The subjects of this study were 52 persons exhibiting symptoms of diabetes with fasting blood glucose higher than 126 mg·dl-1 and vitamin D deficiency symptoms with 25(OH)D3 less than 15 ng·ml^-1. The subjects were elderly women aged 65 and over located in Y-si in G-do. From a total of 60 subjects selected in the early study, 15 people each were initially assigned to (1) the vitamin D intake and circuit training group, (2) the vitamin D intake group, (3) the circuit training group, and (4) the control group. However, because 2 people in the vitamin intake group, 4 people in the circuit training group, and 2 people in the control group dropped out, only the results for 52 people were included in the final analysis.

The researchers explained the purpose and procedures of this study to the subjects through an orientation before conducting the pretest, and took measures to begin the study after receiving test consent forms from the subjects who wanted to voluntarily participate. All plans for this study were approved by the Seoul National University Bundang Hospital Institution Review Board before beginning the study. The physical characteristics of the subjects are shown in Table 1.

The results obtained in this study were analyzed using SPSS PC⁺ for Windows (version 21.0). The descriptive statistics quantity was presented as the mean and the standard error of the mean (SE). In order to simultaneously analyze the average difference of the dependent variables between the four groups and between two tests, repeated two-way ANOVA was used. If the main effect of the group, the main effect of the test, or the interaction between the groups and the tests were significant, the difference between two tests within each group was verified using one-way ANOVA. The least significant difference (LSD) was calculated for post verification, and the difference between two tests within the same group was verified by paired t-test. In order to determine the rate of change of the pre-test and post-test, delta % [(posttest - pretest) / pretest × 100] was presented, and the significant level (α) of all statistical analysis was 0.05.

The main effect of the test on weight, fat mass, percent body fat, and BMI were significant, but significant changes in fat mass were not observed in all groups (Table 3).

The main effect of the test as well as the interaction between the group and the test were significant. However, significant changes were not observed in all variables associated with abdominal fat in the vitamin D intake group and the control group (Table 4).

The main effect of the test as well as the interaction between the group and the test in the TC (p < .001) and HDL-C (p < .001) analyses were significant. The interaction between the group and the test in the TG (p = .031) and LDL-C (p = .001) analyses was also significant (Table 5).

The interaction between the group and the test in the insulin, blood glucose, and HOMA-IR analyses were not statistically significant, but they tended toward improvement in the vitamin D intake group and the circuit training group (Table 6).

It has been reported that the concentration of serum vitamin D has an effect on muscular activity, and also affect the number and size of muscular fibers [12]. It has been reported that in a study investigating the relationship between vitamin D and fat free mass, involving the intake of 1,000 IU of vitamin D a day over 2 years, the size and number of type II muscle fibers increased significantly [13]. Bunout et al. [14] have reported that in a study comparing exercise training and vitamin D intake (400IU/day) over 9 months in elderly women, the fat free mass and the muscle strength in the exercise training group significantly increased, but there were no significant changes in fat free mass and muscle strength in the vitamin D intake group. As described above, contradictory results have been reported for the effect of vitamin D intake on fat free mass. In this study, the change in fat free mass in all groups was not significant. It is judged that the 12 week period of vitamin D intake applied for the elderly human subjects was too short to affect the fat free mass, and it is expected that if the subjects took vitamin D over a longer period in the future, significant changes in fat free mass will be more apparent. It has been reported that unlike vitamin D intake, regular physical activity effects positive changes in body composition [15]. Exercise training targeting patients with diabetes increase their health level and has positive effects on body composition [5]. On the other hand, there have also been research results that significant changes in body composition were not apparent despite participation in exercise training. Kadoglou et al. [16] have reported that in a study that implemented aerobic exercise at 50-75% of VO₂max for 6 months targeting patients with diabetes, a positive change in insulin resistance-related variables was shown, but there was no significant change in body composition. Accordingly, it is considered that exercise at less than medium strength for short periods is somewhat insufficient to induce a change of body composition. Therefore, this study suggests that in the case of diabetic patients, exercise at more than medium strength would be effective to change the body composition.

According to weight and BMI results obtained in this study, the most significant reduction was in the group with vitamin D intake combined with circuit training. However, it is difficult to determine whether a synergistic effect of vitamin D intake and circuit training for body composition was apparent. The mechanism through which the increase in the concentration of vitamin D in the body affects body composition is also difficult to explain, but it is interpreted that the vitamin D receptor (VDR) is activated by the action of 25(OH), and has a positive effect on body composition while improving the homeostasis control ability of calcium [2].

