Efficacy of Modern Diabetes Treatments DPP-4i, SGLT-2i, and GLP-1RA in White and Asian Patients With Diabetes: A Systematic Review and Meta-analysis of Randomized Controlled Trials (original) (raw)

A systematic review and meta-analysis was conducted to assess the impact of ethnicity on the glucose-lowering efficacy of the newer oral agents, sodium–glucose cotransporter 2 inhibitors (SGLT-2i), glucagon-like peptide 1 receptor agonists (GLP-1RA), and dipeptidyl peptidase 4 inhibitors (DPP-4i), using evidence from randomized clinical trials (RCTs).

DATA SOURCES

A literature search was conducted in PubMed of all randomized, placebo-controlled trials of DPP-4i, SGLT-2i, and GLP-1RA. The search strategy was developed based on Medical Subject Headings (MeSH) terms and keywords.

STUDY SELECTION

A total of 64 studies that qualified for meta-analysis after full-text review based on predefined inclusion and exclusion criteria—RCTs with at least 50 patients in each arm, >70% of population from Asian or white group, duration ≥24 weeks, and publication up to March 2019—were selected for systematic review and meta-analysis.

DATA EXTRACTION

DATA SYNTHESIS

Change in HbA1c was evaluated by computing mean differences and 95% CIs between treatment and placebo arms.

LIMITATIONS

The study is based on summarized data and could not be separated based on East Asians and South Asians.

CONCLUSIONS

The glucose-lowering efficacy of SGLT-2i, and to a lesser extent DPP-4i, was greater in studies of predominantly Asian ethnicity compared with studies of predominantly white ethnicity. There was no difference seen by ethnicity for GLP-1RA.

Introduction

Type 2 diabetes presents a global threat to health. According to the International Diabetes Federation 2017 report, China has the highest number of people (114.4 million) with diabetes in the age-group 20–79 years. This is closely followed by India (72.9 million), which is projected to have the highest number of people with diabetes by 2045 (134.3 million) (1). Yet, most studies of diabetes are undertaken in Western populations (2), and treatment guidelines do not take ethnicity into account. The latest consensus report by the American Diabetes Association and the European Association for the Study of Diabetes recommends the use of relatively newer drugs such as sodium–glucose cotransporter 2 inhibitors (SGLT-2i), glucagon-like peptide 1 receptor agonists (GLP-1RA), and dipeptidyl peptidase 4 inhibitors (DPP-4i) in combination with metformin and lifestyle adjustments (3), unless cost is an issue. Nearly one-half of subjects with diabetes fail to achieve the recommended treatment target, and the reasons for this are multifactorial (4,5).

The pathophysiology and metabolic phenotype of type 2 diabetes differ markedly by ethnicity. For example, Southeast Asians and South Asians develop type 2 diabetes at younger age and lower BMI than whites (6–8), and β-cell deficiency has been reported to be a feature of Asian diabetes (6). These pathophysiological differences may impact on treatment efficacy, as most diabetes therapies largely target the underlying pathophysiological defects.

Even though a large number of studies have been carried out to measure safety and efficacy of antidiabetes agents, only a few report on these measures in those of different ethnicity. Previous meta-analysis on DPP-4i (reported in 2013) and GLP-1RAs (reported in 2014) based on ethnicity have reported that Asians responded better than non-Asians (9–11). These studies defined a population as Asian if it was >50% Asian and white if the population was <50% Asian and included studies of short duration. There are no previous studies reporting efficacy of SGLT-2 inhibitors based on ethnicity. Thus, we performed a systematic review and meta-analysis to comprehensively assess the impact of ethnicity on the glucose-lowering efficacy of relatively newer antidiabetes agents—SGLT-2i, GLP-1RA, and DPP-4i—using published evidence from randomized clinical trials (RCTs) reported up to 31 March 2019.

Methods

Participants of three groups were considered for the study: 1) those receiving DPP-4i alone or in combination with other drugs, 2) those receiving SGLT-2i alone or in combination with other drugs, and 3) those receiving GLP-1RA alone or in combination with other drugs.

Data Sources and Searches

The meta-analysis was carried out using methods proposed in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (12). A literature search was conducted in PubMed for studies published up to 31 March 2019 by two independent investigators (S.G. and A.Y.D.) of all randomized, placebo-controlled trials of DPP-4i, SGLT-2i, and GLP-1RA. The search strategy was developed based on Medical Subject Headings (MeSH) terms and keywords. The search algorithm is presented in detail in Supplementary Material.

Study Selection

A title-based search was conducted followed by abstract screening. A full paper search of potentially eligible studies was also performed based on the predefined inclusion and exclusion criteria. Any discrepancies in selection were resolved by a third researcher (E.R.P.).

Inclusion Criteria

RCT on adult, nonpregnant participants aged ≥18 years with type 2 diabetes.
The efficacy of the drug was the primary outcome of the study.
Study reported the effect of drug versus placebo on HbA1c in participants who were either drug naïve or on background therapy.
Study reported outcome by ethnicity, and one ethnic group constituted at least 70% or more participants.
Studies were filtered on the basis of human subjects, clinical trials, and age >19 years.
Study written in English.

Exclusion Criteria

Study duration was <24 weeks or >52 weeks.
The study had <50 participants in each study arm.
Participants were on insulin as background therapy.
Studies that were extensions of previous RCTs.
Studies that included participants in inpatient care.
Non-RCT studies and reviews.

Data Extraction

Data extraction was done for aggregated study-level data by two independent researchers. Absolute changes in HbA1c (%) from baseline to 24 weeks between the drug and placebo were considered as the primary end point of the study. If data were not available for 24 weeks, 52 weeks was considered. Studies with duration >52 weeks were excluded, as these were open-label extensions. A standardized prepiloted form was used to extract data from the included studies for assessment of study quality and synthesis. Extracted information included author, year of publication, sample sizes, participant demographics and baseline characteristics, interventions, and HbA1c outcomes of the lowest dose (in case multiple doses were reported and placebo). Further, the studies were classified by ethnicities provided in terms of whether percentage of participants in a particular ethnic group was >70%. A hierarchical approach was adopted to decide the relevance of studies based on title, abstract, and full manuscript. If a study had more than two relevant arms, each arm was treated separately. Selection process of relevant studies retrieved from databases was shown in a PRISMA-compliant flowchart.

