Association between type 2 diabetes and skeletal muscle quality assessed by abdominal computed tomography scan

1 INTRODUCTION

The prevalence of type 2 diabetes has increased significantly worldwide and has become a major global health burden.1, 2 Thus, it is important to identify the modifiable risk factors for prevention as well as early diagnosis of type 2 diabetes. The major contributing risk factors such as visceral obesity and metabolic syndrome have been well-described and a plethora of studies have focussed on the fat and adipose tissues3-7; in contrast, it remains unclear whether the amount and quality of skeletal muscles are also related to the incidence of type 2 diabetes.8-13

Skeletal muscle is the largest insulin-sensitive organ in the body; accordingly, it accounts for the majority of glucose uptake and plays a critical role in maintaining glucose homoeostasis.14 Low skeletal muscle mass is associated with reductions in insulin-mediated glucose disposal and contributes to increases in insulin resistance and dysglycemia, thereby leading to increased risk of type 2 diabetes.8-10 However, studies on the association between the incidence of type 2 diabetes and the quantity of skeletal muscles measured by computed tomography (CT) scan at the lumbar vertebrae showed inconsistent results.11-13 Paradoxically, increases in skeletal muscle mass showed positive associations with type 2 diabetes or prediabetes.11-13 It is possible that the reason for such results was due to not considering myosteatosis, which refers to ectopic fat accumulation in the cytoplasm of myocytes as well as intermuscular adipocytes.15, 16

CT is one of the most widely used imaging tools to indirectly assess myosteatosis.15, 17 A previous study showed that skeletal muscle attenuation (MA) determined by CT was associated with muscle lipid content determined in muscle biopsy specimen.18 Low radiation attenuation (Hounsfield Units [HU]) indicates a higher degree of myosteatosis, whereas high attenuation indicates low muscle fat infiltration (good quality muscle).15, 17

Previous studies have used average muscle density (HU) or intermuscular adipose tissue area (IMAT) measured by CT scan to evaluate the degree of myosteatosis and showed that myosteatosis is associated with a greater prevalence of type 2 diabetes and insulin resistance.7, 19-21 However, only few studies have investigated the association between muscle quality assessed by segmentation of the muscle area (i.e., normal attenuation muscle area [NAMA] and low attenuation muscle area [LAMA]) and type 2 diabetes.22 No studies to date have investigated the association between good quality muscle and type 2 diabetes. Therefore, we performed a large-sized cross-sectional study to assess the associations between higher proportion of normal attenuation muscle (good quality muscle) measured by abdominal CT and type 2 diabetes.

2 MATERIALS AND METHODS 2.1 Study population and definition of type 2 diabetes

We performed a cross-sectional study based on subjects aged 30 to 79 years who underwent abdominal CT scan as part of routine health check-up at the Health Screening and Promotion Center of Asan Medical Center (Seoul, Republic of Korea) between 1 January 2012 and 31 December 2013 (n = 22,051). Information on medication, previous medical or surgical diseases, history of type 2 diabetes in first-degree relatives, and habits on drinking, smoking, and exercise were obtained from each subject using a standard questionnaire. Drinking habits were calculated as grams per day and smoking habits were categorised as never, previous, or current. Regular aerobic exercise was defined as engaging in moderate-intensity aerobic activity for a minimum of 30 min for 5 days per week or vigorous-intensity aerobic activity for a minimum of 20 min for 3 days per week. Regular resistance exercise was defined as engaging in resistance training sessions for at least 3 days per week. We excluded 1026 subjects due to the absence of data or the presence of any pathological disorders including cancer, liver cirrhosis, chronic renal insufficiency, over thyroid dysfunction, severe anaemia or polycythaemia, and those currently taking glucocorticoids. Additionally, we excluded individuals using insulin as an antidiabetic therapy to eliminate probable cases of type 1 diabetes or latent autoimmune diabetes in adults (LADA; n = 39). Finally, a total of 20,986 subjects were included in this study (Figure S1). Type 2 diabetes was determined by fasting plasma glucose (FPG) ≥ 126 mg/dL, HbA1C ≥ 6.5%, or the use of antidiabetic medications as indicated on a questionnaire. The study protocol was approved by the Institutional Review Board of Asan Medical Center (IRB No. 2018-0917) and has therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.

