Social Science & Medicine 68 (2009) 2240–2246

Contents lists available at ScienceDirect

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

Drivers of tuberculosis epidemics: The role of risk factors and social determinants
Knut Lönnroth*, Ernesto Jaramillo, Brian G. Williams, Christopher Dye, Mario Raviglione
World Health Organization, Geneva, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: The main thrust of the World Health Organization’s global tuberculosis (TB) control strategy is to ensure
Available online 23 April 2009 effective and equitable delivery of quality assured diagnosis and treatment of TB. Options for including
preventive efforts have not yet been fully considered. This paper presents a narrative review of the
Keywords: historical and recent progress in TB control and the role of TB risk factors and social determinants. The
Tuberculosis review was conducted with a view to assess the prospects of effectively controlling TB under the current
Disease control
strategy, and the potential to increase epidemiological impact through additional preventive interven-
Social determinants
Risk factors
tions. The review suggests that, while the current strategy is effective in curing patients and saving lives,
Review the epidemiological impact has so far been less than predicted. In order to reach long-term epidemio-
Epidemiology logical targets for global TB control, additional interventions to reduce peoples’ vulnerability for TB may
WHO therefore be required. Risk factors that seem to be of importance at the population level include poor
living and working conditions associated with high risk of TB transmission, and factors that impair the
host’s defence against TB infection and disease, such as HIV infection, malnutrition, smoking, diabetes,
alcohol abuse, and indoor air pollution. Preventive interventions may target these factors directly or via
their underlying social determinants. The identification of risk groups also helps to target strategies for
early detection of people in need of TB treatment. More research is needed on the suitability, feasibility
and cost-effectiveness of these intervention options.
Ó 2009 Elsevier Ltd. All rights reserved.

Introduction frameworks alive (Amrith, 2002; Rosen, 1974). Several countries

The Commission on Social Determinants for Health (CSDH), set
up by the World Health Organization (WHO) in 2005, has
attempted to revitalize the debate and actions to improve health
through addressing the ‘‘causes of the causes’’ of ill health. The
commission has developed action frameworks in several fields to
address the social determinants of health including early childhood
development, globalization, urbanisation, employment conditions,
social exclusion, etc. A key message from the CSDH is that public
health achievements will largely depend on actions outside the
health care sector (CSDH, 2008).
The work of the commission builds on a long social medicine
tradition, from Virchow and the public health oriented social and
environmental interventions that he and others inspired from the
mid-19th century onwards (Rosen, 1974), through the social model
of health adopted by WHO in the late 1940s (Chisholm, 1948), to the
Health for All concept in the 1970s (WHO, 1978). It has not always
been easy to keep the social medicine concepts and action
* Corresponding author. World Health Organization, Stop TB Department, 20
Avenue Appia, CH-1211 Geneva, Switzerland. Tel.: þ41 22 7911628, fax: þ41 22
7914199.
E-mail address: lonnrothk@who.int (K. Lönnroth).
have tried consistently to pursue social policies aimed to promote
equity in health, some with a great deal of success (CSDH, 2008).
However, in most countries, little of the social model of health
advocated by WHO and others was translated into practice. As
a consequence, much preventable ill health and health inequity
prevails today (CSDH, 2008).
The last 2–3 decades have seen reinvigorated attempts to fight
diseases with medical technologies alone. Key concepts in today’s
discourse on disease control include: evidence-based medicine;
health systems strengthening; and community involvement. The
latter two focuses mainly on the capacity to deliver medical tech-
nologies effectively. Meanwhile, social medicine has often been
reduced to actions to identify risk groups that need to be targeted
with medical interventions and attempts to influence individuals’
risk behaviour (Porter, 2006).
The current global tuberculosis (TB) control paradigm mainly
focuses on cutting transmission through early case detection and
effective treatment. Medical interventions are at the core of the
global strategy (Lönnroth & Raviglione, 2008; WHO, 2006).
However, historically, TB has been used as a prime example of
a ‘‘social disease’’, the control of which requires social, economic
and environmental interventions. Furthermore, the need to control
TB has often been used as an argument for improving living

0277-9536/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.socscimed.2009.03.041
 K. Lönnroth et al. / Social Science & Medicine 68 (2009) 2240–2246
conditions and reduce inequity (Chisholm, 1948; Rosen, 1974). After
the second World War, a medically oriented TB control model
emerged (Amrith, 2002). Much hope was initially placed on mass
vaccination with BCG (Brimnes, 2007), hopes which were later
shattered when it was demonstrated that the protective effect and
epidemiological impact were limited (Rieder, 1999). When effective
chemotherapy for TB became available in the end of the 1940 and
beginning of 1950s, the control model switched to an almost
completely curative focus. ‘‘Prevention starts with cure’’ became
a slogan for the global role out of effective treatments. It was pre-
dicted that good coverage of effective treatment would result in
a rapid decline in TB incidence (Styblo & Bumgarner, 1991).
However, some recent observations have indicated that the impact
of the present TB control strategy has been less than expected
(WHO, 2008a). If these observations are confirmed, there may be
a need to again broaden the strategy to include more preventive
efforts (Farmer, 1997; Jaramillo, 1999). One avenue for improved TB
prevention is the ongoing quest for better TB vaccines and better
chemotherapy for preventive treatment. Another avenue leads
through actions to address the social determinants of TB, as well as
the more proximate risk factors (the physical and biomedical
factors that directly influence the mechanisms that govern expo-
sure to tuberculosis, risk of acquiring tuberculosis infection, and
risk of progression from tuberculosis infection to active tubercu-
losis disease).
In order to assess the relevance and potential benefits of the
latter, a review was conducted to identify key social determinants
of TB and the possible causal pathways through which they operate.
We also reviewed the historical and recent progress in global TB
control with a view to assess the prospects of effectively controlling
TB in the future with the current global strategy, and the potential
to increase epidemiological impact through adding actions on
social determinants and risk factors.
Methods
A literature review was undertaken to identify historic and
recent trends in TB burden; social determinants of TB; and the most
important proximate TB risk factors. We searched PubMed, the
send-list ‘‘TB-Related News and Journal Items Weekly Update’’
(Centre for Disease Control and Prevention, Atlanta, USA), and
a private data-base on tuberculosis publications from the past 40
years (created by Dr Hans Rieder). The analysis was carried out by
the Stop TB Department at WHO, in collaboration with other rele-
vant WHO Departments, including those responsible for tobacco
control, nutrition, alcohol and substance abuse, environmental
health, diabetes, and HIV/AIDS. The Priority Public Health Condition
Knowledge Network of the CSDH provided technical support and
feedback.
