Cohort comparisons in resources and functioning among centenarians: Findings from the Georgia Centenarian Study

Abstract

The purpose of this study was to examine cohort comparisons in levels of resources (e.g., mental health, physical functioning, economic and social resources, and cognitive functioning) for 211 community-dwelling centenarians (whose Mini-Mental Status Examination score was 23 or higher) of phases I and III of the Georgia Centenarian Study. The earlier cohort was defined as those born between 1881 and 1895 (part of phase I) and the later cohort included persons born between 1901 and 1907 (part of phase III). Five specific domains were compared: mental health; mental status; physical functioning; social resources; and economic resources. Results showed that there were significant cohort comparisons in five domains: mental health; mental status; physical functioning; social networks; and economic resources. Findings suggest that the later centenarian cohort was more satisfied with life, felt less depressed, showed less positive or negative emotion, had higher scores on perceived economic status, and higher levels of physical and cognitive functioning when compared to the earlier cohort. In conclusion, our findings suggest that recent cohorts of centenarians may be better off than previous ones with respect to several areas of individual resources. This study suggests that, even as the number of centenarians increases and some proportion of centenarians experience severe deterioration at the end of their life, there are improvements of functioning, health, and overall life quality among advanced older adults.

Keywords

centenarians cohort differences oldest-old adults

The proportion of oldest-old adults (those 85 years and older) is increasing rapidly and continuously in the United States (US) population. There was more than a 300% increase in those over 85 from 1960 to 2000 (US Bureau of the Census, 2004). In addition, the rapid growth of the oldest-old population and increment in exceptional longevity has occurred not only in Western countries, but also all over the world (Christensen, Doblhammer, Rau, & Vaupel, 2009; Engberg, Christensen, Andersen-Ranberg, Vaupel, & Jeune, 2008). The number of centenarians is expected to increase from 37,000 in 1990 to 850,000 in 2050 in the US. Growing numbers of centenarians may have some disabilities and lower levels of health, and there may be a need for medical care and treatment at the end of life (Andersen-Ranberg, Schroll, & Jeune, 2001; Engberg, Christensen, Andersen-Ranberg, Vaupel, et al., 2008). However, even though the number of centenarians and the likelihood to be dependent on caregivers increases due to considerable decline, particularly of functional health, there are several positive prospects for future generations of centenarians. Jeune (2002), for example, noted that we could expect the proportion of autonomous centenarians to increase in the future with the advancement of better medical treatment allowing centenarians to function better physically and cognitively and to continue with an active life style (Jeune, 2002).

It is widely documented that people encounter a number of physical limitations and declining cognitive abilities as they progress through later life (e.g., George, Landerman, Blazer, & Anthony, 1991; Krause, 2005; Nuland, 1994). However, only a few studies have investigated cohort comparisons of physical and cognitive functioning, and little research has explored cohort comparisons on other quality of life aspects such as mental health or social resources. In the present study, therefore, we examined cohort comparisons on several centenarians’ resources in two birth cohorts of centenarians born in the late 1800s and early 1900s. These two cohorts are defined by their birth years. Our work is consistent with life course perspectives that focus on age cohorts moving through historical time. Each cohort has its own diverse life experiences and is shaped by the circumstances encountered earlier in life (Elder, 1978; Hareven, 1978, 1994). Based on the life course perspective, this study extends prior research by: (1) examining levels of physical and cognitive functioning between two cohorts; (2) by comparing two cohorts on levels of mental health and perceived economic status in addition to physical and cognitive function; and (3) by highlighting mean cohort comparisons of levels of social support and networks.

A number of studies have assessed centenarians’ functional abilities and various characteristics such as decline of physical functioning, cognitive functioning, deterioration in housing, economic disadvantages, or nutritional risks (Bayer, 2000; Goetting, Martin, Poon, & Johnson, 1996; Holtsberg, Poon, Noble, & Martin, 1995; Payette, Coulombe, Boutier, & Gray-Donald, 1999; Poon et al., 2000). As such, most studies have focused on physical functioning and health related to mortality and morbidity and showed that there was less decline in these areas (functioning and health) in recent cohorts (Crimmins, 2004; Engberg, Christensen, Andersen-Ranberg, Vaupel, et al., 2008; Freedman, Martin, & Schoeni, 2002; Manton, Gu, & Lamb, 2006; Parker & Thorslund, 2007; Robine, 2006; Zunzunegui, Nunez, Durban, García de Yébenes, & Otero, 2006). However, there are different findings in physical functioning among the oldest-old population. Parker, Ahacic, and Thorslund (2005) found that there was a significant increase in self-reported diseases and symptoms among Swedish oldest-old people. Winblad, Jääskeläinen, Kivelä, Hiltunen, and Laippala (2001) compared the prevalence of disability among three different birth cohorts (born 1903, 1913, and 1923), but there was no significant cohort effect on activities of daily living (ADL) limitations. To our knowledge there are only two studies examining cohort comparisons in physical functioning among centenarians. Suzuki, Akisaka, Ashitomi, Higa, and Nozaki (1995/2008) noted an increase in disabilities among Okinawan centenarians from the 1970s through the 1990s (Suzuki et al., 1995/2008). In contrast, Engberg, Christensen, Andersen-Ranberg, Vaupel, et al. (2008) showed that, among Danish centenarians, later cohorts displayed better self-reported ADLs compared to earlier cohorts.

