Associations between older adults physical fitness level and their engagement in different types of physical activity: cross-sectional results from the OUTDOOR ACTIVE study

Strengths and limitations of this study

A strength of this study is the objective assessment using validated methods of all primary outcomes (handgrip strength, lower muscle strength, aerobic endurance, lower flexibility, upper flexibility) and variables for adjusting (total physical activity and body mass index).

Engagement in specific types of physical activities was assessed using a self-reported questionnaire, which might have led to under-reporting of activities.

A selection bias is possible, however, as sensitivity analyses revealed similar results this probably did not have an impact.

The cross-sectional design of the study does not allow any conclusions regarding causality.

Background

As part of the demographic change, the proportion of older adults in the population is increasing. In Germany, almost one in four people is already 60 years or older.1 The demographic change has a massive impact on the healthcare system and is expected to impose a high economic burden on society.2 This demonstrates the importance of healthy ageing and sustaining physical independence into old age. One of the key determinants of physical independence is a high level of physical fitness (PF).3 The multidimensional construct of PF includes aerobic endurance, muscular endurance, muscular strength, body composition and flexibility.4 Previous research has shown a positive association between PF and health-related quality of life in older adults.5 Moreover, older adults with a low PF level have an increased risk of falling6 7 and frailty.8

Interindividual differences in PF levels are highly dependent on age and sex.9 The ageing process is accompanied by a reduction in muscle strength, aerobic endurance and flexibility.10 11 This leads to an overall decline in PF, which accelerates with age.3 PF differences regarding sex are well studied. Research shows that, in general, men reach higher values in strength and endurance dimensions while women are more flexible.10 12 Furthermore, genetic factors13 14 and dietary intake15–17 appear to be relevant for PF. The main behavioural determinant of PF is physical activity (PA). From a theoretical perspective, an increase in PA leads to an adaptation of the body, resulting in improved PF.18 In older adults, PA is indeed associated with the maintenance and increase of PF11 and older adults who engage in regular PA show a higher level of PF compared with their rather inactive counterparts.19 20 PA cannot completely prevent the reduction of PF through ageing,21 but it can considerably reduce the negative trend.3 While any engagement in PA is better than none,11 the extent of the impact is dependent on PA intensity, frequency and duration. Moreover, specific types of PA are relevant for different PF dimensions. For example, studies found that dancing improved muscle strength, aerobic endurance and flexibility,22 23 Pilates24 and water aerobics25 improved muscle strength and flexibility, cycling26 and Nordic walking27 improved muscle strength and aerobic endurance and soccer improved muscle strength28 in older adults. These findings are based on experimental data and therefore, activities are performed in a rather controlled and artificial setting. Observational evidence for the impact of habitual engagement in a wide range of physical activities on PF dimensions in older adults is currently missing.

Simultaneously, the literature on the prevalence of specific types of PAs in older adults in Germany is scarce. Previous studies focused mainly on engagement in different domains of PA.29 30 Moschny et al and Stalling et al report that most older adults engage in home-based activities (79.1%–86.6%).29 30 While the prevalence of engagement in activities for transport is similar (approximately 84.0 %),30 less older adults engage in leisure activities (50.0–66.3%).29 30

The present study addresses these gaps and can elucidate if the relationship between types of PAs older adults regularly participate in and PF dimensions are consistent with previous experimental findings. Results may also help to specify PA recommendations to improve their potential in sustaining and increasing PF in older age.

This quantitative study aims (1) to describe the prevalence of different types of PAs and (2) to explore the association between engagement in these PAs and performance in the PF dimensions among older adults living in Bremen, Germany.

MethodsStudy design and population

AEQUIPA (Physical activity and health equity: primary prevention for healthy ageing) is a prevention research network that investigates the role of PA as a key determinant of healthy ageing.31 OUTDOOR ACTIVE is one of the six subprojects of AEQUIPA. It aimed to develop and implement a community-based outdoor PA promotion programme using a participatory approach.32 33 OUTDOOR ACTIVE was based in the Free Hanseatic City of Bremen, which is located in northwestern,1 Germany. As of 2021 Bremen had around 560 000 inhabitants with 21.1% being 65 years or older.34 The city is divided into 88 subdistricts, which are highly heterogeneous regarding their history, land-use mix and socioeconomic situation of inhabitants.35 With approximately 25 %, Bremen is the major city with the highest share of bicycle traffic in Germany and ranks third in Europe.36 Moreover, Bremen is considered the most bicycle friendly city of Germany.37 In 2014, 22% of the population aged 60 years and older were members of at least one of the 355 sports clubs.38

