Everyday Mobility and Health E-Book
142,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Geisteswissenschaft
- Serie: ISTE Consignment
- Sprache: Englisch
Everyday mobility is neither favorable nor unfavorable to health. While it can facilitate social interactions, increase access to remote services, or encourage physical activity, it can also generate pollution, promote the spread of epidemics or cause traffic accidents.
This book presents different facets of the relationship between daily mobility and health, focusing on the environments (geographical, social and political) that people live and move around in. It analyzes the role of mobility in the mechanisms of environmental exposure and diffusion, as well as the resulting health inequalities. It deals with active modes of travel (mainly walking and cycling) and the local contexts that are conducive to them. Finally, it offers a critical reading of the place given to everyday mobility in policies to combat obesity and rationalize regional healthcare provision.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 392
Veröffentlichungsjahr: 2024
Ähnliche
Table of Contents
Cover
Table of Contents
Title Page
Copyright Page
Introduction
Part 1: Diffusion and Exposure Mechanisms Related to Daily Mobility
1 Daily Mobility and the Spread of Communicable Infectious Diseases
1.1. Introduction
1.2. Analyzing the role of daily mobility in the spread of epidemics: concepts, studies and tools
1.3. Modeling epidemics and daily mobility
1.4. The importance of mobility data in the analysis and propagation modeling of urban epidemics
1.5. Conclusion
1.6. References
2 Methodological Challenges of the Cross-Analysis Between Daily Mobility, Environments and Health
2.1. Acknowledging daily mobility in health studies: the concept of activity space
2.2. Activity space: from the concept to its measurement
2.3. Towards a more precise estimation of the relationship between the environment and health
2.4. The importance of the “selective daily mobility bias” or the impact of individual preferences in daily mobility
2.5. Conclusion: towards new methodological developments
2.6. References
3 Daily Mobility and Social Inequalities in Health: A Conceptual Framework and Application
3.1. Introduction
3.2. Social inequalities in health: definition, extent and progression
3.3. Conceptual framework
3.4. A contemporary example: the Covid-19 pandemic
3.5. Conclusion
3.6. References
Part 2: Everyday Environments and Active Modes of Travel
4 Walking in Everyday Life: The Built Environment and Pedestrian Insecurity
4.1. The dimensions of walkability
4.2. The experience of insecurity: when walking is a source of fear
4.3. Road safety and walking: potential danger when crossing
4.4. What interventions are needed to reduce collisions involving pedestrians and the severity of their injuries?
4.5. Avenues of research to improve pedestrian mobility
4.6. Conclusion
4.7. References
5 The Geographical Dimension of Daily Active Mobility
5.1. Introduction
5.2. Environment and active mobility: what is the relationship?
5.3. Territorial heterogeneity of the relationship between living environment and active mobility: the importance of context effects
5.4. From the promotion of active mobility in public health to health-enhancing urban planning
5.5. Conclusion
5.6. References
Part 3: Daily Mobility and Public Health Policies
6 A Critical Analysis of Policies Promoting Physical Activity and Active Mobility
6.1. Introduction
6.2. Historical context of recommendations for daily physical activity
6.3. Policies and actions to promote physical activity in Canada and France
6.4. The implicit values underlying policies promoting physical activity and active mobility: what questions should be asked?
6.5. Discussion
6.6. References
7 Rationalization of the Healthcare Provision and Mobility
7.1. Introduction
7.2. Spatial accessibility at the heart of healthcare rationalization policies
7.3. Density indicators: mobility left aside
7.4. The distance indicators: the concept of accessibility is introduced by traveling
7.5. The use of travel flows to define planning territories
7.6. Floating sector methods: greater integration of mobility
7.7. Rationalization policies take a new direction in France and turn away from mobility
7.8. Conclusion: limitations of the methods used by the various healthcare rationalization and territorial regulation policies
7.9. References
List of Authors
Index
End User License Agreement
List of Tables
Introduction
Table I.1. Synthesis of the themes developed in the seven chapters of the book
Chapter 4
Table 4.1. Haddon matrix for senior pedestrians
List of Illustrations
Chapter 1
Figure 1.1. Mobility and types of pathogen transmission
Figure 1.2. Illustration of the principle of spatialized compartmental models,...
Figure 1.3. Complex host–pathogen system of dengue fever
Figure 1.4. Representation of the activity space within a health context
Figure 1.5. Location of a telephone near relay antennas a) Exact telephone loca...
Chapter 2
Figure 2.1. Representation of the activity space at different ages
Figure 2.2. Spatial representations of the activity space
Chapter 4
Figure 4.1. Walkability dimensions and main indicators
Figure 4.2. Fatality distribution by type of road user in different WHO region...
Figure 4.3. Impact of speed on braking distance and risk of death.
