| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
Correspondence: Address correspondence to Daniela Jopp, Fordham University, Department of Psychology, 441 East Fordham Rd, Bronx, NY, 10953. E-mail: jopp{at}fordham.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Valuation of life • Well-being • Young-Old • Oldest Old
It is important to note that VOL goes beyond extant QOL concepts. VOL captures not only the sense of "enjoyment and the absence of distress, but also hope, futurity, purpose, meaningfulness, persistence, and self-efficacy" (Lawton et al., 2001, p. 407), thus extending primarily emotion-focused QOL measures. Arguing that QOL in old age is too often conceptualized in the context of health-dominated decrement models, Lawton (1999) claimed that VOL would reflect that there is more than health issues to be considered, such as positive emotion and a strong potential to adapt. In line with this notion, Lawton and colleagues (2001) found VOL to be positively related to mental health indicators (e.g., Ryff's well-being scales) as well as to psychological constructs such as hardiness and mastery. Although VOL was developed in sharp contrast to health-related QOL and "conceptually should be unrelated to physical health" (Lawton, Moss, Winter, & Hoffman, 2002, p. 541), it cannot be expected that the empirically emerging correlations between VOL and physical health indicators are zero; nevertheless, these relations should be much weaker than those between traditional QOL measures and health. Furthermore, findings demonstrate that VOL has predictive qualities that go beyond those of established well-being constructs. Whereas general and health-related QOL did not sufficiently explain the extent to which an individual was attached to life, VOL predicted how long individuals wanted to live in varying hypothetical end-of-life situations (Lawton, Moss, et al., 1999).
Although Lawton's VOL concept has received quite some interest and has generated many scholarly discussions, only a few studies have empirically investigated the predictors of VOL. This is surprising because examining what contributes to VOL may not only increase researchers' knowledge of factors responsible for an individual's attachment to life but may also have strong implications for professionals dealing with older individuals at the end of life.
What are the factors that make a life worth living? According to Lawton (e.g., 2000), VOL is determined by a mixture of QOL aspects, including objective (e.g., behavioral competence, health, environment) and subjective (e.g., mastery, positive affect, quality of time use) factors. Furthermore, VOL is assumed to be predicted primarily by positive features, less so by negative ones. For instance, individuals should value their lives more strongly if they have positive social relations and pursue activities important to them. Severe health restrictions may also play a role, as these restrict participation in desired activities, but their effects should be less strong compared to those of positive factors. Being associated with a worsening health condition, age may also show a negative relation to VOL (e.g., Lawton, Winter, Kleban, & Ruckdeschel, 1999).
The few studies addressing predictors of VOL support most of the theoretical assumptions. In a study with 600 older adults (aged 70+ years), age had a negative but weak correlation to VOL (r = –.10); other sociodemographic characteristics played no role (Lawton et al., 2002). Social and health-related variables were significantly linked to VOL. Confirming Lawton's prediction, VOL was influenced more strongly by positive factors than by poor health and depression. In particular, quality of relationship with friends, quality of time use, and positive emotions were strongly linked to VOL (cf., Lawton, 1999). Health was significantly related to VOL, but its link was rather small in size (β = –.15; Lawton et al., 2002). In some studies, the independent effect of physical health disappeared when researchers controlled for positive life features (cf. Lawton, 1999).
There is also empirical evidence for the importance of idiosyncratic factors. A large study with older individuals demonstrated a strong positive relationship between personal goals/projects (i.e., actions planned and chosen by the individual) and VOL (Lawton et al., 2002). VOL was correlated with the total number of personal projects as well as specific activities including other-directed activities (e.g., helping others, volunteering), intellectual activities (e.g., educational activities, reading), home planning (e.g., housing adjustment, home maintenance), and spiritual/moral projects (e.g., identity maintenance, religion). At the same time, these activities were independent of positive affect. Thus, the differential relations seem due to the specific aspects captured by VOL, namely meaning and purpose, futurity, agency, and perseverance, which go beyond pure positive emotional states.
