Association between physical activity and quality of life in Japanese adults aged 85 to 89 years: a cross-sectional study | BMC Geriatrics

This study examined the association between physical activity and QOL among people aged 85–89 in Japan. Our study suggests that PAI is associated with better HRQOL in this very old population. In particular, time spent on MVPA was positively associated with better HRQOL. However, no association was found between the PAI and psychological well-being in this population. Furthermore, more than half of the participants engaged in some type of sport or exercise as a leisure activity for their health. Compared with the less active participants, the more active participants performed several kinds of sports or exercises in addition to calisthenics and had a longer walking duration per week.

In Japan, the population of adults aged 85 years and older was 6.16 million in 2020 and is estimated to exceed 10 million by 2030 [23]. Approximately 58% of the population require nursing care [24]. The study participants were older adults aged 85–89 years without limitations in basic activities of daily living; thus, the study population was healthier than the general Japanese population aged 85–89 years. Indeed, the median EQ-5D index score in this study (0.895) exceeded the Japanese population norms in the 80–89-year age group (0.821 in male and 0.774 in female) [25], which is consistent with our previous findings in a smaller sample [16]. To investigate the factors of healthy aging, descriptive analyses of healthier older adults are important. The primary objective of KAWP is to explore trajectories of functional decline, frailty, and cognitive impairment and identify factors that delay or modify this deteriorating process, and substantial evidence has been reported from KAWP [15, 16, 26,27,28,29]. Our preliminary report documented that drug burden exerted a significant negative association with EQ-5D score [16]. Further investigations are warranted to enrich the evidence regarding the methods to promote healthy aging.

The results of our study conducted among older adults in their 80s were consistent with the reports of previous studies conducted among those in 60s and 70s, not only for the EQ-5D index score but also the EQ-5D dimension score (e.g. mobility, usual activities) [5, 18]. Our results suggest that physical activity is necessary to maintain better HRQOL, even in adults aged 85 years and older. In addition, the time spent on MVPA was positively associated with HRQOL, whereas the time spent on SB or LPA was not. Conflicting findings have been reported in several studies in terms of the time spent performing different intensities of physical activity. One study reported that less time spent on SB was associated with better QOL [11], while another study reported that more time spent on LPA was associated with better QOL [30]. Therefore, only the time spent performing intensive physical activity (i.e. MVPA) was associated with HRQOL, considering that the participants in this study were healthy people with higher QOL and the MVPA time exhibited substantial variability. In addition, factors other than physical activity (e.g., hobbies) also contribute to QOL in older adults [31], which may have a greater impact on QOL than time spent on SB and LPA. Owing to the potential risk of falls or injuries when older adults, especially those who are relatively inactive, perform high levels of MVPA, further investigations are warranted to determine the optimal frequency, intensity, duration, and type of physical activity in this population.

PAI and the number of comorbidities were strong predictors of EQ-5D index scores; however, with an R2 value of less than 10%, it was suggested that the magnitude of the influence of these factors on variability in HRQOL among individuals may be less significant. While it is important to assess physical activity objectively, it may also be necessary to investigate further aspects that cannot be captured by the PAI, such as the quality and context of physical activity, including activities of daily living.

In this study, no association was found between the PAI and psychological well-being in the 80s. A systematic review reported inconsistent results, indicating an intermediate-to-consistent association between physical activity and mental health or psychological well-being [5]. Physical activity has been reported to reduce anxiety in older population [32]. Another study reported that functional decline had less impact on psychological well-being of older adults in their 80s compared with older adults in their 60s and 70s [33]. According to gerotranscendence theory, very older adults shift their mindset, and consequently, factors other than physical activity may have stronger associations with psychological well-being. Psychological well-being has six core dimensions (purpose in life, autonomy, personal growth, environmental mastery, positive relationships, and self-acceptance), and physical activity is associated with life satisfaction and self-esteem [34]. In our study, PAI was associated with better answers to two questions regarding positive emotion (i.e., “I have felt active and vigorous.” and “My daily life has been filled with things that interest me.”) among the WHO-5 questions, indicating that PAI partially contributes to the psychological well-being of older adults in their 80s.

