Abstract
Background & aims
The Strength, Assistance with Walking, Rising from a Chair, and Falls (SARC-F) questionnaire was proposed to screen sarcopenia. However, its limited sensitivity may compromise its prognostic value. Since its introduction, several adapted SARC-F-based approaches have been proposed, although their ability to predict clinical outcomes among older hospitalized individuals remains underexplored. This study aimed to investigate the prognostic value of SARC-F and six modified versions in a cohort of hospitalized individuals.
Methods
Cohort study with prospective data collection. Older adults (N = 236, 69.8 ± 7.5 years old) hospitalized in medical or surgical wards were included. Outcomes of interest were 1-year mortality and prolonged length of hospital stay (LOS). The original SARC-F was applied. Additional adapted approaches were explored: (1) SARC-CalF (including calf circumference, CC); (2) BMI−adjusted SARC-CalF (using BMI-adjusted CC); (3) SARC-F + mid-upper arm circumference (MUAC); (4) SARC-F + age and body mass (SARC-F EBM); (5) SARC-CalF + MUAC; and (6) BMI−adjusted SARC-CalF + MUAC. Suggestive signs of sarcopenia were SARC-F scores ≥ 4, SARC-F EBM scores ≥ 12, and scores ≥ 11 for all other SARC-based approaches.
Results
SARC-F EBM was the approach identifying the lowest rate of suggestive signs of sarcopenia (25.9%), while BMI−adjusted SARC-CalF + MUAC identified the highest rate (79.2%). All SARC-F-based approaches were associated with higher odds of prolonged LOS (OR ranged from 2.05 to 3.31). SARC-CalF model had the highest adjusted AUC for LOS prediction (0.64). Except for SARC-F and SARC-F + EBM, all approaches were associated with a higher risk of 1-year mortality (HR adjusted ranged from 2.07 to 11.34). BMI−adjusted SARC-CalF + MUAC exhibited the highest mortality prediction based on the C-index (0.75).
Conclusion
Our study demonstrates that all SARC-F-based approaches are linked to prolonged LOS, and almost all approaches were also linked to 1-year mortality. We also introduce the novel BMI−adjusted SARC-CalF + MUAC version, pending future validation.
Keywords: Aging, Anthropometry, Arm circumference, Calf circumference, Sarcopenia screening, Survival.
Introduction
Sarcopenia is a skeletal muscle disease characterized by the accelerated loss of muscle strength and mass, primarily associated with aging, although not limited to it [1–3]. It is a well-established condition linked to impairments in functional capacity and quality of life, as well as an increased risk of adverse outcomes, including falls, fractures, and higher mortality rates [4–7]. Although our conceptual understanding of sarcopenia and its diagnostic approaches has evolved since the term was first coined in 1989 [1], challenges in its screening and assessment in clinical practice persist [8].
In this context, the SARC-F (Strength, Assistance with walking, Rising from a chair, Climbing stairs, and Falls) questionnaire has been developed as a potential rapid diagnostic tool for sarcopenia. It assesses five key components: physical function (strength), physical performance (autonomy in walking, rising from a chair, and climbing stairs), and the occurrence of falls within the past year [9, 10]. The SARC-F has demonstrated relevant prognostic value for predicting functional and clinical outcomes, including poor motor performance, prolonged length of hospital stay (LOS), and hospital readmissions [10–12]. Although promising as a prognostic tool due to its time- and user-friendly design, SARC-F has shown poor sensitivity for detecting sarcopenia [13, 14].
Years after the initial proposal of the SARC-F, calf circumference (CC) was incorporated into the tool, now termed SARC-CalF, improving its sensitivity for sarcopenia screening [14]. This modification also enhanced its prognostic value, including better survival predictions across different conditions, such as cancer and patients undergoing hemodialysis [15, 16]. Subsequently, additional SARC-F modified versions have been proposed, including SARC-F combined with mid-upper arm circumference (MUAC) [17], incorporating both CC and MUAC [17], and SARC-F combined with age and body mass (termed SARC-F EBM) [18]. Given that a body mass index (BMI) adjustment has been recommended for CC in patients with excess weight, a BMI-adjusted version of the SARC-CalF has also been explored [19, 20]. Taken together, these adaptations share a common feature: the addition of anthropometric measures to the original SARC-F, primarily serving as proxies for muscle mass, a key component in the diagnosis of sarcopenia.
Despite the availability of several modified versions of the SARC-F, their prognostic value remains unexplored among older hospitalized patients with diverse clinical conditions. A comprehensive analysis comparing all available modified versions of the SARC-F may provide valuable insights into the potential additive prognostic value of incorporating different measures into this original tool. The findings of this analysis can contribute to the growing body of evidence supporting the refinement of the SARC-F tool. Therefore, this study aimed to investigate the prevalence of suggestive signs of sarcopenia using the SARC-F and six modified versions incorporating age and/or anthropometric measure(s), as well as to evaluate their predictive value for adverse clinical outcomes among older hospitalized patients.
