Despite two decades of research on HIV-1 mother to child transmission (MTCT), there has been little study of the risks of multiple pregnancy in women with HIV infection. Twin pregnancies are associated with an increased risk of complications of pregnancy such as preterm labour and premature rupture of the membranes, which in turn are known to be risk factors for MTCT [1,2]. The only study to address the issue, performed in the 1980s, reported no significant difference in the MTCT rate between twin and singleton pregnancies, but only 22 twin pairs were enrolled . In an international registry of twin deliveries in HIV-infected women, which did not include a control group of singletons, MTCT was found to be more frequent in the first-born twin than in the second [4–6], but this was not confirmed in a study from Malawi . These studies were conducted in mothers and infants receiving no antiretroviral therapy (ART). More recently, with the use of antiretroviral drugs and other methods of prevention, the incidence of MTCT in industrialized countries has been reduced from 15–25% before 1994, to < 2% since 1997 [8–12]. To our knowledge, there has been no study of twins born to mothers receiving ART.
The issue is important because of the increasing incidence of multiple pregnancies in industrialized countries during the last 20 years, mostly due to advancing maternal age at pregnancy and the use of assisted reproduction techniques such as artificial insemination and in vitro fertilization . In France, the twinning rate increased from 0.89 in 1972 to 1.4 per 100 pregnancies in 1998 [14,15]. Increasing numbers of women with HIV infection are enrolled in assisted reproduction programs , and therefore require updated information on the risks involved in case of twin pregnancy. We performed a study of MTCT in twins, in comparison with singletons, born to HIV-1 infected mothers in the French Perinatal HIV Cohort.
Materials and methods
The French Perinatal HIV Cohort (EPF)
The French Perinatal HIV Cohort has prospectively collected data on HIV-infected pregnant women and their children in 90 centers throughout France since 24 September 1986. Informed consent was obtained from all mothers, during the pregnancy or at delivery. The cohort study was approved, according to French laws, by the Cochin Hospital Institutional Review Board and the French computer database watchdog commission (Commission Nationale de l'Informatique et des Libertés).
Children were followed up according to recommended standards of care, including clinical and biologic examination at birth, 1, 3, 6, 12 and 18–24 months, as previously reported . No specific recommendations for HIV treatment or obstetric care was made for women enrolled in the cohort, but investigators were encouraged to follow French national guidelines for prevention of MTCT, as regularly published and updated [18–22].
The changing management can be grouped into three major periods [18–22]. Until 1994, ART was not routinely available. From 1994 to 1997, the standard of care was prophylaxis with zidovudine monotherapy. Since 1997, the standard of care is combination therapy. Elective Caesarean section was not recommended until 1997, recommended systematically from 1998 to 2002, and recommended for women with detectable plasma viral load since 2002. In all periods, intrapartum and neonatal prophylaxis with zidovudine, as well as refraining from breastfeeding, were recommended.
All consecutive HIV-1 infected women enrolled in the French Perinatal HIV Cohort who delivered between 1984 and 2004, in mainland France (n = 9262) were included, except for a case of triplets, with three uninfected children.
An infant was considered as infected if HIV1 was detected by viral tests on two separate samples [HIV1-PCR DNA or HIV RNA or peripheral blood mononuclear cell (PBMC) viral culture or p24 antigenemia] or if anti-HIV-1 antibodies detected by ELISA and Western Blot persisted after 18 months of age. An infant was considered as non-infected if viral tests were negative on two separate samples, at least one of which was taken after termination of the neonatal prophylactic treatment or, if serological testing was negative after 18 months. Laboratory tests were done on site. HIV-1 RNA quantification in plasma were assessed by either the Roche Amplicor Monitor Test version 1.5 (Roche Diagnostic Systems, Basel, Switzerland) or branched DNA (Quantiplex, Versant, Bayer HealthCare, Tarrytown, New York, USA); HIV-1 PCR DNA in PBMC were performed using the Roche Amplicor Monitor Test with modifications as described , or using the real time PCR following the ANRS method . Lastly, for few specific cases, PBMC viral culture was performed as described . The timing of transmission was considered to be in utero when a first HIV PCR or culture was positive in the first 3 days after birth, and intrapartum if the first test was negative; timing was undetermined when no PCR or culture was available within the first 3 days.
