The mediating role of muscle strength and body composition on the association between movement behaviours and health-related quality of life in Spanish older adults: the IMPACT65 + study

Abstract

Physical activity can improve health-related quality of life (HRQoL); however, it is not clear whether this association is direct or indirect. The aims of this study were to examine the associations of daily movement behaviours with physical and mental HRQoL in older adults, and the mediating role of muscle strength and body composition on it. Three hundred thirty-four community-dwelling older adults wore the Intelligent Device for Energy Expenditure and Activity for two consecutive days to quantify time spent in sedentary behaviour (SB), light physical activity (LPA), and moderate-to-vigorous physical activity (MVPA). HRQoL was assessed using the 12-Item Short-Form Health Survey, yielding physical (PCS) and mental (MCS) component scores. For muscle strength, a handgrip dynamometer and the Sit-to-Stand test were used. Body mass index (BMI), waist circumference, and estimates of skeletal muscle mass and fat mass by bioelectrical impedance were included as body composition variables. Association analyses were examined using linear regression models, and for mediation analysis, structural equation modelling was used. MVPA was positively associated with PCS [β (95%CI): 0.39 (0.19, 0.59)], whereas no significant associations were observed between MVPA, LPA, or SB and MCS. The association between MVPA and PCS was partially mediated by lower-limb muscle strength and fat mass percentage [8.5% (z = 1.753, p = 0.080) and 28.9% (z = 1.912, p = 0.056), respectively], and fully mediated by skeletal muscle mass normalized by BMI [28.2% (z = 2.016, p = 0.044)].Regular engagement in MVPA is positively associated with the physical component of HRQoL in older adults, both directly and indirectly through muscle strength and skeletal muscle mass.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s11357-026-02121-1.

Introduction

Ageing is a biological process that is progressing more rapidly now than ever before. By 2050, one in six people will be aged over 65 years, and by the late 2070 s, the global population of adults aged ≥ 65 years is projected to surpass that of children [1, 2]. These changes in longevity generate a great need to focus on promoting healthy ageing to prepare for the increase in older adults’ population [1].

Ageing is linked to numerous detrimental conditions, including multimorbidity, functional limitations, and cognitive impairment, all of which adversely affect health-related quality of life (HRQoL) in older adults [3–7]. Furthermore, considering that this population spends most of their time in sedentary behaviours (SB) [8, 9] which are associated with a higher risk of morbidity and mortality [10, 11], the prognosis is not encouraging. Conversely, an expanding body of evidence has confirmed that regular physical activity (PA) can be an effective, non-pharmacological, therapy-like approach to improve a wide range of conditions that impair physical and mental aspects of HRQoL [12–15]. On the other hand, muscle strength and body composition, as health-related fitness components [16–19], exert a notable impact on overall wellbeing in older adults and are also consistently related to higher levels of mental and physical health status in this population [1, 7, 20, 21].

Traditionally, the beneficial association between PA and health outcomes has been explained by the effect that this behaviour has on health-related fitness components, and how this generates a health status enhancement [16, 22]. Nowadays, the possibility that regular PA may have a direct effect on health outcomes is being considered [14]. Thus, regarding the HRQoL of older adults, this uncertainty about whether the effect of PA is direct or indirect could be rectified using activity monitors to assess daily physical activity characteristics instead of self-reported instruments to avoid its limitations [10, 23, 24]. Therefore, the aims of this study were (i) to examine the independent associations of pattern-recognition-derived daily movement behaviours with the physical and mental components of HRQoL in older adults, and (ii) to assess the mediating role of muscle strength and body composition in these associations.

