• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jurbhealthspringer.comThis journalToc AlertsSubmit OnlineOpen ChoiceThis journal
J Urban Health. Nov 2008; 85(6): 938–951.
Published online Oct 18, 2008. doi:  10.1007/s11524-008-9325-4
PMCID: PMC2587653

Quantification of Urbanization in Relation to Chronic Diseases in Developing Countries: A Systematic Review

Abstract

During and beyond the twentieth century, urbanization has represented a major demographic shift particularly in the developed world. The rapid urbanization experienced in the developing world brings increased mortality from lifestyle diseases such as cancer and cardiovascular disease. We set out to understand how urbanization has been measured in studies which examined chronic disease as an outcome. Following a pilot search of PUBMED, a full search strategy was developed to identify papers reporting the effect of urbanization in relation to chronic disease in the developing world. Full searches were conducted in MEDLINE, EMBASE, CINAHL, and GLOBAL HEALTH. Of the 868 titles identified in the initial search, nine studies met the final inclusion criteria. Five of these studies used demographic measures (such as population density) at an area level to measure urbanization. Four studies used more complicated summary measures of individual and area level data (such as distance from a city, occupation, home and land ownership) to define urbanization. The papers reviewed were limited by using simple area level summary measures (e.g., urban rural dichotomy) or having to rely on preexisting data at the individual level. Further work is needed to develop a measure of urbanization that treats urbanization as a process and which is sensitive enough to track changes in “urbanicity” and subsequent emergence of chronic disease risk factors and mortality.

Electronic supplementary material

The online version of this article doi:10.1007/s11524-008-9325-4 contains supplementary material, which is available to authorized users.

Keywords: Urbanization, Chronic disease, Systematic review, Developing countries

Introduction

During and beyond the twentieth century, the proportion of the world’s population living in urban areas has grown from 14% to over 50%.1 This transition was most keenly observed in developing countries; in South Korea, for example, there has been a fivefold increase in the number of people living in urban areas in the past 30 years.2 These demographic shifts are associated with many other changes including improved public hygiene, environmental sanitation, greater access to health care, increased individual wealth, changing employment and work force structures, and shifts in dietary and physical activity patterns.3,4 These fundamental changes have clear implications for disease patterns and particularly for the emergence of noncommunicable diseases (NCD).5 Between 1990 and 2000, the population prevalence of NCDs rose from 47% to 56% in developing countries. It is predicted that NCDs will account for 69% of all deaths in developing countries by 2020 and that cardiovascular disease will become the leading cause of mortality.6 The burden of this transition in the disease epidemiology will be heavier in the developing world compared to the developed because, in the developing world, the majority of sufferers are expected to be relatively young, of lower socioeconomic status, and to suffer from a more severe and premature onset of disease.7

The MESH term urbanization is used to describe “The process whereby a society changes from a rural to an urban way of life. It refers also to the gradual increase in the proportion of people living in urban areas” (see http://www.ncbi.nlm.nih.gov/sites/entrez).

Despite this, most studies that examine the relationship between urbanization and chronic disease use a national level rural/urban dichotomy to summarize urbanization. Vlahov and Galea show that for the 228 countries for which the United Nations (UN) collect data, almost half use a basic administrative definition of urban (e.g., living in the capital city), around a quarter define urbanization using population measures (e.g., size and density), and one in eight use functional characteristics (e.g., economic activity).8 In a cross-country analysis of over 100 countries, Ezatti et al. found that body mass index (BMI) and cholesterol increased rapidly with national income and level of urbanization.1 Reporting results from the Chinese Health and Nutrition Survey and data from the UN Food and Agriculture Organization, Mendez and Popkin show that urbanization and globalization enhance access to nontraditional foods resulting in less healthy dietary patterns.9 A large risk factor surveillance study conducted in India found that the prevalence of diabetes was two and a half times higher in urban areas when compared to rural areas.10

The broad measures of urbanization described above have been useful in establishing associations between level of urbanization, chronic disease risk factors and chronic diseases, also known as the urban health penalty. These studies are limited, however, because they lack the specificity to identify the exact changes within the urbanization process that are responsible for the emergence of chronic disease risk factors. Research about the structure of modern urban living and the ways that this can influence health have received scant academic attention,8 and this scarcity of research makes it impossible to examine temporal changes in the urban environment and subsequently changes in patterns of disease. A second complication is the disparate academic traditions involved in this type of research including geography, epidemiology, sociology, and urban planning.11 Many define urbanization using a simple dichotomy (urban/rural) or even a single continuous variable (population density). This removes the potential to understand the specific changes within the process of urbanization that lead to changes in risk and disease. Furthermore, the idea of a “threshold” delineating an urban area as opposed to other types of living environment ignores the graduation in disease states between and within geographical areas and assumes that populations remain static.

