Does concurrency explain the heterosexual HIV epidemic in Sub-Saharan Africa? Lessons from sero-discordant couples.
Paul Arora, Centre for Global Health Research, St. Michael's Hospital, University of Toronto
28 October 2010
A recent paper in this journal critically reviewed the theory that the heterosexual epidemic in Sub-Saharan Africa (SSA) is explained by high levels of concurrent partnerships in SSA[1]. Much of this theory is based on questionable data, unrealistic computer simulations and the almost impossible task of distinguishing concurrency in the narrow sense, i.e. simultaneous long term partnerships, and concurrency in the wider sense, i.e. any form of infidelity. A better understanding of the HIV epidemic in SSA can be achieved by referring to more reliable biological data. Here, using such data from the recent “Partners Study”, we aim to illustrate the importance of concurrency/infidelity to the HIV epidemic in SSA by comparing it to the counterfactual situation in which concurrency is absent, i.e. serial monogamy.
The average duration from HIV infection to death from AIDS (in the absence of treatment) is approximately 10 years. Thus, for an HIV epidemic to occur in a serially monogamous population, transmission between discordant couples would need to be at least 10% per person-year.
Most estimates of transmission risk from discordant couples studies are much lower. The recent “Partners Study” among discordant couples in seven SSA countries found that the risk of infection of the uninfected partner was only 2.7 per 100 person years, and only two-thirds (84/132) of the incident infections were attributable to transmission from the infected partner[2]. Undoubtedly, follow-up of couples selected for being discordant may underestimate the risk of transmission. However, in a review of sero-discordant couples from five cross-sectional demographic health surveys from various parts of SSA, only about 25% of all infected couples were concordant positive, and not all concordance may have occurred as a result of transmission from the infected spouse[3]. Although we do not know the duration of exposure of the uninfected partners, these numbers strongly suggest that the average transmission risk is much less than the 10% per person year required for sustaining epidemics.
If average HIV transmission risk is less than 10% per person year, what explains the generalized heterosexual epidemic in SSA? The answer is suggested by the Partners Study. Of the 132 HIV infections in uninfected partners, over one third (48), were not virologically linked to the infected partner. The Rakai study similarly reported a large percentage of infections originating outside of partnership, although interpretation is complicated by large numbers of indeterminate viral linkages[4]. If we assume that the sexual behaviour of uninfected partners is roughly representative of all uninfected individuals in a stable partnership then, in the populations in which the Partners Study was carried out, such individuals have an exposure to HIV approximately equal to half the exposure of a faithful uninfected partner of an HIV-infected person. Assume for the sake of the argument that HIV prevalence in the population from which participants of the Partners Study were recruited is 10% (in fact 2007 Adult HIV prevalence estimates for the 7 participating countries range from 2.8% to 23.9% with an unweighted mean of 11.5%[5]) and that everybody has a steady partner (spouse). Then over 80% of the newly infected acquired their infection from someone other than their spouse. The importance of this extra-marital transmission becomes even more apparent if transmitters are used as the denominator. Then these 10% account, (in addition to the risk they pose to their spouses), for ALL the extra-marital risk in the 90% of uninfected individuals. That is, these 10% of infected individuals are effectively participating in approximately five concurrent discordant relationships simultaneously.
Clearly, while it is impossible to identify the duration of the relationships in which this extra-marital transmission took place, infidelity, i.e. concurrency in the wider sense, does appear to play a key role.
Nico Nagelkerke 1,2, Paul Arora 3, Prabhat Jha 3
1. Department of Community Medicine, United Arab Emirates University, Al Ain, UAE 2. Department of Public Health, Erasmus Medical Centre, Rotterdam, the Netherlands 3. Centre for Global Health Research, University of Toronto, Toronto, Canada
References 1. Sawers L, Stillwaggon E: Concurrent sexual partnerships do not explain the HIV epidemics in Africa: a systematic review of the evidence. J Int AIDS Soc 2010, 13:34. 2. Celum C, Wald A, Lingappa JR, Magaret AS, Wang RS, Mugo N, Mujugira A, Baeten JM, Mullins JI, Hughes JP, Bukusi EA, Cohen CR, Katabira E, Ronald A, Kiarie J, Farquhar C, Stewart GJ, Makhema J, Essex M, Were E, Fife KH, de Bruyn G, Gray GE, McIntyre JA, Manongi R, Kapiga S, Coetzee D, Allen S, Inambao M, Kayitenkore K, Karita E, Kanweka W, Delany S, Rees H, Vwalika B, Stevens W, Campbell MS, Thomas KK, Coombs RW, Morrow R, Whittington WL, McElrath MJ, Barnes L, Ridzon R, Corey L, Partners in Prevention HSV/HIV Transmission Study Team: Acyclovir and transmission of HIV-1 from persons infected with HIV-1 and HSV-2. N Engl J Med 2010, 362:427-439. 3. DeWalque D: Sero-Discordant Couples in Five African Countries: Implications for Prevention Strategies. Population and Development Review 2007, 33:501-523. 4. Hollingsworth TD, Laeyendecker O, Shirreff G, Donnelly CA, Serwadda D, Wawer MJ, Kiwanuka N, Nalugoda F, Collinson-Streng A, Ssempijja V, Hanage WP, Quinn TC, Gray RH, Fraser C: HIV-1 transmitting couples have similar viral load set-points in Rakai, Uganda. PLoS Pathog 2010, 6:e1000876. 5. Health Topics: HIV/AIDS [http://www.who.int/topics/hiv_aids/en/].
