Only 42% of older Australians (Australian Bureau of Statistics, 2020–2021) and 20%–60% of older adults worldwide (Sun et al., 2013) are reported to meet recommended levels of physical activity and sedentary behavior (World Health Organization, 2020). Physical activity is defined as any bodily movement produced by skeletal muscles that results in energy expenditure (Caspersen et al., 1985). For older adults, physical activity is associated with a reduction in the risk of fractures, falls, dementia, depression, cancer, and cardiovascular and all-cause mortality (Cunningham et al., 2020). Older adults who are sufficiently physically active also experience better physical and cognitive functioning, improved quality of life, and “healthier aging” (Cunningham et al., 2020; Daskalopoulou et al., 2017). Barriers to physical activity in older adults have been extensively studied and include fatigue, injury, pain, poor health, lack of time, expense, and a lack of access to, and availability of, suitable and affordable physical activity options and exercise programs (Baert et al., 2011; Booth et al., 2002; Franco et al., 2015; Rona et al., 2014).
To address program access and availability issues for older adults, Exercise and Sports Science Australia, Australia’s peak professional organization for university-trained exercise and sports science practitioners (Exercise and Sports Science Australia, 2023b), sought and received funding from the Australian Government for the Exercise Right for Active Ageing program (Australian Sports Commission, 2023; Exercise and Sports Science Australia, 2023a). The program aimed to deliver subsidized community-based group exercise classes for older adults, aged 65 years and older, across Australia. Classes were taught by accredited exercise scientists (AESs) and physiologists (AEPs) within community-based facilities, such as fitness centers and community health centers, in metropolitan, regional, and remote areas to promote widespread access (Exercise and Sports Science Australia, 2023a).
The program was intended to run throughout 2020 and 2021, with a recruitment target of 14,600 older Australians. However, with the introduction of nationwide COVID-19 lockdowns in March 2020, many Australians were unable to attend exercise classes in person. The state of Victoria, Australia’s second most populous state, was particularly impacted by COVID-19, experiencing 267 days of lockdown throughout 2020 and 2021 (Boaz, 2021). COVID-19 lockdowns not only affected participation in group exercise classes but also led to a reduction in community mobility and general physical activity in older adults during this time (Goethals et al., 2020; Gough et al., 2023; Hoffman et al., 2022; Oliveira et al., 2022; Public Health England, 2021). There is also evidence that lockdowns were associated with anxiety, depression, sleep disturbances, and loneliness in older adults (Christensen et al., 2022; Gough et al., 2023; Hoffman et al., 2022; Sepúlveda-Loyola et al., 2020).
Engaging in exercise, that is, activity that is planned, structured, and repetitive, with the objective of improving or maintaining physical fitness (Caspersen et al., 1985), contributes to overall physical activity levels (Langhammer et al., 2018). It is important to understand older adults’ levels of participation in exercise classes during the pandemic and also determine who was most affected based on age, gender, health, and geography. Understanding the contribution of participation in exercise classes to overall physical activity levels in this population is also important. Therefore, the primary aim of this study was to determine factors associated with participation in community-based exercise classes of older Australians during the COVID-19 pandemic. A secondary aim was to investigate the association between participation in exercise classes and changes in physical activity levels during this time.
Methods
Design
This quasi-experimental pre- and poststudy was reported in accordance with Transparent Reporting of Evaluations with Nonrandomized Designs guidelines (Haynes et al., 2021). Ethics approval was granted by the Monash University Human Research Ethics Committee (Project ID: 21550) and all participants provided written, informed consent. The study was registered with the Australian New Zealand Clinical Trials Registry (Registration number: ACTRN12623000483651).
Participants
A total of 215 exercise providers across all states and territories of Australia were involved in recruitment of older adult participants. Exercise providers included AESs and AEPs employed within fitness centers, community health centers, and private clinics. A nationwide marketing campaign run by Exercise and Sports Science Australia was launched to promote the Exercise Right for Active Ageing study and drive recruitment, targeting older Australians. The campaign included the establishment of a website (Exercise and Sports Science Australia, 2023a), which included a search function to find a provider, television advertisements for waiting rooms in general practices, Facebook advertising, local community activation events (fetes, shows, seniors weeks, etc.), conference presentations, and hard copy marketing materials distributed to health professionals and community organizations. Potential participants generally self-referred to the provider after exposure to marketing materials.
Community-dwelling older adults (aged 65+ years) were eligible for inclusion. Participants were excluded if they were unable to participate in a low- to moderate-intensity exercise program, which was determined prior to recruitment via completion of the Adult Pre-Exercise Screening System V2 (Stage 1; Supplementary Material [available online]; Exercise and Sports Science Australia, 2019). Individuals at a higher risk of an adverse event due to exercise were advised to seek guidance from an appropriate allied health professional or medical practitioner prior to participating in the study. Those eligible for the study were invited to participate in a free trial class prior to consenting to participate.
Intervention
The intervention consisted of 12 exercise classes, generally delivered at a frequency of one class per week. However, to allow for instructor or participant unavailability, participants were given up to a maximum of 16 weeks to complete their 12 classes. Classes were led by a university-trained AES or AEP who had completed additional project induction coursework and training modules on engaging older adults in physical activity, local community volunteering and engagement, and farmer health.
The content and delivery of the programs were at the discretion of providers but generally consisted of a group-based class of low- to moderate-intensity exercise suitable for older adults. Class types included falls prevention, strength, and general fitness classes, among others (Table 1). The sessions generally ran for 1 hr, in person, at a community-based exercise facility (e.g., fitness center), although online delivery options were also available, particularly during COVID-19 lockdowns. Online classes were delivered live by the usual provider across a range of freely available video conferencing platforms (e.g., Zoom Video Communications Inc.). Adherence was incentivized via a subsidy program. Participants paid $7.27 excluding goods and services tax (GST) ($8.00 including GST) per class, with $9.92 (excluding GST) paid to the provider via project funding.
