Upload
conrad-rose
View
216
Download
4
Embed Size (px)
Citation preview
Presenter:
Andrew W. Brown, PhDNutrition & Obesity Research Center - Office of
EnergeticsUniversity of Alabama at Birmingham
[email protected] Moderator:
James M. Rippe, MD – Leading cardiologist, Founder and Director, Rippe Lifestyle Institute
Approved for 1 CPE (Level 2) by the Commission on Dietetic Registration, credentialing agency for the Academy of Nutrition and Dietetics.
NUTRI-BITES®
Webinar Series
In the Eye of the Beholder:Critical Evaluation of Nutrition
Research
Original recording of the March 12, 2015 webinar and PDF download of presentation available at:www.ConAgraFoodsScienceInstitute.com
March 12, 2015
Based on this webinar the participant should be able to: State potential influences of biases in nutrition
research Discuss the importance of critically evaluating
new research (i.e. whether it confirms or refutes standard clinical practice or commonly held beliefs)
Describe steps to minimize misinterpretation of research
Identify strategies health professionals can use to objectively translate scientific knowledge to clinical practice
NUTRI-BITES®
Webinar Series
Critical Evaluation of Nutrition Research
• How do we know about nutrition?• Critically evaluating research to minimize
misinterpretation• What exactly was studied?• How exactly was it studied?• How does that compare to how it was communicated?
• Translating science to clinic or policy
Outline
Nutrition Quadrilateral
Research
How do we ‘know’ things in Nutrition Science?
Reason
Tradition ExperienceThe quadrilateral requires of a [nutrition scientist] no more than what he or she might reasonably be held accountable for: which is to say, a familiarity with [scientific literature] that is both critical and faithful; plus, an acquaintance with the wisdom of [nutrition science history]; plus, a taste for logical analysis as something more than a debater’s weapon…
adapted from Outler. Wesleyan Theological Journal. 1985;20:1,p17
How Researchers Define Snacks and Meals• By time: 8-10AM, 12-2PM, and 6- 8PM = meals; Other times = snacks
• By food composition/type: Based on ‘taxonomy’ of food, or calories in eating occasion
(Gregori et al, 2011; Gregori, & Maffeis, 2007)
How Individuals Define Snacks and MealsMeal Related-Perceptions Snack Related-Perceptions
Eating with family vs. Eating alone
Cloth napkin vs. Paper napkin
Sitting while eating vs. Standing while eating
Expensive vs. Inexpensive
Prepared food vs. Packaged food
‘Healthy’ food vs. ‘Unhealthy’ food(Adapted from Wansink et al, 2010. Appetite. 54(1), 214-16)
Abstract concepts
orThe Tale of Two Cheese Sandwiches
What exactly are we comparing?
“Whole” Food Processed Food
Bread Multi-grain bread with whole sunflower seeds and whole-
grain kernelsWhite bread
Cheese Cheddar cheese Processed cheese product
Fat 17.5 g 14.5 g
Protein 20 g 15 g
Carbohydrates 40 g 49.5 g
Sandwich 2 slices of bread2 slices of cheese
3 slices of bread2.28 slices of cheese
“We … offer the contrary view that [self-report measures of EI] are so poor as measures of actual EI … that they no longer have a justifiable place in scientific research aimed at understanding actual EI...” – N.V. Dhurandhar et al., Int J Obes (Lond). 2014 Nov 13
Normalweight Overweight0
1000
2000
3000
4000
Self-reported and Observer-es-timated Energy Intake
1 day food record Research dietary history
kcal
/d
“It appears, therefore, that, unless special precautions are applied to the study of the fourth of the adult population which is overweight, any data collected on the caloric intake of populations by the record method is likely to be an underestimate.”
Are the methods good enough?
JADA, January 1953
Spin perpetuates throughout the reporting
Spin: specific reporting strategies, intentional or unintentional, emphasizing the beneficial effect of the experimental treatment
Bias Presented to the Public
Abstracts were categorized based on results and conclusions about breakfast and obesity
Breakfast was more likely to be mentioned in conclusions if results were pro-breakfast (p=0.0492)
Biasing Interpretations of Own Results
Brown A W et al. Am J Clin Nutr 2013;98:1298-1308
Selective Reporting
22%
78%
Not Pro-Breakfast Results
65%35%
Pro-Breakfast Results
Mentioned in Conclusions
Not mentioned in conclusions
CONCLUSIONS: “These schoolchildren are exposed to an obesogenic environment, and it is not surprising that in this situation, many of these children are already overweight and will likely become obese as adults.”
RESULTS: “Based on our observations, it appears that those who have higher BMIs are less likely to consume fast food as often.”
Conclusions not matching results
PMID:22721691
Adapted from: Brown A W et al. Am J Clin Nutr 2013;98:1298-1308
Being cognizant of our own humanity
Food X is Bad!
Mere Exposure Effect
Food X is Bad!
Food Xis
BAD!
Food X is OKAY
Cognitive Dissonance
Confirmation Bias
0 10 20 30 40 50 60 70 80 90 1000
20
40
60
80
100
Strength of BeliefStrength of Evidence
Studies
Cert
aint
y (%
)
Discarded information
Often impossible to tell if something ‘worked’ for an individual (e.g., responders vs non-responders)• Improvements could have been spontaneous• Improvements may have been better with another
option• Impairments may have been mitigated
Scientific investigation tells us whether, on average, a group does better under one condition than another• RCTs in particular tell us whether a change in
exposure causes a change in outcome
The plural of anecdote is not ‘data’
Anecdote vs Data