Version classiqueVersion mobile

Essays on population and space in India

Christophe Z. Guilmoto
Alain Vaguet

Part IV. Health and epidemics in Cities

9. Diffusion of HIV in Mumbai

Emmanuel Eliot

Texte intégral

1Asia is thought to have been the last continent to be touched by the pandemic of the Human Immunodeficiency Virus (HIV) and yet, today, some of its countries are reporting incidence rates of the virus that are amongst the highest in the world, sometimes even exceeding those of sub-Saharan Africa.

2According to UNAIDS at the end of 1998, two countries of this region seem to be particularly contaminated : Thailand (26,0000 declared cases of AIDS and 780,000 HIV positive people) and India (estimations of 350,000 AIDS cases and 1,750,000 seropositive cases). Yet, epidemiological reports far too often veil the complex nature of the epidemic. At the same time they mask problems of under-registration that are very common in the developing countries.

3This article takes a critical look at the available figures and attempts to analyse the diffusion of HIV at different scales in India (national, regional, urban and intra-urban), and in particular in Mumbai (formerly Bombay). In this city, not all areas have been equally affected by the virus and not all the groups have been infected in the same way. The spread of HIV does not seem to be ubiquitous. In Mumbai, for example, areas known as “pathogenic areas” for decades seem to have been particularly favourable localities for its spread.

4Three major types of figures have been utilized :

5-Annual reports of the Indian Ministry of Health. The one used is dated February 1998. The first tests were carried out in August 1986 in Chennai (Madras) and Mumbai.

6-Figures from the Directorate of Health Services of Maharashtra state. They are available for each testing centre from 1991 to 1996.

  • 1 Carried out in winter 1995 and the summers 1995 and 1996. Statistical sources used in this paper al (...)
  • 2 Researcher at the Tata Institute of Social Sciences, Mumbai.

7 - A personal survey1 carried out in collaboration with Dr S. Bharat2 in Mumbai. This took the form of a sampling of 15 of the 20 most important public and private HIV testing centres in this city.

Figure 9.1 : Prevalence of HIV in India (1986/1998), by state and territory

Figure 9.1 : Prevalence of HIV in India (1986/1998), by state and territory

Figures and methodology

Official sources (National AIDS Control Programme).

8These are the results of numerous serological samplings carried out on representative samples of the population at large. Sero-epidemiology is certainly helpful above all for studying modes of transmission, but it also makes it possible to present a valid picture of the diffusion of HIV at any particular moment.

9However, this type of figures raises two problems. On the one hand, not all the cases enumerated have been confirmed by a Western Blot test— guarantee of the reliability of the data. In fact, since the beginning of 1995, only two Elisa tests have been carried out with two different sets of equipment in the public testing centres. The cost of the Western Blot — almost Rs. 1500—had become too high for the Indian government. Nevertheless, the new generation of Elisa gives rise to less risk of mistaken identification than the previous one.

10On the other hand, apart from temporal errors of record, the main problem remains under-registration. The figures from testing centres are sent neither regularly nor correctly to the central Ministry of Health in New Delhi.

11Nevertheless, this information constitutes the only available sources regarding the prevalence of HIV in the 24 states and 7 territories of the Indian Union.

State sources : the example of the Directorate of Health Services, Maharashtra.

12In this state, 28,463 cases were HIV positive from the 274,417 individuals tested between August 1986 and June 1996, a number five times greater than the 6,494 recorded by the central government for the same period. Discussions with health official reveal that the problem of under-registration persists, although it is less “chronic” than at the all-India level.

A personal survey called for

13To mitigate the risk of incorrect reporting of figures, it appeared to be of primary importance to gather information in the places where HIV is being tested. This enquiry was carried out in the form of a sampling in Mumbai, the economic capital of the country. It was done for the period 1988-1994, because from 1986 to 1988 the testing system was not fully effective (tests wrongly carried out, records improperly kept). Only the largest establishments in the city, and those agreeing to the survey, were included in the enquiry. In each centre, the sex, age, place of the residence, type of transmission and occupation of the individuals found to be HIV-positive were collected. Moreover, only cases confirmed by a Western Blot test or two Elisa tests have been taken into consideration. This second option made it possible to reduce the risk of incorrect reporting. In fact, some blood samples found positive by Elisa were not always forwarded to the centres performing the Western Blot confirmation test.

