8,456 peer-reviewed studies, each with a short note written from its published abstract: what kind of study it was, who took part, what was measured and what the authors reported. Observational studies show associations, not proof.
People assigned to a diet, food or program and measured.
2002 / Randomized controlled trial
Twenty-eight men with type 2 diabetes and serum ferritin >200 ng/ml were randomized to iron depletion (three 500-ml blood extractions at 2-week intervals) or observation, with vascular reactivity evaluated at baseline, 4 months, and 12 months. Endothelium-dependent vasodilation remained essentially unchanged in both groups. In contrast, vasodilation induced by glyceryl trinitrate (GTN) improved significantly after iron depletion (P = 0.006) alongside decreases in transferrin saturation index and HbA1c (-0.6%, P < 0.05). Changes in endothelium-independent vasodilation correlated with the change in serum ferritin (r = -0.45, P = 0.04), and change in HbA1c was the best predictor of these modifications. At 12 months, transferrin saturation index and GTN-induced vasodilation returned to values similar to baseline.
Fernández-Real JM, Peñarroja G, Castro A, et al. Diabetes Care, 2002. PMID 12453969
2002 / Randomized crossover trial
Ten healthy post-menopausal women completed four two-meal studies in random order. A high-fat breakfast (40 g fat) rich in SFA, n-6 PUFA, long-chain n-3 PUFA, or MUFA was followed 5 hours later by an identical low-fat, high-carbohydrate lunch. Plasma insulin response was significantly higher following the SFA meal than the other meals after both breakfast and lunch (P<0.006), with no effect on glucose concentrations. Postprandial insulin sensitivity was significantly lower after lunch than breakfast for all meals (P=0.019), following the rank order SFA < n-6 PUFA < n-3 PUFA < MUFA.
Robertson MD, Jackson KG, Fielding BA, et al. Br J Nutr, 2002. PMID 12493085
2002 / Pilot, randomized, double-blind study
In a pilot, double-blind study, 36 patients with dyslipidemia and type 2 diabetes on a cholesterol-lowering diet were randomized to receive policosanol (10 mg/day) or lovastatin (20 mg/day) for 8 weeks. Policosanol significantly lowered LDL-C by 29.9%, total cholesterol by 21.1%, triglycerides by 13.6%, and increased HDL-C by 12.5%, outperforming lovastatin in raising HDL-C and reducing lipid ratios (p < 0.05). Policosanol also improved lag time of LDL peroxidation by 20.9% and plasma antioxidant activity by 24.2%. Both treatments reduced lipid peroxidation markers and were well tolerated, though the authors note further studies with larger sample sizes are required.
Castaño G, Menéndez R, Más R, et al. Int J Clin Pharmacol Res, 2002. PMID 12837046
2001 / Randomized controlled trial
In a 3-month trial, 162 healthy adults received isoenergetic diets high in saturated fat or monounsaturated fat, with added fish oil or placebo. Saturated fat impaired insulin sensitivity by 10% (p = 0.03), while monounsaturated fat showed no significant change (+2%, p = 0.05 between diets). Beneficial effects occurred only in participants consuming below the median fat intake (37% of energy), where monounsaturated fat increased sensitivity by 8.8% and saturated fat decreased it by 12.5% (p = 0.03).
Vessby B, Uusitupa M, Hermansen K, et al. Diabetologia, 2001. PMID 11317662
2001 / Randomized controlled trial
A randomized trial assigned 522 overweight adults with impaired glucose tolerance to an individualized diet and exercise intervention or a control group for a mean follow-up of 3.2 years. The intervention group lost significantly more weight than controls at year one (4.2 vs. 0.8 kg) and year two (3.5 vs. 0.8 kg; p < 0.001). The cumulative four-year diabetes incidence was 11% (95% CI, 6-15%) with intervention versus 23% (95% CI, 17-29%) in controls, representing a 58% risk reduction (p < 0.001).
Tuomilehto J, Lindström J, Eriksson JG, et al. N Engl J Med, 2001. PMID 11333990
2001 / Controlled exercise intervention study
Seven subjects with type 2 diabetes and eight matched controls completed 45 minutes of cycle ergometer exercise at 70% of maximum workload, with vastus lateralis biopsies taken before, during, and after exercise. Blood glucose decreased from 7.6 to 4.77 mmol/l with 45 min of exercise in the diabetic group, with no change in controls. Exercise significantly increased AMPK alpha2 activity 2.7-fold over basal at 20 min in both groups, without affecting AMPK alpha1 activity. Protein expression of AMPK alpha1, alpha2, and beta1 in muscle was similar between groups.