Findings
History of TB epidemiology and social change in industrialized
countries
Epidemiological TB data before the 20th century is imprecise.
However, some broad trends have been identified with a reason-
able level of certainty. The conventional wisdom is that the inci-
dence of TB increased in industrialized countries in the 17th to 18th
centuries, peaking at different times in different places from the
middle of the 1700s in Great Britain to the beginning of the 1900s in
Japan. From these trends, a temporal association has been sug-
gested between increased TB incidence and rapid industrialization
and urbanisation. TB rates were often particularly high in urban
areas. A plausible explanation is that transmission increased due to
2241
increased population density and crowded living conditions, while
poor nutritional status and other risk factors increased the risk of
break-down from infection to active disease in vulnerable groups,
such as among the urban poor (Aparicio, Capurro, & Castillo-
Chavez, 2002; Grundy, 2005; Rieder, 1999; Shimao, 2005). At the
peak of the epidemic in Europe TB death rates were close to 1% per
year in some urban areas, several times higher than the current
rates in countries with high HIV prevalence (Fig. 1).
Throughout the 20th century, even before the introduction of
effective chemotherapy, TB incidence declined steadily in most
industrialized countries although it did increase temporarily during
the two World Wars (Fig. 2). This was a period of economic growth,
social reform, poverty reduction and improved living conditions as
well as important advances in medicine and public health. The
relative importance of the factors that may have contributed to the
decline in TB in the 20th century is still a matter of debate.
McKeown and Record (1962) suggested that before the 1950s the
decline was due to improved nutritional status and to some extent
improved living conditions, while medical and public health
interventions had only a marginal impact before chemotherapy
became available. Others have convincingly argued that the intro-
duction of sanatoria and other mechanisms to isolate infectious
cases as well as pasteurization of milk also had a significant impact
on trends in the incidence of TB (Fairchild & Oppenheimer, 1998;
Grundy, 2005; Lienhardt, 2001; Wilson, 2005). Natural selection
may also have played a role (Davies, Tocque, Bellis, Rimmington, &
Davies, 1999).
The discovery of the TB bacillus in 1882 was an outstanding
advance in the understanding of transmissible diseases, and
marked the beginning of the germ theory era. As a result of this
discovery, a TB control paradigm based mostly on the biological
understanding of the disease gradually emerged. It received a final
boost with the discovery in the 1940s and 1950s of drugs that could
cure the disease (Amrith, 2002). The expanded pharmacopeia of
anti-TB drugs in the 1950s and 1960s helped to sustain and perhaps
accelerate the decline in TB incidence (Fig. 2). But this was not only
a period of rapid medical and health care advances. It was also
a time of both rapid economic growth and accelerated welfare
reforms in many industrialized countries (Navarro et al., 2006).
Progress in TB control in the industrialized countries over the past
centuries was thus brought about by advances on several fronts at
the same time – medical, public health, economic and social.
The evolution of the WHO’s TB control strategy
After decades of TB control neglect, with sustained high TB
burden in many low- and middle income countries and a resur-
gence of TB in several high-income countries in the 1980s, global
Fig. 1. Modelled tuberculosis deaths rates or selected European cities 1750–1950
(source: Rieder, 1999).
2242 K. Lönnroth et al. / Social Science & Medicine 68 (2009) 2240–2246
2000
First anti-TB
1500
Deaths/million/year
drug available
1000
500
World War 1 World War 2
0 vulnerable groups (Stop TB Partnership, 2006; WHO, 1994, 2002,
1900 1920 1940 1960
Fig. 2. Decline in TB mortality in England and Wales, and its association in time with
the two World Wars, and the introduction of chemotherapy against TB.
efforts to control TB were reinvigorated in 1991, when a World
Health Assembly resolution recognized TB as a major global public
health problem (WHO, 1991). The ‘‘DOTS’’ strategy was developed
by WHO and partners as a response in the 1990s (Raviglione & Pio,
2002). Realizing that essential medical technologies for appropriate
diagnosis and treatment were not used optimally in most parts of
the world, the response was to devise a strategy that takes into
account the basic health care elements required to effectively
deliver those technologies. After a decade of DOTS implementation,
the new ‘‘Stop TB Strategy’’ was launched in 2006 in response to
a number of challenges that had not been sufficiently tackled
through DOTS, notably the challenges of multi-drug resistant
(MDR) TB and the intersecting epidemics of TB and HIV, as well as
the challenges of weak health systems, insufficient engagement of
private providers and communities, and reliance on imperfect
medical technologies (Raviglione & Uplekar, 2006; WHO, 2006).
The medium-term goal of global TB control is to halve TB
prevalence and death rates by 2015 as compared to 1990 levels, and
to achieve a decline in incidence, as established in Millennium
Development Goal (MDG) 6 (Stop TB Partnership, 2006; WHO,
2006). A longer-term target is to ‘eliminate’ TB by reducing the
incidence to less than 1 new case per million population per year by
2050. Mathematical models suggest that the medium-term goal
could be met by detecting 70% or more of all incident cases of
infectious TB and curing at least 85% of them (Borgdorff, Floyd, &
Broekmans, 2002; Dye, Garnett, Sleeman, & Williams, 1998; Styblo
& Bumgarner, 1991; Suarez et al., 2001) and these are the targets
against which the performance of National TB Programmes (NTP)
are conventionally measured (WHO, 2008a).
The first priority for NTPs is thus to cure patients with active
disease so as to cut transmission, reduce suffering, avert mortality,
and avoid creation of drug resistance. It has been demonstrated
that TB treatment according to the DOTS principles is a highly
cost-effective public health intervention with regards to disability-
adjusted life years saved (Dye & Floyd, 2006). Furthermore, cost–
benefit calculations show that curative TB control is a good societal
investment (Laxminarayan et al., 2007).
The focus of the strategy is thus on diagnosis and curative
treatment rather than on TB prevention sensu lato. The few
preventive aspects of the strategy – isoniazid preventive therapy
for selected risk groups and BCG vaccination – are biomedical
interventions. The mathematical model underpinning the current
approach does not make explicit assumptions about other risk
factors that may drive the TB epidemic, apart from HIV which has
been included in some of the modelling exercises (Dye et al., 1998;
Stop TB Partnership, 2006). Implicitly, the current strategy does not
rely on reducing vulnerability to TB infection and disease. While it
does not suggest that preventive approaches are un-important it
does suggest a way forward that is independent of them.
