Motivated by a vision for a “new gerontology” that focuses on optimal functioning rather than losses and declines, Rowe and Kahn (Reference Rowe and Kahn1987) argued that many of the age-related declines that define usual aging are due to modifiable factors such as diet, exercise, personal habits such as smoking, and psychosocial factors. Thus, individuals may engage in lifestyle modifications to avoid diseases and, consequently, enhance their chance of aging well. Three domains of outcomes were said to jointly define successful aging: (a) low probability of disease and disease-related disability, (b) high cognitive and physical functioning, and (c) engagement with life (namely in the form of social connectedness and productive activities; Rowe and Kahn, Reference Rowe and Kahn1997). All three criteria are meant to be relative (i.e. without absolute thresholds), although subsequent studies tend to adopt a threshold approach (see Depp and Jeste, Reference Depp and Jeste2006). The three criteria are somewhat hierarchical, in that avoiding diseases (and associated risk factors) makes high physical and cognitive functioning possible, which in turn enables the individual to engage in social and productive activities (Rowe and Kahn, Reference Rowe and Kahn1997). However, remaining free of disease/disability and maintaining high physical and cognitive functioning are less likely as one ages, whereas participation in social and even productive activities may not be affected until significant physical and cognitive declines have occurred (McLaughlin et al., Reference McLaughlin, Connell, Heeringa, Li and Roberts2010; Yaffe et al., Reference Yaffe2010; Hsu and Jones, Reference Hsu and Jones2012).
There were a number of engaging debates about the merits of this model. Perhaps the most common critique concerns the “biomedicalization” of aging, the image of “physiologic agelessness” seemingly portrayed, the branding of the majority who cannot achieve this standard as “losers,” and the bias toward those who are genetically advantaged (Strawbridge et al., Reference Strawbridge, Wallhagen and Cohen2002; Moody and Sasser, Reference Moody and Sasser2012).
Some scholars have argued to let older persons define for themselves whether they are aging well. Not surprisingly, self-rated successful aging, typically on the basis of a single item like “I am aging well,” results in much higher proportions of people in the “successful” category, with one study reporting as much as 92% being successful (Montross et al., Reference Montross2006). While people may be reluctant to label themselves as being “unsuccessful” and may adopt different frames of mind in the exercise, the measure would be of little practical or theoretical value if almost everyone rates himself or herself the same way. Pruchno et al. (Reference Pruchno, Wilson-Genderson and Cartwright2010), however, argued that both objective (or researcher-defined) and subjective criteria should be incorporated into measuring successful aging, and their structural equation model showed a high correlation between objective and subjective successful aging at the latent variable level (r = 0.65). Another approach is to ask lay people to define the contours of successful aging. Aside from attributes that are generally favored regardless of age, such as good health, financial status, enjoyment, and social activities (Bowling, Reference Bowling2006), themes about adaptation to challenges, positive attitudes, acceptance, contentment, and realistic self-appraisal (von Faber et al., Reference von Faber2001; Phelan et al., Reference Phelan, Anderson, LaCroix and Larson2004; Reichstadt et al., Reference Reichstadt, Depp, Palinkas, Folsom and Jeste2007) are informative, as discussed below. Another notable finding is that ordinary people do not regard absence of diseases and disability to be an essential component of successful aging as many with health problems self-identify as successful agers (Strawbridge et al., Reference Strawbridge, Wallhagen and Cohen2002; Montross et al., Reference Montross2006; Romo et al., in press).
Furthermore, in contrast to Rowe and Kahn's “outcome-oriented” approach, a number of “process-oriented” models have been proposed (Baltes and Baltes, Reference Baltes, Baltes, Baltes and Baltes1990; Schulz and Heckhausen, Reference Schulz and Heckhausen1996; Brandtstädter and Rothermund, Reference Brandtstädter and Rothermund2002). The emphasis is on the management of declines and losses while sustaining the motivation for self-maintenance or improvement as well as stabilizing subjective self-appraisal. Most notably, the model of selective optimization and compensation (SOC) argues that with age-related declines and losses, individuals need to disengage from unattainable goals. In order to optimize performance, they need to concentrate resources selectively on attainable goals and employ compensatory strategies when necessary. Kahn (Reference Kahn, Poon, Gueldner and Sprouse2003) considered Baltes and Baltes’ model to be complementary to the one by Rowe and Kahn.
I contend that much of the controversy in this field can be resolved by taking the idea of relativity more seriously, by distinguishing between pathways and end states, and by expanding the contours of successful aging to include psychological functioning. In the following, I will discuss these issues and a revised conceptual model in which age-graded physical, cognitive, psychological, and social functioning together characterize successful aging.
Disentangling means and outcomes
Outcomes refer to criteria of success. Means are mechanisms and pathways by which the achievement of outcomes is enhanced. Distinguishing means from outcomes can have far-reaching implications for the concept of successful aging.