Does concurrency explain the heterosexual HIV epidemic in Sub-Saharan Africa? Lessons from sero-discordant couples.
28 October 2010
A recent paper in this journal critically reviewed the theory that the heterosexual epidemic in Sub-Saharan Africa (SSA) is explained by high levels of concurrent partnerships in SSA[1]. Much of this theory is based on questionable data, unrealistic computer simulations and the almost impossible task of distinguishing concurrency in the narrow sense, i.e. simultaneous long term partnerships, and concurrency in the wider sense, i.e. any form of infidelity. A better understanding of the HIV epidemic in SSA can be achieved by referring to more reliable biological data. Here, using such data from the recent “Partners Study”, we aim to illustrate the importance of concurrency/infidelity to the HIV epidemic in SSA by comparing it to the counterfactual situation in which concurrency is absent, i.e. serial monogamy.
The average duration from HIV infection to death from AIDS (in the absence of treatment) is approximately 10 years. Thus, for an HIV epidemic to occur in a serially monogamous population, transmission between discordant couples would need to be at least 10% per person-year.
Most estimates of transmission risk from discordant couples studies are much lower. The recent “Partners Study” among discordant couples in seven SSA countries found that the risk of infection of the uninfected partner was only 2.7 per 100 person years, and only two-thirds (84/132) of the incident infections were attributable to transmission from the infected partner[2]. Undoubtedly, follow-up of couples selected for being discordant may underestimate the risk of transmission. However, in a review of sero-discordant couples from five cross-sectional demographic health surveys from various parts of SSA, only about 25% of all infected couples were concordant positive, and not all concordance may have occurred as a result of transmission from the infected spouse[3]. Although we do not know the duration of exposure of the uninfected partners, these numbers strongly suggest that the average transmission risk is much less than the 10% per person year required for sustaining epidemics.
If average HIV transmission risk is less than 10% per person year, what explains the generalized heterosexual epidemic in SSA? The answer is suggested by the Partners Study. Of the 132 HIV infections in uninfected partners, over one third (48), were not virologically linked to the infected partner. The Rakai study similarly reported a large percentage of infections originating outside of partnership, although interpretation is complicated by large numbers of indeterminate viral linkages[4]. If we assume that the sexual behaviour of uninfected partners is roughly representative of all uninfected individuals in a stable partnership then, in the populations in which the Partners Study was carried out, such individuals have an exposure to HIV approximately equal to half the exposure of a faithful uninfected partner of an HIV-infected person. Assume for the sake of the argument that HIV prevalence in the population from which participants of the Partners Study were recruited is 10% (in fact 2007 Adult HIV prevalence estimates for the 7 participating countries range from 2.8% to 23.9% with an unweighted mean of 11.5%[5]) and that everybody has a steady partner (spouse). Then over 80% of the newly infected acquired their infection from someone other than their spouse. The importance of this extra-marital transmission becomes even more apparent if transmitters are used as the denominator. Then these 10% account, (in addition to the risk they pose to their spouses), for ALL the extra-marital risk in the 90% of uninfected individuals. That is, these 10% of infected individuals are effectively participating in approximately five concurrent discordant relationships simultaneously.
Clearly, while it is impossible to identify the duration of the relationships in which this extra-marital transmission took place, infidelity, i.e. concurrency in the wider sense, does appear to play a key role.
Nico Nagelkerke 1,2, Paul Arora 3, Prabhat Jha 3
1. Department of Community Medicine, United Arab Emirates University, Al Ain, UAE
2. Department of Public Health, Erasmus Medical Centre, Rotterdam, the Netherlands
3. Centre for Global Health Research, University of Toronto, Toronto, Canada
References
1. Sawers L, Stillwaggon E: Concurrent sexual partnerships do not explain the HIV epidemics in Africa: a systematic review of the evidence. J Int AIDS Soc 2010, 13:34.
2. Celum C, Wald A, Lingappa JR, Magaret AS, Wang RS, Mugo N, Mujugira A, Baeten JM, Mullins JI, Hughes JP, Bukusi EA, Cohen CR, Katabira E, Ronald A, Kiarie J, Farquhar C, Stewart GJ, Makhema J, Essex M, Were E, Fife KH, de Bruyn G, Gray GE, McIntyre JA, Manongi R, Kapiga S, Coetzee D, Allen S, Inambao M, Kayitenkore K, Karita E, Kanweka W, Delany S, Rees H, Vwalika B, Stevens W, Campbell MS, Thomas KK, Coombs RW, Morrow R, Whittington WL, McElrath MJ, Barnes L, Ridzon R, Corey L, Partners in Prevention HSV/HIV Transmission Study Team: Acyclovir and transmission of HIV-1 from persons infected with HIV-1 and HSV-2. N Engl J Med 2010, 362:427-439.
3. DeWalque D: Sero-Discordant Couples in Five African Countries: Implications for Prevention Strategies. Population and Development Review 2007, 33:501-523.
4. Hollingsworth TD, Laeyendecker O, Shirreff G, Donnelly CA, Serwadda D, Wawer MJ, Kiwanuka N, Nalugoda F, Collinson-Streng A, Ssempijja V, Hanage WP, Quinn TC, Gray RH, Fraser C: HIV-1 transmitting couples have similar viral load set-points in Rakai, Uganda. PLoS Pathog 2010, 6:e1000876.
5. Health Topics: HIV/AIDS [http://www.who.int/topics/hiv_aids/en/].
Competing interests
We declare no competing interests.