Dr P von Dadelszen, 2H30-4500 Oak Street, Vancouver, BC, Canada, V6H 3N1. Email email@example.com
Objective Pre-eclampsia involves a maternal inflammatory response that differs from both normal pregnancy and normotensive intrauterine growth restriction (IUGR). Our objective was to examine neutrophil Toll-like receptor (TLR), cryopyrin, nuclear factor-κB (NF-κB) subunit and interleukin-1β (IL-1β), and inflammatory cytokine profiles in women with pre-eclampsia or normotensive IUGR, as well as in normal pregnancy and non-pregnancy controls.
Design and method A case–control study was performed. We examined the messenger RNA (mRNA) and protein expressions of TLR4 and TLR2, mRNA levels of cryopyrin, IL-1β, NF-κB subunits p50 and p65, as well as maternal serum inflammatory cytokine profiles (IL-2, IL-6, tumour necrosis factor-α [TNF-α], interferon-γ [IFN-γ] and IL-10) in women with and without pre-eclampsia using real-time reverse transcription polymerase chain reactions, flow cytometry and multiplex immunoassays.
Setting A single tertiary maternity hospital in Vancouver, Canada.
Population Women with early-onset pre-eclampsia (<34 weeks of gestation, n = 25), women with late-onset pre-eclampsia (≥34+0 weeks of gestation, n = 25), women with normotensive IUGR (n = 25), women with normal pregnancy (n = 75) and non-pregnancy (n = 25) controls.
Results Women with pre-eclampsia (as a single combined group of early- and late-onset, and particularly in women with early-onset pre-eclampsia) had increased TLR2 and TLR4 mRNA and protein expressions elevated cryopyrin, NF-κB subunit, and IL-1β mRNA expression, and TNF-α:IL-10 and IL-6:IL-10 ratios compared with other groups.
Conclusions These data suggest that TLRs and cryopyrin may modulate the innate immune response of the maternal syndrome of pre-eclampsia, and might also trigger the differential inflammatory response existing between early onset pre-eclampsia and normotensive IUGR.
Pre-eclampsia is not only a major cause of maternal morbidity and mortality but is also associated with increased risk for developing cardiovascular disease later in life.1 It is a placenta-dependent disorder and is characterised by proteinuric gestational hypertension after 20 weeks of gestation.2 The activation of the innate immune response observed in normal pregnancy is exaggerated in pre-eclampsia.3,4
The origins of pre-eclampsia are complex and appear to lie in a mismatch between fetoplacental demands and uteroplacental supply,4 a situation which also occurs in normotensive intrauterine growth restriction (IUGR).5 Why there is a different maternal response to the similar underlying pathology remains unclear. When pre-eclampsia occurs early in pregnancy, it is particularly dangerous and more frequently associated with IUGR;6 whereas, at term, there is a predominance of fetal macrosomia.6
Endotoxin or lipopolysaccharide (LPS)-mediated inflammation is believed to be of critical importance in systemic infectious and non-infectious disorders.7,8 In an animal model, Faas et al.9 demonstrated that infusion of low-dose LPS induces a pre-eclampsia-like syndrome, including hypertension, proteinuria and glomerular endotheliosis. Lipolysaccharide leads to release of pro-inflammatory cytokines through activation of two major pattern recognition receptors present on innate immune cells, including the extracellular Toll-like receptor 2 (TLR2) and TLR4,10 and the intracellular cryopyrin, a nucleotide-binding oligomerisation domain protein, also known as the caspase-activating recruitment domain.11–13 Cryopyrin modulates LPS-induced caspase-1 activation and interleukin-1β (IL-1β) secretion.14 The TLRs and cryopyrin have been related to an attenuated immune response to LPS; this results in the production of inflammatory cytokines, in particular tumour necrosis factor-α (TNF-α) and IL-6, through the nuclear factor-κB (NF-κB) pathway.15,16 However, the role of peripheral blood neutrophil TLRs and cryopyrin in either normal or pre-eclampsia-complicated pregnancies has not been clarified.
The induction of signal transduction pathways is necessary for the activation of the innate immune response in pre-eclampsia.17 A probable pathway involves NF-κB; NF-κB exists as a heterodimer composed of p50 and p65 subunits bound to IκB in the unstimulated state. Upon activation, IκB is phosphorylated and degraded, causing the release of p50/p65 components of NF-κB.18 The active p50/p65 heterodimer translocates to the nucleus and initiates the transcription of genes involved in the regulation of leucocyte responses, such as pro-inflammatory cytokines and cell surface adhesion molecules.19 However, the expression of neutrophil NF-κB subunits in pre-eclampsia has not been previously investigated.
