The effect of chamomile consumption on glycemic markers in humans and animals: a systematic review and meta-analysis

Abstract

Purpose

The use of natural and herbal products as alternative therapies, in conjunction with blood glucose-lowering medications, is on the rise for patients with diabetes. Our objective was to conduct a systematic review and comprehensive meta-analysis of both human and animal models to investigate the impact of chamomile consumption on glycemic control.

Methods

A systematic search was conducted on all published papers from January 1990 up to January 2022 via Scopus, PubMed/Medline, Google Scholar, and ISI Web of Science. Human and animal articles evaluating the effect of chamomile on serum glycemic markers were included. We used the random-effects model to establish the pooled effect size. The dose-dependent effect was also assessed.

Results

Overall, 4 clinical trials on human and 8 studies on animals met the inclusion criteria. With regard to RCTs, a favorable effect of chamomile consumption on serum fasting blood glucose (Standardized Mean Differences (SMD): -0.65, 95% CI: -1.00, -0.29, P < 0.001; I² = 0%) and hemoglobin A1C (HbA1C) levels (SMD: -0.90, 95% CI: -1.39, -0.40, P < 0.001; I² = 45.4%) was observed. Considering animal studies, consumption of chamomile extracts significantly reduced serum blood glucose (SMD: -4.37, 95% CI: -5.76, -2.98, P < 0.001; I² = 61.2%). Moreover, each 100 mg/d increase in chamomile extract intervention resulted in a significantly declined blood glucose concentrations (MD: -54.35; 95% CI: -79.77, -28.93, P < 0.001; I² = 94.8).

Conclusion

The current meta-analysis revealed that chamomile consumption could exert favorable effects on serum blood glucose and HbA1C. However, additional randomized controlled trials are needed to further confirm these findings.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40200-023-01345-8.

Introduction

Diabetes (DM) is a serious worldwide health problem in recent decades, considered the ninth important cause of death, which approximately 463 million adults being affected in year 2019 [1–3]. About 90% of patients with DM are type 2 DM (T2DM) [4], a chronic and multifactorial metabolic disorder characterized by persistent hyperglycemia due to insulin resistance in target tissues and/or dysfunction of pancreatic β-cells [5]. Prolonged exposure to chronic hyperglycemia leads to micro and macrovascular complications including nephropathy, retinopathy, and neuropathy (microvascular) and cerebrovascular disease, ischemic heart disease and causes tissue damage in one-half of patients with T2DM [6, 7]. Therefore, diabetes management can prevent or reduce these complications.

The main strategies for managing T2DM include drug therapies, lifestyle modification, and dietary changes [8, 9]. Apart from these main strategies, the use of supplements, alternative therapies and herbal medicines for the prevention and management of DM complications has received much attention recently [10]. Chamomile (Matricaria chamomilla L.) is a well-known and widely used medicinal plant that belongs to the Asteraceae family [11]. Various beneficial effects for chamomile such as anti-microbial, antioxidant [11], anti-inflammatory [12] antibacterial [13], anti- cancer [14] and sedative properties have been reported. The principal ingredient of chamomile are coumarins, polystyrene, and flavonoids, which play key roles in various biological activities [14]. The protective effects of this plant against hyperglycemia, insulin resistance, dyslipidemia and their complications have been investigated in animal and human studies [15]. For example, a study by Khan et al. on animal models indicated consuming chamomile for two months in alloxan-induced diabetic rats could significantly decrease serum concentrations of postprandial glucose, fasting blood glucose (FBG), and hemoglobin A1C (Hb1AC) [16]. Another study showed that aqueous extract of Chamaemelum nobile had a decreasing effect on blood glucose levels, but had no significant effect on fasting insulin levels in streptozotocin-induced diabetic rats [17]. A number of studies in the human population have been conducted on this issue. For instance, Rafraf et al. showed that consuming chamomile tea for 8 weeks had favorable effects on serum levels of HbA1C and insulin in T2DM patients [18]. To date, we are aware of no systematic review and meta-analysis has been performed to summarize the effect of chamomile extract on glycemic control on diabetic human or animal models. Therefore, we attempted to conduct a comprehensive systematic review and meta-analysis on human and animal models to assess the effect of chamomile consumption on glycemic control including FBG, fasting insulin, HbA1c, and homeostatic model assessment of insulin resistance (HOMA-IR).