We have used the classification of “Asian” or “white” of each study as it was reported in the manuscript. Where studies were conducted in relatively homogenous populations in Asian countries (e.g., Korea, China, Japan, and Taiwan), we have considered the participants to be in the “Asian” group. We have also followed the definitions of “East Asians” (China, Japan, Mongolia, North Korea, and South Korea) and “South Asians” (Afghanistan, Bangladesh, Bhutan, India, Maldives, Mauritius, Nepal, Pakistan, and Sri Lanka) from previous reports (13,14).

To ensure more robust ethnicity-specific outcomes, we required >70% of the population to be Asian or white for the study to be allocated to that ethnic group, and to ensure more robust treatment effects we limited our studies to those where the study duration was ≥24 weeks and where there were >50 participants in the intervention and comparison arms.

Quality Assessment

The quality of eligible studies was evaluated by two independent researchers using the Cochrane Collaboration’s risk-of-bias tool (15) for assessing the design, execution, and reporting of the included RCTs. Risk of bias was assessed in random-sequence generation and allocation concealment (selection bias), blinding of participants and personnel (performance bias), incomplete outcome data (attrition bias), and selective reporting (reporting bias). The risk of bias was classified as high, low, or unclear. Funnel plot and Egger test were carried out to assess the publication bias of the overall studies for each drug.

Data Synthesis and Analysis

For each drug group, a meta-analysis was performed with the combined data and stratified analysis by ethnic group using the meta package in RStudio (version 1.0.153). Change in HbA1c was evaluated by computing mean differences (MDs) and 95% CIs between treatment and placebo arms. The MDs were calculated as the change from baseline to end point. When SD was not reported, it was estimated by the formula provided by the Cochrane Handbook for Systematic Reviews of Interventions (16). Forest plots for white- and Asian-dominant groups were constructed using RStudio. Higgins _I_2 statistic was used to evaluate between study heterogeneity and classified as low (<25%), moderate (25–75%), or high (>75%) (16). The Q statistic was used as a test of heterogeneity. τ2 was estimated by the DerSimonian-Laird estimation method (17). A random effects model was used to estimate the pooled effect. Statistical significance was considered as P value <0.05. Tests for subgroup differences were carried out to check whether there were any significant differences between ethnic groups. Meta-regression analyses were performed to determine whether estimates of treatment effects were associated with prespecified clinical characteristics such as age, duration of disease, percentage of specific ethnic group, percentage of men, baseline BMI, and baseline HbA1c. A sensitivity analysis was conducted considering study duration from 12 weeks to 52 weeks for all of the three drug groups, with all other criteria kept the same.

This review is registered with PROSPERO (CRD42019133587).

Results

DPP-4i

Search Results and Study Characteristics

Initially, 1,411 articles were identified from the database, 12 articles were identified from references of other articles, and 26 articles were included in the meta-analysis. A total of 26 comparison pairs were retrieved, which satisfied the selection criteria (Supplementary Fig. 1_A_). The total number of study participants was 8,531, of whom 4,728 were randomized to treatment arms and 3,803 to the placebo arms. The white-dominant group (17 studies) consisted of 5,185 participants, of whom 3,051 participants were randomized in the treatment group and 2,134 participants in the placebo group. The Asian-dominant group (nine studies) consisted of 3,346 participants, of whom 1,677 participants were randomized in the treatment group and 1,669 participants in the placebo group (Fig. 1). The summary of included studies is shown in Supplementary Table 1.

Figure 1

Forest plot for white- and Asian-dominant groups for DPP-4i. See references 24–49 for complete citations.

Figure 1

Forest plot for white- and Asian-dominant groups for DPP-4i. See references 24–49 for complete citations.

Close modal

Quality of Included Studies and Publication Bias Assessments

For the adequacy of sequence generation, 13 studies were categorized as unclear and 13 studies were categorized as low risk. All the included studies achieved the double blinding for the participants and the personnel. The allocation concealment was unclear in 16 studies, and 10 studies were at low risk. There was no particular indication of incomplete data, selective reporting, or other biases in any of the included studies (Supplementary Table 1_A_ and Supplementary Fig. 1_B_). The Egger test and funnel plot suggested that there was no asymmetric pattern and no particular concern regarding publication bias (P = 0.626) (Supplementary Fig. 1_C_ and D).

Efficacy Outcomes

HbA1c data were pooled from the 26 comparison pairs from 26 studies. Overall, the difference between the treatment and placebo groups was −0.53 (95% CI −0.62, −0.44; _I_2 = 78%) favoring treatment (Fig. 1). In the white-dominant group, the difference between treatment and placebo groups was −0.49 (95% CI −0.59, −0.38; _I_2 = 74%) favoring treatment. In the Asian-dominant group, the difference between treatment and comparison groups was −0.62 (CI −0.80, −0.45; _I_2 = 84%) favoring treatment (Fig. 1). The test for subgroup differences (random effects model) showed no difference (P = 0.1919) between the two groups (Supplementary Table 2_A_).

Exploratory Analysis

The median duration of diabetes was 6.1 years (range 2.9–12.2) in the white-dominant group and 6.4 years (0.97–8.15) for the Asian-dominant group. The median HbA1c at baseline was 8.065% (7.8–8.6) in the white-dominant group and 8.5% (7.9–9.4) in the Asian-dominant group. The median BMI at baseline was 31.7 kg/m2 (28.1–32.90) in the white-dominant group and 25.9 kg/m2 (25.30–27.90) in the Asian-dominant group.

Meta-Regression

The univariate meta-regression analysis showed that age (P = 0.33), percentage participants who were white (P = 0.12), HbA1c at baseline (P = 0.98), duration of diabetes (P = 0.22), percentage of participants who were men (P = 0.60), and BMI at baseline (P = 0.38) were not associated with the change in HbA1c from baseline (Supplementary Table 3 and Supplementary Fig. 4).

