Rapid spread of HIV among injecting drug users in north-eastern states of India


Prevalence of intravenous drug users
Description of intravenous drug users
Risk behaviours
Trend of seroprevalence of HIV amongintravenous drug users
Spread of HIV among sexual partners of intravenous drug users
Role of awareness vis-a-visskills in harm minimization
Role of the legal system and the media


Author: S. SARKAR , N. DAS and S. PANDA , T. N. NAIK , K. SARKAR , B. C. SINGH , J. M. RALTE , S. M. AIER , S. P. TRIPATHY
Pages: 91 to 105
Creation Date: 1993/01/01

Rapid spread of HIV among injecting drug users in north-eastern states of India *

S. SARKAR ** Indian Council of Medical Research Unit for Research on AIDS in North-Eastern States of India, Salt Lake, Sector II, Calcutta, India
N. DAS and S. PANDA Field Unit, Indian Council of Medical Research Unit for Research on AIDS in North-Eastern States of India, Lamphelpat, Manipur, India
T. N. NAIK National Institute of Cholera and Enteric Diseases, Calcutta, India
K. SARKAR Indian Council of Medical Research Unit for Research on AIDS in North - Eastern States of India, Salt Lake, Sector 11, Calcutta, India
B. C. SINGH Department of Microbiology, Regional Medical College, Imphal, Manipur, India
J. M. RALTE Department of Psychiatry, Civil Hospital, Aizwal, Mizoram, India
S. M. AIER Directorate of Health Services, Kohima, Nagaland, India
S. P. TRIPATHY Indian Council of Medical Research, Ansari Nagar, New Delhi


Manipur, a north-eastern state of India bordering Myanmar, has experienced very rapid transmission of the human immunodeficiency virus (HIV) among its vast drug-injecting population. Seroprevalence among intravenous drug users increased from 0 per cent in September 1989 to 50 per cent within six months. With a minimum injecting population of 15,000 and seropositivity of over 50 per cent, the infection quickly spread to the population at large. One per cent of antenatal mothers tested seropositive by 1991. Forming part of the area of South-East Asia known as the Golden Triangle, and producing opium and its derivatives, Myanmar shares a long international border with four States of the region, and populations with a common language and culture move freely across borders. Two other north-eastern states of India bordering Myanmar have faced a similar epidemic within a short period of time. As a result of serosurveillance for HIV since 1986, the epidemic could be detected at an early stage. The present paper provides an account of the results of ongoing comprehensive studies conducted in the north-eastern states of India on drug -related HIV infection, already a serious problem, but possibly still restricted to that region of the country. The prevalence of intravenous drug users, their HIV serological status, the demographic profile, risk behaviour, the spread of the infection to other groups and the problems of harm minimization are also covered.

*The present paper is dedicated to the memory of S. C. Pal, founder-director of the National Institute of Cholera and Enteric Diseases (NICED) and the Indian Council of Medical Research (ICMR) Unit for Research on AIDS in North-Eastern States of India, who initiated the study.

**The valuable contributions of Jo Kittlesen, consultant to the World Health Organization, G. B. Nair, Senior Research Officer, NICED, and Mita Mukerjee, Research Associate, ICMR United for Research on AIDS in North-Eastern States of India, in the preparation of the present paper are gratefully acknowledged.


Injecting drug use for non-medical purposes increased rapidly during the last decade, and was a major contributing factor in the outbreak of the HIV epidemic in several parts of the world, including Asia. As a consequence of large production of opium and its derivatives and of the dislocations caused by the Viet Nam war, South and South -East Asia have experienced a dramatic rise in the use of injectable drugs, facilitating very high and efficient HIV transmission.

India has been categorized as having a Pattern III spread for HIV, where the major mode of transmission was thought to be heterosexual intercourse. The first HIV-seropositive individual in India was identified in 1986 among the commercial sex workers of the city of Madras, following which serosurveillance was initiated in all the states of India [ 1] . The current HIV prevalence rate among the general population in India is presumably less than 0.1 per cent, and that of the high-risk group of sex workers varies between 1 and 30 per cent in different cities [ 1] , [ 2] . States of India bordering Myanmar, such as Manipur, Mizoram and Nagaland (see figure I) account for less than 1 per cent of the total population of the country, but contribute nearly 16 per cent of the total number of persons testing seropositive for HIV -1 in India, according to the Ministry of Health. That is mainly because of the high prevalence of injectors in those states, leading to a sharp rise in HIV among them. Reports of HIV - seropositive intravenous drug users are still almost totally restricted to the north -eastern region. Less than 10 cases have so far been reported in the rest of the country [ 3] . The authors have conducted comprehensive studies on the prevalence of intravenous drug users, their risk behaviours and trends in seroprevalence over the last five years, and have collected the baseline data required for community-based control programmes in the region. The results of the studies are covered in the present paper. The spread of infection from intravenous drug users to their sexual partners and the general population is also documented.

