Table 3. Dietary intake of macro-nutrients and DR.
Author, year | Association | Study Design | Quality | Dietary Factor | Sample Size | DR outcome type | Confounders adjusted for | Statistical methods | Main Findings | |
---|---|---|---|---|---|---|---|---|---|---|
Dietary Fats / lipids | ||||||||||
Mono-Unsaturated Fatty Acids (MUFA) | ||||||||||
Alcubierre, 2016 | Protective | Case-Control | 10 | MUFA | Case: 146 Ctrl: 148 |
Prevalence | Age, gender, diabetes duration, energy intake, educational level, physical activity, waist circumference, systolic BP, HDL cholesterol & diabetes treatment | Multivariable Logistic Regression | High MUFA consumption vs Low MUFA consumption, OR: 0.42 (0.18–0.97) | |
Cundiff, 2005 | Risk | Prospective | 8 | MUFA | 1412 | Progression | Energy Intake | Spearman Correlation | MUFA in %/kcal against DR progression rate, r = 0.12 (p = 0.001) | |
Roy, 2010 | NS | Prospective | 9 | MUFA | 469 | Progression & Incidence | Total caloric intake, total fat, sat fat, oleic acid, linoleic acid, protein, fiber, cholesterol & sodium intakes | Multivariable Logistic Regression | No significant associations with DR (Data not reported) | |
Sasaki, 2015 | NS | Cross Sectional | 10 | MUFA | 379 | Prevalence | Age, gender, HBA1C, mean arterial pressure & diabetes duration | Multivariable logistic regression models | Per 10 energy-adjusted g/d increase, OR: 1.19 (0.74–1.92) | |
Roy, 1989 | NS | Cross-Sectional | 5 | MUFA | 34 | Prevalence | Energy Intake | t-test | No significant associations with DR (Data not reported) | |
Poly-Unsaturated Fatty Acids (PUFA) | ||||||||||
Sala-Vila, 2016 | Protective | Prospective | 9 | PUFA (LCw3) | 3482 | Incidence | Age, gender, BMI, intervention group, yeasr after diagnosis of diabetes, use of insulin, use of oral hypoglycemic agents, smoking, systolic BP, hypertension, physical activity, adherence to meddiet. | Cox Proportional Hazard Model | >500mg/d Vs <500mg/d, HR: 0.52 (0.31–0.88) | |
Sasaki, 2015 | Protective for well controlled diabetics | Cross Sectional | 10 | PUFA | 379 | Prevalence | Age, gender, HBA1C, mean arterial pressure & diabetes duration | Multivariable logistic regression models |
All subjects: Per 10 energy-adjusted g/d increase, OR: 0.67 (0.37–1.20) Well controlled Diabetics: Per 10 energy-adjusted g/d increase, OR: 0.18 (0.06–0.59) |
|
Cundiff, 2005 | Risk | Prospective | 8 | PUFA | 1412 | Progression | Energy Intake | Spearman Correlation | PUFA in %/kcal against DR progression rate, r = 0.09 (r = 0.004) | |
Roy, 2010 | NS | Prospective | 9 | PUFA | 469 | Progression & Incidence | Total caloric intake, total fat, sat fat, oleic acid, linoleic acid, protein, fiber, cholesterol & sodium intakes | Multivariable Logistic Regression | No significant associations with DR (Data not reported) | |
Alcubierre, 2016 | NS | Case-Control | 10 | PUFA | Case: 146 Ctrl: 148 |
Prevalence | Age, gender, diabetes duration, energy intake, educational level, physical activity, waist circumference, systolic BP, HDL cholesterol & diabetes treatment | Multivariable Logistic Regression | High PUFA consumption vs Low MUFA consumption, OR: 0.99 (0.69–1.41) | |
Roy, 1989 | NS | Cross-Sectional | 5 | PUFA | 34 | Prevalence | Energy Intake | t-test | No significant associations with DR (Data not reported) | |
Oleic Acid | ||||||||||
Alcubierre, 2016 | Protective | Case-Control | 10 | Oleic Acid | Case: 146 Ctrl: 148 |
Prevalence | Age, gender, diabetes duration, energy intake, educational level, physical activity, waist circumference, systolic BP, HDL cholesterol & diabetes treatment | Multivariable Logistic Regression | High Intake Tertile (T3) vs Lowest Intake Tertile (T1), OR: 0.37 (0.16–0.85) | |