Chapter 5
Figure 5.1. Examples of urban signage systems dedicated to active mobility on ...
Figure 5.2. a) Development of the experimental cycling network in Montpellier,...
Chapter 7
Figure 7.1. Schematic geography of non-compulsory travel in an Île-de-France d...
Guide
Cover Page
Table of Contents
Title Page
Copyright Page
Introduction
Begin Reading
List of Authors
Index
WILEY END USER LICENSE AGREEMENT
Pages
iii
iv
xi
xii
xiii
xiv
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
109
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
231
232
233
234
235
236
SCIENCES
Geography and Demography, Field Director – Denise Pumain
Infrastructure and Mobility Networks Geography,Subject Heads – Hadrien Commenges and Florent Le Néchet
Everyday Mobility and Health
Coordinated by
Julie Vallée
First published 2024 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.
Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:
ISTE Ltd27-37 St George’s RoadLondon SW19 4EUUK
www.iste.co.uk
John Wiley & Sons, Inc.111 River StreetHoboken, NJ 07030USA
www.wiley.com
© ISTE Ltd 2024The rights of Julie Vallée to be identified as the author of this work have been asserted by her in accordance with the Copyright, Designs and Patents Act 1988.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s), contributor(s) or editor(s) and do not necessarily reflect the views of ISTE Group.
Library of Congress Control Number: 2023949778
British Library Cataloguing-in-Publication DataA CIP record for this book is available from the British LibraryISBN 978-1-78945-109-2
ERC code:SH2 Institutions, Values, Environment and Space SH2_8 Energy, transportation and mobility SH2_11 Human, economic and social geographySH3 The Social World, Diversity, Population SH3_9 Health, ageing and society
Introduction
Julie VALLÉE
Géographie-cités, CNRS, Paris, France
From the individual to the geographical space
In the field of health, the daily mobility of individuals can be considered through the lens of the advantages it provides (reduction of social isolation, access to services and equipment far from places of residence, physical activity, etc.), or in terms of the inconveniences it causes (exposure to pollution, traffic accidents, strenuous commuting, etc.). On the contrary, we can also analyze the way in which the physical and mental health of an individual promotes or hinders their daily travels, whether in terms of the distance covered or the availability of modes of transport, and how these are actually used. These different approaches focus on the individual as the entry and exit point; the challenge is to study the influence of the individual’s daily mobility on health, as well as the influence of the individual’s health on mobility.
However, this relationship between everyday mobility and health does not take place at the level of independent individuals: the daily mobility of some people has an impact on the daily mobility of others (e.g. traffic jams can lead some individuals to shift their departure time, to modify their journey or to change their modes of transport), whereas the health of some people has an impact on the health of others (by contagion effect). Therefore, both mobility and health are the product of the interactions between individuals. But individuals are not the only actors involved in this system: social and political structures are also present. Public actors organize the mobility and healthcare provision throughout the local and national territory and promote their uses among individuals. Not only do social structures affect the mobility and health behaviors of individuals, but also public health and regional planning policies.
The role of geographical space within this system is multifaceted: it influences the daily mobility of individuals through the transport system and the built environment. Depending on geographical places, people’s living conditions may vary, together with the local health norms or the availability and quality of the healthcare provision. It would nonetheless be simplistic to imagine that geographical space has a compartmentalized impact on people’s daily mobility, on the one hand, and on their health, on the other hand. In fact, it is the geographical space that structures their relationship. To illustrate this mechanism, we can observe that the distribution of the healthcare provision throughout the territory is not independent from the mobility of individuals, or that the trips of some transform the quality of the living environments of others, and as such, their health.
Organization of the book
Without claiming to exhaust the richness and complexity of the subject, the purpose of this work is to present the different aspects of the relationship between everyday mobility and health, taking into account the social, political and geographical components of the system in which this relationship thrives. This book is made up of seven chapters structured into three main parts.
Part 1 is devoted to the diffusion and exposure mechanisms related to the daily mobility of individuals. Chapter 1, written by Alexandre Cebeillac and Éric Daudé, illustrates how daily mobility – which modifies the exposure of individuals to pathogens – renews research on the spatial and temporal diffusion of communicable diseases. Chapter 2, written by Camille Perchoux, presents the methodological issues related to taking into account people’s daily mobility when measuring environmental effects on health. These first two chapters complement one another: they discuss the methodological issues relating to the consideration of everyday mobility in the diffusion of health problems within the geographical space and in the exposure of individuals to geographical spaces. Chapter 3, written by Martine Shareck, also addresses the exposure of individuals to the environment, but specifically calls into question the social inequalities that socially differentiated mobility induces in terms of exposure and health.
