Effects of museum-based art activities on older community dwellers’ physical activity: the A-health randomized controlled trial results

Because museum-based art activities practiced in a group setting improve the physical health of older adults, we suggest that they may also improve their physical activity through an increase in daily step count. A greater daily step count was observed in individuals who practiced museum-based art activities compared to those who did not practice these activities. Museum-based art activities improved daily physical activity in older community dwellers, confirming health benefits and suggesting the potential of museums in health promotion and disease prevention. Museum-based art activities have demonstrated health benefits in older adults. Few clinical trials, however, have examined physical health benefits specifically. This randomized controlled trial (RCT) aims to compare changes in daily step count over a 3-month period in older adults participating in museum-based art activities and their control counterparts. Using a subset of 53 participants recruited in the A-health RCT, the daily step count of 28 participants in the intervention group (age 70.5 ± 4.9 and 92.0% female) and 25 in the control group (age 71.5 ± 5.3 and 78.6% female) were recorded using a Fitbit Alta HR. Weekly art activities were carried out at the Montreal Museum of Fine Arts (MMFA, Quebec, Canada) over a 3-month period. The outcomes were the mean step count per active hours (i.e., between noon and 6 pm), inactive hours (i.e., between midnight and 6 am) and over the full day (i.e., 24 h) and the change in step count following the 3-month (M3) art-based intervention at the MMFA. The intervention group had a greater daily step count compared to the control group at M3, regardless of the step parameters examined (P ≤ 0.026). Linear regressions showed that the change in daily step count for the full day (P ≤ 0.010) and active hours (P ≤ 0.026) increased significantly with the MMFA art-based activities. MMFA-based art activities improved daily physical activity in older community-dwellers who participated in the RCT, confirming health benefits and suggesting the potential of museums in health promotion and disease prevention.


Introduction
In a 2019 scoping review, the World Health Organization reported that artistic and cultural activities may improve mental and physical health, especially in older adults [1].For instance, it has been demonstrated that museum-based art activities improve quality of life and well-being in the older population [3][4][5].Few clinical trials, however, have examined the physical health benefits of such interventions [1][2][3][4][5].Two recent randomized controlled trials (RCTs) demonstrate that museum-based art activities can improve the physical health of older community dwellers [4,5].A significant limitation of these RCTs, however, was the use of self-reported questionnaires to measure physical health.Self-reported questionnaires are associated with different biases, like the desirability or response bias, which may result in invalid answers [6].This highlights the need for more objective measures, like quantifiable outcomes recorded with devices, to confirm the physical health benefits for older adults of museum-based art activities.
The prevalence and incidence of non-communicable chronic diseases, such as cardiovascular diseases, increase with age [7][8][9].Accumulation of these chronic morbidities impacts physical ability [10].Geriatric medicine has foregrounded the need to switch from a disease-oriented to a functional approach when taking care of older adults [11].Physical impairment due to morbidities is highly prevalent in the older population and acting on it may prevent or delay its adverse consequences [9,12,13].Physical activity is defined as any bodily movement produced by skeletal muscle that requires energy expenditure and may be easily assessed with the number of daily steps [7,8].Adequate physical activity may prevent non-communicable chronic diseases and delay their functional impairments [12,13].Despite this, a sedentary lifestyle is highly prevalent in the Canadian population.For instance, nearly half of Canadian adults are not physically active enough to benefit their health, and this prevalence increases with age [13].Even if people are aware of the health benefits of regular physical activity, reversing a sedentary lifestyle may be difficult, particularly for older adults [7-9, 12, 13].Studies show that older adults prefer to engage in social activities (compared to physical exercise) in their leisure time [14].Such social engagement has both mental and physical health benefits for community-dwelling older adults [15].Regular participation in community activities may promote this social engagement, which the World health organization deems a key part of healthy ageing in its 2015 report on the subject [16].Because museum-based art activities practiced in a group setting have been shown to improve physical health in older adults, we suggest that they may also improve physical activity levels, as reflected in an increased daily step count.
We had the opportunity to test this hypothesis using the daily step count of participants of the A-Health study [4].This study is an RCT which examined the health benefits of art activities carried out at the Montreal Museum of Fine Arts (MMFA; Quebec, Canada) [4].We hypothesized that participants in the intervention group of this RCT would have a higher daily step count than those in the control group.The aim of this RCT was thus to compare changes over of a 3-month period in the daily step count of older adults who participated in the MMFA-based art activities and in their control counterparts.