2.2 Anthropometric, body composition, and laboratory measurements

Trained nurses measured the height and weight of subjects wearing light clothing without shoes. Body mass index (BMI) was calculated as the weight in kilogrammes divided by the square of the height in metres. Waist circumference (WC) was measured in a horizontal plane at the midway point between the inferior margin of the last rib and the superior iliac crest. Blood pressure was measured using an automatic manometer on the right arm with an appropriate cuff size after a resting period of ≥5 min.

Body composition was measured with direct segmental multifrequency bioelectrical impedance analysis using the InBody 720 (InBody Co., Ltd.). Appendicular skeletal muscle mass (ASM) was calculated as the sum of the lean muscle mass in the bilateral arms and legs.

After overnight fasting, early morning blood was drawn from the antecubital vein into vacuum tubes and analysed by a certified, central laboratory at Asan Medical Center. Glucose was measured using the hexokinase method with an autoanalyzer (Toshiba 200 FR Neo autoanalyzer; Toshiba Medical System Co.), and HbA1c level was measured by ion-exchange high-performance liquid chromatography using an automated system (BioRad Laboratories, Inc.) certified by the National Glycohemoglobin Standardization Program and aligned to the Diabetes Control and Complications Trial reference method. The levels of fasting total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, and triglyceride were measured using the autoanalyzer (Toshiba). Serum insulin concentrations were obtained via immunoradiometric assay (TFB). The homoeostasis model assessment of insulin resistance (HOMA-IR), an index of insulin resistance, was calculated as the FPG (mg/dl) multiplied by the fasting insulin (μIU/ml) divided by 405.

2.3 CT image acquisition, assessment of skeletal muscle area and quality

We used a standardised CT acquisition protocol for health check-ups as previously described.23 Body composition was evaluated with abdominopelvic CT scans using an automated artificial intelligence software developed using a fully convolutional network segmentation technique.24 The software automatically selected the axial CT slices at the L3 vertebrae inferior endplate level, which were then automatically segmented to generate the boundaries of total abdominal muscle area (TAMA), visceral fat area (VFA), and subcutaneous fat area (SFA). The TAMA included all muscles on the selected axial images (i.e., psoas, paraspinal, transversus abdominis, rectus abdominis, quadratus lumborum, and internal and external oblique muscles). An image analyst and a radiologist, both of whom were blinded to the clinical information, reviewed all selected CT slice and segmented areas. For muscle quality evaluation, the TAMA was divided into three areas according to the CT density as follows: (1) IMAT (−190 to −30 HU; apparent fat tissues between muscle groups and muscle fibres), (2) NAMA (+30 to +150 HU; healthy muscles with little intramuscular fat), and (3) LAMA (−29 to +29 HU; unhealthy muscles with intramuscular lipid pools).15 The skeletal muscle area (SMA; −29 to +150 HU) referred to the combined area of the NAMA and LAMA. The NAMA/TAMA index was calculated as NAMA divided by TAMA and multiplied by 100. The VFA and SFA were also demarcated using fat tissue thresholds (−190 to −30 HU).17

2.4 Statistical analysis

Statistical analyses were performed separately for each sex because of the differences in muscle mass and attenuation value.25-27 Continuous variables with normal distributions are expressed as mean ± SD, and those with skewed distributions are expressed as median and interquartile range. Categorical variables are expressed as percent (%). Continuous variables were analysed with Student's t-test or Mann-Whitney U test, and categorical variables were analysed with χ2 test. The NAMA/TAMA index was categorised into sex-specific quartiles (Q1–Q4) based on the levels of the whole study population, including those with type 2 diabetes cases and healthy controls. The ranges of each quartile of the NAMA/TAMA index were Q1 < 73.2, 73.2 ≤ Q2 < 78.6, 78.6 ≤ Q3 < 83.0, Q4 > 83.0 for men and Q1 < 66.9, 66.9 ≤ Q2 < 73.9, 73.9 ≤ Q3 < 79.3, Q4 > 79.3 for women. Demographic and biochemical characteristics of the study population sorted according to the quartiles of the NAMA/TAMA index were compared using one-way analysis of variance or Kruskal-Wallis test for continuous variables and χ2 test for categorical variables. Multivariate logistic regression models were carried out to determine the odds ratios (ORs) for type 2 diabetes. A receiver operating characteristic (ROC) curve analysis was used to estimate the cutoff values for the NAMA/TAMA index, and the Youden's index was used to identify the best cutoff values. The ROC curves were analysed using MedCalc® version 14.8.1.0 for Windows (MedCalc Software) according to the method described by DeLong et al.28 All statistical analyses were performed using IBM SPSS Statistics version 20.0 (IBM Corp.). P values < 0.05 were considered statistically significant.

