Share this article on:

The origin and emergence of an HIV-1 epidemic: from introduction to endemicity

Bruhn, Christian A.W.a,b; Audelin, Anne M.c; Helleberg, Maried,e; Bjorn-Mortensen, Karenf; Obel, Nielsd; Gerstoft, Jand; Nielsen, Clausc; Melbye, Madsf; Medstrand, Patrikb; Gilbert, M. Thomas P.a; Esbjörnsson, Joakimb,g,h,i

doi: 10.1097/QAD.0000000000000198
Epidemiology and Social

Objectives: To describe, at patient-level detail, the determining events and factors involved in the development of a country's HIV-1 epidemic.

Design: Clinical information for all recorded Greenlandic HIV-1 patients was analysed and correlated with both novel and previously analysed pol sequences, representing more than half of the entire Greenlandic HIV-1 epidemic. Archival blood samples were sequenced to link early infection chain descriptions to the subsequent epidemic.

Methods: In-depth phylogenetic analyses were used in synergy with clinical information to assess number of introductions of HIV-1 into Greenland, the source of geographic origin, time of epidemic introduction and its epidemiological characteristics such as initial transmission chain, geographic dispersal within Greenland, method of infection, cluster size, sociological and behavioural factors.

Results: Despite its small population size and isolated geographic location, data support at least 25 introductions of HIV-1 into Greenland. Only a single of these led to an epidemic. This introduction occurred between 1985 and 1986, and the epidemic cluster is still active. Facilitating factors for the emergence and spread of the epidemic cluster include a rapid transition from MSM to heterosexual spread, high prevalence of other sexually transmitted diseases, rapid dispersal to larger cities and early emergence in a distinct subpopulation with high-risk behaviour including disregard for condomizing.

Conclusions: The synergistic use of disparate data categories yields such unique detail, that the Greenland epidemic now serves as a model example for the epidemic emergence of HIV-1 in a society. This renders it suitable for testing of present and future sequence-based epidemiological methodologies.

aCentre for GeoGenetics and Section for Evolutionary Genomics, Natural History Museum of Denmark, University of Copenhagen, Copenhagen, Denmark

bDepartment of Laboratory Medicine Malmö, Lund University, Malmö, Sweden

cMicrobiological Diagnostics and Virology, Statens Serum Institut, Copenhagen

dDepartment of Infectious Diseases, Copenhagen University Hospital, Rigshospitalet

eFaculty of Health and Medical Sciences, University of Copenhagen

fDepartment of Epidemiology Research, Statens Serum Institut, Copenhagen, Denmark

gNuffield Department of Medicine, NDM Research Building, University of Oxford, Oxford, UK

hDepartment of Microbiology, Tumor and Cell biology (MTC), Karolinska Institute, Stockholm, Sweden

iREGA Institute, Katholieke Universiteit, Leuven, Belgium.

Correspondence to Christian A.W. Bruhn, Centre for GeoGenetics, Oester Voldgade 5-7, 1350 Copenhagen K, Denmark. E-mail:

Received 3 November, 2013

Revised 19 December, 2013

Accepted 2 January, 2014

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (

Back to Top | Article Outline


Characterization of the natural history of HIV-1 in Greenland lends itself particularly well to understanding the emergence dynamics of this virus. The importance of the Greenlandic epidemic in this respect rests with the possibility of obtaining unprecedented detail of the emergence, and its determinants, covering an entire population. The Greenlandic population is geographically isolated and dispersed, and has a small and stable population size (between 55 000 and 57 000 for the past 20 years [1]). Indicators for levels of risk behaviour are available through decades of data from notifiable sexually transmitted diseases (STDs) [1–5], particularly gonorrhoea (see Figure, Supplemental Digital Content 1,, showing comparative gonorrhoea incidence over time). As an autonomous country within the Kingdom of Denmark, clinical and epidemiological information for all Greenlandic HIV-1 patients is furthermore available through the Danish HIV Cohort Study (DHCS) [6], and detailed records of the early part of the epidemic have been described in the literature [2,3,7] or can be found in archival records. Importantly, molecular sequences have also been collected from a large subset of the Greenlandic patients across a wide time span (Fig. 1). This makes the Greenlandic HIV-1 epidemic particularly suitable for studying factors important for the founding and emergence of an HIV-1 epidemic.

