Genetic influences on the immunologic pathogenesis of encephalomyocarditis (EMC) virus-induced diabetes mellitus.

DBA/2 and Balb/cBY mice were infected with approximately 30 plaque-forming units of the M-variant of encephalomyocarditis (EMC-M) virus. Seven days after inoculation the majority of the animals of both strains were hyperglycemic. A significant correlation between increased concentrations of virus in the pancreas and hyperglycemia was found among individual DBA/2 animals, but not among Balb/cBY mice. T-lymphocyte depletion of DBA/2 mice before infection failed to alter the incidence or severity of hyperglycemia in comparison to intact animals. Conversely, hyperglycemia in T-lymphocyte-depleted Balb/cBY mice was reduced substantially in comparison to infected immunocompetent animals. There appears to be at least two genetically influenced pathogenic mechanisms of diabetes in EMC-M virus-infected mice. In some strains of animals, hyperglycemia results exclusively from viral infection and the consequent injury to the beta cells, whereas in other animals, viral damage to the islets is compounded by immunologic events.

The epidemiology of AIDS in the Vellore region, southern India.

OBJECTIVE: To describe the epidemiology of patients with AIDS in Vellore region, Southern India. DESIGN AND METHODS: Sixty-one patients with AIDS were diagnosed and treated between July 1987 and June 1992. Information on their demographic characteristics and probable modes of acquiring HIV infection was collected at interviews with them and their spouses. RESULTS: There was a progressive increase in the number of patients seen over the 5 years. The mean ages of the 51 men and the 10 women were 33 and 29 years, respectively. Of the 44 patients from our district (population, 5 million), 28 were from Vellore town and 10 from rural areas. Forty-seven (92%) men had frequently used prostitutes. Of the women, four were prostitutes, one had had multiple sex partners and five had not had extramarital sexual contact. One man and one woman had no other risk factor except blood transfusion. Thirty-one (51%) patients had died by August 1992. CONCLUSION: The AIDS epidemic in this region is in its early ascending phase, with a doubling time of approximately 1 year. Most men with AIDS were infected by heterosexual contact with prostitutes, while some women were prostitutes themselves. Together with the male-to-female ratio of 5:1, these results suggest that the male population at risk has sex with a much smaller population of female prostitutes, constituting the major chain of transmission. HIV infection is occurring in both urban and rural populations.

PIP: Between July 1987 and June 1992 physicians diagnosed AIDS in 61 patients at the Christian Medical College Hospital in Vellore, India. They noted that there were 200 AIDS cases reported in the same period in the region, but they extrapolated AIDS cases to be 7656. Yet even the figure of 7656 was likely to be an underestimate, because this hospital does not treat most patients in the district or in the town. The number of cases at this hospital doubled each year. The mean age of the 51 men was 33 years, and 29 years for the 10 women (male-to-female ratio - 5:1). 23% of cases were from rural areas. Most cases (71% for men and 80% for women) were married and lived with their spouses. 47 (92%) of the men had had sexual intercourse with prostitutes, suggesting that this was the leading means of HIV transmission. 4 married women were prostitutes and their husbands knew and/or encouraged them to work as prostitutes. 4 other married women had had no extramarital sexual affairs and were infected by their HIV positive husbands who had had sexual intercourse with prostitutes. The only risk factor for 2 patients was blood transfusion. The most common signs and symptoms of AIDS included considerable weight loss (62%), fever for more than 30 days (56%), tuberculosis (52%), oral candidiasis (41%), and chronic diarrhea (31%). By August 1992, 31 (51%) patients had died.

The initial detection of human immunodeficiency virus 1 and its subsequent spread in prostitutes in Tamil Nadu, India.

To determine the prevalence of HIV-1 antibody and risk factors associated with a positive test in a heterosexually promiscuous female population, we initially screened 412 prostitutes in remand homes in three cites and three towns in Tamil Nadu state (southern India) and then tested all new entrants to one home in Vellore from 1986 to 1990. The proportion of women infected (10 of 102) from the port city of Madras was greater than from all other cities or towns combined (four of 310, p = 0.0002). The only significant risk factor for development of HIV-1 antibody was exposure to foreigners (odds ratio: 7.71; 95% confidence interval 4.2-11.2; p = 0.0004), after correcting for the influence of city. In Vellore the prevalence of HIV-1 antibody increased from 1.8% in 1986 to 28.6% in 1990, with a doubling time of 0.95 years. We hypothesize that HIV-1 infection has been introduced into India relatively recently into the heterosexually promiscuous population, where there has been some spread. Preventive measures should be urgently directed toward this population to prevent spread in the general population.

