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.
1993 / Population-based case-control study
A population-based case-control study examined breast cancer risk among 597 cases and 966 controls of Chinese, Japanese, and Filipino descent aged 20-55 years in California and Hawaii. Asian-American women born in the West had a 60% higher breast cancer risk than those born in the East, with a sixfold gradient across migration patterns. Among women born in the West, having three or four grandparents born in the West was linked to a 50% higher risk than having all grandparents born in the East. For Eastern-born migrants, living in the West for a decade or longer was associated with an 80% higher risk compared to more recent migrants, and urban origin was linked to a 30% higher risk than rural origin.
Shows an association, not proof of cause and effect.
Ziegler RG, Hoover RN, Pike MC, et al. J Natl Cancer Inst, 1993. PMID 8230262
1993 / Population-based case-control study
A population-based case-control study in Utah evaluated associations between cigarette smoking, alcohol, coffee, tea, caffeine, and theobromine intake and prostate cancer risk. Pack-years of cigarettes, alcohol, coffee, tea, and caffeine were not associated with prostate cancer risk. Compared with very low intake, older men consuming 11 to 20 mg/day and over 20 mg/day of theobromine had higher risks for all prostate tumors (OR = 2.06, 95% CI = 1.33-3.20 and OR = 1.47, CI = 0.99-2.19, respectively) and for aggressive tumors (OR = 1.90, CI = 0.90-3.97 and OR = 1.74, CI = 0.91-3.32, respectively). The authors noted this finding requires further exploration in studies with a wider range of theobromine exposures and more aggressive tumors.
Shows an association, not proof of cause and effect.
Slattery ML, West DW Cancer Causes Control, 1993. PMID 8280834
1993 / Cross-sectional survey study
This study evaluated differences in pelvic screening examination frequency between obese and non-obese women and the influence of patient and physician attitudes toward obesity. A total of 291 women and 1,316 physicians participated. As women's weight increased, negative opinions regarding appearance and reluctance to undergo pelvic exams increased, while the likelihood of having annual pelvic exams decreased. Additionally, 17% of physicians reported reluctance to perform pelvic exams on obese patients, and 83% reported reluctance to perform them on reluctant patients.
Shows an association, not proof of cause and effect.
Adams CH, Smith NJ, Wilbur DC, et al. Women Health, 1993. PMID 8372479
1993 / Retrospective case series and case-control analysis
A limited number of verified case histories were analyzed to evaluate a very low fat, moderately high fiber, and moderately reduced calorie diet in patients with primary pancreatic cancer, metastatic stage D2 prostate cancer, and other nutritionally linked cancers. In pancreatic cancer patients, a retrospective study showed 1-year survival was higher among those who modified their diets compared with those without evidence of diet alteration. For metastatic stage D2 prostate cancer patients, a case-control study found a statistical association between dietary modification and both longer survival and improved quality of life.
Shows an association, not proof of cause and effect.
Carter JP, Saxe GP, Newbold V, et al. J Am Coll Nutr, 1993. PMID 8409076
1993 / Ecological study
An ecological study evaluated factors related to esophageal cancer mortality variation across 21 communes in Shanxi Province, China, observing 148,928 individuals over 1983 to 1988. A significant positive relation was found between mortality rates and dietary consumption of corn and wheat flour, while dietary sorghum and millet showed a significant inverse relation. Compared to the lowest quartile, age- and sex-adjusted mortality-rate ratios for esophageal cancer were 1.4 (95% CI: 1.1-2.0) for the third quartile and 3.2 (95% CI: 2.5-4.2) for the highest quartile of corn and wheat-flour consumption. Other dietary and environmental factors studied did not contribute to the mortality variation.
Shows an association, not proof of cause and effect.
Chen F, Cole P, Mi Z, et al. Int J Cancer, 1993. PMID 8473048
1992 / Ecological study, cross-sectional cohort, and meta-analysis
Data on stool weight were gathered from 20 populations across 12 countries, alongside detailed measurements in 220 healthy U.K. adults and a meta-analysis of 11 controlled-diet studies (n = 206). In the U.K. cohort, median daily stool weight was 106 g/day and whole-gut transit time was 60 hours. Across global populations, average stool weight varied from 72 to 470 g/day and was inversely related to colon cancer risk (r = -0.78). In the controlled-diet meta-analysis, nonstarch polysaccharide intake correlated significantly with mean daily stool weight (r = 0.84).
Shows an association, not proof of cause and effect.