Characteristics of Participants Included in Primary (n = 6,626) and Secondary Analysisa (n = 3,504) and Those Lost to Follow-Up (n = 3,122)
| Characteristics | Primary analysis | Secondary analysis | ||
|---|---|---|---|---|
| Included n (%) | Included n (%) | Lost to follow-up n (%) | p | |
| Age (years) | ||||
| Median (IQR) | 73 (69–77) | 72 (69–77) | 73 (69–77) | <.01** |
| Age group (years) | ||||
| 65–69 | 2,019 (30.5) | 1,090 (31.1) | 929 (29.8) | <.01** |
| 70–74 | 2,085 (31.5) | 1,139 (32.5) | 946 (30.3) | |
| 75–79 | 1,457 (22.0) | 743 (21.2) | 714 (22.9) | |
| 80–84 | 731 (11.0) | 383 (10.9) | 348 (11.2) | |
| 85+ | 334 (5.0) | 149 (4.3) | 185 (5.9) | |
| Gender | ||||
| Men | 1,512 (22.8) | 797 (22.8) | 715 (22.9) | .88 |
| Women | 5,111 (77.1) | 2,705 (77.2) | 2,406 (77.1) | |
| Nonbinary | <5 | <5 | <5 | |
| State or territoryb | ||||
| Australian Capital Territory | 124 (1.9) | 49 (1.4) | 75 (2.4) | <.001*** |
| New South Wales | 1,999 (30.2) | 968 (27.6) | 1,031 (33.1) | |
| Northern Territory | <5 | 0 (0.0) | <5 | |
| Queensland | 2,152 (32.5) | 1,076 (30.7) | 1,076 (34.6) | |
| South Australia | 348 (5.3) | 198 (5.7) | 150 (4.8) | |
| Tasmania | 208 (3.1) | 125 (3.6) | 83 (2.7) | |
| Victoria | 593 (9.0) | 289 (8.3) | 304 (9.8) | |
| Western Australia | 1,187 (17.9) | 798 (22.8) | 389 (12.5) | |
| ARIAc | ||||
| Major cities | 3,914 (59.1) | 2,181 (62.3) | 1,733 (55.6) | <.001*** |
| Inner regional | 2,149 (32.4) | 1,058 (30.2) | 1,091(35.0) | |
| Outer regional | 457 (6.9) | 216 (6.2) | 241 (7.7) | |
| Remote/very remote | 96 (1.5) | 44 (1.3) | 52 (1.7) | |
| Comorbidities | ||||
| Low (<2) | 2,561 (38.7) | 1,433 (40.9) | 1,128 (36.1) | <.001*** |
| High (2+) | 4,065 (61.4) | 2,071 (59.1) | 1,994 (63.9) | |
| Class typed | ||||
| Aerobics/cardiovascular | 182 (2.8) | 88 (2.5) | 94 (3.0) | <.001*** |
| Aqua aerobics/hydrotherapy | 106 (1.6) | 47 (1.3) | 59 (1.9) | |
| Falls prevention/balance | 765 (11.6) | 372 (10.6) | 393 (12.6) | |
| Pilates equipment/matwork | 224 (3.4) | 127 (3.6) | 97 (3.1) | |
| Yoga/flexibility/mobility | 42 (0.6) | 30 (0.9) | 12 (0.4) | |
| Strength | 463 (7.0) | 225 (6.4) | 238 (7.6) | |
| Circuit class | 842 (12.7) | 399 (11.4) | 443 (14.2) | |
| Clinical program/condition specific | 601 (9.1) | 377 (10.8) | 224 (7.2) | |
| Functional fitness | 244 (3.7) | 138 (3.9) | 106 (3.4) | |
| General fitness | 775 (11.7) | 431 (12.3) | 344 (11.0) | |
| “Group class” | 1,398 (21.1) | 746 (21.3) | 652 (20.9) | |
| Gym-based program | 799 (12.1) | 446 (12.7) | 353 (11.3) | |
| Othere | 184 (2.8) | 78 (2.2) | 106 (3.4) | |
| Attended free trial classf | ||||
| No | 3,357 (50.7) | 1,741 (53.0) | 1,616 (54.3) | .30 |
| Yes | 2,901 (43.8) | 1,543 (47.0) | 1,358 (45.7) | |
| Class deliveryg | ||||
| In person | 6,451 (97.4) | 3, 416 (97.5) | 3,035 (97.2) | .54 |
| Online | 174 (2.6) | 88 (2.5) | 86 (2.8) | |
| Total | 6,626 | 3,504 (52.9) | 3,122 (47.1) | |
Note. ARIA = Accessibility/Remoteness Index of Australia; IQR = interquartile range.
aFollow-up data collected for at least one physical activity outcome. bMissing data, n = 11. cMissing data, n = 10. dMissing data, n = 1. e“Other” includes walking groups, Tai-Chi, chair-based, small equipment, bushwalking, and low-impact classes. fMissing data, n = 368. gMissing data, n = 1.
*p < .05. **p < .01. ***p < .001.
Measurements
At the initial assessment, demographic details were collected by providers, including date of birth, gender, postcode of residence, and self-reported comorbidities (yes/no), including arthritis, asthma, cancer, dementia, depression, diabetes, heart disease, hypertension, osteoporosis, and prostate issues (males only; Supplementary Material [available online]). Postcode of residence was mapped to the Accessibility/Remoteness Index of Australia (a geographical index of remoteness). Total reported comorbidities were categorized as <2 or ≥2 for analysis.