Population and design
The individuals selected for this present study were participants of the A-health RCT who had their step count recorded over the 3-month period [4].The details of this study have been previously described [4].In brief, it was an RCT in two parallel groups (intervention versus control).Participants in the intervention group were not blinded (compared to investigators).The CONSORT guidelines were applied for this RCT [17].MMFA visitors who were aged 65 years and over and who lived at home or in an autonomous residence with Internet access were enrolled in the A-Health study.A total of 209 individuals registered to participate in the A-health RCT.Among them, 198 (94.7%) were recruited: 100 in the intervention group and 98 in the control group.A subset of 60 (30.3%) participants had their step count recorded: 30 in the intervention group and 30 in the control group.This number of participants was chosen for two reasons: first, the RCT was an explanatory pilot study; and second, the cost of Fitbit device was prohibitive.Participants of the present study were selected using a second randomization that keeps the subsets as similar as possible to control for selection bias.This second randomization adopts the process used during the first randomization.After the second randomization was completed, the project manager responsible provided selected participants with a Fitbit device, an assessment schedule and instructions.A total of 7 (11.7%)participants were excluded from the present study because step count collection failed due to a technical issue.Thus, the step count of a total of 53 (88.3%) participants (28 in the intervention group and 25 in the control group) was recorded over the 3-month RCT.The flow diagram in Fig. 1 illustrates the participant selection and follow-up process.

Intervention
The intervention, consisting of participatory art activities carried out at the MMFA, has been previously described in detail [4].In summary, weekly 2 h workshops were performed over a 3-month period consecutively.All workshops were conducted by two arts and culture facilitators and performed in a group setting.At the beginning of each workshop, the session's objective and medium were presented.The participatory art-based activities were interactive arts and crafts activities targeting creativity (e.g., abstract painting/life drawing), craft skills (e.g., bookbinding/mini fanzine making) or fine motor skills (e.g., rolled paper and stained-glass painting).

Assessment and outcomes
Information about age, sex, ethnicity (i.e., Caucasian versus non-Caucasian) and the Centre of Excellence Self-AdMinistered questionnaire (CESAM) were collected before the first workshop (M0) and after the last workshop at 3 months (M3) [18].CESAM is a digitalized questionnaire (accessible on a web platform) which collects a range of health information using 20 close-ended questions.More specifically, CESAM is composed of the following items: home support (i.e., receiving help from family, friends or a professional for daily living activities), polypharmacy (i.e., number of therapeutic classes taken on a daily basis ≥ 5), activities of daily living (ADL) and instrumental activities of daily living (IADL), mood, physical activity and history of falls in the past 6 months.CESAM was filled out by the participants under the supervision of Principal Investigator representatives.In addition, participants were separated into two groups based on their total number of steps per day.Those with less than 6000 steps were classified as inactive [19].
Step count was recorded using a Fitbit Alta HR, which is a watch-like device that continuously tracks step count using a three-axis accelerometer [20].The measurements were taken using an application provided by Fitbit Inc (San Francisco, CA, USA).Participants wore the watch on the of their choice 24 h per day over a consecutive 5-day period.For this study, we used the mean step count of the two days before the first workshop (M0) and the mean step count of the two days following the last workshop (M3).Participants were asked to not to change wrists and were only to remove the device in situations where it may get wet (e.g., washing dishes or showering).We calculated 3 types of mean daily step count: 1) per active hours (i.e., between noon and 6 pm), per inactive hours (i.e., between midnight and 6 am) and per full day (i.e., 24 h).In addition, we calculated the difference between the first (M0) and second assessments (M3) for each step parameter using the following formula: [(M3 value-M0 value)/((M3 value + M0 value)/2) × 100].

Ethical considerations
Enrolled participants provided their written informed consent for the study.The Jewish General Hospital Ethics Committee (Montreal, Quebec, Canada) approved the project (2019-1493).

Statistics
Mean values, standard deviations (SD), frequencies and percentages were used to describe the participants' characteristics at M0 and M3.Comparisons between intervention and control groups were performed with Mann-Whitney t-tests, chi-squared or Fisher's Exact tests, as appropriate.The association of changes in step count parameters between M0 and M3 (used as dependent variables, with a separated model for each heart rate parameter) and MMFA-based art activity (used as an independent variable) was examined using linear regressions.Adjustments were made for participants' baseline characteristics for each linear regression.The threshold of significance for P-values was fixed at ≤ 0.05.

