The effects of dietary/herbal supplements and the serum levels of micronutrients on the healing of diabetic foot ulcers in animal and human models: a systematic review

Abstract

Purpose

Diabetic Foot Ulcer (DFU) is one of the common and serious complications in patients with Diabetes Mellitus (DM) worldwide. Given the considerable tendency of patients suffering from DFU to use the complementary therapies, the objectives of this study were to: (i) summarize the effects of dietary and herbal supplements on DFU characteristics and metabolic parameters in both animal models and clinical trials, and (ii) evaluate any links between the serum levels of micronutrients and DFU in observational studies.

Methods

A systematic search in five electronic databases including PubMed/Medline, Scopus, Web of Science, Embase, and Cochrane Library was conducted to find relevant English language published from 1990 until 31 December 2018.

Results

Of a total of 8603 studies, 30 eligible papers including animal studies (n = 15), clinical trials (n = 7), and observational works (n = 8) were included in the systematic review. We found that some dietary/herbal supplements and micronutrients had positive effects on the wound healing. However, limited evidence is existed. Also, lower serum levels of vitamin D, C, vitamin E, and selenium in patients with DFU were likely to increase the risk of DFU, leading to impaired wound healing.

Conclusion

Findings suggested that some dietary and herbal supplements such as Vitamin D, Magnesium, Vitamin E, Probiotic, Zinc, and Pycnogenol would be effective on wound healing of DFUs. However, further high-quality randomized controlled clinical trials and prospective cohort studies are needed to clarify the roles of micronutrients and other dietary and herbal supplements on the progress and treatment of DFU.

Introduction

Diabetic foot is a common and serious complication in patients with diabetes mellitus (DM) worldwide [1]. According to the World Health Organization, Diabetic foot ulcer (DFU) is featured as the foot with ulceration, infection and/or destruction of deep tissues, accompanied by neurological dysfunction and various degrees of peripheral vascular problems in the lower limb of patients with DM [2]. Every year, more than 1 million patients with diabetes lose at least a part of their leg due to the DM complications. In other words, every 20 seconds a lower limb is lost because of DM somewhere in the world [3].

DFU imposes a considerable economic and psychological burden on patients, their families, and society [4–6]. Some previous studies have examined various methods to define the risk of the development and the progression of DFU [7–10].

A wide spectrum of adjuvant therapies including oxygen therapies, negative pressure wound therapy, energy-based therapies and oral and topical medications are applied along with standard treatment to accelerate wound healing in patients with DFU [11–14]. However, it is essential to find and develop more effective methods such as complementary medicine for healing wound. As oxidative stress, inflammation, and insulin resistance involve in the pathogenesis of DFU [15–17], it is expected that methods to improve such parameters can achieve the therapeutic goals. On the other hand, wound healing is complicated cellular and biochemical processes depending on nutritional substrates such as energy, protein and micronutrients and can affect enzyme activities contributing to the healing process [18]. Therefore, malnutrition, mal-absorption and insufficient micronutrients can have negative effects on this process. Some studies investigating the association between serum levels of micronutrients [19, 20] and DFU showed that there is an inverse link between low micronutrient concentrations and high risk of DFU [21–24]. Furthermore, the effects of a number of dietary supplements have been evaluated in both animal models [25–29] and clinical trials [30–35] which some of them reported positive effects [25, 26, 30, 31].

Several works have also examined the effects of the medicinal herbs as complementary therapies on diabetic models [25, 36–40]. Some studies reported positive effects [25, 36, 38], whereas the findings were inconsistent. According to the above-mentioned evidence, taking some dietary and herbal supplements along with standard therapy can be helpful. In a meta-analysis performed by Ye et al., on all types of chronic wounds, they found that dietary supplements cannot significantly improve DFU [41]. However, they did not cover publications on neither medical herbs nor the serum levels of micronutrients. In a systematic review by Haiyan et al., positive effects of some dietary supplements on DFU were shown. However, it covered publications up to 2011 and the quality of studies was not examined [42]. Shuo et al., also reported that traditional Chinese medicinal herbs can be effective for healing of DFU. However, they only focus on herbs and dietary supplements were not considered [43]. To the best of our knowledge, no systematic reviews have examined the effects of both dietary supplements and medicinal herbs on DFU. Therefore, the objectives of this study were to (i) summarize the effects of dietary and herbal supplements on DFU characteristics and metabolic parameters in both animal models and clinical trials, and (ii) evaluate any links between the serum levels of micronutrients and DFU in observational studies.

Materials and methods

Search strategy and selection studies

The present study was designed and reported based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline [44].

A systematic search in the five electronic databases such as PubMed/Medline, Scopus, Web of Science, Embase, and Cochrane Library was conducted to find relevant English papers published from 1990 until 31 December 2018. Both MeSH and non-MeSH key terms were considered in syntax. Search Strategies were designed by a librarian (R.A) and adopted designed strategy was used for each database. The study protocol was registered at the International Prospective Register of Systematic Reviews (PROSPERO): CRD42018116728.

