Iliopsoas muscle to visceral fat ratio on CT predicts Cushing’s syndrome in elderly females with adrenal tumors

Abstract

There is no computed tomography (CT)-based numerical index for predicting Cushing’s syndrome (CS) in patients with adrenal incidentalomas. We tested the hypothesis that the iliopsoas muscle (Ip-M) to visceral fat (V-fat) ratio (IVR) on CT may predict CS in elderly female patients with adrenal tumors. We examined the V-fat area, subcutaneous fat (S-fat) area, Ip-M area, V-fat/S-fat ratio, and IVR at the third lumbar vertebra (L3) level using abdominal CT in female patients aged ≥50 years with cortisol-producing adrenal tumor diagnosed with CS or non-functioning adrenal tumor (NFT) in the derivation cohort. We performed receiver operating characteristic (ROC) analysis to evaluate the diagnostic value of the V-fat/S-fat ratio and IVR for predicting CS. We assessed the usefulness of the IVR in a separate validation cohort. In the derivation cohort, the IVR was significantly lower in the 9 patients with CS than in the 15 patients with NFT (p < 0.001). In ROC analysis with a cut-off value of 0.067, the IVR showed a sensitivity of 100%, specificity of 80.0%, positive likelihood ratio (PLR) of 5.000, and negative likelihood ratio (NLR) of 0.000. The area under the curve was significantly higher for the IVR than for the V-fat/S-fat ratio (0.933 vs. 0.704, respectively, p = 0.036). In 23 patients in the validation cohort, the IVR demonstrated a PLR of 5.714 and an NLR of 0.327. The novel IVR index, based on single-slice CT at the L3 level, predicted CS in elderly female patients with adrenal tumors.

ADRENAL INCIDENTALOMAS are discovered in 1–5% of abdominal computed tomography (CT) scan series [1, 2], and include functional adrenal tumors such as Cushing’s syndrome (CS), pheochromocytoma, and primary aldosteronism, with prevalences of 10.5%, 8.5%, and 5.1%, respectively [3]. Chronic hypercortisolism in patients with CS manifests as proximal muscle weakness and visceral fat obesity [4].

CT scans in CS show increased visceral fat (V-fat) and decreased abdominal muscle mass [5]. The ratio of the area of V-fat to that of subcutaneous fat (S-fat) (V-fat/S-fat ratio) is increased in CS compared to non-cushingoid control patients [5, 6]. The V-fat/S-fat ratio was shown to differ between overt CS and adrenal incidentaloma when using a serum cortisol cut-off value of 1.8 μg/dL after a dexamethasone suppression test (DST), but did not differ when using cut-off values of 3.0 μg/dL or 5.0 μg/dL [7]. This suggested that the V-fat/S-fat ratio had poor predictive value for differentiating overt CS from adrenal incidentaloma.

Recently, the area on CT of the iliopsoas muscle (Ip-M) at the third lumbar vertebra (L3) level was defined as the psoas muscle index and used to diagnose sarcopenia, a condition characterized by a progressive and generalized loss of skeletal muscle mass and strength [8]. We used the Ip-M and V-fat areas, both of which are affected by chronic hypercortisolism, to establish a new index to identify CS. We hypothesized that the Ip-M/V-fat ratio (IVR) could differentiate CS from adrenal incidentaloma.

In this study, we evaluated the diagnostic value of IVR to predict CS in a group of individuals with CS caused by cortisol-producing adrenal tumor or non-functioning adrenal tumor (NFT). The subjects were limited to elderly females to reduce the anti-anabolic effects of sex hormones and growth hormone on both adipose tissue and skeletal muscles [9-13]. We also evaluated chronological changes in the IVR among patients with CS who underwent adrenalectomy or medical therapy with steroidogenesis inhibitors.

