The document discusses models for understanding disease causation and predicting the impact of interventions. For communicable diseases, the traditional pathogenic model linking biological causes to disease outcomes works well. However, for non-communicable diseases, this model is insufficient as social and behavioral factors must be integrated into causation models. Additionally, predictive models of intervention for non-communicable diseases must view causal relationships as complex networks rather than simple linear paths and conceptualize the role of social factors in terms of their functions within people's lives.

1. The integration of
social and biological mechanisms
for healthcare prediction and intervention
A follow up from:
The integration of social, behavioural and biological mechanisms
in models of pathogenesis
Mike Kelly, Rachel Kelly, and Federica Russo

2. Aetiology and prediction:
the difference between
pathogenesis and prevention
Mike Kelly & Federica Russo

3. Overview
The pathogenic approach for communicable diseases
Causal models of disease and Predictive models of interventions
Non-communicable diseases
Why the pathogenic model does not work
The contribution of ‘the social’
The role of human behaviour in disease aetiology
Predictive models of intervention
Regress analysis and the means-end relation
3

4. THE PATHOGENIC MODEL
FOR COMMUNICABLE DISEASES
4

5. Causes and mechanisms
The conceptualisation of disease
The outcome of exposure to a pathogen or other noxious factor
Pathogens
Cause disease
Initiate complex mechanisms that lead to disease
Complications
Multiple pathogens at work
Factors that mediate interactions
Individuals experience multiple morbidities
etc
5

6. Intervening on the pathogens
T1: enough knowledge about good health state, biopathogenesis of
disease, risk of getting disease, etc
Action A: treatment of disease, alleviation, protection from risk
protecting people from microorganisms through isolation,
providing clean water,
removing sewage,
immunisation and improving nutritional status and housing conditions
T2: predict evolution of disease, prevention, etc.
Underlying conception:
Necessary and sufficient conditions
6
T1  A  T2

7. NON-COMMUNICABLE DISEASES
7

8. Why the pathogenic model
does not work
NCDs: non-infectious, non-transmittable among people
T1  A  T2 often fails
Actions:
reduce exposure to some environmental factors;
advice about physical activity, nutrition, smoking habits, …
How much control do we have?
On environmental factors – to some extent
On human behaviour – much less
8

9. Asymmetry between
aetiology and prediction in NCDs
Aetiology
Biopathogenesis of CDs
Biological causes and
mechanisms
Behaviour does contribute to
risk in NCDs
Aetiology: bio-psycho-social
pathogenesis
Prediction
Public health interventions
T1AT2 model has been
largely successful
Intervention models did not
shift to a bio-psycho-social
approach
Or, if if it did, it happened very
late
9

10. THE CONTRIBUTION OF
‘THE SOCIAL’ TO AETIOLOGY
10

11. Sociology. And health.
Sociology attempts to explain and predict
human behaviour
Societies manifest observable patterns of change
Humans are thinking acting beings
Their thought and action take place within the constraints imposed by
social structures
What links behaviour and health?
11

12. Social causes are proximal
The proximal – distal distinction
Biological causes are proximal, social causes are distal
Distal causes do not exert direct influence on health
Hence, social causes are at best ‘classificatory devices’,
but not active causes in disease aetiology
Against the proximal – distal distinction
12

13. The ‘lifeworld’
Relationships with significant
others, neighbours, friends
Local services, shops
Communities and workplaces
The immediate physical and
microbiological
environments
Mediates exposure to toxins,
hazards, pathogens, etc
Drives health states of
individuals and populations
Is the product of the
interaction between human
agency and social structure
13

14. An integrated pathogenic approach
The ‘social’ and the ‘biological’ are integrated in the
aetiology
Behavioural factors are active parts of disease
mechanisms
An integrated pathogenic approach leads us
to rethink models of intervention
14

15. PREDICTIVE MODELS OF
INTERVENTION
15

16. Communicable diseases
Causal model of disease Predictive Model of intervention
Recovery /
improved health
status
Biological
mechanisms
Treatment /
public health
intervention
Exposure to
pathogens
Biological
mechanisms
Disease
Cause-effect
Means-end
16

17. Non-Communicable diseases
Causal model of disease Predictive model of intervention
Recovery /
improved health
status
Biological
mechanisms
Public health
intervention
Life world
Bio-psycho-
social
mechanisms
Disease
Multiple
bio-
psycho-
social
paths
17
Multipl
e
means-
end
relation
s

18. Some remarks
The ‘bio-psycho-social paths’ and ‘means-end relations’
complex networks
not linear causal relations
Models of interventions are conceptualised
in terms of means-end:
Identify the function of a psycho-social factor
Intervening on the function may lead to intervene on something
different than the corresponding cause
The function of psycho-social factor is highly context dependent
18

19. What is function?
In the context of a causal mechanism:
Functions are role-functions
The theoretical underpinnings of causal factors
They are part of the description of the functioning
of a component part of a mechanism
A strong conceptual link between functions and causes
19

20. Example:
alcohol consumption,
the lifeworld, and interventions

21. In the pathogenic approach
Alcohol consumption is a
‘single’, ‘homogeneous’
behaviour
Reduce exposure to the
pathogen ( = ethanol)
To reduce liver diseases,
cancer, obesity, accidents,
injury, violence
Actions: change in prices,
licensing regimes, education
campaign
21
T1  A  T2

22. Alcohol consumption is
a social structure
It varies across friends,
family, social groups,
populations, age
groups, etc
Alcohol consumption is
part of the lifeworld
of individual and of
groups
Targeted groups
Function of alcohol
consumption in their
lifeworld
Targeted interventions
22
In an integrated pathogenic approach

23. TO SUM UP
23

24. For communicable diseases
The pathogenic approach is largely
successful
Causal model of disease
Predictive model of intervention
For non-communicable diseases
The pathogenic approach is wanting
on both sides
Causal model of disease
Integrate bio-social mechanisms
Predictive model of intervention
Recast causal paths in terms of
means-end relation, according to
the functions of social factors in the
lifeworld
24

25. Classic Pathogenic Approach
Causal model of disease Predictive model of intervention
Recovery /
improved health
status
Biological
mechanisms
Treatment /
public health
intervention
Exposure to
pathogens
Biological
mechanisms
Disease
Cause-effect
Means-end

26. Complex Pathogenic Model
Causal model of disease Predictive model of intervention
Recovery /
improved health
status
Biological
mechanisms
Public health
intervention
Life world
Bio-psycho-
social
mechanisms
Disease
Multiple
bio-
psycho-
social
paths Multiple
means-
end
relations