It has been reported that, in some patients, diabetes treatments with subcutaneous injections induce "non-palpable subcutaneous changes", which are detected only under ultrasound examinations as subcutaneous adipose tissue changes without showing any subcutaneous induration during inspection and palpation. In this study, we examined the influence of the non-palpable or palpable subcutaneous changes induced by such injections on the glycemic control. Changes in subcutaneous adipose tissue were identified on ultrasonography in 15 patients, but no induration was apparent on inspection or palpation in 6 of these patients (40 %). Continuous glucose monitoring showed markedly high postprandial glucose levels not only in patients with palpable induration but also in those with non-palpable subcutaneous changes. We also reported a case of type 1 diabetes with recurrent hypoglycemia and sustained high HbA1c levels. In this case, non-palpable subcutaneous changes were observed widely in the abdomen. By avoiding injections inducing non-palpable subcutaneous, the amount of insulin required was reduced, with a low frequency of hypoglycemia. These findings suggest that we should examine subcutaneous changes by ultrasonography in addition to inspection and palpation in cases of unexpected hyperglycemia or hypoglycemia following insulin treatment. Avoiding injections after detecting subcutaneous changes may therefore be useful for improving the glycemic control in diabetic patients.

Dietary monitoring is a technically complex, costly and time-consuming activity. The monitoring of the carbohydrate intake is desirable for patients with lifestyle-related diseases, including type 2 diabetes. We developed a simple tool for monitoring the daily carbohydrate intake (STMC) of healthy Japanese students. The tool is a one-page interview sheet that helps to simplify the calculation of the average carbohydrate intake (g/day) on the basis of the respondent's self-reported dietary intake disaggregated by 8 food groups. The validity of the STMC was tested in a study conducted on university students (n=10, 60 % female), by comparing the dietary intake estimated based on 24-hour recall. The mean carbohydrate intake as assessed by the STMC and 24-hour recall was 275.3±60.3 and 263.0±78.0 g/day, respectively. There was a positive correlation between the estimated carbohydrate intake as assessed by the STMC and 24-hour recall, with a Spearman rank correlation coefficient (r) of 0.709 (p = 0.022). With regard to the intake of different food groups, rice, which was the students' main source of carbohydrates, showed a significant correlation with the estimated carbohydrate intake (r = 0.730, p = 0.017). The STMC showed reasonable validity in the estimation of the carbohydrate intake and may have clinical applications.