Sensitivity Analysis

Here, we included additional studies with a shorter duration (from 12 weeks to 52 weeks). Out of the 1,411 articles identified from the database, an additional 7 studies were identified, totaling 33 studies included in the sensitivity analysis. Overall, the difference between the treatment group and comparison group was −0.59 (95% CI −0.70, −0.48; _I_2 = 94%) favoring treatment (Supplementary Fig. 1_E_). In the white-dominant group, the difference between treatment group and comparison group was −0.49 (95% CI −0.59, −0.39; _I_2 = 73%) favoring treatment. In the Asian-dominant group, the difference between treatment and comparison groups was −0.73 (95% CI −0.88, −0.57; _I_2 = 94%) favoring treatment (Supplementary Fig. 1_E_). Testing for subgroup differences (random effects model) showed a greater response in the Asians compared with the white-predominant group (P = 0.0098) (Supplementary Table 2_B_).

SGLT-2i

Search Results and Study Characteristics

Sixteen articles were included in the study from the 555 articles that were identified from the database (Supplementary Fig. 2_A_). The total number of study participants was 4,189, of whom 2,178 were randomized in the treatment group and 2,011 were from placebo group. The white-dominant group (nine studies) consisted of 3,015 participants, of whom 1,515 participants were randomized in the treatment group and 1,500 participants in the placebo group. The Asian-dominant group (seven studies) consisted of 1,174 participants, of whom 663 participants were randomized in the treatment group and 511 participants in the placebo group. The summary of included studies is shown in Supplementary Table 1.

Quality of Included Studies and Publication Bias Assessments

All of the studies were double-blind for the participants and personnel. For the adequacy of sequence generation, five studies were categorized as unclear, eight studies were categorized as low risk, and three studies were categorized as high risk. The allocation concealment was unclear in 4 studies, 11 studies were at low risk, and 2 studies were at high risk. There was no particular indication of incomplete data, selective reporting, or other biases in any of the included studies (Supplementary Table 1_B_ and Supplementary Fig. 2_B_). The Egger test and funnel plot suggested that there was no asymmetric pattern and no particular concern regarding a publication bias (Supplementary Fig. 2_B_ and C).

Efficacy Outcomes

HbA1c data were pooled from the 16 comparison pairs from 16 studies. Overall, the difference between the treatment group and comparison group was −0.79 (95% CI −0.91, −0.66; _I_2 = 80%) favoring treatment (Fig. 2). In the white-dominant group, the difference between treatment group and comparison group was −0.64 (95% CI −0.74, −0.53; _I_2 = 44%) favoring treatment. In the Asian-dominant group, the difference between treatment and comparison groups was −0.96 (95% CI −1.10, −0.82; _I_2 = 66%] favoring treatment. Testing for subgroup differences (random effects model) showed a significant difference (P = 0.0003) between the two groups (Supplementary Table 2_A_).

Figure 2

Forest plot for white- and Asian-dominant groups for SGLT-2i. See references 50–65 for complete citations.

Figure 2

Forest plot for white- and Asian-dominant groups for SGLT-2i. See references 50–65 for complete citations.

Close modal

Exploratory Analysis

The median duration of diabetes was 8.3 years (range 5.64–12.3) in the white-dominant group and 7.49 years (4.72–11.6) for the Asian-dominant group. The median HbA1c at baseline was 8.17% (7.8–9.3) in the white-dominant group and 8.18% (7.9–8.45) in the Asian-dominant group. The median BMI at baseline was 31.9 kg/m2 (31.2–33.3) in the white-dominant group and 25.59 kg/m2 (25.07–26.0) among the Asian-dominant group.

Meta-Regression

The univariate meta-regression analysis showed that percentage of participants who were white (P < 0.01) and BMI at baseline (P = 0.01) are associated with change in HbA1c from baseline. On the other hand, age (P = 0.38), HbA1c at baseline (P = 0.80), duration of diabetes (P = 0.85), and percentage of participants who were men (P = 0.1) were not associated with change in HbA1c from baseline (Supplementary Table 3 and Supplementary Fig. 5).

Sensitivity Analysis

Including five additional studies between 12 and 24 weeks’ duration, we pooled HbA1c data from 21 comparison pairs from 21 studies. Overall, the difference between the treatment group and comparison group was −0.70 (95% CI −0.82, −0.58; _I_2 = 84%) favoring treatment (Supplementary Fig. 2_E_). In the white-dominant group, the difference between treatment group and comparison group was −0.57 (95% CI −0.69, −0.44; _I_2 = 69%) favoring treatment. In the Asian-dominant group, the difference between treatment and comparison groups was −0.85 (95% CI −1.03, −0.66; _I_2 = 87%) favoring treatment (Supplementary Fig. 2_E_). Test for subgroup differences (random effects model) showed a difference (P = 0.0182) between the two groups (Supplementary Table 2_B_).

GLP-1RA

Search Results and Study Characteristics

A total of 1,481 articles were identified from the database and 4 articles were identified from references of other articles, from which 22 articles were included in the meta-analysis (Supplementary Fig. 3_A_). A total of 23 comparison pairs were retrieved, which satisfied the selection criteria. The total number of study participants was 6,559, of whom 3,608 were randomized in the treatment group and 2,951 were from placebo group. The white-dominant group (19 studies, 20 arms) consisted of 5,682 participants, of whom 3,608 participants were randomized in the treatment group and 2,951 participants in the placebo group. The Asian-dominant group (three studies) consisted of 877 participants, of whom 438 participants were randomized in the treatment group and 439 participants in the placebo group. A summary of included studies is shown in Supplementary Table 1.

Quality of Included Studies and Publication Bias Assessments

All the 22 of the studies achieved the double blindness for the participants and the personnel. For the adequacy of sequence generation, 14 studies were categorized as unclear and 8 studies were categorized as low risk. The allocation concealment was unclear in 6 studies, and 16 studies were at low risk. There was no particular indication of incomplete data, selective reporting, or other biases in any of the included studies (Supplementary Table 1_C_ and Supplementary Fig. 3_B_). The Egger test and funnel plot suggested that there was no asymmetric pattern and no particular concern regarding a publication bias (Supplementary Fig. 3_C_ and D).

Efficacy Outcomes

HbA1c data were pooled from the 23 comparison pairs from 22 studies. Overall, the difference between the treatment group and comparison group was −0.78 (95% CI −0.88, −0.69; _I_2 = 79%) favoring treatment (Fig. 3). In the white-dominant group, the difference between treatment group and comparison groups was −0.79 (95% CI −0.89, −0.69; _I_2 = 77%) favoring treatment. In the Asian-dominant group, the difference between treatment and comparison groups was −0.76 (95% CI −1.19, −0.33; _I_2 = 90%) favoring treatment (Fig. 3). Test for subgroup differences (random effects model) showed no statistically significant difference (P = 0.8957) between the two groups (Supplementary Table 2_A_).

