Spain is one of the countries with the highest AIDS incidence rate and prevalence in women in Western Europe. Within Spain, the Autonomous Community of Madrid (CAM) is the area most affected by HIV-infection, with a total of 3606 women with AIDS up to June 2007. Most of these cases occurred by sharing injection material  especially in Madrid during the ‘epidemic of heroin’ in the 1990s that mainly affected the southern districts of the city reflecting social deprivation.
The high HIV prevalence in the female population had a direct impact on the spread of the infection in infants. In the CAM, 233 AIDS cases caused by mother-to-child transmission (MTCT) were reported between 1981 and 2007 .
Health inequalities between social classes are known to exist in the UK  and also in Spain [3–5]. A previous study highlighted geographic differences in socioeconomic status within Madrid . To our knowledge, there has been no study addressing the effect of social segregation on the epidemiology of HIV vertical transmission in Spain. Moreover, social segregation, along with immigration, is a phenomenon that is gaining in magnitude in recent decades and is an issue in the big European urban agglomeration. This is the first time that the spatial trend of viral spread in the paediatric population has been examined in a social context. Our objective was to assess the possible association between demographic and socioeconomic indicators of the residential area in Madrid and the geographical distribution of the HIV-1 MTCT cases.
Study population and methods
The CAM HIV Paediatric Cohort was established in 1995 as an open cohort of HIV vertically infected children. We assumed that vertical transmission occurred on the date of birth . The cohort included all HIV-1 positive children identified in a multicentre network of nine referral paediatric hospitals based in the CAM: Hospital General Universitario ‘Gregorio Marañón’, Hospital Universitario ‘Doce de Octubre’, Hospital Universitario ‘La Paz’, Hospital Universitario ‘Carlos III’, Hospital Infantil Universitario ‘Niño Jesus’, Hospital Universitario ‘Príncipe de Asturias’, Hospital General de Móstoles, Hospital ‘Severo Ochoa’ and Hospital Universitario de Getafe. Children were enrolled either retrospectively or prospectively from 1980 up to the current date. Informed consent was obtained from all mothers, and full ethical approval for the survey was obtained from the Ethical Committees of all hospitals.
HIV testing during pregnancy was offered to all women until 1998, when routine testing was introduced for all pregnant women. According to published guidelines, children were actively followed up every 3–6 months . The diagnosis of HIV infection was made as previously reported . Paediatricians administered the appropriate regimens according to international guidelines [8,10,11].
The municipality of Madrid had a population of 3 128 600 inhabitants in 2006 [12,13] and is divided into 21 districts, which represented the geographical unit of analysis.
The Institute of Statistics of the CAM supplied the cartographic base; data of the resident population of Madrid were collected from the Population Council Register. The addresses of the patients were collected from their medical history.
We incorporated in our analysis five indicators of social inequalities available in the Population Council Registry. These were ethnic density: proportion of non-Spanish people; educational level: percentage of illiterate persons older than 9 years (between 1981 and 1996) or older than 17 years in 2001 according to the highest academic qualification declared by the inhabitants; economic activity: proportion of unemployed people older than 15 years on the total of the active population, this variable was stratified by sex; economic level: annual per capita income (€ and index); health service availability: number of hospital beds per 1000 inhabitants, calculated as the ratio between the sum of the hospital beds of all the districts pertaining to the same ‘health area’ and the total population of these districts.
Statistical and geographic analysis
Within the cohort, we analysed data from all infected children born between 1980 and 2006 in Madrid. The study period was divided in five calendar periods on the basis of the changing HIV treatment . Within each period, we chose 1 or 2 years of reference (1981, 1986, 1991, 1996, 1997, 2001 and 2006) accordingly to the available updates of the Population Council Registry.
We georeferenciated the HIV MTCT cases at district level. Then, we elaborated maps of the HIV MTCT prevalence per 100 000 women living in each district between 15 and 49 years of age. We analysed the possible differences of the demographic and socioeconomic indicators at district level. Consequently, we performed the Spearman rank correlation test to analyse the correlation between HIV MTCT prevalence and each indicator from 1986 to 2006. Finally, we performed an ecological analysis by univariate and multivariate Poisson regression to study the association of each indicator with the probability of MTCT. A P value of less than 0.05 was considered statistically significant. The software ArcView GIS 3.1, Statistical Package for the Social Sciences 14 and STATA 9.1 were used.
Overall, 235 patients were born in Madrid between 1980 and 2006. Of these, we were able to georeferenciate a total of 224 children at district level.
Geographic distribution of MTCT prevalence
We observed a general increase in HIV MTCT prevalence in 1991 (Fig. 1), but the districts where we registered the higher prevalence were mainly in the south–south-east. This part of the city continued to be the most affected by the epidemic in 1996 and 2001. In 2006, we highlighted a general decrease in the prevalence, but the south–south-east districts continued to be the most affected.