Figure I. North-eastern states of India

Full size image: 100 kB, Figure I. North-eastern states of India

Prevalence of intravenous drug users

Data obtained from interviews of key informants show that injectors are mainly concentrated in the cities and peri- urban areas [ 4] . Secondary data sources such as detoxification centres, the Excise Department, prisons and various treatment agencies also reveal a similar distribution of prevalence of the injecting population. Community -based studies have been carried out in urban areas of the States of Manipur, Mizoram and Nagaland [ 4-6] . Snowballing has been done to estimate the prevalence of injectors in selected localities of urban areas with known population denominators. Such anonymous surveys show that the prevalence of intravenous drug users varies between 1 and 2 per cent of the general population in the states concerned. A conservative estimate is arrived at when the same rate is extrapolated by a multiplier technique on a similar population of urban areas in those states. Thus, Manipur accounts for at least 15,000 intravenous drug users in the entire state, Nagaland 1,500 and Mizoram 2,800 in surveyed urban areas [ 4-6] . In another study, the prevalence of intravenous drug users has been found to correlate well with the path of national highway number 39, which originates directly from a town bordering Myanmar and cuts across the urban areas of Manipur to reach Kohima, the capital of Nagaland [ 7] . The prevalence of intravenous drug users has been found to be 0.2 per cent in remote areas and 0.9 per cent in areas less well connected, and 1.3 per cent in areas well connected, to national highway number 39. Urban areas in districts having international borders with Myanmar and without a national highway connection have also reported a high injecting addict population. Thus, the availability of relatively cheap heroin (costing 15 to 20 United States dollars per gram) has been found to be the most important correlate for heroin injecting behaviours. A reduction in the supply of heroin through active vigilance by the Excise Department and the enforcement of stringent legal measures on the drug user population in at least one statehas resulted in a threefold to fourfold increase in the price of heroin, bringing an eventual decline in the absolute number of exclusive heroin users, who have switched to synthetic opiates. Thus, overall risk behaviour has not changed [ 6] . However, the role of subcultures, economic conditions and specific occupations in influencing prevalence in different parts of a city and in rural and urban areas has not been properly ascertained.

Description of intravenous drug users

The pattern of intravenous drug use in the north-eastern states of India is more or less the same as that of Bangkok and Myanmar [ 8] , [ 9] . The injectors are mostly male and between 15 and 30 years of age. In the community-based studies [ 4-6] , males were found to have accounted for 95 per cent of the total injectors. Injecting females were mainly commercial sex workers, as found in studies conducted during 1990 in Manipur. Recent studies of groups of injectors at Manipur suggest that female injectors are increasing. However, the number of female injectors who reported for self -testing for. HIV from 1990 to 1992 has remained stable, accounting for 5 to 8 per cent of intravenous drug users. The occupational distribution of injectors reveals that 53 per cent of them were unemployed, including 34 per cent of students who received economic support from their families. Until recently, intravenous drug users have been staying in their families, mainly with parents who try to cope with the social and economic pressures thus created within the home. It remains to be seen whether that structure will disintegrate, and whether there is a possibility of involving the families of intravenous drug users in individual or socially targeted interventions and large-scale reduction of risk behaviours. The educational status of the addict population has been favourable to communication through the written media, as most of them have had a secondary level of education. That is an important factor in states such as Nagaland and Manipur, where illiteracy is high (40 to 60 per cent) among the general population. Data from Mizoram, where literacy is very high, reveal, however, that neither high literacy nor religion have had a protective effect in relation to the incidence of drug abuse. Unlike Manipur, the majority of the population in Mizoram is Christian.