Fig. 1

Fig. 1

Along with the first HIV-1-positive tests recognized in the Greenlandic community in 1984 and 1985 [2], Greenland's extreme epidemiological figures for other STDs motivated a coordinated surveillance and prevention programme that was implemented in the late 1980s, with the principal aim of averting these and further introductions of HIV-1 from escalating into an epidemic [2,3,8,9]. For example, in 1982, the incidence of gonorrhoea in Greenland among adults between 15 and 59 years of age was 378.5 per 1000 individuals [3]. Initiated in 1986, the programme grew to include systematic contact tracing, HIV-test implementation, surveillance and statistics, information and prevention initiatives and use of several markers for trends in sexual behaviour [2,3,8–11]. By its conclusion in 1992, the project had uncovered the first larger transmission cluster in Greenland [2,3,7], and in the mid-1990s, other monitoring efforts identified further clusters within Greenland's largest towns, Sisimiut and Nuuk [12].

The study reports the amalgamation of data from all previous studies on HIV in Greenland, together with new information obtained from archival blood samples, archival clinical records and previously unpublished database sequences. When utilizing the synergy between thorough molecular analytical methods and comprehensive clinical and epidemiological information, this dataset both enables a detailed picture of emergence of HIV-1 in Greenland, and clearly defines the factors involved in the establishment of the epidemic. More generally, this contributes to our wider-scale understanding of the emergence dynamics of HIV-1.

Back to Top | Article Outline

Materials and methods

Study population and sequences

All patients who have been diagnosed with HIV and who resided in Greenland during a period between 1980 and 2011 were included in the study (n = 174; Fig. 1). Clinical information was obtained via the DHCS [6] and historical records. From this patient group, 98 pol sequences (norm length 1302 bases, HXB2 positions 2253–3554, which includes a short gag overlap) were compiled (Fig. 1) by including sequences from as many patients as possible, and letting them fulfil the following criteria: they were each from a unique patient; in cases where patients had sequences from multiple samples available, only the sequence with the oldest sampling date was included. Of the 98 sequences, 80 had been used in previous studies (sequence names within ‘GRL2-GRL128’, accession AM285216-AM285281 and AM937019-AM937032) [13,14], 15 were available through resistance testing databases located at Statens Serum Institut (SSI), Copenhagen, Denmark (sequences names GRL201, GRL203-GRL209 and GRL211-GRL217), and the final 3 sequences (GRL202, GRL210, GRL218) were acquired through location of archival blood samples at SSI from Greenlandic patients with early HIV-positive tests (before 1990). Additionally, a set of 1787 pol sequences from unique patients of the Danish SERO project [15] (a national surveillance project collecting samples from all newly diagnosed treatment-naive patients in mainland Denmark), spanning the years 2001–2012, were included for use as a local Basic Local Alignment Search Tool (BLAST) database [16]. All sequences were obtained as previously described [15].

