Obesity Surgery (2022) 32:625–633 https://doi.org/10.1007/s11695-021-05815-x ORIGINAL CONTRIBUTIONS The Impact of Bariatric Surgery on the Muscle Mass in Patients with Obesity: 2-Year Follow-up Marta Comas Martínez1,2 · Enzamaria Fidilio Meli1 · Fiorella Palmas Candia1 · Francesca Filippi3 · Ramon Vilallonga4 · Efrain Cordero1 · Irene Hernández1 · Alba Zabalegui Eguinoa1 · Rosa Burgos1 · Anna Vila2 · Rafael Simó1,4,5 · Andreea Ciudin1,4,5 Received: 9 September 2021 / Revised: 11 November 2021 / Accepted: 14 November 2021 / Published online: 30 November 2021 © The Author(s) 2021 Abstract Purpose Bariatric surgery (BS) induces a significant and sustained weight loss in patients with severe obesity (SO). Nev- ertheless, apart from significantly reducing body fat, fat-free mass (FFM) might also be lost. At present, there is little and controversial data in the literature regarding the impact of BS on FFM. In recent years, bioimpedance (BIA) has emerged as a reliable test to assess body composition easily to use in the daily clinical practice. On the bases, the aim of the present study is to evaluate the impact of BS on the FFM, evaluated by means of BIA. Material and Methods This is a prospective, observational study, including consecutive patients with SO that underwent BS between February 2018 and February 2019 at our center. At baseline, 1, 6, 12, and 24 months after the BS, all the patients underwent complete medical history, physical and anthropometric evaluation, and body composition assessment by means of BIA (using Bodystat QuadScan4000®). Results Eighty-five patients with SO were recruited, 72.9% females, aged 45.54 ± 9.98 years, pre-BS BMI 43.87 ± 6.52 kg/ m2. FFM significantly decreased continuously after BS at all timepoints. The loss of FFM 24 months post-BS accounted for approximately 21.71 ± 13.9% of the total weight loss, and was independent of BS technique or protein metabolism. Pre-BS HOMA-IR and FFM were independent predictors of FFM at 24 months. Conclusions Significant and early loss of FFM in patients with SO that undergo BS was seen, not related to protein metabo- lism parameters or the BS technique used, suggesting an independent mechanism. Keywords Morbid obesity · Fat-free mass, Bariatric surgery · Sarcopenic obesity Introduction Key Points • Significant muscle loss occurs starting early after The prevalence of obesity has increased worldwide over the bariatric surgery. last 50 years, reaching pandemic levels, in particular severe • The muscle loss is independent on the bariatric surgery obesity (SO). Obesity represents a significant public health technique. • The muscle mass is related to resting energy expenditure * Rafael Simó of Health Sciences and Welfare, University of Vic - Central rafael.simo@vhir.org University of Catalonia, Barcelona, Spain 3 * Andreea Ciudin Clinical Pharmacology Department, University Hospital Vall aciudin@vhebron.net d’Hebron, Barcelona, Spain 4 1 Endocrine, Metabolic and Bariatric Unit, General Surgery Endocrinology and Nutrition Department, Hospital Department, Hospital Universitari Vall d’Hebron, Universitat Universitari Vall d’Hebron, Diabetes and Metabolism Autònoma de Barcelona, Barcelona, Spain Research Unit, Vall d’Hebron Institut de Recerca (VHIR), 5 Universitat Autònoma de Barcelona, Pg Vall d’Hebron Centro de Investigación Biomédica en Red de Diabetes 119‑129, 08035 Barcelona, Spain y Enfermedades Metabólicas Asociadas (CIBERDEM), 2 Instituto de Salud Carlos III (ISCIII), Madrid, Spain Research Group M3O, Methodology, Methods, Models and Outcomes of Health and Social Sciences, Faculty 13 Vol.:(0123456789) 626 Obesity Surgery (2022) 32:625–633 challenge, and it is associated with a significant economic On these bases, the aims of the present study were (a) to burden on the health systems of developed countries, mainly evaluate the impact of BS on FFM in patients with MO and due to the associated comorbidities, in particular type 2 dia- (b) to explore biomarkers for FFM evolution after the BS. betes (T2D) [1–3]. Additionally, the loss of muscle mass and/or function, also known as sarcopenia, was related to metabolic disorders, like T2D, aging, and poor quality of Material and Methods life. [4, 5]. Sarcopenic obesity (OS) is the combination of low muscle mass and strength with increased fat mass, and We performed a prospective, observational study, includ- it has been associated with adverse health outcomes [6]. A ing patients with SO, consecutively attended at the Mor- recent meta-analysis found that the presence of OS was asso- bid Obesity Unit of our centre, that underwent BS between ciated with a higher risk (OR 1.38, 95% CI [1.27–1.50]) of February 2018 and February 2019. The present study, part