In addition to the cohort effect on physical functioning of centenarians, there was only one study that demonstrated cohort differences in cognitive function among Danish centenarians (Engberg, Christensen, Andersen-Ranberg, & Jeune, 2008). The authors showed that, although there were no significant differences in the cognitive functioning score (MMSE) between the two centenarian birth cohorts, modest tendencies were found towards better cognitive performance for the centenarians in the 1905 birth cohort living at home compared to the community-dwelling ones in the 1895 birth cohort, and worse cognitive functioning was found for the centenarians in the 1905 birth cohort living in nursing homes when compared to the 1895 birth cohort who lived in nursing homes (Engberg, Christensen, Andersen-Ranberg, & Jeune, 2008).

Even though no study has examined cohort differences in mental health or perception of economic status, post-materialism can explain cohort differences in mental health and perceived economic status. Post-materialism perspectives lend support to the hypothesis that a small increase in happiness among more recent cohorts (Rodgers, 1982; Yang, 2008) with more education and income have become less concerned with basic survival needs and more concerned with higher level needs such as social relations problems (Rodgers, 1982; Yang, 2004).

In terms of social support and social networks, few studies have examined cohort differences in social resources among oldest-old adults. Cohort differences in social resources might emerge for two different reasons. First, not only centenarians, but also persons in their social support network, have benefited from an overall increase in life expectancy. The life expectancy at the turn of the 20th century in the US was 47.3 years, and the expected life expectancy was 77.9 years in 2007 (National Center for Health Statistics, 2011). Furthermore, we could expect life expectancy at age 75 to be a further 11.7 years in 2007 (National Center for Health Statistics, 2011). On the one hand, higher life expectancy for later-born cohorts may suggest support for longer periods. On the other hand, the decline of fertility is a general trend in developed countries (Galasso, Gatti, & Profeta, 2009) and this may affect social support available in later life. The total fertility rate was 3.2 in 1920 and dropped to 2.1 in 2000 (Boldrin, De Nardi, & Jones, 2005; Galasso et al., 2009). Traditionally, children took a role of support systems for old age (Galasso et al., 2009). A decline in fertility may be a reason for smaller networks and support for older adults. It remains to be seen which of these possible trends have an effect on social resources of centenarians.

Taken as a whole, previous research leads to the following study hypotheses:

The levels of physical functioning and cognitive functioning in later cohorts will be higher than the levels of physical and cognitive functioning in earlier cohorts.

The levels of mental health and perceived economic status in later cohorts will be higher than the levels of mental health and perceived economic status in earlier cohorts.

Although there is conflicting evidence about cohort differences in social resources, we hypothesized that the levels of social networks and social support in later cohorts will be lower than in earlier cohorts.

Method

Participants

The study was based on two phases (phase I and phase III) of the Georgia Centenarian Study (GCS; phase I – Poon et al., 1992; phase III – Poon et al., 2007). Participants of phase I were required to be community-dwelling and cognitively intact (i.e., a score of 23 or higher on the MMSE; Folstein, Folstein, & McHugh, 1975; or 2 or lower on the Global Deterioration Scale; Reisberg, Ferris, de Leon, & Crook, 1982). The purpose of phase I was to investigate cognitive performances with reliance on self-report in conjunction with adaptation characteristics (e.g., personality, health, dietary measures), so the selection of cognitively intact participants was purposeful (Poon et al., 2007). This selection criterion resulted in the inclusion of individuals who were not community dwelling or were cognitively impaired, thus these individuals were not included in the final sample. Centenarians were recruited from a wide network of community contacts (including the offices of the governor and secretary of state) and local agencies, churches, television, and print media. A sample of 127 centenarians of phase I was included in this study.

In terms of phase III data, the sampling frame of the GCS (phase III; Poon et al., 2007), which targeted a representative sampling of centenarians and near centenarians (i.e., 98 years and older), had two components: using census proportions and the date of birth information in voter registration files. Based on these two components and five different characteristics (geographic, age, gender, race, and type of residence) a sample of centenarians and near centenarians was drawn for this study (Poon et al., 2007). A sample of 84 centenarians and near centenarians who were cognitively intact (MMSE score was 23 or higher) and community dwelling were included in this study.

Although the sampling strategies for the early and late cohort were somewhat different, it must also be noted that centenarians belong to a rare but typically well-known population. Centenarians are known particularly in smaller communities and often readily participate in our studies. We are therefore confident that, in spite of the different sampling strategies, we generated a somewhat similar group of participants.

