SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In the nonpregnant population, there is extensive evidence of a systemic low-grade inflammatory status in relation to excess adipose tissue. Less is known about the relation during pregnancy.

Objective:

Our main objective was therefore to explore the effect of pregnancy on adiposity-related systemic inflammation.

Design and Methods:

This study is a longitudinal cohort study of 240 pregnant women of Scandinavian heritage at Oslo University hospital—Rikshospitalet, Norway from 2002 to 2005. The inflammatory markers (C-reactive protein [CRP], Interleukin-6 [IL-6], monocyte chemoattractant protein 1 [MCP-1], IL1-Ra, tumor necrosis factor receptor II, and IL-10) were measured at four timepoints during pregnancy and analyzed by enzyme immuno-assay. The women were categorized based on BMI at inclusion (BMI <25, 25–30, and >30 kg/m2). Data were analyzed by Friedman-test, Wilcoxon signed rank test, or Kruskal–Wallis test as appropriate.

Results:

Maternal adiposity was associated with significantly higher circulatory levels of several inflammatory markers (CRP, MCP-1, IL-6, and IL-1Ra). However, this proinflammatory upregulation was not evident toward the end of pregnancy, as levels of CRP, MCP-1, and IL-6 were not any longer significantly different between the BMI categories.

Conclusions:

Although normal pregnancy exhibits proinflammatory features, this does not seem to have additive or synergistic effects on the inflammation associated with adiposity. On the contrary, we found that the BMI-dependent increase in proinflammatory markers was not evident at the end of pregnancy.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

There is evidence that pregnancy involves a state of enhanced inflammation (1, 2). The concept is based on studies evaluating changes in inflammatory surface markers and gene expression of inflammatory cells as well as circulating levels of soluble inflammatory markers like acute phase reactants and cytokines (3, 4). The pregnancy-specific activation of inflammatory mechanisms is reasonably attributed to the presence of trophoblastic tissue (2).

Increased amount of body adipose tissue is also generally associated with activation of inflammatory pathways and adiposity is accordingly considered as a state of low-grade inflammation (5, 6). Thus, in the natural course of pregnancy, both the placenta and adipose tissue may contribute to systemic inflammation reflected in increased circulating levels of inflammatory markers.

Maternal overweight and obesity are well-known risk factors of pregnancy complications like pre-eclampsia and gestational diabetes (GDM) (7). In the pathogenesis of both these complications, inflammatory mechanisms are heavily involved (8–10). An intuitive reasoning is therefore that adiposity is a risk factor for pregnancy complications because the gestational and adipose tissue-mediated inflammatory responses potentiate each other. The longitudinal course of proinflammatory and anti-inflammatory factors in normal pregnancy is, however, not well characterized, let alone the role of maternal body composition on the circulating levels of these factors (9, 11–13).

The purpose of this study was therefore to explore how adiposity-related systemic low-grade inflammation is modified by pregnancy in a longitudinal design.

We selected three markers with predominantly proinflammatory action; Interleukin-6 (IL-6), monocyte chemoattractant protein 1 (MCP-1), and the acute phase reactant C-reactive protein (CRP). We also analyzed the soluble receptor tumor necrosis factor receptor II (sTNF-RII) and IL-1 receptor antagonist (IL1-Ra), both with counter-regulatory effects. Interleukin-10 (IL-10), a known anti-inflammatory cytokine, was also determined. These are all upstream markers shown to be produced both by the placenta and by adipose tissue and to reflect both proinflammatory and anti-inflammatory action (6).

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Because of the biological traits of cytokines, short half-time, and considerable variation in circulating levels, the assessment of normal physiological changes requires a strict design. Our study therefore included a high number of participants and followed a longitudinal design with standardized procedures.

Study design and subjects

This work was performed in a subcohort (n = 240) of the STORK study (details have been published previously) (14, 15). The STORK-cohort consists of healthy Caucasian women (n = 1,031) who registered for obstetric care at Oslo University hospital—Rikshospitalet from 2001 to 2008. Exclusion criteria were multiple pregnancies, known pregestational diabetes, and severe chronic diseases (lung, cardiac, gastrointestinal, or renal).

