Abstract

Objectives. This study examines whether aspects of social relations at baseline are related to functional decline at 5-year follow-up among nondisabled old men and women.

Methods. The investigation is based on baseline and follow-up data on 651 nondisabled 75-year-old persons in Jyväskylä (Finland) and Glostrup (Denmark). The analyses are performed separately for men and women. Possible selection problems were considered by using three outcome measures: first, functional decline among the survivors (n = 425); second, functional decline, including death, assuming that death is part of a general decline pattern (n = 565); and third, mortality (n = 651). Social relations were measured at baseline by several items focusing on the structure and function of the social network.

Results. In men, no weekly telephone contact was related to functional decline and mortality. Among women, less than weekly telephone contact, no membership in a retirement club, and not sewing for others were significantly related to functional decline and mortality. The associations were stronger when the dead were included in the outcome measure.

Discussion. The results point to the importance of social relations in the prevention of functional decline in older adults.

THE protective influence of social relationships on health is widely recognized in gerontological and public health research. During recent decades, findings from population studies have been remarkably consistent in demonstrating a beneficial effect of social relations on various health outcomes, including survival, morbidity, survival with serious illness, recovery from disability after acute medical conditions, depression, dementia, and well-being (for a review, see Seeman, 2000). Although this large body of evidence has caused a massive interest in the area of social relations, there is little consensus about the conceptualization of the term.

It is important to clarify and separate the concepts in this area of research because theories and empirical findings indicate that different mechanisms are responsible for the influence of social relations on health. In accordance with the conceptual framework by Due, Holstein, Lund, Modvig, and Avlund (1999), we define social relations as the main concept and the structure and function of social relations as subconcepts. The structure of social relations is defined as the individuals with whom one has an interpersonal relationship and the linkage between these individuals. This is generally studied in terms of the number of social relations that people have, the frequency of seeing other people, the diversity of social relations, and the reciprocity of social relations. The function of social relations is defined as the interpersonal interactions within the structure of the social relations. The function covers the qualitative and behavioral aspects of the social relations and comprises social support, social anchorage, and relational strain, typically measured as the relational content of social interactions (e.g., emotional or instrumental support). Relational strain is defined as the extent to which functions of social relationships cause emotional or instrumental strain and is thus the negative dimension of the functional aspect of social relations.

In this article we examine the degree to which social relations are related to functional decline, a critical measure of disability in older adults that has become an important issue in gerontological and public health research (Stuck et al., 1999). A limited number of studies have investigated the influence of structural and functional aspects of social relations on functional decline among well-functioning community-dwelling elderly people, although with contradictory results: Social network size has been identified as a protective factor in some (Mendes de Leon et al., 1999; Unger, McAvay, Bruce, Berkman, & Seeman, 1999), but not all (Liu, Liang, Muramatsu, & Sugisawa, 1995), studies of functional decline. Social participation has more consistently indicated a protective effect on functional ability (e.g., Lee 2000). Finally, Boult, Kane, Louis, Boult, and McCaffrey (1994) identified social support as a protective factor against functional decline, whereas Seeman and colleagues (1995) demonstrated that social support had a negative effect on function, and that emotional social support was protective only among those with little instrumental support. Mendes de Leon and colleagues, (1999) found no effect of social support at all. These inconsistent patterns may reflect different measures of social relations, gender differences, and selection bias that are due to loss to follow-up.

Measures used in the aforementioned studies reflect limited aspects of both structural and functional components of social relations. Other aspects of social relations, including telephone contact, diversity in social relations, and reciprocity of social relations, are also important and require study. The content of telephone contact may vary from that of face-to-face contacts (Johnson & Troll, 1994). Among older adults it is well known that social relations may create severe strain (Okun & Keith, 1998). It is possible that the issues discussed over the telephone are different or cause less strain than those taken up face-to-face. However, the effect of telephone contact on functional decline is unknown. Other studies have shown that diversity in social relations is a stronger predictor of morbidity and mortality than the number of friends in the network and the frequency of contact with those friends (Vogt, Mullooly, Ernst, Pope, & Hollis, 1992). In other words, what is important for health is not how often one interacts with people, or how many people one interacts with, but rather the number of different arenas of interaction represented in the network. This suggests that a critical issue in surviving serious illness may be the degree to which different types of resources are available and that having an appropriate resource for solving a problem is critical to recovery from illness. However, the influence of diversity in social relations on functional decline is not known. Antonucci and Akiyama (1987) have shown that the give and take relationship is important to both men and women and that both men and women prefer to provide more than they receive. They also showed that reciprocity of social relations was protective of morbidity, but to our knowledge no studies have investigated whether this factor influences functional decline.

Most studies have included gender as a covariate in the analysis (e.g., Mendes de Leon et al., 1999) and have not considered interactions. However, this approach can mask differential patterns of risk between men and women. In analyses of data from the MacArthur Studies of Successful Aging, Seeman, Bruce, and McAvay (1996) and Unger and colleagues (1999) demonstrated that poor social relations had a stronger effect on functional decline among men than among women. However, to our knowledge, no other studies have examined these associations separately for men and women.

