Sex Transm Infect 85:82-87 doi:10.1136/sti.2008.032300
  • Clinical

Pregnancy outcome in women infected with HIV-1 receiving combination antiretroviral therapy before versus after conception

  1. E S Machado1,
  2. C B Hofer1,
  3. T T Costa2,
  4. S A Nogueira3,
  5. R H Oliveira2,
  6. T F Abreu2,
  7. L A Evangelista2,
  8. I F A Farias2,
  9. R T C Mercadante2,
  10. M F L Garcia2,
  11. R C Neves2,
  12. V M Costa2,
  13. J S Lambert4
  1. 1
    Serviço de Doenças Infecciosas e Parasitárias, Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro (UFRJ), Brazil
  2. 2
    Instituto de Puericultura e Pediatria Martagão Gesteira (IPPMG), UFRJ, Brazil
  3. 3
    Secretaria Municipal de Petrópolis, Faculdade de Medicina de Petrópolis, Rio de Janeiro, Brazil
  4. 4
    Mater University Hospital, Dublin, Ireland
  1. Elizabeth S Machado, Serviço de Doenças Infecciosas e Parasitárias, Hospital Universitário Clementino Fraga Filho – HUCFF – UFRJ, Av. Professor Rodolpho Paulo Rocco, n° 255, 5° andar. Cidade Universitária, Ilha do Fundão, Rio de Janeiro, Brazil; emachado{at}
  • Accepted 30 September 2008
  • Published Online First 5 November 2008


Objective: The potential adverse effects of antiretroviral drugs during pregnancy are discrepant and few studies, mostly from Europe, have provided information about pregnancy outcomes of those already on treatment at conception. The aim of this study was to investigate the impact of antiretrovirals (ARVs) on pregnancy outcome according to the timing of treatment initiation in a cohort of pregnant women from Brazil infected with HIV.

Methods: A prospective cohort of 696 pregnant women followed up in one single centre between 1996 and 2006 was studied. Patients who had ARV treatment before pregnancy were compared with those treated after the first trimester. The outcomes evaluated were preterm delivery (PTD) (<37 weeks), severe PTD (<34 weeks), low birth weight (LBW) (<2500 g) and very LBW (<1500 g).

Results: Patients who were using ARVs pre-conception had higher rates of LBW (33.3% vs 16.5%; p<0.001) and a similar trend for PTD (26.3% vs 17.7%; p = 0.09). Stratification by type of therapy (dual vs highly active antiretroviral therapy (HAART)) according to timing of initiation of ARVs showed that patients who use HAART pre-conception have a higher rate of PTD (20.2% vs 10.2%; p = 0.03) and LBW (24.2% vs 10.2%; p = 0.002). After adjusting for several factors, HAART used pre-conception was associated with an increased risk for PTD (AOR 5.0; 95% CI 1.5 to 17.0; p = 0.009) and LBW (OR 3.6; 95% CI 1.7 to 7.7; p = 0.001).

Conclusions: We identified an increased risk for LBW and PTD in patients who had HAART prior to pregnancy.

Over the last decade, treatment for the prevention of mother to child transmission (MTCT) of HIV has moved from zidovudine (AZT) monotherapy to the use of highly active antiretroviral therapy (HAART) resulting in transmission rates of 1–2%.17 In Brazil, vertical transmission of HIV-1 has been responsible for almost 11 000 cases of AIDS in children younger than 13 years between 1996 and 2005; however, less than 350 cases were notified during the year of 2006.8

Discrepant results on the risk of adverse pregnancy outcomes in women infected with HIV treated with HAART have been reported. European studies have shown that exposure to any combination therapy increases the risk of premature delivery (PTD), with the use of protease inhibitors being the highest risk,911 and also an increased risk for low birth weight (LBW) and stillbirths.12 In contrast with these findings, several studies in the USA and Latin America1316 failed to show a higher risk of adverse outcomes, although a report from one single site in the USA showed a correlation between PTD and protease inhibitor use.17

The impact of antiretroviral (ARV) drugs administered prior to conception has not been fully explored. The European Collaborative Study10 18 found a twofold increased risk of premature delivery and a fourfold increased risk for severe premature delivery in women who started combination therapy pre-conception. More recently, a 3.4-fold in the risk of PTD in patients treated with HAART was seen in a German/Austrian cohort where 52% of the patients were already being treated with HAART at conception.19

There has been an increase in pregnant women who are HIV positive receiving antenatal therapy. This might be an effect of the increasing age in the population of pregnant women with HIV, repeated pregnancies or changes in CD4 threshold for starting treatment in patients infected with HIV. The aim of this study was to investigate the impact of ARVs on preterm delivery and LBW according to the timing of the initiation of the therapy in pregnancy.