Figure 3

Forest plot for white- and Asian-dominant groups for GLP-1RA. See references 47, 65–85 for complete citations.

Figure 3

Forest plot for white- and Asian-dominant groups for GLP-1RA. See references 47, 65–85 for complete citations.

Close modal

Exploratory Analysis

The median duration of diabetes was 7.6 years (range 2.8–13.6) in the white-dominant group and 9.3 years (4.0–13.7) for the Asian-dominant group. The median HbA1c at baseline was 8.1% (7.5–9.3) in the white-dominant group and 8.54% (7.95–8.6) in the Asian-dominant group. The median BMI at baseline was 32.94 kg/m2 (29.9–36.9) in the white-dominant group and 25.4 kg/m2 (25.3–26.8) in the Asian-dominant group.

Meta-Regression

The univariate meta-regression analysis showed age (P = 0.98), percentage of participants who were white (P = 0.99), HbA1c at baseline (P = 0.99), duration of diabetes (P = 0.54), percentage of participants who were men (P = 0.59), and BMI at baseline (P = 0.94) were not significantly associated with change in HbA1c from baseline (Supplementary Table 3 and Supplementary Fig. 6).

Sensitivity Analysis

Including four additional studies that were identified between 12 and 24 weeks’ duration, we pooled HbA1c data from 27 comparison pairs from 26 studies. Overall, the difference between the treatment group and comparison group was −0.79 (95% CI −0.88, −0.70; _I_2 = 76%) favoring treatment (Supplementary Fig. 3_E_). In the white-dominant group, the difference between treatment group and comparison group was −0.79 (95% CI −0.89, −0.70; _I_2 = 75.1%) favoring treatment. In the Asian-dominant group, the difference between treatment and comparison groups was −0.79 (95% CI −1.03, −0.54; _I_2 = 82%) favoring treatment (Supplementary Fig. 3_E_). Testing for subgroup differences (random effects model) showed no statistically significant difference (P = 0.9657) between the two groups (Supplementary Table 2_B_).

Discussion

The current systematic review and meta-analysis focuses on the HbA1c-lowering efficacy of DPP-4i, SGLT-2i, and GLP-1RA in ethnically white and Asian participants. Compared with whites, Asians respond better to SGLT-2i. Even though the primary analysis of DPP-4i showed no difference in response to DPP-4i between the two groups, the sensitivity analysis including shorter duration studies did show that Asians respond better to DPP-4i than whites, in keeping with previous reports. No difference was found in the response to GLP-1RA between Asians and whites.

This is the first meta-analysis that reports glycemic response to SGLT-2i by ethnicity. Our results showed that the Asians respond better to SGLT-2i compared with the white-dominant group. In the meta-regression, the percentage of the population who were white and BMI at baseline were associated with HbA1c reduction, whereas baseline HbA1c was not correlated. A recent meta-analysis showed that efficacy and safety of SGLT-2i were favorable in East Asian patients with type 2 diabetes (13). It is interesting to know that SGLT-2i also show greater, albeit nonsignificant, cardiovascular risk reduction in Asians compared with other ethnic groups (18).

Although no significant difference was found between the two groups for DPP-4i, the reduction in HbA1c levels at study end point was greater for the Asian-dominant studies (between-group difference P = 0.1919). However, our sensitivity analysis that included studies of shorter duration (from 12 weeks) did show a −0.11% significantly greater reduction in HbA1c in the predominantly Asian group compared with the predominantly white population. A previous meta-analysis by Kim et al. (9) reported that DPP-4i showed greater HbA1c-lowering effect in Asian-dominant studies than in non–Asian-dominant studies (between-group difference −0.18%, P = 0.006). In another review, Ito et al. (19) hypothesized that DPP-4i had greater efficacy among East Asian participants than their white counterparts due to the different pathophysiology of type 2 diabetes between the two ethnic groups. Interestingly, these two meta-analyses also included studies of 12 weeks’ duration or longer. Finally, in an individual-level analysis of Trial Evaluating Cardiovascular Outcomes With Sitagliptin (TECOS), Davis et al. (20) showed that the greatest initial reduction of HbA1c was observed in East Asians on sitagliptin. It is not clear why our primary analysis based upon studies of at least 24 weeks showed a smaller difference between ethnic groups; there was no obvious difference or bias introduced in the shorter-duration studies, but the results would suggest that the difference seen at 12 weeks does not persist to 24 weeks or longer. Previous studies on sitagliptin and vildagliptin have reported that the clinical pharmacokinetic characteristics of DPP-4i were not different among Asians, blacks, Hispanics, and whites, suggesting that differences in response are more likely to reflect phenotypic or pathophysiological differences (21,22). Lower BMI has been reported to be associated with a better glycemic response (23), yet in the current study, the meta-regression showed no correlation between age, BMI at baseline, percentage of participants who were white, HbA1c at baseline, duration of diabetes, percentage of participants who were men, and HbA1c reduction. This meta-regression suggests that the greater HbA1c reduction seen in Asians in our meta-analysis is not driven by the higher HbA1c at recruitment in the Asian populations. There may, however, be other differences between ethnic groups that are not captured in the recorded baseline characteristics, such as adherence, that could have contributed to differences in results. The use of other glucose-lowering agents in the trials was quite similar between Asian- and white-dominant studies and, hence, is unlikely to explain ethnic differences in treatment efficacy (Supplementary Table 4).

In the present meta-analysis, no difference in efficacy was found between the white- and Asian-dominant groups (P = 0.8957) among the studies of GLP-1RA. In a previous meta-analysis conducted on glucagon-like peptide 1 analogs by Kim et al. (10), it was found that HbA1c reduction from baseline was greater in Asian-dominant groups than in non–Asian-dominant groups. The fact that we show no evidence of difference in response between Asians and whites is at odds with this previous report. However, unlike for DPP-4i, for GLP-1RA, inclusion of shorter-duration studies in our analysis, similar to that of Kim et al., did not make any difference to the estimate of efficacy difference. Even though our study had only three Asian studies included, these differed from that of the three Asian studies included by Kim et al. in their meta-analysis. Thus, our lack of replication of the previous meta-analysis may reflect this small number of studies and heterogeneity between studies in the Asian population.

