Herpes simplex virus type 2 (HSV-2), the main cause of genital herpes, facilitates sexual acquisition of HIV-1 [1–3] and may also facilitate sexual transmission of HIV-1 from an HIV-1 positive to negative partner . To date there is little evidence as to whether maternal HSV-2 co-infection increases risk of intra-partum mother to child transmission (MTCT) of HIV-1. Such excess risk might be mediated by HSV-2 co-infection causing greater HIV-1 genital shedding, as has been reported in some studies [5,6]. Recently, Nagot and colleagues reported that administration of valaciclovir (which suppresses subclinical and clinical reactivation of HSV-2) to HIV-1 positive women not only reduced the frequency and magnitude of HIV-1 genital shedding but also modestly reduced HIV-1 plasma viral load , providing another potential mechanism through which HSV-2 could exacerbate intra-partum HIV-1 MTCT. Finally, one study from the US reported that clinical genital herpes during pregnancy (but not asymptomatic HSV-2 seropositivity) was associated with a greater risk of HIV-1 MTCT .
It is likely that genital ulcer disease caused by syphilis also facilitates sexual transmission of HIV-1 [9,10], but little is known about whether or not active maternal syphilis increases MTCT.
The ZVITAMBO study was a randomized placebo-controlled trial that enrolled post-partum mothers and their neonates in Harare, Zimbabwe to assess the impact of post-partum maternal and neonatal vitamin A supplementation on several health outcomes [11–13]. There was no effect of vitamin A on HIV-1 MTCT overall or during the intra-partum period . The ZVITAMBO study generated a specimen archive that provided an opportunity to examine whether maternal HSV-2 and syphilis infection is a risk factor for intra-partum MTCT of HIV-1.
The specific objectives of this study were to determine whether the risk of intra-partum transmission of HIV-1 was increased among HIV-1 positive women who: (i) had HSV-2 antibodies at delivery (prevalent HSV-2 infection); (ii) acquired HSV-2 antibodies within 6 weeks of delivery (incident HSV-2 infection); or (iii) had serological evidence of active syphilis at delivery.
The ZVITAMBO study recruited 14 110 mother–infants pairs within 96 h post-partum between 1997 and 2000; 4495 of women tested HIV-1 positive at recruitment using an algorithm of two ELISA assays run in parallel with discordant samples resolved by Western blot . HIV-1 positive women were retested for HIV-1 at their next blood draw to confirm the diagnosis. At birth, CD4 cells were counted using flow cytometry; haemoglobin and plasma HIV-1 viral load were measured among subsets of women. HIV-1 positive mothers and their babies had blood taken at birth, 6 weeks, 3 months, and then 3 monthly. Maternal sera were archived at each time point. Plasma and cell pellets were prepared from infant blood at all visits and stored at −70°C. After follow-up was completed, the last available specimen from each infant was tested (plasma by ELISA for samples collected ≥18 months; or cell pellets by DNA PCR for samples collected <18 months). If this sample was HIV-1 negative, the infant was classified as negative; if it was HIV-1 positive, samples collected at younger ages were tested to determine timing of infection. Infants testing PCR-negative at birth and PCR-positive at 6 weeks were classified as having been infected during the intra-partum period.
A nested case–control design was used. Cases (n = 509) comprised all HIV-1 positive women enrolled in the ZVITAMBO study whose baby became HIV-1 infected during the intra-partum period. Controls (n = 2148) were all HIV-1 positive women whose babies remained HIV-1 uninfected to ≥ 12 months of age. For each case, two potential controls (n = 1018) were selected by simple random sampling.