In the present study, we postulated that upregulation of TLRs and cryopyrin would relate to systemic inflammation in women with pre-eclampsia. A case–control study was designed to examine TLR2, TLR4 and cryopyrin expression, to assess NF-κB subunit changes, and to examine inflammatory cytokine profiles (IL-6, TNF-α, interferon-γ [IFN-γ], IL-2 and IL-10) among women with early- or late-onset pre-eclampsia, normotensive IUGR, women with matched normal pregnancy and non-pregnancy controls.
In this case–control study, we collected blood from women following informed consent. Samples were collected between September 2004 and March 2008. Ethics approval was granted by the University of British Columbia and the Children’s and Women’s Health Centre of British Columbia (C&W). Research laboratory staff were blinded to the clinical characteristics of the cases and controls.
Pre-eclampsia was defined as hypertension (blood pressure [BP] ≥140/90 mmHg, taken twice a day more than 4 hours apart after 20 weeks of gestation) and proteinuria (≥0.3 g/day, ≥2 + dipstick reading for proteinuria, or ≥30 mg protein/mmol creatinine).20 The blood pressure readings were taken in a semi-recumbent position, with a supported arm and appropriately-sized cuff, using a manual mercury sphygmomanometer, with Korotkoff V used to determine diastolic pressure.20
Participants comprised (1) 25 women with early-onset pre-eclampsia (clinical onset <34+0 weeks of gestation), (2) 25 women with late-onset pre-eclampsia (same criteria, at ≥34+0 weeks of gestation) and (3) 25 women with normotensive IUGR (identified by an antenatal ultrasound measurement of the fetal abdominal circumference below the fifth centile for gestational age and gender) and confirmed as either below the fifth centile at birth or below the tenth centile at birth (with either uterine artery Doppler notching at 22–24 weeks of gestation, absent/reversed umbilical artery end diastolic flow, or oligohydramnios [amniotic fluid index <50 mm] documented during the index pregnancy).21 None of these women had either hypertension or proteinuria before 20 weeks of gestation.
The controls were 75 women who ultimately delivered after normal pregnancies at term (>37 weeks without documented concerns about hypertension, proteinuria, gestational diabetes, or IUGR during their pregnancy) and 25 non-pregnant women. Normal pregnancy controls were matched for maternal age (±5 years), gestation age (±2 weeks) and parity (0, 1, ≥2). There were 25 matched women with normal pregnancies for women with each of (1) early-onset pre-eclampsia, (2) late-onset pre-eclampsia and (3) normotensive IUGR (i.e. 25 per group). Non-pregnant controls were women aged 20–40 years old, not using systemic hormonal contraception. Many of the observations in this present study had not been made previously in either normal or complicated pregnancies. The effects of pregnancy per se on these aspects of innate immune activation were considered important.22 All pregnancies were singleton gestations.
Venous blood was taken antenatally. Serum was prepared by centrifugation and specimens were frozen at −80°C until analysis. Case–control pairs were batch assayed to minimise any effect of inter-assay variability.
Detailed laboratory methods
These are described in the supplementary Text S1.
In brief, neutrophils were isolated from venous blood by 6% Dextran sedimentation, Ficoll-Paque differential centrifugation, and NHCl4 red cell lysis.
RNA preparation and synthesis of first-strand complementary DNA
Neutrophil RNA was extracted with an RNeasy minikit (Qiagen, Valencia, CA, USA) within 60 min of phlebotomy. Intact total RNA extracts (0.5–5 μg) were reverse transcribed into complementary DNA according to the manufacturer’s instructions (ThermoScript RT-PCR System, Invitrogen, Carlsbad, CA, USA).
Relative quantitative SYBR Green real-time polymerase chain reaction
Messenger RNA (mRNA) expressions of TLR2, TLR4, cryopyrin, IL-1β, NF-κB, and p65 were detected by quantitative SYBR Green real-time polymerase chain reaction (PCR) on a Sequence Detection System (ABI Prism 7300, Applied Biosystems). Primers were titrated to check for amplification efficiency (i.e. no primer dimerisation, low CT, and doubling of the amplicon each cycle). The relative level of the TLR2, TLR4, cryopyrin, NF-κB p50, NF-κB p65, IL-1β gene expression in clinical samples was compared with a calibrator and 18s ribosomal RNA. (for details [including primer sequences], see supplementary Text S1).