The primary outcome was the number of classes attended out of a maximum of 12. Participants’ attendance at each class (yes/no) was recorded by providers as soon as possible after each class via the electronic project portal. Physical activity levels were self-reported using a questionnaire (Supplementary Material [available online]) adapted from the International Physical Activity Questionnaire Short Form, the most widely used self-reported measure of physical activity and sitting time (Lee et al., 2011). Although the agreement between the International Physical Activity Questionnaire Short Form and device-based measurement is generally reported as moderate to poor, it demonstrates sufficient test–retest reliability in older adults, making it suitable for pre- and poststudies (Lee et al., 2011). The adapted questionnaire consisted of five questions relating to the past week, assessing (a) minutes spent walking continuously for 10 min or more and (b) participation in moderate- and (c) vigorous-intensity physical activity (MPA and VPA, respectively), (d) the total number of active days (i.e., spent walking or performing MPA or VPA), and (e) average daily sitting time. Consistent with the wording of the questions and International Physical Activity Questionnaire scoring protocols, walking time <10 min was recoded as zero. Sitting time was truncated at a maximum of 16 hr per day to reflect sitting time during waking hours. Weekly minutes of walking, MPA, and VPA were converted into metabolic equivalents (METs), an indicator of the energy expenditure required for each activity, by multiplying the time spent performing a particular activity by a predetermined weighting factor (Walking: 3.3 METs, MPA: 4.0 METs, VPA: 8.0 METs) (Ainsworth et al., 2011). Weekly MET minutes of walking, MPA, and VPA were summed to produce total weekly MET minutes.
The physical activity questionnaire was administered at the initial and postintervention assessments. Scores were uploaded by the provider to the project portal immediately after testing and then locked for editing. When performing postintervention assessments, providers and participants were not blinded to preintervention scores. Participants in major cities paid $16.81 excluding GST ($18.50 including GST) for the initial assessment, with the remaining grant subsidy of $22.68 (excluding GST) paid to the provider via project funding. To incentivize study participation in rural and remote areas, participants in these areas paid only $5.00 excluding GST ($5.50 including GST) for the initial assessment with the remaining $34.50 (excluding GST) paid to the provider via project funding. Posttest assessments were free for participants, with a grant subsidy of $60.00 (excluding GST) paid to providers. Participants did not receive any financial support for traveling to or from classes or assessments.
Deviations From Protocol and Other Complications
Due to COVID-19 lockdowns preventing attendance at exercise classes, recruitment and program delivery were suspended in several states and territories for certain periods throughout 2020 and 2021. Recruitment in the states of Victoria and New South Wales (NSW) was particularly affected, and although some providers converted to online delivery, most classes were suspended until lockdowns were lifted. As shown in Figure 1, there was a reduced rate of recruitment following the start of the initial 6-week nationwide lockdown on March 23, 2020, and again from July to October 2021, when the majority of states entered lockdowns of varying duration.
—Recruitment plot (n = 6,949).
Citation: Journal of Aging and Physical Activity 32, 3; 10.1123/japa.2023-0199
Statistical Analysis
For the primary analysis (class participation), all participants were included (including those lost to follow-up) provided that they attended at least one class. Characteristics of these participants were summarized using counts and percentages for categorical variables and medians and interquartile ranges (for continuous characteristics with skewed outcomes). For secondary analyses (change in physical activity levels), only participants with at least partial follow-up (i.e., predata and postdata available for at least one physical activity measure) were included. Characteristics of these participants were compared with those lost to follow-up using Chi-square or Wilcoxon rank-sum analyses.
To determine factors associated with the primary outcome (class participation), negative binomial regression (with a robust variance estimator) was used to account for potential overdispersion (where the variance is greater than mean), with the outcome as the number of classes attended and the exposure as the total number of classes offered (n = 12). Incidence rate ratios (IRRs) and coefficients were estimated for a range of potential explanatory variables, including age group, gender, comorbidity, state of residence, Accessibility/Remoteness Index of Australia classification of residence, attendance at free trial class (yes/no), class type, and online delivery (yes/no). Covariates showing a significant (p < .2) association on univariable analyses were entered into each multivariable model (Hosmer et al., 2013). Multicollinearity between covariates was investigated using variance inflation factor—with a threshold for inclusion of <5 (Sheather, 2009). In alignment with the general guideline for sample size in relation to independent variables in multivariable regressions (50 + 8[independent variables]), our sample exceeded the minimum number of respondents required for the primary analysis (50 + 8 × 8 = 114) (Tabachnick & Fidell, 2014).
For the secondary analysis, three physical activity outcomes were investigated: (a) total weekly MET minutes (based on minutes of walking, MPA, and VPA), (b) active days per week (out of seven), and (c) average daily sitting time (in hours). To assess the change from pretest to posttest, these outcomes were modeled using mixed-effects linear regression with patient as a random effect. To assess the change in outcomes according to class attendance, an interaction term between the number of classes attended (out of 12) and time (i.e., pretest and posttest) was included in linear mixed models, with the adjusted mean change representing the improvement in outcome for every additional class attended. For all analyses, residual plots were inspected to evaluate model assumptions (i.e., normal distribution of residuals and equal variances). All analyses were performed using Stata (version 17) with a significance level of .05.
Results
Out of 7,104 participants who were initially registered for participation in an Exercise Right for Active Ageing exercise program, 6,949 met the inclusion criteria for this study and participated in a pretest assessment (Figure 2). Of the 6,949 participants recruited (from August 2019 to June 2022), 6,626 attended at least one class and were included in the primary analysis (95%), and 3,504 completed or partially completed follow-up testing and were included in secondary analyses (53%).
—Flowchart of participant inclusion and follow-up.
Citation: Journal of Aging and Physical Activity 32, 3; 10.1123/japa.2023-0199
Of participants included in the primary analysis (n = 6,626), the majority were women (77%), were aged between 65 and 74 years of age (62%), resided in major cities (59%), and reported two or more comorbidities (61%; Table 1). The highest number of participants was from Queensland and NSW.
Type of class was most commonly described as a “group class” (21% of participants; although it is important to note that all classes were delivered in group settings), followed by circuit classes (13%). Only 3% of classes were delivered online, with the remainder delivered in person.
For the secondary analysis, a total of 3,122 participants (47%) were lost to follow-up, and they differed significantly to included participants (Table 1). Notably, participants lost to follow-up were slightly older, came from the Australian Capital Territory, NSW, Queensland, or Victoria, were residing in inner/outer regional or remote areas, and had more comorbidities.
