Epidemiology and Social
Can migrants from high-endemic countries cause new HIV outbreaks among heterosexuals in low-endemic countries?
Xiridou, Mariaa; van Veen, Maaikea; Coutinho, Roela,b; Prins, Mariab,c
aCentre for Infectious Diseases Control, National Institute of Public Health and the Environment, Bilthoven, The Netherlands
bAcademic Medical Center, University of Amsterdam, The Netherlands
cAmsterdam Municipal Health Service, Amsterdam, The Netherlands.
Received 30 November, 2009
Revised 23 March, 2010
Accepted 25 March, 2010
Correspondence to M. Xiridou, National Institute of Public Health and the Environment, P.O. Box 1, 3720 BA Bilthoven, The Netherlands. Tel: +31 30 2743057; fax: +31 30 2744409; e-mail: email@example.com
Objectives: To investigate how the sexual behaviour of migrants originating from HIV-endemic countries affects the spread of HIV among heterosexuals in low-endemic countries.
Methods: A mathematical model is developed describing the transmission of HIV in heterosexual partnerships between African migrants, Caribbean migrants, and local natives. The model accounts for infection of migrants before migration and during trips to their home country. The model is parameterized using data from the Netherlands.
Results: Among new and newly imported, heterosexually acquired, infections in 2010 in the Netherlands, the individual acquiring HIV is an African in 53% of cases, a Caribbean in 26% of cases, and a Dutch native in 21% of cases. The percentage of new infections acquired outside the Netherlands is 40% among African migrants and 32% among Caribbean migrants; these are mostly acquired before migration to the Netherlands. The prevalence of HIV in the Netherlands is hardly affected by changes in risk behaviour of migrants during trips to their home country after migration. If migrants mix more with the Dutch in forming partnerships, then HIV prevalence among migrants will decrease. The more initiating antiviral therapy is delayed among migrants, the higher the resulting prevalence in their own ethnic group and among the Dutch.
Conclusion: The serostatus of individuals migrating to low-prevalence countries as well as their sexual behaviour in the country of residence affect considerably the spread of HIV. Preventive measures should focus on targeted interventions, promoting safe sex practices, HIV testing, and entry to specialized HIV care among migrants.
Individuals originating from HIV-endemic countries comprise a relatively large proportion of heterosexually acquired HIV infections in Europe [1–3]. Migrants may engage in unsafe sexual practices in their country of residence, but also in their country of origin, while visiting their family and friends [4,5]. In addition, some migrants may have concurrent partners [6–8]. Most partners of migrants are individuals from their own country, but sometimes also natives of their country of residence [9,10]. If migrants originate from a country with higher HIV prevalence than that in their country of residence, they may form a bridge population that imports new infections from the high-endemic to the low-endemic country. Furthermore, via their sexual partners in the country of residence, they may spread the infection to the general indigenous population. Molecular studies have shown that indeed import of heterosexual HIV infections from high-endemic countries has occurred . It is imperative, therefore, to understand the contribution of migrants originating from high-endemic countries to the spread of HIV in lower-endemic countries, such that prevention efforts can be targeted at the populations most at risk for infection.
To investigate that, a mathematical model has been developed, that describes the formation of heterosexual partnerships between local natives and migrants. The model is parameterized using data from the Netherlands [10,12–14]. The heterosexual epidemic in the Netherlands serves as an example of a European country with low HIV prevalence, concentrated mainly in specific risk groups (such as men who have sex with men, injecting drug users, and migrants from HIV-endemic countries). In the Netherlands, there are two large groups of migrants from countries with medium to high HIV prevalence: migrants from sub-Saharan Africa and migrants from Suriname, the Netherlands Antilles, and Aruba (former Dutch colonies), comprising 1 and 3% of the adult population, respectively . The prevalence of HIV was estimated at 0.2% in 2005 in the whole Dutch population , which is lower than that in the Caribbean (1.1%) and in sub-Saharan Africa (5%) [16,17]. We examine to what extent migrants ‘drive’ the epidemic, what fraction of new and prevalent heterosexual infections can be attributed to migrants, and how behavioural and treatment changes among migrants could affect the heterosexual transmission of HIV. The model includes migration of infected individuals and travel-related infection during visits of migrants to their country of origin. We show that public health measures focusing on the sexual risk behaviour of migrants, facilitating HIV testing and access of migrants to antiretroviral therapy can considerably contribute to reducing HIV transmission and the burden of disease in migrant communities.
