Introduction
Evidence-based nutrition uses the best available scientific evidence, professional expertise and relevant individual or population circumstances to guide nutrition decisions. Nutrition recommendations should be supported by reliable research rather than isolated claims, marketing messages or personal anecdotes.
Foundations
| Foundation | Explanation |
|---|---|
Scientific evidence | Recommendations should be based on appropriately designed studies and a critical evaluation of their results. |
Professional interpretation | Nutrition scientists and qualified healthcare professionals evaluate evidence, limitations and clinical relevance. |
Individual and population context | Recommendations must account for age, health status, culture, food availability and the needs of the target population. |
Transparency | Scientific guidelines should explain their evidence base, assumptions, uncertainties and intended population. |
Types of nutrition research
| Study type | Description | Strengths | Limitations |
|---|---|---|---|
Randomized controlled trial | Participants are assigned to different interventions or comparison groups using randomization. | Can help evaluate causal effects of a dietary intervention when well designed and conducted. | Dietary interventions can be difficult to control, and results may not represent long-term effects or every population. |
Prospective cohort study | Researchers follow a group of participants over time and examine relationships between dietary exposures and health outcomes. | Useful for studying long-term dietary patterns and health outcomes. | Observational associations can be affected by confounding factors and measurement errors. |
Case-control study | Researchers compare people with a particular outcome with people without that outcome and investigate previous exposures. | Can be useful for studying uncommon diseases or outcomes. | Recall bias and selection bias may influence findings. |
Cross-sectional study | Dietary exposures and health characteristics are measured at a particular point or period in time. | Useful for describing population patterns and estimating prevalence. | Usually cannot establish which factor occurred first or demonstrate causation. |
Systematic review | Researchers use predefined methods to identify, select and critically evaluate studies addressing a specific research question. | Provides a structured summary of relevant research. | The conclusions depend on the quality, consistency and completeness of the available studies. |
Meta-analysis | A statistical method that combines results from multiple sufficiently comparable studies. | Can improve statistical precision and investigate patterns across studies. | Combining biased or highly heterogeneous studies does not automatically produce reliable conclusions. |
Evidence evaluation
Important questions
Is the research question clearly defined?
Is the study design appropriate for the question?
Was the sample sufficiently large and representative?
Were dietary intake and health outcomes measured reliably?
Were important confounding variables considered?
Are the findings statistically and clinically meaningful?
Are the results consistent with other high-quality research?
Can the findings reasonably be applied to the target population?
Evidence hierarchy
Different study designs provide different forms of evidence. A randomized trial may be particularly useful for evaluating an intervention, while well-conducted observational research is often essential for studying long-term dietary patterns. Evidence quality depends on methods and context, not only on the study's label.
Development of dietary recommendations
| Step | Name | Explanation |
|---|---|---|
1 | Define the target population | Identify the population, life stage, health context and dietary issue being addressed. |
2 | Formulate the research question | Specify the nutrient or dietary pattern, relevant comparison and health outcome. |
3 | Collect scientific evidence | Search relevant scientific literature, dietary reference reports and authoritative public-health information. |
4 | Evaluate evidence quality | Assess study design, bias, consistency, precision, relevance and uncertainty. |
5 | Interpret nutritional significance | Determine whether the observed effects are meaningful for health and applicable to the target population. |
6 | Develop recommendations | Translate findings into clear, practical and culturally appropriate dietary guidance. |
7 | Implement and communicate | Share recommendations through healthcare, education, food systems and public-health programmes. |
8 | Monitor and update | Evaluate implementation and revise recommendations when new reliable evidence becomes available. |
Authoritative guidance
| Organization | Role |
|---|---|
World Health Organization | Publishes global public-health guidance on healthy diets, malnutrition and nutrition-related health issues. |
Food and Agriculture Organization | Provides scientific and technical resources on food systems, food security and nutrition. |
National Academies | Publishes scientific reports and dietary reference frameworks for nutrient requirements. |
National health authorities | Develop national dietary guidance based on scientific evidence and population-specific circumstances. |
Recognizing unreliable claims
Claims that one food or supplement cures many unrelated diseases.
Recommendations based only on personal testimonials.
Studies involving very small samples presented as universal proof.
Correlation presented as proof of causation.
Research results exaggerated beyond the population or conditions actually studied.
Commercial claims that do not disclose relevant evidence or conflicts of interest.
Limitations and uncertainty
Nutrition research is challenging because people consume complex dietary patterns rather than isolated nutrients.
Long-term dietary intake is difficult to measure accurately.
Observational studies may be affected by confounding variables.
Individual responses to dietary patterns can vary.
Scientific recommendations may change as stronger evidence becomes available.