OUTDOOR ACTIVE was carried out in two stages: the pilot study (February 2015–January 2018) and the cluster-randomised trial (February 2018–December 2022). The eligibility criteria for both stages were (1) being between 65 and 75 years of age, (2) being non-institutionalised and (3) living in specific subdistricts in Bremen, Germany (pilot study: Arbergen, Hastedt, Hemelingen, Mahndorf, Sebaldsbrück; cluster-randomised trial: Blumenthal, Burg-Grambke, Gete, Lehe, Lehesterdeich, Neustadt, Ohlenhof, Ostertor). For the recruitment of participants in both stages, address data from all residents in the selected subdistricts aged 65–75 years were obtained from the registry office of Bremen. Eligible individuals were initially contacted by letter, followed by a phone contact if the number was listed in one of the available registers.

The baseline assessment consisted of (1) a questionnaire on intrapersonal, interpersonal and environmental determinants of PA, (2) a health examination including a short physical examination and a fitness test and (3) a 7-day accelerometer measurement. The follow-up assessment took place 1–3 years after baseline, yet, the data were not included in the following analyses. Participants were free to choose in which parts of the survey they wanted to partake.

In total, 11 079 individuals meeting the age criteria were registered in the study regions of the pilot study and the cluster-randomised trial. Of those, 125 are deceased and another 461 were not able to participate due to acute health problems. A total of 450 individuals moved outside the study region and 77 were not able to participate because of language barriers. Of the remaining 9966 confirmed eligible individuals, 3425 were never reached after the initial invitation letter and 4247 refused to participate. Furthermore, 151 individuals of the subdistrict Lehesterdeich were never contacted because the end of the Lehesterdeich survey period was reached and the actual sample size of the subdistrict already exceeded the calculated sample size. A total of 2143 individuals participated in at least one part of the pilot study or the cluster-randomised trial. All participants who filled out the questionnaire and completed at least one PF measurement were included in the following analyses (n=1583).

All participants provided written informed consent. Both the pilot study (number: 2015–6) and the cluster-randomised trial (number: 2018–06) were approved by the ethics committee of the University of Bremen.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.

MeasurementsPhysical activity

A self-developed questionnaire, based on the German Physical Activity Questionnaire 50+,39 was used to assess participants’ regular engagement in specific types of PA. In a first step, participants were asked if they engaged in home-based activities (housework, gardening) and activities for transport (cycling, walking). Specific types of housework and gardening were not assessed, but examples for these activities were given (housework: eg, vacuum cleaning, doing the dishes, repair works; gardening: eg, weeding, planting, digging). In a second step, participants were asked to list all further regular PAs using an open text format. Moreover, they were asked to state the duration of engagement in all respective activities (in hours per week) for summer and winter, which were later summarised by calculating the mean. All mentioned activities were coded and then categorised based on similarity following an inductive approach. Two main categories of activities emerged: leisure activities and sports therapy (eg, specific sport groups for individuals with cardiovascular or pulmonary diseases). Leisure activities were further categorised into the following groups: hiking/running, gym training, aerobics, water aerobics, swimming, dancing, ball sports and other sports. The ‘other sports’ group included leisure activities that were only rarely mentioned and did not fit in any of the other groups (eg, boating, martial arts, horseback riding).

Total PA was objectively assessed via accelerometry. Participants were asked to wear an ActiGraph wGT3X-BT accelerometer (ActiGraph LLC, Pensacola, Florida, USA) on their non-dominant wrist for seven consecutive days at day and night.40 41 Sampling frequency was set to 30 Hz.41 Accelerometer data were downloaded and processed using ActiLife (V.6.13.3, ActiGraph LLC, Pensacola, Florida, USA).

Physical fitness

Handgrip strength was measured using a Saehan DHD-3 digital hand dynamometer SH1003 (Saehan Corporation, Changwon, South Korea). The measurement was conducted in a standing position, upper arm close to the upper body and elbow flexed in a 90° angle. Maximum isometric strength was measured twice for both hands and the overall maximum was used for the analyses.42

All other PF measurements were conducted according to the test protocols in the Senior Fitness Test Manual, 2nd edition.43 Lower muscle strength was measured with the 30 s-chair stand test. For this, the participant must stand up from a seated position and sit back down as often as possible in 30 s. The 2 min-step test was conducted for assessing aerobic endurance. This test requires the participant to step in place for 2 min with both knees reaching a required height. The sit-and-reach test was chosen to measure lower body flexibility. For this test, the participant sits on a chair with one leg extended and must reach towards his toes. Upper body flexibility was determined using the back scratch test. Participants are asked to try to touch the middle fingers of both hands behind the back with one hand reaching over the shoulder and the other up the middle of the back. For both flexibility tests, the distance (negative values) or overlap (positive values) was measured.