Musi N, Fujii N, Hirshman MF, et al. Diabetes, 2001. PMID 11334434
2001 / Cross-sectional comparison with phlebotomy intervention sub-study
Insulin sensitivity and iron status were compared between 30 lacto-ovo vegetarians and 30 meat-eaters, all of whom were lean, healthy, and glucose tolerant. Lacto-ovo vegetarians demonstrated greater insulin sensitivity, with a steady-state plasma glucose of 4.1 mmol/l (95% CI 3.5, 5.0) versus 6.9 mmol/l (95% CI 5.2, 7.5) in meat-eaters. Vegetarians also showed lower body iron stores, with serum ferritin of 35 microg/l (95% CI 21, 49) compared to 72 microg/l (95% CI 45, 100) in meat-eaters. Lowering body iron via phlebotomy in six male meat-eaters to levels similar to vegetarians resulted in an approximately 40% enhancement of insulin-mediated glucose disposal.
Hua NW, Stoohs RA, Facchini FS Br J Nutr, 2001. PMID 11591239
2001 / Experimental meal challenge study
Ten healthy volunteers consumed a high-glycaemic index white bread meal or low-glycaemic index spaghetti meal with milk (200 or 400 ml) or water (400 ml) after an overnight fast. Adding milk did not affect postprandial glucose AUC for either the bread or spaghetti meals. However, insulin AUC was significantly higher when 400 ml milk was added to the bread meal (+65%) and when 200 ml or 400 ml milk was added to the spaghetti meal (+300%) compared with water.
Liljeberg Elmståhl H, Björck I Eur J Clin Nutr, 2001. PMID 11641749
2001 / Experimental infusion and dietary intervention study
Researchers evaluated the impact of a 6-hour lipid infusion during a hyperinsulinemic-euglycemic clamp (n=12 men) and a 3-day high-fat versus low-fat diet (n=12 men) on IMCL and insulin sensitivity. During infusion, IMCL rose significantly after 2 hours, reaching 120.8 +/- 3.4% of baseline in soleus and 164.2 +/- 13.8% in tibialis anterior muscle after 6 hours (both P < 0.05), while glucose infusion rate dropped to 60.4 +/- 5.4% of baseline. Following the high-fat diet, tibialis anterior IMCL increased to 148.0 +/- 16.9% of baseline (P = 0.005) and insulin sensitivity decreased to 83.3 +/- 5.6% (P = 0.033), with no significant changes observed after the low-fat diet.
Bachmann OP, Dahl DB, Brechtel K, et al. Diabetes, 2001. PMID 11679437
2001 / Randomized controlled clinical trial
Seventy-six male CAD patients were randomly assigned to consume either a daily whole-grain meal (whole grain and legume powder replacing refined rice) or a control diet for 16 weeks. In the whole-grain group, serum glucose and insulin decreased by 24% and 14%, respectively, and areas under the curve for glucose and insulin were reduced. Plasma malondialdehyde, homocysteine, and urinary 8-epi-prostaglandin F(2alpha) concentrations each decreased by approximately 28%, while lipid-corrected concentrations of alpha-carotene, retinol, tocopherols, and lycopene increased by 11% to 40%.
Jang Y, Lee JH, Kim OY, et al. Arterioscler Thromb Vasc Biol, 2001. PMID 11742886
2000 / Interventional crossover trial (before-and-after diet switch)
Eleven type 2 diabetic patients switched from their usual linoleic-acid-rich diet to an oleic-acid-rich diet for 2 months. Insulin-stimulated glucose transport in isolated adipocytes was significantly greater on the oleic-acid-rich diet (0.56+/-0.17 vs. 0.29+/-0.14 nmol/10^5 cells/3 min, p<0.0001). Endothelium-dependent flow-mediated vasodilatation (FMD) was also significantly greater on the oleic-acid-rich diet (6.12+/-1.36% vs. 3.90+/-0.97%, p<0.0001). There was a small but significant decrease in fasting glucose/insulin, and adipocyte membrane oleic/linoleic acid ratio correlated positively with endothelium-dependent FMD (r=0.61, p<0.001).
Ryan M, McInerney D, Owens D, et al. QJM, 2000. PMID 10700478
2000 / Acute interventional challenge study
Blood samples were collected from 14 normal subjects prior to and at 1, 2, and 3 hours after ingesting 75 g of glucose. Reactive oxygen species (ROS) generation by polymorphonuclear leucocytes and mononuclear cells increased to a peak of 244 +/- 42% and 233 +/- 34% of basal levels, respectively, at 2 hours. Mononuclear cell p47phox levels rose significantly at 2 and 3 hours, while alpha-tocopherol levels significantly decreased at 1, 2, and 3 hours.