Methods

Design and participants

A cross-sectional study was conducted with 334 high-functioning community-dwelling older adults, aged 65 to 92 years, participating in the IMPACT65 + study, which aims to examine the association of objectively measured PA with frailty, HRQoL, and other health indicators in Spanish adults aged 65 and older [25–29]. Data collection was carried out from April 2015 to June 2017. From an initial cohort of 607 participants, 175 with missing or insufficient accelerometer data were removed. An additional 98 individuals with incomplete information on health outcomes, mediator variables, or covariates were excluded, yielding a final analytical sample of 225 women and 109 men with complete data. Furthermore, we considered a subsample of 192 participants (117 women and 75 men) with bioelectrical impedance data to perform secondary body composition analysis (Online resource 1). The Ethics Committee of the Autonomous University of Madrid (Madrid, Spain) approved the study. All participants were informed of the nature of the study and provided written informed consent prior to data collection.

Movement behaviours

The multisensor, pattern-recognition Intelligent Device for Energy Expenditure and Activity (IDEEA, Minisun, Fresno, CA, USA), which has been previously validated for accuracy and reliability [30, 31], was used in this study to assess movement behaviours. The IDEEA consists of a small and lightweight data recorder device (70 × 44 × 18 mm, 59 g) that can be worn on the belt and connected by wires and wirelessly to five sensors, which are biaxial accelerometers detecting angular acceleration and displacement in two orthogonal planes. Acceleration and position information obtained from the sensors is integrated into the microprocessor within the recorder device. Participants were asked to wear the IDEEA for two consecutive days [30] without removing the device. The sensors were adhered with medical tape to the anterior sternum (just below the sternal angle), the anterior side of each thigh (midpoint between the anterior superior iliac spine and knee joint), and the plantar surface under each foot (under the foot arch). Devices were fitted and removed by trained personnel and were initialized and individually calibrated following the IDEEA software instructions [27]. For this study, sleep and awake periods were identified using a validated automated algorithm [26] and all movement and posture recorded were merged into major daily activities (i.e., lying, sitting, standing, walking at different paces, and jumping, among others), which, in turn, were classified into three common intensity categories (Online resource 2): (i) SB; (ii) light physical activity (LPA); and (iii) moderate-to-vigorous physical activity (MVPA). Valid data for analysis were defined as at least one day of wear time with a minimum of 10 waking hours [32].

Health outcomes

HRQoL was measured using the 12-Item Short-Form Health Survey. This self-reported instrument yields two composite scores: (i) the Physical Component Summary (PCS), encompassing general health, physical functioning, role-physical, and bodily pain; and (ii) the Mental Component Summary (MCS), which includes vitality, social functioning, role-emotional, and mental health [33]. The psychometric properties of both component scores have been previously confirmed in the Spanish population, showing satisfactory reliability and construct validity [34].

Health-related fitness components

Muscle strength

The handgrip strength (HGS) was measured with a TKK 5101 grip dynamometer (Takey, Tokyo, Japan) using the non-dominant hand. The handgrip was adjusted for each participant's hand size (i.e., with the index finger’s middle phalanx positioned at 90°). For the test, participants were instructed to remove rings from the evaluated hand and to keep the elbow extended, avoiding any contact between the body and the dynamometer. Participants pressed the handgrip as strongly as possible for three seconds. The test was performed twice on non-consecutive trials, and the highest value (kg) was considered for analysis [35]. For lower limbs' muscle strength, the Sit-to-Stand test was used. Participants were instructed to stand up from a chair and sit down repeatedly for 30 s. The test started with the participant seated with the back straight, feet on the ground, and arms crossed on the chest. During the performance, participants had to achieve complete knee extension when standing and then return to a seated position. The total number of repetitions performed in 30 s was recorded for analysis [36].

Body composition

Height and weight were measured by a stadiometer (Seca 225) and a scale (Seca 861), respectively. Body mass index (BMI) was calculated as weight (kg) divided by squared height (m²). Waist circumference (WC) was measured (cm) with a nonelastic tape at the midpoint between the lowest rib and the iliac crest after normal expiration [28].