Archived maternal sera collected at recruitment were tested for HSV-2 antibodies using HerpeSelect HSV-2 ELISA (Focus Diagnostics, Cypress, California, USA). Samples that were indeterminate (index 0.9–1.1) and low positive (index 1.1–3.5) were retested in duplicate using HerpeSelect and the final result was declared by two of the three ELISAs having a concordant result. For women testing HSV-2 seronegative at birth, the 6-week sample was tested for HSV-2 antibodies as previously outlined to identify those who seroconverted around delivery. Dried blood spots (DBS) of 144 samples (132 baseline samples plus 12 6-week samples) that remained indeterminate or low positive were sent to University of Washington, USA for testing using a Western blot assay . Western blot resolved only 19.4% of the samples, failing to confirm or refute infection in the remainder. All indeterminate and low positive samples were therefore re-tested using the BioElisa HSV-2 IgG type specific assay (Biokit, Barcelona, Spain). The final HSV-2 sero-status was determined according to the following strategy formulated by the University of Washington (Fig. 1): (i) if the HerpeSelect result was positive or negative then the final HSV-2 result was declared as such; (ii) if the HerpeSelect result was either low positive, indeterminate or missing, but with a Western blot result positive or HSV-2 atypical, then the final result was declared as positive irrespective of the BioElisa result; (iii) if the HerpeSelect result was either low positive, indeterminate or missing, but with a Western blot result negative, indeterminate or missing, then the final result was declared using the BioElisa; (iv) if the HerpeSelect result was either low positive, indeterminate or missing and there was no result for both Western blot and BioElisa then the sero-status remained unresolved.
Inter-assay reliability was assessed using a random sample of 20 positives (index > 3.5) and 20 negatives (index < 0.9) on HerpeSelect which were retested using both Western blot and the BioElisa assays. Intra-assay reproducibility was assessed only for the HerpeSelect test using 256 samples which were repeated in duplicate because their initial index value was either in the indeterminate or low positive range (0.9–3.5). Kappa values were calculated for both inter- and intra-assay agreement using methods described by Baker and Freedman .
Plasma HIV-1 RNA viral load was measured using Roche Amplicor HIV-1 Monitor test version 1.5 (Roche Diagnostics, Alameda, California, USA) on a random subsample of 444 HIV-positive women in the larger trial, 163 of whom were selected for this study. A further 165 women in this study had viral load tested as part of other substudies and therefore may not be representative of all HIV-positive women in the study. Infant plasma was tested using GeneScreen HIV1/2 (Sanofi Diagnostics Pasteur, Johannesburg, South Africa) and infant cell pellets were tested using Amplicor HIV-1 DNA test version 1.5 (Roche Diagnostics).
In mothers enrolled from 1 October 1998 to the end of the study (∼60% of total study population), haemaglobin concentrations were measured using a hemaglobinometer (HemoCue, Mission Viejo, California, USA) .
Active maternal syphilis at delivery was determined by testing the 6-week sample for RPR (SPINREACT RPR-Carbon Slide Agglutination, Santa Coloma, Spain). Positive samples were confirmed using TPHA (SPINREACT TPHA, Santa Coloma, Spain). All women with evidence of active syphilis at 6 weeks (RPR titres >1: 8 and positive TPHA) were assumed to have had incubating or active syphilis at delivery.
It was anticipated that at least 80% of women in the control group were co-infected with HSV-2. The study had over 90% power to detect a 50% increase in the risk of intra-partum transmission associated with maternal HSV-2 infection with 500 cases and 1000 controls at 5% significance.  If risk of maternal acquisition of HSV-2 in the peri-partum period were 1% in the control arm (equates with annual incidence of 6%) the study has 80% power to show a 300% increase in risk of intra-partum HIV-1 transmission associated with incident HSV-2 infection (P < 0.05).
With regards to syphilis, it was anticipated that 5% of women in the control group would have evidence of syphilis at the time of delivery. The study had 90% power to detect a twofold increase in the risk of intra-partum transmission associated with maternal syphilis infection with 500 cases and 1000 controls at 5% significance .