Clinical and laboratory profile of AIDS in India.

The clinical features and results of laboratory investigations of the first 19 Indian patients with AIDS seen in our hospital are presented. Weight loss, fever, and diarrhea were the most common symptoms. Tuberculosis (TB) was the most common secondary infectious disease; among 13 patients, seven had only pulmonary TB, five had pulmonary and extrapulmonary TB, and one had only extrapulmonary TB. Oropharyngeal candidiasis was found in 11 patients. Other secondary infections were predominantly by virulent bacteria. Opportunistic infections other than candidiasis were infrequent; one patient had cryptococcosis, two had symptomatic cryptosporidiosis, one had noncoagulase-positive staphylococcus septicemia, and one had cytomegalovirus retinitis. Reduced lymphocyte counts (particularly of the CD4 subset), anemia, hypoalbuminemia, hyperglobulinemia, and elevated liver enzyme levels were frequent laboratory findings. Six patients are under follow-up, two are lost to follow-up, and 11 have died. Lymphocyte counts less than 500/mm3 were only seen in those patients who subsequently died. Response to antituberculosis therapy was good in several patients. Thus, the clinical profile of Indian patients with AIDS is not different from the common picture of patients of low socioeconomic and poor hygienic standards; patients presented with TB, undernutrition, and multiple infections. Therefore, a large population of patients with AIDS in India will not be recognized unless they are tested for evidence of HIV infection.

Detection of HIV type 1 env subtypes A, B, C, and E in Asia using dried blood spots: a new surveillance tool for molecular epidemiology.

Global surveillance of HIV-1 subtypes for genetic characterization is hampered by the biohazard of processing and the difficulties of shipping whole blood or cells from many developing country regions. We developed a technique for the direct automated sequencing of viral DNA from dried blood spot (DBS) specimens collected on absorbent paper, which can be mailed unrefrigerated in sturdy paper envelopes with low biohazard risk. DBS were collected nonrandomly from HIV-1-infected, mostly asymptomatic, patients in five Asian countries in 1991, and shipped via airmail or hand carried without refrigeration to Bangkok, and then transshipped to North America for processing. After more than 2 years of storage, including 6 months at ambient temperatures, proviral DNA in the DBS was amplified by nested PCR, and a 389-nucleotide segment of the C2-V3 env gene region was sequenced, from which 287 base pairs were aligned and subtyped by phylogenetic analysis with neighbor-joining and other methods. From southern India, there were 25 infections with subtype C and 2 with subtype A. From Myanmar (Burma), we identified the first subtype E infection, as well as six subtype BB, a distinct cluster within subtype B that was first discovered in Thailand and that has now appeared in China, Malaysia, and Japan. From southwest China, one BB was identified, while a "classical" B typical of North American and European strains was found in Indonesia. From Thailand, five DBS of ambiguous serotype were identified as three B, one BB, and one E. A blinded control serotype E specimen was correctly identified, but a serotype BB control was not tested. Most HIV-1 in southern India appears to be env subtype C, with rare A, as others have reported in western and northern India. The subtypes BB and E in Myanmar, and the BB in China, suggest epidemiological linkage with these subtypes in neighboring Thailand. DBS are a practical, economical technique for conducting large-scale molecular epidemiological surveillance to track the global distribution and spread of HIV-1 variants.

PIP: The global surveillance of HIV-1 subtypes for genetic characterization is frustrated by the danger of processing and the difficulties of shipping whole blood or cells from many developing country regions. The authors therefore developed a technique for the direct automated sequencing of viral DNA from dried blood spot (DBS) specimens collected on absorbent paper. Such specimens can be mailed unrefrigerated in paper envelopes with low biohazard risk. 51 DBS specimens were collected nonrandomly from HIV-1-infected, mostly asymptomatic individuals in India, Myanmar, China, Indonesia, and Thailand in 1991, then shipped via airmail or hand carried without refrigeration to Bangkok from where they were forwarded to North America for processing. After more than 2 years in storage, including 6 months at ambient temperatures, proviral DNA in 42 of the DBS was successfully amplified by nested polymerase chain reaction, and a 389-nucleotide segment of the C2-V3 env gene region was sequenced, from which 287 base pairs were aligned and subtyped by phylogenetic analysis with neighbor-joining and other methods. From southern India, there were 25 infections with subtype C and two with subtype A; the first subtype E infection was identified from Myanmar, as well as six subtype B(B); one B(B) was identified from southwest China; subtype B was identified from Indonesia; and five DBS of ambiguous serotype classified as three B, one B(B), and one E were identified from Thailand. DBS can be used as a practical, cost-effective way of tracking the global distribution and spread of HIV-1 variants.