Study population

A prospective cohort of 899 pregnant women followed up between 1996 and 2006 in a single HIV reference centre in Rio de Janeiro (Program of Integral Assistance for HIV-Infected Pregnant Women of the Federal University of Rio de Janeiro, Brazil) was studied. Treatment of pregnant women uses a multidisciplinary team approach. ARVs are offered free of charge. Mono and dual therapy was offered until 1998 when protease inhibitors became available in the public healthcare system. All women are counselled not to breastfeed and formula, free of charge, is offered for a year. Elective caesarean section is usually performed at 38 weeks gestation for those with a viral load (VL) >1000 copies/ml or an unknown VL and also in cases of severe immunosupression (CD4+ T cell counts <200 cells/μl). Patients who have a spontaneous loss of a pregnancy before 24 weeks were excluded. Data were collected during the follow-up and completed during the postpartum period.

Study variables

The following pregnancy outcomes were evaluated: PTD <37 weeks’ gestation; severe PTD <34 weeks’ gestation; LBW <2500 g at birth; very LBW <1500 g at birth.

Gestational age was estimated based on the date of the last menstrual period or the earliest ultrasound date available.

Co-variables studied were age, use of tobacco, alcohol and illicit drugs. CD4+ T cell counts and VL were evaluated at initiation of prenatal care and before delivery. Symptomatic disease was defined as any clinical condition classified as category B or C (1993 revised Centers for Disease Control and Prevention classification system) occurring during the current pregnancy or in the past. ARV therapy was categorised as monotherapy, dual therapy (2 nucleoside reverse transcriptase inhibitors (NRTIs)) or HAART (2 NRTIs + protease inhibitors or non-nucleoside reverse transcriptase inhibitors (NNRTIs)). Patients who were taking ARVs before the estimated time of conception were included in the group of ARVs pre-conception. Use of ARVs was categorised as prophylaxis (if CD4 at entry ⩾200 cell/μl) or treatment (CD4 at entry <200 cells/μl or ARV pre-conception use).

Obstetric history included multiparity (defined as women who have given birth at least once) and mode of delivery (vaginal, emergency or elective caesarean section). Any previous history of miscarriage or PTD was considered as a single variable (previous adverse pregnancy) because information about the exact week of gestation in which the event occurred as well as LBW, fetal abnormalities, stillbirth or perinatal/neonatal mortality was not collected. Presence of hypertension was defined as any patient with chronic hypertension or development of hypertension, pre-eclampsia or eclampsia during the present pregnancy.

Sexually transmitted diseases (STDs) were considered positive if patients had a diagnosis during the present pregnancy of trichomoniasis, human papillomavirus (HPV) infection, syphilis, herpes infection or culture of vaginal swabs positive for Gardnerella vaginalis.

Patients with a hepatitis B surface antigen (HBSAg) or hepatitis C virus (HCV) serology positive were considered as co-infected.

Statistical analysis

Data analysis was performed using Stata (v9.0) statistical software (Stata Corp, College Station, Texas, USA). Univariate analyses were performed using independent t tests (for variables with a normal distribution) or Wilcoxon (Mann-Whitney) two-sample tests (for variables that did not follow the normal distribution). The χ2 test was used to evaluate associations for categorical variables (or Fisher’s exact test if 20% or more cells on a table had an expected value of ⩽5). Variables with p⩽0.15 were included in the multivariate analysis.

A main-effects logistic regression model was fitted. Interactions were assessed using the −2 log likelihood ratio test comparing models with and without interactions. The Pearson’s χ2 goodness of fit test, as well as the Hosmer-Lemeshow test, was used to evaluate fitness of the model.


We excluded 203 patients: 150 subsequent pregnancies, 13 because of missing data on weight of the newborn and 40 who started treatment ⩽2 weeks before delivery (30 patients who received ⩽1 week of ARVs, 10 patients who started ARVs at 36 weeks of gestation). A total of 696 pregnancies were analysed.

Women already on ARV treatment at the time of conception

A total of 130 patients (18.7%) conceived while on ARVs. Altogether, 30 patients were on dual therapy (2 NRTIs), 47 (36.2%) on NNRTI-based HAART and 53 (40.7%) on protease inhibitor based HAART. Pre-conception ARV use was more frequent between 2001 and 2006 (82.3%). ARV treatment during prenatal care for this group was two NRTIs (22 patients; 16.9%), triple therapy with a protease inhibitor (73 patients; 56.2%) or nevirapine (35 patients; 26.9%). VL at the beginning of prenatal care was undetected in 32 patients, median VL at entry was 4900 copies/ml (19 patients with a VL⩽1000 copies/ml) in 88 patients and unavailable in 10 patients. Pre-conception therapy was changed in 10 patients with an undetectable VL for substitution of unwanted combinations and in 58 patients with a detectable VL to increase viral suppression.

