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.
2006 / Meta-analysis of randomized controlled trials
A meta-analysis of 41 randomized controlled trials evaluated the effect of isolated soy protein supplementation on serum lipid levels in adults. Soy protein supplementation was associated with a significant reduction in mean serum total cholesterol (-5.26 mg/dl, 95% CI -7.14 to -3.38), LDL cholesterol (-4.25 mg/dl, 95% CI -6.00 to -2.50), and triglycerides (-6.26 mg/dl, 95% CI -9.14 to -3.38), alongside an increase in HDL cholesterol (0.77 mg/dl, 95% CI 0.20 to 1.34). Meta-regression analyses showed a dose-response relation between soy protein and isoflavone supplementation and net changes in serum lipids.
Reynolds K, Chin A, Lees KA, et al. Am J Cardiol, 2006. PMID 16923451
2006 / Meta-analysis of cohort studies
A meta-analysis of nine cohort studies examined 91,379 men and 129,701 women with 5,007 coronary events. Each extra daily serving of combined fruits and vegetables was linked to a 4% lower risk of coronary heart disease (RR 0.96, 95% CI 0.93-0.99), and fruits alone to a 7% lower risk (RR 0.93, 95% CI 0.89-0.96). Vegetable intake showed heterogeneous associations, while possible publication bias suggested relative risks might be overestimated.
Shows an association, not proof of cause and effect.
Dauchet L, Amouyel P, Hercberg S, et al. J Nutr, 2006. PMID 16988131
2006 / Systematic review and meta-analysis of randomized controlled trials
A systematic review of 43 randomized controlled trials evaluated exercise interventions for weight loss in 3,476 adults with overweight or obesity. Exercise combined with diet resulted in greater weight loss than diet alone (WMD -1.1 kg; 95% CI -1.5 to -0.6), and increasing exercise intensity increased weight loss (WMD -1.5 kg; 95% CI -2.3 to -0.7). As a sole intervention, exercise significantly reduced diastolic blood pressure (WMD -2 mmHg), triglycerides (WMD -0.2 mmol/L), and fasting glucose (WMD -0.2 mmol/L). Significant heterogeneity limited pooling across some analyses, and no data were identified on adverse events, quality of life, morbidity, costs, or mortality.
Shaw K, Gennat H, O'Rourke P, et al. Cochrane Database Syst Rev, 2006. PMID 17054187
2005 / Systematic review and meta-analysis of randomized controlled trials
A meta-analysis of 52 randomized controlled trials (4596 patients) evaluated the efficacy and safety of plant sterols and stanols versus policosanol. Weighted percent changes in LDL were -11.0% for plant sterol/stanol esters (3.4 g/day) versus -2.3% for placebo, compared with -23.7% for policosanol (12 mg/day) versus -0.11% for placebo (both p<0.0001 vs placebo). The net LDL reduction relative to placebo was significantly greater with policosanol than plant sterols and stanols (-24% vs -10%, p<0.0001). Policosanol also improved total cholesterol, HDL, triglycerides, and the LDL:HDL ratio more favorably.
Chen JT, Wesley R, Shamburek RD, et al. Pharmacotherapy, 2005. PMID 15767233
2004 / Meta-analysis of randomized controlled trials
A meta-analysis evaluated 16 randomized controlled trials with parallel or crossover designs published between 1981 and 2003 examining low-GI versus high-GI diets. Low-GI diets significantly reduced fructosamine by -0.1 mmol/l (95% CI -0.20, 0.00; P=0.05), HbA1c by 0.27% (95% CI -0.5, -0.03; P=0.03), and total cholesterol by -0.33 mmol/l (95% CI -0.47, -0.18; P<0.0001). They also tended to reduce LDL-cholesterol in type 2 diabetic subjects by -0.15 mmol/l (95% CI -0.31, -0.00; P=0.06), with no changes observed in HDL-cholesterol and triacylglycerol concentrations.
Opperman AM, Venter CS, Oosthuizen W, et al. Br J Nutr, 2004. PMID 15469640
2003 / Meta-analysis of controlled trials
A meta-analysis of 60 controlled trials assessed how dietary fats and carbohydrates affect lipid markers. Substituting saturated fats with cis unsaturated fatty acids decreased the ratio of total to HDL cholesterol, whereas carbohydrate substitution caused no change. Replacing trans fatty acids with cis unsaturated fats and carbohydrates produced nearly double the reduction seen with saturated fat replacement, while replacing fats with carbohydrates raised fasting triacylglycerols.
