Table 3.

Studies of antioxidant supplementation in patients with diabetic retinopathy included in our review.

Author/Year/Country/Reference	Study	Study Focus	Antioxidant Composition per Pill	Trade Name Dose	N per Group: Supplemented (S) and Control (C). Mean Age (Years)	Follow-Up Time in Months	Clinical Findings	Biochemical Findings
Sanz-González 2020 Spain [145]	Case-control study	Type 2 DM with and without DR	Oil as a source of PUFAs: 400 mg Omega-3 (ω3): DHA 140 mg Vitamin C 80 mg Vitamin D 5 µg Vitamin B 20.1 mg Vitamin E 12 mg Lutein 6 mg Zeaxanthin 0.3 mg Glutathione 1 mg Hydroxytyrosol 0.75 mg Zinc 7.5 mg Copper 1 mg Selenium 55 µg Manganese 2 mg Dosage = 1 tablet/day: Supplement or Placebo	Nutrof Omega® (Thea SA, (Barcelona, Spain)	N = 365 225 T2DM −With DR: 100 −Without DR: 125 140 healthy controls Mean Age: T2DM: 60 Controls: 55	38	The placebo group was more representative in subjects with T2DM in whom DR progressed. NS differences in IOP and CMT	The A/ω3 regime significantly reduced the pro-oxidants (p < 0.05) and augmented the antioxidants (p < 0.05).
Moon 2019 Korea [146]	Randomised (1:2:2), double-blind controlled trial	Type 2 DM with NPDR 40–80 y.o. AV > 0.5 Without laser or intravitreal therapy or intraocular surgery in the previous 6 months	S group 1: 50 mg—Grape seed proanthocyanidins extracts (GSPE) (Vitis vinifera extract) S group 2: 250 mg of calcium dobesilate (CD) C group.	GSPE: Entelon (Hanlim Pharm, Seoul, South Korea) CD: Doxium (Ilsung Pharm, Seoul, South Korea).	N = 86 3 tablets 3 times daily S1: GSPE (150 mg/day): 32 S2: CD (750 mg/day): 35 Placebo: 19	12	Hard exudates severity improvement: higher in GSPE (43.9%) vs. CD (14.29%) and vs. placebo (8%) (0.0007) NS differences between OCT parameters (CMT, TVM) GSPE TVM significantly decreases with respect to baseline.	NS differences with regard to vital signs and laboratory results between groups.
Lafuente 2019 Spain [129]	Randomised Single-Blind Controlled Trial	T2DM adults with decreased vision due to central-involved DME	Omega-3 Fatty Acids DHA 350 mg EPA 42.5 mg DPA 30 mg Vitamin C 26.7 mg Vitamin E 4 mg B vitamins 7.3 mg Lutein 3 mg Zeaxanthin 0.3 mg Glutathione 2 mg Zinc 1.66 mg Copper 0.16 mg Selenium 9.16 µg Manganese 0.33 mg	Brudyretina® 1.5 g (Brudy Lab S.L Barcelona, Spain) 3 capsules of 1.5 g once daily	N = 55 (69 eyes) S + Ranibizumab * n = 26 (31 eyes) C: Only Ranibizumab * n = 29 (38 eyes) All patients with four monthly doses of ranibizumab followed by pro re nata basis.	36	VA: NS difference in ETDRS letters. Gains of >5 and >10 letters significantly higher in S group. CMT: Significant decrease in S group vs. C group (275 ±50 µm vs. 310 ± 97 µm) Number of Ranibizumab injections: NS differences between groups.	Significant differences in HbA1c, plasma total antioxidant capacity values, erythrocyte DHA content and IL-6 levels in favour of S group.