The women were scheduled for four examinations (visits 1–4) at gestational weeks 14–16, 22–24, 30–32, and 36–38. Gestational age was based on ultrasound measures made at weeks 17–19. Maternal height was obtained at the first visit and weight at each visit. BMI (kg/m2) was calculated based on first visit measurements. Subcutaneous fat at the triceps, subscapular, and iliac sites were estimated using a Holtain caliper (Holtain, Crymych, UK). Fasting glucose was measured at weeks 14–16 and 30–32. Data on age, parity, educational level, smoking status, and pregestational BMI were registered. Data on pre-eclampsia and hypertension were obtained from hospital records. Written informed consent was obtained from the participants. The study was approved by the Regional Ethics Committee, South-Norway.

Laboratory methods

The blood samples were drawn in the morning, between 07:30 and 08:30 after an overnight fast at all four visits. The samples used for analysis were obtained from venipuncture into vacutainertubes. Vacutainer® tubes on ice containing EDTA were centrifuged at 3,000g for 25 min at 4°C, and plasma was frozen within 1 h of collection and stored in aliquots at −80°C until analyzed. Before analysis, the aliquots were thawed and mixed well. IL-6 (UltraSensitive), IL-10 (UltraSensitive), and IL-1Ra were measured by enzyme immuno-assay (EIA) using commercially available kits (BioSource International Inc., CA). The detection limit was 0.1, 0.2, and 4 pg/ml, respectively. Plasma levels of TNF-RII and MCP-1 were measured by EIA using reagents from R&D Systems (Minneapolis, MN). Plasma CRP levels were quantified by EIA with antibodies from DakoCytomation, Carpinteria, CA as described by others (16). The lowest detectable concentrations were 8 pg/ml for TNF-RII and 8 pg/ml for MCP. The minimum detectable concentration for CRP was 0.001 mg/l.

All samples were analyzed in duplicate from a given participant in the same microtiter plate to minimize run-to-run variability. Intra-assay and interassay coefficients of variation were < 10% for all assays. All assays were performed according to the manufacturer's instructions.

Plasma glucose was measured immediately in EDTA blood, by Accu Chek Glucose Test strips (Roche Diagnostics, Basel, Switzerland) (17).

Statistical analysis

Descriptive statistics are presented as mean and standard deviation, frequency, and percentage (%), or median and interquartile ranges (Q1 and Q3). To exclude possible cases of unrecognized infection or other specific inflammatory disease known to affect cytokine levels, we excluded all cases with CRP values above 10 mg/l (18). Fifty-two cases were excluded due to elevated or missing values of CRP. In total, 188 women formed the basis for further analysis.

The women were categorized based on BMI measured at 14–16 weeks of gestation according to WHO classification, BMI < 25, 25–30, and >30 kg/m2. Group differences, including baseline characteristics, were analyzed by one-way ANOVA or Kruskal–Wallis test as appropriate. Caliper measurements were categorized into tertiles (data not shown).

The longitudinal changes in markers of inflammation were evaluated by the nonparametric Friedman-test and pair-wise comparisons from visit to visit (v1–2, v2–3, and v3–4) were performed using Wilcoxon signed rank test (with Bonferroni correction).

Pearson's or Spearman's correlation coefficients were used to assess associations between continuous variables.

Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS, Version 18.0) for Windows (SPSS Inc., Chicago, IL). A P-value < 0.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The demographic and clinical details of the cohort and the BMI categories are presented in Table 1. The women included in this study had a mean BMI of 25 kg/m2 and 47.5% were primipara. Maternal body weight, caliper measurements, birth weight, and placental weight were significantly different between the BMI groups (P < 0.001). The highest BMI category also had significantly higher fasting glucose values and higher systolic and diastolic blood pressure, compared to lower categories (P < 0.001). No significant differences between the groups were noted regarding height, age, and weight gain (0.09 ≤ P ≥ 0.26).

Table 1. Demographic, anthropometric, and clinical characteristics of study subjects and according to BMI category
Cohort characteristics at inclusionWhole cohort (n = 188)BMI <25 (n = 109)BMI 25–30 (n = 65)BMI > 30 (n = 14)
  • a

    Based on WHO criteria (75 g oral glucose tolerance test and corresponding 2 h-glucose value ≥ 7.8 mmol/l).