Anthropometry

Basic anthropometry was taken without shoes and with all outer garment removed. Body weight was measured with a Kern MPC 250K100M personal floor scale (Kern & Sohn GmbH, Ballingen, Germany). Standing height was measured with the head in a Frankfort plane44 with a Seca 217 mobile stadiometer (Seca GmbH & Co. KG, Hamburg, Germany).

Other variables

Information on educational years, net household income, employment history, medication intake and sports membership were collected in the questionnaire. Educational years (school years and training years combined), net household income and employment history were used to determine participants’ socioeconomic status (for details, see Stalling et al30). Participants were asked to list all medications they take on a daily basis. This list was used to determine the individual sum of daily medications. The questionnaire also asked whether the participants were members of a sports club, gym, or sports group.

Statistical analyses

For determining the body mass index (BMI), the quotient of body weight (in kg) and the squared height (in m) was calculated. BMI was classified into underweight (<18.5 kg/m2), normal weight (18.5 to <25 kg/m2), overweight (25 to <30 kg/m2) and obesity (≥ 30 kg/m2).45

Accelerometer data were reintegrated to epoch lengths of 60 s. Non-wear time was excluded based on the algorithm by Troiano et al.46 Participants had to wear the accelerometer for at least four valid days (10 hours per day).41 The mean of vector magnitude counts per minute (VM counts) for all valid days was calculated to be included in the analyses.

Performance in the PF measurements was classified according to previously published sex-specific and age-specific normative values.12 Participants’ scores falling within the first quartile were defined as being under norm, scores in the second and third quartile as in norm and scores in the fourth quartile as over norm.

For descriptive statistics, absolute and relative frequencies were determined for engagement in different PAs. For those individuals participating in an activity, medians and IQRs of engagement duration (in hours per week) for the respective activity were calculated. Absolute and relative frequencies were also calculated for BMI, socioeconomic status and membership in a sports club, gym or sports group. Medians and IQR were determined for participants’ age, all PF dimensions, total PA and daily medications. These analyses were done stratified by sex. To investigate the association between engagement in different types of PA and level in the PF dimensions, first, relative frequencies of participants in or over norm in the PF dimensions were determined for all PAs. In a second step, logistic models were fitted. All variables regarding engagement in PAs (binary: yes/ no) were included as independent variables in the models. Being in or over norm in the PF dimensions was included as outcome variable in separate models. Logistic models were adjusted for total PA (in average vector magnitude counts per minute) and BMI (in kg/m2) as both variables are highly associated with the independent and dependent variables. Age, sex and number of daily medications were not included for adjustment, as these variables did not affect the models. All selected variables were simultaneously inserted in the models (‘Enter’ method in SPSS Statistics). Only participants with accelerometer data were included in the regression analyses. A sensitivity analysis without adjustment for total PA was conducted. All analyses were performed in SPSS Statistics V.22.0 (IBM Corp., Armonk, NY, USA).

Results

The characteristics of the 1583 participants (53.1 % women) are displayed in table 1. The median age was 69.0 years (IQR: 5.0 years) for both women and men.

Table 1

Characteristics of the study population, stratified by sex (n=1583)

The most common types of regular PA were home-based activities (overall: 97.9% of women vs 95.1% of men; housework: 96.7% of women vs 91.4% of men; gardening; 73.5% of women vs 76.1% of men) and activities for transport (overall: 96.3% of women vs 96.2% of men; walking: 85.9% of women vs 83.6% of men; cycling: 80.1% of women vs 88.8% of men). 69.0% of women and 62.3% of men participated in leisure activities. In women, the most common leisure activities were aerobics (37.9 %), hiking/ running (19.6 %), and gym training (18.2 %). In men, the most common leisure activities were gym training (25.3 %), ball sports (17.3 %), and aerobics (15.9 %). Among those who participated in the respective activities, the duration in activity was higher for home-based activities and activities for transport compared with leisure activities and sports therapy.

