Background

The prevalence of age associated physical and cognitive decline is markedly rising due to the ageing population and places an enormous strain on health and social care in the UK [1]. One facet of ageing for many older adults is a tendency towards physical frailty, which, among other things, is underpinned by a deterioration in muscle function. The estimated annual loss of muscle size is 0.5-1% per year, with muscle strength lost at 1–5% per year in older age, with this process starting anywhere between 45 and 60 years of age [2, 3]. Eventually the continued loss of muscle strength results in the tasks of daily living becoming too physically strenuous to be managed safely, greatly increasing fall risk and compromising individuals’ ability to live independently [4]. In the UK, muscle weakness in and of itself is estimated to result in excess healthcare costs of £2.5 billion annually [5], with the total annual cost of falling in the UK estimated at £4.4 billion, including £1.1 billion for social care [6]. Indeed, falls are the number one reason for hospitalisation in UK older adults, with at least 1 third of all community-dwelling people aged 65 years or older falling each year [7].

Older people with cognitive impairment are markedly more likely to have reduced physical function and strength, and ultimately a greater fall risk [8, 9]. Up to a third of emergency hospital admissions occur in an older person with dementia, and of these admissions, over half are associated with a fall [10]. Finding ways to engage and support older adults identified as being at increased risk of falls to maintain or even improve their strength and balance will have enormous benefits for both the individuals and wider society. The recently published World Falls Guidelines stress that even in low-risk groups exercise interventions are to be recommended, but it is acknowledged that adherence is particularly challenging [11]. Maintaining strength and balance is key to protecting against a loss of mobility, retaining the confidence and ability to undergo aerobic physical activity and social engagement, and preventing further decline in health [7, 12]. Progressive strength training is undisputedly the most effective countermeasure to age associated declines in muscle mass and function [13].

There is promising evidence that exercise programmes may improve the ability of people with dementia to perform activities of daily living [14]. The UK CMO recommends that older adults undertake exercise sessions (e.g., resistance training) to build muscle strength and balance on two days each week [15]. Recent data however suggests that only 12% of adults aged 65 + and 5% aged 75 + years meet these guidelines [16]. Research suggests that the most common barriers to exercise in older age groups include a perceived lack of time, expertise (i.e., for resistance training), poor access to facilities, low self-confidence for attending gym settings or fear of injury/falling [17,18,19,20]. Moreover, cognitive impairment is a potential barrier to engagement and adherence in exercise interventions. As such, there is a need for interventions that are accessible, and to which people can adhere to achieve at least the minimum required dose to promote improvements in functional capacity. Innovative but simple and inexpensive programmes such as homebased ‘exercise-snacking’ have been designed to overcome these barriers, while also providing the functional improvements that would benefit older adults.

Exercise snacking is a mode of exercise that aims to have maximum impact on physical function at minimum ‘cost’ to the patient in terms of time, money and no requirement for specialist exercise facilities or equipment [21, 22]. In other words, the exercise snacks, which are movements designed to increase muscle strength and balance, are to be performed in the home environment over very short periods of time that fit with the current lifestyle of the patient [23]. A novel format of home-based exercise-snacking was designed and refined by researchers at the University of Bath in an attempt to overcome key barriers to engagement of older adults in exercise, by providing a safe, effective, programme that requires no specialist equipment or membership of a gym [18]. Furthermore, the simplicity of the programme and ease of which it is implemented should help empower and boost the self-efficacy of older adults to undertake regular strength and balance exercise [24].

Key to the success of any [new] intervention is that it is centred around the specific needs of the target population [25]. Early indications are that this exercise snacking intervention is safe, beneficial and acceptable to healthy older adults recruited from the general population [21]. However, no research exists examining its potential in outpatient populations who present with increased risk, such as pre-frail older adults who attend a memory clinic. Understanding the perspectives of these target users on the usefulness and ease of implementation is an important first step in ensuring the programme maximises the chance of engaging and benefitting patients. Contemporary research postulates the construct of ‘acceptability’ to be multidimensional, covering elements such as whether a user finds a given intervention to be enjoyable, aligned to their values and identity, easy to understand; not burdensome or displacing of other positive aspects of their life; perceptively effective; and self-efficacy endorsing [26, 27]. Additionally, best practice guidance on the development of health interventions recommends the understanding of the process variables derived to form a theoretical basis and logic model that examine the mechanisms of change or effect [28]. To this end, the present pilot study had two key aims:

  1. 1)

    to examine the acceptability of the exercise snacking intervention guided by the Theoretical Framework of Acceptability (TFA) and identify areas that may need to be improved to optimise it for attendees of a memory clinic.