Risk behaviours

The data on risk behaviours have been collected through anonymous interviews of street addicts and those attending various treatment centres and prisons [ 4-6] . The practice of injecting apparently started in 1983 in Manipur and Nagaland, where the main substance was pethidine and morphine. Heroin as the substance of choice became popular in 1985, and the usual frequency of administration was three or four times a day.

Figure 11. Improvised syringe made of ink-dropper fitted with needle

Full size image: 13 kB, Figure 11. Improvised syringe made of ink-dropper fitted with needle

The injectors are overwhelmingly intravenous. In areas with a high price of heroin, injectors have switched to dextropropoxyphen and other synthetic analgesics, with a similar frequency of intravenous injection per day. The use of the improvised "ink dropper" (figure II) as a syringe (with a possibly greater amount of blood left out) has not been substantiated as a risk factor for higher seroprevalence in Manipur, as it was observed that the seroprevalence among injectors in Nagaland who used plastic syringes was also over 50 per cent, as in Manipur. The average number of regular needle-sharing partners in those groups has remained relatively stable, varying between 3 and 5. In each state, the injectors are reported as sharing their injecting equipment regularly (in 83 to 96 per cent of cases), mainly with partners of the same sex, with rates varying between 66 and 90 per cent. The cleaning of needles is not practiced at all by 62 per cent of the injectors in Manipur, and 77 per cent in Mizoram. Methods used for cleaning needles were washing with ordinary water and, rarely, boiling water. Injectors were unaware of bleach as a disinfectant until mid- 1991, and the product is not commonly available on the market as a household item. Of the injectors in each state, 50 to 70 per cent had had sexual experiences. Nearly three fourths of the intravenous drug users in Manipur had experienced decreased libido, 4 per cent had increased sexual desire, and the rest were indifferent to sexual activity. The median number of heterosexual partners was three during the past five years in Manipur, where the drug- user population was the least active sexually among the three states concerned. Condoms are generally unpopular among the population at large, and only a small proportion of injectors (3 to 5 per cent) have reported even occasional use of them. The median age for the onset of injecting drug use is early (15 years). The most important reasons cited by 60 to 90 per cent of injectors in the three states indicates that peer pressure and the influence of friends were major factors in initiating drug abuse.

Trend of seroprevalence of HIV among intravenous drug users

Manipur has been a classic example of a sharp rise in seroprevalence among injectors within a short span of time. Serosurveillance started in Manipur in 1986, when a total of 2,322 persons, including 707 intravenous drug users, were screened, and the first seropositive intravenous drug user was detected in October 1989 [ 10] , [ 11] . The number of injectors screened between 1986 and 1989 are shown in table 1.

Table 1. Result of screening for HIV among intravenous drug users in Manipur between 1986 and 1992


Number screened

Number testing positive

90 percent confidence interval (percentage)

1986 128 0
1987 165 0
1988 207 0
1989 245 21 (8.6)
1990 1762 1050 (59.6
1991 447 244 (54.6)
1992 a/
250 168 (67.2)

Note: Numbers in parentheses are percentages.

a/Up to September 1992.

Since the first HIV- positive intravenous drug user was detected, the quarterly prevalence has increased sharply from 0 per cent (true prevalence not known) to 56 per cent within two consecutive quarters, and appears to have stabilized around 55 per cent (figure III) in 1990/91. The estimates for seroprevalence with a 90 per cent confidence interval show that the lowest figure (52.4 per cent, with a confidence interval of 44.3 to 60.6) in the second quarter of 1991 did not vary significantly from that of the highest figure recorded in the third quarter of 1991 (53.8 to 74.1). The percentage of seroprevalence in 1992 again shows an upward trend, with the highest prevalence for 1992 (73.1 per cent, with a confidence interval of 62.9-83.2) still comparable to the peak recorded in the third quarter of 1991, and significantly different from the lowest figure from the second quarter of 1991. Similar data with a quarterly breakdown are not available from other states. However, data from Kohima, the capital of Nagaland, will possibly yield similar curves, as the seroprevalence figure for the two consecutive years in that district has remained apparently stable around 50 per cent [ 6] , [ 12] . In a study conducted in 1990, HIV seroprevalence among intravenous injectors of synthetic analgesics in Mizoram was found to be 8 per cent. The movement of population along the border between Mizoram and Myanmar appears to he less than that between Manipur and Myanmar. It is still difficult to explain why the rate (as observed in 1990 and 1991) has stabilized at between 6 and 10 per cent in Mizoram, whereas it has increased sharply from less than 1 per cent to 16 per cent, and finally to more than 64 per cent, in three consecutive quarters in Manipur. The reason why the HIV infection plateau is at a certain prevalence in one area and a higher prevalence in another is not known [ 13] . Whether the seroprevalence of synthetic-analgesics users with similar risk behaviours stabilizes at a low level remains to be seen. For further insight into the dynamics of transmission of HIV infection, there is a need for data, often very difficult to collect, on the incidence of HIV among intravenous drug users. Otherwise, an investigation and comparative analysis of the prevalence of HIV among new injectors could be carried out. The peak of HIV infection in 1989/90 in the north-eastern states of India follows that of Thailand in 1987/88 and Myanmar in 1988/89 [ 9] , [ 14] . There is other epidemiological evidence to suggest that the infection spread from Thailand to Myanmar and India. Recent studies conducted at the National Institute of Cholera and Enteric Diseases on the molecular epidemiology of the prevalent HIV strains might throw more light on the above issue.