Back to Top | Article Outline

Phylogenetic analysis

To estimate the origin and number of HIV-1 introductions into Greenland and their origin, the 98 pol sequences were compared against both a local database consisting of the 1787 Danish SERO study pol sequences [15], and GenBank [17] using BLAST [16]. In both cases, the 50 best matches for each query sequence were retained, ensuring that all available sequences related to the Greenland epidemic were included. Duplicate hits were removed and to reduce the risk of obtaining intra-host sequences, the dataset was further reduced using the program skipredundant.exe (threshold 99%) from the European Molecular Biology Open Software Suite (EMBOSS) package [18]. Four maximum likelihood phylogenetic trees were constructed, one for the full set (n-total = 1228, n-Greenland = 98), and three based on the subtypes present (subtype B: n-total = 838, n-Greenland = 94; subtype A1: n-total = 206, n-Greenland = 2; CRF11_cpx and CRF13_cpx: n-total = 94, n-Greenland = 2). Alignments were constructed using Multiple Alignment using Fast Fourier Transform (MAFFT) v. 7.017 [19] in Geneious v. 6.0.3 [20], with manual post-editing leaving all of them 1302 nucleotides in length and virtually un-gapped (except end-gaps). Maximum likelihood trees were constructed in Genetic Algorithm for Rapid Likelihood Inference (GARLI) v. 2.0 [21] using the GTR + I + G model with four gamma categories (3–6 search replicates). Support for clusters in the GARLI tree was estimated using Shimodaira–Hasegawa-like support [22] in PhyML v. 20110526 [23] using a fixed tree topology with branch length optimization. Clusters with Shimodaira–Hasegawa support values above 0.90 were considered well supported.

Next, the number of phylogenetically determined HIV-1 introductions was used to estimate the total number of introductions into Greenland by extrapolation. A cluster identity was assigned to each of the 98 pol sequences (resulting in 17 different cluster identities). This distribution was used to randomly sample (with replacement) a simulated dataset of cluster identities. A gradually increasing number of samples, from 1 to 98 in increments of 1, were drawn from the distribution. This was repeated 100 times at each increment allowing estimation of the average number of introductions at each increment. These estimates could then be used to fit a curve (log–log fit linear regression on log-transformed values) and estimate the total number of introductions (based on the 174 patients diagnosed with HIV-1).

Back to Top | Article Outline


Number of introductions and their origin

In total, 17 introductions were identified (Table 1, Fig. 2), 14 subtype B introductions, two of subtype A1, and one of the circulating recombinant form 13_cpx (CRF13_cpx). Eleven of the introductions consisted of single patients, five introductions had 2–5 patients and the remaining introduction consisted of 70 of the 98 total sequences [hereafter referred to as the Greenland epidemic cluster (GEC)] (Figs 1–3). In addition to the 70 Greenlandic sequences, two sequences collected on the Danish mainland from patients of Greenlandic origin fell in the GEC (information from the DHCS). This cluster alone represents the epidemic aspect of HIV-1 in Greenland. On the basis of these results, the actual number of introductions can be estimated statistically by extrapolation assuming knowledge of all HIV-1 infections. Taking into account the potential for a rarefaction effect on finding new introductions while sampling, we performed a simulation analysis based on the distribution of introductions among the 98 collected sequences. Extrapolating the fitted curve to the total number of HIV-1-infected individuals in Greenland resulted in a total estimate of 25 separate HIV-1 introductions [95% confidence interval (CI) 24–26]. Alternatively, patients’ self-reported location of infection – a category of the clinical information held at the DHCS [6] – can be used as a proxy for both number of introductions and their origin. This can be reported for the entire dataset of 174 patients or, because we can cross-reference identities, for the unsequenced set of 76 (174–98) patients alone. The distribution by DHCS geographical categories are the following (number of patients in full set/number of patients in unsequenced set): ‘Denmark’ (17/8), ‘Europe’ (3/–), ‘Africa’ (5/4), ‘Asia’ (4/4), ‘Other’ (1/–), ‘Unknown’ (8/8), information missing (3/3). The remaining (133/49) patients report ‘Greenland’ as the location of infection. Counting ‘Unknown’, missing information and ‘Greenland’ as zero yields an estimated 30 introductions for the full set and 16 for the unsequenced. Adding the latter to the result from the phylogeny yields a combined hybrid estimate of 33 (17 + 16) introductions, representing a third alternative estimate for the full set (see Discussion section). The preponderance of Danish introductions could also be seen in the phylogeny (Table 1, Fig. 2), where all introductions with strongly supported geographical associations were from Denmark. This was also the case for the GEC, which clustered basally, and was highly supported (Shimodaira–Hasegawa value 1.0), with a Danish cluster of heterosexual (HTX), MSM and intravenous drug using (IVDU) patients (Fig. 2b top cluster, Fig. 3 cluster marked ‘Denmark’). This establishes a phylogenetic association, which clearly demonstrates a close link between the Greenlandic epidemic and a distinct group within the epidemic on the Danish mainland. However, the original directionality of the link cannot be established from the tree alone.