Almost half (49%) of participants attended the full 12 classes offered (Figure 3). In univariable negative binomial regression models, six out of eight variables were associated with class attendance (state, Accessibility/Remoteness Index of Australia classification, comorbidity, class type, attendance at a free trial class, and attendance at online classes), and these were retained in the final multivariable model (Table 2). Relative to the reference category (Western Australia), there was a significantly lower rate of attendance in all states except South Australia and Tasmania, with the lowest rates of attendance in the Northern Territory (64% lower; adjusted IRR [95% confidence interval (CI)]: 0.36 [0.21, 0.64]), the Australian Capital Territory (24% lower; adjusted IRR [95% CI]: 0.76 [0.70, 0.83]), and Victoria (18% lower; adjusted IRR [95% CI]: 0.82 [0.78, 0.86]). Relative to major cities, the attendance rate was 5% lower (adjusted IRR [95% CI]: 0.95 [0.93, 0.98]) in inner regional areas. Relative to participants with fewer than two comorbidities, individuals with two or more comorbidities had a 3% lower rate of attendance (adjusted IRR [95% CI]: 0.97 [0.95, 0.99]). Relative to yoga/flexibility/mobility classes, there was a lower rate of attendance for strength, circuit, functional fitness, and “other” types of classes. Participants who attended a free trial class had a 5% higher rate of attendance (adjusted IRR [95% CI]: 1.05 [1.03, 1.08]) than those who did not, and participants attending online classes had a 19% higher rate of attendance (adjusted IRR [95% CI]: 1.19 [1.11, 1.26]) than those attending in-person classes.
—Distribution of number of classes attended (n = 6,626 participants).
Citation: Journal of Aging and Physical Activity 32, 3; 10.1123/japa.2023-0199
Factors Associated With Class Attendance (n = 6,626)
| Characteristics | Median (IQR) attendance | Unadjusted IRR [95% CI] | p | Adjusted IRRa [95% CI] | p |
|---|---|---|---|---|---|
| Age group (years) | |||||
| 65–69 | 11 (7–12) | 1.0 (ref.) | |||
| 70–74 | 11 (7–12) | 1.00 [0.97, 1.03] | .98 | ||
| 75–79 | 11 (7–12) | 0.97 [0.95, 1.00] | .08 | ||
| 80–84 | 12 (7–12) | 0.99 [0.96, 1.03] | .72 | ||
| 85+ | 11 (6–12) | 0.96 [0.91, 1.00] | .08 | ||
| Gender | |||||
| Men | 12 (7–12) | 1.0 (ref.) | |||
| Women | 11 (7–12) | 1.00 [0.97, 1.02] | .74 | ||
| Nonbinary | 12 (2–12) | 0.93 [0.57, 1.53] | .79 | ||
| State or territoryb | |||||
| Western Australia | 12 (10–12) | 1.0 (ref.) | 1.0 (ref.) | ||
| Australian Capital Territory | 9 (4–12) | 0.75 [0.69, 0.82] | <.001*** | 0.76 [0.70, 0.83] | <.001*** |
| New South Wales | 11 (6–12) | 0.88 [0.86, 0.91] | <.001*** | 0.89 [0.86, 0.92] | <.001*** |
| Northern Territory | 1 (1–7) | 0.37 [0.21, 0.64] | .001** | 0.36 [0.21, 0.64] | <.001*** |
| Queensland | 11 (7–12) | 0.89 [0.87, 0.92] | <.001*** | 0.89 [0.86, 0.92] | <.001*** |
| South Australia | 11 (8–12) | 0.93 [0.88, 0.97] | <.01** | 0.94 [0.88, 1.00] | .06 |
| Tasmania | 12 (8–12) | 0.94 [0.89, 1.00] | .06 | 0.96 [0.90, 1.03] | .24 |
| Victoria | 10 (5–12) | 0.82 [0.79, 0.86] | <.001*** | 0.82 [0.78, 0.86] | <.001*** |
| ARIAc | |||||
| Major cities | 12 (8–12) | 1.0 (ref.) | 1.0 (ref.) | ||
| Inner regional | 11 (6–12) | 0.93 [0.91, 0.95] | <.001*** | 0.95 [0.93, 0.98] | <.001*** |
| Outer regional | 11 (6–12) | 0.95 [0.91, 0.99] | .01* | 0.98 [0.94, 1.02] | .36 |
| Remote/very remote | 11 (8–12) | 1.02 [0.93, 1.11] | .71 | 0.99 [0.89, 1.10] | .87 |
| Comorbidities | |||||
| Low (<2) | 12 (8–12) | 1.0 (ref.) | 1.0 (ref.) | ||
| High (2+) | 11 (7–12) | 0.96 [0.94, 0.98] | <.001*** | 0.97 [0.95, 0.99] | .01* |
| Class typed | |||||
| Yoga/flexibility/mobility | 12 (9–12) | 1.0 (ref.) | 1.0 (ref.) | ||
| Aerobics/cardiovascular | 11 (7–12) | 0.86 [0.75, 0.99] | .03* | 0.87 [0.75, 1.01] | .07 |
| Aqua aerobics/hydrotherapy | 11 (7–12) | 0.88 [0.76, 1.02] | .09 | 0.89 [0.77, 1.04] | .14 |
| Falls prevention/balance | 11 (7–12) | 0.87 [0.77, 0.99] | .04* | 0.89 [0.78, 1.01] | .08 |
| Pilates equipment/matwork | 12 (9–12) | 0.95 [0.83, 1.08] | .42 | 0.89 [0.77, 1.03] | .12 |
| Strength | 11 (6–12) | 0.85 [0.75, 0.97] | .02* | 0.86 [0.75, 0.98] | .03* |
| Circuit class | 11 (6–12) | 0.84 [0.74, 0.95] | <.01** | 0.84 [0.73, 0.96] | .01* |
| Clinical program/condition specific | 12 (9–12) | 0.94 [0.82, 1.06] | .32 | 0.96 [0.84, 1.10] | .55 |
| Functional fitness | 12 (6–12) | 0.88 [0.77, 1.00] | .06 | 0.86 [0.75, 0.99] | .04* |
| General fitness | 11 (6–12) | 0.87 [0.77, 0.99] | .03* | 0.88 [0.77, 1.01] | .07 |
| “Group class” | 12 (8–12) | 0.91 [0.80, 1.03] | .14 | 0.89 [0.78, 1.01] | .07 |
| Gym-based program | 11 (7–12) | 0.87 [0.77, 0.99] | .04* | 0.90 [0.78, 1.02] | .11 |
| Othere | 11 (7–12) | 0.86 [0.75, 0.98] | .03* | 0.85 [0.74, 0.99] | .03* |
| Attended free trial classf | |||||
| No | 11 (6–12) | 1.0 (ref.) | 1.0 (ref.) | ||
| Yes | 12 (8–12) | 1.06 [1.04, 1.09] | <.001*** | 1.05 [1.03, 1.08] | <.001*** |
| Class deliveryg | |||||
| In person | 11 (7–12) | 1.0 (ref.) | 1.0 (ref.) | ||
| Online | 12 (10–12) | 1.14 [1.08, 1.22] | <.001*** | 1.19 [1.11, 1.26] | <.001*** |
Note. ARIA = Accessibility/Remoteness Index of Australia; CI = confidence interval; IRR = incidence rate ratio; IQR = interquartile range; Akaike information criterion = 35092.29; Bayesian information criterion = 35274.24.
