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.
Patterns in how people eat and what happens to them. Shows an association, not proof.
2016 / Two-sample Mendelian randomization study
Researchers evaluated the effects of genetically predicted coffee consumption on type 2 diabetes mellitus (T2DM), depression, Alzheimer's disease, ischemic heart disease (IHD), and related risk factors using two-sample Mendelian randomization, with childhood cognition as a negative control. Genetically predicted coffee consumption was not associated with T2DM (OR 1.02, 95% CI 0.76 to 1.36), depression (OR 0.89, 95% CI 0.66 to 1.21), Alzheimer's disease (OR 1.17, 95% CI 0.96 to 1.43), IHD (OR 0.96, 95% CI 0.80 to 1.14), lipids, glycemic traits, adiponectin, or childhood cognition. The authors concluded that, contrary to observational findings, coffee may not have beneficial effects on T2DM, depression, or Alzheimer's disease.
Shows an association, not proof of cause and effect.
Kwok MK, Leung GM, Schooling CM Sci Rep, 2016. PMID 27845333
2015 / Cross-sectional analysis
A cross-sectional analysis evaluated cardiovascular risk factors across dietary patterns in 592 black participants from the Adventist Health Study-2. Compared with non-vegetarians, vegetarian/vegans had lower adjusted odds for hypertension (OR 0.56, 95% CI 0.36, 0.87), diabetes (OR 0.48, 95% CI 0.24, 0.98), high total cholesterol (OR 0.42, 95% CI 0.27, 0.65), high LDL-cholesterol (OR 0.54, 95% CI 0.33, 0.89), obesity (OR 0.43, 95% CI 0.28, 0.67), and abdominal obesity (OR 0.54, 95% CI 0.36, 0.82). Pesco-vegetarians also had lower odds of obesity (OR 0.47) and abdominal obesity (OR 0.50). Further adjustment suggested BMI acts as an intermediary variable between diet and hypertension and diabetes, though the authors note cross-sectional analysis cannot establish cause.
Shows an association, not proof of cause and effect.
Fraser G, Katuli S, Anousheh R, et al. Public Health Nutr, 2015. PMID 24636393
2015 / Prospective cohort study
In a prospective cohort of 2,939 mother-child pairs, maternal diet was assessed in the third month of lactation using a validated FFQ. Among children with HLA-conferred susceptibility, 172 developed preclinical and 81 developed clinical type 1 diabetes over an average follow-up of 7.5 to 7.7 years. Maternal intake of fatty acids was not associated with type 1 diabetes. However, maternal total consumption of red meat and meat products was associated with increased risk for preclinical (HR 1.19, 95% CI 1.02-1.40) and clinical type 1 diabetes (HR 1.27, 95% CI 1.06-1.52), driven particularly by processed meat (HR 1.23, 95% CI 1.02-1.48). Maternal vegetable oil use was also associated with increased risk for preclinical diabetes (HR 1.21, 95% CI 1.03-1.41).
Shows an association, not proof of cause and effect.
Niinistö S, Takkinen HM, Uusitalo L, et al. Acta Diabetol, 2015. PMID 25563476
2015 / Prospective cohort study
Researchers analyzed 815 non-shift workers from the Dunedin Longitudinal Study to evaluate whether social jetlag was associated with clinically assessed metabolic phenotypes and obesity-related disease indicators. Social jetlag was associated with multiple clinically assessed measures of metabolic dysfunction and obesity. Higher levels of social jetlag were observed in metabolically unhealthy obese individuals compared to metabolically healthy obese individuals. Among metabolically unhealthy obese individuals, social jetlag was additionally linked to elevated glycated hemoglobin and an indicator of inflammation.
Shows an association, not proof of cause and effect.
Parsons MJ, Moffitt TE, Gregory AM, et al. Int J Obes (Lond), 2015. PMID 25601363
2015 / Cross-sectional survey
A cross-sectional study evaluated awareness, barriers, and promoters of plant-based diets among 98 patients and 25 healthcare providers at a regional diabetes clinic. While only 9% of patients currently followed a plant-based diet, 66% reported willingness to follow one for 3 weeks. Common barriers included family eating preferences and meal planning skills. Additionally, 72% of healthcare providers reported knowledge of plant-based diets for diabetes management, though practice levels were low.