T2D than with each condition separately (obesity or sarcope- of the PREDIBAR trial (NCT 03784508), was approved by nia) [7]. The complete underlying mechanism is still unclear, the local Ethics Committee (PR(AG)320/2018) and carried but it seems that there is a bi-directional relationship, having out in accordance with the Declaration of Helsinki. All the as main factors chronic inflammation and insulin resistance patients signed the informed consent form before the inclu- (IR). sion in the study. Physiologically, the maximum level of skeletal muscle Inclusion criteria are as follows: (a) age between 18 and mass and strength is reached between 30 and 50 years of age, 60 years, (b) SO fulfilling criteria for BS according to our and starting from this point, the muscle mass is decreasing protocol (BMI > 40 kg/m2 regardless of the comorbidities or as part of the aging process [8]. Nevertheless, several con- BMI > 35 kg/m2 with at least 1 comorbidity related to obe- ditions can alter the physiological evolution of the muscle sity, (c) previous accomplishment of the preoperative pro- mass at earlier ages, such as associated obesity, rapid weight tocol for BS at our site, (d) written informed consent form. loss after diet, and physical inactivity. Exclusion criteria are as follows: (a) patients undergoing The use of rapid weight-loss diets, such as very-low- evaluation for a second-step surgery; (b) unable to perform calorie diet (VLCD), has been shown to have a significant the post-BS follow-up at our site at least during 2 years; (c) weight loss effect and a significant impact on muscle mass subjects unable perform BIA (e.g. limb amputation, unwill- reduction. The impact on muscle mass has been seen to be ing, and unable to fast for more than 8 hours); (d) presence greater than the impact on fat mass [9]. of other conditions that can affect the muscle mass as per Different methods are available to assess muscle mass investigator criteria (e.g. immobilization, myopathies, and and body composition such as bioimpedance analysis (BIA), endocrinopathy such as Cushing disease); (e) severe con- dual-energy X-ray absorptiometry (DXA), computed tomog- comitant pathology (cardiovascular, cerebrovascular, pul- raphy (CT), and magnetic resonance imaging (MRI). DXA monary, renal, or neoplastic) that can limit the participation is still considered the gold standard method in clinical in the study as per investigator criteria; (f) the use of drugs practice and investigation—and used the diagnosis criteria that can affect the muscle mass (e.g. corticosteroids); (g) EWGSOP2 for sarcopenia [10]. However, DXA is expen- active abuse of drugs or alcohol; (h) uncontrolled psychiatric sive, needs particular space, has subject-related limitations illness or eating disorders. (maximum weight 160 kg), and provides only quantitative All patients underwent baseline, at 1, 6, 12, and 24 skeletal muscle mass evaluation. On the other hand, BIA is months after the BS performance: complete medical history, a relatively simple, quick, and non-expensive method for physical, anthropometric, as per preoperative protocol for assessing body composition [11–13] and provides data on BS at our site. Additionally, biochemical analysis was per- the muscle quantity and estimated quality. formed including HbA1c, insulin levels, homeostatic model Bariatric surgery (BS) is at present the most successful assessment for insulin resistance (HOMA-IR), and sensitive treatment for SO, in terms of significant and sustainable parameters from protein metabolism (transthyretin). weight loss, leading to an important improvement in obesity- related comorbidities and quality of life [14]. At present, • HOMA-IR was calculated using the formula: [Fasting there is little and controversial data regarding the relation- glucose (mmoL/L) * fasting insulin (μUI/mL)]/22.5 ship between bariatric surgery and skeletal muscle mass. • BMI was calculated by the formula: weight (kg)/height Furthermore, the data and the methods that were used in the ­(m2) different studies are heterogeneous (some used muscle mass • %Excess of weight loss (%EWL) was calculated by the estimation based on mathematical equations, while others formula: (W (kg) initial − W (kg) final)/(W (kg) initial use more direct methods) [15, 16]. − W (kg) ideal) * 100 13 Obesity Surgery (2022) 32:625–633 627 • %Total weight loss (%TWL) was calculated by the for- Table 1  The baseline characteristics of the patients included in the mula: (W (kg) initial − W (kg) final)/(W (kg) initial) * study 100 N 85 Body muscle mass was evaluated by means of BIA [17]. Gender (females %) 62 (72.9%) The BIA device used in our study was Bodystat QuadS- Age (years) mean ± SD 45.54 ± 3.38 can4000®, a