Therefore, the total number of centenarians and near centenarians (i.e., 98 years and older) of this study was 211: 127 from phase I and 84 from phase III. Whereas the earlier cohort was defined as the participants in phase I who were born from 1881 to 1895, the later cohort was defined as persons in phase III who were born from 1901 to 1907. In this study, 76.8% of participants were women. About 85% were widowed, and 45.7% did not complete high school. There was a significant difference in education levels between the two cohorts, χ²(1,208) = 17.46, p < .05. The later cohorts had higher education levels than the earlier cohorts did.

Compared to the entire pool of participants from both phases, participants included in this study had similar characteristics. There was only one significant difference between the excluded samples and the study sample in education, χ²(7, N = 278) = 15.95, p < .05. Almost three-quarters of the excluded sample (74.3%) had a high school graduation or less, whereas only 58.6% of the sample used in this study had a high school graduation or less. There was no significant difference in other characteristics. A summary of demographic characteristics by cohorts is presented in Table 1.

Table 1.

Summary of demographic characteristics (N = 211)

	Earlier cohort		Later cohort		χ²
Demographic characteristics	n	%	n	%
Age	100–101^a	(101.66)^b	98.11–101.89^a	(99.38)^b
Gender
Female	97	(76.4)	65	(77.4)	.03
Male	30	(23.6)	19	(22.6)
Marital status
Currently married	5	( 4.0)	6	( 7.1)	2.95
Divorced	3	( 2.4)	4	( 4.8)
Widowed	108	(86.4)	71	(84.5)
Never married	9	( 7.2)	3	( 3.6)
Ethnicity
White/Caucasian	96	(75.6)	71	(84.5)	2.45
Black/African American	31	(24.4)	13	(15.5)	2.45
Education
0–4 years	14	(11.1)	1	( 1.2)	17.46*
5–8 years	38	(30.2)	15	(18.3)
Some high school	14	(11.1)	13	(15.9)
High school diploma	13	(10.3)	15	(18.3)
Trade school or vocational degree	8	( 6.3)	2	( 2.4)
Some college	18	(14.3)	13	(15.9)
College degree	12	( 9.5)	14	(17.1)
Graduate degree	9	( 7.1)	9	(11.0)

Note. *p < .05.

^a Range of age

^b Mean of age

Measures

Life satisfaction

Liang’s (1985) adaptation of the Life Satisfaction Index A was used to examine the levels of life satisfaction among centenarians in both cohorts. This scale is composed of seven items. The answer is scaled so that −1 = disagree; 0 = in between; or 1 = agree. The seven items were summed and the range of summary score is from −7 to 7. Internal consistency of this scale for all participants is α = .72. Higher scores indicated higher levels of life satisfaction.

Depression

The Geriatric Depression Scale (GDS; Yesavage et al., 1983) was used to assess depressive symptoms reported by centenarians. The short form of the GDS used in this study consisted of 15 items. All answers were coded as dichotomous variables (1 = yes, 0 = no). Fifteen items were summed and the range of summary score is from 0 to 15. Cronbach’s α of this scale is .63 for all participants. Higher scores indicated higher levels of depressive symptomatology.

Positive and negative affect

The Bradburn Affect Balance Scale (Bradburn, 1969) was used to examine positive and negative affect. Two dimensions (positive affect and negative affect) composed this scale. Five positive affect items (α = .52) and five negative affect items (α = .66) were used in this study. Answer categories included “not at all (= 1),” “once (= 2),” “several times (= 3),” and “often (= 4).” The ranges of each dimension were from 5 to 20. Higher scores indicated higher levels of positive and negative mood.

Physical functioning

Physical functioning was assessed with the self-care capacity (ADL) scales from the Duke Older Americans Resources and Services procedures (OARS; Fillenbaum, 1988). These 13 self-report items assess the difficulty participants have with instrumental (e.g., shopping, cooking, and cleaning) and physical tasks (e.g., bathing or showering, dressing, eating, getting in and out of bed or a chair, walking, and getting outside). All 13 items were scaled so that 2 = without help (e.g., can clean floors, etc.); 1 = with some help (e.g., can prepare some things but unable to cook full meals themselves); or 0 = completely unable (e.g., not able to prepare any meals). The internal consistency estimate of this scale was .82 for all participants in this study. The summary score of 13 items ranges from 0 to 26. Physical functioning was scaled so that higher scores indicate higher levels of self-care capacity.

Social networks

The Duke OARS (Fillenbaum, 1988) was used to assess social networks and social supports. Participants assessed their social networks (e.g., “How many people do you know well enough to visit with in your home or in their homes?”) and social support (“How many times during the past week did you spend some time with someone who does not live with you; that is, you went to see them or they came to visit you, or you went out to do things together?”). The two questions were used separately as single items. Higher scores indicated higher levels of social networks and social support.

Perceived economic status

Perception of economic status adequacy was assessed with the economic resources scale items from the Duke OARS (Fillenbaum, 1988). Participants assessed their economic resources (e.g., “Are your assets and financial resources sufficient to meet emergencies?,” “How well does the amount of money you have take care of your needs?,” “Are your expenses so heavy that you cannot meet the payments, or can you barely meet the payments, or are your payments no problem?,” and “Do you usually have enough to buy those little extras; that is, those small luxuries?”). The four items yielded a reliability of α = 0.76 in this study. The summary score ranged from 1 to 6. Higher scores indicate perceptions of higher economic status.