Eligibility criteria

Papers were selected if they met the following criteria: (i) either clinical trials (parallel or crossover design), observational or animal studies, (ii) examining dietary or herbal supplements alone or in combination with other components, or (iii) examining the serum levels of nutrients, and (iv) reporting DFU as a primary outcome. Case report, case series, and grey literatures including interviews, conference papers and theses, unpublished data and topical treatments such as ointments were excluded from the study. To avoid missing any eligible articles, reference lists of all included papers and review studies were hand-searched.

After collecting publications and removing duplications, screening based on title and abstract were performed by two independent investigators (MR.A, M.A). Then, the full-texts of remaining publications from the previous step were evaluated based on the eligibility criteria. The full- texts of possible relevant papers were obtained from some cases via contacting their corresponding authors. Any disagreements between two investigators were resolved by the third investigator (N.N).

Data extraction

From each eligible paper, data were extracted using a pre-defined data extraction form. The primary exposure was dietary/ herbal supplements and the main outcome was DFU. The following data were extracted by two independent reviewers (M.S, N.N) for both human and animal studies (M.R.A, M.A) based on study design: the first author’s last name, publication year, study design, the participants’ characteristics (mean age, gender), sample size at baseline and at the end of study, name of intervention, dosage, duration of intervention, mode of intervention, ulcer characteristics and their changes, biochemical parameters, comparison groups, and controlled confounders. If a study reported the effects of both topical and oral treatments, only findings related to non-topical therapies were extracted. In the case of any discrepancy between the two investigators, the third investigator (E.N) was consulted.

Quality assessments

The quality of the included clinical trials was assessed using the Jadad scale (3 main items) [45]. The maximum score was five and each study obtained minimum three scores was considered as high quality. For observational studies, Newcastle-Ottawa Scale (NOS) was used (three main items) [46]. Based on this scale, a maximum of nine/ten scores can be dedicated to each study and a minimum score of five were considered as high quality. Two investigators (N.N, M.S) independently examined the quality of human studies and then compared their assessments with each other.

For animal studies, SYRCLE checklist (10 items) was used [47]. This tool is based on the Cochrane risk of bias checklist. It has been adjusted for various aspects of bias that play pivotal roles in animal interventions. Each item should be identified by YES/NO/ UNCLEAR. The quality of each animal study was checked by two independent reviewers (MR.A, M.A) and any disagreement was resolved by the third investigator (E.N).

Quantitative data synthesis

Given limited studies conducted on each dietary/herbal supplement and serum levels of micronutrients, quantitative synthesis was not possible and we only summarized the results qualitatively for all nutrients.

Results

Literature search

In total, 8603 publications including PubMed (n = 1049), Scopus (n = 3006), Web of Science (n = 1498), and Cochrane library (n = 379), Embase (n = 2671) were found. Of a total of 8603 studies, 3896 were duplicates. Screening was performed on 4707 publications. As depicted in Fig. 1, after the removal of duplicates and screening based on titles and abstracts, 150 publications were considered potentially relevant. With the precise assessment based on the full-texts of papers, publications were excluded due to either not being relevant or the lack of eligibility criteria. Finally, 30 eligible papers including animal studies (n = 15) [25–29, 37–40, 48–53], clinical trials (n = 7) [30, 34, 35, 54–57], observational works (n = 8) [23, 24, 58–62] were included in the systematic review.

Fig. 1.

Paper selection for the present systematic review

Characteristics of the included publications

Animal studies

A total of 15 animal studies conducted on dietary supplements (n = 6) [25, 26, 29, 48, 50, 52] and medicinal herbs (n = 9) were included in the study [25, 27, 28, 37–40, 50, 51]. Of these, 11 papers [26–28, 38–40, 49–53] assessed oral intake and the remaining studies evaluated the impact of intravenous injection of supplements. The effects of each supplement were examined in only one study; just supplementation with hesperidin was explored in two papers [27, 28].

The impacts of vitamin C, Zinc (Zn), vitamin D, vitamin E, amino acids (Argintinand proline), and mixture of nutrients were examined and six types of medicinal herbs (alone or mixture) including Angelica Dahurica, San-huang-sheng-fu, Aloe vera, Radix Astragali & Rehmanniae, and Chinese 2-herb formulas were examined. In addition, three effective components of natural products including hesperidin, naringin, and genistein were also assessed in animal models. All the aforementioned supplements except Zn [48] showed positive effects on wound healing and other metabolic parameters as shown in Table 1. Based on SYRCLE checklist, all animal studies had high risk of bias (Appendix 1).

Table 1.

