| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
1 The Finnish Centre for Interdisciplinary Gerontology and 2 Department of Health Sciences, University of Jyväskylä, Finland.
Address correspondence to Tiina-Mari Lyyra, The Finnish Centre for Interdisciplinary Gerontology, P.O. Box 35 (Viveca), FIN-40014 University of Jyväskylä, Finland. E-Mail: tiina-mari.lyyra{at}sport.jyu.fi
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. We measured perceived social support in 206 Finnish men and women aged 80 years old by using the Social Provision Scale, which consists of six dimensions: attachment, social integration, opportunity for nurturance, reassurance of worth, reliable alliance, and guidance.
Results. By using a theoretical framework that divided perceived social support into assistance-related and non-assistance-related support, we found that the risk of death was almost 2.5 times higher in women in the lowest tertile of non-assistance-related social support (comprising infrequent experiences of reassurance of worth, emotional closeness, sense of belonging and opportunity for nurturance) than in women in the highest tertile. The risk remained strong even when we controlled for the indicators of baseline sociodemographics and psychological and physiological health and functioning. Among men, none of the perceived social support dimensions showed a significant association with mortality.
Discussion. The results of this study present a challenge for society to find and develop new social innovations and interventions in order to promote a sense of emotional social support in older people, thereby contributing to their health and welfare.
A GROWING number of studies have documented the beneficial effect of social support on various health outcomes, including survival. This beneficial effect has been confirmed not only in relation to all-cause mortality, but also in relation to several causes of death, including cancer, coronary heart disease, and other cardiovascular diseases (Avlund, Damsgaard, & Holstein, 1998
; Berkman & Syme, 1979; Blazer, 1982; Brummett et al., 2005; Ceria et al., 2001; Kaplan et al., 1988; Maier & Smith, 1999; Murberg, 2004; Seeman et al., 1993; Seeman, Kaplan, Knudsen, Cohen, & Guralnik, 1987; Temkin-Greener et al., 2004). Although the evidence for the protective effects of social support on health outcomes is growing, the important question of the relevant mechanisms and pathways remains unclear. In addition, it has proved difficult to determine exactly which aspects of social support have been responsible for the beneficial effects on mortality (Walter-Ginzburg, Blumstein, Chetrit, & Modan, 2002).
There are wide-ranging variations in the existing definitions and measures of social support. Generally, a distinction is drawn between structural and functional social support (Avlund et al., 1998; Avlund et al., 2004; Due, Holstein, Lund, Modvig, & Avlund, 1999): The former focuses on the existence of interconnection between social ties, whereas the latter focuses more on the specific functions that relationships serve (Uchino, 2004). Barrera (1986) proposed three kinds of social support: social embeddedness (e.g., indicators assessing the frequency of contacts), received support (e.g., measures of the amount of tangible help received), and perceived support (which he defined as subjective evaluations of supportive exchanges, such as satisfaction with social support). As Norris and Kaniasty (1996) also stated, perceived support exerts the strongest effects on health and well-being in old age.
Weiss (1973) proposed a theoretical model describing the provision of social relationships. He identified and described six categories of social provision based on the individual's assessment of how his or her social relationships worked. The six categories are attachment, social integration, reassurance of worth, opportunity for nurturance, guidance, and reliable alliance. The theory implies that all six categories of social provision are needed, although there may be some differences in their relative importance in different circumstances or in different phases of the life course. Attachment is provided by relationships whereby a person gains a sense of safety and security. Lack of a feeling of attachment may result from loneliness or emotional isolation. Social integration is provided by a network of relationships through which an individual shares common interests and concerns. The absence of these kinds of relationships may lead to social isolation. Reassurance of worth is related to relationships that acknowledge a person's skills and abilities. The fourth category, opportunity for nurturance, differs from the others: Here the individual is the provider, rather than the recipient, of assistance. The feeling given is one of being needed by others. In the conceptual framework developed by Cutrona and Russell (1987), these four categories are non-assistance-related, whereas Weiss's remaining two categories are assistance-related. These final two categories are guidance (more related to relationships with people who can provide knowledge, advice, and expertise) and reliable alliance (a person has people in his or her life who can be counted on or who can provide assistance in certain circumstances).
Broadly, there are two hypotheses relating to the possible causal pathways by which the protective effects of social support may act to reduce mortality. These are known as the stress-related and direct effect models. The stress-related model states that social support is protective because it diminishes the deleterious effects of stress in a person's life (Uchino, 2004). The direct effect hypothesis, in turn, works in two ways. First, environments are seen to influence attitudes and behaviors—including the regulation of health behavior and access to health care—by providing informational resources relating, for example, to economic aid and transportation (Berkman & Syme, 1979; Penninx et al., 1997). Secondly, there is an impact on self-esteem that is bolstered by feelings of belonging and also by a sense of security (Thoits, 1983; Van Baarsen, 2002). It has sometimes been difficult to evaluate theories linking social support to health outcomes, as conflicting findings may reflect differences in how one has defined and measured social support (Uchino, 2004).
