Ageing is associated with a variety of underlying changes at the biological level that drive the development of chronic diseases [1]. Adults over their lifespan typically gain adipose tissue due to the combined effects of underlying cellular changes, such as reduced mitochondrial volume and oxidative capacity, hormonal changes, and decreased energy expenditure [2]. This increased adiposity, compounded by genetic and environmental influences, can lead to the development of obesity as adults age [3]. Obesity rates in older adults aged ≥65 years exceed 40%, and are strongly associated with functional decline, mobility disability, nursing home placement, and death [4–7]. However, efforts that seek to address obesity purely by lowering body weight in this population are problematic. This is a result of sarcopenia, which is the physiologic, age-associated loss of muscle mass and strength and function, that is often accelerated by the development of chronic diseases. Thus, conventional weight loss interventions that purely focus on reducing caloric input may risk worsening sarcopenia, which itself is strongly associated with functional decline [8].

In recent years, the syndrome of sarcopenic obesity has been formally characterized as the co-existence of increased adipose tissue (obesity) and decreased muscle mass or strength (sarcopenia) [9]. Sarcopenic obesity has been postulated to synergistically lead to a wide variety of adverse outcomes, more so than obesity or sarcopenia alone [10]. People with sarcopenic obesity have an increased risk of cardiometabolic disorders, mobility disability, functional impairment, and mortality [11, 12]. However, efforts to study sarcopenic obesity and its treatment outcomes have historically been limited given the discrepancies in definitions used by both clinicians and researchers. Previous definitions of sarcopenic obesity have included a range of characteristics using various measures of body imaging, muscle indexes, and body fat percentage. Additionally, definitions are heavily reliant on their different reference populations [13–15]. As such, the prevalence rate ranges considerably - one study noted up to an 18-fold difference depending on the definition used [16].

The absence of a consensus definition for sarcopenic obesity led to recent guidelines put forth by the European Association for the Study of Obesity (EASO) and the European Society for Clinical Nutrition and Metabolism (ESPEN) [17]. This consensus holds promise for characterizing the definitions of sarcopenic obesity by standardizing its nomenclature. As a result of the challenges in the definition, there has been a paucity of studies exploring the differences in treatment outcomes in this population. As was highlighted in the EASO/ESPEN guidelines, there is concern that the treatment of sarcopenic obesity may require nuanced and personalized approaches that differ from those with obesity alone without sarcopenia. Weight loss may risk the loss of muscle mass which can then lead to worsening of physical functioning in adults with sarcopenic obesity. As the interplay between fat and muscle impact both chronic inflammation and myokine signaling, identifying the response to conventional interventions can help us advance precision nutrition and exercise interventions to maximize their impact and improve physical function. Therefore, our goal in this study was to apply these new definitions in a retrospective analysis of older adults with obesity who participated in a multicomponent, technology-based, weight loss intervention [18]. To our knowledge, there are no studies that have assessed whether older adults with obesity and sarcopenia respond differently to diet and exercise interventions compared to those persons without sarcopenic obesity using this updated definition. Thus, as a secondary analysis of existing data, we aim to provide formative data as to whether the presence of sarcopenia in the context of obesity leads to differences in weight loss, physical function changes, or body composition changes.

The overall group consisted of 53 adults with a mean age of 72.9±3.9 (Table 2) years of which 69.8% were female (100% White). We compared baseline characteristics of the participants completing the intervention (n=44) to those who did not (n=9) in Table 2. Completers were more likely to have Medicare insurance (p=0.03) and a diagnosis of diabetes (p=0.05) than the non-completers. For our primary outcomes of the entire intervention cohort in our parent study, we saw a significant decrease in weight (Δ=−4.6 kg [95%CI: −3.6, −5.6]; p<0.001) and a significant improvement in 30-second STS repetitions (Δ=+3.1 repetitions [1.9, 4.3]; p<0.001). This was also observed in this current secondary analysis in individuals with and without sarcopenic obesity between baseline and follow-up. However, we found no significant differences between groups in the change in variables between baseline and follow-up (Table 3 and Supplemental Table 1).

For our secondary outcomes, there was no difference in the pre/post measurements of gait speed for any group except using Definition 5 nor were there differences by sarcopenia status for any of the definitions. There were differences observed by sarcopenia status in the 6MWT that varied by definition. A significant post-intervention increase was observed in LLFDI score and REAP-S diet quality score for each definition irrespective of sarcopenia status. There were no significant differences in the change in LLFDI in the sarcopenic vs. non-sarcopenic group. There was however a significant difference in the change in REAP-S for those with sarcopenic obesity and those with non-sarcopenic obesity for some but not all definitions. There was no significant change in measured caloric intake using ASA-24 in either those with sarcopenic obesity or those with sarcopenic obesity.