In the model, three ethnic groups are distinguished African migrants, Caribbean migrants, and the remaining ‘general’ Dutch population. The last group will be referred to as Dutch natives to contrast with the group of migrants and includes indigenous Dutch those who themselves and their parents were born in the Netherlands, migrants from countries with HIV prevalence lower than that in the Netherlands, and other small migrant groups (not from Africa or the Caribbean). The group of migrants comprises of first and second generation migrants, meaning those who were themselves or one of their parents born in the specific country. The heterosexual population of the Netherlands is, therefore, stratified into six subgroups, according to sex (male, female) and ethnicity (Dutch, Africans, and Caribbeans). Each of the six subgroups is further divided into five classes, according to HIV status such as uninfected, acute HIV infection, chronic HIV infection untreated, pre-AIDS untreated, and HIV treated (see Fig. S1 in Supplemental Digital Content, for a schematic diagram of the model, http://links.lww.com/QAD/A58). Among those with chronic HIV, a proportion initiates combination antiretroviral therapy (cART) and progresses to the treated class, whereas the rest remains in the chronic class and progresses to the pre-AIDS class. From the pre-AIDS and treated classes, individuals may progress to AIDS.
Individuals get infected via unprotected sexual intercourse in the Netherlands. Migrants can also get infected via sexual contacts in their country of origin, while visiting these countries. Some migrants report having concurrent steady partners or casual partners along with a steady partner . To account for that, we model two kinds of sexual partnerships namely main and secondary partnerships. Main partnerships are of long duration with many sexual contacts per partner; secondary partnerships are of shorter duration, with less sexual contacts per partner than main partnerships, and include a second (concurrent) steady partner or casual partners. The population includes all sexually active adult heterosexuals, 15–69 years old. Dutch heterosexuals enter the population uninfected. Among migrants, a large fraction is born in their country of origin and enter the Netherlands as adults . Therefore, a fraction of those migrating to the Netherlands is already infected (for simplicity assumed to be with chronic HIV) and the remaining individuals are uninfected.
Data sources and parameter estimates
Parameters relating to sexual behaviour of migrants living in the Netherlands were estimated from data from the ‘HIV survey among migrants in Amsterdam’ . Parameters relating to sexual behaviour of Dutch natives were estimated using data from two national surveys, the second PIENTER study  and the surveys of the Rutgers-NISSO Group . In the Supplemental Digital Content more information about the data sources and about the parameter estimation is provided, http://links.lww.com/QAD/A103. The values of the parameters are summarized in Table 1 and in Tables S1-S3 in the Supplemental Digital Content, http://links.lww.com/QAD/A61, http://links.lww.com/QAD/A62, http://links.lww.com/QAD/A66. These values will be referred to as the basic scenario.
Initial conditions for numerical results
From data from the HIV Monitoring Foundation  the numbers of individuals living with HIV/AIDS and of those receiving cART were estimated for the whole country for each year since the beginning of the epidemic in the Netherlands in 1980. From these numbers, we calculated the percentage of those living with HIV/AIDS receiving cART. This percentage started increasing from 1996 steadily every year; since 2002, it has been around 75–80% . Because the data on sexual behaviour are also for the same time period (2003–2006), we initiated the numerical results at the year 2002. It was assumed that since that year, cART administration has been constant . The prevalence of HIV among migrants from Suriname and the Netherlands Antilles in the ‘HIV survey among migrants in Amsterdam’ was 0.4% in 2003–2004 . The prevalence of HIV among women undergoing standard screening during pregnancy was 0.025% in 2006 . For 2008, the prevalence of HIV in the Netherlands was estimated at 0.2% in the whole population; excluding men who have sex with men and splitting the analysis according to ethnic origin of those living in the Netherlands, the prevalence was estimated at 3.44% among African migrants, 0.35% among Caribbean migrants, and 0.04% in the remaining (mostly) indigenous heterosexual population . Therefore, for the numerical results the initial prevalence used was 3.5% for Africans, 0.5% for Caribbeans, and 0.05% for Dutch natives.