Shye, Mullooly, Freeborn, and Pope (1995) argued that, despite the overall evidence that social support is associated with increased mortality risk, questions regarding different subgroups remained, particularly with regard to gender. Most studies have included gender and age as covariates in their analyses. In the few studies where analyses have been performed separately for men and women, the results have been conflicting. Jylhä and Aro (1989) and Seeman and colleagues (1993) found social support to be a significant predictor of survival in both men and women. In some studies, using a wide age range, the health effects were stronger in men than in women (House, Robbins, & Metzner, 1982; Kaplan et al., 1988). In the present study, we constructed separate models for men and women, and the people in the study population represent the same age cohort.
The objective of the present study was to examine the association between all-cause mortality and different dimensions of perceived social support at a 10-year follow-up. The underlying hypothesis of this study was that a lack of perceived social support would be associated with mortality in the 80-year-old Finnish population.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Eligible participants consisted of all 80-year-old residents (born in 1910) of Jyväskylä, who were alive at the beginning of the year 1990. Of these, 91.9% of the total sample took part in home interviews. The sample for the present study numbered 206 (61 men and 145 women). Of the 80-year-old men who answered the questions of the Social Provision Scale (SPS), only 31.% lived alone. Among the women, the situation was the reverse: More than 70% lived alone. The most common civil status among men was married (70.%); among women, widowed (59%). The most common reason for refusal to participate (ascertained by telephone) was illness. Dates of deaths obtained from the population register included all the deaths that occurred between January 1990 and March 2000.
Measures
Perceived social support was measured with the SPS (Cutrona & Russell, 1987; Mancini & Blieszner, 1992; Russell, Cutrona, Rose, & Yurko, 1984; Weiss, 1973), which measures the kind of support a person receives from his or her social network. The scale includes 24 items divided equally among six subscales (attachment, social integration, opportunity for nurturance, reassurance of worth, reliable alliance, and guidance), resulting in two items measuring the presence and two measuring the absence of each of the six dimensions. Respondents assessed on a scale from 1 (strongly disagree) to 4 (strongly agree) how each statement described their current social relationships. Russell and colleagues (1984) reported a reliability of 0.92 for the SPS, with subscale reliabilities ranging from 0.76 to 0.84. The stability and clear structure of the SPS have also been confirmed recently by Törmäkangas, Heikkinen, and Ilmarinen (2003). In addition to analyzing the individual subscales, we used sum scores derived from alliance, reassurance of worth, social integration, and opportunity for nurturance subscales (representing non-assistance-related social support) and the guidance and alliance subscales (representing assistance-related social support) in our analyses.
We calculated survival time in days from the last baseline examination date (January 31, 1990) to either the date of death or the end of the follow-up period (March 31, 2000).
We measured education by using self-reports of years of formal schooling.
We assessed depressiveness by using the Center for Epidemiologic Studies–Depression scale, which was developed for the purpose of screening for depressive symptomatology in community and epidemiological studies (Radloff, 1977). The questionnaire consists of 20 items, each rated on a 4-point scale indicating the frequency of mood problems during the past week. We used a sum score of the items as a continuous variable. A higher score indicated more depressive symptoms.
We examined cognitive performance by using three tests: Digit Symbol, Digit Span, and Word Fluency. Digit Symbol and Digit Span are subscales of the Wechsler Adult Intelligence Scale and measure level of concentration, learning and visuospatial abilities, and visual–motor coordination. Word Fluency is a measure of verbal fluency and speed of mental processes (Steen, Fromholt, Äystö, & Berg, 1997).
We measured mobility by using a stopwatch to record maximum walking speed over 10 meters in the laboratory corridor (Era & Rantanen, 1997).
In order to measure the prevalence of serious diseases, a physician first ascertained dichotomized diagnoses (0 = no, 1 = yes). Then we counted those that we considered fatal as a sum score. These included ischemic heart disease, cardiac insufficiency, heart infarction, cerebral infarction, chronic obstructive pulmonary disease, and cancers.
Statistical Analysis
We used Student's t test to compare the differences in means of the continuous baseline measurements between the nonsurvivors and survivors. For the dichotomous variables, we used the chi-square test. We studied the association between the covariates and mortality by using Cox regression models (Collett, 2003; Cox & Oakes, 1987). The few missing values were imputed by the regression method, which replaces missing values with the linear trend for that point. The estimated values were based on regression analysis of the existing ones.
In the multivariate Cox regression analyses, we considered diverse indicators of baseline sociodemographics and psychological and physiological health and functioning that might account for possible associations between perceived social support and mortality. We selected these indicators from those suggested by earlier studies as possible predictors of mortality or potential mediating factors. We added the terms into the core model in four combinations. We performed analyses by using SPSS 12.0 (SPSS, Chicago, IL).