Skeletal muscle mass (SMM) was estimated using the equation developed by Janssen and colleagues [37]:

$$SMM[kg]=\left(\frac{{\mathit{Height}}^2\left[{\mathit{cm}}^2\right]}{resistance\left[o,h,m,s\right]},,\ast ,0.401\right)+\left(S,e,x,,\ast ,3.825\right)+(Age[years]\ast -0.071)+5.102$$

where sex was coded as 0 for women and 1 for men. Resistance data and fat mass (relative to total body weight) were obtained from bioelectrical impedance analysis (Tanita BC418MA).

Covariates

Sex, age, and the highest educational level attained were recorded with participants classified as low (i.e., illiterate or primary studies) or high (i.e., secondary education, vocational training, university studies) education. Tobacco and alcohol consumption were reported as current, former, or never. Information on the following chronic conditions, diagnosed by a physician and self-reported by participants, was also recorded: coronary heart disease, stroke, rheumatism, hip fracture, diabetes, cancer at any site, Parkinson’s disease, and dementia/Alzheimer’s disease. Responses were tabulated into two categories: having one or more chronic conditions or having none.

Statistical analyses

Descriptive characteristics of the study sample are presented as means and standard deviations (SD) for continuous variables and as percentages for categorical variables. Pearson’s correlations were used to assess bivariate relationships between study variables. Preliminary analyses showed no significant interaction effect for sex; therefore, men and women were treated as a whole sample in order to increase the statistical power.

Associations of daily movement behaviours (i.e., SB, LPA, and MVPA) and potential mediators (i.e., muscle strength and body composition indicators) with health outcomes were examined using multiple linear regression models. Three partition models were fitted with progressive adjustment for potential confounders. Model 1 was adjusted for age (years), sex (woman, man), and device wear time (h/day). Model 2 was further adjusted for educational level (low, high), tobacco and alcohol consumption (current, former, never), and chronic conditions (none, one or more). Model 3 included all covariates in model 2 and additionally controlled for time spent in MVPA (h/day) when SB and LPA were the exposures, and for time spent in SB (h/day) when MVPA was the exposure. Associations were reported as unstandardized regression coefficients (B) and 95% confidence intervals (95% CI).

To examine the mediating role of muscle strength and body composition variables in the association of daily movement behaviours with HRQoL, structural equation modelling (SEM) was performed. Thus, Equation c regressed the dependent variable on the independent variable. Equation a regressed the mediators (i.e., HGS, lower limb muscle strength, muscle mass, BMI, WC or fat mass percentage) on the independent variable. Equation b regressed the dependent variable on the mediator, and equation c’ regressed the dependent variable on the independent variable while adjusting for the mediator to account for the indirect (mediating) pathway. Mediation was established when coefficients in Equations a, b, and c were statistically significant, and when the direct association between the independent and dependent variables was attenuated (partial mediation) or became nonsignificant (full mediation) after inclusion of the mediator (equation c’). Thus, the indirect effect was computed as the product of the unstandardized coefficients from Equations a and b, and its statistical significance was tested using the Sobel test. The proportion of the total effect explained by mediation was calculated as (a × b)/c. Bootstrap resampling with 5000 iterations was applied to obtain bias-corrected 95% CIs for the indirect effects. Mediation analyses were adjusted as Model 3 described above. All analyses were performed using STATA version 15.1 (StataCorp LLC, College Station, TX, USA), setting the significance level at p < 0.05.

Results

The main characteristics of the study participants are presented in Table 1 (n = 334). Most individuals had low educational levels (54.49%) and no chronic condition (56.89%). Moreover, most participants had never consumed tobacco (55.69%) but currently consumed alcohol (66.17%). The mean device wear time was 23.21 h/day, with 8.34 h/day spent sleeping. During waking hours, the most common behaviour was SB (7.15 h/day), while MVPA was the least prevalent (0.79 h/day).

Table 1.