Cognitive functioning

Cognitive functioning was assessed via the MMSE (Folstein et al., 1975). The MMSE is composed of five sections: orientation; registration; attention and calculation; recall; and language. The internal consistency estimate of MMSE in this sample is .76. The score ranges from 23 to 30 in the analysis sample with higher scores on the MMSE indicative of better mental status.

Data analysis

Comparisons of mean differences in resources and functioning variables between the two cohorts are shown in Figure 1. Differences between the cohorts on nine variables were tested using analysis of covariance (ANCOVA) with education as a covariate. A p value with a Bonferroni correction was considered to indicate statistical significance. All raw scores were converted to standardized scores (z-scores). Data were analyzed using the SPSS Statistical Software Package (version 18.0).

Figure 1.

Mean comparisons for resources and adaptation between earlier cohort and later cohort. Note. Values are z-scores. LS = life satisfaction; DP = depressive symptoms; PA = positive affect; NA = negative affect; MMSE = mini-mental status exam; ADL = activities of daily living; SN1 = number of people to know and to visit; SN2 = spending time with someone; PES = perceived economic status.

Results

The goal of this study was to assess comparisons in resources and functioning such as mental health, physical functioning, mental status, social resources, and economic status between two centenarian cohorts. Given the significant cohort difference in education, ANCOVA was conducted to examine cohort differences in resources and functioning variables with education as a covariate. ANCOVA showed that there were eight significant differences: life satisfaction, depression, positive affect, negative affect, mental status, ADL, perceived economic status, and spending time with someone (Table 2 and Figure 1). First, in terms of mental health, there were significant differences in four measures. The mean difference for life satisfaction was −.56, F(1, 194) = 16.13, p < .001. The mean difference for depression was .76, F(1, 190) = 31.34, p < .001. The mean difference for positive affect was .30, F(1, 188) = 3.91, p < .05. The mean difference for negative affect was .60, F(1, 199) = 18.71, p < .001. Therefore it appears that the later cohort was more satisfied with life, less depressed, emotionally less positive and less negative than the earlier cohort.

Table 2.

Examination of cohort differences in resources and adaptation

Dependent variables^1,2	Mean Differences³	Std. Error	95% CI	F	Cohen’s f
Mental health
Life satisfaction	−.56^***	.14	[−.84, −.29]	16.13	.29
Depressive symptoms	.76^***	.14	[.50, 1.03]	31.34	.41
Positive affect	.30^*	.15	[.00, .60]	3.91	.14
Negative affect	.60^***	.14	[.33, .88]	18.71	.31
Mental functioning
Mini-mental status exam	−.47^**	.14	[−.74, −.21]	12.14	.24
Physical functioning
Activities of daily living	−.30^*	.15	[−.59, −.01]	4.15	.15
Social networks
Number of people to know and to visit	−.13	.16	[−.41, .16]	.76	.06
Spending time with someone	−.61^***	.14	[−.89, −.34]	19.36	.31
Economic resources
Perceived economic status	−.35^*	.15	[−.65, −.06]	5.52	.18

¹All dependent variables were z-scores.

²Education was included as a covariate in all models.

³Bonferroni adjustment was applied for statistical significance.

Note. *p < .05; **p < .01; ***p < .001.

Second, with regard to mental status, the MMSE showed a significant difference between the two cohorts. The mean difference for MMSE was −.47, F(1, 205) = 12.14, p < .01. Furthermore, there was a significant mean difference in ADL measuring physical functioning. The mean difference for ADL was −.30 F(1, 192) = 4.15, p = .04. Thus the later cohort reported better cognitive and physical functioning than the earlier cohort.

Lastly, in terms of economic resources and social networks, there were significant differences in perceived economic status and in the frequency of spending time with someone with whom centenarians did not live. The mean difference for perceived economic status was −.35, F(1, 172) = 5.52, p = .02; for the frequency to spend time with someone, −.61, F(1, 201) = 19.36, p < .001. This suggests that later cohorts perceived their economic status as higher but spent less time with other people than earlier cohorts did. However, no significant difference was found in the overall social network, F(1, 203) = .76, p = .36.

Overall, the first hypothesis that the levels of physical functioning and cognitive functioning in the later cohorts would be higher than the levels of physical and cognitive functioning in the earlier cohort was supported with this study. In addition, the second hypothesis that the levels of mental health and perceived economic status in the later cohort would be higher than the levels of mental health and perceived economic status in the earlier cohort was supported by the results. Even though there were significant differences between the two cohorts, only the levels of life satisfaction in the later cohort were higher than those in the earlier cohort. The earlier cohort had higher levels on the other three measures (i.e., depressive symptoms, positive affect, and negative affect) than the later cohort. The last hypothesis that the levels of social networks in the earlier cohorts would be higher than the levels of social networks in the later cohorts was partially supported. Specifically, there was a significant difference in the frequency of time spent with others, but a significant difference in social network was not found.

