Myostatin is a myokine primarily produced in skeletal muscle where it suppresses cellular growth and differentiation thereby inhibiting muscle mass growth. Low levels of myostatin are also detected in other tissues such as adipose and cardiac [1, 2]. Myostatin has emerged as a potential mediator of sarcopenia and is negatively related to muscle function and strength [3,4,5,6]. Myostatin concentrations are elevated in sarcopenic obesity, negatively associated with insulin sensitivity indices and positively with measures of insulin resistance [7, 8]. Sarcopenia is primarily a disease of the older adult population and is an important determinant of muscle strength and physical performance. Additionally, older adult populations within the United States are highly susceptible to metabolic disorders such as obesity and type-2 diabetes . Because older adults display co-existing factors related to sarcopenia, obesity and diabetes, they are particularly vulnerable to the negative effects of myostatin .
With aging body composition changes occur that include increases in body fat with concurrent decreases in muscle mass and strength. Moreover, the increase in body fat is distributed predominately within the abdominal area increasing the risk for metabolic disorders such as insulin resistance, hyperglycemia, visceral adiposity, dyslipidaemia and hypertension [11,12,13]. Cardiometabolic disease describes the clustering of such metabolic abnormalities and adults aged 65 years and older are at increased risk for cardiometabolic disease due to experiencing multiple related metabolic disorders thereby enhancing the onset of cardiovascular disease and type 2 diabetes .
What is known about the relationship between diet and myostatin is from rodent studies that involve high-fat diets, myostatin inhibitors and knock-out mouse models [15,16,17]. The impact of diet on changes in myostatin concentrations and parallel changes in muscle and metabolic health in humans remains unexplored. In a controlled-feeding dietary intervention study in which adults aged 65 years and older consumed a calorie-restricted Dietary Approaches to Stop Hypertension (DASH) diet for 12-weeks, we observed improved changes in body composition and biomarkers of cardiometabolic health characterized by decreases in waist circumference and fat mass, muscle strength maintenance with an increase in strength-to-weight ratio, reduced cholesterol and improved insulin sensitivity [18, 19]. Extending the scope of these findings and given the role that myostatin plays in muscle and metabolic health, we sought the following: (i) to evaluate the changes in circulating myostatin levels in response to a calorie-restricted DASH diet in older adults; (ii) to assess associations between myostatin, body composition and cardiometabolic biomarkers in this cohort of older adults; and (iii) considering the role that follistatin plays as an antagonist of myostatin, we assessed the changes in follistatin concentrations in response to the diet intervention. Considering that the relationship between diet and myostatin in humans remains relatively unknown, this study aims to contribute to this unexplored gap in knowledge.
Participants and methods
Subject characteristics, recruitment, and study diet have been previously reported [18, 19]. Briefly, older sedentary adults (> 65-years) were recruited from Brookings, South Dakota between June 2017 to August 2018. A questionnaire that included date of birth, medication use, vitamin/mineral use, and drug & alcohol use was completed by participants prior to the start of the study. Inclusion on this study was based upon: 1) age; 2) mobile ability; 3) consumption of one meal per day at the study location; 4) no comsumption of foods or beverages outside of those provided by research personnel; and 5) provide fasted blood samples at 5 timepoints throughout the intervention period. Individuals with physical and/or mobility impairments, under the age of 65 years, or could not maintain the dietary regimen/protocol were unable to participate in the study. A full characterization of body composition and cardiometabolic outcomes have been previously published [18, 19]. The study was conducted in accordance with the Declaration of Helsinki. The protocol was reviewed and approved by the Institutional Review Board for Human Study Participant Use at South Dakota State University (Approval #: IRB-1712006-EXP) and informed consent was obtained from all participants before entry into the study.
Study design and sample collection
As previously described, this was a parallel designed controlled-feeding diet intervention study [18, 19]. Upon entry into the study, females (n = 17) and males (n = 11) were assigned to consume either 85 g (3 oz.; n = 15) or 170 g (6 oz.; n = 13) of lean fresh beef per day within a standardized DASH-like diet . Beef intake assignment for each participant was determined by random number generator (random.org) and assigned by a study investigator. The daily caloric intakes were determined using the 2015–2020 Dietary Guidelines for Americans for caloric intake in sedentaty older adults . The composition of the study diet has been previously reported [18, 19] and was created using Nutritionist Pro software (Axxya Systems, Redmond, WA, US).