Mohanty P, Hamouda W, Garg R, et al. J Clin Endocrinol Metab, 2000. PMID 10946914
1999 / Euglycemic-hyperinsulinemic clamp trial
In seven healthy men, skeletal muscle metabolism was tracked via 31P nuclear magnetic resonance spectroscopy during euglycemic-hyperinsulinemic clamps under high, basal, and low free fatty acid (FFA) levels. Whole-body glucose uptake was approximately 20% lower after 140 minutes with basal and high FFAs (42.8 +/- 3.6 and 41.6 +/- 3.3 vs. control 52.7 +/- 3.3 micromol x kg(-1) x min(-1), P < 0.05). Intramuscular glucose-6-phosphate (G-6-P) was blunted at 45 minutes of high FFA exposure (184 +/- 17 vs. control: 238 +/- 17 micromol/l, P = 0.008) and remained lower at 180 minutes with high and basal FFAs versus control (P < 0.05). The authors conclude these early G-6-P patterns suggest FFAs primarily inhibit glucose transport/phosphorylation prior to reducing whole-body glucose disposal.
Roden M, Krssak M, Stingl H, et al. Diabetes, 1999. PMID 10334314
1999 / Interventional oral glucose challenge study
Researchers evaluated flow-mediated endothelium-dependent vasodilation, TBARS, and nitrite/nitrate during an oral glucose tolerance test in 58 participants with normal glucose tolerance (NGT), impaired glucose tolerance (IGT), or type 2 diabetes (DM). Flow-mediated vasodilation decreased after glucose loading at 1 and 2 hours in all groups (NGT: 7.53+/-0.40% fasting to 4.24+/-0.28% and 6.35+/-0.40%; IGT: 6.50+/-0.48% to 1.40+/-0.41% and 4.00+/-0.47%; DM: 4.77+/-0.37% to 1.35+/-0.38% and 1.29+/-0.29%). TBARS concentrations increased in parallel with plasma glucose in each group, while nitrite/nitrate concentrations did not change in any group.
Kawano H, Motoyama T, Hirashima O, et al. J Am Coll Cardiol, 1999. PMID 10400004
1998 / Randomized crossover trial
Ten healthy volunteers consumed a white bread reference meal or an identical meal supplemented with vinegar in random order after an overnight fast. The addition of acetic acid as vinegar significantly lowered postprandial glucose (glycemic index = 64) and insulin responses (insulinemic index = 65). Lowered postprandial paracetamol levels suggested that delayed gastric emptying rate is the probable mechanism.
Liljeberg H, Björck I Eur J Clin Nutr, 1998. PMID 9630389
1998 / Uncontrolled community-based lifestyle intervention
The community-based Coronary Health Improvement Project (CHIP) enrolled 304 participants at elevated coronary disease risk in a 30-day, 40-hour educational program encouraging 30 minutes of daily exercise and a largely unrefined plant-food-centered diet. A total of 288 participants graduated. At 4 weeks, overall improvements in laboratory test results, blood pressures, weights, and body mass indexes were highly significant (p <0.001). Triglyceride levels decreased significantly (p <0.05) in participants with elevated baseline triglycerides (>200 mg/dL in men, 200-299 mg/dL in women).
Diehl HA Am J Cardiol, 1998. PMID 9860383
1997 / Clinical intervention study
Researchers evaluated the oral absorption and renal clearance of dietary AGEs in 38 diabetic patients (with or without kidney disease) and 5 healthy controls following a single meal of egg white cooked with fructose (AGE-diet) or without fructose (CL-diet). The AGE-diet distinctly elevated serum AGE levels in direct proportion to the amount ingested (r = 0.8, P < 0.05). Serum area under the curve correlated directly with the severity of kidney disease, while renal excretion of dietary AGE correlated inversely with albuminuria and directly with creatinine clearance (r = 0.8, P < 0.05), dropping to <5% in diabetics with renal failure. Post-meal serum showed increased crosslinking activity (two times above baseline), which was inhibited by aminoguanidine.
Koschinsky T, He CJ, Mitsuhashi T, et al. Proc Natl Acad Sci U S A, 1997. PMID 9177242
1997 / Comparative feeding study
Thirty-eight common foods in 1000-kJ portions were tested in groups of 11-13 healthy subjects to determine postprandial insulin scores relative to white bread (100%). Glucose and insulin scores were strongly correlated overall (r = 0.70, P < 0.001), though protein-rich foods and bakery products produced disproportionately high insulin responses compared to their glycemic responses. Total carbohydrate (r = 0.39, P < 0.05) and sugar (r = 0.36, P < 0.05) contents were positively correlated with insulin scores, while fat (r = -0.27, NS) and protein (r = -0.24, NS) were not significantly related.
