Current study is single center, retrospective observational study. We evaluated 74 females aged above 18 years, who performed CACS measurement, hair mineral analysis, BMD and calcium intake measurement at one medical checkup at Ajou University Health Promotion Center, Suwon, Korea from January 1st 2007 to June 30th 2011. Usually, major companies include these tests in their medical checkup program for board members or managers. If, there were more than two checkup data with these four tests during study period, only initial data were used. Exclusion criteria were three. First, individuals who have diseases or conditions that influence calcium metabolism, such as small intestine resection status, chronic renal failure and liver cirrhosis. Second, patients who have disease associated with bone mineralization, such as hyperparathyroidism, Cushing syndrome, adrenal insufficiency, hyperthyroidism, hypothyroidism, and rheumatoid arthritis. Third, participants who were taking any medications that can affect BMD, such as steroid hormones, thyroid hormones or anti-thyroid drugs, anticonvulsants, male and female hormone therapy including oral contraceptives, and calcium or vitamin D medications. Finally, 24 women excluded and 50 women were included. The Institutional Review Board of Ajou University Hospital in Suwon, South Korea, approved this study (AJIRB-MED-MDB-11-323).

CACS was represented by Agatston calcium score, which is suggested by Agatston et al.[8] at 1990. And, it is measured by 64-slice MDCT scanner (Brilliance 64; Philips Medical Systems, Best, the Netherlands). CACS was computed as the sum of all Hounsfield units (HU) in a lesion multiplied by the voxel volume in mm. Coronary artery calcification (CAC) was defined as a hyperattenuating focus (Four contiguous pixels, and was considered as tomographic numbers >130 HU for each pixel). A score for each focus of CAC was calculated by multiplying the focus area (mm²) by a density measurement defined by the peak CT number in the focus. The total CACS was calculated as the sum of scores for all foci in the epicardial arteries.

BMD was measured by using dual energy X-ray absorptiometry (DXA; Lunar Prodigy Advance, GE Lunar, Medison, WI, USA). Bone mineral content (BMC; g), area (cm²) and T-score were measured at the 1st, 2nd, 3rd, and 4th lumbar spine (L1-4), femoral neck and total femur. BMD (g/cm²) was calculated by dividing area (cm²) by BMC (g).

Hair samples (80 mg) were collected from four different points of occipital scalp. The samples were cut proximal portion of hair (3-4 cm from skin) with stainless steel sampling scissors. The participants were asked not to chemically process their hair for at least 2 weeks, such as dying, perm, straightening, or frosting. The hair also had to be free of all gels, oils, and hair creams before sample collection. Measurements were performed using a microwave temperature-controlled digestion technique and Perkin-Elmer Mass Spectrometer in a licensed and certified clinical laboratory that undergoes regular inspections with the Clinical Laboratory Division of the Department of Health and Human Services (Trace Elements Inc., Addison, TX, USA). The spectrophotometers of minerals were reported as unit of mg% (mg/100 g of hair).

Dietary calcium intake measured by 24-hr recalls method. The type and amount of food was recorded by direct interview with a nutritionist. Each mineral in daily diet was processed and evaluated using dietetic computer program, Computer Aided Nutritional analysis Program, version 3.0 (CAN-Pro; The Korean Nutrition Society, Seoul, Korea).

All participants underwent a full physical examination and anthropometric measurements. The anthropometric measurements were conducted with wearing light gown and no shoes, and after an overnight fasting. Height was measured to the nearest 0.1 cm, and weight was measured to the nearest 0.1 kg, by automatic height and weight measuring machine. Body mass index (BMI) was calculated by dividing weight (kg) by height squared (m²). Waist circumference was measured with a flexible tape, at the level of the umbilicus, and was recorded to the nearest millimeter. Systolic and diastolic blood pressures and pulse rates were measured by automatic blood pressure measuring machine at sitting position. Body fat percentage were measured at the fasting status, by Inbody 720 (Biospace Inc., Seoul, Korea), which is body composition estimator based on bioelectrical impedance analysis.

Blood samples were also collected after an overnight fasting. Measurements of serum glucose, total cholesterol, triglyceride, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, high-sensitivity C-reactive protein, homocysteine, calcium, blood urea nitrogen, and creatinine were performed.

All continuous variables were expressed as mean±standard deviation. One-sample Kolmogorov-Smirnov test was performed for normality test of all continuous variables. In normality test, CACS was non-normal distribution, and other major variables (hair and serum calcium level, dietary calcium intake and BMD) and age were normally distributed. Pearson's correlation test (in case of nonparametric variables, Spearman's correlation test) was performed for elucidating relationships between age and major variables, such as CACS, BMD, hair and serum calcium level and dietary calcium intake. All statistical analyses were two-tailed, and a P-value<0.05 was considered statistically significant. Data were analyzed by using SPSS 19.0 software (SPSS Inc., Chicago, IL, USA).

Baseline characteristics of study subjects were presented in Table 1. The mean age was 50.7±8.2 years (range, 40-75 years). The mean value of CACS was 0.52±2.55. Average hair calcium concentration was 101.24±56.27 mg%, average serum calcium level was 9.04±0.43 mg/dL, average serum creatinine concentration was 0.82±0.12 mg/dL, and average calcium intake was 809.53±253.26 mg/day. The mean BMD and T-score of L1-4 vertebrae were 1.124±0.168 g/cm² and -0.1±1.4, and these of total femur were 0.959±0.122 g/cm² and 0.0±1.1, respectively.

Age was positively correlated with CACS and dietary calcium while it was negatively correlated with lumbar and femoral BMD (Table 2). CACS showed negative correlation with BMD (r=-0.280, P=0.049 with L1-4, r=-0.310, P=0.028 with femur neck, r=-0.333, P=0.018 with femur total) and hair calcium level (r=-0.352, P=0.012), while CACS did not showed any relationship with dietary calcium intake and serum creatinine level (Table 3). Serum calcium level was negatively correlated with lumbar BMD. Lumbar and femoral BMD were significantly correlated each other. Dietary calcium intake showed no significant relationship with other variables (Table 2, 3). Other serum variables were not correlated with CACS, BMD, and hair calcium level. Figure 2 showed summary of our study results.