Table 1

Table 1

Fig. 2

Fig. 2

Fig. 3

Fig. 3

Back to Top | Article Outline

Establishing the founding event of the epidemic

One of the archival blood samples (GRL218) belonged to a patient with the earliest HIV-1-positive date of all sequenced Greenlandic patients (DHCS). The sequence from this sample is a highly supported sister sequence to all other sequences in the GEC (Figs 2 and 3). Archival records and use of the DHCS allowed us to confirm that this person is identical to a patient mentioned in the existing literature, as the second link in the first major transmission chain established by contact tracing in Greenland [2,3,7] (Fig. 3). Obtaining the original contact tracing records tied phylogenetic, clinical and epidemiological information, and established exactly how, when and where the GEC started (Fig. 3). Inference based on included sequences, incidence data, patient information and phylogenetics shows that all later described larger transmission clusters (number of patients >5) [12] are descendants of this single introduction.

Back to Top | Article Outline

History of the Greenland epidemic cluster

Given the present results, the history of the epidemic cluster can now be retraced as follows (Fig. 3): an HIV-1 subtype B strain was introduced to Greenland no later than 1986 by a male patient to a township on its east coast. The patient had been homosexually infected in Denmark in 1985 and went on to infect two male contacts in Greenland, one between November and December 1986, and another between November 1986 and April 1987. Both of these men infected women, from which point the epidemic shifted to become HTX dominated. One of these women, having likely been infected in the summer of 1987, infected two men, one of whom was infected no later than November 1989 (Fig. 3). This man was from western Greenland, and thus the epidemic made an important geographic dispersal to the west coast establishing in Greenland's two largest cities, Sisimiut, and the capital of Nuuk. A distinction between Nuuk and Sisimiut clusters lies deep in the phylogeny but with indications of continuing interchange of infections between the two cities (Figs 2 and 3 and DHCS). Data indicate that prior to this, several other independent introductions had already occurred in west Greenland, predominantly Nuuk (DHCS). During 1992 and 1993, few new cases were reported (Fig. 1). Then clusters of infection were discovered in Sisimiut 1994/1995 (11 patients) and Nuuk 1995 (8 patients) and both cities 1997 (12 patients) [12]. Correlating data show that roughly half of these patients are included in the phylogenetic analysis, and that they belong to the GEC (Figs 1 and 2, and DHCS). These patients belonged to or had contact with a distinct social environment (see Discussion section). From around 1997, a steady rise in the number of deaths among sequenced GEC patients is seen; together with emigration and falling incidence this leads to a levelling out (and later a drop) in prevalence (not shown). Thus, the end of the 1990s marks a transition of the GEC between being epidemic and becoming endemic. Although HIV is almost absent from the east coast of Greenland at the current time, the GEC has continued to spread on the west coast (Fig. 2 and DHCS). A noteworthy recent development (within the last roughly 7 years) is the continuation of the GEC amongst younger patients in Nuuk (maximum 30 years of age at the time of diagnosis). These patients represent a group, which is a generation younger than the above-mentioned distinct GEC group (DHCS); whether they also represent a different social demographic is currently unknown.

Back to Top | Article Outline


There are two main findings of this study: exactly how, when, where and also from where a particular HIV-1 introduction established and subsequently rose to become epidemic in a country; and furthermore, how this introduction relates to other introductions into the same country, particularly their numbers.