Methods
Study design and participants
This was a secondary pooled analysis of cohort studies with prospective data collection. Data collection occurred between May 2021 and August 2023 at Hospital das Clínicas de Pernambuco, Recife, PE, Brazil. This is a general tertiary hospital with approximately 400 beds. Not all beds are dedicated to older adults. Intensive care, obstetric, and pediatric units were excluded from consideration. Of the total, 11 wards are allocated for the hospitalization of adult and older patients, covering medical or surgical specialties. Older adults (aged 60 years and older) hospitalized in these medical or surgical (preoperative or late postoperative) wards were eligible. Patients undergoing hemodialysis, those hospitalized for postoperative trauma, or admitted to the intensive care unit (ICU) were excluded due to their influence on anthropometric markers. Individuals with conditions that impaired anthropometric assessments (e.g., edema in both legs, ascites, or anasarca), unable to sit, or those with cognitive impairment that prevented adequate interview responses were also excluded. Except for outcomes, all assessments were conducted within 72 h of hospital admission by five expert-trained clinical dietitians, as part of their practical routine.
The study was approved by the Institutional Ethics Committee (Protocols: 60681122.3.0000.8807, 42739120.5.0000.8807), in compliance with the Declaration of Helsinki (1964), its subsequent revisions, and Resolution no. 466/2012 of the National Health Council. All patients provided their written informed consent. This manuscript was reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement guidelines [21].
Covariates and outcomes
Age, sex, and clinical characteristics were assessed through bedside interviews and cross-verified using electronic medical records. Clinical characteristics included comorbidities (systemic arterial hypertension and diabetes mellitus) and the primary clinical diagnoses responsible for hospitalization. Clinical diagnoses were categorized as oncological, cardiovascular, elective surgical procedures (e.g., hernia or cholelithiasis), genitourinary conditions, and other diseases (including infectious, parasitic, neurological, and rheumatological conditions).
Patients were followed throughout their hospitalization and up to 12 months after discharge to assess clinical outcomes. The primary outcome of interest was 1-year mortality rate. In-hospital mortality was rare (n < 5), precluding a specific analysis of this outcome. The secondary outcome was prolonged LOS, defined as hospitalization longer than the median duration (> 8 days). Outcomes were primarily assessed using electronic medical records. For mortality rate, additional attempts were made to contact patients or their relatives by telephone to cross-verify the medical records; however, successful contact was achieved for only approximately 50% of the cohort.
Anthropometry
Body weight (kg) and height (m) were measured to calculate BMI, kg/m2, classified according to the World Health Organization criteria: underweight if < 18.5 kg/m2, normal range if 18.5–24.9 kg/m2, overweight if 25–29.9 kg/m2, obesity ≥ 30 kg/m2 [22]. CC was measured using an inelastic tape at the point of maximum circumference. Patients were positioned seated with both knees and ankles flexed at a 90-degree angle [22]. Measurements were consistently taken on the right side; in cases of mild edema, the non-edematous calf was measured instead. For patients with excess weight, CC measurements were corrected by subtracting 3 cm for BMIs 25–29.9 kg/m2, 7 cm for BMIs 30–39.9 kg/m2, and 12 cm for BMIs ≥ 40 kg/m2 [23]. MUAC was measured on the patient’s right arm, at the midpoint between the acromion and the olecranon, with the arm relaxed alongside the body [22]. In case of venous access, the other arm was measured. Cutoff values applied for CC and MUAC will be described in the next subsection.
SARC-F and its modified versions incorporating anthropometric measure(s)
The original SARC-F and its six modified versions incorporating anthropometric measure(s) were included in this exploratory study. As earlier mentioned, the SARC-F questionnaire consists of five questions related to strength, assistance with walking, rising from a chair, and falls, with scores ≥ 4 indicating suggestive signs of sarcopenia [9]. SARC-CalF incorporates CC into the SARC-F by adding 10 points for patients with low CC values (≤ 34 cm for males and ≤ 33 cm for females) [14]. In this SARC-CalF version, scores ≥ 11 indicate suggestive signs of sarcopenia [14]. For SARC-F + MUAC, cutoffs used were: ≤ 29.5 cm for males, ≤ 28.4 cm for females, and scores ≥ 11 were also suggestive signs of sarcopenia. For SARC-F + EBM, including age ≥ 75 and/or BMI ≤ 21 kg/m2, scores ≥ 12 indicate suggestive signs of sarcopenia [10, 14, 17, 18]. For surgical patients, SARC-F questions were referred to their usual status before surgery. A detailed description of each approach is provided in Fig. 1.
Fig. 1.
Detailed graphical summary of the SARC-F and its six modified versions used in this study. Cutoffs are from [10, 14, 17, 18]. Abbreviations: BMI: body mass index; EBM: “elderly + body mass”; MUAC: mid-upper arm circumference
Data analyses
Data were analyzed using Stata version 15 (StataCorp LLC® Texas, USA). Sample’s characteristics were described as absolute (N) and relative (%) frequencies (categorical variables) and were compared using Pearson’s or Likelihood ratio χ2 tests. Kaplan-Meier (KM) curves were used to illustrate and assess survival probabilities (%) across all methods during follow-up, and differences between groups were assessed using the log-rank test. Univariate survival and LOS predictive power for SARC-F and modified versions were estimated using a Receiver Operating Characteristic (ROC) test. Performance accuracy was categorized as excellent for area under the ROC (AUROC) values: 0.90 and 1.00, good: 0.80 and 0.90, fair: 0.70 and 0.80, poor: 0.60 and 0.70, and failed: 0.50 and 0.60 [24].