Data management and statistical analysis
All data from the microplate ELISA reader were entered into a dBase PLUS 2.01 database by direct transfer. Western blot results were captured manually. Range and consistency checks were performed. Data was analysed using STATA version 8.0. The unadjusted odds ratios (OR) for intra-partum HIV-1 transmission by prevalent and incident maternal HSV-2 status and active maternal syphilis were calculated using univariate logistic regression. Adjusted OR were calculated using forward step-wise multivariate logistic regression with other risk factors found to be significantly associated with intra-partum HIV-1 transmission offered in the model and retained if α ≤ 0.05. The adjusted odds were calculated twice, first including all 328 viral loads and then once restricting the analysis to the 163 randomly selected samples. A further adjusted analysis was carried out in which both HSV-2 and syphilis were included in the same model. Sensitivity analyses were performed to explore the implications of assigning HSV-2 antibody sero-status for low positive and indeterminate samples according to different testing algorithms: (i) using HerpeSelect and Western blot results only; (ii) using HerpeSelect and BioELisa results only; and (iii) using all three assays.
The population attributable fraction is defined as the proportion of disease cases over a specified time that would be prevented following elimination of the exposures, assuming the exposures are casual,  and for a single binary exposure risk factor PAF is defined as:
where P x is the proportion of exposure in the population and RR is the risk ratio associated with the risk factor. For case–control studies with appropriate design P x is estimated from control subjects while RR is replaced with the OR . In this study, PAF, confidence intervals and P-values were calculated using the Stata macro aflogit [20,21], which uses the above Eq. (1).
Role of the funding source
Donors were not involved in the design, conduct, analysis, or reporting of the study.
When cases were compared to controls they were significantly less educated and had more signs of advanced HIV-1 disease (lower CD4 cell count, lower haemoglobin, higher viral load, smaller arm circumference – factors that were associated with mortality among HIV-positive women in ZVITAMBO) (Table 1).  Infants of cases had significantly lower birth weight and gestational age compared to those of controls.
Figure 1 outlines the results of HSV-2 testing at various stages in the testing algorithm. The final HSV-2 results at delivery were 1195 positive, 253 negative, 4 indeterminate, 34 low positive and 41 women had no HSV-2 result. The 79 women without clear positive or negative HSV-2 results at delivery were excluded from the HSV-2 prevalence analysis leaving 1448, comprising 478 cases and 970 controls.
The HerpesSelect ELISA intra-assay proportion of agreement was 84.8%, kappa value (κ) 0.72 (SE, 0.051) and P < 0.0001. For inter-assay agreement among the three assays, the mean κ was 0.84 and P < 0.0001; agreement on HSV-2 positives was κ 0.83, P < 0.000 and HSV-2 negatives κ 0.87, P < 0.000. The proportion of agreement between HerpeSelect ELISA and Western blot was 92.5%, κ 0.85 (SE, 0.16) and P < 0.0001. Between HerpeSelect ELISA and BioELISA the proportion of agreement was 90% (SE, 0.80) P < 0.0001. Agreement between BioELISA and Western blot was 92.5%, κ 0.85 (SE, 0.15) and P < 0.0001. These kappa statistics demonstrate that both intra-assay and inter-assay agreements were substantial to almost perfect .
Effect of prevalent HSV-2 infection on risk of intra-partum HIV-1 MTCT
Of the 1448 women with HSV-2 results, 1195 tested positive (82.5% [95% confidence interval (CI), 80.6–84.5)]. Cases were more likely to be HSV-2 infected than controls: (unadjusted OR, 1.49; 95% CI, 1.10–2.02; P = 0.010) (Table 2). This increased risk remained the same after controlling for other factors significantly associated with intra-partum transmission (adjusted OR, 1.50; 95% CI, 1.09–2.08; P = 0.014). When the model was re-run with only the randomly selected viral loads included (n = 163) the adjusted odds were unaltered (adjusted OR, 1.48; 95% CI, 1.07–2.05; P = 0.018). When the model was re-run including syphilis as a potential confounder, again the model was not appreciably altered (adjusted OR, 1.48; 95% CI, 1.07–2.04). The proportion of HIV-1 intra-partum transmission potentially attributable to prevalent maternal HSV-2 infection at time of delivery was 28.4% (95% CI, 7.3–44.7).