Neutrophil surface TLR2 and TLR4 protein expression was examined by flow cytometry (BD FACS Calibur System, Mississauga, ON, Canada). Phycoerythrin (PE)-labelled antibodies against TLR2 and TLR4 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) and isotype control (PE-labelled mouse immunoglobulin G2a, k, BD Biosciences Pharmingen, Mississauga, ON, Canada) were used for neutrophil staining.
Cytokine measurements using multiplex immunoassays
Circulating inflammatory (TNF-α, IL-2, IL-6 and IFN-γ), and anti-inflammatory IL-10 cytokine profiles were measured by a multiplexed fluorescent microsphere immunoassay using the Luminex 100 System (Luminex Corporation, Austin, TX, USA). Multiplex bead kit-based assays were performed in duplicate according to the manufacturer’s protocol (LINCO Research, Inc., St Charles, MO, USA). Thresholds for detection were 0.1 pg/ml for all assays.
Data were analysed using Prism 4.0 software (GraphPad, San Diego, CA, USA), comparing results across all groups, and then between individual groups. In most analyses, early-onset pre-eclampsia was set as the comparator group for presentation (see tables and figures). For continuous variables, Kruskal–Wallis analysis of variance, Dunn’s post-test, and Mann–Whitney U test were used, as appropriate. A P-value <0.05 was considered statistically significant.
The maternal clinical characteristics are shown in Table 1.
Table 1. Maternal clinical characteristics
EOPET (n = 25)
Ctrl (EOPET) (n = 25)
LOPET (n = 25)
Ctrl (LOPET) (n = 25)
nIUGR (n = 25)
Ctrl (nIUGR) (n = 25)
Nonpreg (n = 25)
P-value (χ2 or KW)
AST, aspartate transaminase; Ctrl, matched normal pregnancy control; EO, early onset (<34+0 weeks); GA, gestational age; KW, Kruskal–Wallis analysis of variance; LO, late onset (≥34+0 weeks); MAP, mean arterial pressure = diastolic blood pressure + (pulse pressure/3); nIUGR, normotensive intrauterine growth restriction; PET, pre-eclampsia; SGA, small for gestational age.
For SGA, the pregnancy was deemed to have achieved the outcome if any one fetus was born less than fifth percentile for gestational age and gender using multiethnic Canadian birthweight charts.21
Data expressed as median [range] or n (%).
Data were first compared across all columns to calculate the Kruskal–Wallis ANOVA or χ2P-value. Between-column comparisons were performed with early-onset pre-eclampsia, late-onset pre-eclampsia and nIUGR as the comparator groups (as outlined below). *P < 0.05; **P < 0.001 (Dunn’s test, versus late-onset pre-eclampsia); †P < 0.05; ††P < 0.01; †††P < 0.001 (Dunn’s test, versus early-onset pre-eclampsia); §P < 0.05; §§P < 0.01; §§§P < 0.001 (Dunn’s and Fisher’s exact tests, versus nIUGR).