  2. 2)

    to explore and characterise the potential impact of this intervention on the physical function for this population on outcomes relating to physical function, health, and exercise cognitions.

Methods

Study design

This pilot study used a single group, pre-test-post-test design to assess the acceptability homebased ‘exercise snacking’ in older adult patients attending the memory clinic at the Research Institute for the Care of Older People (RICE) in Bath, UK. All participants were asked to undertake 28 days of twice daily exercise snacks, with baseline measures of physical and cognitive function, patient reported health, wellbeing, and psychological process outcomes relating to exercise behaviour, recorded on the day before the intervention and follow-up measures scheduled within 7-days of the final day of the intervention. The primary outcome of intervention acceptability was measured by self-report questionnaire, with participants also invited to participate in a qualitative interview of their experience of the intervention.

Participants

Outpatients attending the memory clinic at RICE who were identified as being potentially eligible to participate in the present pilot study were invited to participate by RICE gerontology clinicians. Interested individuals were referred to study staff, with clinicians taking no further roll in study procedures. Eligibility requirements were age > 65 years; diagnosed mild cognitive impairment only (Mini-Mental State Examination (MMSE) [29] score ≥ 20); mildly impaired physical function (Short Physical Performance Battery (SPPB) [30] score 3–8 but not scoring zero on any individual component of the test); ability to safely perform 28 consecutive days of homebased exercise snacking; have someone present in the house during exercise snacks who would be capable of calling for help in the event of an emergency; and not currently engaging in regular exercise or participating in another research study with measures or interventions that might interfere with the present study outcomes. Potential participants with contraindications to exercise or co-morbidity preventing participation in exercise (severe breathlessness, pain, psychosis, Parkinson’s, Dementia with Lewy Bodies, or other severe neurological disease), or current musculoskeletal condition that could be made worse through exercise were excluded from participation. The inclusion and exclusion criteria were assessed with a bespoke self-report health screen questionnaire, and eligible participants provided written informed consent. The protocol was approved by the National Health Service (NHS) South West – Frenchay Research Ethics Committee (Ref: 22/SW/0084), conducted in accordance with the Helsinki Declaration of 1975, as revised in 2000, and registered on ClinicalTrials.gov (Identifier: NCT05439252, 30/06/2022). All participants taking part in the study provided written informed consent.

With intervention acceptability as a primary outcome measure, there is a lack of comparable literature to quantitatively estimate a required sample size in this population, given adherence rates to the exercise snacking intervention are yet to be established. As such, for reasons of pragmatism and a fixed duration of study funding, a target of 20 participants to complete follow-up measurements in a three-month testing window was set. Participants who had not adhered to the intervention were included within this target sample size, so long as they would attend follow-up assessment. This approach was taken in an attempt to avoid a withdrawal bias, i.e., only participants who adhered to the intervention returning for follow-up assessment thus positively skewing interpretation of intervention acceptability. A rolling recruitment strategy was employed whereby up to 10 extra participants could be recruited to replace participants who withdrew from the study and were not willing or able to attend follow-up assessments, provided there was sufficient time to enrol and complete follow-up assessments of the replacement participants in the three-month testing period.

Assessments

Intervention acceptability was assessed by a theoretical framework of acceptability questionnaire on participant experiences of exercise snacking completed at the follow-up assessment only. The seven domains (Affective attitude; Burden; Ethicality; Intervention Coherence; Opportunity Cost; Perceived effectiveness; and Self-efficacy) of the TFA along with overall acceptability were measured using an 8-item questionnaire [27]. An optional qualitative interview to gather further data on the participant’s experience of the exercise snacking intervention was offered to all participants.

