Family Support on EWB

Family members are core providers of material aid and instrumental support, who also take on important roles for the EWB of the elderly [12], [13]. On one hand, numerous studies found family support has a strong effect on reducing older adults' negative affect, and this effect is even stronger than it does for friend support. Those elderly women with lower perceived family support were associated with higher negative affect and feeling of loneliness, but their levels of friend support or objective network embedment were not relevant to the negative affect [4]. Similar results were found among Chinese elderly participants, showing satisfaction with family support as the primary predictor for less depressive symptoms compared to objective network relations [22], [23]. It was further discovered that material aid and instrumental support are more important in ameliorating negative affect when provided by family members than provided by friends or acquaintances [24]–[27]. In a longitudinal study among Chinese elderly, those with more depressive symptoms are associated with increased family support but decreased friend support three years after [23]. Such results added weights to the importance of family support on decreasing the negative affect of the elderly.

On the other hand, other studies suggested that family support is helpful in increasing positive affect of the elderly. One American study found that contacting more frequently with family members had a beneficial but weak effect on morale, although relatively less so compared with the effect of interacting with friends [28]. In China, family support similarly was associated with perception of happiness among aged people [29], [30]. Both in western and eastern cultures, nevertheless, the influences of family support on positive affect were less studied.

To better discern the effect of family support on EWB, the characteristic of family support may be of relevance. Family relationship is continuously bonded by aid, affect, and affirmation and is difficult to break apart across lifespan. Empirical data confirmed that intimate, attachment relationships were most frequently occurred within family members [7], [30]–[34], therefore they are often considered as the first choice of the elderly when requesting tangible aid [11], [12].

To sum up, family support is closely associated with the EWB of the elderly [11], [12], [30]–[34]. It is noted that many studies considered family support being of particular relevance to the amelioration of negative affect, and a few studies also showed its relevance to the enhancement of positive affect. In particular, compared with other members of a family, spouse is the core supporter and plays the most important role for providing a buffer for the elderly to reduce negative affect.

Friend Support on EWB

Apart from family members, friends also play non-negligible roles on the elderly's EWB. While most studies on friend support were concerned with positive affect, relatively fewer studies have focused on negative affect of the elderly.

In terms of positive affect, past studies showed frequencies of contacting with friends were more strongly associated with life satisfaction, happiness, and self esteem of the elderly compared with children support and served as a primary predictor for morale than family support [2]–[4], [28], [35], [36]. In terms of quality of relationships, studies again suggested that older adults reported higher levels of enjoyment and better self-esteem when interacted with friends compared with family members [37]–[39].

When it comes to negative affect, still some studies revealed that support from friends prevents or reduces the feeling of loneliness [36], [40], [41]. Perceived support from both family members and friends were associated with less depressive symptoms [1]. In the context of China, friend support was found unrelated to depression levels, it was suggested that perhaps Chinese older people less relied on friend support for ameliorating negative affect when facing difficulties [22], [42].

Results of friend support on EWB may also be influenced by the characteristic of friend interactions. Compared to family support, friend support is generally considered as great sources for those aspects of EWB related to current state of enjoyment [2]. Networking with friends tends to be a matter of choice, which is spontaneous and reciprocal [35], [36], [43], rather than predetermined as in kinship relations [11], [12], [44]. As birds of a feather flocked together [45]–[47], friends may be more open to leisure activities and conversations of mutual interests, which are oriented towards the enhancement of positive affect [14], [48].

Related Theories

To clarify whether different sources of social support have the same or different effect on either positive or negative affect, the hierarchical compensatory model and the task specificity model provided two guiding perspectives useful for establishing frameworks in studying effects of social support on different facets of the EWB [49].

According to the hierarchical compensatory model [13], [50], people seek social support from others based on a set of ordered preferences rather than the types of assistance required. Once the initially preferred sources of support are absent or insufficient, individuals from the lower hierarchies would replace them in a compensatory manner [13], [50], [51]. This description bears similarity to the convoy model that the levels of social support reduces as one moved from circles that are closer and more stable (i.e., spouse and children) to ones that are less so (i.e., friends, neighbors, and colleagues) [52]. In contrast, the task specificity model proposed that different groups of people work on specifically tailored tasks or provide different sorts of social support, and if the elderly are deprived of support from certain groups (e.g., friends), their corresponding tasks cannot be substituted or can only be partly supplemented by others.

