There are common stereotypes of elderly individuals - older adults are "senile" and their memories are failing, their personalities are becoming more rigid, they are often depressed, they are lonely and isolated, and their adult children have abandoned them or "warehoused" them into nursing homes. These stereotypes may cause harm in different ways, including the following:
* When older adults are treated as if the stereotypes were true.
* When older adults accept stereotypes as true, and thus restrict their lives unnecessarily.
* When families lower their expectations of an elderly parent's role in the family.
Inaccurate stereotypes pose harm even to helping professionals who may treat old individuals in ways that are based on their assumed but unfounded expectations for that age, resulting in unnecessarily difficult interactions and possibly incorrect diagnoses. This report will show that many of these stereotypes are false.
The "oldest-old," generally defined as those age 85 and older, are the fastest-growing age group in the United States. This trend was first identified by Mantón and Stallard (1990) and Siegel and Taeuber (1986), and is still a demographic imperative today (Martin, 1999). As more and more individuals survive to advanced old age, their influence on society has increased, challenging existing attitudes related to the delivery of health care services. How individuals who are the oldest-old differ from the oldold (age 75 to 84) and the young-old (65 to 74) has been a matter of considerable attention for more than a decade.
Figure 1. An over-time comparison of Wechsler Adult Intelligence Scale verbal and performance scores for the old-old and oldest-old groups. Note: Numbers at the top of the graphs indicate higher IQ scores.
Most studies of older adults are cross-sectional, that is, at a single time period a group of individuals from several age groups are examined and the age groups are compared. For some aspects of aging (e.g., housing choices), this is the most efficient research method. However, in a cross-sectional study there is no way to determine whether observed age differences result from age itself or are an effect of membership in different birth cohorts. For example, the Seattle Longitudinal Study of Aging (Schaie, 1996) has shown convincingly that what had appeared to be age decline in IQ (when measured cross-sectionalIy) was in fact the result of measuring different cohorts. Historically, older adults had less education and possibly less adequate nutrition than younger individuals, hence their IQ scores were lower. But when the same individuals are measured longitudinally, remarkably little decline was found. Longitudinal studies, in which by definition the same individuals are examined over two or more measurement periods, can provide valuable information about continuity or change over time in the same individuals.
This report is a longitudinal comparison of several aspects of the lives of the individuals who constitute the Berkeley Older Generation Study. At Tune I1 most of the sample (more than 85%) were young-old (age 60 to 74); at Tune 2, 13 years later, two thirds (69%) were old-old (age 73 to 84), and one third (31%) were oldest-old (age 85 to 93). The focus of this arricie is on the trajectories of change for those who were (at Time 2) old-old, as compared with those who were oldestold. The following aspects of their lives will be summarized: intellectual functioning, personality, family relations, and beyond-family social relations. Implications will be extracted for nursing and other professions who provide services to older adults.
The Berkeley Older Generation Study began in 1929 with a representative group of men and women who were first interviewed as young-old residents of Berkeley, California, and who have been re-interviewed at periodic intervals since that time by the Institute of Human Development, University of California at Berkeley (Bayley, 1933; Maas & Kuypers, 1974; Macfarlane, 1938). At the beginning of the study there were 470 young-old adults; 142 were interviewed in 1968/1969 (which for this study is lime 1), and 94 survived to be interviewed in 1982/1983 (Time 2). Birth years of survivors ranged from 1890 to 1910, with a mean birth year of 1900. At the Time 2 follow-up, 61% of the sample were widowed (only 4% divorced); 67% were women; and they had 1 to 6 children (mean = 2.2) and O to 16 grandchildren (mean = 5.9). Their education ranged from fewer than 7 years to professional degrees, and occupations ranged from unskilled laborer to high-level professional (Field, Schaie, & Leino, 1988).
A comparison of the characteristics of the Berkeley group with the national cohort in the United States (Longino, 1986) found that the two groups are similar in all demographic characteristics measured by the 1980 census, except for education and, to a lesser extent, financial status. In both cases, the Berkeley group was somewhat higher.
Standard Error of Measurement [SEM]. Reliable change in a variable was measured with the SEM, which identifies reliable change over time using the standard deviation at Tune 1 and the correlation between lime 1 and Time 2 (Dudek, 1979; Schaie, 1989). This method identifies those individuals whose IQ scores changed reliably - scores that had either increased or decreased.
