Association of raised titres of antibodies to Chlamydia pneumoniae with a history of pre-eclampsia
Dr C. Williamson, Institute of Reproductive and Developmental Biology, Imperial College London, Hammersmith Hospital Campus, Du Cane Road, London, UK.
Objective To establish the prevalence of Chlamydia pneumoniae (C. pneumoniae) infection in a pregnant UK population and to investigate whether C. pneumoniae infection is more common in women with a previous history of pre-eclampsia.
Design Prospective study.
Setting Academic Hospital.
Population Ninety-one pregnant women (54 parous and 37 nulliparous) at 16–22 weeks of gestation were studied. Of the parous women, 32 had a previous history of pre-eclampsia.
Methods Peripheral blood was drawn for C. pneumoniae antibodies between 16–22 and 28–40 weeks of gestation. C. pneumoniae antibodies were measured using a solid-phase enzyme immunoassay. According to pregnancy outcome, women were categorised into normal, gestational hypertension and pre-eclampsia groups.
Main outcome measures Serum levels of IgG, IgA and IgM C. pneumoniae antibodies.
Results Prevalence of seropositivity to C. pneumoniae was 77%. Parous women had significantly higher levels of IgA and IgM C. pneumoniae antibodies than nulliparous women (P < 0.04). Parous women with previous pre-eclampsia were found to have higher levels of antibodies than parous women with a normal obstetric history (P≤ 0.003). There was no difference in the antibody levels in women with different pregnancy outcomes.
Conclusions The longitudinal data do not indicate an association between C. pneumoniae infection and pre-eclampsia. However, the subgroup analysis of parous women demonstrated raised C. pneumoniae antibodies in the women with previous pre-eclampsia, and therefore suggests that there may be an association between C. pneumoniae and the disease in this group.
Pre-eclampsia is a multisystem disorder that affects approximately 3% of primigravidae in the UK and is a major cause of maternal and perinatal mortality worldwide. Hypertension and proteinuria developing in the second half of pregnancy are defining clinical features. However, pre-eclampsia is a multisystem disorder characterised by widespread endothelial cell dysfunction.1 The aetiology of pre-eclampsia is multifactorial. Pre-eclampsia can be familial, is associated with uteroplacental ischaemia, and is more common in women with pre-existing endothelial cell damage.2
Biochemical evidence for endothelial dysfunction includes increased circulating levels of fibronectin, factor VIII antigen, endothelin and reduced maternal plasma levels of 6-keto-prostaglandin F1α.3 Pharmacological evidence includes impaired endothelium-dependent relaxation in arteries from women with pre-eclampsia compared with normotensive pregnant women.4 Morphological evidence of damaged endothelium is found in the spiral arteries of the placental bed and in glomerular capillaries.5
It has been postulated that the endothelial dysfunction of pre-eclampsia is part of a generalised intravascular inflammatory reaction.6 This hypothesis is supported by the demonstration of activated circulating blood leucocytes in normal pregnancy and, to an even greater extent, in pre-eclampsia.7 Furthermore, inflammatory cytokines such as tumour necrosis factor α (TNF-α), its soluble receptors and interleukin 6 (IL-6) are found to circulate at higher concentrations in women with pre-eclampsia than in those with normal pregnancy.8 In pre-eclampsia there is also an increase in intracellular reactive oxygen species.7
One reason that women destined to develop pre-eclampsia might have a more intense inflammatory response than healthy pregnant women would be a coincidental or preceding inflammatory stimulus, such as infection. There have been reports of an association between pre-eclampsia and urinary tract infection9 and malaria10 but no studies have demonstrated a causative mechanism. During healthy pregnancy there is a shift in the maternal T-lymphocyte response to infection. The possible consequences are a more favourable environment for the implanting embryo at the uteroplacental interface, but at the expense of an increased maternal vulnerability to intracellular pathogens.
