Background

The ageing process is characterized by a change in body composition with a progressive increase of fat mass, in particular visceral fat, and an ectopic distribution of fat in non-fat tissues, which also impairs organ functions [1]. In addition of being associated with many metabolic and cardiovascular diseases [2], the excess of body fat and obesity have been linked to impaired physical performance, mobility limitations and risk of developing disability in older people [3].

With ageing, there is a reduction of muscle mass too, which is associated with a decrease of muscle strength and power.

Although a generally accepted threshold for the age associated physiological reduction of muscle mass and function is still lacking [4], there is consensus that an important decline, known as sarcopenia [5] may lead to adverse functional and clinical outcomes such as institutionalization, mortality, and increased length of hospital stays [6, 7]. Sarcopenia is also closely related to physical frailty [8, 9].

Frailty is a condition characterized by a reduction in physiological reserves, leading to increased vulnerability to stressors and the risk of adverse consequences [10]. This condition comprises both physical and general frailty. The first one is well described by the frailty phenotype (FP) which focuses on physical aspects of frailty and on the concept of energy loss. FP identifies individuals as frail if they exhibit at least three of the following criteria: low muscle strength, slow gait speed, reduced physical activity, unintentional weight loss, and exhaustion [11]. General frailty is instead well described by the cumulative deficit model, also known as the frailty index (FI). The FI provides a broader perspective on frailty compared to the FP, considering social, psychological, cognitive, and functional determinants in addition to physical aspects and comorbidities [12]. According to the FI model, the higher the number of health deficits present, the greater the vulnerability and frailty of the individual.

When sarcopenia and obesity occur together, they create a vicious cycle that synergistically enhance both conditions [13]. On the one hand the increased fat tissue promotes oxidative stress, inflammation and the development of insulin resistance, which negatively act on the trophism and function of muscles [14]. On the other hand, sarcopenia reduces total energy expenditure [15] leading to fat accumulation [16]. Obese individuals tend also to have a high burden of chronic conditions [17, 18] which negatively affect muscle metabolism. Finally, sedentary life, which is both a cause and a consequence of sarcopenia and obesity, favours the progression of both conditions [19,20,21].

The co-occurrence of sarcopenia and obesity is named sarcopenic obesity (SO) [13]. The risk of developing adverse clinical outcomes derived from the two conditions cumulates in SO [22,23,24,25,26,27,28,29]. SO has been linked to physical frailty in the Asian [30, 31] and Brazilian [32] populations. Also, two small French studies described an association between SO and physical frailty [33] and SO and physical function [34] in middle-aged and older women. A more general association between body composition and physical frailty, but without a specific reference to sarcopenic obesity has been described in the American [35,36,37], Norwegian [38] and Portuguese [39] populations. A link between general frailty and body composition measure with body mass index (BMI) was highlighted in the Survey of Health, Ageing and Retirement in Europe (SHARE) [40]. Here, we analysed the relative contribution of body composition and SO, measured with a functional perspective according to ISarcoPRM guide [41], to general frailty, measured with the cumulative deficit model in a population of community dwelling Italian older volunteers.

Materials and methods

Study participants

The Frailty and Sarcopenia Network (FRASNET) study is a cross-sectional multicentre observational cohort study involving community-dwelling healthy volunteers and istitutionalized patients [42]. The study was approved by the San Raffaele Scientific Institute ethical board (24/INT/2017). Participants signed a written informed consent before taking part in the study. The enrolment was performed between the 1st April 2017 and the 16th October 2020.

Inclusion/exclusion criteria

The inclusion criteria were: (i) being aged 65 years old or older, (ii) being able of walking more than 500 m without assistance; (iii) life expectancy > 6 months. The exclusion criteria for our study were as follows: (i) Severe Cognitive Impairment: participants with a Mini-Mental State Examination (MMSE) score of less than 18/30 were excluded. This threshold was chosen to ensure that all participants could understand and comply with the study procedures and to minimize the risk of inaccurate self-reported data due to cognitive deficits. (ii) Inability to Provide Informed Consent: individuals who were unable to sign a written informed consent were excluded. This criterion was essential to comply with ethical standards and ensure that all participants voluntarily agreed to partake in the study with a clear understanding of its purpose and procedures. (iii) Severe Health Problems: participants with severe health issues were excluded to avoid confounding effects on the study outcomes. Specific conditions for exclusion from the study were uncontrolled hypertension defined as consistently high blood pressure despite medication, which could interfere with physical performance and other study measures. Recent fractures: individuals with upper or lower extremity fractures within the past six months were excluded to prevent pain or mobility limitations from affecting the results. Recent myocardial infarction: those who had experienced a heart attack within the past year were excluded due to potential restrictions on physical activity and overall health instability. (iv) Residence in Retirement Homes: patients recruited from retirement homes were excluded to ensure a homogenous sample. This decision was made to avoid variability in care levels and living conditions that might influence the study outcomes. (v) Incomplete Data: participants with missing data necessary for the computation of frailty or lacking information on body composition were excluded. Complete datasets were required to accurately assess frailty and its associations with other variables, ensuring the reliability and validity of the study findings. These exclusion criteria were carefully selected to maintain the integrity of the study sample, ensure participant safety, and enhance the reliability and applicability of the results.