14From the 15 establishments investigated, 10,209 cases have been recorded, and 88,7 per cent of these were resident in Mumbai. This information certainly refers only to detected cases, but at present, it constitutes the only basic available data to study diffusion of HIV in the capital of Maharashtra. It is obvious that these figures constitute “the tip of the iceberg” (P. Gould) of the epidemic. Nevertheless, it proved extremely difficult, if not impossible, to find out how many tests had been carried out. Records were in fact too often badly kept, and some centres recorded only cases that were found positive.

Figure 9.2 : Prevalence of HIV in Maharashtra (1991/1995), by district

Figure 9.2 : Prevalence of HIV in Maharashtra (1991/1995), by district

Diffusion of HIV in Mumbai

15Three sets of areas in India seem particularly severely affected by the virus (Figure 9.1).

16The north-eastern states (Manipur and Nagaland) have the highest rates of prevalence in the entire country, almost 100 cases per 1 000. They are situated on the margins of the famous drug-trading “Golden-Triangle.” The greater part of the infections there is linked with intravenous consumption of heroin. Regions of conflict—Assam and the Punjab—form the second set. It seems as if times of war or inter-communal tension may be favourable to the spread of the epidemic (Vaguet, 1990). In this case, perhaps Kashmir should be added to this group ? In the Punjab, tensions persist but are not as violent as during the 1980s. Moreover the small size of the sample tested—780— should be noted, as it may cast some doubt on the high rate of prevalence.

17Lastly, the western part of the country, and particularly the states of Goa and Maharashtra, seem to be very much affected by HIV. In the latter state, analysis of the prevalence of the virus between 1991 and 1995 reveals a very heterogeneous and complex pattern of contamination (Figure 9.2). The urban districts of the two major economic centres of the north-west of the subcontinent, Pune and Mumbai, are highly infected. But their industrial fringes (Thane and Ahmednagar) register the highest levels of HIV incidence. The southern borderline districts of Kholapur and Sangli-major road and rail junctions to south India have comparable rates. National Highway 4, from Mumbai to Bangalore, the “Silicon Valley” of India, is much used by lorry drivers, and connections between these two major economic centres of the country are particularly close. Prostitutes from nearby towns and villages move towards communication axes to exercise their profession, because of the size of the clientele. And it seems that these groups are playing a role in the diffusion of HIV throughout the country (Eliot, 1997).

18Mumbai is the capital of Maharashtra and the first largest city in India, with almost 13 million inhabitants. It constitutes one of the major economic centres, and a “window” for the country. According to Indian officials, it is also the city that has been most affected by HIV.

  • 3 In the case of Mumbai, municipal and district boundaries coincide.

19Perhaps, it is also one of the most thoroughly tested. In fact, one of the first recorded cases of the country was reported in this city in 1986. This was an Indian businessman who had received a transfusion of contaminated blood when undergoing an open-heart surgery in the United States. A system of testing was quickly set up throughout Maharashtra, and the number of tests carried out in this region is one of the highest in the whole country. In fact, the Maharashtrian capital appears to be a highly infected district.3 It is therefore a very suitable location for analyzing the diffusion of the virus.

HIV in Mumbai : long-standing “pathogenic areas ”

20The various stages in the industrialization of the city should be seen alongside the settlement of populations (Figure 9.3). The wealthier classes have settled in the west of the town, to avoid pollution from the textile spinning mills of the 19th and early 20th centuries, and today’s petrochemical industries. The centre and east have received the very poor, the working classes and the middle classes. There is a striking contrast in structures between the villas of the rich on Malabar hill, the worker’s flats (chawls) in the centre, an inheritance from the British era, and the middle-classes apartment blocks. In the south, the business quarter between Nariman Point and Fountain is located near to the historic core of the town, the Fort.

Figure 9.3 : Urban morphology and population : Central and South Mumbai

Figure 9.3 : Urban morphology and population : Central and South Mumbai

Figure 9.4 : Number of HIV positive detected in Mumbai by Wards (1988/1994)

Figure 9.4 : Number of HIV positive detected in Mumbai by Wards (1988/1994)

Figure 9.5 : Main HIV testing centres, cases in Mumbai (1994)

Figure 9.5 : Main HIV testing centres, cases in Mumbai (1994)
  • 4 In Mumbai, the hierarchy of administrative units, in decreasing order of size, is as follows: Wards (...)