Chlamydia pneumoniae (C. pneumoniae) is a gram-negative intracellular bacterium, which was first identified 13 years ago. It is a recognised respiratory tract pathogen and has been associated with vascular disease. It has been hypothesised that C. pneumoniae acts in a similar fashion to other chlamydial species, causing chronic inflammation, which can result in clinical disease developing years following the primary infection.11 There is both direct and indirect evidence linking C. pneumoniae infection with vascular disease. Sero-epidemiological studies have shown the presence of anti-C. pneumoniae antibodies to be significantly correlated with the presence of coronary artery12 and cerebrovascular disease.13 The organism has been found in diseased cardiovascular tissue significantly more often than in control vascular tissue,14 and there is evidence of the presence of C. pneumoniae in atherosclerotic vascular tissue throughout the arterial system.15 Using animal models, it is possible to show that C. pneumoniae can be disseminated haematogenously following pulmonary infection, and that it shows a tropism for vascular tissue,16 where it has the capacity to infect and multiply in cells of the arterial wall.17 These cells include endothelial cells, smooth muscle cells and macrophages. There is one study linking C. pneumoniae infection with essential hypertension in the non-pregnant state.18C. pneumoniae infection causes similar changes in inflammatory markers to those seen in pre-eclampsia, and there has been one study that provided evidence for an association between pre-eclampsia and C. pneumoniae infection,19 although another study did not confirm this.20 One recent study reported a trend for increased levels of IgG antibodies against C. pneumoniae in early-onset pre-eclampsia, but this was not statistically significant.21
In this study, all three classes of serum antibodies were detected: specific IgM indicating acute infection, specific IgG indicating past infection and specific IgA indicating chronic infection.
We hypothesised that the presence of C. pneumoniae infection could play a role in the aetiology of pre-eclampsia. The aims of this study were to establish the seroprevalence of C. pneumoniae infection in a UK population, and to investigate whether evidence of previous C. pneumoniae infection is more common in women with a history of pre-eclampsia.
Women were recruited prospectively at St Thomas' Hospital and the Chelsea and Westminster Hospital, London, for a separate study relating to the prevention of pre-eclampsia.22 In that study, women were identified as being at increased risk of pre-eclampsia on the basis of (i) abnormal uterine artery Doppler waveform (defined as a resistance index ≥95th centile for gestation and/or the presence of an early diastolic notch) or (ii) previous history of pre-eclampsia requiring delivery before 37 weeks of gestation. The risk of pre-eclampsia in a subsequent pregnancy for this group is approximately 10%, that is several fold higher than other parous women. At recruitment (between 16 and 22 weeks of gestation), women were randomly allocated to receive either combined treatment with vitamins C and E, or placebo (Fig. 1).
Pre-eclampsia was defined according to the guidelines of the International Society for the Study of Hypertension in Pregnancy (ISSHP). Pre-eclampsia was defined as two recordings of diastolic blood pressure (Korotkoff V) ≥90 mmHg at least 4 hours apart in association with proteinuria in a previously normotensive woman. Proteinuria was defined as ≥300 mg/L in a 24-hour collection, or two readings of 2+ or higher on dipstick analysis of midstream or catheter urine specimens if a 24-hour collection was not available. Gestational hypertension was defined as two recordings of diastolic blood pressure ≥90 mmHg at least 4 hours apart without proteinuria.
Blood specimens were taken prior to randomisation (16–22 weeks of gestation) and prior to delivery (between 28 and 40 weeks of gestation). C. pneumoniae antibodies were measured using a solid-phase enzyme immunoassay (EIA) [Labsystems IgG (61-11-300), IgA (61-11-310), IgM (61-11-320)] for C. pneumoniae. For C. pneumoniae infection detection, it has a sensitivity of 97%, positive predictive value of 99%, negative predictive value of 98%, within-run reproducibility (CV) of 4–6% and between-run reproducibility (CV) of 12–18%.
Phase 1 (analysis of pre-randomisation samples)
Phase 1A. C. pneumoniae has been shown to be more common in the inhabitants of households with young children. Therefore, the levels of IgG, IgA and IgM anti-C. pneumoniae antibodies were compared separately between all multiparous women (those with a previous pregnancy resulting in a live birth or with delivery after 24 weeks of gestation) and all primiparous women.
Phase 1B. To investigate whether C. pneumoniae infection is associated with a previous history of pre-eclampsia, the IgG, IgA and IgM C. pneumoniae antibody levels in the pre-randomisation sample from the 32 women with a previous history of pre-eclampsia were compared with 22 multiparous women with a normal obstetric history. To ensure that the observed differences in C. pneumoniae antibodies were not a result of differences in total immunoglobulin levels, the levels of total IgG, IgA and IgM were also measured in the 54 multiparous women using an immuno-rate-turbidometric method.
Phase 2 (analysis of longitudinal samples)
To investigate whether the levels of C. pneumoniae IgG, IgA and IgM change from mid-gestation to term in women with differing pregnancy outcomes, the absolute levels of antibodies between the paired blood specimens from women with pre-eclampsia, gestational hypertension and normal pregnancy outcomes were compared. Only women in the placebo arm of the study were studied.