It is known that accumulation of abdominal fat is the main cause of insulin resistance and metabolic syndrome. It has been reported that the accumulation of abdominal fat increases the secretion of adipocytokine, and has a negative effect on diabetes, hypertension, and cardiovascular diseases [17]. In previous studies, it was reported that vitamin D deficiency often appears in cases of obesity, and the concentration of 25(OH)D shows a negative correlation with abdominal fat [18]. However, there were no significant changes in abdominal fat-related variables in the vitamin D intake group of this study. Similarly, Wamberg et al. [19] have reported that as a result of obese middle-aged women taking vitamin D at 7,000 IU per day over 26 weeks, their blood 25 (OH) D concentrations increased to a sufficient level, but there were no significant changes in subcutaneous fat mass and visceral fat mass.

It has been reported that abdominal fat tissue is significantly reduced by exercise training that increases physical activity. In this regard, Bacchi et al. [20] have reported that for middle-aged men and women with diabetes conducting aerobic exercise using a tread mill or cycle at 60-65% of HRR, 60 minutes per day, 3 times per week for 16 weeks, the abdominal visceral fat area and subcutaneous fat area were significantly reduced. In this study, the TFA and SFA were significantly reduced only in two groups conducting circuit training, and these results are similar to the results of previous studies that conducted aerobic exercise. Collectively, the abdominal fat results indicate that the increase of fat metabolism and the reduction of abdominal fat are effectively accomplished when combining exercise that increases direct physical activity with vitamin D intake rather than only independent treatment of vitamin D intake.

It has been reported that vitamin D plays an essential role in the homeostasis control of calcium. Vitamin D was closely connected to the metabolic mechanism of calcium, and increased blood calcium concentration has an effect on blood lipid concentration and insulin resistance [19]. In general, it has been reported that intake of vitamin D decreases LDL-C and TG [21], and increases HDL-C [22]. In this study, blood lipid analysis of the vitamin D intake group showed that HDL-C was significantly decreased, which contrasted with the above study. Heikkinen et al. [23] have reported that as a result of menopausal women taking 300 IU of vitamin D every day for 3 years, their TG and LDL-C concentrations were significantly increased, while TC and HDL-C concentrations were significantly reduced. Major et al. [24] also reported that in obese middle-aged women taking 600 mg of calcium and 200 IU of vitamin D per day for 15 weeks, TC, LDL-C, and HDL-C concentrations were significantly reduced. Unlike the changes in blood lipid according to general estimates of vitamin D intake, the study that reported the decrease of HDL-C also reported that recording the dietary daily log by a recall method can be inaccurate even when the dietary intake was prepared. It was therefore concluded that difficulty with dietary intake control and alcohol intake may have an effect on blood lipid.

In this study, the change in blood lipid was clearly shown in the two groups undertaking circuit training. Sung & Bae [25] reported that in diabetic elderly men participating in walking exercise at 65-75% of HRmax for 50 minutes a day, 3 times a week for 24 weeks, TC and TG were significantly reduced. Greene et al. [26] also reported that obese middle-aged men and women participating in endurance exercise at 70% of VO₂max, 3 times per week for 12 weeks had significant increases in HDL-C and significant decrease in LDL-C levels. The circuit training performed in this study is the form involving aerobic exercise. Since the result of this study showed that the effect of fat mass and abdominal fat on the blood lipid of the two groups that underwent circuit exercise were significantly reduced, it is considered that circuit training is an effective exercise to induce positive changes in blood lipid.

In recent years, a number of studies to improve insulin resistance targeting diabetic patients have reported differences in the insulin resistance index according to vitamin D concentration [27]. Talaei et al. [28] reported that comparing insulin resistance indices according to the concentration of blood 25(OH) D in middle-aged women with diabetes showed significant differences in blood glucose, insulin, and HOMA-IR between people who are normal (＞30 ng·ml^-1), people who have insufficiency (＜30 ng·ml^-1), and people with deficiency (＜20 ng·ml^-1). In this regard, VonHurst et al. [29] have reported that as a result of taking 4,000 IU of vitamin D every day for 26 weeks, the insulin and blood glucose levels of patients with diabetes was significantly reduced. These results suggest that vitamin D intake increases the secretion and sensitivity of insulin, and has an effect on the maintenance of homeostasis [4].

It has been reported that regular physical activity is the most effective method for improving insulin resistance [30]. Umpierre et al. [31] reported that conducting a meta-analysis of 23 research results that involved an aerobic exercise showed that aerobic exercise is effective in improving insulin resistance, regardless of type, strength, time and frequency of training. It could be concluded from these results that aerobic exercise has a positive effect on the insulin resistance index by inducing the oxidation of fatty acids as well as an increase in insulin sensitivity.

Based on the insulin resistance index obtained in this study, it was determined that the results were not statistically significant. The group that combined vitamin D intake and circuit training showed the most positive changes, and the reduction of fat mass and abdominal fat had an effect on improving the insulin resistance indices. It was determined that combining the two kinds of treatments rather than a single treatment involving only vitamin D intake or exercise training is more effective.