The median handgrip strength of women was 25.2 kg (IQR: 6.6 kg) and of men 42.3 kg (IQR: 10.5 kg). Regarding lower muscle strength, women reached 13.0 n in 30 s (IQR: 4.0 n in 30 s) and men reached 13.0 n in 30 s (IQR: 4.0 n in 30 s) in median. In the 2 min-step test for aerobic endurance, women scored 87.0 n in 2 min (IQR: 25.0 n in 2 min) and men scored 88.0 n in 2 min (IQR: 22.0 n in 2 min) in median. The median result for lower body flexibility was 3.3 cm (IQR: 10.0 cm) in women and −2.0 cm (IQR: 14.5 cm) in men. Regarding upper body flexibility, the median result of women was −3.0 cm (IQR: 12.5 cm) and of men −11.0 cm (IQR: 20.0 cm).

For total PA estimation, women reached 1794.6 VM counts (IQR: 563.6 VM counts) and men reached 1482.1 VM counts (IQR: 518.0 VM counts). In median, both women and men took one medication per day (IQR: 3.0 medications). About 42.7% of women and 27.8% of men had normal weight, 56.2% of women and 72.2% of men were overweight or obese. With regard to socioeconomic status, most participants belonged to middle class (women: 61.5 %, men: 60.9 %). More men (26.1 %) than women (18.1 %) were part of the upper class. About 70.1% of women and 59.9% of men had a membership in either a sports club, gym or sports group.

Table 2 displays the relative frequencies of participants classified as in or over norm in the separate PF dimensions by PAs. The highest percentages of participants in or over norm were seen in those regularly participating in hiking/running, dancing and ball sports. In contrast, water aerobics, sports therapy, and housework had the lowest percentages of participants in or over norm.

Table 2

Relative frequencies of participants in or over norm in physical fitness dimensions by engagement in physical activities (n=1583)

Table 3 shows the regression results for the association between engagement in PAs and being in or over norm in the PF dimensions, adjusted for total PA and BMI. Being in or over norm in handgrip strength was positively associated with cycling (OR: 1.56, 95% CI: 1.13 to 2.15), hiking/running (OR: 1.50, 95% CI: 1.05 to 2.16) and other sports (OR: 3.22, 95% CI: 1.37 to 7.56). Being in or over norm in lower muscle strength was positively associated with cycling (OR: 1.91, 95% CI: 1.37 to 2.65), gym training (OR: 1.62, 95% CI: 1.16 to 2.26) and dancing (OR: 2.15, 95% CI: 1.00 to 4.61). Being in or over norm in aerobic endurance was positively associated to cycling (OR: 1.90, 95% CI: 1.37 to 2.65), gym training (OR: 1.68, 95% CI: 1.20 to 2.36), aerobics (OR: 1.64, 95% CI: 1.19 to 2.26), dancing (OR: 2.62, 95% CI: 1.10 to 6.22) and ball sports (OR: 2.07, 95% CI: 1.30 to 3.29). Being in or over norm in lower body flexibility was not significantly associated with any of the PAs. Being in or over norm in upper body flexibility showed a negative association with housework (OR: 0.39, 95% CI: 0.19 to 0.78).

Table 3

Association between physical fitness dimensions and engagement in different physical activities; adjusted for total physical activity and body mass index (n=1400)

Discussion

In this study on older adults in Germany, home-based activities (housework, gardening) and activities for transport (walking, cycling) were performed by nearly all the participants, while leisure activities were less prevalent. The most common leisure activities were aerobics, hiking/ running and gym training in women and gym training, ball sports and aerobics in men. Several specific types of PAs were associated with muscle strength and aerobic endurance, independent of total PA and BMI. Apart from housework, none of the investigated activities was associated with flexibility dimensions.

Muscle strength dimensions were positively associated with specific activities. Older adults engaging in cycling or hiking/running had a 56% or 50% increased chance of being in or over norm in handgrip strength, respectively. Both activities are not intuitively associated with handgrip strength. A randomised-controlled trial on an indoor cycling intervention in older adults did indeed reveal no significant effects on handgrip strength.47 However, comparability between indoor and outdoor cycling is limited, at least for recreationally trained48 and professional cyclists.49 Outdoors, cyclists must constantly adapt to the changing environment (eg, uneven ground, other road users). This could lead to a greater strain on handgrip strength when cycling outdoors compared with when cycling indoors. For lower muscle strength, associations with gym training, cycling and dancing were statistically significant with a 62%, 91% and 115% increased chance of being in or over norm, respectively.