At delivery, 47% (46/98) had an undetectable VL: 4 patients were treated with dual therapy, 18 patients treated with NNRTI-based HAART (nevirapine) and the rest with protease inhibitor-based HAART. Median gestational age when the change of therapy occurs was 15 weeks' gestation.

Women who started ARVs after the first trimester of pregnancy

A total of 566 women started ARVs after the first trimester of pregnancy (monotherapy with AZT: 179 (31.6%); 2 NRTIs: 182 (32.2%); HAART: 205 (36.2%)). HAART with protease inhibitors was used in 140 patients and nevirapine in 65 patients. Median gestational age at the beginning of ARV treatment was 24 weeks. VL at delivery was available for 336 patients and 44% achieved an undetectable VL of which 28 patients were treated with zidovudine, 41 with 2 NRTIs, 26 with NNRTI-based HAART and 53 with a protease inhibitor-based HAART.

The characteristics of both groups are described in table 1. Patients already on ARVs tended to be older, had a higher frequency of STDs, were more frequently multiparous, were commonly using HAART and had a higher frequency of VL ⩾10 000 copies/ml at delivery. Although symptomatic disease was more prevalent in this group, most of the reported infections were related to a past event. The occurrence of opportunistic infections during the current pregnancy was similar in both groups studied: 5% (pre-conception ARVs) and 3.3% (post-conception ARVs) and the difference was not statistically significant.

Table 1 Characteristics of 696 pregnant women according to timing of ARVs

Most babies were delivered by caesarean section (69.7%). The method of delivery was similar for both groups as was the median gestation age at delivery (38.5 weeks). The HIV status of 541 children was available (101 children born to ARV pre-conception mothers and 433 children from ARV post-conception mothers). Seven children (1.3%) were infected of whom only one mother was already using ARVs.

Adverse outcomes

The overall frequency of PTD and LBW in our cohort was 80 (11.5%) and 90 (12.9%), respectively (table 2). In univariate analyses, both outcomes were associated with a VL ⩾10000 copies/ml at delivery and pre-conception use of ARVs. Hypertension was also a risk factor for LBW. Patients treated with monotherapy had a trend towards a lower rate of LBW (odds ratio (OR) 0.54; 95% CI 0.29 to 1.00).

Table 2 Univariate analysis for preterm delivery and low birth weight

Both outcomes were more frequent during the period of 2001–2006 when there was an increase in the use of HAART pre-conception and post-conception. HAART use increased from 9% (19/207; 8 pre-conception) to 60% (293/489; 92 before pre-conception) when comparing the two periods.

Babies born to pre-conception ARV women were lighter (2879 g) than those born to post-conception ARV mothers (3077 g): p = 0.001. We noted a higher rate of LBW (30/130 (23%) vs 60/566 (10.6%); p<0.001) in this group and also a trend for PTB (21/130 (16.1%) vs 59/566 (10.4%); p = 0.09).

Very LBW was seen in nine cases with no difference between groups (p = 0.67). Severe prematurity was responsible for 23.7% of the cases of PTD (19 cases). Despite the trend in relation to use of ARV pre-conception (p = 0.06), no other risk factor was found probably due to the small sample size.

In order to explore the impact of pre-conception use of dual therapy and HAART on both outcomes, we compared patients treated exclusively with dual therapy or HAART. We excluded eight patients who were taking two NRTIs at conception and changed to HAART after the diagnosis of pregnancy and one patient with HAART pre-conception who changed to dual therapy. Only use of HAART pre-conception was significantly associated with PTD and LBW (table 3).

Table 3 Preterm delivery and low birth weight rates for dual therapy and highly active antiretroviral treatment (HAART) according to timing of initiation of antiretrovirals (ARVs)

The HAART treated group was then split into two categories: those treated exclusively with NNRTI-based HAART or protease inhibitor-based HAART (pre-conception and post-conception). We excluded 24 patients who were taking NNRTI-based HAART (19 patients) or protease inhibitor-based HAART (5 patients) at conception and had their regimen switched to protease inhibitor-based and NNRTI-based HAART, respectively. In patients treated with NNRTI-based HAART, PTD and LBW for those treated pre-conception or post-conception were 10/47 (17.8%) versus 5/65 (7.7%) (p = 0.16) and 5/28 (17.8%) versus 8/65 (12.3%) (p = 0.52), respectively. For the protease inhibitors treated group, PTD and LBW were 9/47 (19.1%) versus 16/140 (11.4%) (p = 0.27) and 12/47 (25.5%) versus 13/140 (9.3%) (p = 0.009), respectively.