Cummings JH, Bingham SA, Heaton KW, et al. Gastroenterology, 1992. PMID 1333426
1992 / Prospective cohort study
Researchers evaluated 508 adolescents (ages 13 to 18) from the Harvard Growth Study of 1922 to 1935 to examine long-term morbidity and mortality based on adolescent BMI. After 55 years of follow-up, adolescent overweight was linked to increased all-cause mortality (RR 1.8; 95% CI, 1.2 to 2.7; P = 0.004) and coronary heart disease mortality (RR 2.3; 95% CI, 1.4 to 4.1; P = 0.002) in men, but not in women. In both sexes, adolescent overweight was associated with higher morbidity from coronary heart disease and atherosclerosis, as well as colorectal cancer and gout in men, and arthritis in women. These adverse health outcomes were independent of adult weight.
Shows an association, not proof of cause and effect.
Must A, Jacques PF, Dallal GE, et al. N Engl J Med, 1992. PMID 1406836
1991 / Nested case-control study within a prospective cohort
In a Finnish cohort of 36,265 adults followed for an average of 8 years, 766 incident cancer cases were identified and compared with 1,419 matched controls using stored serum samples. Individuals with low serum alpha-tocopherol levels had approximately a 1.5-fold higher risk of cancer compared to those with higher levels. The association varied by cancer site, showing the strongest links with certain gastrointestinal cancers and cancers unrelated to smoking. The relationship was also most pronounced among nonsmoking men and women with low serum selenium concentrations.
Shows an association, not proof of cause and effect.
Knekt P, Aromaa A, Maatela J, et al. Am J Clin Nutr, 1991. PMID 1985399
1991 / Comparative ecological study of vital statistics
Vital statistics from 1955 through 1985 for Japanese natives and United States whites were evaluated to examine changes in cancer mortality alongside lifestyle patterns. As the Japanese lifestyle and diet became more westernized, rates of malignancies of the breast, ovary, corpus uteri, prostate, pancreas, and colon continued to rise. Lung cancer rates accelerated rapidly among Japanese males, and oropharyngeal cancer rose alongside alcohol and tobacco use. Mortality patterns for laryngeal, esophageal, and urinary bladder cancers were discrepant with established risk factor associations, suggesting differences in risk factor exposure between the nations.
Shows an association, not proof of cause and effect.
Wynder EL, Fujita Y, Harris RE, et al. Cancer, 1991. PMID 1985768
1990 / Population-based matched case-control study
A population-based matched case-control study analyzed 191 histologically confirmed thyroid cancer cases and 441 controls across five ethnic groups in Hawaii. Among women, intake of seafood (especially shellfish), harm ha, and dietary iodine was associated with increased risk, whereas consumption of goitrogenic (primarily cruciferous) vegetables was linked to decreased risk. Odds ratios were 4.8 (95% CI = 1.2-19.2) for combined high iodine intake and first-pregnancy miscarriage, and 7.3 (95% CI = 1.5-34.5) for high iodine intake and fertility drug use. Non-dietary risk factors in women included miscarriage, fertility drug use, family history of thyroid disease, obesity, and work as a farm laborer; in men, positive associations were found for obesity, farm labor, and history of benign thyroid disease.
Shows an association, not proof of cause and effect.
Kolonel LN, Hankin JH, Wilkens LR, et al. Cancer Causes Control, 1990. PMID 2102295
1990 / Cross-sectional comparative study
Serum oestrone (E1), oestradiol (E2), and SHBG levels were measured in 91 postmenopausal rural Japanese women and 38 postmenopausal American white women. E1 levels were 47% greater and E2 levels were 36% greater in the American women; after adjustment for the lower weight of the Japanese women, differences remained at 43% and 27%, respectively (all statistically highly significant). Little difference was observed in SHBG levels between the two groups.
Shows an association, not proof of cause and effect.
Shimizu H, Ross RK, Bernstein L, et al. Br J Cancer, 1990. PMID 2206953
1990 / Cross-sectional survey
This study assessed knowledge of dietary recommendations for chronic disease prevention among 2,305 individuals from the public, health care workers, graduate students, and health club attendees. Over 90% of participants were unaware of recommendations for calcium, salt, vitamin A, and fiber, as well as the fiber content in high-fiber cereal. Approximately 80% were unaware of fat recommendations and could not calculate the fat content of a food product, while nearly half took a daily vitamin pill. Subgroups expected to be more informed, such as health care professionals or those with higher education, did not demonstrate satisfactory knowledge.
