Causal Inference

Dr. Amna Rehana Siddiqui

Department of Family and Community Medicine

Objectives:

Explain basic models of disease causation.Explain basic models of disease causation.

To understand concepts related to scientific inference for cause effect relation

To understand the applicability of causal criteria as applied to epidemiological studies

Approach to etiology To see whether a certain substance is an

agent / microorganism; a controlled laboratory experiment can be done by Exposing animals to organism Setting the exposure dose Monitoring environmental conditions Selecting genetic factors Minimum loss to follow up Species differ in response

Observations in Human populations

Cannot randomize human beings for harmful substances

Depend on nonrandomized observations Important populations – occupational cohorts Natural experiments

Residents of Hiroshima and Nagasaki Residents of Bhopal

Stages of disease and Levels of prevention Susceptibility

Pre-symptomatic

Clinical

Disability or Recovery

Primary prevention

Secondary prevention(Screening)

Tertiary prevention

Tertiary prevention

Development of Disease Combination of events

A harmful agent A susceptible host An appropriate environment

7

In epidemiology, there are several models of disease

causation that help understand disease process.

The most widely applied models are:

The epidemiological triad (triangle),

the wheel, and

the web. And

The sufficient cause and component causes models (Rothman’s

component causes model)

General Models of Causation

8

The Epidemiologic Triad

HOST

AGENT ENVIRONMENT

Agent factorsAgent factors

•Infectious agents: agent might be microorganism—virus, Infectious agents: agent might be microorganism—virus, bacterium, parasite, or other microbes. e.g. polio, measles, bacterium, parasite, or other microbes. e.g. polio, measles, malaria, tuberculosis Generally, these agents must be present malaria, tuberculosis Generally, these agents must be present for disease to occur. for disease to occur.

•Nutritive: excesses or deficiencies (Cholesterol, vitamins, Nutritive: excesses or deficiencies (Cholesterol, vitamins, proteins)proteins)

•Chemical agents: (carbon monoxide, drugs, medications)Chemical agents: (carbon monoxide, drugs, medications)

•Physical agents (Ionizing radiation,…Physical agents (Ionizing radiation,…

9

Host factorsHost factors

•Host factors are intrinsic factors that influence an individual’s Host factors are intrinsic factors that influence an individual’s

exposure, susceptibility, or response to a causative agent. exposure, susceptibility, or response to a causative agent.

•Host factors that affect a person’s risk of exposure to an agent:Host factors that affect a person’s risk of exposure to an agent:

•e.g. Age, race, sex, socioeconomic status, and behaviors e.g. Age, race, sex, socioeconomic status, and behaviors

(smoking, drug abuse, lifestyle, sexual practices and eating (smoking, drug abuse, lifestyle, sexual practices and eating

habits) habits)

•Host factors which affect susceptibility &response to an agent:Host factors which affect susceptibility &response to an agent:

•Age, genetic composition, nutritional and immunologic status, Age, genetic composition, nutritional and immunologic status,

anatomic structure, presence of disease or medications, and anatomic structure, presence of disease or medications, and

psychological makeup.psychological makeup.10

Environmental factorsEnvironmental factors

Environmental factors are extrinsic factors which affect the agent Environmental factors are extrinsic factors which affect the agent

and the opportunity for exposure. and the opportunity for exposure.

Environmental factors include: Environmental factors include:

physical factors such as geology, climate,.. physical factors such as geology, climate,..

biologic factors such as insects that transmit an agent; and biologic factors such as insects that transmit an agent; and

socioeconomic factors such as crowding, sanitation, and the socioeconomic factors such as crowding, sanitation, and the

availability of health services.availability of health services.

11

12

Agent

Host Environment

Vector

MalariaMalaria

13

Agent:Amount, infectivity, pathogenicity, virulence, chemical composition,

cell reproduction

Environment:Physical, biological, social

Host:Intrinsic factors, genetic, physiologic factors,

psychological factors, immunity

Health

or

Illness

?

The epidemiologic triad ModelThe epidemiologic triad Model

14

Web of CausationWeb of Causation

There is no single cause

Causes of disease are interacting

Illustrates the interconnectedness of

possible causes

The Web of causation

Developed to de-emphasis agent Chain of causation Complexity of origin is web Multiple factors promote or inhibit Emphasizes multiple interactions between

host and environment

Web of Causation

Disease

behaviourUnk

nown f

acto

rsgenes

phenotype

workplace

soci

al o

rgan

izat

ion

microbes

environment

16

Web of Causation - CHD

RS Bhopal

Disease

smokingUnk

nown f

acto

rsgender

genetic susceptibility

inflamm

ation

med

icat

ions

lipids

physical activityblood pressure

stress

17

18

Example of a Web of Causation

Susceptible Host Infection Tuberculosis

Vaccination Genetic

Overcrowding Malnutrition

Tissue Invasion and Reaction

Exposure to Mycobacterium

19

The Wheel of CausationThe Wheel of Causation

The Wheel of Causation de-emphasizes the

agent as the sole cause of disease,

It emphasizes the interplay of physical,

biological and social environments. It also brings

genetics into the mix.