Since the vast majority of people with TB are from the poorer
segments of society, the global TB control targets cannot be met
unless these poor segments are reached with quality health
services (WHO, 2005). Early diagnosis and treatment are also
important to reduce adverse social and financial consequences of
the disease for TB patients and their families (Dye & Floyd, 2006;
Hanson, Floyd, & Weil, 2006). Consequently, no influential policy
document on TB control fails to highlight that it is essential to
device strategies that ensure good access to quality TB diagnosis
and treatment for the entire population, especially the most
1980 2005, 2006).
In summary, the Stop TB Strategy aims to control TB mainly by
cutting the transmission chain, and this should be achieved by early
and effective treatment of all people with TB. The strategy
acknowledges that various social factors put certain vulnerable
groups at high risk of developing TB and makes recommendations
on ways to provide these groups with effective TB treatment.
However, it does not explicitly address the underlying factors that
make these groups more vulnerable to TB in the first place.
Epidemiological impact of DOTS
Successful DOTS implementation was associated with a decline
in TB incidence and prevalence in Peru, Cuba, China, parts of India,
and the USA (China TB Control Collaboration, 2004; Frieden, Fuji-
wara, Washko, & Hamburg, 1995; Gonzalez, Armas, & Llanes, 2007;
Suarez et al., 2001; Subramani et al., 2006). In some other countries
in Latin America, including Chile and Uruguay, TB declined rapidly
from the 1960s to the 1980s, which coincided with implementation
of TB programmes that had the basic DOTS elements in place
(Rodriguez De Marco, Sanches, & Alvarez Goya, 2007; Zuniga &
Rojas, 2002). However, where there has been a demonstrable
decline in incidence it has been difficult to separate out the effects
of DOTS from those of social and economic development.
Recent country-level investigations of the impact of DOTS pro-
grammes have shown that, after several years of apparent
successful implementation (as measured by high case detection
and treatment success), incidence is not falling as rapidly as was
expected. Vietnam appears to have reached the targets for case
detection and treatment success since 1997, but the case notifica-
tion rate has remained approximately constant over the last decade
(Huong et al., 2006). The explanation for this is unclear. However,
a recent study showed that a decrease in notification rates among
older people has been offset by an increase among young people,
especially young men in urban areas and remote rural areas (Vree
et al., 2007). It is possible that internal migration and/or higher
exposure to various risk factors for TB infection and disease in these
young men has balanced out the impact of the control programme.
In Morocco also, the reduction in notification rates has been less
than expected considering the successful implementation of the
DOTS programme for many years. Here too the problem seems to
be related to increasing rates of TB among young men in urban
areas (Dye, Ottmani, Laasri, & Benchelkh, 2007). Similar patterns
are also seen in the routine surveillance data from Sri Lanka and
Myanmar (WHO, 2007). Several States of India have been
successfully implementing DOTS since 1998 but there has been no
detectable decline in case notification rates (Watt, Hosseini, Lönn-
roth, Williams, & Dye, 2008). These observations raise the possi-
bility that the impact of improved diagnostic and treatment
services is offset by maintained or increased vulnerability to TB
infection and disease among certain subgroups of the population.
 K. Lönnroth et al. / Social Science & Medicine 68 (2009) 2240–2246
The slower-than-expected rate of decline might also be because
many people with TB are diagnosed so late that they have already
infected many people before effective treatment starts (Porter &
Grange, 1999; WHO/EMRO, 2006; WHO/WPRO, 2004).
The epidemiological challenge ahead: the role of prevention
The uncertainty about the impact of DOTS on the burden of TB is
reinforced by a recent analysis of predictors of TB case notification
trends between 1997 and 2006 (Dye, Lönnroth, Jaramillo, Williams,
& Raviglione, in press). This analysis suggests that the variation in
TB trends is more strongly associated with biological, social and
economic factors than with NTP performance. Moreover, recent
calculations suggest that, although prevalence and death rates are
falling globally, the rate of decline is not fast enough to meet the
MDG related targets to halve TB prevalence and death rates by 2015
compared to 1990 levels (WHO, 2009). The biggest challenges in
cutting TB burden are in sub-Saharan Africa and Eastern Europe.
The countries of sub-Saharan Africa and in Eastern Europe and the
former Soviet Union showed striking increases in TB burden during
the 1990s, for somewhat different reasons: in Eastern Europe and
former Soviet Union countries, the resurgence of TB (after decades
of steady decline before the 1990s) can be explained by economic
decline and the failure of TB control and other health services since
1991, along with other factors such as social deprivation, poor living
conditions, alcoholism, and the mixing of prison and civilian pop-
ulations (Leon et al., 1997; Pesut, Gledovic, Grgurevic, Nagorni-
Obradovic, & Adzic, 2008). In Africa the increase was largely due to
the rapid spread of HIV in the 1980s and 1990s (Corbett et al.,
2003). Reducing the burden of TB in these two regions will depend
to a great extent on dealing with the factors that drove the burden
of TB up so rapidly in the 1990s (WHO, 2008a).
In 2006–2007, the decline in global TB incidence was estimated
at less than 1% per year (WHO, 2009). To eliminate TB by 2050 the
incidence rate must fall by an average of 16% annually over the next
40 years (Fig. 3). Recent modelling suggests that even if the Global
Plan to Stop TB is successfully implemented and results in the
expected rate of reduction in incidence of about 6%, the global
incidence rate in 2050 would still be of the order of 100 per million
population, i.e. about 100 times greater than the elimination target
(Watt et al., 2008). An incidence decline of >15% per year might be
achieved temporarily through massive efforts to scale up of
10000
Incidence/1 million/yr
1000
100
10
Elimination -16%/yr
Global Plan -6%/yr
Current trajectory -1%/yr
1 Rieder, 1999), diabetes (Stevenson et al., 2007), malignancies,
2000 2010 2020 2030 2040
Year
Fig. 3. Scenarios for the trend of global TB incidence 2007–2050, based on: (a)
extrapolation of current rate of decline (1% per year); (b) predicted decline with full
implementation of the Global Plan to Stop TB 2006–2015 (6% per year); and (c)
desired trend for reaching the TB elimination target by 2050 (16% per year).
