4. Discussion
The aim of the present study was to investigate the association of maternal viral infections during pregnancy with pregnancy complications and delivery outcomes. The results are based on the comparison of mothers of cases and uninfected controls.
The major findings of our study are that viral infections during pregnancy are associated with higher incidences of threatened miscarriage, threatened preterm delivery and anaemia. On the contrary the risk of gestational diabetes was decreased in infected mothers. Gestational age was longer in the infected group as compared to controls. Birth weight and the risk of preterm birth or IUGR did not differ significantly between the two study groups. Hepatitis B infection during pregnancy was associated with shorter gestational age and lower birth weight.
Miscarriage occurs in about 20% of pregnancies and may result in severe psychological and physiological issues for the patient. A thorough analysis of the available literature has found that influenza infection was associated with an elevated risk of spontaneous abortion, while the effect of hepatitis B and herpes simplex virus remained controversial.12 Our data also support the potential deleterious effects of viral infections on the incidence of threatened miscarriage. However, the database of the HCCSCA is not suitable for analysing miscarriage data since it only contains records of preterm and term new-borns.
Preterm birth affects approximately 12-15% of pregnancies and in more than 50% of the cases no risk factors are known. Chronic hepatitis B virus infection was found to increase the risk of preterm labour and birth in pregnant women.13,14 Increased rates of preterm birth are also reported in pregnant patients hospitalized with influenza virus infection, however, the limited amount of data does not permit firm conclusions.15,16 In our dataset we were not able to confirm an elevated risk of preterm delivery in mothers suffering from virus infections during pregnancy. A shorter gestational age and lower birth weight was verified only in the case of hepatitis B infection; however, these alterations did not reach the limits of preterm birth. On the other hand, the prevalence of threatened preterm labour (preterm uterine contractions or ultrasonographic signs exhibited by the cervix and lower uterine segment) was increased significantly among mothers with viral infections. The pathomechanism of viral infections in inducing preterm birth is still debated. Viral infections of the placenta may act as a factor sensitizing women to intrauterine bacterial infection, resulting in an inflammatory response to even low concentrations of bacteria.17 Intrauterine bacterial infection may then lead to preterm uterine contractility.
Gestational diabetes mellitus (GDM) affects 1-22% of all pregnancies globally, depending on the population and diagnostic criteria used.18 It is largely associated with severe pregnancy complications like prenatal morbidity, preterm labour, dystocia, etc. In our dataset GDM was found to be significantly less frequent among mothers affected by viral infections as compared to the control group. Regarding this finding we were not able to identify substantive publications in the literature. In the background of the development of GDM, the role of many cytokines and other immunological factors have been investigated.19 Pro-inflammatory cytokines and other inflammatory markers have been shown as predictors of diabetes.20,21 We only hypothesize that viral infections may alter the activity of some anti-inflammatory cytokines resulting in a protection against insulin resistance. On the other hand, this finding of ours may also be incidental since the number of investigated cases is relatively low (8 in 2,238 infected mothers vs. 458 in 54,993 controls).
In most countries of the world, the prevalence of anaemia in pregnancy is over 20%.22 It has been associated with prematurity, low birth weight, hypertensive disorders and adverse pregnancy outcomes.23,24 In our study viral infections during pregnancy were associated with a significantly elevated prevalence of maternal anaemia as compared to uninfected control mothers. We were unable to find significant data in the literature, that could explain this relatedness. In non-pregnant individuals autoimmune haemolytic anaemia has been reported in a number of viral infections.25 Moreover, parvovirus infection has been associated with red cell aplasia.26 In infants a mild viral infection was found to induce a significant decrease in haemoglobin that may persist for 14 to 30 days and may be difficult to distinguish from iron deficiency.27 In a rodent model it was suggested that viruses may trigger autoantibody-mediated anemia by activating macrophages through gamma-interferon production.28 On the other hand, anaemia may also act as a causative factor, enhancing susceptibility of pregnant mothers to viral infections.29
A further interesting finding of our study is that maternal smoking was significantly more frequent among mothers affected by viral infections as compared to the uninfected control group. Smoking is a well-established risk factor of various infections,30therefore, this finding is not surprising. However, it is worth mentioning that although maternal smoking is associated with unfavourable pregnancy outcome,31,32 this correlation could not be verified in the present study.
The strengths of our study are related to the large population-based data set of the HCCSCA in an ethnically homogeneous Hungarian (Caucasian) population. Pregnancy complications, gestational age and birth weight can be estimated with high accuracy since data were medically recorded. This study included all viral infections during pregnancy and these exposure data were based on multiple sources including prenatal maternity logbooks which provided prospective medically recorded data. Exposure time and potential confounders were also documented.
However, there are also some limitations of our study. The major weakness of our dataset is that the diagnosis of viral infections was based on reported data without serological evidence of the virus infection. Maternal smoking data were based partly on retrospective maternal information burdened by recall bias.33 A previous validation study of our group showed the low reliability of retrospective maternal self-reported information regarding smoking and alcohol drinking during pregnancy.34 Another weakness of our study is that only cases born between 1980 and 2009 were evaluated, thus the results of the recent decade could not be analysed in this field.
In conclusion the findings of this study suggest that viral infections during pregnancy do not exert a deleterious effect on birth outcomes. The incidences of threatened miscarriage threatened preterm birth and maternal anaemia are higher in pregnancies affected by viral infections. The incidence of gestational diabetes is lower after viral infections during pregnancy. Maternal smoking is more frequent in pregnancies with viral infections.