Asthma epidemiology Flashcards
(29 cards)
What is ‘descriptive epidemiology’?
first stage of epidemiological investigation
focuses on describing disease distributions by place, time and person
often based on mortality rates
What levels can geographic variation in health and disease occur?
Between countries
Between regions
Between boroughs
Between neighbourhoods
What is an ‘ecological study’?
a study in which the unit of observation is the population or community
NOT the individual
When does an ecological study work best?
when there are differences in exposure and disease risks
What factors ARE NEEDED for geographic studies studying disease?
causal factor should remain relatively constant in one location
e.g dietary fat, weather, air pollution
How do migration studies inform geographic variation for disease?
examine whether disease risk reflects the adult environment
or
if its interplay between genes and environment
What age demographic usually migrates?
people in their 20s
For migration studies, which individuals are studies for disease risk?
- first generation immigrants
- second generation immigrants
What migration studies were performed for asthma?
- people moving from low to high asthma prevalence locations
- MIGRANTS: prevalence of asthma increases with duration of time spent in high asthma adoption country to reach levels of host population
- some studies, prevalence of asthma does not reach host population level until second generation
- IMPLICATION: strong environmental influence,
key expo may take an extended period to exert effects
What are the pitfalls of investigating geographic variation for disease risk?
Are the differences real?
1. differences in ASCERTAINMENT between countries (willingness to seek medical Rx, ability to obtain it)
- differences in DIAGNOSIS
- local customs or fashion
- diagnostic equipment - differences in RECORDING info
- differences in POPULATION STRUCTURE
- age
- gender etc
In geographic studies, how are problems in ascertainment of data overcome?
SOLUTION:
- population wide survey,
- using objective measures of disease risk
In geographic studies, how are problems in diagnosis or recording overcome?
SOLUTION:
- standardise diagnostic criteria e.g. WHO definitions, ICD
In geographic studies, how are problems in population structure overcome?
SOLUTION:
- take account of age and gender
- stratify data to adjust for these variable to prevent confounding
What are the special features of the ISAAC asthma study 1998?
prevalence of asthma Sx in questionnaire in 14yo
- population based study
- standard diagnostic criteria
- narrow age range
What is are the pitfalls of ecological studies?
analysis based on population groups not individuals
Assumption: expo and outcome are also related in individuals, which may not be true = ‘ECOLOGICAL FALLACY’
- Confounding: since groups may differ from one another in many ways
What are the pitfalls of migration studies?
- SELECTION BIAS: are the people who migrated the same as the people left behind?
- INFO BIAS: do immigrants provide info consistent with non-migrants in surveys?
- STRESS MIGRATION: can this contribute to disease patterns observed?
What should be considered if the geographic differences in disease are real?
- affect prevalence, incidence and mortality?
- reflect differences in causal factors?
- reflect differences in clinical prevention or in Rx?
e. g. screening programmes
What levels do time-trends in disease occur at?
- long term trends
- short term (seasonal for eg)
What clues can we gain for disease risk from time trends?
distinguish whether:
- recent cause (period effect0
- early life cause (cohort effect)
What is the period effect in disease risk?
usually seen in mortality rates
change in all age-groups occurs at the same time point
suggests that there is a cause operating at this time
What is the cohort effect in disease risk?
usually seen in mortality rates
change in all age-groups related to year at birth
suggests a cause operating from early life
How do time trends helot to inform disease risk?
assuming trends are real:
- tell us disease is not inevitable
- clues for disease causes
- testing out whether potential causes are important
What are the time time trends for asthma in UK?
- do not know why asthma incidences and admissions increased and then fell again 1980s-2000s
- but we DO know why there was an increase in asthma mortality in 1960s
- > increase in aerosol inhalers correlates with increased mortality
What are the best ways of overcoming confounding in time trend studies?
issues: confounding and analyses based on population not individuals
solution: run comparative studies in parallel (expo vs. outcome in individuals)
e. g. cohort, case-control, RCT