Gerontological research suggests that old age does not represent a homogenous life phase characterized by continuous quantitative change but that there are (at least) two phases distinguishable by differing qualities, namely the third and the fourth ages (M. M. Baltes, 1999; P. B. Baltes & Smith, 2003; Neugarten, 1974). Thus, one may wonder whether examinations of VOL should consider both periods separately rather than investigating a combined group of elders. Individuals in their third age, often referred to as old or the young-old, are mostly in good health and cognitively well functioning and are socially embedded and engaged in many activities. Individuals in the fourth age, referred to as very old or the old-old, are likely to suffer from chronic conditions and health restrictions, have often seen their spouses and friends passing away, and face increasing limitations to their independence (e.g., Antonucci, 2001; P. B. Baltes & Smith, 2003; Bosworth, Schaie, & Willis, 1999; Lindenberger & Baltes, 1997; Smith, Borchelt, Maier, & Jopp, 2002). Thus, when it comes to what makes a life worth living, the question about whether there are differences between the third and the fourth ages seems to suggest itself due to the discrepant internal and external conditions in both ages. Another important factor at play may be age-related adaptation processes (e.g., Boerner & Jopp, 2007). As individuals find ways to adapt to age-related changes, it may be the case that factors essential in the young-old age become less important in old-old age. Prior findings on predictors of well-being support the assumption of age-differential patterns. Isaacowitz and Smith (2003) reported that cognition and social relationships lost their predictive value for positive affect in old-old compared to young-old individuals. The first empirical hints that the factors related to VOL may differ between old and very old individuals originated from a study investigating VOL in centenarians (Rott, Jopp, d'Heureuse, & Becker, 2006). Centenarians who were highly impaired in terms of health (but not severely demented) had significantly lower levels of VOL compared to the 70- to 85+-year-old persons studied by Lawton, Moss, and colleagues (1999). The absolute difference was notable but less substantial than expected (d =.31), which provided support for the VOL concept by demonstrating that very old individuals are able to maintain high levels of VOL even under adverse life circumstances (i.e., substantial health restrictions and dependence). Regarding predictors of VOL, results mostly converged with Lawton's theoretical expectations, but there were also differences that may have been due to the specific life situations of the centenarians. As expected, physical and cognitive functioning had no explanatory value. Extraversion was a strong predictor, which is in line with the idea that social activity generates value. By contrast, the prediction of one variable came rather as a surprise. Instrumental activities of daily living (IADL), usually considered indicators of health, were highly predictive of VOL. To some extent, the IADL finding may parallel Lawton and colleagues' (2002) results on the links between activities and VOL mentioned earlier. Rott and colleagues (2006) attempted to explain the finding by drawing on the motivational and emotional incentives that very old individuals may experience when performing activities such as cooking or using the phone by themselves rather than being dependent on others for such tasks (e.g., Lawton, 1993). In sum, findings provide a first indication that old-old individuals endorse VOL quite strongly, but maybe to a lesser extent than the young-old, and that there may be similarities as well as differences between both groups regarding predictors of VOL. Further empirical efforts seem necessary to draft a more complete picture about the extent to which levels of VOL vary between the third and the fourth age and about which factors contribute to VOL in both phases of later life.
The present study investigated VOL and its predictors in young-old and old-old individuals. Three research questions were of main interest. First, we tested mean-level differences in VOL between age groups. We hypothesized that young-old and old-old individuals would show, on average, high levels of VOL. Based on Lawton's theoretical framework (e.g., Lawton et al., 2002) and the centenarian findings (Rott et al., 2006), we expected that VOL levels would differ significantly between young-old and old-old individuals but that the absolute difference would be small. Second, we examined the predictive value of sociodemographic, social, and health-related factors for VOL. In order to relate our findings to prior results, we first investigated VOL predictors within the total sample and then separately for both age groups. Following Lawton's argument that VOL should be mostly determined by positive aspects of life rather than by health, we assumed that social predictors would have the strongest predictive value for VOL; we expected sociodemographic factors to have only limited effects. Third, we explored whether specific predictors have age-differential predictive validity. Because VOL is the result of a cognitive–affective evaluation process, we assumed that young-old and old-old individuals would profit from different factors. Because they have adapted to health restrictions while advancing into very old age, we assumed that health would play less of a role for VOL for the old-old compared the young-old. At the same time, we assumed social factors would become more important because old-old individuals may experience contact with their loved ones as more important than in earlier years.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
=.87, young-old: =.80, old-old: =.90).