Cox regression analysis was performed to evaluate the associations of suggestive signs of sarcopenia by SARC-F and its modified versions with 1-year mortality, estimating hazard ratios (HR) and their 95% confidence intervals (CI). A satisfactory prognostic value was achieved if HR from Cox regression analysis was > 2, as previously recommended by a nutrition society [25]. Proportional hazards assumption was not violated for any of the SARC-based approaches, as assessed by Schoenfeld residuals: SARC-CalF (P = 0.8317), SARC-F (P = 0.5751), BMI−adjusted SARC-CalF (P = 0.5888), SARC-F + MUAC (P = 0.4808), SARC-F + EBM (P = 0.1122), SARC-CalF + MUAC (P = 0.2617), and BMI−adjusted SARC-CalF + MUAC (P = 0.2906). Model discrimination (primary measure) for survival outcomes was evaluated using Harrell’s C index, with values > 0.70 considered satisfactory for predictive accuracy.
Adjustment covariates were chosen based on well-established prognostic factors available in our dataset and included age, clinical diagnoses, and collected comorbidities (i.e., systemic arterial hypertension and type 2 diabetes mellitus). Binary logistic regression assessed the associations between SARC-F approaches and prolonged LOS, estimating the odds ratio (OR) and its 95% CIs. Adjustment factors were the same aforementioned, and models’ predictive accuracy was assessed using AUC from logistic regression results. There was no missing data, and statistical significance was defined as P < 0.05 for all tests.
Sample size
Considering this is a secondary analysis, no formal sample size calculation was performed before the study. All patients who met the eligibility criteria during the study period were included, ensuring the use of the available data.
Results
A total of 236 older patients were included in this study (57.6% male; 57.6% aged 60–69 years, mean age: 69.8 ± 7.5 years old). Systemic arterial hypertension was the most frequent comorbidity (72%), while cancer (36.9%) and cardiovascular diseases (19.9%) were the leading causes of hospitalization. Overall, 55.6% of the sample had excess weight by BMI and had their CC adjusted accordingly. Low unadjusted CC was identified in 58.9% of patients, increasing to 86.4% after BMI adjustment (BMI−adjusted CC). Low MUAC was identified in 51.3% of the cohort. Prolonged LOS was observed in 46.2% of the sample (n = 109), and the 1-year mortality rate was 18.2% (n = 43). Excepting SARC-F (non-significant), the 1-year mortality rate was higher across all SARC-based approaches. Almost all patients who died were classified with suggestive signs of sarcopenia using the BMI−adjusted SARC-CalF + MUAC approach, excepting one (P = 0.001). Differences were also observed for all SARC-based approaches in relation to prolonged LOS. Detailed clinical characteristics of the cohort and comparisons based on outcomes are presented in Table 1.
Table 1.
Patients’ characteristics by outcomes (N = 236)
| Variables | Total | Survivors (n = 193) |
Non-survivors (n = 43) |
P | LOS ≤ 8 days (n = 127) |
LOS > 8 days (n = 109) |
P |
|---|---|---|---|---|---|---|---|
| Males, n % | 136 (57.6) | 112 (58.0) | 24 (55.8) | 0.790 | 75 (59.1) | 61 (55.9) | 0.632 |
| Age categories, n % | 0.170 | 0.337 | |||||
| 60–69 | 136 (57.6) | 116 (60.1) | 20 (46.5) | 76 (59.8) | 60 (55.1) | ||
| 70–79 | 75 (31.8) | 60 (31.1) | 15 (34.9) | 41 (32.3) | 34 (31.2) | ||
| ≥ 80 | 25 (10.6) | 17 (8.8) | 8 (18.6) | 10 (7.9) | 15 (13.8) | ||
| Systemic arterial hypertension, n % | 170 (72.0) | 143 (74.1) | 27 (62.8) | 0.135 | 92 (72.4) | 78 (71.6) | 0.880 |
| Type 2 diabetes mellitus, n % | 102 (43.2) | 86 (44.6) | 16 (37.2) | 0.379 | 54 (42.5) | 48 (44.0) | 0.815 |
| Diagnoses, n % | 0.001* | 0.005 | |||||
| Oncological | 87 (36.9) | 60 (31.1) | 27 (62.8) | 47 (37.0) | 40 (36.7) | ||
| Cardiovascular | 47 (19.9) | 39 (19.7) | 9 (20.9) | 15 (11.8) | 32 (29.4) | ||
| Elective surgeries | 18 (7.6) | 18 (9.3) | 0 (0.0) | 14 (11.0) | 4 (3.7) | ||
| Genitourinary | 23 (9.8) | 20 (10.4) | 3 (7.0) | 15 (11.8) | 8 (7.3) | ||
| Others | 61 (25.8) | 57 (29.5) | 4 (9.3) | 36 (28.4) | 25 (22.9) | ||
| Nutritional features | |||||||
| BMI (kg/m2), n % | 0.091* | 0.032 | |||||
| < 18.5 | 16 (6.8) | 10 (5.2) | 6 (14.0) | 5 (3.9) | 11 (10.1) | ||
| 18.5–24.9 | 89 (37.7) | 70 (36.3) | 19 (44.2) | 40 (31.5) | 49 (45.0) | ||
| 25-29.9 | 76 (32.2) | 63 (32.6) | 13 (30.2) | 46 (36.2) | 30 (27.5) | ||
| 30-39.9 | 50 (21.2) | 45 (23.3) | 5 (11.6) | 33 (26.0) | 17 (15.6) | ||
| ≥ 40 | 5 (2.2) | 5 (2.6) | 0 (0.0) | 3 (2.4) | 2 (1.8) | ||
| Low CC, n % | 139 (58.9) | 107 (55.4) | 32 (74.4) | 0.022 | 62 (48.8) | 77 (70.6) | 0.001 |
