This study was conducted from June 2021 to July 2022. We used whole cluster random sampling to collect data on adults aged 55–90 years from multiple Chinese communities and nursing facilities across Hebei Province, China. The study protocol was approved by the Ethics Committee of the First Hospital of Hebei Medical University (Trial Registration: Chinese Clinical Trial Registry ChiCTR2200061824). All subjects provided written informed consent before data collection. To ensure the consistency of the questionnaires, all investigators received standardized training before the trial.

The inclusion criteria were as follows: (1) Aged 55–90 years. (2) Ability to communicate without barriers. (3) Capable of independently performing actions such as “standing up and sitting down” and “walking” without external assistance. The exclusion criteria were as follows: (1) Physical disabilities. (2) Presence of specific diseases (including Congestive Heart Failure, Chronic Obstructive Pulmonary Disorder, Chronic Renal Failure, cancer, etc.). (3) Installed with a cardiac pacemaker. (4) The presence of severe cognitive impairment or inability to communicate. (5) Activity limitations due to injury. (6) Refusal to sign or inability to understand the informed consent form.

To investigate the relationship between daily BCAAs intake and sarcopenia, we used the Recommended Nutrient Intake (RNI) to determine whether BCAAs intake of each subject was inadequate. RNIs for leucine, isoleucine, and valine are 39, 20, and 26 mg/kg, respectively according to the standards issued by Joint WHO/FAO/UNU Expert Consultation (25). We applied a 64-item food frequency questionnaire (FFQ) to assess the usual free-living dietary intakes of subjects. We primarily investigated the dietary intake of the subjects over the past 12 months, including 15 major food categories (refined grains, whole grains, fried foods, vegetables, fungi and algae, fruits, nuts, legumes and soy products, pickled foods, meat and eggs, seafood, dairy products, sugary beverages, alcohol, and drinking). We also investigated the household consumption of cooking oils and condiments to better estimate daily dietary intake.

All subjects were asked to answer the following questions regarding food consumption quantity: (1) Record the frequency of consumption over the past year, with units of: times/year, times/month, times/week, times/day; (2) Record the average amount consumed per instance. Daily intake of each food category is calculated as:

Body composition was determined by bioelectrical impedance analysis (BIA) using an InBody S10 analyzer (BioSpace Co., Ltd., South Korea), including skeletal muscle mass (SMM), appendicular skeletal muscle mass (ASM), body fat mass (BF), and phase angle (PhA), muscle-to-fat ratio (MFR), etc. We assess subjects’ muscle mass using the appendicular skeletal muscle mass index (ASM/Height², ASMI), where low muscle mass is defined as <7.0 kg/m² for males and <5.7 kg/m² for females. All measurements were performed in the morning for all participants, and they were asked not to change their usual dietary pattern and daily physical activity and to refrain from vigorous activities the day before the trial started.

The diagnosis of sarcopenia for the included participants is conducted according to the AWGS2019 guidelines (2). In that guideline, sarcopenia has been defined as the existence of both low muscle mass and low muscular strength or low physical performance.

All subjects were asked to place their feet flat on the ground and muscle strength was assessed by testing HGS. The test arm should hang down, with the forearm positioned at an angle of no more than 15° from the body. The angle of wrist flexion is 0–30° and the upper arm is pressed against the chest. HGS was measured twice for each hand (once for each right and left hand), with a 20 s interval between measurements. We used the maximum value obtained in each measurement as the subject’s HGS value. AWGS2019 recommends the following diagnostic thresholds for sarcopenia based on HGS measurements: for males, < 28.0 kg; for females, < 18.0 kg.

The Five-Times Sit-to-Stand Test (FTSST) was used to evaluate the physical function. Subjects were asked to sit on a chair with a height of 46 cm and no armrests, with both feet flat on the ground and hands crossed over the chest or placed on the shoulders. When the tester gives the “start” command, the subjects must stand up and sit down as quickly as possible without using their arms for support, completing five repetitions. The test ends after the fifth repetition when the participant’s body makes contact with the chair, and the overall time taken for the test is recorded. A total test time of >12 s is considered the cutoff value indicating a decline in physical function.

GS was measured over a 6 m straight walking course using a stopwatch. All subjects were asked to walk a distance of 6 m at a normal GS. The measurement was repeated twice, and the physical function was evaluated using the faster value as the evaluation index. A GS <1 m/s was determined to indicate low physical function.

Blood samples were drawn after overnight fasting. Clinical and blood biochemical assessments was measured by Blood Cell Analyzer DxH800 Serie and Coulter Chemistry analyzer AU5800 Serie (UniCel, United States). Blood concentration for amino acid was measured using high-performance liquid chromatography and AccQ Tag Fluor Reagent Kit (Waters, United States).

We also used the Mini Nutritional Assessment (MNA) scale to evaluate the nutritional status of the subjects. The Montreal Cognitive Assessment (MoCA) scale was employed to assess the cognitive function of the subjects, and the mobility was evaluated using the Activities of Daily Living Assessment Scale (ADL) as well. Required data on other variables such as demographic characteristics (including age, gender, education, and occupation), past medical history and medication use, alcohol intake, and smoking status were collected via pretested questionnaires.