The study received ethical approval from the ethical committees of the hospitals where the women were recruited (St Thomas' Hospital and the Chelsea and Westminster Hospital). All women who participated in the study gave informed consent.
Data are expressed as median (interquartile range). Mann–Whitney U test was used to identify differences in the mean values between two groups, and Kruskal–Wallis H test when there were more than two groups. Statistical analysis of the data was performed using the software SPSS for Windows 11, SPSS.
1A. Of 130 subjects studied for evidence of C. pneumoniae infection, 100 (77%) were seropositive. Multiparous women had significantly higher levels of C. pneumoniae antibodies (enzyme immunoassay units [EIU]) than primiparous women for IgA [10.9 (18.8) vs 6.1 (12.6), Mann–Whitney U, P= 0.009] and IgM [0.4 (0.5) vs 0.3 (0.4), P= 0.04], but not for IgG [114.9 (149.1) vs 72.8 (162.3), P= NS].
1B. Details of the 54 multiparous women are given in Table 1. Women with a previous history of pre-eclampsia were found to have significantly higher levels of IgG [177.3 (190.4) vs 62.8 (101.1)], IgA [17.8 (22.9) vs 5.5 (5.9)] and IgM [0.6 (0.7) vs 0.2 (0.2)] than women with a normal obstetric history (Mann–Whitney U, P= 0.003, P= 0.002 and P < 0.001, respectively) (Table 2). There were no differences between the two groups in the total immunoglobulin levels (g/L) [IgG 8.0 (3.9) vs 8.9 (3.0), IgA 1.8 (1.1) vs 1.8 (1.2) and IgM 1.5 (0.6) vs 1.2 (1.2), P= NS].
Table 1. Subject characteristics (multiparous women only). Data are expressed as median (interquartile range) or as absolute numbers [percentage %]. Body mass index (BMI) has been calculated using the formula: BMI = (weight in kilograms)/(height in meters)2.
|1st sample||20 (1)||20 (0)||NS|
|2nd sample||36 (4)||36 (2)||NS|
|Age (years)||32 (10)||33 (6)||NS|
|Ethnic origin|| || ||NS|
|Caucasian||12 ||22 || |
|Afro-Caribbean||9 ||10 || |
|Other||1 ||0 || |
|Parity|| || ||NS|
|1||14 ||24 || |
|>1||8 ||8 || |
|Smoking status|| || ||NS|
|Non-smoker||21 ||30 || |
|Smoker||1 ||2 || |
|Body mass index||26.0 (8.4)||25.1 (4.0)||NS|
Table 2. C. pneumoniae antibody values in the pre-randomisation blood specimen from parous women with a normal obstetric history or with previous pre-eclampsia. Data are expressed as median (interquartile range).
|IgG (EIU)||62.8 (101.1)||177.3 (190.4)||0.003|
|IgA (EIU)||5.5 (5.9)||17.8 (22.9)||0.002|
|IgM (EIU)||0.2 (0.2)||0.6 (0.7)||0.001|
When the absolute levels of C. pneumoniae antibodies were compared prospectively in women with different outcomes in the index pregnancy (pre-eclampsia, gestational hypertension, normal), no significant differences were seen except for the pre-delivery IgM antibody levels in primiparous women. The results for the 32 multiparous women are summarised in Table 3, and those for the 37 primiparous women in Table 4. In addition, there were no differences in the C. pneumoniae antibody levels between the pre-randomisation and the pre-delivery specimens. C. pneumoniae antibodies were detected in 80% of multiparous women with a normal pregnancy outcome, compared with 80% of women with gestational hypertension and 92% with pre-eclampsia. The corresponding figures for primiparous women were 70%, 75% and 83%, respectively.
Table 3. C. pneumoniae antibody levels according to pregnancy outcome in multiparous women. Data are expressed as median (interquartile range).
|Pre-randomisation||99.4 (192.1)||234.7 (202.3)||116.6 (145.6)||NS|
|Pre-delivery||85.5 (181.1)||227.5 (221.9)||85.7 (162.0)||NS|
|Pre-randomisation||7.4 (19.6)||24.9 (21.8)||11.6 (24.2)||NS|
|Pre-delivery||8.3 (18.8)||24.9 (18.8)||13.4 (28.3)||NS|
|Pre-randomisation||0.4 (0.2)||0.6 (1.2)||0.4 (0.9)||NS|
|Pre-delivery||0.4 (0.5)||0.2 (0.8)||0.4 (0.9)||NS|
Table 4. C. pneumoniae antibody levels according to pregnancy outcome in primiparous women. Data are expressed as median (interquartile range).