  • b

    Blood pressure ≥ 140/90 with or without proteinuria (≥1 + on dipstick).

Age (years)31 (4)31 (4)31 (4)32 (3)
Education ≥ 12 yr84%88%81%64%
Primigravidae50%51%53%29%
Smokers2.1%0.9%4.8%0%
Height (cm)168 (5)168 (5)168(5)169 (3)
Weight (kg)70 (11)63 (6)75 (5)96 (10)
Body mass index (kg/m2)25 (4)22 (2)27 (1)36 (6)
Subcutaneous fat triceps (mm)21 (7)17 (5)24 (6)32 (7)
Subcutaneous fat subscapular (mm)20 (8)16 (5)24 (7)35 (7)
Subcutaneous fat iliaca (mm)23 (9)20 (8)27 (8)34 (5)
Gestational weight gain (kg) (visit 1–visit 4)10.8 (3.7)10.9 (3.5)10.6 (4.0)10.0 (4.5)
Systolic blood pressure (mm Hg)110 (11)107 (11)111 (8)120 (12)
Diastolic blood pressure (mm Hg)65 (8)63 (7)67 (7)70 (5)
Fasting glucose visit 1 (mmol/l)4.2 (0.5)4.1 (0.4)4.3 (0.5)4.4 (0.3)
Fasting glucose visit 3 (mmol/l)4.4 (0.5)4.3 (0.4)4.5 (0.5)4.6 (0.5)
Gestation at delivery40 (1)40 (1)40 (1)40 (1)
Birth weight (g)3,749 (448)3,643 (383)3,871 (468)4,040 (612)
Placental weight (g)752 (151)729 (142)760 (144)853 (186)
Pregnancy complications (n)
 GDMa7.7% (14)473
 Hypertensive complicationsb2.1% (7)250

The gestational changes in inflammatory markers are shown in Table 2. MCP-1, TNF-RII, IL-6, and IL1-Ra all showed significant overall increase throughout pregnancy, however not between all visits (Table 2). CRP was the only marker that displayed a fall during pregnancy. The anti-inflammatory IL-10 did not change during pregnancy.

Table 2. Longitudinal changes in inflammatory markers median (Q1 and Q3) according to gestational age and BMI category
Inflammatory markerGestational weeks 14–16 visit 1Gestational weeks 22–24 visit 2Gestational weeks 30–32 visit 3Gestational weeks 36–38 visit 4Significant changes overalla and from visit to visit (v)Significant differences between BMI categories at each visitb
  • a

    Friedman and post hoc test Wilcoxon signed rank test: v1–2 represent a significant change from visit 1 to visit 2, v2–3 represent a significant change from visit 2 to visit 3, and v3–4 represent a significant change from visit 3 to visit 4.

  • b

    Kruskal–Wallis test.

  • BMI < 25; n = 109.

  • BMI 25–30; n = 61.

  • BMI > 30; n = 14.