Continuous variables are presented as means and standard deviations (SD) or median (inter-quartile range, IQR). The independent-sample t-test or Wilcoxon rank-sum test was used for comparisons. We used linear regression to analyze factors that may affect ASM in aged people.

To examine the association between BCAA intake and sarcopenia, we used binary logistic regression in three different models:

The first tertile (T1) of BCAA intake was defined as the reference category and odds ratios (OR) and 95% Confidence intervals (CIs) in the second/third tertiles (T2/T3) were computed. A two-tailed p-value of less than 0.05 was considered statistically significant between T1 and T3. Binary logistic regression models were used and adjusted for age, BMI, gender and food sources when used to explore the association between food-based BCAAs and sarcopenia in older adults.

A total of 367 subjects participated in the study, including 112 males and 255 females aged over 55. The baseline characteristics of study subjects were provided in Table 1. The overall prevalence of sarcopenia was 20.7% (76 in 367), with no significant difference (p = 0.539) between males (18.8%, 21 in 112) and females (21.6%, 55 in 255). There were no significant differences in age, BMI or HC between males and females, but body fat mass was higher in females and muscle strength and quality indicators was higher in males (p < 0.001).

Table 2 showed that the sarcopenia group had higher averages in age, BMI, HGS, and SMM than the non-sarcopenia group. The mean daily energy intake, protein, fat, and BCAAs (including leucine, isoleucine, and valine) were significantly lower in the sarcopenia group than in the non-sarcopenia. The daily protein and fat intake in the sarcopenia group was higher than that in the non-sarcopenia group (p < 0.001), and there was no statistically significant difference in carbohydrate intake (p = 0.600). The non-sarcopenia group had higher per capita income (p < 0.001), MFR (p = 0.023) and better cognitive functioning (p < 0.001). However, we did not find significant differences in the serum indicators.

Pearson correlation analysis was performed to study the association between dietary intake, body composition, and serum BCAAs in sarcopenia adults in Figure 1. Dietary carbohydrate (r = 0.297, p = 0.011) and protein (r = 0.384, p < 0.001) intakes were related to ASM in the elderly, and the intake of dietary BCAAs showed a positive trend of ASM (r = 0.358, p = 0.002). However, serum BCAAs and dietary fat intake were more correlated (r = 0.370, p = 0.003), higher than dietary protein and carbohydrates intake (r = 0.201, p = 0.114; r = −0.047, p = 0.715).

The linear regression in Table 3 showed that, variables such as age, BMI, income per capita, and smoking may have influences on ASM in older adults, The prevalence of sarcopenia is associated with a lower intake of total energy, dietary fiber, protein, and BCAAs in the diet. We also observed an association between the prevalence of sarcopenia and low carbohydrate intake by adjusting for energy intake. Dietary factors were observed to have a large effect on muscle mass, strength, and mobility in older adults. After adjuested for age, gender, BMI and energy intake, dietary carbohydrate intake negatively correlated with muscle mass in older adults.

The compliance rate for daily BCAAs intake (including leucine, isoleucine, and valine) among all subjects was 19.02%, 26.43%, and 17.44% (70 in 367; 97 in 367; 64 in 367), respectively. Binary logistic regression was used in three different models. The first tertile was defined as the reference category and ORs and 95% CIs in the second and third tertiles were computed. Based on the tertiles of the intake levels of the BCAAs, the subjects were divided into the lowest intake group (T1, n = 122), the moderate intake group (T2, n = 122), and the highest intake group (T3, n = 123).

Crude and multivariable-adjusted ORs and 95% CIs for sarcopenia and its components by tertiles of dietary leucine, isoleucine and valine are shown in Table 4. BCAAs intake in T3 were less likely to be sarcopenia than those in T1 in crude model (OR: 0.250, 95% CI: 0.148–0.425; OR: 0.271, 95% CI: 0.161–0.454; OR: 0.307, 95% CI: 0.184–0.511, p < 0.001). After adjustment for age, sex and daily energy intake, the association strengthened (OR: 0.115, 95% CI: 0.043–0.308, p < 0.001; OR: 0.153; 95% CI: 0.059–0.394, p < 0.001; OR: 0.190; 95% CI: 0.073–0.495, p = 0.001, respectively). Further adjustment for other potential confounders did not alter the association (Model 2, Model 3). With the increasing intake of BCAAs, the components of sarcopenia, including low HGS and low GS, also showed significant association.

Table 5 shows the ORs for sarcopenia and dietary sources of BCAAs. In the crude model, dietary intake of animal-derived BCAAs was negatively associated with sarcopenia (OR: 0.766, 95% CI: 0.0.691–0.849, p < 0.001), whereas we did not observe an association between intake of plant-derived BCAAs and sarcopenia (OR: 0.952, 95% CI: 0.898–1.010, p = 0.101). This trend remained unchanged after further adjusted for age, gender, BMI and daily animal-based foods (OR: 0.819, 95% CI: 0.675–0.955, p = 0.044; OR: 0.903, 95% CI: 0.742–1.098, p = 0.305, respectively).