The estimates of the total number of introductions are naturally prone to various forms of error and bias. Common to all these estimates given in ‘Results section’ is that they assume that all HIV-positive patients are known. For Greenland, this is not unreasonable, due to the history of HIV testing in this small population. There is a strong reason to believe that a very large fraction of all HIV-infected Greenlanders are known, and that any unknown infections would tend to have occurred recently [2,3,6,24]. Like several other sources of uncertainty, including the ‘Unknown’ and missing information categories in patients self-reporting, this can only lead to an underestimate. There are other errors tied to self-reporting; however, for location of infection (unlike, e.g. method of infection), there is no obvious reason to assume a strong bias in either direction. The estimate based on extrapolating the phylogenetic result is likely to be conservative because it assumes that the 98 sequences are a representative sampling. But in fact, the dataset includes sequences from all patients of the DHCS for whom one was in any way obtainable. Because of historical factors such as introduction of routine antiretroviral resistance testing, this leads to bias towards sampling patients with later HIV-1-positive dates (see Fig. 1). At any low prevalence stage, including very early in an epidemic, any new infection will have an increased expectation of being due to an introduction. This is also evident in the reporting. It furthermore explains why the hybrid estimate of 33 is higher because it combines the high level of reports of infection abroad amongst the early primarily unsequenced patients with the phylogenetic analysis’ ability to catch an infection as part of a separate introduction, even though the introducing patient is unaccounted for. In general, the differences between the uncertainties in reporting and phylogenetics mean that there is not a perfect 1 : 1 relationship between the results. Taking the above and the results in concert means that we are confident in stating that 25 is a minimum assessment for the number of introductions of HIV into Greenland.

The Greenlandic HIV-1 epidemic clearly demonstrates that even for a population with a history of high-risk sexual behaviour, the risk of an HIV-1 introduction establishing itself as an epidemic can be smaller than or equal to 1 in 25. Related to this observation are several key issues. Firstly, this estimate represents an average over roughly the past 3 decades, during which several key parameters have changed. Secondly, a population with high prevalence and incidence figures for any of several STDs is at increased risk for an HIV epidemic emerging for at least two significant and distinct reasons: it is indicative of high levels of sexual risk behaviour [2,3,25,26], and it increases the population average level of transmission risk [27–35]. Using gonorrhoea incidence as an indicator, it is clear that such parameters can vary dramatically in time, for Greenland, by a factor of almost 50 (1982 vs. 1995; see Figure, Supplemental Digital Content 1,, showing comparative gonorrhoea incidence over time), and hence overall HIV emergence risk can vary along with them. Thirdly, the nature of the given transmission network [36–41], including its associated geographical and demographic parameters, will be expected to affect the ratio between number of introductions and founding events. Here, Greenland is characterized by not having an IVDU structure and by a small and widely dispersed population.

