Effect of Nutrition Education Using Self-Monitoring of Blood Glucose (SMBG) on Glycemic Control in Non-Insulin-Treated Obese Type 2 Diabetes Patients ()
1. Introduction
The objective of diabetes treatment is to inhibit the development and progression of complications. The importance of glycemic control for inhibition of complications has been investigated in various large-scale clinical studies [1], and the close relationship between postprandial hyperglycemia and arteriosclerotic diseases, such as myocardial and cerebral infarction, has been clarified [2,3]. In response to these reports, the International Diabetes Federation (IDF) published the “Guideline for Management of Postmeal Glucose” in September 2007 [4], in which the target blood glucose level 2 hours after a meal was specified to below 140 mg/dl for diabetes patients, and selfmonitoring of blood glucose (SMBG) was recommended as a useful tool to achieve that target. The American Association of Diabetes Educators (AADE) also stated that“regular monitoring is an essential component of any diabetes management program” [5]. Studies on the educational effect of SMBG for non-insulin-treated type 2 diabetes patients have progressed and its effectiveness has been evaluated [6-9]. The conclusions of a systematic clinical study on blood glucose were as follows: “Our findings demonstrate that appropriate use of SMBG in poorly controlled, insulin-naive type 2 diabetic patients can be efficacious and clinically meaningful.” [10].
In our previous study [11], we verified that nutrition education using SMBG 2 hours after supper twice a week was effective for glycemic control in non-insulin-treated type 2 diabetes patients. We also showed that the effect may have been due to improved eating ability by selflearning the association between meals and blood glucose through SMBG. This study suggested that SMBG is a necessary tool for patients to learn the association between postprandial blood glucose levels and an ingested meal.
In addition, it has been reported that SMBG and intervention based on the educational programme, based on face-to-face encounters every 3 months and additional telephone contacts every month, significantly improved glycemic control [12]; however, no tool to educate the association between postprandial hyperglycemia and meals was used. Moreover, the Guidelines for Self-Monitoring of Blood Glucose in Non-Insulin Treated Type 2 Diabetes published by the IDF did not mention nutrition education on the association between postprandial hyperglycemia and ingested meals [13].
Thus, we investigated the effects of nutrition education using SMBG and newly developed educational tools on glycemic control in non-insulin-treated obese type 2 diabetes patients.
2. Materials and Methods
2.1. Procedure
2.1.1. Study Design
A randomized controlled study was performed.
2.1.2. Subjects
36 males and 25 females non-insulin-treated obese (Body mass index (BMI) over 25 kg/m2), type 2 diabetes patients aged 30 - 69 years with HbA1c of 6.9% - 9.3% and treated in the outpatient clinic in Niigata Medical Center, Nagaoka Chuo General Hospital, and Kariwagun General Hospital in Niigata prefecture, Japan, who gave consent for this study were registered. Those who had changed medication 6 months prior to the time of intervention (March 2010) were excluded. Sixty-one patients at the 3 hospitals were stratified according to the hospital, gender, age, BMI, and HbA1c and randomly allocated to intervention (n = 30) and control (n = 31) groups. To estimate the target sampling size, the endpoint was set to HbA1c reduction, and the reduction by intervention and that in the control group were assumed to be –0.5 and 0%, respectively, referring to a previous study [11]. Setting the approximate population variance at 0.3%, significance level at 5%, and power at 90%, the sample size for 2-sided analysis was estimated to be 60 or more in the 2 groups.
2.1.3. Intervention Method
The assessment period was set between September 2010 and February 2011. In the intervention group, patients measured the blood glucose level 2 hours after supper twice a week (weekdays and holidays) for 6 months, and recorded the following items on an “SMBG record form”: the content of the meal and blood glucose level 2 hours after the meal, place where the meal was eaten, presence or absence of exercise after the meal, and self-evaluation of the association between the meal and blood glucose level. Patients attended nutrition education concerning the association between blood glucose and meals using educational tools when they visited the hospital once every 2 months (3 times in total). The control group attended nutrition education without SMBG using the same educational tools once every 2 months. Target BMI and HbA1c after 6 months were set in the first session of nutrition education in both groups, but physicians were not informed of the target values, registered dietitians managed these values. Nutrition education was performed by the same registered dietitian at Niigata Medical Center and 4 registered dietitians each at the 2 other hospitals, using the same educational media and content, and the frequencies of assignment to the intervention and control groups were adjusted to be the same by each registered dietitian. The SMBG procedure was explained by a clinical technologist at Niigata Medical Center, a nurse at Nagaoka Chuo General Hospital, and a registered dietitian at Kariwagun General Hospital.