Sepahi 2018 Iran [147]	Phase 2 randomised, double-blind, placebo-controlled trial.	Refractory to conventional DME therapy in type 1 or 2 diabetes Refractory therapy including: macular photocoagulation and intravitreal injection of bevacizumab with or without triamcinolone	S1: Crocin tablet 15 mg S2: Crocin tablet 5 mg	Crocin tablet Pharmaceutical laboratory of School of Pharmacy, Mashhad University of Medical Science, Mashhad, Iran 1 tablet per day (15 mg, 5 mg or placebo)	N = 60 patients (101 eyes) S 1: 20 (33 eyes) S 2: 20 (34 eyes) C: 20 (34 eyes) Age: 41–82	Supplementation: 3 Follow-up: 6	VA: LogMAR: S1 significantly improved compared to S2 (p < 0.05) and to C (p = 0.02). CMT: S1 significantly improved compared to S2 (p < 0.05) and to C (p = 0.005). S2 NS improvement compared to C.	HbA1c and FBS: S1 and S2 significantly better than C.
Zhang 2017 China [148]	Randomised, double-blind, placebo-controlled trial	NPDR mild or moderate stages Type 2 diabetes Exclusion criteria: DME, other eye disorders other than mild or moderate NPDR	Lutein 10 mg Placebo capsule	Lutein 10 mg 1 capsule once a day (1 capsule of placebo once a day if C) Lutein Pharmaceutical Co Ltd. (Guangzhou, China)	N = 30 patients S: 15 C: 15 Mean age: 60.2., SD: 10.3	9	VA: slight NS improvement in S (p = 0.11) Contrast sensitivity: S: significant increase in 3 cycles/° by 0.16 (p = 0.02) ANOVA analysis showed differences between S and C. NS in 6.12 and 36 cycles/°. Glare sensitivity: NS differences.
Lafuente 2017 Spain [149]	Randomised Single-Blind Controlled Trial	Type 2 diabetes adults with decreased vision due to central-involved DME.	Omega-3 Fatty Acids DHA 350 mg EPA 42.5 mg DPA 30 mg Vitamin C 26.7 mg Vitamin E 4 mg B vitamins 7.3 mg Lutein 3 mg Zeaxanthin 0.3 mg Glutathione 2 mg Zinc 1.66 mg Copper 0.16 mg Selenium 9.16 µg Manganese 0.33 mg	Brudyretina® 1.5 g (Brudy Lab S.L Barcelona, Spain) 3 capsules of 1.5 g once daily	N = 76 eyes S + Ranibizumab * n = 34 C: Only Ranibizumab * n = 42 All patients with four monthly doses of ranibizumab followed by pro re nata basis.	24	VA: NS difference in ETDRS letters. Gains of >5 letters significantly higher in S group (p = 0.044), NS for gains of >10 letters. CMT: Significant decrease in S group (95% CI 7.20–97.656; p = 0.024) Number of Ranibizumab injections: NS differences between groups.	Significant increase in TAC (total antioxidative capacity) in S group (p < 0.001) Significant reduction in the erythrocyte membrane content of ω-6 arachidonic acid in the S group (p < 0.05) NS differences in HbA1c levels
Rodriguez-Carrizalez 2016 Mexico [150]	Randomised, controlled, phase IIa clinical trial	T2DM with NPDR, but without DME	S1: Ubiquinone 400 mg Dosage 1 tablet/day S2: Vitamin C 180 mg Vitamin E 30 mg Lutein 10 mg Astaxanthin 4 mg Zeaxanthin 1 mg Zinc 20 mg Dosage 1 tablet/day C: Placebo tablet	Noncommercialised supplement	N = 60 patients S1: N = 20 S2: N = 20 C: N = 20 Mean age S1: 58.5 ± 1.9 S2: 62.1 ±1.1 C: 57.8± 1.9	6	VA: NS changes	S1 and S2 Significant decrease in lipid peroxidation products, NO metabolites, catalase and glutathione peroxidase (p < 0.0001) Increased TAC (p < 0.0001) Vs. C group NS changes in HbA1c%, cholesterol and triglyceride levels between groups