Predictors
As sociodemographic predictors we used age (in years), gender (0 = male, 1 = female), being married (1 = yes, 0 = no) or widowed (1 = yes, 0 = no), and education (1 = less than 9 years, 2 = 9–11 years, 3 = 12 years or more). We also asked whether participants had worked most of their lifetime (job: 1 = yes, 0 = no). We assessed income with five categories (1 = less than 
500, 2 = 500–1000, 3 = 1000–1500, 4 = 1500–2000, 5 = more than 2000). In order to specify social predictors, we used seven indicators. These included the total number of both living children and living grandchildren. We assessed general social contact, representing how often participants had been in recent contact with significant others (i.e., relatives, friends), and number of phone contacts with these persons with the categories 0 = not at all, 1 = once a week, 2 = 2 to 6 times a week, and 3 = once a day or more. The latter two items derived from the Duke Older Americans Resources and Services Procedures (OARS; Fillenbaum, 1988). In addition, we asked whether the participants had a confidant or were volunteering (1 = yes, 0 = no for both variables). We also assessed whether participants had contact with young individuals (i.e., children, adolescents) not belonging to their family (1 = yes, 0 = no). Six items represented health-related predictors. These included a mobility indicator (range = 0–11) that was computed based on how many out of 11 selected facilities in Arheilgen the individual was able to walk to (e.g., grocery, post office, civil service, church, restaurant, cinema). Individuals also rated their visual capacity (with corrective means such as glasses or contact lenses) on a 4-point scale (1 = excellent to 4 = poor). The same categories applied for measuring hearing capacity. Using an item from the OARS (Fillenbaum, 1988), we also asked whether health conditions restricted activities the individual wanted to do. Answers were 0 = not at all, 1 = a little, and 2 = a great deal. As additional health indicators, we assessed ADLs (eating, dressing, personal care, walking, getting into/standing up from bed, showering, using toilet) and IADLs (shopping, talking on the phone, cooking, performing household chores, taking medications, organizing finances) with the OARS (Fillenbaum, 1988). Answer options were 2 = without help, 1 = with some help, 0 = no longer possible, resulting in a score of 0 (completely dependent) to 14 (fully independent) for each scale. With the exception of health restrictions, for which higher numbers indicated higher restrictions, we recoded health indicators so that higher values always indicated more positive health resources.
Analytic Procedure
First, we determined differences in mean levels of VOL and its predictors using ONEWAY analysis. Second, we described relations between VOL and predictors based on zero-order Pearson correlations. Third, we conducted a set of regression analyses to determine whether sociodemographic, social, and health predictors separately predicted interindividual differences in VOL in the total sample as well as in both age groups. Fourth, we carried out a set of hierarchical regressions using all sociodemographic, social, and health predictors concurrently to examine the predictive value of each predictor group for the total sample and for the age groups. Finally, a last set of regression analyses explored the statistical reliability of the differential predictive patterns found in both age groups by testing Age x Predictor interaction effects.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
Correlations Between Sociodemographic, Social, and Health Resources and VOL
VOL was positively related to several factors (see Table 2). All health indicators were linked to VOL, mostly depicting stronger correlations than sociodemographic and social indicators. The strongest correlation emerged for IADLs (r =.55), followed by ADLs (r =.45) and vision (r =.42; ps <.01). Social resources varied in their relation to VOL between r = –.04, ns (grandchildren) and r =.29, p <.01 (phone contact). Among the sociodemographic variables, age was most strongly linked to VOL (r = –.32, p <.01). Being married (r =.11, p <.05) and having a higher education (r =.23, p <.01) were also related to higher VOL. Having lost a spouse (r = –.16) and being female (r = –.11; ps <.05) were associated with lower VOL.
|
|
The model focusing on health factors predicted 35% of the variance in VOL, which was substantively more than the models that included sociodemographic or social predictors. IADLs had the strongest link to VOL (β =.36), followed by visual capacity (β =.20; ps <.01). Mobility, hearing capacity, activity restrictions, and ADLs were without independent predictive value.
When we investigated the domain-specific predictor models separately for young-old and old-old individuals, similarities as well as differences emerged. Overall, prediction of the sociodemographic model was rather similar in both groups, with a somewhat better prediction for the old-old (young-old = 9%; old-old = 11%). Age was a significant negative predictor in the old-old (β = –.22, p <.01) but was only marginally significant in the young-old (β = –.12, p =.09). Education had a positive effect that was limited to the young-old (β =.15, p <.05). In the old-old group, being married was strongly negatively linked to VOL (β = –.51, p <.05), but there was no such effect for the young-old.
In the social models, the proportion of explained variance was identical in both age groups (16%). In the young-old, phone contacts (β =.16) and volunteering (β =.15; ps <.05) explained unique variance in VOL. General social contacts, having a confidant, and contact with youth had marginal positive effects (all βs =.13, ps =.09). Having grandchildren had a negative relation (β = –.25, p <.05). In the old-old, by contrast, only phone contact had an independent effect (β =.38, p <.01).
