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.
1999 / Population-based case-control study
A nationwide population-based case-control study in Sweden evaluated the relationship between gastroesophageal reflux symptoms and esophageal or gastric cardia tumors. Compared with persons without recurrent reflux symptoms, those with recurrent symptoms had odds ratios of 7.7 (95% CI, 5.3 to 11.4) for esophageal adenocarcinoma and 2.0 (95% CI, 1.4 to 2.9) for adenocarcinoma of the cardia. For persons with long-standing and severe symptoms, odds ratios reached 43.5 (95% CI, 18.3 to 103.5) for esophageal adenocarcinoma and 4.4 (95% CI, 1.7 to 11.0) for cardia adenocarcinoma. Reflux symptoms were not associated with esophageal squamous-cell carcinoma (OR 1.1, 95% CI, 0.7 to 1.9).
Shows an association, not proof of cause and effect.
Lagergren J, Bergström R, Lindgren A, et al. N Engl J Med, 1999. PMID 10080844
1999 / Prospective cohort study
Researchers examined the relationship between total fluid intake and bladder cancer risk over 10 years among 47,909 men in the Health Professionals Follow-up Study, identifying 252 newly diagnosed cases. Total daily fluid intake was inversely associated with bladder cancer risk, with a multivariate relative risk of 0.51 (95% CI, 0.32 to 0.80) for the highest quintile (>2531 ml/day) compared to the lowest quintile (<1290 ml/day). Consumption of water (RR 0.49; 95% CI, 0.28 to 0.86 for >=1440 ml/day vs. <240 ml/day) and consumption of other fluids (RR 0.63; 95% CI, 0.39 to 0.99 for >1831 ml/day vs. <735 ml/day) were also linked to lower risk.
Shows an association, not proof of cause and effect.
Michaud DS, Spiegelman D, Clinton SK, et al. N Engl J Med, 1999. PMID 10228189
1999 / Comparative cross-sectional study
Researchers compared dietary intake, colonic fermentation, and rectal epithelial proliferation indices between black South Africans (colon cancer prevalence <1:100,000) and white South Africans (prevalence 17:100,000). Epithelial proliferation was significantly lower in rural and urban blacks than in whites. Diets of black subgroups featured low animal product and high boiled maize-meal intake, with below-RDA quantities of fiber (43% of RDA), vitamin A (78%), C (62%), folic acid (80%), and calcium (67%), while whites consumed excess animal protein (177% of RDA) and fat (153%). Fasting and food-induced breath methane production was two to three times higher in blacks.
Shows an association, not proof of cause and effect.
O'Keefe SJ, Kidd M, Espitalier-Noel G, et al. Am J Gastroenterol, 1999. PMID 10235221
1999 / Prospective cohort study
Researchers prospectively followed 384 men diagnosed with prostate cancer to evaluate the link between dietary fat intake and survival over a median follow-up of 5.2 years. During follow-up, 32 patients died of prostate cancer and 39 died of other causes. After controlling for grade, clinical stage, initial treatment, age, and total energy intake, saturated fat intake was significantly associated with disease-specific survival (p = 0.008). Compared to men in the lowest tercile, those in the highest tercile of saturated fat intake had three times the risk of dying from prostate cancer (hazards ratio 3.13, 95% CI 1.28-7.67).
Shows an association, not proof of cause and effect.
Fradet Y, Meyer F, Bairati I, et al. Eur Urol, 1999. PMID 10325493
1999 / Cohort study (with cross-sectional data)
Dietary associations with chronic disease were summarized in a cohort of 34,192 California Seventh-day Adventists. Beef intake was associated with fatal IHD in men (RR = 2.31 for >=3 times/wk vs. vegetarians), whereas consuming nuts >=5 times/wk was tied to lower fatal and nonfatal IHD risk (RR approximately 0.5 vs. <1 time/wk). Colon and prostate cancers were significantly more likely in nonvegetarians (RR of 1.88 and 1.54, respectively), and legumes and fruit were negatively associated with certain cancers. Cross-sectional data suggest vegetarian Adventists have lower risks of diabetes mellitus, hypertension, and arthritis than nonvegetarians.
Shows an association, not proof of cause and effect.
Fraser GE Am J Clin Nutr, 1999. PMID 10479227
1998 / Prospective cohort study
A prospective study followed 12,395 California Seventh-Day Adventist men who reported their soy milk intake in 1976, identifying 225 incident cases of prostate cancer. Frequent consumption of soy milk (more than once a day) was associated with a 70 percent reduction in prostate cancer risk (relative risk = 0.3, 95% confidence interval 0.1-1.0, p-value for linear trend = 0.03). This association remained after extensive adjustments were performed.
Shows an association, not proof of cause and effect.