20

The Wheel of Causation

Social Environment

Genetic Core

Physical Environment

Biological Environment

Host (human)

21

Association Vs. Causation

Association refers to the statistical dependence between two variables

The presence of an association…in no way implies that the observed relationship is one of cause and effect

22

Types of causes

Sufficient causes:a set of conditions without any one of which the disease

would not have occurrednot usually a single factor, often several

Necessary cause:must be present for disease to occur, disease never develops

in the absence of that factor.a component cause that is a member of every sufficient

cause

The sufficient cause and component causes model The sufficient cause and component causes model Rothman’s component causes modelRothman’s component causes model

23

24

Necessary and sufficient causesNecessary and sufficient causes

A A necessarynecessary cause is a causal factor whose presence is cause is a causal factor whose presence is

required for the occurrence of the effect.required for the occurrence of the effect. If disease does If disease does

not develop without the factor being present, then we term not develop without the factor being present, then we term

the causative factor the causative factor ““necessarynecessary”.”.

SufficientSufficient cause is a “minimum set of conditions, factors or cause is a “minimum set of conditions, factors or

events needed to produce a given outcome.events needed to produce a given outcome.

The factors or conditions that form a sufficient cause are The factors or conditions that form a sufficient cause are

called called componentcomponent causes.causes.

25

Example

The tubercle bacillus is required to cause The tubercle bacillus is required to cause

tuberculosis but, alone, does not always tuberculosis but, alone, does not always

cause it, cause it,

so tubercle bacillus is a so tubercle bacillus is a necessarynecessary,, not a not a

sufficient, cause.sufficient, cause.

26

Rothman'sRothman's model has emphasised that the causes of disease model has emphasised that the causes of disease

comprise a collection of factors. comprise a collection of factors.

These factors represent pieces of a pie, the whole pie These factors represent pieces of a pie, the whole pie

((combinations of factors) are the the sufficientsufficient causes for a causes for a

disease.disease.

It shows that a disease may have more that one sufficient It shows that a disease may have more that one sufficient

cause, with each sufficient cause being composed of several cause, with each sufficient cause being composed of several

factors.factors.

Rothman’s Component Causes and Component Causes and Causal Pies ModelCausal Pies Model

27

The factors represented by the pieces of the pie in this model

are called componentcomponent causes.

Each single component cause is rarely a sufficientsufficient cause by

itself, But may be necessarynecessary cause.

Control of the disease could be achieved by removing one of

the components in each "pie" and if there were a factor

common to all "pies“ (necessary cause) the disease would be

eliminated by removing that alone.

Rothman’sComponent Causes and Causal PiesComponent Causes and Causal Pies

28

Exercise Exercise

Some of the risk factors for heart disease are smoking,

hypertension, obesity, diabetes, high cholesterol, inactivity,

stress, and type A personality.

- Are these risk factors necessary causes, sufficient causes,

or component causes?

Causal pies representing all sufficient causes of a particular disease

29

Types of Associations

Real: probability depends upon the occurrence of one or more other events, characteristics, or other variables

Spurious: Non causal associations depend on bias, chance, failure to control for extraneous variables (confounding)

Percentage of pregnancies (n=50,267) with infant weighing < 2500 g by maternal cigarette smoking category (peri-natal mort study Comm Vol 1, 1967

4.7

7.7

12

0

2

4

6

8

10

12

14

Non smoker < 1 pack >=1 pack

% less than 2500 g

Percentage of LBW infants by smoking status of their mothers (Yerushalmay J, Am J Obs & Gynecol)

5.3

9.5 8.9

6

0

2

4

6

8

10

Non Smoker Non Smoker Smoker Smoker All pregnancies Future All Preg Future Smoker Ex smoker

% of LBW infants

33

“Is there an association between an exposure and a disease?”

IF SO…. Is the association likely to be due to chance? Is the association likely to be due to bias? Is the association likely to be due to

confounding? Is the association real/causal?

34

Establishing the cause of diseaseAssociation?Association?

Chance?Chance?

Bias ?Bias ?

Confounding?Confounding?

