A Strong Correlation of Glucose Variability in Non-Insulin-Treated Patients With Type 2 Diabetes Evaluated Between Continuous Glucose Monitoring and Seven-Point Testing

Dear Editor,

Although glucose variability is associated with the development of microvascular complications in type 2 diabetes, ¹ no consensus exists for its assessment. Self-monitoring of blood glucose with 7-point testing daily (SMBG-7) is assigned as an intensive testing method for patients with type 2 diabetes. ² On the other hand, continuous glucose monitoring (CGM) samples interstitial glucose every 5 minutes. It seems more feasible for measuring glucose variability than SMBG-7. However, both of these 2 testing have seldom been evaluated in patients with type 2 diabetes treated with oral agents alone. We therefore conducted a study to evaluate the relationship between SMBG-7 and CGM.

We enrolled 5 female patients with a median age of 45 years (32-66 years), a median baseline A1C level of 7.5% (5.4-8.8%), and a median diabetic duration of 6 years (3-19 years), and who were treated with 2 to 3 kinds of oral antidiabetic drugs (metformin 100%, sulfonylurea 80%, dipeptidyl peptidase-4 inhibitor 80%). Accu-Chek® Performa meter and fresh supply of test strips were provided. They should perform 7-point testing each day for at least 3 consecutive days and concomitantly wear a CGM device (CGMS® Gold™, MiniMed Medtronic). They were instructed to perform 4 glucose calibration measurements (before each meal and at bedtime) by pressing the input button on the CGM device.

The mean of overall glucose measurements, standard deviation (SD) of overall glucose measurements, mean amplitude of glucose excursion (MAGE), M-value, and J-index, which could be calculated based on both testing methods, ³ were used to assess glucose variability. MAGE was calculated by measuring the arithmetic mean of the differences in consecutive peaks and nadirs, which were taken into consideration only if they exceeded 1 SD from the mean. ⁴ M-value and J-index were calculated by the automated algorithm EasyGV on the website (www.easygv.co.uk).

Each patient performed SMBG-7 for a mean of 4.2 days (3-5 days), and the mean frequency of testing each day was 6.8 times. Overall full compliance was 80%. Patients wore a CGMS for a mean of 3.9 days (3-5 days). The medians of the mean glucose level (157.4 mg/dL vs 151.3 mg/dL), SD (46.1 mg/dL vs 39.6 mg/dL), MAGE (83.7 mg/dL vs 81.4 mg/dL), M value (5.8 vs 4.3), and J-index (39.7 vs 32.0) were similar for both SMBG-7 and CGM data. The correlations of the aforementioned parameters from the 2 testing methods were high (Figure 1). The highest concordance was mean glucose level (r = .969, P < .001).

Figure 1.

Comparison of (A) SD (standard deviation), (B) MAGE (mean amplitude of glucose excursion), (C) M value, (D) J-index from SMBG-7 (self-monitoring of blood glucose with 7-point testing daily) and CGMS (continuous glucose monitoring system) for each participant. r, Pearson’s correlation coefficient.

Our study is the first report comparing SMBG-7 with CGM as a means to evaluate glucose variability in diabetic patients treated with oral agents only. SMBG-7 might provide information about the mean glucose level and glucose variability in patients treated with oral antidiabetic drugs alone similar to that provided by CGM. Due to budget restrictions, SMBG-7, a less expensive alternative than CGM, can be considered as a reasonable measure for assessing the glucose variability for non-insulin-treated patients with type 2 diabetes under reasonable good glycemic control.