Characteristics of the included animal studies in the systematic review

Author (year)	Type of animal/sample size	Comparative groups	Intervention	Dosage	Duration	Type of intervention (oral, injection, etc)	Findings	Quality score
SusamaChokesuwattanaskul (2018)	nude mice / 48	1-Control 2-Diabetes Mellitus [DM] 3-Diabetes Mellitustreated with mesenchymal stem cells [MSCs] 4-DM treated with VitC 5-DM treated with MSCs and VitC	Vitamin C	- Vit. C (1.5 g/l)	7 days	Oral	In the 4th Group: - ↑Angiogenesis -Accelerate diabetic wound healing	High RoB
Yusom Shin (2017)	Rat / 20	1-Non-diabetic group 2-A nontreated 3-A nerve-block 4-A nerve-block and IV zinc group	Zinc	0.1 ml zinc sulphate hydrate	10 days	IV	Among the three DFU groups: The wound-surface-area differences were not statistically significant.	High RoB
Xiao-na Zhang, (2017)	Rat / 45	1-Diabetic treated with ADEE 2-Diabetic without drugtreatment 3-Non-diabetic rats served as normal control	Angelica Dahurica Ethanolic Extract	1.1 ml/0.2 kg body weight once daily	14 days	Oral	-Correcting impaired Wound Healing -Accelerating wound closure -↑ newly formed blood vessels -Improving diabetes-impaired wound healing	High RoB
Wenbin Li 2018	Rat / 30	1-Non-diabetic and non-wounded animals 2-Non-diabetic and wounded animals 3-Diabetic and wounded animals	Hesperidin	25, 50 and 100 mg/kg,	21 days	Oral	-↑inpercent Wound closure and serum insulin level. -↑Lipid peroxidation (MDA) and Nitric Oxide levels. -↑mRNA expressions	High RoB
YiFeng Yuan et al. 2018	Mice /45	1-Normal group 2- Diabetic group 3-Vitamin D treatment group	Vitamin D	100 ng/ kg per day	14 days	IV	-↓Level tumor necrosis factor-α, interleukin 6,1β) -Suppressed NF-κB pathway activation -Improves impaired wound	High RoB
Li Wang, 2018	Rat / 90	1-Non-diabetic rats and nonwounded got sterile water for injection (WFI) 2-Non-diabetic rats and wounded received sterile WFI 3-Diabetic rats & wounded (DW) administered with sterile WFI 4-DW rats Administered with aqueous suspension of HSP (10 mg/kg). 5- DW rats Administered with aqueous suspension of HSP (20 mg/kg). 6- DW rats Administered with aqueous suspension of HSP (40 mg/kg). 7- DW rats Administered with aqueous suspension of HSP (60 mg/kg) 8- DW rats Administered with aqueous suspension of HSP (80 mg/kg) 9- DW rats Administered with insulin (10 IU/kg)	Hesperidin (HSP) Suspension And Insulin	HSP: 10, 20, 40, 60, and 80 mg/kg And Insulin 10 IU/kg	20 days	Hesperidin: Oral And Insulin: subcutaneous	↑ Expression of VEGF and stimulation of angiogenesis ↓ expression of inflammatory mediators ↑ Expression of growth factors statistically significant (P < 0.05) improvement in wound dimension	High RoB
Amit D. 2014	Rat /?	Group I: Normal non-diabetic (ND): without wound received double distilled water (10 mg/kg, p.o.) Group II: non-diabetic Normal wound control: single injection of citrate buffer (vehicle), and then double distilled water (10 mg/kg, p.o.) Group III: diabetic wound control: double distilled water (10 mg/kg, p.o.) Group IV: diabetic animals with wound received naringin (20 mg/kg, p.o.) in double distilled water. Group V: diabetic animals with wound received naringin (40 mg/kg, p.o.) in double distilled water Group VI: diabetic animals with wound received naringin (80 mg/kg, p.o.) in double distilled water Group VII: diabetic animals with wound received insulin (10 IU/kg, s.c.). Group VIII: TCDO: diabetic animals with wound received TCDO(1 drop, twice a day, topically).	Naringin And Insulin And tetra chlorodecaoxide	Naringin (20, 40 and 80 mg/kg, p.o.), Insulin (10 IU/ kg, s.c.) tetra chlorodecaoxide (TCDO) (1 drop, twice a day, topically)	16 days	Naringin: injection Insulin: injection TCDO topical	↓ wound area ↑ rate of wound contraction ↑Angiogenesis ↓ Inflammatory mediator expression ↑growth factor	High RoB
