DOI: https://doi.org/10.21203/rs.3.rs-536402/v1
Background: Dipeptidyl peptidase-4 inhibitors (DPP-4i) provide a unique anti-hyperglycemic effect through regulating incretin peptides in type 2 diabetes mellitus (T2DM) patients that are inadequately controlled with insulin therapy. The aim of this study was to investigate the impact of DPP-4i on leptin concentrations in subjects with T2DM.
Methods: Randomized controlled trials (RCTs) with comparators were identified through systematically searching PubMed, Embase, and Cochrane library. Quantitative analysis was performed with a fixed or random-effects model according to heterogeneity. Publication bias was evaluated by using the standard methods for sensitivity analysis.
Results: Ten trials with 698 patients with T2DM were included. Pooled analysis demonstrated that DPP-4i did not significantly change leptin concentrations (1.31 ng/mL, 95% CI, -0.48 to 3.10). DPP-4i exerted no stronger effect on modulating leptin levels compared to active comparators (0.21 ng/mL, 95% CI, -1.37 to 1.78). Meta-analysis was powerful and stable after sensitivity analysis.
Conclusions: DDP-4i did not modulate leptin concentrations and exerted no stronger effect than traditional antidiabetic agents.
Type 2 diabetes mellitus (T2DM) is a common metabolic disease characterized by hyperglycemia and often accompanied by obesity. Reports indicate that obesity is a promoter of T2DM and obesity in childhood increases the risk of T2DM in adults[1]. An obesity forecast study based on the nonlinear regression model suggested that 51% of the population in the world will be obese by 2030[2]. In the obese state, excessive visceral fat accumulation causes adipose tissue dysfunctionality that contributes to the onset of obesity-related comorbidities[3]. Adipose tissue exhibits its function not only as storing energy but also synthesizes and secretes adipocytokines. These adipokines play distinct roles in physiological and pathophysiological conditions. Among them, leptin is an adipokine mainly secreted from adipose tissue, serving as an afferent signal in maintaining homeostasis of adipose tissue mass[4]. Data have suggested that dysregulated leptin is usually associated with metabolic disease including obesity and T2DM[5]. Apart from obesity, hyperleptinemia is associated with insulin resistance and hypertension. On the other hand, leptin could independently reduce blood glucose levels, particularly in hyperglycemic models of insulin deficiency[6]. Currently, a number of studies have shown that different antidiabetic agents modulate serum leptin concentrations in physiological and pathophysiological conditions[7–9].
Novel glucose-lowering drugs including sodium-glucose cotransporter 2 (SGLT2) inhibitors, glucagon-like peptide 1 (GLP-1) receptor agonists, and DPP-4 inhibitors have become available. These agents provide protective effects via reducing blood glucose levels and improving insulin resistance[10, 11]. The effective glucose control from these new agents significantly improves long-term microvascular and macrovascular complications in patients with T2DM[10, 12]. Among them, dipeptidyl peptidase 4 inhibitors exert their effect on lowering blood glucose by inhibiting the inactivation of GLP-1[13]. The efficacy and safety of DPP-4 inhibitors have been proved by several randomized controlled trials, demonstrating improved glucose control with a low risk of hypoglycemia[14]. It remains unknown whether DPP-4 inhibitors could modulate leptin and to what extent compared to other antidiabetic agents. Based on the increased incidence between obesity and diabetes, leptin might be recognized as a surrogate cardiovascular biomarker for antidiabetic medications. Therefore, the current study aimed to help demonstrate the impact of DPP-4i on leptin levels in T2DM.
We searched PubMed, Embase, and the Cochrane Library for randomized controlled trials (RCTs) published in English from inception until 30 March 2021. Key terms were used for “sitagliptin” OR “vildagliptin” OR “teneligliptin” OR “saxagliptin” OR “linagliptin” OR “alogliptin”. Two authors independently performed the literature search.
All RCTs lasting at least 4 weeks and reporting data on leptin outcome were included. A study was identified if it was a randomized controlled study comparing DPP-4i with placebo or other antidiabetic agents, if it reported leptin levels with DPP-4i treatment, if it was conducted in patients with T2DM. A study was excluded if it was conducted in healthy participants, it was non-human designed, if it was a narrative review, or only an abstract paper. Reference lists of eligible studies as well as systematic review and meta-analysis of DPP-4i were hand-searched for additional relevant studies. Corresponding authors were contacted if missing information was relevant. Inclusion and exclusion criteria were evaluated objectively by two reviewers.
Two authors independently extracted data by using standardized predefined forms on: first author name, year of article publication, country origin, sample size, gender distribution, body mass index, mean age, diabetes duration, DPP-4 inhibitor(s), and comparator(s), therapy duration, baseline HbA1c, and serum leptin concentrations. The primary outcome measure was a change in the leptin concentrations. When studies reported leptin data for different treatment durations, the longest was used.