Methods

The current meta-analysis addressing the effects of chamomile on glycemic control in patient with T2DM, was performed based on the preferred reporting items for systematic review and meta-analysis (PRISMA) guideline Supplemental Table 4 [19]. We registered the protocol of this meta-analysis in the center for Open Science Framework (OSF) database (https://www.osf.io, ID: 10.17605/OSF.IO/YGWFT).

Search strategy

We searched all published papers from January 1990 up to January 2022 through ISI Web of Science, Scopus, PubMed/Medline, and Google Scholar with no restriction on language. For this purpose, the following keywords including MESH and non-MESH terms were used: “chamomile” AND “diabetes mellitus” or “type 2 diabetes” or “type II diabetes” or “diabetic” or “diabetes” or “T2DM” or “noninsulin-dependent diabetes mellitus” or “NIDDM” or “hyperglycemia” or “diabetic” or “FBS” or “fasting blood sugar” or “hyperglycemic” or “glycemic outcomes” or” fasting blood glucose” or “HOMA-IR” or “glycemic”. The reference list of all eligible clinical trials was reviewed. In addition, manual searches of reference lists of related previous review papers were also performed to identify other eligible studies. To find new papers that might related to our systematic search, we have activated PubMed’s e-mail alert service.

Inclusion criteria

Original articles were included in this systematic review if they met the following inclusion criteria: (1) Publications with either parallel or cross-over randomized clinical trials (RCT) design; (2) studies carried out on human or animal models; (3) investigating the impact of chamomile on glycemic control, and (4) reporting adequate data on baseline and final trials of fasting blood glucose or/and serum blood glucose or/and HBA₁C in both chamomile and control groups. If multiple papers with the same population were identified, only the most comprehensive findings were included.

Exclusion criteria

Publications were excluded if they: (1) did not have any control group; (2) published without sufficient information on the outcomes in chamomile and control groups; for example, (a) not reporting the intervention duration; (b) taking other supplements besides chamomile. Cohort studies, case series, cross-sectional studies, letters, short communications, comments, case reports, conference abstracts, reviews, and meta-analyses were omitted as well.

Data extraction

Two independent investigators (AB and PA) carried out the study selection and extraction whereas a third investigator (MMA) was also present to resolve any discrepancies. The pertinent data was extracted from each eligible paper and listed in Table 1: 1) study characteristics: first authors’ last name, publication year, country, study design, intervention and control group size; 2) human participant characteristics: mean age, weight, height, gender, BMI; 3) animal characteristics: mean age, weight, source, species, strain, gender, method used to induce diabetes; 4) intervention characteristics: chamomile’ dosage, type of control, intervention duration; 5) Outcome characterizes: means and standard deviations of chamomile and control groups in baseline and end of intervention and list of confounders controlled for.

Table 1.

Characteristics of included RCTs in the meta-analysis

Study, year	Country	Design	Blinding	Age (mean) years	Sex	Participants	Number of I/C	Intervention	Control	Duration (weeks)	Outcomes
Heidari et al. 2017 [32]	Iran	RCT, Parallel	No	I = 32.66  C = 36.20	F	Overweight/Obese with impaired glucose intolerance	9/10	Brewed chamomile contains 1500 mg chamomile and exercise	Exercise	8	FBG, insulin and HOMA-IR
Kaseb et al. 2018 [30]	Iran	RCT, Parallel	Single	I = 55.33  C = 55.22	M/F	T2DM	22/22	Standard treatment with orally 200 ml/day of brewed chamomile (10 g/100 mL boiling water) twice a day before meals (lunch and dinner)	Standard therapy	4	FBG
Kermanian et al. 2018 [31]	Iran	RCT, Parallel	No	I = 51.95  C = 52.30	M/F	T2DM with depression	32/32	3 cups of chamomile tea daily containing 2500 mg chamomile	3 cups of black tea daily half an hour after meals	12	HbA1c
Rafraf et al. 2014 [18]	Iran	RCT, Parallel	Single	I = 50.19  C = 51.9	M/F	T2DM	32/32	Chamomile tea (3 g/150 ml hot water) three times per day	A water regimen	8	FBG, insulin, HbA1c and HOMA-IR