HIV prevalence among adult heterosexuals
From the model, we calculated the prevalence and incidence during the first 20 years after cART administration reached the current levels (years 2002–2022). The prevalence among Dutch natives (calculated as the number of Dutch infected divided by the total number of Dutch) slightly decreases from 0.050% in 2002 to 0.043% in 2022 (Fig. 1b). Among African migrants the prevalence increases from 3.5 to 4.4%, among Caribbean migrants from 0.5 to 0.7% (Fig. 1a), and in the total heterosexual population from 0.098 to 0.106% (Fig. 1b) in the same period. Among all the infected individuals being alive in 2010, 39% originate from Africa, 17% from the Caribbean, and 44% from the Netherlands.
New (incident) HIV infections
Incidence among adult heterosexuals in 2010 was calculated at 1.50 new infections (occurring via sexual contacts in the Netherlands or during trips of migrants to their home country) per 100 000 individuals per year. Including also those migrating to the Netherlands (in 2010) being already infected, results in a total of 2.11 ‘new’ infections per 100 000 individuals per year (here ‘new’ refers to infections ‘new to the country’, because they are acquired abroad, but they are not necessarily recent infections). Among these ‘new’ infections, 70.3% occurred via sexual contacts in the Netherlands, 0.7% during visits of migrants to their home country, and 29% were individuals who migrated in 2010 to the Netherlands being already infected (Fig. 1c). The individual acquiring HIV was an African migrant in 53% of ‘new’ infections (32% via sex in the Netherlands, 0.5% during trips to home country, and 21% infected before migration); or a Caribbean migrant in 26% of ‘new’ infections (18% via sex in the Netherlands, 0.2% during trips to home country, and 8% infected before migration); or a Dutch native in 21% of ‘new’ infections (Fig. 1c). Among infections via sexual contacts in the Netherlands, in 53% the infectious partner was an African migrant, in 22% a Caribbean migrant, and in 25% a Dutch native (Fig. 1d). One in five newly infected Dutch natives acquired their infection from a migrant and 80% from another Dutch native (Fig. 1e). Finally, we investigated whether migrants were mostly infected in the Netherlands or abroad. In 2010, 40% of new infections among African migrants and 32% among Caribbean migrants were acquired outside the Netherlands (see Fig. S2 in Supplemental Digital Content, http://links.lww.com/QAD/A59).
Sexual behaviour in country of origin
As only a few infections occur during trips of migrants to their home country, moderate variations in the factors determining infections abroad have only a minor impact on HIV prevalence in the Netherlands. We calculated HIV prevalence in the different ethnic subpopulations in the Netherlands, under different scenarios for the level of sexual behaviour during trips of migrants to their home country. By increasing, for instance, the number of partners, or decreasing condom use, only minor changes were observed in the prevalence in the Netherlands.
Sexual behaviour of migrants while in the Netherlands
We investigated the impact of changing the number of partners of migrants on the prevalence of HIV (see Fig. S3 in Supplemental Digital Content, http://links.lww.com/QAD/A60 ). If Caribbean or African migrants have more partners, the prevalence of HIV increases considerably in their own community, but only slightly in the other ethnic groups. Specifically among Dutch natives and in the total heterosexual population, the prevalence is hardly affected. Decreasing condom use or increasing the frequency of sexual intercourse of migrants have the same effect (results not shown).
Ethnic mixing in sexual relationships in the Netherlands
Figure 2 shows the effect of increasing mixing, such that migrants form more partnerships with Dutch natives and less with individuals of their own ethnic background (the mixing between Africans and Caribbeans was kept as estimated from the data– see Table 1). Looking at the first 20 years of such a change, the prevalence among migrants decreases, whereas that among the natives slightly increases. This happens because the prevalence among natives is much lower than that among migrants and since more ‘mixed’ partnerships are being formed, it becomes more likely for an infected migrant to have a Dutch partner and spread the infection to the Dutch population rather than to the migrant population; also it becomes more likely for an uninfected migrant to have a Dutch partner and hence avoid getting infected.