Materials and Methods

Study design

This study was approved by the ethics board of the University of Miyazaki (#O-0735, #O-0991). The selection period for both the derivation and validation cohorts was from January 2005 to December 2019. Patients’ medical records and other information were anonymized and de-identified before analysis. The inclusion criteria were (i) female sex; (ii) age ≥50 years; (iii) diagnosis of unilateral or bilateral adrenal tumor based on CT; and (iv) hospitalization for 24-hour urinary free cortisol (UFC) measurement. The exclusion criteria were (i) diagnosis of pheochromocytoma; (ii) primary aldosteronism; (iii) ACTH-producing pituitary tumor; (iv) use of steroid-containing medications; (v) ectopic adrenocorticotropic hormone syndrome; (vi) subclinical CS; (vii) active malignant neoplasm; (viii) chronic kidney disease; (ix) metastatic adrenal tumor; (x) adrenal neuroendocrine tumor; (xi) adrenal carcinoma; (xii) adrenal lymphoma; (xiii) myelolipoma; (xiv) chronic heart failure; (xv) chronic obstructive pulmonary disease; (xvi) liver cirrhosis; (xvii) collagen disease; (xviii) inflammatory bowel disease; (xix) neuromuscular disease; (xx) ascites; (xxi) scoliosis; (xxii) alcohol abuse; and (xxiii) being bedridden (Fig. 1). The diagnosis of subclinical CS was based on the Japan Endocrine Society guidelines [14].

Fig. 1

Enrollment flowchart.

Abbreviations: CS, Cushing’s syndrome; NFT, non-functioning adrenal tumor.

Patients at the University of Miyazaki who were diagnosed with CS (n = 9) or NFT (n = 15) were enrolled in the derivation cohort. All enrolled patients with CS had clinical features of chronic hypercortisolism and undetectable plasma ACTH concentrations. In the derivation study, a receiver operating characteristic (ROC) analysis was performed to detect the cut-off values of the IVR and V-fat/S-fat ratio, and the diagnostic values of the two parameters were compared using area under the curve (AUC) analysis. The external validation study was conducted in elderly female patients aged ≥50 years who were seen at Kurume University Hospital and diagnosed with CS (n = 7) or NFT (n = 16), using the same inclusion and exclusion criteria as in the derivation study. The diagnostic values of the IVR were evaluated using the cut-off value determined in the derivation study (Fig. 1).

To assess chronological changes in the IVR among CS patients after treatment, six CS patients (four in the derivation cohort and two in the validation cohort) with complete follow-up CT scan data were analyzed. Five of the six patients underwent adrenalectomy and one was treated with the steroidogenesis inhibitor metyrapone.

Definitions of CS and NFT

The definitions of CS and NFT were based on the diagnostic criteria for CS [15]. The circadian rhythm of plasma cortisol levels, 1-mg DST results, and 24-hour UFC levels were examined. Plasma and urinary cortisol concentrations were determined by commercial laboratories (Supplemental Table 1). CS was defined by clinical features of chronic hypercortisolism (central obesity, moon face, buffalo hump, thinning skin with easy bruising, or striae) and the presence of two or more of the following: (i) plasma cortisol concentration ≥5.0 μg/dL at 2300 h; (ii) plasma cortisol concentration ≥5.0 μg/dL after 1-mg DST; and (iii) 24-hour UFC ≥80.3 μg/day. NFT was defined as the absence of clinical features of chronic hypercortisolism and the presence of two or more of the following: (i) plasma cortisol concentration <5.0 μg/dL at 2300 h; (ii) plasma cortisol concentration <1.8 μg/dL after 1-mg DST; and (iii) 24-hour UFC <80.3 μg/day.

