Association between metabolic phenotypes and incident pre-sarcopenia: 3 years follow-up of Tehran Lipid and Glucose Study | BMC Public Health

The study participants were recruited from the Tehran Lipid and Glucose Study (TLGS). The TLGS is a long-term community-based research initiative aimed at identifying and preventing noncommunicable disorders. It is located in district No. 13, covering an area of approximately 13 km², and is situated in the eastern part of Tehran City. The study was conducted with the support of Shahid Beheshti University of Medical Sciences and Health Services. In this region, three medical health centers were chosen, each with data on over 90% of the families in the area. Initial measurements were recorded, and participants were followed up for three years. An initial sample of 15,005 participants aged ≥ 3 years was selected by a multistage stratified cluster sampling method for the TLGS [11]. Among 6561 subjects aged ≥ 40 years, the participants who had anthropometric and metabolic information in phase VI (2015–2018) and BIA data in phase VII (2019–2021) were selected via simple random sampling.

The participants were chosen through meticulous inclusion and exclusion criteria, taking into consideration diverse factors, including health status. We conducted a thorough assessment of the demographic information to acquire a comprehensive understanding of the participants’ overall health and well-being. Furthermore, emphasis was placed on including only individuals with good health, while excluding those with any underlying health conditions that could potentially introduce confounding factors into the results. On the other hand, participants were excluded from the study if they had missing data, conditions such as heart failure or renal failure (with a GFR less than 60), a history of cancer, were pregnant or lactating, had a BMI less than 18.5, a history of diuretic or glucocorticoid use, or had implanted heart pacemakers, platinum or metal prostheses, or Holter devices, as Bioelectrical Impedance Analysis (BIA) is not recommended for these individuals. Ultimately, the final analysis included 2257 participants (Fig. 1). Approval for this study was obtained from the Ethics Committee of the Research Institute for Endocrine Sciences (RIES) at the Shahid Beheshti University of Medical Sciences (code: IR.SBMU.ENDOCRINE.REC.1402.053). All participants provided written informed consent prior to participating in the study.

The flowchart of recruitment

Measurements of anthropometric and body composition indices

Participants who were invited to participate in the TLGS were subsequently referred to trained physicians after signing an informed consent form. During the anthropometric measurements, the participants were attired in light clothing and without shoes. Weight and height were assessed using a digital electronic weighing scale (Seca 707; range 0.1–150 kg; Seca, Hanover, MD) with a precision of up to 100 g and a tape meter stadiometer, respectively. Body mass index (BMI) was calculated by dividing weight (in kilograms) by the square of height (in meters). Waist circumference (WC) was measured in centimeters at the level of the umbilicus.

Body composition was assessed using a portable multi-frequency bioelectrical impedance analyzer (BIA) device (Model: InBody 570, InBody Co., Ltd. Seoul, KOREA). The BIA technique offers uncomplicated, secure, and reliable assessment of skeletal muscle mass, with validation for the measurement of appendicular skeletal muscle mass (ASM) in large study groups [1]. InBody 570 is widely acknowledged for its credibility and consistency across diverse populations, although its precision may be influenced by variables such as hydration status, body temperature, and traits specific to certain populations. Participants followed precise preparation guidelines, which involved a 2-hour fasting period, abstaining from caffeine for 2 h, refraining from exercise for 4–6 h, and dressing in lightweight clothing without shoes and socks. The participants were required to initiate the test following an overnight fasting period and maintain a seated position for 5 min before the measurement.

It is important to highlight that BIA is not recommended for individuals with heart pacemakers, platinum, metal prostheses, or Holter devices implanted in their bodies. Participants were directed to remove any metals or jewelry they had before undergoing the BIA measurement. This precautionary measure was implemented to minimize potential interference and enhance the reliability of the BIA results. After wiping the palm and sole, each subject was instructed to stand barefoot with their feet positioned symmetrically on the foot electrodes in an upright stance. Simultaneously, they were guided to keep their arms straight down, with their hands gripping onto the hand electrodes, as prompted by the instrument. Bioelectrical impedance analysis with eight electrodes assesses different segmental impedances (i.e., the trunk, right and left arms, and right and left legs) at 5, 50,500 kHz employing eight electrodes in a tetrapolar arrangement, and the device output included parameters composed of fat mass, fat free mass, ASM, trunk muscle mass, protein, mineral, total body water, intracellular water, extracellular water, and visceral fat area. Ultimately, the Skeletal Muscle Index (SMI) was computed by dividing the ASM by the square of the height in meters. Additionally, other data, including sex, height, weight, and age were recorded.

The intraclass correlation coefficient (ICC) was used to assess the reproducibility of the measurements obtained by the BIA device within each group [12]. The ICC is a statistical metric that measures the consistency or reproducibility of measurements, encompassing both technical reproducibility and daily biological variations. A sample comprising 15 women and 16 men was chosen, adhering to the pertinent criteria. The same operator conducted body composition analyses on each group twice, with a three-day interval between sessions. The average age of men was 24 ± 6.4 years, whereas women had a mean age of 35 ± 10.8 years. The ICC values and 95% confidence intervals (CI) were calculated using the SPSS software version 20. The ICCs and 95% CIs calculated for PBF and FFM were 0.996 (0.991–0.998) and 0.998 (0.997–0.999), respectively.

The mean differences for the two measurements of FM and FFM were (0.04 ± 1.11) and (0.10 ± 1.04), respectively. These values, being close to zero, suggest reliability.