To the best of our knowledge, this is the first study to reveal the types of sports and exercises performed in this very old population. Calisthenics were the most commonly selected sports and exercises by both male and female in this population. Participants in this study performed calisthenics anywhere, including at home, in a park, and at a rehabilitation center. In Japan, various calisthenic programs are available via TV and radio, and calisthenics are easy to start anywhere, anytime, and with anyone. Thus, calisthenics may be one of the best options for people who do not perform any exercise to start some kind of exercise. The secondary- and tertiary-selected sports and exercises differed by sex; male participants preferred resistance training and ground golf, whereas female participants preferred stretching/yoga and dancing. This difference may be partially due to gender preference; it has been reported that older Japanese female prefer to participate in regular group activities compared with older Japanese male [35]. Previous studies have reported that these sports and exercises are also associated with better HRQOL in the older population [36,37,38,39]. Based on the findings of this study and those of previous studies, these types of exercises should be considered for inclusion in health promotion programs for this population. Although no significant difference was found in the proportion of participants performing calisthenics regardless of PAI, more active participants tended to perform several sports or exercises in addition to calisthenics. Walking is the most frequently performed sport/exercise among older Japanese adults [40], with more active participants walking for longer durations than the inactive participants. Furthermore, more active participants may walk longer to attend sports or exercise programs. To promote physical activity and maintain QOL even at an advanced age, it is essential to cultivate an environment and community in which individuals can easily explore sports or exercises that interest them and begin exercising in a manner suitable for this generation.

Strength and limitations

The major strengths of our study are as follows: first, this study used a large community-based cohort with an older population aged between 85 and 89 years; second, physical activity was objectively assessed using an accelerometer to minimize recall bias, overestimation, and underestimation [41] and was also subjectively assessed using a questionnaire to provide qualitative insights into specific types of physical activities. However, this study has several limitations. First, causal relationships could not be established due to the cross-sectional study design. Further investigation using follow-up data from the KAWP is warranted. Second, the study was conducted among residents of Kawasaki City who visited the study site and did not require nursing care. As the study participants represented a healthier population with a higher QOL, the generalizability of the results warrants further investigation. Third, not all variables, including unobserved ones, were fully adjusted in this study. Residual confounding factors, including sleep and diet, may have existed and were not adjusted for in this model. Sleep and diet are important in assessing the relationship between PA and QOL. However, in this study, the accelerometer was worn during waking hours, which precluded the inclusion of sleep data in the analysis. While diet is also important, this study primarily focused on the relationship between PA and QOL, and specific types of PA in this very old population. The impact of diet should be considered in future studies. Polypharmacy was not adjusted in the model; instead, it was adjusted for the number of comorbidities, as older patients received standard medical care, including drugs, under the Japanese universal care system. Fourth, although the validity of the accelerometer was confirmed in the older population, the mean errors and degree of underestimation were greater in the older population, especially for higher-intensity activities, than in the younger population [42]. Fifth, owing to the use of complete case analysis, the study findings were limited to participants with no missing data. In the complete case analysis dataset (n = 876) and the excluded individuals (n = 38), age, alcohol intake, smoking status, years of education, perceived household income level, BMI, and the number of comorbidities were similar. The percentages of female participants (73.7% vs. 49.5%) and those living alone (44.0% vs. 26.3%) were higher in the excluded group than in the group with complete case analysis dataset. This could have led to an overestimation of the association between PA and QOL, but given the small percentage of excluded individuals, we believe that the likelihood of such bias is low. The median EQ-5D index score (0.895 vs. 0.895), median WHO-5 total score (19.0 vs. 20.0), and mean PAI (21.5 METs*h/day vs. 22.2 METs*h/day) were similar among the complete case analysis dataset and the excluded groups.