Part 2 is dedicated to daily environments and active transportation (mainly walking and cycling) at a time when this trend is generating great interest: on the one hand, because it induces increased physical expenditure (and in this way can reduce the risk of obesity and associated chronic pathologies), and on the other hand, because it limits the carbon emissions from car travel, contributing to the fight against global warming. The use of active transportation is eased (or hindered) by certain local characteristics of the built environment (see the notions of walkability and bikeability), which will be detailed and discussed in the fourth and fifth chapters of this book. Chapter 4, co-written by Marie-Soleil Cloutier and Karine Lachapelle, specifically addresses walking and insists on the various sources of insecurity that pedestrians may encounter in the geographical space. In Chapter 5, Thierry Feuillet and Hélène Charreire revisit the local contexts that are more or less favorable to active transportation, with a special focus on the territorial heterogeneity of relationships between the living environment and active mobility. These two chapters present several interventions intended to promote (and secure) the use of active transportation and discuss the difficulty in evaluating the effectiveness of such experiments.
Addressed by the two previous chapters in connection with active mobility, the question of policies – and the place they give to daily mobility – is at the heart of Part 3 of the book. Chapter 6, written by Stephanie Alexander, offers a critical analysis of the actions and policies that promote physical activity and active mobility by detailing their underlying assumptions and values around public health. Chapter 7, co-authored by Véronique Lucas-Gabrielli and Catherine Mangeney, discusses the role of daily mobility in the policies for rationalizing the healthcare provision. It presents the way in which everyday mobility is taken into account in measures of spatial accessibility to healthcare and in the delimitation of priority areas with insufficient healthcare provision.
Following this three-part structure is only a key to reading the book. The seven chapters can also be read according to the health indicators considered, the ways of analyzing daily mobility, the mechanisms involving the geographical space and the living environments or the challenges (methodological, political or contemporary) raised and debated (see Table I.1).
Table I.1.Synthesis of the themes developed in the seven chapters of the book
Part
1
2
3
Chapter
1
2
3
4
5
6
7
Health indicators
Communicable diseases
•
Obesity (and associated chronic diseases)
•
•
Road accidents
•
Access to healthcare
•
Diverse and varied indicators
•
•
Ways to analyze daily mobility
Spatio-temporal dynamics
•
Activity space
•
•
Social and spatial justice
•
•
Accessibility to services
•
•
•
Injunctions; constrained versus chosen mobility
•
•
•
Benefits and risks associated with active mobility
•
•
•
Geographical space and the living environments…
in which pathogens spread
•
to which individuals are exposed
•
•
•
•
•
•
where the services are localized
•
•
•
favorable to active mobility
•
•
•
Challenges…
in methodology (data and measures)
•
•
•
•
•
for regional planning
•
•
•
for public health
•
•
•
•
•
•
•
of today’s world (Covid-19…)
•
•
•
•
PART 1Diffusion and Exposure Mechanisms Related to Daily Mobility
1Daily Mobility and the Spread of Communicable Infectious Diseases
Alexandre CEBEILLAC1,2 and Eric DAUDÉ1
1 IDEES, CNRS, Normandie Université, Rouen, France
2 Institut Pasteur, Paris, France
1.1. Introduction
Communicable infectious diseases are caused by microorganisms, called pathogens (viruses, fungi, bacteria, parasites) that infect a living organism, called the host. Pathogens are transmitted from an infected host to a healthy host through various channels. A first category of transmission channels involves direct, physical contact (saliva, blood, sexual contact: AIDS, hepatitis B, syphilis, etc.) or the air or respiratory tract (cough: flu, tuberculosis, pertussis, nasopharyngitis, etc.). A second category involves an indirect mode of transmission, either by ingestion (food, water: hepatitis A, cholera, rotavirus, poliomyelitis) or through a vector (mosquito, tick, flea: dengue fever, Zika, Chikungunya, Lyme). Whether direct or indirect, this transmission pattern implies the proximity of hosts moving around the same geographic space. Mobility therefore becomes a key element for understanding the spread of epidemics.
An epidemic corresponds to a rapid increase in the number of cases of a disease in a local or national area – a pandemic is its global expression. Identifying the emergence of the pathogen responsible for a disease as well as its modes of propagation should enable a better understanding, or even anticipation, of its geographical diffusion. The example of the spread of Covid-19 after the first cases were identified in China evidences the weight of international (and then national) mobility in this propagation. At a local scale, in everyday spaces, the role and weight of mobility is more complex to disentangle, because mobility is concomitant with many other factors which, combined, lead to local disease diffusion mappings: local disease surveillance and control actions, characteristics of the environments visited, behavior of the populations at risk, etc. This complexity partly explains why local daily mobility has long remained understudied in retrospective analyses. In the majority of cases, individual exposure factors are solely deduced from the characteristics of the living space of sick people, a bias partly related to the recording of the case at their place of residence. However, this obstacle tends to be partially surmounted thanks to digital tools and data. As it modifies a static vision of the exposure to pathogens, daily mobility renews research on the spread of communicable diseases.