Holt SH, Miller JC, Petocz P Am J Clin Nutr, 1997. PMID 9356547
1996 / Controlled crossover metabolic infusion study
In nine healthy subjects under 6-hour euglycemic-hyperinsulinemic clamp conditions, muscle metabolism was monitored via 13C/31P NMR spectroscopy during low or high plasma free fatty acid levels. Lipid infusion decreased whole-body glucose uptake to approximately 46% of control values after 6 hours (P < 0.00001) and caused an approximately 40% reduction in oxidative glucose metabolism starting at hour 3. Muscle glycogen synthesis also decreased to approximately 50% of control after 3 hours (4.0 +/- 1.0 vs. 9.3 +/- 1.6 mumol/[kg.min], P < 0.05). This reduction was preceded by a fall in muscle glucose-6-phosphate concentrations starting at approximately 1.5 hours (195 +/- 25 vs. control 237 +/- 26 mM; P < 0.01), showing lipid-induced insulin resistance starts with inhibition of glucose transport/phosphorylation.
Roden M, Price TB, Perseghin G, et al. J Clin Invest, 1996. PMID 8675698
1996 / Controlled dietary intervention trial
Researchers evaluated the effect of a 7-day avocado-enriched diet (high in monounsaturated fatty acids) compared to an isocaloric control diet in 30 healthy volunteers and 37 patients with mild hypercholesterolemia. In healthy individuals, total cholesterol decreased by 16% on the avocado diet but rose on the control diet (p < 0.001 between diets). In hypercholesterolemic subjects, the avocado diet led to a significant (p < 0.01) decrease in total cholesterol (17%), LDL-cholesterol (22%), and triglycerides (22%), along with an 11% increase in HDL-cholesterol, whereas no significant changes occurred with the control diet.
López Ledesma R, Frati Munari AC, Hernández Domínguez BC, et al. Arch Med Res, 1996. PMID 8987188
1995 / Crossover trial
Eight healthy subjects consumed two equivalent solid meals differing only in naturally present dietary fibre content (high fibre 20 g vs low fibre 4 g per 1000 kcal). Total gastric emptying time was significantly shorter after the low fibre meal compared to the high fibre meal (186.0 [15.6] vs 231.7 [17.3] minutes, p < 0.05). Blood glucose and the area under the glycaemic curve were significantly higher after the low fibre meal (226 [23.1] vs 160 [20.0] mmol/min/dl-1, p < 0.05). Satiety and fullness did not differ, but hunger returned more rapidly after the low fibre meal.
Benini L, Castellani G, Brighenti F, et al. Gut, 1995. PMID 7615267
1995 / Randomized crossover trial
In a randomized crossover trial, five healthy subjects consumed six test meals in random order, with some meals followed by white bread. Adding acetic acid as vinegar depressed the post-meal glucose response by 31.4% (P = 0.0228) compared to neutralized vinegar or olive oil alone, and markedly reduced blood acetate response. In four additional subjects, gastric emptying times did not differ between native and neutralized vinegar meals.
Brighenti F, Castellani G, Benini L, et al. Eur J Clin Nutr, 1995. PMID 7796781
1995 / Within-subject feeding study
Ten fasted healthy men consumed mixed test meals containing 150 g or 250 g of microwaved spinach (providing 4.3 g and 7.2 g fiber) or a spinach-free control meal. The largest spinach portions were associated with higher fullness scores and lower postprandial blood glucose through 3.5 hours. Overall satiety scores correlated positively with vegetable fiber and water content, whereas vegetable structure was not linked to satiety.
Gustafsson K, Asp NG, Hagander B, et al. Int J Food Sci Nutr, 1995. PMID 8574859
1994 / Randomized crossover trial
Six healthy subjects ingested test meals of glucose alone or glucose with 15 g of wheat bran, sugar beet, or ispaghula fibres in a randomized order over 4 days. Duodeno-jejunal motility was evaluated electromyographically alongside glucose and insulin levels for 180 minutes. Sugar beet and ispaghula fibres inhibited stationary contractile activity, increased propagation length and velocity, and reduced the blood glycaemic response, whereas wheat bran had no effect on these measures. Postprandial blood glucose levels were significantly correlated with the total motility index (r = 0.82) and stationary activity (r = 0.79).
Cherbut C, Bruley des Varannes S, Schnee M, et al. Br J Nutr, 1994. PMID 8054323