French Guiana is a French territory in South America that shares borders with Brazil and Suriname. For cultural reasons, it has the highest birth rate in Latin America (26.8 per 1,000 persons) (1). The HIV prevalence and incidence are the highest of all French territories. The prevalence of HIV in French Guiana in the 15–49 year-old age group is estimated at 1.6% and incidence has substantially declined since the advent of highly active antiretroviral therapy (2). The drivers of the epidemic are: frequent multiple sexual partnerships, transactional sex, and crack cocaine-related risky sex behaviors (3). These drivers are often related to social precariousness. Because French Guiana has the highest Gross Domestic Product (GDP) and health expenditure per capita in Latin America, it attracts numerous vulnerable immigrants who arrive in hope of better economic prospects. However, when they arrive, immigrants often live in great poverty. While 29% of the total population and nearly half of adults are foreigners, for persons living with the Human Immunodeficiency Virus (HIV), 80% are foreigners (3). Although many immigrants with HIV come from HIV endemic countries, we have shown that over half acquire the infection early after arrival in French Guiana, presumably because of enhanced sexual vulnerability linked to social precariousness (4, 5). In this context, the transmission mode of HIV is by far mostly heterosexual, with a balanced number of males and females. The health system is the French universal health care system, which provides mostly free health care to all (genotyping, viral load and CD4 lymphocyte measurements, and antiretrovirals, among others), including undocumented immigrants, most of whom benefit from residence permits for health reasons. However, for persons who have a difficult life, access to care is often complicated or secondary to the immediate pressing needs of survival. This is particularly obvious for pregnant women, over a third of whom have a poor pregnancy follow-up (6). Preterm delivery is double that of mainland France (13% vs. 6.9%, respectively) (7). Stillbirths and infant mortality are both 2–3 times greater than in France (8). The prevalence of HIV among pregnant women in French Guiana has exceeded 1% for over 3 decades. Since 2008, following the national French guidelines, all pregnant women receive a triple antiretroviral combination therapy. HIV during pregnancy has been associated with a number of adverse outcomes (9–12). The role of antiretrovirals has yielded contrasting results: some finding no difference in pregnancy outcomes (13) whereas others, notably a case control study in French Guiana, found antiretroviral treatment was associated with preterm delivery (12).

In this context of high HIV prevalence, frequent poverty, poor pregnancy follow-up, frequent preeclampsia, frequent preterm delivery, we hypothesized that describing pregnancy outcomes and their trends would provide precious information to optimize the care of women with HIV. We thus aimed to describe the characteristics and outcomes of 1,774 pregnancies among a historical hospital cohort of 881 women living with HIV in French Guiana. The Amazonian context of the study may provide useful information for other countries in the region.

The study was a retrospective cohort conducted on the French Guiana data included in the French Hospital Database for HIV (FHDH), a national cohort that prospectively records all events and characteristics of persons with HIV. HIV-infected pregnancies (confirmed infections by western blot analysis) enrolled between January 1st 1992 to 31st July 2022 were considered. While HIV was the exposure the main outcomes of interest were pregnancy outcome, transmission, obstetrical complications. Patient data are routinely entered in the database by physicians as they consult and trained research technicians who control for exhaustivity. The main study variables were age, country of birth, date of HIV discovery and reason for testing, initial CD4 count, CDC stage, viral load, antiretroviral treatment, opportunistic infections, and pregnancy outcomes (obstetrical complications, labor and delivery complications, early neonatal outcomes, complications related to puerperium). Because HIV records did not have precise information about term at delivery, between 2008 and 2022 research technicians compiled this information in Cayenne.

Overall, there were 1,774 pregnancies in 881 women living with HIV. The mean age at the time of pregnancy was 30.3 years (SD = 6.6). Most pregnant women were from Haiti (39.2%), Suriname (26%), and France (15.4%), as in the overall cohort.

Table 1 shows HIV characteristics. Briefly, 69% of women had first been screened for HIV during pregnancy at the initiation of a physician. Only 14.9% of women had CD4 counts <200 per mm³ and 17% were at the Acquired Immuno Deficiency Syndrome (AIDS) stage. Between 1993 and 2002, the mean nadir CD4 count was 579 per mm³, between 2003 and 2012 it was 375 per mm³, and between 2013 and 2022 it was 509 per mm³.