Stationary Component Analysis [SCA]. An SCA, devised specifically for longitudinal study, was used to combine a large number of individual ratings into a small number of personality traits. This is an extension of the usual principal-component analysis, with the added feature that the components are identical across all time periods. This method provides a common frame of reference to study change, using commonly interpretable variables (Field & Millsap, 1991).
ANOVA, Two separate analyses of variance with repeated measures on time were used to compare old-old and oldest-old individuals on the trajectories of IQ and of health (Field et al., 1988) and measures of relations with family members and friends (Field & Minkler, 1988).
Continuity and change in intellectual functioning
The first comparison focuses on longitudinal tests of intellectual functioning, as measured by the Wechsler Adult Intelligence Scale [WAIS] (Wechsler, 1955), a well-respected individually administered test of intelligence (IQ).
Figure 2. Reliable change in Wechsler Adult Intelligence Scale verbal and performance scores over 13 years of late life.
Comparison of the old-old and oldest-old groups. Longitudinal analyses were performed using the WAIS scores for those individuals in the Berkeley Older Generation Study who had been tested in young-old age and were then retested when oldold or oldest-old (Field et al., 1988). There were different trajectories of change for the younger from that of the older group. As Figure 1 shows, both verbal and performance scores decline significantly for both groups, but the declines are greater for the oldest-old. This is what one would expect, based on cross-sectional studies. However, these comparisons are of the age groups as a whole, but do not apply to every individual.
Individual differences. When one examines the scores of individuals, one finds that some change, while others do not. Using the SEM, the authors discovered that some individuals increase reliably in their WAIS scores, some decline reliably, and some scores change less than the SEM and thus are judged to be merely random fluctuations about the mean.
Figure 2 shows that 62% of the oldold and 44% of the oldest-old do not decline reliably in WAIS Verbal scores. In fact, some individuals actually increased reliably in verbal scores over the course of 13 years of old age, including 8% of the oldest-old age group.
The profile of performance scores is somewhat different. Most of the WAIS Performance Intelligence subtests are timed tests, and it is known that old individuals do have trouble with tests in which speed is a factor (Schaie, 1996). Even so, 15% of the old-old and 6% of the oldest-old do not decline even in their performance subtests, as shown in Figure 2.
How do we resolve these seemingly contradictory findings? It is true that the groups as a whole did decline during the course of 13 years, yet only a third of the old-old individuals and just more than half of the oldest-old decline reliably in verbal intelligence. Therefore, while many maintain the same level of intellectual functioning, the declines for other individuals are so great that their scores pull down the means of an entire group. This decline was first attributed to "terminal decline" by Kleemeier (1962), and later by Riegel and Riegel (1972), who hypothesized a period of about 5 years as the period of terminal decline. Recently, Small and Bäckman (1999) reviewed research, finding there is little agreement about the timing of terminal decline. Since a larger proportion of individuals older than age 85 would naturally be in terminal decline, the mean of the oldest-old group reflects this decline. Results similar to those of the Berkeley study were found in the Duke Longitudinal Study of Aging (Nowlin, 1985) and the Seattle Longitudinal Study of Aging (Bosworth, Schaie, Willis, Oc Siegler, 1999).
Shattering a myth. Many of the results presented in this article are so different from the stereotypes of old age that it is valuable to reflect on these findings as evidence of the remarkable resiliency of some individuals. For example, what happened to those old individuals whose measured intelligence increased reliably during 13 years of old age? There are several explanations: for example, in some cases health improved, and so did IQ scores.
Most interesting, however, is the finding that shows the impact of caregiving. A number of women were serving as primary caregiver to an ill spouse at lime 1. At lime 2 the spouse was gone, and some of the wives appear to have blossomed. One might speculate that these women, now free from the strain of intense caregiving, took up new interests and new activities - and their IQs increased reliably, as identified by the SEM. It has been shown that caregiving can affect the health of a caregiver (Salariano, Minkler, & Langhauser, 1984; Schulz, O'Brien, Bookwala, & Fleissner, 1995; Wright, Clipp, & George, 1993), but these data suggest that the stress of caregiving may actually depress the measured IQ of the caregiver (Field et al., 1988).
