Understanding Nutrition Research
Evidence matters, but it's complex. Some nutrition interventions have decades of rigorous research showing they work. Others are promising but understudied. Some work brilliantly in controlled settings but fail when scaled. Context always matters.
This hub helps you:
- Understand what "strong evidence" actually means
- Know which interventions have the best research support
- Identify important evidence gaps
- Think critically about research claims
- Navigate the difference between "works in trials" and "works at scale"
⚠️ Important Caveat
Research evidence is necessary but not sufficient. Even interventions with strong evidence require adaptation to local context, political feasibility, implementation capacity, and community acceptance. Don't let "evidence-based" become an excuse for ignoring on-the-ground reality.
The Evidence Pyramid
Not all research is created equal. Here's how to assess evidence quality:
💡 Using the Pyramid
Start at the top, move down as needed. If systematic reviews exist, start there. If not, look for RCTs. But don't dismiss lower-level evidence—it's often the only evidence available for emerging interventions or specific contexts. The key is knowing what each level can and can't tell you.
What We Know Works
These interventions have strong research support across multiple contexts:
🔍 Critical Evidence Gaps
Important questions we don't have good answers for:
🤔 Thinking Critically About Research
Ask these questions when reading research or hearing claims:
- Who funded the study? Industry-funded research often shows more favorable results than independent research.
- Was it efficacy or effectiveness? Efficacy = ideal conditions. Effectiveness = real-world conditions. Big difference.
- What was the comparison group? Comparing to "no intervention" shows bigger effects than comparing to "standard care."
- How long was follow-up? Short-term effects often disappear. Beware studies with <6 month follow-up.
- What was the attrition rate? If >20% of participants dropped out, results may be biased.
- Is it externally valid? Study in urban Kenya may not apply to rural Bangladesh. Context matters immensely.
- What was measured? Weight gain is easier to detect than reducing stunting. Absence of evidence ≠ evidence of absence.
- Was it adequately powered? Small sample sizes can miss real effects (Type II error).
- Were harms reported? Studies often emphasize benefits and underreport harms or unintended consequences.
- Who's making the claim? Is it the researchers, or someone interpreting their work? Read the actual paper if possible.
Finding Reliable Research
From Research to Action
✅ Don't Wait for Perfect Evidence
Perfect evidence rarely exists. Some of the most impactful public health interventions (sanitation, vaccination, clean water) were implemented before RCTs existed. Use best available evidence, acknowledge uncertainty, monitor outcomes, and adjust.
The ethical question isn't "Do we have perfect evidence?" It's "Given what we know and don't know, what's the most responsible action to take?"
Practical Steps
- Start with systematic reviews if they exist for your intervention of interest
- Look for evidence from similar contexts to yours—not just any context
- Assess implementation feasibility separately from efficacy. Something can "work" but be impossible to implement
- Pilot before scaling when evidence is limited or context is very different from study settings
- Build in monitoring and learning systems from day one. You're contributing to evidence even if not doing formal research
- Be transparent about uncertainty with stakeholders, beneficiaries, and funders