Demographics (age, geographical origin, and drug use) were recorded, as were HIV care (knowledge of HIV infection status, ART) and obstetric care (parity, gestational age at booking in the obstetrical centre, gestational age at delivery, mode of delivery, administration of intrapartum zidovudine, neonatal prophylaxis, breastfeeding). Premature rupture of the membranes and zygocity were recorded only before 2000. The type of ART was defined by the last treatment received before delivery, and classified as monotherapy with a nucleoside reverse transcriptase inhibitor (NRTI), dual-NRTI therapy, or HAART (combinations of three or more drugs). The infant prophylactic treatment was classified as none, NRTI monotherapy or combination with two or more drugs. Maternal plasma HIV-1 RNA viral load and CD4 cell counts were recorded nearest to the time of delivery.
We first estimated changes in the incidence of twin pregnancies over the 1984–2004 period, using the χ2 test for trends. A logistic regression with multiple pregnancy as a dependant variable was performed to study relation with time, as a quantitative variable, after adjustment for geographical origin, maternal age, and parity.
Secondly, we compared demographic characteristics, obstetric and HIV management of twins and singleton pregnancies, overall and according to the three periods, and according to standards of ART: before 1994, between 1994 and 1996, and from 1997 to 2004. Since a quarter (24%) of mothers were enrolled in the cohort at least twice for successive pregnancies, we considered only the last pregnancy globally, and for each period.
We then considered all pregnancies with HIV status available for the child to study the association of twin pregnancy with mother-to-child HIV-1 transmission and the effect of birth order. We estimated the transmission rate globally and for each period, considering pregnancy as the unit of observation, i.e., the mother was considered to have transmitted the virus if at least one of the children was infected and not to have transmitted if both children were uninfected. In two cases in which twin B was stillborn, we used the HIV transmission status of twin A. We also calculated MTCT rates taking all children, and then only first twins, as the unit of observation. Percentages were estimated with exact 95% confidence intervals (CI), and medians with interquartile range (IQR). We used the χ2 test or two-sided Fisher exact test to compare percentages, and t Student or Wilcoxon test to compare means of continuous variables. We compared the proportion of transmission in first and second twins, using the McNemar test, and we estimated the children pair agreement for HIV status, using Kappa coefficients . To study the independent association of multiple pregnancy with the probability of MTCT, we performed a logistic regression adjusted for variables which differed between twins and singletons, gestational age, parity and geographic origin, as well as others found to be associated with transmission risk. Because there was only one case of transmission among twins in the period from 1997 onwards, we restricted the multivariable analysis to the period 1984–1996.
A P-value < 0.05 was considered for statistically significant. Statistical analyses were performed using the STATA software (Stata Statistical Software: Release 8.0; 2003, StataCorp, College Station, Texas, USA).
Trend over time of twinning
Overall, 192 of the 9262 deliveries were twin deliveries, an incidence of 2.1% (95% CI, 1.8–2.4). The proportion increased over time (χ2 for trends, 9.049; P = 0.003) (Fig. 1). It was 0.6% (4/320) between 1984 and 1987, reached 1.7% (26/1514) between 1992 and 1995 and 2.5% (84/3482) in 2000–2004. Over the 20-year study period, there were increases in maternal age, and in the proportion of African mothers; both factors were associated with an increased twinning rate (Table 1). After adjustment in logistic regression for maternal age, geographical origin and parity, the adjusted odds ratio associated with each 4-year increment was was 1.02 (95% CI, 0.90–1.22; P = 0.497).
The incidence of preterm delivery, i.e., before 37 weeks, was higher in twins than in singletons (P < 0.01; Table 1). Overall, severe prematurity (< 33 weeks) in twins was fourfold that of singletons. The mode of delivery differed significantly (P < 0.01), mostly regarding the rate of emergency Caesarean section, which was higher in twins than in singletons. None of the twins was breastfed.