Maternal age (years)
35 [19, 43]
33 [24, 41]
34 [23, 41]
33 [24, 39]
34 [25, 42]
33 [29, 43]
32 [21, 43]
GA at sampling (weeks)
32.1** [22.7, 33.8]
30.3 [21, 33.7]
37.1 [34.4, 40]
32.7** [21.4, 39]
32 [21, 38.3]
GA at delivery (weeks)
32.9 [26.3, 37.1]
38.7††† [36.4, 41]
37.6††† [34.4, 40.1]
39.6 [37.9, 40.9]
35.7†† [24.6, 39.1]
38.9 [37.3, 40.9]
127§§ [110, 139]
85†† [80, 92]
118§§ [84, 139]
86** [81, 103]
95 [78, 107]
101 [100, 116]
Uric acid (μm)
347§§ [249, 521]
384§§§ [298, 464]
271 [194, 343]
195 [88, 263]
150 [137, 170]
176 [77, 277]
210 [151, 265]
211 [125, 346]
32.5§§ [27, 133]
29§ [20, 237]
21 [18, 30]
1480 [395, 3385]
3318††† [2730, 4615]
2702† [1480, 4400]
3378 [2885, 4270]
1733 [360, 2645]
3460 [2750, 4420]
SGA (<5th centile)
Neutrophil TLR2, TLR4, cryopyrin, IL-1β, NF-κB p50 and NF-κB p65
Pre-eclampsia (as a single combined group of early- and late-onset) had increased TLR2 (median 0.669 [interquartile range (IQR) 0.390, 1.621] versus 0.348 [0.173, 0.625]; Mann–Whitney U (MWU) P < 0.0001) and TLR4 (0.921 [0.571, 1.572] versus 0.502 [0.245, 0.697]; MWU P < 0.0001) mRNA expressions, TLR2 (1633 [1063, 2950] versus 1205 [928, 1069]; MWU P = 0.0010) and TLR4 (0.921 [0.501, 1.572] versus 0.502 [0.245, 0.697]; MWU P = 0.0389) protein expressions as well as elevated cryopyrin mRNA (0.182 [0.083, 0.581] versus 0.040 [0.018, 0.115]; MWU P = 0.0003) expression compared with combined matched normal pregnancy controls, respectively. These differences were more marked in early-onset pre-eclampsia (Figures 1 and 2). We have also observed that mRNA expressions of cryopyrin and IL-1β varied among study groups and were elevated in early-onset pre-eclampsia (Figure 2). In addition, compared with normotensive IUGR, women with early-onset pre-eclampsia had upregulated mRNA levels of TLR2, TLR4, cryopyrin and NF-κB p65 (TLR2: MWU P < 0.05; TLR4: P < 0.001; cryopyrin: P < 0.01; NF-κB p65: P < 0.05). Furthermore, when combined into a single group, women with a normal pregnancy had higher mRNA expressions of cryopyrin (median [IQR] 0.048 [0.028, 0.154] versuss 0.032 [0.015, 0.050]; MWU P = 0.0118) and NF-κB p50 (7.08 [3.64, 11.32] versus 2.81 [2.10, 4.87]; MWU P = 0.0097) than non-pregnant controls.
Serum TNF-α:IL-10 and IL-6:IL-10
The ratios of serum TNF-α:IL-10 and IL-6:IL-10, which differed among groups, were increased in early-onset pre-eclampsia (Table S1 [online]). No significant difference was detected in ratios of serum IL-2:IL-10 and IFN-γ:IL-10 among study groups (Table S1). For individual cytokines, IL-6, IL-10 and TNF-α varied among groups (Kruskal–Wallis P = 0.0089, 0.0186, 0.0192, respectively), but not IL-2, and IFN-γ Kruskal–Wallis P = 0.6221, 0.3034, respectively) Pre-eclampsia (as a single combined group of early- and late-onset) had increased serum TNF-α (median [IQR] 4.23 pg/ml [3.86, 5.35] versus 3.11 pg/ml [2.35, 4.39; MWU P = 0.0080) and IL-6 levels (24.2 pg/ml [10.6, 37.9] versus 9.35 pg/ml [5.78, 15.580; MWU P = 0.0018) compared with combined normal pregnancy controls.
To our knowledge, this is the first study to report an association between early-onset pre-eclampsia and upregulated expressions of neutrophil TLR2, TLR4 and cryopyrin, increased mRNA expressions of NF-κB subunits p50 and p65, as well as imbalanced pro-inflammatory and anti-inflammatory cytokine expression, all in a single case–control series. These differences were not noted for women with late-onset pre-eclampsia and normotensive IUGR. New observations were made of neutrophil TLR2, TLR4 and cryopyrin upregulation in normal pregnancy.
