Table 2.
Dietary intake of micronutrients and diabetic retinopathy.
| Study, Year Study Design Sample Size (n) |
Quality Score |
Dietary Factor and Its Association with DR |
Adjustment/Matched | Statistical Method Analysis | Key Findings |
|---|---|---|---|---|---|
| Antioxidants | |||||
| Carotenoids | |||||
| Tanaka et al., 2013 Prospective n = 978 |
10 | Carotenoids Protective |
Sex, age, BMI, HbA1c, diabetes duration, insulin treatment, oral hypoglycaemic agents without insulin treatment, systolic blood pressure, LDL and HDL cholesterol, triglycerides, physical activity alcohol, smoking, total energy intake, proportions of dietary protein, fat, carbohydrate, saturated fatty acids, omega-6 PUFA and omega-3 PUFA and sodium |
Multivariate Cox regression | Highest intake Q4 vs. lowest Intake Q1, HR: 0.52 (0.33–0.81) p < 0.01 |
| Sahli et al., 2016 Cross-sectional n = 1430 |
9 | Lutein carotenoids NS |
Diabetes duration, HbA1c, blood pressure, race, total energy consumption, and study center | Multivariable logistic regression | Intake Q4 vs. Q1, OR: 0.89 (0.31–2.50), p = 0.72 |
| Mayer-Davis et al., 1998 Cross-sectional n = 387 |
9 | Beta-Carotene NS |
Age, gender, ethnicity, diabetes duration, HbA1c, hypertension, caloric intake, and insulin use | Multivariable logistic regression | No significant associations with DR (data not shown) |
| Zhang et al., 2019 Case–control Type2 DM-86 control-40 |
8 | Retinol carotenoids Protective |
Age, sex, smoking, BMI and alcohol consumption | Logistic regression | Intake of retinol (100 μg/day) on DR (OR: 0.88, 95%CI, 0.79–0.98, p = 0.025) |
| Sasaki et al., 2015 Cross-sectional n = 379 |
8 | Beta-carotene NS |
Intake of energy | Data not shown | No significant associations with DR (data not shown) |
| Shalini et al., 2021 Cross-sectional n = 495 |
7 | Carotenoids Protective |
Nil | One-way analysis of variance F test with a post hoc test of least significant difference | The plasma concentration of carotenoids was significantly lower in the DR group compared to no DR patients and healthy controls (p < 0.001) |
| Vitamin C | |||||
| Tanaka et al., 2013 Prospective n = 978 |
10 | Vitamin C Protective |
Sex, age, BMI, HbA1c, diabetes duration, insulin treatment, oral hypoglycaemic agents without insulin treatment, systolic blood pressure, LDL and HDL cholesterol, triglycerides, physical activity alcohol, smoking, total energy intake, proportions of dietary protein, fat, carbohydrate, saturated fatty acids, omega-6 PUFA and omega-3 PUFA and sodium |
Multivariate Cox regression | Intake Q4 vs. Q1, HR: 0.61 (0.39–0.96), p = 0.03 |
| Mayer-Davis et al., 1998 Cross-sectional N = 387 |
9 | Vitamin C Risk |
Age, gender, ethnicity, diabetes duration, HbA1c, hypertension, caloric intake, and insulin use | Multivariable logistic regression |
Intake 9th decile vs. 1st quintile, OR: 2.21, (p = 0.01) |
| She et al., 2020 Cross-sectional n = 455 |
8 | Vitamin C NS |
Sex, race, insulin use, HbA1c, hypertension, exercise | Binomial logistic regression multivariate analysis | No significant association with DR (p = 0.413) |
| Sasaki et al., 2015 Cross-sectional n = 379 |
8 | Vitamin C NS |
Intake of energy | Data notshown | No significant association with DR (data not shown) |
| Millen et al., 2004 Cross-sectional n = 1353 |
8 | Vitamin C NS |
Race, BMI, diabetes duration, serum glucose, total energy intake, hypertension, waist–hip ratio, smoking, alcohol, drinking status, plasma cholesterol, hematocrit value, prevalent coronary heart disease, plasma triacylglycerol, diabetes treatment group, and oral hypoglycaemic treatment or insulin treatment | Multivariable logistic regression | Intake Q4 vs. Q1, OR: 1.4 (0.8–2.4), p = 0.19 |
