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.
2011 / Cross-sectional analysis
Using NHANES 1999-2004 data from 13,292 adults, researchers evaluated dietary quality and nutrient intake among lacto-ovo vegetarians and nonvegetarians, including those consuming weight-loss diets. Mean intakes of fiber, vitamins A, C, and E, thiamin, riboflavin, folate, calcium, magnesium, and iron were higher in vegetarians than nonvegetarians. Specifically, intakes of vitamin E (8.3 ± 0.3 vs 7.0 ± 0.1 mg), vitamin A (718 ± 28 vs 603 ± 10 μg), and magnesium (322 ± 5 vs 281 ± 2 mg) exceeded those of nonvegetarians, though intakes remained below desired levels in both groups. The Healthy Eating Index score did not differ significantly between all vegetarians and all nonvegetarians (50.5 ± 0.88 vs 50.1 ± 0.33, P = 0.6).
Shows an association, not proof of cause and effect.
Farmer B, Larson BT, Fulgoni VL, et al. J Am Diet Assoc, 2011. PMID 21616194
2011 / Cross-sectional survey
A 2008 nationally representative survey of 1,211 U.S. primary care physicians (response rate 64.5%) examined practices regarding diet, physical activity, and weight control. Half of PCPs (49%) reported regularly recording BMI, and fewer than 50% reported always providing specific guidance on diet, physical activity, or weight control. Less than 10% always referred patients for evaluation/management and <22% always systematically tracked weight or weight-related behaviors over time, though >70% reported ever using pharmacologic treatments and 86% had referred for obesity surgery. PCPs were more likely to counsel on physical activity than on diet or weight control (p's < 0.05).
Shows an association, not proof of cause and effect.
Smith AW, Borowski LA, Liu B, et al. Am J Prev Med, 2011. PMID 21665061
2011 / Cross-sectional analysis of survey data
Using 1999-2004 NHANES 24-hour dietary recalls, researchers assessed the link between candy consumption and health parameters in 11,181 youth. Total candy consumers had higher total energy (2248.9±26.8 vs 1993.1±15.1 kcals, p<0.0001) and added sugars (27.7±0.44 vs 23.4±0.38 g, p<0.0001) compared to non-consumers. Mean HEI-2005 scores were lower in chocolate candy consumers (46.7±0.8 vs 48.3±0.4, p=0.0337) but not different for total or sugar candy consumers. Weight, BMI, and waist circumference were lower in candy consumers, who were 22% and 26% less likely to be overweight and obese, respectively. Blood pressure and lipids did not differ between consumers and non-consumers, though sugar candy consumers had lower C-reactive protein levels.
Shows an association, not proof of cause and effect.
O'Neil CE, Fulgoni VL, Nicklas TA Food Nutr Res, 2011. PMID 21691462
2011 / Repeated cross-sectional survey analysis
Using nationally representative US survey data from 1977-78 to 2003-06, researchers mathematically decomposed changes in total daily energy intake (TE) into contributions from portion size (PS), energy density (ED), and eating occasions (EO). Between 1977-78 and 2003-06, TE increased by +570 kcal/d and daily EOs increased by +1.1. Between 1977-78 and 1989-91, annualized changes in PS contributed nearly 15 kcal/d/y to TE increases, while EO accounted for 4 kcal/d/y. Between 1994-98 and 2003-06, changes in EO accounted for 39 kcal/d/y of increase, while PS changes accounted for a 1 kcal/d/y decline.
Shows an association, not proof of cause and effect.