Effect of cART uptake
Migrants frequently appear late to health services for testing or for treatment [21,22], due to fear of their partners or their community. In addition, several migrant communities suffer from underreporting and/or large number of illegal migrants who do not consult the local health authorities. It is possible, therefore, that the levels of cART administration that we obtained from the HIV registries are rather over-optimistic. In order to investigate the effect of this, we calculated the prevalence of HIV, with lower levels of cART administration and longer interval until initiation of cART among migrants (Fig. 3). For Dutch natives, cART administration was as in the basic scenario (see Table S3 in Supplemental Digital Content, http://links.lww.com/QAD/A66). We see here that the higher the fraction of infected migrants not receiving cART and the longer they remain untreated, the higher the resulting prevalence in all ethnic groups. As for the previous results, this change is more profound among migrants. For instance, if 50% of migrants initiate cART one year after infection, then HIV prevalence among African and Caribbean migrants is 3.90% and 0.57%, respectively, but the prevalence is 4.76% and 0.74%, if cART is initiated five years after infection. Among Dutch, there is only a minor change in prevalence (0.043% or 0.044% if cART is initiated one or five years after infection).
The present study shows that the heterosexual epidemic in the Netherlands is relatively low and stable, despite the presence of migrant groups from HIV-endemic countries. The epidemic is kept at this level and does not further diminish, partly due to the immigration of individuals being already infected and the relatively high prevalence among migrants living in the Netherlands. Nevertheless, due to the small number of migrants, their relatively moderate sexual risk behaviour and low mixing with the Dutch, migrants cannot easily trigger major rises in HIV among heterosexuals. This can be observed from the fact that behavioural and treatment changes among migrants have a major impact in the migrant communities, but hardly affect the prevalence among Dutch natives. The sexual contacts that migrants have during trips to their home country result in only a very small number of new infections. On the contrary, immigration of infected individuals accounts for a considerable fraction of the current HIV prevalence. Consequently, the sexual behaviour of migrants during trips to their home country hardly affect the heterosexual epidemic in the Netherlands, as their sexual behaviour in the Netherlands has a major impact on HIV prevalence. Earlier HIV-diagnosis and initiation of cART among migrants can also result in considerable reductions in HIV transmission.
These findings are applicable not only to the Netherlands but other European and high-income countries as well. In England, for instance, there are also large migrant communities from sub-Saharan Africa and the Caribbean. In 2007, the prevalence of diagnosed HIV in black African and black Caribbean communities in England was estimated at 3.7 and 0.4%, respectively, as among the indigenous population at 0.09% . In addition, the proportion of late diagnoses was 42 and 27% among African and Caribbean migrants, respectively . A recent study in the United Kingdom indicated that for 25–33% of infected black-African migrants, the infection was acquired in the United Kingdom . In Ireland  and in Sweden , high rates of heterosexual HIV transmission among migrants have also been reported. We expect that in other European countries, as for the Netherlands, the travel-behaviour of migrants is not important for the heterosexual HIV epidemic, but their sexual risk behaviour in the country of residence is important. In addition, as shown in our study, other European countries also report a large fraction of migrants acquiring their infection before migration to Europe . Furthermore, our model can be easily used to investigate the spread of HIV among heterosexuals in other countries, by adjusting appropriately the parameters relating to sexual risk behaviour.
The present study can be extended to investigate the impact of other factors relating to migration, such as the changing size of migrant populations (if a country becomes more or less ‘popular’ to migrate to or if stringent policies limit the influx of migrants) and the fact that migrants may stay in a new country only for short periods of time. The model can also be extended to account for the presence of coinfections with other STI and investigate their impact on the spread of HIV. The presence of other STI may increase susceptibility to HIV infection and infectiousness of those with the coinfection . Migrants from the Caribbean and sub-Saharan Africa living in the Netherlands have been shown to have higher prevalence of STI than the general Dutch population . Furthermore, it would be interesting to include in the model other risk groups, such as sex workers or injecting drug users, which also ‘mix’ with the general heterosexual Dutch population, and hence could account for a fraction of HIV transmission; the contribution of these risk groups to the heterosexual epidemic in the Netherlands could then be investigated.
