Prognostic effectiveness of interactive vs. non-interactive mobile app interventions in type 2 diabetes: a systematic review and meta-analysis

Abstract

Background

Mobile app interventions are emerging as significant tools in managing the prognosis of type 2 diabetes, demonstrating progressively greater impacts. The effectiveness of these interventions needs further evidence-based support.

Objective

This study conducted a systematic review and meta-analysis of randomized controlled trials to evaluate the effectiveness of mobile app interventions in improving prognosis for patients with type 2 diabetes.

Methods

We searched PubMed, Cochrane, Embase, and Web of Science for relevant studies published from inception to 18 April 2024, adhering to the Cochrane Handbook guidelines. The quality of the included studies was assessed using the Cochrane risk of bias tool. Primary outcomes measured were changes in glycated hemoglobin (HbA1c) and diabetes self-management (DSM). Secondary outcomes included changes in diastolic blood pressure (DBP), systolic blood pressure (SBP), triglycerides(TG), total cholesterol(TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL), lipid profiles, fasting plasma glucose (FPG), body mass index (BMI), and Steps outcomes. Subgroup analyses were performed for the primary outcomes and for high-density lipoprotein (HDL), low-density lipoprotein (LDL), diastolic blood pressure (DBP), and systolic blood pressure (SBP). Interventions with or without interactions were also used as a basis for subgrouping.

Results

A total of 15 eligible articles involving 17 studies with 2,028 subjects (1,123 in the intervention group and 1,020 in the control group) were included in the synthesis. Interactive mobile app interventions significantly reduced HbA1c levels (SMD − 0.24; 95% CI, -0.33 to -0.15; P < 0.00001) and significantly improved diabetes self-care (SMD 0.71; 95% CI, 0.21 to 1.21; P = 0.005). Secondary outcomes, including FPG, LDL, DBP, and SBP, showed varying degrees of improvement. Subgroup analyses indicated that the intervention effect was more pronounced and less heterogeneous in the short-term (≤ 3 months) for younger Asian individuals (< 55 years) who used an interactive mobile app.

Conclusion

Interactive mobile app interventions effectively improve HbA1c levels and diabetes self-care competencies in patients with type 2 diabetes. These interventions offer supportive evidence for their clinical use in managing and prognosticating type 2 diabetes.

Systematic review registration

CRD42024550643.

Introduction

Diabetes is a serious, chronic disease characterized by elevated blood glucose levels. As of 2021, 529 million individuals worldwide have been diagnosed with diabetes, with 95% of cases being type 2 diabetes [1]. Additionally, approximately 100 million individuals are affected by prediabetes [2]. Diabetes poses a substantial threat to human life and health while also placing a considerable burden on the global economy [3]. During the 2019 coronavirus disease pandemic, individuals with diabetes exhibited at least a twofold increased risk of developing serious illness or succumbing to it, significantly impacting the healthcare system and economy [4]. This stark reality underscores the devastating impact of non-communicable chronic diseases on human health.

Once diagnosed with diabetes, lifelong care is required. Routine care and consultation are typically conducted via outpatient clinics and at home, with patients choosing their own convenient times to visit the clinic. This often results in untimely visits due to scheduling conflicts. Furthermore, the expense of travel is an additional burden and inconvenience for patients with disabilities [5, 6]. Consequently, diabetes self-management becomes a crucial aspect of care for patients with diabetes mellitus [2, 7]. A significant number of patients lack an understanding of basic nursing principles related to self-management awareness, blood glucose monitoring, and healthy living instructions. This lack of knowledge contributes to the suboptimal self-management of diabetes among patients. Digital health represents a viable and effective means of promoting health education and supporting the self-management of type 2 diabetes [8]. As of 2023, it is estimated that 69.2 billion people (86% of the world’s population) use smartphones [9]. The incorporation of health management technology into smart apps is becoming increasingly prevalent. Among the most common health management-related apps are those designed to assist with diabetes management [10]. Patients with diabetes can use intelligent mobile applications to aid in self-management of their condition. These applications offer various features, including video introductions, exercise guidance, glucose monitoring, and communication between patients and healthcare professionals [11,12,13]. With the advancement of technology, remote healthcare has evolved from its initial form, which relied on mobile phones and text messaging, to encompass two distinct categories: healthcare-related mobile applications on mobile devices and hardware for data collection and analysis [14,15,16]. Mobile app interventions represent a promising solution to the challenges posed by scheduling conflicts, lack of timely communication, and transportation issues in traditional routine care. These interventions are particularly relevant for rural and remote populations with limited access to health services and areas with a high prevalence of diabetes. Mobile apps can enhance patients’ awareness of diabetes and provide effective guidance on self-management. Additionally, doctors can use mobile apps to monitor patients’ conditions and provide timely advice. The high penetration of mobile apps enables interventions to reach diverse populations. In 2021, Eva Hilmarsdóttir et al. demonstrated that HbA1c levels were significantly lower in the intervention group than in the control group following an intervention using the SidekickHealth smart app [17]. In 2023, Adrian H. Heald et al. compared the results of an intervention using the “Healum” smart app with those of a control group [18]. The intervention group exhibited greater reductions in BMI and HbA1c, as well as higher scores on the EQ-5D-5 L questionnaire, a general measure of health, compared to the control group. These results indicate that mobile apps can improve certain indicators in patients with type 2 diabetes and enhance their quality of life.