multi-frequency device. The measurement is Pre-BS BMI (kg/m2) mean ± SD 43.87 ± 6.52 performed by placing two electrodes on the wrist and hand Roux-en-Y gastric bypass (RYGB) (%) 48 (57%) and two electrodes on the foot and ankle on the same side of Sleeve gastrectomy (SG) (%) 36 (43%) the body. BIA can be carried out both in the outpatient clinic HbA1c (% DCCT) mean ± SD 5.89 ± 0.96 HOMA-IR mean ± SD 6.13 ± 4.26 and in the hospitalization floors since these devices can be Type 2 diabetes (%) 40 (47%) transported easily [18]. The patients included in the study Arterial hypertension (%) 37 (43.5%) were asked to fulfil the standardized conditions to perform Hypercholesterolemia (%) 22 (25.9%) a BIA: avoid physical exercise the previous 8 h; fasting 6–8 Sleep apnea (%) 50 (58.8%) h before the measurement, including water; have previously removed the hair from the limbs and all metal objects that BMI, body mass index; BS, bariatric surgery; HOMA-IR, homeostatic may interfere with the measurement; if the patient wears model assessment for insulin resistance; HbA1c (% DCCT), glycated a prosthesis or implant, the measurement is made on the haemoglobin (% Diabetes Control and Complications Trial) opposite side [19]. The variables that were collected from the BIA evaluation evolution of weight, body composition evaluated by BIA were fat mass (FM) (kg), fat-free mass (FFM) (kg), fat-free and biochemical analysis after BS, is shown in Table 2. mass index (FFMI) (kg/m2), and phase angle (PA) (°). The All the patients presented remission of T2D, according to PA is considered an indicator of cellular integrity; it allows ADA guidelines [21] after the BS and was maintained at the interpretation of the capacity of the muscle cell to trans- 24 months. mit the electrical stimulus produced by the BIA apparatus. The FFM significantly decreased 1 month after the BS Low PA values have been correlated with a worse prognosis and continued to drop at 24 months. It should be noted that and higher comorbidities [17] [20]. starting from 1 month until the end of the follow-up, FFM loss represented > 20% of the TWL (Table 2). FM also Statistical Analysis BM SPSS statistical software version significantly decreased starting from 1 until 12 months 24 was used. Continuous variables are expressed as means after the BS. Between months 12 and 24 after the BS, ± standard deviation (SD) for normal distributed variables FM stabilized and presented a slight tendency to increase, and median ± interquartile range (IQR) for non-normal dis- although without reaching statistical significance, as tributed variables. Categorical variables are expressed with reflected by Fig. 1. percentages. For differences between groups in continuous Additionally, basal metabolism rate (BMR), measured by variables ANOVA, Student’s t test or Mann-Whitney U test BIA, significantly decreased 1 month after the BS compared was used while χ2 was used for categorical variables. Spear- to baseline and continued to decrease during the follow-up man correlation and regression logistics analysis were used (Table 2). A positive correlation was found between FFM to explore the relation between different variables. A p-value and the BMR at all timepoints, respectively. By contrast, < 0.05 was considered statistically significant. FM showed no relation with the BMR as reflected by Fig. 2. Regarding the two BS techniques that were used in the study (RYGB and SG), as expected, RYGB was associated Results with greater global EWL and TWL than SG. No significant differences were seen between the two techniques in terms A total of 90 consecutive patients with SO undergoing of body composition and biochemical analysis (HOMA-IR preoperative evaluation for BS at our site were recruited. and transthyretin), except for the PA that was significantly Eighty-five patients underwent BS and had at least 24 lower after the RYGB, as reflected by Table 2. We found months of follow-up at our site and were included in the significantly lower levels of PA after the BS in all patients, study. The baseline characteristics of the patients are shown when compared to data in the literature from subjects with in Table 1. normal weight, of similar age and gender [22] (Table 3). Patients with T2D were treated with diet (100%), met- A multiple regression analysis, including variables as formin (90%), aGLP-1 (62.4%), iSGLT-2 (13.2%), iDPP4 age, HbA1c levels, baseline FFM, baseline FM, baseline (45.1%), and insulin (2.3%). For the calculation of HOMA- transthyretin, and HOMA-IR, was performed. We found that IR, the patients treated with insulin were excluded. The HOMA-IR and baseline FFM were the only independent 13 Table 2  Evolution after bariatric surgery 628 