Discussion

The purpose of this study was to highlight cohort comparisons of centenarians with regard to mental health, physical and cognitive functioning, and social resources. This study was based on the life course perspective which emphasizes that each cohort has its own diverse life experiences and is shaped by circumstances encountered earlier in life (Elder, 1978; Hareven, 1978, 1994). A major finding indicates that there were cohort differences with respect to mental health, physical and cognitive functioning, perceived economic status, and the frequency of time spent with others. Compared to the earlier cohort, the later cohort was more satisfied with life, felt less depressed, reported less negative emotions, greater perceived economic status, and higher levels of physical and cognitive functioning, despite less reported positive affect and more opportunities to spend time with others when compared to the earlier cohort. This finding is noteworthy because it demonstrates for the first time in centenarian studies that cohort differences may exist among centenarians, not only in terms of physical and cognitive functioning, but also in mental health and social resources.

The results of this study are consistent with previous research. First, in terms of physical functioning, our later birth cohort achieved higher levels of physical functioning (i.e., activities of daily living [ADL]) than the earlier birth cohort. This is similar to the findings by Engberg, Christensen, Andersen-Ranberg, Vaupel, et al. (2008), whose Danish 1905 cohort (later cohort) reported better ADL capability compared to their 1895 cohort (earlier cohort). Engberg and colleagues’ work was based on specific ADL items, whereas our results were based on a summary score. Nonetheless, our later cohort born from 1901 to 1907 also showed significantly higher scores of ADL compared to the earlier cohort born from 1881 to 1895. Second, in terms of cognitive functioning, the results of this study are again consistent with Engberg, Christensen, Andersen-Ranberg, and Jeune (2008)’s work. Even though Engberg et al. did not find significant cohort differences in cognitive functioning, (i.e., MMSE scores), they reported modest tendencies that later cohorts were functioning better than the earlier cohorts of community-dwelling centenarians. In the same vein, our later cohort of community-dwelling centenarians achieved higher levels of MMSE scores than the earlier cohort of community-dwelling centenarians.

One explanation for better physical and cognitive functioning among the later cohort may be differences in advances in medical sciences, in particular improvement in assistive devices, which were not available to the same extent to the earlier born cohort. This explanation could be partly based on the life course perspective, in particular historical effects. The two cohorts experienced historical events of the 20th century at different ages. For example, while the earlier cohort experienced World War I (WWI) in their 20s and 30s, the later cohort members were still children. The earlier cohort experienced the Great Depression in their 30s and 40s, whereas the later cohort experienced the Great Depression in their 20s and 30s. Of course, the later cohort could also take advantage of an improved health care system and earlier support through Social Security and the Medicare and Medicaid system. Therefore, with help from assistive devices, later cohorts might maintain higher levels of independence (Engberg, Christensen, Andersen-Ranberg, Vaupel, et al., 2008).

It could be argued that these factors helped our later cohort of centenarians to be less frail and intact. Alternatively, some studies suggested that physical hardship experienced by the pre-1900 birth cohorts contributed to more positive health assessments than for the after-1900 birth cohorts (Cain, 1967; Idler, 1993; Rodgers, 1982). It is possible that earlier cohorts were hardier than those later born. As Rodgers (1982) mentioned, however, physical hardship made it harder to overcome disabilities and discomforts, and produced lower levels of happiness.

In addition to the effect of historical trend, the context of culture may explain the inconsistency between Suzuki et al. (1995/2008) and the findings of this study. Suzuki et al. (1995/2008) found that an earlier cohort (1870s) had better physical functioning than a later cohort (1890s) among Okinawan centenarians. It may be plausible to explain the inconsistency with a different context of culture (e.g., differences in development of health care systems and experiences in wars such as WWI and World War II [WWII]).

In addition to physical functioning and cognitive functioning, two other worthy findings were useful as a contribution to centenarian research. First, there were significant differences in mental health between the two centenarian birth cohorts. Specifically, the later birth cohort born from 1901 to 1907 reported higher levels of life satisfaction and lower levels of depressive symptoms and negative affect despite slightly lower levels of positive affect than the earlier birth cohort born from 1881 to 1895. Second, the later cohort perceived their economic status more positively than the earlier cohort did. Taken together, this finding implies that later cohorts of centenarians may have more positive perceptions of their life and are functioning better physically and cognitively.

We found significant differences in four measures of mental health, cognitive and physical functioning, perceived economic status, and frequency of time spent with someone with whom centenarians did not live. However, one factor, number of people to know and to visit, did not show a significant difference between the two cohorts. One plausible explanation is based on the socio-emotional selectivity perspective which suggests that, when people construct their social networks, they could be influenced by changes in time perspective and related to changes in social preferences (Löckenhoff & Carstensen, 2004). Therefore, centenarians who lived longer than others experienced the loss of family members or friends through their long life regardless of when they were born. It appears to be more significant how much time they spend with people rather than the frequency of time they spend with acquaintances.