The first section of this chapter (section 1.2) will summarize the evolution of knowledge on the contagiousness of diseases and their modes of transmission in an epidemic context, emphasizing the link between daily mobility and epidemiology. The modeling of daily mobility will be addressed in section 1.3. We will discuss different modeling methods and the concepts used in an epidemic context. Special emphasis will be placed on the most recent classes of models, namely individual-based models (IBMs). These models integrate geographic information as well as social and epidemiological knowledge into a single formalism, enabling exploration of a wide variety of scenarios. Finally, a last section (section 1.4) will prompt a selective literature review, dedicated to the data currently in use to better understand daily mobility and to calibrate increasingly detailed epidemiological models.
1.2. Analyzing the role of daily mobility in the spread of epidemics: concepts, studies and tools
Understanding the mechanisms responsible for the spread of an epidemic is a prerequisite for any action attempting to prevent it. Spatial epidemiology now plays a major role in this fight.
1.2.1. The path of daily mobility in the geographical reflection of infectious diseases
When it reached Europe in 1829, little medical controversy was stirred up in identifying Asiatic cholera morbus. The symptoms described by colonial doctors stationed in India matched the symptoms observed at Europe’s doorstep: “rice water” stools, profuse vomiting of liquid matter, spasms, horrible cramps, imperceptible pulse, bluish tints in the extremities and around the lips (hence the name “Blue Death”), body cooling, collapse, etc. and, above all, a mortality rate exceeding 50% in many cities. Though the poison had a name and a geographical origin, it was still necessary to determine its medical origin, its modes of propagation, the means to protect against it and the treatment to eradicate it. These questions fueled part of the medical research of the 19th century when two currents of thought collided, the insufficiency of medical knowledge leading to different interpretations of identical statistics and cartographic documents (Eliot et al. 2012).
Doctors then became divided between two schools: contagionists and anti-contagionists. In the 19th century, the sense of the term “contagion” implied that the disease was transmitted from person to person, via contact. This excluded diseases transmitted by air or water from the list of contagious diseases. One of the measures against cholera for the doctors who classified it under contagious diseases involved the isolation of any infected person in view of protecting the population: quarantine. On the other hand, anti-contagionists believed that disease was transmitted via the interactions between human beings and the environment – chiefly as a result of miasmatic exhalations. On the one hand, favorable conditions to the development and propagation of epidemics could be environment-related (climate, atmospheric density, degree of humidity, prevailing winds, the term “malaria” being derived from the Latin Mal’aria – which means “bad air” – whereas the term “palu” means a swamp) and, on the other hand, they could be owing to the type of habitat (cleanliness of houses and clothes, as well as inhabitants’ mores).
Overcoming this opposition between contagionists, anti-contagionists and supporters of the miasma theory, John Snow (1813–1858) shed new light on the understanding of the modes of propagation of cholera. As Europe experienced a new epidemic onslaught and the “Blue Death” spread across the cities and the countryside, Snow identified the waterborne nature of the “poison” that was wreaking havoc from Bengal to America, and ravaging Europe.
In his famous treatise On the Mode of Communication of Cholera published in 1855, supplemented with maps and statistics, the English doctor offered an original demonstration to support his hypotheses on the disease’s mode of propagation (Snow 1855). The singularity of Snow’s work was his inclusion on medical topography maps – which occasionally recorded deaths and people contaminated by cholera – of water points by neighborhood, notably that of Broad Street in London. He proved that most of the people who died of cholera lived near dirty water points.
Apart from drawing up these maps, Snow sought to understand why certain houses in the neighborhood – for instance, the workhouse on Poland Street or the brewery on Broad Street – were spared the disease. It turned out that the former had its own water point, whereas the workers of the latter simply did not drink water because they were entitled to rations of malt1 liquor. Another interesting fact: these surveys revealed that the deaths of certain persons living far from Broad Street were due to the fact that they preferred to drink water from this point, considering it less dirty than the source near their home2.
When he analyzed the mortality data by gender in London in 1848, he also realized that the epidemic outbreak affected men more than women, but in time, those numbers tended to equilibrium. The interpretation – paraphrased in modern terms – is related to gender mobility trends; men having a greater ability to disperse across the city because of their work, whereas women generally stayed at home. Men had a wider variety of sources to meet their water needs and therefore, a higher risk of contamination during an epidemic outbreak. At a second stage, when cholera spread throughout the city, men and women became equally at-risk3 populations. Snow also noted that valets and servants were the least affected by the disease because these workers used to live at the home of their “masters” in wealthy neighborhoods with better sanitary infrastructure. In this sense, they were affected by cholera less than their employers, whose daily routine was marked by numerous visits across the city.