Further, all longitudinal studies in populations of older people have large losses to follow-up because of death or nonparticipation for other reasons (Diehr et al., 1995). If loss to follow-up is not random, it may result in biased measures of associations. This is likely to happen in studies of social relations and functional decline, as both factors are related to mortality (Avlund, Schultz-Larsen, & Davidsen, 1998; Lund, Modvig, Due, & Holstein, 2000) and may be associated with nonparticipation (Hoeymans, Feskens, Bos, & Kromhout, 1998; Shaw, Cronan, & Christie, 1994). Consequently, it might be of interest to include the dead and nonparticipants in the analyses of functional decline. Nearly all studies evaluating the associations between social relations and functional decline have restricted their analyses to persons who survived and participated at follow-up (e.g., Seeman et al., 1996). Other studies exclusively studied the effect of social relations on mortality and showed that poor social relations were related to mortality, also in older populations (e.g., Lund et al., 2000). To our knowledge, no studies have included the dead and the nonparticipants in the analyses of social relations and functional decline.

The objective of the present study is to examine whether aspects of social relations at baseline are related to functional decline at 5-year follow-up among nondisabled old men and women. In this study we address some of the issues already discussed. First, we used measures of social relations that were not previously used in studies of functional decline (telephone contact, diversity in social relations, and reciprocity of social relations) as well as more traditional measures (club membership, social participation, and instrumental social support). Second, we performed the analyses separately for men and women. Third, we considered possible selection problems by using four outcome measures: (a) functional decline among the survivors, (b) functional decline, including death, assuming that death is part of a general decline pattern, (c) mortality, and (d) nonparticipation for reasons other than death.

Methods

Study Population

The data for this study were collected as part of the NORA study (Nordic Research on Ageing; Schroll, Steen, Berg, Heikkinen, & Viidik, 1993). The baseline study, performed in 1989–1990, included all citizens born in 1914 in Jyväskylä, Finland, and a random sample of persons born in 1914 who were living in Glostrup, a suburban area of Copenhagen (N = 835). Follow-up data were collected after 5 years. To select an initially nondisabled cohort, we excluded 184 persons reporting need of help in at least one of the following daily activities: combing hair, washing upper and lower body, using the toilet, dressing upper and lower body, and cutting fingernails and toenails (n = 651). At 5-year follow-up, 140 persons had died and 86 surviving subjects did not want to participate. After exclusions, the study population for the present investigation consisted of 425 older adults (participation rate 83%) in the analyses among the survivors, 565 persons in the analyses including the deceased as part of a disability pattern, and 651 persons in the analyses with mortality and nonparticipation for other reasons as outcomes (see Table 1).

Variables

Four outcome measures were used. The first was functional decline among the survivors (n = 425), based on a validated scale (the PADL-H Scale) about need of help in combing hair, washing the upper and lower body, using the toilet, dressing the upper and lower body, and cutting fingernails and toenails. Functional decline was measured both as a continuous and a dichotomized scale (sustained no need of help during follow-up vs. persons who developed need of help in at least one activity). This cut point was chosen because of the clinical relevance for older adults of being able to manage without help (Guralnik et al., 1993). Reliability tests on the PADL-H Scale showed agreement percentages from 98.1 to 1.0 and kappa values from 0.73 to 1.0 for the included items on intrarater and interrater tests (Avlund, Thudium, Davidsen, & Fuglsang-Sørensen, 1995). Item bias analyses showed that all participants in the baseline study at a given index score on the PADL-H Scale had the same probability of scoring on an item, regardless of gender and geographic locality (p <.01; Avlund, Era, Davidsen, & Gause-Nilsson, 1996). Further, functional ability as measured by the scale was strongly associated with isometric muscle strength (p <.01 in relation to hand grip, arm flexion, knee extension, and body extension), simple function tests (p <.05 in relation to walking speed and step test; Avlund, Schroll, Davidsen, Løvborg, & Rantanen, 1994) and postural balance (p <.05 in relation to tests for speed of anteroposterior and mediolateral movement of the center of pressure and mean moment of velocity; Era et al., 1997).

The second measure was functional decline for persons from age 75 to 80, including the dead (n = 565). Persons with decline are defined as those who had become in need of help or died. Death is thus regarded as part of a functional decline pattern.

The third measure was mortality (n = 651), measured by all-cause mortality obtained from the Civil Registration Office 5 years after the end of the baseline study, in November 1994.

The fourth measure was nonparticipation for other reasons than death (n = 651).

Measures of Social Relations

The structure of social relations was measured by five aspects: first, telephone contact with children and friends (weekly and less often); second, membership in a club for retired people (yes or no); third, social participation (three items about paying visits to others, receiving visits at home, and participating in social activities outside the home: range 0–3); fourth, diversity of social relations (number of different categories with whom one has personal contact at least once a week: children, grandchildren or great-grandchildren, relatives, confidants, acquaintances, and neighbors: range 0–6); and fifth, reciprocity of social relations by helping others (five items about help from the participants to their social network during the last month: take care of; have on vacation; sew, knit, or repair clothes; make repairs; and do housework).

The function of social relations was measured by instrumental social support (support to housework, cooking, shopping, and repairs in the house by relatives or friends: range 0 = no support to 8 = much support). A test–retest study of the included items of social relations showed that the agreement percentages were between 72 and 100 and the kappa values were between 0.501 and 1.0 for all items, with the structural items having the highest reliability. Further, in-depth interviews demonstrated high face and content validity of the social relations measures used (Due et al., 1999).

Covariates

The following factors known to be associated with both the determinant and outcome measures were included as potential confounders, all measured at baseline. Live Alone (yes or no; Sarwari, Fredman, Langenberg, & Magaziner, 1998; Wu & Pollard, 1998) was one such factor. Number of Chronic Diseases (Boult et al., 1994; Vogt et al., 1994) was measured in connection with the medical examination by an open-ended question: “Do you suffer from any longstanding illness?” If the answer was positive, questions were asked about type, localization, duration, medical diagnosis, and the like. The answers were coded in accordance with WHO ICD-8, at the end of the examination day (Schroll et al., 1993). There was a range of from zero to seven diseases.

