**1. Introduction**

Type 2 diabetes mellitus (T2DM) is a chronic condition frequently occurring in old age [1,2]. T2DM is associated with higher risk of negative outcomes, including disability and mortality [3]. T2DM-related complications are especially prevalent in older adults and account for the increasing costs of T2DM [4]. Frailty defines a geriatric syndrome characterized by reduced ability to cope with life stressors and increased risk of adverse events (e.g., falls, delirium, loss of independence, mortality) [5,6]. T2DM and frailty are intimately related and share common features, including complex pathophysiology and heterogeneous phenotypes [7], that challenge their managemen<sup>t</sup> [5,8].

Muscle failure, both in its metabolic and functional manifestations, is a hallmark of T2DM and frailty [7,9]. The progressive and generalized loss of muscle mass, strength, and function with age, termed sarcopenia, fuels a self-reinforcing cycle in which structural, metabolic, and endocrine perturbations in muscle exacerbate T2DM-related signs and symptoms [10]. T2DM further promotes the decline in muscle mass and function [11,12] which, in turn, aggravates functional impairment [13].

The central role of muscle wasting in frailty and T2DM may guide the identification of novel biomarkers and possibly new treatment targets for the two conditions [14–16]. In this context, circulating amino acids are promising candidates given their sensor-transducer-e ffector role in systemic metabolism, muscle homeostasis, and physical function [17–20]. Moreover, amino acids are involved in processes critical to the development and progression of frailty and T2DM, such as inflammation, glucose homeostasis, and redox regulation [21–23].

Targeted metabolomics allowed identifying specific amino acid profiles that were associated with insulin resistance and risk of developing T2DM in independent cohorts across US, Europe, and China [24–27]. A plasma amino acid signature, together with specific circulating lipid species, was linked to glucose dyshomeostasis and impaired insulin sensitivity in older adults from the Baltimore Longitudinal Study of Aging (BLSA) [28]. In addition, distinct patterns of circulating amino acids were associated with frailty and/or muscle-related parameters (mass, turnover, performance) in older individuals at risk for frailty [29–32]. Finally, within the "BIOmarkers associated with Sarcopenia and PHysical frailty in EldeRly pErsons" (BIOSPHERE) study, a combination of serum amino acids and derivatives was identified that characterized the metabotype of older adults with physical frailty and sarcopenia (PF&S) [17].

Here, we sought to define the circulating amino acid profile of frail/pre-frail older adults with T2DM (F-T2DM). Our approach, described in the context of the "Metabolic biomarkers of frailty in older people with type 2 diabetes mellitus" (MetaboFrail) study, coupled targeted metabolomics with a chemometric modeling strategy [16]. Through this innovative biomarker discovery strategy, we identified a specific profile of serum amino acids in F-T2DM older people. These findings may o ffer new insights into the metabolic perturbations associated with the disabling cascade in older persons with T2DM.

#### **2. Materials and Methods**

## *2.1. Study Population*

MetaboFrail was developed as an ancillary study of the "Multi-modal Intervention in Diabetes in Frailty" (MID-Frail) project [16,33,34]. The latter was a cluster-randomized multicenter clinical trial that evaluated the e ffectiveness of a multicomponent intervention (mainly based on resistance exercise and lifestyle counseling) on improving physical performance compared with usual care in F-T2DM older adults from seven European countries (ClinicalTrials.gov identifier: NCT01654341) [16,33]. For the present study, a subgroup of MID-Frail participants recruited in Spanish and French study centers were enrolled. The main eligibility criteria were: (a) age at screening 70 years or older; (b) T2DM diagnosis from at least two years; and (c) being pre-frail or frail according to Fried's criteria [35]. The main exclusion criteria were: (a) poor cognition operationalized as a Mini Mental State Examination score <20 [36]; (b) severe disability defined as a Barthel index score <60 [37]; critical conditions and/or major illnesses with a life expectancy <6 months; (c) inability or unwillingness to provide informed consent. Control participants were enrolled at the Università Cattolica del Sacro Cuore (Rome, Italy) and had the following characteristics: 70+ years of age, no T2DM, and no functional impairment. The study protocol was approved by local ethics committees according to both national and international laws. Prior to enrolment, all participants provided written informed consent. The study was conducted in agreemen<sup>t</sup> with legal requirements and international norms (Declaration of Helsinki, 1964).

#### *2.2. Blood Collection and Determination of Serum Concentrations of Amino Acids and Derivatives*

Blood samples were collected after overnight fasting. For serum separation, blood samples were kept on ice for about 30 min until clotting and were subsequently centrifuged at 1000× *g* for 10 min at 4 ◦C. Serum samples were eventually aliquoted and stored at −80 ◦C until analysis.

Concentrations of 37 amino acids and derivatives were determined in serum by ultraperformance liquid chromatography/mass spectrometry (UPLC/MS), as described previously [17]. Briefly, 50 μL of sample was added to 100 μL 10% (w/v) sulfosalicylic acid containing an internal standard mix (50 μM; Cambridge Isotope Laboratories, Inc., Tewksbury, MA, USA) and centrifuged at 1000× *g* for 15 min. Ten μL of the resulting supernatant were mixed with 70 μL of borate buffer and 20 μL of AccQ Tag reagents (Waters Corporation, Milford, MA, USA) and heated at 55 ◦C for 10 min. Samples were eventually loaded onto a CORTECS UPLC C18 column 1.6 μm 2.1 × 150 mm (Waters Corporation) for chromatographic separation (ACQUITY H-Class, Waters Corporation, Milford, MA, USA). Elution was performed at 500 μL/min flow rate with a linear gradient (9 min) from 99:1 to 1:99 water 0.1% formic acid/acetonitrile 0.1% formic acid. Analytes were detected on an ACQUITY QDa single quadrupole mass spectrometer equipped with electrospray source operating in positive mode (Waters Corporation, Milford, MA, USA). Amino acid controls (MCA laboratory of the Queen Beatrix Hospital, Winterswijk, The Netherlands) were used to monitor the analytic process.