Although many preliminary trials of mobile app interventions for type 2 diabetes have been conducted, there appears to be an insufficient quantity and quality of relevant summary articles. There is a lack of supporting evidence for the prognostic effectiveness of interventions based on mobile apps for people with type 2 diabetes. The principal evaluation criteria of relevant reviews are primarily limited to HbA1c levels [19,20,21], with less attention paid to other forms of validity data such as blood pressure, lipids, and DSM(diabetes self-management). Systematic evaluations have focused on various combinations of digital health technologies, predominantly SMS and web-based platforms, and have relied extensively on low-quality evidence from single-arm or before-and-after studies [22, 23]. Furthermore, there is a paucity of literature examining the specific intervention modalities of mobile apps, particularly whether mobile app-based digital interventions with or without interaction influence post-intervention outcomes in both scenarios. Given the current deficiencies in the literature, this study conducted a systematic review of the available evidence on the effectiveness of mobile app-based digital interventions in improving the prognosis of patients with type 2 diabetes. Subgroup analyses were also performed to explore whether significant improvements in the prognosis of patients with type 2 diabetes were associated with mobile app-based digital interventions, with or without interactions.

Methods

Literature search

The systematic review was performed following the Cochrane Handbook guidelines [24] and was registered prospectively on PROSPERO (CRD42024550643). A comprehensive literature search was performed using four databases: PubMed, Cochrane, Embase, and Web of Science. The search spanned from the inception of the databases to 18 April 2024. A combination of Medical Subject Headings (MeSH) terms, subject terms, and free terms was utilized. The following terms were employed: “type 2 diabetes”, “mobile apps”, and “randomised controlled trials”. The detailed search strategy is presented in Supplementary Table S1. Additionally, the reference lists of all eligible studies were manually reviewed. The included studies were independently searched and evaluated by two researchers (JL and ZY), and any discrepancies were resolved by consensus with a third researcher (ZT).

Identification of eligible studies

Studies meeting the following criteria were included: (1)Patients with a confirmed diagnosis of T2DM(type 2 diabetes mellitus).(2)Randomised controlled trials (RCTs).(3)Intervention group using only a mobile app intervention.(4)Control group using a routine clinic in an offline face-to-face format.(5)Published in English.

The following studies were excluded: (1)Studies lacking clinical data, including reviews, letters, conference abstracts, case reports, editorial commentaries, study protocols, unpublished articles, other systematic reviews, and meta-analyses.(2)Articles for which the full text or extracted data were unavailable.(3)Duplicates.(4)Studies involving interventions in addition to mobile apps, such as text messaging, phone calls, and wearable devices.

Data extraction

Two researchers (JL and ZY) independently extracted the following study characteristics from each included trial using a standardized table in Microsoft Excel (2016):(1)Trial characteristics: first author, year of publication, study period, study area, included population, registration number, sample size.(2)Patient characteristics: mean age, sex ratio, length of trial, and changes in patient outcomes from baseline to the end of the trial in the intervention and control groups.(3)Characteristics of the intervention: mobile app with or without interaction. Any differences were adjudicated by a third researcher (ZT). If the RCT included two intervention groups, data were divided into two distinct groups with separate controls. In cases of missing or unavailable data, authors were contacted in advance to obtain the necessary information.

Quality assessment

In accordance with the Cochrane Collaboration’s Risk of Bias Tool [24], two researchers (HL and LZ) independently evaluated the quality of evidence and risk of bias of the included trials across seven dimensions: sequence generation, allocation concealment, blinding of subjects and healthcare providers (HCPs), blinding of outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias. Discrepancies were resolved through discussion. Each aspect was scored as low risk, high risk, or unclear risk. Risk of bias summaries were generated using Review Manager 5.4 (Cochrane Collaboration, Oxford, UK) software.

Statistical analysis

Data presentation in this study used the mean ± SD format. If the included data were SE rather than SD, they were converted using the formula SD = SE × √n, and to SD if confidence intervals rather than SD were reported [24]. After harmonizing the data forms, we imported the data differences into Review Manager 5.4 and Stata v.15 SE (College Station, TX, USA) for meta-analysis. The standardized mean difference (SMD) with 95% confidence intervals (CI) was employed for continuous variables. The heterogeneity of the studies was assessed by the inconsistency index (I²). P < 0.05 or I² >50% indicated significant heterogeneity, prompting the use of a random effects model. Conversely, in the absence of heterogeneity (I² ≤50%), a fixed-effects model was employed. Additionally, we conducted a unidirectional sensitivity analysis to assess the stability of our findings.We utilized Stata v.15. SE to conduct Egger’s regression test on the outcomes of studies that included three or more trials, thereby visually assessing the potential for publication bias. A P-value of less than 0.05 was considered statistically significant for publication bias. When each subgroup had a minimum of three trials reporting significant heterogeneity, we conducted a univariate subgroup analysis. In addition to these, we also analyzed common outcome measures (DBP, SBP, HDL and LDL). The classifications of these analyses are as follows: (1) Population characteristics.(2) Intervention measures.