HyeyoonEo, 2016	Mice	1) CON: Nondiabetic mice were fed AIN-93G rodent diet without genistein supplementation. [2] DMC: Diabetic control mice were fed AIN-93 Grodent diet without genistein supplementation. [3] LG: Diabetic mice were fed AIN-93G rodent diet with 0.025% genistein dietary supplementation. [4] HG: Diabetic mice were fed AIN- 93G rodent diet with 0.1% genistein dietary supplementation.	Genistein	0.025% and 0.1% genistein dietarysupplementation.	2 weeks	Oral	Genistein supplementation: ↓Inflammation and oxidative stress The wound closure rates of LG and HG were accelerated significantly.	High RoB
Jacqueline ChorWing Tam, 2014	Rat/?	1:diabetic rats without ischemia and wound induction (Sham); 2: diabetic rats with ischemia surgery and without wound induction(I only); 3: diabetic rats with both ischemia and wound induction followed by water treatment (I + W); 4: diabetic rats with both ischemia and wound induction followed by NF3 treatment (I + W + NF3)	(NF3)* *[Astragali Radix (AR) and Rehmanniae Radix (RR) in the ratio of 2 to 1]	NF3 (at clinical relevant dose = 0.98 g/kg)	7 DAYs	IV	↓ Wound area of the diabetic foot ulcer	High RoB
Wan Y, et al. 2016	Rat / 25	G1: non-diabetes (N), gavaged with saline and applied with sesame oil on pelma of both hind limbs G2: diabetes and sham treatment (DS),gavaged with saline and applied with sesame oil on pelma of both hind limbs G3: metformin (M), gavaged with diluted M and applied with sesame oil on pelma of both hind limbs G4: S, gavaged with saline and applied with S onpelma of both hind limbs G5: combined treatment (CT) gavaged with diluted M and applied with S onpelma of both hind limbs	San-huang-sheng-fu oil	-	35 days	Oral	-↓ Plantar temperature and pain dysesthesia. -↓ Cyclooxygenase-2 (COX-2) - ↑ Vascular endothelial growth factor (VEGF). The wound areas of rats with diabetes in the 4 groups were close to those in group N (with t values from 0.111 to 1.476, P values above 0.05). On PTD (post treatment day) 21. The wound area of rats in group DS was significantly larger than that in group N (t = 5.502, P < 0.01), The wound areas of rats with diabetes in the other 3 groups were close to the area in group N (with t values from 0.544 to 1.676, P values above 0.05).	High RoB
Musalmah M, et al. 2002	Rats/ 24	1-Normal rats Untreated group -α-tocopherols group 2- Diabetic rats -Untreated group -α-tocopherols group	α-tocopherols (Vit E)	200 mg/kg	10 days	Oral	-↓ plasma malondialdehyde levels, -↑ lutathione peroxidase activity -Accelerated the rate of wound closure in treated rats	High RoB
T.W. Lau, et al. 2008	Rats / 164	1- Control group (water treatment) 2-Intervention group (Herbal treatment)	Radix Astragali and Radix Rehmanniae	-RA or RR 3.7 g/kg	7 days	Oral	Radix Astragali: There was no significant difference in the rate of change of wound area between the two groups [control & RA] (p = 0.94) Radix Rehmanniae: There was a significant difference in the rate of change of wound area between the two groups[control & RR] (p = 0.04).	High RoB
A. Raynaud-Simon et al. 2012	Rats / 17	1-Standard Diet 2- Arginine + proline Diet 3-Iso nitrogenousisoenergetic control Diet	Arginine + proline	120 ml / daily	5 DAYs	Oral	-Significantly improved wound repair (by angiogenesis)	High RoB
Na-Young Park 2011	Mice	1-CON (non-diabetic control mice), 2-DM (diabetic mice), 3-VCE (diabetic mice fed 0.5% vitamin C and 0.5% vitamin E supplemented diet), 4-Comb (diabetic mice fed 0.5% vitamin C, 0.5% vitamin E, and 2.5% N-acetylcysteine supplemented diet).	Vit. C + Vit. E + N-acetylcysteine	0.5% vitamin C, 0.5% vitamin E, 2.5% N-acetylcysteine	10 days	Oral	-Improved blood glucose levels and wound closure rate. -Wound closure rate of the VCE and the Comb groups were accelerated significantly as compared to those of the DM group.	High RoB
Mohan Daburkar et al. 2014	Rats	1-Untreated control 2-Untreated DFUs 3-DFUs treated with Aloevera gel ethanolic extract 4-DFUs treated with topical A. vera gel 5-DFUs treated with A. vera gel ethanolic extract and topical A. vera gel	Aloe vera gel ethanolic extract	Aloevera gel ethanolic extract [300 mg/kg, per os (p.o.), twice daily]	9 days	Oral	↓ wound size (P < 0.01)	High RoB