Overall, the risk of bias for the domains included in the Cochrane risk assessment tool was judged to be low. There was an unclear risk of bias in some items, including allocation concealment, blinding of the outcome, and participants. Two studies had detection bias based on blinding of outcome assessment, and three studies had performance bias due to the absence of implementation of blind methods. All the studies were randomly designed and details on the items of bias criteria across the studies were displayed (Fig. 2).
From 12,459 identified records we excluded non-human and observational studies, leaving 10 for full-text assessment. After systematic selection (Fig. 1), ten RCTs fulfilled the inclusion criteria (Table 1). RCTs published between 2015 and 2021 included 698 participants. Of these, 348 were treated with a DPP-4 inhibitor (175 with sitagliptin, 72 with vildagliptin, 21 with saxagliptin, 42 with alogliptin, and 38 with linagliptin), monotherapy or in addition to metformin or other antidiabetic agents, and 350 were treated with placebo or control therapy. Follow-up time ranged from 1 to 13 months. In the largest study, 241 subjects were included, while the smallest one recruited 20 subjects. Most patients received sitagliptin, while the remaining studies compared vildagliptin, saxagliptin, alogliptin, and linagliptin with placebo or traditional antidiabetic agents, respectively.
| Study, year | Location | Treatment arm (n) | HbA1c (%) | Male (n) | Age (years) | BMI (kg/m2) | Diabetes duration (years) | Treatment duration (months) | leptin (ng/mL) |
|---|---|---|---|---|---|---|---|---|---|
| Takeshita, 2015a[37] | Japan | Sitagliptin:28 mitiglinide:29 | 6.7 ± 0.6 6.9 ± 0.8 | 18 19 | 61.0 ± 13.8 65.8 ± 9.7 | 24.5 ± 3.8 24.2 ± 4.6 | 86.4 ± 90 145.2 ± 122.4 | 4 | 8.7 ± 6.5 10.5 ± 13.4 |
| Takeshita, 2015b[38] | Japan | Vildagliptin:53 liraglutide:49 | 8.1 ± 1.2 8.0 ± 0.9 | 36 35 | 64.5 ± 12.7 64.9 ± 1.9 | 24.5 ± 4.6 25.4 ± 4.8 | NS | 3 | 8.1 ± 6.9 6.9 ± 5.7 |
| Kato,2015[39] | Japan | Sitagliptin:10 glimepiride:10 | 7.2 ± 0.2 7.3 ± 0.2 | 6 5 | 62 ± 4.7 55 ± 6.7 | 25.6 ± 2.6 26.6 ± 2.5 | NS | 6 | 12.6 ± 2.3 10.3 ± 3.0 |
| Matsushima,2016[40] | Japan | Sitagliptin: 120 voglibose:121 | 7.9 ± 1.0 7.8 ± 0.8 | 72 71 | 63.2 ± 13.8 63.2 ± 11.6 | 25.0 ± 4.5 25.1 ± 4.5 | NS | 3 | 8.3 ± 6.9 9.0 ± 9.3 |
| Dore,2018[41] | American | Saxagliptin:21 Placebo:21 | 7.0 ± 0.8 6.6 ± 0.5 | 10 14 | 58.3 ± 5.7 56.4 ± 8.5 | 32.3 ± 4.2 31.5 ± 4.8 | ༜120 | 3 | 19.4 ± 3.7 14.1 ± 2.1 |
| Takihata,2019[42] | Japan | Sitagliptin:17 luseogliflozin:17 | 10.0 ± 1.4 10.4 ± 1.0 | 14 15 | 52.8 ± 15.5 52.1 ± 15.3 | 26.8 ± 5.1 26.4 ± 4.8 | NS | 3 | 9.1 ± 6.7 7.2 ± 4.7 |
| Takeshita,2019[43] | Japan | Alogliptin:42 metfomin:42 | 7.5 ± 1.0 7.4 ± 1.2 | 29 29 | 63.8 ± 10.5 63.1 ± 13.1 | 25.4 ± 6.1 24.4 ± 4.0 | 122.4 ± 124.8 169.2 ± 156 | 3 | 11.2 ± 12.8 8.4 ± 10.7 |
| Schiapaccassa,2019[44] | Brazil | Vildagliptin:19 Metformin:19 | 8.0 ± 1.8 7.9 ± 2.0 | 0 0 | 39.1 ± 5.3 39.8 ± 7.7 | 36.0 ± 4.0 38.5 ± 6.1 | NS | 1 | 21.9 ± 19.4 25.4 ± 13.3 |
| Awal,2020[45] | American | Linagliptin:14 Placebo:17 | 7.1 ± 0.7 7.4 ± 1.0 | 11 7 | 61.0 ± 5.0 63.0 ± 6.0 | 31.2 ± 4.4 30.6 ± 2.9 | ≤ 180 | 3 | 21.7 ± 22.8 22.5 ± 12.6 |
| Komorizono,2020[46] | Japan | Linagliptin:24 Metformin:25 | 7.0 ± 0.5 7.2 ± 0.8 | 10 9 | 49.4 ± 10.8 55.6 ± 10.2 | 29.7 ± 4.9 27.9 ± 4.1 | NS | 13 | 17.7 ± 9.3 18.5 ± 8.2 |
