Interventions Promoting Physical Activity Among Older Adults: A Systematic Review and Meta-Analysis

A Systematic Review and Meta-Analysis

Guilherme D. Grande, MSc; Crystian B. Oliveira, MSc; Priscila K. Morelhão, MSc; Catherine Sherrington, PhD; Anne Tiedemann, PhD; Rafael Z. Pinto, PhD; Marcia R. Franco, PhD


Gerontologist. 2020;60(8):e583-e599. 

In This Article

Abstract and Introduction


Background and Objectives: Frequent participation in physical activity (PA) has benefits across the lifespan but is particularly important for older adults. PA levels are either measured by objective or self-reported survey methods. Objective PA measurement is used to increase accuracy. This systematic review investigated the effect of physical activity-based interventions on objectively measured PA levels among community-dwelling adults aged 60 years and older.

Research Design and Methods: Literature searches were conducted in five electronic databases and four clinical trial registries. Randomized controlled trials investigating the effect of physical activity–based interventions on objectively measured PA levels (e.g., accelerometers or pedometers) in community-dwelling adults aged 60 years and older compared with no/minimal intervention were considered eligible. Data were pooled using the most conservative estimates reported from each study using the standardized mean difference (SMD). Grading of Recommendations Assessment, Development, and Evaluation (GRADE) was used to evaluate the overall quality of the evidence.

Results: Fourteen published trials and 3 ongoing trials were identified. There were significant effects favoring physical activity–based interventions compared with minimal intervention at short-term (less than or equal to 3 months) (SMD: 0.30, 95% CI: 0.17 to 0.43) and intermediate-term (more than 3 months and less than 12 months; SMD: 0.27, 95% CI: 0.06 to 0.49) follow-ups. The quality of evidence was moderate according to GRADE (downgraded for risk of bias).

Discussion and Implications: Our findings suggest that physical activity–based interventions may increase objectively measured PA levels in community-dwelling older adults. Further studies are still needed to identify the optimal dose, intensity, and mode of delivery of physical activity–based interventions.


The population is rapidly aging worldwide. The older population aged 60 years or older was estimated to be 962 million in 2017 which is more than double the population compared with 1980. Similarly, this number is expected to double again by 2050 reaching nearly 2.1 billion of older adults worldwide (Nations, U., 2007; Nations, U., 2017). The prevalence of physical inactivity also increases substantially with increasing age. According to the World Health Organization, physical activity (PA) is defined as any bodily movement produced by a contraction of skeletal muscles and that increases energy expenditure. The recommended PA levels for older adults (≥65 years old) are similar to adults (from 18 to 64 years old), that is, 150 min of moderate PA intensity per week, 75 min of vigorous PA intensity per week, or a combination of moderate and vigorous PA intensity plus strength training twice per week (WHO, 2010). Around 45% of people aged over 60 do not meet the recommended PA level (Hallal et al., 2012). For those aged 75 and older, the proportion of people not meeting the recommended PA levels increases to nearly 75% (Australian Institute of Health and Welfare, 2014). The inability to increase PA, despite being willing to do so, is common among community-living older people who have mobility problems and who report negative environmental features in their neighborhood, such as lack of resting places and distances perceived to be too long, noisy traffic, dangerous crossroads, and streets in poor condition. Mobility promotes healthy aging as it relates to the basic human need for physical movement (Rantakokko et al., 2010).

The growing older population brings challenges to the capacity of public health systems and governments in delivering high quality health services as the risk of chronic disease onset and disability rises in older age (Prince et al., 2015). Conservatively estimated, physical inactivity cost healthcare systems international $ (INT$) 53.8 billion worldwide in 2013, of which $31.2 billion was paid by the public sector, US$12.9 billion by the private sector, and $9.7 billion by households. In addition, physical inactivity–related deaths contribute to $13.7 billion in productivity losses, and physical inactivity was responsible for 13.4 million disability-adjusted life-years (DALYs) worldwide (Ding et al., 2016). Compelling evidence shows that PA can provide primary and secondary prevention of chronic disease (Warburton, Nicol, & Bredin, 2006), prolong years of active life (Clark et al., 2012), reduce the risk of early mortality (Löllgen, Bockenhoff, & Knapp, 2009), reduce the risk of falls (Gillespie et al., 2012) and improve functional performance and quality of life among older adults (Sun, Norman, & While, 2013). Despite the wide ranging benefits of regular PA, participation levels particularly among older adults are still low (Matthews et al., 2012; Nelson et al., 2007).

PA levels are commonly measured by self-reported and objective methods, with the self-reported measures including mainly self-reported questionnaires and diaries such as Physical Activity Scale for the Elderly (PASE), Physical Activity Questionnaire for Older Adults, Yale Physical Activity Survey (YPAS), and Incidental and Planned Activity Questionnaire (IPEQ). Although these questionnaires have some evidence of validity and reliability (Delbaere, Hauer, & Lord, 2010; Moore et al., 2008; Silva et al., 2019; Washburn, McAuley, Katula, Mihalko, & Boileau, 1999), their use among the older population is challenging due to changes in cognitive abilities and in recall bias, especially when considering recall over long periods of time (Kowalski, Rhodes, Naylor, Tuokko, & MacDonald, 2012; Shephard & Vuillemin, 2003). In addition, aging and disability modify the metabolic cost of activities, so standard tables and equations used to determine the energy expenditure of common activities that have been developed in younger populations may be inaccurate for use with older adults (Kowalski et al., 2012; Rikli, 2000).

Since self-reported methods have some limitations, objective measures of PA using technology to measure and record in real-time biomechanical and/or physiological consequences of physical activities are commonly used to increase accuracy. Objective measures provide more accurate estimates of energy expenditure and eliminate response biases. These consist of, for example, motion sensing and monitoring devices (accelerometers, pedometers, and heart-rate monitors), physiological markers (cardiorespiratory fitness and biomarkers), and calorimetry (Prince et al., 2008).

Randomized clinical trials (RCTs) and systematic reviews investigating PA-based interventions have identified positive results for prevention of falls (Sherrington et al., 2019), reduction of cognitive decline (Olanrewaju, Kelly, Cowan, Brayne, & Lafortune, 2016), and improvement of balance (Howe, Rochester, Neil, Skelton, & Ballinger, 2011), among others. However, to date, there is limited systematic review evidence of the effect of interventions for increasing objectively measured PA among the older population. Previous systematic reviews (Chase, 2015; Conn, Valentine, & Cooper, 2002; Sun et al., 2013) found a small effect favoring physical activity–based intervention over the control intervention. However, given that these reviews included studies that used objective and self-reported PA measures, there is less clarity about the effect of physical activity–based interventions on objective PA measures alone. In addition, previous reviews have summarized the evidence from different study designs and not RCTs alone. We would argue that, when available, RCTs should be used as this is the most robust research design for assessment of the relative effects of intervention (Chandler et al., 2019). Other methodological flaws in previous systematic reviews include high heterogeneity in the meta-analysis (Chase, 2015), search restriction with regard to the date of publication (Sun et al., 2013), and the lack of assessment of risk of bias and overall quality of evidence (Chase, 2015; Conn et al., 2002; Sun et al., 2013). Therefore, the primary objective of this systematic review was to investigate the effect of physical activity–based interventions on objectively measured PA levels of older adults. Secondly, we also aimed to investigate the effect of physical activity–based interventions on mobility in this population.