Socioeconomic, cultural, political, and environmental factors have a powerful influence on population health-related behaviors and subsequent health outcomes. A number of authors propose a framework for understanding the relationship between the social and physical environments that define “urban” as being influenced by municipal factors including government, civil society, and national and global trends.9,10 This framework posits that the physical, social, economic, and political elements of the urban environment have some effect on the health of all residents. In addition, there are a number of upstream influences on living conditions including municipal (governments, markets, etc.), national, and international trends that may effect health. This taxonomy describes how population composition relates to the demographic characteristics of the population living within a city including fertility, migration, and immigration. The social environment describes properties of the urban community such as socioeconomic status, crime and violence, the population mix and attendant risk factors, and so on. In practice, elements of interest in the physical environment may include features of the built environment or levels of air, water, and noise pollution. For public health and chronic disease, areas of interest might include access to sanitation, supermarkets, paved roads, etc. A second important consideration is the provision of health and social services, which in turn is interrelated to the physical and social environment, health insurance arrangements, and availability, equity, and quality of services.

We know that “urbanization” is associated with the emergence of NCDs. We do not know what this process of “urbanization” entails nor do we know which of the elements of “urbanization” are associated with which NCDs. An understanding of the individual elements of “urbanization” and how these lead to the development of NCD risk is an important first step in identifying potential sites for intervention at a population level. This review is the first step in developing a more detailed picture of the ways in which previous peer-reviewed research has shown how “urbanization” contributes to the emergence of chronic diseases in the developing world.

In this paper, we set out to identify how urbanization has been measured in studies that examine the relationship between urbanization and chronic diseases in developing countries. We aim to list the elements of urbanization that have been studied in relation to chronic diseases and their risk factors. In response to this aim, our research question was:

How has urbanization been defined and measured in previous observational studies of urbanization and chronic disease in developing countries?

Methods

Search Strategy

The full systematic review of urbanization and chronic disease was preceded by a pilot phase in order to develop the correct search terms for our full review. In April 2007, we searched PUBMED for studies under the MESH terms “urbanization,” “chronic disease,” or “coronary heart disease” using the appropriate truncation and wildcards. This initial search returned 83 papers, of which 21 were review papers and two were primary studies reporting some quantification of urbanization. We categorized all of the MESH terms from all 83 papers (n = 492) in this pilot set into four subcategories: (1) urbanization as an exposure; (2) other exposures and outcomes; (3) outcomes related to a specific chronic disease; and (4) participant demographics. Full results of this pilot phase are available from the authors upon request. A full search strategy was developed by two of the researchers (SA and CF) using this set of 492 MESH terms as the basis for the search strategy used in the full review.

Terms used for urbanization included “acclimatization,” “urbanization,” and “modernization.” Terms used for chronic disease included “chronic disease,” “cardiovascular disease,” and “cancer.” Chronic disease is a root term that includes specific diseases such as cardiovascular disease, obesity, and diabetes. Terms used for developing countries included “developing countries,” regional descriptors (such as Africa), and the full list of developing countries and regions listed by the World Bank.12 Truncation was used to overcome differences between American and English spelling. The full search strategy is available online from the journal website.

We searched the MEDLINE, EMBASE, CINAHL, and GLOBAL HEALTH electronic databases for English language studies published after January 1990 and before August 2007. We also searched reference lists of reviews found from the pilot search study, from the main search, and from the final included studies.

Study Selection and Inclusion Criteria

We included any study that aimed to examine the relationship between urbanization and chronic disease in a developing country (Table 1). The study also had to report details on the construction of the urbanization variable used and had to have at least one chronic disease as an outcome. Once duplicate studies were removed from the total hits (n = 1,285), two reviewers (SA, LH) independently assessed the remaining titles and abstracts (n = 868) for inclusion. The full paper was assessed against the inclusion criteria where ambiguity remained after reading the title or abstract. Any disagreements at this stage were resolved by a third reviewer (CF). A number of studies were excluded because they did not include a measure of urbanization and several others because they reported data from a developed country.

Table 1
Results of search and application of inclusion criteria

Data Extraction

We extracted data from the final included studies using two independent reviewers, and the discrepancies in final data extraction were resolved by discussion with the third reviewer. We extracted all available data from the publication reporting each study. We categorized the studies into high or low quality based on the level of detail reported by the paper on the method of development of a measure of urbanization. We adopted this approach because the focus of our review was to describe the methods used to quantify urbanization rather than presenting results against criteria for internal or external validity.