Data on leptin were available from all RCTs. The pooled analysis of the effect of DPP-4i on leptin concentrations was 1.31 ng/mL (95% CI, -0.48 to 3.10, P = 0.95, I2 = 0%) compared to placebo, and 0.21ng/mL (95% CI, -1.36 to 1.78, P = 0.16, I2 = 33%) compared to traditional antidiabetic agents (Fig. 3). The pooled estimate of the modulating effect of DPP-4i on leptin was credible in the leave-1-out sensitivity analysis (WMD 0.42ng/mL, 95% CI -0.54, 1.39, N = 10 studies, heterogeneity P = 0.39; Fig. 4). This confirmed that the effect across the studies is an overall effect of all the identified studies. In subgroup analysis, baseline HbA1c, leptin, BMI, length of follow-up, or age parameters did not influence the impact of DPP-4i on leptin (see supplementary Fig. 1–5). No publication bias was suggested by Begg’s test (P = 0.93) and Egger’s test (P = 0.95) across the 10 studies (Fig. 5).
The present study aimed to evaluate the effect of DPP-4 inhibitors on serum levels of leptin in T2DM. Leptin is associated with metabolism, insulin sensitivity, and diabetes. This meta-analysis demonstrated that DPP-4 inhibitors exerted no significant effect on changing circulating leptin levels in T2DM patients compared to placebo or active drugs.
Leptin is a critical metabolic hormone that plays a key role in regulating the physiologic switch between the fed and starved states. Since its discovery in 1994, leptin provides deep insights into the regulation of central nervous system energy balance circuits[15]. It is a protein containing 167 amino acid mainly secreted by white adipose tissue into the blood and can be transported across the blood-brain barrier[16]. This adipokine regulates metabolic homeostasis by inhibiting food intake and increasing energy expenditure. Leptin significantly reduced blood glucose in mouse models of insulin-deficient diabetes, suggesting that leptin modulated glucose homeostasis independently of insulin[17, 18]. Although its exact mechanism of lowering glucose levels remains unknown, data have shown that leptin decreases appetite, suppresses insulin secretion, and increases insulin sensitivity. Overall, the identification of leptin has provided a framework for studying the pathogenesis of obesity in the diabetic population. Decreases in the sensitivity to leptin might contribute to the development of T2DM[19].
Leptin therapy has been found to effectively reverse hyperglycemia and prevent mortality in mouse models of diabetes[20]. The pathogenesis of obesity is analogous to diabetes and can result from either leptin hyposecretion or leptin resistance. The former type of obesity is characterized as low endogenous plasma leptin levels who respond to leptin therapy, while the latter form describes most obese subjects, who are leptin resistant but might respond to leptin therapy in combination with other drugs such as leptin sensitizers[21]. In states of leptin resistance such as obesity and T2DM, leptin action is decreased in the brain parenchyma and vessels, despite its elevated concentrations in the plasma[18]. In patients with T2DM, elevated levels of leptin are often linked with increased cardiovascular risk, as well as with the presence of macro- and microvascular complications. Treatment of diabetes in human beings might benefit from correction of leptin resistance as well as insulin resistance[20]. In patients with severe coronary artery disease, abdominal obesity is commonly related to increased leptin levels and decreased adiponectin concentrations. Leptin/adiponectin imbalance might mediate the increased risk of developing T2DM and cardiovascular disease associated with abdominal obesity[22]. Our team previously found that DPP-4 inhibitors increased serum adiponectin levels in T2DM[23]. In the current study, we demonstrated that DPP-4 inhibitors did not significantly change serum leptin concentrations, suggesting that these drugs provided a neutral effect without aggravating leptin resistance in the diabetic state.
