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.
2013 / Case-control study
A case-control study assessed the association between dietary behaviours and colorectal cancer (CRC) in 250 cases and 250 matched controls in Greece. More daily meals were associated with lower CRC likelihood (OR = 0.74, 95% CI 0.61, 0.89), while coffee intake (OR = 3.27, 95% CI 1.09, 9.8) and non-stick cookware use (OR = 1.57, 95% CI 1.02, 2.4) were positively associated with CRC; however, these associations slightly lost significance after accounting for Mediterranean diet adherence. A 1/75 increase in the modified-MedDietScore combined with these nutritional behaviours was linked to 13% lower odds of CRC (95% CI 0.83, 0.91, P < 0.001).
Shows an association, not proof of cause and effect.
Kontou N, Psaltopoulou T, Soupos N, et al. Public Health Nutr, 2013. PMID 22874008
2013 / Cross-sectional study
Researchers analyzed 138 subjects with erosive esophagitis (EE), 148 with non-erosive reflux disease (NERD), and 565 controls based on FSSG scores and endoscopy findings. Factors significantly linked to NERD versus controls were egg intake (OR 1.89, 95% CI 1.01-3.50), sleep shortage (OR 2.44, 95% CI 1.54-3.88), and strong psychological stress (OR 1.77, 95% CI 1.18-2.62). In men, current smoking was associated with both NERD and EE. Among subjects with hiatal hernia, alcohol >200 kcal/day (OR 3.99) and eating a large quantity of food at supper (OR 7.85) were risk factors for EE, while a large supper was more common in NERD than EE (OR 2.09).
Shows an association, not proof of cause and effect.
Matsuki N, Fujita T, Watanabe N, et al. J Gastroenterol, 2013. PMID 22911169
2013 / Prospective cohort study
Researchers examined the association between dietary patterns (non-vegetarian, lacto, pesco, vegan, and semi-vegetarian) and cancer incidence among 69,120 participants in the Adventist Health Study-2, identifying 2,939 incident cases. Overall, vegetarians had a lower risk of overall cancer compared with non-vegetarians (multivariate HR, 0.92; 95% CI, 0.85-0.99) and lower risk of gastrointestinal tract cancers (HR, 0.76; 95% CI, 0.63-0.90). Vegan diets were associated with lower overall cancer incidence (HR, 0.84; 95% CI, 0.72-0.99) and lower female-specific cancers (HR, 0.66; 95% CI, 0.47-0.92), while lacto-ovo-vegetarian diets were associated with reduced risk of gastrointestinal cancers (HR, 0.75; 95% CI, 0.60-0.92).
Shows an association, not proof of cause and effect.
Tantamango-Bartley Y, Jaceldo-Siegl K, Fan J, et al. Cancer Epidemiol Biomarkers Prev, 2013. PMID 23169929
2013 / Observational study
Fifty patients who underwent Roux-en-Y gastric bypass were evaluated using a dumping syndrome questionnaire and the Three-Factor Eating Questionnaire (TFEQ). Dumping syndrome prevalence was 42%, and nondumpers had a greater mean decrease in BMI than dumpers at 1 and 2 years (18.5 and 17.8 vs. 14.4 and 13.7, respectively). While dumping syndrome had no definite relationship with eating behavior, cognitive restraint scores over 80% were associated with significantly greater average BMI decreases at 1 and 2 years (19 and 20.8 vs. 14.6 and 12.4; p = 0.01 and p = 0.03, respectively).
Shows an association, not proof of cause and effect.