DM: Diabetes Mellitus; Vit C: Vitamin C; MSCs: Mesenchymal stem cells; RoB: Risk of Bias; DFU: Diabetic Foot Ulcer;ADEE:Angelica DahuricaEthanolic Extract; WFI: Water for injection; DW: Diabetic rats& wounded; HSP: Hesperidin; VEGF: Vascular endothelial growth factor; MDA: Malondialdehyde; ND:Non-diabetic;TCDO: Tetra chlorodecaoxide; VEGF: Vascular endothelial growth factor; CT: combined treatment

Observational studies

Eight observational studies [23, 24, 58–63] were finally included in the systematic review (Table 2). All articles except three papers [23, 61, 62] were conducted in Asian countries. Design of the included observational studies was prospective/retrospective cohort (n = 3) [23, 24, 59] and case-control and cross-section (n = 5) [58, 60–63]. In total, four paper evaluated the serum levels of vitamin D [23, 24, 59, 61] and the remaining studies were conducted on vitamin C, vitamin E, and selenium (Se), magnesium (Mg), copper (Cu) and Zn. All the included papers obtained more than half of maximum score of Newcastle Ottawa Quality (NOS) checklist and considered as high quality (Table 2).

Table 2.

Study characteristics of the included observational studies

Author (year)	Location	Study design	Measured nutrient	Comparative groups	Sample size/ sex (M/F)	Mean age (year)	Type of diabetes	Baseline BMI (kg/m2)	Baseline FBS (mg/dL)/HbA1c (%)	Mean Duration of diabetes (year)	Findings	Quality score
Greenhagen 2019	USA	Retrospective Cohort	Serum vitamin D levels Albumin creatinine	Diabetic patients with and without CN*, PAD*, DFI*, DFU, DPN*	50 Non-CN 28/22	55.7	T1DM	30.9	176/7.7	–	-Vitamin D with and without CN Lower levels of vitamin D: ↑PAD ↑DFI ↑DFU Lower levels of albumin and higher serum Creatinine: ↑PAD, DFI, DPN, and DFU.	8/10
					61 Non- DFI 29/32	55.9		33.8	163/7.7
					39 DFI 27/ 12	56.2		31.8	184/7.7
					79 Non- PAD 39/40	55.6		33.3	169/7.8
					22 PAD 17/5	58.5		33.0	179/7.5
					11 Non- DPN 7/4	52.7		30.2	141/6.9
					89 DPN 37/52	56.2		32	175/7.8
					46 No DFU 17/29	55.5		32.6	159/7.5
					54 DFU 39/15	56.1		32.2	182/7.9
					CN	55.9		33.8	175/7.7
Sever Cağlar 2018	Turkey	Prospective study	Vitamin D OPG* CRP Creatinine63.6 51.4	Patients with DFU (study group)	Total= 105 58 DFU (42/16)		T2DM	–	7.8	14.3 4.2	lower OPG levels: ↑DFU lower vitamin D D levels: ↑DFU OPG level: ↑CRP (r = 0.379, p < 0.05). OPG level: ↑Creatinine (r = 0.562, p < 0.05)	7/9
				Newly diagnosed T2DM (control group)	47 (27/20)				8
Joachim Feldkamp 2018	Germany	Cross-sectional	Vitamin D Calcium, Creatinine	Controls	99	69.6	T2DM	Not reported	–	–	↑Armstrong classification Vitamin D level↓ (r = −0.241; p < 0.01) Vitamin D level: Controls: 27.3 ± 12.2 Diabetes type 2: 19.0 ± 14.4 Diabetic foot syndrome: 11.8 ± 11.3 Inpatients: 10.8 ± 9.5 Outpatients: 13.4 ± 13.1	8/9
				T2DM	103	69.4			60.65	16.8
				Diabetic foot syndrome	104	70.2			59.56	17.5
				Inpatients	63	70.2			65.36	18.1
				Outpatients	41	69.6			51.91	17.4
Bolajokol 2016	Nigeria	Cross sectional	vitamin C, vitamin E, copper, selenium, zinc	DM + FU/ non-diabetic	70	–	T2DM	–	12.98 mmol/L/ 8.63%	–	↑DM + FU: Lower Vitamin C, Vitamin E, Selenium, -DM + FU: Copper and Zinc Lower vitamin C: ↑FPG: (r = 0.250, p = 0.037) Lower Cu: ↑HbAIc: (r = 0.131, p = 0.365)	7/9
					50				5.09 mmol/L/4.08%
ŞakirzgürKeşkek 2013	Turkey	cross-sectional	Magnesium	Patients with T2DM and DFU	Total= 147 F = 24	56.6	T2DM	–	221.5/10.0	12.4	Lower magnesium levels DFU ↑ (p < 0.001)	7/9
				Patients with T2DM without DFU	F = 24	53.4			184.9/8.5	12.1
				Healthy subjects	F = 26	53.1			90.1/5.5	–
Zubair 2013	India	Prospective cohort hospital based study	Vitamin D	Patients with Ulcer (DFU)	103/59	46.29	T2DM	24.84	9.6	<10 years DFU:111(68.5) non-DFU: 123(75.9) >10 years DFU: 51(31.4) non-DFU: 39(24.0)	↓Vitamin D DFU: 8.4(7.1–9.2) non-DFU: 29.8 (15.6–44.2) <0.005	7/9
				Patients without Ulcer (non-DFU)	102/58	47.10		24.03	7.9
Singh 2008	India	Cross-sectional	Vitamin E Malondialdehyde SOD*	DFU	Total= 32 (26/6)	53.81	T2DM	–	–	–	Vitamin E: Group A vs B= 5.35 < 0.001 B vs C= 2.34 < 0.05	7/9
				Non-DPN and Non-PAD	15	53.4
				Healthy control	15	50.66
Martha Rodríguez-Morán2001	México	Cross-sectional	Magnesium	Diabetic patients with and without ulcer	With DFU: 33 (17/16)	62.3	T2DM	26.1	291.8	13.9	Lower levels ofMagnesium: ↑DFU OR: 2.9, 95%CI: 1.7–6.8 Hypomagnesemia: With DFU: 49 (74.2%) Without DFU: 31 (93.9%)	7/9
					Without DFU: (31/66)	61.2		26.7	273.2	14.2