When interpreting the present results, potential limitations of this study should be considered. First, there was a difference in sampling of participants between phase I and phase III. The centenarians who were included for the earlier cohort (phase I) were limited to those who lived in households and who were cognitively intact, whereas the centenarians who were recruited for the later cohort (phase III) were based on a population-based sampling (Poon et al., 2007). To correct for the different sampling strategies, we excluded centenarians from phase III who did not match the inclusion criteria of phase I. Because we excluded centenarians who were cognitively impaired and who did not live in the community, we do not know whether it is more or less likely for centenarians of either cohort to be cognitively impaired or to live in long-term care facilities.

The differing recruitment strategies could be responsible for the obtained cohort differences. More specifically, positive cohort differences could be due to the fact that the later cohort was a more positively selected oldest-old group, given that centenarians belonging to the earlier cohort were recruited from a wide network of community contacts such as local agencies, churches, television, and print media. In contrast, voter registration files were used for the later cohort. These lists may be biased by a number of demographic criteria (e.g., education, race, etc.). Even though we used some of the demographic variables as covariates, these variables may only in part explain differences in selectivity. We also need to caution that the later cohort was recruited from a more restricted area in Georgia (the northern 44 counties), whereas the earlier cohort was recruited from the entire state of Georgia. The two samples also differed somewhat in age, given that only centenarians were included in the earlier cohort, whereas the later cohort included participants who were 98 years and older. Finally, the intervals within cohorts were 14 years for the earlier cohort and 6 years for the later cohort. There were only 6 years separating the two groups. The differences between two cohorts should therefore be interpreted with caution. Even though it turned out there were significant differences in a number of domains, a span of 6 years may not be enough to detect cohort differences in social networks among centenarians (Engberg, Christensen, Andersen-Ranberg, & Jeune 2008).

However, taking these alternative explanations into consideration, it is useful to contemplate historical changes that may account for cohort differences as well. One historical change that possibly differentiates the two cohorts includes the child labor law, enacted in the early 1900s, that could have contributed to significant differences in education and early work experiences between the earlier and later cohort (Schaie, Willis, & Pennak, 2005). The earlier cohort might have been employed earlier in the life course than the later cohort, providing an important source for family income, whereas the later cohort was more likely to attend school at similar ages. Third, the participants of this study were recruited from only one geographic region of the US. Individuals from other parts of the US or other countries might exhibit different patterns of cohort difference on centenarians’ resources. For example, Branscum, Martin, MacDonald, Margrett, and Poon (2009) showed that there were significant differences in items of ADL, depression, and economic burden scales between US centenarians residing in Georgia and Iowa. Branscum and colleagues did not compare cohort differences by region; however, their results imply that researchers should be careful not to generalize the findings of this study to the whole population of centenarians.

Despite several limitations, our study has important strengths. Even though there are different levels in functioning among the centenarian population, both phases included sizeable numbers of centenarians who were community dwelling and cognitively intact. Furthermore, we considered cohort differences among centenarians in several domains. Several studies have focused on cohort differences in one or two domains, in particular in cognitive and physical functioning of centenarians (Engberg, Christensen, Andersen-Ranberg, & Jeune 2008; Engberg, Christensen, Andersen-Ranberg, Vaupel, et al., 2008; Parker et al., 2005; Suzuki et al., 1995/2008). Otherwise, sequential designs have been used to detect inter-individual differences in specific domains (e.g., cognition) in later life (Gerstorf, Ram, Hoppmann, Willis, & Schaie, 2011). This study, as “an initial step toward a systemic-multivariate perspective” (Gerstorf et al., 2011, p. 1038) for centenarian studies, extends previous research on centenarian cohort differences by comparing mental health, perceived economic status, and social supports among centenarians using time-lagged comparisons. The results provide evidence for the notion of compression of morbidity (Fries, 1980). Comparing two cohorts of those who lived for nearly 100 years or longer, who were cognitively intact, and who lived independently may allow future cohorts to maximize their life span and to achieve long life with high quality.

In conclusion, our findings suggest that recent cohorts of centenarians are likely to be better off than previous ones with respect to several domains of functioning and individual resources. The findings of this study suggest that, even though the number of centenarians is increasing and some proportion of centenarians experience severe deterioration at the end of their life, there are improvements of functioning, health, and overall life quality among advanced older adults in later cohorts.

Footnotes

Acknowledgements

The Georgia Centenarian Study (Leonard W. Poon, PI) was funded by 1P01-AG17553 from the National Institute on Aging, a collaboration among The University of Georgia, Tulane University Health Sciences Center, Boston University, University of Kentucky, Emory University, Duke University, Wayne State University, Iowa State University, Temple University, and University of Michigan. Authors acknowledge the valuable recruitment and data acquisition effort of M. Burgess, K. Grier, E. Jackson, E. McCarthy, K. Shaw, L. Strong and S. Reynolds, data acquisition team manager; S. Anderson, E. Cassidy, M. Janke, and J. Savla, data management; M. Poon for project fiscal management. Additional investigators include S. M. Jazwinski, R. C. Green, M. Gearing, W. R. Markesbery, J. L. Woodard, J. S. Tenover, I. C. Siegler, W. L. Rodgers, D. B. Hausman, C. Rott, A. Davey, and J. Arnold.