We compared the use of ART between twins and singletons in the period since 1994 (Table 2). Lack of ART was more exceptional in twins than in singletons, but there was no difference in the proportions on ART at conception, median gestational age at the start of antenatal ART, type of ART, the proportion with no intrapartum therapy, or prophylactic treatment of the neonate. The median duration of ART during pregnancy was 4 weeks shorter in twin pregnancies (13 versus 17 weeks, P < 0.01), relative to the lower gestational age at delivery. The median maternal plasma viral load at delivery did not differ between twin and singleton deliveries (111 copies/ml versus 156 copies/ml, P = 0.3), nor did the proportions below 500 copies/ml (Table 2).
Mother-to-child transmission of HIV-1
HIV status was documented for both children in 164 of 192 twin deliveries from 1984 to 2004. Neither child was infected in 90.2% (148) of cases, twin A only was infected in 7.3% (12), twin B only in 1.2% (2), and both were infected in 1.2% (2). The risk of transmission to twins A and B were correlated (P = 0.04, exact test), with a low level of agreement (Kappa coefficient, 0.19). The transmission rate for twin A was threefold that for twin B (8.5% versus 2.4%; MacNemar χ2 test, P = 0.008). For the two twin deliveries in which twin B was stillborn, twin A was infected in one case and uninfected in the other.
The timing of transmission was considered to be intrapartum in twin A for four of the five cases in which an HIV PCR or culture at birth was available, and was in utero in twin B for three of the four cases. Of note, both cases in which twin B, but not twin A, was infected, involved in utero transmission. Although zygocity was unknown in nearly one half of cases, presumably in dichorionic pregnancies with same sex, there was only one case of discordant transmission (to the first twin) in a monochorionic pregnancy.
When compared with the MTCT rate in singleton pregnancies (6.2%, 503/8127), HIV transmission occurred in 16 of 164 twin pregnancies, considered as the unit of observation (9.8%, P = 0.06), 15 of 166 first twins (9.0%, P = 0.13), and 19 of the total of 330 twins (5.8%, P = 0.60).
The MTCT rate sharply decreased over time, with a significant interaction between multiple pregnancies and the three treatment periods (P < 0.001). The MTCT rate in twin pregnancies (34.4%, 11/32) was twice that in than in singleton pregnancies (17.4%, 336/1929) in the period before 1994 (P = 0.01), as well as in the period between 1994 and 1996 [19.1% (4/21) in twins versus 6.7% (78/1148) in singletons; P = 0.05]. In the period since 1997, the rate was similarly low in both groups [0.9% (1/111) in twin pregnancies versus 1.8% (89/5050) in singletons; P = 1.0]. Similar trends were observed for the transmission to twin A or when restricting the analysis to the last pregnancy for each woman enrolled in the cohort. The only case of transmission after 1997 concerned a mother who started prenatal care in the third trimester of the pregnancy, began HAART at 36 weeks, and delivered by emergency Caesarean section at 38 weeks with an HIV-1 RNA concentration of 8115 copies/ml.
We studied risk factors for MTCT in the period up to 1996 (Tables 3 and 4). The difference in MTCT between twin and singleton pregnancies remained significant in term deliveries, among multiparas, following premature rupture of membranes, or emergency Caesarean (Table 3). After adjustment in logistic regression for gestational age, parity, geographical origin, maternal CD4 cell count and mode of delivery, multiple pregnancy remained associated with MTCT, with an adjusted odds ratio of 2.3 (95% CI, 1.1–4.4; P = 0.03; Table 4). The association was especially strong in case of premature rupture of the membranes (50% in twins versus 17.4% in singletons, P = 0.01), with an adjusted odds ratio of 4.5 (95% CI, 1.2–16.9; P = 0.03). The first-born twin was infected in all six cases of MTCT which occurred in the 28 twin pregnancies with premature rupture of membranes, and the second twin was also infected in one case. In the absence of premature rupture of the membranes, MTCT was similar in both groups (14.7% versus 11.9%; adjusted odds ratio, 1.4; 95% CI, 0.5–4.0; P = 0.55). Only 3.5% (2/55) of first-born twins delivered by elective Caesarean section were infected versus 10.3% (6/58) of those born delivered vaginally.