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
. We then divided these sum scores into tertiles. Figure 1 presents the results of the univariate Cox regression models for men and women. The sum score of assistance-related social support derived from guidance (advice or information) and reliable alliance (assurance that there are people who can be counted on or who are able to provide assistance in certain circumstances) had no significant association with mortality in the 80-year-old women (hazard ratio [HR] = 1.43, 95% confidence interval [CI] = 0.88–2.37). Neither assistance-related nor non-assistance-related social support were significantly associated with mortality in the men (HR = 1.37, 95% CI = 0.69-2.71; and HR = 0.61, 95% CI = 0.29–1.30, respectively).
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
In a conceptual framework supported by initial analyses, the items of Weiss's SPS were divided into two categories: assistance-related and non-assistance–related perceived social support (Cutrona & Russell, 1987). This study yielded three important results. First, it provides evidence of the beneficial effect of non-assistance-related perceived social support (consisting of reassurance of worth, emotional closeness, a sense of belonging, and an opportunity for nurturance) on the survival of older women. The women in the tertile that received the least non-assistance-related social support had a risk for death almost 2.5 times higher than that of the women in the highest tertile. The risk remained high when we controlled for the effects of sociodemographics, health, and functioning. Second, we found that assistance-related social support (consisting of guidance [advice or information] and reliable alliance [assurance that there are people who can be counted on in certain circumstances]) was not significantly related to mortality in the older women. Third, in this study neither assistance-related nor non-assistance-related perceived social support showed a significant association with mortality in men.
Identity theorists argue that being embedded in a social network is protective because it gives the individual meaningful roles; these, in turn, generate self-esteem and a sense of purpose in life (Thoits, 1983; Uchino, 2004). The absence of close attachments and recognition of worth causes emotional loneliness and depressiveness (Heikkinen & Kauppinen, 2004; Taylor & Lynch, 2004). It has been suggested that there is also a link between self-esteem and mortality (Stamatakis et al., 2004). The feeling of being needed and valued is important; it gives one the strength to take care of oneself. Those who are alone and forgotten, despite having formal support, are at a higher risk for death. In this study, non-assistance-related social support (consisting of feelings of worth, emotional closeness, belonging, and an opportunity for nurturance) was strongly related to mortality in older women. Our finding that this is a more important predictor of survival than assistance-related social support is new. The dimensions of social support included in non-assistance-related social support, as defined here, have a strong emotional component.
Our study showed no significant association with mortality in any of the dimensions of social support among the men. This is contradictory to the findings of earlier studies; for example, House and colleagues (1982) and Kaplan and associates (1988) found the reverse to be true. However, the ages of the participants in these studies varied widely, and studies with exclusively older samples have found social support to have a beneficial effect on mortality for both genders (Jylhä & Aro, 1989; Seeman et al., 1993). Men tend to maintain close, intimate relationships with only a few people, primarily with their spouses. Being married has been found to be more protective for men than for women (Berkman & Syme, 1979). In the present study, the majority of the men were married, whereas most of the women were widowed and lived alone. For these women who lived alone, the quality of social relationships seems to have been more important than it was for married men.
We are aware of some limitations in the present study. First, the study sample was quite small. Studying cause-specific mortality might have been more informative; however, we chose to use all-cause mortality as an end point, partly because it would have been difficult to ascertain the exact causes of death for the elderly participants.
Despite these potential limitations, the study population was well suited for these analyses. This was a prospective epidemiological cohort study, in which the study population comprised the entire age cohort born in 1910 and living in the same town; the rate of participation was high. A sufficient number of deaths occurred during the relatively long follow-up period to enable the required statistical analyses to be performed. Peduzzi, Concato, Kemper, Holford, and Feinstein (1996) demonstrated the Rule of 10 by showing that at least 10 events per parameter are necessary in order to obtain reliable estimates of regression coefficients and their standard errors. Also, because of the infrastructure of Finnish society, the data on the dates of death were reliable and conclusive.
Many models exist on the possible relationship between social support and mortality, and they cannot all be studied here. For instance, future research might test how social support influences access to health and social services, thereby influencing health and survival. More research is also needed on the linkages between social support, psychological processes, and relevant health outcomes. For instance, investigating different coping mechanisms and the meaningfulness of life could be a promising way to approach the link between perceived social support and mortality in elderly populations.
In conclusion, this study found a strong association between non-assistance-related perceived social support (consisting of feelings of worth, emotional closeness, belonging, and an opportunity for nurturance) and survival in older women. These associations held when controlled for a broad set of demographic and physical and psychological health and functioning variables. The results of this study present a challenge for society to find and develop new social innovations and interventions in order to promote a sense of emotional social support in older people, thereby promoting their health and welfare. It is important to make sure that older women have access to various kinds of social relationships in which they can experience a sense of proximity and togetherness, as well as respect.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received for publication February 11, 2005. Accepted for publication December 13, 2005.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T.-M. Lyyra, T. M. Tormakangas, S. Read, T. Rantanen, and S. Berg Satisfaction with present life predicts survival in octogenarians. J. Gerontol. B. Psychol. Sci. Soc. Sci., November 1, 2006; 61(6): P319 - P326. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
2 Tests.