Implications. There are four important implications. First is the confirmed and urgent need to change the mind-set held by society that older adults inevitably become demented. Instead, if society embraces the idea that older adults maintain their cognitive competence, they may be capable to live at a higher level of competence (Dawson, 1997; LeSage, Slimmer, Lopez, & Ellor, 1989). If individuals are continually treated as if they are demented, they will act demented - even worse, they may come to believe it themselves. When their competence is acknowledged, they are likely to exercise it.
Second, the findings related to timed performance should not be considered of great importance. Timed intelligence tests are not necessary in everyday life. Although tests for rapid reflexes are appropriate before licensing car drivers, few other aspects of life depend on quick responses. Thus, a more accurate measure of an older adult's mental ability may be obtained by reducing time pressures in any assessment situation. Nowhere is this more necessary than in the admission procedures of health care institutions, as described by Gueldner (1997).
Third, the majority of older adults were found to retain their intellectual abilities into their 80s and 90s. This finding demands that health-care professionals, before they too casually label older adults demented, look more closely for other reasons for mental decline, such as medication interactions and side effects, dehydration, and sensory deficits. Fourth, the association between stress on a caregiver and decreased intellectual function emphasizes the urgency of providing more help for caregivers.
Continuity in Personality
Some longitudinal studies have shown that personality is very stable in adulthood (Costa & McCrae, 1988). Others, however, have found individual differences in stability (Shanan, 1991). Block (1981), for example, noted that some individuals are "impressively predictable" and others "unrecognizable" over long time spans.
In the Berkeley Older Generation Study, SCA analyses yielded five traits that are similar to the "Big Five" personality traits found in younger individuals (John & Srivastava, 1999). The trait "intellect" refers to social intelligence, because the ratings were made based on interviews. There is significant decline over time in intellect, but there is no difference between old-old and oldest-old in the amount of decline. The trait "agreeableness" does not decline over time, nor are there significant age differences. There are no significant time or age group differences in "satisfaction." "Extraversión" declined significantly but equally for both age groups. These four traits seem to identify enduring characteristics because their correlations in 13 years of old age are high (r ranges from = .45 to .53). The fifth trait, "energetic," is not as stable (r = .26). It may reflect responses to external circumstances, including physical health (Field & Millsap, 1991). Furthermore, the trait "energetic" shows significant decline over time and differences between age groups. There were significant gender differences in two traits, intellect and satisfaction, in both of which men scored higher.
Three traits have significant ageby-time interactions. In two of these traits, intellect and energetic, the oldest-old declined while the old-old remained steady. In a third trait, agreeableness, the oldest-old were steady and the old-old increased. This is not what would be expected by the prevailing stereotypes, as will be discussed in terms of implications.
The most unexpected findings regarding personality are twofold. First, there is no decline for the oldest-old or the old-old in satisfaction and there is no difference between the age groups in mean levels of satisfaction. These findings challenge the myth that all old individuals become lonely or depressed - quite the contrary. The remarkable continuing satisfaction of old individuals, even in the face of the many losses of spouse, friends, and roles, has been noted (Diener Si Diener, I996).
Second, the trait agreeableness appears to resemble the integrity aspect of the last stage of Erikson's (1950) eight stages of life - ego integrity vs. despair. Fifty years ago Erikson theorized that as older adults approach their final years, their major life task is to come to terms with their lives, to accept what their lives have been, and to accept their own part in this outcome. If this coming-to-terms does not happen, then the old individual is awash in despair.
Much later, Erikson, Erikson, and Kivnick (1986) interviewed many of the same old individuals who had contributed to the formulation of his 1950 theory of eight stages of life, partly to search for a ninth stage, but none was found (Erikson, E., personal communication, April, 1983). As for the stage of ego integrity shown by the Berkeley parents, the oldestold remained steady, while the younger old-old increased in agreeableness. This suggests that at Time 1 the younger group was still in the process of coming-to-terms, while the older group had for the most part completed this process. At Time 2, the oldest-old remained high in agreeableness, while the old-old increased to the level of the oldest-old.
Implications. These personality findings contrast sharply with current societal beliefs. For example, the stereotype abounds that old individuals become irascible, as illustrated in the popular film "Grumpy Old Men" (Petrie, 1993). Older adults deserve to be recognized for who they are, not as stereotypical "rigid" or "grumpy" individuals. Again, these are self-fulfilling prophecies: If, as a society, we expect our oldest members to behave unpleasantly, this is likely to actually happen.