Seroconversion to HSV-2 antibody and its risk of intra-partum HIV-1 MTCT
Of the 193 women who were HSV-2 antibody negative at delivery, 158 had a 6-week sample available for testing: 129 remained HSV-2 negative, 2 were HSV-2 low-positive and 27 (17.3%, 95% CI 11.3–23.3) were positive, possibly due to recent acquisition of the infection in some women. Among the 158 baseline HSV-2-negative women with an available 6-week sample, the women who transmitted HIV-1 to their babies during the intra-partum period appeared more likely to have become HSV-2 antibody positive around the time of delivery than controls, although this difference was not statistically significant (unadjusted OR, 1.59; 95% CI, 0.67–3.73; P = 0.29; adjusted OR, 1.44; 95% CI, 0.57–3.69; P = 0.44) (Table 3).
If final results of HSV-2 antibody testing were based on HerpeSelect and Western blotting only, prevalent HSV-2 infection still increased the odds of vertical transmission but this increase was not statistically significant (unadjusted OR, 1.27; 95% CI, 0.91–1.77; P = 0.15; adjusted OR, 1.28; 95% CI, 0.90–1.83; P = 0.17). If the HSV-2 results were based on HerpeSelect and BioElisa only, prevalent HSV-2 infection increased odds of vertical transmission with borderline significance (unadjusted OR, 1.36; 95% CI, 1.03-1.80; P = 0.03; adjusted OR, 1.31; 95% CI, 0.97–1.76; P = 0.08).
Effect of Syphilis on HIV-1 vertical transmission
Fifty-two of 1289 women with available specimens had evidence of active syphilis at 6 weeks (4.0%; 95% CI, 3.0–5.1). The unadjusted OR of intra-partum transmission associated with syphilis was 0.89 (95% CI, 0.49–1.59; P = 0.68) and adjusted OR was 0.63 (95% CI, 0.34–1.20; P = 0.16). When the model was re-run including HSV-2 as a potential confounder the adjusted OR did not change appreciably (adjusted OR, 0.92; 95% CI, 0.50–1.70) (Table 4).
We believe this is the first study to show that maternal HSV-2 infection is associated with an increased risk of intra-partum HIV-1 MTCT. HSV-2 infection was very common among this population of HIV-1 infected women, over 80% being co-infected. Our findings suggest that over a quarter of all intra-partum HIV-1 transmissions are potentially attributable to maternal HSV-2 co-infection, presumably through increased HIV-1 genital shedding at the time of delivery in women co-infected with HSV-2. Current diagnostic techniques do not perfectly discriminate late in utero, intra-partum, and early post-natal MTCT. In this study, cases were defined as women whose babies were HIV DNA PCR negative at birth but became HIV DNA PCR positive at 6 weeks. While some of these babies will have acquired HIV-1 very late in utero or through early breast feeding rather than during the intra-partum period, this would probably reduce the observed association between maternal HSV-2 infection and risk of intra-partum HIV-1 MTCT rather than exaggerate it. An alternative explanation, given the finding in other studies, that suppressing genital herpes infection in HIV infected women reduces HIV plasma viral load in addition to genital HSV-2 shedding, is that HSV-2 is associated with intra-partum transmission through its effect on plasma viral load across the peri-partum period. However, in this study, we controlled for plasma HIV-1 RNA levels in the analysis. HSV-2 remained an independent predictor of MTCT. Of note, participants did not undergo genital examination so the risk of MTCT can not be related to clinical disease, level of genital HSV-2 DNA or to other reproductive tract infections. It is likely that if HSV-2 infection is causally associated with MTCT, it will be more strongly associated with levels of genital HSV-2 shedding than presence or absence of antibody.