Cognitive Function (Avlund, Fromholt, Berg, & Davidsen, 1997; Fratiglioni, Wang, Ericsson, Maytan, & Winblad, 2000) was measured by a trained psychologist and included the intelligence and memory tests Digit Span, Digit Symbol, Word Fluency, Visual Reproduction, and Raven's Progressive Matrices. These five variables were combined into one measure of cognitive function based on factor analysis (Avlund, Fromholt, et al., 1997), which expresses the mean factor score for the cognitive function tests. These tests were included to measure performance “to the limit” by stressing speed, accuracy, and span. We found that, in particular, two instrumental activity of daily living (IADL) tasks (use public transportation and manage economy) were significantly related (p <.05) to the used measure of cognitive performance (Avlund & Fromholt, 1998). These two IADL tasks are complex and involve orientation and procedural skills that might be affected very early in the dementing process. Thus we feel confident that the used measure of cognitive function may be a relevant measure in the present population.

The Depressive Symptoms factor (Penninx, Leveille, Ferrucci, van Eijk, & Guralnik, 1999) was measured by use of the Center for Epidemiological Studies Depression scale (CES-D; Radloff, 1977) based on 20 items, each rated from 0 (rarely or none of the time) to 3 (most of the time). The total score is a sum of all items, ranging from 0 to 60. The internal consistency for the CES-D scale was high (Cronbach's alpha above 0.80) in both localities and across genders (Heikkinen, Berg, & Avlund, 1997). We also included Nordic Locality (Jyväskylä or Glostrup) as a covariate because patterns of associations may be different in the two localities (Avlund, Holstein, Heikkinen, & Schroll, 1997).

Statistical Analysis

All analyses were performed separately for men and women in several steps. The first step was to test whether the social relations variables were associated with the outcome measures at the bivariate level by use of chi-square tests. The multivariate logistic regression analyses included several steps (see Table 2). Initially, all analyses were performed by using all categories for each measure of social relations. Then, if the results were in the same direction for some of the categories, and if there were few persons in some of the categories, we combined them as described in Table 3. The SAS PROBIT procedure was used. Finally, multiple linear regression analyses were used to test whether the results were the same when the outcome measure was used as a continuous scale. The SAS PROG REG procedure was used.

Results

Approximately half of the men and women had sustained good functional ability from baseline to follow-up (Table 1). More women than men survived with functional decline, and more men than women died. A large proportion of men and women had weekly telephone contacts, high social participation, and large diversity in the social network, and were members of a retirement club at baseline (Table 3). In general, women had stronger social relations than men. The distributions of social relations were approximately the same in a study population restricted to those who survived and participated in the follow-up study.

Men

The bivariate analyses showed that only structural components of social relations (less than weekly telephone contact, and not being member of a club) were associated with one of the outcome measures (p <.2; Table 4). Consequently, these variables were incorporated in the multivariate logistic regression analyses (Table 5). Less than weekly telephone contact was significantly related to decline, both when the dead were included as part of the functional decline measure and when mortality was used as a separate outcome measure. This result was in the same direction, but far from significant when the dead were excluded from the analysis. In addition, the odds ratios (ORs) of no club membership were 2.1 (confidence interval or CI = 0.8–5.7) on functional decline without the dead, 1.6 (CI = 0.9–3.0) when the dead were included in the functional decline measure, and 1.1 (CI = 0.6–1.9) on mortality. The linear regression analyses showed that some of the associations disappeared when functional decline was entered as a continuous variable. The ORs of no club membership and low social participation on nonparticipation among the survivors were 2.0 (CI = 0.8–4.6) and 2.3 (CI = 0.9–5.8), whereas the associations between the other social relations variables and nonparticipation were even more insecure.

Women

The result of the bivariate analyses among women was that both structural (less than weekly telephone contact, not being member of a club, low social participation) and functional components of social relations (no social support, not taking care of others, and not sewing for others) were associated with one of the outcome measures (p <.2; Table 4). Consequently, these variables were incorporated in the multivariate logistic regression analyses (Table 6). Only one social relations variable, not sewing for others, was significantly related to functional decline when the dead were excluded from the analyses. When the dead were included as part of the functional decline measure, less than weekly telephone contact, no membership in a retirement club, and not sewing for others were significantly related to decline. These results remained significant when mortality was used as a separate outcome measure. In addition, not taking care of others was significantly related to death. These results persisted when adjusted by the covariates and when the functional decline measures were included as continuous variables, except for no club membership among the survivors and not sewing for others. A logistic regression analysis with nonparticipation for reasons other than death as outcome showed that no club membership (OR = 3.5; CI = 1.4–8.3) and no social support (OR = 2.7; CI = 1.3–5.3) were significantly related to nonparticipation.