With the exception of cycling, aerobic endurance only showed positive associations with leisure activities. These associations were stronger for activities that typically have a high impact on the cardiorespiratory system (dancing, ball sports) compared with activities that often combine resistance and endurance exercise (gym training, aerobics). In contrast, a systematic review on same-session resistance and endurance training reported that combined training might be slightly more beneficial for maximal aerobic capacity than endurance training alone.50 Non-leisure activities might not be strenuous enough to initiate endurance improvements. In line with that, a study by Silva et al reported that only moderate-to-vigorous PA was positively associated with aerobic endurance but not PA at light intensity.51 Another study found that while older adults engaged in lower amounts of sports activity than habitual activity, the relationship between sports activity and overall PF was substantially higher than between habitual activity and overall PF. The authors propose that the unsystematic character of habitual activity is not sufficient for PF benefits.9

Older adults who engaged in housework had a 61% decreased chance of being in or over norm in upper body flexibility. Otherwise, flexibility dimensions did not show any significant associations with the PAs. Research shows that flexibility can be improved even in older age by specific exercises, such as yoga,52 Pilates,24 53 54 Tai Chi55 and dancing.22 23 56 57 The reason why we did not observe similar results in our study remains unclear. Housework was negatively associated with upper body flexibility. With 96.7% of women and 91.4% of men, the vast majority of the study population engaged in housework. This indicates that functionally impaired adults who might not engage in many further activities, still engage in housework.

From a public health perspective, the potential of specific PAs for health promotion of a population is not only dependent on the extent of PF benefits but also on the prevalence of these activities.58 In our study, most participants regularly engaged in home-based activities and activities for transport. This has also been shown in other studies.29 59 60 Even though the prevalence was high, home-based activities do not seem suitable for improving PF in older adults as these activities were not positively associated with any of the PF dimensions. This is also the case for walking as one of the activities for transport. On the other hand, cycling was positively associated with several PF dimensions which—together with the high prevalence—highlights its potential for health promotion in older adults. While the prevalence of cycling in older adults was high in Bremen, this is not the case for every region. As Bremen is considered to be the most bicycle-friendly city in Germany,37 improving the traffic situation for cyclists is a promising approach to increase the prevalence of cycling and, therefore, the PF level of the population.

Overall, specific leisure activities were less prevalent than home-based activities and activities for transport. Among leisure activities, the highest prevalence was observed for predominantly fitness-oriented and health-oriented activities (eg, hiking/running, gym training, aerobics). These were also associated with muscle strength dimensions and aerobic endurance, indicating moderate prevention potential. Dancing had the overall highest ORs, even though not reaching statistical significance in all PF dimensions. However, prevalence was quite low with 6.1% of female and 3.6% of male participants. The promotion of dancing for older adults could increase its prevention potential for the population.

The present study has several limitations that need to be addressed. Results were derived from a cross-sectional study. Therefore, the directions of associations are unclear and longitudinal studies are required for statements regarding causality. Since participants were free to choose in which parts they wanted to partake, a selection bias is possible. In the questionnaire, participants of the PF test were less likely to report poor subjective health compared with the other survey participants (14.7% vs 26.4 %). This is a well-known limitation in research studies.61 At the same time, a comparison of participants who wore and those who did not wear an accelerometer showed a comparable proportion regarding poor subjective health (14.6% vs 14.2 %). Sensitivity analyses of the models without adjustment for total PA revealed similar results (see online supplemental table S1). Also, controlling for age, sex or number of daily medications did not affect the models, which is why we chose not to include them for adjustment. Engagement in PAs was assessed via self-reported questionnaire. This might have led to under-reporting of activities. However, this enabled us to analyse specific activities and not only PA frequency, amount and estimated intensity as usually assessed via objective methods. Besides the PAs, only objective measurements were included in the logistic models.

Conclusions

This study provides an overview of the prevalence of older adults’ engagement in different types of PAs. Moreover, study results indicate that the association between engagement in PAs and PF is dependent on the observed PF dimension. While muscle strength dimensions and aerobic endurance were positively associated with engagement in several PAs, flexibility dimensions were associated with almost none of the investigated activities. Overall, results indicate that home-based activities and walking are not strenuous enough to provide sufficient stimulus for PF improvements in older adults. Instead, the potential of cycling and leisure activities (eg, hiking/running, gym training, aerobics, dancing) for health promotion in older adults is highlighted. These findings should be considered for the specification of PA recommendations and the development of PA promotion programmes for older adults.

Data availability statement

Data are available upon reasonable request. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics statementsPatient consent for publicationEthics approval

The OUTDOOR ACTIVE study was approved by the ethics committee of the University of Bremen (pilot study: 2015-6; cluster-randomised trial: 2018-06). Participants gave informed consent to participate in the study before taking part.

留言 (0)

沒有登入
gif