2.1.4. Evaluation Method
Prior and post surveys were performed in September 2010 and March 2011, respectively. The primary outcome was HbA1c, and the secondary outcome was BMI. In addition, the stages of change for eating the appropriate supper amount (the stages of change) were investigated to verify the process of the educational effect, and satisfaction with diabetes treatment and well-being were investigated to verify the continuity of treatment. Survey items were height, weight, and HbA1c on blood testing, satisfaction with diabetes treatment, well-being, the stages of change, and execution of exercise. Satisfaction with diabetes treatment was assessed using Bradley’s questionnaire (8 items) translated into Japanese by Ishii [14]. There were 7 choices of answer from “fully satisfied: 6 points” through to “not satisfied at all: 0 points”, showing that the evaluation is more favorable as the score increases. Well-being was also assessed using Bradley’s questionnaire (12 items) translated into Japanese by Ishii [15]. There were 4 choices of answer from “always: 1 point” through to “not at all: 4 points”, showing that the evaluation is more favorable as the score increases. In summation of the scores of satisfaction with diabetes treatment and well-being, the scores of negative items were inversed. The total score was analyzed. Regarding the stages of change, that of the important item of this study, “ingestion of an appropriate amount of supper (desirable calories and food balance)”, was surveyed using a 5-stage rating: continued for more than half a year, 1 point; continued for less than half a year, 2 points; thinking of starting within one month of preparation, 3 points; thinking of starting within half a year, 4 points; and not thinking of starting within half a year, 5 points [16].
For SMBG, ONE TOUCH Ultra Vue (Johnson & Johnson Co., Ltd.) was used, and the device was provided to patients out of our research funds. SMBG was performed at the patients’ home, and blood testing, questionnaires, and nutrition education took place at each hospital. Survey results were analyzed at the Diabetes Center of Niigata Medical Center. HbA1c was measured using an automatic glycohemoglobin measurement device, ADAMS A1c HA-8170, at Niigata Medical Center and Kariwagun General Hospital, and JCA-BM2250 (Nihon Denshi) at Nagaoka Chuo General Hospital. HbA1c is presented as National Glycohemoglobin Standardization Program (NGSP) values (=Japan Diabetes Society (JDS + 0.4%) [17].
In statistical analysis, after the test for normality, changes in HbA1c and BMI after intervention were analyzed using the paired t-test, and differences in changes after intervention between groups were analyzed using the unpaired t-test. The Wilcoxon signed-ranks test was employed for analysis of changes after intervention in the scores of attitude and behavior, and between-group differences in changes after intervention were analyzed using the Mann-Whitney test. For analysis of betweengroup differences in the change in exercise after intervention, the χ2 test of independence was employed. SPSS 19.0 for Windows was used for data analysis.
2.1.5. Subjects Analyzed
Subjects included in analysis are shown in Figure 1. Intention-to-treat analysis (ITT analysis) and per-protocol analysis excluding dropouts (pp analysis) were performed. Excluding 4 patients unable to follow due to transfer and admission, 5 dropouts who performed SMBG only 24 times or less, which is less than half of the target frequency, and 2 patients who changed to DPP-4 inhibitor treatment, in total 11 patients (36.7%), 19 patients were included in pp analysis. In the control group, excluding 2 admitted, 3 with a change in DPP-4 inhibitor treatment, one with additional treatment beside this intervention, one dropout, and one with discontinuation of medication, in total 8 patients (25.8%), 23 patients were included.
