How to Read Health Research Without Being Misled

Health headlines often reduce a complicated study to “works” or “doesn't work”. A better reading asks what was studied, who took part, how large the result was and whether it matters to the decision in front of you.

Start with the question

Look for the population, the exposure or intervention, the comparison and the outcome. A study in a narrow group cannot automatically answer a question about everyone else.

What type of study was it?

A laboratory experiment, observational study, randomised trial and systematic review answer different questions. Study type alone does not guarantee quality. The methods, missing data, comparison group and risk of bias still matter.

How large was the result?

Statistical significance does not show whether a difference is important in everyday life. Check the size of the effect and its uncertainty, not just whether a result crossed a statistical threshold.

Absolute risk or relative risk?

An older NHMRC handbook on preparing consumer publications recommended usually presenting absolute risk. A relative change can sound dramatic when the starting risk is small. Look for the number of people affected in each group, not only a percentage increase or reduction.

Who funded it, and what did the authors declare?

Funding does not automatically invalidate research, but it is relevant context. Check author affiliations, declared conflicts, whether the protocol was registered and whether all planned outcomes were reported.

What are the limitations?

One study rarely settles a broad health question. Small samples, short follow-up, selective reporting and results that have not been replicated all reduce certainty. A result may be real but too small, too uncertain or too specific to guide your decision.

A practical reading check

Read beyond the headline. Find the original source, check who was studied, compare absolute numbers, read the limitations and look for a systematic review or current guideline. Discuss personal health decisions with a qualified professional rather than applying a study directly to yourself.

References

• National Health and Medical Research Council. How to Present the Evidence for Consumers: Preparation of Consumer Publications. 2000 (endorsed November 1999). The guide covers study level, quality, effect size, relevance and clear risk communication. Checked 2 September 2026.

This article provides general information about reading research. It does not assess a specific paper or replace advice from a qualified health professional.