Evaluating Evidence Flashcards
(14 cards)
Why is establishing diet-disease relationships complex?
Multiple factors contribute to disease risk, including protective lifestyle factors (e.g., diet, exercise) and harmful ones (e.g., drinking, smoking). Confounders, long latency periods, and the need for large, long-term studies add to the complexity.
What is a confounder, and how can it influence diet-disease studies?
A confounder influences both the exposure and the outcome. For example, smoking may affect fruit and vegetable consumption (exposure) and lung cancer risk (outcome), creating an apparent but potentially misleading association.
How do long latency periods affect diet-disease studies?
Long latency periods require extended study durations to observe effects, which can be challenging due to adherence issues and changing participant conditions over time.
Why are large sample sizes essential for diet-disease studies?
They are needed to detect meaningful differences, especially for uncommon outcomes, ensuring statistical and clinical significance.
What challenges arise in assessing nutrient intake in studies?
Challenges include inaccuracies in dietary assessment tools (e.g., FFQs, 24-hour recalls), potential misreporting, reliance on nutrient databases, and the invasive nature of quantitative methods like blood sampling.
What are the 5 key types of study designs used in nutrition research?
Ecological studies: Indicate trends but cannot establish causation.
Cross-sectional studies: Measure exposure and outcome simultaneously.
Case-control studies: Start with the outcome and trace back to the exposure.
Cohort studies: Begin with exposure and follow participants for outcomes.
Experimental studies (RCTs): Manipulate the independent variable to test causation
What makes randomized controlled trials (RCTs) the gold standard?
They minimize bias and confounding through randomization, double-blinding, and placebo control, allowing stronger causation inferences.
What are Hill’s 4 criteria for causation, and how do they apply to nutrition studies?
Consistency: Evidence is replicated across studies.
Strength of relationship: A strong association (e.g., 2-fold increase) is required.
Dose-response: Observes the effect of increasing exposure.
Biological plausibility: Mechanisms linking exposure to outcome must be reasonable.
How is evidence for diet-disease relationships categorized?
Convincing: Consistent, high-quality evidence from multiple study types.
Probable: Strong evidence but less robust than “convincing.”
Limited-suggestive: Evidence is emerging but not strong enough for firm conclusions.
Limited-no conclusion: Insufficient evidence.
What factors are considered in the hierarchy of evidence for clinical decisions?
Systematic reviews and meta-analyses rank highest, followed by RCTs, cohort studies, case-control studies, and in vitro research.
What is the significance of relative risk (RR) in nutrition studies?
RR compares the likelihood of an outcome in exposed vs. unexposed groups. An RR > 1 indicates a positive association, while RR < 1 indicates a negative association. Statistical significance occurs when the 95% CI does not include 1.
What is the role of a forest plot in meta-analyses?
It visually represents study results, with boxes showing individual study effects, their confidence intervals, and a diamond summarizing the overall effect.
How do confidence and benefit-risk ratios differ between nutrition and pharmacological studies?
Nutrition studies generally have lower confidence and benefit-risk ratios due to challenges in isolating dietary effects, while pharmacological studies are more controlled but may involve significant side effects.
What are 5 alternative explanations for associations between variables?
Cause-effect: The exposure directly influences the outcome.
Chance/random error: The association occurred by luck.
Bias/systematic error: Flaws in study design or execution.
Effect-cause: The outcome influences the exposure.
Confounding: A third variable affects both exposure and outcome.