Three manual immunoassays for the rapid detection of hepatitis B surface antigen.

We have evaluated 3 commercial test kits for the detection of hepatitis B surface antigen (HBsAg) which do not require elaborate instrumentation and are relatively inexpensive. These included the Immunocomb-II (Orgenics, Israel), Austragen latex agglutination (Span, India) and Daina Screen (Abbott, USA) tests. A panel of 98 sera tested by IMx HBsAg assay (Abbott, USA) were used to evaluate the tests. Sensitivities of 96.6%, 86.2% and 96.6%, and specificities of 95.7%, 94.2% and 100%, respectively, were obtained. We therefore recommend the Immunocomb-II or Daina Screen tests for HBsAg detection in small laboratories where facilities to conduct automated enzyme-linked immunosorbent assays are not available.

Comparison of the indirect immunofluorescence assay with western blot for the detection of HIV antibody.

With the increase in the prevalence of human immunodeficiency virus (HIV) infection in India, there is an urgent need to explore cost-containing methods of HIV antibody detection. We compared the indirect immunofluorescence assay (IFA) with the widely used but expensive confirmatory test, the Western blot (WBT). A panel of sera comprising 42 WB positive, 46 antibody negative (27 ELISA negative and 19 ELISA reactive but WB negative) and 58 sera with indeterminate WB profiles were tested by a commercial IFA kit. The sensitivity of IFA was found to be 97.6 per cent and specificity 97.8 per cent. Of the indeterminate sera, 18 (31%) were positive in IFA while 26 (45%) were negative and 14 (24%) were equivocal. In general, the IFA positivity was associated with the presence of the gag and env antibodies of HIV while negativity associated with the presence of lone antibodies. The cost of testing by IFA is less than 20 per cent of that of WB testing. Thus IFA appears to be a cost-containing alternative to WB in centres in India where immunofluorescent microscopy is routinely done.

Prevalence of human immunodeficiency virus infection among voluntary blood donors.

HIV infection was found among voluntary blood donors at the Christian Medical College Hospital, Vellore (south India), the prevalence rate of infection among these donors being 1.63 per 1000. Based on these findings, the authors recommend that appropriate steps be taken immediately by all blood transfusion centres in India to evaluate the prevalence of HIV infection among blood donors and to prevent the transmission of the infection through blood transfusion.

The detection of HIV-2 infection in southern India.

To determine the prevalence of HIV-2 infection in southern India, we tested two sets of sera, selected from among the samples which had been collected between January 1988 and October 1990 from high risk subjects and tested for HIV-1 antibody. They were screened for HIV-2 antibody by ELISA and repeatedly reactive sera confirmed by HIV-2 Western blot and line immunoassay. In the first set of 604 sera, only one (0.16%) was positive for HIV-2. In the second set of 24 sera, selected on the basis of having indeterminate HIV-1 Western blot profiles, again one (4%) was positive for HIV-2. The two HIV-2 infected subjects were residents of Madras or Visakhapatnam. Residents of Vellore region constituted 88 and 75 per cent of the two sets of subjects; none was positive for HIV-2. Our results show the prevalence of HIV-2 in the port-cities of southern India. Since it will spread to other regions continuous monitoring for this infection is essential in order to determine when to establish HIV-2 screening in addition to the existing HIV-1 screening of donor blood for transfusion.

PIP: While HIV-1 is the principal etiologic agent of AIDS throughout the world, the genetically distinct HIV-2 has been recently credited with causing AIDS especially in the countries of West Africa. HIV-2 infection was first reported in India in the city of Bombay in 1991. This paper reports findings from an HIV-2 seroprevalence study of blood sera in southern India. Sera sampled were selected from among those collected from January 1988 to October 1990 from high-risk subjects and tested for antibody to HIV-1. Sera were screened for HIV-2 with ELISA and repeatedly reactive sera were confirmed with Western blot for HIV-2 and line immunoassay. 1 serum (0.16%) of the 604 sera initially tested was positive for HIV-2. Another serum was found positive in the 2nd set of 24 sera (4%). The 2 infected subjects were residents of Madras or Visakhapatnam. Although residents of the Vellore region respectively comprise 88% and 75% of the 2 sets of subjects, none was found to be positive for HIV-2. Since these findings confirm the existence of HIV-2 in India, the authors stress the need to monitor its spread in order to determine when blood from donors should start being screened for both HIV-1 and HIV-2.