WG: Wagner grade; CN: Charcot neuroarthropathy; PAD: peripheral arterial disease; DFI: Diabetic Foot infection; DPN: Peripheral neuropath; OPG: Osteoprotegerin; SOD: Superoxide dismutase enzyme

Greenhang et al., compared the serum levels of vitamin D in patients suffering from DM and without Charcot neuroarthropathy (CN), infection, DFU, peripheral arterial disease and peripheral neuropathy in a retrospective cohort study. They found that 78% of patients had insufficient/deficient vitamin D and no significant difference were found between those with and without CN [23]. However, serum vitamin D concentrations in patients with DFU was significantly lower than those without this complication (p = 0.04) [23]. Bolajokol et al., also indicated that serum levels of vitamin C, vitamin E, and Se in diabetic patients with DFU were considerably lower than those without DM. However, no differences were found between two the groups with respect to the levels of Cu and Zn [58]. Feldkamp et al., concluded that the serum levels of vitamin D in 55.8% of patients with DFU was below 10 ng/mL. In addition, they reported a positive link between DFU and vitamin D deficiency [61].Singh et al., also reported that patients with DFU had lower serum levels of vitamin E compared to those with DM and healthy subjects (control group) [60].

Clinical trials

As presented in Table 3, seven clinical trials [30, 34, 35, 54–57] were considered for the current systematic review. All the included trials were double blind placebo controlled clinical trials. All publications except two papers were conducted in Iran. In their study, Afzali et al., showed that Magnesium oxide (250 mg/day) combined with Vitamin E (400 IU/day) could decrease length, depth and width of ulcer compared to placebo after 12 weeks in patients with grade III DFU [56]. The study of Mohseni et al., demonstrated that daily supplementation with probiotic for 12 weeks had positive effects on wound healing and it reduced length, depth and width of ulcer as well as improvement in FBS and insulin sensitivity in diabetic patients [55]. In another study, 50,000 IU vitamin D (every two weeks)also suggested an improvement in DFU after 12 weeks [34]. The double blind clinical trial performed by Momen-heravi et al., showed that 220 mg/day zinc sulfate decreased the dimension of ulcer as well as improvement in metabolic status [54]. Razzaghi et al., indicated positive effects of 250 mg/day magnesium on diabetic ulcer and glydemic status after 12 weeks [35]. A mixture of Arginine (7 g), glutamine (7 g), and calcium b-hydroxy-b-methylbutyrate (1.5 g) daily showed beneficial effects on wound healing after the 16-week intervention [30]. Furthermore, other clinical trials showed the positive impact of Arginine supplement on DFU [64, 65].

Table 3.