| Low BMI−adjusted CC, n % | 204 (86.4) | 166 (86.0) | 38 (88.4) | 0.682 | 106 (83.4) | 98 (89.9) | 0.149 |
| Low MUAC, n % | 121 (51.3) | 90 (46.6) | 31 (72.1) | 0.003 | 57 (44.9) | 64 (58.7) | 0.034 |
| Suggestive signs of sarcopenia, n % | |||||||
| SARC-F | 90 (38.1) | 68 (35.2) | 22 (51.2) | 0.052 | 37 (29.1) | 53 (48.6) | 0.002 |
| SARC-CalF | 100 (42.4) | 74 (38.3) | 26 (60.5) | 0.008 | 38 (29.9) | 62 (56.9) | < 0.001 |
| BMI−adjusted SARC-CalF | 147 (62.3) | 114 (59.1) | 33 (76.7) | 0.031 | 68 (53.5) | 79 (72.5) | 0.003 |
| SARC-F + MUAC | 88 (37.3) | 63 (32.6) | 25 (58.1) | 0.002 | 37 (29.1) | 51 (46.8) | 0.005 |
| SARC-F + EBM | 61 (25.9) | 42 (21.8) | 19 (44.2) | 0.002 | 22 (17.3) | 39 (35.8) | 0.001 |
| SARC-CalF + MUAC | 144 (61.0) | 107 (55.4) | 37 (86.05) | < 0.001 | 64 (50.4) | 80 (73.4) | < 0.001 |
| BMI−adjusted SARC-CalF + MUAC | 187 (79.2) | 145 (75.1) | 42 (97.7) | 0.001 | 91 (71.7) | 96 (88.1) | 0.002 |
Pearson’s or Likelihood ratio χ2 tests. Bold is for significant P-values
Abbreviations: BMI body mass index, CC calf circumference, LOS length of hospital stay, MUAC mid-upper arm circumference
* Not ideal P-values estimations due to zero frequency in cells
SARC-F and SARC-CalF identified suggestive signs of sarcopenia in 38.1% and 42.4% of the patients, respectively. After BMI adjustment, the SARC-CalF identified 62.3% of patients. The SARC-F + EBM approach showed the lowest identification rate (25.9%), whereas the BMI−adjusted SARC-CalF + MUAC approach identified the highest proportion (79.2%). These findings are illustrated in Fig. 2.
Fig. 2.
Prevalence of suggestive signs of sarcopenia across SARC-F methods (N = 236). Abbreviations: adj: adjusted; BMI: body mass index; CalF: SARC-F + calf circumference; EBM: SARC-F + “elderly” and body mass; MUAC: SARC-F + mid-upper arm circumference
Figure 3 and Table 2 complement each other by presenting survival curves and survival probabilities (%) across all SARC-based methods. Patients with suggestive signs of sarcopenia consistently demonstrated lower survival during follow-up. Among the approaches, patients identified by the SARC-F + EBM method had the lowest survival probability (68.5%, P = 0.002), although survival probabilities were relatively similar across all methods. This finding was supported by the AUC values for mortality prediction: all approaches demonstrated poor to fair discriminatory power, with no significant global or individual differences between methods (Chi2 = 5.04, P = 0.5392; see Fig. 4). Similar trends were observed for prolonged LOS, with all SARC methods showing poor to fair AUC values and no significant global differences (Chi2 = 5.05, P = 0.5378).
Fig. 3.
Kaplan-Meier curves. Log-survival probabilities across SARC-F methods (N = 236). Blue lines: indicate no sign of sarcopenia; red lines indicate suggestive signs of sarcopenia. Abbreviations: adj: adjusted; BMI: body mass index; CalF: SARC-F + calf circumference; EBM: SARC-F + “elderly” and body mass; MUAC: SARC-F + mid-upper arm circumference
Table 2.
1-year survival probabilitiesa comparing groups with and without suggestive signs of sarcopenia based on different SARC-F approaches (N = 236)
| Method | No sign (%) | Suggestive signs of sarcopenia (%) | P |
|---|---|---|---|
| SARC-F | 84.9 | 74.5 | 0.041 |
| SARC-CalF | 87.0 | 72.6 | 0.006 |
| BMI−adjusted SARC-CalF | 88.4 | 76.5 | 0.032 |
| SARC-F + MUAC | 87.3 | 70.2 | 0.002 |
| SARC-F + EBM | 85.5 | 68.5 | 0.002 |
| SARC-CalF + MUAC | 93.3 | 72.9 | 0.002 |
| BMI−adjusted SARC-CalF + MUAC | 97.9 | 76.5 | 0.002 |
Abbreviations: BMI body mass index, CalF SARC-F + calf circumference, EBM SARC-F + “elderly” and body mass, MUAC SARC-F + mid-upper arm circumference
a Survival probabilities (%) from Kaplan-Meier estimates; Bold is for significant P-values
Fig. 4.
Area under the ROC curve for SARC-F approaches concerning (A) 1-year mortality and (B) prolonged LOS (N = 236). Abbreviations: adj: adjusted; BMI: body mass index; CalF: SARC-F + calf circumference; EBM: SARC-F + “elderly” and body mass; LOS: length of stay; MUAC: SARC-F + mid-upper arm circumference. Note ROC curves: mortality chi2(6) = 5.04, P = 0.5392; LOS: chi2(6) = 5.05, P = 0.5378 → no global differences among approaches
Except for SARC-F and SARC-F + EBM, suggestive signs of sarcopenia across approaches were associated with higher hazards for 1-year mortality; however, the BMI−adjusted SARC-CalF + MUAC approach showed the highest predictive accuracy (C-index = 0.75). All methods were associated with higher odds of prolonged LOS. For this outcome, the model including SARC-CalF demonstrated the highest AUC (AUC = 0.64). Table 3 presents detailed results.