3 RESULTS 3.1 Clinical characteristics and CT measurements

A total of 20,986 subjects were analysed, of whom 13,007 (62%) were men. The mean age was 53.0 ± 8.6 years in men and 53.3 ± 8.5 years in women. The mean BMI was 24.6 ± 2.8 kg/m2 in men and 22.7 ± 3.0 kg/m2 in women. The characteristics of the subjects with type 2 diabetes and healthy controls are summarised in Table 1. The prevalence of type 2 diabetes was 13.9% in men and 6.9% in women. Compared with healthy controls, subjects with type 2 diabetes were older, more obese (specifically, visceral obesity), and more likely to receive lipid-lowering medication; those with type 2 diabetes also had less metabolically favourable laboratory findings and higher prevalence of hypertension and family history of type 2 diabetes. Among men, TAMA and SMA were not significantly different between controls and those with type 2 diabetes; among women, however, TAMA and SMA were significantly higher in those with type 2 diabetes. In both men and women, subjects with type 2 diabetes had significantly lower values of NAMA and the NAMA/TAMA index and significantly higher values of LAMA.

TABLE 1. Demographic, anthropometric characteristics and CT measurements of the study population Men p-value Women p-value Healthy control Type 2 diabetes Healthy control Type 2 diabetes (n = 11,200) (n = 1807) (n = 7430) (n = 549) Demographic and anthropometric data Age (years) 52.5 ± 8.6 56.0 ± 8.3 <0.001 52.8 ± 8.4 59.5 ± 8.3 <0.001 Height (m2) 170.9 ± 5.6 169.9 ± 5.7 <0.001 158.4 ± 5.2 156.3 ± 5.3 <0.001 Weight (kg) 71.7 ± 9.7 73.2 ± 10.7 <0.001 56.7 ± 7.5 60.1 ± 9.5 <0.001 BMI (kg/m2) 24.5 ± 2.8 25.3 ± 3.1 <0.001 22.6 ± 2.9 24.6 ± 3.5 <0.001 WC (cm) 87.2 ± 7.6 90.2 ± 8.3 <0.001 78.4 ± 8.1 85.1 ± 8.6 <0.001 ASM (kg) 23.8 ± 2.9 23.6 ± 3.0 <0.001 16.1 ± 2.0 16.0 ± 2.3 0.69 Body fat (%) 21.7 ± 5.0 23.4 ± 5.5 <0.001 29.1 ± 6.2 32.6 ± 6.2 <0.001 Biochemical and clinical data SBP (mmHg) 124.6 ± 13.0 127.2 ± 14.1 <0.001 117.2 ± 14.3 125.7 ± 15.6 <0.001 DBP (mmHg) 80.3 ± 10.2 80.6 ± 10.6 0.26 73.0 ± 10.5 76.3 ± 10.0 <0.001 Glucose (mg/dl) 97.4 ± 9.7 135.9 ± 31.3 <0.001 94.2 ± 9.1 133.7 ± 29.5 <0.001 Cholesterol (mg/dl) 194.8 ± 33.7 179.0 ± 38.9 <0.001 199.2 ± 33.8 188.3 ± 41.1 <0.001 HDL cholesterol (mg/dl) 51.9 ± 12.9 48.7 ± 12.5 <0.001 62.1 ± 14.9 53.4 ± 12.9 <0.001 LDL cholesterol (mg/dl) 124.3 ± 30.0 109.5 ± 34.2 <0.001 123.4 ± 30.7 116.2 ± 35.1 <0.001 Triglyceride (mg/dl) 113.0 (82.0, 159.0) 129.0 (91.0, 185.0) <0.001 84.0 (62.0, 115.0) 114.0 (83.0, 149.5) <0.001 HbA1c (%) 5.5 ± 0.3 7.0 ± 1.1 <0.001 5.5 ± 0.3 7.0 ± 1.0 <0.001 Insulin (mIU/ml) 4.8 (2.8, 7.0) 6.4 (3.8, 9.3) <0.001 4.1 (2.5, 6.3) 6.5 (4.0, 9.1) <0.001 HOMA-IR 1.14 (0.66, 1.72) 2.10 (1.22, 3.17) <0.001 0.94 (0.57, 1.49) 1.96 (1.25, 3.13) <0.001 hsCRP (mg/dl) 0.05 (0.03, 0.11) 0.07 (0.03, 0.16) <0.001 0.04 (0.02, 0.08) 0.06 (0.03, 0.14) <0.001 Current smoker, n (%) 3671 (32.8) 663 (36.7) <0.001 256 (3.4) 24 (4.4) 0.30 