CRP0.98 (0.5, 2.5)1.11 (0.4, 2.0)0.79 (0.3, 1.6)0.81 (0.3, 1.6)a0.000 v2–3 
 BMI < 250.82 (0.4, 2.0)0.87 (0.3, 1.7)0.72 (0.3, 1.4)0.78 (0.3, 1.4) b0.001 (v1–v3)
 BMI 25–301.04 (0.5, 2.8)1.15 (0.5, 2.0)0.69 (0.4, 1.6)0.78 (0.3, 1.6)  
 BMI > 302.72 (1.9, 6.2)2.66 (1.1, 3.8)1.99 (1.3, 6.1)1.69 (0.5, 4.2)  
MCP-167 (44, 100)61 (37, 87)70 (50, 95)79 (58, 113)a0.000 v1–2, v2–3, v3–4 
 BMI < 2556 (40, 91)50 (36, 79)62 (48, 88)77 (57, 108)b0.000–0.01 (v1–v3)
 BMI 25–3078 (49, 109)72 (45, 94)85 (60, 109)82 (63, 119) 
 BMI > 30118 (83, 175)76 (40, 116)86 (74, 131)88 (54, 149)  
IL-60.11 (0.1, 0.2)0.12 (0.1, 0.2)0.18 (0.1, 0.3)0.28 (0.2, 0.5)a0.000 v2–3, v3–4 
 BMI < 250.10 (0.1, 0.2)0.10 (0.1, 0.2)0.16 (0.1, 0.3)0.28 (0.2, 0.5)b0.001 (v 1)
 BMI 25–300.13 (0.1, 0.2)0.17 (0.1, 0.3)0.23 (0.1, 0.4)0.28 (0.2, 0.5)  
 BMI > 300.26 (0.2, 0.3)0.20 (0.1, 0.3)0.24 (0.2, 0.4)0.45 (0.2, 06)  
TNF RII4,520 (3,885, 5,254)5,,104 (4,417, 5,920)5,332 (4,610, 6,372)5,293 (4,443, 6,466)a0.000 v1–2, v2–3 
 BMI < 254,502 (3,844, 5,156)5,069 (4,339, 5,618)5,282 (4,321, 6,258)5,162 (4,269, 6,332)bns
 BMI 25–304,637 (3,871, 5,513)5,178 (4,456, 6,049)5,440 (4,671, 6,352)5,433 (4,575, 6,483)  
 BMI > 304,394 (3,959, 6,450)5,581 (4,231, 6,269)6,103 (5,284, 6,622)5,614 (4,393, 7,694)  
IL-1Ra140 (109, 186)142 (115, 188)161 (132, 209)1,867 (151, 252)a0.000 v2–3, v3–4 
 BMI < 25127 (100, 166)126 (107, 160)149 (120, 182)168 (140, 211)b0.000 (v1–v4)
 BMI 25–30163 (120, 227)154 (130, 206)198 (152, 238)232 (173, 298)  
 BMI > 30229 (147, 322)209 (164, 299)275 (193, 398)297 (236, 478)  
IL-100.26 (0.2, 0.4)0.24 (0.2, 0.4)0.24 (0.2, 0.4)0.23 (0.2, 0.4)0.103a 
 BMI < 250.27 (0.2, 0.5)0.23 (0.2, 0.5)0.25 (0.2, 0.4)0.23 (0.2, 0.4) bns
 BMI 25–300.27 (0.2, 0.4)0.25 (0.2, 0.3)0.20 (0.2, 0.3)0.26 (0.2, 0.5)  
 BMI > 300.21 (0.2, 0.3)0.22 (0.2, 0.3)0.29 (0.2, 0.3)0.20 (0.2, 0.4)  

The levels of MCP-1, IL-1Ra, and CRP were significantly different between BMI categories, the differences in IL-6 levels were only significant at visit 1, while TNF-RII and IL-10 did not vary across BMI categories (Table 2). The differences in CRP and MCP-1 between BMI categories were no longer significant at visit 4. The box plots show the concentrations of the biomarkers in BMI categories at different gestational ages during the study period (Figure 1).

thumbnail image

Figure 1. Concentration of inflammatory markers according to gestational age and BMI category. The horizontal lines represent the median, the top and bottom of boxes represent quartiles, and the t-bars denote 10th and 90th percentiles. Open circles represent single values.

Download figure to PowerPoint

We also repeated the analysis using BMI quartiles, tertiles, and two categories as well as BMI categories based on reported prepregnancy BMI and caliper measurements. We then repeated the analysis excluding cases of hypertensive pregnancy complications and gestational diabetes. Similar results were obtained in all the analyses, indicating robustness in our data (data not shown).

Furthermore, we examined the correlations between weight gain and changes in circulatory levels of inflammatory markers. We found no significant correlation between maternal weight gain and changes in inflammatory parameters (neither from visit to visit or visits 1–4). Finally, there were no significant correlations between parity and inflammation.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In this study, we found a proinflammatory effect of increasing maternal adipose tissue (BMI) in pregnancy as previously reported in nonpregnant individuals. However, importantly, this effect was not evident toward the end of pregnancy, based on the finding that the levels of CRP, MCP-1, and IL-6 were no longer significantly different between BMI categories at 36–38 weeks of gestation in contrast to earlier stages of pregnancy. The counter-regulatory marker IL-1Ra, however, remained significantly different between BMI categories throughout pregnancy.