The presence of HSV-2 antibody was detected by HerpeSelect EIA, but with the cut-off for diagnosing a positive result increased above that recommended by the manufacturers . The concordance of Focus test results with comparator tests (Western blot) is very high for index values lower than 0.9 and higher than 3.5. Intermediate range (low positive) results with index values ranging from 1.1 to 3.5 are more likely to be falsely positive . Thus a second test, the BioElisa (which used the same antigen as used in the Biokit test), and Western blot, an accepted gold standard test for specificity, were applied to samples with intermediate index values. If positive by Western blot the outcome was considered positive (due to the high inherent specificity of Western blot); if negative by BioElisa or indeterminate by Western blot, the outcome was considered to be negative (due to the inherent sensitivity of BioElisa) . Although the HerpeSelect EIA assay has had an high false positivity rate in sera from some countries in Africa  it has been validated using Zimbabwean sera with 100% sensitivity and specificity .
It is possible that maternal HSV-2 infection may simply reflect a higher risk of sexual exposure to MTCT co-factors e.g., HIV superinfection. HSV-2 infected women did report riskier sex behaviour during the post-partum period. However, including variables reflecting sex risk in the models for intra-partum transmission did not change the effect of HSV-2 on predicting intra-partum transmission at all (data not shown).
We observed a very high rate of HSV-2 seroconversion (17%) during the peri-partum period, higher than expected from what is known about HSV-2 incidence in non-pregnant women [26,27]. Several groups have shown that HSV-2 antibody levels are depressed in late pregnancy [28,29]. However, it is likely, that some of these women acquired HSV-2 during the peri-partum period possibly indicating that women are particularly susceptible to HSV-2 acquisition at this time, either for biological or behavioural reasons. Data from Rakai has recently shown that pregnancy increases susceptibility to HIV-1 infection . No other data on HSV-2 incidence in the peri-partum period from sub-Saharan Africa are available, however incidence in pregnant women in the USA is considerably lower than found in this study .
While seroconverting to HSV-2 antibody appeared to be associated with an increased risk of MTCT this was not statistically significant. It is likely that the strength of the association between maternal incident HSV-2 infection and MTCT was diluted by the over-estimation of incident infections, both because of antibody depression in late pregnancy and because only incident HSV-2 infection prior to delivery would impact MTCT, but incident cases in this study could have acquired HSV-2 either 3 weeks before or after delivery. Incident HSV-2 infection is associated with much higher rates of sexual transmission than prevalent infection and the same may be true for MTCT. The effect of maternal incident HSV-2 infection on MTCT needs to be further explored.
Known risk factors for intra-partum HIV-1 MTCT include advanced maternal HIV-1 disease (e.g., higher viral load, lower CD4 cell count) and obstetric factors. Several studies have shown that HIV-1 genital shedding is higher in the presence of genital herpes lesions [2,32,33] and asymptomatic HSV-2-infection . Additionally, HIV-1 infection increases both frequency and severity of HSV-2 genital lesions , creating a vicious cycle in which each infection exacerbates the other. Thus, HSV-2 co-infection may increase intra-partum HIV-1 MTCT both by increasing the likelihood and amount of genital HIV-1 shedding at delivery and the HIV-1 plasma viral load. It may alter the ratio of genital to plasma HIV-1 viral load, which could in turn explain why the relationship between maternal plasma HIV-1 viral load and intra-partum transmission is strong but not absolute. Finally there is some evidence to suggest that treating HSV-2 infection could reduce intra-partum transmission of HIV-1: in three studies, treating symptomatic genital herpes reduced HIV-1 genital shedding [32,34], and in another, suppressing genital herpes with antiviral drugs reduced plasma HIV-1 viral load .
In the absence of antiretroviral therapy the risk of HIV-1 MTCT varies around the world, with transmission rates in the developed world around half those from developing countries. These differences have largely been attributed to differences in rates of transmission through breastfeeding. Sero-epidemiological studies of HSV-2 show that the age-specific prevalence of HSV-2 is at least twice as high in sub-Saharan Africa than in the USA and Europe [36–39], with, for example, 40% of Tanzanian women infected with HSV-2 by the age of 20 years . Thus, variation in HSV-2 seroprevalence appears to mirror the variation in rates of HIV-1 MTCT and may explain at least some of the geographic differences.