Descriptive statistics of the study sample

	All
n	334
Age, years (SD)	71.77 (5.27)
Sex, n (%)
Women	225 (67.37)
Men	109 (32.63)
Educational levela, n (%)
Low	182 (54.49)
High	152 (45.51)
Tobacco consumption, n (%)
Current	19 (5.69)
Former	129 (38.62)
Never	186 (55.69)
Alcohol consumption, n (%)
Current	221 (66.17)
Former	11 (3.29)
Never	102 (30.54)
Chronic conditions, n (%)
None	190 (56.89)
One or more	144 (43.11)
Weight, kg (SD)	71.28 (12.83)
Height, cm (SD)	157.30 (8.70)
BMI, kg/m2 (SD)	28.80 (4.67)
Waist circumference, cm (SD)	101.47 (10.99)
IDEEA-derived variables, h/day (SD)
Wear time	23.21 (1.34)
Sleep time	8.34 (1.46)
Sedentary behaviours	7.15 (1.73)
Light PA	6.93 (2.04)
MVPA	0.79 (0.58)
HRQoL, Score (SD)
SF-12 PCS	−0.47 (0.99)
SF-12 MCS	0.11 (0.99)

Values are mean (SD) or percentages

^aLow: illiterate or primary studies; high: secondary studies, medium university studies or vocational training, or higher university studies

Abbreviations: BMI, body mass index; IDEEA, intelligent device for energy expenditure and activity; PA, physical activity; MVPA, moderate-to-vigorous PA; SF-12 PCS, physical health-related component summary score of the 12-Item Short-Form Health Survey; SF-12 MCS, mental health-related component summary score of the 12-Item Short-Form Health Survey

Bivariate correlations among study variables ranged from − 0.43 to 0.73; the strongest relationships were observed between MVPA and PCS, as well as between MVPA and most health-related fitness component variables (Online resource 3). PCS was positively associated with total time spent in MVPA [B (95%CI): 0.394 (0.194, 0.594)]; however, there was no significant association between total time spent in MVPA, LPA, or SB and MCS (Table 2). All health-related fitness component variables, except for HGS, were significantly related to PCS (Table 3). Lower limb muscle strength and skeletal muscle mass normalized by BMI were directly associated with PCS [B (95%CI): 0.051 (0.019, 0.083) and 1.958 (0.608, 3.309), respectively], whereas BMI, WC, and fat mass percentage showed inverse associations [B (95%CI): − 0.030 (− 0.053, − 0.008); − 0.012 (− 0.022, − 0.002), and − 0.029 (− 0.052, − 0.006), respectively]. On the other hand, only WC was directly associated with MCS [B (95%CI): 0.012 (0.002, 0.021)].

Table 2.

Independent associations of pattern-recognition-measured movement behaviours with the health outcome variables

	Model 1 (95% CI)	Model 2 (95% CI)	Model 3 (95% CI)
SF-12 PCS
Sedentary Behaviour	− 0.03 (− 0.10, 0.04)	− 0.03 (− 0.10, 0.04)	0.01 (− 0.06, 0.07)
Light physical activity	− 0.01 (− 0.06, 0.05)	− 0.01 (− 0.07, 0.05)	− 0.00 (− 0.06, 0.05)
Moderate-to-vigorous Physical activity	0.41 (0.22, 0.60) ***	0.39 (0.20, 0.58) ***	0.39 (0.19, 0.59) ***
SF-12 MCS
Sedentary Behaviour	− 0.00 (− 0.07, 0.06)	− 0.01 (− 0.08, 0.06)	0.00 (− 0.07, 0.07)
Light physical activity	− 0.04 (− 0.09, 0.02)	− 0.03 (− 0.09, 0.02)	− 0.03 (− 0.09, 0.05)
Moderate-to-vigorous Physical activity	0.10 (− 0.10, 0.29)	0.11 (− 0.09, 0.31)	0.11 (− 0.10, 0.32)

Values are unstandardized regression coefficients (95% Confidence Interval)

Model 1 was adjusted for sex (women, men), age (years), and device wear time (h/day); Model 2 was adjusted as model 1, plus educational level (low, high), tobacco consume (current, former, never), alcohol consume (current, former, never) and chronic conditions (none/one or more); Model 3 was adjusted as model 2, plus time (h/day) in moderate-to-vigorous physical activity for sedentary behaviour and for light physical activity, and plus time (h/day) in sedentary behaviour for moderate-to-vigorous physical activity

Abbreviations: SF-12 MCS, mental health-related component summary score of the 12-Item Short-Form Health Survey; SF-12 PCS, physical health-related component summary score of the 12-Item Short-Form Health Survey

^***p < 0.001

Table 3.

