Successful aging defined by health-related quality of life and its determinants in community-dwelling elders

The proportion of the elderly population aged ≥65 years has continued to dramatically increase worldwide from 8% in 1950 to 11% in 2009 [1]. A similar tendency has occurred in Taiwan, as the number of elderly people increased by approximately 2% each year throughout the 1990s. The proportion of elderly Taiwanese is projected to further increase almost threefold from 10% in 2009 to 36% by 2050. Because the aging process is associated with increased susceptibility to chronic conditions, disabilities, psychosocial problems and comorbidities [2, 3], it is important that this aging population is able to enter late stages of life in relatively good health.

An extensive review of 28 quantitative studies reported that there are several methods to operationally define and measure successful aging [4]. The prevalence and definitions of successful aging were found to vary among studies [4]. As the number of included domains increased, definitions of successful aging tended to become more complex [5]; there was, however, a consensus regarding the multidimensional nature of successful aging [4‐6]. The Study of Health Assessment and Risk in Ethnic groups also reported a cross-national variation in the prevalence of successful aging among 14 European countries. The authors suggested that self-reported health measures may be the most appropriate measures for comparing cross-national differences to avoid diagnostic differences across countries [7]. Multidimensional, standardized, and self-rated instruments are more practical and are more easily comparable among different countries and populations.

The Short Form-36 (SF-36) is a multidimensional scale and a well-established instrument that measures health concepts and self-reported health-related quality of life [8]. Using SF-36 to define successful aging is acceptable because it represents overall health status in two key measures—the physical component summary (PCS) and the mental component summary (MCS) scores. These two scores reflect the status of physical, mental, and social well-being that are generally used to define health status [9]. The aim of this study was to explore the prevalence of successful aging, as gauged by SF-36, as well as to identify determinants of successful aging among community-dwelling elders.

Of the 903 elders (mean age, 73.90 years) enrolled in this study, 10.4% were identified as successful agers. Successfully aging elders had significantly higher scores in eight SF-36 dimensions than those who were not successfully aging; they also had the most favorable health status (with scores of 100) in the RP, SF, and RE variables (Table 1).

Older elders, females, and elders without enough or with just enough money for personal use had a lower prevalence of successful aging. The unadjusted analysis showed a higher prevalence of successful aging among elders with ≥7 years of education, with general or good visual capacity, with good hearing capacity, who regularly exercised, and who were able to see relatives and friends whenever they wished (Table 2).

Among the associations with chronic health problems, increased odds of successful aging were found in non-frail elders and in those without hypertension, heart disease, hyperuricemia, arthritis, cataracts, fall histories, pain problems, or sleep disorders (Table 3).

Nested logistic regression analysis was used to further examine the interrelatedness of factors significantly associated with successful aging. Among socioeconomic and psychosocial support factors, the odds of successful aging were significantly lower for the very elderly and for those with insufficient income. After accounting for the effects of visual capacity, hearing capacity, and exercise habits on successful aging, the significance of age remained unchanged while the strength of the financial resources association diminished. Meanwhile, we observed increased odds of successful aging for elders with general or good visual capacity and for those who exercised regularly. After chronic illness and history of falling were added into the regression model, successfulness of aging was statistically associated with an age range of 71–75 years; regular exercise; and the absence of heart disease, arthritis, and fall histories.

The associations of regular exercise and chronic diseases with successful aging disappeared, however, after adding the factors of pain, sleep disorders and frailty in the fourth model. In the final model, the odds of successful aging decreased in elders aged 71–75 years relative to those ≤70 years (adjusted odds ratio [OR]: 0.27; 95% confidence interval [CI]: 0.13–0.58) and in those in pre-frail states. An increase in the odds of successful aging was observed in elders who were able to visit relatives and friends whenever they wished (OR: 3.86; 95% CI: 1.09–13.61) and in those without a history of falling (OR: 4.95; 95% CI: 1.79–13.74), pain (OR: 4.04; 95% CI: 2.18–7.50), or sleep disorders (OR: 2.36; 95% CI: 1.30–4.27) (Table 4).

The sensitivity analysis results showed that age, fall history, and pain were still significantly associated with successful aging when using the highest cutoff value at the seventieth percentile for both the PCS and the MCS. Successful aging associations with sleep disorders and visiting relatives and friends were strong but insignificant. Using the highest quartile as a cut-off value, only fall history and pain were significantly associated with successful aging. All elders who were frail were found to be unsuccessfully aging; however, the factor of frailty could not be added to the final multivariate model due to problems with statistical convergence.

The prevalence of successful aging, as was defined using SF-36, was 10.4% in community-dwelling elders in a metropolitan population of Taiwan. Elders with chronic diseases or other health-related problems had a lower prevalence of successful aging compared with those without such health issues. All elderly people who had had a stroke, for example, were considered to be unsuccessfully aging. Elders more likely to be successfully aging, however, were ≤70 years old; were are able to visit relatives or friends whenever they wished; and had no history of falling, pain, or sleep disorders.