Shows an association, not proof of cause and effect.
Schapira DV, Kumar NB, Lyman GH, et al. Prev Med, 1990. PMID 2320559
1989 / Cross-sectional comparative tissue study
Beta-carotene levels were measured by high-pressure liquid chromatography in uterine leiomyomas and adjacent normal myometrium from 18 hysterectomy patients. Beta-carotene concentration was significantly lower in fibroid tissue than in normal myometrium (p = 0.0013). Concentrations were also assayed in cancers of the cervix, endometrium, ovary, breast, colon, lung, liver, and rectum compared with adjacent normal sites, and levels were lower across all cancer tissues. The authors suggest tissue deficiency may play a role in pathogenesis, though the mechanism remains unknown.
Shows an association, not proof of cause and effect.
Palan PR, Mikhail M, Romney SL Am J Obstet Gynecol, 1989. PMID 2603922
1989 / Hospital-based case-control study
A hospital-based case-control study assessed diet using a quantitative food frequency questionnaire in 203 colorectal cancer cases and 425 controls among Chinese in Singapore. For colon and rectum cancers combined, high cruciferous vegetable intake was associated with reduced risk (OR = 0.50, p < 0.01), whereas a high meat/vegetable consumption ratio was associated with increased risk (OR = 1.77, p < 0.05). Low coffee consumption was linked to a higher risk (OR = 1.59, p < 0.05 for trend), and history of cholecystectomy increased risk of combined cancers (OR = 3.43, p < 0.05). No consistent trends were observed for fat or fibre intakes across multiple regression models.
Shows an association, not proof of cause and effect.
Lee HP, Gourley L, Duffy SW, et al. Int J Cancer, 1989. PMID 2731998
1989 / Prospective cohort study
A cohort of 20,341 California Seventh-day Adventist women completed a lifestyle questionnaire in 1976 and was followed for 6 years, during which 215 histologically confirmed primary breast cancer cases were detected over approximately 115,000 person-years. Increasing consumption of high-fat animal products was not associated with increased risk of breast cancer in a consistent fashion. Furthermore, childhood and early teenage vegetarian versus nonvegetarian dietary habits, as well as a derived adult index of percent of calories from animal fat, were not significantly related to risk after controlling for potential confounders.
Shows an association, not proof of cause and effect.
Mills PK, Beeson WL, Phillips RL, et al. Cancer, 1989. PMID 2743252
1989 / Prospective cohort study
A cohort of approximately 14,000 Seventh-day Adventist men completed a lifestyle questionnaire in 1976 and were followed through 1982, during which 180 prostate cancer cases were detected. Increasing intake of beans, lentils and peas, tomatoes, raisins, dates, and other dried fruit were all linked to significantly decreased prostate cancer risk. A history of prostate trouble was associated with a highly significant 60% increase in risk. Suggestive links between animal product consumption and increased risk did not persist after accounting for fruit and vegetable intake, and childhood vegetarianism and BMI were not related to risk.
Shows an association, not proof of cause and effect.
Mills PK, Beeson WL, Phillips RL, et al. Cancer, 1989. PMID 2743254
1989 / Prospective cohort study
Bowel function was studied in relation to 123 breast cancer cases among 7,702 women participating in the US NHANES I Epidemiologic Follow-up Study. The results suggest a slight increased risk of breast cancer for decreased frequency of bowel movements (relative risk = 1.5, 95% confidence interval = 0.8, 2.7) and firm stool consistency (RR = 1.8, 95% CI = 1.0, 3.2). These findings are consistent with an hypothesized link between constipation and elevated breast cancer risk.
Shows an association, not proof of cause and effect.
Micozzi MS, Carter CL, Albanes D, et al. Am J Public Health, 1989. PMID 2909187
1988 / Prospective cohort study
A prospective study tracked 34,000 California Seventh-day Adventists between 1976 and 1983, observing 40 deaths from pancreas cancer. Meat, egg, and coffee consumption showed suggestive associations with increased risk, but were not significantly related after controlling for smoking. Conversely, increasing consumption of vegetarian protein products, beans, lentils, peas, and dried fruit showed highly significant protective relationships with pancreas cancer risk. A prior history of diabetes and surgery for peptic or duodenal ulcer were also associated with increased fatal pancreas cancer risk.
Shows an association, not proof of cause and effect.