Mensink RP, Zock PL, Kester AD, et al. Am J Clin Nutr, 2003. PMID 12716665
2002 / Meta-analysis of prospective cohorts
In a collaborative meta-analysis of individual data from one million adults across 61 prospective studies (12.7 million person-years), about 56,000 vascular deaths occurred. Within each decade of age, each 20 mm Hg lower usual systolic blood pressure (or 10 mm Hg diastolic) was linked to more than a twofold difference in stroke, ischemic heart disease, and other vascular death rates at ages 40-69 years. These proportional risk differences remained consistent down to at least 115/75 mm Hg.
Shows an association, not proof of cause and effect.
Lewington S, Clarke R, Qizilbash N, et al. Lancet, 2002. PMID 12493255
2001 / Meta-analysis of randomized controlled trials
A meta-analysis evaluated 14 randomized clinical trials examining coffee intake and serum lipid levels. A dose-response relationship was observed between coffee intake and increases in both total cholesterol and LDL cholesterol (p < 0.01). Larger lipid increases occurred in hyperlipidemic participants and trials using boiled or caffeinated coffee, whereas filtered coffee trials demonstrated minimal impact.
Jee SH, He J, Appel LJ, et al. Am J Epidemiol, 2001. PMID 11207153
2001 / Meta-analysis of controlled trials
A meta-analysis analyzed 17 controlled trials with 556 participants to evaluate the effect of dietary cholesterol on blood lipids. An additional 100 mg of dietary cholesterol daily increased the ratio of total to HDL cholesterol by 0.020 units (95% CI: 0.010 to 0.030). It also increased total cholesterol by 0.056 mmol/L (95% CI: 0.046 to 0.065) and HDL cholesterol by 0.008 mmol/L (95% CI: 0.005 to 0.010).
Weggemans RM, Zock PL, Katan MB Am J Clin Nutr, 2001. PMID 11333841
2001 / Systematic review and meta-analysis of randomized controlled trials
A systematic review evaluated 27 randomized controlled trials encompassing 30,901 person-years to determine the impact of dietary fat reduction or modification lasting at least 6 months. Dietary fat interventions did not significantly affect overall mortality (rate ratio 0.98, 95% CI: 0.86 to 1.12) or cardiovascular mortality (rate ratio 0.91, 95% CI: 0.77 to 1.07). Cardiovascular events were significantly lower (rate ratio 0.84, 95% CI: 0.72 to 0.99), though this became non-significant in sensitivity analysis.
Hooper L, Summerbell CD, Higgins JP, et al. Cochrane Database Syst Rev, 2001. PMID 11687015
2000 / Systematic review and meta-analysis of randomized placebo-controlled trials
A systematic review and meta-analysis evaluated 18 randomized placebo-controlled trials comprising 823 participants with type 2 diabetes receiving fish oil doses of 3 to 18 g/day for a mean of 12 weeks. Pooled data showed fish oil significantly lowered triglycerides (-0.56 mmol/l [95% CI -0.71 to -0.41]) and raised LDL cholesterol (0.21 mmol/l [0.02 to 0.41]). No statistically significant effect was observed for fasting glucose, HbA1c, total cholesterol, or HDL cholesterol. Effects on triglycerides and LDL cholesterol were most pronounced in trials using higher doses and enrolling hypertriglyceridemic subjects, which also explained observed heterogeneity.
Montori VM, Farmer A, Wollan PC, et al. Diabetes Care, 2000. PMID 10977042
1998 / Collaborative analysis of prospective cohorts
In a pooled analysis of 76,172 men and women followed for a mean of 10.6 years, 8,330 deaths occurred before age 90. Vegetarians had a 24% lower mortality from ischaemic heart disease compared to non-vegetarians (death rate ratio 0.76, 95% CI 0.62-0.94), with stronger inverse associations observed at younger ages. No significant differences were observed between vegetarians and non-vegetarians for other causes of death.
Shows an association, not proof of cause and effect.
Key TJ, Fraser GE, Thorogood M, et al. Public Health Nutr, 1998. PMID 10555529
1998 / Systematic review and meta-analysis of randomized controlled trials
A systematic review of 19 randomized controlled trials evaluated the efficacy of dietary advice to modify fat intake in free-living subjects. Dietary guidance reduced blood total cholesterol by 5.3% (95% CI 4.7% to 5.9%) after at least six months of intervention. Diets matching the American Heart Association Step 2 approach lowered total cholesterol more effectively than Step 1 diets (6.1% versus 3.0%), though target dietary changes were rarely achieved.