Parameters Baseline 1 month after BS 6 months after BS 12 months after BS 24 months after BS Global RYGB SG Global RYGB SG Global RYGB SG Global RYGB SG Global RYGB SG 13 Fat-free mass 66.47 ± 65.68 ± 66.61 ± 62.03 ± 62.12 ± 61.91 ± 58.94 ± 59.12 ± 58.68 ± 56.87 ± 56.37 ± 57.53 ± 55.48 ± 54.34 ± 55.66 ± (kg)‡ 14.49 4.09 15.46 12.84 a 11.96 14.11 11.89a,b 11.18 12.99 12.60a,b,c 2.02 13.53 13.97a,b,c,e 15.04 12.91 Fat mass (kg)‡ 54.97 ± 52.84 ± 57.21 ± 46.42 ± 44.64 ± 48.79 ± 28.60 ± 26.27 ± 31.89 ± 23.68 ± 21.44 ± 26.64 ± 25.43 ± 8.89a,b,c 24.39 ± 27.78 ± 0.23 15.86 11.32 19.92 13.89a 11.42 16.52 12.68a,b 10.31 14.97 11.99a,b,c 9.77 14.06 7.83 Phase angle (°)‡ 6.79 ± 0.86 6.79 ± 0.91 6.96 ± 0.74 6.26 ± 6.08 ± 6.57 ± 5.54 ± 5.39 ± 0.87 5.76 ± 5.68 ± 1.10a 5.438 ± 5.86 ± 5.78 ± 0.96a 5.63 ± 6.00 ± 0.94h 0.86a 0.84 0.82h 0.78a,b 0.59h 1.04h 0.68h 0.96 Basal metabo- 1960.51 ± 1928.49 ± 1984.25 ± 1838.84 ± 1838.4 ± 1799.6 ± 1777.33 ± 1767.83 ± 1790.74 ± 1726.68 ± 1707.05 ± 1752.61 ± 1722.86 ± 1712.05 ± 1727.06 ± lism rate 378.49 366.01 402.88 361.50a 309.61 425.15 303.71a 282.93 334.79 318.84a,b,c 302.09 343.61 349.15a,b,c 75.20 323.26 (kcal)‡ Fat-free mass NA NA NA 30.88 ± 29.56 ± 32.64 ± 26.46 ± 26.37 ± 26.58 ± 19.47 ± 19.98 ± 20.09 ± 21.71 ± 23.20 ± 20.94 ± loss from 20.74 18.95 23.06 8.76b 10.06 6.79 12.63b,c 10.30 15.19 13.90c,e,f 9.96 17.60 TWL (%)‡ BMI (kg/m2)‡ 43.82 ± 42.47 ± 45.85 ± 38.43 ± 37.71 ± 40.45 ± 31.81 ± 30.38 ± 33.81 ± 29.47 ± 27.89 ± 31.57 ± 28.14 ± 8.51a,b 25.75 ± 31.59 ± 6.26h 6.35 5.06h 7.78h 8.48a 4.84 9.69 5.92a,b 4.43h 7.14h 6.16a,b,c 4.42h 7.48h 9.16h %EWL‡ NA NA NA 24.25 ± 26.85 ± 20.78 ± 65.61 ± 69.59 ± 60.15 ± 79.80 ± 82.07 ± 75.31 ± 78.67 ± 82.29 ± 73.77 ± 14.86 11.58 17.94 22.03b 21.79h 21.48h 24.30b,c 25.14 26.49 26.98b,c,d 24.04h 30.59 %TWL‡ NA NA NA 10.15 ± 10.86 ± 9.20 ± 27.30 ± 28.46 ± 25.7 ± 33.84 ± 7.43b,c 33.84 ± 31.58 ± 33.50 ± 9.57b,c,d 34.39 ± 32.30 ± 4.95 3.46 6.37 6.34b 5.89h 6.66h 7.46 7.88 8.50h 11.01 HbA1c (% 5.97 ± 1.10 6.01 ± 1.17 5.72 ± 0.53 5.69 ± 5.78 ± 5.57 ± 5.26 ± 5.26 ± 0.51 5.26 ± 0.43 5.32 ± 0.47a,b 5.29 ± 0.48 5.32 ± 0.45 5.42 ± 0.64a,b 5.40 ± 5.44 ± 0.57 DCCT) 0.89a 1.09 0.52 0.47a,b 0.69 HOMA-IR 6.31 ± 4.00 5.63 ± 3.34 6.85 ± 5.28 1.43 ± 1.45 ± 2.25 ± 1.68 ± 1.31 ± 2.13 ± 1.59 ± 1.15a,c 1.16 ± 2.32 ± 1.69 ± 1.34a,d 1.45 ± 1.92 ± 1.33 0.48a 0.43 0.92 1.31a,b 0.64h 1.71h 0.53g 1.83g 1.34 Transthyretin 23.65 ± 23.32 ± 23.97 ± 18.89 ± 18.89 ± 20.63 ± 19.86 ± 18.54 ± 21.36 ± 22.56 ± 21.59 ± 23.63 ± 25.24 ± 24.24 ± 26.47 ± 3.89 (mg/dl) 4.60 4.92 4.44 4.58a 4.58h 3.13h 3.86a 4.13g 2.93g 3.64a,b,c 3.86 3.12 4.66a,b,c,d 5.07 (20–40) Albumin (g/dl) 4.14 ± 0.51 4.20 ± 0.21 4.16 ± 0.26 4.22 ± 0.30 4.19 ± 4.25 ± 4.08 ± 4.06 ± 0.26 4.10 ± 0.29 4.06 ± 0.46 4.03 ± 0.60 4.09 ± 0.24 5.22 ± 5.99 6.07 ± 4.17 ± 0.28 (3.5–5.2) 0.22 0.38 0.28b 8.01 FFMI (kg/ ­m2)‡ 23.79 ± 3.2 23.19 ± 24.61 ± 22.30 ± 21.83 ± 22.93 ± 21.30 ± 20.88 ± 21.90 ± 20.75 ± 20.12 ± 21.57 ± 20.20±3.33a,b,c,d 19.54 ± 21.04 ± 3.87 2.74 3.64 2.85a 2.39 3.31 2.63a,b 2.22 3.05 2.98a,b,c 2.55 3.33 2.71 BS, bariatric surgery; BMI, body mass index; EWL, excess of weight loss; TWL, total weight loss; HOMA-IR, homeostatic model assessment for insulin resistance; HbA1c, glycated haemoglobin; FFMI, fat-free mass index a Significantly different with respect to baseline at p < 0.001 b Significantly different with respect to 1 month at p < 0.001 c Significantly different with respect to 6 months at p < 0.001 d Significantly different with respect to 12 months at p < 0.001 e Significantly different with respect to 12 months at p < 0.05 f Significantly different with respect to 1 month at p < 0.05 g Significantly different between bariatric surgery techniques at p < 0.001 h Significantly different between bariatric surgery techniques at p < 0.05 ‡Repeated measures ANOVA, p < 0.05 Obesity Surgery (2022) 32:625–633 Obesity Surgery (2022) 32:625–633 629 Fig. 1  Evolution of the FFM and FM during 24-month follow-up. FFM fat-free mass, FM fat mass, M months—reviewer 2 Fig. 2  Correlation between basal metabolism rate and body composition. BMR basal metabolism rate, BS bariatric surgery predictors of FFM at 24 months after the BS as reflected continued to significantly drop while FM presented a ten- by Table 4. dency to increase, confirming the actual data in the literature [23–25] regarding the lack of precision of BMI to properly monitor the evolution after the BS. Discussion Recently, our group showed that BS induces a significant reduction in basal metabolism rate starting 1 month after the We showed in the present study that BS induces a significant BS in patients with extreme obesity that was associated with and early loss of FFM after the BS, independent of protein weight