Duffey KJ, Popkin BM PLoS Med, 2011. PMID 21738451
2011 / Cross-sectional analysis of survey data
Using NHANES 1999-2004 data from 13,292 adults, researchers assessed associations between dried fruit consumption (≥1/8 cup-equivalent/day, reported by 7% of participants) and diet quality or adiposity. Dried fruit consumers had significantly higher intakes of fiber (+6.6 g/d), vitamins A, E, C, K, calcium, phosphorus, magnesium, and potassium (all P < .01). Consumers also had a higher total Healthy Eating Index 2005 score (59.3 ± 0.5 vs 49.4 ± 0.3, P < .01). Covariate-adjusted weight (78.2 ± 0.6 vs 80.7 ± 0.3 kg), BMI (27.1 ± 0.2 vs 28.1 ± 0.2), and waist circumference (94.0 ± 0.5 vs 96.5 ± 0.2 cm) were significantly lower in consumers than nonconsumers (P < .01).
Shows an association, not proof of cause and effect.
Keast DR, O'Neil CE, Jones JM Nutr Res, 2011. PMID 21745628
2011 / Cross-sectional study
Fifty-one subjects (41 female, 10 male) were evaluated for oral sensitivity to oleic acid (3.8 mM) and classified as hypersensitive (3/3 correct identifications) or hyposensitive (<3/3). Overall, 75% (n = 38) were classified as hyposensitive. Compared with hypersensitive subjects, hyposensitive subjects consumed significantly more energy, fat, saturated fat, and fatty foods (butter, meat, dairy), had greater BMI, and were less perceptive of small changes in the fat content of custard (all P < 0.05).
Shows an association, not proof of cause and effect.
Stewart JE, Newman LP, Keast RS Clin Nutr, 2011. PMID 21757270
2011 / Cross-sectional survey
A cross-sectional survey of 1,165 U.S. adults evaluated 36 weight-control practices used in the past week to see if practices linked to weight loss differed from those linked to weight-loss maintenance. Only 8 of 36 practices (22%) were associated with both outcomes, reflecting poor overall agreement (kappa=0.22). Following a consistent exercise routine (OR 1.97; 95% CI 1.33, 2.94) and eating plenty of low-fat protein (OR 1.76; 95% CI 1.25, 2.50) were linked to weight-loss maintenance but not loss. Conversely, doing different kinds of exercises (OR 2.56; 95% CI 1.44, 4.55) and planning meals ahead (OR 1.68; 95% CI 1.03, 2.74) were linked to weight loss but not maintenance.
Shows an association, not proof of cause and effect.
Sciamanna CN, Kiernan M, Rolls BJ, et al. Am J Prev Med, 2011. PMID 21767723
2011 / Repeated cross-sectional customer surveys
Cross-sectional surveys evaluated lunchtime fast food purchases before (2007) and after (2009) calorie labeling regulation across 11 chains in NYC. For the full sample, unadjusted mean calories purchased did not change from before to after regulation (828 vs 846 kcal, P = 0.22), though an adjusted model showed a modest decrease (847 vs 827 kcal, P = 0.01). Significant reductions occurred at McDonald's (829 vs 785 kcal), Au Bon Pain (555 vs 475 kcal), and KFC (927 vs 868 kcal), while Subway saw an increase (749 vs 882 kcal). In 2009, 15% of customers reported using the calorie information and purchased 106 fewer kilocalories than those who did not (757 vs 863 kcal, P < 0.001).
Shows an association, not proof of cause and effect.
Dumanovsky T, Huang CY, Nonas CA, et al. BMJ, 2011. PMID 21791497
2011 / Prospective cohort study
In the Chicago Western Electric Study, 1,730 employed white men aged 40 to 55 years were followed from 1958 to 1966 to examine the relation of baseline protein intake to overweight and obesity. After adjusting for potential confounders, the odds ratio for obesity comparing the highest to lowest quartile of animal protein was 4.62 (95% CI: 2.68 to 7.98; P for trend < 0.01). Conversely, the odds ratio was 0.58 (95% CI: 0.36 to 0.95; P for trend = 0.053) for participants in the highest versus lowest quartile of vegetable protein intake.
Shows an association, not proof of cause and effect.