Jacobsen BK, Knutsen SF, Fraser GE Cancer Causes Control, 1998. PMID 10189040
1998 / Prospective cohort study
In a 6-year prospective study of 32,051 Adventist Health Study cohort members, researchers evaluated diet in relation to incident colon cancer. Total meat intake of 1 or more times per week versus none was positively associated with colon cancer (RR = 1.85, 95% CI 1.19-2.87), as was red meat (RR = 1.90, 95% CI 1.16-3.11) and white meat (RR = 3.29, 95% CI 1.60-6.75). Legume intake of more than 2 times per week versus less than once per week was inversely associated with risk (RR = 0.53, 95% CI 0.33-0.86). Subjects with high red meat intake, low legume intake, and high body mass had a more than threefold elevation in risk compared with other patterns.
Shows an association, not proof of cause and effect.
Singh PN, Fraser GE Am J Epidemiol, 1998. PMID 9786231
1998 / Nested case-control study
A nested case-control study assessed meat doneness levels and breast cancer risk among women from the Iowa Women's Health Study cohort. A dose-response relationship was observed between meat doneness and breast cancer risk. Adjusted odds ratios for very well-done versus rare or medium-done meat were 1.54 (95% CI=0.96-2.47) for hamburger, 2.21 (95% CI=1.30-3.77) for beef steak, and 1.64 (95% CI=0.92-2.93) for bacon. Women who consistently consumed all three meats very well done had a 4.62 times higher risk (95% CI=1.36-15.70) compared with those who ate them rare or medium done.
Shows an association, not proof of cause and effect.
Zheng W, Gustafson DR, Sinha R, et al. J Natl Cancer Inst, 1998. PMID 9827527
1998 / Ecological observational study
Investigators examined mortality data for >50 diseases along with blood, urine, food samples, and dietary data from 50 adults across 130 villages in rural mainland China compared to US data. In rural China, fat intake was less than half that in the US, fiber intake was 3 times higher, animal protein was roughly 10% of US intake, and mean serum total cholesterol was 127 mg/dL versus 203 mg/dL in the US. CAD mortality was 16.7-fold higher for US men and 5.6-fold higher for US women. In rural China, combined CAD mortality rates were inversely associated with green vegetable intake and erythrocyte monounsaturated fatty acids, and positively associated with salt intake plus urinary sodium and plasma apolipoprotein B. Apolipoprotein B was positively linked to animal protein and meat intake, and inversely associated with plant protein, legumes, and light-colored vegetables.
Shows an association, not proof of cause and effect.
Campbell TC, Parpia B, Chen J Am J Cardiol, 1998. PMID 9860369
1997 / Comparative multicenter case-control and prospective study
Serum antioxidant vitamins and minerals were evaluated in 170 children with newly diagnosed malignancy at diagnosis and 6 months after starting treatment, and compared with 632 cancer-free controls. At diagnosis, age- and sex-adjusted serum concentrations of retinol, beta-carotene, zinc, and alpha-tocopherol were significantly inversely associated with cancer. At 6 months of treatment, no significant decreases in mean values were observed except for the alpha-tocopherol-to-cholesterol ratio in bone tumors and serum zinc in bone and central nervous system tumors. An increase during treatment was observed for retinol and selenium in leukemia patients, while beta-carotene remained at pre-therapy concentrations.
Shows an association, not proof of cause and effect.
Malvy DJ, Arnaud J, Burtschy B, et al. Med Pediatr Oncol, 1997. PMID 9212847
1997 / Comparative cross-sectional study
Gas chromatography-mass spectrometry measured lignans and isoflavonoids in plasma and prostatic fluid from Portuguese, Chinese, and British men. Mean enterolactone in prostatic fluid was 31 ng/ml in Hong Kong, 162 ng/ml in Portugal (with some > 600 ng/ml), and 20.3 ng/ml in Britain. Mean daidzein in prostatic fluid was 70 ng/ml in Hong Kong versus 4.6 ng/ml in Portugal and 11.3 ng/ml in Britain, with equol reaching 3270 ng/ml in a Hong Kong sample. Mean plasma daidzein was higher in Hong Kong (31.3 ng/ml) than Portugal (1.3 ng/ml) and Britain (8.2 ng/ml), while mean plasma enterolactone levels were comparable across centres (6.2, 3.9, and 3.9 ng/ml).
Shows an association, not proof of cause and effect.
Morton MS, Chan PS, Cheng C, et al. Prostate, 1997. PMID 9215400
1997 / Prospective cohort study with nested case-control analysis
A total of 654 residents aged 30 or older from three arseniasis-hyperendemic villages in Taiwan were followed for skin lesions. During follow-up, 33 newly diagnosed skin cancer cases were identified (incidence 14.74 per 1,000 person-years). Skin cancer cases had significantly lower baseline serum beta-carotene levels than matched healthy controls. A higher proportion of monomethylarsonic acid (MMA) in total urinary arsenic was linked to increased skin cancer risk. Individuals with cumulative arsenic exposure >=20.0 mg/liter-year and urinary MMA >26.7% had a multivariate-adjusted odds ratio of 20.91 (95% CI, 2.63-166.5) compared to those with lower exposure and MMA levels.