There was no significant effect of gender on transmission. For the first-born, 8.0% (5/62) of boys were infected versus 10.5% (9/86) of girls (P = 0.7). For the second-born, these proportions were respectively 1.5% (1/66) versus 3.7% (3/82; P = 0.4).
In the French Perinatal HIV Cohort, the incidence of twin deliveries increased over 20 years. The increases in maternal age and in the proportion of African women, factors know to be associated with multiple gestation [27–29], largely accounted for this trend. We cannot exclude that a few pregnancies may have resulted from assisted reproductive techniques, but the cohort protocol started recording this data after the study period.
The greatest difference in pregnancy outcome was the high rate of preterm delivery in twins versus singletons, consistent with population-based data from France [29,30] and elsewhere . Over one half of the twins were delivered before 37 weeks, and severe prematurity (< 33 weeks) in twins was fourfold that in singletons (12.5% versus 3.1%). The other important differences were a higher rate of emergent Caesarean section in twins, without a difference in the elective Caesarean section rate, and a trend towards more premature rupture of the membranes. This trend did not reach statistical significance, but data on membrane rupture was not available for the period 2001–2004. Neither the types of ART regimens used, nor the maternal plasma viral load at delivery differed between twins and singletons, however twins had a shorter median duration of ART, which reflected the higher incidence of preterm delivery.
The MTCT rate in our study was significantly higher in twin pregnancies than in singleton pregnancies, in the period before any antiretroviral prevention was available, as well as in the period of zidovudine monotherapy. In the HAART era, i.e., since 1997, we found that twins were no longer at a significantly increased risk of HIV transmission [0.9% (95% CI, 0.02–4.9) for twins versus 1.8% (95% CI, 1.4–2.2] for singletons. It is unlikely that the higher transmission rate in twin pregnancies could be accounted for simply by the fact that twice as many children are exposed, because when considering the first-born child the transmission rate remained significantly higher in multiple pregnancies than in single pregnancies. To our knowledge, the only previous cohort study comparing transmission in twin and singleton pregnancies performed in Italy in the 1980s and including only 22 twin pairs, did not observe an increased risk for twins . Twins may be exposed to an increased transmission risk for several reasons, such as premature rupture of the membranes and preterm delivery [1,2].
We found that transmission of HIV occurred more frequently to the first-born twin than to the second twin. This result is consistent with data from an international registry on 115 twin deliveries , whereas no difference in MTCT was found in a large study from Malawi . The greater risk of transmission to first twin is indirect evidence for ascending route or birth canal exposure [31,32]. Other indirect evidence is the higher risk of infection for the first-born twin in case of premature rupture of the membranes, as well as the protective effect of elective Caesarean section, i.e., before labour and with intact membranes, which we observed in the period up to 1996. The impact of premature rupture of the membranes has been previously reported for singletons [1,2], and the role of the mode of delivery has been widely published previously [12,33,34]. The international twin registry also found that Caesarean section was associated with a lower rate of transmission to both the first twin, and to a lesser degree the second twin. However, because over one-half of twin deliveries occur preterm, Caesarean section cannot be planned in most cases.
The data that is most relevant to current obstetric and medical care and counselling concerns the current HAART era. Our findings indicate that twins, as well as singletons, are at low risk of MTCT HIV transmission when the mother has an undetectable viral load with HAART. Although the statistical power was limited due to the small number of children who were infected during this period, this offers some reassurance for women with twin pregnancies. Nonetheless, women considering infertility therapy, which increases the likelihood of multiple pregnancy, can be counselled that such an occurrence should be avoided if possible. In cases of multiple pregnancies in HIV-infected women, management must take into account the risk of preterm premature rupture of the membranes and preterm delivery. In particular, we would recommend starting effective ART no later than beginning of the second trimester.
This study was supported by the French National Agency for AIDS Research (ANRS), Paris, France. We thank Valerie Benhammou, Yassine Benmebarek, Corinne Laurent, Elisa Ramos, Marlène Peres, Nacima Chernai, Nancy Zeller, Aziz Diop, Paulette Huyhn, Leila Ouzrout, Thierry Wack.