Results

Descriptions of studies

A systematic literature search yielded a total of 1,172 articles from PubMed (n = 179), Cochrane (n = 668), Embase (n = 35), and Web of Science (n = 290) databases. After removing 332 duplicates, 840 titles and abstracts were initially selected. Finally, 15 articles were included, reporting 17 studies [17, 18, 25,26,27,28,29,30,31,32,33,34,35,36,37]. The literature search process is illustrated in Fig. 1.

Fig. 1

Flow diagram for articles included in interactive vs. non-interactive mobile app interventions in type 2 diabetes (searched up to april 18, 2024)

Appendix table 3 provides a summary of the details pertaining to the 17 trials, with a total of 2,028 subjects (1,123 in the intervention group and 905 + 71 + 44 in the control group) included in the combined analysis. Geographically, 3 studies were conducted in China, 2 in Korea and Indonesia, and 1 each in Colombia, India, Sweden, the United Kingdom, Iceland, Switzerland, Iran, and Norway. The largest number of studies were conducted in Asia, totaling 9. Regarding the duration of publication, the years spanned from 2014 to 2024, with the highest number of publications recorded in 2019 and 2022 (four articles in each year). Over the past five years, 10 articles have been published. In terms of intervention duration, the range was between 3 and 6 months, with a mean duration of 4.35 months. Concerning the nature of the interventions, 7 of the study interventions were interactive, as detailed in Appendix table 3.

Risk bias

In terms of the method of randomisation, nine articles were deemed to have a low risk of bias, while only one article was considered to be at high risk. The method of randomisation was unclear in five articles. Regarding the concealment of allocation, eleven articles were unclear, one article was rated as high risk, and three articles were rated as low risk. Given the specificity of mobile app digital interventions, the majority of studies were unable to be blinded to researchers, subjects, or measurers. Consequently, for “blinding of investigators and subjects,” there were two low-risk, two high-risk, and eleven unclear cases. For “blinding of measurers,” there were two low-risk, one high-risk, and twelve unclear cases.In terms of selective reporting of findings, only one article was identified as unclear, while the remainder were classified as low risk. Similarly, all articles were assigned a low-risk rating for completeness of outcome data and other sources of bias. Figures 2 and 3 illustrate the risk of bias results.

Fig. 2

Risk of bias summary

Fig. 3

Risk of bias graph

Results of meta-analysis of clinical indicators

Effect on HbA1c (glycated hemoglobin)

Sixteen studies reported HbA1c intervention data, including 1,034 cases in the intervention group and 940 cases in the control group. Meta-analysis indicated that, compared with the control group, the intervention group showed a significant reduction in HbA1c levels (SMD − 0.24; [95% CI, -0.33, -0.15]). The results were statistically significant. Qualitative analysis (I² = 48%, P < 0.00001) (Fig. 4A) and visual assessment of the funnel plot showed no publication bias, with the Egger test yielding p = 0.251, indicating no statistically significant publication bias (Fig. 5A).

Fig. 4

Forest plots of (a) hba1c, b fpg, c bmi for interactive vs. non-interactive mobile app interventions in type 2 diabetes

Fig. 5

Funnel plots of (a) hba1c, b fpg, c bmi for interactive vs. non-interactive mobile app interventions in type 2 diabetes

Although no obvious heterogeneity was observed in HbA1c, we conducted a subgroup analysis using it as the main evaluation criterion (Table 1). The table reveals that mobile application interventions, whether interactive or non-interactive, resulted in a statistically significant reduction in HbA1c levels. However, the heterogeneity increased for non-interactive mobile applications (I² = 55%).Regarding the duration of intervention, both groups showed a statistically significant reduction, with a more pronounced effect observed when the intervention duration was ≤ 3 months. Nevertheless, the heterogeneity increased significantly (SMD − 0.37; [95% CI, -0.53, -0.21], I² = 76%, P < 0.00001). In terms of region, the intervention effect in Asia was more significant (SMD = -0.27), although the heterogeneity increased (I²=61%). The intervention in Europe showed no statistical significance. Considering average age, the intervention effect was more significant for individuals with an average age < 55 years (SMD = -0.30) compared to those with an average age ≥ 55 years (SMD = -0.19), with no observed heterogeneity.