Figure 3. Beyond -fami Iy: Voluntary organizations. Note: Each Figure plots the entire range of scores for a variable at Time 1 and Time 2. Numbers at the top of the graph indicate more activity and involvement.
Social Relations Among the Old-old and Oldestold
An important area of social gerontology concerns the social relationships of elderly individuals, especially the role of the family in caregiving and support. Twenty years ago, Berkman and Syme (1979) reported that a lack of social relations predicts early mortality, and this is still true. The remainder of this review will focus on family relation and beyond-family relations.
Family relations: Contact. There is no significant change in amount of contact older adults have with their children. As for spouses, a larger number of study members were widowed at Time 2 (only 39% were still married) compared with 73% married at Time 1. Naturally, more oldest-old than old-old individuals were widowed; thus, the support derived from a marriage partner was not available to most of them (Weishaus & Field, 1988).
A decline in contact with grandchildren occurred for both age groups. This decline likely reflects the fact that between Time 1 and Time 2, most grandchildren had grown up, gone to college, and often married and started jobs (Field Sc Minkler, 1 988). Similar results have been found in other longitudinal studies, such as the Bonn Longitudinal Study of Aging (Olbrich Si Lehr, 1976).
Finally, there is a decline for both groups in contact with siblings. Many individuals lost siblings during those 1 3 years of old age. However, for those individuals who still had living siblings, there was no significant change over time in amount of contact.
Family relations: Emotional involvement. An important aspect of family relations that is seldom studied is older adults' feelings about their families. Emotional involvement with family members can contribute greatly to quality of life in advanced old age. Feelings of emotional involvement with family members did not change. The old individuals felt just as satisfied as ever with their family members.
Implications. The stereotype persists that adult children "abandon" their elderly parents, despite unassailable findings that most care for old individuals is provided by family members (Kane, 1994; National Alliance for Caregiving & AARP, 1997; Shanas & Maddox, 1976). The present results replicate this finding, because contact with children remains steady for both oldold and oldest-old study members. As for satisfaction with family members, this too is strong evidence that the parents have no feeling that their children had abandoned them. Health care professionals must recognize that the family is an important resource, without which their work would be far more difficult.
Social Relations Beyond the Family
Fnends. Contrary to common myths, the amount of contact and involvement with good friends continues unabated in older adults, but contact and involvement with casual friends decline over time. Most often casual friends are seen only at group activities, whereas close friends will exchange confidences, share interesting experiences and thoughts, and help each other. Three measures of involvement with friends decline for the oldest-old but not for the old-old. As also found by Johnson and Troll (1994), close friends are there for confidences and sympathy, companionship, advice, long-term memories, and recreation rather than for practical assistance (Field, 1999). There are, however, gender differences in relations with friends. Men decline in number of new friends, in their desire for close friendships, in the less intimate nature of their interactions with friends, and in involvement in beyond-family activities, while women do not change in any of these activities.
Although almost all of these old individuals had lost friends due to deaths or to mobility problems, most were able to find another individual who could be upgraded to the status of close friend. As long as an individual has one friend remaining, that individual can become a "multipurpose" friend, performing many of the functions of friendship that were once spread out over a larger network (except, probably, that of a confidant).
Beyond-family relations: Voluntary associations. Figure 3 shows significant differences between the age groups, and significant age-by-time interactions for club and church attendance and involvement beyond the family show greater decline for the oldest-old. Attendance at senior centers reflects historical times. There were almost no senior centers in existence from 1968 to 1969 (Time 1). Olbrich and Lehr (1976) reported similar results. Members of the Bonn Longitudinal Study of Aging simply did not give up their ties to family members or to close friends. Casual friends were given up only with great reluctance, but decline did occur in ties with clubs and other voluntary associations.
Implications. Even though they have lost many close friends and relatives, most individuals in this sample, even the oldest-old, still retain ties with important others. Social agencies should be vigilant in their search for environmental adaptations that could provide elderly individuals with assistance in maintaining contact with friends.
Instead of the Stereotypie decline described in the introduction to this article, the authors found continuity, not decline, and in a few fortunate cases, increases were found. No longer can all older adults be considered homogeneous, as the individual differences described here attest.