Discussion

This article focused on the impact of structural and functional characteristics of social relations on functional decline at 5-year follow-up for a cohort of relatively well-functioning older men and women who were aged 75 at baseline. The first important result is the beneficial effect among both men and women of having weekly telephone contact. This supports other research of the protective effect of the structural aspects of social relations and is generally consistent with findings from several previous studies of the role of social ties in disability (Mendes de Leon et al., 1999; Unger et al., 1999). It is of interest that the findings showed more protective effects of weekly telephone contacts than of face-to-face contacts. This is in agreement with a qualitative study among individuals aged 85 years and older that showed that elderly people no longer required face-to-face contacts in their friendships in order to sustain feelings of closeness (Johnson & Troll, 1994). It is possible that face-to-face contacts, such as frequent visits by a family member, could result in a decline in function, because these family members perform tasks for the old person instead of encouraging the person to do the activities themselves. In the long run this would result in a subsequent loss of function. Another explanation may be that many face-to-face contacts may be an early indicator of difficulties performing activities of daily living tasks, so that someone often comes to help. In contrast, it is possible that telephone contacts cause less strain and include more positive informational support by verbal encouragement, thereby boosting self-efficacy expectations and encouraging behavior (Cresci, 2001).

The second important result is that indicators of poor social participation (no membership in a retirement club and not sewing for others) are related to functional decline among women. Other studies have shown protective effects of group membership (e.g., Wolinsky, Stump, Callahan, and Johnson, 1996), and one study has shown the protective effects of reciprocity of social relations (Antonucci & Akiyama, 1987). The explanation may be that women who are club members or help others actually have a greater number of roles, thereby having a greater level of social anchorage. Moen, Dempster-McClain, and Williams (1989) explained this by means of the role enhancement perspective, which points to the way in which multiple roles augment an individual's power, prestige, resources, and emotional gratifications, including social recognition and a heightened sense of identity. Moen and associates (1989) and Adelman (1994) showed that loss of roles affected both physical and psychological well-being.

Along with Akiyama, Elliott, and Antonucci (1996) we found that men's social relations in general were smaller than women's, but the results also indicate that more aspects of social relations are related to functional decline among women than men. This is in contradiction to the studies by Seeman and colleagues (1996) and Unger and colleagues (1999). The present findings may be partly due to the social relations variables available and used in the study. Akiyama and associates (1996) showed that men and women serve intricately different functions in the various types of relations and situational circumstances that evolve around aging individuals. It is thus possible that the use of other indicators of social relations would show different patterns.

The third important result is that the associations between social relations and functional decline are in the same direction with regard to all three outcome measures, with only one exception (take care of others, among the women). The associations were stronger when the dead were included in the analyses of functional decline than when they were excluded from the analyses. These results are in accordance with Lee (2000), who found that persons with low social activity were at larger risk of both functional decline among the survivors over 6 years and mortality over 7 years. However, to our knowledge, no other studies of this kind have had a special focus on the influence of various aspects of social relations on functional decline and death as in the present study.

Some methodological aspects should be considered in the interpretation of our results. The response rate at follow-up was relatively high (83%). However, men and women with poor social relations were more likely to be nonparticipants than others. We do not think that this has resulted in bias, as social relations are related to functional decline, mortality, and nonparticipation. As a consequence of the longitudinal design based on nondisabled persons at baseline, we can exclude the possibility of health selection, that is, that persons in poor health are more likely to have poor social relations.

With regard to loss to follow-up because of deaths, we found that the associations between social relations and the outcomes were in the same direction for functional decline among the survivors, in the group with functional decline or death and among the dead. This means that exclusion of the dead from the analyses does not result in bias, but in insecure estimations of the associations.

It is a strength that the measures of functional ability and social relations have been validated in several ways and that the covariates included objective measures of chronic diseases (self-reported in a clinical examination and coded by a physician), cognitive function (measured by a trained psychologist), and a validated scale of depressive symptoms (the CES-D scale). However, it is a weakness of the study that we do not have information about the quality of social relations, for example, about emotional support and whether the social exchanges with key relationships were positive or negative. It is possible that these variables exert additive effects on functional decline or that negative social exchange with some persons are buffered by positive social interactions with other persons (Okun & Keith, 1998). The present data did not include information about possible changes in social relations preceding the baseline study, and thus it is not known whether changes in social relations over time accompany functional decline.

The present study underlines the importance of social relations for maintaining a good functional ability in old age. Consequently, a focus on the health-protective aspects of social relations in health promotion among older adults is recommended.

Decision Editor: Charles F. Longino, Jr., PhD

Table 1.

Functional Decline, Death, and Nonparticipation From Baseline to 5-Year Follow-Up.

 Men  Women   
Follow-Up Status n n p 
Sustained good functional ability 140 51 193 52  
Functional decline 27 10 65 17  
Death 76 27 64 17  
Nonparticipation 34 12 52 14 .001 
 Men  Women   
Follow-Up Status n n p 
Sustained good functional ability 140 51 193 52  
Functional decline 27 10 65 17  
Death 76 27 64 17  
Nonparticipation 34 12 52 14 .001 

Notes: Functional decline, death, and nonparticipation are among nondisabled men (n = 277) and women (n = 374); p = Pearson's chi-square test for equal distribution between men and women.

Table 2.

Variables Entered in Each Model of the Multivariate Logistic Regression Analyses.

Crude analysis Only variables that were related to at least one of the outcome measures in the bivariate analyses (see Table 3; p <.2) 
Model 1 Social relations variables that were related to the specific outcome measures in the crude analyses (p <.2) 
Model 2 Social relations variables that were related to the specific outcome measures in Model 1 (p <.2) 
Model 3 The variables from Model 2 and live alone 
Final model The variables from Model 3, if related to the outcome measure (p <.2) and chronic diseases, cognitive function, depressive symptoms, and locality 
Crude analysis Only variables that were related to at least one of the outcome measures in the bivariate analyses (see Table 3; p <.2) 
Model 1 Social relations variables that were related to the specific outcome measures in the crude analyses (p <.2) 
Model 2 Social relations variables that were related to the specific outcome measures in Model 1 (p <.2) 
Model 3 The variables from Model 2 and live alone 
Final model The variables from Model 3, if related to the outcome measure (p <.2) and chronic diseases, cognitive function, depressive symptoms, and locality 
Table 3.