Shows an association, not proof of cause and effect.
Lee V, McKay T, Ardern CI J Nutr Metab, 2015. PMID 25802755
2015 / Prospective cohort study
Researchers prospectively followed 367,442 participants from the NIH-AARP Diet and Health Study for an average of 14 years, documenting 46,067 deaths. Compared to those with the lowest intake, participants with the highest intake of whole grains had a 17% (95% CI, 14-19%) lower risk of all-cause mortality and an 11-48% lower risk of disease-specific mortality. Highest intake of cereal fiber was associated with a 19% (95% CI, 16-21%) lower risk of all-cause mortality and a 15-34% lower risk of disease-specific mortality. When adjusted for cereal fiber, associations of whole grains with CVD, respiratory disease, and infection mortality became non-significant, while associations with all-cause, cancer, and diabetes mortality were attenuated but remained significant.
Shows an association, not proof of cause and effect.
Huang T, Xu M, Lee A, et al. BMC Med, 2015. PMID 25858689
2015 / Cross-sectional analysis of survey data
Researchers analyzed 2009-2010 NHANES data to evaluate associations between nighttime fasting duration (calculated from 24-hour food records) and glycemic biomarkers among women. Each 3-hour increase in nighttime fasting was associated with a 4% lower 2-hour glucose measurement (β, 0.96; 95% CI, 0.93-1.00; P < 0.05) and roughly a 20% lower odds of elevated HbA1c (OR, 0.81; 95% CI, 0.68-0.97; P < 0.05). Associations with continuous HbA1c and odds of elevated 2-hour glucose were not statistically significant.
Shows an association, not proof of cause and effect.
Marinac CR, Natarajan L, Sears DD, et al. Cancer Epidemiol Biomarkers Prev, 2015. PMID 25896523
2015 / Prospective cohort study
In the EPIC-Norfolk study, beverage consumption among 25,639 UK adults was assessed via 7-day food diaries, with 847 incident type 2 diabetes cases verified over 10.8 years of follow-up. Positive associations per serving/day were found for soft drinks (HR 1.21; 95% CI 1.05, 1.39), sweetened-milk beverages (HR 1.22; 95% CI 1.05, 1.43), and artificially sweetened beverages (ASB: HR 1.22; 95% CI 1.11, 1.33), though the ASB association attenuated after adjusting for adiposity (HR 1.06; 95% CI 0.93, 1.20). No associations were observed for sweetened tea/coffee or fruit juice. Substituting one serving/day of water or unsweetened tea/coffee for soft drinks or sweetened-milk beverages reduced incidence by 14% to 25%, while substituting ASB for SSB did not reduce incidence when accounting for energy intake and adiposity.
Shows an association, not proof of cause and effect.
O'Connor L, Imamura F, Lentjes MA, et al. Diabetologia, 2015. PMID 25944371
2015 / Mendelian randomization and observational cohort study
Researchers analyzed 93,179 individuals from two general population cohorts and up to 78,021 additional individuals to assess observational and genetic relationships between coffee intake and metabolic outcomes. Observationally, high coffee intake was associated with a lower risk of obesity, metabolic syndrome, and type 2 diabetes. However, genetically derived high coffee intake (via variants near CYP1A1, CYP1A2, and AHR) was not convincingly associated with obesity, metabolic syndrome, type 2 diabetes, or related metabolic markers. Per-allele meta-analyzed odds ratios for type 2 diabetes ranged from 0.98 to 1.01 across the evaluated variants.
Shows an association, not proof of cause and effect.
Nordestgaard AT, Thomsen M, Nordestgaard BG Int J Epidemiol, 2015. PMID 26002927
2015 / Comparative risk assessment modeling study
Researchers modeled global, regional, and national burdens of disease related to sugar-sweetened beverage (SSB) intake in 2010 using pooled dietary surveys, cohort effect estimates, and GBD 2010 mortality data. Worldwide, an estimated 184,000 deaths per year (95% UI, 161,000-208,000) were attributable to SSBs: 133,000 from diabetes mellitus, 45,000 from CVD, and 6,450 from cancers. Middle-income nations accounted for 70.9% of these deaths, high-income for 24.1%, and low-income for 5.0%. Mexico had the highest absolute (405 deaths/million adults) and proportional (12.1%) deaths among the 20 most populous nations. Additionally, 8.5 million DALYs were associated with SSB consumption.