Alcohol consumption (g/day) 14.0 (3.4, 43.5) 15.2 (3.2, 50.7) 0.05 0.41 (0.0, 2.25) 0 (0, 0.75) <0.001 Regular aerobic exercise, n (%) 1034 (9.2) 197 (10.9) 0.03 613 (8.3) 48 (8.7) 0.69 Regular resistant exercise, n (%) 1840 (16.4) 308 (17.0) 0.52 708 (9.5) 48 (8.7) 0.60 Menopause, n (%) - - - 3782 (50.9) 386 (70.3) <0.001 Hypertension, n (%) 4203 (37.5) 1040 (57.6) <0.001 1590 (21.4) 278 (50.6) <0.001 Taking anti-lipid drug, n (%) 1249 (11.2) 472 (26.1) <0.001 721 (9.7) 169 (30.8) <0.001 Family history of type 2 diabetes, n (%) 2210 (19.7) 698 (38.6) <0.001 1560 (21.0) 216 (39.3) <0.001 CT measurement data TAMA (cm2) 169.4 ± 22.4 170.0 ± 24.0 0.28 112.5 ± 13.7 115.9 ± 16.2 <0.001 SMA (cm2) 163.8 ± 21.7 163.5 ± 23.0 0.62 107.1 ± 12.9 108.7 ± 15.4 0.02 NAMA (cm2) 131.7 ± 21.0 127.5 ± 22.3 <0.001 81.6 ± 13.5 77.4 ± 15.7 <0.001 LAMA (cm2) 30.3 (24.1, 38.2) 33.6 (26.7, 42.4) <0.001 24.1 (19.0, 30.4) 29.4 (23.4, 38.1) <0.001 IMAT (cm2) 4.31 (2.77, 6.98) 4.85 (3.07, 8.33) <0.001 4.20 (2.58, 6.80) 5.68 (3.51, 9.18) <0.001 NAMA/TAMA index 78.9 (73.6, 83.3) 76.6 (70.4, 81.3) <0.001 74.2 (67.4, 79.6) 68.5 (60.5, 74.9) <0.001 SFA (cm2) 118.9 (94.5, 147.4) 117.0 (93.5, 147.7) 0.43 145.2 (116.3, 179.3) 158.8 (129.0, 200.5) <0.001 VFA (cm2) 132.2 (92.1, 173.9) 160.7 (118.6, 206.3) <0.001 63.7 (38.6, 95.4) 111.0 (75.1, 146.2) <0.001 Visceral-to-subcutaneous fat ratio 1.05 (0.79, 1.37) 1.31 (0.99, 1.67) <0.001 0.42 (0.29, 0.59) 0.64 (0.48, 0.84) <0.001 Note: Data are presented as mean ± SD, median (interquartile range) or n (%). Abbreviations: ASM, appendicular skeletal muscle mass; BMI, body mass index; DBP, diastolic blood pressure; HbA1c, glycated haemoglobin; HDL, high-density lipoprotein; HOMA-IR, homoeostasis model assessment of insulin resistance; hsCRP, high-sensitivity C-reactive protein; IMAT, intermuscular adipose tissue area; LAMA, low attenuation muscle area; LDL, low-density lipoprotein; NAMA, normal attenuation muscle area; SBP, systolic blood pressure; SFA, subcutaneous fat area; SMA, skeletal muscle area; TAMA, total abdominal muscle area; VFA, visceral fat area; WC, waist circumference. 3.2 Prevalence of type 2 diabetes and comparison of characteristics according to the quartiles of the NAMA/TAMA index

The prevalence of type 2 diabetes and the characteristics of the subjects according to quartiles of the NAMA/TAMA index are shown in Tables 2 and 3. The prevalence of type 2 diabetes showed a significantly decreasing tendency (p < 0.001) according to higher quartiles of the NAMA/TAMA index. Regardless of sex, compared with subjects in the lowest NAMA/TAMA quartile, subjects in the highest quartile tended to be younger and less obese, and had less fat mass, less VFA, lower blood pressure, lower glucose level, and favourable metabolic laboratory values except for total and LDL cholesterol in men. In contrast, as the quartiles of the NAMA/TAMA index increased, ASM and TAMA were decreased in both men and women.