Predictive models, considering impedance, age, sex, height, and weight, were employed to estimate body composition parameters. These models were developed through regression analysis using data from diverse populations, including those evaluated using more precise methods such as Dual-Energy X-ray Absorptiometry (DXA). The accuracy of BIA depends on the quality of the device, accuracy of the reference data, and characteristics of the population under consideration. Basal Metabolic Rate (BMR) values were approximated using BIA software, which utilizes prediction equations based on factors such as age, weight, height, and gender. It is essential to recognize that these values represent calculated estimates and should not be construed as direct measurements of the BMR.

Measurements of metabolic indices

Blood samples were collected from all study participants between 7:00 am and 9:00 am, following an overnight fast lasting 12–14 h. Fasting glucose levels were determined by glucose oxidase and enzymatic colorimetry. Serum total cholesterol (TC) and triglycerides (TGs) levels were assessed using the enzymatic colorimetric method, employing cholesterol esterase, cholesterol oxidase, and glycerol phosphate oxidase, respectively. High-density lipoprotein cholesterol (HDL-C) was quantified following the precipitation of apolipoprotein B-containing lipoproteins with phosphotungstic acid. All these biochemical tests were performed on the day of sampling using commercial kits from Pars Azmoon, Inc., Tehran, Iran, and conducted using the Selectra 2 auto-analyzer from Vital Scientific, Spankeren, The Netherlands. Analyses were performed on all the samples once quality control was established. Both inter- and intra-assay coefficients of variation (CVs) were < 2.3% for glucose, < 2% for TC, < 2.1% for TG, and < 3% for HDL-C.

All measurements were performed simultaneously at the RIES research laboratory.

After the subjects had rested for 15 min, a qualified physician measured their systolic blood pressure (SBP) and diastolic blood pressure (DBP) twice while they were in a seated position. The initial measurement was used to ascertain the peak inflation level using a mercury sphygmomanometer. In this study, the blood pressure of the participants was computed as the average of two measurements.

Physical activity assessment

Physical activity was assessed using the modifiable activity questionnaire (MAQ), a validated tool that measures the frequency and duration of physical activity across different domains, including walking, moderate-intensity, and vigorous-intensity activities over the past seven days. Participants reported the time spent in each activity, which was then used to calculate the total physical activity level in Metabolic Equivalent of Task (MET) minutes per week. Accordingly, participants were classified into three categories: Low Physical Activity (< 600 MET-minutes/week), Moderate Physical Activity (600–2,999 MET-minutes/week), and High Physical Activity (≥ 3,000 MET-minutes/week).

Definition of variables and outcomes

Definition of obesity phenotypes

Utilizing BMI ≥ 25 kg/m² as a threshold to define overweight/obesity appears to be a more reasonable approach [13]. Abnormal metabolic components were delineated according to the criteria outlined in the Joint Interim Statement (JIS) [14], which are indeed a well-established method for classifying metabolic abnormalities [15]. These criteria include: (i) serum TG ≥ 150 mg/dL or taking lipid-lowering drugs; (ii) HDL-C < 40 mg/dL in men and < 50 mg/dL in women, or taking lipid-lowering drugs; (iii) SBP ≥ 130 mmHg or DBP ≥ 85 mmHg, or taking antihypertensive drugs; and (iv) fasting blood glucose ≥ 100 mg/dL or undergoing treatment for diabetes. Participants with fewer than two JIS components were classified as metabolically healthy, whereas the metabolically unhealthy group comprised individuals meeting two or more criteria. Due to its high correlation with BMI, WC was excluded from the definition of metabolically unhealthy status [16]. Subsequently, the individuals were categorized into four groups according to their BMI and metabolic conditions: [1] metabolically healthy normal weight (MHNW) defined as BMI < 25 kg/m² and unhealthy metabolic status; [2] metabolically healthy overweight/ obese (MHO) defined as BMI ≥ 25 kg/m² and healthy metabolic status; [3] metabolically unhealthy normal weight (MUNW) defined as BMI < 25 kg/m² and unhealthy metabolic status; [4] metabolically unhealthy overweight/obese (MUO) defined as BMI ≥ 25 kg/m² and unhealthy metabolic status.

Definition of pre-sarcopenia

Appendicular skeletal muscle mass (ASM) was assessed using BIA and was determined as the total lean soft-tissue mass in the arms and legs [8]. Then, the skeletal muscle index (SMI) was determined by dividing ASM by the square of the individual’s height in meters. Pre-sarcopenia was characterized as having low muscle mass without any discernible effect on muscle strength or physical performance, as outlined by the Asian Working Group for Sarcopenia (AWGS) criteria [17] pertaining to conceptual staging. According to the AWGS, low muscle mass is characterized by SMI values below 7.0 kg/m² in men and below 5.7 kg/m² in women [1].

Statistical analysis

For data with a normal distribution, the mean and standard deviation were employed, whereas for skewed distributions, the median [25th and 75th percentiles] was utilized. Categorical variables were expressed as numerical values accompanied by their respective percentages. Differences in continuous variables were assessed using either one-way analysis of variance or Kruskal–Wallis one-way analysis of variance, depending on the nature of data distribution. Group comparisons for categorical variables were performed using the chi-square test or Fisher’s exact test, depending on the specific circumstances. The association between metabolic phenotypes and onset of pre-sarcopenia was investigated using multiple logistic regression analysis. A two-tailed P-value of less than 0.05 was deemed statistically significant. All statistical analyses were conducted using Stata version 15.1 statistical software (StataCorp LLC, Texas, USA).