Table 3.
Associations of suggestive signs of sarcopenia by different SARC-F methods with outcomes: 1-year mortality and prolonged LOS (N = 236)
| 1-year mortality | Prolonged LOS | |||||
|---|---|---|---|---|---|---|
| Suggestive signs of sarcopenia by | HR adjusted (95% CI)a | P | C-index | OR adjusted (95% CI)b | P | AUC |
| SARC-F | 1.90 (0.99 to 3.64) | 0.052 | - | 2.30 (1.31 to 4.05) | 0.004 | 0.62 |
| SARC-CalF | 2.07 (1.09 to 3.92) | 0.025 | 0.72 | 2.68 (1.53 to 4.66) | 0.001 | 0.64 |
| BMI−adjusted SARC-CalF | 2.08 (1.00 to 4.33) | 0.050 | 0.72 | 2.24 (1.27 to 3.94) | 0.005 | 0.62 |
| SARC-F + MUAC | 2.19 (1.16 to 4.13) | 0.015 | 0.73 | 2.05 (1.16 to 3.63) | 0.013 | 0.61 |
| SARC-F + EBM | 2.20 (0.98 to 4.88) | 0.053 | - | 3.31 (1.53 to 7.14) | 0.002 | 0.62 |
| SARC-CalF + MUAC | 3.86 (1.59 to 9.35) | 0.003 | 0.74 | 2.70 (1.52 to 4.82) | 0.001 | 0.63 |
| BMI−adjusted SARC-CalF + MUAC | 11.34 (1.54 to 83.19) | 0.017 | 0.75 | 2.83 (1.39 to 5.78) | 0.004 | 0.62 |
Abbreviations: BMI body mass index, CalF SARC-F + calf circumference, EBM SARC-F + “elderly” and body mass, LOS length of hospital stay, MUAC SARC-F + mid-upper arm circumference
aCox regression analysis: each row represents one statistical model sex-stratified, adjusted for age, clinical diagnosis, and collected comorbidities (i.e., systemic arterial hypertension and type 2 diabetes mellitus)
bLogistic binary regression: each row represents one statistical model sex-stratified, adjusted for age, clinical diagnosis, and collected comorbidities (i.e., systemic arterial hypertension and type 2 diabetes mellitus); Bold is for significant P-values
Discussion
To our knowledge, this is the first study to compare the prognostic value of SARC-F and its six modified versions incorporating age and/or anthropometric measure(s) in older inpatients. Our main findings demonstrated that suggestive signs of sarcopenia identified by almost all SARC-F versions were associated with a higher risk of 1-year mortality, except for SARC-F and SARC-F + EBM, which demonstrated borderline significance. Moreover, all SARC-F versions were associated with higher odds of prolonged LOS, with SARC-CalF model demonstrating the highest predictive performance for this outcome, based on adjusted AUC values, despite with fair values.
The finding that BMI−adjusted SARC-CalF + MUAC approach demonstrated the most robust mortality risk prediction, based on C-index results, provides important insights into the potential of combining upper- and lower-limb anthropometric proxies for muscle mass to enhance sarcopenia screening within the SARC-F framework. This approach was also the top one method in identifying a greater proportion of patients with suggestive signs of sarcopenia (79.2%).
Based on our findings, we advocate for the importance of applying BMI-adjustment factors to CC among patients with excess weight in the context of SARC-CalF, given its relevant prognostic value and potential to more accurately identify patients with suggestive signs of sarcopenia [19, 20]. This was reinforced by our C-index results, in which this approach had the highest value, as earlier discussed. However, we expected that the unadjusted approach (i.e., SARC-CalF + MUAC) could potentially demonstrate superior prognostic value/prediction for mortality. This was based on the rationale that even patients with excess weight can present with low CC without the need for BMI adjustment, potentially reflecting a more severe catabolic state and, therefore, representing a worst-case scenario, as we have previously discussed [26, 27].
The original SARC-F did not reach the HR threshold (> 2) typically considered valid for mortality prognosis, based on malnutrition expert consensus [25], although this threshold remains an ongoing debate. On the other hand, SARC-F demonstrated the lowest risk ratio among all approaches, and nonsignificant results. This reinforces the ongoing discussion regarding the clinical value of SARC-F as a suggestive marker of sarcopenia [28]. Based on our critical review of the five questions comprising SARC-F, it may be more effectively interpreted as a tool to identify impaired muscle function – one key component of the current operational definition of sarcopenia [29, 30]. It further supports the need to combine SARC-F with anthropometric proxies of muscle mass.
For prolonged LOS, the SARC-CalF model exhibited the highest adjusted AUC values (from logistic regression), indicating higher predictions, although with modest values (i.e., fair discrimination). This finding differs from a previous study that found no prognostic value of SARC-CalF in predicting clinical outcomes, including prolonged LOS [31]. This difference may be partially explained by methodological differences, as the previous study defined prolonged LOS as > 10 days, whereas we used > 8 days. Additionally, the previous cohort included adults of all ages, which may have obscured the true prognostic value of SARC-CalF. Although SARC-F and SARC-CalF can theoretically be applied across all age groups, both methods were originally developed for older adults, and their components are closely related to age-associated functional decline [9, 14].