Accepting the above notions, Greenland was highly susceptible to an introduction of HIV leading to an epidemic in the 1980s, based on the extreme incidence figures from notifiable gonorrhoea cases (see Figure, Supplemental Digital Content 1,, showing comparative gonorrhoea incidence over time). This began changing from the mid-1980s, with incidence dropping almost continually until reaching a low in 1995, at a level almost 50 times below the one in 1982 and 1983 – possibly as a consequence of intense anti-HIV campaigns. It is thus noteworthy that our analyses demonstrate that the introduction, which subsequently seeded the epidemic, occurred between 1985 and 1986, and was firmly established by 1990, when gonorrhoea incidence was still very high. It also follows that introductions into Greenland in the 1990s had a significantly reduced risk of establishing an epidemic, and indeed, the analyses show that none of them have (Table 1, Fig. 2). Why then, did only a single introduction, of the several known to have taken place in the 1980s, become epidemic? One explanation could be that a significant component of the earliest HIV-positive tests in Greenland is from patients reported to be MSM infected abroad (DHCS, [6]). Even though the first onwards transmissions of the epidemic cluster within Greenland occurred by MSM contact (Fig. 3), it likely required bisexual behaviour for any larger emergence to establish, simply because of the smaller size of the MSM community in Greenland. This could be a determining reason for why the majority of introductions in the 1980s did not escalate. Thus, because the Greenlandic HTX population was itself a high-risk population in general at the time (see Figure, Supplemental Digital Content 1,, showing comparative gonorrhoea incidence over time), the transition from MSM to HTX was a significant early event in the history of the Greenland epidemic cluster (Fig. 3). A second important event was the geographical dispersal of infection to the largest towns in Greenland (Fig. 3). The urban relevance is substantiated by the fact that east Greenland, with its significantly smaller communities, where the GEC was introduced, has virtually no HIV cases today, whereas the GEC is still ongoing in those two largest Greenlandic towns (DHCS). The introduction into Nuuk and Sisimiut also occurred in a distinct social group characterized by unemployment, poor or non-existing living quarters, alcohol abuse, partial isolation from the rest of society, recent or ongoing histories for other STDs, a close-knit sexual and social network, initially poor or no knowledge of HIV/AIDS, and failure to heed advice on barrier protection [12,42]. These aspects tempt the explanation that the introduction into this group set the GEC apart from other HIV introductions in Greenland based on risk behaviour at that point in time when indication from gonorrhoea incidence shows the continuation of a strong downward trend in sexual risk behaviour in the general population. Such points highlight the chance aspect between outcomes of individual introductions. The fewer steps in a transmission network an introduction has to take to reach any demographic, which can maintain an effective reproductive number above 1, the more likely it will be to emerge.

Importantly, the unique detail level of the Greenland dataset now renders it suitable to test methods for sequence-based epidemiological parameter estimates, in order to assess the impact of existing model assumptions on datasets where detailed clinical and epidemiological information is scarce. It can thus serve as a strong supplement to simulation in this regard.

In conclusion, this description of the Greenland epidemic and the GEC underlines that HIV-1 evolutionary and epidemiological dynamics are governed by bottleneck effects at multiple levels. From the emergence of HIV from Simian Immunodeficiency Virus (SIV), to the geographical founding events of the subtypes, of which the GEC is a small but now uniquely well described and documented example, over the determining jumps between demographic groups – the bottleneck is a unifying characteristic. Thus, in a similar manner to the increasingly acknowledged importance of bottlenecks at the time of infection and its potential for prophylactic strategies, the sometimes narrow links between determining demographic groups deserve a significant amount of attention when formulating an overall strategy against HIV/AIDS.

Back to Top | Article Outline


M.T.P.G., C.A.W.B. and J.E. devised the study, with additional input from all authors. A.M.A. and C.N. provided archival blood samples, performed sequencing and provided SERO sequences. A.M.A., M.H., K.B.M., N.O., J.G., C.N. and M.M provided necessary clinical and epidemiological information. C.A.W.B. led and performed most of the analyses, with additional assistance from J.E. and P.M. The extrapolation analysis was designed and conducted by J.E. with assistance from Dr Anders Kvist and Per-Erik Isberg. C.A.W.B wrote the manuscript with M.T.P.G. and J.E. All authors performed critical review of the manuscript.

DHCS data have been accessed under approval 2012-41-0067 from the Danish Data Protection Agency.

Back to Top | Article Outline

Conflicts of interest

Source of funding: C.A.W.B. and M.T.P.G. are supported by Lundbeckfonden grant R52-A5062 for this work ( J.E. is supported by the Swedish Research Council. For the remaining authors none were declared.