Active syphilis at the time of delivery was not associated with an increased risk of perinatal HIV-1 transmission. This finding is consistent with data from the Rakai trial. In that trial, presumptive treatment of sexually transmitted infection (STI) in pregnant women significantly reduced maternal rates of bacterial STI but this reduction was not associated with a reduction in perinatal HIV-1 transmission .
Although WHO recommends that countries strive to provide daily zidovudine from ∼28 weeks plus single-dose nevirapine during labour and 7 days of zidovudine/lamivudine post-partum to as many HIV-1 positive women as possible, many countries in sub-Saharan Africa are still only able to offer single dose nevirapine to prevent peri-natal transmission. Treatment of HSV-2 infection with antiviral drugs, such as acyclovir, is cheap, safe and easy to take. It is possible that giving acyclovir to pregnant women in the peri-partum period, in addition to single dose nevirapine may further reduce their risk of HIV-1 transmission. A randomized trial of single dose nevaripine with or without acyclovir from 36 weeks is now warranted.
Anne Cent who undertook confirmatory HSV western blot analyses on a sub-set of the samples tested.
Sponsorship: The ZVITAMBO trial was supported by the Canadian International Development Agency (CIDA) (R/C Project 690/M3688), United States Agency for International Development (USAID) (cooperative agreement number HRN-A-00-97-00015-00 between Johns Hopkins University and the Office of Health and Nutrition - USAID), the Bill and Melinda Gates Foundation, Seattle Washington, USA, Rockefeller Foundation (New York, USA) and BASF (Ludwigshafen, Germany). Funding for this sub-study was from GlaxoSmithKline R&D.
F.C., J.H. and P.I. designed the study. F.C. and J.H. wrote the first draft of the paper on which all authors commented. R.N. undertook the statistical analyses. K.M. undertook and oversaw all laboratory aspects of the project. RMA provided expert advice on HSV-2 serological testing algorithms and on validating locally derived HSV-2 results.
1. Fleming DT, Wasserheit J. From epidemiological synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV. STI 1999; 75:3–17.
2. Mbopi-Keou F-X, Gresenguet G, Mayaud P, Weiss H-A, Gopal R, Matta M, et al
. Interactions between herpes simplex virus type 2
and human immunodeficiency virus type 1 infection in African women: opportunities for intervention. Journal of Infectious Diseases 2000; 182:1090–1096.
3. Freeman E, Weiss HA, Glynn JR, Cross P, Whitworth J, Hayes RJ. Herpes simplex virus type 2
infection increases risk of HIV acquisition among men and women: systematic review and meta-analysis of longitudinal studies. AIDS 2005; 20:73–83.
4. Gray R, Wawer M, Brookmeyer R, Sewankambo NK, Serwadda D, Wabwire-Mangen F, et al
. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet 2001; 357:1149–1153.
5. Mbopi Keou FX, Legoff J, Gresenguet G, Si-Mohammed A, Matta M, Mayaud P, et al
. Genital shedding of Herpes simplex virus type 2
DNA and HIV-1 RNA and proviral DNA in HIV-1 Herpes simplex virus type 2
co-infected women. JAIDS 2003; 33:121–124.
6. McClelland RS, Wang CC, Overbaugh J, Richardson BA, Corey L, Ashley RL, et al
. Association between cervical shedding of herpes simplex virus and HIV-1. AIDS 2002; 16:2425–2430.
7. Nagot N, Ouedraogo A, Foulongne V, Konate I, Weiss HA, Vergne L, et al
. Reduction of HIV-1 RNA levels with therapy to suppress herpes simplex virus. New Engl J Med 2007; 359:790–799.
8. Chen KT, Segu M, Lumey LH, Kuhn L, Carter RJ, Bulterys M, et al
. Genital herpes simplex virus infection and perinatal transmission of human immunodeficiency virus. Obs Gyn 2005; 186:1341–1348.