Bujnowski D, Xun P, Daviglus ML, et al. J Am Diet Assoc, 2011. PMID 21802560
2011 / Repeated cross-sectional survey analysis
Data from 29,217 US children aged 2 to 18 years across national surveys between 1977 and 2006 were analyzed for trends in daily energy intake by eating location and food source. An increase in children's total energy intake (+179 kcal/day) from 1977 to 2006 was linked to an increase in energy eaten away from home (+255 kcal/day), with the share of daily energy eaten away from home rising from 23.4% to 33.9%. Between 1994 and 2006, fast food grew to become the largest contributor to foods prepared away from home, and store-bought food became the largest source of energy eaten away from home. Fast food eaten at home and store-bought food eaten away from home increased significantly.
Shows an association, not proof of cause and effect.
Poti JM, Popkin BM J Am Diet Assoc, 2011. PMID 21802561
2011 / Prospective multi-center observational cohort study
In the Longitudinal Assessment of Bariatric Surgery (LABS) across 10 centers, prospective data were collected for surgeries between March 2005 and April 2009, with deaths independently adjudicated. Among 6,118 patients undergoing primary bariatric surgery, 18 deaths (0.3%) occurred within 30 days. Sepsis was the most common cause of death (33%), followed by cardiac causes (28%) and pulmonary embolism (17%). In one patient, the cause of death could not be determined.
Shows an association, not proof of cause and effect.
Smith MD, Patterson E, Wahed AS, et al. Obes Surg, 2011. PMID 21866378
2011 / Simulation modeling study
A simulation model projected the health and economic consequences of rising obesity over the next two decades in the USA and the UK. The model projects 65 million more obese adults in the USA and 11 million more in the UK by 2030. Combined, these trends are projected to result in an additional 6-8.5 million diabetes cases, 5.7-7.3 million heart disease and stroke cases, 492,000-669,000 cancer cases, and 26-55 million quality-adjusted life years forgone. Annual medical treatment costs are estimated to rise by $48-66 billion in the USA and by £1.9-2 billion in the UK by 2030.
Shows an association, not proof of cause and effect.
Wang YC, McPherson K, Marsh T, et al. Lancet, 2011. PMID 21872750
2011 / Case report
A 30-year-old male was admitted with general weakness and a drowsy mental status after eating only 3-4 spoons of brown rice and fresh vegetables without salt for 3 months to treat a tic disorder. He was bed-ridden and experienced a 14 kg weight loss over the 3-month period. The patient developed hyponatremia, metabolic acidosis, subcutaneous emphysema, mediastinal emphysema, pneumothorax, and pancytopenia. The authors also review literature on preventing refeeding syndrome and initiating feeding in malnourished patients.
Shows an association, not proof of cause and effect.
Park SW, Kim JY, Go GJ, et al. Electrolyte Blood Press, 2011. PMID 21998605
2011 / Cross-sectional comparative study
Researchers analyzed whether fat and carbohydrate intake modified the association of FTO gene variants with BMI in 1,069 GOLDN participants and 1,094 BPRHS participants. SFA intake significantly interacted with FTO on BMI in both cohorts (categorical P-interaction = 0.007 in GOLDN, 0.014 in BPRHS). Homozygous participants for the FTO-risk allele had a higher mean BMI than other genotypes only with high SFA intake above the population mean (29.7 ± 0.7 vs. 28.1 ± 0.5 kg/m² in GOLDN, P = 0.037; 33.6 ± 0.8 vs. 31.2 ± 0.4 kg/m² in BPRHS, P = 0.006), with no association found at lower SFA intakes. No significant interactions were observed with carbohydrate intake.
Shows an association, not proof of cause and effect.
Corella D, Arnett DK, Tucker KL, et al. J Nutr, 2011. PMID 22049296
2011 / Cross-sectional institutional-based study
A cross-sectional study screened 325 pregnant Indian women without previously known diabetes using a 75 g, 2-hour oral glucose tolerance test. The authors evaluated statistical associations between various factors and gestational diabetes mellitus (GDM). Results indicated that a bad obstetrics history, obesity, a high-calorie diet (especially non-vegetarian), and lower physical activity were associated with higher susceptibility to GDM.