21According to the figures collected in the 15 testing centres, the central wards4 (C, E, F/North, F/South, D, G/South) concentrate the highest number of HIV positive cases (Figure 9.4). The north of the city seems less affected. Is this linked to the fact that the great majority of the establishments where the testing is done are located in the heart of Mumbai ? In fact, the major health services are concentrated in this part of the town, and people therefore have to travel there to get a blood test done (Figure 9.5). Thus, while the number of recorded cases of HIV is perhaps incomplete in the northern wards of the city due to the absence of major testing centre, the one of the centre and south would seem to be indicative of the spread of the virus. In fact, almost all the biggest centres where the test is carried out have been investigated in this part of the town.

Figure 9.6 : Diffusion of HIV in “pathogenic” wards

Figure 9.6 : Diffusion of HIV in “pathogenic” wards

22Apart from the analysis of the concentration of HIV cases in the metropolis, a major element seems to be the link between the prevalence of HIV in 1994 and the mortality rates in 1986 at the ward scale. The question is whether HIV is propagated in locations that are special from the health point of view. The year 1986 was chosen because it is the date of the first tests carried out in India. The analysis of the correlation coefficient (r) between these two sets of data seems to be the most interesting tool (Figure 9.6). According to the quality of the figures, the coefficient is quite high (r = 0.6). Thus, the wards that had a high mortality rate in 1986—in the order of 10.5 to 11.5 per thousand—were the ones that in 1994 were most highly affected by HIV. Although the ward E has not been included in the analysis, it belongs to this group with a prevalence of 4.6 per thousand in 1994 and a mortality rate close to 11 per thousand in 1986. This correlation however is not perfect. Some wards with high mortality rates do not show a significant HIV prevalence. Nevertheless, the retrovirus seems to have spread first in “pathogenic areas. ” But this low level of salubrity in the central parts of Mumbai is not a recent thing.

  • 5 Health Officer’s report 1892. See Ramasubban and Crook (1996).

23Already at the end of the 19th century, British health officers were describing the situation of ward E for example as “intolerable.”5 In 1892, the mortality rates in some parts of this ward were almost 50 per thousand, whereas those in the south or further north in the city were in the order of 9 to 16 per thousand. The bad condition of ward E has persisted through the various stage of the industrialization of Mumbai. This part of the town has always been an area for the very poor, refugees and the homeless. Today, conditions have improved a little. This ward still contains a very high majority of people at a low or medium socio-economic level. It also seems to be one of the most congested areas of the city. The housing is very old and sometimes borders the insalubrity. Moreover, the communal strife between Hindus and Muslims in 1992-93 after the Ayodhya events was particularly violent in this part of Mumbai.

24Therefore, it seems that HIV appeared first in wards where health conditions have been substandard for decades. Although mortality rates have not increased everywhere since 1986, it appears that the part of deaths due to tuberculosis is increasing in an alarming way. The tuberculosis/HIV nexus constitutes a dangerous combination. The WHO estimated in 1990 that more than 75 per cent of carriers of the bacteria were living in developing countries. In 1991, moreover, the number of HIV carriers suffering from tuberculosis was estimated at 4.6 million. The tuberculosis/HIV nexus is already having dramatic consequences in developing countries where, according to some sources, 95 per cent of HIV positive individuals are suffering from tuberculosis.

25In Mumbai, only to wards K/East and B show a decrease or a relative stagnation in their rate of annual increase of deaths due to tuberculosis, with -2.0 and -7.8 per cent respectively, between 1986 and 1994. Whatever the reason, in the wards most affected by HIV the rates are among the highest in the city. For example, in ward E the number of deaths due to tuberculosis doubled between 1986 and 1994.

26However, in this part of Mumbai, which has the highest number of HIV positive cases (2,173), not all localities have been affected equally. The most affected sections are in the north and the south, Byculla and Kamathipura. Moreover the latter houses more than 37 per cent (803 cases detected) of all the HIV positive cases in the ward.