Radiological studies

CT was performed with a 64-slice scanner (SOMATOM Definition AS; Siemens Healthineers, Erlangen, Germany) or 128-slice scanner (SOMATOM Definition AS+; Siemens Healthineers). Cross-sectional CT images at the L3 level were used to measure the areas of V-fat, S-fat, and Ip-M (cm²) (Fig. 2). Ip-M consisted of both psoas major muscles. The V-fat/S-fat ratio and IVR were calculated by division of the relevant areas. The V-fat and S-fat areas were obtained from cross-sectional CT scans at the L3 level using an image processing workstation (Ziostation; Ziosoft Corp, Tokyo, Japan). A single trained observer manually corrected the tissue boundaries if necessary. The Ip-M area was evaluated by tracing the muscles manually as previously reported [16, 17]. An independent radiological technologist measured each area in CT scans. CT scan data were assessed using the same methods at both Kurume University School of Medicine and University of Miyazaki.

Fig. 2

Areas of fat and muscle on CT scan at the L3 level in a 68-year-old female patient with an NFT.

Solid lines represent V-fat (brown) boundaries. Dashed lines represent S-fat boundaries. Dotted lines represent Ip-M (blue) boundaries.

Abbreviations: CT, computed tomography; Ip-M, iliopsoas muscle; NFT, non-functioning adrenal tumor; S-fat, subcutaneous fat; V-fat, visceral fat.

Statistical analysis

Numerical results are expressed as medians (interquartile range). Patients with missing data were omitted from the analysis. Continuous variables were non-parametrically analyzed using the Mann-Whitney U test. In the derivation study, ROC analysis was performed to detect the cut-off values based on the Youden index. The AUC, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (PLR), and negative likelihood ratio (NLR) to diagnose CS were calculated. DeLong’s test was used to compare the AUC of the IVR with that of the V-fat/S-fat ratio. In the validation study, sensitivity, specificity, PPV, NPV, PLR, and NLR to diagnose CS were calculated using the cut-off value of IVR determined in the derivation study. All analyses were performed using R version 3.6.3. p < 0.05 was considered statistically significant.

Results

The baseline characteristics of CS and NFT patients who were enrolled in the derivation cohort are shown in Table 1. One of 9 CS patients in the derivation cohort was diagnosed with bilateral macronodular adrenal hyperplasia (BMAH). As measures of adrenocortical function, 24-hour UFC and plasma cortisol concentrations both at 2300 h and following the 1-mg DST were significantly higher in CS patients than in NFT patients. Additionally, compared to NFT patients, CS patients had a significantly higher V-fat area and V-fat/S-fat ratio, and a significantly lower Ip-M area and IVR (Table 1 and Fig. 3).

Table 1 Characteristics of CS and NFT patients in the derivation cohort

Variable	CS patients (n = 9)	NFT patients (n = 15)	p value
Clinical characteristics
Age (years)	60.0 (56.0–66.0)	67.0 (60.5–77.0)	0.068
BMI (kg/m2)	26.7 (24.0–28.9)	23.6 (22.0–25.7)	0.043
24-hour UFC (μg/day)	139.5 (49.0–234.0)	26.6 (18.2–31.7)	0.003
Plasma cortisol (2300 h, μg/dL)	14.4 (13.0–16.0)	3.7 (3.0–5.2)	<0.001
Plasma cortisol (1-mg DST, μg/dL)	16.1 (12.6–20.6)	1.0 (0.9–1.0)	<0.001
Structure areas
V-fat (cm2)	179.1 (167.3–292.4)	110.2 (87.0–131.9)	0.008
S-fat (cm2)	151.7 (130.0–187.3)	120.9 (99.3–150.1)	0.074
Ip-M (cm2)	6.7 (5.5–7.0)	10.4 (8.2–11.0)	<0.001
Area ratios
V-fat/S-fat ratio	1.15 (0.85–1.93)	0.87 (0.69–1.13)	0.043
Ip-M/V-fat ratio (IVR)	0.04 (0.02–0.04)	0.10 (0.07–0.11)	<0.001

Data are presented as medians (interquartile range). p < 0.05 in bold was considered significant.

Abbreviations: BMI, body mass index; CS, Cushing’s syndrome; DST, dexamethasone suppression test; Ip-M, iliopsoas muscle; NFT, non-functioning adrenal tumor; S-fat, subcutaneous fat; UFC, urinary free cortisol; V-fat, visceral fat.