In the study population as a whole, maternal CRP levels decreased with increasing gestational age while the two other proinflammatory markers, IL-6 and MCP-1 showed a moderate increase. In parallel, the counter-regulatory markers of inflammation IL-1Ra and TNF-RII exhibited significant increases throughout pregnancy.

In early pregnancy, local inflammation and inflammatory cytokines play a pivotal role for successful placentation and pregnancy outcome and the placenta has been shown to be able to produce virtually all known cytokines (19). To what extent these mediators of inflammation act merely locally or induce a more systemic response by being released into circulation remain more elusive and there is a lack of consensus regarding the normal physiological changes in inflammatory markers during pregnancy (12, 20).

We therefore wanted first to explore longitudinally the normal physiological changes in six proinflammatory and anti-inflammatory markers during pregnancy, including soluble receptors and receptor antagonists playing a role in modulation of inflammatory response. An expected feature of all the cytokines was a marked variation, with extreme values and a skewed distribution despite a homogenous study population and strict procedures for collecting and assaying the samples. Still, the overall analyses showed significant changes (P < 0.001) over the course of gestation for all markers except IL-10.

The plasma concentration of inflammatory markers tended to increase with advancing gestation while CRP showed a significant decrease from second to third trimester. This latter finding is consistent with some, although not all previous reports (12, 21–24). So in contrast to the notion that pregnancy is a progressive low-grade inflammatory state, we find evidence of the opposite when considering systemic levels of CRP, a marker considered reliable of systemic inflammation (18). However, IL-6, another proinflammatory marker and a known stimulator for hepatic CRP production showed a gradual increase over gestation, a finding in accordance with some studies, but in contrast to others (11, 12, 25). Our data thus may indicate a decreased response of CRP to IL-6 as pregnancy progresses. Still, these two markers were significantly correlated throughout pregnancy (Spearman r2 0.16–0.35).

The proinflammatory MCP-1 is a member of the chemoattractant cytokine group and in this study displayed a j-shaped curve through gestation. Our finding is in contrast to a previous study suggesting a fall over gestation (4). The different findings may be due to the cross-sectional design of the latter study. The two measured soluble receptors also showed significant increases with advancing gestation. IL1-Ra is considered an anti-inflammatory cytokine, but also reflects proinflammatory action and activation of the IL-1 system. To our knowledge, there are few reports on this cytokine during pregnancy, one reporting no significant gestational change (26) and another reporting an increase (27). Upregulation of soluble cytokine receptors like IL-1Ra might reflect a counter-regulatory response to systemic inflammatory activation. Treatment of rheumatic diseases, like rheumatoid arthritis, with IL-1 receptor antagonists (Kineret®) is now well established and it has been speculated in whether the gestational increase in this cytokine might offer part of a molecular basis for the observed spontaneous remission during pregnancy (26). Treatment with Kineret® has also been shown to reduce CRP levels and the physiological increase in IL-1Ra might thus be responsible for the observed fall in CRP levels (28). Likewise, TNF-RII is believed to reflect the activation of the tumor necrosis factor system, but at the same time has anti-inflammatory properties (29). Data concerning the temporal pattern of TNF-system during pregnancy are partly conflicting (11, 26, 30, 31), but our data indicate a moderate but significant upregulation during pregnancy. In contrast, IL-10, an anti-inflammatory cytokine, did not change significantly over the course of gestation, neither overall nor from visit to visit. Again there is conflicting data on IL-10 during pregnancy (12, 25, 32).

The sources of the gestational-specific changes of inflammatory markers are elusive, but the fall in cytokine levels postpartum is indicative of a placental contribution (26). However, one study found no significant differences when comparing TNFα in peripheral blood and uterine vein samples indicating that other sources rather than the placenta contribute to systemic levels (33). In support of a placental contribution, we found overall no significant correlation between maternal weight gain and changes in inflammatory parameters (data not shown). This finding is in line with another study reporting no association between CRP and gestational weight gain (34).