The only remaining differences were that in our study design, to ensure separation between studies reporting efficacy in whites versus Asians, we defined a population cutoff of 70%, whereas Kim et al. used a 50% cutoff, and we only included studies that had 50 patients per treatment arm.

There are some limitations to this study. First, this systematic review and meta-analysis is based on summarized data of RCT studies. Further investigation of individual-level trial data based on ethnicity is required to confirm the reported differences for SGLT-2i and DPP-4i. Second, due to lack of enough studies that satisfied our inclusion and exclusion criteria, studies could not be separated based on South Asians and East Asians. It is known from previous reports that South Asians and East Asians are ethnically heterogeneous, and this demands that more studies be conducted in the South Asian region (14).

In conclusion, the glucose-lowering efficacy of SGLT-2i and DPP-4i was higher in the Asian-dominant group compared with the white-dominant group but not for and GLP-1 analogs. Our data suggest that, if our results are replicated by individual-level patient analyses from clinical trials, ethnicity should be incorporated into the treatment guidelines.

Further studies would also be warranted as to the physiological or pharmacological basis of these differences, given the reported β-cell deficiency and visceral adiposity in Asians.

See accompanying article, p. 1687.

Article Information

Funding. The research was commissioned by the National Institute for Health Research (NIHR) using Official Development Assistance funding (INSPIRED 16/136/102). E.R.P. holds Wellcome Trust New Investigator Award 102820/Z/13/Z.

Duality of Interest. E.R.P. has received honoraria for speaking from Eli Lilly. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. S.G. researched data and wrote the manuscript. A.Y.D. researched data and reviewed and edited the manuscript. L.A.D. contributed to the discussion and reviewed and edited the manuscript. A.T.N.N. researched data and reviewed the manuscript. C.N.A.P. reviewed the manuscript. V.M. reviewed and edited the manuscript. E.R.P. researched data and reviewed and edited the manuscript. E.R.P. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

References

Cho

Shaw

Karuranga

, et al.

IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045

Diabetes Res Clin Pract

2018

;

138

271

–

281

Afroz

Ali

Karim

, et al.

Glycaemic control for people with type 2 diabetes mellitus in Bangladesh - an urgent need for optimization of management plan

Sci Rep

2019

;

10248

Davies

D’Alessio

Fradkin

, et al.

Management of hyperglycemia in type 2 diabetes, 2018. A consensus report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD)

Diabetes Care

2018

;

2669

–

2701

Jude

Nixon

O’Leary

, et al.

Evaluating glycemic control in patients with type 2 diabetes suboptimally controlled on basal insulin: UK ATTAIN Real-World Study

Diabetes Ther

2019

;

1847

–

1858

Brown

Nichols

Perry

The burden of treatment failure in type 2 diabetes

Diabetes Care

2004

;

1535

–

1540

Sattar

Gill

JMR

Type 2 diabetes in migrant south Asians: mechanisms, mitigation, and management

Lancet Diabetes Endocrinol

2015

;

1004

–

1016

Unnikrishnan

Gupta

Mohan

Diabetes in South Asians: phenotype, clinical presentation, and natural history

Curr Diab Rep

2018

;

Staimez

Weber

Ranjani

, et al.

Evidence of reduced β-cell function in Asian Indians with mild dysglycemia

Diabetes Care

2013

;

2772

–

2778

Kim

Hahn

Kwak

Park

Differences in the glucose-lowering efficacy of dipeptidyl peptidase-4 inhibitors between Asians and non-Asians: a systematic review and meta-analysis

Diabetologia

2013

696

–

708

Kim

Hahn

Park

Cho

Differences in the HbA1c-lowering efficacy of glucagon-like peptide-1 analogues between Asians and non-Asians: a systematic review and meta-analysis

Diabetes Obes Metab

2014

;

900

–

909

Singh

Incretin response in Asian type 2 diabetes: Are Indians different

Indian J Endocrinol Metab

2015

;

–

Liberati

Altman

Tetzlaff

, et al.

The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration

J Clin Epidemiol

2009

;

Yang

Zhang

Efficacy and safety of sodium-glucose cotransporter 2 inhibitors in East Asians with type 2 diabetes: a systematic review and meta-analysis

Diabetes Ther

2019

;

1921

–

1934

Nanditha

Ramachandran

, et al.

Diabetes in Asia and the Pacific: implications for the global epidemic

Diabetes Care

2016

;

472

–

485

Higgins

Altman

Gøtzsche

, et al.

The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials

BMJ

2011

;

343

d5928

The Cochrane Collaboration

Cochrane Handbook for Systematic Reviews of Interventions [Internet]

2011

Higgins

JPT

Green

, Eds.

Deeks

Altman

Bradburn

Statistical methods for examining heterogeneity and combining results from several studies in meta-Analysis

In Systematic Reviews in Health Care: Meta-Analysis in Context

London

BMJ Publishing Group

2001

, p.

285

–

312

Zinman

Wanner

Lachin

, et al.;

EMPA-REG OUTCOME Investigators

Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes

N Engl J Med

2015

;

373

2117

–

2128

Ito

Ambe

Kobayashi

Tohkin

Ethnic difference in the pharmacodynamics-efficacy relationship of dipeptidyl peptidase-4 inhibitors between Japanese and non-Japanese patients: a systematic review

Clin Pharmacol Ther

2017

;

102

701

–

708

Davis

TME

Mulder

Lokhnygina

, et al.;

TECOS Study Group

Effect of race on the glycaemic response to sitagliptin: insights from the Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS)

Diabetes Obes Metab

2018

;

1427

–

1434

Yin

Deckert

, et al.

Pharmacokinetics and pharmacodynamics of vildagliptin in healthy Chinese volunteers

J Clin Pharmacol

2009

;

–

European Medicines Agency

Scientific discussion [Internet]

2007

Monami

Dicembrini

Martelli

Mannucci

Safety of dipeptidyl peptidase-4 inhibitors: a meta-analysis of randomized clinical trials

Curr Med Res Opin

2011

;

(

Suppl. 3

–

Dou

Lui

, et al.