While all participants had BIA performed at baseline for body composition, the unfolding of the COVID-19 pandemic during the study period precluded the collection of BIA for all participants at follow-up. In the subset of participants with pre- and post-intervention body composition measures (n=30), we observed significant decreases in fat mass % across all definitions with and without sarcopenia (Supplemental Table 1). There was a significant difference in pre/post values for skeletal muscle mass/weight for all definitions but not in the differences across groups by skeletal muscle mass/weight (Supplemental Table 1). There was a significant difference in VAT for some but not all sarcopenic obesity definitions and sarcopenic or non-sarcopenic obesity groups, but there was no difference in the differences across groups. Additional exploratory analyses compared the effect of the intervention by weight loss response status. Overall, we observed few differences in the changes by response status (loss of >5% of body weight) (Tables 4).

Our secondary analysis of this pilot study may suggest that a technology-based weight loss intervention utilizing exercise and nutritional interventions for adults with obesity may provide similar benefits across weight loss, physical function, and functional status domains for both participants classified with and without sarcopenic obesity. Our preliminary results are based on the recent EASO/ESPEN consensus definitions for sarcopenic obesity. An additional major finding is that even in persons classified as having sarcopenic obesity, there were no adverse changes in skeletal muscle mass/weight. This has major implications on the content and design of future interventions.

The weight loss observed in our overall study cohort approaches the threshold for clinically significant weight loss of 5% for both those with and without sarcopenic obesity [34]. The corresponding improvement in 30 second STS is also clinically relevant across sarcopenic obesity status groups and has been previously associated with improved quality of life and physical function [35]. Together, these suggest that weight loss was achieved while simultaneously improving physical function.

Yet, traditional weight loss interventions focusing on caloric restriction have a strong potential of reducing muscle mass. Minimizing dietary intake can lead to a decrease in muscle protein synthesis and an increase in muscle proteolysis [36, 37]. Previous research in this population group has shown dietary restriction without concomitant resistance exercise can lead to loss of muscle mass and decreased handgrip strength [38]. Beyond muscle loss, caloric restriction has been linked to alterations in bone metabolism and possible decreases in bone mineral density, potentially putting patients at higher risk of fractures [39]. Interventions that combine exercise with caloric restriction may mitigate this risk [40, 41]. The benefits of resistance training for treating sarcopenic obesity have been well established, with several studies suggesting improved muscle mass and function in older adults [42–44]. Additionally, nutrition interventions that increase protein intake are associated with increased muscle mass and decreased total body fat in individuals with sarcopenic obesity [45, 46]. Our pilot findings of preserved skeletal muscle mass suggest our multi-component telemedicine intervention may have avoided the pitfall of worsening muscle loss and functional decline while losing weight risked by caloric restriction alone. These foundational results suggest that even when losing weight with this delivery modality, muscle preservation may be possible.

One important aspect to highlight is that participants with sarcopenic obesity were heavier than those without sarcopenia. While participants with sarcopenic obesity were heavier at baseline, they had no significant differences in skeletal muscle mass/weight. Using cutpoints that are not normalized for height can lead to challenges when using straight cutpoints to ascertain sarcopenia.

Notably, previous studies have used varied definitions of sarcopenic obesity, which makes intra-study comparison difficult. Our study serves as one of the first to utilize the recent EASO/ESPEN consensus definitions for sarcopenic obesity. Use of this definition holds promise for improving standardization of the term “sarcopenic obesity”. We used multiple definitions for sarcopenic obesity based on the consensus definition with most of our findings demonstrating similar results irrespective of the definition used. While we acknowledge this may increase the risk of type II error, gaining an understanding and trends on the variability of the impact on the definitions can be helpful to the consortium’s goal of streamlining definitions. Future studies may benefit from consolidating to a single one of these definitions to further limit additional variability between studies. Lastly, our sub-group analyses compared outcomes for responders and non-responders to losing weight.

We acknowledge that our study had a number of limitations that limit the ability to generalize and make solid causal inferences. First, the cohort lacked racial diversity with a strong female predominance preventing us from identifying racial- or sex-differences. Second, the small number of participants as part of this feasibility study suggested that our analyses may have lacked sufficient power to detect significant differences, particularly in the responder vs non-responder subgroups. Future, adequately powered trials are critically needed. Third, this feasibility study’s design lacked randomization, blinding, and did not have a control group. These can introduce bias into the interpretation of our results. Fourth, the long-term implications of the study are also limited given the fixed six-month duration of the study. It is unclear whether these persistent changes in weight loss or function will be maintained. Fifth, our limited sample size prevents us from controlling for potential confounders. Sixth, we acknowledge that we were unable to determine all possible combinations proposed by EASO/ESPEN as a result of only having BIA data. However, this consortium is proposing additional, streamlined definitions in the coming years that may overcome this limitation. While we acknowledge these limitations, our analysis utilizes contemporary definitions and permits us to gain an early understanding of precision medicine – tailoring the right intervention to the right person. We recognize that certain individuals did not lose weight or improve function. Future work should focus on the heterogeneity of response in this patient population.

We provide foundational evidence that this multicomponent weight loss intervention for older adults with obesity demonstrated similar weight loss and improvements in physical functioning with maintained muscle mass in individuals with and without sarcopenic obesity using the new EASO/ESPEN consensus definitions. Our findings suggest that multicomponent weight loss interventions may potentially confer benefit for older adults with obesity irrespective of sarcopenic obesity status.