Tang JL, Armitage JM, Lancaster T, et al. BMJ, 1998. PMID 9552999
1997 / Meta-analysis of metabolic ward studies
A meta-analysis evaluated 395 dietary experiments (median duration 1 month) among 129 groups of healthy volunteers from metabolic ward studies on solid food diets. Replacing saturated fats with complex carbohydrates for 10% of dietary calories decreased total cholesterol by 0.52 (SE 0.03) mmol/l and LDL cholesterol by 0.36 (0.05) mmol/l. Replacing complex carbohydrates with polyunsaturated fats for 5% of calories further lowered total cholesterol by 0.13 (0.02) mmol/l and LDL cholesterol by 0.11 (0.02) mmol/l, while monounsaturated fats had no significant effect. Avoiding 200 mg/day of dietary cholesterol further reduced total cholesterol by 0.13 (0.02) mmol/l and LDL cholesterol by 0.10 (0.02) mmol/l.
Clarke R, Frost C, Collins R, et al. BMJ, 1997. PMID 9006469
1997 / Meta-analysis of dietary intervention studies
A meta-analysis analyzed 224 intervention studies published between 1966 and 1994, encompassing 8,143 subjects and 878 diet-blood lipid comparisons, using weighted multiple-regression analysis. Regression models for total cholesterol, triacylglycerol, LDL, HDL, and VLDL cholesterol showed multiple correlations of 0.74, 0.65, 0.41, 0.14, and 0.34, respectively. Study characteristics, initial intakes, and baseline concentrations had little effect on these models. Predictions indicated that following recommendations of 30% energy from fat, <10% saturated fat, and <300 mg cholesterol/day will lower plasma total and LDL-cholesterol by approximately 5% compared to an average American diet.
Howell WH, McNamara DJ, Tosca MA, et al. Am J Clin Nutr, 1997. PMID 9174470
1992 / Meta-analysis
A meta-analysis of 70 studies evaluated the influence of diet-induced weight loss on lipid concentrations. Weight loss correlated significantly with declines in total cholesterol (r = 0.32), LDL-C (r = 0.29), VLDL-C (r = 0.38), and triglycerides (r = 0.32; all P <= 0.05). Each kilogram of weight lost increased HDL-C by 0.009 mmol/L during weight stability but decreased it by 0.007 mmol/L during active weight loss.
Dattilo AM, Kris-Etherton PM Am J Clin Nutr, 1992. PMID 1386186
1992 / Meta-analysis of controlled trials
Researchers analyzed 76 data points from 27 controlled feeding trials supplying diets through metabolic kitchens. A nonlinear regression model accounting for baseline intake alongside added dietary cholesterol yielded a good fit to serum cholesterol changes (P < 0.0005, r = 0.617 between observed and predicted values). The mathematical equation described a hyperbolic shape for the relationship between added intake and serum cholesterol response.
Hopkins PN Am J Clin Nutr, 1992. PMID 1534437
1991 / Analysis of crossover and randomized controlled trials
An analysis assessed 68 crossover trials and 10 randomized trials of dietary salt reduction. Reductions in systolic blood pressure closely matched population predictions in 33 trials lasting five weeks or longer, whereas shorter trials showed smaller decreases. In adults aged 50 to 59 years, lowering daily sodium by 50 mmol was estimated to decrease systolic pressure by 5 mm Hg overall and 7 mm Hg in hypertensive individuals.
Law MR, Frost CD, Wald NJ BMJ, 1991. PMID 1827353
2025 / Global Burden of Disease modeling and trend analysis
Using the Global Burden of Disease 2021 database covering 204 countries and territories, researchers analyzed lower extremity peripheral arterial disease (LE-PAD) burden from 1990 to 2021. From 1990 to 2021, global LE-PAD prevalence and incidence increased significantly, with higher prevalence in females and higher mortality in males under 75. High processed meat and low whole-grain intake were major dietary contributors to the disease burden. Population growth drove the increased burden, which was partially mitigated by epidemiological changes.
Shows an association, not proof of cause and effect.