There is much to distinguish the oldest-old from younger old individuals - they are members of different cohorts, they have lived through somewhat different historical times, and therefore they grew up with different experiences. Over long lifetimes they have been faced with problems (e.g., the Great Depression, World War II, the social disruptions of the 1960s), introduced at different life stages (Elder, 1974). Indeed, it is possible to discover that some of the differences between the cohorts compared in this study are based more on historical differences, and less on ageor time-related differences.
There is much more to learn about the lives and the experiences of the oldest-old. Because they are the fastest-growing group in the United States (Martin, 1999; Siegel & Taeuber, 1986), the oldest-old will become more and more important to us all - both as professionals and as concerned citizens. The authors suggest that the small group of survivors of this very long-term longitudinal study can serve as a guide to interesting new questions about social relationships, personality, and cognitive functioning among very old individuals, which can then be replicated in larger samples.
Although cross-sectional studies can provide important information about the aged, longitudinal studies can reveal the most salient patterns of aging. For that reason, it is important that disciplines and social agencies providing care for elderly individuals understand these patterns. With the projected growth in the elderly population, social policy and services must make use of all available information to craft a climate conducive to successful aging. It is important, however, to exercise some caution before generalizing from the findings of this study, because the Berkeley participants are more favored in terms of education than are others of similar age (Longino, 1986). Education alone is likely to be a help to those who are navigating the sometimes treacherous shoals of old age.
These data demand that society change its view of the oldest-old. Members of this age group must be seen as individuals with healthy potential, even when they exhibit considerable disability. In fact, health care professionals must envision the oldest-old as the largest market for life-long health educational programs.
- Bayley, N. (1933). Mental growth during the first three years: A developmental study of sixty-one children by repeated tests. Genetic Psychology Monographs, 14(1), 1-92.
- Berkman, L.E, & Syme, S.L. (1979) Social networks, host resistance, and mortality: A nine-year follow-up study of Alameda County residents. American Journal of Epidemiology, 109(2), 186-204.
- Block, J. (1981). Some enduring and consequential structures of personality. In A.I. Rabin (Ed.), Further explorations in personality. New York: wiley, 27-43.
- Bosworth, H.B., Warner Schaie, K., Willis, S.L., & Siegler, I.C. (1999). Age and distance to death in the Seattle longitudial study. Research on Aging, 21(6), 723-738.
- Costa, P.T., Bc McCrae, R.R. (1 988). Personality in adulthood. Journal of Personality and Social Psychology, 54(5), 853-863.
- Dawson, P. (1997). Nurses aides decrease !earned helplessness in nursing home residents. Perspectives, 2/(4), 9.
- Diener, E, & Diener C. (1996). Most people in the United States experience positive subjective well-being. University of Illinois.
- Dudek, EJ. (1979). The continuing misinterpretation of the standard error of measurement. Psychological Bulletin, 86(2), 335-337.
- Elder, G. (1974). Children of the Great Depression. Chicago: University of Chicago Press.
- Erikson, E.H. (1950). Childhood and society. New York: Norton.
- Erikson, E.H., Erikson, J.M., & Kivnick, H-Q. (1986). Vital involvement in old age. New York: W.W. Norton.
- Field, D. (1999). Continuity and change in friendships in advanced old age: Findings from the Berkeley Older Generation Study. international Journal of Aging and Human Development, 48(4), 325-346.
- Field, D-, & Millsap, R.E. (1991). Personality in advanced old age: Continuity or change? Journal of Gerontology: Psychological Sciences, 46(6), P299-P308.
- Field, D, & Minkler, M. (1988). Continuity and change in social support between youngold and old-old or very-old age. Journal of Gerontology: Psychological Sciences, 43(4), P100-P106.
- Field, D., Schaie, K. W., & Leino, VE. (1988). Continuity in intellectual functioning: The role of self-reported health. Psychology and Aging, 3(4), 385-392.
- Gueldner, S.H., (1997). Creating an elder sensitive acute care climate: A health care imperative. Journal of Gerontologtcal Nursing, Z3(4), 7-9.
- John, O.P., & Srivastava, S. (1999). The big-five trait taxonomy: History, measurement, and theoretical perspectives. In L.A. Pervin & O.P. John (Eds.), Handbook of personality: Theory and research (2nd éd. pp. 102-138) New York: Guilford.