Social Relations at Baseline Among Nondisabled Men and Women.

 Men  Women  p
Aspects of Social Relations n n  
At least weekly telephone contact 266 30 359 67 .001 
Member of retirement club 277 37 373 36 ns 
High social participation 277 17 374 26 .05 
High social diversity (4–5) 277 373 11 ns 
No instrumental social support 277 81 374 45 .001 
Take care of others 274 16 373 24 .05 
Have had grandchildren, etc. on vacation 274 16 374 20 ns 
Have helped others by sewing and repairing clothes 274 372 26 .001 
Have made repairs for others 274 35 374 11 .001 
Have helped others with housework 275 15 374 24 .01 
 Men  Women  p
Aspects of Social Relations n n  
At least weekly telephone contact 266 30 359 67 .001 
Member of retirement club 277 37 373 36 ns 
High social participation 277 17 374 26 .05 
High social diversity (4–5) 277 373 11 ns 
No instrumental social support 277 81 374 45 .001 
Take care of others 274 16 373 24 .05 
Have had grandchildren, etc. on vacation 274 16 374 20 ns 
Have helped others by sewing and repairing clothes 274 372 26 .001 
Have made repairs for others 274 35 374 11 .001 
Have helped others with housework 275 15 374 24 .01 

Notes: p = Pearson's chi-square test for equal distribution of the social relations variables between men and women; ns = nonsignificant.

Table 4.

Social Relations Among Nondisabled Participants at Baseline and the Proportions Who Declined in Functional Ability or Died During Follow-Up.

 Men      Women      
 Functional Decline—Alive  Functional Decline Including the Dead  Mortality  Functional Decline—Alive  Functional Decline Including the Dead  Mortality  
Aspects of Social Relations n n n n n n 
Structural             
    Live alone 31 16 41 37 50 20 166 25 202 39 230 16 
    Live with others 136 16 ns 202 44 ns 227 29 ns 92 25 ns 120 43ns 144 19 ns 
    Weekly telephone contact 56 13 72 32 80 20 182 23 213 34 240 13 
    Less than weekly 103 17 ns 163 48* 186 32 ns 66 32* 95 53** 119 24** 
    Club member 68 12 95 37 103 26 115 22 129 30 136 10 
    Not club member 99 19 (*148 46 (*174 28 ns 142 28 ns 192 47** 237 21** 
    High social diversity (2–6) 92 13 136 41 154 29 173 25 208 38 233 15 
    Low social diversity (0–1) 75 20 ns 107 44 ns 123 26 ns 85 25 ns 114 44 ns 141 21 
    High social participation (2–3) 72 17 99 44 106 25 142 26 166 44 184 13 
    Low social participation (0–1) 95 16 ns 144 39 ns 171 29 ns 116 24 ns 156 37 190 21* 
Functional             
    Social support 34 17 49 41 52 29 155 25 190 38 207 17 
    No social support 133 15 ns 194 43 ns 225 27 ns 103 26 ns 132 42** 167 17 ns 
    Taken care of others 27 26 42 52 45 33 77 23 84 30 91 
    Have not 138 14 (*198 40 ns 229 26 ns 180 26 ns 237 43** 282 20** 
    Had on vacation — — 39 46 43 33 56 25 65 35 74 12 
    Had not — — 201 42 ns 231 26 ns 202 25 ns 257 41 ns 300 18 ns 
    Sewn for others — — — — — — 74 16 83 25 95 
    Have not — — — — — — 182 29* 237 46** 277 20 ns 
    Repaired for others 60 18 87 44 97 28 — — 36 44 40 20 
    Have not 105 15 ns 153 42 ns 177 27 — — 286 40 ns 334 17 ns 
    Helped with housework — — 35 46 40 28 64 27 81 42 89 19 
    Have not — — 206 42 ns 235 27 ns 194 25 ns 241 39 ns 285 16 
 Men      Women      
 Functional Decline—Alive  Functional Decline Including the Dead  Mortality  Functional Decline—Alive  Functional Decline Including the Dead  Mortality  
Aspects of Social Relations n n n n n n 
Structural             
    Live alone 31 16 41 37 50 20 166 25 202 39 230 16 
    Live with others 136 16 ns 202 44 ns 227 29 ns 92 25 ns 120 43ns 144 19 ns 
    Weekly telephone contact 56 13 72 32 80 20 182 23 213 34 240 13 
    Less than weekly 103 17 ns 163 48* 186 32 ns 66 32* 95 53** 119 24** 
    Club member 68 12 95 37 103 26 115 22 129 30 136 10 
    Not club member 99 19 (*148 46 (*174 28 ns 142 28 ns 192 47** 237 21** 
    High social diversity (2–6) 92 13 136 41 154 29 173 25 208 38 233 15 
    Low social diversity (0–1) 75 20 ns 107 44 ns 123 26 ns 85 25 ns 114 44 ns 141 21 
    High social participation (2–3) 72 17 99 44 106 25 142 26 166 44 184 13 
    Low social participation (0–1) 95 16 ns 144 39 ns 171 29 ns 116 24 ns 156 37 190 21* 
Functional             
    Social support 34 17 49 41 52 29 155 25 190 38 207 17 
    No social support 133 15 ns 194 43 ns 225 27 ns 103 26 ns 132 42** 167 17 ns 
    Taken care of others 27 26 42 52 45 33 77 23 84 30 91 
    Have not 138 14 (*198 40 ns 229 26 ns 180 26 ns 237 43** 282 20** 
    Had on vacation — — 39 46 43 33 56 25 65 35 74 12 
    Had not — — 201 42 ns 231 26 ns 202 25 ns 257 41 ns 300 18 ns 
    Sewn for others — — — — — — 74 16 83 25 95 
    Have not — — — — — — 182 29* 237 46** 277 20 ns 
    Repaired for others 60 18 87 44 97 28 — — 36 44 40 20 
    Have not 105 15 ns 153 42 ns 177 27 — — 286 40 ns 334 17 ns 
    Helped with housework — — 35 46 40 28 64 27 81 42 89 19 
    Have not — — 206 42 ns 235 27 ns 194 25 ns 241 39 ns 285 16 