The following persons and institutions participated in the ANRS French Perinatal Cohort (EPF):
Aix-en-Provence, Brusquet Y., Opimel P., Tadrist B., Thevenieau D., Tramier D.; Amiens, Boulanger J.C., Douadi Y., Gondry J., Horle B., Pautard B., Roussel C., Schmidt J.L., Smail A., Vergne C.; Angers, Achard, Binelli C., Chennebault J.M., Fournie A., Grosieux P., Rialland X.; Argenteuil, Allizy C., Brault D., Genet P., Piquet, Rischebe, Tordjeman N.; Aubervilliers, Rozan M.A.; Basse-Terre, Couchy B., Sibille G., Sid Elmrabt S.; Bastia, Bastien, Belgodere, Colombani D., Lonrenzi, Pincemaille O., Salvetti A., Turquini; Bayonne, Bonnal F., Cayla C., Chabanier C., Guerre P., Hernandorena X.; Besançon, Bassignot A., Bettinger Lab M., Chirouze C., Estavoyer J.M., Leroy J., Maillet R., Schaal J.P.; Blanc Mesnil, Bajer A., Balde P.; Bobigny, Deny P.; Bondy, Benoist L., Carbillon, François F., Jeantils V., Lachassinne E., Rodriguez A., Uzan-Cohen M.; Bordeaux, Beylot J., Brun J.L., Douard D., Elleau C., Fleury H., Guyon F., Horovitz J., Lacoste, Leng J.J., Masquelier B., Morlat P., Pontgahet M., Ragnaud, Roux D., Schaeffer V.; Boulogne Billancourt, Gilles I., Zenaty D.; Bourg La Reine, Gantzer A.; Bullion, Colin-Gorki A.M.; Caen, Barjot P., Brouard J., Freymuth F., Goubin P., Muller G., Petit J., Six M.; Cayenne, Delattre P., Elenga N., Magnien C., Patient G.; Clamart, Bornarel D., Chambrin V., Clech L, Dehan M., Dommergues, Foix L., Frydman R., Keros L., Vial M.; Clichy, De Curtis A., Levardon M., Mazy F.; Colombes, Crenn-Hebert C., Engelmann P., Ferreira C, Floch-Tudal C., Gaba S., Joras M., Mandelbrot L., Marty L, Mazy F., Meier F.; Compiègne, Coicaud M., Lagrue A., Meriem D.; Corbeil Essonnes, Blasquez, Daveau C., Devidas A., Lotfy N.; Courbevoie, Botto, Bourdon P.; Creil, Cesbron P., Cordier F., Devulder G., Duval-Arnould M., Kingue-Ekollo C.; Creteil, Cortey A., Delacroix I., Elharrar B., Garrait V., Huraux-Rendu C., Paniel J.B., Touboul C.; Dijon, Buisson M., Guerin M.N., Kohli, Martha S., Reynaud I., Rousseau T., Sagot P.; Dourdan, Guth, Ercoli V.; Drancy, Boddaert M.; Dreux, Denavit M.F., Garnier J., Roudiere, Tribalat S.; Elbeuf, Lahsinat K., Paquet M., Pia P.; Evreux, Allouche C., Elhaik, Pascal C., Toure K.; Evry, Farvacque R., Grall F., Khanfar, May A., N'guyen R.; Fontainebleau, All-Issa K., Cosnefroy, Cote, Dallot M.C., Fillipini, Kalengi, Lhuillier P., Routier C.; Fort de France, Cabie A., Cecile W., Hatchuel Y., Mezin R., Ouka M., Sainte-Rose D.; Gonesse, Balde P., Dauptain G., Lobut J.B. Paindaveine; Ivry, Gervais A., Jault T., Jrad I., Pathier D., Lagny sur Marne, Agbo P., Algava G., Chalvon Dermesay A., David Ouaknine F., Gourdel B., Lanty C., Lerudulier C., Pfeiffer F.; Le Chesnay, Beal G., carre N, Harzic M., Hentgen, Jacquemot M.C., Lasfargues P., Messaoudi F., Teboul; Le Kremlin Bicêtre, Bader-Meunier