Despite SARC-CalF’s superior results for LOS, all other SARC-based approaches also exhibited significant associations with this outcome. A potential mechanism explaining these associations with prolonged LOS is that higher SARC-F-based scores reflect impaired muscle strength and function, which may predispose patients to slower recovery trajectories and functional decline during hospitalization. While our data do not allow us to directly test this pathway, it represents a plausible explanation that warrants further investigation in future studies.
Our additional results showed that SARC-F classified only 38.1% of patients with suggestive signs of sarcopenia, ranking ahead of only SARC-F + MUAC (37.2%) and SARC-F + EBM (25.9%). The approach with the highest identification rate was the BMI−adjusted SARC-CalF + MUAC, identifying 79.2% of patients with suggestive signs of sarcopenia. Despite promising, its potential superiority in identifying at-risk patients does not equate to greater diagnostic accuracy. Validation against a reference standard for sarcopenia diagnosis is required to assess its sensitivity, specificity, and overall diagnostic performance. Therefore, while the approach may offer better prognostic value, it demands confirmation in future studies designed to test its accuracy metrics. In the current context, where there is growing emphasis on proactive rather than reactive screening of nutrition-related abnormalities [32, 33], the BMI−adjusted SARC-CalF + MUAC approach may represent a promising, early tool to help address gaps in timely identification and individualized intervention, pending further validations.
Although a 79.2% frequency of patients with suggestive signs of sarcopenia may seem high, the characteristics of our cohort help explain these results. The study population consisted of older adults with acute-on-chronic conditions, most hospitalized with potentially catabolic diagnoses, at a public hospital where socioeconomic inequities are prevalent. These factors contribute to the accumulation of risk factors for nutritional abnormalities, including sarcopenia. These findings are also consistent with our previous study, which demonstrated higher malnutrition rates when appropriate cutoffs were applied to existing tools [27].
Despite being a promising approach, the BMI−adjusted SARC-CalF + MUAC method still presents relevant challenges, particularly due to the lack of previously validated MUAC cutoffs, which may have influenced our findings. The MUAC cutoffs applied in this study were derived from the previous cohort that proposed the SARC-F + MUAC and SARC-CalF + MUAC approaches, based on a Chinese population [17], which may differ substantially from our cohort in terms of epidemiology and body composition profile. Future studies are needed to better elucidate these aspects and to establish population-specific MUAC cutoffs, particularly the newly proposed cutoff values for MUAC as a marker of muscle mass [34].
Our study is not without limitations. Its single-center design, the heterogeneity of patient diagnoses (although reflecting a real-world setting), and the absence of a priori sample size calculation warrant that the results be interpreted with caution. The relatively small sample size, with some results presenting high HR + wide 95% CI is also a relevant limitation. Additionally, not all potential factors influencing sarcopenia and outcomes were assessed or controlled for, including physical activity levels, dietary intake, and frailty assessments. The use of electronic medical records for outcome ascertainment and the inability to successfully contact all patients/relatives for cross-verification may also have introduced outcome bias (underreports). Lastly, logistical and resource constraints precluded us from accurately assessing sarcopenia diagnosis and cross-validating all seven SARC-based methods within our cohort.
Limitations offer opportunities for future research to address the gaps and methodological challenges we identified. Future studies with larger sample sizes, including cohorts of older adults with specific clinical diagnoses as well as adults across all age groups, are needed to further explore, confirm, or refute our findings. In particular, future research should aim to assess the concurrent validity of all SARC-based approaches, especially the novel BMI−adjusted SARC-CalF + MUAC, against the “gold standard” criteria for sarcopenia diagnosis.
Future studies can also explore the novel MUAC cutoff values derived from NHANES dataset, using the T-score approach, as a marker of muscle mass [34], as earlier mentioned. Furthermore, these studies could also assess the prognostic impact of these adapted versions in shorter-term outcomes (e.g., 3- and 6-month mortality).
Conclusion
Our study demonstrates that almost all SARC-F-based approaches were associated with 1-year mortality. Furthermore, all approaches were predictors of prolonged LOS. BMI−adjusted SARC-CalF + MUAC potentially exhibited the most robust prognostic value for mortality prediction, while SARC-CalF was the strongest predictor of prolonged LOS. These findings highlight the potential of refining sarcopenia screening methods by combining upper- and lower-limb anthropometric proxies for muscle mass (i.e., MUAC and CC), particularly when incorporating BMI-adjustment factors, pending future validation against the “gold-standard” sarcopenia diagnosis. Given the lower prevalence observed with SARC-F, SARC-F + MUAC, and SARC-F + EBM in our findings, these approaches warrant caution and critical consideration when used as standalone screening methods in older hospitalized populations. In this context, studies are also needed to confirm their effectiveness.
Acknowledgements
None.
Clinical trial number
Not applicable.
Authors’ contributions
TGN and JPCP contributed to the conception and design of the research. TGN, JPCP, JFMC, and JZSF acquired data; JPCP conducted data analysis. TGN and JPCP wrote the manuscript. APTF, FMS, ECC, CPSPR, PCC, RPPS, and VSL critically revised the manuscript. All the authors critically reviewed, interpreted, and approved the final version of the manuscript.
Funding
TGN and JPCP were partially funded by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil (Finance Code 001). JPCP is partially funded by the ESPEN Broaden Your Horizons Grant 2025. APTF, FMS, and ECC received a productivity scholarship from the Brazilian National Council for Scientific and Technological Development (CNPq). The supporting sources have no involvement or restrictions on this publication.