2243
curative services, which would reduce transmission of TB.
However, it is unlikely to be sustainable without additional efforts
to prevent progression from infection to disease. Currently, more
than 2 billion people are infected with Mycobacterium tuberculosis.
As transmission falls, a growing proportion of cases arise from the
reactivation of latent infections, rather than from progression of
recent infections. A recent analysis of the reasons for slow decline
in TB incidence in Hong Kong, despite good TB diagnosis and
treatment services for several decades, suggests that high preva-
lence of TB infection among elderly combined with very high risk of
reactivating contributes to a sustained high incidence (Vynnycky,
Borgdorff, Leung, Tam, & Fine, 2008).
In summary, to achieve the long-term targets for TB control,
efforts to cut transmission need to be complemented with inter-
ventions that reduce progression to active TB disease. Such inter-
ventions may include better preventive chemotherapy, a new
vaccine that prevents progression from infection to disease, as well
as interventions that reduce exposure to various social, environ-
mental and biological risk factors for TB.
Proximate risk factors and upstream determinants
The socio-economic gradient
The TB burden follows a strong socio-economic gradient
between countries, within countries, and within communities, and
the poorest have the highest risk (D’Arcy Hart, 1934; Hanson et al.,
2006; Harling, Ehrlich, & Myer, 2008; Hinman, Judd, Kolnik, &
Daitch, 1976; Holtgrave & Crosby, 2004; Lopez De Fede, Stewart,
Harris, & Mayfield-Smith, 2008; Muniyandi et al., 2007; Van Rie
et al., 1999). Studies assessing the burden of TB in specific vulner-
able populations such as prisoners, the homeless, and certain
ethnic minorities also show that there is a strong association
between social deprivation and TB risk (Bobrik, Danishevski, Ero-
shina, & McKee, 2005; Buskin, Gale, Weiss, & Nolan, 1994; Fried-
man, Williams, Singh, & Frieden, 1996; Keppel, 2007; Mori,
Leonardson, & Welty, 1992; Rieder, 1999). However, the causal
pathways linking poverty and low socio-economic status to
increased risk of TB are not fully understood.
Proximate risk factors
Proximate risk factors include those that directly increase
exposure to infectious droplets. A necessary risk factor for TB
infection is contact with a person with active disease. The likeli-
hood of having such a contact is determined by the underlying
disease burden in the community. People living or working in
environments where TB prevalence is particularly high are obvi-
ously at high risk of infection, for example prison staff and inmates
(Bobrik et al., 2005) and certain health care workers (Menzies,
Joshi, & Pai, 2007). The risk of exposure is also determined by the
physical environment in which the contact takes place, including
aspects of crowding, air flow and humidity (Rieder, 1999).
Proximate risk factors also include those that impair the host
defence against infection and break-down to disease, such as HIV
infection (Corbett et al., 2003), malnutrition (Cegielski & McMurray,
2004), tobacco smoke (Lin, Ezzat, & Murray, 2007, Slama et al.,
2007), indoor air pollution caused by burning of solid fuels (Lin
et al., 2007; Rhefuess, 2006), alcohol abuse (Lönnroth, Williams,
Stadlin, Jaramillo, & Dye, 2008), silicosis (Corbett et al., 2000;
2050 a wide range of chronic systemic illnesses, and immunosuppressive
treatment (Rieder, 1999). There is some evidence that outdoor air
pollution is a risk factor for TB (Cohen & Mehta, 2007; Tremblay,
2007). Depression and stress can have negative effect on the cell-
mediated immune system and could therefore in theory increase
the risk of TB (Prince et al., 2007).
2244 K. Lönnroth et al. / Social Science & Medicine 68 (2009) 2240–2246
In an analysis applied to the 22 High TB Burden Countries (HBC,
countries that together suffer 80% of the estimated global TB burden),
the population attributable fraction (PAF) for selected TB risk factors
for impaired host defence was roughly estimated (Lönnroth &
Raviglione, 2008). This analysis, which was recently updated based
on new estimates, suggests that malnutrition (PAF, 27%), smoking
(23%), HIV (19%), diabetes (6%), harmful alcohol use (13%) and indoor
air population (26%) are all important risk factors globally (though
the evidence base for indoor air pollution is still weak). Their relative
importance depends on the prevalence of exposure and is therefore
different for different countries and regions. For example, a region-
alized PAF analysis suggests, as expected, that HIV is a much more
important risk factor in the Africa (PAF, >50%) than elsewhere due to
the high prevalence of HIV.
Links between socio-economic status and proximate risk factors
It is reasonable to assume that the higher risk of TB among
people in low socio-economic status (SES) groups is largely an effect
of their greater exposure to some or all of the risk factors discussed
above. A recent multi-level analysis of TB risk factors from South
Africa found an association between the risk of TB and smoking,
alcohol and under-nutrition. These factors partly explained the
underlying association between low SES and TB risk, suggesting
that these proximate risk factors are on the causal pathway between
poverty and TB risk (Harling et al., 2008). People from low SES
groups have on average: more frequent contact with people with
active TB disease; higher likelihood of crowded and poorly venti-
lated living and working conditions; limited access to safe cooking
facilities; more food insecurity; lower levels of awareness and/or
less power to act on existing knowledge concerning healthy
behaviour (e.g. safe sex, smoking, diet and alcohol use); and limited
access to high quality health care (FAO, 2006; Gilson, Doherty,
Loewenson, & Francis, 2007; Kjellström et al., 2007).
Malnutrition and indoor air pollution are direct markers of
poverty (FAO, 2006; Rhefuess, 2006). Smoking prevalence is
consistently higher among the lower SES groups than among
higher SES groups in all regions of the world and smoking preva-
lence is increasing rapidly in low income countries, while it is
decreasing in high-income countries (WHO, 2008b). For HIV,
alcohol and diabetes, the picture is more complex. On average, HIV
prevalence is higher in poor countries than in rich countries and the
prevalence is higher in countries with more unequal wealth
distribution. However, there is wide variation in national HIV
prevalence across countries with similar level of national wealth
(Piot, Greener, & Russell, 2007; Reidpath, 2008). Furthermore, on an
individual level the association between SES and HIV prevalence is
less clear (Humphrey, Nathoo, & Hargrove, 2007; Mishra et al.,
2007, Wojcicki, 2005). Prevalence of alcohol dependence is higher
in rich than in poor countries. However, within countries, at least
for men, harmful drinking patterns and alcohol related morbidity
and mortality follow a reversed socio-economic gradient: those
with low SES are at highest risk (Schmidt, Mäkelä, Rehm, & Room,
2008). The association between SES and diabetes is also complex.
