Testosterone is associated with abdominal body composition … – Nature.com

Study design

We used a cross sectional design that is appropriate to confirm the relationship between testosterone and abdominal body composition based on large data from Health Promotion Centre.

We retrospectively obtained data from 1612 participants aged20years (range: 2284years) who underwent self-referral APCT and testosterone level measurement as a part of routine check-ups at the Health Promotion Centre, Ulsan University Hospital, between March 2014 and June 2019. Ulsan is an industrial city in Korea with many large companies, which provide health check-ups every 2years for their employees. The exclusion criteria were as follows: (1) presence of chronic diseases affecting muscle mass, such as stroke, tuberculosis, chronic kidney disease, chronic liver disease, and cancer, and (2) insufficient medical records. Finally, 1453 subjects were included in the analysis (Fig.1). Clinical and laboratory variables were collected using the clinical data warehouse platform in conjunction with electronic medical records at the Ulsan University Hospital. This study was approved by Institutional Review Board of Ulsan University Hospital (No. 2021-11-033); it conformed to the principles outlined in the Declaration of Helsinki. The need for informed consent was waived by Institutional Review Board of Ulsan University Hospital owing to the retrospective nature and the anonymization of the data included in the study.

Overview of the study population.

Data on clinical factors (e.g., comorbidities, such as hypertension, diabetes, dyslipidemia, and cardiovascular disease) were obtained from systemized self-reported questionnaires issued to the participants prior to their check-up, as described previously5,6. Height and weight were collected while the participants wore light clothing without shoes. Body mass index (BMI) was calculated as the weight (kg) divided by the square of the height (m2). Waist circumference (cm) was measured midway between the costal margin and the iliac crest at the end of a normal expiration. Blood pressure was checked on the right arm after a 5 or more min rest using an automatic manometer with an appropriate cuff size. After overnight fasting, morning blood samples were collected and were subsequently analyzed at the central laboratory of our hospital. Fasting blood glucose, hemoglobin A1c (HbA1c), albumin, C-reactive protein (CRP), total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and triglycerides were measured. CRP was used for analysis as a categorical variable because it was impossible to measure less than 0.042mg/dL, and the cutoff value was set at 2.0mg/dL reported in the JUPITER trial10. Serum testosterone level was measured by competitive immunoassay using direct chemiluminescent technology on ADVIA Centaur XP system (Siemens Healthcare Diagnostics, Malvern, PA, USA).

MetS was defined based on the revised National Cholesterol Education Program criteria proposed by the American Heart Association/National Heart, Lung, and Blood Institute11. MetS requires the presence of at least three of the following five components: (1) abdominal obesity (waist circumference90cm for Asian men and80cm for Asian women), (2) triglyceride level150mg/dL, (3) HDL cholesterol level<40mg/dL for males or<50mg/dL for females or those receiving drug treatment, (4) systolic/diastolic blood pressure130/85mmHg or receiving drug treatment, and (5) fasting plasma glucose concentration100mg/dL or receiving drug treatment.

All CT images were obtained using the SOMATOM Definition Flash system (Siemens Healthcare, Erlangen, Germany), as described previously5,6. Enhanced images were obtained after a 80s delay after contrast injection. The scanning parameters were as follows: beam collimation, 1280.6mm; beam pitch, 0.6; gantry rotation time, 0.5s; field of view to fit, 100 kVp. An automatic exposure control system (CARE Dose 4D, Siemens Medical Solutions, Erlangen, Germany) was used.

Body composition was evaluated with APCT using the Asan-J software, which was developed based on ImageJ (NIH, Bethesda, MD, USA), as described previously5,6,12. Two consecutive axial CT images at the inferior endplate of the L3 lumbar vertebra were captured for each patient (Appendix S1). Using the Asan-J software, we calculated the total abdominal muscle area (TAMA) (cm2), including all muscles in the field (psoas, paraspinal, transversus abdominis, rectus abdominis, quadratus lumborum, and internal/external obliques), with predetermined Hounsfield unit (HU) thresholds on CT. The TAMA was divided into a low-attenuation abdominal muscle area (LAMA) and a normal-attenuation abdominal muscle area (NAMA) based on HUs on CT (TAMA, 29150 HU; LAMA, 2929 HU; NAMA, 30150 HU)13,14 LAMA implies lipid-rich skeletal muscle, which has more fat elements between and inside the muscle fibers. However, NAMA indicates lipid-poor skeletal muscle, which includes less fat between and inside the muscles15. Furthermore, the visceral fat area (VFA) (cm2) and the subcutaneous fat area (SFA) (cm2) were evaluated using adipose tissue thresholds on CT (190 to 30 HU)16,17. We adjusted the cross-sectional areas of the abdominal fat and muscles by BMI based on the Foundation for the National Institutes of Health Sarcopenia Project recommendation18; these were named index such as the TAMA index (TAMAi) (TAMAi=TAMA [cm2]/BMI [kg/m2]), LAMA index (LAMAi) (LAMAi=LAMA [cm2]/BMI [kg/m2]), NAMA index (NAMAi) (NAMAi=NAMA [cm2]/BMI [kg/m2]), IMFA index (IMFAi) (IMFAi=IMFAi [cm2]/BMI [kg/m2]), VFA index (VFAi) (VFAi=VFA [cm2]/BMI [kg/m2]), and SFA index (SFAi) (SFAi=SFA [cm2]/BMI [kg/m2]).

Clinical characteristics were summarized as frequency (percentage) for categorical variables and as meanstandard deviation (SD) for continuous variables. The coefficient of variation was used to show the extent of variability of the continuous variables in this cohort. Before the main analysis, we examined distributions for all continuous variables and found some variables were severely skewed to the right; thus, we conducted log transformation to obtain more stable analysis results (Appendix S2). For example, the testosterone level underwent loge-transformation. To determine whether there was an association between testosterone and abdominal body compositions or metabolic factors, we conducted several linear regression model analyses. First, we fitted univariable and multivariable regression models. The effect of abdominal body compositions was then adjusted for important clinical factors such as age, albumin, and five metabolic syndrome components in the multivariable regression analysis. A p value less than 0.05 was considered statistically significant (two-tailed). All data analyses were performed using R software version 4.1.2.

Read the original:
Testosterone is associated with abdominal body composition ... - Nature.com