Our findings indicate that the heterosexual transmission of HIV in the Netherlands is mainly occurring within migrant communities. Therefore, limiting migration or imposing travel restrictions would hardly affect the prevalence of HIV in the country. Prevention should focus on the sexual behaviour of migrants while in the Netherlands, and less on their behaviour during trips to their home country. Moreover, it is essential to detect those entering the country being infected as soon as possible. Timely initiation of cART is extremely important not only for the infected individual [28–30], but also for limiting the spread of HIV to the community [31,32]. Socioeconomic status, HIV stigma, along with cultural and language barriers might impair the uptake of HIV testing and treatment, and compliance with cART. Policy making should focus on targeted interventions avoiding stigmatization and on facilitating testing and referral of infected migrants to specialized HIV care, in order to reduce the burden of disease in migrant communities.
The authors would like to thank the WHO Regional Office for Europe for partially funding this project and in particular Martin Donoghoe, Annemarie Rinder Stengaard, and Stine Nielsen (currently at the Koch Institute, Berlin, Germany) for their assistance and suggestions. The authors are grateful to Liesbeth Mollema and Fiona van der Klis for providing the PIENTER data and to Martijn van Rooijen and Han Fennema (Municipal Health Service Amsterdam) for providing the data from the Anonymous Unlinked HIV Testing Survey in Amsterdam. The authors thank Eline Op de Coul, Merlijn Kramer, and Ingrid van de Broek for information and assistance with the data; Marianne van der Sande, Jacco Wallinga, and Mirjam Kretzschmar for useful comments and suggestions. The anonymous referees are also thanked for several constructive comments that improved the manuscript. The participants of all the studies from which data were used are gratefully acknowledged.
1. European Centre for Disease Prevention and Control/WHO Regional Office for Europe. HIV/AIDS surveillance in Europe 2007
. Stockholm: European Centre for Disease Prevention and Control/WHO Regional Office for Europe; 2008.
2. van den Broek I, Koedijk F, van Veen M, Op de Coul E, van Sighem A and van der Sande M. Sexually transmitted infections including HIV in the Netherlands in 2007.
Report of the National Institute of Public Health and the Environment, Number 210261004/2008,2008.
3. The UK Collaborative Group for HIV and STI Surveillance. A complex picture of HIV and other sexually transmitted infections in the United Kingdom, 2006.
London: Health Protection Agency, Centre for Infections; 2006.
4. Sadler K, McGarrigle C, Elam G, Ssanyu-Sseruma W, Davidson O, Nichols T, et al
. Sexual behaviour and HIV infection in black-Africans in England: results from the Mayisha II survey of sexual attitudes and lifestyles. Sex Transm Infect 2007; 83:523–529.
5. Kramer M, van Veen M, Op de Coul E, Geskus R, Coutinho R, van de Laar M, et al
. Migrants travelling to their country of origin: a bridge population for HIV transmission? Sex Transm Infect 2008; 84:554–555.
6. Kocken P, van Dorst A, Schaalma H. The relevance of cultural factors in predicting condom-use intentions among immigrants from the Netherlands. Health Educ Res 2006; 21:230–238.
7. van Veen M, Schaalma H, van Leeuwen A, Prins M, de Zwart O, van de Laar M, et al. Concurrent partnerships and sexual risk taking among African and Caribbean migrant populations in the Netherlands
. Int J STDAIDS
8. Adimora A, Schoenbach V, Doherty I. Concurrent sexual practices among men in the United States. Am J Public Health 2007; 97:2230–2237.
9. Ford K, Sohn W, Lepkowski J. American adolescents: sexual mixing patterns, bridge populations, and concurrency. Sex Transm Dis 2002; 29:13–19.
10. van Veen M, Kramer M, Op de Coul E, van Leeuwen A, de Zwart A, van de Laar M, et al
. Disassortative sexual mixing among migrant populations in the Netherlands: a potential for HIV/STI transmission? AIDS Care 2009; 21:683–691.
11. Op de Coul E, Coutinho R, van der Schoot A, van Doornum G, Lukashov V, Goudsmit J, et al
. The impact of immigration on env
HIV-1 subtype distribution among heterosexuals in the Netherlands: influx of subtype B and non-B strains. AIDS 2001; 15:2277–2286.
12. van der Klis F, Mollema L, Berbers G, de Melker H, Coutinho R. Second national serum bank for population-based seroprevalence studies in the Netherlands. Neth J Med 2009; 67:301–308.