9. Wasserheit J. Epidemiological synergy: interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sex Trans Dis 1992; 19:61–77.
10. Serwadda D, Gray RH, Sewankambo NK, Wabwire-Mangen F, Chen MZ, Quinn TS, et al
. Human immunodeficiency virus acquisition associated with genital ulcer disease and Herpes simplex virus type 2
infection: a nested case control study in Rakai, Uganda. Journal of Infectious Diseases 2003; 188:1492–1497.
11. Humphrey JH, Iliff PJ, Marinda E, Mutasa K, Moulton LH, Chidawanyika H, et al
. Impact of single large doses of vitamin A given during the postpartum period to HIV-infected women and their neonates on breastfeeding-associated mother-to-child-transmission of HIV and infant mortality in Zimbabwe. Journal of Infectious Diseases 2006; 193:860–871.
12. Humphrey JH, Hargrove J, Malaba LC, Iliff PJ, Moulton LH, Zvandasara P, et al
. HIV incidence among postnatal women in Zimbabwe: Risk factors and effect of vitamin A supplementation. AIDS 2006; 20:1437–1446.
13. Zvandasara P, Hargrove J, Ntozini R, Chidawanyika H, Iliff PJ, Moulton LH, et al
. Mortality among postpartum HIV-positive and HIV-negative women in Zimbabwe: risk factors, causes, and impact of single-dose postpartum vitamin A supplementation. JAIDS 2006; 43:107–116.
14. Ashley RL, Militoni J, Lee F, Nahmias A, Corey L. Comparison of western blot (immunoblot) and glycoprotein G-specific immunodot enzyme assay for detecting antibodies to herpes simplex virus types 1 and 2 in human sera. J Clin Microbiol 1988; 26:662–667.
15. Baker S, Freedman L, Parmar M. Using replicate observations in observer agreement studies with binary assessments. Biometrics 1991; 47:1327–1348.
16. Miller MF, Stoltzfus RJ, Mbuya NV, et al
. Total body iron in HIV positive and HIV negative Zimbabwean newborns strongly predicts anemia throughout infancy and is predicted by maternal hemaglobin concentration. Journal of Nutrition 2003; 133:3461–3468.
17. Breslow NE, Day NE. Statistical Methods in Cancer Research: volume II – the design and analysis of cohort studies. Oxford: Oxford University Press; 1987.
18. Rockhill B, Newman B, Weinberg C. The use and misuse of population attributable fraction. Am J Pub Health 1998; 88:15–19.
19. Pearce N. Analytical implications of epidemiological concepts of interaction. Int J Epidemiol 1989; 18:976–980.
20. Brady AR. Adjusted population attributable fractions from logistic regression. College Station Texas: StataCorp; 1998.
21. Greenland S, Drescher K. Maximum likelihood estimates of the attributable from logistic models. Biometrics 1993; 49:865–872.
22. Landis JR, Kock G. The measurement of observer agreement for categorical data. Biometrics 1976; 33:159–174.
23. Hogrefe W, Su X, Song J, Ashley R, Kong L. Detection of Herpes simplex virus type 2
specific immunoglobulin G antibodies in African sera by using recombinant gG2, western blotting and gG2 inhibition. J Clin Microbiol 2002; 40:3635–3640.
24. Van Dyck E, Buve A, Weiss HA, Glynn JR, Brown DWG, de Deken B, et al
. Performance of commercially available immunoassays for detection of antibodies against Herpes Simplex virus type 2
in African populations. J Clin Microbiol 2004; 42:2961–2965.
25. Morrow R, Freidrich D, Meier A, Corey L. Use of biokit
HSV-2 rapid assay to improve the positive predictive value of Focus HerpeSelect HSV-2 ELISA. BMC Infect Dis 2005; 5:84.