Mills PK, Beeson WL, Abbey DE, et al. Cancer, 1988. PMID 3365678
1987 / Consecutive autopsy study
Breasts from 110 consecutive medicolegal autopsies of women aged 20-54 years were examined using extensive histopathology and specimen radiography. Malignancy was identified in 22 women (20%), including 2 with invasive breast cancer (2%) and 20 with in situ carcinoma (18%). Breast malignancy was significantly more frequent among women older than 40 years, with late age at first full-term pregnancy, with alcohol abuse, and with steatosis or cirrhosis of the liver.
Shows an association, not proof of cause and effect.
Nielsen M, Thomsen JL, Primdahl S, et al. Br J Cancer, 1987. PMID 2829956
1986 / Cross-sectional population study
Faecal pH and dietary fibre intake were examined across four South African ethnic populations. Mean faecal pH values were 6.12 and 6.15 for rural and urban blacks, 6.21 for Indians, and 6.29 for coloureds, which were significantly lower (p < 0.01) than the 6.88 observed in whites. Corresponding mean dietary fibre intakes among children's mothers (or associates) were roughly 25 g, 18 g, 20 g, 21 g, and 23 g, respectively. Colon cancer frequency was very low in rural and urban blacks, low in Indians and coloureds, and much higher in whites, indicating that low disease frequency was associated more with low faecal pH than with dietary fibre intake.
Shows an association, not proof of cause and effect.
Walker AR, Walker BF, Walker AJ Br J Cancer, 1986. PMID 3011051
1986 / Cross-sectional comparative study
Urinary lignans (enterolactone, enterodiol) and isoflavonic phytoestrogens (daidzein, equol, O-desmethylangolensin) were measured across 94 72-hour urine samples from 53 women adhering to omnivorous, lactovegetarian, or macrobiotic diets in Boston and Helsinki. Total mean 24-hour excretion of enterolactone was 17,680 nmol in macrobiotics, 4,170 nmol in Boston lactovegetarians, 3,650 nmol in Helsinki lactovegetarians, 2,460 nmol in Helsinki omnivores, and 2,050 nmol in Boston omnivores. Other diphenols followed a similar pattern.
Shows an association, not proof of cause and effect.
Adlercreutz H, Fotsis T, Bannwart C, et al. J Steroid Biochem, 1986. PMID 3027456
1986 / Prospective cohort study
In 15 cohorts of the Seven Countries Study, 2,288 of 11,579 men died over 15 years. Death rates were positively associated with average percentage of dietary energy from saturated fatty acids and negatively associated with monounsaturated fatty acids, while unrelated to polyunsaturated fatty acids, proteins, carbohydrates, and alcohol. All death rates were negatively related to the monounsaturated to saturated fatty acid ratio, and including this ratio with age, blood pressure, cholesterol, and smoking accounted for 85% of all-cause and 96% of coronary heart disease death rate variance. All-cause and coronary heart disease death rates were low in cohorts with olive oil as the main fat.
Shows an association, not proof of cause and effect.
NANO counterpoint →
Keys A, Menotti A, Karvonen MJ, et al. Am J Epidemiol, 1986. PMID 3776973
1984 / Case-control study
A French case-control study analyzed the link between alcohol intake and breast cancer among 1,010 cases and 1,950 controls. The relative risk of breast cancer for women consuming alcoholic beverages with meals compared with nondrinkers was 1.47 (p = 10(-4)) after allowing for classical risk factors. In a subset of 500 cases and 945 controls with detailed intake data, risks were elevated for women drinking beer and wine versus nondrinkers, and risks increased with greater amounts of beer, wine, and total alcohol in grams.
Shows an association, not proof of cause and effect.
Lê MG, Hill C, Kramar A, et al. Am J Epidemiol, 1984. PMID 6475912
1984 / Cross-sectional comparative study
Dietary data were collected via a 3-day weighed food diary and a 24-hour recall from 50 matched sets of Seventh-day Adventist lacto-ovo-vegetarians, Seventh-day Adventist nonvegetarians, and general population nonvegetarians, along with 18 unmatched pure vegetarians. All groups showed adequate or excess intakes of calories, protein, and fat, with higher calorie intake in nonvegetarians explained by higher fat and protein intake. Vitamin A, B, and C levels were adequate with and without supplements across groups. Calcium intake was much below recommended levels for pure vegetarian females, and iron intake was low for all females regardless of heme iron sources.
Shows an association, not proof of cause and effect.
Calkins BM, Whittaker DJ, Nair PP, et al. Am J Clin Nutr, 1984. PMID 6486098