During eligibility screening, baseline characteristics were collected (Table 1), and participants completed a health screen questionnaire, the MMSE, and the SPPB. Height (Seca 213 stadiometer, Hamburg, DE) and weight (Seca 799 electronic scales, Hamburg, DE) were measured with participant’s shoes off, and their ability to safely perform exercise snacking was assessed by the trained research psychologist, using a subjective assessment of the participants’ stability and self-reported confidence of completing the assessment.

At baseline and follow-up assessments, participants completed questionnaires on their attitude towards the benefits of exercise using the Multidimensional Outcome Expectancy for Exercise Questionnaire [31], self-confidence for exercise using the Barrier Self-efficacy Scale [32], and satisfaction of the psychological needs for autonomy competence and relatedness in an exercise context using the Psychological need satisfaction for exercise questionnaire [33]. The Patient Health Questionnaire [34] and Generalised Anxiety Disorder Assessment [35] were used to assess mental health, while general health and quality of life was measured using the Short Form Health Survey [36], the Subjective Vitality Scale [37], and the Life Satisfaction Scale [38]. Cognitive function was assessed with the Montreal Cognitive Assessment [39] and self-reported indices of frailty were assessed with the Groningen Frailty Index [40]. The questionnaires were administered by a research psychologist.

Alongside the SPPB, physical function was also assessed by a 60-second sit-to-stand test, in which participants were instructed to complete as many repetitions as possible in 60 s from a hard backed chair with seat height of 45 cm. Participants were permitted to use their arms for balance, but not to use their arms to propel or push themselves out of the chair. To avoid any self-pacing, participants were positioned out of view of a clock and the researcher did not count the repetitions out loud. A single attempt was permitted, and rating of perceived exertion (RPE [41]) was taken immediately after the sit-to-stand test as a secondary measure of test difficulty to capture potential improvements in the event of number of repetitions reaching a physical ceiling for a given participant. As a further test of physical function, the Timed-Up-and-Go (TUG [42],) was attempted twice with the best effort reported. Maximum duration standing balance tests were attempted with feet in semi- and full tandem positions, followed by single leg balance assessment. Balance tests were capped at 60 s, with up to two attempts per position and the best effort reported if the first attempt did not achieve the 60 s of balancing [43].

To characterise habitual physical activity and assess intervention adherence, a wearable physical activity monitor (ActivPAL4, PAL Technologies ltd, Glasgow, UK) was worn on the thigh for the seven consecutive days after the baseline assessment. The device was mounted under an adhesive waterproof dressing so it could be worn continuously, including during water-based activity. Participants were provided with a pre-paid and addressed envelope in which to return the device following seven days of wear.

Intervention

The exercise snacking intervention consisted of five exercises, each performed for one minute interspersed with one minute of seated rest, with the aim being to complete as many repetitions as possible of each exercise in the given minute. The total length of each exercise ‘snack’ was therefore nine minutes. Participants were instructed to complete the exercises twice a day at home, and only to perform exercise snacks when there was someone else present in the home with capacity to call for help in the event of an emergency. The exercises were sit-to-stand from a chair, seated lateral arm raising to overhead, marching on the spot with high knees (aiming for thigh parallel to the floor), seated arm crossing from arms down by the sides to touching opposite shoulders, and seated calf raises. Participants were provided with a set of written and pictorial instructions on how to complete the exercises along with a logbook in which to record the number of sit-to-stand repetitions completed in each exercise snack, along with a rating of perceived exertion (1–10 scale) for the whole exercise snack. The logbook included space to record minor adverse events or notes from the participant, and clear instruction for reporting serious adverse events to study staff. This logbook was to be returned following the intervention to assess intervention adherence. See Additional file 1 for logbook with exercise instructions.