Women with early- and late-onset pre-eclampsia and normotensive IUGR share the same intrauterine pathology, namely uteroplacental mismatch based on either incomplete placentation (early-onset pre-eclampsia and most cases of normotensive IUGR5,6) or exuberant fetoplacental growth (i.e. late-onset disease complicating macrosomia or multiple pregnancy6). Uteroplacental mismatch can evolve into either the systemic disorder of early- and late-onset pre-eclampsia (with or without IUGR) or the isolated fetal syndrome normotensive IUGR. These data reinforce the conjecture that both the origins (i.e amount of circulating placental debris) and maternal inflammatory response of early-onset pre-eclampsia may explain the differential maternal response and risks of early-onset versus both late-onset pre-eclampsia and/or normotensive IUGR.22–24
We chose to study neutrophils because they play a key role in innate immune defence25 and there is evidence of their role in pre-eclampsia.3,22,26,27 In addition, neutrophils express extracellular receptors, TLR2 and TLR4, and the intracellular cryopyrin inflammasome, which are crucial to inflammatory cytokine production.28–30
Toll-like receptors play a critical role in the induction of the immune response to invading pathogens. To date, 10 human TLRs have been identified and designated, TLR1 to TLR10.31 Although each extracellular TLR is distinct in its specificity, all receptors signal to a common intracellular pathway. Following ligand-binding, TLRs signal through the adapter molecule MyD88 to activate the NF-κB pathway; this results in an immune response characterised by the production of cytokines, antimicrobial products, and the regulation of co-stimulatory molecules.32 TLR2 identifies specific cell-wall components of Gram-positive and Gram-negative bacteria, fungi and viruses.31,32 Toll-like receptor 2 interactions with trophoblast at the fetal and maternal interface are responsible for infection-induced immune responses.33 TLR4 recognises specific components of Gram-negative bacterial LPS.31 Upregulated TLR4 protein expression was reported in interstitial trophoblasts in women with pre-eclampsia.34 In addition to the increased expression of neutrophil TLR2 and TLR4 we observed in this study, for women with early-onset pre-eclampsia we have recently determined that the presence of TLR2 (RR 2.57 [95% CI 1.31, 5.05]) and TLR4 (RR 2.06 [1.16, 3.67]) single nucleotide polymorphisms aggregated with early-onset (n = 42), but not late-onset (n = 52), pre-eclampsia compared with women with normal pregnancy outcomes (n = 176).35 Through synthesis of these and the other relevant published data,33,35–37 TLR2 and TLR4 polymorphisms appear to lower thresholds for early-onset and severe pre-eclampsia, but not late-onset or mild disease.35 Data from this cohort of women contributed to that case–control series. Therefore, the current body of evidence suggests that TLR2 and TLR4 may contribute to the neutrophil activation of pre-eclampsia.
Cryopyrin (NALP3), a member of the nucleotide binding and oligomerisation domain-like receptor family, serves as an activator in response to specific toxins, endogenous danger signals, or microbial pathogens.38 Independent of TLR signalling, cryopyrin has been shown to form the assembly of the inflammasome, a crucial part of the innate immune response and a cytosolic complex of proteins that activates caspase-1 to process the pro-inflammatory cytokine IL-1β, with NALP1 and NALP2.38 IL-1β activation may lead to delayed neutrophil spontaneous apoptosis,39 which has been reported in early-onset pre-eclampsia.22 Therefore, our study also suggests a role for cryopyrin in the neutrophil activation of pre-eclampsia.
Within leucocytes, nuclear translocation of NF-κB results in elevated synthesis of inflammatory cytokines, such as IL-6, and alterations in cell surface adhesion molecule expression.18 We have found that early-onset pre-eclampsia was associated with an increased mRNA expression of NF-κB p50 compared with non-pregnancy. Similarly, there was elevated NF-κB p65 as compared with matched normal pregnancy. Although we were unable to detect neutrophil NF-κB activation, pre-eclampsia may be characterised by NF-κB pathway activation in peripheral blood mononuclear cells, when compared with non-pregnancy, and with women with uncomplicated pregnancies at term.17
We examined serum cytokine profiles in women with established, but different onsets of pre-eclampsia, and observed an association between increased serum ratios of TNF-α:IL-10 and IL-6:IL-10 and early-onset pre-eclampsia. For individual cytokines, we observed increased TNF-α and IL-6 and reduced IL-10 in the combined pre-eclampsia cohort compared with matched normal pregnancy controls; these are consistent with other published reports.40–46 The imbalance between inflammatory and anti-inflammatory cytokines probably contributes to endothelial dysfunction,47 and is therefore linked with the pathogenesis of pre-eclampsia
In this study, we did not detect any difference in IL-2, or IFN-γ levels between pre-eclampsia-complicated and normal pregnancies, which is consistent with other published observations.41,43,48,49. As the present study was undertaken in women with established pre-eclampsia, it is not possible to determine whether the altered cytokine levels represent a cause or consequence of the disease. It would be interesting, therefore, to determine whether serum levels of cytokines are altered before the emergence of the maternal syndrome of pre-eclampsia. Further studies are required to determine whether measurement of serum cytokines in the first and second trimesters could predict pre-eclampsia.