Data collected from each paper were methods, definition and measurement of urbanization, sources of data, sampling frame, individual or area level data, and statistical and analytical methods for creating summary variables. The final data corpus of included studies is presented in Table 2.

Table 2
Summary of characteristics of included studies by quality and publication date

Results

Of 868 titles and abstracts identified by the searches, nine studies remained after application of the exclusion criteria. These studies are presented (in quality and publication date order) in Table 2. Five studies were conducted in a single developing country while the remainder was conducted in at least two or more countries.

We divided the studies into two types: studies that examined the relationship between urbanization and health at a single time point and those that used repeated measures over time. The most common comparisons were made between populations living in urban and rural areas across a range of outcomes. These outcomes included measures of behaviors for communicable and noncommunicable diseases (e.g., physical activity, calorific intake, and sexual/reproductive health), anthropometric data (weight, body mass index, and blood pressure), morbidity (diabetes), mortality and sociodemographic data (income and education).

We found considerable differences in the definitions and measurements of urbanization. Studies adopted either a simple definition of urbanization using demographic data or a complex one that summed categories of variables into an urbanization index. Table 3 presents the different components of measures of urbanization by included studies in quality and publication date order.

Table 3
Components of measures of urbanization by included studies in quality and publication date

Studies using a Single Demographic Measure to Define Urbanization

In this first group, we found four studies that defined urbanization as a simple product of population density within an area.2 This definition was based on the UN dichotomy of urban/rural and was obtained from secondary data sources (e.g., population census data, aggregated to an area level measure). Dahly and Adair have described a number of problems with the UN dichotomy, including differing definitions and measurement across different countries.13 One additional study classified their sample into urban or rural groups using the proportion of the respondent’s life spent in an urban environment.14 This data was collected at an individual level from the study population using an unspecified measure of urbanization.

Studies using Composite Measures to Define Urbanization

The remaining four studies defined urbanization from a range of different types of variables using either individual level data or secondary data collected at an area level. One South African study categorized respondents into one of five groups based on where they lived and their occupations.15 This study categorized people according to whether they lived in rural traditional villages, worked on commercial farms, lived in informal housing and whether they were laborers in industry or professionals. An earlier South African study constructed an urbanization score from measures of place of birth, number of years spent in urban area since leaving school, home ownership, land, livestock, and presence of family in the area.16

Two studies created multivariate indices of urbanization by using measures taken at both an individual and area level. Mendez and Popkin created an urbanization index using ten measures: population size, population density, access to markets, transportation, communications/media, economic factors, environment/sanitation, health, education, and housing quality.9 The index was developed using data from community surveys, supplemented with household level information. The authors classified areas into different levels of urban and rural according to scores.

Dahly and Adair13 used an approach based on Mendez and Popkin9 to define urbanization using a multivariable scale. Their “urbanicity scale” was based on two criteria: content validity (an a priori assessment, based on authoritative sources such as peer-reviewed literature, of whether an item truly reflects urbanicity) and the availability of relevant data. The authors did provide an a priori definition of urbanization mimicking the variable used previously. However, they were unable to obtain the same data sources for comparable community level data on housing infrastructure, economic indices, and sanitation as Mendez and Popkin.9 They were forced to develop their scale on a mix of measures thought appropriate by the researchers and the data available.

The “urbanicity scale” used the following items: population size; population density; communications (e.g., the presence of phone service, mail, newspapers, the internet, cable TV, and cellular phones); transportation (e.g., the density of paved roads and the availability of public transportation); educational facilities (e.g., the presence of educational institutions, including primary and secondary schools, colleges, and vocational schools); health services (e.g., the presence of health services, including hospitals, medical clinics, maternal health clinics, family planning clinics, and community health centers); and markets (e.g., the number of sari-sari stores [small, retail shops] and the presence of drug stores, grocery stores, and gas stations).

Each of the seven components was scored to create a value from 0 to 10, resulting in a scale ranging from 0 to 70 across all items. The authors argued that using a continuous measure of urbanicity allowed for better illustrations of the relationships between urbanicity and health. Their scale allowed comparisons of similar measures over time, and analyses of its components could identify specific factors which vary consistently across urbanizing environments. Their new scale was a better measure of urbanicity than the traditionally used urban/rural dichotomy. The scale allowed a better reflecting of differences between urban and rural areas as well as heterogeneity across areas. These differences might not exist within a developed country as all participants might share similar access to services or goods.