I, intervention; C, control; RCT, randomized controlled trial; T2DM, type 2 diabetes mellitus; M, male; F, female; FBG, fasting blood glucose; HOMA-IR, homeostatic model assessment for insulin resistance; Hb1c, glycated hemoglobin

Quality assessment of studies

In order to appraise risk of bias for human RCTs, the principles proposed by the Cochrane Cooperation’s tool was applied [20]. It comprises the following seven fields: (1) random sequence generation, (2) allocation concealment, (3) reporting bias, (4) performance bias, (5) detection bias, (6) attrition bias, and (7) other sources of bias. Studies were appraised to have a high or low risk of bias, regarding to the guideline proposed by the Cochrane Handbook. Moreover, to evaluate risk of bias in animal studies, SYRCLE’s tool was applied [21]. There was a total of ten biases in this evaluation tool, which included baseline characteristics, random sequence generation, random housing, allocation concealment, random outcome assessment, blinding of participants and personnel, blinding of outcome assessment, selective reporting, incomplete outcome data, and other bias.

Statistical analysis

Mean change and standard deviation (SD) of the outcomes were used to obtain the overall effect size. If the mean changes were not reported, they were calculated by considering the changes in each outcome throughout the study. When a study provided standard error (SE), SD was calculated by using this equation: (SD = SE × square root [number of participants]). Studies which reported medians and interquartile ranges, the median was considered as the mean, and SDs were calculated by dividing interquartile ranges by 1.35 [22]. Furthermore, we used the WebPlotDigitizer version 4.4 to extract data from studies that reported outcomes in the graphical form [22]. We obtained the overall effect size by using a random-effects model, which takes between-study variation into account [23]. The degree of heterogeneity across selected studies was calculated by I² (> 50% as significant heterogeneity) and Cochrane’s statistics (P _{heterogeneity} > 0.10).

It should be noted that we performed meta-analysis for the effect of chamomile on fasting insulin, FBG, HOMA-IR, and HbA1c on human and animal studies. Standardized mean differences were applied for human and animal studies. Then, a series of pre-defined subgroup analyses on animal studies were conducted to find inter-study heterogeneity for animal studies based on diabetic animal model (streptozotocin / alloxan), dosage of intervention (≤ 100 mg/kg / >100 mg/kg), duration (≥ 3 weeks / < 3 weeks). Sensitivity analysis was applied find the influence of each study on the overall effect size. Publication bias was tested using Egger’s test, Begg’s test, as well as by visually inspecting funnel plots.

Considering the dose-response analysis, the method provided by Crippa and Orsini [24] was used to calculate MD and SD of change in FBG ratio for each 100 mg/d addition in chamomile intake. Then, the specific effect sizes were pooled by a random-effects model. Finally, a nonlinear dose-response meta-analysis was performed to identify the shape of the effect of different dosage of chamomile on FBG [24]. Statistical analyses were performed using STATA software version 16. A two-tailed P value < 0.05 was considered significant.

Results

The number of papers identified via databases and reference lists was 378. Of them, 282 articles remained for more examination after removing duplicates. After the titles and abstracts screening, 261 papers were excluded. Then, 21 papers were fully reviewed for eligibility. Of them, we excluded 8 studies that did not relevant to our meta-analysis. Moreover, we included one [18] out of two studies [18, 25] that were conducted on the same population and data. Finally, 12 articles were included for the present meta-analysis [17, 26–28, 16, 18, 29–34] (Supplemental Fig. 1).

Study characteristics

Out of 12 studies, 4 RCTs (n = 191 participants) provided information for outcomes including fasting insulin, FBG, HOMA-IR, and HbA1c [18, 30–32]. Human populations included adults with type 2 diabetes [18, 30], depressed patients with diabetes [31], and overweight or obese adults with impaired glucose tolerance [32]. Three studies provided data for the change of FBG [18, 30, 32], two studies reported HBA1c [18, 31], and two studies measured HOMA-IR and fasting insulin [18, 32]. The dose of intervention ranged from 400 mg/day to 2500 mg/day and the duration of intervention ranged from 4 weeks to 8 weeks. All RCTs were conducted in Iran. Characteristics of eligible RCTs are provided in Table 1.