The percentage of immigrants in Madrid increased rapidly after 1996 (Fig. 2a). Until 2001, immigrants were concentrated more in the northern area of the city as compared with the southern, whereas in recent years, this trend was opposite (mean 2.6 versus 1.2% in 1996; 14.5 versus 16.1% in 2006). We observed a significant negative correlation between the percentage of immigrants and the prevalence of HIV MTCT cases for the years 1996 (ρ = −0.635; P = 0.002) and 2001 (ρ = −0.621; P = 0.003) (Fig. 2a).
The percentage of illiterates was disturbingly high in the 1980s but halved in the course of the study period (Fig. 2b). The southern area of the city had a higher mean percentage of illiterates as compared with the northern (38.4 versus 28.0% in 1986; 15.5 versus 7.8% in 2001). The correlation between the HIV MTCT prevalence and percentage of illiterates was positive (ρ = 0.523; P = 0.015 in 1996; ρ = 0.635; P = 0.002 in 2001) (Fig. 2b).
The mean percentage of unemployed women was higher than men and it was higher in the southern area than the northern (data not shown). The HIV MTCT prevalence was positively associated with the percentage of unemployed women in 1996 (ρ = 0.540; P = 0.012) and 2001 (ρ = 0.639; P = 0.002) (Fig. 2c). However, we did not observe any correlation with the percentage of unemployed men.
The northern area was richer than the southern (mean annual income index of 116.8 versus 84.6 in 1996; 123.3 versus 80.2 in 2000). The HIV MTCT prevalence was negatively correlated to the mean annual income in 1996 (ρ = −0.582; P = 0.006) and 2000 (ρ = −0.558; P = 0.009) (Fig. 2d). Overall, the mean number of hospital beds decreased from 5.7 in 1987 to 4.5 in 2001. It was at least double in the northern area of Madrid than the southern (8.7 versus 3.7 in 1987; 6.6 versus 2.6 in 2001). Any correlation between the prevalence of HIV cases and this variable was not observed.
Univariate analysis showed that the percentage of immigrants as well as the annual per capita income were associated with a lower MTCT prevalence [relative risk (RR), 0.80; 95% confidence interval (CI), 0.75–0.85; P < 0.001 and RR, 0.80; 95% CI, 0.76–0.85; P < 0.001, respectively], whereas the percentage of illiterates and unemployed women was associated with a higher MTCT prevalence (RR, 1.04; 95% CI, 1.03–1.05; P < 0.001 and RR, 1.08; 95% CI, 1.06–1.09; P < 0.001, respectively). We observed an increase in the risk in the periods up to 1996 as compared with the calendar period 1980–1989, whereas the risk decreased afterwards (Fig. 2e). After adjustment in Poisson regression for calendar period and district, the risk was especially higher in the districts of Usera [absolute relative risk (ARR), 11.4; 95% CI, 2.6–49.5; P = 0.001], Puente de Vallecas (ARR, 14.0; 95% CI, 3.4–57.9; P < 0.001) and San Blas (ARR, 12.5; 95% CI, 2.9–53.6; P = 0.001) in comparison with the district of Salamanca, where a lower MTCT prevalence was registered. These three districts were between the urban areas where illiterates were mainly represented, and this variable was the only statistically significant one after adjustment for the indicators of social inequalities stratified by calendar period (Fig. 2e).
Our manuscript highlights the association between low socioeconomic status and spatial pattern of HIV MTCT epidemic in Madrid, with the support of GIS. Nowadays, paediatric HIV cases are still being diagnosed in industrialized countries  despite the efficacy of MTCT prophylactic treatment [16–18]. In fact, in 2006, two vertically infected children were diagnosed in Madrid. We estimated the correlation between indicators of social inequalities and the geographic distribution of the HIV MTCT comparing two urban areas, north and south. We observed the emergence of core areas of transmission in the south–south-east boundaries that remained the more affected areas of the city since recent years. Core areas were between the more poor and disadvantaged areas. Thus, Usera, Puente de Vallecas and San Blas had the highest concentration of illiterates. These three districts and all the south-east area had the highest rate of unemployment and received the lowest annual per capita income. Due to the immigration phenomenon, Spain has become in a very multiethnic society during the present decade. We observed an important change in the migratory flux. Although immigrants were more concentrated in the northern districts until the end of the 1990s, they were more represented in the southern areas in recent years, with the exception of the districts Centro and Tetuan. This pattern of residential segregation can be accounted by the fact that in the past, immigrants were generally well-off people coming from rich countries, whereas nowadays, they originate mainly from developing countries .