Discussion

Statement of Principal Findings

We identified nine studies that quantified urbanization in relation to chronic disease or chronic disease risk factors. Most studies relied on secondary demographic and geographic measures of urbanization comparing populations living in urban and rural areas across a range of outcomes. Four studies used either secondary or primary data or combinations of the two at area and individual level to develop summary scores from multiple proxy measures of an urban environment.

Area level measures include population size, population density, proportion of population living in urban areas, health, education, housing quality, economic factors, environment/sanitation, and access to markets, transportation, and communications/media.

Individual measures include place of residence, number of years spent in urban area since leaving school, proportion of life spent in an urban environment, occupation, home, land and livestock ownership, and proximity of family. None of these studies provided an objective validation of these measures of urbanization.

Strengths and Weaknesses of the Study in Relation to Other Studies, Discussing Particularly any Differences in Results

This is the first review of its kind which takes a systematic approach to identifying the ways in which urbanization is measured as an exposure variable in relation to chronic disease. The strengths of the approach include the systematic nature of the review. The initial pilot phase made it possible to generate a broad range of search terms before a full search strategy was built. The application of inclusion criteria ensures that only those studies relevant to the research question as posed are included in the review.

xStringent inclusion criteria meant we conducted a focussed review but that the number of papers (n = 9) included was limited. The data corpus could be expanded by including those studies without an explicit definition of urban. Such an exercise could compare and contrast the relationships between urban and whichever outcome is of interest to the paper with an examination of how comparable outcomes varied by the measure of urban utilized by the study authors. This would not suit the current research question, however, as we were interested specifically in how these studies had defined “urban.”

Some of these studies appeared to be of low quality and the quality measure described in the methods was an attempt to take this into account within the review. The majority of studies reviewed provided poor explanations of the methods for deriving the reported urbanization measure. This happened most often in those studies using secondary measures (e.g., population density).2,17

Our initial pilot study led us to believe that the MESH term urbanization is being misused. The majority of studies reviewed does not consider “urbanization” as a process but rather compare geographical or administrative locations based on urban/rural dichotomies. We have tried to overcome this problem by expanding the terms used to represent urbanization to include modernization, etc. Identifying the possible misuse of “urbanization” proved an invaluable result of the pilot study and should be an important step of any systematic review in an untried area.

Finally, our requirement that papers be published in English may have limited the number of papers returned. This may be particularly so for studies reporting data from developing countries who do not have English as their first language.

Meaning of the Study: Possible Mechanisms and Implications for Clinicians or Policymakers

While most authors agree that urbanization refers to the process of becoming urban, there is little agreement about how these phenomena can be measured or studied in association with population health or other outcomes.18 The majority of studies reviewed relied on preexisting measures of urbanicity using an urban rural dichotomy, identifying that urban settings adversely affect health but not how or why such settings may affect health.8 A subset of these studies use arbitrary cut points to create a third intermediate level of urbanization. These measures are, therefore, crude and may mask variations (intra—within country and inter—between countries) across population areas because of the scale used in classifying population density.

Using this crude dichotomy means that it is not possible to understand why urbanization affects chronic disease risk. It also makes it impossible to track changes that may occur in the urban environment or indeed identify components of the urbanization process that may affect disease risk.

There are other, more recent measures, which may serve as surrogates for urbanization. At an area level, this list might include the proportion of households living below the poverty line, the human development index, etc. The measures have their own limitations and as with those found in this review have not been validated as measures of the process of “urbanization.” They may be hampered by their timeliness and by the pace of urbanization outstripping the collection and attribution of these measures.

Unanswered Questions and Future Research

Our study suggests that we need definitions of “urbanization” that are culturally and country-specific to study the effects of urbanization both within and between developing countries. For researchers planning studies taking urbanization as an exposure, it is important to consider urbanization as a process rather than a static measure. Any scale measure attempting to capture proxies for “urbanization” should ensure that individual and area level measures are included. Any subsequent analysis should then consider multilevel models providing comparisons at individual and area level. Similarly, outcome measures might be considered both at population and individual level. Area level measures that should be considered include those used traditionally such as population size and density but other measures should also be considered including quality of roads, distances to markets, types of markets available, transport options, types and accessibility of employment, health services, and so on. Other methods of data collection such as geographical information systems should also be considered.

At a macro level, we suggest that comprehensive measure of urbanization should be constructed using qualitative and quantitative data. The definition of urbanization should be validated where appropriate at an area level (defined by the population studied) using ground truthing methods (environmental audits). Ideally, data informing this process should be contemporaneous and have temporal and seasonal stability at baseline and follow-up points.