B., Desfraissy J.F., Goujard C., Fridmann S., Peretti D.,; Le Lamentin, Chout, Monlouis M.; Lille, Bocket L., Codaccioni X., D'Angelo S., Delmas S., Hammou Y., Mazingue F., Vanderstichele S.; Limoges, Alain J., Denis F., Rogez S., Tabaste J.L., Venot C.; Longjumeau, Abbara A., Bailly Salin P., Blanchard I., Bronstein R., Dalvilic S., De San Pedro, Lemercier Y., Seaume H.; Lyon, Andre, Bertrand Y., Brochier C., Communal P.H., Cotte L., Kebaili K., Raudrant D., Ritter J., Roussouly M.J., Tardy, Thouarain V.; Mantes La Jolie, Berardi J.C., Botto C., DeLanete A., Doumet A., Furioli J., Granier F., Grise, Salomon J.L., Wipff P.; Marseille, Blanc B., Cravello L., Deboisse P., Gallais H., Michel G., Pelloux, Tamalet C., Thuret I.; Meaux, Crumiere C., Demachy M.C., Karaoui L., Lefevre V., Michel F., Morel B.; Melun, Kleitz, Le Lorier B., Pauly-Ravelly I., Ponge B.; Meulan, Robichez, Seguy D.; Montfermeil, Camus M., Dehlinger M., Echard M., Mullard C., Rideau F., Ropert, Talon P.; Montpellier, Benos P., Boulot P., Bouzinger, Dechaud H., Guigue N., Laffargue F., Lalande M., Nicolas J., Reynes J., Segondy B., Sobierajski J., Vendrel; Montreuil, Heller-Roussin B., Saint-Leger S., Winter C.; Nancy, Delaporte M.O., Hubert C., May T., Neimann L., Schweitzer M.; Nanterre, Karoubi P., De Sarcus B.; Nantes, Auger M., Billaudel S., Brunet-François C., Boog G., Ferre V., Mechinaud F., Reliquet V., Winer N.; Neuilly sur Seine, Berterottiere D., Boto; Nice, Allione J., Bongain A., Cottalorda J., Couderc A., Deville A., Durant J., Galiba E, Gillet J.Y., Monpoux F.; Nîmes, Arnaut A., Barbuat C., Carles M.J., Dendale J., Ferrer C., Rouannet I., Sotto A.; Orléans, Arsac P., Barthez, Bondeux D., Mesnard L., Tescher M., Werner E.; Orsay, De Gennes C., Devianne, Drisset, Isart V., Razon; Perpignan, Bachelard G., Bachelard B., Medus M., Roudil; Pointe-à-Pitre, Bardinet F., Bataille, Duffilot D., Samar K; Sow M.T., Strobel M.; Poissy, Nisand I., Pavard, Rousset M.C.; Pontoise, Blum L., Danne O., Hervio P., Mouchnino G., Muray; Reims, Beguinot I., Graesslin O., Ingrand D., Munzer M., Quereux C., Remy G., Rouger C., Saniez D.; Rouen, Borsa-Lebas F., Brossard V., Buffet-Janvresse C., Clavier B., Debab Y., Marpeau L., Vannier J.P.; Saint-Etienne, Berger C., Billiemaz K., Fresard A., Pozzetto B., Stephan J.L., Varlet M.N.; Saint-Denis, Allemon M.C., Ekoukou D., Ghibaudo N., Khuong M.A., Luzolo A., Mechali D., Normand V., Poupard M.P., Retbi J, Retbi J.M., Rotten D., Seffert; Saint-Germain en Laye, Guyot B., Michelon, Narcy P.; Saint-Martin, Bissuel F., De Caunes F., Elouedghiri, Laborde O., Walter V.; Saint-Maurice, Jeny R., Robin; Sèvres, Belaisch-Allart J., Segard L.; Strasbourg, Brettes, Cheneau C., David, Dreyfus M., Entz-Werle N., Favreau J.J., Fischer P., Lang J.M., Langer B., Lutz P., Nisand I., Partisani M., Ritter, Treisser