Shows an association, not proof of cause and effect.
Singh GM, Micha R, Khatibzadeh S, et al. Circulation, 2015. PMID 26124185
2015 / Retrospective cohort study
Researchers examined non-cancer mortality from 1950 to 2006 among 10,701 workers in supermarket meat and delicatessen departments who had occupational exposure to animal microorganisms and wrapping machine fumes. Compared with the US population, the cohort or specific race/sex subgroups showed significantly increased deaths from diabetes, ischemic heart disease, pulmonary embolism, chronic bronchitis, peritonitis, intracranial and intraspinal abscess, and other bacterial diseases. Conversely, significantly decreased deaths were observed for diffuse connective tissue diseases, functional heart diseases, intracerebral hemorrhage, occlusion/stenosis of precerebral/cerebral arteries, and various accidents.
Shows an association, not proof of cause and effect.
Jadhav S, Chedjieu IP, Faramawi MF, et al. Environ Res, 2015. PMID 26160045
2015 / Cross-sectional online survey
A cross-sectional online survey evaluated plain water intake and type 2 diabetes (T2D) risk in 138 adults from England. Differences in mean plain water intake across T2D risk categories showed a non-significant trend (P = .084). However, plain water intake was significantly negatively correlated with T2D risk score (τ = -0.180, P = .005), with each 240-mL cup of water consumed per day linked to a 0.72-point reduction in risk score (B = -0.03, 95% CI = -0.06 to -0.01, P = .014).
Shows an association, not proof of cause and effect.
Carroll HA, Davis MG, Papadaki A Nutr Res, 2015. PMID 26255759
2015 / Prospective cohort study
This longitudinal study evaluated 1,141 participants from the Tehran Lipid and Glucose Study to examine the link between baseline dietary phytochemical index (DPI), assessed via a 168-item semi-quantitative FFQ, and insulin-related outcomes. After 3 years of follow-up, the risk of hyperinsulinemia significantly decreased by 65% (OR = 0.35, 95% CI = 0.21-0.60) and 86% (OR = 0.14, 95% CI = 0.07-0.29) in the third and fourth quartiles of DPI, respectively. Higher DPI was also associated with a lower occurrence of insulin resistance (OR = 0.48, 95% CI = 0.25-0.93) and insulin insensitivity (OR = 0.11, 95% CI = 0.05-0.24).
Shows an association, not proof of cause and effect.
Bahadoran Z, Mirmiran P, Tohidi M, et al. Int J Food Sci Nutr, 2015. PMID 26600067
2015 / Prospective cohort study
In a prospective cohort of 90,317 US adults followed for 805,644 person-years, 8,718 deaths occurred. Compared with nondrinkers, hazard ratios for overall mortality were 0.99 (95% CI: 0.92, 1.07) for <1 cup/day, 0.94 (95% CI: 0.87, 1.02) for 1 cup/day, 0.82 (95% CI: 0.77, 0.88) for 2-3 cups/day, 0.79 (95% CI: 0.72, 0.86) for 4-5 cups/day, and 0.84 (95% CI: 0.75, 0.95) for ≥6 cups/day. Similar findings were observed for decaffeinated coffee and coffee additives. Inverse associations were seen for deaths from heart disease, chronic respiratory diseases, diabetes, pneumonia and influenza, and intentional self-harm, but not cancer.
Shows an association, not proof of cause and effect.
Loftfield E, Freedman ND, Graubard BI, et al. Am J Epidemiol, 2015. PMID 26614599
2015 / Case-control study
A case-control study compared 464 consecutive NAFLD patients with 181 controls in India using lifestyle/dietary questionnaires, anthropometrics, and biochemical assays. NAFLD patients had significantly higher BMI (26.25 ± 3.80 vs 21.46 ± 3.08 kg/m2, P = 0.000), fasting blood sugar, triglycerides, and HOMA-IR (2.53 ± 2.57 vs 1.16 ± 0.58, P = 0.000). Dietary factors linked to NAFLD included a non-vegetarian diet (35% vs 23%, P = 0.002), fried food (35% vs 9%, P = 0.000), spicy foods (51% vs 15%, P = 0.001), and tea (55% vs 39%, P = 0.001). A majority (90.2%) of NAFLD patients were sedentary, and family history of metabolic syndrome, diabetes, hypertension, snoring, and sleep apnea were also associated.