The model including health indicators explained somewhat more variance in VOL in the young-old (33%) compared to the old-old (29%) group. In both age groups, vision and IADLs were positively linked to VOL (vision: young-old: β =.23, p <.01, vs. old-old: β =.19, p <.05; IADLs: young-old: β =.39, p <.01, vs. old-old: β =.31, p <.05). There was one additional significant predictor in the young-old, namely activity restrictions (β = –.21, p <.05), indicating that VOL was lower for those young-old individuals who reported more activity restrictions.
Predictors of VOL: Combined Models
The next set of regressions examined the predictive value of sociodemographic, social, and health indicators concurrently (see Table 4). The hierarchical regression conducted for the full sample explained 46% of interindividual differences in VOL. In this model, sociodemographic predictors were entered in the first step, predicting 15% of simple variance. Social predictors were entered in the second step, predicting an additional 10% of the variance. Entering health predictors in the last step explained an additional 20% of the variance. Follow-up analysis determining the proportion of independent (unique) variance by varying the sequence of predictor blocks showed that sociodemographic indicators had no independent explanatory value (2%, ns), whereas social indicators predicted only half as much of the independent variance compared to health indicators (9% and 20%, respectively).
|
Conducting the same hierarchical regression analysis only for the young-old group explained 46% of the variance in VOL. Most of the effects observed were identical to the findings reported for the total sample. Sociodemographic indicators explained 8% of the simple variance when entered first, social indicators added 12%, and health resources added 27%. As shown by follow-up analysis, sociodemographic variables made no independent contribution, but health indicators explained 27% of the unique variance, which was 3 times more than that explained by social indicators (9%). It is interesting that the effects of the specific health indicators were also much stronger compared to the total sample. IADLs was the strongest predictor, followed by vision and activity restriction. We found marginal effects for most social predictors, including number of children, general social contacts, phone contacts, having a confidant, and having contact with youth. Being widowed had a marginal negative effect.
For the old-old group, the identical regression analysis explained 45% of the variance in VOL. Although sociodemographic indicators explained simple variance when entered in the first step, their block explained no independent variance when entered last, as shown by follow-up analysis. The independent variance explained by the social variables was 11%, which was slightly higher compared to in the young-old. The unique contribution made by the health indicators was 18%, which was, by contrast, substantially less than in the young-old. It was, nevertheless, still the block of predictors explaining the most unique variance in the old-old age group. Inspection of the beta weights revealed significant positive effects of IADLs and phone contact. Gender also had a significant effect, with women having lower levels of VOL. Follow-up analysis indicated an interaction between gender, being married, and age. Very old women depicted substantially lower VOL levels if they were married compared to married men.
In order to ensure the reliability of the reported effects, we also ran a hierarchical regression that included only those variables that had been found to have some relation to VOL in the previous domain-specific regression analyses; i.e., p <.10 in either the full sample or the age-group analyses. This predictor selection resulted in a more favorable cases–predictor ratio. Findings were identical to those reported earlier, strengthening the reliability of the results. This is especially important in light of the difference in sample size between young-old and old-old. Although the overall explained variance was identical in both groups, which speaks for equivalent model conditions in both groups, the smaller number of old-old individuals could have gone along with reduced statistical power due to the less favorable case predictor ratio, which could have resulted in less of a chance of detecting single predictor effects. Nevertheless, we found no evidence that this was the case.
Because the direct comparison of the regression coefficients of two groups only pointed to differences but did not determine whether the differences were significant, we conducted some follow-up analyses to underscore our findings. As group differences in predictions can be explored with interaction terms, we ran the above-described regression analysis for the total sample by adding interaction terms for indicators whose predictions seemed to differ between both groups: Gender x Age, Phone Contact x Age, Vision x Age, and Activity Restriction x Age. The interaction including phone contacts was significant (p =.01), and the interaction including activity restriction approached significance (p =.12), but there was no indication that gender (p >.55) and vision (p >.65) related differentially to VOL within both age groups. Thus, follow-up findings provided additional evidence for reliable age-group differences in the predictive value of phone contacts and activity restrictions.