Based on Cochrane Cooperation’s tool two studies had high quality [18, 30] and one study had medium quality [31] and one is poor [32]. All four studies had a low risk of bias in random sequence generation. Two papers had a high risk of bias in allocation concealment [31, 32]. All studies had a low risk of bias in selective reporting. One had other sources [32] and two had a high risk of bias for blinding participants, outcome and personnel assessment [31, 32]. One had incomplete outcome data bias [30] (Supplemental Table 1).

Eight studies conducted on animal models including: 6 animal studies in rats (n = 80 rats) [16, 27–29, 34, 35], one study in mice (n = 12 mice) [26], one in rabbits (n = 10 rabbits) [33]. Among animal model studies, 8 papers reported FBG [35, 26–28, 16, 29, 33, 34], and 3 articles measured fasting insulin [29, 34, 35]. The dose of intervention ranged from 20 mg/kg to 500 mg/kg and the duration of intervention ranged from 14 days to 10 weeks. One study was conducted in England [28], two in Morocco [26], two in Egypt [29, 34], one in Turkey [27], one in Iran [33], and one in Pakistan [16]. Characteristics of eligible animal studies are provided in Table 2.

Table 2.

Characteristics of included animal studies in the meta-analysis

Study, year	Country	Animal model	Number of I/C	Dose (mg/kg)	Diet	Duration (weeks)	Outcomes
Ahmed et al. 2007 [26]	Morocco	Streptozotocin-induced diabetic mice	6/6	20	Food ad libitum	15 days	Glucose
Cemek et al. 2008 [27]	Turkey	Streptozotocin-induced diabetic rat	9/9	100	Chow and water	14 days	Glucose
Eddouks et al. 2005 [17]	Morocco	Streptozotocin-induced diabetic rat	6/6	20	Standard diet and water ad libitum	15 days	Glucose and insulin
Kato et al. 2008 [28]	England	Streptozotocin-induced diabetic rat	4/4	500	Basal 18% casein diet	21 days	Glucose
Rajagopalan et al. 2017 [29]	Egypt	Alloxan- induced diabetic rat	5/5	300	Standard diet and water ad libitum	3 weeks	Glucose and insulin
Saeed Khan et al. 2014 [16]	Pakistan	Alloxan monohydrate rat	10/10	100	Standard pellets diet and water ad libitum	53 days	Glucose
Saghahazrati et al. 2020 [33]	Iran	Streptozotocin-induced diabetic rabbit	5/5	100	Standard chow pellets and water ad libitum	21 days	Glucose
Soliman et al. 2020 [34]	Egypt	Streptozotocin-induced diabetic rat	6/6	400	Food ad libitum	10 weeks	Glucose and insulin

I, intervention; C, control I, intervention; C, control

Based on SYRCLE’s risk of bias tool, 6 out of 8 studies stated random sequence generation. Baseline characteristics of the outcomes were demonstrated in seven articles. All included papers did not state anything about blinding of participants and personnel, random housing, allocation concealment, and random outcome assessment. All studies stated complete outcome data and clarified other biases did not occur. The detailed results of SYRCLE’s risk of bias tool are shown in Supplemental Table 2.

Human studies

Three RCTs with 127 participants provided data for the meta-analysis of FBG. Pooled effects demonstrated that chamomile yielded a significant reduction in fasting blood glucose (SMD: -0.65, 95% CI: -1.00, -0.29, P < 0.001) and the between-study heterogeneity was low (I² = 0%, P = 0.806) (Fig. 1).

Fig. 1.

