Introducing 

Prezi AI.

Your new presentation assistant.

Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.

Loading…
Transcript

Relative risk (risk ratio)

What is it?

An example

In a cohort study of 50,277 female nurses aged 30-55 years who were non-obese in 1992 were followed over 6 years. At cohort enrolment 7,474 reported watching television for >20 hours each week. By 6 year follow-up 616 of these participants had become obese (BMI >= 30 kg/m2). Of the 42,803 nurses who watched TV less frequently at baseline, 3,141 developed obesity by 6 year follow-up.

Is watching TV for >20 hr/wk associated with an increased risk of becoming obese? How strong is the association?

Source: Hu et al. JAMA 2003 https://www.ncbi.nlm.nih.gov/pubmed/12684356

Check your learning

The 2x2 table and how to calculate the relative risk

  • A measure of the direction and strength of association between a risk factor and a disease (or other outcome) of interest

  • Commonly used in cohort studies

  • Values range from 0 to infinity

  • Values >1 = a 'positive' association (i.e. the presence of the risk factor is associated with a higher risk of the disease)

  • Values <1 = a 'negative' association (i.e. the presence of the risk factor is associated with a lower risk of the disease - it is protective)

The relative risk (risk ratio) is a measure of association between a risk factor and an outcome of interest

In cohort studies can you recognise which is the exposure (risk factor) and which is the outcome?

Can you correctly describe the basic design of a cohort study?

Can you calculate a relative risk?

Can you correctly interpret a relative risk?

Interpretation

Relative risk* = absolute risk of outcome among exposed / absolute risk of

outcome among non-exposed

= (616/7474) / (3141/42803)

= 0.082 / 0.073

= 1.12

The risk of developing obesity is 1.12 times higher in adults who watch TV for more than 20 hours each week, than adults who watch TV less than this, i.e. regular TV watching in adulthood appears to be a risk factor for becoming obese.

But how precise is this estimate? What is the confidence interval?

In this example the 95% confidence interval is (1.03, 1.22). The increased risk of obesity associated with regular TV watching is statistically significant.

But still need to consider the role of bias and confounding

* This is the crude or unadjusted relative risk (i.e. before adjusting for potential confounders)

Learn more about creating dynamic, engaging presentations with Prezi