examine the association of disease with past
exposure(s) (e.g. abuse as a child and severe
depression as an adult).
can be completed quickly (unlike cohort studies where
people often need to be followed for several years to see if
certain conditions develop)
reasonably economical, and the study
population can be small
used to study relatively rare diseases.
identify the cause of food-related illness
after customers have eaten in a restaurant
(i.e. What did the people who got sick eat
[cases], that the people who did not get sick
[controls] refrained from eating?).
Results expressed as
results are expressed as odds ratios (OR).
estimate the increased (or decreased)
risk of disease for those in the exposed
group
Order of inquiry
disease status is measured first, then participants are
followed up (backwards in time) until exposure status is
established, and the two groups are compared.
E.g
1.Are oral contraceptives
associated with breast
cancer?
Cross sectional studies
provide data from a single 'snapshot'
in time and are often used to study
prevalence (current disease or
outcome)
used to measure variables such as health-related
attitudes and behaviours in a population or sample
of the population.
Analytical cross-sectional studies/surveys
'getting the temporal direction of cause-effect
correct', and cross-sectional studies cannot, in
themselves, establish this.
does excessive television
watching cause obesity, or
are these two variables
simply associated?
either the entire population or a
sample from that population is
selected
individuals' data are collected using records
or a questionnaire (or some type of measuring
instrument) that focuses on the area of
interest.
Examples
1.What is the prevalence of
stroke in Victoria in 2012?
2.Are young
people aware of,
and involved in,
tobacco marketing?
Results
expressed in terms frequencies and proportions (%)
statistical approaches like Chi-square (x2)and p-values.