In sum, regression findings support the existence of age-differential prediction patterns. For the young-old, health predictors played a substantial role for predicting differences in VOL as indicated by the unique variance explained by all health indicators together. For the old-old, health predictors had less importance for VOL, suggesting adaptation processes. Only IADL capacity remained a significant health predictor, depicting a comparable effect in both age groups. Social predictors were somewhat more important in the older group, but differences were not strong. Considering the specific predictors, social activities contributed differentially to VOL in both groups. Sociodemographic variables explained no independent variance. Notably, health and social predictors absorbed the total amount of age-related variance.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Sociodemographic, social, and health factors were associated with attachment to life, as indicated by zero-order correlations. For instance, we found that being married, not being widowed, being male, and being highly educated was each positively related to VOL. Assuming that individuals derive meaning from having acquired education and from having a spouse (and that they have difficulty maintaining meaning if the partner dies), the observed links are straightforward. Compared to sociodemographic factors, social resources were more strongly linked to VOL, which is also in accordance with theory. Individuals indicating more frequent general social contact and phone contact reported higher levels of VOL, which confirm findings on the value of social resources for VOL (e.g., Lawton, 1983). Our study also shows that individuals profit from volunteering, replicating results on personal projects (Lawton et al., 2002). Complementing those results, having contact with young individuals outside of the family emerged as an additional social factor related to higher VOL. This latter finding seems especially interesting in light of socioemotional selectivity theory (Carstensen, 1995), which would predict that older individuals prefer close social partners (i.e., family) to optimize emotion regulation; attachment to life may benefit from interactions with both close and more distal partners.
In contrast to theory, we found that health factors had the strongest zero-order relations to VOL. This is noteworthy because, conceptually, VOL is assumed to be independent of physical health influence (e.g., Lawton et al., 2001), although there is empirical evidence for at least a weak link between health and VOL (e.g., Lawton, Moss, et al., 1999; Lawton et al., 2001, 2002). Maybe even more noteworthy is that social factors were less powerful in predicting VOL than health, which also contradicts Lawton's theoretical framework.
The regression analyses further highlighted the strong role of health predictors. One reason for the dominance of health over social factors may be that social contacts are not always positive but can be ambivalent (e.g., Fingerman, Hay, & Birditt, 2004), which also is responsible for low correlations between social indicators and subjective well-being. In Lawton's work (e.g., 1999), quality of social relations was assessed, which may have contributed to higher prediction of the social domain. Alternatively, one could assume that examining old and very old individuals together has contributed to blurring the relation between health and VOL. Also, health indicators in the present study were individual ratings characterized by a varying degree of subjectivity. As subjective health judgments are more likely to be related to VOL than objective health indicators, our findings may be the result of our measurement selection. However, the predictive value seems not to differ between health indicators that are easily accessible by the individual and therefore less likely to be biased (i.e., IADLs) and indicators that involve a more extensive, bias-prone evaluation process (i.e., activity restrictions). Nevertheless, when we compare our findings with those of earlier studies, it becomes clear that a systematic evaluation of objective and subjective health predictors of VOL would be helpful to better understand the differences in results.
Concerning differences in the predictive value of sociodemographic, social, and health factors in the young-old and the old-old, findings suggest that individuals in the fourth age have engaged in multiple adaptation processes. Whereas attachment to life in the young-old is substantially related to visual capacity, activity restrictions, and IADLs, the contribution of health factors is limited to IADL capacity in the old-old. At the same time, phone contacts emerged as a significant social predictor in the older group. We suggest that both findings (i.e., the reduced prediction of health indicators and the enhanced role of phone contacts) are the result of adaptation processes to age-associated resource restrictions in very old age. This is in accordance with earlier work on adaptation to aging that found that very old individuals profited from giving up blocked goals in terms of well-being (e.g., Jopp & Smith, 2006). When one is no longer able to visit significant others due to health and mobility limitations, it may be adaptive to give up the wish of visiting friends or family and rather use the phone to keep in touch. This interpretation is also underscored by the substantially higher frequency of daily phone contacts in the very old group. In that sense, we interpret the lack of unique prediction of the established social resources in very old age but the increase of the importance of phone contacts as an instance of positive adaptation, focusing on those things that are still possible given age-related restrictions. A comparable argument may hold for the age-differential prediction of activity restrictions: Very old individuals may give up activities no longer performable, or they may downgrade their importance, which reduces reports of activity restrictions and their link to VOL. Thus, our findings seem to illustrate nicely Lawton and colleagues' (2001) claim that VOL reflects dynamic accommodation and assimilation processes with the help of which individuals deal with illness and decline.