Shows an association, not proof of cause and effect.
Singh SP, Singh A, Misra D, et al. J Clin Exp Hepatol, 2015. PMID 26900270
2014 / Case-control and cross-sectional study
Researchers evaluated 73 adults with newly diagnosed NAFLD and compared 58 of them with 58 matched healthy controls to explore the role of adherence to the Mediterranean diet (MedDietScore). MedDietScore did not differ between NAFLD patients and controls, but among patients, it was negatively correlated with serum alanine aminotransferase (p = 0.03), insulin levels (p = 0.001), insulin resistance index (p = 0.005), and steatosis severity (p = 0.006), and positively with adiponectin levels (p = 0.04). Patients with NASH had lower adherence than those with simple fatty liver (29.3 ± 3.2 vs. 34.1 ± 4.4, p = 0.004). After adjustment for sex and abdominal fat level, each one-unit increase in MedDietScore was associated with a 36% lower likelihood of having NASH (OR: 0.64, 95% CI: 0.45-0.92).
Shows an association, not proof of cause and effect.
NANO counterpoint →
Kontogianni MD, Tileli N, Margariti A, et al. Clin Nutr, 2014. PMID 24064253
2014 / Cross-sectional study
In a cross-sectional study of 1997 women aged 18-76 years, habitual intakes of flavonoid subclasses were assessed via food frequency questionnaires and evaluated against biomarkers of insulin resistance and inflammation. Higher intakes of anthocyanins and flavones were associated with significantly lower HOMA-IR (Q5 to Q1 = -0.1, P-trend = 0.04 for both) and decreased insulin concentrations (Q5-Q1 = -0.7 μU/mL, P-trend = 0.02 for anthocyanins; Q5-Q1 = -0.5 μU/mL, P-trend = 0.02 for flavones). Higher anthocyanin intake was linked to lower hs-CRP (Q5-Q1 = -0.3 mg/L, P-trend = 0.04), while higher flavone intake was linked to improved adiponectin (Q5-Q1 = 0.7 μg/L, P-trend = 0.01). No significant associations were observed for total flavonoids or other subclasses.
Shows an association, not proof of cause and effect.
Jennings A, Welch AA, Spector T, et al. J Nutr, 2014. PMID 24336456
2014 / Cross-sectional study
Researchers examined whether adherence to 'Blood-Type' diets was associated with cardiometabolic biomarkers and if ABO genotype modified these associations in 1,455 participants. Adherence to the Type-A diet was associated with lower BMI, waist circumference, blood pressure, serum cholesterol, triglycerides, insulin, HOMA-IR, and HOMA-Beta (P<0.05). The Type-AB diet was linked to lower levels of these biomarkers (P<0.05) except BMI and waist circumference, while the Type-O diet was associated with lower triglycerides (P<0.0001); no significant associations were found for the Type-B diet. Matching the diets to an individual's corresponding blood group did not change effect sizes, failing to support the 'Blood-Type' diet hypothesis.
Shows an association, not proof of cause and effect.
Wang J, García-Bailo B, Nielsen DE, et al. PLoS One, 2014. PMID 24454746
2014 / Cross-sectional study
The association between diet and diabetes or impaired fasting glucose (IFG) was evaluated in 4,384 Taiwanese Buddhist volunteers. Crude diabetes prevalence in vegetarians versus omnivores was 0.6% vs 2.3% in pre-menopausal women, 2.8% vs 10% in menopausal women, and 4.3% vs 8.1% in men. After multivariable adjustment, a vegetarian diet was negatively associated with diabetes in men (OR 0.49, 95% CI: 0.28-0.89) and menopausal women (OR 0.25, 95% CI: 0.15-0.42), and with IFG in men (OR 0.66, 95% CI: 0.46-0.95) and menopausal women (OR 0.73, 95% CI: 0.56-0.95). In pre-menopausal women, adjusted ORs were 0.26 (95% CI: 0.06-1.21) for diabetes and 0.60 (95% CI: 0.35-1.04) for IFG.