Characteristics of the included clinical trials in the systematic review

Author (year)	Location	Study design (randomization/ blindness)	Sample size/ Characteristics of participants	Intervention	Dosage	Duration (week)	Baseline BMI (kg/m2)	Baseline FBS/HbA1c In intervention group	Mean Duration of diabetes (year)	Findings	Jadad score
Afzali et al. (2019)	Iran	Double blind	57/ patients with grade 3 DFU (45:m, 12:f)	Magnesium oxide + Vitamin E	250 mg + 400 IU	12	30	FBS: 173 HbA1c:7.4	–	↓Ulcer length	4
										↓Ulcer width
										↓Ulcer depth
										↓FBS
										↓Insulin
										IR
										HbA1c
										↑Insulin sensitivity
										↓TG
										↓LDL-C
										↑VLDL
										↓ESR
										↓hs-CRP
										↓MDA
										↑HDL-C
										↑TAC
Mohseni et al. (2018)	Iran	Double blind	60 (20:m, 10:f)	Probiotic (Lactobacillus acidophilus, Lactobacillus casei,Lactobacillus FermentumandBifidobacteriumbifidum	2 × 109 CFU/g each	12	25	FBS:225 HbA1C:8	–	↓Ulcer length	4
										↓Ulcer width
										↓Ulcer depth
										↓FBS
										↓Insulin
										IR
										HbA1c
										↑Insulin sensitivity
										↓TG
										↓LDL-C
										↑VLDL
										↓TC
										↓hs-CRP
										↓MDA
										↑HDL-C
										↑TAC
										↓MDA
										-GSH
Razzaghi et al. (2018)	Iran	Double blind	70 patients with grade 3	Magnesium	250 mg	12	27	FBS: 226 HbA1c: 8.3		↓Ulcer length	4
										↓Ulcer width
										↓Ulcer depth
										↓FBS
										↓Insulin
										↓IR
										↓HbA1c
										↑Insulin sensitivity
										-HOMA-IR
										-TG, TC, LDL-C, HDL-C
Razzaghi et al. (2017)	Iran	Double blind	60 patients with grade 3	Vitamin D	50,000	12				↓Ulcer length	4
										↓Ulcer width
										↓Ulcer depth
										↓FBS
										↓Insulin
										↓IR
										↓HbA1c
										↑Insulin sensitivity
										-TG
										↓LDL-C
										-VLDL
										↓ESR
										↓hs-CRP
										↓MDA
										-HDL-C
										↓Erythema rate
										-GSH
Momen-Heravi et al. (2017)	Iran	Double blind	60 patients with grade 3 (42:m,18 = f)	Zinc sulphate	220 mg	12	25.8	FBS:190 HbA1c:7.9	–	↓Ulcer length	5
										↓Ulcer width
										↓Ulcer depth
										↓FBS
										↓Insulin
										↓IR
										↓HbA1c
										↑Insulin sensitivity
										-TG
										↓LDL-C
										-VLDL
										↓ESR
										↓hs-CRP
										↓MDA
										↑HDL-C
										↑TAC
										-GSH
Armstrong et al. (2014)	USA	Double blind	270	Mixture of Arginine (7 g), glutamine (7 g), calcium b-hydroxy-b-methylbutyrate (1.5 g) contained 79 kcal drink	Twice/day	16	32	NR	15	In patients with Alb≤40 g/L and/or ankle-brachial index <1: significant wound healing -No differences in healing in non-ischaemic or those with normal Alb	5
Gianni Belcaro et al. (2006)	Germany	Double blind	16 diabetic patients with severe microangiopathy causingfoot ulcerations	Oral treatment: 150 mg/d Pycnogenol (50 mg)	3 times daily	6 weeks	NR	NR	NR	Healing rate: 85%	2

NR: Not Reported; FBS: fasting blood glucose; IR: Insulin Resistance; TG: triglyceride; LDL-C low-density lipoprotein; VLDL-C very low-density lipoprotein; ESR: Erythrocyte sedimentation rate; hs-CRP: high-sensitivity C-reactive protein; MDA: malondialdehyde; HDL-C: high-density lipoprotein; TAC: total antioxidant capacity; GSH: glutathione; HOMA: Homeostatic Model Assessment for Insulin Resistance; MDA: Malondialdehyde

Discussion

The present systematic review showed the positive effects of dietary and herbal supplements with anti-oxidant properties on the process of wound healing in both animal models and clinical trials. Additionally, it was revealed that serum levels of vitamin D, C, vitamin E, and Se in subjects with DFU were lower than those in either healthy or patients without DFU. However, there are limited investigations in this area.

To the best of our knowledge, no systematic reviews have examined the link between serum levels of nutrients and DFU in observational studies. In their study, Dai et al., examined the association between serum levels of vitamin D and DFU in a systematic review and meta-analysis and found that severe vitamin D deficiency increased the risk for DFU by 3.2 [66]. In addition, a literature review performed by Macido et al., pointed out an inverse link between 25(OH) D3 levels and the presence of ulcer in patients with diabetes mellitus [67].

Based on previous works, malnutrition can play a key role in the progress of DFU and infection [68]. It has been shown that in patients with DFU, apart from body mass index (BMI), serum albumin, other metabolic parameters, and infection, nutritional status affect wound healing and metabolic profile [68, 69]. Apart from macronutrients, several micrunitrients including vitamins and minerals may involve in the treatment or the progress of DFU [70].

Limited observational studies which mostly had case-control design were included in the study. As such types of studies cannot reveal any cause and effect link between serum levels of nutrients and DFU, we cannot make a definitive decision regarding possible supplementations which are effective in DFU healing.

Although there were few studies on each specific nutrient, based on quality assessment tool all the included studies had high quality. Therefore, risk of bias seems not to be able to affect the results. We found that the levels of several nutrients with anti-oxidative properties in patients with DFU were lower than those in controls. Some studies compared the levels of nutrients in patients suffering from DFU with non-diabetic subjects with any ulcers. As there is a mutual link between DM and some nutrients, it seems reasonable to compare people with DM with non-diabetic subjects.