Banerjee A, Ding Y, Mikami DJ, et al. Surg Endosc, 2013. PMID 23233009
2013 / Observational twin study
Researchers analyzed habitual dietary intake, clinical data, quantitative PCR, and DGGE to assess stool microbiota in Finnish monozygotic twins using hierarchical linear mixed models. Bacterial abundance and diversity did not differ by weight status (normal-weight, overweight, obese). Intakes of energy, MUFAs, n3 PUFAs, n6 PUFAs, and soluble fiber had significant associations with stool bacterial counts; for example, increased energy intake was associated with reduced Bacteroides spp. Co-twins with identical energy intake had more similar numbers and DGGE-profile diversities of Bacteroides spp., while twins consuming equal amounts of saturated fatty acids showed very similar Bacteroides spp. profiles.
Shows an association, not proof of cause and effect.
Simões CD, Maukonen J, Kaprio J, et al. J Nutr, 2013. PMID 23343669
2013 / Prospective twin study with gnotobiotic mouse transplantation
Researchers studied 317 Malawian twin pairs during their first 3 years of life to investigate the gut microbiome's role in kwashiorkor. Half of the twin pairs remained well nourished, 43% became discordant, and 7% showed concordance for acute malnutrition. Both children in discordant pairs received a peanut-based ready-to-use therapeutic food (RUTF), which produced a transient maturation of metabolic functions in kwashiorkor gut microbiomes that regressed once RUTF was stopped. Fecal transplants into gnotobiotic mice showed marked weight loss and metabolic perturbations under a Malawian diet that were only transiently improved with RUTF.
Shows an association, not proof of cause and effect.
Smith MI, Yatsunenko T, Manary MJ, et al. Science, 2013. PMID 23363771
2013 / Population-based case-control study
In a population-based case-control study of 989 cases and 1,033 healthy controls, participants completed a food frequency questionnaire with a meat-specific module. Significant positive associations (P-trends < 0.05) were identified for DiMeIQx with colorectal, distal colon, and rectal tumors; MeIQx with colorectal and colon cancer; nitrites/nitrates with proximal colon cancer; and PhIP and benzo[a]pyrene with rectal cancer. Red processed meat was linked to proximal colon cancer and pan-fried red meat to colorectal cancer, while unprocessed poultry and certain cooked poultry items showed inverse associations across several subsites (P-trends < 0.05).
Shows an association, not proof of cause and effect.
Miller PE, Lazarus P, Lesko SM, et al. Nutr Cancer, 2013. PMID 23441608
2013 / Retrospective cross-sectional study
A cross-sectional study compared 148 vegetarian Buddhist priests with age- and sex-matched controls undergoing health checkups. The prevalence of reflux esophagitis was 12.2% in Buddhist priests versus 21.6% in the control group, despite priests having higher body mass index and higher prevalence of metabolic syndrome (30.4% vs 17.6%). In multivariate analysis, male sex (OR = 3.44; 95% CI, 1.698-6.970), non-vegetarianism (OR = 2.08; 95% CI, 1.086-3.974), and negative H. pylori IgG antibody status (OR = 1.96; 95% CI, 1.039-3.712) were significantly associated with reflux esophagitis.
Shows an association, not proof of cause and effect.
Jung JG, Kang HW, Hahn SJ, et al. Dig Dis Sci, 2013. PMID 23508985
2013 / Cross-sectional comparative study
Researchers surveyed fecal, oral, and skin microbiota from 60 families to assess microbial sharing among cohabitants. Household members, particularly couples, shared more microbiota than people from different households, with stronger effects observed on skin than on oral or fecal microbiota. Dog ownership significantly increased shared skin microbiota in cohabiting adults, and adults shared more skin microbiota with their own dogs than with other dogs. The authors noted it remains unknown whether these shared microbes occupy a true niche or reflect transient detection from direct contact.
Shows an association, not proof of cause and effect.
Song SJ, Lauber C, Costello EK, et al. Elife, 2013. PMID 23599893
2013 / Cross-sectional comparative study
Researchers compared colonic microbes and metabolites in fecal samples from 12 healthy African Americans (high colon cancer risk) and 12 age- and sex-matched rural native Africans (low risk). Native Africans showed a predominance of Prevotella (enterotype 2), significantly higher total bacteria and major butyrate-producing groups, and higher fecal short-chain fatty acids. African Americans had a predominance of Bacteroides (enterotype 1), greater abundance of microbial genes encoding secondary bile acid production, and higher fecal secondary bile acid concentrations.