Economic development, reduced poverty and improved food
security can lead to increased diabetes prevalence. The prevalence
is generally higher among the better off than among the worse off
in poor countries. However, in middle- and high-income countries
the reverse is true. Here, people from low SES groups have higher
prevalence, and this is probably linked to a less healthy diet and less
physical activity (Irwin & Whiting, 2008; Smith, 2007).
The role of urbanisation
Rapid urbanisation can create ideal conditions for TB epidemics to
flourish, unless accompanied by good urban planning, social reforms,
environmental protection, and a strong and well-coordinated urban
health system (Kjellström et al., 2007). TB incidence is generally
higher in urban than in rural areas (Lönnroth, Zignol, & Uplekar, 2006),
and, as discussed above, there are signs that urban TB control is
particularly challenging (Dye et al., 2007; Watt et al., 2008; WHO,
2007). The tendencies for the burden of TB to be higher in urban than
in rural areas may be due to high population density, crowded living
and working conditions, as well as lifestyle changes associated with
urban living. Exposure to certain TB risk factors such as smoking,
alcohol abuse, unsafe sex, and unhealthy diet, may increase when
absolute poverty falls at the same time as rapid socio-cultural tran-
sitions lead to changed health behaviour patterns (Kjellström et al.,
2007). Such changes are partly the result of globalization (Raviglione,
2007). For example, globalization of nutrition makes food with high
calorific value but a low nutritional value available at relatively cheap
prices (Labonté, Blouin, & Chopra, 2007). Changed nutritional patterns
combined with less physical exercise due to poor urban planning and
lack of health promoting policies can explain why diabetes can coexist
with malnutrition in poor urban settings (Kjellström et al., 2007). HIV
prevalence is on average 1.7 times higher in urban than rural areas
(UNAIDS & WHO, 2006), and this is linked to various social determi-
nants of unsafe sex. The potential increase in risk increase due to
clustering of these TB risk factors is often further augmented by
fragmented health care systems in urban areas, and poor health care
access among urban slum dwellers (Lönnroth et al., 2006).
Discussion
This review suggests that, in order to reach long-term TB control
targets, the current TB control strategy needs to be complemented
with efforts to address TB risk factors and social determinants. Fig. 4
presents a framework for identifying entry points for such inter-
ventions. In principle, reducing the prevalence of the proximate risk
factors will reduce TB incidence. This may be achieved by tackling
these risk factors directly, or the social determinants – on the
individual, community, national and international level – that lay
behind them.
We know from the history of TB that general socio-economic
development can contribute substantially to TB control by
improving living and conditions, nutrition status, and health care
services. However, economic development may also be associated
with increased risk of TB, for example due to urbanisation and life-
style changes. The dynamics of the relationship between socio-
economic development and TB epidemiology, and the causal path-
ways that link SES with risk of TB are not fully understood. Fig. 4 is
therefore indicative only, and may be used also as a framework for
further research. The following areas of research require attention:
Fig. 4. Framework for proximate risk factors and upstream determinants of TB.
 K. Lönnroth et al. / Social Science & Medicine 68 (2009) 2240–2246
 Basic epidemiological research to further establish causality
and the strength of association for proximate TB risk factors.
 Refined country-specific analyses of population attributable
fractions of different risk factors, accounting for interaction and
heterogeneity across countries.
 Multi-level analyses of causal pathways linking low SES with
higher risk of TB.
 Analysis of factors determining variation in TB burden and
historic change in TB burden across countries and across
geographical areas within countries.
 Modelling of the epidemiological impact of different scenarios
for socio-economic change and change in risk factor exposure
in populations.
While waiting for such analyses, available data points to the
following actions:
First, the sine qua non of TB care and control remains high quality
diagnostic and treatment. Additional efforts can be made to ensure
that the most vulnerable groups of the population are reached with
available medical technologies in order to stop transmission where
it is most deleterious and in order to make available preventive
chemotherapy to those at highest risk of progressing from latent
infection to active TB disease. This will require mapping of high-risk
groups and access barriers on national and sub-national levels, and
tailoring of interventions strategies accordingly.
Second, NTPs could strengthen collaboration with other public
health programmes in order to contribute to the prevention,
treatment and management of HIV, malnutrition, smoking-related
conditions, diabetes, and alcohol abuse. Frameworks for such work
are already well established for HIV (WHO, 2006), and on the way
for smoking (IUATLD & WHO, 2007). Good collaboration between
public health programmes ultimately requires a well-functioning
primary health care system. Contribution to health systems
strengthening, already on the agenda for TB programmes (WHO,
2006), can help not only to improve diagnosis and treatment of TB,
but also help to address these TB risk factors.
Third, and perhaps most challenging, interventions outside the
health sector will have to be strengthened. Priority public condi-
tions share many ‘‘upstream’’ social determinants. The further
upstream the entry point for intervention, the more widespread
the effect. The CSDH has developed frameworks for action to
address a wide range of upstream social determinants (Blas &
Sivasankara, 2009; CSDH, 2008). NTPs and the international tech-
nical partners supporting them need not re-invent the wheel but
should support the full implementation of these frameworks.
Political commitment – the first element of the Stop TB Strategy –
should not only concern commitment from Governments to invest
in and support TB diagnosis and treatment programmes, but also
commitment by all political actors, including civil society, to
address the upstream drivers of the TB epidemic.
Competing interests
The authors are staff members of the World Health Organiza-
tion. The authors alone are responsible for the views expressed in
this publication and they do not necessarily represent the decisions
or policies of the World Health Organization. No external funding
was provided for this research.
References
Amrith, S. (2002). Plague of poverty? The World Health Organization, Tuberculosis and
International Development 1945–1980. Cambridge: King’s College, University of
Cambridge.
Aparicio, J. P., Capurro, A. F., & Castillo-Chavez, C. (2002). Markers of disease evolution:
the case of tuberculosis. Journal of Theoretical Biology, 215, 227–237.
2245
Blas, E., & Sivasankara, A. K. (Eds.). (2009). Priority public health conditions: From
learning to action on social determinants of health. Geneva: World Health
Organization.