The Present Study

Previous literature so far looked at either different source of social support on the same facet of the EWB (e.g., family and friend support on negative affect) or specific type of social support on different facets of EWB (e.g., friend support on positive and negative affect). Despite a few inconsistent results, most studies seemingly suggested that family support outdoes friend support in predicting negative affect levels; and friend support are more important than family support in predicting positive affect levels.

To our knowledge, nevertheless, few except one survey assessed family and friend support simultaneously among 60 older women, and found that family (both spouse and children) support and perceived friend support were predictive of positive affect, and family quality was predictive of negative affect [53]. Yet the survey was limited in sample size and representativeness. Therefore, the question how sources of social support would influence different aspects of the EWB needs to be reassessed through the lens of the hierarchical compensatory model and the task specificity model.

Moreover, from perspectives of cultural norms, family plays a pivotal role in supporting the elderly in China compared with other sources of social support [5], [22], [42]. On the other hand, influenced by filial piety, the preferred living arrangement in China for the elderly remained to be co-residence with children, especially when an older adult's spouse was no longer available [30], [31]. Therefore, children ranked just below spouse in providing critical instrumental support for the elderly [21], [54], [55]. Based on the hierarchical compensatory model, it is of interest to analyze how children support was related to EWB among older adults without spouse than those with spouse in the context of Chinese culture.

The purpose of this study is to compare the relative importance of spouse, children and friend support on positive and negative affect of the EWB. Under the guidance of the theories presented above and based on previous studies, we considered that to be consistent with either the hierarchical compensatory model or the task specificity model, results would likely to be different.

If results were consistent with the hierarchical compensatory model, support from spouse, children, and friends would likely follow a descending order in their associations with both positive and negative affect; and when the spouse was unavailable or absent, it would be replaced by children and friend support in sequence.

Alternatively, if results were consistent with the task specificity model, spouse, children and friend support may be related to positive and negative affect differently; and the effect of spouse, children, and friend support on positive and negative affect may or may not follow a descending order.

Ethics Statement

This research was approved by the Ethics Committee of the Institute of Psychology, Chinese Academy of Sciences. All participants provided written informed consent to participate in the study.

Participants and Procedures

We finally recruited 700 older adults based on convenient sampling, who came from communities of Haidian (342 persons), Dongcheng (173 persons) and Fengtai (185 persons) districts in Beijing. The voluntary nature of the participation and the confidentiality of their data were assured for the participants. Of those who agreed to participate, another 49 participants were removed from the analyses, since they did not complete the survey. The completion rate was 92.8%. In addition, five respondents were excluded since they were never married, hence cannot be assessed in terms of spouse or children support. We excluded them from the analyses and confirmed that our results were not sensitive to this choice. Interviews were conducted with the participants either at their homes or in the community center. Participants answered the questionnaire by themselves or under the instruction of the investigators. The survey typically took 30 to 45 minutes to complete. Each participant was offered a small token for their participation.

The sociodemographical data of this sample were presented in Table 1. Briefly, the sample was predominantly young old adults, relatively well-educated but financially deprived, which was typical of the urban population in China. Although 13.1% of the participants reported more than two chronic diseases and 13.9% reported in bad or very bad health status, most of the participants were physically healthy and no one was confined to their homes. Most of the participants lived with spouse and/or children, and did not experience any negative life events during the last two weeks. In addition, 21.9% of the participants were widowed/divorced, including 19.9% widowed and 2.0% divorced. Widowed/divorced participants were older, less educated, and poorer in financial status.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive Characteristic of Participants with Different Marital Status.

https://doi.org/10.1371/journal.pone.0090051.t001

Measurement

Family and Friend Support: Social support was measured by three questions as below, namely (1) how frequently have you contact with him/her/them? (2) how extensive have he/she/they provide support to you? (3) how extensive have you provide support to him/her/them? Participants were asked to rate the contact frequency and mutual support under (a) spouse, (b) children, (c) friends/neighbors/colleagues. Questions were rated on a 5-point Likert scale, ranging from 1 (very rarely) to 5 (very frequently). Participants without spouse or children rated as 0 (not available) in questions regarding to spouse or children support respectively. In this study, social support was represented as the average score of these three questions.