|Pre-randomisation||110.3 (169.3)||57.4 (121.7)||131.0 (158.4)||NS|
|Pre-delivery||102.4 (173.7)||33.3 (12.8)||100.0 (118.5)||NS|
|Pre-randomisation||6.3 (14.8)||5.9 (6.5)||11.1 (20.1)||NS|
|Pre-delivery||6.8 (23.1)||2.7 (1.4)||9.7 (9.6)||NS|
|Pre-randomisation||0.3 (0.5)||0.2 (0.2)||0.2 (0.6)||NS|
|Pre-delivery||0.3 (0.4)||0.1 (0.0)||0.2 (0.4)||0.02|
The first study that provided evidence of an association between C. pneumoniae infection and pre-eclampsia reported that IgG antibodies to C. pneumoniae at a titre of at least 1:16 were more common in women with pre-eclampsia (25 of 37) than in women without (15 of 37) (odds ratio 3.1; 95% confidence interval 1.2–7.9).19 In a similar study, the odds ratio for the proportion of pre-eclamptic women (9 of 11) versus controls (52 of 73) with a positive C. pneumoniae titre was 1.8 (95% confidence interval 0.4–9.1).20 Finally, in a recent study, the authors reported a trend for women with early-onset pre-eclampsia (<34 weeks of gestation) to have significantly higher levels of IgG antibodies to C. pneumoniae compared with women with late-onset pre-eclampsia (=34 weeks of gestation), normotensive intrauterine growth retardation or normal pregnancy, but this was not statistically significant.21 In our study, we have shown that women with a history of pre-eclampsia have higher titres of all classes of C. pneumoniae antibodies than women without a history of pre-eclampsia, whereas their total immunoglobulin levels remain unchanged. These data provide additional evidence that there may be an association between C. pneumoniae infection and pre-eclampsia. However, the immunoglobulin titres were not different when samples were studied longitudinally from women with different pregnancy outcomes. Therefore, it is possible that the association observed in parous women is an epiphenomenon. Women with previous pre-eclampsia may also be more susceptible to C. pneumoniae infection in the postnatal period, or may have higher exposure, possibly as a result of spending time on the neonatal unit.
In Phase 2, the prospective analysis of the variation in antibody levels from mid-gestation to term, we found no association between pregnancy outcome and the levels of C. pneumoniae antibodies, nor significant differences in the C. pneumoniae antibody levels between the pre-randomisation and the pre-delivery specimens. However, the surprisingly high background prevalence of C. pneumoniae infection in the study population meant that the study did not have sufficient power to detect such a change with the number of subjects used. We estimate that, in order to establish a significant difference between the pregnancy outcome subgroups (pre-eclampsia, gestational hypertension, normal), assuming a difference in mean values of 20 EIU, a standard deviation (SD) of 100 EIU, a study power of 80% and a type I error of 0.05, we will need a study group of 800 patients and a similar number of controls. In addition, the comparison of paired pre-randomisation and pre-delivery samples may not be appropriate, as the disease process in pre-eclampsia starts in early pregnancy, many weeks before the clinical features are apparent. Thus, if C. pneumoniae plays a role in the aetiology of pre-eclampsia, it may be that the presence of chronic infection at the time of conception is more important than subsequent reactivation of the infection later in pregnancy.
We selected an indirect solid-phase enzyme immunoassay for a semi-quantitative determination of C. pneumoniae antibodies. The serologic methods used routinely for diagnosis of C. pneumoniae infection are complement fixation (CF) and micro-immunofluorescence assay (mIF).23 Although the CF test detects primary infection, it is unable to distinguish between antibodies to the three Chlamydial species (C. pneumoniae, C. trachomatis and C. psittaci). The mIF test is species specific and sensitive, however, it is technically demanding and it is not suitable for large volume testing. The solid-phase enzyme immunoassay has been reported to show no cross-reactivity with C. trachomatis; the specificity for C. pneumoniae having been confirmed by the manufacturers of the kit using sera with known antibody to C. trachomatis. There have been no reports of an association between pre-eclampsia and either C. trachomatis or C. psittaci. However, pregnant women exposed to C. psittaci can contract gestational psittacosis with atypical pneumonia, sepsis and placental insufficiency resulting in preterm birth or miscarriage.24
C. pneumoniae infection is associated with inflammatory changes similar to those reported in pre-eclampsia3,6–8 and it is possible that chronic infection or reactivation of a persistent low grade infection could further exacerbate the inflammatory changes of normal pregnancy, resulting in pre-eclampsia.