regain at 5-year follow-up [26]. It should be noted metabolism, without significant differences between the BS that in the study performed by Fidilio et al. [26], the BMR techniques (RYGB and SG). In our study, BMI stabilized was measured by means of indirect calorimetry in patients after 12 months post-BS. In exchange, body composition with extreme obesity (BMI > 50 kg/m2) and study of body showed different profiles between 12 and 24 months: FFM composition was not performed. In the present study, we 13 630 Obesity Surgery (2022) 32:625–633 Table 3  The BS impact on the Phase angle levels in all patients Baseline 1 month 6 month 12 months 24 months Normal values [22] p‡ Females 18–20 years 7.86 ± 1.06 6.46 ± 0.3a 6.25 ± 0.59a,b 6.15 ± 0.6a,b,c 6.13 ± 0.38a,b,c,d 7.04 ± 0.85 0.001 21–29 years 7.4 ± 0.74 6.14 ± 0.85a 5.80 ± 0.55a,b 5.9 ± 0.76a,b,c 5.68 ± 0.53a,b,c,d 6.98 ± 0.92 0.0021 30–39 years 6.6 ± 0.59 5.96 ± 0.76a 5.39 ± 0.66a,b 5.15 ± 0.77a,b,c 5.07 ± 0.80a,b,c,d 6.91 ± 0.85 < 0.001 40–49 years 6.12 ± 0.9 5.7 ± 0.12a 4.89 ± 0.85a,b 4.85 ± 1.04a,b,c 5.07 ± 0.16a,b,c,d 6.87 ± 0.84 < 0.001 Males 18–20 years NA NA NA NA NA 7.9 ± 0.47 NA 21–29 years 7.7 ± 0.3 7.3 ± 1.12a 7.05 ± 0.3a,b 7.15 ± 0.4a,b,c 6.35 ± 0.91a,b,c,d 8.02 ± 0.75 < 0.001 30–39 years 7 ± 0.56 6.7 ± 0.01a 6.45 ± 1.47a,b 6.25 ± 0.75a,b,c 6.14 ± 0.54a,b,c,d 8.01 ± 0.85 < 0.001 40–49 years 6.75 ± 0.79 6.85 ± 0.46a 5.63 ± 0.68a,b 5.58 ± 0.61a,b,c 5.73 ± 0.70a,b,c,d 7.76 ± 0.85 < 0.001 a Significantly different respect to baseline at p < 0.001; bsignificantly different with respect to 1 month at p < 0.001; csignificantly different with respect to 6 months at p < 0.001; dsignificantly different with respect to 12 months at p < 0.001; ‡repeated measures ANOVA, p < 0.05 Table 4  Independent parameters related to the FFM at 24 months after the BS Coefficientsa Model Unstandardized coef- Standardized t Sig. 95.0% confidence interval for B Collinearity statistics ficients coefficients B Std. error Beta Lower bound Upper bound Tolerance VIF 1 (constant) − 8.750 8.300 − 1.054 .300 − 25.725 8.225 Age .009 .076 .006 .116 .908 − .147 .164 .884 1.131 FFM (kg) .843 .047 .933 17.966 .000 .747 .939 .813 1.230 Transthyretin before − .024 .145 − .008 − .162 .872 − .320 .273 .812 1.231 HOMA-IR .409 .192 .102 2.132 .042 .017 .802 .966 1.036 HbA1c levels .800 .610 .065 1.313 .200 − .447 2.048 .891 1.122 FM (kg) .041 .056 .040 .722 .476 − .075 .729 .729 1.371 a Dependent variable FFM (kg) 24 months FFM fat-free mass, FM fat mass, M months found also a significant and early reduction in BMR start- = FFM > 15% of TWL. In our study, we found reduction in ing 1 month after the BS, although measured by BIA that FFT > 20% of TWL at all timepoints, reaching 25% after was significantly correlated with FFM at all timepoints and 24 months, independently of the BS technique (RYGB or showed no relation with the FM. These findings suggest SG) which according to present published data is considered the role of FFM in the weight homeostasis and metabolism significant muscle loss. and point out to the importance of centring the attention Besides FFM and FM parameters that are usually evalu- on the FFM and a more personalized approach of patients ated in body composition studies, recently, a new parameter with obesity that are proposed for weight loss interventions. has emerged as a more sensitive biomarker of muscle mass, Nevertheless, at present, there is no data in large cohorts of in particular quality: the phase angle (PA) [27]. PA makes patients on the role of the FFM in the evolution after weight possible to interpret the muscle cell’s capacity to transmit loss interventions, since body composition is not usually the electrical current produced by the BIA device accord- taken into consideration in the daily clinical practice. This ing to the cell’s quality. PA is an attractive index, because fact represents an important gap in the management of the it is independent of body weight and was associated with patients with obesity that urgently needs to be filled. poor health outcomes, quality of life, and mortality [28]. It Additionally, there is very scarce data regarding the should be noted that at present, there is very few data in the impact of BS on the FFM evolution. Vaurs et al. [16] identi- literature regarding PA normal cut-offs. Most of the studies fied two phenotypes of FFM evolution after BS and defined were performed including patients in a critical state in the as “muscle spare” = FFM < 15% of TWL and “muscle loss” intensive care units or with cancer, showing levels around 13 Obesity Surgery (2022) 32:625–633 631 4–5° [28, 29]. In a study of 1967 healthy subjects from the the same study, treatment with rosiglitazone (a