Back to Top | Article Outline


1. Statistics Greenland. The StatBank. [Accessed 28 August 2013]
2. Melbye M, Misfeldt J, Olsen J. AIDS/HIV Situationen i Grønland 1989: Status, målsætninger og anbefalinger [The AIDS/HIV situation in Greenland 1989: status, objectives and recommendations] [report]. Nakorsaaneqarfik, Landslægeembedet: Atuakkiorfik; 1989.
3. Misfeldt J, Olsen J, Melbye M. AIDS/HIV Situationen i Grønland 1991: Status rapport pr. 30 September 1991 [The AIDS/HIV situation in Greenland: status report per September 30th 1991] [report]. Nakorsaaneqarfik, Landslægeembedet: Atuakkiorfik; 1991.
4. Statens Serum Institut. Infektionssygdomme i Grønland, Del II [Infectious diseases in Greenland, Part II] [Newsletter]. In EPI-NYT week 8, 2003.∼/media/Indhold/DK-dansk/Aktuelt/Nyhedsbreve/EPI-NYT/EPI-NYT-Arkiv/2003/2003pdf/EPI-NYT-2003-uge8.ashx.
5. Government of Greenland. Årsberetning Landslægeembeddet (2001–2010) [National yearly health report].
6. Obel N, Engsig FN, Rasmussen LD, Larsen MV, Omland LH, Sorensen HT. Cohort profile: the Danish HIV Cohort Study. Int J Epidemiol 2009; 38:1202–1206.
7. Moi H, Misfeldt JC, Olsen J, Melbye M. Kaposis-sarcoma in HIV-1-positive heterosexual eskimo. Lancet 1993; 342:1298–11298.
8. Misfeldt J. Sexually transmitted diseases in Greenland. A statement with proposals for new strategies. Arctic Med Res 1988; 47 (Suppl 1):675–678.
9. Misfeldt JC. Dramatic decline in the incidence of gonorrhea and syphilis in Greenland: result of an intervention strategy?. Ugeskr Laeger 1994; 156:4690–4694.
10. Misfeldt JC, Senderovitz F, Melbye M, Olsen J. Sexual behavior and knowledge about AIDS, an investigation among young people in Greenland Arctic Ocean. Ugeskr Laeger 1989; 151:2715–2719.
11. Misfeldt JC, Werdelin L, Senderovitz F, Melbye M, Olsen J. The sexual habits of young Greenlanders and their knowledge about AIDS, an investigation among students in vocational colleges in 1989. Ugeskr Laeger 1990; 152:237–239.
12. Winthereik M. The spread of HIV in Greenland. Heterosexual epidemic--risk or reality? A 10-year review of HIV transmission and preventive care. Ugeskr Laeger 1998; 160:2851–2855.
13. Madsen TV, Leitner T, Lohse N, Obel N, Ladefoged K, Gerstoft J, et al. Introduction of HIV type 1 into an isolated population: molecular epidemiologic study from Greenland. AIDS Res Hum Retroviruses 2007; 23:675–681.
14. Madsen TV, Lohse N, Jensen ES, Obel N, Ladefoged K, Gerstoft J, et al. High prevalence of drug-resistant human immunodeficiency virus type 1 in treatment-naive patients in Greenland. Aids Res Hum Retroviruses 2008; 24:1073–1077.
15. Audelin AM, Gerstoft J, Obel N, Mathiesen L, Laursen A, Pedersen C, et al. Molecular phylogenetics of transmitted drug resistance in newly diagnosed HIV type 1 individuals in Denmark, a nation-wide study. Aids Res Hum Retroviruses 2011; 27:1284–1291.
16. Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997; 25:3389–3402.
17. Benson DA, Cavanaugh M, Clark K, Karsch-Mizrachi I, Lipman DJ, Ostell J, et al. GenBank. Nucleic Acids Res 2013; 41 (Database issue):D36–D42.
18. EMBOSS skipredundant. [Accessed 16 October 2012]
19. Katoh K, Misawa K, Kuma K, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res 2002; 30:3059–3066.
20. Biomatters. Geneious version 6.0.3.
21. Zwickl DJ. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion [PhD dissertation]. The University of Texas at Austin; 2006.
22. Shimodaira H, Hasegawa M. Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Mol Biol Evol 1999; 16:1114–1116.
23. Guindon S, Gascuel O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol 2003; 52:696–704.