Shows an association, not proof of cause and effect.
Hsueh YM, Chiou HY, Huang YL, et al. Cancer Epidemiol Biomarkers Prev, 1997. PMID 9264271
1997 / Cross-sectional dietary and stool collection study
Fifty-three Australians consuming their usual diet completed a 7-day weighed food record and a 3-day stool collection to assess dietary carbohydrate intake and faecal markers. Mean intake was 14 +/- 7 g/d non-starch polysaccharides (NSP) and 5 +/- 2 g/d resistant starch (RS), with a daily faecal output of 127 +/- 70 g. Stool RS was positively correlated with faecal butyrate concentration (r = 0.34, P < 0.05) and stool output (r = 0.64, P < 0.01), and negatively related to faecal ammonia (r = -0.40, P < 0.01). Individuals who consumed more NSP (19 +/- 7 g/d) excreted over 150 g faeces daily and had faster transit times, higher concentrations of short-chain fatty acids, and lower concentrations of ammonia and phenols.
Shows an association, not proof of cause and effect.
Birkett AM, Jones GP, de Silva AM, et al. Eur J Clin Nutr, 1997. PMID 9306090
1997 / Comparative observational and in vitro laboratory study
Researchers investigated whether tobacco and tobacco smoke contain reactive glycation products capable of increasing advanced glycation end product (AGE) formation. Reactive glycotoxins in aqueous extracts of tobacco and tobacco smoke rapidly reacted with proteins to form AGEs, a process inhibited by aminoguanidine. In human samples, AGE-apolipoprotein B and serum AGE levels were significantly higher in cigarette smokers compared with nonsmokers.
Shows an association, not proof of cause and effect.
Cerami C, Founds H, Nicholl I, et al. Proc Natl Acad Sci U S A, 1997. PMID 9391127
1996 / Case-control study
A case-control study in Uruguay evaluated dietary patterns using a 64-item food frequency questionnaire to examine meat and fat intake in relation to breast cancer risk. After adjustment for confounders including family history, menopausal status, BMI, total energy, and alcohol intake, increased risk was observed for total meat, red meat, total fat, and saturated fat intake. The strongest association was observed for red meat intake in the upper quartile (OR 4.2, 95% CL 2.3-7.7) after controlling for protein and fat intake.
Shows an association, not proof of cause and effect.
Ronco A, De Stefani E, Mendilaharsu M, et al. Int J Cancer, 1996. PMID 8575853
1996 / Ecological study
This ecological study examined standardized 24-hour urine sodium and nitrate levels from INTERSALT participants aged 20-49 years across 24 countries in relation to national stomach cancer mortality rates. Pearson correlation with stomach cancer mortality for sodium was 0.70 in men and 0.74 in women (both P < 0.001), and for nitrate was 0.63 in men (P = 0.001) and P < 0.005 in women. Multiple regression using sodium and nitrate yielded an adjusted R2 of 0.61 in men and 0.54 in women (both P < 0.001); adding an interaction term increased adjusted R2 to 0.77 in men and 0.63 in women. In this model, the importance of nitrate as a risk factor increased markedly with higher sodium levels, though the relationship with sodium was always stronger than with nitrate.
Shows an association, not proof of cause and effect.
Joossens JV, Hill MJ, Elliott P, et al. Int J Epidemiol, 1996. PMID 8671549
1996 / Observational cohort study
An observational study followed 10,771 vegetarians and health-conscious UK individuals for a mean of 16.8 years, recording 1,343 deaths before age 80. Overall mortality in the cohort was roughly half that of the general population. Within the cohort, daily fresh fruit consumption was significantly associated with reduced smoking-adjusted mortality from ischaemic heart disease (rate ratio 0.76; 95% CI 0.60 to 0.97), cerebrovascular disease (0.68; 95% CI 0.47 to 0.98), and all causes combined (0.79; 95% CI 0.70 to 0.90).
Shows an association, not proof of cause and effect.
Key TJ, Thorogood M, Appleby PN, et al. BMJ, 1996. PMID 8842068
1995 / Prospective cohort study
In the Health Professionals Follow-up Study, dietary intake among 47,894 men was tracked from 1986 to 1992, during which 812 new prostate cancer cases occurred. Intakes of beta-carotene, alpha-carotene, lutein, beta-cryptoxanthin, and retinol were not consistently linked to risk, but lycopene intake was associated with lower risk (RR = 0.79; 95% CI, 0.64-0.99 for high vs. low quintile). Combined consumption of tomatoes, tomato sauce, tomato juice, and pizza was inversely associated with total prostate cancer (multivariate RR = 0.65; 95% CI, 0.44-0.95 for >10 vs. <1.5 servings/week) and advanced stages (multivariate RR = 0.47; 95% CI, 0.22-1.00).