The waterborne origin of cholera was thus established, as well as the first reflections on the role of mobility and the living spaces of populations, their exposure to pathogens and the spread of the epidemic. Based on these first results on cholera, Snow came to the conclusion that plague and yellow fever spread in the same way4. It only took a few more decades for research in epidemiology to show that both cases were vector-borne diseases. It is likely that the co-occurrence of mosquitoes with water points and the swarming of rats in slums could have misled him.
At the beginning of the 20th century, while the pathogens responsible for the main deadly diseases of the time and their mode of transmission were known, understanding the emergence and diffusion of new pathogens in epidemic form represented a challenge. This was the case of AIDS in the 1980s, of dengue in the 1990s and of SARS-CoV2 after 2019. Even if the main mechanisms contributing to the transmission and propagation of communicable diseases were known, their combinations in heterogeneous territories and within host populations with varied susceptibilities was quite complex. This required acknowledging the environmental and social criteria in which pathogens develop, the socio-demographic characteristics and behavior of hosts (in particular daily mobility and interactions) as factors of direct or indirect exposure to pathogens.
1.2.2. The role of mobility in the spread of epidemics
The spatial diffusion of an infectious and communicable disease implies that a pathogen multiplies and spreads from one host to another by means of a propagation channel (e.g. by airborne transmission or an animal vector). It is necessary to identify the actors involved in the transmission cycle (the pathogen’s potential hosts and transmission channels) as well as the processes interacting within the cycle, namely the pathogen’s propagation mechanisms, which lead to host contamination (Langlois and Daudé 2007).
The contamination process acts on the hosts: it represents the invasion and development of the pathogen within the host. Associated with infection, this process can alter the host’s health condition after the amount of pathogens present reaches a certain threshold. Different health conditions can be defined for the host, such as:
contamination, moment of the pathogen invasion;
infection, when the pathogen develops within the host;
viremia or contagiousness; the period during which the host emits a sufficient amount of pathogens to contaminate another susceptible person (who has not developed immune defenses capable of fighting the infection).
These health conditions may or may not be associated with symptoms, and the infection may end in partial or total recovery, with or without immunity to the pathogen, or result in death. This process can be globally represented by the number of hosts contaminated by the pathogen and the number required to qualify the disease as an epidemic. In fact, a certain density threshold of pathogens must be reached within an individual to trigger the disease, and reach a sufficient volume of infected individuals for an epidemic to break out.
The propagation process represents the pathogen’s “behavior” emitted from each infected host. The pathogen’s survival outside the host and before the invasion of a new host depends on environmental conditions. The influenza A virus, for example, survives between one and two days on hard surfaces and less than 12 hours on fabric. We can globally represent the propagation process by the number of active pathogens in the propagation channel (in this case, the environment). When this number is very low, particularly during the pre-epidemic phase, epidemic success is a stochastic process which depends on social and environmental conditions: high concentrations of people, absence of barrier measures, unfavorable environmental conditions. This is when highly targeted control measures can prove effective for preventing the spread of the disease. Pathogen viability emitted from the infected host also depends on the behavior of its possible vector. For instance, dengue viruses transmitted by mosquitoes can pose a risk to humans throughout the infected mosquito’s life cycle. When assessing epidemic risk, the number of pathogens can then be confused with the population of infected vectors.
Figure 1.1.Mobility and types of pathogen transmission
(source: Daudé and Eliot (2005))
By multiplying and differentiating the possible areas of contamination and propagation, the daily mobility of populations contributes to the pathogen’s spatial diffusion. This offers risk reduction strategies: ceasing mobility reduces the number, nature and places for interactions between hosts and can stop – if not slow down – the spread of the disease (the intention behind quarantines and lockdowns). Knowledge about host mobility can be useful for predicting the next contamination places or for tracing contamination sources a posteriori, in view of discovering the index case (the first contaminated person). Finally, epidemiological studies in places in which populations interact (markets, schools, places of work, etc.) or in places in which hosts and vectors co-exist (gardens, parks, public transport, etc.) can be useful for identifying potential sites for the emergence of pathogens or contamination and for reducing risk at the source (school closures, vector control in gardens).