27It is also notorious in the town, as one of the main prostitution centres, and the best-known one, because of its central location. Prostitutes seem to be one of the major vectors of the diffusion of HIV in India (Eliot, 1997). Therefore, it would seem particularly worthwhile to analyse the spread of the virus in Kamathipura. This investigation is made easier by the fact that this quarter is marked by a strong spatial segregation between the locations reserved for the sex trade and those devoted to commercial and residential purposes.

Diffusion of HIV in Kamathipura

28Almost half of the best known centres for prostitution are situated in central Mumbai. Kamathipura is perhaps the most famous of them (Figure 9.7). Because of its central location in the city, it receives a very numerous clientele. It is located near Mumbai Central railway and bus terminals, and close to major communication nodes serving the north, south, east and west of the Maharashtrian capital. These axes are all concentrated in the northern part of the area.

Figure 9.7 : Kamathipura : a red-light area in the heart of Mumbai

Figure 9.7 : Kamathipura : a red-light area in the heart of Mumbai

29HIV seems to have affected red-light lanes first and more strongly than the others. In 1994, these streets contained most of the recorded cases : 497 as against 147 in the commercial and residential parts of Kamathipura. However, the entire red-light area does not seem to have been infected in a homogenous way. The north and the south of it were particularly affected. In fact, this spatial difference in the diffusion of HIV points to a rather distinct inter-communal segregation. Indians (mostly Hindu or Muslim) are concentrated in the northern lanes. Groups of Nepali origin, as well as the “Zenana” (transvestites) and “Hijra” (eunuchs) communities, have settled more in the centre and the south. Moreover in this part of the area the number of HIV positive cases detected seems to be lower (Figure 9.8).

30Prostitution is a business activity of the service type. It is therefore related to access. The streets nearest the communication nodes (station, bus-stands...) seem more contaminated by HIV. Meetings with the directors of various non-governmental organizations working with the prostitutes of this quarter confirm this : the most accessible streets, the ones in the north, are more frequently visited by the male clients. Nevertheless, the proximity of important communication nodes can not be the only reason for the relatively high number of HIV positive cases in the lanes occupied by the Hijra and Zenana communities. Here it seems necessary to call in a cultural factor. In fact this community is visited by Indians and even more so by Arabs from Gulf countries. There are even specialized travel agencies organizing “sex trips” between the Emirates and India.

31Nowadays the prostitutes of Kamathipura have the reputation of being highly infected by HIV. Many alarmist newspaper articles mention that 85 to 90 per cent of the prostitutes in certain lanes of this area are HIV positive. Although these percentages should be regarded with caution, clients are turning to other communities, the Hijras and Zenanas, as well as the Nepali girls, whose skin is fairer. Thus, in addition to an increased demand for young virgins related to the fear of contamination, fairness is presently becoming a mark of freedom from HIV infection. A fair complexion has always been a criterion of beauty in India. Today it is becoming in addition a “hypothetical assurance” of non-contamination by HIV.

32Therefore, a migratory move has been set in train within the boundaries of Kamathipura. Indian prostitutes with lighter skins, or with resemblance to the Tibeto-Nepali type, have moved to work in the central streets. This means that very soon this part of the district too will be as contaminated as the areas north and south of it.

Figure 9.8 : HIV diffusion in Kamathipura (by lane)

Figure 9.8 : HIV diffusion in Kamathipura (by lane)

33In the commercial and residential areas, the highest number of cases are located near to those streets devoted to the sex-trade that are most affected by HIV. However, it should be noted that the clients and people living in the eastern part of Kamathipura have perhaps not been tested as thoroughly as the prostitutes have. In fact, the first tests and epidemiological samplings were carried out among this group.

34Finally, analysis of HIV diffusion in Kamathipura, in addition to a high level of contamination among prostitutes, shows a lack of homogeneity in propagation of the virus, linked to a communal rationale of a kind of sexual tourism (Figure 9.9).

Figure 9.9 : The spatial dynamic linked with HIV in Kamathipura

Figure 9.9 : The spatial dynamic linked with HIV in Kamathipura

Governmental action against HIV.