Independent associations of health-related fitness component variables with the health outcome variables

	Model 1 (95% CI)	Model 2 (95% CI)
SF-12 PCS
Handgrip strength	0.02 (− 0.00, 0.04)	0.02 (− 0.00, 0.04)
Lower limbs muscle strength	0.06 (0.03, 0.09) **	0.05 (0.02, 0.08) **
Skeletal muscle massa	2.04 (0.71, 3.37) **	1.96 (0.61, 3.31) **
Body mass index	− 0.04 (− 0.06, − 0.01) **	− 0.03 (− 0.05, − 0.01) **
Waist circumference	− 0.01 (− 0.02, − 0.01) **	− 0.01 (− 0.02, − 0.00) *
Fat mass percentage	− 0.03 (− 0.05, − 0.01) *	− 0.03 (− 0.05, − 0.01) *
SF-12 MCS
Handgrip strength	0.02 (− 0.01, 0.04)	0.02 (− 0.01, 0.04)
Lower limbs muscle strength	0.03 (− 0.01, 0.06)	0.03 (− 0.01, 0.06)
Skeletal muscle massa	− 0.57 (− 1.89, 0.74)	− 0.65 (− 1.98, 0.69)
Body mass index	− 0.00 (−0.02, 0.02)	0.00 (−0.02, 0.03)
Waist circumference	0.01 (0.00, 0.02) *	0.01 (0.00, 0.02) *
Fat mass percentage	0.01 (− 0.01, 0.03)	0.01 (− 0.02, 0.03)

Values are unstandardized regression coefficients (95% Confidence Interval)

Model 1 was adjusted for sex (women, men), age (years), and device wear time (h/day); Model 2 was adjusted as model 1, plus educational level (low, high), tobacco consume (current, former, never), alcohol consume (current, former, never) and chronic conditions (none, one or more)

^aSkeletal muscle mass normalized by body mass index

Abbreviations: SF-12 MCS, mental health-related component summary score of the 12-Item Short-Form Health Survey; SF-12 PCS, physical health-related component summary score of the 12-Item Short-Form Health Survey

^*p < 0.05; **p < 0.01

Mediation analyses were conducted only for associations that reached statistical significance in the fully adjusted linear regression models, that is between MVPA and the PCS (Figs. 1 and 2). This analysis, including muscle strength and body composition variables, revealed that the association between MVPA and PCS was partially mediated by lower limb muscle strength and fat mass percentage. The adjusted mediation effect accounted for 8.5% and 28.9%, respectively, with borderline significances in the Sobel test [(z = 1.753, p = 0.080) and (z = 1.912, p = 0.056), respectively] (Figs. 1b and 2a). Furthermore, the association between MVPA and PCS was fully mediated by skeletal muscle mass normalized by BMI, accounting for 28.2% of the effect (z = 2.016, p = 0.044) (Fig. 2b).

Fig. 1.

Simple mediation models of the association between moderate-to-vigorous physical activity and the Physical Component Summary score of the SF-12, showing mediation by: (a) handgrip strength, (b) lower-limbs muscle strength, (c) body mass index, and (d) waist circumference. Values are unstandardized regression coefficients (95% Confidence Interval). Analyses were adjusted for sex (women, men), age (years), device wear time (h/day), educational level (low, high), tobacco consume (current, former, never), alcohol consume (current, former, never), chronic conditions (none, one or more), and the time spent in SB (h/day)

Fig. 2.