If our analysis is not biased by any potential confounders, it clearly masks different patterns of the HIV epidemic in the infant population in Madrid. At the same time, it shows a correlation between prevalence of HIV MTCT infection and a measure of low socioeconomic status. Previous studies highlighted the association between the risk of sexually transmitted diseases and socioeconomic deprivation in Western countries [20–22]. Anyway, a serious disadvantage of our study was the assumption that the demographic and socioeconomic information at district level was representative of the HIV-infected children living in these areas. This approach could generate bias , but its important advantage was the availability of data. Another major shortcoming could be represented by the possible geographic heterogeneity of factors affecting the HIV MTCT prevalence and the indicators of social inequalities that we did not take into account, which could introduce ecologic confounding. In particular, we might have analysed the use of injecting drugs that represented, especially in the initial years of the spread of the virus, the main mode of transmission among women. In our cohort, 63% of the women acquired HIV-1 through injecting drug use. Unfortunately, this data was not available at district level. However, we believe that this public health problem can be related to the socioeconomic status of the population, so that even if indirectly, we have taken it into account. Moreover, the validity of our maps depends upon the quality of the reported HIV MTCT cases and on the georeferentiation procedure. Missing data accounted for only 11 of 235 patients.
Our findings highlight the presence of core areas of HIV vertical transmission reinforcing the core group hypothesis [24,25]. Moreover, on the basis of previous publications [26,27], we hypothesize that the same areas could be important for the spread of other sexually transmitted diseases.
The more relevant conclusion is that our study could be extrapolated to other industrialized cities. More interestingly, core areas of MTCT could exist also in other industrialized cities where inhabitants have to face poverty and disadvantaged life conditions. Our obligation should consist of focusing attention on this phenomenon and improving prevention programs within this part of the population.
This work has been supported by grants from Fundación para la Investigación y Prevención del SIDA en España, FIPSE (36514/05, 36536/05, 24632/07, 36644/07), Fundación Caja Navarra, Red Temática de Investigación Cooperativa Sanitaria ISCIII (RED RIS RD06/0006), Fondo de Investigación Sanitaria (PI061479, PI070236), Comunidad Autónoma de Madrid (SAL/2001/2004) and Network of Excellence TEDDY (Task-Force in Europe for the Drug Development in the Young) supported by the EC-Sixth Framework Program (Contract no. 0005216 LSHBCT-2005-005126) to MAMF. Claudia Palladino is supported by grant from Istituto Pasteur, Fondazione Cenci-Bolognetti, Università degli Studi di Roma ‘La Sapienza’, P.le Aldo Moro 5, 00185 Rome, Italy. The authors wish to thank Maria Jesus Mazario Sopeña of the Madrid city council for her kindness and helpfulness.
Authors' contributions: Claudia Palladino had primary responsibility for protocol development, carried out the epidemiological survey and the analysis of the demographic and socioeconomic characteristics of the population, elaborated the maps and the majority of the figures of the manuscript and performed part of the statistical analysis. Jose María Bellón participated in the development of the protocol and had primary responsibility in analytic framework and database maintenance. Santiago Perez-Hoyos performed a major role in carrying out the Poisson statistical analysis. Rosa Resino collaborated in data gathering and had primary responsibility for geographic study. Sara Guillén was responsible in data gathering in the participant hospitals. Dolores García participated in protocol development and had primary responsibility for diagnostics of HIV infection, viral load and CD4 cell analysis. Ma Dolores Gurbindo, Ma Isabel de José, José Tomás Ramos, Ma Josefa Mellado were the clinicians responsible for patient screening and follow-up in the participating hospitals and also contributed to the writing of the manuscript. Dra. Ma Angeles Muñoz-Fernández had full access to all the data in the study and supervised its design and execution, performing the final data analyses and writing the manuscript.
Spanish group of paediatric HIV infection: participating hospitals and personnel staff in this paper are as follows: Hospital General Universitario ‘Gregorio Marañón’: C. Palladino, R. Resino, J.M. Bellón, D. García Alonso, M.D. Gurbindo, M.L. Navarro and M.A. Muñoz-Fernández; Hospital Universitario ‘La Paz’: M.I. Isabel de José, B. Larru; Hospital Universitario ‘12 Octubre’: S. Guillen, I. Tomé, P. Carreño, J. Ruiz, J. Clemente; Hospital Universitario ‘Carlos III’: P. Martín-Fontelos, M.J. Mellado, J. Villota; Hospital Universitario de Getafe: J.T. Ramos; Hospital Infantil Universitario Niño Jesús: L. Ciria, J. Martínez Pérez; Hospital Universitario Príncipe de Astúrias: J. Beceiro and Hospital Severo Ochoa: C. Calvo.
There are no conflicts of interest.
2. Department of Health and Social Security. Inequalities in health: report of a research working group
. London: DHSS; 1980.
3. Navarro V, Benach J, the Scientific Commission for the study of health inequalities in Spain. Social inequalities in health in Spain
: Ministerio de Sanidad y Consumo; 1996.