In addition, it is interesting that IADLs was the factor that remained significant in predicting VOL in young-old and old-old participants, a finding that also emerged in the centenarian study by Rott and colleagues (2006). IADL capacity, which includes extended ADLs such as calling people on the phone, shopping, and taking care of financial business, may reflect more of a motivational rather than a functional health aspect and may be a proxy for meaningful activities rather than basic competence. Regression findings showed that IADLs predicted VOL over and above basic ADL competence. ADLs itself was without any predictive value, which parallels Lawton and colleagues' (2002) findings that ADL-related personal projects were unrelated to VOL. Nevertheless, we need to note that both ADLs and IADLs share a substantial amount of variance. As IADLs capture to a stronger extent than ADLs an individual's cognitive capacity, more IADLs may also go along with being better able to perform the more complex activities that may be valued more strongly. Furthermore, mean levels of ADLs and IADLs were fairly high in both groups, which may have contributed to the present findings.
The gender effect in the old-old group was also surprising. We are not aware that gender differences in mean levels of VOL have been reported as yet. Follow-up analyses showed an interaction between gender, being married, and age, with old-old women having the least attachment to life when being married. It could be the case that these very old married women provide care for a disabled spouse, which would parallel work on depression (Prince, Harwood, Blizard, Thomas, & Mann, 1997). Such differential findings underscore further that examining VOL separately for young-old and old-old individuals adds important aspects to researchers' understanding of VOL.
Limitations
Some limitations need to be considered. First, as our study was the first one to examine age differences in VOL within a sample of older adults, our set of predictor variables was vast, but other variables may also matter. For instance, a larger array of time use indicators or personality factors such as extraversion could represent important additional predictors. It may also be fruitful to dedicate more attention to the process of jointly evaluating positive and negative aspects of life and to the shifting of priorities in order to better capture how old and very old individuals' efforts result in specific levels of attachment to life.
Second, we measured sociodemographic, social, and health variables concurrently with VOL, which restricted our analysis to describing interrelations rather than causal prediction (despite using the term prediction due to regression analysis conventions). Longitudinal analyses are important for examining whether sociodemographic, social, and health factors precede attachment to life and whether changes in these factors result in subsequent VOL changes. More complex models may also address the interdependence between predictors and allow for exploration of mediation and moderation effects.
A third limitation is that we had no information about the participants' mental status, which may be problematic due to the age range under examination. The sample was based on old and very old individuals who volunteered to share ideas about aging in their community in a mail-out study. Contact with the participants was limited to letters and phone calls. Thus, it could be that the sample included persons with mild cognitive impairment, but we doubt that such limitations would have influenced their answers substantially. It also does not seem likely that individuals with stronger cognitive limitations would have participated and provided coherent data (e.g., we excluded participants with more than 50% missing data). Nevertheless, some differences in findings may have occurred due to the fact that in Lawton's studies, participants were interviewed in person, which may have allowed clarifying participants' questions as well as their contributions.
Conclusion
Findings from the present study support the claim that community-dwelling older individuals value life highly and remain greatly attached to life despite increasing negative conditions. Nevertheless, average levels of attachment to life were substantially lower in old-old compared to young-old individuals, mirroring the fact that it may not be so easy to keep one's spirits up when advancing into very old age. Although researchers are still far from understanding the process of how people integrate positive and negative features of life which determines their attachment to life, the present study illuminates which factors might be important in the mix for both age groups. Within each age group, the pattern of findings corroborates that VOL is the result of a complex process of balancing positive (e.g., young-old: vision; old-old: phone contacts) and negative (e.g., young-old: activity restrictions; old-old: gender) aspects. Notably, health aspects seem to be more salient than theoretically expected, which is of importance in terms of theory development, but they seem to become less important in very old age as specific social factors gain in relevance.
Finally, we want to highlight the fact that health and social factors explained the total amount of age-related variance. Qualifying the notion of the incomplete architecture of the human ontogenesis (P. B. Baltes, 1997), this finding demonstrates the integrity of the human psychological architecture in contrast to the physical architecture in very old age. This is promising, because it could mean that if experts succeed in developing interventions for the young-old and the old-old with the goal of enhancing the positive determinants of VOL and mitigating the negative ones, a long and valuable life—which is not necessarily a healthy life—may be possible for the majority of elders.
|
Footnotes |
|---|
1 School of Psychology, Georgia Institute of Technology, Atlanta.
2 Institute of Gerontology, University of Heidelberg, Germany.
3 Department of Psychological Ageing Research, Institute of Psychology, University of Heidelberg, Germany.
Decision Editor: William J. McAuley, PhD
Received for publication October 2, 2007. Accepted for publication December 10, 2007.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