an = 6,242. bMissing data, n = 11. cMissing data, n = 10. dMissing data, n = 1. e“Other” includes walking groups, Tai-Chi, chair-based, small equipment, bushwalking, and low-impact classes. fMissing data, n = 368. gMissing data, n = 1.
*p < .05. **p < .01. ***p < .001.
For all physical activity outcomes, there was a statistically significant improvement from pretest to posttest (p < .001; Table 3). On average, participants who attended one or more exercise class increased their total physical activity by 447.0 (95% CI [417.6, 476.3]) MET minutes per week, were active on 0.9 (95% CI [0.8, 1.0]) days more per week, and reduced their sitting time by 0.5 (95% CI [−0.6, −0.5]) hr per day from pretest to posttest. For each additional class attended, there was an increase in total physical activity from pretest to posttest of 25.2 (95% CI [2.4, 48.0]) MET minutes per week (Table 4). There was no association between the number of classes attended and the improvement in active days or sitting time.
Pretest and Posttest Physical Activity Values and Unadjusted Mean Differences (n = 3,504)
| Pretest | Posttest | β [95% CI] for unadjusted mean difference | p | |
|---|---|---|---|---|
| Total physical activity (MET minutes/week) | ||||
| Mean (SD) | 739.0 (809.4) | 1,186.0 (959.0) | 447.0 [417.6, 476.3] | <.001*** |
| Range | 0–7413 | 0–7330 | ||
| Active days/week | ||||
| Mean (SD) | 4.1 (2.2) | 5.0 (1.8) | 0.9 [0.8, 1.0] | <.001*** |
| Range | 0–7 | 0–7 | ||
| Sitting time (hr/day) | ||||
| Mean (SD) | 5.5 (2.5) | 5.0 (2.3) | −0.5 [−0.6, −0.5] | <.001*** |
| Range | 0–16 | 0–16 |
Note. CI = confidence interval; MET = metabolic equivalent.
***p < .001.
Association Between the Number of Classes Attended and the Change in Physical Activity Values From Pretest to Posttest (n = 3,504)
| β [95% CI] for adjusted mean difference | pa | |
|---|---|---|
| Total physical activity (MET minutes/week) | ||
| Per class attended | 25.2 [2.4, 48.0] | .03* |
| Active days/week | ||
| Per class attended | 0.1 [0.0, 0.1] | .05 |
| Sitting time (hr/day) | ||
| Per class attended | 0.0 [−0.1, 0.1] | .94 |
Note. CI = confidence interval; MET = metabolic equivalent.
ap value for the interaction with time (i.e., pretest and posttest).
*p < .05.
Discussion
This study aimed to determine factors associated with older adults’ participation in community-based exercise classes during the COVID-19 pandemic and to investigate the association between participation in exercise classes and changes in physical activity levels during this time. Overall, participation in classes was high, with over 95% of all participants attending at least one class and almost 50% attending all 12 classes offered. Participation was associated with state of residence, regionality, comorbidity, the type of class offered, attendance at a free trial class, and whether classes were offered online. For those who completed follow-up testing, participation in classes was associated with a significant increase in self-reported physical activity levels and active days and a significant decrease in self-reported sitting time.
The high participation rates suggest that the Exercise Right for Active Ageing program was well received by older Australians. However, certain challenges to recruitment and retention of older adults were evident throughout the COVID-19 pandemic. Participant recruitment was lower than expected in NSW and Victoria, the most populous states of Australia. Given the timing of slowed recruitment in these states, it is likely that recruitment was affected by COVID-19 lockdowns (Australian Bureau of Statistics, 2021). This was particularly evident in Victoria, where recruitment was minimal throughout the longest lockdown phases (March to October 2020 and August to October 2021) (Boaz, 2021). With regard to attendance, Western Australia, South Australia, and Tasmania had the best rates of class attendance and were also the states with the fewest number of days spent in lockdown (Australian Bureau of Statistics, 2021). The NT and Australian Capital Territory had the poorest rates of attendance. However, in line with their small populations, these territories had very few participants overall, and confidence intervals were wide. Once again, Victoria was highly impacted, with the next lowest rate of attendance. These findings are consistent with those reported internationally wherein attendance at exercise classes was significantly reduced during the COVID-19 pandemic (Christensen et al., 2022; Mañago et al., 2021; Nygård et al., 2022). It was notable that although very few participants (<3%) attended classes online, those who did had better class attendance than those who attended in person. It is unclear whether the low attendance of online classes was due to participant or provider preferences. However, in line with previous research, it is possible that a lack of assistance with technology or suitable digital equipment, for example, laptops with webcams and high-speed internet access, or safety issues were barriers for some older adults (Gell et al., 2021; Mañago et al., 2021; Owen et al., 2022; Paul et al., 2023).
