EPIDEMIOLOGIC
APPROACH TO CAUSATION
Dr Anthony Kapesa MD,MMED
Epidemiology
A study of the distribution of disease
frequency in human population and the
determinants of that distribution
Epidemiologists are not concerned with an
individual’s disease as clinicians do, but with a
population’ distribution of the disease
Distribution of disease by person, place, time
Assumption:
Disease does not occur randomly
Disease has identifiable causes
which can be altered and therefore
prevent disease from developing
Ecology of disease and models of
disease causation
Necessary and sufficient
causal
In principle, a cause can be necessary – without it
the effect will not occur – and/or sufficient – with
it the effect will result regardless of the presence or
absence of other factors.
Factors Influencing Disease
Transmission
Agent Environment
• Infectivity • Weather
• Pathogenicity • Housing
• Virulence • Geography
• Immunogenicity • Occupational
setting
• Antigenic
stability • Air quality
• Survival • Food
• Age
•Biological
Host
• Sex
•Physical
• Genotype
•Chemical
• Behaviour
•Nutritive
• Nutritional status
• Health status-immunity
Validity of Estimated
Association and Causation
Smoking Lung Cancer
OR = 7.3
True association Bias?
causal Confounding?
non-causal
Chance?
6
The Role of Bias, Confounding, and
Chance in The Estimated Association
absent
Association ?
present
present
Selection Bias and
Information Bias? False
absent
association
likely
Confounding ?
unlikely
likely
Chance ?
unlikely
7 True association
BIAS
Systematic errors in selection of study
subjects, collecting or interpreting data
such that there is deviation of results or
inferences from the truth.
Selection bias: noncomparable procedure used to
select study subjects leading to noncamparable study
groups in their distribution of risk factors. Example.
Community prevalence of HIV using VCT clinics
Information bias: bias resulting from measurement
error/ error in data collection (e.g. faulty instrument,
differential or non-differential misclassification of disease
and/ or exposure status. Example: interviewer bias,
recall bias)
Confounding
1. A mixing of effects
between the exposure, the disease, and a
third factor associated with both the
exposure and the disease
such that the effect of exposure on the
disease is distorted by the association
between the exposure and the third factor
2. This third factor is so called
confounding factor
Cases of Down syndroms by birth order
Cases per 100 000
live births
180
160
140
120
100
80
60
40
20
0
1 2 3 4 5
Birth order
Confounding
Observed (but spurious) association,
presumed causation
Birth Order Down’s
syndrome
Unobserved True
association association
Maternal age
Confounding
[Biomedical Bestiary: Michael, Boyce & Wilcox, Little Brown. 1984]
Observed (but spurious) association,
presumed causation
Gambling Cancer
Unobserved Smoking,
True
association Alcohol, association
other
Factors
Hill’s Criteria for Causation
1. Strength of association
2. Specificity
3. Temporal sequence
4. Biologic gradient (dose-response
relationship)
5. Biologic plausibility
6. Consistency
7. Coherence
8. Experimental study
9. Analogy
1: Strength of Association. The
stronger the relationship between the
independent variable and the dependent
variable, the less likely it is that the
relationship is due to an extraneous
variable.
2: Temporality.
The ability to establish that the cause in fact preceded
in time the presumed effect.
3: Consistency. Multiple observations,
of an association, with different people
under different circumstances and with
different measurement instruments
increase the credibility of a finding.
Replication of the findings by different investigators,
at different times, in different places, with different
methods and the ability to convincingly explain
different results.
4: Theoretical Plausibility. It is easier
to accept an association as causal when
there is a rational and theoretical basis
for such a conclusion.
Biologic Plausibility
Is consistent with current biological and
medical common knowledge.
Smoking
Ingesting of chemicals and known
carcinogens
DNA mutations
lung cancer
5: Coherence. A cause-and-effect interpretation
for an association is clearest when it does not
conflict with what is known about the variables
under study and when there are no plausible
competing theories or rival hypotheses. In other
words, the association must be coherent with
other knowledge.
6: Specificity in the causes. In the ideal
situation, the effect has only one cause. In other
words, showing that an outcome is best
predicted by one primary factor adds credibility
to a causal claim.
Specificity of the
association
Strengthens evidence if the cause has
ONLY one effect.
abestosis
Asbestos exposure mesothelioma
lung cancer
7: Dose Response Relationship.
There should be a direct relationship
between the risk factor (i.e., the
independent variable) and people’s
status on the disease variable (i.e., the
dependent variable).
8: Experimental Evidence. Any
related research that is based on
experiments will make a causal
inference more plausible.
The demonstration that under controlled conditions
changing the exposure causes a change in the
outcome is of great value, some would say
indispensable, for inferring causality.
9: Analogy. Sometimes a commonly
accepted phenomenon in one area can
be applied to another area.
If a similar agent exerts similar effects, it
is
more likely that the association is
causal.