Notes: ns = nonsignificant.

(*) p <.20; * p <.05;

** p <.01.

Table 5.

Social Relations as Determinant of Functional Decline: Men.

 Functional Decline—Alive     Functional Decline or Dead     Mortality  
 Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  
Aspects of Social Relations Crude Final Modela β SE p Crude Final Modela β SE p Crude Final Modela 
Live alone 1.0 (0.4–2.9) —    0.8 (0.4–1.5)     0.6 (0.3–1.3)  
No weekly telephone contact 1.5 (0.6–3.8) 2.3 (0.8–7.1) −0.070 0.191 .714 2.0 (1.1–3.5) 2.5 (1.3–4.9) −0.113 0.538 .040 1.9 (1.0–3.6) 2.1 (1.0–4.2) 
Not member of retirement club 1.8 (0.7–4.3) 2.1 (0.8–5.7) −0.299 0.181 .100 1.5 (0.9–2.5) 1.6 (0.9–3.0) −0.392 0.495 .429 1.1 (0.6–1.9)  
Do not take care of others 0.5 (0.2–1.3) —    0.9 (0.4–1.8) —    0.7 (0.4–1.4)  
 Functional Decline—Alive     Functional Decline or Dead     Mortality  
 Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  
Aspects of Social Relations Crude Final Modela β SE p Crude Final Modela β SE p Crude Final Modela 
Live alone 1.0 (0.4–2.9) —    0.8 (0.4–1.5)     0.6 (0.3–1.3)  
No weekly telephone contact 1.5 (0.6–3.8) 2.3 (0.8–7.1) −0.070 0.191 .714 2.0 (1.1–3.5) 2.5 (1.3–4.9) −0.113 0.538 .040 1.9 (1.0–3.6) 2.1 (1.0–4.2) 
Not member of retirement club 1.8 (0.7–4.3) 2.1 (0.8–5.7) −0.299 0.181 .100 1.5 (0.9–2.5) 1.6 (0.9–3.0) −0.392 0.495 .429 1.1 (0.6–1.9)  
Do not take care of others 0.5 (0.2–1.3) —    0.9 (0.4–1.8) —    0.7 (0.4–1.4)  

Notes: Significant associations (p <.05) are shown in bold type. Functional decline—alive, n = 167; functional decline or dead, n = 243; mortality, n = 277. OR = odds ratio; CI = confidence interval.

a Final model after the steps described in Table 2. Adjusted by the variables that in the intermediate models were related to the outcome (p <.20) and by chronic diseases, cognitive function, depressive symptoms, and locality.

b Adjusted by chronic diseases, cognitive function, depressive symptoms, and locality.

Table 6.

Social Relations as Determinant of Functional Decline: Women.

 Functional Decline—Alive     Functional Decline or Dead     Mortality  
 Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  
Aspects of Social Relations Crude Final Modela β SE p Crude Final Modela β SE p Crude Final Modela 
Live alone 1.0 (0.6–1.8) — —   0.9 (0.5–1.4)     0.8 (0.5–1.3)  
No weekly telephone contact 1.6 (0.9–3.0) 1.5 (0.8–3.0) −0.223 0.157 .153 2.2 (1.3–3.6) 1.8 (1.1–3.1) −0.879 0.396 .027 2.2 (1.2–3.8) 1.8 (0.95–3.3) 
Not member of retirement club 1.4 (0.8–2.5) 1.5 (0.8–2.8) 0.048 0.137 .728 2.0 (1.3–3.3) 2.1 (1.2–3.6) −0.966 0.362 .008 2.3 (1.2–4.4) 2.4 (1.2–4.8) 
Low social participation 1.1 (0.9–1.6) —    1.3 (0.9–2.1) —    1.8 (1.0–3.1)  
Do not take care of others 1.1 (0.6–2.1)     1.8 (1.1–3.1) 1.5 (0.8–2.8) −1.089 0.425 .011 3.0 (1.3–6.9) 2.7 (1.1–6.7) 
Do not sew for others 2.1 (1.1–4.3) 2.7 (1.2–5.7) −0.13 0.161 .428 2.5 (1.4–4.3) 2.5 (1.4–4.7) −0.509 0.435 .243 2.4 (1.1–5.0) 1.9 (0.9–4.2) 
 Functional Decline—Alive     Functional Decline or Dead     Mortality  
 Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  Linear Regression Adjustedb   Logistic Regression OR (95% CI)  
Aspects of Social Relations Crude Final Modela β SE p Crude Final Modela β SE p Crude Final Modela 
Live alone 1.0 (0.6–1.8) — —   0.9 (0.5–1.4)     0.8 (0.5–1.3)  
No weekly telephone contact 1.6 (0.9–3.0) 1.5 (0.8–3.0) −0.223 0.157 .153 2.2 (1.3–3.6) 1.8 (1.1–3.1) −0.879 0.396 .027 2.2 (1.2–3.8) 1.8 (0.95–3.3) 
Not member of retirement club 1.4 (0.8–2.5) 1.5 (0.8–2.8) 0.048 0.137 .728 2.0 (1.3–3.3) 2.1 (1.2–3.6) −0.966 0.362 .008 2.3 (1.2–4.4) 2.4 (1.2–4.8) 
Low social participation 1.1 (0.9–1.6) —    1.3 (0.9–2.1) —    1.8 (1.0–3.1)  
Do not take care of others 1.1 (0.6–2.1)     1.8 (1.1–3.1) 1.5 (0.8–2.8) −1.089 0.425 .011 3.0 (1.3–6.9) 2.7 (1.1–6.7) 
Do not sew for others 2.1 (1.1–4.3) 2.7 (1.2–5.7) −0.13 0.161 .428 2.5 (1.4–4.3) 2.5 (1.4–4.7) −0.509 0.435 .243 2.4 (1.1–5.0) 1.9 (0.9–4.2) 

