Block 15 Flashcards
What types of study designs can be used to evaluate the affect of diet on cancer?
- cross-sectional observational studies of different countries diet and cancer risk
- cross-sectional migrant studies
How much of cancer is preventable by appropriate food, nutrition, physical activity, and body fatness?
- total for all cancers = 26% in UK
- but obviously different for different cancer sites = as high as 75% for oesophageal cancer in the UK
Which exposures are most attributable to cancer cases in the UK?
- tobacco 19.4%
- diet 9.2%
- overweight and obese 5.5%
- alcohol 4%
What are the different types of population research?
Descriptive
- survey
- case report
- case series
Analytic
- experimental (trials)
- observational (cohort, case-control)
What are the problems with using case-control designs to study diet?
- variety of accuracy in recall of diet between cases and controls (cases are more motivated to accurately report exposures than controls)
- possible early impact of disease on diet (pre-symptomatic disease state may have influenced dietary choices - studies often disregard 1-2years before diagnosis)
- measurement of diet is difficult and often ambiguous
- bias
- confounding
What are the problems with using cohort designs to study diet?
- hard to accurately measure diet in large cohorts
- difficult to maintain follow-up over the long periods of time that are necessary to power the study
- bias
- confounding
Why is it difficult to accurately measure and study diet?
- random error (diet varies and people may make mistakes in their reporting)
- homogeneity of exposure (at a population level, e.g. UK students, diet is actually largely the same)
- bias
- confounding
How can you measure diet?
- food disappearance data
- household surveys
- individual surveys
- 24 hour recall
- food frequency (FFQ)
- diet diary
- biomarkers
What are the pros and cons of using food frequency questionnaires in studying the effects of diet on health?
Pros:
- captures usual diet
- less work to code of compute
Cons:
- doesn’t record actual diet as eaten
- overestimates fruit and veg
- poor measure of energy intake
- less flexible
What are the pros and cons of using diet diaries in studying the effects of diet on health?
Pros:
- records diet as eaten (over a limited period)
- more flexible
- better estimate of energy and absolute intake
Cons:
- requires effort to complete
- expensive and time-consuming to code
- alters diet when the diary is being completed
What are the main associations between oesophageal cancer and diet?
Increased risk:
- Convincing evidence for alcohol (squamous) and obesity (adenocarcinoma)
What are the main associations between stomach cancer and diet?
Increased risk:
- Probable evidence for body fatness, alcohol, salted preserved food, meats and pickles
What are the main associations between pancreatic cancer and diet?
Increased risk:
- Convincing evidence for obesity
- Probable evidence for height
What are the main associations between hepatic cancer and diet?
Increased risk:
- Convincing evidence for aflatoxin contamination, alcohol, body fatness
Decreased risk:
- Probable evidence for coffee
What are the main associations between colorectal cancer and diet?
Increasing risk:
- Convincing evidence for processed meat, alcohol, body fatness, central obesity, height
- Probable evidence for red meat
Decreased risk:
- Convincing evidence for more physical activity
- Probable evidence for dietary fibre, wholegrain, dairy, calcium supplementation
What are the main associations between pre-menopausal breast cancer and diet?
Increased risk:
- Convincing evidence for height
- Probable evidence for alcohol and high birthweight
Decreased risk:
- Probable evidence for more physical exercise, overweight/obesity, lactation
What are the main associations between post-menopausal cancer and diet?
Increased risk:
- Convincing evidence for alcohol, body fatness, height, adult weight gain, central obesity
Decreased risk:
- Probable evidence for more physical activity, body fatness in young adulthood, lactation
What are the main associations between lung cancer and diet?
Increased risk:
- Convincing risk for arsenic in drinking water, high dose beta-carotene supplementation
What are the main associations between prostate cancer and diet?
Increased risk:
- Probable evidence for body fatness, height
What are the main associations between renal cancer and diet?
Increased risk:
- Convincing evidence for overweight
- Probable evidence for adult height
Decreased risk:
- Probable evidence for alcohol
What are the main associations between endometrial cancer and diet?
Increased risk:
- Convincing evidence for overweight/obesity
- Probable evidence for glycemic load
Decreased risk:
- Probable evidence for physical activity and coffee
Why “five a day” ?
- evidence that less than 200g/day fruit/veg is associated with increased cancer risk
- little benefit observed beyond 400g/day
- average portion size = 80g, 400g/80g = 5 portions
- however continuing benefit shown for increased fruit/beg intake in reducing CVD risk
What are the health promotion messages relevant to cancer?
- increase levels of physical exercise
- don’t put on weight in adulthood
- aim for a BMI between 18-25
- maintain safe levels of alcohol intake
- increase fruit/veg intake to at least 400g/day
- limit intake of preserved and red meat
What are clinical decision support systems?
- systems designed to aid clinical decision making, not replace it
- provide clinicians with patient-specific assessments or recommendations to aid clinical decision making
- different types available for different purposes:
- computerises (including apps)
- paper based
- reminder systems/prompts
Outline the potential uses of reminder systems:
- screening
- vaccination
- testing
- medication use
- identification of risky behaviour
- e.g. recall system, allergy pop-ups, reminders to consent for online services
Give examples of clinical decision support systems:
- reminder systems
- decision systems (diagnoses and treatment)
- prescribing systems
- condition management
What are decision systems (for diagnoses and treatment)?
- model individual patient data against epidemiological data (which may be held in a computer database)
- often match patient signs and symptoms to database to provide hypotheses or estimates of probability of different potential diagnoses
- e.g. Ottawa ankle rules