Although the cardiovascular safety of DPP-4 inhibitors has been proven in T2DM, the net effect of these drugs on leptin concentrations in obesity-related disease remains unclear[24]. In Kitamura’s study, leptin sensitivity was enhanced after anagliptin treatment in high-fat diet fed mice[25]. In the study evaluating the inhibitory effect of vildagliptin on fibrosis markers on white adipose tissue of high-fat diet-induced obese mice, vildagliptin prevents the increase of fibrosis markers and reduces leptin levels[26]. The effect of sitagliptin on reducing BMI and the occurrence of hypoglycemia in obese patients with insulin treatment-induced diabetes mellitus might be correlated with decreased leptin levels and increased adiponectin levels[27]. Add-on therapy with anagliptin in Japanese T2DM patients treated with metformin for 52 weeks also reduced leptin concentrations[28]. Our study included most studies with relatively shorter treatment durations lasting from 1 month to 6 months, with only 1 study lasting for 13 months. Further studies with a longer duration and a larger number of participants will be needed to illuminate the effect of DPP-4 inhibitors on leptin concentrations and leptin sensitivity.
T2DM is associated with metabolic dysregulation and chronic inflammation. Data emerged from the research of leptin in diabetes has been discussed as an inflammatory mediator sustaining multifactorial diseases[29]. Leptin induces tumor necrosis factor-α (TNF-α)-dependent inflammation in acquired generalized lipodystrophy disease[30]. Statin[31] and antidiabetic agents[32] including sitagliptin, metformin, pioglitazone, liraglutide, and empagliflozin exhibit certain effects on inflammation. Sitagliptin ameliorated diet-induced metabolic syndrome and fatty liver via regulation of adipose tissue inflammation and hepatic adiponectin/ leptin levels[33]. Another study also proved that the novel DPP-4 inhibitor teneligliptin prevents high-fat diet-induced obesity accompanied by increased energy expenditure in mice[34]. DPP-4 inhibitor anagliptin exerts anti-inflammatory effects on macrophages, adipocytes, and mouse livers by suppressing NF-kB activation[35]. We also found that the inflammatory marker of C-reactive protein was effectively reduced after DPP-4 inhibition[36]. Further data should be reviewed regarding the role of leptin in inflammation, and the role of inflammation on the development of leptin resistance and obesity.
Although there were some reports investigating the effect of antidiabetic agents including DPP-4 inhibitors, there were no confirming answers to draw on whether DPP4 inhibitors could modulate leptin. We could not get further information for other comparisons for modulating leptin levels in T2DM. In the absence of comparative evidence between DPP-4i and other anti-diabetic medications, this meta-analysis added detailed illustration on adipokine of leptin levels. In this pooled analysis, comparisons of DPP-4i therapy and other treatment for type 2 diabetes (with 10 included trials) were performed, providing evidence that DPP-4i treatment was not significantly associated with changing leptin levels in participants from different regions in comparison with placebo. This effect of DPP-4i on leptin levels had not been changed by potential variables of treatment duration, age, and baseline HbA1c.
This study is the first meta-analysis demonstrating the effect of DDP-4i on serum leptin concentrations in T2DM. It suggested that DPP-4i did not exert an effect on leptin resistance in T2DM in patients with diabetes and obesity-associated cardiovascular diseases. It also provides insights into the therapeutic implications of obesity-related atherosclerotic disease in humans for the potentially protective effects on leptin sensitivity. Secondly, the pooled results suggest that leptin potentially serves as an effective cardiovascular biomarker in T2DM. Thirdly, subgroup analysis has been performed to explore the effect of therapy duration, diabetes duration, dosage, and age.
This study also has some limitations needed to be disclosed. Firstly, only literatures published in English were searched, which could inevitably generate publication bias and unstable estimates of treatment effects. Secondly, the pooled analysis should be interpreted with consideration for the moderate heterogeneity across identified studies, although measures had been taken to alleviate it by performing the sensitivity analysis and subgroup analysis. Thirdly, the follow-up periods were relatively short, and evaluating the long-term effect of DPP-4i treatment is necessary.
DPP-4i exerted no off-target effect on modulating leptin concentrations in patients with T2DM. Data on long-term effects are needed to perform in patients with T2DM with risks of obesity and cardiovascular disease.
Ethics approval and consent to participate
Not applicable
Availability of data and materials
All data generated or analyzed during this study are included in this published article [and its supplementary information files].
Competing interests
The authors declare that they have no competing interests.
Author contributions
X.W. and Y.B. participated in the design, conduct, and collection of this meta-analysis. X.W., Y.B., Z.W. and X.H.Z. searched the literature, extracted the data, and evaluated the risk of bias. X.W. and Y.B. took part in writing the manuscript. X.L., and Z.N.J. have full access to all the data in the study and take full responsibility for the integrity of the data analysis. All authors read and approved the final manuscript.
Funding
This work was supported by grants from the Key Specialty of Beijing.
Consent for publication
If the manuscript is accepted, we approve it for publication in Diabetology & Metabolic Syndrome.
Acknowledgements
Not applicable.