Notes: Significant associations (p <.05) are shown in bold type. Functional decline—alive, n = 258; functional decline or dead, n = 322; mortality, n = 374. OR = odds ratio; CI = confidence interval.

a Final model after the steps described in Table 2. Adjusted by the variables that in the intermediate models were related to the outcome (p <.20) and by chronic diseases, cognitive function, depressive symptoms, and locality.

b Adjusted by chronic diseases, cognitive function, depressive symptoms, and locality.

Support for this research was provided by a research grant from The Danish Medical Research Council, the Danielsen Foundation, the Wedell-Wedellsborg Foundation, the Academy of Finland, the Ministry of Education in Finland, the Ministry of Social Affairs and Health in Finland, the Social Insurance Institution in Finland, and the City of Jyväskylä.

References

Adelman, P. K.
1994
. Multiple roles and psychological well-being in a national sample of older adults.
Journal of Gerontology: Social Sciences
 ,
49
,
S277
-S285.
Akiyama, H., Elliott, K., Antonucci, T. C.
1996
. Same-sex and cross-sex relationships.
Journal of Gerontology: Psychological Sciences
 ,
51B
,
P374
-P382.
Antonucci, T. C., Akiyama, H.
1987
. An examination of sex differences in social support among older men and women.
Sex Roles
 ,
17
,
737
-749.
Avlund, K., Era, P., Davidsen, M., Gause-Nilsson, I.
1996
. Item bias in self-reported functional ability among 75-year-old men and women in three Nordic localities.
Scandinavian Journal of Social Medicine
 ,
3
,
206
-217.
Avlund, K., Fromholt, P.
1998
. Instrumental activities of daily living: The relationships to self-rated memory and cognitive performance among 75-year-old men and women.
Scandinavian Journal of Occupational Therapy
 ,
5
,
93
-100.
Avlund, K., Fromholt, P., Berg, S., Davidsen, M.
1997
. Cognitive function and functional ability in daily life. Facts, Research and Intervention in Geriatrics, Supplementum, 93–106.
Avlund, K., Holstein, B. E., Heikkinen, R.-L., Schroll, M.
1997
. Social relations and functional ability. Facts, Research and Intervention in Geriatrics, Supplementum, 27–40.
Avlund, K., Schroll, M., Davidsen, M., Løvborg, B., Rantanen, T.
1994
. Maximal isometric muscle strength and functional ability in daily activities among 75-year-old men and women.
Scandinavian Journal of Medicine and Science in Sports
 ,
4
,
32
-40.
Avlund, K., Schultz-Larsen, K., Davidsen, M.
1998
. Tiredness in daily activities at age 70 as a predictor of mortality during the next 10 years.
Journal of Clinical Epidemiology
 ,
51
,
323
-333.
Avlund, K., Thudium, D., Davidsen, M., Fuglsang-Sørensen, B.
1995
. Are self-ratings of functional ability reliable?
Scandinavian Journal of Occupational Therapy
 ,
2
,
10
-16.
Boult, C., Kane, R. L., Louis, T. A., Boult, L., McCaffrey, D.
1994
. Chronic conditions that lead to functional limitation in the elderly.
Journal of Gerontology: Medical Sciences
 ,
49
,
M28
-M36.
Cresci, M. K.
2001
. The relationship between informational support and post-injury functional status in older women recovering from a hip fracture.
Educational Gerontology
 ,
27
,
281
-295.
Diehr, P., Patrick, D., Hedrick, S., Rothman, M., Grembowski, D., Raghunathan, T. E., et al.
1995
. Including deaths when measuring health status over time.
Medical Care
 ,
33
,
AS164
-AS172.
Due, P., Holstein, B., Lund, R., Modvig, J., Avlund, K.
1999
. Social relations: Network, support and relational strain.
Social Science and Medicine
 ,
48
,
661
-673.
Era, P., Avlund, K., Jokela, J., Gause-Nilsson, I., Heikkinen, E., Steen, B., et al.
1997
. Postural balance and self-reported functional ability in 75-year-old men and women—A cross-national comparative study.
Journal of the American Geriatrics Society
 ,
45
,
21
-29.
Fratiglioni, L., Wang, H.-X., Ericsson, K., Maytan, M., Winblad, B.
2000
. Influence of social network on occurrence of dementia: A community-based longitudinal study.
The Lancet
 ,
355
,
1315
-1319.
Guralnik, J. M., LaCroix, A. Z., Abbott, R. D., Berkman, L. F., Satterfield, S., Evans, D. A., et al.
1993
. Maintaining mobility in late life.
American Journal of Epidemiology
 ,
137
,
845
-857.
Heikkinen, R. L., Berg, S., Avlund, K.
1995
. Depressive symptoms in late life.
Journal of Cross-Cultural Gerontology
 ,
10
,
315
-330.
Hoeymans, N., Feskens, E. J. M., van den Bos, G. A. M., Kromhout, D.
1998