Shows an association, not proof of cause and effect.
Jali MV, Desai BR, Gowda S, et al. Eur Rev Med Pharmacol Sci, 2011. PMID 22195364
2011 / Cross-sectional study
Using 1999-2004 NHANES data from 13,292 adults aged 19+ years, researchers assessed the relationship between nut consumption (≥1/4 ounce/day) and health risk factors. Nut consumers had significantly lower BMI (27.7 ± 0.2 vs 28.1 ± 0.1 kg/m²), waist circumference (95.6 ± 0.4 vs 96.4 ± 0.3 cm), and systolic blood pressure (121.9 ± 0.4 vs 123.20 ± 0.3 mmHg), along with lower rates of hypertension and low HDL-C. Tree nut consumers had lower body weight (78.8 ± 0.7 vs 80.7 ± 0.3 kg), a lower prevalence of four metabolic syndrome risk factors, and lower overall prevalence of metabolic syndrome (21.2% ± 2.1% vs 26.6% ± 0.7%, p < 0.05).
Shows an association, not proof of cause and effect.
O'Neil CE, Keast DR, Nicklas TA, et al. J Am Coll Nutr, 2011. PMID 22331685
2010 / Cross-sectional study
This study evaluated the prevalence of human adenovirus-36 (Ad-36) infection and its correlation with BMI z-score and other obesity measures in obese Korean children (age 14.8 +/- 1.9 years). Blood samples taken at annual school physical exams or clinic visits were analyzed using a serum neutralization assay alongside routine serum chemistry values. Among subjects, 30% were Ad-36 positive (N = 25) and 70% were negative (N = 59). Infected children showed significantly higher BMI z-scores (1.92 vs. 1.65, p < 0.01) and waist circumferences (96.3 vs. 90.7 cm, p = 0.05) compared with uninfected children, while cardiovascular risk factors showed no significant difference.
Shows an association, not proof of cause and effect.
Atkinson RL, Lee I, Shin HJ, et al. Int J Pediatr Obes, 2010. PMID 19593728
2010 / Cross-sectional study
A cross-sectional study of 54 young adults aged 18-30 evaluated the relationship between a dietary phytochemical index (PI) score—calculated from 3-day food records—and markers of adiposity, oxidative stress, and inflammation. PI values were higher in the overweight-obese group. The PI score significantly correlated with body mass index, waist circumference, waist-to-hip ratio, and plasma oxidative stress, and was a significant contributor to yearly weight gain. The PI score did not correlate with any measured cytokine levels (interleukin-1beta, interleukin-6, or C-reactive protein).
Shows an association, not proof of cause and effect.
Vincent HK, Bourguignon CM, Taylor AG J Hum Nutr Diet, 2010. PMID 19735350
2010 / Cross-sectional study
Ninety-five community-dwelling older females (ages 52-92 years) underwent high-resolution MRI scanning and extensive neuropsychological testing to evaluate associations among BMI, brain volumes, and cognition. Higher BMI was associated with decreased gray matter volumes in several regions (including the left orbitofrontal gyrus) and increased white matter volumes in frontal, temporal, and parietal lobes, even when hypertension was considered. Obese women performed poorer on executive functioning tests compared to normal weight women, and smaller left orbitofrontal gray matter volume was linked to lower executive functioning.
Shows an association, not proof of cause and effect.