Fig. 3

Comparison of V-fat, S-fat, and Ip-M areas between CS and NFT patients.

Plots of V-fat (A), S-fat (B), and Ip-M (C) areas in CS (n = 9) and NFT (n = 15) patients. Horizontal lines represent medians. P values were calculated using the Mann-Whitney U test for CS versus NFT patients.

Abbreviations: CS, Cushing’s syndrome; Ip-M, iliopsoas muscle; NFT, non-functioning adrenal tumor; S-fat, subcutaneous fat; V-fat, visceral fat.

A ROC analysis was performed to assess the ability of the IVR and V-fat/S-fat ratio to predict CS (Fig. 4). The AUC of the IVR was 0.933 (95% confidence interval [CI], 0.838–1.000) when using a cut-off value of 0.067, with a sensitivity of 100% (95%CI, 55.5–100.0), specificity of 80.0% (95%CI, 54.4–96.0), PPV of 75.0% (95%CI, 42.8–94.5), NPV of 100% (95%CI, 64.0–100.0), PLR of 5.000 (95%CI, 1.817–13.757), and NLR of 0.000 (95%CI, 0.000–not applicable) (Table 2). The AUC of the V-fat/S-fat ratio was 0.704 (95%CI, 0.474–0.933) when using a cut-off value of 1.378, with a sensitivity of 44.4% (95%CI, 13.7–78.8) and a specificity of 93.3% (95%CI, 68.1–99.8) (Table 2 and Fig. 4). The IVR demonstrated a significantly higher AUC compared to the V-fat/S-fat ratio (p = 0.036, Fig. 4). The AUC of IVR was higher than that of Ip-M, but it did not reach statistical significance (Supplemental Fig. 1).

Fig. 4

ROC analysis for predicting CS in the derivation cohort. ROC curves for the Ip-M/V-fat ratio (solid line) and the V-fat/S-fat ratio (dashed line), with AUCs [95% confidence interval], were calculated under the non-parametric assumption. The black dots represent cut-offs closest to (1,1). P values were calculated using DeLong’s test.

Abbreviations: AUC, area under the curve; CS, Cushing’s syndrome; Ip-M, iliopsoas muscle; ROC, receiver operating characteristic; S-fat, subcutaneous fat; V-fat, visceral fat.

Table 2 Diagnostic values of the Ip-M/V-fat ratio (IVR) and V-fat/S-fat ratio in CS and NFT patients in the derivation cohort

Diagnostic value	IVR	V-fat/S-fat ratio
Cut-off value	0.067	1.378
Sensitivity (%)	100.0 [55.5–100.0]	44.4 [13.7–78.8]
Specificity (%)	80.0 [54.4–96.0]	93.3 [68.1–99.8]
Positive predictive value (%)	75.0 [42.8–94.5]	80.0 [28.4–99.5]
Negative predictive value (%)	100.0 [64.0–100.0]	73.7 [48.8–90.9]
Positive likelihood ratio	5.000 [1.817–13.757]	6.667 [0.876–50.736]
Negative likelihood ratio	0.000 [0.000–NA]	0.595 [0.327–1.084]

Data are presented as variables [95% confidence interval].

Abbreviations: CS, Cushing’s syndrome; Ip-M, iliopsoas muscle; NA, not applicable; NFT, non-functioning adrenal tumor; S-fat, subcutaneous fat; V-fat, visceral fat.

The characteristics of CS and NFT patients in the validation cohort are shown in Table 3. One of 7 CS patients in the validation cohort was diagnosed with BMAH. Ip-M and IVR were significantly lower in the CS patients than in the NFT patients. Using the IVR cut-off value of 0.067 obtained in the derivation cohort, the diagnostic parameters obtained in the validation cohort were as follows: sensitivity, 71.4% (95%CI, 29.0–96.3); specificity of 87.5% (95%CI, 61.7–98.4), PPV of 71.4% (95%CI, 29.0–96.3), NPV of 87.5% (95%CI, 61.7–98.4), PLR of 5.714 (95%CI, 1.440–22.680), and NLR of 0.327 (95%CI, 0.100–1.069) (Table 4).