Table 1 Subgroup analysis of self-management in patients with type 2 diabetes with telemedicine intervention

Effect on FPG (fasting plasma glucose)

In the context of the FPG meta-analysis, a total of six studies reporting results were included, involving 472 cases in the intervention group and 403 cases in the control group. Compared with the control group, the intervention group showed a significant reduction in FPG levels (SMD − 0.21; [95% CI, -0.35, -0.08]). The results were statistically significant and showed no significant heterogeneity (I² = 24%, P = 0.002) (Fig. 4B). Although there was a slight publication bias in the visual assessment of the funnel plot, the Egger test yielded a p-value of 0.132, indicating no statistically significant publication bias (Fig. 5B).

Effect on BMI (body mass index)

A total of nine studies reported BMI data, which included 680 cases in the intervention group and 573 cases in the control group. The meta-analysis revealed no significant difference in BMI between the two groups (SMD − 0.01; [95% CI, -0.12, 0.10]), indicating that the intervention did not have a statistically significant effect on BMI. Additionally, there was no heterogeneity observed among the studies (I² = 0%, P = 0.87) (Fig. 4C). Although the visual assessment of the funnel plot suggested slight publication bias, the Egger’s test showed a p-value of 0.225, indicating that there was no statistically significant publication bias (Fig. 5C).

Effect on TG (Triglyceride)

There were five studies investigating changes in TG, with 286 cases in the intervention group and 229 cases in the control group. The meta-analysis revealed no statistically significant difference between the two groups in terms of triglyceride levels (SMD − 0.02; [95% CI, -0.20, 0.15]), indicating that the intervention had no effect on triglycerides. Furthermore, there was no observed heterogeneity among the studies (I² = 0%, P = 0.81) (Fig. 6A). Although the visual assessment of the funnel plot suggested slight publication bias, the Egger’s test showed a p-value of 0.284, indicating that there was no statistically significant publication bias (Fig. 7A).

Fig. 6

Forest plots of (a) tg, b tc, c hdl, d ldl for interactive vs. non-interactive mobile app interventions in type 2 diabetes

Fig. 7

Funnel plots of (a) tg, b tc, c hdl, d ldl for interactive vs. non-interactive mobile app interventions in type 2 diabetes

Effect on TC (Total cholesterol)

Five studies investigated changes in TC, involving 286 patients in the intervention group and 229 patients in the control group. The meta-analysis revealed no statistically significant difference between the two groups in terms of total cholesterol levels (SMD 0.03; [95% CI, -0.15, 0.20]), indicating that the intervention had no effect on total cholesterol. Additionally, there was no observed heterogeneity among the studies (I² = 0%, P = 0.76) (Fig. 6B). The visual assessment of the funnel plot indicated a slight publication bias (Fig. 7B), though the Egger’s test showed a p-value of 0.211, suggesting that there was no statistically significant publication bias.

Effect on HDL (high-density lipoprotein)

In the context of the meta-analysis of HDL, a total of six studies were included, which comprised 316 participants in the intervention group and 259 participants in the control group. The meta-analysis indicated that there was no statistically significant difference in HDL levels between the two groups (SMD 0.27; [95% CI, -0.03, 0.58]), and there was significant heterogeneity among the studies (I² = 60%, P = 0.08) (Fig. 6C). Visual inspection of the funnel plot suggested the presence of publication bias (Fig. 7C), but the Egger’s test showed that this bias was not statistically significant (p = 0.691).Sensitivity analysis (Fig. 8A) revealed that the study by Bonn, S.E. (25) et al., published in 2024, was a major contributor to the instability of the overall HDL results. After excluding this study, the meta-analysis showed a statistically significant higher HDL level in the intervention group compared to the control group, and the heterogeneity was reduced (SMD 0.36; [95% CI, 0.00, 0.72], I² = 55%, P = 0.05). This suggests that the data from Bonn, S.E. may have been a significant source of heterogeneity in HDL studies.

Fig. 8

Sensitivity analysis plots for interactive vs. non-interactive mobile app interventions in type 2 diabetes: a hdl, b diabetes self-care, c steps

We performed subgroup analyses based on intervention modality, intervention duration, intervention region, and age group of the participants. We found differences primarily in the subgroup with a mean age of < 55 years. The results of the subgroup analyses are as follows: (1) Intervention Modality: The mobile app group without interaction (SMD 0.04; [95% CI, -0.20, 0.27], I² = 0%, P = 0.77) and the mobile app group with interaction (SMD 0.64; [95% CI, -0.13, 1.41], I² = 85%, P = 0.10) were both ineffective at significantly increasing HDL levels compared to the control group. (2) Intervention Duration: For interventions lasting > 3 months, there was no heterogeneity and no statistical significance (SMD 0.03; [95% CI, -0.22, 0.28], I² = 0%, P = 0.80). Conversely, for interventions lasting ≤ 3, there was significant heterogeneity and no statistical significance (SMD 0.48; [95% CI, -0.08, 1.03], I² = 74%, P = 0.09). (3) Publication Region: In Asia, the results showed significant heterogeneity and no statistical significance (SMD 0.48; [95% CI, -0.08, 1.03], I² = 74%, P = 0.09). In Europe, the results showed no heterogeneity but were not statistically significant (SMD 0.03 mmol/L; [95% CI, -0.22, 0.28], I² = 0%, P = 0.80). (4) Average Age: For studies where the average age was ≥ 55 years, there was significant heterogeneity and no statistical significance (SMD 0.55; [95% CI, -0.50, 1.60], I² = 83%, P = 0.30). In contrast, for studies where the average age was < 55 yeed to the control group, with no heterogeneity (SMD 0.28; [95% CI, 0.04, 0.53], I² = 0%, P = 0.02). The remaining subgroup analyses are detailed in Table 1.