Some review studies summarized the effects of supplements on the DFU. A recent systematic review included only three clinical trials, two on protein-based supplementation and one on medicinal herb. In a study conducted by Ye et al., no nutrient supplements were included. They reported that nutritional supplementations had no positive effect on the DFU [41].

In the present study, all animal and clinical trials assessing the effects of non-topical dietary and herbal supplements were included. It seems that only three studies had examined the effect of herbs orally or intravenously on animal models. All studies conducted on herbal supplements showed positive effects on DFU and some metabolic parameters. Antioxidant effects of mentioned plants seem to be the main reason for such positive effects [71].

However, given few animal studies conducted on each nutrient and herb as well as low quality of methodology, we have to interpret results with great caution. In the present systematic review, we used SYRCLE checklist that had been developed in 2014 [47]. Therefore, it seems reasonable that papers published before this date, did not obtain acceptable score. This systematic review revealed that high-quality animal studies are needed to make a decision regarding the efficacy of supplements. In addition, to examine the novel supplements, more attention should be paid for the methodology and study design to minimize bias.

Wound healing is a complicated process containing four steps as follows: hemostasis, inflammation, proliferation and tissue remodeling [72]. One of possible mechanisms through which nutrients affect DFU is reducing free radicals and oxidative parameters followed by inducing a balance between oxidant and antioxidant components [73, 74]. Production of reactive oxygen species (ROS) play a crucial role in the etiology of complications related to diabetes including DFU. Following hyperglycemia, the levels of free radicals increase and can damage tissue, leading to non-healing DFU and infection [75]. Accordingly, anti-oxidant ingredients can be helpful to equilibrate oxidant and anti-oxidant components.

Another pathway contributing to wound healing is reducing systemic inflammatory parameters [76]. Anti-inflammatory effects of several nutrients including vitamin D have been proposed [77]. Hyperglycemia can disturb normal procedures for the production of cytokines. Therefore, uncontrolled glucose level can provide conditions for increasing inflammation, ulcer, and infection [78]. Duration of diabetes, age and BMI impact upon inflammatory cytokine production [78]. Therefore, apart from dosage, duration of intervention by a specific supplement with anti-inflammatory effects, the aforementioned factors can lead to different findings following the supplementation.

Vitamin D as an immune modulator plays a pivotal role in the activation of T and B cells by macrophages. Besides, it can decrease inflammatory parameters including C-reactive protein (CRP), and tumor necrosis- alpha (TNF- alpha) [23]. It has been reported that insufficient vitamin D increase the release of inflammatory cytokine (IL-1beta, IL-6, TNF-alpha) in subjects with DFU and infection, particularly when serum levels of 25 (OH) D3 is lower than 10 ng/mL [78]. Other proposed mechanisms for vitamin D are as follows: (i) improvement of glycemic status following supplementation with vitamin D, (ii) effects on insulin resistance, and (iii) beta-cell function [66]. In the present systematic review, studies examining the effects of vitamin D on DFU are greater than other nutrients and this issue highlighted the possible significant role of vitamin D on DFU. However, more high-quality studies are needed to clarify its efficacy.

Not only did some nutrients including Mg, vitamin D, vitamin E, and vitamin C have been shown positive effects on DFU, but also improved glycemic status, lipid profile, inflammatory and oxidative stress. The mentioned nutrients would be involved in several biological functions and therefore the low levels of such necessary nutrients cause to increase insulin resistance, beta-cell dysfunction, and deterioration of metabolic status.

Limitations of the study

Main limitations of this systematic review would include the following: First, given few studies conducted on each nutrient and supplement we were not able to do a quantitative synthesis. Second, studies published in languages other than English were not included. Third, due to few cohort studies, cause and effect relationship between nutrients and DFU remained unclear.

Strengths of this systematic review are as follows: (i) to the best of our knowledge this is the first systematic review to assess the effects of serum levels of nutrients on DFU healing (ii), it examined methodological quality for both animal and human studies and, (iii) address all types of supplements including medicinal herbs and dietary supplements.

Conclusion

We found that some dietary and herbal supplements can shorten the process of wound healing in both animal models and clinical trials, but the relevant studies included in the review are limited. Also, we revealed that lower serum levels of vitamin D, C, vitamin E, and Se in patients with DFU were likely to increase the risk for DFU, whereas any cause and effect relationship remained unclear due to limited cohort studies. Therefore, it would be recommended that further high-quality randomized controlled clinical trials and prospective cohort studies are needed to shed light the impacts of micronutrients and other dietary and herbal supplements in the progress and treatment of DFU.

Appendix

Table 4.

SYRCLE’s tool for assessing risk of bias

Funding

Endocrinology and Metabolism Research Center, Endocrinology and Metabolism Clinical Sciences Institute, Tehran University of Medical Sciences, Tehran, Iran

Declarations

Ethics approval

Not Applicable.

Conflict of interest

The authors declared that they have no conflict of interest.