24. Bjorn-Mortensen K, Ladefoged K, Obel N, Helleberg M. The HIV epidemic in Greenland: a slow spreading infection among adult heterosexual Greenlanders. Int J Circumpolar Health 2013; 72:19558.
25. Hewett PC, Mensch BS, Ribeiro M, Jones HE, Lippman SA, Montgomery MR, et al. Using sexually transmitted infection biomarkers to validate reporting of sexual behavior within a randomized, experimental evaluation of interviewing methods. Am J Epidemiol 2008; 168:202–211.
26. Misfeldt JC, From E, Olsen J, Melbye M. The incidence of gonorrhea as an indicator of AIDS prophylaxis in Greenland. Ugeskr Laeger 1990; 152:2171–2174.
27. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75:3–17.
28. Galvin SR, Cohen MS. The role of sexually transmitted diseases in HIV transmission. Nat Rev Microbiol 2004; 2:33–42.
29. Malott RJ, Keller BO, Gaudet RG, McCaw SE, Lai CCL, Dobson-Belaire WN, et al. Neisseria gonorrhoeae-derived heptose elicits an innate immune response and drives HIV-1 expression. Proc Natl Acad Sci U S A 2013; 110:10234–10239.
30. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li CJ, Wabwire-Mangen F, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. N Engl J Med 2000; 342:921–929.
31. Cohen MS, Hoffman IF, Royce RA, Kazembe P, Dyer JR, Daly CC, et al. Reduction of concentration of HIV-1, in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV-1. Lancet 1997; 349:1868–1873.
32. Castel AD, Befus M, Willis S, Griffin A, West T, Hader S, et al. Use of the community viral load as a population-based biomarker of HIV burden. AIDS 2012; 26:345–353.
33. Das M, Chu PL, Santos GM, Scheer S, Vittinghoff E, McFarland W, et al. Decreases in community viral load are accompanied by reductions in new HIV infections in San Francisco. PLoS One 2010; 5:e11068.
34. Das M, Raymond HF, Chu P, Nieves-Rivera I, Pandori M, Louie B, et al. Measuring the unknown: calculating community viral load among HIV-infected MSM unaware of their HIV status in San Francisco from national HIV behavioral surveillance, 2004–2011. J Acquir Immune Defic Syndr 2013; 63:E84–E86.
35. Miller WC, Powers KA, Smith MK, Cohen MS. Community viral load as a measure for assessment of HIV treatment as prevention. Lancet Infect Dis 2013; 13:459–464.
36. Gupta S, Anderson RM, May RM. Networks of sexual contacts: implications for the pattern of spread of HIV. AIDS 1989; 3:807–817.
37. Morris M. Sexual networks and HIV. AIDS 1997; 11:S209–S216.
38. Ghani AC, Swinton J, Garnett GP. The role of sexual partnership networks in the epidemiology of gonorrhea. Sex Transm Dis 1997; 24:45–56.
39. Aral SO, Hughes JP, Stoner B, Whittington W, Handsfield HH, Anderson RM, et al. Sexual mixing patterns in the spread of gonococcal and chlamydial infections. Am J Public Health 1999; 89:825–833.
40. Garnett GP, Anderson RM. Contact tracing and the estimation of sexual mixing patterns - the epidemiology of gonococcal infections. Sex Transm Dis 1993; 20:181–191.
41. Hue S, Pillay D, Clewley JP, Pybus OG. Genetic analysis reveals the complex structure of HIV-1 transmission within defined risk groups. Proc Natl Acad Sci U S A 2005; 102:4425–4429.
42. Lohse N, Ladefoged K, Pedersen L, Jensen-Fangel S, Sorensen HT, Obel N. Low effectiveness of highly active antiretroviral therapy and high mortality in the Greenland HIV-infected population. Scand J Infect Dis 2004; 36:738–742.

epidemics; epidemiology; founder effect; Greenland; molecular epidemiology; natural history; surveillance

Supplemental Digital Content

Back to Top | Article Outline
© 2014 Lippincott Williams & Wilkins, Inc.