Efficacy and safety of saxagliptin in combination with metformin as initial therapy in Chinese patients with type 2 diabetes: results from the START study, a multicentre, randomized, double-blind, active-controlled, phase 3 trial

Diabetes Obes Metab

2018

;

590

–

598

Wang

Ning

, et al.

A randomized clinical trial of the safety and efficacy of sitagliptin in patients with type 2 diabetes mellitus inadequately controlled by acarbose alone

Curr Med Res Opin

2017

;

693

–

699

Yang

Xing

, et al.

Vildagliptin added to sulfonylurea improves glycemic control without hypoglycemia and weight gain in Chinese patients with type 2 diabetes mellitus

J Diabetes

2015

;

174

–

181

Lukashevich

Prato

Araga

Kothny

Efficacy and safety of vildagliptin in patients with type 2 diabetes mellitus inadequately controlled with dual combination of metformin and sulphonylurea

Diabetes Obes Metab

2014

;

403

–

409

Pan

Xing

Han

, et al.

Efficacy and tolerability of vildagliptin as add-on therapy to metformin in Chinese patients with type 2 diabetes mellitus

Diabetes Obes Metab

2012

;

737

–

744

Pan

Yang

Tou

Gause-Nilsson

Zhao

Efficacy and safety of saxagliptin in drug-naive Asian patients with type 2 diabetes mellitus: a randomized controlled trial

Diabetes Metab Res Rev

2012

;

268

–

275

Yang

Pan

Tou

Zhao

Gause-Nilsson

Efficacy and safety of saxagliptin added to metformin in Asian people with type 2 diabetes mellitus: a randomized controlled trial

Diabetes Res Clin Pract

2011

;

217

–

224

Yang

Guan

Shentu

, et al.

The addition of sitagliptin to ongoing metformin therapy significantly improves glycemic control in Chinese patients with type 2 diabetes

J Diabetes

2012

;

227

–

237

Kadowaki

Inagaki

Kondo

, et al.

Efficacy and safety of teneligliptin added to canagliflozin monotherapy in Japanese patients with type 2 diabetes mellitus: a multicentre, randomized, double-blind, placebo-controlled, parallel-group comparative study

Diabetes Obes Metab

2018

;

453

–

457

Gantz

Sokolova

Jain

, et al.

Use of Prohibited medication, a potentially overlooked confounder in clinical trials: omarigliptin (once-weekly DPP-4 inhibitor) monotherapy trial in 18- to 45-year-olds

Clin Ther

2017

;

2024

–

2037

Tinahones

Gallwitz

Nordaby

, et al.

Linagliptin as add-on to empagliflozin and metformin in patients with type 2 diabetes: two 24-week randomized, double-blind, double-dummy, parallel-group trials

Diabetes Obes Metab

2017

;

266

–

274

Matthaei

Catrinoiu

Celiński

, et al.

Randomized, double-blind trial of triple therapy with saxagliptin add-on to dapagliflozin plus metformin in patients with type 2 diabetes

Diabetes Care

2015

;

2018

–

2024

Barnett

Huisman

Jones

von Eynatten

Patel

Woerle

H-J

Linagliptin for patients aged 70 years or older with type 2 diabetes inadequately controlled with common antidiabetes treatments: a randomised, double-blind, placebo-controlled trial

Lancet

2013

;

382

1413

–

1423

Strain

Lukashevich

Kothny

Hoellinger

Paldánius

Individualised treatment targets for elderly patients with type 2 diabetes using vildagliptin add-on or lone therapy (INTERVAL): a 24 week, randomised, double-blind, placebo-controlled study

Lancet

2013

;

382

409

–

416

Derosa

Carbone

Franzetti

, et al.

Effects of a combination of sitagliptin plus metformin vs metformin monotherapy on glycemic control, β-cell function and insulin resistance in type 2 diabetic patients

Diabetes Res Clin Pract

2012

;

–

Taskinen

M-R

Rosenstock

Tamminen

, et al.

Safety and efficacy of linagliptin as add-on therapy to metformin in patients with type 2 diabetes: a randomized, double-blind, placebo-controlled study

Diabetes Obes Metab

2011

;

–

Rosenstock

Aguilar-Salinas

Klein

Nepal

List

Chen

CV181-011 Study Investigators. Effect of saxagliptin monotherapy in treatment-naïve patients with type 2 diabetes

Curr Med Res Opin

2009

;

2401

–

2411

DeFronzo

Hissa

Garber

, et al.;

Saxagliptin 014 Study Group

The efficacy and safety of saxagliptin when added to metformin therapy in in patients with inadequately controlled type 2 diabetes with metformin alone

Diabetes Care

2009

;

1649

–

1655

Nauck

Ellis

Fleck

Wilson

Mekki

Efficacy and safety of adding the dipeptidyl peptidase-4 inhibitor alogliptin to metformin therapy in patients with type 2 diabetes inadequately controlled with metformin monotherapy: a multicentre, randomised, double-blind, placebo-controlled study

Int J Clin Pract

2009

;

–

Garber

Schweizer

Baron

Rochotte

Dejager

Vildagliptin in combination with pioglitazone improves glycaemic control in patients with type 2 diabetes failing thiazolidinedione monotherapy: a randomized, placebo-controlled study

Diabetes Obes Metab

2007

;

166

–

174

Bosi

Camisasca

Collober

Rochotte

Garber

Effects of vildagliptin on glucose control over 24 weeks in patients with type 2 diabetes inadequately controlled with metformin

Diabetes Care

2007

;

890

–

895

Rosenstock

Brazg

Andryuk

Stein

Sitagliptin Study 019 Group

Efficacy and safety of the dipeptidyl peptidase-4 inhibitor sitagliptin added to ongoing pioglitazone therapy in patients with type 2 diabetes

Clin Ther

2006

;

1556

–

1568

Gomis

Espadero

R-M

Jones

Woerle

Dugi

Efficacy and safety of initial combination therapy with linagliptin and pioglitazone in patients with inadequately controlled type 2 diabetes: a randomized, double-blind, placebo-controlled study

Diabetes Obes Metab

2011

;

653

–

661

Bergenstal

Forti

Chiasson

J-L

Woloschak

Boldrin

Balena

Efficacy and safety of taspoglutide versus sitagliptin for type 2 diabetes mellitus (T-Emerge 4 Trial)

Diabetes Ther

2012

;

–

Pratley

Reusch

JE-B

Fleck

Wilson

Mekki

Efficacy and safety of the dipeptidyl peptidase-4 inhibitor alogliptin added to pioglitazone in patients with type 2 diabetes: a randomized, double-blind, placebo-controlled study

Curr Med Res Opin

2009

;

2361

–

2371

Pratley

Kipnes

Fleck

Wilson

Mekki

;

Alogliptin Study 007 Group

Efficacy and safety of the dipeptidyl peptidase-4 inhibitor alogliptin in patients with type 2 diabetes inadequately controlled by glyburide monotherapy

Diabetes Obes Metab

2009

;

167

–

176

Han

K-A

Chon

Chung

, et al.