Mobility acquires a different status depending on the mechanism underlying the transmission from pathogen to host or vector. On the one hand, mobility can be an intermediary in the transmission of the infectious agent from place to place. The agent is transmitted from one place to another from an infection source (outbreak). Here, mobility is a vector of contamination (Figure 1.1, (1) “vector-borne mobility”). This is the case, for example, when a contagious person struck down with influenza introduces the virus into the workplace, by sharing a meal in a communal room with several susceptible colleagues. On the other hand, mobility can lead to situations featuring exposure to pathogens. In this case, contamination is the result of mobility that exposes the host through the places they frequent (Figure 1.1, (2) “acquiring-mobility”). Here, it is the susceptible colleague who visits the communal room at the workplace who is exposed to the pathogen, whereas at home, there was potentially no level of risk. Movement is “at the origin” of the agent’s contamination. Lastly, mobility can constitute a contamination “channel”. It is during transportation that the host is contaminated by a pathogen and introduces it into the place of destination, in which case we speak of “capturing-mobility” (Figure 1.1, (3)). This is the moment when public transport becomes a key contamination place among traveling hosts.
These three basic forms of contamination are exclusive in the propagation of a pathogen at the individual-host level, but not at the group scale, and even less so at the scale of a whole population. This makes it all the more complex to identify the most vulnerable places exposed to contamination, the riskiest journeys, the duration and number of contacts most likely to spread a given disease during trips. Field studies are a first attempt to appraise the role of these different forms of contamination underlying pathogen diffusion.
1.2.3. Case studies: epidemic dynamics based on the location of proven cases
One major difficulty when trying to assess the role and weight of daily mobility in the spread of an epidemic lies in identifying the circumstances, place and time of contamination. In the majority of cases, public health data and epidemiological surveys only identify sick people at their place of residence. Trying to reproduce the spread of the disease from clusters of patients can be misleading – a cluster of residents in a neighborhood does not necessarily convey that there is a high risk of residential contamination. By the same token, an absence of clusters based on the place of residence does not necessarily mean that another type of cluster would not have emerged if the cases had been located otherwise (e.g. at the workplace or a recreational area).
1.2.3.1. Looking for contact cases
To overcome this limitation, further research may be conducted in order to collect ad hoc data and try to reconstruct the person’s itinerary a few days or weeks prior to the illness. The search for contact cases begins with a questionnaire administered to the hospitalized or sick person to accurately trace their movements after the appearance of the first symptoms, or a few days earlier, depending on viremia. Family members, acquaintances and colleagues are also questioned in order to identify the nature and duration of contacts with the sick person.
This method varies depending on the pathogen considered, as well as its contagiousness: a highly contagious virus will require extending the field of research to the greatest number of potential contact cases, with all the inaccuracies this entails. Asking a person how long they spent near another person and at what distance at a supermarket queue systematically induces a bias which over-represents risk. Applications for detecting proximity contacts using Bluetooth technology embedded in mobile phones were developed and widely distributed during the Covid-19 pandemic to partially address this issue. Nevertheless, there is currently little feedback from experience or studies clearly defining the contributions and limitations of those methods.
1.2.3.2. Spatio-temporal cluster analysis
In addition to the monitoring of individual trajectories, spatio-temporal cluster analysis assumes that detected cases are related both in time (duration of the contagion period and exposure duration) and in space (co-presence of people during contamination). But given the fact that determining the place of contamination is difficult, these studies are often conducted taking into account the places of residence of infected people. This boundary is even more significant for vector-borne diseases, as it dissociates the co-presence of healthy and infected hosts in the contamination/propagation process. An infected host living in district A can contaminate a vector during their morning visit to a park in district B. A few days later, this same vector can contaminate a host in district B (but be living in neighborhood C) during their nightly jog. Assessing the share of contamination that occurs while traveling outside the place of residence is a real methodological challenge. Phylogeography, which studies the distribution of the different genetic lineages of a virus, is an approach that partly makes it possible to overcome this obstacle.
Viruses are organisms that tend to mutate relatively quickly, and slight mutations accompany variations in infections and transmissions. The more differences there are between two viruses affecting two people, the smaller the chances for these individuals to belong to the same contamination chain. Here lies the origin of phylogeography, a science that studies the dynamics of viruses and analyzes the evolution of their genome, producing new variants. This approach is widely used at a global scale, particularly in the context of influenza, in order to estimate which strain will be dominant and which vaccine should be recommended.
This method was used by Salje et al. (2017) to determine the contamination chains of dengue fever in Bangkok. To do this, they first developed a phylogenetic tree including 640 genetic sequences of dengue viruses, drawn from blood samples of children having suffered from dengue fever between 1994 and 2010, and whose home address was known. Two individuals belong to the same transmission chain when the differences between the genomes of the same virus are relatively small. They found that 60% of the cases in the same chain had homes less than 200 m apart, which underlines the importance of residential neighborhoods in the spread of the epidemic. Despite this, we should observe that the subjects studied by Salje et al. (2012, 2017) were children, whose mobility is narrower than that of adults, often limited to the neighborhood and to schools, in most cases, located near their homes.