Results
There was no significant difference between the groups at baseline assessment for the clinical characteristics (Table 1).Mean daily step counts and changes between M0 and M3 per full day (P ≤ 0.040), active hours (P ≤ 0.030) and inactive hours (P ≤ 0.040) were significantly greater in the intervention group compared to the control group (Table 2).The linear regression showed that changes in daily step count between M0 and M3 per full day and per active hours were positively associated with participatory art-based activity (P = 0.010 and P = 0.026 for full set of participants; P < 0.001 and P = 0.001 for inactive participants) (Table 3).

Discussion
The results demonstrate that the MMFA-based art activities were associated with an improvement in the daily step count of older community dwellers.Compared to the control group, the intervention group showed a significant increase in mean step count at M3, regardless of the step parameters examined.The change between M0 and M3 was positively associated with the MMFA-based art activity for both the full day and active hour mean step count.This effect was greater in inactive participants compared to the other participants.No significant association was shown with the inactive hour mean step count.
Overall, the results point to an increase in the intervention group's physical activity compared to their control counterparts.These results are consistent with previous studies demonstrating the health benefits of museum-based art activities [1][2][3][4][5].Walking is a central component of physical activity and public health promotion efforts [21,22].Daily step count is a surrogate marker of physical activity levels in adults, with an increase in step count indicating an increase in physical activity and thus better health [22,23].This in turn reduces sedentary behavior and the related risk of cardiovascular diseases, sarcopenia, and mental illnesses [21][22][23].Thus, a higher step count may be assimilated as a health benefit.The mechanism of the increase in daily step count in our study likely has different complementary components.First, we previously demonstrated that MMFAbased art activities are associated with an improvement in participants' mental health.Thus, this positive mental health benefit may encourage the participants to be more physically active.It is known from the literature that positive affect positively influences physical health [1][2][3][4][5].Second, MMFA activities were performed in groups and therefore, provide an opportunity for the participants to break out of social isolation, which is also associated with sedentary behavior.Third, art activities were carried out on-site at the MMFA.This encouraged participants to get out of the house and thus partly explains the increase in step count.
We observed mixed results for mean step count per inactive hours.If there was a significant increase in the mean step count at M3 in the intervention group compared to the control group, the change between M0 and M3 was not associated with MMFA art-based activities.A lack of data may explain this non-conclusive result.The number of participants in this RCT was low, and compared to the full day and active hour parameters, the number of steps collected during inactive hours was also low.The inactive hours represent resting hours, which means reduced activity levels are normal for this period.Since the average number of steps during inactive hours shows an increase in physical activity during this period, we suggest that it may be possible to obtain a significant result with a larger sample of participants.
This study has a number of strengths and limitations.First, it uses an RCT design in two parallel groups.Second, the measures obtained are objective, quantitative measures.Third, there were no significant differences in participant characteristics at baseline.Some limitations, however, should be mentioned.First, the sample size was limited to 53 participants.Second, the study was a unicentric RCT conducted exclusively at the MMFA.Third, the control group could have been exposed to activities that may have had an influence on the outcomes.Fourth, recruiting participants from an existing pool of museum visitors may be lead to bias because these participants were already interested in museum-based art activities, which may have limited the effects of the intervention.The mixed results for mean step count per inactive hours may be explained by this limitation of our RCT.In addition, because of participants' preexisting interest in art, the observed benefits of art-based activities may be limited to older people who are interested in art.Fifth, a bias due to missing data is also possible.Around 12% of participants were excluded because of a failure in data collection due to a technical issue.This lack of this data may introduce a systematic error or distortion in the obtained results.Sixth, going to the museum is a physical activity which may increase the number of daily steps in the intervention group.However, participants went to the museum using their car or public transportation, which limits the potential to impact the number of steps in a large proportion.In addition, there was no recommendation limiting the practice of physical activity in both the intervention and control groups.Furthermore, the proportion of participants who practiced a physical activity at baseline was similar in intervention and control groups.Finally, the participants of the RCT had a specific phenotype.They were Caucasian females who had a habit of regular physical activity.Thus, the results are not generalizable to other population groups.
Table 3 Linear regressions showing the association between the change in step parameters between baseline and the end of the intervention at 3 months (used as dependent variables, with a separated model for each parameter) and the art-based activities (used as an independent variable) in the participants (n = 53) β Coefficient of regression beta, CI confident interval; All models adjusted on baseline characteristics of participants (Please see In conclusion, MMFA-based art activities improved daily physical activity of the older community-dwellers who participated in the RCT, confirming the health benefits of artbased activities and suggesting museums' potential as key actors in health promotion and disease prevention.

Fig. 1
Fig. 1 Flow diagram of the selection and follow-up of participants in the RCT who had their step count recorded.M0 baseline, M3 3 months