drug that USA, the PA of age-matched individuals was 6.96° ± 1.10° decreases IR) induced only a partial recovery of the muscle for males and 5.97° ± 0.83° for females [30]. A larger study mass and the authors hypothesized that maybe because it did published by Bosy-Westphal et al. [22], including data not interfere with protein synthesis. from 210,000 healthy German individuals, proposed the These previous data in murine models explain our find- cut-off for PA of 6.01° ± 0.75° for males and 5.59° ± 0.72° ings in the present study. Pre-BS IR might have altered the for females, with tenth percentile values of 5.14 and 4.79, muscle mass quantity and quality and the impact of signifi- respectively. A value below this cut-off is considered sar- cant rapid metabolic and body weight changes induced by copenia and low muscle mass and also quality. It should the BS have led to a significant and continued muscle mass be noted that none of these studies included patients with loss. Furthermore, as seen in the murine model [33], the obesity or overweight. muscle mass was not recovered after the normalization of At present, there is no data regarding the PA in subjects the IR. We observed the same evolution in our study, where with obesity and/or the impact of BS on the muscle quality the FFM continued to drop out at 24 months even if the evaluated by this parameter. As reflected by Table 3 from the HOMA-IR normalized after the first month and remained results, in our study, we found PA levels at baseline compa- within normal limits for Spanish population [34] during the rable to those reported in the German study [22], although follow-up. Our findings suggest that the mechanisms of sig- slightly lower. Nevertheless, our study was not designed to nificant muscle mass loss after BS seem independent on the evaluate sarcopenia pre-BS and we cannot make any hypoth- BS techniques or protein metabolism parameters, like tran- esis on the pre-BS FFM status of our patients. Actually, at sthyretin, but are far from being elucidated. This is the first present, this point represents another important gap in the study that shows a significant FFM loss after BS maintained management of the patients with obesity, since there is no after 24 months, having independent predictors pre-BS FFM reliable data regarding the prevalence of sarcopenia in these and IR. The main limitations of our study are the lack of a subjects, especially the younger ones that are usually candi- control group of subjects with normal weight and the sample dates for BS. Application of different criteria to identify sar- size. Nevertheless, the patients represented their own control copenia associated to obesity may therefore currently lead to and the study was aimed to evaluate the evolution during substantially and clinically unacceptable variable prevalence follow-up. For this reason, we consider that the main limita- levels (ranged from 8 to over 50%) due the heterogeneity tions had no significant impact on the results. in the methods that were used [6]. In order to fill this gap These findings point to the urgent need to take into con- further, large case control studies are needed in order to dis- sideration changing the actual clinical guidelines for the pose of reliable data. In exchange, the PA was significantly management of patients with obesity, especially those that lower after the BS at all timepoints compared to baseline are candidate for BS, by incorporating body composition and significantly below the normal cut-off for both females studies (both quantitative and qualitative) in the routinary and males proposed by Bosy-Westphal et al. [22], suggesting preoperatory and follow-up evaluations. BIA is a rapid, reli- sarcopenia. Furthermore, PA was significantly lower in the able, repeatable, and non-expensive test that can be easily patients that underwent BYGR, despite similar FFM with implemented in the daily clinical practice, as part of the SG, suggesting that RYBG might have a greater impact on evaluation of patients with obesity. This action will allow to the muscle quality after the BS. implement more personalized approach, and design-specific As reflected by the logistic regression analysis, the impor- physical exercises, diet, and pharmacological therapy aimed tant loss of FFM after the BS was independent of the age, at improving the pre-BS muscle mass and insulin resistance. gender, protein metabolism, or BS techniques. The only parameters associated with the FFM at 24 months after the BS in our study were pre-BS FFM and pre-BS HOMA-IR, Conclusions suggesting that pre-BS conditions have an important influ- ence on the evolution after the BS. Previous data in the lit- There is a significant and