Efficacy and safety of ipragliflozin as an add-on therapy to sitagliptin and metformin in Korean patients with inadequately controlled type 2 diabetes mellitus: a randomized controlled trial

Diabetes Obes Metab

2018

;

2408

–

2415

Kawamori

Haneda

Suzaki

, et al.

Empagliflozin as add-on to linagliptin in a fixed-dose combination in Japanese patients with type 2 diabetes: glycaemic efficacy and safety profile in a 52-week, randomized, placebo-controlled trial

Diabetes Obes Metab

2018

;

2200

–

2209

Kadowaki

Inagaki

Kondo

, et al.

Efficacy and safety of canagliflozin as add-on therapy to teneligliptin in Japanese patients with type 2 diabetes mellitus: results of a 24-week, randomized, double-blind, placebo-controlled trial

Diabetes Obes Metab

2017

;

874

–

882

Kashiwagi

Kazuta

Goto

Yoshida

Ueyama

Utsuno

Ipragliflozin in combination with metformin for the treatment of Japanese patients with type 2 diabetes: ILLUMINATE, a randomized, double-blind, placebo-controlled study

Diabetes Obes Metab

2015

;

304

–

308

Kaku

Watada

Iwamoto

, et al.;

Tofogliflozin 003 Study Group

Efficacy and safety of monotherapy with the novel sodium/glucose cotransporter-2 inhibitor tofogliflozin in Japanese patients with type 2 diabetes mellitus: a combined phase 2 and 3 randomized, placebo-controlled, double-blind, parallel-group comparative

Cardiovasc Diabetol

2014

;

–

Inagaki

Kondo

Yoshinari

Takahashi

Susta

Kuki

Efficacy and safety of canagliflozin monotherapy in Japanese patients with type 2 diabetes inadequately controlled with diet and exercise: a 24-week, randomized, double-blind, placebo-controlled, phase III study

Expert Opin Pharmacother

2014

;

1501

–

1515

Seino

Sasaki

Fukatsu

Ubukata

Sakai

Samukawa

Efficacy and safety of luseogliflozin as monotherapy in Japanese patients with type 2 diabetes mellitus: a randomized, double-blind, placebo-controlled, phase 3 study

Curr Med Res Opin

2014

;

1245

–

1255

Mathieu

Herrera Marmolejo

González González

, et al.

Efficacy and safety of triple therapy with dapagliflozin add-on to saxagliptin plus metformin over 52 weeks in patients with type 2 diabetes

Diabetes Obes Metab

2016

;

1134

–

1137

Rodbard

Seufert

Aggarwal

, et al.

Efficacy and safety of titrated canagliflozin in patients with type 2 diabetes mellitus inadequately controlled on metformin and sitagliptin

Diabetes Obes Metab

2016

;

812

–

819

Mathieu

Ranetti

, et al.

Randomized, double-blind, phase 3 trial of triple therapy with dapagliflozin add-on to saxagliptin plus metformin in type 2 diabetes

Diabetes Care

2015

;

2009

–

2017

Matthaei

Bowering

Rohwedder

Sugg

Parikh

Johnsson

;

Study 05 Group

Durability and tolerability of dapagliflozin over 52weeks as add-on to metformin and sulphonylurea in type 2 diabetes

Diabetes Obes Metab

2015

;

1075

–

1084

Jabbour

Hardy

Sugg

Parikh

Dapagliflozin is effective as add-on therapy to sitagliptin with or without metformin: a 24-week, multicenter, randomized, double-blind, placebo-controlled study

Diabetes Care

2014

;

740

–

750

Wilding

JPH

Charpentier

Hollander

, et al.

Efficacy and safety of canagliflozin in patients with type 2 diabetes mellitus inadequately controlled with metformin and sulphonylurea: a randomised trial

Int J Clin Pract

2013

;

1267

–

1282

Rosenstock

Seman

Jelaska

, et al.

Efficacy and safety of empagliflozin, a sodium glucose cotransporter 2 (SGLT2) inhibitor, as add-on to metformin in type 2 diabetes with mild hyperglycaemia

Diabetes Obes Metab

2013

;

1154

–

1160

Bode

Stenlöf

Sullivan

Fung

Usiskin

Efficacy and safety of canagliflozin treatment in older subjects with type 2 diabetes mellitus: a randomized trial

Hosp Pract (1995)

2013

;

–

Jabbour

Frías

Hardy

, et al.

Safety and efficacy of exenatide once weekly plus dapagliflozin once daily versus exenatide or dapagliflozin alone in patients with type 2 diabetes inadequately controlled with metformin monotherapy: 52-week results of the DURATION-8 randomized controlled trial

Diabetes Care

2018

;

2136

–

2146

Yu Pan

Han

Lui

, et al.

Lixisenatide treatment improves glycaemic control in Asian patients with type 2 diabetes mellitus inadequately controlled on metformin with or without sulfonylurea: a randomized, double-blind, placebo-controlled, 24-week trial (GetGoal-M-Asia)

Diabetes Metab Res Rev

2014

;

726

–

735

Seino

Min

Niemoeller

Takami

;

EFC10887 GETGOAL-L Asia Study Investigators

Randomized, double-blind, placebo-controlled trial of the once-daily GLP-1 receptor agonist lixisenatide in Asian patients with type 2 diabetes insufficiently controlled on basal insulin with or without a sulfonylurea (GetGoal-L-Asia)

Diabetes Obes Metab

2012

;

910

–

917

Kaku

Rasmussen

Clauson

Seino

Improved glycaemic control with minimal hypoglycaemia and no weight change with the once-daily human glucagon-like peptide-1 analogue liraglutide as add-on to sulphonylurea in Japanese patients with type 2 diabetes

Diabetes Obes Metab

2010

;

341

–

347

Lingvay

Desouza

Lalic

, et al.