Participants living in inner regional areas also had poorer attendance than those living in cities or in more remote areas. In Australia, residents of regional areas have previously been reported to have lower physical activity levels than people living in cities, potentially due to having poorer health and greater levels of socioeconomic disadvantage, which may have made classes more unaffordable, even with the subsidies provided (Brown et al., 2013; Darcy et al., 2022; Eime et al., 2015; Smith et al., 2008). A lack of financial support for travel to and from classes may also have impacted attendance levels and follow-up. It was notable that participants from outer regional and remote areas had better attendance than those from inner regional areas. The reasons for this are unclear. However, it may have related to these outer regional and remote areas being less affected by lockdowns. Similar to previous studies, participants with a higher number of comorbidities also had lower class attendance, which could be due to physical limitations or a fear of becoming unwell by attending in-person classes during the pandemic (Murphy et al., 2011; Oliveira et al., 2019, 2022). It was notable that alongside yoga/flexibility/mobility classes, clinical programs and condition-specific classes had the highest rates of attendance. This may suggest that these types of classes are preferable for older adults, particularly those with health conditions, who may want gentler or more tailored options (Bock et al., 2014). Finally, it is important to note that for exercise professionals seeking to recruit and retain participants in future programs, older adults who attended a free trial class prior to commencing the program had better overall attendance.
Participants who attended any of the classes offered, and completed follow-up testing, reported an increase in physical activity equivalent to >100 min of MPA per week, almost one additional active day per week, and a half-hour reduction in daily sitting time. Although self-reported physical activity levels often represent an overestimation of true levels (Cerin et al., 2016), these findings suggest that participation in the Exercise Right for Active Ageing program boosted overall physical activity levels and that there was additional time spent engaging in physical activity beyond just the time spent in classes. This is likely to have led to health benefits for participants, with evidence that in older adults, an additional 15 min of daily physical activity can reduce all-cause mortality by 4% (Wen et al., 2011). It is important to investigate whether physical activity levels have been sustained since the end of the subsidized program. Also, given the likely reduction in class participation and physical activity levels in areas under lockdown, it is important to collect empirical data on the physical health effects of COVID-19 lockdowns on older adults. Because the physical health-promoting effects of physical activity may take some time to be realized, such data may only be available in the long term. However, U.K. modeling predicts an increase in the total number of falls by 6.3% in men and 4.4% in women over coming years without mitigation of physical activity levels lost during the pandemic (Public Health England, 2021). Furthermore, this is projected to cost England’s health and social care system an additional £211 million over a two-and-a-half-year period.
This nationwide, multisite study is one of the largest to investigate older adults’ participation levels in AES- and AEP-led exercise classes. Although COVID-19 restrictions disrupted recruitment and retention of study participants, the timing of the pandemic also enabled an indirect investigation of the potential association between lockdowns and class participation. However, there are some study limitations that should be acknowledged. First, the absence of a control group in this quasi-experimental study means that a causal link between program participation and physical activity levels is less certain. Second, the large loss to follow-up for secondary analyses and significant differences between participants lost to follow-up and those included are a threat to the generalizability of these findings. Although we did not collect data on the reasons for loss to follow-up, participants who were lost to follow-up had similar characteristics to participants with lower class attendance, suggesting disengagement with the overall program. It is possible that participant drop out related to financial issues, poor health, and/or travel limitations, which may have been exacerbated by the COVID-19 pandemic and related restrictions. Third, the nonblinding of assessors and participants may have affected the trustworthiness of follow-up physical activity data, with participants potentially attempting to improve their pretest scores, thereby inflating the effect of the intervention. Fourth, and as discussed previously, although suitable for pre- and poststudies, self-reported physical activity levels are prone to overestimation and self-reported sedentary time to underestimation (Cerin et al., 2016; Chastin et al., 2014). Therefore, self-reported data should not be compared or combined with device-measured data. Finally, the type of classes delivered was decided upon by individual providers based on their skills and preferences. It was not the intention of the study to analyze the impact of specific types of classes. However, by combining all class types in this study, there was a loss of information about the specific effects of different types of exercise. In future research, it would be beneficial to conduct subgroup analysis of the impact of different types of exercise classes included in this study.
Despite the challenges faced by participants and providers during the COVID-19 pandemic, our findings have implications of for promoting older adults’ participation in group exercise classes. Given the high rates of class attendance in this study, certain characteristics of our program could be considered in the design of future exercise programs, including offering free trial classes and other financial subsidies, designing classes specifically for older adults that can accommodate physical impairments and disability and reduce social stigma, providing online class options, and enlisting trained health professionals to ensure safe delivery. However, given our findings of lower class attendance in certain subgroups, including older adults from inner regional areas and those reporting comorbidities, it is important to investigate what kinds of support are needed from health professionals, organizations, and policymakers to help these people to participate in community-based exercise classes.
Conclusions
The latest World Health Organization Global Action Plan on Physical Activity includes a key recommendation that nations should work to strengthen physical activity programs and services for older adults (World Health Organization, 2018). This was the key driver of the Exercise Right for Active Ageing program, which sought to provide subsidized AES- and AEP-led exercise classes for community-dwelling older adults across Australia. This study aimed to determine factors associated with older adults’ participation in the Exercise Right for Active Ageing program during the COVID-19 pandemic and to investigate the association between program participation and changes in physical activity levels during this time. Overall, classes were well attended, particularly in states and territories less impacted by COVID-19 lockdowns, and participation in classes was associated with a significant increase in physical activity levels. It is important to investigate the health and physical performance effects of program participation and, now that the subsidized program has ceased, whether the program has led to sustained improvements in exercise and physical activity behaviors in older Australians.