Forest plot for the effects of chamomile consumption on fasting blood glucose in human studies. Horizontal lines represent 95% CIs. Diamonds represent pooled estimates from random-effects analysis. SMD: Standardized mean difference, CI: confidence interval

In addition, the meta-analysis of two studies with 83 participants revealed a significant reduction in HBA1c in participants receiving chamomile compared with controls (SMD: -0.90, 95% CI: -1.39, -0.40, P < 0.001) and the between-study heterogeneity was moderate (I² = 45.4%, P = 0.176) (Fig. 2).

Fig. 2.

Forest plot for the effects of chamomile consumption on HbA1C in human studies

Meta-analysis of two studies with 128 Participants showed insignificant reduction of HOMA-IR (SMD: -1.06, 95% CI: -3.29, 1.17, P = 0.352) and fasting insulin (SMD: -1.88, 95% CI: -5.7, 1.94, P = 0.336) in chamomile group compared to control group (Supplemental Figs. 2 and 3, respectively). The between-study heterogeneity was high for HOMA-IR (I² = 94%, P < 0.001) and fasting glucose (I² = 97.4%, P < 0.001).

Sensitivity analysis revealed that the overall size effect regarding the effects of chamomile on FBG did not depend on a single study. No publication bias was observed by Begg and Egger’s test for FBG (P = 0.296 and P = 0.133, respectively).

Animal studies

Eight animal studies provided information for the analysis of blood glucose at baseline and after the intervention. Pooled effects of the included studies indicated that chamomile decreased significantly blood glucose (SMD: -4.37, 95% CI: -5.76, -2.98, P < 0.001) and the between-study heterogeneity was moderate (I² = 61.2%, P = 0.012) (Fig. 3).

Fig. 3.

Forest plot for the effects of chamomile extract on serum blood glucose in animal studies

We performed subgroup analysis for studies that administered chamomile extract. This analysis was done based on diabetic animal model (streptozotocin/alloxan), the dosage of intervention (≤ 100 mg/kg / >100 mg/kg), duration (≥ 3 weeks / < 3 weeks), following which we found that dose and duration of intervention were factors responsible for between-study heterogeneity (Supplemental Table 3).

In addition, the meta-analysis of three animal studies revealed an insignificant increasing in insulin level in animals receiving chamomile compared with controls (SMD: 0.44, 95% CI: -2.30, -3.17, P = 0.755) and the between-study heterogeneity was high (I² = 88.4%, P < 0.001) (Supplemental Fig. 4).

Sensitivity analysis showed that the overall size effect regarding the effects of chamomile on all examined outcomes did not depend on a single study. No publication bias was observed by Begg and Egger’s test for Glucose (P = 0.063 and P = 0.018, respectively) and insulin (P = 0.602 and P = 0.396, respectively).

Dose-dependent effect of chamomile intervention on serum blood glucose in animals

Nonlinear dose-response analysis effects of chamomile intervention on serum blood glucose in animals are indicated in Fig. 4.

Fig. 4.

Non-linear dose-response the effects of chamomile extract (mg/d) on serum blood glucose in animal studies. The 95% CI is demonstrated in the dotted lines

Serum blood glucose concentrations were significantly decreased with chamomile intake at the dosage of 50–500 mg/day. Following linear dose-response evaluation, we found that each additional 100 mg/d of chamomile intervention resulted in a significant declined blood glucose levels (MD: -54.35; 95% CI: -79.77, -28.93, P < 0.001; I² = 94.8) (Fig. 5).

Fig. 5.

Linear dose-response the effect of each 100 mg/d additional chamomile extract (mg/d) on serum blood glucose in animal studies

Discussion

The present systematic review and meta-analysis of RCTs and animal studies is the first study that critically evaluated the effects of supplementation with chamomile on blood glucose and indices of insulin resistance and glycemic control. This review demonstrated that consumption of chamomile extract significantly reduced the FBG and HbA1c compared with controls. Also, the meta-analysis of animal studies showed a significant reduction in blood glucose concentration. The dose-response analysis revealed for every 100 mg increase in chamomile, blood glucose decreases by -54.35 mg/dl.