early loss of FFM in patients with erature showed that insulin resistance (IR) has an indirect SO who undergo BS, which is not related to the parameters relationship with the muscle mass in humans [31]. From a of protein metabolism or the surgical technique used, sug- mechanistic point of view, interestingly, in murine models, gesting the existence of an independent mechanism. Ostler et al. [32] demonstrated that IR, in a context of leptin signalling impairment and inflammation (that are also pre- sent in humans with obesity), induces a significant decrease Author Contribution Conceptualization, A.C.; methodology, A.C, M.C., A.V.; software, M.C., E.F.; validation, A.C. and R.S.; formal in muscle size and quality. Additionally, another study in analysis, M.C., E.F., A.C.; investigation, M.C., E.F., F.P., F.F., R.V., murine models performed by Wang et al. [33] showed that E.C., I.H., R.B., A.Z., A.V.; resources, A.C., R.B., R.V.; data cura- IR causes muscle protein degradation. Furthermore, in tion, M.C., E.F., A.C.; writing—original draft preparation, M.C.; 13 632 Obesity Surgery (2022) 32:625–633 writing—review and editing, A.C., R.S.; visualization, A.C., R.S., Topinková E, Maurits Vandewoude MZ. Sarcopenia: con- A.C.; supervision, A.C.; project administration, A.C.; funding acqui- senso europeo sobre su definición y diagnóstico. Age Ageing. sition: A.C. All authors have read and agreed to the published version 2010;39(4):412–23. of the manuscript. 11. Pateyjohns IR, Brinkworth GD, Buckley JD, Noakes M, Clifton PM. Comparison of three bioelectrical impedance Funding Open Access Funding provided by Universitat Autonoma de methods with DXA in overweight and obese men. Obesity. Barcelona. This study was supported by grants from the Instituto de 2006;14(11):2064–70. Salud Carlos III (Fondo de Investigación Sanitaria, PI20/01806). The 12. Ruiz De Eguilaz MH, Martínez De Morentín B, Pérez-Diez S, funders had no role in study design, data collection and analysis, deci- Navas-Carretero S, Martínez JA. Estudio comparativo de medi- sion to publish, or preparation of the manuscript. das de composición corporal por absorciometría dual de rayos X, bioimpedancia y pliegues cutańeos en mujeres. An la Real Acad Nac Farm. 2010;76(2):209–22. Declarations 13. Lee K, Shin Y, Huh J, Sung YS, Lee IS, Yoon KH, et al. Recent issues on body composition imaging for sarcopenia evaluation. Conflict of Interest The authors declare no competing interests. Korean J Radiol. 2019;20(2):205–17. 14. Sjöström L et al. for the SOSS. Effects of bariatric surgery on Open Access This article is licensed under a Creative Commons Attri- mortality in Swedish obese subjects. 2007. bution 4.0 International License, which permits use, sharing, adapta- 15. Mastino D, Robert M, Betry C, Laville M, Gouillat C, Disse E. tion, distribution and reproduction in any medium or format, as long Bariatric surgery outcomes in sarcopenic obesity. Obes Surg. as you give appropriate credit to the original author(s) and the source, 2016;26(10):2355–62. provide a link to the Creative Commons licence, and indicate if changes 16. Vaurs C, Diméglio C, Charras L, Anduze Y, Chalret du Rieu M, were made. The images or other third party material in this article are Ritz P. Determinants of changes in muscle mass after bariatric included in the article's Creative Commons licence, unless indicated surgery. Diabetes Metab [Internet] 2015;41(5):416–421. Available otherwise in a credit line to the material. If material is not included in from: https://​doi.​org/​10.​1016/j.​diabet.​2015.​04.​003 the article's Creative Commons licence and your intended use is not 17. García Almeida JM, García García C, Bellido Castañeda V, Bel- permitted by statutory regulation or exceeds the permitted use, you will lido GD. Nuevo enfoque de la nutrición. Valoración del estado need to obtain permission directly from the copyright holder. To view a nutricional del paciente: función y composición corporal. Nutr copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. Hosp. 2018;35(3):1–14. 18. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM, et al. Bioelectrical impedance analysis - part I: review of principles and methods. Clin Nutr. 2004;23(5):1226–43. 19. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, References Gómez JM, et al. Bioelectrical impedance analysis - part II: utili- zation in clinical practice. Clin Nutr. 2004;23(6):1430–53. 1. Stensel D. Obesity and diabetes. Exerc Physiol Spec Popul. 20. da Silva BR, Gonzalez MC, Cereda E, Prado CM. Exploring the 2008;13(2):21–49. potential role of phase angle as a marker of oxidative stress: a 2. Mc Farlane SI. Obesity, obstructive sleep apnea and type 2 diabe- narrative review. Nutrition [Internet]. 