| Vitamin E | |||||
| Tanaka et al., 2013 Prospective n = 978 |
10 | Vitamin E NS |
Sex, age, BMI, HbA1c, diabetes duration, insulin treatment, oral hypoglycaemic agents without insulin treatment, systolic blood pressure, LDL and HDL cholesterol, triglycerides, physical activity alcohol, smoking, total energy intake, proportions of dietary protein, fat, carbohydrate, saturated fatty acids, omega-6 PUFA and omega-3 PUFA and sodium |
Multivariate Cox regression | Intake Q4 vs. Q1, HR: 0.84 (0.51–1.40), p = 0.51 |
| Mayer-Davis et al., 1998 Cross-sectional N = 387 |
9 | Vitamin E Risk (in non-insulin taking subjects) |
Age, gender, ethnicity, diabetes duration, HbA1c, hypertension, caloric intake, and insulin use | Multivariable logistic regression | No association found in insulin subjects and in non-insulin taking subjects: Intake 10th decile vs. 1st quintile, OR: 3.79, (p < 0.02) |
| She et al., 2020 Cross-sectional n = 455 |
8 | Vitamin E Protective |
Sex, race, insulin use, HbA1c, hypertension, exercise | Binomial logistic regression multivariate analysis | Intake in DR vs. No DR (OR: 0.97, 95%CI: 0.95–1.00, p = 0.036) |
| Granado-Casas et al., 2018 Cross-sectional n = 243 |
8 | Vitamin E Protective |
Age, sex, educational level, smoking, physical activity, BMI, dyslipidemia, hypertension, diabetes duration, HbA1c | Multivariable conditional logistic regression models | Intake of Vitamin E on DR (OR: 0.85 [0.77–0.95], p = 0.006) |
| Sasaki et al., 2015 Cross-sectional n = 379 |
8 | Vitamin E NS |
Intake of energy | Data notshown | No significant associations with DR (data not shown) |
| Millen et al., 2004 Cross-sectional n = 1353 |
8 | Vitamin E NS |
Race, BMI, diabetes duration, serum glucose, total energy intake, hypertension, waist–hip ratio, smoking, alcohol, drinking status, plasma cholesterol, hematocrit value, prevalent coronary heart disease, plasma triacylglycerol, diabetes treatment group, and oral hypoglycaemic treatment or insulin treatment | Multivariable logistic regression | Intake Q4 vs. Q1, OR: 1.4 (0.8–2.3), p = 0.76 |
| Selenium | |||||
| She et al., 2020 Cross-sectional n = 455 |
8 | Selenium Protective |
Sex, race, insulin use, HbA1c, hypertension, exercise | Binomial logistic regression multivariate analysis | Intake in DR vs. No DR (OR: 0.98, 95%CI: 0.96–1.00, p = 0.017) |
| Riboflavin | |||||
| She et al., 2020 Cross-sectional n = 455 |
8 | Riboflavin NS |
Sex, race, insulin use, HbA1c, hypertension, exercise | Binomial logistic regression multivariate analysis | No significant association with DR (p > 0.05) |
| Vitamin D | |||||
| Millen et al., 2016 Cross-sectional n = 1305 |
9 | Vitamin D NS |
Race, duration of diabetes, HbA1c and, hypertension |
Multivariable logistic regression | Intake Q4 vs. Q1, OR: 1.20 (0.76–1.89), p trend = 0.740 |
| Alcubierre et al., 2015 Case–control Case:139 Ctrl:144 |
8 | Vitamin D NS |
NIL | Chi-squared | No significant associations with DR (p = 0.93) |
| Choline | |||||
| Liu et al., 2021 Cross-sectional n = 1272 |
9 | Choline Risk in female |
Age, race, diabetes duration, glycaemic control, CVD, CKD * results analyzed in individual sex groups | Multivariable logistic regression | High intake vs. low intake (OR: 2.14, 95%CI: 1.38–3.31; p = 0.001) |
| Calcium | |||||
| Chen et al., 2022 Cross-sectional n = 5321 |
9 | Calcium Protective |
Age, sex, race, smoking, serum glucose, serum laboratory data, hemoglobin | Multivariable logistic regression | High intake vs. low intake OR: 0.70, 95%CI: 0.54–0.90, p = 0.05) |
| Alcubierre et al., 2015 Case–control Case:139 Ctrl:144 |
8 | Calcium NS |
NIL | Chi-squared | No significant associations with DR (p = 0.65) |
| Potassium | |||||