It will not be sufficient only to observe the problem that urbanization is associated with an increasing risk of morbidity and mortality. This situation offers little in terms of public health interventions to combat the inevitable changes that will happen within developing countries. This transition is happening in the developing world. More precise and validated measures of urbanization are, therefore, needed urgently to identify which aspects of urbanization are associated with different health and social outcomes within which population groups and communities. Identifying these “tipping points” for the emergence of chronic disease risk opens the potential for policy and regulatory intervention within the development cycle.

Conclusion

Our systematic review of studies that measure urbanization as an exposure for chronic disease has found that there are limited empirical studies. Most studies rely on secondary data providing population level summaries at an area level. A more comprehensive measure is used that is sensitive and specific to the process of urbanization as it affects chronic disease risk.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Appendix 1(91K, pdf)

Search strategy (PDF 91kb)

Footnotes

Electronic supplementary material

The online version of this article doi:10.1007/s11524-008-9325-4 contains supplementary material, which is available to authorized users.

References

1. Ezzati M, Vander Hoorn S, Lawes CMM, et al. Urbanisation and health. Clin. Med. 2005;5:137–141.
2. Choi JY, Min Cho Y, Ku Park C, et al. Rapidly increasing diabetes-related mortality with socio-environmental changes in South Korea during the last two decades. Diabetes Res Clin Pract. 2006;74:295–300. [PubMed]
3. Murray CJL. Rethinking the “diseases of affluence” paradigm: global patterns of nutritional risks in relation to economic development. PLoS Med. 2005;2:404–412. [PMC free article] [PubMed]
4. Popkin BM. The nutrition transition and its health implications in lower income countries. Public Health Nutr. 1998;1:5–21. [PubMed]
5. World Health Organization. The World Health Report 2007: Global Public Health Security in the 21st Century—A Safer Future. Geneva: World Health Organization; 2007.
6. Boutayeb A, Boutayeb S. The burden of non communicable diseases in developing countries. Int J Equity Health. 2005;4:2. [PMC free article] [PubMed]
7. Reddy KS, Yusuf S. Emerging epidemic of cardiovascular disease in developing countries. Circulation. 1998;97:596–601. [PubMed]
8. Vlahov D, Galea S. Urbanization, urbanicity, and health. J Urban Health. 2002;79(4):S1–S12. [PMC free article] [PubMed]
9. Mendez MA, Popkin BM. Globalization, urbanization and nutritional change in the developing world. Electron J Agric Dev Econ. 2004;1:220–241.
10. Mohan V, Mathur P, Deepa R, et al. Urban rural differences in prevalence of self-reported diabetes in India—The WHO-ICMR Indian NCD risk factor surveillance. Diabetes Res Clin Pract. 2008;In Press. [PubMed]
11. Galea S, Freudenberg N, Vlahov N. Cities and population health. Soc Sci Med. 2005;60:1017–1033. [PubMed]
12. World Bank. World Bank list of developing countries. Available at: http://web.worldbank.org/WBSITE/EXTERNAL/DATASTATISTICS/0,,contentMDK:20421402~pagePK:64133150~piPK:64133175~theSitePK:239419,00.html. Accessed on: April 1, 2007.
13. Dahly DL, Adair LS. Quantifying the urban environment: a scale measure of urbanicity outperforms the urban-rural dichotomy. Soc Sci Med. 2006;64:1407–1419. [PMC free article] [PubMed]
14. Steyn K, Kazenellenbogen JM, Lombard CJ, Bourne LT. Urbanization and the risk for chronic diseases of lifestyle in the black population of the Cape Peninsula, South Africa. J Cardiovasc Risk. 1997;4(2):135–42. [PubMed]
15. Vorster HH, Venter CS, Wissing MP, Margetts BM. The nutrition and health transition in the North West Province of South Africa: a review of the THUSA (Transition and Health during Urbanisation of South Africans) study. Public Health Nutr. 2005;8(5):480–90. [PubMed]
16. Pick W, Cooper D. Urbanisation and women’s health in South Africa. Afr J Reprod Health. 1997;1(1):45–55 March. [PubMed]
17. Sobngwi E, Mbanya JC, Unwin NC, et al. Exposure over the life course to an urban environment and its relation with obesity, diabetes, and hypertension in rural and urban Cameroon. Int J Epidemiol. 2004;33:769–776. [PubMed]
18. Galea S, Vlahov D. Urban health: evidence, challenges, and directions. Annu Rev Public Health. 2005;26:341–365. [PubMed]

Articles from Journal of Urban Health : Bulletin of the New York Academy of Medicine are provided here courtesy of New York Academy of Medicine
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...