Simple mediation models, in the study subsample, of the association between moderate-to-vigorous physical activity and the Physical Component Summary score of the SF-12, showing mediation by: (a) fat mass (percentage of total body weight); (b) skeletal muscle mass. Values are unstandardized regression coefficients (95% Confidence Interval). Analyses were adjusted for sex (women, men), age (years), device wear time (h/day), educational level (low, high), tobacco consume (current, former, never), alcohol consume (current, former, never), chronic conditions (none, one or more), and the time spent in SB (h/day)

Discussion

The present study examined the associations between movement behaviours—objectively assessed using a pattern-recognition accelerometer—and both physical and mental components of HRQoL in community-dwelling Spanish older adults, and the potential mediator role of muscle strength and body composition in these associations. Our results show that total time spent in MVPA, lower limb muscle strength, and skeletal muscle mass are directly associated with the physical HRQoL component, whereas BMI, WC, and fat mass percentage are inversely associated with this outcome. In addition to these independent associations, our results show that the association between MVPA and PCS is partially mediated by lower limb muscle strength and fat mass percentage, and fully mediated by skeletal muscle mass. There were no associations between total time spent in SB or LPA and the health outcomes, and only WC presented a direct, albeit weak, association with the mental HRQoL component.

According to the World Health Organization, the ability to preserve our physical and mental capacities is a key determinant of healthy ageing [15]. Exercise interventions have been shown to be effective in improving physical function and attenuating age-related decline in older adults [15, 38–40], and our findings are consistent with this evidence. Previous studies have reported that higher amounts of PA and lower SB are associated with greater muscle strength and power, with MVPA emerging as a strong determinant and predictor of health outcomes [38]. The positive statistical association between MVPA and PCS observed in the present study reinforces this notion, suggesting that, theoretically, a greater engagement in MVPA may positively impact physical function in older adults. This aligns with the findings of Sunde et al., who conducted a randomized controlled trial and found that a high-intensity exercise programme improved the physical component, but not the mental component, of HRQoL in older adults [41]. For this population, mobility plays a central role in maintaining an independent lifestyle and preserving both physical and psychological health [40]. Therefore, our findings may have clinical applications in promoting MVPA daily activities in cases where the benefits outweigh the risks, particularly given that this population is predominantly sedentary [9].

In this study, when lower limb muscle strength was included in the model as a mediator variable, equations a and b result was statistically significant, and the association between MVPA and PCS (equation c’) remained significant although attenuated. This was interpreted as a partial mediation from lower limbs’ muscle strength; besides, the indirect effect (a × b) was not statistically significant, although borderline in the Sobel test. Moreover, Equation c’ became non-significant and the indirect effect becomes significant in the Sobel test when skeletal muscle mass was included as a mediator variable, which is considered as a full mediation, highlighting the role of skeletal muscle in the physical component of HRQoL. Skeletal muscle is now recognized as an endocrine organ capable of participating in metabolic regulation, immune response, and inter-organ communication, with effects on neurological function, bone metabolism, and cardiovascular health, among others [42, 43]. It has been documented that myokines, bioactive proteins secreted by skeletal muscle in response to contraction, may have a positive impact on ageing-related processes [43]. Thus, it is reasonable to hypothesize that the health benefits of regular PA initiate with its inherent muscle contractions that trigger the endocrine muscle response, which in turn contributes to improvements in HRQoL, as reported in this study. Therefore, in this line, the practice of physical activity of moderate-to-vigorous intensity oriented to the skeletal muscle mass and strength improvement in older adults seems to be appropriate to preserve their physical functioning and quality of life.

Muscle strength, the capacity to generate force through muscle contraction [44], is also associated with diverse health outcomes in older adults. Positive associations between handgrip strength and outcomes such as grey matter volume, cognitive functions, and healthy ageing index scores have been reported [1, 6]. Therefore, muscle strength may play a protective role against ageing-associated neurodegeneration [7] and may delay cognitive decline and dementia [20]. Low muscle strength is associated with cardiovascular complications in patients with type 2 diabetes mellitus [45], while sarcopenia, the age-related decline in muscle strength and mass, is a risk factor for falls and fractures and predicts functional decline in activities of daily living [46]. Our findings support this evidence, considering that the statistical analyses revealed that lower limb muscle strength and skeletal muscle mass were directly associated with the physical health-related component of quality of life and mediated the association between this health outcome and PA.