Our data reinforce the view that early-onset pre-eclampsia differs more from normal pregnancies of similar gestation than does late-onset pre-eclampsia.24 Clinically, early-onset pre-eclampsia represents additional maternal risk, as maternal mortality is some 20-fold higher when pre-eclampsia develops at <32 weeks’ gestation than when pre-eclampsia develops near term.50 Our findings add further evidence that early-onset pre-eclampsia, in particular, is a form of systemic inflammation.3,4
Our study has several limitations. First, all cases were identified in a single tertiary referral centre, and our sample size is relatively small. Second, although we tested several sources of commercial cryopyrin antibody, cryopyrin protein expression could not be detected in study groups. To our knowledge, there are no reports of the use of cryopyrin protein antibodies in the studies of either human or animal neutrophils. Hence, neutrophil cryopyrin protein expression is under-explored. Also, we omitted to examine serum IL-1β concentrations to correlate with our mRNA observations. Similarly, we omitted to examine the mRNA concentrations of the cytokines examined in maternal serum. Third, in the present study, we did not estimate neutrophil NF-κB activation among study groups. Electrophoretic mobility shift assays are required to confirm whether the neutrophil NF-κB pathway is activated in women with pre-eclampsia. Fourth, the patient population served by our centre consists of three predominant ethnic groups, Caucasian, East Asian, and South Asian, and 40% of the adult population was born outside Canada. Also, approximately 20% of our maternity population is of mixed ethnicity, which confounds any interpretation of the role of ethnicity in our centre. For these reasons, we were not able to determine whether either ethnic background or country of birth might modify the association between TLRs and the cryopyrin inflammasome and the maternal syndrome of early-onset pre-eclampsia. Fifth, our normal pregnant population is not a random sample. This is an intrinsic problem with a case–control study design that could be addressed through a future prospective, population-based, nested case–control cohort. Therefore, our findings need to be confirmed in a study with a larger sample size and more separation of ethnic groups.
In summary, we report an association between early-onset pre-eclampsia and neutrophil TLR2, TLR4 and cryopyrin. In addition, we observed that early-onset pre-eclampsia is related to an imbalance between circulation inflammatory and anti-inflammatory patterns. These changes were not noted for women with either late-onset pre-eclampsia or normotensive IUGR. These findings provide new insights into possible roles for TLRs and cryopyrin in host innate immune defence mechanisms and the pathogenesis of systemic inflammation in both normal pregnancy and, in an exaggerated form, early-onset pre-eclampsia. Through further investigation of the immune system receptors and transcription pathways (Figure 3), it may be possible to identify potential targets for pre-eclampsia intervention.
Disclosure of interests
L.A.M. and P.vD. have received fees for opinions in legal cases related to pre-eclampsia. P.vD. is a principal investigator for an investigator-initiated safety and efficacy trial of a possible disease-modifying therapy for pre-eclampsia sponsored by Eli Lilly, Canada, and is a site investigator for a pre-eclampsia prediction study sponsored by Ortho-Clinical Diagnostics.
Contribution to authorship
F.X. performed the experiments as part of her PhD studies, under the direct supervision of Y.H., L.A.M., E.T., S.E.T. and P.vD. All authors, other than F.X., Y.H. and S.E.T., were co-applicants for project funding and contributed to the study design. P.vD. was primarily responsible for data analyses. D.M.P. provided statistical input. All authors contributed to the manuscript.
Details of ethics approval
This study was approved by the University of British Columbia Clinical Research Ethics Board and the Children’s and Women’s Health Centre of British Columbia Research Review Board.
This study was funded by a New Investigator Pilot Project Grant from the Canadian Institutes of Health Research (CIHR) Institute of Infection and Immunity (to P.vD.); F.X. is the recipient of a studentship from the CIHR-funded Interdisciplinary Training Program in Women’s Reproductive Health, Child and Family Research Institute. L.A.M., S.E.T. and P.vD. are the recipients of salary awards from the Michael Smith Foundation for Health Research (L.A.M., S.E.T. and P.vD.), CIHR (P.vD.), and the Child and Family Research Institute (P.vD.).
Direct research assistance was received from Pamela Lutley, Tara Morris, Céline Basque and Monica Pearson. We gratefully acknowledge the support of Dr David P. Speert with the original funding application and study design, and Dr Keith R. Walley for access to the Luminex technology.