As previously described, five fasting blood samples were collected throughout the 12 week intervention period . Blood was collected into EDTA-coated tubes (Pulmolab) and serum separator clot activator tubes (SST Vacutainer; Pulmolab). The SST tubes were kept at room temperature, allowed to clot, and centrifuged at 650×g for 15 min at room temperature. The EDTA-coated tubes were put on ice directly after blood collection and centrifuged within 90 min at 1055×g for 15 min at 4 °C. All samples were aliquoted into 1.8-mL cryostat vials (CryoTube; NUNC) and stored at − 80 °C.
Myostatin and follistatin analysis
Quantification of myostatin and follistatin were performed by the Human Nutritional Chemistry Service Laboratory at Cornell University (Ithaca, NY) . Myostatin was measured using the human quantikine myostatin immunoassay solid phase enzyme-linked immunosorbent assay (ELISA; R&D systems, Minneapolis, Minnesota, USA). The human follistatin quantikine ELISA kit (R&D systems, Minneapolis, Minnesota, USA) was used to measure follistatin concentrations. Myostatin intra- and interassay CV was 3.02 and 3.32%, respectively. The intra- and interassay CV for follistatin was 2.98 and 3.43%, respectively.
Body composition and cardiometabolic measurements
Body composition and cardiometabolic measurements were previously detailed and reported [18, 19]. Briefly, abdominal waist circumference was measured using a Gulick tape. Bioelectrical impedance (InBody 270, InBody USA, Cerritos, California) was used to measure absolute fat mass and skeletal muscle mass. The maximum grip force of the right and left hand using a hand-held dynamometer (Smedley III analog) was used to quantify handgrip strength.
Quantification of total cholesterol, low density lipoprotein cholesterol (LDL-C) and insulin were performed by the Human Nutritional Chemistry Service Laboratory at Cornell University (Ithaca, NY) and has been previously described . Briefly, the Dimension Xpand plus integrated chemistry automated analyzer (Siemens Healthineers, Malvern, Pennsylvania) was used to measure total cholesterol and LDL-C. The Immulite 2000 automated immunoassay system (Siemens Healthineers, Malvern, Pennsylvania) was used to measure insulin. Concentrations of the above biomarkers in response to the intervention have been reported .
The formula used to calculate the homeostatic model assessment of insulin resistance (HOMA-IR) is: fasting plasma glucose (mmol/l) times fasting serum insulin (μIU/mL) divided by 22.5 .
The statistical analysis plan for the present study is an extension of the original studies [18, 19] (ClinicalTrials.gov; Identifier: NCT04127240). An Independent Samples T test was used to assess differences in baseline characteristics, and myostatin, follistatin, and the myostatin:follistatin ratio between males and females. Differences between beef intake groups at week 12 was determined by Independent Samples T test. A linear mixed models analysis was applied to determine changes in the primary outcome variables across the intervention. A random intercept for each participant and Time (week 0, 3, 6, 9, and 12) as the fixed effect was used. The primary outcome of interest was the difference between baseline and week 12 for plasma concentrations of myostatin, follistatin, and the myostatin:follistatin ratio. When indicated by a significant Time effect, the Bonferroni adjustment for multiple comparisons was used to determine pairwise differences at specific time points. To adjust for the influence of changes in body weight across the intervention on the primary outcome variables, we repeated the linear mixed model analyses by including body weight as a covariate. A sensitivity analyses was also completed with the exclusion of the three normal weight subjects. Data for males and females are pooled but also displayed separately by sex. Relations between the change from baseline in myostatin concentrations and cardiometabolic variables, and body composition variables were determined by Pearson’s correlation coefficient. To identify the independent determinants of the change from baseline in myostatin levels, we performed stepwise multiple regression analysis. In each multiple regression model, variables with a related probability of greater than 0.10 were removed. Statistical significance was set at p < 0.05. Data are presented as means (SD) and analyzed with SPSS version 24 (IBM Inc., Armonk, NY, USA).
Sample size calculations for this study are based on the original study  which was to investigate muscle strength with varying amounts of meat. Additional studies, such as the present analysis are extensions of the original study (ClinicalTrials.gov; Identifier: NCT04127240). Thus, sample size was estimated based on an expected 22% improvement in the sit-to-stand test following 6-weeks of a high protein diet in 16 elderly adults. Using these estimates, sample sizes were calculated based on 80% power at an alpha level of 0.05 to detect an absolute mean improvement of 4 ± 3 repetitions in the 30-s sit-to-stand test. Resulting sample sizes were 14 participants per group, respectively.