Procedures

Participants underwent multidimensional geriatric assessments, that included the collection of demographic and psychosocial data through self-administered questionnaire, the assessment of comorbidities and pharmacological therapies, of the number of falls and emergency department accesses in the year preceding the evaluation, anthropometric measurements (weight, height, waist circumference, body mass index (BMI) calculation). Cognition was screened with the Mini Mental State Examination (MMSE) [43], mood with the Geriatric Depression Scale (GDS) 15 items [44], exhaustion with the Fatigue Severity Scale (FSS) [45], quality of with life the Short Form 36 (SF-36) Health Survey [46] and the level of physical activity with the Physical Activity Scale for Elderly (PASE) questionnaire [47].

Body composition was determined thought the Full Body Sensor Body Composition Monitor and Scale (OMRON) balance [48]. This balance uses an extremely weak electrical current (50 kHz and less than 500µA) that passes through the body to estimate with the bioelectrical impedance method body composition. According to the ISarcoPRM guide [41] muscle performance was assessed with gait speed and the Short Physical Performance Battery (SPPB) [49] and muscle strength with the chair-stand subtest of the SPPB [50]. Values > 15 s were considered indicative of reduced muscle strength and therefore of pre-sarcopenia [5, 41]. People with pre-sarcopenia were further classified as obese and non-obese according to the percentage of fat mass measured though the bioelectrical impedance balance (obese if fat mass ≥ 30% in men and ≥ 42% in women) [48].

In this older population, the diagnosis of sarcopenia was performed according a functional approach as the one described by the ISarcoPRM [41] and the European Working Group on Sarcopenia [5]. Due to the absence of muscle echography for assessing muscle mass, we used data on muscle mass percentage derived from bioelectrical impedance analysis. Sarcopenia was defined by the presence of both reduced muscle strength (SPPB chair test > 15 s) and mass (< 32.9% in men and < 23.9% in women) [48].

Sarcopenic obesity was defined according to the ESPEN and EASO Consensus Statement [13] as altered skeletal muscle strength (i.e. SPPB chair test > 15 s), reduced muscle mass (< 32.9% in men and < 23.9% in women) and increased fat mass % (≥ 30% in men and ≥ 42% in women) [48].

Individuals with a normal muscle mass and strength were considered robust. People with impaired muscle mass but normal muscle strength configured the group of people with reduced muscle mass but normal muscle strength.

All study participants were further classified as obese and non-obese according to the percentage of fat mass measured though the bioelectrical impedance balance [48]. Thus, we finally had eight body composition classes: robust non-obese, robust obese, pre-sarcopenic non-obese, pre-sarcopenic obese, sarcopenic non-obese, sarcopenic obese, individuals with reduced muscle mass and normal muscle strength obese and non-obese.

Frailty was assessed through a 49-items FI created by using the criteria defined by Searl et al. [51].

The 49 variables used to compute the FI were also retrieved from the multidimensional geriatric evaluations (Table S1). Each deficit included in the FI was scored 0 when absent, and 1 when present. In cases of missing data, the FI was calculated by using an adequately reduced denominator excluding the items for whom data were missing [51]. Participants having more than 20% of missing variables were excluded from the computation of the FI [51]. The score of the FI ranges from 0 to 1, with lower levels identifying fitter individuals. A cut-off point of ≥ 0.25 defines ‘frail’ individuals [51].

The data were collected and reported in an electronic case report form (eCRF) created using FileMaker Pro v. 11.

Statistical analyses

Descriptive statistics were used to show the baseline characteristics of the study population. Continuous variables were presented as mean and standard deviations (SD), when normally distributed, or with median and interquartile range (IQR), when data had a skewed distribution. Dichotomous variables were presented as number (N) and percentage (%). Comparison of the characteristics among obese and non-obese sarcopenic, obese and non-obese pre-sarcopenic, obese and non-obese robust and obese and non-obese participants with a reduced muscle mass but a normal muscle strength was performed with the U Mann Whitney test for continuous variables and with the Chi Squared test for categorical variables.

Linear regression analyses were used to assess the association between frailty and different body composition classes. Analyses were also adjusted for age and sex.

Classification and regression tree (CRT) modelling was used to assess the predictors of frailty. Only predictors not included in the computation of the FI were considered: age, gender, civil status, BMI, percentage of fat mass, visceral fat, muscle mass, muscle strength (assessed though the SPPB chair test) and waist circumference.

All statistical analyses were performed with SPSS version 25.0 (SPSS Inc., Chicago, IL, USA).

Results

Among the participants enrolled in the FRASNET study 1114 were included in the present analysis. Only 10 institutionalized patients, 91 individuals who missed more than 20% of variables for the computation of the FI and 26 participants who missed data on body composition were excluded from the present analysis.

The study sample had a median age of 72 years and was composed for the 39.5% by males; 234 individuals (21%) were frail according to the FI with a median score of 0.11 (IQR 0.07–0.20). A FI higher score indicates greater vulnerability to stressors for the individual. Health deficits include comorbidities, symptoms, disabilities, or altered physical or mental signs.