35Most of the actions have concerned ward E. Prevention campaigns have been set up in the chawls, especially in Byculla, and red-light areas. Most public actions have focused on prostitutes. At the instigation of certain institutions, the extremist Hindu alliance between BJP (Bharathiya Janata Party) and Shiv Sena that used to form the government tried to legalize prostitution. The main objective of this was to be able to test prostitutes and assess the extent of the epidemic. Lamps were placed at the entrance to brothels. These were supposed to be a guarantee for clients that the prostitutes there were under regular medical supervision, not HIV positive and using condoms during intercourse. However, it was very easy for brothel-owners to obtain false licences and it became impossible to identify those which were visited by medical practitioners and those which were not.

36Another course of action was set in motion by the government to “clean up” this area, which was becoming highly infected by HIV. Immigrant prostitutes found to be HIV positive were sent back to their native regions, but some Maharashtrian women carrying the virus were not always expelled from Kamathipura. These actions can be seen in conjunction with the government’s policy of “Marathi nationalism”, giving preferential employment to native Maharashtrian Hindus.

37Thus, these policies focused primarily upon the prostitutes, who are accused of being the sole vectors for the diffusion of the virus. In spite of the increasing number of HIV positive cases, and the alarming rates of infection in Mumbai and all over India, the vast majority of official interventions have been limited to discriminatory practices. Being found HIV positive in Mumbai means, according to many state health officials, that you are a prostitute, poor, and/or a client of prostitutes. The reality, however, is very different. Although indeed prostitutes and their customers seem to have been the first to be affected by HIV, today the virus is spreading throughout the entire population, and is not confined to so-called “high risk” groups.

38Finally, because of Mumbai’s importance as a major economic centre, its population is composed of people of very heterogeneous origin. Migrants from the entire sub-continent, especially from the central and southern parts of the country, arrive in large numbers. When they return home, these workers may very well be carrying not only money but also the virus. Mumbai is interfacing not only with Maharashtra but also with the rest of the country and with the world, especially the countries of the Persian Gulf. The segregationist actions of the Maharashtrian and Indian Governments may give rise to very great tensions between the different communities living in the city, especially with the Muslims, who have often been accused of spreading diseases. Investigations in Mumbai and Hyderabad have shown that some Hindus consider Muslims to blame for propagating the virus in some parts of the sub-continent, by their sexual practices and “impure” way of life.


1 Carried out in winter 1995 and the summers 1995 and 1996. Statistical sources used in this paper also include : UNAIDS, 1998. Epidemiological Fact Sheet on HIV/AIDS. Geneva ; Bombay Municipal Corporation, 1986 and 1994 (unpublished), “Annual Reports of the Executive Health Officer.” ; NACO, 1993. Government of India, National AIDS Control Programme India : Country Scenario : an Update ; Ministry of Health and Family Welfare, 28/02/1998, Government of India, Monthly Update on HIV Infection in India ; Directorate of Health Services, Government of Maharashtra, ZBTCwise, Yearwise Blood Sample Screening and Positivity.

2 Researcher at the Tata Institute of Social Sciences, Mumbai.

3 In the case of Mumbai, municipal and district boundaries coincide.

4 In Mumbai, the hierarchy of administrative units, in decreasing order of size, is as follows: Wards, sections, circles, blocks.

5 Health Officer’s report 1892. See Ramasubban and Crook (1996).

Table des illustrations

Titre Figure 9.1 : Prevalence of HIV in India (1986/1998), by state and territory
Fichier image/jpeg, 53k
Titre Figure 9.2 : Prevalence of HIV in Maharashtra (1991/1995), by district
Fichier image/jpeg, 71k
Titre Figure 9.3 : Urban morphology and population : Central and South Mumbai
Fichier image/jpeg, 116k
Titre Figure 9.4 : Number of HIV positive detected in Mumbai by Wards (1988/1994)
Fichier image/jpeg, 36k
Titre Figure 9.5 : Main HIV testing centres, cases in Mumbai (1994)
Fichier image/jpeg, 27k
Titre Figure 9.6 : Diffusion of HIV in “pathogenic” wards
Fichier image/jpeg, 38k
Titre Figure 9.7 : Kamathipura : a red-light area in the heart of Mumbai
Fichier image/jpeg, 75k
Titre Figure 9.8 : HIV diffusion in Kamathipura (by lane)
Fichier image/jpeg, 59k
Titre Figure 9.9 : The spatial dynamic linked with HIV in Kamathipura
Fichier image/jpeg, 55k


Universite de Rouen, France.

© Institut Français de Pondichéry, 2000

Licence OpenEdition Books

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search