Acknowledgments
Leanne Evans, Dr. Sharon Hetherington, Dr. Channa Marsh, and Jeffrey Allen from Exercise and Sports Science Australia (ESSA) are thanked for their assistance with this project. We gratefully acknowledge the participants of this research and the exercise providers for contributing their time and effort. This project was supported by the Australian Government and managed by the Australian Sports Commission, through the Participation Grants Program. The funder had no involvement in the study design, data collection, analysis and interpretation of data, the writing of the report, or the decision to submit the article for publication. Ayton was supported by a National Health and Medical Research Council of Australia (NHMRC) Investigator Grant (APP1195357).
References
Ainsworth, B.E., Haskell, W.L., Herrmann, S.D., Meckes, N., Bassett, D.R., Jr., Tudor-Locke, C., Greer, J.L., Vezina, J., Whitt-Glover, M.C., & Leon, A.S. (2011). 2011 compendium of physical activities: A second update of codes and MET values. Medicine & Science in Sports & Exercise, 43(8), 1575–1581.
Australian Bureau of Statistics. (2020–2021). Physical activity. https://www.abs.gov.au/statistics/health/health-conditions-and-risks/physical-activity/latest-release
Australian Bureau of Statistics. (2021). Impact of lockdowns on household consumption – Insights from alternative data sources. Commonwealth of Australia. https://www.abs.gov.au/articles/impact-lockdowns-household-consumption-insights-alternative-data-sources
Australian Sports Commission. (2023). Move it AUS grant programs. Australian Government. https://www.sportaus.gov.au/grants_and_funding/moveitaus-grant-programs
Baert, V., Gorus, E., Mets, T., Geerts, C., & Bautmans, I. (2011). Motivators and barriers for physical activity in the oldest old: A systematic review. Ageing Research Reviews, 10(4), 464–474.
Boaz, J. (2021). Melbourne passes Buenos Aires’ world record for time spent in COVID-19 lockdown. ABC News Online. https://www.abc.net.au/news/2021-10-03/melbourne-longest-lockdown/100510710
Bock, C., Jarczok, M.N., & Litaker, D. (2014). Community-based efforts to promote physical activity: A systematic review of interventions considering mode of delivery, study quality and population subgroups. Journal of Science and Medicine in Sport, 17(3), 276–282.
Booth, M., Bauman, A., & Owen, N. (2002). Perceived barriers to physical activity among older Australians. Journal of Aging and Physical Activity, 10(3), 271–280.
Brown, W.J., Burton, N.W., Sahlqvist, S., Heesch, K.C., McCarthy, K.B., Ng, N., & van Uffelen, J.G. (2013). Physical activity in three regional communities in Queensland. Australian Journal of Rural Health, 21(2), 112–120.
Caspersen, C.J., Powell, K.E., & Christenson, G.M. (1985). Physical activity, exercise, and physical fitness: Definitions and distinctions for health-related research. Public Health Reports, 100(2), 126–131. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1424733/
Cerin, E., Cain, K.L., Oyeyemi, A.L., Owen, N., Conway, T.L., Cochrane, T., Schipperijn, J., Mitáš, J., Toftager, M., Aguinaga-Ontoso, I., & Sallis, J.F. (2016). Correlates of agreement between accelerometry and self-reported physical activity. Medicine & Science in Sports & Exercise, 48(6), 1075–1084.
Chastin, S.F., Culhane, B., & Dall, P.M. (2014). Comparison of self-reported measure of sitting time (IPAQ) with objective measurement (activPAL). Physiological Measurement, 35(11), 2319–2328.
Christensen, A., Bond, S., & McKenna, J. (2022). The COVID-19 conundrum: Keeping safe while becoming inactive. A rapid review of physical activity, sedentary behaviour, and exercise in adults by gender and age. PLoS One, 17(1), Article e0263053.
Cunningham, C., O’ Sullivan, R., Caserotti, P., & Tully, M.A. (2020). Consequences of physical inactivity in older adults: A systematic review of reviews and meta-analyses. Scandinavian Journal of Medicine and Science in Sports, 30(5), 816–827.
Darcy, M., Parkinson, J., McDonald, N., Moriarty, S., Kadariya, S., & Sapkota, D. (2022). Geographic remoteness and socioeconomic disadvantage reduce the supportiveness of food and physical activity environments in Australia. Australian and New Zealand Journal of Public Health, 46(3), 346–353.
Daskalopoulou, C., Stubbs, B., Kralj, C., Koukounari, A., Prince, M., & Prina, A.M. (2017). Physical activity and healthy ageing: A systematic review and meta-analysis of longitudinal cohort studies. Ageing Research Reviews, 38, 6–17.
Eime, R.M., Charity, M.J., Harvey, J.T., & Payne, W.R. (2015). Participation in sport and physical activity: Associations with socio-economic status and geographical remoteness. BMC Public Health, 15(1), Article 434.
Exercise and Sports Science Australia. (2019). Adult pre-exercise screening system V2. https://www.essa.org.au/Public/ABOUT_ESSA/Pre-Exercise_Screening_Systems.aspx
Exercise and Sports Science Australia. (2023a). Exercise right for active ageing. https://exerciseright.com.au/betterageing/
Exercise and Sports Science Australia. (2023b). Who we are. https://www.essa.org.au/Public/About/Who_we_are/Public/ABOUT_ESSA/Who_we_are.aspx?hkey=db18d152-8af4-4376-984d-1a2c2890498f
Franco, M.R., Tong, A., Howard, K., Sherrington, C., Ferreira, P.H., Pinto, R.Z., & Ferreira, M.L. (2015). Older people’s perspectives on participation in physical activity: A systematic review and thematic synthesis of qualitative literature. British Journal of Sports Medicine, 49(19), 1268–1276.