Reduction of the cost of testing for antibody to human immunodeficiency virus, without losing sensitivity, by pooling sera.

The sensitivity of testing pooled sera instead of individual sera for antibody against human immunodeficiency virus (HIV) was evaluated using a non-competitive enzyme-linked immunosorbent assay (ELISA). For this purpose, 42 HIV antibody positive sera were titrated and introduced into 42 sets of pools of 2, 4, 8, 16, 32 or 64 sera in such a manner that each pool had one positive sample and the rest, HIV antibody negative sera. When the pools were tested in ELISA, all pools with high titred antibody positive sera were reactive irrespective of pool size, while some of the pools containing medium or low titred sera were non-reactive when pool size exceeded 16. Subsequently the pool size was limited to 16. When 208 previously unscreened samples were tested in 52 pools of 4, 26 pools of 8 or 13 pools of 16 sera, or individually, 6 antibody positive sera were correctly identified. Thus, it was found that the pooling method did not reduce the sensitivity of ELISA test, whereas the cost was reduced to less than half of that of individual testing.

Rising trend in the prevalence of HIV infection among blood donors.

The prevalence of HIV infection among blood donors was analysed, for the five year period from April 1988 to March 1993. All donors were patients' relatives or volunteers; no paid/commercial donors were accepted. Each year between 14,084 and 15,544 donor blood samples were screened by ELISA and those found reactive were tested by Western blot. Western blot positive samples were considered to be infected with HIV. The prevalence rates were 1.5 per 1000 (1988-89), 1.1 per 1000 (1989-90 and 1990-91) 1.9 per 1000 (1991-92) and 3.1 per 1000 (1992-93). When the mean prevalence rate over the first three years [1.3 per 1000 (1988-89 to 1990-91)] was compared to the prevalence in 1991-92 (1.9/1000), increase was statistically significant (P < 0.05). The prevalence rate in 1992-93 (3/1000) was significantly higher than that of the previous year (P < 0.01). These data suggest that the prevalence of HIV infection in blood donors is increasing and this could be a reflection of the rising prevalence of HIV infection in the general population.

HIV seropositivity among patients with sexually transmitted diseases in Vellore.

In a prospective, systematic investigation, 2215 new patients with sexually transmitted diseases (STD) were screened for human immuno deficiency virus (HIV) infection during April 1986 to December 1988. Among them 9 were positive for HIV antibody giving a rate of 4/1000. In addition, 5 patients had been referred to our clinic with a diagnosis of HIV infection and STD. Among these 14, there was one married couple. Of the remaining 5 married individuals, 4 were men; the wives of 3 of them were infected but asymptomatic, and were not included in this report. The rate of conjugal infection was 71 per cent. Most (57%) patients belonged to poor socio-economic background; 78 per cent of patients were young, in the age group 15-25 yr. The presenting STD were syphilis, venereal warts, gonorrhoea, chancroid, genital herpes, non-gonococcal urethritis and candidiasis. It appeared that the course of associated STD had not been altered in these HIV infected individuals.

Lack of evidence for seroconversion to human immunodeficiency virus in recipients of anti-D immunoglobulin containing HIV antibody.

Twenty nine Rh negative women who had received injections of anti-D (Rho) immunoglobulin preparations (from an Indian firm) contaminated with antibody to human immunodeficiency virus (HIV), were tested for seroconversion to HIV, 14-38 wk after injections. All 29 were negative for HIV antibody, indicating that infectious HIV or immunogenic concentrations of inactivated HIV were not present in the investigated immunoglobulin product.

High prevalence of HTLV antibodies in wild-caught bonnet monkeys in southern India.

The prevalence of human T-lymphotropic virus (HTLV) type-1 antibodies was determined in the bonnet monkeys, living naturally, within about 30 km radius of Vellore (south India). Sera from 157 animals, collected between January 1982 and May 1993 were screened for the presence of HTLV-I infection by a particle agglutination test (PAT). When sera repeatedly reactive in PAT were subjected to indirect immunofluorescence and western blot tests, 63 (40%) were confirmed to be positive for HTLV-1 antibody. These findings are significant in the light of recent reports that HTLV infection is endemic to southern India.

Diagnosis of typhoid fever by the detection of anti-LPS & anti-flagellin antibodies by ELISA.