. Non-response bias in a study of cardiovascular diseases, functional status and self-rated health among elderly men.
Age and Ageing
 ,
27
,
35
-40.
Johnson, C. L., Troll, L. E.
1994
. Constraints and facilitators to friendships in late late life.
The Gerontologist
 ,
34
,
79
-87.
Lee, Y.
2000
. The predictive value of self assessed general, physical, and mental health on functional decline and mortality in older adult.
Journal of Epidemiology and Community Health
 ,
54
,
123
-129.
Liu, X., Liang, J., Muramatsu, N., Sugisawa, H.
1995
. Transitions in functional status and active life expectancy among older people in Japan.
Journal of Gerontology: Social Sciences
 ,
50B
,
S383
-S394.
Lund, R., Modvig, J., Due, P., Holstein, B.
2000
. Stability and change in structural social relations as predictor of mortality among elderly women and men.
European Journal of Epidemiology
 ,
16
,
1087
-1097.
Mendes de Leon, C. F., Glass, T. A., Beckett, L. A., Seeman, T. E., Evans, D. A., Berkman, L. F.
1999
. Social networks and disability transitions across eight intervals of early data in the New Haven EPESE.
Journal of Gerontology: Social Sciences
 ,
54B
,
S162
-S172.
Moen, P., Dempster-McClain, D., Williams, R. M.
1989
. Social integration and longevity: An event history analysis of women's roles and resilience.
American Sociological Review
 ,
54
,
635
-647.
Okun, M. A., Keith, V. M.
1998
. Effects of positive and negative social exchanges with various sources on depressive symptoms in younger and older adults.
Journal of Gerontology: Psychological Sciences
 ,
53B
,
P4
-P20.
Penninx, B. W. J. H., Leveille, S., Ferrucci, L., van Eijk, J. T. M., Guralnik, J. M.
1999
. Exploring the effect of depression on physical disability: Longitudinal evidence from the Established Populations for Epidemiologic Studies of the Elderly.
American Journal of Epidemiology
 ,
89
,
1346
-1352.
Radloff, L. S.
1977
. The CES-D scale: A self-report depression scale for research in the general population.
Applied Psychological Measurement
 ,
1
,
358
-401.
Sarwari, A. R., Fredman, L., Langenberg, P., Magaziner, J.
1998
. Prospective study on the relation between living arrangement and change in functional health status of elderly women.
American Journal of Epidemiology
 ,
147
,
370
-378.
Schroll, M., Steen, B., Berg, S., Heikkinen, E., Viidik A.
1993
. NORA—Nordic Research on Ageing: Functional capacity of 75-year-old men and women in three Nordic localities.
Danish Medical Bulletin
 ,
40
,
618
-624.
Seeman, T. E.
2000
. Health promoting effects of friends and family on health outcomes in older adults.
American Journal of Health Promotion
 ,
14
,
362
-370.
Seeman, T. E., Berkman, L. F., Charpentier, P. A., Blazer, D. G., Albert, M. S., Tinetti, M. E.
1995
. Behavioral and psychosocial predictors of physical performance: MacArthur Studies of Successful Aging.
Journal of Gerontology: Medical Sciences
 ,
50A
,
M177
-M183.
Seeman, T. E., Bruce, M. L., McAvay, G. J.
1996
. Social network characteristics and onset of ADL disability: MacArthur Studies of Successful Aging.
Journal of Gerontology: Social Sciences
 ,
51B
,
S191
-S200.
Shaw, W. S., Cronan, T. A., Christie, M. D.
1994
. Predictors of attrition in health intervention research among older subjects with osteoarthritis.
Health Psychology
 ,
13
,
421
-431.
Stuck, A. E., Walthert, J. M., Nikolaus, T., Büla, C. J., Hohmann, C., Beck, J. C.
1999
. Risk factors for functional status decline in community-living elderly people: A systematic literature review.
Social Science and Medicine
 ,
48
,
445
-469.
Unger, J. B., McAvay, G., Bruce, M. L., Berkman, L., Seeman, T.
1999
. Variation in the impact of social network characteristics on physical functioning in elderly persons: MacArthur Studies of Successful Aging.
Journal of Gerontology: Social Sciences
 ,
54B
,
S245
-S251.
Vogt, T. M., Mullooly, J. P., Ernst, D., Pope, C. R., Hollis, J. F.
1992
. Social networks as predictors of ischemic heart disease, cancer, stroke and hypertension: Incidence, survival and mortality.
Journal of Clinical Epidemiology
 ,
45
,
659
-666.
Wolinsky, F. D., Stump, T. E., Callahan, C. M., Johnson, R. J.
1996
. Consistency and change in functional status among older adults over time.
Journal of Aging and Health
 ,
8
,
155
-182.
Wu, Z., Pollard, M. S.
1998
. Social support among unmarried childless elderly persons.
Journal of Gerontology: Social Sciences
 ,
53B
,
S324
-S335.