Shows an association, not proof of cause and effect.
Ou J, Carbonero F, Zoetendal EG, et al. Am J Clin Nutr, 2013. PMID 23719549
2013 / Prospective cohort study
Researchers prospectively analyzed data from a mother-infant nutrition study to evaluate the relationships between perinatal pet exposure, wheezy bronchitis (WB) by 24 months, and infant gut microbiota at 1 month assessed by quantitative PCR. None of the 30 infants exposed to pets developed WB by 24 months, compared to 15 of 99 (15%) nonexposed infants (P = 0.03). Counts of Bifidobacterium longum were higher in nonwheezing infants with pet exposure (n = 17) than in wheezing infants without pet exposure (n = 10) (median/upper limit of range: 8.59/10.44 vs 5.94/9.86 bacteria(log)/g of stool; P = 0.02). Conversely, B. breve was more abundant in the wheezing infants (P = 0.02).
Shows an association, not proof of cause and effect.
Nermes M, Niinivirta K, Nylund L, et al. ISRN Allergy, 2013. PMID 23724248
2013 / Multi-institution retrospective cohort study
Using a multi-institution inflammatory bowel disease cohort of 3,217 patients, researchers evaluated the association between plasma 25(OH)D levels and subsequent risk of surgery or hospitalization. In Crohn's disease (CD), multivariable analysis showed 25(OH)D <20 ng/mL was linked to increased risk of surgery (OR 1.76; 95% CI, 1.24-2.51) and hospitalization (OR 2.07; 95% CI, 1.59-2.68) compared to levels ≥30 ng/mL, with similar estimates in ulcerative colitis. Furthermore, patients with CD who normalized their 25(OH)D levels from <30 ng/mL had a lower likelihood of surgery compared with those who remained deficient (OR 0.56; 95% CI, 0.32-0.98).
Shows an association, not proof of cause and effect.
Ananthakrishnan AN, Cagan A, Gainer VS, et al. Inflamm Bowel Dis, 2013. PMID 23751398
2013 / Mendelian randomization and prospective cohort study
A study of 77,679 individuals from the general population with up to 34 years of follow-up (4,106 symptomatic gallstone cases) evaluated observational and Mendelian randomization associations using three BMI-related genetic variants. Observationally, fifth versus first BMI quintile had an adjusted HR of 2.84 (95% CI: 2.32-3.46) overall, 3.36 in women, and 1.51 in men. Genetically, carrying 6 versus 0-1 BMI-increasing alleles had an HR of 1.43 (95% CI: 0.99-2.05) overall and 1.54 (95% CI: 1.00-2.35) in women. Instrumental variable analysis for a 1 kg/m² increase in genetically determined BMI showed an estimated causal OR of 1.17 (95% CI: 0.99-1.37) overall and 1.20 (95% CI: 1.00-1.44) in women, which the authors concluded is compatible with a causal association.
Shows an association, not proof of cause and effect.
Stender S, Nordestgaard BG, Tybjaerg-Hansen A Hepatology, 2013. PMID 23775818
2013 / Case-control study
Researchers compared the gut microbiomes of 20 neurotypical children without GI symptoms to 20 autistic children mostly presenting GI symptoms using pyrosequencing of fecal bacterial 16S rDNA. Autistic symptoms, rather than GI symptom severity, were associated with less diverse gut microbiomes. Significant reductions in the relative abundances of Prevotella, Coprococcus, and unclassified Veillonellaceae were observed in autistic children. Multivariate analyses showed that these microbiome alterations correlated with autistic symptoms but not with diet patterns.
Shows an association, not proof of cause and effect.