Shows an association, not proof of cause and effect.
Giovannucci E, Ascherio A, Rimm EB, et al. J Natl Cancer Inst, 1995. PMID 7473833
1995 / Cross-sectional study
Researchers evaluated the relationship between egg consumption and intermediate biomarkers of crypt cell proliferation in 75 patients undergoing colonoscopy. There was no significant difference in egg consumption between patients with adenomatous polyps (n = 53) and those without (n = 22). Furthermore, frequency of egg consumption did not significantly correlate with crypt length, frequency of mitosis per crypt, or spatial distribution of mitoses in the overall group or among patients with polyps analyzed separately.
Shows an association, not proof of cause and effect.
NANO counterpoint →
Matthew JA, Johnson IT Eur J Cancer Prev, 1995. PMID 7496330
1995 / Cross-sectional comparative study
Fecal floras were compared among polyp patients and Japanese-Hawaiians (high colon cancer risk groups), rural native Japanese and rural native Africans (low risk groups), and North American Caucasians (intermediate flora composition). Fifteen bacterial taxa were significantly associated with high risk of colon cancer, while five were significantly associated with low risk. Total concentrations of Bacteroides species and Bifidobacterium species were generally positively associated with increased risk of colon cancer. In contrast, some Lactobacillus species and Eubacterium aerofaciens were most closely associated with low risk of colon cancer.
Shows an association, not proof of cause and effect.
Moore WE, Moore LH Appl Environ Microbiol, 1995. PMID 7574628
1995 / Population-based matched case-control study
A population-based matched case-control study investigated the relation between sunscreen use, host factors, and malignant melanoma among 400 patients and 640 healthy controls aged 15-75 years. Sunscreen use did not protect against melanoma; instead, almost always versus never using sunscreens was associated with an increased risk (OR 1.8), with a tentative dose-response relation. Trunk lesions were associated with sunscreen use in females (adjusted OR = 3.7), while extremity or head and neck lesions were associated in males (adjusted OR = 3.2). Major constitutional risk factors included raised naevi on the left arm (OR = 3.9 for more than three vs none) and freckling (OR = 1.4).
Shows an association, not proof of cause and effect.
Westerdahl J, Olsson H, Måsbäck A, et al. Melanoma Res, 1995. PMID 7734957
1994 / Prospective cohort study
Researchers examined the relationship between fried meat intake, estimated via a dietary history interview covering the previous year, and cancer risk among 9,990 Finnish men and women followed for 24 years. During follow-up, 853 cancer cases occurred. Intake of fried meat was positively associated with combined female-hormone-related cancers (breast, endometrium, and ovary), with a multivariate-adjusted relative risk of 1.77 (95% CI: 1.11-2.84) comparing the highest to the lowest tertile. Non-significant suggestive associations were seen for pancreatic and nervous system cancers, while no associations were found for other single sites or all cancers combined.
Shows an association, not proof of cause and effect.
Knekt P, Steineck G, Järvinen R, et al. Int J Cancer, 1994. PMID 7989114
1994 / Ecological / international comparative study
Nutrient intakes of starch, non-starch polysaccharides (NSPs), protein, and fat were compared with colorectal cancer incidence across 12 populations worldwide. Starch consumption showed strong inverse associations with large bowel cancer (r = -0.70) and colon cancer (r = -0.76). There was no significant relation with NSPs alone, but combining NSP with resistant starch (RS) to estimate fermentable carbohydrate yielded significant inverse associations (large bowel r = -0.52, colon r = -0.60). These relationships remained statistically significant after adjusting for fat and protein intakes.
Shows an association, not proof of cause and effect.
Cassidy A, Bingham SA, Cummings JH Br J Cancer, 1994. PMID 8180027
1993 / Comparative cross-sectional study
Researchers measured four isoflavonoids in the plasma of 14 Japanese men and 14 Finnish men. Geometric mean plasma total individual isoflavonoid levels were 7 to 110 times higher in the Japanese men compared to Finnish men. Genistein had the highest concentration among Japanese men, with a geometric mean of 276 nmol/L. The authors hypothesize that these high phyto-oestrogen levels may inhibit prostatic cancer growth and potentially explain lower prostate cancer mortality in Japan.
Shows an association, not proof of cause and effect.
Adlercreutz H, Markkanen H, Watanabe S Lancet, 1993. PMID 7901532