As Rowe and Kahn (Reference Rowe and Kahn1998) had commented, “The MacArthur Study . . . was rooted in our concept of ‘successful aging’ – that is, the many factors which permit individuals to continue to function effectively [italics added], both physically and mentally, in old age” (p. xii). It is the present author's view that the subsequent focus of the field on the avoidance of diseases and disease-related risk factors as a goal, and its seeming impossibility, is misplaced. Avoiding risk factors and diseases is a pathway to effective functioning, but certainly not the only pathway. I have no intention to downplay the role disease conditions have on various domains of functioning and the importance, from a public health perspective, to avoid diseases and their risk factors. For instance, prevention of obesity alone can lead to significant reductions in cardiovascular diseases, diabetes, cancer, and thus healthcare costs. However, one would not say that an older person is unsuccessful simply because he or she has abdominal obesity. Likewise, Strawbridge and colleagues (Reference Strawbridge, Wallhagen and Cohen2002) commented that “most gerontologists are not ready to call someone unsuccessful merely because he or she is . . . diagnosed with diabetes” (p. 728). True, but this debate lingers on only because absence of diseases is treated as an outcome; when it is recognized as a means to achieve effective functioning, its role in successful aging is duly recognized without arousing debates as to whether it is realistic and whether successful aging is a biomedical paradigm. To classify individuals as successful or unsuccessful on the basis of risk factors and diseases may have confused means with meaningful outcomes.
It should be noted that not all diseases have direct bearing on disability or physical functioning. Some common chronic diseases, such as hypertension, hyperlipidemia, and atherosclerosis, are for the most part asymptomatic unless they become very serious. With advances in medicine and assistive technologies, many chronic conditions may have little impact on day-to-day functioning, especially when they are properly managed. Other conditions, especially mental illnesses and dementia, are defined partly on the basis of functional loss (in non-physical domains) and so their diagnoses already imply compromises in functioning. Thus, not all diseases impair functioning, which may be why many lay people with chronic illnesses consider themselves successful, nonetheless (Strawbridge et al., Reference Strawbridge, Wallhagen and Cohen2002; Montross et al., Reference Montross2006; Romo et al., in press). In this connection, it is noteworthy that the availability of neuroimaging techniques to detect Alzheimer-related pathological changes in the brain long before symptom onset, leading to proposals, though controversial (Chiu and Brodaty, Reference Chiu and Brodaty2013), to diagnose such individuals without symptoms as having the disease (Dubois et al., Reference Dubois2010; Reference Dubois, Gauthier and Cummings2013), presents further problems with the disease-free criterion.
Also concerned with pathways to successful aging are the process models. Let us take a look again at a well-known example cited by Baltes (Reference Baltes1993) in support of his model, “When Rubinstein was asked, in a television interview, how he managed to remain such a successful pianist in his old age [italics added], he mentioned three strategies: (1) in old age he performed fewer pieces [selection], (2) he now practiced each piece more frequently [optimization], (3) he introduced more ritardandos in his playing before fast segments so that the playing speed sounded faster than it was in reality [compensation]” (p. 590). Would Baltes have offered this example had Rubinstein's performance been unsuccessful? If a process model has to be validated by outcome, then the value of the model in describing the state of successful aging would be limited in the sense that it cannot do so independent of outcome. This in no way undermines the value of SOC and other theoretical perspectives in articulating the mechanisms and pathways by which the chance of achieving successful aging may be enhanced. These mechanisms are as important as the lifestyle factors emphasized by Rowe and Kahn; they all represent behaviors and strategies that are modifiable even in later life. In the realm of cognitive aging, active and sustained participation in physical and intellectually stimulating activities in order to build up cognitive reserve (Fratiglioni and Wang, Reference Fratiglioni and Wang2007; Stern, Reference Stern2012) cannot be ignored.
Having disentangled means from outcomes, a fundamental issue remains: What constitutes success or effective functioning? Any criteria of success are value judgments, which would likely vary by culture and settings. Rowe and Kahn (Reference Rowe and Kahn1997) argued that the criteria should be relative, with the MacArthur studies mapping participants in the top, middle, and bottom tertile in the population across a number of indicators as successful, usual, and pathological agers, respectively (Berkman et al., Reference Berkman1993). Likewise, researchers have classified individuals regardless of age, leading to age being an important determinant of successful aging (Depp and Jeste, Reference Depp and Jeste2006). However, why should someone at the age of 65 be regarded as successful simply because he or she is superior to an 85-year-old? Lay people certainly would not compare themselves against younger ones when making self-assessment of successful aging. The lack of reference to age-graded criteria by Rowe and Kahn might have been an unintended consequence of their formulation (Kahn, Reference Kahn, Poon, Gueldner and Sprouse2003), stemming from their attempt to separate aging from pathology and success. However, in terms of their vision to engage the population to make the necessary behavioral changes in order to function effectively, this is counter-productive. When the bar is set too high, it is in fact adaptive to disengage; yet such individuals would be labeled as unsuccessful. Guidelines and expectations must be age-appropriate before older adults would identify with them and use them to guide development. The differences between process and outcome models would also be partly reconciled as the former is predicated on the assumption that goals selected must be realistic.
