Lack of association between circulating apelin level and frailty-related functional parameters in older adults: a cross-sectional study

Frailty is a geriatric condition characterized by progressive physiologic decline across multiple body systems, leading to increased vulnerability to adverse outcomes and death [1]. Because the number of older persons is rapidly increasing worldwide, frailty is likely to become a significantly more serious public health concern. Despite the lack of gold standard measurement, two major frailty models have been well accepted and widely used in aging research and practice [2]. The first concept proposed by Fried et al. [3] is “physical frailty” that views frailty as a biologic syndrome of decreased reserve and resistance to stressors. The other concept proposed by Rockwood et al. [4] is “cumulative deficit frailty,” which is also termed as “frailty index (FI)”, based on the hypothesis that the accumulation of health, functional, psychological, and cognitive problems serves as an indicator of an individual’s aging-related health status. Although each assessment tool has strengths and weaknesses in terms of clinical application [2], both are useful to identify older adults at high risk for mortality [5]. Importantly, frailty is a heterogeneous condition resulting from various causes that might be potentially reversible [6]. Therefore, continuous efforts to identify factors affecting functional changes with aging are critical to achieve the goal of expanding health span.

Apelin is an active endogenous peptide that acts via G protein-coupled apelin receptor (APJ) to exert diverse functional effects [7]. This pleiotropic molecule is involved in multiple biologic processes, including cardiovascular regulation, neuroprotection, immunity, and glucose metabolism [8‐10]. Apelin has been recently receiving great attention as a promising target for antiaging intervention, primarily based on results from genetically altered mice [11]. For example, the expression of apelin and APJ was downregulated in an age-dependent manner [12, 13]. Importantly, apelin-deficient mice exhibited accelerated the onset of sarcopenia and senescence, whereas apelin treatment in aged mice enhanced muscle strength and physical activity and rejuvenated behavioral and circadian phenotypes [12, 13]. However, despite the clear implications of apelin in physiologic alterations with aging, clinical studies relating circulating apelin concentration to frailty are limited. With the aim to validate previous in vitro and in vivo data and investigate the possible role of apelin in humans, this study examined serum apelin level in relation to frailty and its associated parameters in a cohort of older adults.

Baseline characteristics of 80 study participants are listed in Table 1. Among the 21 (26.2%) robust, 47 (58.8%) prefrail, and 12 (15.0%) frail older adults based on Fried’s criteria, 8 (38.1%), 28 (59.6%), and 7 (58.3%) were women, respectively. The mean ages of the robust, prefrail, and frail groups were 67.5 ± 6.5, 69.7 ± 6.2, and 70.8 ± 5.0, respectively. There were no significant differences in weight, height, BMI, serum albumin level, time to complete 5 chair stands, ASMI, and prevalence of polypharmacy, multimorbidity, and ADL disability among the three groups. Frail participants had lower grip strength, gait speed, SPPB score, muscle mass, and mini-cognitive score and higher FI, social frailty score, and PHQ-2 score than robust and/or prefrail participants. Furthermore, frail participants were more likely to have sarcopenia, IADL disability, cognitive dysfunction, and depression than robust and/or prefrail participants.

Differences in serum apelin level according to the phenotypic frailty status were assessed using analysis of covariance. Before and after adjustment for sex, age, and BMI, there were no significant differences in serum apelin level between the three groups (Fig. 1).

Linear regression analyses were performed to examine the association between serum apelin level and frailty-related parameters with normal distribution (Table 2). However, serum apelin concentration was not associated with FI, grip strength, gait speed, time to complete 5 chair stands, muscle mass, and ASMI, regardless of the adjustment models.

To determine whether the association between serum apelin level and frailty-related parameters may have a threshold effect rather than a gradual effect, we divided the participants into quartiles according to serum apelin level (Fig. 2). Before and after adjustment for sex, age, and BMI, the groups did not differ significantly in terms of FI, grip strength, gait speed, time to complete 5 chair stands, SPPB score, muscle mass, ASMI, social frailty score, mini-cognitive score, and PHQ-2 score. Furthermore, logistic regression analyses showed that the ORs for phenotypic frailty according to serum apelin quartiles were not statistically significant, regardless of the adjustment models (Fig. 3).