Because splitting the HAART group yielded a low number of patients and outcomes, our multivariate analysis included the whole group who was treated with HAART (table 4). Prematurity was associated with pre-conception use of HAART (adjusted odds ratio (AOR) 5.06; 95% CI 1.5 to 17.0) and a VL⩾10 000 copies/ml (AOR 5.5; 95% CI 1.0 to 30.8). LBW was associated with HAART pre-conception (AOR 3.6; 95% CI 1.7 to 7.7) and hypertension (AOR 3.8; 95% CI 2.24 to 8.26).

Table 4 Unadjusted and adjusted odds ratios for preterm delivery and low birth weight in patients treated exclusively with highly active antiretroviral treatment (HAART)


The analyses of our cohort of pregnant women with HIV followed up in a single hospital with a uniform standard of care have shown rates of LBW and PTD of 12.9% and 11.5%, respectively, among patients treated with ARV therapy. This is higher than the average rate of LBW (9.1%) and PTD (7.8%) in public hospitals in Rio de Janeiro.20

Our ratios are similar to a previous report in a cohort of pregnant women with HIV in Brazil15 where there was a slightly higher prevalence of LBW (14.3%) compared with PTD (10.6%). The lower rates of PTD and LBW in Latin American cohorts when compared with those in Europe and the USA could be related to a lower prevalence of illicit drug use and to the absence of women who did not received ARVs during pregnancy in the studies.

Stratification of our cohort by type and time of initiation of ARV revealed that HAART use pre-conception is associated with 3.6-fold increased risk for LBW and a fivefold increased risk for PTD. LBW in this cohort was also associated with hypertension, which is a well-known risk for this outcome.21

The higher frequency of PTD and LBW among women treated with ARV pre-conception, seen in our cohort, could be related to a more advanced stage of disease. However, inclusion of a present or past history of symptoms did not show a significant statistical association with any adverse outcome.

Stratification of our cohort by time of initiation of ARVs and by type of therapy (dual therapy vs HAART) revealed that HAART use pre-conception was associated with higher rates of LBW and PTD compared with post-conception. A VL⩾10 000 copies/ml at the time of delivery was found to be an independent risk factor for PTD. A study from one single centre in the UK22 failed to show any increased risk for adverse outcomes in women treated with ARVs before pregnancy. The low median VL of these patients (49 copies/ml) could explain this discrepancy. In our cohort, a high VL was only marginally significant as a predictive factor for PTD and it should be carefully interpreted because VL at delivery was missing in 38% of the population studied.

Our study had some limitations, such as the absence of data about ethnicity, socioeconomic status, route of transmission (although more than 95% were due to heterosexual transmission), a detailed differentiation of previous adverse pregnancy outcomes and body mass index. A low body mass index has been shown to increase two to three times the risk for both LBW and PTD.15 23 24 Another limitation was the lack of data on the use of alcohol, tobacco and illicit drugs, which were missing for 35% of our cohort, but the prevalence of these behaviours were overall very low and would not have a great influence in our results. Finally, we were unable to evaluate the risk for individual ARV class.

As ARV use pre-conception cannot be stopped in women who are being treated for their own health, it will be important to continue monitoring adverse pregnancy outcomes in this population to assess if additional or different findings from cohorts with different geographies and demographics arise over time. As the number of patients who become pregnant while on ARV is increasing in Brazil and worldwide, detecting risk factors for adverse outcomes early in gestation will clearly be important to improve management for such women.

Key messages

  • Use of antiretrovirals (ARVs) pre-conception was associated with an increased risk of preterm delivery (PTD) and low birth weight (LBW) in a cohort of 696 pregnant women with HIV in Brazil during the period 1996–2006.

  • Stratification of our cohort by time of initiation of ARVs and by type of therapy (dual therapy vs highly active antiretroviral therapy (HAART)) revealed that HAART use pre-conception was associated with higher rates of PTD and LBW compared with post-conception. Hypertension was also an independent risk factor for LBW.

  • As ARV use pre-conception cannot be stopped in women who are being treated for their own health, it will be important to continue monitoring adverse pregnancy outcomes in this population to assess if additional or different findings from cohorts with different geographies and demographics arise over time.

  • Detection of risk factors for adverse outcomes early in gestation in this population will clearly be important to improve management for such women.


We thank Andreia Fiorani for proofreading.


  • Funding: Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) research grant for ESM. Ministério da Saúde do Brazil, Programa DST/AIDS (TC 238/07) for CBH.

  • Competing interests: None.

  • Ethics approval: Approved by the Universidade Federal do Rio de Janeiro Ethical Committee.

  • Contributors: ESM designed the study and was the lead author of the paper. CBH contributed to the design of the study and performed the statistical analysis. ESM, CBH, TPC, SAN, RHO, TFA, IFAF, RTCM, MFLG, RCN and VMC contributed in the follow-up of the patients and children. JSL provided revision of the manuscript.


Free sample

This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of STI.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.