The effect of green tea on patients with type 2 diabetes mellitus: A meta-analysis

Abstract

Background:

To systematically evaluate the effect of green tea on patients with type 2 diabetes.

Methods:

A computer search Cochrane, PubMed, Embase, CNKI, Wanfang, VIP, and other Chinese and English databases were conducted for randomized controlled trials of green tea in the treatment of patients with type 2 diabetes. The duration of these trials spanned from the establishment of the database to January 10, 2024. The obtained data were subjected to meta-analysis using Stata15.1 software. A total of 15 articles were included, encompassing 722 patients.

Results:

The meta-analysis results showed that compared to the control group, green tea intervention significantly improved the improvement of fasting blood glucose (SMD = −0.41, 95% CI: −0.67 to −0.19, P = .001), glycated hemoglobin (SMD = −0.68, 95% CI: −1.15 to 0.21, P = .004) and insulin resistance index (SMD = −0.70, 95% CI: −1.18 to −0.22, P = .005) in the experimental group compared to the control group. The differences were statistically significant (P < .05).

Conclusion:

Green tea significantly improves fasting blood glucose, glycated hemoglobin and insulin resistance in patients with type 2 diabetes.

1. Introduction

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disease that typically develops after the age of 35 to 40 years and accounts for more than 90% of diabetic patients. The incidence of T2DM is increasing year by year, with a prevalence of about 10.6% in China. It is a heterogeneous group of diseases characterized by elevated blood glucose levels.^[1] Ultimately, the overall metabolic disorders manifest as disruptions of fat, sugar and protein metabolism. In the short term, these metabolic disruptions can lead to a range of associated symptoms, while long-term glucose metabolism disorders can result in macrovascular and microvascular complications.^[2] Therefore, good blood glucose management is crucial for T2DM patients to reduce complications and lower the risk of death. Long-term drug therapy has limitations, such as high medical costs and side effects. In recent years, non-pharmacological interventions have gradually become the preferred treatment for glycemic control in patients with T2DM.^[3]

Green tea is an unfermented tea that retains the green tone of fresh tea leaves in its finished products and brewed tea. Its impact on T2DM glycemic control is mainly attributed to components like tea polyphenols, caffeine and theophylline, which help lower glucose levels, improve insulin resistance, and control blood glucose levels. Moderate consumption of green tea can assist in lowering blood glucose. However, current research results on this topic are inconsistent.^[4–6] Therefore, this study conducted a meta-analysis on the glycemic control status of T2DM patients with green tea to provide corresponding evidence-based basis for clinical practice.

2. Information and collection methods

2.1. Search strategy

Cochrane, PubMed, Embase, China Knowledge, Wanfang, Wipo and other Chinese and English databases were searched using a computer, and the search period spanned from the establishment of the databases to January 2, 2024. The literature search was conducted using a combination of “subject + free word” terms, and references were traced back to expand the scope of the search. The search terms included: green tea, type 2 diabetes, and fasting blood glucose.

2.2. Inclusion and exclusion criteria

Inclusion criteria: research type: randomized controlled trials (RCTs) on the effect of green tea on glycemic control of patients with T2DM; research object: patients with T2DM; intervention method: green tea for the treatment group and conventional care for the control group; outcome indicators: fasting blood glucose, glycated hemoglobin, insulin resistance, BMI, etc.

Exclusion criteria: non-RCT studies, noncompliance with the criteria for enrollment of patients with T2DM,and the intervention method; unclear definition of the outcome indicators, misclassification, and errors in statistical methods; studies from which relevant data could not be extracted; studies with a bias grade of C.

2.3. Literature screening and data extraction

The database was searched independently by 2 researchers. The titles and abstracts of the articles were screened according to the inclusion and exclusion criteria. For those that met the criteria, the full text of the literature was further reviewed. Consensus was required for the inclusion of the articles; if consensus could not be reached, a third researcher resolved the disagreement. Data extracted from each study mainly included: included studies, source of subjects, sample size, and outcome indicators.

2.4. Quality evaluation

The Cochrane Evaluation Handbook 5.1.0 criteria for evaluating the risk of bias^[7] were used to evaluate the included literature. This evaluation included 7 entries: generation of randomized sequences, allocation concealment, blinding, description of missing visits, completeness of outcome data, selective reporting of outcomes, and other biases. Entries were graded as follows: Grade A for low-risk of bias, Grade B for moderate risk of bias, and Grade C for high risk of bias. Seven entries were graded as Grade A, suggesting a low-risk of bias, and some were graded as Grade B, suggesting a moderate risk of bias. None of the entries were graded as Grade C, suggesting a high risk of bias.

2.5. Statistical methods

Stata 15.1 software was used for statistical analysis. The I² and Q tests were used to test the heterogeneity among the studies. When I² > 50% and P < .1, it suggests that the studies are heterogeneous, and a meta-analysis was performed with a random-effects model. Conversely, when P > .1 and I² ≤ 50%, the studies were considered to be homogeneous, and fixed-effects model was used. Sensitivity analysis was performed on the sources that could potentially lead to heterogeneity (α = 0.05).Publication bias was tested using funnel plots and Egger method.