References

Andersen-Ranberg

Schroll

Jeune

(2001). Healthy centenarians do not exist, but autonomous centenarians do: A population-based study of morbidity among Danish centenarians. Journal of the American Geriatrics Society, 49, 900–908. doi: 10.1046/j.1532-5415.2001.49180.x

Bayer

A.H.

(2000). Fixing to stay: A national survey on housing and home modification issues. Washington, DC: AARP.

Bradburn

N. M.

(1969). The structure of psychological well-being. Chicago, IL: Aldine.

Branscum

A. Y.

Martin

MacDonald

Margrett

Poon

(2009). Regional differences in functional domains: A comparison between Georgia and Iowa centenarians. The Gerontologist, 49, 69.

Boldrin

De Nardi

Jones

L. E.

(2005). Fertility and social security. Retrieved from SSRN: http://ssrn.com/abstract=1174402

Cain

L. D.

(1967). Age status and generational phenomena: The new old people in contemporary America. The Gerontologist, 7, 83–92.

Christensen

Doblhammer

Rau

Vaupel

J. W.

(2009). Ageing populations: The challenges ahead. Lancet, 374, 1196–1208. doi: 10.1016/S0140-6736(09)61460-4

Crimmins

E. M.

(2004). Trends in the health of the elderly. Annual Review of Public Health, 25, 79–98. doi: 10.1146/annurev.publhealth.25.102802.124401

Elder

G. H.

(1978). Family history and the life course. In Hareven

T. K.

(Ed.), Transitions: The family and the life course in historical perspective (pp. 17–64). New York, NY: Academic.

10.

Engberg

Christensen

Andersen-Ranberg

Jeune

(2008). Cohort changes in cognitive function among Danish centenarians–A comparative study of 2 birth cohorts born in 1895 and 1905. Dementia and Geriatric Cognitive Disorders, 26, 153–160. doi: 10.1159/000149819

11.

Engberg

Christensen

Andersen-Ranberg

Vaupel

J. W.

Jeune

(2008). Improving activities of daily living in Danish centenarians–But only in women: A comparative study of two birth cohorts born in 1895 and 1905. Journal of Gerontology: Medical Sciences, 63A, 1186–1192.

12.

Fillenbaum

G. G.

(1988). Multidimensional functional assessment of older adults: The Duke older Americans resources and services procedures. Hillsdale, NJ: Lawrence Erlbaum Associates.

13.

Folstein

S. E.

McHugh

P. R.

(1975). Mini-mental state: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 2, 189–198.

14.

Freedman

V. A.

Martin

L. G.

Schoeni

R. F.

(2002). Recent trends in disability and functioning among older adults in the United States: A systematic review. The Journal of the American Medical Association, 288, 3137–3146.

15.

Fries

J. F.

(1980). Aging, natural death, and the compression of morbidity. New England Journal of Medicine, 303, 130–135. doi: 10.1056/NEJM198007173030304

16.

Galasso

Gatti

Profeta

(2009). Investing for the old age: Pensions, children and savings. International Tax and Public Finance, 16, 538–559. doi: 10.1007/s10797-009-9104-5

17.

George

L. K.

Landerman

Blazer

D. G.

Anthony

J. C.

(1991). Cognitive impairment. In Robins

L. N.

Regier

D. A.

(Eds.), Psychiatric disorders in America: The epidemiologic catchment area study (pp. 291–327). New York, NY: Free Press.

18.

Gerstorf

Ram

Hoppmann

Willis

S. L.

Schaie

K. W.

(2011). Cohort differences in cognitive aging and terminal decline in the Seattle longitudinal study. Developmental Psychology, 47, 1026–1041. doi: 10.1037/a0023426

19.

Goetting

Martin

Poon

L. W.

Johnson

M. A.

(1996). The economic well-being of community-dwelling centenarians. Journal of Aging Studies, 10, 43–55.

20.

Hareven

T. K.

(1978). Historical changes in the life course and the family. In Yinger

J. M.

Cutler

S. J.

(Eds.), Major social issues: Multidisciplinary view (pp. 338–345). New York, NY: Free Press.

21.

Hareven

T. K.

(1994). Aging and generational relations: A historical and life course perspective. Annual Review of Sociology, 20, 437–461.

22.

Holtsberg

P. A.

Poon

L. W.

Noble

C. A.

Martin

(1995). Mini-mental state exam status of community dwelling, cognitively intact centenarians. International Psychogeriatrics, 7, 417–427.

23.

Idler

E. L.

(1993). Age differences in self-assessments of health: Age changes, cohort differences, or survivorship? Journal of Gerontology: Social Sciences, 48, S289–S300.

24.

Jeune

(2002). Living longer—but better? Aging Clinical and Experimental Research, 14, 72–93.

25.

Krause

(2005). Exploring age differences in the stress-buffering function of social support. Psychology and Aging, 20, 714–717. doi: 10.1037/0882-7974.20.4.714

26.

Liang

(1985). A structural integration of the affect balance scale and the life satisfaction index A. Journal of Gerontology, 40, 552–561. doi: 10.1093/geronj/40.5.552

27.