Effect on LDL (low-density lipoprotein)

Through the screening of studies, a total of eight studies on LDL were included, which comprised 479 participants in the experimental intervention group and 401 participants in the control group. The results of the meta-analysis indicated that the intervention group achieved a significantly greater reduction in LDL levels compared to the control group. This result was statistically significant, with no observed heterogeneity (SMD − 0.14; [95% CI, -0.27, -0.01], I² = 0%, P = 0.04) (see Fig. 6D). Additionally, visual inspection of the funnel plot did not reveal obvious publication bias (Fig. 7D), and the Egger’s test showed no statistically significant publication bias (p = 0.346). Analysis of LDL subgroups revealed no significant heterogeneity across any of the subgroups.

Effect on SBP (Systolic Blood Pressure)

Five studies reported data on the effects of SBP, including 363 patients in the intervention group and 307 patients in the control group. For the effect size of SBP, we observed: SMD − 0.16; [-0.32, -0.01], which is a clinically meaningful effect size and the intervention group was better than the control group in redu`cing SBP levels; in addition, we also observed I² = 24%, P = 0.04, this report has no significant heterogeneity and is statistically significant (Fig. 9A). There was a slight publication bias in the visual analysis of the funnel plot, p = 0.854 by Egger test, and there was no publication bias statistically (Fig. 10A). Subgroup analysis is shown in Table 2.

Fig. 9

Forest plots of (a) sbp, b dbp, c diabetes self-care (d) steps for interactive vs. non-interactive mobile app interventions in type 2 diabetes

Fig. 10

Funnel plots of (a) sbp, b dbp, c diabetes self-care, d steps for interactive vs. non-interactive mobile app interventions in type 2 diabetes

Table 2 Subgroup analysis of self-management in patients with type 2 diabetes with telemedicine intervention

Effect on DBP (diastolic blood pressure)

A total of five studies reported on DBP, including 363 patients in the intervention group and 307 patients in the control group. The meta-analysis results showed that DBP levels in the intervention group were significantly lower compared to those in the control group (SMD − 0.17; [95% CI, -0.32, -0.01], I² = 30%, P = 0.03) (Fig. 9B). Although the funnel plot (Fig. 10B) suggested a slight possibility of publication bias, the Egger’s test indicated that there was no statistically significant publication bias (p = 0.970). Subgroup analysis results are presented in Table 2.

Results of DSM (diabetes self-management)

The meta-analysis of DSM included seven studies, with a total of 290 participants in the intervention group and 295 participants in the control group. The results demonstrated a significant improvement in self-management scores in the intervention group compared to the control group (SMD 0.71; [95% CI, 0.21, 1.21], P = 0.005), although the heterogeneity among the studies was high (I² = 87%) (Fig. 9C). The funnel plot analysis indicated a potential risk of publication bias (Fig. 10C), but Egger’s test showed that this bias was not statistically significant (p = 0.08). Sensitivity analysis did not reveal any individual study that significantly affected the overall results (Fig. 8B).

Subgroup analysis revealed that the use of an interactive mobile app significantly increased self-management scores and reduced heterogeneity (SMD 0.33; [95% CI, 0.07, 0.59], I² = 36%, p = 0.01). Additionally, interventions lasting > 3 months showed a statistically significant improvement in self-management scores with reduced heterogeneity (SMD 0.32; [95% CI, 0.09, 0.55], I² = 9%, p = 0.007). Furthermore, interventions conducted in Europe and for participants ≥ 55 years also demonstrated increased self-management scores with decreased heterogeneity (Table 2).

Effect on steps

Three studies reported data on steps(the outcomes related to physical activity levels, measured in step), with 187 participants in the intervention group and 169 participants in the control group. The results indicated that there was substantial heterogeneity among the studies (SMD 0.23; [95% CI, -0.16, 0.61], I² = 66%, p = 0.24) (Fig. 9D), and the intervention had no statistically significant effect on step count. The funnel plot (Fig. 10D) revealed visual evidence of publication bias, and the Egger’s test confirmed a significant publication bias (p = 0.02). However, sensitivity analysis demonstrated that the results were stable (Fig. 8C).