A., Vayssiere C., Weill M.; Suresnes, Clement, Colau J.C.; Toulouse, Armand, Berrebi A., Cohen M., De Coster P., Lecuyer I., Massip P., Perez-Baronne, Puel J., Reme J. M., Tricoire J.; Tours, Bansard H, Bastides F., Besnier J.M., Borderon J.C., Barin F., Lansac, Lionnet C., Marchand S Nau P, Pascale, Perrotin F., Potin J., Sigogneau H; La Trinité, Hugon N.; Vandoeuvre Les Nancy, Finance, Le Faou; Villejuif, Dussaix E.; Villeneuve Saint Georges, Bantsimba J., Bonnard C, Camuss A, Chace A., Guillot F., Jubin C., Maria B., Montaland F, Patey O., Richier L, Stampf F., Tran Van P.; Villepinte, Boulanger M.C., Broyard A., Caruge, Caubel P., Chitrit Y., Delassus J.L., Goldenstein, Le Pennec M., Poulen, Scart G., Zakaria A.; Vitry sur Seine, Lacroix-Coutry A., Sturbois G. Paris: ASE St Vincent de Paul, Commeau A.; Centre d'Hémobiologie Périnatale, Parnet Mathieu F.; Clinique Notre Dame de Bon Secours, Ayral D., De Kermadec S.; Groupe Hospitalier Cochin Tarnier Port-Royal, Brival, Cabrol D., Clement D., Compagnucci A., Desfeux P., Fikenstein, Firtion G., Goupil I., Henrion R., Launay O., Mandelbrot L., Pannier E., Pons J.C., Taulera,; Hôpital Boucicaut, Bardin C., De Bievre P., Gras V., Labussiere E., Lafay Pillet M.C., Parat S., Taurelle R.; Hôpital Robert Debré, Bensaid Ph., Blot P., Boissinot Ch., Cotten G., Faye A., Levine M., Ottenwalter A.,Oury JF, Schaller F., Vilmer E.; Hôpital Trousseau, Chenon, Dollfus C., Gabarg-Chenon A., Tabone, Vaudre G., Institut de Puériculture Brune, Dubois M., Hôpital Lariboisière, Brunner C., Ciraru-Vigneron N., Peynet J., Colonna R., Sanson-Lepors, Truc; Hôpital des Métallurgistes, Cheynier J.M., Hatem-Gantzer G., Heller-Roussin B., Rami M.; Hôpital Rothschild, Fritel, Lebrette M.G., Nicolas J.C., Uzan M., Wallet A.; Hôpital Saint-Antoine, Bouillie J., Bui E., Carbonne B., Meyohas M.C., Milliez J., Rodriguez J., Schrub S.; Centre Hospitalier Pitié Salpetrière, Agut H., Bricaire F., Daher S., Darbois Y., De Montgolfier, Dermer E., Deville-Chabrol A., Dommergues M., Huraux J.M., Noseda G., Pauchard M., Tubiana R., Hôpital Saint-Vincent de Paul, Boccara J.F., Francoual C., Krivine A., Lebon P.; Institut Mutualiste Montsouris, Carlus Moncomble C., Cohen H., Groupe Hospitalier Necker, Benachi A., Blanche S., Burgard M., Parat S., Rouzioux C., Viard J.P.; Hôpital Saint-Michel, Aufrant C.; Hôpital Bichat-Claude Bernard, Allal, Bastian H., Batallan A., Brun-Vezinet F., Bernard, Damond F., Darai E., Faucher Ph., Longuet P., Madelenat P, Matheron S., Mazy F, Moreau G., Proust A., Rajguru M, Simon F.; Hôpital Tenon, Berkane N., Chaux M.C., Herve F., Lebrette M.G., Uzan S.
Sponsorship: The EPF (Enquete Perinatale Francaise) Cohort was funded and supported by the Agence de Recherches sur le Sida et les Hepatites virales (ANRS), Paris, France.
© 2007 Lippincott Williams & Wilkins, Inc.