In a cohort of ambulatory, community-dwelling older adults, we observed that serum apelin level was not significantly associated with both the phenotypic frailty and the functional parameters. Additionally, serum apelin level was not associated with the risk of phenotypic frailty before and after adjustment for sex, age, and BMI.

Many lines of evidence from in vitro and animal experiments indicate that apelin may be implicated in various age-related disorders. While mice lacking apelin and its receptor showed markedly decreased muscle strength and physical performance, chronic apelin supplementation attenuated muscle aging [8, 13]. The beneficial effects of apelin on muscle metabolism are probably mediated by the stimulation of AMP-activated protein kinase and Akt and the resultant mitochondriogenesis [13, 27]. Downregulation of apelin in mice exhibited age-dependent decline in multiple organs, whereas systemic restoration of apelin improved cognitive and cardiovascular dysfunction, resulting in increased mammalian health span [8, 9, 11, 12, 28]. Because these physiologic alterations including musculoskeletal weakness and neurodegeneration are major manifestations of frailty, apelin could be expected not only as a therapeutic target to slow or reverse aging process but also as a clinical biomarker to assess the frailty status.

Despite the definite role and plausible mechanisms of apelin as an antiaging factor in animal models, its effects on human health remain unclear. Because it is impossible to conduct interventional trials in humans without safety verification from preclinical data, the roles of candidate factors on human metabolism must be extrapolated from the observational results, leading us to conduct this clinical study. However, different from the experimental results, we did not observe a significant association between circulating apelin level and frailty-related parameters in older adults. Although we cannot determine the exact reason for this discrepancy, several explanations might be hypothesized. First, although mice are widely used to study human illness mechanisms because of biologic similarity, they are not the same. In detail, the comparison research focusing on the genetic and biochemical processes regulating genome activity in humans and mice by the Mouse ENCODE Consortium revealed a large degree of divergence of sequences involved in transcriptional regulation, chromatin state, and higher-order chromatin organization [29, 30]. This could answer the question why certain processes and systems in mice, such as the immune system, metabolism, and stress response, are significantly different to those in humans and why drugs that have worked in animal experiments are not equally successful in humans. The lack of associations between apelin and frailty in humans, unlike in mice, could also be attributable to these differences in two mammals. Second, physiologic apelin level observed in humans may not be high enough to exert the phenotypic changes. The apelin expression and production are almost completely blocked in apelin-deficient mice, and the concentration of apelin treated into mice is relatively high with a long period [12, 13]. On the contrary, because the difference in apelin mean values between the lowest (Q1) and highest (Q4) quartiles in our study was only 0.466 ng/mL, we may have not observed significant changes of functional parameters. Third, apelin is synthesized as a 77-residue preproapelin and cleaved into the 55-residue proapelin and subsequently into 13–36-residue active isoforms [31]. Therefore, diverse isoforms, such as apelin-13, apelin-17, and apelin-36, could be present in biological fluids, including blood, and there is a possibility that each isoform plays a different role as a biomarker to assess frailty. This issue needs to be clarified through further studies using more accurate techniques.

The major strengths of our study are that we adapted both frailty operational definitions that have been well validated and included various functional parameters, including grip strength, gait speed, time to complete 5 chair stands, SPPB score, social frailty score, mini-cognitive score, PHQ-2 score, as well as muscle mass, contributing to increased reliability of our results. However, despite these strengths, several limitations should also be considered when interpreting the results. First, because this was a cross-sectional study, we could not determine a causal relationship between the variables. Second, because of difficulties in simultaneously obtaining the information about frailty status and the blood samples, the number of enrolled participants with consent was relatively small. However, we believe that this preliminary analysis might provide an important background for further studies. Third, our study population was exclusively Koreans. Thus, we cannot be certain that our results are applicable to other populations. Fourth, we could not adjust for various confounding factors because of a relatively small sample size in the analysis. Finally, we cannot exclude the possibility that any biased information or uncontrolled factors that affect apelin and frailty could affect the conclusion.

In summary, serum apelin levels did not differ significantly according to the phenotypic frailty status and was not associated with FI and functional parameters in older adults, despite the beneficial effects of apelin against age-related physiologic decline that is observed primarily in animal models. Additional well-designed large-scale longitudinal studies are necessary to understand the definite role of circulating apelin in frailty risk assessment.