Figure III. Quarterly prevalence of HIV among intravenous drug users in Manipur, October 1989-September 1992

Full size image: 11 kB, Figure III. Quarterly prevalence of HIV among intravenous drug users in Manipur, October 1989-September 1992

Spread of HIV among sexual partners of intravenous drug users

The VDRL-positivity (2 to 3 per cent) among injectors has not varied significantly between HIV - positive and HIV - negative intravenous drug users [ 15] , which indicates risk-taking sexual behaviour of HIV-positive injectors. From 10 to 16 per cent of the addicts have a history of exposure to commercial sex workers. As a result, HIV infection should spread from intravenous drug users to their sexual partners and the population at large. The prevalence of seropositivity among sexual partners of injectors and among antenatal mothers attending hospital out-patient departments is shown in table 2. It can be seen that within one year of high seroprevalence (around 50 per cent) among intravenous drug users the infection could significantly reach the general population, as I per cent of antenatal mothers have already tested positive for HIV in an unlinked anonymous survey conducted in 1991, indicating a very rapid spread. However, the spouses of the HIV- positive intravenous drug users represent a self -selected group that had voluntarily reported for testing, and the detailed sampling characteristics are not known.

Table 2. Sexual transmission of HIV in Manipur


Number tested

HIV prevalence (percentage)

90 percent confidence interval (percentage)

Period of data collection

Spouse of HIV-positive intra-venous drug user
68 5.9
Antenatal mother
200 2

Role of awareness vis-a-vis skills in harm minimization

A study [ 4] of awareness among the 450 injectors in Manipur has shown that knowledge of the mode of transmission of HIV was higher among the addict population than among college students who had a higher level of formal education. More recent data from Manipur [ 16] and Nagaland reveal that the level of information on the mode of transmission of AIDS is also high among the general population. However, a negligible portion of the population has experienced contact with clinically overt AIDS patients, which could be an important reason for unchanged practices of risk behaviour in spite of high awareness. The stage of the epidemic in the region is quite early, and therefore the possibility of an encounter with a friend who had AIDS was obviously less in 1990 and 1991. Data from ongoing studies in 1992, when analysed, would therefore yield useful clues to a better understanding of the situation. The major source of awareness among the injecting population is through discussions among peers, which suggests an important potential role of peer pressure in prevention of high-risk behaviour among the concerned group. There are more than 33 tribes and subtribes with different dialects in Manipur, which makes the preparation of information, education and communication (IEC) materials a difficult task. The role of counselling and testing in changing high-risk behaviours of injectors is yet to be seen, with the efforts of counselling still at a very rudimentary stage. A recent analysis from the longitudinal data on knowledge, attitudes and practices(KAP) of nearly 500 injectors (with a seroprevalence rate of more than 50 per cent) in Manipur showed that only a small percentage received any kind of counselling, and a repeat survey of nearly 100 injectors revealed no change in behaviour of those who had been counselled. Given the lack of a social support system and the extremely meagre detoxification facilities, counselling is not expected to be of practical significance. Until the huge requirement for counsellors and trained personnel is met, counselling cannot be viewed as an effective tool for "individual targeted intervention" in the near future. A global review [ 17] of the effects of counselling and testing on the population of drug addicts up to mid-1990 has yielded inconclusive results. The findings of a study on discordant couples [ 18] showing a change of behaviour through testing and counselling may not be applicable to intravenous drug users, a group in which dynamics and social norms play a more important role in providing motivation for reducing risk behaviours.