Discussion

In recent years, as the economy has grown, unhealthy habits such as smoking, alcohol consumption, sugar addiction and low physical activity have led to an alarming increase in the prevalence of diabetes [38]. According to the latest literature, it is projected that over 1.31 billion people worldwide will be living with diabetes by 2050 [39], with type 2 diabetes accounting for approximately 90% of these cases [40]. The World Health Organization has recently reported that nearly 1.55 million people will die of diabetes, making it the ninth leading cause of death worldwide [41]. This will place an even heavier burden on the global health economy. Recent studies have shown that type 2 diabetes prevalence is higher in populations with lower economic and educational levels [42]. Therefore, there is a need to reduce the risk of developing type 2 diabetes through lifestyle interventions as a basic measure to guide diabetic patients in their daily diet and exercise. Mobile apps have emerged as a promising and convenient tool for this purpose, offering timely doctor-patient communication, easy access for patients to seek help, and a lower financial burden. These apps are widely used to support behavior change and improve outcomes in patients with chronic diseases. During the COVID-19 pandemic, DCT applications (applications for tracking contacts of infectious disease cases) have played a crucial role in early disease detection, reducing the costs associated, and curbing the spread of the pandemic [43]. Similarly in type 2 diabetes, mobile app interventions have been have been shown to significantly reduce HbA1c levels [17] and SBP [30]compared to usual care.

We included 17 trials involving 2028 patients to evaluate the effectiveness of mobile apps as interventions for type 2 diabetes outcomes. Our review found that mobile app interventions led to statistically significant improvements in key clinical parameters including HbA1c levels, FPG, LDL, DBP, SBP and DSM.

For HbA1c levels, we observed that SMD = − 0.24, indicating a significant reduction in HbA1c levels in the intervention group compared to the control group. This effect size is clinically meaningful and aligns with findings from previous reviews [20, 44, 45]. The analysis also revealed no significant heterogeneity (I² = 48%), and sensitivity analysis confirmed the stability of these results. Regarding FPG levels, the intervention group showed a significant reduction with an SMD = − 0.21 (P = 0.002) and low heterogeneity (I² = 24%). The forest plot confirmed a decrease across all studies, and sensitivity analysis supported the robustness of this finding.In contrast, our meta-analysis found no clinically significant changes in BMI, consistent with similar previous reviews [14]. While mobile app interventions did not lead to significant improvements in TC, TG, or HDL, there was a reduction in LDL levels with an SMD = -0.14 and no observed heterogeneity (I² = 0%). This discrepancy may be attributed to variations in the quantity and quality of LDL-related reports among the 17 included studies. Sensitivity analyses revealed that one study [28] caused heterogeneity in TG outcomes, which was resolved upon its removal, resulting in I² = 0%. For HDL, a sensitivity analysis indicated that one study [37] might have contributed to the higher heterogeneity observed. TC results indicated no heterogeneity among studies, suggesting that mobile app interventions do not improve TC levels in T2DM patients.

Regarding DBP and SBP, mobile app interventions resulted in significant reductions in both DBP (SMD − 0.16) and SBP (SMD − 0.17). However, sensitivity analysis identified potential instability in these outcomes, possibly due to inadequate intervention depth, a small number of patients with concurrent hypertension, or reliance on antihypertensive medications rather than specific diabetes-related treatments. These factors might contribute to the observed instability in results.

In assessing DSM, we aggregated evaluation criteria from multiple studies, including “self-efficacy (DMSES),” “quality of life (HRQoL),” and “diabetes self-management ability (Cronbach score),” into a unified measure of DSM. Compared to the control group, the intervention group showed a significant improvement in self-management scores with an SMD of 0.71, although there was high heterogeneity among studies (I² = 87%). This high heterogeneity may stem from varying scoring standards across studies. In terms of step counts, the findings were insignificant, likely due to the limited number of relevant studies and the infrequent use of steps as an exercise measure.

Given the diverse characteristics of the interventions analyzed, we conducted a subgroup analysis to better understand the effectiveness of mobile app interventions and achieve precision in diabetes management. The subgroup analyses are summarized below:

First, based on different intervention methods, we observed that both interactive and non-interactive mobile apps can potentially lower HbA1c levels. Furthermore, interactive mobile apps appeared to exhibit more notable improvements and reduced heterogeneity in the studies. We posit that mobile apps offer advantages in terms of convenience, speed, and cost-effectiveness compared to traditional outpatient care. In cases where patients encounter common questions or unclear nursing methods during the intervention process, mobile apps may facilitate simpler access to remote guidance. Given that diabetes patients often experience complications such as hypertension and obesity, interactive mobile applications can customize individualized nursing plans for type 2 diabetes patients to comprehensively adjust prognosis when unconventional problems arise. In a study examining the link between user experience, perceived usefulness, and the intent to use Diabetes Care and Treatment (DCT) application, Tim Schrills [46] found that the supportive experience of users significantly influences their intention to engage with such apps. Their research indicated that proactive user behavior and a sense of support in their experience directly affect the efficacy of DCT applications. Additionally, it was identified that users primarily seek decision-related communication from the system. Consequently, we can conclude that in the context of healthcare actions, the quality of user experience and the interactive communication with the system are indeed pivotal factors affecting the effectiveness of digital interventions through mobile applications.