Roy, 2010 | NS | Prospective | 9 | Oleic Acid | 469 | Progression & Incidence | Total caloric intake, total fat, sat fat, oleic acid, linoleic acid, protein, fiber, cholesterol & sodium intakes | Multivariable Logistic Regression | No Significant associations with DR (Data not reported) | |
Interventional Studies | ||||||||||
Houtsmuller, 1979 | Protective | Interventional | High Bias | Unsaturated Fats | 96 | Progression | Matched for gender | Saturated Fat Diet Vs Unsaturated Fat Diet Males (n = 52, 26 on each diet) P<0.001 Females (n = 44, 22 on each diet) P<0.025 |
||
Howard-williams, 1985 | NS | Interventional | High Bias | PUFA | 149 | Incidence | Matched for age, sex & BMI | Persons on modified fat diet (PUFA: saturated fat ratio, 0.3) vs persons on low carb diet (PUFA: Saturated fat ratio, 0.9) All patients (n = 149) No difference between two groups (chi-squared, p = 0.69) Dietary compliers (n = 58) No difference between two groups (chi-squared, p = 0.13) |
||
Carbohydrates | ||||||||||
Cundiff, 2005 | Protective | Prospective | 8 | Carbohydrates | 1412 | Progression | Energy Intake | Spearman Correlation | Carbohydrates in %/kcal against DR progression rate, r = -0.11 (p<0.001) | |
Roy, 1989 | Protective | Cross-Sectional | 5 | Carbohydrates | 34 | Prevalence | Energy Intake | t-test | Persons without retinopathy vs Persons with retinopathy (p<0.05) | |
Horikawa, 2017 | NS | Prospective | 10 | Carbohydrates | 978 | Incidence and Progression | Age, sex, BMI, HbA1C, Diabetes Duration, systolic BP, LDL-cholesterol, HDL-cholesterol, triglycerides, treatment by insulin, treatment by antihypertensive agents, treatment by lipid-lowering agents, current smoker, alcohol intake, energy intake & physical activity | Multivariable Cox Regression Models | Highest Intake Tertile (T3) vs lowest Intake Tertile (T1), HR: 1.00 (0.72–1.38) | |
Roy, 2010 | NS | Prospective | 9 | Carbohydrates | 469 | Progression & Incidence | Total caloric intake, total fat, sat fat, oleic acid, linoleic acid, protein, fiber, cholesterol & sodium intakes | Multivariable Logistic Regression | No significant associations with DR (Data not reported) | |
Alcubierre, 2016 | NS | Case-Control | 10 | Carbohydrates | Case: 146 Ctrl: 148 |
Prevalence | Age, gender, diabetes duration, energy intake, educational level, physical activity, waist circumference, systolic BP, HDL cholesterol & diabetes treatment | Multivariable Logistic Regression | High Intake Tertile (T3) vs lowest intake tertile (T1), OR: 1.18 (0.45–3.09) | |
Sasaki, 2015 | NS | Cross Sectional | 8 | Carbohydrates | 379 | Prevalence | Energy Intake | Chi-Squared | No significant associations with DR (data not reported) | |
Protein | ||||||||||
Cundiff, 2005 | Protective | Prospective | 8 | Protein | 1412 | Progression | Energy Intake | Spearman Correlation | Protein in %/kcal against DR progression rate, r = -0.6 (p = 0.0188) | |
Roy, 1989 | Risk | Cross-Sectional | 5 | Protein | 34 | Prevalence | Energy Intake | t-test | Persons without retinopathy vs Persons with retinopathy (p<0.02) | |
Roy, 2010 | NS | Prospective | 9 | Protein | 469 | Progression & Incidence | Total caloric intake, total fat, sat fat, oleic acid, linoleic acid, protein, fiber, cholesterol & sodium intakes | Multivariable Logistic Regression | No Significant associations with DR (Data not reported) | |
Alcubierre, 2016 | NS | Case-Control | 10 | Protein | Case: 146 Ctrl: 148 |
Prevalence | Age, gender, diabetes duration, energy intake, educational level, physical activity, waist circumference, systolic BP, HDL cholesterol & diabetes treatment | Multivariable Logistic Regression | Highest protein intake tertile (T3) vs lowest protein intake tertile (T1), OR: 1.24 (0.49–3.16) | |
Sasaki, 2015 | NS | Cross Sectional | 8 | Protein | 379 | Prevalence | Energy Intake | Chi-Squared | No Significant associations with DR (Data not reported) |