After overnight fasting, each participant underwent a physical examination and venous blood collection early in the morning. Body measurements were taken in the standing position. Body mass index (BMI) was calculated as weight (kg) / height (m)
2. Body fat, body fat percentage, muscle mass, and skeletal muscle mass were measured by the bioelectrical impedance method using multifrequency BIA (MF-BIA; InBody 770, Cerritos, CA, USA). The height-adjusted skeletal muscle mass index (SMI: skeletal muscle mass index, [kg]/square of height [m
2]) was calculated based on the obtained body weight for statistical analyses [
25]. Collected blood and urine samples were centrifuged and measured by each method. Plasma glucose levels were measured by the glucose oxidase method and HbA1c levels by high performance liquid chromatography (HPLC) on a HLC723-G9 (Tosoh, Tokyo, Japan; range of measurement, 1.9–18.6%; coefficient of variation [CV], 0.58%). Serum and urine creatinine levels were measured by the oxidase method (Sekisui Medical, Tokyo, Japan; limit of detection [LOD], 0.029 mg/dL; limit of quantitation [LOQ-CV10%], 0.205 mg/dL; range of measurement, 0.05–100 mg/dL; CV, 0.4%), and values in mL/min/1.73 m
2 calculated as follows: 194 x Serum creatinine (− 1.094) x Age (− 0.287) × 0.739 (if female) [
24]. Urine albumin was quantified using the immunoturbidimetry method (Nittobo Medical, Tokyo, Japan; LOD, 0.183 μg/mL; LOQ-CV10%, 0.697 μg/mL; range of measurement, 1.0–500 μg/mL; CV, 0.86%). Serum total cholesterol (TC) and triglyceride (TG) levels were assessed with an enzymatic method (Sekisui Medical, Tokyo, Japan; TC: LOD, 0.03 mg/dL; LOQ-CV10%, 0.17 md/dL; range of measurement, 5–1000 mg/dL; CV, 0.46%; TG: LOD, 0.20 mg/dL; LOQ-CV10%, 0.40 md/dL; range of measurement, 3–2000 mg/dL; CV, 0.68%). High-density lipoprotein cholesterol (HDL-C) was measured directly by homogeneous assay (Sekisui Medical, Tokyo, Japan; LOD, 0.07 mg/dL; LOQ-CV10%, 0.13 md/dL; range of measurement, 2–150 mg/dL; CV, 0.59%). Low-density-lipoprotein cholesterol (LDL-C) was determined using the Friedewald equation [
26]. Serum zinc concentration was measured using the ACCURAS AUTO Zn reagent kit (Shino-Test Corporation, Japan; LOD, 1.87 μg/dL; LOQ-CV10%, 5.312 μg/dL; range of measurement, 4.0–500 μg/dL; CV, 1.08%), which can be employed with all auto-analyzers widely used in hospital laboratories and does not need any serum pretreatment [
27]. Serum zinc levels were categorized according to the criteria of the Japanese Society of Clinical Nutrition [
28]: normal, ≥80 μg/dL; subclinical zinc deficiency, ≥60 μg/dL and < 80 μg/dL; and zinc deficient, < 60 μg/dL. DKD was categorized into stages according to the 2014 classification of the Joint Committee on Diabetic Nephropathy [
29]: stage 1 (pre-nephropathy), normoalbuminuria (A1) < 30 mg/gCr and eGFR ≥30 mL/min/1.73m
2; stage 2 (incipient nephropathy), microalbuminuria (A2) 30–299 mg/gCre and eGFR ≥30 mL/min/1.73m
2; stage 3 (overt nephropathy), overt albuminuria (A3) ≥300 mg/gCr and eGFR ≥30 mL/min/1.73m
2; stage 4 (kidney failure), eGFR < 30 mL/min/1.73m
2 and not on dialysis; and stage 5 (dialysis therapy), end-stage renal failure regardless of albuminuria status.