Bobrik, A., Danishes, K., Eroshina, K., & McKee, M. (2005). Prison health in Russia:
the larger picture. Journal of Public Health Policy, 26, 30–59.
Borgdorff, M., Floyd, K., & Broekmans, J. F. (2002). Interventions to reduce tuber-
culosis mortality and transmission in low- and middle-income countries.
Bulletin of WHO, 80, 217–227.
Brimnes, N. (2007). Vikings against tuberculosis: The International Tuberculosis
Campaign in India, 1948–1951. Bulletin of the History of Medicine, 81, 407–430.
Buskin, S. E., Gale, J. L., Weiss, N. S., & Nolan, C. M. (1994). Tuberculosis risk factors in
adults in King County, Washington, 1998 through 1990. American Journal of
Public Health, 84, 1750–1756.
Cegielski, P., & McMurray, D. N. (2004). The relationship between malnutrition and
tuberculosis: evidence from studies in humans and experimental animals. The
International Journal of Tuberculosis and Lung Disease, 8, 286–298.
China Tuberculosis Control Collaboration. (2004). The effect of tuberculosis control
in China. Lancet, 364, 417–422.
Chisholm, B. (1948). Social medicine. Scientific American, 180, 11–15.
Cohen, A., & Mehta, S. (2007). Pollution and tuberculosis: outdoor sources. PLoS
Medicine, 4, e141–e143.
Commission on Social Determinants of Health. (2007). Achieving health equity: From
root causes to fair outcomes – Commission on Social Determinants of Health
interim statement. Geneva: World Health Organization.
Corbett, E. L., Churchyard, G. J., Clayton, T. C., Williams, B. G., Mulder, D., Hayes, R. J., et
al. (2000). HIV infection and silicosis: the impact of two potent risk factors on the
incidence of mycobacterial disease in South African miners. AIDS, 14, 2759–2768.
Corbett, E. L., Watt, C. J., Walker, N., Maher, D., Williams, B. G., Raviglione, M. C., et al.
(2003). The growing burden of tuberculosis: global trends and interactions with
the HIV epidemic. Archives of Internal Medicine, 163, 1009–1021.
CSDH. (2008). Closing the gap in a generation: Health equity through action on the
social determinants of health. Final report of the Commission on Social Determi-
nants of Health. Geneva: World Health Organization.
D’Arcy Hart. (1932). The value of tuberculin test in man, with special reference to
the intracutaneous test. Medical Research Council Special Series, 164, 5–132.
Davies, R., Tocque, K., Bellis, M., Rimmington, T., & Davies, P. (1999). Historical
declines in tuberculosis in England and Wales: improving social conditions or
natural selection? The International Journal of Tuberculosis and Lung Disease, 3,
1051–1054.
Dye, C., & Floyd, K. (2006). Tuberculosis. In T. Jamison, J. G. Breman, & A. R. Measham
(Eds.), Disease control priorities in developing countries (pp. 289–309). WA, DC:
The World Bank.
Dye, C., Garnett, G. P., Sleeman, K., & Williams, B. G. (1998). Prospects for worldwide
tuberculosis control under the WHO DOTS strategy. Directly observed short-
course therapy. Lancet, 352, 1886–1891.
Dye, C., Lönnroth, K., Jaramillo, E., Williams, B. G., & Raviglione, M. Determinants of
trends in tuberculosis incidence: an ecologic analysis for 134 countries. Bulletin
of WHO, in press.
Dye, C., Ottmani, S., Laasri, L., & Benchelkh, N. (2007). The decline of tuberculosis
epidemics under chemotherapy: a case study in Morocco. The International
Journal of Tuberculosis and Lung Disease, 11, 1225–1231.
Fairchild, A., & Oppenheimer, M. (1998). Public health nihilism vs. pragmatism:
history, politics, and the control of tuberculosis. American Journal of Public
Health, 88, 1105–1117.
FAO. (2006). The state of food insecurity in the world 2006. Rome: Food and Agri-
culture Organization of the United Nations.
Farmer, P. (1997). Social scientists and the new tuberculosis. Social Science &
Medicine, 44, 347–358.
Frieden, T. R., Fujiwara, P. I., Washko, R. M., & Hamburg, M. A. (1995). Tuberculosis in
New York Citydturning the tide. The New England Journal of Medicine, 333,
229–233.
Friedman, L. N., Williams, M. T., Singh, T. P., & Frieden, T. R. (1996). Tuberculosis,
AIDS, and death among substance abusers on welfare in New York City. The New
England Journal of Medicine, 334, 828–833.
Gilson, L., Doherty, J., Loewenson, R., & Francis, V. (2007). Knowledge network on
health systems – WHO Commission on Social Determinants of Health. Final report.
London: Centre for Health Policy.
Gonzalez, E., Armas, L., & Llanes, M. J. (2007). Progress towards tuberculosis elim-
ination in Cuba. The International Journal of Tuberculosis and Lung Disease, 11,
405–411.
Grundy, E. (2005). The McKeown debate: time for burial. International Journal of
Epidemiology, 34, 529–533.
Hanson, C., Floyd, K., & Weil, D. (2006). Tuberculosis in the poverty alleviation
agenda. In M. Raviglione (Ed.), TB: A comprehensive international approach (pp.
1097–1114). New York: Informa Healthcare.
Harling, G., Ehrlich, R., & Myer, L. (2008). The social epidemiology of tuberculosis in
South Africa: a multilevel analysis. Social Science & Medicine, 66, 492–505.
Hinman, A. R., Judd, J. M., Kolnik, J. P., & Daitch, P. B. (1976). Changing risks in
tuberculosis. American Journal of Epidemiology, 103, 486–497.
Holtgrave, D. R., & Crosby, R. A. (2004). Social determinants of tuberculosis case
rates in the United States. American Journal of Preventive Medicine, 26, 159–162.
Humphrey, J. H., Nathoo, K. J., & Hargrove, J. W. (2007). HIV-1 and HIV-2 prevalence
and associated risk factors among postnatal women in Harare, Zimbabwe.
Epidemiology and Infection, 135, 933–942.
2246 K. Lönnroth et al. / Social Science & Medicine 68 (2009) 2240–2246
Huong, N. T., Duong, B. D., Co, N. V., Quy, H. T., Tung, L. B., Broekmans, J. F., et al.