3. Results

3.1. Literature search results and literature characteristics

Initially, 762 documents were identified, and 486 documents remained after eliminating duplicates. After preliminary screening by reading the titles and abstracts, 28 documents were selected. Finally, 15 documents were included after further reading the full text.^[8–22] The details are shown in Figure 1. The basic characteristics of the included documents are shown in Table 1.

Figure 1.

Flow chart of study selection for inclusion in meta-analysis.

Table 1.

Basic characteristics of included studies.

Author	Year	Country	Sample size (intervention/control)	Outcome indicator
Ran et al[8]	2019	China	36/34	①③
Quezada et al[9]	2019	Mexico	10/10	①②③
Mahmoud et al[10]	2016	Iran	17/15	②④
Cinnamon et al[11]	2016	Iran	40/39	④
Borges et al[12]	2016	Brazil	23/24	②④
Rafraf et al[13]	2015	Iran	32/32	②③④
Liu et al[14]	2014	China	29/28	①②③④
Lasaite et al[15]	2014	China	17/14	②
Mousavi et al [16]	2013	Iran	24/14	③④
Hsu et al[17]	2011	China	35/33	①②④
Huyen et al[18]	2010	Sweden	11/12	①③
Mirzaei et al[19]	2009	Iran	26/46	①②
Nagao et al[20]	2009	Japan	23/20	②④
Fukino et al[21]	2008	Japan	29/31	①②
Teixeira et al[22]	2006	Brazil	9/9

Note: ①: fasting blood glucose; ②: glycosylated hemoglobin; ③: insulin resistance index; ④: BMI.

3.2. Quality assessment

All 15 papers included in this study were RCTs; 5 papers used the computerized randomized table of numbers method; 5 papers used allocation concealment; and outcome rater blinding was easier to control, so 8 papers used rater blinding. And most of the literature described case loss during the course of the study, with complete outcome data. All studies were comparable at baseline. The overall quality level of this study was moderately high, and the results of the quality assessment are shown in Figures 2 and 3.

Figure 2.

Quality assessment.

Figure 3.

Quality assessment.

3.3. Results of meta-analysis

3.3.1. Effect of green tea on fasting blood glucose in type 2 diabetes mellitus patients:

Eight papers^{[8,9,14,17–19,21,22]} evaluated the effect of green tea on fasting blood glucose levels in T2DM patients. The heterogeneity test P = .002, I² = 65.9%, and a meta-analysis was performed using random-effect model. The results showed that the green tea group was significantly better than the control group in reducing fasting blood glucose levels in T2DM patients (SMD = −0.41, 95% CI: −0.67 to −0.19, P = .001), as shown in Figure 4.

Figure 4.

Forest plot of the effect of green tea on fasting blood glucose in patients with T2DM.

3.3.2. Effect of green tea on glycated hemoglobin in type 2 diabetes mellitus patients:

Ten papers^{[9,10,12–15,17,19–21]} evaluated the effect of green tea on the glycated hemoglobin status of T2DM patients. The heterogeneity test results were P < .001, I² = 84.0%, and a meta-analysis was performed using a random-effects model. The results showed that the green tea group was significantly better than the control group in reducing glycated hemoglobin levels in T2DM patients (SMD = −0.68, 95% CI: −1.15 to −0.21, P = .004), as shown in Figure 5.

Figure 5.

Forest plot of the effect of green tea on glycated hemoglobin in patients with T2DM.

3.3.3. Effect of green tea on insulin resistance in type 2 diabetes mellitus patients:

Five papers^{[8,13,14,16,18]} evaluated the effect of green tea on insulin resistance in patients with T2DM. The heterogeneity test results were P = .014, I²=68.1%, and a meta-analysis was performed using a random-effects model. The results showed that the green tea group was significantly better than the control group in reducing insulin resistance in T2DM patients (SMD = −0.70, 95% CI: −1.18 to −0.22, P = .005), as shown in Figure 6.

Figure 6.

Forest plot of the effect of green tea on insulin resistance in patients with T2DM.

3.3.4. Effect of green tea on BMI in type 2 diabetes mellitus patients:

Nine papers^{[9–14,16,17,20]} evaluated the effects of green tea on BMI in patients with T2DM, with heterogeneity test P = .018, I² = 56.5%, and a meta-analysis was performed using random-effects model. The results showed that the green tea group was significantly better than the control group in reducing BMI in T2DM patients (SMD = −0.11, 95% CI: −0.40 to 0.17, P = .442), as shown in Figure 7.

Figure 7.

Forest plot of the effect of green tea on BMI in patients with T2DM.

3.4. Sensitivity analysis

The meta-analysis of the effect of green tea on the glycemic control status of T2DM patients remained within the normal range when sensitivity analysis was performed by systematically excluding single studies in each group (see Figures S1–S4, Supplemental Digital Content, http://links.lww.com/MD/N838).