Gell, N., Hoffman, E., & Patel, K. (2021). Technology support challenges and recommendations for adapting an evidence-based exercise program for remote delivery to older adults: Exploratory mixed methods study. JMIR Aging, 4(4), Article e27645.
Goethals, L., Barth, N., Guyot, J., Hupin, D., Celarier, T., & Bongue, B. (2020). Impact of home quarantine on physical activity among older adults living at home during the COVID-19 pandemic: Qualitative interview study. JMIR Aging, 3(1), Article e19007.
Gough, C., Barr, C., Lewis, L.K., Hutchinson, C., Maeder, A., & George, S. (2023). Older adults’ community participation, physical activity, and social interactions during and following COVID-19 restrictions in Australia: A mixed methods approach. BMC Public Health, 23(1), Article 172.
Haynes, A.B., Haukoos, J.S., & Dimick, J.B. (2021). TREND reporting guidelines for nonrandomized/quasi-experimental study designs. JAMA Surgery, 156(9), 879–880.
Hoffman, G.J., Malani, P.N., Solway, E., Kirch, M., Singer, D.C., & Kullgren, J.T. (2022). Changes in activity levels, physical functioning, and fall risk during the COVID-19 pandemic. Journal of the American Geriatrics Society, 70(1), 49–59.
Hosmer, D.W., Lemeshow, S., & Sturdivant, R.X. (2013). Model-building strategies and methods for logistic regression. In W.A. Shewhart & S.S. Wilks (Eds.), Applied logistic regression (pp. 89–151). John Wiley & Sons, Inc.
Langhammer, B., Bergland, A., & Rydwik, E. (2018). The importance of physical activity exercise among older people. BioMed Research International, 2018, Article 7856823.
Lee, P.H., Macfarlane, D.J., Lam, T.H., & Stewart, S.M. (2011). Validity of the international physical activity questionnaire short form (IPAQ-SF): A systematic review. International Journal of Behavioral Nutrition and Physical Activity, 8, Article 115.
Mañago, M.M., Swink, L.A., Hager, E.R., Gisbert, R., Earhart, G.M., Christiansen, C.L., & Schenkman, M. (2021). The impact of COVID-19 on community-based exercise classes for people with Parkinson disease. Physical Therapy, 101(11), Article pzab203.
Murphy, C.-L., Sheane, B.J., & Cunnane, G. (2011). Attitudes towards exercise in patients with chronic disease: The influence of comorbid factors on motivation and ability to exercise. Postgraduate Medical Journal, 87(1024), 96–100.
Nygård, A.J., Taraldsen, K., Granbo, R., Selbæk, G., & Helbostad, J.L. (2022). Impacts of COVID-19 restrictions on level of physical activity and health in home-dwelling older adults in Norway. European Review of Aging and Physical Activity, 19(1), Article 30.
Oliveira, J.S., Sherrington, C., Rowling, L., & Tiedemann, A. (2019). Factors associated with ongoing participation in structured exercise among people aged 50 years and older. Journal of Aging and Physical Activity, 27(5), 739–745.
Oliveira, M.R., Sudati, I.P., Konzen, V.M., de Campos, A.C., Wibelinger, L.M., Correa, C., Miguel, F.M., Silva, R.N., & Borghi-Silva, A. (2022). COVID-19 and the impact on the physical activity level of elderly people: A systematic review. Experimental Gerontology, 159, Article 111675.
Owen, P.J., Keating, S.E., Askew, C.D., Clanchy, K.M., Jansons, P., Maddison, R., Maiorana, A., McVicar, J., Robinson, S., & Mundell, N.L. (2022). Impact of the COVID-19 pandemic on exercise physiology services in Australia: A retrospective audit. Sports Medicine – Open, 8(1), Article 94.
Paul, S.S., Hubbard, A., Johnson, J., & Dennis, S.M. (2023). Transition to a virtual model of physiotherapy and exercise physiology in response to COVID-19 for people in a rural Australia: Is it a viable solution to increase access to allied health for rural populations? PLoS One, 18(1), Article e0280876.
Public Health England. (2021). Wider impacts of COVID-19 on physical activity, deconditioning and falls in older adults. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1010501/HEMT_Wider_Impacts_Falls.pdf
Rona, M., Victoria, P., Dafna, M., Andrew, M., Claire, M., Adrian, B., & Hidde, P. (2014). Barriers and enablers to physical activity among older Australians who want to increase their physical activity levels. Journal of Physical Activity and Health, 11(7), 1420–1429.
Sepúlveda-Loyola, W., Rodríguez-Sánchez, I., Pérez-Rodríguez, P., Ganz, F., Torralba, R., Oliveira, D.V., & Rodríguez-Mañas, L. (2020). Impact of social isolation due to COVID-19 on health in older people: Mental and physical effects and recommendations. Journal of Nutrition, Health and Aging, 24, 938–947.
Sheather, S. (2009). A modern approach to regression with R. Springer.
Smith, K.B., Humphreys, J.S., & Wilson, M.G. (2008). Addressing the health disadvantage of rural populations: How does epidemiological evidence inform rural health policies and research? Australian Journal of Rural Health, 16(2), 56–66.
Sun, F., Norman, I.J., & While, A.E. (2013). Physical activity in older people: A systematic review. BMC Public Health, 13, Article 449.
Tabachnick, B.G., & Fidell, L.S. (2014). Using multivariate statistics: Pearson new international edition (6th ed.). Pearson Education Limited.
Wen, C.P., Wai, J.P.M., Tsai, M.K., Yang, Y.C., Cheng, T.Y.D., Lee, M.-C., Chan, H.T., Tsao, C.K., Tsai, S.P., & Wu, X. (2011). Minimum amount of physical activity for reduced mortality and extended life expectancy: A prospective cohort study. The Lancet, 378(9798), 1244–1253.
World Health Organization. (2018). Global action plan on physical activity 2018–2030: More active people for a healthier world.
World Health Organization. (2020). WHO guidelines on physical activity and sedentary behaviour.