(2006). Tuberculosis epidemiology in six provinces of Vietnam after the
introduction of the DOTS strategy. The International Journal of Tuberculosis and
Lung Disease, 10, 963–969.
Irwin, N., & Whiting, D. (2008). The social determinants of diabetes and its impact
on human well-being. In E. Blas, & A. K. Sivasankara (Eds.), Priority public health
conditions: From learning to action on social determinants of health. Geneva:
World Health Organization.
IUATLD, & WHO. (2007). A WHO/The Union monograph on TB and tobacco control –
Joining efforts to control two related global epidemics. WHO/HTM/TB/2007.390.
Geneva: World Health Organization.
Jaramillo, E. (1999). Encompassing prevention with treatment: the path for a long
lasting control of tuberculosis. Social Science & Medicine, 49, 393–404.
Keppel, K. G. (2007). Ten largest racial and ethnic health disparities in the United
States based on healthy people 2010 objectives. American Journal of Epidemi-
ology, 166, 97–103.
Kjellström, T., Mercado, S., Sattherhwaite, D., McGranaham, G., Friel, S., &
Havemann, K. (2007). Our cities, our health, our future: Acting on social deter-
minants for health equity in urban settings. Report to the WHO Commission on
Social Determinants of Health from the Knowledge Network on urban settings.
Kobe: World Health Organization Kobe Centre.
Labonté, R., Blouin, C., & Chopra, M. (2007). Towards health-equitable globalization:
Rights, regulation and redistribution. Final report to the Commission on Social
Determinants of Health from the Globalization Knowledge Network. Ottawa:
Institute of Population Health.
Laxminarayan, R., Klein, E., Dye, C., Floyd, K., Darley, S., & Adeyi, O. (2007). Economic
benefit of tuberculosis control. Policy Research Working Paper 4295. WA: The
World Bank.
Leon, D. A., Chenet, L., Shkolnikov, V. M., Zakharov, S., Shapiro, J., Rakhmanova, G.,
et al. (1997). Huge variation in Russian mortality rates 1984–94: artefact,
alcohol, or what? Lancet, 350, 383–388.
Lienhardt, C. (2001). From exposure to disease: the role of environmental factors in
susceptibility to and development of tuberculosis. Epidemiologic Review, 23,
288–301.
Lin, H., Ezzati, M., & Murray, M. (2007). Tobacco smoke, indoor air pollution and
tuberculosis: a systematic review and meta-analysis. PLoS Medicine, 4, e142.
Lönnroth, K., & Raviglione, M. (2008). Global tuberculosis epidemiology – prospects
for control. Seminars in Respiratory & Critical Care Medicine, 29, 481–491.
Lönnroth, K., Williams, B. G., Stadlin, S., Jaramillo, E., & Dye, C. (2008). Alcohol use as
risk factor for tuberculosis disease – a systematic review. BMC Public Health, 8, 289.
Lönnroth, K., Zignol, M., & Uplekar, M. (2006). Controlling TB in large metropolitan
settings. In M. Raviglione (Ed.), TB: A comprehensive international approach (pp.
1005–1027). New York: Informa Healthcare.
Lopez De Fede, A., Stewart, J. E., Harris, M. J., & Mayfield-Smith, K. (2008). Tuber-
culosis in socio-economically deprived neighborhoods: missed opportunities
for prevention. The International Journal of Tuberculosis and Lung Disease, 12,
1425–1430.
McKeown, T., & Record, R. G. (1962). Reasons for the decline of mortality in England
and Wales during the nineteenth century. Population Studies, 16, 94–122.
Menzies, D., Joshi, R., & Pai, M. (2007). Risk of tuberculosis infection and disease
associated with work in health care settings. The International Journal of
Tuberculosis and Lung Disease, 11, 593–605.
Mishra, V., Assche, S. B., Greener, R., Vaessen, M., Hong, R., Ghys, P. D., et al. (2007).
HIV infection does not disproportionately affect the poorer in sub-Saharan
Africa. AIDS(Suppl. 7), S17–S28.
Mori, M. A., Leonardson, G., & Welty, T. K. (1992). The benefits of isoniazid
chemoprophylaxis and risk factors for tuberculosis among Oglala Sioux Indians.
Archives of Internal Medicine, 152, 547–550.
Muniyandi, M., Ramachandran, R., Gopi, P. G., Chandrasekaran, V., Subramani, R.,
Sadacharam, K., et al. (2007). The prevalence of tuberculosis in different
economic strata: a community survey from South India. The International
Journal of Tuberculosis and Lung Disease, 11, 1042–1045.
Navarro, V., Muntaner, C., Borrell, C., Benach, J., Quiroga, A., Rodrı́guez-Sanz, M., et al.
(2006). Politics and health outcomes. Lancet, 368, 1033–1037.
Pesut, D. P., Gledovic, Z. B., Grgurevic, A. D., Nagorni-Obradovic, L. M., & Adzic, T. N.
(2008). Tuberculosis incidence in elderly in Serbia: key trends in socioeconomic
transition. Croatian Medical Journal, 49, 807–812.
Piot, P., Greener, R., & Russell, S. (2007). Squaring the circle: AIDS, poverty, and
human development. PLoS Medicine, 4, 1571–1575.
Porter, D. (2006). How did social medicine evolve, and where is it heading? PLoS
Medicine, 3, 1667–1672.
Porter, J. D. H., & Grange, J. M. (1999). Tuberculosis – An interdisciplinary perspective.
London: Imperial College Press.
Prince, M., Patel, V., Saxena, S., Maj, M., Maselko, J., Phillips, M. R., et al. (2007). No
health without mental health. Lancet, 370, 859–877.
Raviglione, M. (2007). Infectious diseases and globalization. Rome, Vatican City.
Raviglione, M., & Pio, A. (2002). Evolution of WHO policies for tuberculosis control,
1948–2001. Lancet, 359, 775–780.
Raviglione, M., & Uplekar, M. (2006). The new Stop TB Strategy of WHO. Lancet, 367, 952.
Reidpath, D. (2008). Social determinants of HIV/AIDS. In E. Blas, & A. K. Sivasankara
(Eds.), Priority public health conditions: From learning to action on social deter-
minants of health. Geneva: World Health Organization.
Rhefuess, E. (2006). Fuel for life – Household energy and health. Geneva: World
Health Organization.