Shows an association, not proof of cause and effect.
Chiu TH, Huang HY, Chiu YF, et al. PLoS One, 2014. PMID 24523914
2014 / Pilot cohort study
This pilot cohort study examined the Complete Health Improvement Program (CHIP), a community-based, comprehensive lifestyle modification approach, in an underserved, rural, and vulnerable Appalachian population. Two hundred fourteen participants in CHIP collectively demonstrated significant reductions in body mass index, systolic and diastolic blood pressure, and fasting blood levels of total cholesterol, low-density lipoprotein, and glucose.
Shows an association, not proof of cause and effect.
Drozek D, Diehl H, Nakazawa M, et al. Adv Prev Med, 2014. PMID 24527219
2014 / Prospective cohort study and animal experiments
Researchers examined links between protein intake and mortality in humans over an 18-year follow-up, alongside mouse models. Respondents aged 50-65 reporting high protein intake had a 75% increase in overall mortality and a 4-fold increase in cancer death risk, associations that were abolished or attenuated when proteins were plant-derived. In contrast, high protein intake was associated with reduced cancer and overall mortality in respondents over 65, though it was linked to a 5-fold increase in diabetes mortality across all ages. Mouse studies supported detrimental effects of high protein intake on tumors and detrimental effects of low protein in the very old.
Shows an association, not proof of cause and effect.
Levine ME, Suarez JA, Brandhorst S, et al. Cell Metab, 2014. PMID 24606898
2014 / Case-cohort study
The InterAct case-cohort study examined combined effects of an additive genetic type 2 diabetes (T2D) risk score and lifestyle factors among 12,403 incident cases and a 16,154-person sub-cohort. Relative genetic risk was significantly greater in younger individuals (p = 1.20×10^-4) and leaner participants by BMI (p = 1.50×10^-3) and waist circumference (p = 7.49×10^-9). Across extreme quartiles of genetic score, 10-year cumulative T2D incidence rose from 0.25% to 0.89% in normal weight individuals versus 4.22% to 7.99% in obese individuals. No significant interactions were found between the genetic score and sex, family history, physical activity, or a Mediterranean diet score.
Shows an association, not proof of cause and effect.
Langenberg C, Sharp SJ, Franks PW, et al. PLoS Med, 2014. PMID 24845081
2014 / Repeated measures observational study
Researchers evaluated diurnal variations in fatty acid metabolism at breakfast (7:00 am) and dinner (7:00 pm) in 13 metabolically normal, overweight women. Plasma free fatty acid (FFA) concentration was ~30% higher just before dinner than breakfast (P < .05) and remained higher post-dinner (FFA AUC: 0.88 ± 0.33 vs 0.51 ± 0.09 μmol/mL × 4 h, P = .001), despite no morning-evening difference in adipose lipolytic activity. Skeletal muscle expression of genes regulating fatty acid oxidation was 38% to 82% lower, while genes involved in de novo lipogenesis were 51% to 87% higher before dinner than before breakfast (all P < .05), accompanied by variation in core clock genes.
Shows an association, not proof of cause and effect.
Yoshino J, Almeda-Valdes P, Patterson BW, et al. J Clin Endocrinol Metab, 2014. PMID 24878055
2014 / Simulation modeling study using nationally representative data
Using nationally representative data and models among over 19,000 SNAP participants, researchers examined two proposed SNAP policies: banning sugar-sweetened beverage purchases and providing a 30% subsidy on fruits and vegetables. The ban on sugar-sweetened beverages was projected to significantly reduce obesity prevalence and type 2 diabetes incidence, especially among adults ages 18-65 and certain racial and ethnic minorities. In contrast, the fruit and vegetable subsidy was not projected to have a significant effect on obesity and type 2 diabetes, but could more than double the proportion of SNAP participants meeting federal fruit and vegetable guidelines.
Shows an association, not proof of cause and effect.
Basu S, Seligman HK, Gardner C, et al. Health Aff (Millwood), 2014. PMID 24889953