Löckenhoff

C. E.

Carstensen

L. L.

(2004). Socioemotional selectivity theory, aging, and health: The increasingly delicate balance between regulating emotions and making tough choices. Journal of Personality, 72, 1395–1424. doi: 10.1111/j.1467-6494.2004.00301.x

28.

Manton

K. G.

Lamb

V. L.

(2006). Change in chronic disability from 1982 to 2004/2005 as measured by long-term changes in function and health in the US elderly population. Proceedings of the National Academy of Sciences, 103, 18374–18379. doi: 10.1073/pnas.0608483103

29.

National Center for Health Statistics. (2011). Health, United States, 2010: With special feature on death and dying. Hyattsville, MD: National Center for Health Statistics.

30.

Nuland

S. B.

(1994). How we die: Reflections on life’s final chapter. New York, NY: Knopf.

31.

Parker

M. G.

Ahacic

Thorslund

(2005). Health changes among Swedish oldest old: Prevalence rates from 1992 and 2002 show increasing health problems. Journal of Gerontology: Medical Sciences, 60, 1351–1355. doi: 10.1093/gerona/60.10.1351

32.

Parker

M. G.

Thorslund

(2007). Health trends in the elderly population: Getting better and getting worse. Gerontologist, 47, 150–158.

33.

Payette

Coulombe

Boutier

Gray-Donald

(1999). Weight loss and mortality among free-living frail elders: A prospective study. Journal of Gerontology: Medical Sciences, 54A, M440–M445. doi: 10.1093/gerona/54.9.M440

34.

Poon

Clayton

G. M.

Martin

Johnson

M. A.

Courtenay

B. C.

Sweaney

A. L.

Thielman

S. B.

(1992). The Georgia centenarian study. International Journal of Aging and Human Development, 34, 31–46.

35.

Poon

L. W.

Jazwinski

S. M.

Green

R. C.

Woodard

J. L.

Martin

Rodgers

W. L.

Dai

(2007). Methodological considerations in studying centenarians: Lessons learned from the Georgia Centenarian Studies. In Poon

L. W.

Perls

T. T.

(Eds.), Annual review of gerontology and geriatrics: Vol. 27. Biopsychosocial approaches to longevity (pp. 231–264). New York, NY: Springer.

36.

Poon

L. W.

Johnson

M. A.

Davey

D. V.

Dawson

D. V.

Siegler

I. C.

Martin

(2000). Psycho-social predictors of survival among centenarians. In Martin

Rott

Hagberg

Morgan

(Eds.), Autonomy versus dependence in the oldest old (pp. 77–89). New York, NY: Springer.

37.

Reisberg

Ferris

S. H.

de Leon

M. J.

Crook

(1982). The global deterioration scale for assessment of primary degenerative dementia. American Journal of Psychiatry, 139, 1136–1139.

38.

Robine

J. M.

(2006).Trends in population health. Aging Clinical and Experimental Research, 18, 349–351.

39.

Rodgers

(1982). Trends in reported happiness within demographically defined subgroups, 1957–78. Social Forces, 60, 826–842.

40.

Schaie

K. W.

Willis

S. L.

Pennak

(2005). An historical framework for cohort differences in intelligence research in human development. Research in Human Development, 2, 43–67. doi: 10.1080/15427609.2005.9683344

41.

Suzuki

Akisaka

Ashitomi

Higa

Nozaki

(2008). Chronological study concerning ADL among Okinawan centenarians. In Engberg

Christensen

Andersen-Ranberg

Jeune

(Eds.), Cohort changes in cognitive function among Danish centenarians – A comparative study of 2 birth cohorts born in 1895 and 1905. Dementia and Geriatric Cognitive Disorders, 26, 153–160. (Original work published 1995)

42.

United States (US) Bureau of the Census. (2004). US interim projections by age, sex, race, and Hispanic origin. Retrieved from http://www.census.gov/ipc/www/usinterimproj/

43.

Winblad

Jääskeläinen

Kivelä

S. L.

Hiltunen

Laippala

(2001). Prevalence of disability in three birth cohorts at old age over time spans of 10 and 20 years. Journal of Clinical Epidemiology, 54, 1019–1024. doi: 10.1016/S0895-4356(01)00370-5

44.

Yang

(2008). Social inequalities in happiness in the United States, 1972 to 2004: An age-period-cohort analysis. American Sociological Review, 73, 204–226. doi: 10.1177/000312240807300202

45.

Yesavage

J. A.

Brink

T. L.

Rose

T. L.

Lum

Huang

Adey

M. B.

Leirer

V. O.

(1983). Development and validation of a geriatric depression screening scale: A preliminary report. Journal of Psychiatric Research, 17, 37–49.

46.

Zunzunegui

M. V.

Nunez

Durban

García de Yébenes

M. J.

Otero

(2006). Decreasing prevalence of disability in activities of daily living, functional limitations and poor self-rated health: A 6-year follow-up study in Spain. Aging Clinical and Experimental Research, 18, 352–358.