26. McFarland W, Lovemore G, Bassett MT, Machekano R, Latif AS, Ley C, et al
. Prevalence and incidence of herpes simplex virus type 2
infection among male Zimbabwean factory workers. J Infect Dis 1999; 180:1459–1465.
27. Cowan F, Hargrove J, Langhaug LF, Jaffar S, Mhuriyengwe L, Swarthout TD, et al
. The appropriateness of core group interventions using presumptive periodic treatment among rural Zimbabwean women who exchange sex for gifts or money. J AIDS 2005; 38:202–207.
28. Eskild A, Jeansson S, Hagen J, Jenum PA, Skrondal A. Herpes simplex virus type-2 antibodies in pregnant women: the impact of the stage of pregnancy. Epidemiol-Infect 2000; 125:685–692.
29. Boggess K, Watts H, Hobson A, Ashley RL, Brown Z, Corey L. Herpes simplex type 2 detection by culture and polymerase chain reaction and relationship to genital symptoms and cervical antibody status during the third trimester of pregnancy. Am J Obstet Gynecol 1997; 176:443–451.
30. Gray RH, Wabwire-Mangen F, Kigozi G, Sewankambo NK, Serwadda D, Moulton LH, et al
. Randomized trial of presumptive sexually transmitted disease therapy during pregnancy in Rakai, Uganda. Am J Obstet Gynecol 2005; 185:1209–1217.
31. Brown ZA, Selke S, Zeh J, Kopelman J, Maslow A, Ashley RL, et al
. The Acquisition of herpes simplex virus during pregnancy. N Engl J Med 1997; 337:509–515.
32. Schaker T, Ryncarz A, Goddard J, Diem K, Shaughnessy M, Corey L. Frequent recovery of HIV-1 from genital Herpes simplex virus lesions in HIV-1 infected men. JAMA 1998; 280:61–66.
33. Augenbraun M, Feldman J, Chirgwin K, Zenilman J, Clarke L, DeHovitz J, et al
. Increased genital shedding of herpes simplex virus type 2
in HIV-serpositive women. Ann Intern Med 1995; 123:845–847.
34. Nagot N, Ouedraogo A, Mayaud P, Konate I, Vergne L, Weiss HA, et al
. Impact of HSV2 suppressive therapy on HIV-1 genital shedding and plasma viral load: a proof of concept randomised double-blind placebo controlled trial (ANRS 1285 Trial). Thirteenth Conference on Retroviruses and Opportunistic Infections
. Denver, CO, 5–8 February 2006 [abstract 33LB].
35. Corey L, Celum CL, Quinn TC. The Effects of herpes simplex virus-2 on HIV-1 acquisition and transmission: a review of two overlapping epidemics. J AIDS 2004; 35:435–445.
36. Fleming DT, McQullian GM, Johnson RE, Nahmias AJ, Aral SO, Lee FK, et al
. Herpes simplex virus type 2
in the United States, 1976 to 1994. New Engl J Med 1997; 337:1105–1111.
37. Vyse A, Gay N, Slomka MJ, Gopal R, Gibbs T, Morgan-Capner P, et al
. The burden of infection with HSV-1 and HSV-2 in England and Wales: implications for the changing epidemiology of genital herpes. Sex-Transm-Infect 2000; 76:183–187.
38. Pebody RG, Gopal R, Brown DWG, Miller L, The European HSV seroepidemiology multicentre study group. Multicentre HSV-1 and HSV-2 seroprevalence study. Int J STD AIDS 2001; 12(Suppl 2):148.
39. Weiss HA, Buve A, Robinson NJ, Van Dyck E, Kahindo M, Anagonou S, et al
. The epidemiology of HSV-2 infection and its association with HIV infection in four urban African populations. AIDS 2001; 15(suppl 4):S97–S108.
40. Obasi A, Mosha F, Quigley M, Sekirassa Z, Gibbs T, Munguti K, et al
. Antibody to herpes simplex virus type 2
as a marker of sexual risk behaviour in rural Tanzania. J Infect Dis 1999; 179:16–24.