4. Regidor E, Barrio G, de la Fuente L, Domingo A, Rodríguez C, Alonso J. Association between educational level and health related quality of life in Spanish adults. J Epidemiol Community Health 1999; 53:75–82.
5. Borrell C, Pasarín M. The study of social inequalities in health in Spain: where are we? J Epidemiol Community Health 1999; 53:388–389.
6. Chueca Goitia F, Sambricio C, Capitel A, Ruiz Cabrero G, Hernández de León J. Architecture of Madrid, XX century
: Antonio Camuñas Tanais Foundation; cop. 1999.
7. de Martino M, Tovo PA, Balducci M, Galli L, Gabiano C, Rezza G, et al
. Reduction in mortality with availability of antiretroviral therapy for children with perinatal HIV
-1 infection. Italian Register for HIV
Infection in Children and the Italian National AIDS Registry. JAMA 2000; 284:190–197.
8. CDC. Guidelines for use of antiretroviral agents in pediatric HIV infection
. MMWR Morb Mortal Wkly Rep
9. Resino S, Alvaro-Meca A, de Jose MI, Martin-Fontelos P, Gutierrez MD, Leon JA, et al
. Low immunologic response to highly active antiretroviral therapy in naive vertically human immunodeficiency virus type 1-infected children with severe immunodeficiency. Pediatr Infect Dis J 2006; 25:365–368.
10. CDC. Recommendations of the US Public Health Service Task Force on the use of zidovudine to reduce perinatal transmission of human immunodeficiency virus
. MMWR Recomm Rep
11. Sharland M, Blanche S, Castelli G, Ramos J, Gibb DM. PENTA guidelines for the use of antiretroviral therapy, 2004. HIV
Med 2004; 5(Suppl 2):61–86.
14. Resino S, Resino R, Bellon JM, Micheloud D, Gutierrez MD, de Jose MI, et al
. Clinical outcomes improve with highly active antiretroviral therapy in vertically HIV
type-1-infected children. Clin Infect Dis 2006; 43:243–252.
15. Macassa E, Burgard M, Veber F, Picard C, Neven B, Malhaoui N, et al
. Characteristics of HIV
-infected children recently diagnosed in Paris, France. Eur J Pediatr 2006; 165:684–687.
16. Cooper ER, Charurat M, Mofenson L, Hanson IC, Pitt J, Diaz C, et al
. Combination antiretroviral strategies for the treatment of pregnant HIV
-1-infected women and prevention of perinatal HIV
-1 transmission. J Acquir Immune Defic Syndr 2002; 29:484–494.
17. European Collaborative Study. Mother-to-child transmission of HIV infection in the era of highly active antiretroviral therapy
. Clin Infect Dis
18. Volmink J, Siegfried N, van der Merwe L, Brocklehurst P. Antiretrovirals for reducing the risk of mother-to-child transmission of HIV infection
. Cochrane Database Syst Rev
19. Méril J, Alcolea Moratilla M. Report about social segregation and immigration of Madrid
. Revista Electronica de Medioambiente 2006; 2:73–89.
20. Krieger N, Waterman PD, Chen JT, Soobader MJ, Subramanian SV. Monitoring socioeconomic inequalities in sexually transmitted infections, tuberculosis, and violence: geocoding and choice of area-based socioeconomic measures – the public health disparities geocoding project (US). Public Health Rep 2003; 118:240–260.
21. Monteiro EF, Lacey CJ, Merrick D. The interrelation of demographic and geospatial risk factors between four common sexually transmitted diseases. Sex Transm Infect 2005; 81:41–46.
22. Fleming DT, McQuillan GM, Johnson RE, Nahmias AJ, Aral SO, Lee FK, et al
. Herpes simplex virus type 2 in the United States, 1976 to 1994. N Engl J Med 1997; 337:1105–1111.
23. Greenland S, Robins J. Ecological studies: biases, misconceptions, and counterexamples. Am J Epidemiol 1994; 139:747–760.
24. Bernstein KT, Curriero FC, Jennings JM, Olthoff G, Erbelding EJ, Zenilman J. Defining core gonorrhea transmission utilizing spatial data. Am J Epidemiol 2004; 160:51–58.
25. Zenilman JM, Glass G, Shields T, Jenkins PR, Gaydos JC, McKee KTJ. Geographic epidemiology of gonorrhoea and chlamydia on a large military installation: application of a GIS system. Sex Transm Infect 2002; 78:40–44.
26. Law DC, Serre ML, Christakos G, Leone PA, Miller WC. Spatial analysis and mapping of sexually transmitted diseases to optimise intervention and prevention strategies. Sex Transm Infect 2004; 80:294–299.
27. Rothenberg R, Muth SQ, Malone S, Potterat JJ, Woodhouse DE. Social and geographic distance in HIV
risk. Sex Transm Dis 2005; 32:506–512.