Although WC and BMI were significantly associated with PCS in this study, they did not show a statistical mediating role on the association between this health outcome and MVPA. On the other hand, fat mass percentage appears to be a partial mediator in the association between MVPA and PCS (i.e., Equations a and b statistically significant, Equation c’ non-significant and borderline Sobel test), even though it has an independent significant inverse relationship with this outcome. These results suggest that part of the PCS enhancement associated with the MVPA practice is explained by the biologically plausible mediating role of the fat tissue reduction induced by this behaviour; however, a fat tissue reduction achieved through mechanisms other than PA, such as dietary interventions, pharmacological treatments, or surgery, among others, could also be beneficial to improve quality of life [44, 47].

HRQoL is a multidimensional construct encompassing physical health, personal beliefs, social relationships, and mental health [48]. According to our results, there was no association between PA and the mental component of HRQoL, and only WC showed a small direct association with this outcome. Although we use the 12-Item Short-Form Health Survey, a reliable and valid instrument to assess the mental health component [34], the absence of association could be explained by unmeasured variables known to influence mental health, such as chronic stress and maladaptive coping mechanisms, among others [49]. Therefore, we cannot dismiss the importance of regular PA in improving brain-related health outcomes in older populations, as there is growing evidence of its influence on cognitive decline prevention, neuroplasticity, and memory performance, among others [5, 6]. More sensitive and validated tools to assess mental health disorders (e.g., depression, anxiety, and stress), as well as cognitive performance, could provide deeper insight into the direct and mediating roles of movement behaviours in these health outcomes.

Although this study has important strengths, such as the use of a device with strong concurrent validity to assess physical behaviours [31, 32], several limitations must be acknowledged. Due to storage and battery life restrictions of the IDEEA device, daily activities were monitored for only 48 h, which may not fully capture variability in movement behaviours. The body composition assessment, conducted with bioelectrical impedance analysis, could not be performed in a larger sample. The generalization of our findings is limited to high-functioning community-dwelling older adults and further research is needed among physically vulnerable older adults’ populations, including frail, multimorbid, and institutionalized individuals. The fact that participants were high-functioning older adults most likely influenced the results, given their relatively preserved body composition and physical performance level. Finally, the mediation analysis should be interpreted as exploratory due to the cross-sectional design, which limits our ability to infer causality. Future longitudinal studies are needed to confirm these findings.

In conclusion, our results suggest that total time spent in MVPA is directly associated with improvements in the physical component score of HRQoL, and indirectly through potential effects on muscle strength and body composition fitness components. Lower limb muscle strength acted as a partial mediator, while muscle mass, assessed in a reduced sample, produced full mediation when included in the model. These findings highlight the role of skeletal muscle in the ageing process and the importance of preserving muscle mass and strength in high-functioning community-dwelling older adults to promote higher HRQoL. However, more longitudinal studies and randomized controlled trials are required to strengthen these postulates.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Julia Wiedmaier-Barros analysed and interpreted the data and drafted the manuscript. Sara Higueras–Fresnillo and Verónica Cabanas–Sánchez acquired the data and revised the manuscript for key intellectual content. Kabir Prem Sadarangani revised the manuscript for key intellectual content. David Martínez–Gómez conceived and designed the study, obtained funding, supervised the study, and revised the manuscript for key intellectual content.

Funding

This work was supported by Grants from MINECO I + D + i (DEP2013–47786-R); UAM-Santander (CEAL-AL/2015–20); Real Madrid–UEM (P2016/RM09); and a “Ramon y Cajal” contract (RYC-2016–20546) to D.M.-G.

Declarations

Ethics approval

The Ethics Committee of the Autonomous University of Madrid (Madrid, Spain) approved the study.

Consent to participate

All participants were informed of the nature of the study and provided written informed consent prior to data collection.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.