Causal?Causal?

presentpresent

absentabsent

absentabsent

absentabsent

absentabsent

presentpresent

likelylikely

likelylikely

35

Association Vs. Causation

Association refers to the statistical dependence between two variables

The presence of an association…in no way implies that the observed relationship is one of cause and effect

36

An association rarely reflects a causal

relationship but it may.

Criteria for causality provide a way of

reaching judgements on the likelihood

of an association being causal.

Epidemiological criteria (guidelines) for Epidemiological criteria (guidelines) for causalitycausality

37

Hill’s Criteria for Causal Relation

Strength of association Consistency of findings Specificity of association Temporal sequence Biological gradient (dose-response) Biological plausibility Coherence with established facts Experimental evidence

38

Strength of association association

Does exposure to the cause change disease incidence?

The strength of the association is measured by the relative risk.

The stronger the association, the higher the likelihood of a causal relationship.

Strong associations are less likely to be caused by chance or bias

39

Consistency of findings

Consistency refers to the repeated observation of an Consistency refers to the repeated observation of an

association in different populations under different association in different populations under different

circumstances. circumstances.

Causality is more likely when the association is repeated by Causality is more likely when the association is repeated by

other investigations conducted by different persons in different other investigations conducted by different persons in different

places, circumstances and time-frames, and using different places, circumstances and time-frames, and using different

research designs.research designs.

40

Specificity of association

It means that an exposure leads to a single or characteristic It means that an exposure leads to a single or characteristic effect, or affects people with a specific susceptibilityeffect, or affects people with a specific susceptibility easier to support causation when associations are easier to support causation when associations are

specific, butspecific, but this may not always be the casethis may not always be the case

as many exposures cause multiple diseasesas many exposures cause multiple diseases This is more feasible in infectious diseases than in non-This is more feasible in infectious diseases than in non-

infectious diseases, which can result from different risk infectious diseases, which can result from different risk agents. agents.

41

Temporal sequence (temporality)

Did the cause precede the effect? Did the cause precede the effect?

Temporality refers to the necessity that the cause must Temporality refers to the necessity that the cause must

precede the disease in time. precede the disease in time.

This is the only absolutely essential criterion.  This is the only absolutely essential criterion. 

It is easier to establish temporality in experimental and It is easier to establish temporality in experimental and

cohort studies than in case-control and cross-sectional cohort studies than in case-control and cross-sectional

studies.studies.

42

Biological gradient

Does the disease incidence vary with the level of exposure? Does the disease incidence vary with the level of exposure?

((dose-response relationship)dose-response relationship)

Changes in exposure are related to a trend in relative riskChanges in exposure are related to a trend in relative risk

A dose-response relationship (if present) can increase the A dose-response relationship (if present) can increase the

likelihood of a causal association.likelihood of a causal association.

43

Biological gradient(Dose Response)

44

Age standardized death rates due to bronchogenic carcinoma by current amount of smoking

Dose-response relationship

45

Biological plausibility

Is there a logical mechanism by which the

supposed cause can induce the effect?

Findings should not disagree with established

understanding of biological processes.

46

Coherence

Coherence implies that a cause-and-effect implies that a cause-and-effect

interpretation for an association interpretation for an association

does not conflictdoes not conflict with what is known of the with what is known of the

natural history and biology of the disease natural history and biology of the disease

47

Experimental evidence

It refers to evidence from laboratory It refers to evidence from laboratory

experiments on animal or to evidence from experiments on animal or to evidence from

human experiments human experiments

Causal understanding can be greatly advanced Causal understanding can be greatly advanced

by laboratory and experimental observations.by laboratory and experimental observations.

48

Judging the causal basis of the associationJudging the causal basis of the association

No single study is sufficient for causal inferenceNo single study is sufficient for causal inference It is always necessary to consider multiple alternate It is always necessary to consider multiple alternate

explanations before making conclusions about the explanations before making conclusions about the causal relationship between any two items under causal relationship between any two items under investigation.  investigation. 

Causal inference is not a simple processCausal inference is not a simple process consider weight of evidenceconsider weight of evidence requires judgment and interpretationrequires judgment and interpretation

49

Figure 5.12 The scales of causal judgement

Weigh up weaknesses in data and alternative explanations

Weigh up quality of science and results of applying causal

frameworks

Pyramid of Associations

RS Bhopal

Causal

Non-causal

Confounded

Spurious / artefact

Chance

50

51

Evaluating Evidence of Causal relationship Major Criteriaa. Temporal relationshipb. Biologic plausibilityc. Consistency of Resultsd. Alternative explanationsOther criteriaa. Strength of associationb. Dose-response relationshipc. Cessation of effects