Second, the analysis showed that interventions lasting more than three months had no heterogeneity, while those lasting three months or less achieved a more significant reduction in HbA1c levels (SMD = -0.37). We believe that this phenomenon may occur because the intervention has strong patient compliance in the short term, so the effect is significant. However, the longer the intervention lasts, the worse the patient’s compliance under the guidance of the mobile application, so the above situation occurs.

Third, we observed that interventions in Asia (SMD = -0.27) were more effective than those in Europe (SMD = -0.12). We analyzed this phenomenon as follows. Previously, researchers had believed that diabetes was prevalent among rich and developed countries [47]. However, according to further predictions, 80% of diabetic patients in 2035 will be low- and middle-income people in Asia, Africa and other regions. The sick population is becoming the main body [48]. As a developed region, Europe’s social chronic disease management model has already become mature and stable over hundreds of years of social development. Compared with digital intervention through mobile applications, the European population may be more adaptable to traditional diagnosis and treatment models, it is challenging to accept the new forms of intervention. Asia has a growing population of type 2 diabetes and advanced 5G technology. The prevalence of type 2 diabetes is tilting towards the Asian population. At the same time, China, as a country leading 5G technology, promotes the widespread application of 5G technology in Asia, making mobile applications more convenient and easier to accept. The sick people in Asia will have a higher acceptance of new intervention technologies.

Finally, the analysis showed that participants < 55 years old (SMD = -0.30) had better outcomes for HbA1c levels compared to those > 55 years olde (SMD = -0.19). The findings indicate that younger individuals may be more inclined to embrace mobile apps. It is postulated that younger individuals are more amenable to internet-related information, more facile in the utilisation of associated techniques, and better able to collaborate with researchers. In addition, we of HDL and LDL subgroup analysis showed that the average age of < 55 people, the improvement of HDL was statistically significant, but the improvement is not obvious, the SMD = 0.28, and no heterogeneity; However, no statistically significant improvement in HDL was observed in any of the other subgroups. With regard to the intervention involving an interactive mobile app, the improvement in LDL was more pronounced when the intervention duration was ≤ 3 months and the intervention region was Asia than before the subgroup analysis. This finding provides indirect support for our previous analysis.

In the case of DSM, with interactive mobile app intervention and intervention duration > 3 months, the heterogeneity of subgroups decreased, but the effectiveness also decreased (SMD = 0.33, SMD = 0.32), We believe that the reduced effectiveness of interactive mobile app intervention may be due to the small number of included studies and a certain bias; the impact of intervention duration is consistent with previous analyses. As far as intervention areas are concerned, in Asia SMD = 1.10, which has been significantly improved; there is no statistically significant improvement in Europe, indicating that people in Asia will be more receptive to new intervention methods. The mean age subgroup of the intervention population was not significant for DSM. For SBP and DBP, when the intervention was an interactive mobile app, the intervention region was Asia, and the intervention duration was ≤ 3 months, SBP (SMD=-0.21) and DBP (SMD=-0.26) changed the same, and the improvement was greater. To be obvious. For people with an average age < 55 years old, their SBP improvement was not statistically significant, and DBP (SMD=-0.20) had a more significant improvement with low heterogeneity.

In summary, our study suggests the effectiveness of mobile app intervention on the prognosis of patients with type 2 diabetes. Mobile app intervention may improve the HbA1c, FPG, LDL, DBP, SBP and DSM. Compared with traditional diabetes management, knowledge promotion is mostly leaflets and oral lectures, and nursing methods are mostly face-to-face care through offline expert clinics. This seriously affects the medical needs of patients with type 2 diabetes in terms of time and distance; in the later stages of diabetes, If gangrene makes it difficult to walk, a family doctor may be a good choice, but this will undoubtedly increase the financial burden. With the rapid development of 5G technology, Digital health has obviously become a more convenient and cheaper option. As a type of Digital health, digital health applications can not only eliminate the distance restrictions between doctors and patients, but also enable information sharing. Diabetes-related knowledge can be transferred to patients through videos, games, comics, etc., and patients can also Feedback their questions to the doctor and get timely answers. Mobile app interventions have greatly improved the uneven distribution of medical resources. It also provides services such as scheduled monitoring of relevant indicators, medication reminders, and community communication, detecting and guiding users’ healthcare activities. The mechanisms by which these interventions affect the management of type 2 diabetes are likely to be multifaceted, involving behavioral, psychological, and clinical aspects. Perhaps the following two frameworks can clarify the mechanisms through which application interventions impact diabetes management. (1) Health Belief Model (HBM), The HBM suggests that health behaviors are shaped by one’s perceived vulnerability to illness, the perceived seriousness of the illness, the perceived advantages of preventative measures, and the perceived impediments to these measures [49]. Mobile applications have the potential to enhance these perceptual factors by offering tailored blood glucose monitoring feedback, educational resources specific to diabetes, and notifications to encourage adherence to treatment protocols, which can stimulate proactive health behaviors. (2) Self-Determination Theory (SDT): SDT underscores autonomy, competence, and relatedness as pivotal for behavioral motivation [50]. Mobile applications can bolster these elements by providing a personalized care management platform, celebrating user achievements to enhance self-efficacy, and integrating social networking to foster community ties and professional support.