Statistical analysis

The primary outcome of acceptability of the exercise snacking intervention was assessed descriptively based on the scores from the TFA questionnaire, and qualitatively using framework analysis [44]. Distribution of continuous data from secondary functional outcome measures (60-second sit-to-stand, TUG, and standing balance tests) were assessed with Shapiro-Wilk normality test. The 60-second sit-to-stand data were normally distributed, so a paired T-test was used to assess differences between baseline and follow-up, and data presented as mean ± standard deviation. The standing balance and TUG data were non-normally distributed, so along with SPPB data (considered to be interval level data in the present study), these were assessed for differences from baseline to follow-up assessment with Wilcoxon matched pairs signed rank test, and data reported as median (IQ range). P-values for physical function outcomes are presented for additional context as to the tests that are of interest for further exploration at larger scale and should not be interpreted as conclusive. For secondary outcome measures, standardised effect sizes were calculated (Cohen’s dz [45]), and classed as small (0.2), moderate (0.5), and large (0.8) according to Cohen [46]. Secondary outcomes of cognitive function, patient reported health, wellbeing, and psychological process outcomes relating to exercise behaviour are reported descriptively as mean ± standard deviation. Statistical analysis performed using GraphPad Prism 9.5.0 for Windows (GraphPad Software, San Diego, California USA).

Results

Participant recruitment and retention

Thirty-one participants who attended the memory clinic were screened for inclusion in the present pilot study, of which 21 were included in the study and provided baseline data. The reasons for exclusion at the screening stage are included in the flow diagram in Fig. 1. Baseline characteristics of participants enrolled in the pilot study are displayed in Table 1. Of these, 18 completed the intervention and provided follow-up outcome data (of which one participant attended the follow-up 14 days after the intervention due to illness). Fifteen participants elected to complete the post-study qualitative interview. Reasons for study withdrawal included one participant who experienced a family bereavement, one participant who could not attend follow-up assessment due to logistical issues, and another participant who tragically lost their life owing to an accident unrelated to the study, recorded as an unrelated serious adverse event. Two of the three withdrawals occurred within 28 days of the end of the study timeframe, thus it was not feasible to include replacement participants for these two participants.

Fig. 1
figure 1

Recruitment flow diagram. MMSE; Mini-Mental State Examination, SPPB; Short Physical Performance Battery, SAE; Serious Adverse Event.

Full size image
Table 1 Participant characteristics of N = 21 participants enrolled in pilot study
Full size table

Completeness of data

Of the 18 participants who completed the follow-up assessment, all provided physical function, cognitive function, and health, wellbeing and psychological outcome data across all tests. Sixteen of the 18 participants returned their exercise snacking logbooks, which showed a reported adherence rate of 80% of all exercise snacks completed across the whole group. One participant elected to discontinue the exercise snacking intervention after one week but attended the follow-up assessment and the follow-up interview. Of the 15 participants who attempted 28 days of the intervention, reported adherence was 85%. Half of participants who returned their logbooks completed all exercise snacking sessions over the 28 days. Three accelerometers that were distributed at baseline (n = 21) were misplaced by participants and not returned to the research team, and a further three provided no data owing to technical faults with the device (n = 2) or were removed by the participant. Of the remaining 15 accelerometers, 11 returned complete seven days of valid data, and four had partially complete data (three with six days, and one with three days due to battery fault).

Acceptability

The quantitative scores indicated an overall acceptability rating from the exercise snacking intervention by completers of the study of 4.6 out of 5. The highest rated aspects of the TFA by study participants were self-efficacy (4.3) and affective response (4.1), followed by ethicality (3.9), coherence (3.8) and effectiveness (3.7). Burden (2.1) and opportunity cost (1.8) were scored low (scores closer to 1 represent a favourable appraisal), demonstrating strong all-round acceptability of the intervention. Similarly, the qualitative interviews indicated a generally positive view of the exercise snacking intervention and insight into some improvements that may enhance the acceptability further (Table 2). In sum, most participants found the programme simple to engage with and fit into their lives, enjoyed participating and recognised the benefits the programme was having on both their physical function and, for some, their mental and cognitive wellbeing, with many declaring confidence to keep up the exercise after the study. For more active participants the programme was considered tedious, while those with more severe cognitive impairment struggled to remember to do the programme every day.