Table 3 Characteristics of CS and NFT patients in the validation cohort

Variable	CS patients (n = 7)	NFT patients (n =16)	p value
Clinical characteristics
Age (years)	56.0 (54.5–62.0)	65.5 (57.3–73.0)	0.124
BMI (kg/m2)	18.7 (18.3–20.1)	21.9 (21.1–25.0)	0.030
24-hour UFC (μg/day)	148.0 (116.0–250.0)	32.0 (25.2–37.6)	<0.001
Plasma cortisol (2300 h, μg/dL)	21.2 (17.9–22.9)	2.7 (2.4–3.9)	<0.001
Plasma cortisol (1-mg DST, μg/dL)	21.2 (18.9–24.4)	1.3 (1.0–1.4)	<0.001
Structure areas
V-fat (cm2)	76.0 (63.2–108.0)	77.2 (48.5–98.5)	0.789
S-fat (cm2)	67.0 (53.6–106.0)	108.4 (89.2–160.6)	0.249
Ip-M (cm2)	5.7 (4.5–8.0)	9.9 (9.1–12.4)	0.038
Area ratio
V-fat/S-fat ratio	1.12 (1.03–1.28)	0.78 (0.50–1.02)	0.025
Ip-M/V-fat ratio (IVR)	0.06 (0.04–0.07)	0.15 (0.09–0.19)	0.023

Data are presented as medians (interquartile range). p < 0.05 in bold was considered significant.

Abbreviations: BMI, body mass index; CS, Cushing’s syndrome; DST, dexamethasone suppression test; Ip-M, iliopsoas muscle; NFT, non-functioning adrenal tumor; S-fat, subcutaneous fat; UFC, urinary free cortisol; V-fat, visceral fat.

Table 4 Diagnostic values of the Ip-M/V-fat ratio (IVR) in CS and NFT patients in the validation cohort

Diagnostic value	IVR
Sensitivity (%)	71.4 [29.0–96.3]
Specificity (%)	87.5 [61.7–98.4]
Positive predictive value (%)	71.4 [29.0–96.3]
Negative predictive value (%)	87.5 [61.7–98.4]
Positive likelihood ratio	5.714 [1.440–22.680]
Negative likelihood ratio	0.327 [0.100–1.069]

Data are presented as variables [95% confidence interval].

Abbreviations: CS, Cushing’s syndrome; Ip-M, iliopsoas muscle; NFT, non-functioning adrenal tumor; V-fat, visceral fat.

Chronological studies in six CS patients, including four in the derivation cohort and two in the validation cohort, showed increases in the IVR after treatment (Fig. 5A). A longitudinal CT scan of a 55-year-old female CS patient showed an increased Ip-M area and a decreased V-fat area after treatment, resulting in an increased IVR (Fig. 5B–5D). In the combined derivation and validation cohorts (Supplemental Table 2), IVR showed an AUC of 0.879 (95%CI 0.828–0.987) (Supplemental Table 3 and Supplemental Fig. 2).

Fig. 5

Chronological changes in the IVR among CS patients after treatment.

(A) Plot of the IVR in six CS patients before and after treatment. In the derivation cohort, a 69-year-old woman (open circle, solid line) received metyrapone therapy, and a 66-year-old woman (open square, solid line), a 55-year-old woman (open diamond, solid line), and a 51-year-old woman (open triangle, solid line) underwent adrenalectomy. In the validation cohort, a 68-year-old woman (closed circle, dashed line) and a 56-year-old woman (closed square, dashed line) underwent adrenalectomy. The y-axis shows IVR values plotted on a logarithmic scale. Cross-sectional CT at the L3 level in a 55-year-old female CS patient before adrenalectomy (B), 14 months afterward (C), and 30 months afterward (D). Arrowheads indicate the iliopsoas muscles.