13. Rutgers Nisso Group. Sexual health in the Netherlands 2006
. Eburon, Delft: Rutgers Nisso Group; 2006.
15. Op de Coul E, van Sighem A and van de Laar M. National estimate of people living with HIV/AIDS in 2005 in the Netherlands
. Inf Dis Bull
:398–403. (in Dutch).
16. UNAIDS/WHO. 2008 report on the global AIDS epidemic. Annex: HIV and AIDS estimates and data, 2007 and 2001. http://www.unaids.org
17. Medical and Public Health Service of Curacao. Cumulative registered number of known HIV positives, Netherlands Antilles. 1985–2002.
18. Gras L, van Sighem A, Smit C, Zaheri S, Schuitemaker H and de Wolf F. Monitoring of human immunodeficiency virus (HIV) infection in the Netherlands report 2007.
Amsterdam, The Netherlands: HIV Monitoring Foundation; 2007.
19. van Veen M, Wagemans M, Op de Coul E, Fennema J, van der Helm T, Walter J, et al. HIV-surveys among high-risk groups in Amsterdam, 2003–2004.
Report of the National Institute of Public Health and the Environment, Number 441100021/2005,2005. (in Dutch).
20. van Veen M, Presanis A, Conti S, Xiridou M, van der Sande M, Donoghoe M, et al. National estimate of HIV prevalence in the Netherlands: comparison of different tools. Netherlands Conference on HIV Pathogenesis, Prevention and Treatment.
Amsterdam, 24 November 2009. Abstract no. 24, 2009.
21. Health Protection Agency. Sexually transmitted infections in the black African and black Caribbean communities in the UK: 2008 report
. London: Health Protection Agency, Centre for Infections; 2008.
22. Lanoy E, Lewden C, Lievre L, Tattevin P, Boileau J, Aouba A, et al
. How does loss to follow-up influence cohort findings on HIV infection? A joint analysis of the French hospital database on HIV, Mortalite 2000 survey and death certificates. HIV Medicine 2009; 10:236–245.
23. Burns F, Arthur G, Johnson A, Nazroo J, Fenton K, and on behalf of the SONHIA collaboration group. United Kingdom acquisition of HIV infection in African residents in London: more than previously thought. AIDS 2009; 23:262–265.
24. Hargreaves S. Ireland addresses HIV and migration. Lancet Inf Dis 2008; 8:470.
25. Arneborn M and Blaxhult A. Increase in reported HIV infections in Sweden
. Euro Surveill
26. Padian N, Buve A, Balkus J, Serwadda D, Cates W Jr. Biomedical interventions to prevent HIV infection: evidence, challenges, and way forward. Lancet 2008; 372:585–599.
27. de Boer-van der Kolk I, van de Laar M and Op de Coul E. Surveillance of sexual behaviour within the regular STI surveillance: a pilot. Tijdschrift Gezondheidswetenschappen
:537–541. (in Dutch).
28. Smith D, Berrey M, Robetson M, Mehrotra D, Markowitz M, Perrin L, et al
. Virological and immunological effects of combination antiretroviral therapy with zidovudine, lamivudine, and indinavir during primary human immunodeficiency virus type 1 infection. J Infect Dis 2000; 182:950–954.
29. Lillo F, Ciuffreda D, Veglia F, Capiluppi B, Mastrorilli E, Vergani B, et al
. Viral load and burden modification following early antiretroviral therapy of primary HIV-1 infection. AIDS 1999; 13:791–796.
30. Antiretroviral Therapy Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy in high-income countries: a collaborative analysis of 14 cohort studies
31. Granich R, Gilks C, Dye C, De Cock K, Williams B. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: a mathematical model. Lancet 2008; 373:48–57.
32. Xiridou M, Geskus R, de Wit J, Coutinho R, Kretzschmar M. The contribution of steady and casual partnerships to the incidence of HIV infection among homosexual men in Amsterdam. AIDS 2003; 17:1029–1038.
antiretroviral therapy; heterosexuals; HIV; mathematical model; migrants; sexual risk behaviour
Supplemental Digital Content
© 2010 Lippincott Williams & Wilkins, Inc.
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.