Table 2 Qualitative analysis organised using the theoretical framework of acceptability domains, with implications for future implementation
Full size table

Exploratory analysis of outcome data

Changes were observed from baseline to follow-up for total SPPB score (8 (1) vs. 9 (3), p < 0.01, dz = 1.29), TUG (11.32 (4.02) vs. 9.18 (5.25) seconds, p < 0.01, dz = 0.77) and in the 60-second sit-to-stand test (17 ± 5 vs. 23 ± 7 repetitions, p < 0.01, dz = 0.74), with no significant difference in the RPE of the 60-second sit-to-stand tests (11 (3) vs. 12 (3), p = 0.36, dz = 0.22 ). Most participants reached 60-seconds in both baseline and follow-up assessments of semi- and tandem position balance tests, with no differences observed in median balance time between assessments. Single leg standing balance time on the left leg significantly increased (11.27 (16.83) vs. 20.33 (46.56) seconds, p < 0.01, dz = 0.61), with no change observed single leg standing on the right leg (18.15 (51.90) vs. 23.79 (34.78) seconds, p = 0.93, dz = -0.02). Figure 2 displays the individual data for the four physical function outcomes that saw improvements between baseline and follow-up. Table 3 displays the group level baseline and follow-up scores for all health and exercise cognitions outcomes.

Fig. 2
figure 2

Individual changes in (a) total Short Physical Performance Battery (SPPB) score, (b) 60-second sit-to-stand score, (c) timed up and go (TUG) time following 28-days of exercise snacking, and (d) left leg single balance time (to 60 s). *denotes a significant difference in median (SPPB, TUG, left leg single balance time) or mean (60-second sit-to-stand) from baseline to follow-up (p < 0.01)

Full size image
Table 3 Aggregated descriptive data for secondary outcomes in study completers (n = 18)
Full size table

Discussion

This study provides preliminary evidence that a simple homebased, twice daily exercise snacking programme appears to be acceptable to pre-frail older adult memory clinic attendees. Qualitative data from 15 of the 18 completers suggested that the simplicity of the exercises, and the short, simple to do nature of exercise snacking is a useful feature of this programme, with many non-exercising participants seeing value of continued practice of these exercises. Participants generally found the exercise snacking programme easy to implement, not burdensome, and worthwhile, and for participants who returned follow-up data we observed potential improvements in basic tests of physical function and single leg balance.

Together, these findings offer encouragement for the utility of exercise snacking as an intervention for inactive, older adult, memory clinic attendees, with a large effect size for improvement in the SPPB (dz = 1.29, and moderate effect sizes for improvements in the TUG and 60-second sit-to-stand (dz = 0.77 and 0.74 respectively). In a short period, this simple programme has demonstrated improvements in physical function that for a given individual, would present as a positive step towards preventing falls and maintaining independence [7, 12, 47]. Indeed, seven of the 18 participants who completed follow up assessment scored > 9 on the SPPB after the intervention, effectively moving them out the ’frail or pre-frail’ categories on this metric [48]. It should be noted that such physical assessments may be subject to a learning effect [49], and no familiarisation with physical assessments was undertaken prior to baseline, which along with the lack of control group means caution should be exercised when interpreting these change scores and effect sizes.

Self-reported adherence to the exercise snacking intervention was higher than in other exercise studies in people with cognitive impairment (80% of all participants who completed follow up assessment in the present study compared to typical adherence rates around 70%) [50]. Furthermore, while self-reported adherence could be subject to overestimation through social desirability bias, these data may represent an underestimate in adherence owing to participants forgetting to record completed exercise snacks in their logbooks (Table 2). Further studies would do well to deploy appropriate objective measures of capturing adherence, such as the recording of behaviour through technology. Other exercise trials targeting people with mild cognitive impairment have achieved similar improvements in TUG [51], and it is also encouraging that the increase in the sit-to-stand scores were similar to those previously observed implementing a similar exercise snacking intervention in healthy older adults (32% in present study (dz = 1.29) vs. 31% observed previously (Hedge’s g = 1.40) [21]).