Data availability
The datasets generated and/or analyzed during the current study are not publicly available due ethical restrictions, but are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was approved by the Institutional Ethics Committees (Hospital das Clínicas de Pernambuco) following the Declaration of Helsinki (1964), its subsequent revisions, and Resolution no. 466/2012 of the National Health Council. All patients provided their written informed consent to participate.
Consent for publication
Not applicable.
Competing interests
JPCP has received speaking honoraria and travel support from Fresenius Kabi, travel fees for Speaker engagement from Danone Nutricia Brazil, and speaking honoraria from Prodiet Medical Nutrition. APTF received a grant for research from Prodiet Medical Nutrition. The other authors declare no competing interests. The supporting sources have no involvement or restrictions on this publication.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Sayer AA, Cooper R, Arai H, Cawthon PM, Ntsama Essomba MJ, Fielding RA, et al. Sarcopenia. Nat Rev Dis Primers. 2024;10:68. 10.1038/s41572-024-00550-w. [DOI] [PubMed] [Google Scholar]
- 2.Cruz-Jentoft AJ, Sayer AA, Sarcopenia. Lancet. 2019;393:2636–46. 10.1016/S0140-6736(19)31138-9. [DOI] [PubMed] [Google Scholar]
- 3.Anker SD, Morley JE, Von Haehling S. Welcome to the ICD-10 code for sarcopenia. J Cachexia Sarcopenia Muscle. 2016;7:512–4. 10.1002/jcsm.12147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Beaudart C, Zaaria M, Pasleau F, Reginster J-Y, Bruyère O. Health outcomes of sarcopenia: a systematic review and meta-analysis. PLoS ONE. 2017;12:e0169548. 10.1371/journal.pone.0169548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Beaudart C, Demonceau C, Reginster J, Locquet M, Cesari M, Cruz Jentoft AJ, et al. Sarcopenia and health-related quality of life: a systematic review and meta‐analysis. J Cachexia Sarcopenia Muscle. 2023;14:1228–43. 10.1002/jcsm.13243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bachettini NP, Bielemann RM, Barbosa-Silva TG, Menezes AMB, Tomasi E, Gonzalez MC. Sarcopenia as a mortality predictor in community-dwelling older adults: a comparison of the diagnostic criteria of the European working group on sarcopenia in older people. Eur J Clin Nutr. 2020;74:573–80. 10.1038/s41430-019-0508-8. [DOI] [PubMed] [Google Scholar]
- 7.Au PC-M, Li H-L, Lee GK-Y, Li GH-Y, Chan M, Cheung BM-Y, et al. Sarcopenia and mortality in cancer: a meta-analysis. Osteoporos Sarcopenia. 2021;7:S28–33. 10.1016/j.afos.2021.03.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Prado CM, Landi F, Chew STH, Atherton PJ, Molinger J, Ruck T, et al. Advances in muscle health and nutrition: a toolkit for healthcare professionals. Clin Nutr. 2022;41:2244–63. 10.1016/j.clnu.2022.07.041. [DOI] [PubMed] [Google Scholar]
- 9.Malmstrom TK, Morley JE. SARC-F: a simple questionnaire to rapidly diagnose sarcopenia. J Am Med Dir Assoc. 2013;14:531–2. 10.1016/j.jamda.2013.05.018. [DOI] [PubMed] [Google Scholar]
- 10.Malmstrom TK, Miller DK, Simonsick EM, Ferrucci L, Morley JE. SARC-F: a symptom score to predict persons with sarcopenia at risk for poor functional outcomes: SARC-F. J Cachexia Sarcopenia Muscle. 2016;7:28–36. 10.1002/jcsm.12048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Noda T, Kamiya K, Hamazaki N, Nozaki K, Ichikawa T, Yamashita M, et al. SARC-F predicts poor motor function, quality of life, and prognosis in older patients with cardiovascular disease and cognitive impairment. Exp Gerontol. 2023a;171:112021. 10.1016/j.exger.2022.112021. [DOI] [PubMed] [Google Scholar]
- 12.Fayh APT, Guedes FFDO, Calado GCF, Queiroz SA, Anselmo MGGB, Sousa IMD. Sarc-f is a predictor of longer Los and hospital readmission in hospitalized patients after a cardiovascular event. Nutrients. 2022;14:3154. 10.3390/nu14153154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fu X, Tian Z, Thapa S, Sun H, Wen S, Xiong H, et al. Comparing SARC-F with SARC-CalF for screening sarcopenia in advanced cancer patients. Clin Nutr. 2020;39:3337–45. 10.1016/j.clnu.2020.02.020. [DOI] [PubMed] [Google Scholar]
- 14.Gonzalez B-ST, Menezes AMB, Bielemann RM, Malmstrom TK, Gonzalez MC. Enhancing sarc-f: improving sarcopenia screening in the clinical practice. J Am Med Dir Assoc. 2016a;17:1136–41. 10.1016/j.jamda.2016.08.004. [DOI] [PubMed] [Google Scholar]