3.5. Publication bias

The funnel plot test was carried out on the literature with fasting glucose, glycated hemoglobin, insulin resistance index, and BMI as the outcome indicators. The results showed that the left and right sides of the funnel plot were basically symmetrical (Fig. 8).The risk of publication bias was analyzed for the results of fasting glucose, glycated hemoglobin, insulin resistance index, and BMI in the included literature. The results of Egger test were (t = −0.03, P = .980), (t = −1.28, P = .237), (t = −2.57, P = .082), and (t = −0.44, P = .671), which indicated a small possibility of publication bias, suggesting that the conclusions of the literature were relatively robust.

Figure 8.

Funnel plot of green tea on glycemic control status in T2DM patients.

4. Discussion

T2DM is a heterogeneous group of diseases characterized by elevated blood glucose, and long-term disturbances in glucose metabolism will eventually cause macrovascular and microvascular complications. For patients with T2DM, effective blood glucose management is crucial to reduce complications and lower the risk of death, and green tea as a non-pharmacological intervention has gradually become a therapeutic approach for the treatment of glycemic control in patients with T2DM. In this paper, a meta-analysis was applied to summarize several similar studies and quantitatively merge the study effects. This approach helps to overcome the limitations of individual studies, such as insufficient test efficacy, large randomization error, and the influence of factors like region, population, and time on the conclusions. Only literature in Chinese and English was searched for this study, and literature in other languages was not included. The methodological quality of the included studies varied and most of the literature retrieved in the last 2 years was excluded due to incomplete data profiles and non-RCTs. The overall quality of the 15 studies was moderately high, which was mainly due to the lack of rigor in the study protocols. Only 5 studies used allocation concealment, while outcome rater blinding was easier to control, with 8 studies employing rater blinding. Meta-analysis showed that, compared with the control group, the green tea intervention significantly improved fasting blood glucose (SMD = −0.41, 95% CI: −0.67 to −0.19, P = .001), glycated hemoglobin (SMD = −0.68, 95% CI: −1.15 to −0.21, P = .004), and insulin resistance index (SMD = −0.70, 95% CI: −1.18 to −0.22, P = .005). The differences were all statistically significant (P < .05).The reason may be that, on the 1 hand, tea polyphenols in green tea help lower blood glucose, improve insulin resistance, and also lower blood lipids to prevent cardiovascular diseases. On the other hand, alkaloids in green tea may lead to insulin resistance, affecting the regulation of blood glucose. Additionally, tea polyphenols, with their roles as antioxidants and inhibitors of human fat metabolism, can help lower blood lipids and stop the progression of atherosclerosis.^[23] The results of this study are different from some studies, Quezada et al^[9] found that supplementation with 400 mg of decaffeinated green tea extract per day reduced the central augmentation index in T2DM patients. However, there was no significant decrease in total cholesterol, triglycerides, or glycosylated hemoglobin after 12 weeks of intervention. The reasons for these differences may include insufficient dose: the dose of green tea used in the study may not be sufficient to produce a significant effect, and the active ingredients in green tea may require a higher dose to produce a significant impact lipids and glycated hemoglobin; study design: factors such as the duration of the study, the sample size, and the characteristics of the study population may have affected the reliability of the results; individual differences: different individuals may have different metabolisms and response may vary among individuals, which may also lead to inconsistency in the results of the study; combined effects of interventions: the treatment of patients with T2DM is usually a combination of interventions, including medication, dietary control and exercise, etc. The green tea intervention may be affected by other therapeutic measures. The green tea intervention may be affected by other therapeutic measures, making it more difficult to observe the effects of green tea on lipids and glycosylated hemoglobin alone.

Limitations: only analyzed the effect of green tea on the glycemic control indexes of T2DM patients, but not other indexes such as self-management behaviors, complication rates, cost-effectiveness ratio, etc; part of the data was extracted or converted by software, which may have some deviation from the original data; different green tea intervention time, dosage, and type may result in the heterogeneity of each study.

5. Conclusion

The results of the current study show that green tea intervention, as a low-risk, high-efficiency, low-cost complementary therapy, can effectively improve the fasting blood glucose, glycated hemoglobin and insulin resistance in T2DM patients. However, its effect on BMI needs to be clarified by further research. In the future, more in-depth discussions can be conducted in terms of different intervention durations and dosages, with a view to providing new strategies for the management of clinical T2DM patients.

Acknowledgments

MJ contributed to the concept and design of the study, and performed data management and analysis. YL assisted with data management and statistical analysis of the report. The preparation of the manuscript was accomplished by MJ and XL. DL and HL reviewed the manuscript and provided important feedback to improve and structure the report. All authors have read and agreed to the published version of the manuscript.

Author contributions

Conceptualization: Xing-Ning Liu.

Data curation: Ming-Jie Jia.

Formal analysis: Ming-Jie Jia.

Resources: Ying-Chao Liang.

Writing – original draft: Ming-Jie Jia, Xing-Ning Liu.

Writing – review & editing: De-Liang Liu, Hui-Lin Li.