2022;93:111493. Available tes mellitus: epidemiology and pathophysiologic insights. Sleep from: https://​doi.​org/​10.​1016/j.​nut.​2021.​111493 Med Disord Int J. 2018;2(3):52–8. 21. Riddle MC, Cefalu WT, Evans PH, Gerstein HC, Nauck MA, 3. Kahn BB, Flier JS. On diabetes: insulin resistance Obesity and Oh WK, et al. Consensus report: definition and interpretation of insulin resistance. J Clin Invest. 2000;106(4):473–81. remission in type 2 diabetes. 2021;1–7. 4. Son JW, Lee SS, Kim SR, Yoo SJ, Cha BY, Son HY, et al. 22. Anja BW, Danielzik S, Dörhöfer RP, Later W, Wiese S, Müller Low muscle mass and risk of type 2 diabetes in middle-aged MJ. Phase angle from bioelectrical impedance analysis: popula- and older adults: findings from the KoGES. Diabetologia. tion reference values by age, sex, and body mass index. J Parenter 2017;60(5):865–72. Enter Nutr. 2006;30(4):309–16. 5. Janssen I, Ross R. Linking age-related changes in skeletal muscle 23. Rothman KJ. BMI-related errors in the measurement of obesity. mass and composition with metabolism and disease. J Nutr Heal Int J Obes. 2008;32:S56–9. Aging. 2005;9(6):408–19. 24. Okorodudu DO, Jumean MF, Montori VM, Romero-Corral A, 6. Johnson Stoklossa CA, Sharma AM, Forhan M, Siervo M, Pad- Somers VK, Erwin PJ, et al. Diagnostic performance of body wal RS, Prado CM. Prevalence of sarcopenic obesity in adults mass index to identify obesity as defined by body adiposity: with class II/III obesity using different diagnostic criteria. J Nutr a systematic review and meta-analysis. Int J Obes [Internet] Metab. 2017;2017. 2010;34(5):791–799. Available from: https://​doi.​org/​10.​1038/​ 7. Khadra D, Itani L, Tannir H, Kreidieh D, El MD, El GM. Associa- ijo.​2010.5 tion between sarcopenic obesity and higher risk of type 2 diabetes 25. Segal-Lieberman G, Segal P, Dicker D. Revisiting the role in adults: a systematic review and meta-analysis. World J Diabe- of bmi in the guidelines for bariatric surgery. Diabetes Care. tes. 2019;10(5):311–23. 2016;39(August):S268–73. 8. Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Ceder- 26. Fidilio E, Comas M, Giribés M, Cárdenas G, Vilallonga R, Palma holm T, et al. Sarcopenia: revised European consensus on defini- F, et al. Evaluation of resting energy expenditure in subjects with tion and diagnosis. Age Ageing. 2019;48(1):16–31. severe obesity and its evolution after bariatric surgery. Obes Surg. 9. Sivakumar J, Chong L, Ward S, Sutherland TR, Read M, Hii 2021 MW. Body composition changes following a very-low-calorie 27. Thibault R, Makhlouf AM, Mulliez A, Cristina Gonzalez M, pre-operative diet in patients undergoing bariatric surgery. Obes Kekstas G, Kozjek NR, et al. Fat-free mass at admission predicts Surg. 2020;30(1):119–26. 28-day mortality in intensive care unit patients: the international 10. Cruz Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm prospective observational study Phase Angle Project. Intensive T, Landi F, Martin FC, Michel JP, Rolland Y, Schneider SM, Care Med. 2016;42(9):1445–53. 13 Obesity Surgery (2022) 32:625–633 633 28. Alves FD, Souza GC, Clausell N, Biolo A. Prognostic role of 32. Ostler JE, Maurya SK, Dials J, Roof SR, Devor ST, Ziolo MT, phase angle in hospitalized patients with acute decompensated et al. Effects of insulin resistance on skeletal muscle growth and heart failure. Clin Nutr. 2016;35(6):1530–4. exercise capacity in type 2 diabetic mouse models. Am J Physiol 29. Norman K, Stobäus N, Zocher D, Bosy-Westphal A, Szramek A, - Endocrinol Metab. 2014;306(6):592–605. Scheufele R, et al. Cutoff percentiles of bioelectrical phase angle 33. Wang X, Hu Z, Hu J, Du J, Mitch WE. Insulin resistance acceler- predict functionality, quality of life, and mortality in patients with ates muscle protein degradation: activation of the ubiquitin-pro- cancer. Am J Clin Nutr. 2010;92(3):612–9. teasome pathway by defects in muscle cell signaling. Endocrinol- 30. Barbosa-Silva MCG, Barros AJD, Wang J, Heymsfield SB, ogy. 2006;147(9):4160–8. Pierson RN. Bioelectrical impedance analysis: population ref- 34. Gayoso-Diz P, Otero-González A, Rodriguez-Alvarez MX, Gude erence values for phase angle by age and sex. Am J Clin Nutr. F, García F, De Francisco A, et al. Insulin resistance (HOMA-IR) 2005;82(1):49–52. cut-off values and the metabolic syndrome in a general adult popu- 31. Srikanthan P, Karlamangla AS. Relative muscle mass is inversely lation: effect of gender and age: EPIRCE cross-sectional study. associated with insulin resistance and prediabetes. Findings from BMC Endocr Disord. 2013;13(Cvd). the Third National Health and Nutrition Examination Survey. J Clin Endocrinol Metab. 2011;96(9):2898–903. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 13