Based on our data, exercise snacking may be of particular benefit for those who express common barriers to engagement with more comprehensive exercise provision or formats (such as group-based classes and intensive instructor led interventions), and certainly act as a gateway programme to improve baseline strength and balance and self-efficacy for regular exercise. The process outcomes displayed in Table 3 indicate some of the psychological mechanisms underpinning exercise participation that may potentially be modified by an exercise snacking intervention, namely exercise competence and outcome expectancies [52, 53]. Conversely, scores of other psychological needs and cognitions endorsed by self-determination and social cognitive theory did not change or went down (viz. barrier self-efficacy, autonomy and relatedness), suggesting that additional support may be beneficial for building lasting exercise behaviour change. Of course, given the single-group pilot nature of the present study this exercise snacking programme warrants further testing to evaluate its effectiveness, the processes of change, safety and long-term benefit.

The results indicate several aspects of the protocol and intervention that warrant further consideration prior to full scale testing to improve acceptability, such as the need to develop adaptations of the snacking exercises to make them easier or more challenging as required, or making the logging and performance of exercise snacks simpler and more habitual (Table 2). Furthermore, given challenges in retrieving full data from both exercise logbooks and accelerometers, strategies to aid data collections in cognitively impaired participants requires careful planning and perhaps further patient and public involvement [54].

Consistent with the other evaluations of exercise snacking in general non-patient populations [55], the simple exercise snacking format appears more acceptable to those who do not currently do exercise and ‘tedious’ for those who engage in more physically activity generally. Accordingly, it is proposed that studies evaluating exercise snacking in older adult populations (or practitioners implementing it) target non or low exercisers as those most suited to requiring and receiving such a programme. Consideration of building in progression so that the level of difficulty within the snacking format can be tailored to individual capabilities and exercise experience is also warranted. Secondly, self-monitoring the snacking using logbooks, while considered an important intervention ingredient, was problematic for some memory clinic patients who either forgot to do or log their exercise snacking sessions. Developing innovative yet intuitive ways to prompt the completion and recording of exercise snacks such as the use of digital technologies [56] or aligning to contextual habit-forming cues such as completing an exercise snack after a certain television programme has finished each day [57] may be useful developments of this protocol for those with cognitive impairment. Thirdly, although slight improvements in outcome expectancy was observed from baseline to follow-up, the qualitative acceptability data indicates more effort could be made to improve the rationale by which exercise helps minimise cognitive decline to increase engagement and adherence to the exercise snacking intervention.

There are also several limitations with the study design that warrant consideration. It should be emphasised that the single-group design and lack of a non-exercise control group means that we cannot conclusively state the observed improvements in physical function were a product of the exercise snacking intervention. Whilst the SPPB and TUG are well established with reported reference ranges, the 60-second strength and balance tests used in the present study lack data on their validity and reliability. Furthermore, the present study lacks mechanistic data to explain what physiological adaptations might have contributed to a change in function, for example improved maximum force or power generating capacity [58]. Likewise, the small sample size and analysis of just completers of the intervention means that any statistical results should be interpreted as exploratory and not conclusive. The observed high retention rate, completeness and direction of outcome data and largely positive appraisal of the intervention rate suggest that the study was feasible, acceptable, and potentially efficacious for a pre-frail, mild cognitive impairment population. However, the demographic and health data suggests that the recruited population was fairly homogenous with all participants living in their own home, 20/21 being white British, and having high quality of life and life satisfaction and low depression and anxiety scores (Table 3). Whilst this is likely to be indicative of the population characteristics in the geographical area this study clinic is based, we cannot assume these findings generalise to the wider older adult population in the UK and beyond. The short intervention and lack of further follow-up measurements also means we cannot infer whether any observed changes to the physical function, health or process data, nor adherence to exercise snacking, were sustained in participants.

Considering the findings of this study in light of these limitations, we conclude that the utilisation of a simple exercise snacking intervention could be an acceptable and beneficial way to improve the physical function of pre-frail memory clinic patients and warrants further evaluation to establish the efficacy of this approach. Whilst our data demonstrate an encouraging pattern of overall improvement in functional tests, this is not a robust study design to demonstrate efficacy of the intervention and further research employing a randomised controlled trial study design is required. Nonetheless, these initial findings suggest the exercise snacking format is acceptable, low burden, and engaging in this target population.