In another study and with another method, Vazquez-Prokopec et al. (2010) studied the spatial distribution of homes of confirmed dengue fever cases during the 2003 epidemic, which affected the city of Cairns in Australia. It was the first major epidemic, which means that most patients had no prior experience of the virus (naïve population), and were therefore more susceptible to infection. Out of 383 confirmed cases, the team evaluated the aggregation and/or dispersal levels in spatial point patterns based on distance-methods and Ripley’s K (1977). They showed that in 63% of cases, residences were located less than 800 m from the index case (the first confirmed case of dengue fever). They also applied a spatio-temporal partition method for identifying cases both close in time (less than 20 days) and in space (within 100 m of places of residence), which resulted in 18 independent clusters. Finally, the order of appearance of these clusters revealed that the disease had spread both in the direction of prevailing winds (which can facilitate the movement of mosquitoes) and along the axes of communication.
According to these two illustrative studies, while nearly 60% of the dengue cases studied presented a high probability of local contamination – thus confirming the usefulness of knowing the patients’ place of residence – nearly half of them could not be associated with nearby cases, nor had themselves acted as the source of local contamination. In the absence of data on mobility in retrospective studies, different modeling-based methods have been developed to ascertain the role of mobility in epidemic propagation.
1.3. Modeling epidemics and daily mobility
The spread of an epidemic can be modeled as an expansion process in time and space. This process is activated by the circulation of individuals and/or groups who move differently depending on their socio-demographic profile and who transmit a pathogen following different modalities.
Geographic research on epidemics has identified three propagation modes:
The first mode is based on an ordered sequence of places (
hierarchical diffusion
).
The second one is based on the proximity between places (diffusion by
spatial contiguity
).
The third one is based on the relocation possibilities of a phenomenon (
jump diffusion
).
However old, these works have remained largely confined to the disciplinary universe of geography and have not been mobilized by other disciplines studying the spread of epidemics.
1.3.1. Mobility, largely absent from the first mathematical models
A classical modeling approach in epidemiology assigns different clinical states to a host population and models the transition of individuals from compartment to compartment (Kermack and McKendrick 1927). The first clinical state concerns a fraction of the total population that is susceptible (S) to becoming contaminated. Part of this group may be contaminated and join the infected (I) compartment. After a certain period of time, infected people are cured and may develop (or not) immunity against the pathogen (R, standing for recovered). It is usually described in the form of stocks and their transitions: S ⇒ I ⇒ R model, where S + I + R equals the total population.
The SIR model is formalized as three differential equations, which describe the evolution of each of the S, I and R clinical states over time and according to an infection force β and a recovery rate γ:
During each time step, the number of newly infected people is described as the product of the capacity β (or force of infection) of an infected (I) person to transmit a disease to a susceptible (S) person, depending on the number of possible i interactions between the two subgroups, S and I. After a certain period, the infected people eventually recover (R) following recovery rate γ – which is inversely proportional to the infection’s duration – and acquire immunity. The choice, number and order of the compartments depend on the epidemic studied. For example, clinical state E (Exposed) could be added, characterizing the period after which the host has been infected, has developed a viral load but is not yet contagious. In that case, we speak of the SEIR (Susceptible–Exposed–Infected–Recovered) model. This class of model describes pathogen transmission among the hosts of the same population, while other models make it possible to introduce inter-species interactions.
Ronald Ross was the first to propose a mathematical transmission model for malaria based on a system with two differential equations taking human hosts and vectors into account (Ross 1911a, 1911b). It primarily revealed that below a threshold value of the mosquito population, the malaria epidemic should disappear. The classic deterministic models used in the context of vector-borne diseases thus take up the bases of the SIR or SEIR models and add new compartments describing the different states of the mosquito, their growth and especially their interactions with humans (Derouich et al. 2003). These differential equation systems are solved mathematically and show great sensitivity to the initial parameters. In the case of a highly contagious disease, a high force of infection will lead to rapid contamination of the whole population, whereas a long period of infection implies that more people are sick at the same time. We should also bear in mind that these approaches only model epidemics at a given place, without taking into account human mobility between territories. Nevertheless, the need to integrate statistics on mobility and migration into these models5 quickly emerges (Waite 1910). However the difficulty in obtaining mobility data and the limitations of computing power/capacity at the beginning of the 20th century did not make such integration possible. Thus, the work of Ross and McKendrick – the basis for modern mathematical epidemiology – lies on the postulate of a total intermixing of the population, where all individuals are likely to interact, excluding socio-spatial structures and individual mobility behaviors.