An increasing number of mobile applications are being utilized in healthcare to address various challenges. Studies have demonstrated that closed-loop insulin delivery systems can enhance glycemic control in patients with type 1 diabetes [51], offering a personalized approach to treatment. In the realm of epidemiology, the vast amount of data and rapid updates have rendered traditional analysis methods less efficient. The advent of mobile applications has significantly improved this situation, as exemplified by contact tracing apps that have facilitated the detection of infection sources and effectively implemented isolation measures during the COVID-19 pandemic [52]. Pei-Duo Yu et al. [53] have utilized statistical distance centrality to design an optimal contact tracing algorithm for epidemic source detection, which can accurately identify super-spreaders in the context of epidemics. Mobile app interventions offer a potential solution to address the uneven distribution of healthcare resources. Such interventions may alleviate the pressures of scarce and imbalanced health care resources and relieve the strain on health care systems. During the COVID-19 pandemic, there has been a proliferation of false and misleading information. Ching Nam Hang et al. [54] have proposed MEGA (a framework that combines feature engineering and graph neural networks), which by counting triangle motifs and computing the distance centrality, provides accurate screening for the authenticity of information during the COVID-19 pandemic.

Limitations and strengths

This literature review innovatively summarizes the evidence on the effectiveness of mobile app interventions for improving the prognosis of patients with type 2 diabetes. In addition to the commonly studied HbA1c level, our review also examines several less-explored indicators that affect patient outcomes. While some reviews have analyzed mobile app interventions, this study is the first to perform a comprehensive subgroup analysis distinguishing between interactive and non-interactive mobile app interventions. Our findings demonstrate that different intervention approaches can significantly impact patient prognosis, providing valuable insights for the future development of mobile health applications. We utilized Egger’s test, subgroup analysis, and sensitivity analysis to offer a more comprehensive classification of the levels of evidence. Furthermore, our review did not impose any time limits on the literature included, ensuring that no relevant studies from earlier years were excluded.

However, there are several limitations to this literature review. Firstly, randomized controlled trials (RCTs) of mobile apps are inherently unblinded, making it impossible to eliminate the potential impact of the placebo effect. Given the unique nature of digital interventions through mobile applications, some studies are unable to implement blinding of researchers, participants, and outcome assessors. This limitation may expose the measurement outcomes to subjective influences from these individuals, potentially skewing the true data regarding the efficacy of digital interventions. Secondly, the consolidation of various similar metrics into a unified Diabetes Self-Care format may introduce inconsistencies that could affect the review’s results. Additionally, our review was limited to English-language articles, which may have introduced language and publication biases. Most of the included studies were published in Asia and Europe, which could be subject to biases influenced by local economic policies. Moreover, the demographic data of the participants included in the studies were narrow in scope, lacking racial diversity, which may not accurately reflect the broader global population affected by Type 2 diabetes. In our analysis, we have incorporated a variety of software with differing mechanisms of action and intervention strategies. Given the specialized operational logic inherent in mobile apps, we have not been able to classify the software with precision based on their mechanisms or modes of intervention. Furthermore, some mobile apps rely on users to manually upload monitored values, which may introduce subjective biases affecting the accuracy of the outcomes. This potential source of inaccuracy may warrant additional scrutiny. Finally, despite demonstrating the effectiveness of mobile app-based interventions, the quantity of RCTs is still insufficient, and the quality of these trials is uneven.

Conclusion

Self-intervention and maintaining a healthy lifestyle are both necessary and challenging for people with type 2 diabetes. Our literature review suggests that mobile app-based digital interventions may be effective in improving outcomes for patients with type 2 diabetes, indicating possible improvements in HbA1c levels, FPG, LDL, DBP, SBP, and Diabetes Self-Care. This study may provide reasonably robust evidence supporting the use of precision medicine approaches for self-intervention in type 2 diabetes management. However, it is important to note that there were no significant improvements in some outcomes, indicating that more rigorous, larger-scale, and long-term randomized controlled trials may be needed to fully confirm the effectiveness of these applications. Despite the rapid development of smart applications as a new intervention method for type 2 diabetes and the substantial benefits they bring to patients, this field is still in its early stages and faces several challenges. These challenges include the need for improved interoperability between different applications, ensuring the protection of users’ personal privacy, and providing effective user training for correct application use. Addressing these issues will be crucial for the future advancement and success of mobile app-based interventions in managing type 2 diabetes.