- Johnson, C.L., & Troll, L-E. (1994). Constraints and facilitators to friendships in late late life. The Gerontologia, 34, 79-87.
- Kane, R.L. (1994). A study of post-acute care. Report for the Health Care Financing Administration. Minneapolis: University of Minnesota Institute for Health Services Research.
- Kleemeier, R. W. (1962). Intellectual changes in the senium. Proceedings of the American Statistical Association, 1, 290-295.
- LeSage, J., Summer, L. W., Lopez, M., & EIlor, J.R. (1989). Learned helplessness. Journal of Gerontologi Nursing, Ii(S), 8-15,42-43.
- Longino, C.F. (1986). The oldest Americans: state profiles for data-based planning. Final Report for the A.A.R.P. Andrus Foundation. Coral Gables, FL: Center for Social Research in Aging. 1986.
- Maas, H.B., & Kuypers, J.A. (1974). From thirty to seventy. San Francisco: Jossey-Bass.
- Maci ariane, J. (1938). Studies in child guidance: 1. Methodology of data collection and organization. Monographs of the Society for Research in Child Development, 3(19).
- Manton, K.G., & Stallard, E. (1990). Changes in health functioning and mortality. In S.M. Stähl (Ed.)· The legacy of longevity (pp. 140-162). Newbury Park, CA; Sage.
- Martin, P. (1999). From the Centre's Director. Newsletter, 3. Heidelberg: The German Centre for Research on Aging.
- National Alliance for Caregiving & American Association of Retired Persons. (1997). Family caregiving in the U.S. Bethesda, MD, and Washington, DC.
- Nowlin, J.B. (1985). Introduction. In E. Palmore, E.W. Busse, G.L. Maddox, J.B. Nowlin, & LC. Siegler. (Eds.). Normal Aging III (pp. 1-4). Durham, NC: Duke University Press.
- Olbrich, E., & Lehr, U. (1976). Social roles and contacts in old age: Consistency and patterns of change. In H. Thomae (Ed.) Patterns of aging (pp.113-126). Basel: Karger.
- Petrie, D. (Director). (1993). Grumpy old men [FUm]. (Available from Warner Studios: http://www2.warnerbros.com)
- Riegel, K.F., & Riegel, R.M. (1972). Development, drop, and death. Developmental Psychology, 6(2), 306-319.
- Salariano, W.A., Minkler, M., & Langhauser, C. (1984). The significance of an ill spouse for assessing health differences in an elderly population. Journal of the American Geriatric Society, 32, 187-190.
- Schaie, K-W. (1989). Perceptual speed in adulthood: cross-sectional and longitudinal studies. Psychology and Aging, 4(4), 443-453.
- Schaie, K.W. (1996). Intellectual development in adulthood: The Seattle Longitudinal Study. New York: Cambridge University Press.
- Schulz, R-, O'Brien, A.T., Bookwala, J., & Fleissner, K. (1995). Psychiatric and physical morbidity effects of dementia caregiving: Prevalence, correlates, and causes. The Gerontologist, 35(6), 771-791.
- Shanan, J. (1991), Who and how: some unanswered questions in adult development. Journal of Gerontology: Psychological Sciences, 46, P309-P316.
- Shanas, E., & Maddox, G.L. (1976). Aging, health, and the organization of health resources. In R.H. Binstock Ec E. Shanas (Eds.), Handbook of aging and the social sciences (pp.592-618). New York: Van Nostrand Reinhold.
- Siegel, J.S. & Taeuber, C.M. (1986). Demographic perspectives on the long-lived society. Daedalus, 111(Winter), 77-117.
- Small, B.J., & Backman, L. (1999). Time to death and cognitive performance. Current Directions in Psychological Science, 8(6), 168-172.
- Wechsler, D. (1955). Manual for the Wechsler Adult Intelligence Scale. New York: Psychological Corporation.
- Weishaus, S. & Field, D. (1988). Half century of marriage: continuity or change? Journal of Marriage and the Family, 50(3), 763-774.
- Wright, L.K., Clipp, E.C. & George, L.K. (1993). Health consequences of caregiver stress. Medicine, Exercise, Nutrition, & Health, 2, 181-195.