Chous 2016 USA [144]	Randomised controlled clinical trial	T1 or T2DM without DR or with mild-to-moderate NPDR without CSME	S: Vitamin C 60 mg Vitamin D3 50 mg Vitamin E 40 mg α-Lipoic acid 150 mg Coenzyme Q10 20 mg Omega-3 Fatty Acids EPA 128 mg DHA 96 mg Zeaxanthin 8 mg Lutein 4 mg Zinc oxide 15 mg Benfotiamine N-acetyl cysteine Grape seed extract Resveratrol Turmeric root Extract green tea leaf Pycnogenol (Not specified mg) Dosage = 2 tablets/day C: Placebo tablet	DiVFuSS® (ZeaVision, LLC, Chesterfield, MO, USA)	N = 67 patients S: N = 39 C: N = 28 Mean age S: 53.5 ± 14.6 C: 59.7 ± 10.3	6	VA: NS changes CMT: NS changes RNFL thickness: NS changes Contrast sensitivity, colour error Score, visual field mean sensitivity and MPOD: significant 27% improvement in the S group vs. 2% in the C group. (p values ranging from 0.008 to <0.0001). MPOD (macular pigment optical density)	NS changes in HbA1c, total cholesterol or TNF-α between the groups
Roig-Revert 2015 Spain [151]	Randomised, prospective, multicentre study	T2DM Group 1: NPDR ± DME Group 2: Diabetic patients without DR Healthy subjects	S: Vitamin C 80 mg Vitamin D 5 µg Vitamin B 20.1 mg Vitamin E 12 mg Omega-3: DHA 140 mg Lutein 6 mg Zeaxanthin 0.3 mg Glutathione 1 mg Hydroxytyrosol 0.75 mg Zinc 7.5 mg Copper 1 mg Selenium 55 µg Manganese 2 mg Dosage = 1 tablet/day C: no placebo capsule	Nutrof Omega® (Thea SA, (Barcelona, Spain)	N = 208 patients Group 1 DM DR+ (N = 62) S (n = not specified) C (n = not specified) Group 2 DM DR- (N = 68) S (N = not specified) C (n = not specified) Group 3 Healthy subjects (N = 78) S (n = not specified) C (n = not specified) Mean age DM DR+ 65.1 ± 8.6 DM DR− 62.3 ± 10.1	18	Group 1 DM DR + DR progression: S: 61% C: 91% Group 2 DM DR- DR onset: S: 9% C: 35% RNFLT of the LE was significantly reduced in the S group (p = 0.01)	Significant reduction in TAS in supplemented DMDR+ (p = 0.020) Plasma lipid peroxidation by-products significantly decreased in the DMDR+ supplemented group. NS in terms of HbA1c, HDL/LDL cholesterol and triglycerides.
Domanico 2015 Italy [152]	Randomised prospective study	T2DM showing mild-to-moderate NPDR, without CSME or CVRF	Vitamin E 30 mg Pycnogenol 50 mg Coenzyme Q10 20 mg Dosage = 1 tablet/day C: no placebo capsule	Diaberet® (Visufarma, Rome, Italy)	N = 68 patients (eyes) S: N = 34 C: N = 34 Mean age S: 58.29 ± 12.37 C: 62.29 ± 11.54	6	CMT: significant reduction on the S group (p < 0.01) (–15.44 µm, [95% CI: 3.26, 27.61])	Significant reduction of ROS levels (free oxygen radical test) in the S group (p < 0.001)
Watanabe 2014 Japan [153]	Randomised, prospective study	T2DM patients without DR	2.5 g of goshajinkigan extract three times a day, which included: 4.5 g of the compound extracts of 10 herbal medicines: Rehmanniae radix (5 g), Achyranthis radix (3 g), Corni fructus (3 g), Dioscoreae rhizoma (3 g), Hoelen (3 g), Plantaginis semen (3 g), Alismatis rhizoma (3 g), Moutan cortex (3 g), Cinnamomi cortex (1 g) and heat-processed Aconiti radix (1 g)	TJ-107; Tsumura Co., Tokyo, Japan	N = 116 patients S: N = 74 C: N = 42 Mean age S: 59.4 ± 7.8 C: 60.9 ± 7.4	60	Progression of retinopathy: No differences between S and C. A total of 25 patients had DR at the end of the study. 17.9% in Goshajinkigan group 20.0% in control group p = 0.816	Glycated haemoglobin significantly decreased in the S group at the 60th month. Fasting glucose significantly decreased in the S group beginning at the 36th month. No differences between insulin or oral antidiabetic medications.