Kang DW, Park JG, Ilhan ZE, et al. PLoS One, 2013. PMID 23844187
2013 / Cross-sectional study
A cross-sectional study analyzed 539 individuals with diverticulosis and 1569 controls undergoing colonoscopy who had no prior knowledge of their diverticular disease. Constipation was not linked to an increased risk of diverticulosis; participants with less frequent bowel movements (<7/wk) had reduced odds compared with those with regular movements (7/wk) (OR, 0.56; 95% CI, 0.40-0.80). Hard stools, straining, and incomplete bowel movements also showed no positive association. Comparing the highest to lowest quartiles of dietary fiber intake (mean 25 vs 8 g/day), no association with diverticulosis was found (OR, 0.96; 95% CI, 0.71-1.30).
Shows an association, not proof of cause and effect.
Peery AF, Sandler RS, Ahnen DJ, et al. Clin Gastroenterol Hepatol, 2013. PMID 23891924
2013 / Prospective cohort study
Data from 170,776 women in the Nurses' Health Study were analyzed over 26 years to examine dietary fiber intake and incident Crohn's disease (CD) and ulcerative colitis (UC). Over follow-up, 269 incident cases of CD and 338 cases of UC were confirmed. The highest quintile of energy-adjusted cumulative average dietary fiber intake (median 24.3 g/day) was associated with a 40% reduction in CD risk compared with the lowest quintile (multivariate HR 0.59; 95% CI, 0.39-0.90), with the apparent reduction greatest for fruit-derived fiber. In contrast, total dietary fiber was not significantly linked to UC risk (multivariate HR 0.82; 95% CI, 0.58-1.17), nor was fiber from specific sources.
Shows an association, not proof of cause and effect.
Ananthakrishnan AN, Khalili H, Konijeti GG, et al. Gastroenterology, 2013. PMID 23912083
2013 / Cross-sectional and longitudinal observational study
Researchers analyzed the gut microbial composition in 123 non-obese and 169 obese Danish individuals. Individuals were categorized into two groups differing by gut bacterial richness; those with low bacterial richness (23% of the population) exhibited more marked overall adiposity, insulin resistance, dyslipidaemia, and a more pronounced inflammatory phenotype compared to those with high bacterial richness. Furthermore, obese individuals in the lower bacterial richness group gained more weight over time.
Shows an association, not proof of cause and effect.
Le Chatelier E, Nielsen T, Qin J, et al. Nature, 2013. PMID 23985870
2013 / Cross-sectional analysis with longitudinal and small intervention sub-studies
Researchers analyzed 143 macro- and micronutrients from 3-day diet records and urinary S-(-)equol in 159 healthy adults in the US and Australia. Overall, 29.6% of participants were equol producers. Total protein, carbohydrate, fat, saturated fat, and fiber intakes did not differ between producers and nonproducers; however, equol producers had higher intakes of polyunsaturated fatty acids (P = 0.039), maltose (P = 0.02), vitamin A (P = 0.01), and vitamin E (P = 0.035), and lower total cholesterol intake (P = 0.010). Over 2 years, status remained unchanged in 100% of nonproducers and 80% of producers, while metronidazole abolished production in 20% of 5 tested participants.
Shows an association, not proof of cause and effect.
Setchell KD, Brown NM, Summer S, et al. J Nutr, 2013. PMID 24089421
2013 / Prospective cohort survey study
A survey of 292 patients with IBD evaluated patterns and perceived benefits of marijuana use. Overall, 12.3% were active users, 39.0% past users, and 48.6% never users. Among current and past users, 16.4% used marijuana for disease symptoms, with the majority reporting it was 'very helpful' for relieving abdominal pain, nausea, and diarrhea. On multivariate analysis, chronic abdominal pain was associated with current use (OR 3.5, 95% CI 1.24-9.82; P = 0.02) and medicinal use (OR 4.7, 95% CI 1.8-12.2; P = 0.001), while age showed an inverse association (OR 0.93; 95% CI 0.89-0.97; P < 0.001 for both).
