² | SE | P | 95% CI | Mediation (%) | Ξ² | SE | P | 95% CI | Mediation (%) | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Lower | Upper | Lower | Upper | |||||||||
| WeChat usage--Social participation--Depression | WeChat usage--Social participation--Degree of depression | |||||||||||
| Total effect | ββ0.356 | 0.075 | 0 | ββ0.503 | ββ0.209 | ββ0.787 | 0.153 | 0 | β1.086 | β0.487 | ||
| Direct effect | β0.317 | 0.076 | 0 | β0.466 | β0.167 | β0.716 | 0.155 | 0 | β1.020 | β0.412 | ||
| Indirect effect | β0.040 | 0.013 | 0.003 | β0.066 | β0.014 | 11.11 | β0.071 | 0.028 | 0.01 | β0.125 | β0.017 | 9.01 |
| WeChat usage--Levels of social participation--Depression | WeChat usage--Levels of social participation--Degree of depression | |||||||||||
| Total effect | β0.357 | 0.075 | 0 | β0.505 | β0.210 | β0.787 | 0.153 | 0 | β1.086 | β0.487 | ||
| Direct effect | β0.313 | 0.077 | 0 | β0.463 | β0.162 | β0.692 | 0.157 | 0 | β0.999 | β0.385 | ||
| Indirect effect | β0.045 | 0.018 | 0.013 | β0.080 | β0.009 | 12.49 | β0.095 | 0.035 | 0.007 | β0.164 | β0.026 | 12.04 |
| WeChat usage--Recreation--Depression | WeChat usage--Recreation--Degree of depression | |||||||||||
| Total effect | β0.356 | 0.075 | 0 | β0.503 | β0.209 | β0.787 | 0.153 | 0 | β1.086 | β0.487 | ||
| Direct effect | β0.320 | 0.076 | 0 | β0.469 | β0.171 | β0.733 | 0.155 | 0 | β1.037 | β0.429 | ||
| Indirect effect | β0.036 | 0.013 | 0.007 | β0.062 | β0.010 | 10.08 | β0.054 | 0.027 | 0.048 | β0.107 | β0.001 | 6.82 |
| WeChat usage--Levels of recreation--Depression | WeChat usage--Levels of recreation--Degree of depression | |||||||||||
| Total effect | β0.357 | 0.075 | 0 | β0.504 | β0.209 | β0.787 | 0.153 | 0 | β1.086 | β0.487 | ||
| Direct effect | β0.310 | 0.076 | 0 | β0.460 | β0.160 | β0.709 | 0.156 | 0 | β1.014 | β0.404 | ||
| Indirect effect | β0.047 | 0.015 | 0.002 | β0.077 | β0.017 | 13.17 | ββ0.078 | 0.03 | 0.01 | β0.137 | ββ0.018 | 9.89 |
Discussion
This study concluded that WeChat usage was significantly associated with lower risk of depression and lower levels of depression in middle-aged and older adults, while social participation had mediating effects. Moreover, there was heterogeneity in the mediating effect across groups.
After we used PSM and adjusted for a range of control variables, the study found that WeChat usage was significantly associated with lower rates of depression and lower levels of depression. Social media usage was generally significantly associated with fewer depressive symptoms [48, 49], which is generally consistent with previous studies. There is an explanation that increasing age decreases physical functioning and impedes mobility in older adults, which increases the vulnerability to social isolation and loneliness dilemmas, causing mental health problems [50]. According to a survey, almost half of the elderly in China lived alone or only with their spouses [51]. Due to physical and distance reasons, it is difficult for them to meet with their relatives or friends frequently. Besides the convenience and speed of social media, its real-time updates and interactivity can facilitate the middle-aged and elderly to communicate with others and share their emotions, which may alleviate the problem of social isolation. For example, the use of social media such as WeChat and Facebook could expand the social network of the middle-aged and elderly, enhance their interaction with others, and promote communication and emotional cohesion with family members [52, 53]. So, we think that WeChat, as a comprehensive social media, offers online communication technology that may benefit people who cannot or rarely interact with others face-to-face, and its mental health benefits for middle-aged and older adults should be considered.
Social participation has a great impact on the occurrence of depression. We found that middle-aged and older adults who do not engage in social participation have a higher risk of depression than those who do, and among the four types of social participation activities proposed in this study, recreation could significantly reduce the probability and levels of depression. In line with the results of previous studies, active social participation could significantly reduce the risk of depression [22β24]. Because positive social participation may increase communication and interaction between middle-aged and older adults and help them understand the aging process as well as mitigate the negative effects of aging on mental health [54]. Recreational activities chosen according to their interests can also help them to relieve negative emotions effectively [21]. In this study, we found that recreation (socializing, playing Mahjong, playing chess, dancing, practicing qigong, etc.) was the main social activity of Chinese middle-aged and older adults. Competitive activities, such as playing chess and mahjong, can fully exercise their thinking skills and bring joy as well as a sense of accomplishment when they achieve victory. Leisure activities, such as dancing and qigong, can effectively relax their bodies. These activities are usually participated by many people and they can have sufficient conversations while engaging in recreational activities, which can allow them to relax and release the negative emotions caused by stressful events. Therefore, we hold the opinion that it may be effective to reduce the risk of depression by organizing more forms of recreational activities to meet the needs of the middle-aged and elderly, and encouraging them to try more forms of social participation.