TABLE 2. Prevalence of type 2 diabetes and comparison of various characteristics according to quartiles of normal attenuation muscle area (NAMA)/total abdominal muscle area (TAMA) index in men Quartile 1 Quartile 2 Quartile 3 Quartile 4 p-value (n = 3251) (n = 3253) (n = 3251) (n = 3252) Demographic and anthropometric data Type 2 diabetes, n (%) 626 (19.3) 480 (14.8) 391 (12.0) 310 (9.5) <0.001 Age (years) 56.6 ± 9.0 53.6 ± 8.1 52.1 ± 8.0 49.8 ± 7.9 <0.001 Height (m2) 171.1 ± 5.9 170.8 ± 5.6 170.7 ± 5.5 170.4 ± 5.6 <0.001 Weight (kg) 76.1 ± 11.1 72.7 ± 9.1 70.9 ± 8.5 67.9 ± 8.5 <0.001 BMI (kg/m2) 25.9 ± 3.1 24.9 ± 2.6 24.3 ± 2.4 23.4 ± 2.5 <0.001 WC (cm) 92.5 ± 7.8 88.7 ± 6.6 86.4 ± 6.4 82.9 ± 6.8 <0.001 ASM (kg) 24.1 ± 3.3 23.9 ± 2.9 23.7 ± 2.8 23.5 ± 2.7 <0.001 Body fat (%) 25.3 ± 5.0 22.7 ± 4.4 21.0 ± 4.2 18.6 ± 4.5 <0.001 Biochemical and clinical data SBP (mmHg) 127.6 ± 13.7 124.9 ± 12.6 124.4 ± 13.0 123.0 ± 13.0 <0.001 DBP (mmHg) 81.5 ± 10.6 80.6 ± 9.9 80.1 ± 10.1 79.2 ± 10.2 <0.001 Glucose (mg/dl) 106.1 ± 22.4 103.3 ± 19.1 102.0 ± 19.4 99.6 ± 17.7 <0.001 Cholesterol (mg/dl) 190.1 ± 36.3 192.6 ± 34.7 193.7 ± 34.5 194.1 ± 34.0 <0.001 HDL cholesterol (mg/dl) 50.2 ± 12.3 50.6 ± 12.5 51.4 ± 12.8 53.8 ± 13.5 <0.001 LDL cholesterol (mg/dl) 120.2 ± 32.1 122.2 ± 30.6 123.0 ± 30.4 123.0 ± 30.4 <0.001 Triglyceride (mg/dl) 121.0 (88.0, 169.0) 120.0 (87.0, 169.0) 114.0 (82.0, 161.0) 105.0 (77.0, 150.0) <0.001 HbA1c (%) 5.83 ± 0.84 5.71 ± 0.71 5.65 ± 0.68 5.57 ± 0.63 <0.001 Insulin (mIU/ml) 5.9 (3.6, 8.2) 5.3 (3.2, 7.4) 4.8 (2.8, 7.0) 4.0 (2.3, 6.4) <0.001 HOMA-IR 1.47 (0.89, 2.20) 1.31 (0.77, 1.93) 1.17 (0.67, 1.78) 0.97 (0.55, 1.59) <0.001 hsCRP (mg/dL) 0.07 (0.04, 0.15) 0.06 (0.03, 0.12) 0.05 (0.03, 0.11) 0.04 (0.02, 0.09) <0.001 Current smoker, n (%) 1040 (32.0) 1065 (32.7) 1107 (34.1) 1122 (34.5) 0.19 Alcohol consumption (g/day) 15.0 (3.3, 48.0) 15.2 (4.2, 46.5) 14.2 (3.4, 44.7) 11.4 (3.3, 38.0) <0.001 Regular aerobic exercise, n (%) 298 (9.2) 272 (8.4) 304 (9.4) 357 (11.0) 0.005 Regular resistant exercise, n (%) 448 (13.8) 538 (16.5) 532 (16.4) 630 (19.4) <0.001 Hypertension, n (%) 1717 (52.8) 1353 (41.6) 1176 (36.2) 997 (30.7) <0.001 Taking anti-lipid drug, n (%) 539 (16.6) 475 (14.6) 376 (11.6) 331 (10.2) <0.001 Family history of type 2 diabetes, n (%) 659 (20.3) 742 (22.8) 764 (23.5) 743 (22.8) 0.01

留言 (0)

沒有登入
gif