For those elderly with spouse, family support included spouse support and children support, while for widowed or divorced elderly, family support is represented only by children support. In China, the concept of “friends” is defined as “those who shared a good relationship with each other” according to the Contemporary Chinese Dictionary (the 6th edition, 2013). In fact, most friends “who shared a good relationship with each other” are the colleagues and neighbors of the Chinese elderly. Thus the ambiguous definition of friends was extended to “friends/colleagues/neighbors who shared a good relationship with you” as a specified working definition in the present study. On the other hand, the roles of friends, neighbors, and colleagues to the elderly may be overlapped. It was regarded friend support and neighbor support as a similar support form [56]. For most people, both friends and colleagues could provide companionship and emotional support to the elderly [57], [58]. In the later section of this article, “friends/colleagues/neighbors who shared a good relationship with you” were named after “friends” for short.

The internal consistency of the social support items was adequate in its entire form (Cronbach α = .84) or in terms of spouse, children, and friend support (Cronbach α between 0.83 and 0.98). The test-retest reliability in 44 participants during one month interval was 0.94 (p<0.001).

EWB: The Positive and Negative Affect: The EWB was measured by a 12-item affective wordlist, which was taken from Kahneman et al. [59]. Four word items are related to positive affect and eight word items are related to negative affect. Participants were asked to rate the 12 words or phrases they experienced during the past month, for example, happy, depressed/blue, impatient and so on. The scale rated on a 5-point Likert scale, ranging from 0 (never) to 4 (always). Positive affect is the average of ratings on happy, warm/friendly, enjoyment myself, competent. Higher score represents a more positive affect. Negative affect is the average of ratings on frustrated/annoyed, depressed/blue, hassled/pushed around, angry/hostile, worried/anxious, criticized/put down, impatient, tired. Higher score represents a more negative affect. In this study, the internal consistency (Cronbach α) of positive affect and negative affect items were 0.75 and 0.87 respectively.

Sociodemographic Information and Other Variables: The sociodemographic information (age, sex, marital status, education level, monthly income and income satisfaction), health status (number of chronic diseases, self-rated health status), living arrangement, and life events were included in the survey. Income satisfaction was measured on 5-point scale ranging from very unsatisfactory to very satisfactory. The list of chronic diseases included hypertension, coronary disease, cerebrovascular disease, diabetes, chronic bronchitis, spondylosis or osteoarthrosis, pulmonary disease (asthma, emphysema or cor pulmonale), cancer, and others (if not in the list). Self-rated health status was measured on 5-point scale ranging from very good to very bad. The list of life events included serious disease or injury of oneself, die of relatives, serious disease or injury of family members, children divorce, die of friends, bad relationship with spouse, bad relationship with children, suffering loss by theft or property loss, unforeseen events (unexpected frightening, having an accident, or suffering from natural disaster). In order to avoid collinearity problems, only age, sex, education, self-rated health, and life events were included in the final regression model.

Limitations

Self-report rather than objective measures of social support were used in the current study, thus the perceived social support might not reflect the actual situation. Under the cultural background which emphases family solidarity and filial piety, participants might have elevated the ratings for spouse and children support compared with friend support. When such inflations were removed, the magnitude of the hierarchy might be lessened or even becoming negligible. In addition, we admit that our loose definition of friends was distinctively differed to “close friends” or “best friends” as so defined in some previous studies, thus results cannot be generalized to such alternative definitions. Furthermore, the sample was recruited using convenient sampling from a group of urban dwelling older adults. A related issue was that most of participants were physically healthy, thus the generalizability of our study to China and other societies remains to be investigated. Finally, although the task specificity model was supported from our cross-sectional data, longitudinal data might provide fairer comparisons between the task specificity model and the hierarchical compensatory model.

Implications

Our study discovered that children could substitute the supportive roles for relieving negative affect when the spouses were not available; while friends play more important roles for enhancing positive affect. These results suggested that, in the context of China, advocating filial responsibilities remained to be a feasible option to ease the burden of social endowment; and furthermore, public programs should take considerations of establishing and strengthening community infrastructure and function to make communication more convenient among the elderly community dwellers.