The World Health Organization and other international bodies are planning to support socially targeted interventions such as "peer outreach activities" in the region. Initial group interviews of addicts suggest that washing injecting equipment with ordinary cold water could be more acceptable to injectors than the use of bleach [ 19] . There is a need for sufficient scientific evidence to promote the former as a sound and effective strategy for intervention. Syringe and needle exchange programmes or methadone clinics are not encouraged by current government policy. Moreover, in the current immunization programme and usual hospital practices, insufficient funds are provided for disposable syringes and needles, therefore the social and political barriers against making them available for other programmes are obvious.

Role of the legal system and the media

At present, there are almost no detoxification facilities or any kind of organized support service for intravenous drug users in the north- eastern states of India. In a state such as Manipur, with an injecting addict population of at least 15,000, only a very limited number of beds are available for detoxification facilities. As a result, prisons have been used as detoxification services to which injectors are sent by their parents to keep them off drugs. In a study carried out in 1990, more than 50 per cent of the addicts had been in prison at least once in their lives, and at least half of them had been sent by family members. However, imprisonment has induced no change in risk behaviour among addicts, since abstinence from drugs was the only message given in prison, and no counselling support was available. Of the 450 injectors interviewed, only 2 per cent wished to stop their addiction because of their imprisonment.

The legal system of the country bans the free procurement of injecting equipment, and a needle and syringe exchange programme has not been supported by the federal Government. The Narcotic Drug and Psychotropic Substance Abuse Act does not effectively differentiate between pedlar, trafficker and abuser. The law could thus be easily manipulated to imprison any self-injector otherwise not involved in peddling drugs [ 20] .

The overall message so far conveyed by the media has been designed to encourage abstinence and to provoke fear through the use of frightening images. A survey of the tattoo marks of addicts revealed that the majority (75 per cent) of injectors would prefer the snake, skull or scorpion as their favourite tattoo, with the scare tactics of the media apparently having had no effect on them. The messages do not include any encouragement of harm minimization, for which the social sanction appears doubtful. Recent newspaper reports have suggested that addicts have been killed by certain extremist groups that have also been successful in forcing the Government to close down all liquor shops in at least one state [ 20] . Whether such an effort will in the long run succeed in reducing supply and demand of the substances abused remains to be seen. However, in the absence of any other form of social support, unsafe injecting practices will certainly increase as a result of such activities. Peer outreach projects that could otherwise play a significant role among the addict population can only be implemented successfully if the social milieu created by the police, law and local pressure groups are congenial to such programmes.


The rapid spread of HIV among the intravenous drug users in Manipur is a clear example of the inevitable spread of the HIV epidemic if high-risk behaviour is prevalent. Faced with other pressing public health issues, poor countries such as India, Myanmar and Thailand failed to give due priority to HIV before the epidemic struck them. Despite early detection of a high prevalence of HIV among injectors, its quick spread to heterosexual partners could not be prevented. The huge demand for treatment support, service and counselling is unlikely to be met in the near future. The harm reduction approach as a strategy is possibly the only powerful and cost-effective tool in such a situation, but it has received scant regard in campaigns launched by the states or by pressure groups within the population. Of immense importance to such a programme, and a precondition for its initiation, is the creation of the necessary political will.

In spite of the limited availability of data on HIV among the injecting population in other parts of India, there is a possibility that its prevalence is low. However, there has recently been a phenomenal increase (5 to 13 times in the last two years) in injecting behaviours among the population of drug addicts in cities like Calcutta, where 15 to 46 per cent of substance abusers (mainly heroin smokers) have reportedly switched to injecting behaviours, from less than I per cent in 1986 [ 21] , [ 22] . Similar reports have also come in from Madras, another metropolitan city in southern India. The experience of Australia suggests that the prevalence of HIV could possibly stabilize at a low endemicity (less than 5 per cent) if it does not reach a critical limit, or if a good social support system, resources and skills in inducing behavioural changes among intravenous drug users [ 23] are available. Political and social support must be urgently mobilized for socially targeted intervention in all areas with an injecting population.