The study population main characteristics are illustrated in Table 1. Robust non-obese individuals accounted for the 38.2% of the study sample. Figure 1 illustrates the body composition categories of the study population. Table 2A and 2B show body composition, muscle function and frailty in the different body composition categories. Obese individuals had a lower muscle mass and a higher FI score compared to non-obese participants (Table 2A and 2B). In addition, among robust people and people with reduced muscle mass and normal muscle strength, obese individuals displayed also a lower gait speed (Table 2A and 2B).

Fig. 1
figure 1

Body composition categories in the study population. Robust no obesity: 425 individuals (38.2%), robust with obesity 28 individuals (2.5%), reduced muscle mass normal strength no obesity 138 individuals (12.4%), reduced muscle mass normal strength with obesity 195 individuals (17.5%), presarcopenia no obesity 173 individuals (15.5%), presarcopenia with obesity 18 individuals (1.6%), sarcopenia no obesity 51 individuals (4.6%), sarcopenia with obesity 86 individuals (7.7%)

Full size image
Table 1 Main characteristics of the study population
Full size table
Table 2 Body composition, muscle function and frailty in the different body composition categories
Full size table

In the age and sex adjusted regression analyses being robust and non-obese (B – 0.42, 95% C.I. -0.06 - -0.03, p < 0.001) or being with a reduced muscle mass, a normal muscle strength and non-obese (B – 0.03, 95% C.I. -0.05 - -0.006, p = 0.011) were associated with a reduced probability of being frail. Instead, being pre-sarcopenic, either non-obese (B 0.03, 95% C.I, 0.007–0.044, p < 0.001) or obese (B 0.11, 95% C.I. 0.05–0.16, p < 0.001) or being sarcopenic obese (B 0.074, 95% C.I. 0.05–0.1, p < 0.001) were associated with an increased probability of being frail.

The regression tree model showed that fat mass % predicted frailty in people ≤ 70 years, whereas muscle strength predicted frailty in people between 70 and 81 years (Fig. 2).

Fig. 2
figure 2

Regression tree of the predictors of frailty measured through the frailty index

Full size image

Discussion

In this observational study we found that SO and pre-sarcopenia (either with and without obesity) were associated with frailty measured with the FI whereas being non obese (either robust or with a reduced muscle mass but a normal muscle strength) had a negative association with frailty. Fat mass percentage predicted frailty in people aged 65–70 years while muscle strength predicted frailty in people aged 70–81 years.

Previous studies have demonstrated an association between SO and physical frailty in the Asian [30, 31] and Brazilian populations [32] and a link between BMI and general frailty [40]. Indeed, obesity and frailty share common pathophysiologic mechanisms like inflammation and oxidative stress [52]. Muscle ectopic fat deposition alters mitochondrial function increasing reactive oxygen species production which in turn interfere with muscle protein function and contribute to the manifestation of frailty.

In this work we clarified how different classes of body composition, characterized in term of fat and muscle mass percentage, were associated with general frailty. Moreover, we identified fat mass percentage and muscle function as potential predictors of frailty. Our data on the identification of fat mass thresholds associated with general frailty in subjects aged 65–70 years highlights the importance of screening for general frailty in this demographic with elevated fat mass. Identification of subjects at risk could facilitate the implementation of early preventive interventions, such as weight loss, protein supplementation, and physical activity before actual frailty appears. These measures may mitigate obesity and prevent frailty and sarcopenic obesity before the onset of disability and other adverse consequences. Moreover, our study demonstrates that in individuals aged 70–81 years, the decline in muscle strength plays a significant role in predicting general frailty, thus adding to the available information that might be used to identify subjects in which the intervention could be useful.

Although frailty is the subject of intense and always growing attention from the scientific community several issues remains open, and in particular there is no accepted consensus on the methods that should be used for assessing frailty [53]. Here we add to the commonly used approach to define frailty a better insight on the added value of comprehensively examining body composition through measurements of both general and visceral fat, as well as evaluations of muscle mass and performance. We intentionally evaluated general frailty by constructing a FI that did not initially incorporate measures of body composition as health deficits. Our discovery of a significant association between this index and body fat, muscle mass, and performance reinforces the theory of interdependence among health deficits [54]. According to this theory, the presence of a deficit in one organ or system increases the likelihood of deficits in other related organs or systems. Muscle weakness and excess fat, although not explicitly included in our FI, can indeed be considered as health deficits.

Our findings lay the groundwork for future prospective studies, which could employ more precise methods—such as muscle echography capable of assessing regional muscle masses affected early by sarcopenia—to evaluate whether interventions aimed at improving body composition and muscle function could decrease the incidence of frailty.

However, some limitations of our work warrant mention. Firstly, the regional nature of the study conducted in an urban area may restrict the generalizability of our results. Secondly, the cross-sectional design prevented us from assessing the impact of body composition on frailty and other health outcomes over time. Additionally, the lack of evaluation of local muscle masses through muscle echography is noteworthy, as this could have enabled the early detection of sarcopenia in muscle districts prone to early disease onset including in particular the anterior thigh muscles [55].