A 26-week randomized controlled trial of semaglutide once daily versus liraglutide and placebo in patients with type 2 diabetes suboptimally controlled on diet and exercise with or without metformin

Diabetes Care

2018

;

1926

–

1937

Ludvik

Frías

Tinahones

, et al.

Dulaglutide as add-on therapy to SGLT2 inhibitors in patients with inadequately controlled type 2 diabetes (AWARD-10): a 24-week, randomised, double-blind, placebo-controlled trial

Lancet Diabetes Endocrinol

2018

;

370

–

381

Meneilly

Roy-Duval

Alawi

, et al.

GetGoal-O Trial Investigators. Lixisenatide therapy in older patients with type 2 diabetes inadequately controlled on their current antidiabetic treatment: the GetGoal-O randomized trial

Diabetes Care

2017

;

485

–

493

Dungan

Weitgasser

Perez Manghi

, et al.

A 24-week study to evaluate the efficacy and safety of once-weekly dulaglutide added on to glimepiride in type 2 diabetes (AWARD-8)

Diabetes Obes Metab

2016

;

475

–

482

Nauck

Stewart

Perkins

, et al.

Efficacy and safety of once-weekly GLP-1 receptor agonist albiglutide (HARMONY 2): 52 week primary endpoint results from a randomised, placebo-controlled trial in patients with type 2 diabetes mellitus inadequately controlled with diet and exercise

Diabetologia

2016

;

266

–

274

Reusch

Stewart

Perkins

, et al.

Efficacy and safety of once-weekly glucagon-like peptide 1 receptor agonist albiglutide (HARMONY 1 trial): 52-week primary endpoint results from a randomized, double-blind, placebo-controlled trial in patients with type 2 diabetes mellitus not controlled

Diabetes Obes Metab

2014

;

1257

–

1264

Wysham

Blevins

Arakaki

, et al.

Efficacy and safety of dulaglutide added onto pioglitazone and metformin versus exenatide in type 2 diabetes in a randomized controlled trial (AWARD-1)

Diabetes Care

2014

;

2159

–

2167

Bolli

Munteanu

Dotsenko

, et al.

Efficacy and safety of lixisenatide once daily vs. placebo in people with type 2 diabetes insufficiently controlled on metformin (GetGoal-F1)

Diabet Med

2014

;

176

–

184

Derosa

Franzetti

Querci

, et al.

Variation in inflammatory markers and glycemic parameters after 12 months of exenatide plus metformin treatment compared with metformin alone: a randomized placebo-controlled trial

Pharmacotherapy

2013

;

817

–

826

Pinget

Goldenberg

Niemoeller

Muehlen-Bartmer

Guo

Aronson

Efficacy and safety of lixisenatide once daily versus placebo in type 2 diabetes insufficiently controlled on pioglitazone (GetGoal-P)

Diabetes Obes Metab

2013

;

1000

–

1007

Derosa

Franzetti

Querci

, et al.

Exenatide plus metformin compared with metformin alone on β-cell function in patients with type 2 diabetes

Diabet Med

2012

;

1515

–

1523

Zinman

Gerich

Buse

, et al.;

LEAD-4 Study Investigators

Efficacy and safety of the human glucagon-like peptide-1 analog liraglutide in combination with metformin and thiazolidinedione in patients with type 2 diabetes (LEAD-4 Met+TZD)

Diabetes Care

2009

;

1224

–

1230

Nauck

Frid

Hermansen

, et al.;

LEAD-2 Study Group

Efficacy and safety comparison of liraglutide, glimepiride, and placebo, all in combination with metformin, in type 2 diabetes: the LEAD (Liraglutide Effect and Action in Diabetes)-2 study

Diabetes Care

2009

;

–

DeFronzo

Ratner

Han

Kim

Fineman

Baron

Effects of exenatide (exendin-4) on glycemic control and weight over 30 weeks in metformin-treated patients with type 2 diabetes

Diabetes Care

2005

;

1092

–

1100

Henry

Mudaliar

Kanitra

Woloschak

Balena

Efficacy and safety of taspoglutide in patients with type 2 diabetes inadequately controlled with metformin plus pioglitazone over 24 weeks: T-Emerge 3 trial

J Clin Endocrinol Metab

2012

;

2370

–

2379

Raz

Fonseca

Kipnes

, et al.

Efficacy and safety of taspoglutide monotherapy in drug-naive type 2 diabetic patients after 24 weeks of treatment: results of a randomized, double-blind, placebo-controlled phase 3 study (T-emerge 1)

Diabetes Care

2012

;

485

–

487

Hollander

Lasko

Barnett

, et al.

Effects of taspoglutide on glycemic control and body weight in obese patients with type 2 diabetes (T-emerge 7 study)

Obesity (Silver Spring)

2013

;

238

–

247

2020

6,928 Views

67 Web of Science

Efficacy of Modern Diabetes Treatments DPP-4i, SGLT-2i, and GLP-1RA in White and Asian Patients With Diabetes: A Systematic Review and Meta-analysis of Randomized Controlled Trials (original) (raw)

Introduction

Methods

Data Sources and Searches

Study Selection

Inclusion Criteria

Exclusion Criteria

Data Extraction

Quality Assessment

Data Synthesis and Analysis

Results

DPP-4i

Search Results and Study Characteristics

Quality of Included Studies and Publication Bias Assessments

Efficacy Outcomes

Exploratory Analysis

Meta-Regression

Sensitivity Analysis

SGLT-2i

Search Results and Study Characteristics

Quality of Included Studies and Publication Bias Assessments

Efficacy Outcomes

Exploratory Analysis

Meta-Regression

Sensitivity Analysis

GLP-1RA

Search Results and Study Characteristics

Quality of Included Studies and Publication Bias Assessments

Efficacy Outcomes

Exploratory Analysis

Meta-Regression

Sensitivity Analysis

Discussion

Article Information

References

Email alerts

See also