Drug abuse in general and injecting drug use in particular have always provoked conflict between public health principles and existing sociopolitical structures regarding methods of control. The rapid emergence of HIV and its consequences has sharpened the debate. It is often believed that prevention of transmission of HIV through injecting drug use could remain the most difficult goal to achieve among all possible modes of the spread of the infection. Although that view has not been substantiated by the encouraging figures from Amsterdam, San Francisco and Australia, it might unfortunately be valid for the north- eastern region of India and in neighbouring countries.



Indian Council of Medical Research, "Current dimensions and future implications", ICMR Bulletin, vol. 22, Nos. 11 and 12 (November-December 1992).


A. K. Chakraborty, "A community-based survey of STD/HIV", paper presented at the Second International Congress on AIDS, Asia and the Pacific, held at New Delhi from 8 to 12 November 1992.


National AIDS Control Organization, India, personal communication.


S. Sarkar and others, "Descriptive epidemiology of intravenous drug users - a new risk group for transmission of HIV in India", Journal of Infection, vol. 23, No. 2 (1991), pp. 201-207.


J. M. Ralte, S. Sarkar and S. Panda, "Drug addiction and HIV infection in Mizoram", paper presented at the Second International Congress on AIDS, Asia and the Pacific, held at New Delhi from 8 to 12 November 1992.


S. C. Pal and others, "Explosive epidemic of HIV infection in north-eastern states of India, Manipur and Nagaland", Centre of AIDS Research Calling, vol. 3, 1990, pp. 3-6.


K. Sarkar and others, "Relationship of national highway with intravenous drug abuse of heroin and HIV infection in Manipur", paper presented at the Second International Congress on AIDS, Asia and the Pacific, held at New Delhi from 8 to 12 November 1992.


D. C. Des Jarlais and others, "Risk reduction and stabilization of HIV seroprevalence among drug injectors in New York City and Bangkok, Thailand", in Science Challenging AIDS, G. B. Rossi and others, eds. (Basel, Karger, 1992), pp. 207-213.


Myint Htoo, Drug Dependency Treatment and Research Unit, Psychiatric Hospital, Yangon, personal communication.


T. N. Naik and others, "Intravenous drug users - new risk group for HIV infection in India", AIDS, vol. 5, No. 1 (1991).


S. Sarkar and others, "Rapid spread of HIV among injecting drug users of north-cast India: a new risk group for transmission", paper presented at the Eighth International Conference on AIDS, held at Amsterdam from 19 to 24 July 1992.


National Institute of Cholera and Enteric Diseases, Annual Report (Calcutta, 1991), pp. 27-30.


S. R. Friedman and D. C. Des Jarlais, "HIV among drug injectors: the epidemic and the response", AIDS Care, vol. 3, No. 3 (1991), pp.239-249.


The International Working Group on AIDS and Drug Use, Newsletter, vol. 5, Nos. 2-4 (May 1991), p. 15.


Unpublished data of the National Institute of Cholera and Enteric Diseases, Calcutta.


P. N. Sehgal and K. Singh, Knowledge, Attitude, Beliefs and Practices (KABP) Study related to AIDS in Manipur State, India and Intervention Strategies (New Delhi, Voluntary Health Association of India, 1992).


D. L. Higgins and others, "Evidence for the effects of HIV antibody counselling and testing on risk behaviors", Journal of the American Medical Association, vol. 266, No. 17 (1991), p. 2413.


S. Allen and others, "Effect of serotesting with counselling on condom use and seroconversion among HIV discordant couples in Africa", British Medical Journal, vol. 304, 20 June 1992, p. 1605.


Unpublished data of the Indian Council of Medical Research.


AIDS Anti-Discriminatory Movement, "The sugar is bitter" (New Delhi, May 1992).


A. Basu, "A case for harm reduction", poster presented at the Second International Congress on AIDS, Asia and the Pacific, held at New Delhi from 8 to 12 November 1992.


Indian Chamber of Commerce, "I.V. use on the rise" Combat International, vol. 1, No. 2 (June 1992), p. 13.


J. Kaldor, "Epidemiological pattern of HIV infection in Australia", Journal of Acquired Immune Deficiency Syndrome, vol. 6 (Supplement 1), 1993, pp. S1-S4.