Natural and herbal products are increasingly used as supplements in combination with blood glucose-lowering medications in patients with T2DM [36]. Blood glucose-lowering medications are used sparingly to manage T2DM due to some adverse effects [37]. Many studies have proven that supplementation with some natural and herbal products such as French maritime pine bark extract [38], propolis [39], cinnamon [40], and curcumin [41] may be improved T2DM complications. Chamomile is rich in polyphenols and flavonoids such as apigenin, luteolin, scultin, and quercetin, which have a role in the digestion and absorption of carbohydrates [42]. The effect of chamomile on glycaemic control has been proven in animal studies [26, 43, 44] and RCTs [18, 31]. So far, one systematic review [45] has been conducted in this field, the results of it are consistent with the present study and have shown a positive effect on glycemic control. It should be noted that they did not perform meta-analysis. Moreover, they had included 3 duplicated studies [18, 25, 46] that conducted on the same dataset. Given the systematically search and including animal and human studies in the current meta-analysis, our findings seem more accurate.

The proposed mechanism for the reducing effect of chamomile on blood glucose is related to increasing the capacity of insulin for glucose metabolism [43]. These mechanisms include increasing storage of glycogen in the hepatocyte, reducing blood glucose levels, and blocking of sorbitol in the human erythrocytes [28]. Furthermore, chamomile has a protective effect on β-cells of the pancreas against diabetes-induced oxidative stress [47]. Chamomile suppresses the release of oxidative stress indicators into the circulation in a dose-dependent manner [43]. Therefore, chamomile with antioxidant properties creates cellular protection, and on the other hand, reduces oxidative stress by controlling diabetes. One of the common treatments for insulin resistance and dyslipidemia is pharmacological intervention on the PPARs [48]. Chamomile tea reduces insulin resistance through PPAR family [49]. In human adipocytes cells, administration of chamomile extract resulted in expression of PPARγ target genes [49]. Luteolin, another ingredient of chamomile [50], increases insulin sensitivity and activates PPARγ transcription in adipocyte cells [51]. In addition, luteolin modulates GLUT-4 receptors in vitro, which can decrease insulin resistance. Luteolin also increased adiponectin gene expression [51], which can increase peripheral glucose utilization by activation of the AMPK enzyme and expression of the PPARγ gene in muscle and adipose tissue. Moreover, adiponectin suppresses gluconeogenesis [52, 53].

Chamomile is relatively safe [54]; however, along with the aforementioned health properties of it, some studies reported adverse effects such as infections, headache, dizziness, fatigue, diarrhea, taste lingering, drowsiness, nausea, indigestion, and musculoskeletal problems in long-term use [55, 56].

To the best of our knowledge, this is the first systematic review and meta-analysis on the effects of chamomile supplementation on glycemic control in both human and animal studies. Sub-group analysis was performed to find the sources between study heterogeneities. Moreover, the dose-response analysis is another strength of this review. However, the number of RCTs and sample size was low in this meta-analysis. Also, all the RCTs included in this review have been performed in the population of Iran. Therefore, further clinical trials in different ethnicity are needed to examine the duration and doses of chamomile supplementation and confirm these findings.

Conclusion

The result from the present meta-analysis suggests beneficial effects of the chamomile extract on blood glucose in animal studies. Moreover, the linear dose-response evaluation showed that each additional 100 mg/d of chamomile intervention resulted in a significantly declined blood glucose level. Similarly, such a significant effect was observed for serum FBG and HbA1c in human RCTs; however, no significant effect was seen for HOMA-IR, and insulin. Since the number of RCTs and sample size was low, further studies are still needed to support the results.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Authors’ contributions

Amir Bagheri, Sara Ebrahimi-Mousavi and Mahsa Malekahmadi designed and wrote the study. Amir Bagheri, Camellia Akhgarjand, and Narges Ghorbani Bavani performed the systematic search, data extraction, quality assessment, and statistical analysis. Mohammad Taghi Beigmohammadi and Jalal Moludi critically read the manuscript. All authors read and approved the final manuscript made and also made a significant scientific contribution to the research.

Funding

None.

Data Availability

The data that support the findings of this study will be made available by request to corresponding author.

Declarations

Ethics approval

This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. An ethics statement is not applicable because this study is based exclusively on published literature.

Conflict of Interest

The authors declare that there is no conflict of interest regarding the publication of this paper.