ASSOCIATION
Prof. dr. Nisreen Mohammed
Community project lecture
Fifth year
2019-2020
It is the inference that are gained
from observation of a set of
cases of illness that are
accompanied by a suspected
factor, or it means relation
between two variables.
TYPES OF ASSOCIATION:
1- Direct association (causal association-
causation): It is the causal relationship,
which does not give any doubt. E.g.
measles virus cause measles disease.
2- Indirect association: The effect is due
to another hidden factor, it is most
commonly due to the presence of
confounder. E.g. smoking associated
with ischemic heart diseases but
confounder is coffee drinking.
3- Spurious association: There is no
association between the factor and the
effect, but the study was not properly
conducted. It also called fictitious
association.
4- Artifactual association: It is the
association that appears due to
faulty design or analysis of a study.
THE MAIN CRITERIA (MILL`S
CRITERIA) OF CAUSAL ASSOCIATION
The major criteria:
1. Strength of association: It is measured
by relative risk e.g. relative risk in lung
carcinoma among smokers compared to
non smokers is 8 times this mean high
relative risk then strong association
.
2. Temporal relationship (time
sequence is logical): This mean the
cause precedes the outcome. E.g.
H.pylori cause chronic gastritis then
duodenal ulcer.
3. Specificity of association: the outcome
specifically due to this cause not other, e.g.
the prevalence of H.pylori infection in patients
with duodenal ulcers is 90%-100%.
4. Consistency: This means different studies in
different times on different population by
different approach all give the same finding.
As e.g. about H.pylori and
D.U.
5. Coherence with the existing
Knowledge (biological plausibility):
e.g. the mucosa that is infected with
H.pylori will be weakened and will
become susceptible to the damaging
effects of the gastric acid.
THE FOLLOWING MINOR CRITERIA:
6. Dose response relationship: It means dose of
exposure increases then the risk of disease
also increase. E.g. increase cigarette smoking
increase risk of lung cancer.
7. The reversibility: If a factor is a cause of a
disease, the risk of the disease is expected to
decline if the factor reduced or eliminated.
8. Analogy: We analog the findings together.
THE RELATION BETWEEN FACTOR AND
OUTCOME AFFECTED BY:
1. Chance: This means that there is no relationship
between the factor and the outcome, and what we
had observed was due to chance. This can be
excluded by a proper statistical test which is p-value
and when it is less than 0.05 what does it mean?
This means that the chance does not have a role of
more than 0.05 (5%) in developing the outcome.
2. Confounder: It is a third factor sharing
with exposure that increased or
decreased the probability of outcome
occurrence. It is not an error in the
study.
3. Bias: It is a systematic error not
random in an epidemiological
studies that result in an incorrect
estimate of association between
exposure and outcome
.
Sources of Bias
1: Selection bias: selection of study group
individuals.
2. Observational bias (information): This
include:
2.a. Recall bias: case under study not
remember information as in case-control
study.
2.b. Interviewer bias: It occur in those
collecting data.
2.c. Loss to follow up: Either by
migration, death, or case refuse
continuation in study this happened in
cohort study.
2.d. Misclassification: It occurs when
subjects are wrongly categorized with
respect to either exposure or disease
state.