In a developing country like ours where typhoid fever is endemic and there are very few microbiology laboratories to provide diagnosis by culture, a search for non culture techniques for rapid and reliable diagnosis continues. The Widal test has a low sensitivity. We have attempted to adapt the well established enzyme linked immunosorbent assay (ELISA) technique to provide a serological test of better sensitivity and specificity. The ELISA described here to detect anti-LPS antibodies was found to have a sensitivity of 89.4 per cent and a specificity of 94.9 per cent. The sensitivity for the antiflagellin ELISA was 68.1 per cent and specificity was 97.4 per cent. The likelihood ratio was 17.5 for anti-LPS ELISA and 26.2 for the anti-flagellin ELISA. Considering the positivity of either one ELISA as diagnostic, the sensitivity was 93.6 per cent and specificity was 94.9 per cent.

Importance of RT-PCR for detection of hepatitis C virus among different patient groups.

A pilot study was undertaken to evaluate a sensitive single step reverse transcription polymerase chain reaction (RT-PCR) to detect hepatitis C virus (HCV) RNA in plasma samples as an adjunct to antibody detection. In this study, digoxiginin labelling of the amplicon was carried out during the PCR and the product was detected by an ELISA format (DIG-ELISA) as well as ethidium bromide stained gel detection. DIG-ELISA detection of PCR product compared favourably with gel detection. We tested 42 individuals with either parenteral risk factors or chronic liver disease and found the RT-PCR enhanced the detection rate by 19 per cent. Considering either plasma RNA or antibody positivity, HCV infection was detected in 64.3 per cent of patients.

Increasing prevalence of HIV antibody among blood donors monitored over 9 years in one blood bank.

To determine the time trend of HIV infection among non-professional blood donors at the Christian Medical College and Hospital, Vellore, the annual HIV prevalence among them for the years 1993-1997 was compared with that of the previous 5 yr. Since confirmed number of HIV positive blood donors is required for calculation of prevalence, the serum samples which were reactive for HIV 1/2 antibody by ELISA, were confirmed by Western blot. The annual prevalence of HIV antibody gradually increased from 1.6 per 1000 in 1988-1989, to 3.8 per 1000 in 1996-1997. The mean tri-annual prevalences for the years 1988-91, 1991-94 and 1994-97 were 1.3, 2.7 and 3.6 per 1000 respectively, suggesting a slow but steady increase in HIV prevalence over time. Our findings provide an insight into the dynamics of HIV infection in the general population of our country, whom the nonprofessional blood donors at our hospital represent.

Prevalence of HIV infection in pregnant women in Vellore region.

All pregnant women attending the antenatal clinic of a large hospital in Vellore, India (Christian Medical College Hospital) were screened for HIV infection between October 1987 and June 1992. A total of 36,953 blood samples were thus screened and 20 infected women were identified. Among these 18 women had acquired HIV infection from their husbands, who were also detected to be HIV infected. While these 18 women were monogamous, all the husbands had multiple sex partners. Two of the 20 women in this series were commercial sex workers. Among the 20, 17 (85%) women belonged to low socio-economic status, while three were from well-to-do families, with the husbands being businessmen or teachers. As the overall prevalence (0.054%) of HIV infection among pregnant women was relatively high and equal to or higher than many States in the USA and in areas outside London in the UK, the authors recommend that strict universal precautions be instituted in all obstetric practice in India.

PIP: All pregnant women attending the antenatal clinic at the Christian Medical College Hospital, Vellore, from October 1987 until June 1992 were screened for HIV antibody while testing them for evidence of syphilis using the VDRL test. The screening was done by enzyme linked immunosorbent assay (ELISA) using Vironostika Anti-Uniform, Vironostika HIV Mixt or Wellcozyme test kits. Reactive samples were retested and those confirmed reactive were subjected to Western blot test. The women were also given counselling. The husbands were then interviewed by trained social workers or physicians and tested for HIV antibodies. During these sessions, health education regarding the mode of transmission of HIV infection, risk to the newborn, family planning to prevent further pregnancies, and the use of condoms to prevent further transmission of infection was also elucidated. The women and their husbands were regularly followed up in the HIV clinic. The annual prevalence rate of infection ranged from 0.4/1000 to 1/1000 with no trend of increase over time. Among the 20 HIV seropositive women, 18 were monogamous and had received no blood transfusions. Their husbands were also found to be HIV seropositive. Each of the 18 husbands had had sex with commercial sex workers (CSWs), mostly in Vellore town. The remaining two women were CSWs and had had multiple sexual partners. The women were between 19 and 33 years of age. The husbands of the 18 married women were between 21 and 44 years of age. Seventeen (85%) women, including the two CSWs, were of low socioeconomic status. Three women ( 15%) belonged to middle class families. None of the women had clinical features of AIDS. Seventeen (85%) women had normal vaginal delivery; two (10%) had forceps delivery while one (5%) had caesarian section. The prevalence of HIV infection in pregnant women is already relatively high in the Vellore region and therefore all institutions providing obstetric care should institute strict infection control procedures in order to avoid risk to clients and health workers.