He Y, Zhou M, Tang J, et al. J Health Popul Nutr, 2025. PMID 41063331
2024 / Prospective cohort study
Researchers examined the association between food allergen-specific IgE sensitization and cardiovascular mortality ascertained to 2019 in NHANES 2005-2006 and the MESA cohort. In NHANES, sensitization to at least 1 food was associated with higher cardiovascular mortality (HR, 1.7 [95% CI, 1.2-2.4], P = .005). Milk sensitization was specifically associated with cardiovascular mortality in NHANES (HR, 2.0 [95% CI, 1.1-3.8], P = .026) and replicated in MESA (HR, 3.8 [95% CI, 1.6-9.1], P = .003). Restricting NHANES analyses to consumers of the relevant allergen strengthened associations and revealed shrimp and peanut sensitization as additional risk factors for cardiovascular mortality.
Shows an association, not proof of cause and effect.
Keet C, McGowan EC, Jacobs D, et al. J Allergy Clin Immunol, 2024. PMID 37943208
2024 / Prospective cohort study
Researchers followed 74,563 women and 39,501 men over 31-34 median years of follow-up, documenting 48,193 deaths. Compared to the lowest quarter, the highest quarter of ultra-processed food intake was associated with a 4% higher all-cause mortality (HR 1.04, 95% CI 1.01 to 1.07) and 9% higher other-cause mortality (HR 1.09, 95% CI 1.05 to 1.13), with no associations found for cancer or cardiovascular mortality. Meat/poultry/seafood ready-to-eat products, sweetened beverages, dairy desserts, and ultra-processed breakfast foods were also linked to higher mortality. No consistent associations were observed within quarters of dietary quality assessed by the AHEI-2010 score, while better dietary quality showed an inverse association with mortality across all quarters of ultra-processed food intake.
Shows an association, not proof of cause and effect.
Fang Z, Rossato SL, Hang D, et al. BMJ, 2024. PMID 38719536
2024 / Prospective cohort study
Researchers examined the association between ultra-processed food consumption and incident hypertension in 8,923 middle-aged US adults over a median follow-up of 13 years, during which 79% developed hypertension. Participants in the highest quartile of ultra-processed food intake had a 15% higher risk of incident hypertension compared to the lowest quartile (HR, 1.15 [95% CI, 1.08-1.23]). Highest consumption quartiles for sugar-sweetened beverages, red and processed meat, and margarine were also associated with higher risks (HR 1.16, 1.10, and 1.06, respectively). Conversely, each additional serving of minimally or unprocessed food was linked to a 2% lower risk of incident hypertension (HR, 0.98 [95% CI, 0.98, 0.99]).
Shows an association, not proof of cause and effect.
Rivera N, Du S, Bernard L, et al. J Am Heart Assoc, 2024. PMID 39189486
2024 / Prospective cohort study
Data from 126,842 UK Biobank participants (aged 40-69 years) followed for a median of 9 years were analyzed to examine CVD risk by food origin (plant vs non-plant/animal) and ultra-processing level. Every 10 percentage points increase in plant-sourced non-ultra-processed food (non-UPF) intake was associated with a 7% lower CVD risk (95% CI 0.91-0.95) and a 13% lower CVD mortality risk (95% CI 0.80-0.94). Conversely, every 10 percentage points increase in plant-sourced UPF was associated with a 5% higher CVD risk (95% CI 1.03-1.07) and 12% higher CVD mortality (95% CI 1.05-1.20). Overall UPF intake was linked to higher CVD risk and mortality, while total plant-sourced food contribution showed no evidence of association.
Shows an association, not proof of cause and effect.
Rauber F, Laura da Costa Louzada M, Chang K, et al. Lancet Reg Health Eur, 2024. PMID 39210945
2023 / Prospective cohort study
In the EPIC prospective cohort study of 266,666 participants free of cancer, cardiovascular disease, and type 2 diabetes at baseline, diet was assessed via food-frequency questionnaires and categorized by the Nova classification. Over a median of 11.2 years, 4,461 participants developed multimorbidity of cancer and cardiometabolic diseases. Higher ultra-processed food (UPF) consumption (per 1 SD increment, ~260 g/day without alcoholic drinks) was associated with an increased risk of multimorbidity (HR: 1.09, 95% CI: 1.05, 1.12). Among subgroups, associations were notable for animal-based products (HR: 1.09, 95% CI: 1.05, 1.12) and artificially and sugar-sweetened beverages (HR: 1.09, 95% CI: 1.06, 1.12), while ultra-processed breads and cereals and plant-based alternatives showed no increased risk.
Shows an association, not proof of cause and effect.
Cordova R, Viallon V, Fontvieille E, et al. Lancet Reg Health Eur, 2023. PMID 38115963