The main objective of this study was to explore the combined effect of maternal adiposity and gestational age on circulatory levels of inflammatory markers. In the nonpregnant population, the link between obesity and inflammation is well established. The six chosen markers have all been shown to be central, systemic biomarkers of inflammation and upregulated or downregulated in obesity in the nonpregnant population (5, 6, 35–37). Fewer studies have addressed the association between obesity and inflammation during pregnancy, and the results of these studies indicate, although not unequivocally, that the same association is present in pregnancy as well (11–13, 34, 38–40). Based on these previous reports, it is conceivable that a chronic inflammatory process in excess adipose tissue could exacerbate a physiological low-grade inflammation in pregnancy. However, our data do not support this notion. We did find that median plasma concentration of proinflammatory markers was significantly higher in the higher BMI classes and displayed a dose-dependent pattern. However, while CRP, MCP-1, and IL-6 tended to be higher in the higher BMI categories at beginning of pregnancy, the differences were not statistically significant toward the end of gestation (visit 4). This finding may indicate that adiposity-induced inflammation as reflected in increased CRP, MCP-1, and IL-6 is restrained late in gestation. Accordingly, our findings do not indicate a synergistic or additive effect of gestation and adiposity.

We found no significant differences in weight gain between the BMI groups, so differences in amount of fat mass accretion can therefore hardly explain this observation.

The physiological grounds for the marked fall in CRP in the higher BMI group remain elusive. It may be that physiological anti-inflammatory factors operating in pregnancy in general counteract some of the proinflammatory effects of obesity. IL-1Ra may represent such a factor as we found it to increase as pregnancy progressed. Furthermore, we found IL-1Ra markedly upregulated with increasing maternal BMI, a finding consistent with studies in the nonpregnant population (36). Interestingly, the BMI-related increase in IL1-Ra did not disappear late in pregnancy, in contrast to the proinflammatory markers. For TNF-RII and the anti-inflammatory IL-10, no differences between the BMI categories were found at any time point. Our findings are overall in line with another longitudinal study of circulating levels of TNFα, IL-10, IL-6, and CRP (12).

Our study has some limitations. One is the lack of prepregnancy and postpregnancy samples. However, the correlation coefficient between reported pregravid BMI and measured BMI at weeks 14–16 was 0.9 and thus this is not likely to have affected our results. In addition, maternal BMI is a surrogate for maternal fat mass. However, when grouping maternal adiposity by calliper measurements, the associations with the inflammatory markers remained similar (data not shown). The selection and number of inflammatory markers is arguable. Besides financial restrictions, the volumes of plasma samples available were limited. We acknowledge that if other or more inflammatory markers had been investigated, our findings may have been modified. Our findings do not necessarily imply a general suppression of any inflammatory response. It is of interest to note however that BMI-associated inflammatory markers of different cellular origin (CRP from hepatocytes and IL-6 from macrophages/T-cells) responded to pregnancy similarly as the BMI-mediated enhancement of both these markers was not evident with progression of pregnancy. Our study cohort had a relatively modest degree of obesity. We acknowledge that a study population with more pronounced obesity may have given different results.

We also excluded 42 women with CRP > 10 mg/l. The average BMI of this group was 27 kg/m2 versus 25 kg/m2 of the study group. This exclusion was done a priori. The rational was to exclude individuals with increased CRP due to pathophysiological states (infections and autoimmune diseases) irrelevant to our hypothesis. Despite these limitations, this large longitudinal study with standardized sampling procedures and fasting samples is to our knowledge the first to explore more comprehensively the impact of gestational age and maternal anthropometry on levels of several inflammatory markers.

In conclusion, we did not find an additive or synergistic effect between adiposity and pregnancy in terms of maternal levels of proinflammatory markers. In fact, we found that the enhancement of inflammatory markers associated with increasing BMI was not evident as pregnancy progressed toward term. Further studies are needed to explore the mechanisms by which pregnancy exerts this effect.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank N Voldner and K Frey Frøslie for their valuable contributions. We also thank all sources of financial support: South-Eastern Norway Regional Health Authority, the National Resource Centre for Women's Health, Division of Obstetrics and Gynaecology, Oslo University Hospital, Rikshospitalet, the Department of Obstetrics, Women and Children's Division, Oslo University Hospital, Rikshospitalet, Oslo, Norway, The Faculty of Medicine, Thematic Research Area, University of Oslo, and Aktie Selskanet Freia Chocolade Fabriks Medisinske Fond for their contribution.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References