Shows an association, not proof of cause and effect.
Ravikoff Allegretti J, Courtwright A, Lucci M, et al. Inflamm Bowel Dis, 2013. PMID 24185313
2013 / Prospective observational study
A prospective study examined 72 infants with confirmed cow's milk allergy (CMA) for the presence of anal fissures, assessed by pediatricians. Anal fissures were present in 79% of infants with CMA who had gastrointestinal symptoms and in 83% of those without gastrointestinal symptoms. The authors reported that the prevalence of anal fissure in these infants is significantly higher than in normal infants and suggested it may be a pathognomonic sign of CMA.
Shows an association, not proof of cause and effect.
Jirapinyo P, Densupsoontorn N, Kangwanpornsiri C J Med Assoc Thai, 2013. PMID 24319847
2012 / Cross-sectional case-control study
Fecal samples from 144 vegetarians, 105 vegans, and matched omnivorous controls were analyzed using classical bacteriological methods. Vegan samples had significantly lower counts of Bacteroides spp. (P=0.001), Bifidobacterium spp. (P=0.002), Escherichia coli (P=0.006), and Enterobacteriaceae spp. (P=0.008) compared to controls, with vegetarian counts ranking between vegans and controls. Total microbial counts did not differ among groups. Stool pH was significantly lower in vegan and vegetarian subjects than in controls (P=0.0001) and was significantly correlated with counts of E. coli and Enterobacteriaceae across all subgroups.
Shows an association, not proof of cause and effect.
Zimmer J, Lange B, Frick JS, et al. Eur J Clin Nutr, 2012. PMID 21811294
2012 / Cross-sectional comparative study
Researchers compared colonic contents among 12 African Americans (fat intake 123 ± 11 g/d), 10 Caucasian Americans (fat intake 129 ± 17 g/d), and 13 native Africans on a low-fat diet (38 ± 3.0 g/d). The chief secondary colonic bile acids, deoxycholic acid and lithocholic acid, correlated with fat intake and were 3-4 times higher in the American groups compared to native Africans (p < 0.05). Major SCFAs were significantly lower in African Americans (p < 0.001) and Caucasian Americans (p < 0.001) compared to native Africans, whereas branched-chain fatty acids were higher.
Shows an association, not proof of cause and effect.
Ou J, DeLany JP, Zhang M, et al. Nutr Cancer, 2012. PMID 22136517
2012 / Case report
This case report describes a 3-month-old infant girl who presented to an emergency department with signs and symptoms of toxicity after recent star anise ingestion. Her presentation showed particular gastrointestinal and neurological findings consistent with previous toxicity reports. Chinese star anise is commonly used for infantile colic, but contamination with Japanese star anise has been linked to infant toxicity.
Shows an association, not proof of cause and effect.
Madden GR, Schmitz KH, Fullerton K Pediatr Emerg Care, 2012. PMID 22391927
2012 / Comparative cross-sectional study
This study compared urinary excretion rates of p-cresol sulfate (PCS) and indoxyl sulfate (IS) in 15 vegetarians and 11 individuals consuming an unrestricted diet with normal kidney function across two study periods 1 month apart. Average PCS excretion was 62% lower (95% CI, 15-83) and average IS excretion was 58% lower (95% CI, 39-71) in vegetarians than in participants consuming an unrestricted diet. Lower excretion of PCS and IS in vegetarians was associated with a 69% higher fiber intake (95% CI, 20-139) and a 25% lower protein intake (95% CI, 3-42) based on food records. Excretion rates varied widely among individuals and were not closely correlated with each other, but tended to remain stable over 1 month.
Shows an association, not proof of cause and effect.
Patel KP, Luo FJ, Plummer NS, et al. Clin J Am Soc Nephrol, 2012. PMID 22490877