We adopted the five components of frailty defined by Fried et al. [11] that included objective, performance-based measures. A previous study had reported that frailty in elders is an important clinical state that is distinct from normal aging and a strong predictor for disability [11]. In this study, frailty was found to be an independent determinant of successful aging. Among frail elders, only one (0.93%) was found to be successfully aging; and even after PCS and MCS cutoff values were increased from tertile to a higher percentile, no other frail elders were found to be successfully aging. These results indicated that gauging successful aging using SF-36 is not only easy but also reflects actual health conditions. A noteworthy additional study by Theou et al., however, reported that frailty in elders can be reversed by structured exercise training [13].

The prevalence of successful aging among previous community-based studies ranges from 0.4% to 95% based on various definitions [5, 6, 12, 14‐20]. In general, those studies using a single-item, self-rating scale reported a higher and more variable prevalence of successful aging, ranging from 50.3% to 90.2% [6, 15, 18]. The focus of many studies on successful aging has shifted from disease status and functional decline to multidimensional health status, which encompasses physical, functional, psychological, and social health [21]. Early published reports defined successful aging using only one or two dimensions of health, such as the absence of chronic diseases, longer longevity, independent physical functioning, social life engagement, and mental health [14, 16, 17, 19]. However, more recent studies have focused on the concept that successful aging is more multifaceted and complex [5, 7]. Studies using the absence of major disease as one of the domains to define successful aging in elders reported a lower prevalence of successful aging [15, 18, 20] compared with those not using this domain [5, 7, 12, 15]. To obtain a comparable prevalence of successful aging among elders and to explore the associated determinants, there is a need for a simple, standardized and multidimensional tool to be used across studies. In a study conducted in Hong Kong, higher PCS and MCS scores in SF-36 reflected better overall physical, functional, psychological, and social health in Chinese elders [22]. Consequently, higher PCS and MCS scores in SF-36 among elders were used to define successful aging in this study.

Age is a commonly identified determinant of successful aging. Consistent with the results reported in previous studies [5, 23, 24], younger age was one determinant of successful aging among elders in this study. We also found that the ability to see relatives and friends at will (as a variable related to social support) was an independent determinant for successful aging among the elderly. Pruchno et al. [23] reported that unsuccessfully aging elders have less social support compared with successful agers. Our findings on the associations among sleep disorders, fall histories, and successful aging were similar to those in prior research [25]. Elderly persons with sleep disorders were at a higher risk of unsuccessfully aging compared with those without sleep disorders. This finding provides evidence supporting a previous study by Andrews et al. [25] that defined the concept of high functioning and found that elders without sufficient sleep were less likely to be in the high functioning category. In our findings, a history of falling was associated with a lower likelihood of successful aging, as those with such a history tend to restrict outdoor activity out of a fear of falling [26]. In general, the fear of falling has tended to impede elderly participation in healthy activities and leads to decreased mobility, resulting in diminished quality of life [27, 28].

Previous studies have reported that chronic diseases—such as arthritis, diabetes, stroke, and hypertension—were all associated with successful aging. [5, 16] We found that these diseases were not independent risk factors of unsuccessful aging because their effects were explained by falls, frailty, pain, and sleep problems. Among these chronic diseases, stroke had a major impact on successful aging; all elders who had a history of stroke in this study were found to be unsuccessfully aging. Strokes were not, however, added to the final multivariate model due to problems with statistical convergence. We observed lower physical component scores in elders with stroke histories compared with those without stroke histories; no significant differences were observed in mental component scores. Consistent with the findings of Okonkwo et al. [29], the difference in PCS scores between subjects with and without a stroke history in this study was approximately twofold that in MCS scores. In contrast to our findings on stroke effects, Haley et al. [30] indicated that an incident of stroke has a negative impact in both mental and physical health over a three-year period. One possible explanation for these different findings is that the elders in our present study had already coped with their conditions, allowing for better mental health.

In present study, we used Cohen’s d in the chi-square value formula (d = 4 x 2 / N − x 2, where N is total sample size) [31] to calculate the effect size of significant factors. Our results indicated a mid-size effect of both pain (d = 0.47) and frailty (d = 0.38). The effect sizes of other factors (including age, seeing relatives and friends at will, fall histories, and sleep disorders) were small, between 0.2 and 0.3. The two determinants of pain problems and frailty were found to have greater relative association with successful aging.

The present study has certain limitations that should be taken into account when interpreting the results. First, owing to the nature of the cross-sectional study, we cannot explore the possible causal relationships between the health conditions considered in this study and successful aging. Second, the findings of our work are not generalizable to rural elders because this study sample was a group of metropolitan elders. Finally, the prevalence of successful aging in urban elders may have been overestimated due to the exclusion of elders who were diagnosed with dementia, had cognitive impairment, and were unable complete the SF-36 questionnaire. Our study’s elders reflected the general global measures of physical and mental dysfunction, rather than those of specific disabilities or impairments [32].

The following conclusions can be drawn from the present study: the higher proportion of successful aging in elders was found among those who were relatively younger, had social support, not frail, and had no pain, sleep impairment, and fall history. These findings revealed the importance of considering suitable intervention to improve the frailty, pain, and sleep problems among the elders. However, the causal relationships between these factors and successful aging will require further investigations.