2.1.6. Ethical Considerations
Regarding ethical considerations, this study was performed conforming to the Declaration of Helsinki after examination and approval by the Ethics Committees of Niigata Medical Center and Niigata University of Health and Welfare. Examination and approval were also obtained from the Ethics Committees of collaborating hospitals, Nagaoka Chuo General Hospital and Kariwagun General Hospital. Collaboration by patients was optional, and written informed consent was obtained. Data were managed by assigning IDs to prevent identification of individual patients.
Figure 1. Flowchart of patients in the randomized controlled.
2.2. Nutrition Education Tool
The text of the nutrition education tool was composed of: 1) “To improve postprandial hyperglycemia and achieve appropriate HbA1c”, and 2) “Real-size photographs of dishes to learn the appropriate amounts of single meals”.
1) “To improve postprandial hyperglycemia and achieve appropriate HbA1c”
This section was as follows to facilitate understanding of the association between postprandial blood glucose and meals: (1) Why is postprandial blood glucose important? (2) Postprandial blood glucose and nutrients (including figures presenting “the speeds of conversion and ratios of nutrients to blood glucose) [18]; (3) Important points of meals; (4) Setting a target value of HbA1c; and (5) Lookback check list. The “Look-back check list” is shown in Figure 2. This list was prepared to check whether the measured postprandial blood glucose level was higher than the target and to identify the cause when it was higher. The check points were prepared based on approximately 3 viewpoints. Firstly, 5 important points concerning meals were established (1. Did you eat excessive amounts of staple food? White rice or whole grain?) [19,20]; 2. Did you eat excessive amounts of foods high in carbohydrates, except staple food? [19]; 3. Did you have more than one main dish? [19]; 4. Was your meal lacking in vegetables, seaweed, and mushrooms? [21]; and 5. Did you eat excessive amounts of oil or fat?) [20]. These items were accompanied by notes describing the presence of evidence, and the mechanism was closely explained. Secondly, 5 items concerning meals other than the important points were established (6. Did you eat fruit with your meal? 7. Did you eat snacks before your meal? 8. Did you eat snacks after your meal? 9. Did you eat quickly? 10. Did you drink too much alcohol?). Thirdly, 3 items other than those concerning meals were established (11. Did you forget to take your medicine? 12. Was the interval short between lunch and dinner? 13. Are you sick (fever)?) (13 items in total). Patients checked the list after each blood glucose measurement to confirm problems with their own meals.
2) “Real-size photographs of dishes to learn the appropriate amounts of single meals”
To facilitate the understanding of items 1, 2, 3, and 5 of the “important points of meals”, “the appropriate amounts of single meals” of related dishes were presented as real-size photographs. For item 4, photographs were omitted because it is easy for Japanese to identify appropriate amounts. Six dishes to reduce carbohydrates to a specific amount were presented with regards to staple food, such as those containing vegetables high in carbohydrates, such as potatoes and pumpkin, and gyoza containing wheat flour. Regarding the main dish, 8 dishes were presented, such as mixed proteinaceous foods and vegetables (e.g. boiled chicken and vegetables) and main dishes that tended to be eaten excessively (fried chicken and Chinese dishes). Real-size photographs of meals actually served to diabetes patients admitted to Niigata Medical Center were presented. To facilitate understanding of the amounts in the photographs, not only the energy, but also the weight (g), size, and number of pieces were presented numerically (e.g. 3 cm in length × 3 cm in width), and the approximate amount of 1 unit was also presented. For photographs of main dishes, the ratio of fat in one meal was presented, in addition to nutrition labeling of energy and fat. Tableware size was also presented in each photograph to facilitate understanding of the appropriate amount.
In the intervention group, 1) and 2) were used as nutrition education tools. In the control group, the same tools were basically used, but the “Look-back check list” was omitted from 1).
2.3. Training of Registered Dietitians
To perform identical nutrition education, registered dietitians at the 3 hospitals underwent training in the effective use of nutrition education tools.
3. Results
Table 1 shows patient characteristics at baseline, showing no significant differences between the 2 groups.