This study found that WeChat usage can promote social participation in middle-aged and older adults, which is consistent with previous findings [34, 43, 55]. We further found that WeChat usage can influence depression through social participation. A possible reason is that online socializing can alleviate the shrinking social circle of older adults [15, 56]. Especially for those who have fewer social activities, online communication technologies allow them to gain more opportunities to communicate with others [57]. We think that the photo sharing, voice messaging and video calling technologies of social media have greatly increased the fun of communication for the middle-aged and elderly. It also fulfills their need for emotional expression and sharing their lives. Social media enhances their resistance to depression by increasing their emotional support and social support. On the other hand, social media can have an impact on different groups of people [58]. We think the online social function of social media is applicable not only to individuals, but also to groups. For example, different family WeChat groups can be established among different family members. The middle-aged and elderly can also form WeChat groups with like-minded people to promote further communication among them and facilitate them to organize more offline activities. Both online interaction and offline communication can alleviate the loneliness of the middle-aged and elderly and have a positive effect on mental health.
Our study found that WeChat usage promotes social participation among middle-aged and older adults, but only the recreational activities among them have a mediating effect. In previous studies, voluntary and cultural activities were also beneficial to physical and mental health [30, 46, 59], but in this study, we did not find that they could have a mediating effect. Probably the reason is that the sample of this study had a high participation rate in recreation and a low participation rate in other types of activities, especially cultural activities, which social activity participation status had an impact on the final results. Another possibility is that WeChat usage does not have an impact through these types of activities, which needs to be further investigated.
We also found some differences in the levels of effect of WeChat usage on depression and in the mediating effect of social participation because of differences in age and gender. The positive effect of WeChat usage on depression was better in the elderly than in the middle-aged. Although social participation had mediating effects in most groups, WeChat usage reduced the prevalence of depression through recreational activities only in middle-aged adults and men. Possible explanations are that communication is the most pressing need for WeChat usage [60]. The use of social media such as WeChat can help older adults communicate more easily with family and friends, share their lives and communicate emotionally [15], which can have a more positive effect on mental health. Compared with the elderly, middle-aged people are less likely to face obstacles in action and communication because of better physical health. On the other hand, older adults still face some difficulties in using WeChat, but the middle-aged group has better learning and receptive abilities, so they are more likely to fully utilize social media such as WeChat for recreational activities and thus gain mental health benefits. We propose that the accessibility of social media for older adults should be taken into account to fully utilize the positive effects of social media on mental health.
This study still has some limitations. First, we were using cross-sectional data as the CHARLS question item on WeChat first appeared in 2018, so we were unable to predict the causal relationship between WeChat usage and social participation and depression. In future studies, we will consider using longitudinal data to study this aspect; Second, the effect of WeChat usage on depression may also arise through other mediating variables, but this study only explored the mediating role of social participation. We will look for the possibility of other mediating variables in future studies. Our classification of activities may not be sufficiently objective because we did not find a unified classification standard for social participation activities. Finally, we have adjusted for covariates as much as possible and used propensity score matching to make the results as robust as possible, but the factors affecting depression are complex and there are still some unobserved or unaccounted for biases.
Despite some limitations, this study also has some practical implications. First, we verified that WeChat usage and social participation were protective factors for depression in Chinese middle-aged and elderly people. Second, we found that WeChat usage could reduce the risk and levels of depression through social participation. This would imply that we can further promote social participation rates through social media usage to reduce the risk of depression among the middle-aged and elderly.
Conclusions
Our findings suggested that WeChat usage was significantly associated with lower risk of depression and lower levels of depression in a Chinese middle-aged and elderly population, with a partial mediating effect of social participation and recreational activities. These findings may have important implications for the prevention and treatment of mental health in middle-aged and older populations. Encouraging middle-aged and older adults to actively use social media and participate in social activities may be considered for the prevention and treatment of depression in the future to enhance mental health.
Supplementary Information
Additional file 1: S1. Regression results of WeChat usage and whether depressed in different groups. S2. Regression results of WeChat usage and depression levels in different groups. S3. Regression results for social participation and depression in the middle-aged population. S4. Regression results for social participation and depression levels in the middle-aged population. S5. Regression results for social participation and depression in the elderly. S6. Regression results for social participation and depression levels in the elderly. S7. Regression results for social participation and depression in women. S8. Regression results for social participation and depression levels in women. S9. Regression results for social participation and depression in men. S10. Regression results for social participation and depression levels in men. S11. KHB test in the middle-aged. S12. KHB test in the elderly. S13. KHB test in men.