Rieder, H. (1999). Epidemiologic basis of tuberculosis control. Paris: International
Union Against Tuberculosis and Lung Disease.
Rodriguez De Marco, J., Sanches, D., & Alvarez Goya. (2007). El control de la tuber-
culosis en Uruguay – 25 anos de la implantacion del programa nacional de control
de la tuberculosis. Washington: Pan American Health Organization.
Rosen, G. (1974). From medical police to social medicine: Essays on the history of
health care. New York: Science History Publication.
Schmidt, L. A., Mäkelä, P., Rehm, J., & Room, R. (2008). Alcohol and social deter-
minants of health. In E. Blas, & A. K. Sivasankara (Eds.), Priority public health
conditions: From learning to action on social determinants of health. Geneva:
World Health Organization.
Shimao, T. (2005). Tuberculosis and its control – lessons from the past and future
prospects. Kekkaku, 80, 481–489, (in Japanese).
Slama, K., Chiang, C. Y., Enarson, D., Hassmiller, K., Fanning, A., Gupta, P., et al. (2007).
Tobacco and tuberculosis: a qualitative systematic review and meta analysis. The
International Journal of Tuberculosis and Lung Disease, 11, 1049–1061.
Smith, J. P. (2007). Nature and causes of trends in male diabetes prevalence,
undiagnosed diabetes, and the socio-economic status health gradient. PNAS,
104, 13225–13231.
Stevenson, C. R., Critchley, J. A., Forouhi, N. G., Roglic, G., Williams, B. G., Dye, C., et al.
(2007). Diabetes and the risk of tuberculosis: a neglected threat to public
health? Chronic Illness, 3, 228–245.
Stop TB Partnership. (2006). The Global Plan to Stop TB 2006–2015. WHO/HTM/STB/
2006.35. Geneva: World Health Organization.
Styblo, K., & Bumgarner, J. R. (1991). Tuberculosis can be controlled with existing
technologies: Evidence. The Hague: Tuberculosis Surveillance Research Unit.
Suarez, P. G., Watt, C. J., Alarcon, E., Portocarrero, J., Zavala, D., Canales, R., et al.
(2001). The dynamics of tuberculosis in response to 10 years of intensive
control effort in Peru. Journal of Infectious Diseases, 184, 473–478.
Subramani, R., Santha, T., Frieden, T. R., Radhakrishna, S., Gopi, P. G., Selvakumar, N.,
et al. (2006). Active community surveillance of the impact of different tuber-
culosis control measures, Tiruvallur, South India, 1968–2001. International
Journal of Epidemiology, 36, 387–393.
Tremblay, G. A. (2007). Historical statistics support a hypothesis linking tubercu-
losis and air pollution caused by coal. The International Journal of Tuberculosis
and Lung Disease, 11, 722–732.
UNAIDS, & WHO. (2006). AIDS epidemic update: Special report on HIV/AIDS:
December 2006. Geneva: UNAIDS.
Van Rie, A., Beyers, N., Gie, R. P., Kunneke, M., Zietsman, L., & Donald, P. R. (1999).
Childhood tuberculosis in an urban population in South Africa: burden and risk
factor. Archives of Disease in Childhood, 80, 433–437.
Vree, M., Duong, B. D., Sy, D. N., Co, N. V., Borgdorff, M. W., & Cobelens, F. G. J. (2007).
Tuberculosis trends, Vietnam. Emerging Infectious Diseases, 13, 796–797.
Vynnycky, E., Borgdorff, M. W., Leung, C. C., Tam, C. M., & Fine, P. E. (2008). Limited
impact of tuberculosis control in Hong Kong: attributable to high risks of
reactivation disease. Epidemiology and Infection, 136, 943–952.
Watt, C., Hosseini, M., Lönnroth, K., Williams, B., & Dye, C. (2008). The global
epidemiology of tuberculosis. In Schaaf, & Zumla (Eds.), Tuberculosis. London:
Global Medicine, Elsevier.
WHO. (1978). Declaration of Alma-Ata. International Conference on Primary Health
Care, Alma-Ata, USSR, 6–12 September 1978, Geneva: World health organiza-
tion from :http://www.who.int/hpr/NPH/docs/declaration_almaata.pdf.
WHO. (1991). Forty-fourth World Health Assembly. WHA44/1991/REC/1. Geneva:
World Health Organization.
WHO. (1994). WHO Tuberculosis Programme – Framework for effective tuberculosis
control. WHO/TB/94.179. Geneva: World Health Organization.
WHO. (2002). An expanded DOTS framework for effective tuberculosis control. WHO/
CDS/TB/2002.297. Geneva: World Health Organization.
WHO. (2005). Addressing poverty in TB control – Options for national TB control
programmes. WHO/HTM/TB/2005.352. Geneva: World Health Organization.
WHO. (2006). The Stop TB Strategy – building on and enhancing DOTS to meet the TB-
related Millennium Development Goals. WHO/HTM/TB2006.368. Geneva: World
Health Organization.
WHO. (2007). Global tuberculosis control. WHO/HTM/TB2007.376. Geneva: World
Health Organization.
WHO. (2008a). Global tuberculosis control. WHO/HTM/TB2008.393. Geneva: World
Health Organization.
WHO. (2008b). Tobacco use and the social determinants of health. In E. Blas, &
A. K. Sivasankara (Eds.), Priority public health conditions: From learning to action
on social determinants of health. Geneva: World Health Organization.
WHO. (2009). Global tuberculosis control, 2009. WHO/HTM/TB/2009.411. Geneva:
World Health Organization.
WHO/EMRO. (2006). Diagnostic and treatment delay in tuberculosis. Cairo: World
Health Organization, Regional Office for the Eastern Mediterranean.
WHO/WPRO. (2004). Reaching the poor: challenges for TB programmes in the Western
Pacific Region. WHO/HTM/TB/2004.352. Manila: World Health Organization.
Wilson, L. G. (2005). Medicine, population and tuberculosis. International Journal of
Epidemiology, 34, 521–524.
Wojcicki, J. M. (2005). Socio-economic status as a risk factor for HIV infection in
women in east, central and southern Africa: a systematic review. Journal of
Biosocial Science, 37, 1–36.
Zuniga, M., & Rojas, M. (2002). Programa nacional de control de la tuberculosis ano
2000 – Avances hacia la eliminacion. Revista chilena de enfermedades respira-
torias, 18, 55–63.