- 15.Ururi-Cupi K, Oliva-Zapata F, Salazar-Talla L, Cuba-Ruiz S, Urrunaga-Pastor D, Runzer-Colmenares FM, et al. Sarc-f and sarc-calf scores as mortality risk factors in older men with cancer: a longitudinal study from Peru. J Nutr Health Aging. 2022;26:856–63. 10.1007/s12603-022-1844-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Borges S, Fortes RC, Ferreira Martins T, Lopes Alves M, Cipriano G Jr, França Bernardelli Cipriano G. Performance of SARC-F and SARC-CalF scores to predict risk of mortality in Hemodialysis patients: a cohort study. Sci Rep. 2024;14:23262. 10.1038/s41598-024-74412-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhou J, Li T, Chen X, Wang M, Jiang W, Jia H. Comparison of the diagnostic value of SARC-F and its three modified versions for screening sarcopenia in Chinese Community-Dwelling older adults. J Nutr Health Aging. 2022;26:77–83. 10.1007/s12603-021-1718-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kurita N, Wakita T, Kamitani T, Wada O, Mizuno K. SARC-F validation and SARC-F + EBM derivation in musculoskeletal disease: the SPSS-OK study. J Nutr Health Aging. 2019;23:732–8. 10.1007/s12603-019-1222-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Do Nascimento MK, Costa Pereira JPD, De Araújo JO, Gonzalez MC, Fayh APT. Exploring the role of body mass index-adjusted calf circumference within the SARC-CalF screening tool among older patients with cancer. J Nutr Health Aging. 2024;28:100251. 10.1016/j.jnha.2024.100251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Burgel CF, Oliveira De Carvalho BZ, Milesi BM, Silva FM. SARC-Calf using calf circumference adjusted for BMI predicts 6-mo readmission and mortality in hospitalized patients: a secondary analysis of a cohort study. Am J Clin Nutr. 2025;121:151–7. 10.1016/j.ajcnut.2024.10.010 [DOI] [PubMed]
- 21.Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (Strobe) statement: guidelines for reporting observational studies. Lancet. 2007;370:1453–7. 10.1016/S0140-6736(07)61602-X. [DOI] [PubMed] [Google Scholar]
- 22.De Onis M, Habicht J, World Health Organization Expert Committee. Anthropometric reference data for international use: recommendations from a. Am J Clin Nutr. 1996;64:650–8. 10.1093/ajcn/64.4.650. [DOI] [PubMed] [Google Scholar]
- 23.Gonzalez MC, Mehrnezhad A, Razaviarab N, Barbosa-Silva TG, Heymsfield SB. Calf circumference: cutoff values from the NHANES 1999–2006. Am J Clin Nutr. 2021;113:1679–87. 10.1093/ajcn/nqab029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nahm FS. Receiver operating characteristic curve: overview and practical use for clinicians. Korean J Anesthesiol. 2022;75:25–36. 10.4097/kja.21209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.De Van Der Schueren MAE, Keller H, Cederholm T, Barazzoni R, Compher C, Correia MITD, et al. Global leadership initiative on malnutrition (GLIM): guidance on validation of the operational criteria for the diagnosis of protein-energy malnutrition in adults. Clin Nutr. 2020;39:2872–80. 10.1016/j.clnu.2019.12.022. [DOI] [PubMed] [Google Scholar]
- 26.Costa Pereira JPD, Gonzalez MC, Prado CM, Cabral PC, Nascimento TGD, Nascimento MKD, et al. Body mass index–adjusted calf circumference is associated with mortality in hospitalized older patients with excess weight. Nutrition 2024;125:112505. 10.1016/j.nut.2024.112505 [DOI] [PubMed]
- 27.Costa-Pereira JP, Prado CM, Gonzalez MC, Sousa IM, Ferreira AMJ, Cabral PC, et al. New insights on the use of mini nutritional assessment: impact of alternative calf circumference cutoffs. Clin Nutr. 2025;48:60–9. 10.1016/j.clnu.2025.03.015. [DOI] [PubMed] [Google Scholar]
- 28.Bahat G, Erdoğan T, İlhan B. SARC-F and other screening tests for sarcopenia. Curr Opin Clin Nutr Metabolic Care. 2022;25:37–42. 10.1097/MCO.0000000000000801. [DOI] [PubMed] [Google Scholar]
- 29.Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Cederholm T, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48:16–31. 10.1093/ageing/afy169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kirk B, Cawthon PM, Arai H, Ávila-Funes JA, Barazzoni R, Bhasin S, et al. The conceptual definition of sarcopenia: Delphi consensus from the global leadership initiative in sarcopenia(Glis). Age Ageing. 2024;53:afae052. 10.1093/ageing/afae052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rodrigues FW, Burgel CF, Brito JE, Baumgardt E, De Araújo BE, Silva FM. SARC-CalF tool has no significant prognostic value in hospitalized patients: A prospective cohort study. Nut Clin Prac. 2021;36:1072–9. 10.1002/ncp.10675. [DOI] [PubMed] [Google Scholar]
- 32.De Van Der Schueren MAE, Jager-Wittenaar H. Malnutrition risk screening: new insights in a new era. Clin Nutr. 2022;41:2163–8. 10.1016/j.clnu.2022.08.007. [DOI] [PubMed] [Google Scholar]
- 33.Jager-Wittenaar H, Sealy M, Naumann E, De Van Der Schueren M. Nutritional risk screening: a need to guide Alice in nutritionland. Curr Opin Clin Nutr Metabolic Care. 2024;27:381–6. 10.1097/MCO.0000000000001051. [DOI] [PubMed] [Google Scholar]
- 34.Costa-Pereira JP, Prado CM, Heymsfield SB, Fayh AP, Menezes-Júnior LA, Cabral PC, et al. Arm circumference as a marker of muscle mass: cutoff values from NHANES 1999–2006. Am J Clin Nutr. 2025;S0002916525005854. 10.1016/j.ajcnut.2025.09.034. [DOI] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due ethical restrictions, but are available from the corresponding author on reasonable request.