Haritoglou 2011 Germany [154]	Randomised, prospective, multicentre, study	T2DM showing mild-to-moderate NPDR in at least one eye	S: α-lipoic acid (ALA) 600 mg Dosage 1 tablet/day C: placebo tablet	Noncommercialised supplement	N: = 399 patients S: = 196 C: = 203 Mean age S 58.0 C 57.9	24	CSME debut during follow-up S 26/196 C 30/203 NS reduction in macular oedema development (p = 0.7108)	NS differences in terms of HbA1c levels between groups
García-Medina 2011 Spain [155]	Randomised prospective study	T2DM with NPDR but no CSME	S: Vitamin C 60 mg Vitamin E 10 mg Lutein 3 mg Zinc 13.5 mg Copper 1 mg Selenium 10 µg Manganese 1 mg Niacin 10 mg β-Carotene 3 mg Dosage = 2 tablets/day C: no placebo capsule	Vitalux Forte® (Novartis Pharma AG Ophthalmics, Basel, Switzerland)	N = 97 patients S: N = 56 C: N = 41 Mean age S 53.3 ± 11.9 C 57.0 ± 11.4	60	VA: NS changes DR degree: Significant progression in C group (p < 0.01) vs. non-significant progression in S group	Significant reduced plasma lipid peroxidation end products (MDA) in S vs. increased in C group (p < 0.01) Stable TAS in the S group vs. significant decrease in C group (p = 0.02)
Forte 2011 Italy [156]	Randomised prospective, interventional, controlled study	T2DM and DME without macular thickening at OCT	S = Desmin 300 mg Troxerutin 300 mg C. asiatica 30 mg Melilotus 160 Dosage 1/day C = Placebo capsule	Noncommercialised supplement	N = 40 patients (eyes) S = 20 C = 20 Mean age S 63.6 ± 3.1 C 62.2 ± 3.4.	14	VA: NS differences CMT: NS differences between groups. Five eyes of the S group showed resolution of retinal cysts, in comparison to no changes in the C group RS (dB): S showed a significant increase at month 14 (p < 0.001) (16.43 ± 0.39)	NS differences during follow-up in terms of HbA1c, microalbuminuria or blood pressure
Bursell 1999 USA [157]	Randomised double-masked placebo-controlled crossover trial	T1DM without or with minimal DR	S = Vitamin E 1800 IU C = Placebo capsule Dosage 1800 IU/day	Noncommercialised supplement	N = 45 patients S = 36 (T1DM) C = 9 (ND) 4 months follow-up Crossover S = 9 (ND) C = 36 (T1DM) 4 months follow-up Mean age DM = 31.2 ± 6.8 ND = 31.6 ± 7.1	8	T1DM significant increase in retinal blood flow (p < 0.001) (34.5 ± 7.8 pixel2/s) Retinal blood flow measured by mean circulation times in fluorescein angiography: C: No changes	NS differences in terms of HbA1c between groups Statistically significant creatinine clearance improvement after supplementation in T1DM subjects (p = 0.039). This change reverted after crossover.

C = Control group, CMT = Central Macular Thickness, CSME = Clinically significant macular oedema, DHA = Docosahexaenoic acid, DM = Diabetes Mellitus, DM DR + = Diabetic patients with diabetic retinopathy, DM DR = Diabetic patients without diabetic retinopathy, DME = diabetic macular oedema, DPA = Docosapentaenoic acid, DR = Diabetic Retinopathy, EPA = Eicosapentaenoic acid, ETDRS = Early Treatment Diabetic Retinopathy Study Scale, FBS = fasting blood sugar, HbA1C = glycated haemoglobin, IOP = Intraocular pressure, IU = International Units, MDA = Malondialdehyde, MPOD = macular pigment optical density, NO = nitrogen oxide. NPDR = Nonproliferative diabetic retinopathy, NS = Not statistically significant, PUFA = polyunsaturated fatty acids, RS = retinal sensitivity, S = Supplemented group, T2DM = Type 2 Diabetes Mellitus, TAS = Total Antioxidant Status, TVM = total macular volume, VA = visual acuity. * Ranibizumab dosage: four loading doses followed by pro re nata treatment, both groups.