At the same time, it is noteworthy that the meaning of success to individuals goes beyond performance against some reference points, but is also based on individuals’ assessments of how they have lived their lives, such as the choices made, under different circumstances. This may be why themes related to adaptation and life satisfaction (including absence of regrets) surface so frequently when older adults are invited to talk about what they think of successful aging (von Faber et al., Reference von Faber2001; Phelan et al., Reference Phelan, Anderson, LaCroix and Larson2004; Reichstadt et al., Reference Reichstadt, Depp, Palinkas, Folsom and Jeste2007). In other words, researchers take an interindividual approach whereas lay people may adopt an intraindividual (and life review) approach. Therefore, to challenge researcher-defined criteria using lay people's subjective criteria may not be entirely appropriate due to the different frames of reference involved. Nevertheless, it does not mean that the views of lay people are irrelevant because their identification with expert-defined goals may be crucial for participation in behavioral change programs.
A revised model and concluding comments
The revised model is one in which high physical, cognitive, psychological, and social functioning jointly define successful aging, measured with reference to age-graded norms. All four dimensions are considered equally as building blocks of successful aging (i.e. without any hierarchical organization as in Rowe and Kahn). A substantial body of literature has demonstrated that these four factors interrelate in complex ways (Depp et al., Reference Depp, Vahia and Jeste2010) and there is no strong reason to treat any dimension as more crucial than another. Disability and physical functioning are closely related and researchers often discuss them together (Depp and Jeste, Reference Depp and Jeste2006); hence, they are grouped into one dimension. Because of the focus on functioning rather than disease, a separate definition of successful aging for people with chronic diseases (Young et al., Reference Young, Frick and Phelan2009) may not be necessary. Moreover, social and productive engagements are renamed as “social functioning” for convenience. This multidimensional approach to defining high functioning is consistent with World Health Organization's (2002) concept of holistic health.
Additional comments on the dimension of psychological functioning are warranted as it was not included in Rowe and Kahn's model. Although there were psychologists and psychiatrists serving as core members on the MacArthur studies team, the main purpose appeared to be locating psychosocial factors predicting physical and cognitive functioning, rather than looking at psychological functioning per se as an outcome (Rowe and Kahn, Reference Rowe and Kahn1998). This has led to a relative neglect of psychological well-being in studies on successful aging (Depp and Jeste, Reference Depp and Jeste2006). Note, however, that Havighurst (Reference Havighurst1961) had long ago remarked on the importance of life satisfaction for successful aging, which, together with emotional well-being, are endorsed by the majority of older people as markers of aging well (von Faber et al., Reference von Faber2001; Phelan et al., Reference Phelan, Anderson, LaCroix and Larson2004). Recently, Jeste et al. (Reference Jeste, Depp and Vahia2010) emphasized lack of depression as an indicator of successful emotional aging. A distinction has to be made between clinical depression, the mental illness, and depressive mood as an indicator of psychological functioning in the context of the present model. This distinction is not easily made. As mentioned before, as a mental illness, depression necessarily implies compromises in functioning. Using depression as a criterion also tends to bias against women as they report more depressive symptoms than men (Smith and Baltes, Reference Smith and Baltes1998). For instance, a recent study in Taiwan (Hsu and Jones, Reference Hsu and Jones2012) included depression, along with chronic diseases, physical function difficulties, productive activities, and so on, as criterion and found that 72% of those with successful aging trajectories were men. Aside from the possibility of differential attrition due to mortality over a 14-year interval, leaving healthier men being retained in the sample, preexisting gender difference in depression might have further separated women from men in the identification of latent classes (i.e. groups). In contrast to depression, other indicators of psychological functioning, such as life satisfaction and positive/negative affect, show little or no gender differences (Pinquart and Sörensen, Reference Pinquart and Sörensen2001; Cheng et al., Reference Cheng, Li, Leung and Chan2011). Researchers need to be careful in the selection of measures to avoid biased estimation of gender difference in successful aging.
As mentioned before, the four dimensions do not have similar trajectories with age. Age-related declines in physical and cognitive functioning are evident but social and psychological functioning may change little until very old age. There are virtually universal indicators for each of these dimensions (though cultural adaptation of different measures may be needed). For some of the indicators, such as activities of daily living, Mini-Mental State Examination, and life satisfaction, population data are often available for the construction of age-graded norms. In keeping with Rowe and Kahn's (1997) relativistic approach, successful aging represents those functioning at relatively high levels, which may be empirically established through research. By including measures of psychological well-being like life satisfaction, adopting age-graded expectations, and removing diseases and risk factors from the criteria, the gap between expert and lay definitions of successful aging may be significantly reduced and existing differences between theoretical perspectives may be somewhat reconciled. While this paper has focused on the conceptual issues, let us not forget that the promotion of actual behavioral and lifestyle changes (Lam and Cheng, Reference Lam and Cheng2013) representing the means to successful aging is a more pressing issue in the face of global aging.
Conflict of interest
None.