Walther K, Birdsill AC, Glisky EL, et al. Hum Brain Mapp, 2010. PMID 19998366
2010 / Prospective cohort study
A prospective cohort of 89,432 European participants followed for an average of 6.5 years examined associations between dietary fiber sources and anthropometric changes. A 10-g/d higher total fiber intake was associated with a weight change of -39 g/y (95% CI: -71, -7 g/y) and a waist circumference change of -0.08 cm/y (95% CI: -0.11, -0.05 cm/y). A 10-g/d higher cereal fiber intake was associated with -77 g/y (95% CI: -127, -26 g/y) weight change and -0.10 cm/y (95% CI: -0.18, -0.02 cm/y) waist circumference change. Fruit and vegetable fiber was not linked to weight change, but showed a similar inverse association with waist circumference change.
Shows an association, not proof of cause and effect.
Du H, van der A DL, Boshuizen HC, et al. Am J Clin Nutr, 2010. PMID 20016015
2010 / Case series
Researchers analyzed 8 patients from medical centers and 9 cases from the FDA MedWatch database who developed liver injury after taking Hydroxycut. All 8 center patients were hospitalized, with 3 requiring liver transplantation, and the FDA cases included 1 fatal case of acute liver failure. Causality adjudication using DILIN methodology classified 8 cases as definite, 5 as highly likely, 2 as probable, and 2 as possible. Most patients presented with a hepatocellular pattern of injury, typically experiencing jaundice, fatigue, nausea, vomiting, and abdominal pain.
Shows an association, not proof of cause and effect.
Fong TL, Klontz KC, Canas-Coto A, et al. Am J Gastroenterol, 2010. PMID 20104221
2010 / Cross-sectional analysis
A cross-sectional study evaluated approximately 8,550 four-year-old US children to examine the link between childhood obesity and three household routines: regular family evening meals (>5 nights/week), adequate sleep (>=10.5 hours/night), and limited screen time (<=2 hours/day). Overall, 18% of children were obese. Obesity prevalence was 14.3% (95% CI: 11.3%-17.2%) among children exposed to all 3 routines versus 24.5% (95% CI: 20.1%-28.9%) among those exposed to none. After adjusting for covariates, the odds ratio for obesity was 0.63 (95% CI: 0.46-0.87) for all 3 routines, 0.64 (95% CI: 0.47-0.85) for any 2 routines, and 0.84 (95% CI: 0.63-1.12) for 1 routine compared with none.
Shows an association, not proof of cause and effect.
Anderson SE, Whitaker RC Pediatrics, 2010. PMID 20142280
2010 / Case-control study
Researchers compared 48 adults with bacterial infective endocarditis to 48 age-matched controls, evaluating BMI one month prior to symptom onset and one year post-hospital discharge. Patients treated with vancomycin plus gentamicin experienced a significant increase in BMI (+2.3 ± 0.9 kg/m², p = 0.03), whereas controls and patients given other antibiotics did not. Treatment with vancomycin plus gentamicin independently predicted a BMI increase of ≥10% (adjusted OR 6.7; 95% CI 1.37-33.0; p = 0.02), with five patients developing obesity. Weight gain was especially pronounced in males older than 65 who did not undergo cardiac surgery.
Shows an association, not proof of cause and effect.
Thuny F, Richet H, Casalta JP, et al. PLoS One, 2010. PMID 20161775
2010 / Prospective time series (nonexperimental)
A prospective time series evaluated 2,974 men and women participating in an intensive cardiac rehabilitation program across 24 socioeconomically diverse sites at baseline, 12 weeks, and 1 year. Retention was 88% at 12 weeks and 78.1% at 1 year. Statistically significant improvements were observed at 12 weeks for BMI, triglycerides, LDL cholesterol, total cholesterol, HbA1c, systolic and diastolic blood pressure, depression, hostility, exercise, and functional capacity, with all differences remaining significant at 1 year compared to baseline (except HDL, which was unchanged; p < .005). Significant recidivism occurred between 12 weeks and 1 year for most measures, though further improvements between 12 weeks and 1 year were noted in BMI, HDL cholesterol, functional capacity, and hostility.
Shows an association, not proof of cause and effect.
Silberman A, Banthia R, Estay IS, et al. Am J Health Promot, 2010. PMID 20232608