| Tanaka et al., 2013 Prospective n = 978 |
10 | Potassium NS |
Sex, age, BMI, HbA1c, diabetes duration, insulin treatment, oral hypoglycaemic agents without insulin treatment, systolic blood pressure, LDL and HDL cholesterol, triglycerides, physical activity alcohol, smoking, total energy intake, proportions of dietary protein, fat, carbohydrate, saturated fatty acids, omega-6 PUFA and omega-3 PUFA and sodium |
Multivariate Cox regression | No significant association with DR (p > 0.05) |
| Chen et al., 2022 Cross-sectional n = 5321 |
9 | Potassium Protective |
Age, sex, race, smoking serum glucose, serum laboratory data, hemoglobin | Multivariable logistic regression | High intake vs. low intake OR: 0.761, 95%CI: 0.59–0.97, p = 0.029 |
| Sodium | |||||
| Horikawa et al., 2021 Prospective n = 912 |
10 | Sodium Risk (under low vegetable consumption) |
Age, sex, BMI, HbA1c, diabetes duration, LDL cholesterol, HDL cholesterol, log-transformed triglycerides, insulin treatment, smoking, alcohol, energy intake, physical activity, systolic blood pressure, angiotensin II receptor blocker, angiotensin-converting enzyme inhibitor, calcium channel blocker |
Multivariate Cox regression analyses | Intake for 2nd, 3rd, and 4th quartile vs. 1st quartile, HRs were 0.87 [95%CI, 0.31–2.41], 2.61 [1.00–6.83], and 3.70 [1.37–10.02], respectively p = 0.010. |
| Horikawa et al., 2014 Prospective n = 978 |
10 | Sodium NS |
Sex, age, BMI, HbA1c, duration of diabetes, LDL cholesterol, HDL cholesterol, log-transformed triglycerides, insulin treatment, lipid-lowering agents, smoking, alcohol intake, energy intake, sodium intake, and physical activity |
Multivariate Cox regression | Intake Q4 vs. Q1, HR: 1.10 (0.75–1.61), p = 0.55 |
| Roy et al., 2010 Prospective n = 469 |
10 | Sodium Risk (ForDME) NS for DR |
Age, sex, HbA1c, hypertension, total caloric intake, protein intake, oleic acid intake, physical exercise, and oleic acid intake | Multivariable logistic regression | Intake Q4 vs. Q1, OR: 1.43 (1.10–1.86), p = 0.008 for DME. No significant associations with DR |
| Cundiff et al., 2005 Prospective n = 1412 |
8 | Sodium NS |
Intake of energy | Spearman correlation | Sodium in mg/kcal against DR progression rate, r = 0.02 (p = 0.47) |
| Engelen et al., 2014 Cross-sectional n = 1880 |
7 | Sodium NS |
Sex, age, smoking, BMI, urinary potassium excretion, sat fat intake, protein intake antihypertensive medication, total energy intake, physical activity, fiber intake, and alcohol intake |
Multivariable logistic regression | Per 1g/day increase in dietary salt intake against NPDR OR: 1.00, (0.96–1.04, p = 0.84. PDR OR: 1.02 (0.95–1.08), p = 0.65 |
| Vitamin B6 | |||||
| Horikawa et al., 2019 Prospective n = 978 |
9 | Vitamin B6 Protective |
Age, sex, BMI, HbA1c, diabetes duration, systolic blood pressure, LDL cholesterol, HDL cholesterol, triglycerides, insulin treatment, oral hypoglycemic agents, antihypertensive agents, lipid-lowering agents, urine albumin creatinine ratio, estimated glomerular filtration rate, alcohol, smoking, energy intake, physical activity, retinol, vitamin B1, vitamin B2, vitamin B9, vitamin B12 | Multivariate Cox regression analyses | Intake Q4 vs. Q1 HR: 0.50, 95%CI: 0.30–0.85, p = 0.010) |
BMI—Body mass index, CI–Confidence interval, CVD—Cardiovascular disease, CKD—Chronic kidney disease, CI—Confidence interval, DR—Diabetic retinopathy, DME—diabetic macular edema, DM—Diabetes mellitus, HDL—High-density lipoprotein, HR—Hazard ratio, HbA1c—glycated hemoglobin, LDL—Low-density lipoprotein, NS—Not significant, NPDR—Non-proliferative diabetic retinopathy, OR—Odds ratio, PUFA—Polyunsaturated fatty acid, PDR—Proliferative diabetic retinopathy.