Abbreviations: CS, Cushing’s syndrome; CT, computed tomography; IVR, iliopsoas muscle to visceral fat ratio.

Discussion

This study showed that a novel numerical index, the IVR, was useful for predicting CS in elderly female patients with adrenal tumors. The IVR measured with single-slice CT is a simple and convenient tool to predict the cortisol-producing capacity of adrenal incidentalomas before endocrine tests. The IVR reflected chronic hypercortisolism and was not influenced by clock time or emergency one-time steroid use. CT may help avoid false positives or negatives on endocrine tests in patients with psychiatric disorders or morbid obesity [15]. In addition, IVR is useful when false positives and negatives on DST occur due to drugs such as itraconazole and pioglitazone that affect dexamethasone metabolism by inducing CYP3A4, or if serum cortisol levels are falsely elevated by drugs such as estrogen that increase levels of cortisol-binding globulin [15].

The predictive value of the IVR for CS was superior to the V-fat/S-fat ratio in the derivation cohort (Fig. 4). Useful and simple radiological indicators of CS have been explored in adrenal incidentalomas [2]. CT examinations in CS showed an increased V-fat/S-fat ratio and decreased total abdominal muscle mass [5-7, 17]. In our study, the V-fat/S-fat ratio was increased in CS patients (Table 1), which was consistent with the results of previous studies [5-7]. However, the IVR value was more accurate, and the validation study showed consistent diagnostic performance using a cut-off value of 0.067 (Table 4).

The median age at the diagnosis of adrenal incidentaloma is 58.1 years, and therefore the application of the IVR in elderly female patients may be useful for diagnosing CS [2, 3]. Detecting CS is important because patients with chronic hypercortisolism are at increased risk of hypertension, diabetes, and cardiovascular-related mortality [15]. The IVR should contribute to effectively identifying CS, although adjusting the IVR for age-specific V-fat and Ip-M area data derived from the entire Japanese population might be required before the IVR can be applied to patients with incidentalomas in other age groups. Furthermore, analyses of the duration and amount of cortisol exposure should make it possible to evaluate the IVR in patients with mild autonomous cortisol excess (MACE) [18], cyclic CS, or steroid therapy.

The limitations of our study are as follows. First, the number of enrolled patients was small. Second, the pathological diagnosis in patients with NFT was not confirmed, which means that NFT with low to moderate cortisol hypersecretion, including MACE, could not be ruled out. In our derivation cohort, some patients had a plasma cortisol concentration ≥5.0 μg/dL at 2300 h. However, 22.5% of patients with adrenal incidentaloma and a plasma cortisol concentration <1.8 μg/dL after DST had a plasma cortisol concentration ≥5.0 μg/dL at 2300 h [19], which was consistent with the results of the present study.

In conclusion, the IVR derived from a single-slice CT scan at the L3 level may predict CS in elderly female patients with adrenal incidentalomas. Large prospective studies are required to clarify the efficacy of the IVR. Despite the small number of patients in this study, the IVR is promising as a predictor of CS activity.

Acknowledgments

The authors would like to thank T. Nakamura and M. Azuma in the Department of Radiology, Faculty of Medicine, University of Miyazaki, for their support with radiological analysis.

Disclosure

The authors have nothing to disclose.