Table 2 shows that the HIV diagnosis was known in 75.1% of pregnancies and that, in 67.6% of women, HIV follow-up predated pregnancy. Nearly half of women, 49.4%, only had one pregnancy since having been diagnosed with HIV. Although most women received antiretroviral therapy during pregnancy, only 48.5% of those with the available information had an undetectable viral load at delivery. Between 2012 and 2022, 93% of pregnant women were on antiretroviral therapy.

Overall, 15.3% of pregnancies ended with an abortion but between 2012 and 2022 it significantly declined to 10.4%. Overall, 99.8% of women gave exclusive artificial feeding to their newborn. Since the beginning of data collection, there were a total of 110 newborns infected with HIV representing an overall transmission rate of 6.2% (110/1,771). Between 1993 and 2002, the transmission rate was 34%, between 2003 and 2012 it was 1.3%, and between 2013 and 2022 it was 0.7%.

Figure 1 shows the gradual reduction of the viral load at delivery.

Here, in a territory combining the highest HIV prevalence of all French territories, the highest fertility rate in Latin America, we provide the first epidemiologic overview of HIV-positive pregnancies. Among pregnant women living with HIV, we show that the preterm delivery rate remained particularly high: in 2022 it was 19.2% vs. 13% in the general population of French Guiana, and 6.9% in the general population of mainland France. The proportion of low birth weight births was also very high: 17% vs. 11.9% in the general population of French Guiana, and 7.5% in the general population of mainland France, and 14.6% in the world (14).

Over half (53%) of the general population of French Guiana lives below the poverty threshold, 29% lives in great poverty (15), and persons living with HIV are an even more precarious subpopulation. The link between poor follow-up, psychosocial stressors, and preterm delivery has been previously shown and presumably explain much of this difference (6). However, HIV itself may also contribute to this very high rate of preterm delivery (12). Other analytic studies in territories with widespread poverty have yielded contrasting conclusions: some did not demonstrate significant differences (13) whereas others did (12). Although the present study was only descriptive, the levels of preterm delivery or low birth weight were substantially greater than in the general population of French Guiana and it is likely that there are indeed worse pregnancy outcomes among women living with HIV than in the general female population of French Guiana. Given the progress in testing and treating women, we hypothesize that the remaining challenges regarding preterm delivery and low birth weight would most benefit from optimal obstetrical follow-up.

In French Guiana, health professionals and non-government organizations have been proactive in reaching vulnerable populations. These treatment and prevention efforts have now led to most women delivering with viral loads beneath transmission thresholds and very low residual transmission, usually in cases of women tested late in their pregnancy or lost to follow-up. Between 2012 and 2022, 93% of pregnant women had received ARV treatment, a figure that is only slightly above that of what is reported globally (13). This emphasizes the residual challenges that remain to follow the most vulnerable women, a fraction of whom escape the reach of the health system.

The present study has a number of limitations. The absence of a proper control group in the present study was a notable limitation. The crude comparisons relied on different data sources for the general population of French Guiana. Our analysis was purely descriptive and spans over 3 decades of multiple gradual changes, in terms of social conditions, of HIV care and more generally of the organization of the health system. It is therefore difficult to pinpoint any single factor involved in the positive trends but rather a convergence of proactive efforts to reach and test pregnant women, improvements of the coordination of health care workers, improvement in treatment strategies, improvements in the tolerance and effectiveness of antiretrovirals being a potent combination to reduce vertical transmission of HIV. Another important limitation is that social factors are probably the major determinant of the studied health outcomes, yet we had no direct measurement of precariousness in our data, so we can only assume –not prove—that much of what is shown is detrimentally influenced by precariousness. The information about preterm delivery rests on a subsample from Cayenne's level 3 maternity which may bias preterm delivery estimates upwards. Furthermore, the data on birthweight was also a subsample from Cayenne maternity so results should be taken with caution. Despite the above limitations, the present study has some merits as it is the first study to provide real life data on historical trends and remaining challenges in French Guiana; it also provides important information on various standard aspects of HIV infection and their temporal evolution.

In summary, the present study shows that with time the system has been successful in reducing mother to child transmission and preterm births. However, challenges remain because incidence of preterm delivery and low birth weight are still substantially higher than in the general population of pregnant women in French Guiana. Teasing out what fraction is linked to HIV and what fraction is linked to social precariousness and poor follow-up was not possible in this study but, in practice, it seems crucial to implement to continue early and careful obstetrical follow-up for pregnant women living with HIV. Reducing maternal-fetal transmission of HIV necessarily implies efforts to improve access to care for fragile populations.