The study protocol was designed in accordance with the Declaration of Helsinki and was approved by the ethics committee of the First Affiliated Hospital of Jilin University. This trial was registered at the China Clinical Trial Registration Center (ChiCTR 1800015214). Written informed consent was obtained from all patients prior to their enrollment.

Patients who attended the outpatient endocrinology department of the First Affiliated Hospital of Jilin University between March and September 2016 were eligible for inclusion in this randomized controlled trial.

Inclusion criteria in this study were age older than 65 years, glycated hemoglobin (HbA_1c) level 7.0% to 10.0%, and the ability to use a mobile phone. Exclusion criteria were illiteracy, abnormal liver and kidney function, severe diabetic complications, use of insulin pumps, and participation in other clinical trials.

A total of 91 patients were enrolled: 44 (19 males) in the intervention group and 47 (18 males) in the control group.

Patients were randomly assigned to the intervention and control groups using the random number sequence generated by SPSS Statistics version 17.0 (IBM Corp) in batches of 6 patients at a time. Patients in the intervention group were provided training to independently use the mHealth management app and upload the glucometer data, which was then automatically transmitted to the medical server (glucometer was connected to the mobile phone via Bluetooth). The medical teams logged on to the system and sent medical advice and reminders to patients to monitor their glucose levels via the personal messaging app or telephonically every 2 weeks. Patients in the control group received a free glucometer and were followed up through conventional outpatient clinic appointments. For the control group patients, no limitations were imposed to the number of visits; however, they were instructed to monitor and record their blood glucose data regularly.

The study dietitian offered guidance for blood glucose monitoring and provided dietary advice based on the individual blood glucose levels. Patients in the intervention group used the app-based diet management software to input daily dietary intake. The dietitian received the daily dietary record of each patient via the mHealth app. On the basis of the analysis of this information, once-monthly dietary recommendations were sent from the dietitian to patients in the intervention group. The control group received dietary guidance from dietitians during face-to-face meetings at baseline and at the conclusion of all study-related procedures.

Information pertaining to physical activity (daily calorie expenditure) was obtained from patients in the intervention group via text message. The patients were instructed on how to text pedometer data to the study personnel. This information was analyzed, and each patient in the intervention group was provided with guidance related to aerobic and resistance-based exercises. In the control group, guidance related to exercise was provided during face-to-face dietary counseling session during clinic visits.

All patients were followed up in the outpatient clinic at 3-month intervals. Patients in both groups underwent physical examination, blood biochemical tests, follow-up clinic visits, and ambulatory therapy by the same medical team.

In order to assess the condition of patients, the following data were reviewed at baseline and every 3 months until the end of the experiment (a total of 3 times): medical history, treatment details, physical examination, and laboratory investigations. Patient compliance was assessed by the frequency of uploading blood glucose data in the intervention group. At the end of the experiment, all patients completed a satisfaction questionnaire, which contained 7 questions, with each question awarded a score of 1 and the highest possible total score being 7 points. Higher total score indicated better satisfaction.

Data were processed using SPSS Statistics version 17.0 (IBM Corp). Normally distributed variables are presented as mean and standard deviation; nonnormally distributed variables are presented as median and interquartile range. Between-group differences to normally distributed variables were assessed using an independent sample t test, whereas those to nonnormally distributed variables were assessed using a Mann-Whitney U test. For intragroup comparison, normally distributed variables were tested by paired t test and nonnormally distributed variables were tested by Wilcoxon rank-sum test. P<.05 was considered indicative of a statistically significant difference. Study figures were created using SigmaPlot (Systat Software Inc).

