Food and Cancer - L7 Flashcards
In the world-wide exercise what did “test of reproducibility” mean ?
all the areas studying the effects undertook the same methodology to ensure data findings could be replicated
Why did the WCRF do a world-wide study ?
because there was a huge increase in the number of papers from 1997-2007
What is a meta-analyses?
looking at lots of different tides with their own risk estimates. The data is then pooled together to form an overall risk estimate
In the stomach cancer review what different types of studies were used in the meta-analysis ?
11 RCTs 94 prospective studies 197-case control studies 10 cross sectional studies 135 ecological studies 7 case to case studies (genetic) 18- case- series studies
What is a forest plot?
it is plot used for meta-analysis studies
it is a typical plot for cohort or case control analysis
What do the lines and squares on the forest plot represent ?
each of the lines and squares represents a separate study -the size of the square indicates the population size of each study
estimate and confidence interval of each study is highlighted
What is the middle line of the plot and what does the pooled estimate show?
the middle line is the line of no effect - if an estimate lies on this line then there is no increase or decreased risk of that study
if the pooled estimate went across the line of no effect then it would not be statistically significant
the pooled estimate indicates the level of risk depending on which side on the line it lies
What is the link between the squares and lines associated with studies ?
even if the square of a study is indicating an increased or decreased risk, if the confidence interval crosses the line of no effect then it may not be statistically sginifcant due to lots of variation between studies
What does it mean by forest plots look at heterogeneity ?
they look at how variable the outcome are from different studies
- if very different studies had been pooled together then it would be meaningless and this would be shown in the meta-analysis plot
When looking at a forest plot how can you determine whether the results are similar across the studies ?
1) Eyeball test: do they look similar
2) Test o “null hypothesis” of no variation (p-value)
3) proportion of variation not due to chance (I>2)- important to check because it may have been manipulated . The higher it is= less confidence of pooling data appropriately
In the world-wide study of diet and cancer how did the experts draw conclusions from the information ?
CONVINCING- indicates component of diet was causal
-evidence from more than 1 study type and at least 2 cohort studies
-no substantial unexplained heterogeneity, measurement error or confounding factors
-confirmatory experimental evidence
PROBABLE
-evidence from at least 2 cohort studies or five case-control studies
-no substantial unexplained heterogeneity, measurement error or bias
-biological plausability
What are the WCRF Uk’s recommendations for cancer prevention ?
1) be as lean as possible without being underweight
2) be physically active for at least 30 minutes everyday
3) limit consumption of energy dense foods
4) eat more of a variety of plant foods
5) limit consumption of red meats and avoid processed meats
6) limit alcoholic drinks to 2 for men and 1 for women per day
7) limit consumptions of salty foods
8) dont use supplements to protect against cancer
What are the additional WCRF’s recommendations for women ?
Mothers to breastfeed and children to be breastfed
- action protects both mother and child
What is the link between red meat/processed meat and cancer ?
strong evidence that red and processed meat are causes of bowel cancer and that no amount of processed meat can be confidently shown not to increase risk
Why is processed meat worse than red meat ?
because during the processing process chemicals are added