Increasing prevalence of human immunodeficiency virus infection among patients attending a clinic for sexually transmitted diseases.

Of 5883 patients screened during 1986 to 1993 in the sexually transmitted diseases (STD) clinic at a hospital in Vellore, south India, 105 (1.79%) were positive for human immunodeficiency virus (HIV) antibody. The prevalence of HIV infection increased from 0.26 per cent in 1986 to 2.64 per cent in 1993. In 1992 the prevalence was even higher, namely 3.94 per cent. The overall prevalence in these 8 yr was 1.98 per cent among men (n = 4392) and 1.21 per cent among women (n = 1491), the difference being statistically significant (P < 0.05). The prevalence in men rose from 0.36 per cent in 1986 to 4.24 per cent in 1992, and in women from 0 to 3.04 per cent during the same period. The major source of infection in men was female commercial sex workers, but among women it was mostly their husbands. The increase in prevalence of HIV infection is alarming, and indicates that the massive educational campaigns and other preventive measures have not resulted in decreased transmission.

PIP: In Vellore, India, health workers collected blood samples from 5883 new attendees at the sexually transmitted disease (STD) clinic of the Christian Medical Hospital to screen for HIV infection. 105 (1.79%) STD patients were confirmed positive for HIV infection (Western blot test). HIV prevalence increased between 1986 and 1992 (0.26-3.94%). It was 2.64% in 1993. The male:female ratio among the HIV-positive STD patients was 4.8:1 (1.98% vs. 1.21%; p 0.5). The age of the 87 HIV-positive men ranged from 19 to 55. All but three had unprotected sex with multiple partners who were commercial sex workers. The other three men were homosexuals. All but two of the HIV-positive women had sex only with their husbands. Clinicians and paramedical staff in STD clinics have already been trained in counseling and in screening contacts. They should continue to apply this training. They should strongly promote condom use among STD patients, especially since it is becoming a more and more acceptable method of preventing STDs. Early diagnosis and treatment and control of STDs would also be an intervention to control HIV infection.

Evidence for HTLV-III infection in prostitutes in Tamil Nadu (India).

PIP: Blood samples were collected from 102 female prostitutes housed in a custodial care institution in Tamil Nadu, India, to determine the presence of antibodies to human T-lymphotropic virus type III (HTLV-III). Both social and sexual histories were taken from 101 of the 102 women. Commercial test kits were used to test sera for antibody to HTLV-III. Reactive sera were tested for a 2nd time by the enzyme-linked immunoabsorbent assay (ELISA). Those repeatedly reactive sera were transported to the US, the National Institutes of Health, for western blot analysis. The sera from 11 of the study subjects were found to be repeatedly reactive in ELISA, and 10 were confirmed to have specific antibody to the virus by western blot analysis. Both infected and uninfected women were similar in age and of low socioeconomic status. The risk ratio for HTLV-III antibody was 8.2 in those women who had had sexual exposure to foreigners. None of the women were intravenous drug abusers, and all denied oral or rectal intercourse. On the basis of the stringent criteria used in the western blot analysis, it is believed that the 10 women have HTLV-III antibody. This emerges as the 1st report of evidence for HTLV-III infection in India. 10-40% of prostitutes in North America and Europe have HTLV-III antibody; the risk factors for infection appear to be intravenous drug use and penis-rectal intercourse. 54-88% of prostitutes in Central Africa have HTLV-III antibody, and the frequency of sexual contact with different partners is more important here as a risk-factor than the type of intercourse. As the prostitutes in this study in Indian did not use intravenous drugs and did not practice penis-rectal or penis-oral intercourse and had been prostitutes for shorter periods of time than the noninfected women and had fewer contacts, it is believed that HTLV-III infection has been introduced only recently into prostitutes in India. Sexual exposure to foreigners was a significant factor in the infected women.^ieng