Inflammation and infection are well-documented risk factors in vascular disease such as myocardial infarction and stroke, but increasing attention is being paid to specific bacterial and viral species as candidates that can worsen the atherosclerotic disease process. It has been hypothesised that the presence of C. pneumoniae within the endothelial smooth muscle cells and macrophages of the arterial wall acts as a stimulus to chronic inflammation. It is possible that the organism does not have to be viable, but that highly immunogenic epitopes persisting within macrophages after death could initiate and maintain the chronic inflammatory state. In vitro work has shown that human monocytes infected with C. pneumoniae produce cytokines as part of the cell-mediated immune response. These include TNF-, IL-6 and IL-1β,25 which are also raised in pre-eclampsia. These cytokines increase expression of leucocyte adhesion molecules and activation of T cells, increasing the immune response. In the circulation, they can promote a hypercoagulable state by activating platelets,26 increasing hepatic synthesis of acute phase proteins, including fibrinogen, and inducing tissue factor release from endothelium with subsequent activation of the coagulation cascade.27
This study revealed a higher prevalence of C. pneumoniae than has previously been reported in Western populations. In Western countries, C. pneumoniae rarely causes infections before school age, but it is an important childhood pathogen in tropical urban areas. Epidemics of C. pneumoniae occur every five to seven years with primary infection occurring in teenage years and the prevalence rising with age.16 Cycles of reinfection are common28 and serum antibodies do not appear to have a protective effect. The prevalence of serum antibodies starts to rise steadily at school age in Western countries, and it is estimated that between 50% and 70% of adults are seropositive.9 A recent study has reported the prevalence of IgG seropositivity for C. pneumoniae as 60% in postmenopausal North American women.29 The median age of the women in this study was 32 years and the prevalence of C. pneumoniae infection was 77%. This high prevalence may be largely due to the selection criteria used for this study, as the majority of subjects were from an inner city population and had an abnormal uterine artery Doppler waveform at 24 weeks of gestation. Alternatively, the high background level could in part be attributed to the improved detection of antibody by the EIA method. It would be of interest to compare the prevalence of seropositivity in a matched, non-pregnant UK urban population in subsequent studies. In a Canadian population, von Dadelszen et al.21 found a 53% prevalence of IgG C. pneumoniae seropositivity for normal pregnancy compared with 67% for early-onset pre-eclampsia, 38% for late-onset pre-eclampsia and 30% for normotensive intrauterine growth retardation. Higher titres of C. pneumoniae antibodies have been found in subjects who are older,30 in smokers and in those of Afro-Caribbean origin,31 but there were no significant differences in our study populations for these variables.
Despite the significant differences in C. pneumoniae antibody levels found between multiparous women with previous pre-eclampsia and those with a normal obstetric history (Phase 1B), the majority of women in both groups have levels of C. pneumoniae antibodies consistent with either chronic or acute infection. Many women with previous C. pneumoniae infection did not develop pre-eclampsia in the index pregnancy. Therefore, if there is an association between C. pneumoniae infection and pre-eclampsia, it is not necessarily simple and may not be causative. High antibody levels could simply be a consequence of the altered T-cell response during pregnancy, which increases maternal vulnerability to intracellular infections. However, this would not explain the higher levels of C. pneumoniae antibodies in women with a history of pre-eclampsia. As pre-eclampsia is undoubtedly of multifactorial aetiology, it is possible that C. pneumoniae infection plays a role in the origin of the disease in only a subgroup of predisposed women.
The normal obstetric history group for Phase 1 was recruited on the basis of an abnormal uterine artery Doppler waveform at 24 weeks of gestation. It is possible that this group has abnormal endothelial function compared with women with a normal waveform. However, their normal obstetric history suggested that endothelial function in previous pregnancies may not have been impaired.
In conclusion, this study has demonstrated that women with a history of pre-eclampsia have higher titres of all classes of C. pneumoniae antibodies, compared with women without such a history. However, when the levels of C. pneumoniae antibodies were compared in primiparous and multiparous women with different pregnancy outcomes, we did not find any differences. The study was performed in an inner city population. It is therefore important that it is repeated with larger numbers in different populations. If C. pneumoniae infection is found to play a role in the aetiology of pre-eclampsia, this will have implications for the investigation and treatment of affected women.