1.3.2. The first steps of human mobility in the modeling of epidemics
While conducting epidemiological research on AIDS, Peter Gould (1993) clearly showed the limitations of these models, which were misdirected, partial and based on a series of population dynamic models that acknowledged the temporal dimension but not the spatial one. Since the end of the 1960s, so-called metapopulation models have integrated mobility, primarily as population fluxes between the cities and across the country (Baroyan and Rvachev 1967). In the context of epidemics, these are more or less complex models (SIR, SEIR, etc.) that are able to take into account subgroups with different serological characteristics as well as the commuting of individuals between different geographical areas (Sattenspiel and Dietz 1995; Arino and van den Driessche 2003). In concrete terms, the SIR model is replicated according to the number of cities in the urban system. The population volumes for each city are then used for calibrating each of these sub-models (Figure 1.2).
Interaction rules are integrated into the equations to convey the arrival to/departure from a city towards another city. Inter-city population movements among the infected host compartments are explored in depth. These journeys, processed in the form of mobility flow (or travel probabilities), are generally represented by origin/destination matrices. These metapopulation models, still widely used today, provide powerful simulations at the macroscale because national (motorway flow) and international transport (air flow) are the main modes of propagation of a pathogen at this scale (Sattenspiel and Dietz 1995; Arino 2005). Like any model, they require good estimates of the flow matrices between the different zones, and depend on good quality mobility data (Tizzoni et al. 2014).
Simplified versions, with a smaller number of differential equations, have also been developed to reflect the arrival/departure of populations from a territory. The dengue epidemic that took place in Madeira in 2012 was simulated using air traffic towards the island as a proxy for the importation of the epidemic (Lourenço and Recker 2014). Another study adopted a twofold approach: first, an epidemic was simulated in the area of Karachi (Pakistan) – where dengue fever is endemic. This stage made it possible to estimate the number of infected persons. They then introduced infected people into other areas in the north of the country (where dengue fever is seasonal), matching the likelihood of interactions between different country sub-districts (Wesolowski et al. 2015).
Figure 1.2.Illustration of the principle of spatialized compartmental models, with transport infrastructure (gray links) and inter-city (ellipses) flow (black dotted lines) (inspired by Pastor-Satorras et al. (2015))
There are solid arguments for choosing differential equation modeling:
Firstly, it offers the possibility of understanding the global properties of a system through a small number of macroscopic variables (parsimony principle).
Secondly, all other things being equal, it helps estimate the maximum number of infected persons during an episode if current trends continue.
Thirdly, it helps to identify threshold effects in the dynamics of a phenomenon. Threshold effects are particularly interesting when several strategies for fighting an epidemic are in competition and when striving to find the path offering the most effect and the fewest constraints.
Yet, these models are hampered by significant barriers to understanding epidemics. First and foremost, the thorough and homogeneous mixing of the population, which results in a lack of consideration for social and spatial structures. This leads to an extremely simplified vision of the characteristics and behaviors of individuals in the population. Even if it is possible to multiply the compartments to represent the different population categories, and to replicate these compartments to signify different territories, calibrating all the parameters that govern these differential equation systems quickly becomes a challenge.
A second limitation of these models: they cannot track isolated individuals and tend to overestimate the propagation phase and epidemic peaks (Ajelli et al. 2010). While these models may be effective in the middle of an epidemic, they struggle to represent and simulate the emergence of epidemics, at the moment when “everything is at stake” by only relying on stochastic interactions between a few localized hosts. They also have difficulty in explaining the forms of diffusion, largely dependent on the spatial structures that existed prior to the epidemic. A necessary complement would require studying the ordinary dynamics of the city and the spaces involving population mobility, alongside inter-host contagion mechanisms, to better understand the geography and local spread of communicable diseases.
1.3.3. Deciphering activity spaces and daily mobilities
Most individuals travel on a daily basis to fulfill their obligations. They do so regularly (work, education, purchase of goods and services) or occasionally (health, administration). They also travel to fulfill their life plans (socialization, leisure), and all of these journeys are subject to a series of constraints (accessibility, alternative choices in more or less constrained schedules) and opportunities (concentration of activities made possible by urban densification). Depending on the nature of the places frequented, these mobilities can increase or decrease the likelihood of exposure to pathogens.
1.3.3.1. Complex host–pathogen system
When stepping foot in different territories, individuals can be exposed to risky environments (Figure 1.1). The pathogenic areas described are in line with the work on pathogen complexes by Sorre (1933). The endemicity of dengue fever in Indian cities is related to environmental conditions (monsoon period, favorable temperatures, availability of breeding sites), letting mosquitoes survive and spread the virus within the population residing within and nearby that pathogenic area. But the environment frequented by individuals is not the only contributing factor to the geography of diseases. Their origin also lies in social factors such as lifestyle or work, individual and collective behavior, resource levels, etc. Disease is thereby considered as the by-product of multiple interacting factors. Through his analysis of the host–pathogen system, Picheral (1983) conceptualized the intertwining of the links between factors provoking, conditioning and predisposing exposure to certain diseases. Health risks are explored as a complex articulation of factors within a complex host–pathogen system.
The concept of complex pathogenic system