References

• 1   Young  WF Jr (2007) Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 356: 601–610.
• 2   Sherlock  M,  Scarsbrook  A,  Abbas  A,  Fraser  S,  Limumpornpetch  P, et al. (2020) Adrenal incidentaloma. Endocr Rev 41: 775–820.
• 3   Ichijo  T,  Ueshiba  H,  Nawata  H,  Yanase  T (2020) A nationwide survey of adrenal incidentalomas in Japan: the first report of clinical and epidemiological features. Endocr J 67: 141–152.
• 4   Orth  DN (1995) Cushing’s syndrome. N Engl J Med 332: 791–803.
• 5   Rockall  AG,  Sohaib  SA,  Evans  D,  Kaltsas  G,  Isidori  AM, et al. (2003) Computed tomography assessment of fat distribution in male and female patients with Cushing’s syndrome. Eur J Endocrinol 149: 561–567.
• 6   Delivanis  DA,  Iñiguez-Ariza  NM,  Zeb  MH,  Moynagh  MR,  Takahashi  N, et al. (2018) Impact of hypercortisolism on skeletal muscle mass and adipose tissue mass in patients with adrenal adenomas. Clin Endocrinol (Oxf) 88: 209–216.
• 7   Debono  M,  Prema  A,  Hughes  TJ,  Bull  M,  Ross  RJ, et al. (2013) Visceral fat accumulation and postdexamethasone serum cortisol levels in patients with adrenal incidentaloma. J Clin Endocrinol Metab 98: 2383–2391.
• 8   Hamaguchi  Y,  Kaido  T,  Okumura  S,  Kobayashi  A,  Hammad  A, et al. (2016) Proposal for new diagnostic criteria for low skeletal muscle mass based on computed tomography imaging in Asian adults. Nutrition 32: 1200–1205.
• 9   Lizcano  F,  Guzmán  G (2014) Estrogen deficiency and the origin of obesity during menopause. Biomed Res Int 2014: 757461.
• 10   Collins  BC,  Laakkonen  EK,  Lowe  DA (2019) Aging of the musculoskeletal system: how the loss of estrogen impacts muscle strength. Bone 123: 137–144.
• 11   Franco  C,  Brandberg  J,  Lönn  L,  Andersson  B,  Bengtsson  BA, et al. (2005) Growth hormone treatment reduces abdominal visceral fat in postmenopausal women with abdominal obesity: a 12-month placebo-controlled trial. J Clin Endocrinol Metab 90: 1466–1474.
• 12   Kraemer  WJ,  Ratamess  NA,  Hymer  WC,  Nindl  BC,  Fragala  MS (2020) Growth hormone(s), testosterone, insulin-like growth factors, and cortisol: roles and integration for cellular development and growth with exercise. Front Endocrinol (Lausanne) 11: 33.
• 13   Kelly  DM,  Jones  TH (2015) Testosterone and obesity. Obes Rev 16: 581–606.
• 14   Yanase  T,  Oki  Y,  Katabami  T,  Otsuki  M,  Kageyama  K, et al. (2018) New diagnostic criteria of adrenal subclinical Cushing’s syndrome: opinion from the Japan Endocrine Society. Endocr J 65: 383–393.
• 15   Nieman  LK,  Biller  BM,  Findling  JW,  Newell-Price  J,  Savage  MO, et al. (2008) The diagnosis of Cushing’s syndrome: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab 93: 1526–1540.
• 16   Dallaway  A,  Kite  C,  Griffen  C,  Duncan  M,  Tallis  J, et al. (2020) Age-related degeneration of the lumbar paravertebral muscles: systematic review and three-level meta-regression. Exp Gerontol 133: 110856.
• 17   Miller  BS,  Ignatoski  KM,  Daignault  S,  Lindland  C,  Gauger  PG, et al. (2011) A quantitative tool to assess degree of sarcopenia objectively in patients with hypercortisolism. Surgery 150: 1178–1185.
• 18   Elhassan  YS,  Alahdab  F,  Prete  A,  Delivanis  DA,  Khanna  A, et al. (2019) Natural history of adrenal Incidentalomas with and without mild autonomous cortisol excess: a systematic review and meta-analysis. Ann Intern Med 171: 107–116.
• 19   Akehi  Y,  Kawate  H,  Murase  K,  Nagaishi  R,  Nomiyama  T, et al. (2013) Proposed diagnostic criteria for subclinical Cushing’s syndrome associated with adrenal incidentaloma. Endocr J 60: 903–912.