We completed a 6-month, prospective, randomized controlled trial of the mHealth telemedicine system in patients with T2DM aged over 65 years. The results showed that the PBG and HbA_1c levels in the intervention group were significantly lower than those in the control group; our results are very similar to those reported by Lim et al [16] and Egede et al [17]. At the completion of 3 months, the PBG level in the intervention group was 1.02 mmol/L lower than that at baseline; by the completion of 6 months, the PBG levels showed a progressive decrease of approximately 1.21 mmol/L relative to that at the completion of 3 months. After 6 months, we also observed a significant difference between the intervention and control groups in this respect. At the completion of 3 months, the HbA_1c level in the intervention group had decreased by 1% as against 0.66% in the control group. Although the between-group difference was not statistically significant, the HbA_1c levels in the intervention group at 6 months showed a further decline of 0.13% as against an increase of 0.04% in the control group. After the sixth month, the intervention group showed continuous improvement in HbA_1c levels and the between-group difference was statistically significant; this finding is consistent with the results of Cho et al [18], who also demonstrated the efficacy of telemedicine interventions after a certain period of time. These findings suggest that older patients require time to familiarize themselves with the mHealth system. However, after self-training and remote support from the medical team, the patients started independent use of the portable smart device, which reflected in the positive effects [19]. In this research, both groups showed improved blood glucose and HbA_1c levels. This may be attributable to personalized medicine and dietary and exercise plans provided to all subjects [20]. However, without remote supervision and ongoing support, it is difficult to achieve sustained efficacy in the long term; thus, long-term follow-up is essential for older patients with diabetes [21]. During the study, 13 subjects in the control group (7 with hyperglycemia and 6 with hypoglycemia) and 5 subjects in the intervention group (4 with hyperglycemia and 1 with hypoglycemia) required adjustment of drug dosage for titration of glycemic control; however, none of the patients in either group experienced any serious adverse events or aggravation of complications. The intervention group had significantly fewer hypoglycemic events compared with the control group. This was likely attributable to prompt identification of the risk of hypoglycemia in the intervention group by the medical team via the mHealth platform and the consequent implementation of timely corrective actions [10,22,23]. After the trial, over 89% of patients in the intervention group continued to measure their blood glucose level 2 to 3 days each week. Intervention group satisfaction survey responses indicated that frequent communication with the medical team via the mHealth platform enhanced patient understanding of diabetes, increased their awareness, and helped alleviate depressive symptoms [8,24].

Previous studies have shown that telemedicine interventions can improve blood glucose control in patients with diabetes [25,26]; however, the applicability of telemedicine for older patients has rarely been discussed.

The study by Quinn et al [24] and Kim et al [27] showed that middle-aged and older patients with diabetes have good interaction in a mobile phone–based diabetes education environment and that it significantly improves the self-management of blood glucose levels. However, the study was conducted over a period of 1 month, which is too short to determine the compliance of older patients over the long term with remote intervention. However, Egede et al [17] conducted a 12-month-long study involving remote psychotherapy intervention for older diabetic patients. Older diabetic patients not only maintained good compliance but also achieved long-term glycemic control. However, the intervention involved only psychotherapy, and there was no routine medication-, diet- and exercise-related intervention. Therefore, the study was not designed to determine the advantages of remote intervention with regular outpatient treatment. Cho et al [18] performed a 6-month comparative study of telemedicine and traditional outpatient treatment; although the benefits of telemedicine were not found at 3 months, HbA_1c levels were significantly improved at 6 months and the benefit was mainly found among women aged over 40 years. Williams et al [28] conducted a 12-month long-distance interventional clinical trial among African Americans aged over 21 years; the results suggested that long-term remote interventions can improve long-term glycemic control. However, all the above studies involved remote interventions in a wide range of age groups. Since the cognitive ability of older patients is relatively low, the operation interface used by middle-aged patients cannot be expected to be equally effective in older patients. Therefore, we greatly simplified the user interface of our telemedicine system to make it suitable for use by older patients. This enhanced the confidence of patients and their ability to follow the advice and provided us with valuable data that can be analyzed.

In general, telemedicine facilitates good glycemic control in older diabetic patients. In this study, the personal and family medical history, smoking history, history of alcohol intake, birth history, history of drug allergy, and personal living environment were not included in the analysis [6]. However, these factors can potentially affect the nutritional status and function of major organs; in addition, this information is important for the assessment of the quality of life of patients [29,30]. Moreover, data collected from dietary caloric intake and expenditure are not as accurate as blood glucose data; therefore, the effect of dietary and exercise-related guidance on glycemic control was not reliably measured; it is necessary to develop an accurate data collection method for calorie intake and consumption [31]. When this was achieved, telemedicine assisted the medical team and allowed the team to provide timely warnings of the risk of hypoglycemia or hyperglycemia as well as encouraged patients to continue their diet and exercise plan.

In this study, PBG level in the intervention group was significantly lower than that in the control group after the first 3 months. The improvement in glycemic control was sustained after 6 months and showed a significant difference from that in the control group. Our results suggest that the improved glycemic control in the intervention group was attributable to improved communication between doctors and patients with real-time tracking of older diabetic patients by the mHealth system and improved patient compliance after implementation of mHealth monitoring. On the basis of our findings, we can conclude that telemedicine is effective and safe for older diabetic patients.