DOI: 10.1111/1471-0528.12375 www.bjog.

org

Systematic review

Pre-eclampsia is associated with, and preceded by, hypertriglyceridaemia: a meta-analysis

ID Gallos,a K Sivakumar,b MD Kilby,c A Coomarasamy,c S Thangaratinam,d M Vatisha,b
a Nufeld Department of Obstetrics and Gynaecology, University of Oxford, Oxford Radcliffe Hospitals NHS Trust, Oxford, UK b Clinical Sciences Research Institute, Warwick Medical School, Coventry, UK c School of Clinical and Experimental Medicine (Reproduction, Genes and Development), College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK d Womens Health Research Unit, Centre for Primary Care and Public Health, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, UK Correspondence: M Vatish, Nufeld Department of Obstetrics and Gynaecology, University of Oxford, Oxford Radcliffe Hospitals NHS Trust, Oxford, OX3 9DU, UK. Email manu.vatish@obs-gyn.ox.ac.uk

Accepted 22 May 2013. Published Online 17 July 2013.

Background Elevated triglycerides are a feature of the metabolic

Data collection and analysis We collected and meta-analysed the

syndrome, maternal obesity, maternal vasculitis (i.e. systemic lupus erythematosus) and diabetes mellitus. These conditions are all known risk factors for pre-eclampsia. Hypertriglyceridaemia therefore may be associated with pre-eclampsia and indeed this may precede the presence of overt disease.
Objective In this study we determine the association between

weighted mean differences (WMDs) of triglyceride levels from individual studies using a random effects model.
Main results We found strong evidence from meta-analysis of 24 casecontrol studies (2720 women) that pre-eclampsia is associated with higher levels of serum triglycerides (WMD 0.78 mmol/l, 95% condence interval 0.60.96, P < 0.00001). This nding is also conrmed in ve cohort studies, that recruited 3147 women in the second trimester before the onset of pre-eclampsia, which proves that hypertriglyceridaemia precedes the onset of pre-eclampsia (WMD 0.24 mmol/l, 95% condence interval 0.130.34, P < 0.0001). Authors conclusions Hypertriglyceridaemia is associated with and

hypertriglyceridaemia and pre-eclampsia in pregnant women.

Search strategy We searched MEDLINE, EMBASE, Web

of Science, Excerpta Medica Database, ISI Web of Knowledge, Cumulative Index to Nursing and Allied Health Literature, Cochrane Library from inception until June 2012 and reference lists of relevant studies.
Selection criteria Two reviewers independently selected studies on

pregnant women where triglycerides were measured and women were followed up until the development of pre-eclampsia or selected on the basis of presence of pre-eclampsia and compared with controls.

precedes the onset of pre-eclampsia. Further research should focus on dening the prognostic accuracy of this test to identify women at risk and the benecial effect of triglyceride-lowering therapies in pregnancy.
Keywords Meta-analysis, predictive marker, pre-eclampsia, systematic review, triglycerides.

Please cite this paper as: Gallos I, Sivakumar K, Kilby M, Coomarasamy A, Thangaratinam S, Vatish M. Pre-eclampsia is associated with, and preceded by, hypertriglyceridaemia: a meta-analysis. BJOG 2013;120:13211332.

Introduction
Pre-eclampsia is a multi-organ disorder of pregnancy that manifests after 20 weeks of gestation with new onset hypertension and proteinuria. Pre-eclampsia is dened as blood pressure 140 mmHg systolic and 90 mmHg diastolic diagnosed for the rst time after 20 weeks of gestation together with >300 mg proteinuria/24 hours as dened in the proceedings of the 16th Scientic Study Group of the Royal College of Obstetricians and Gynaecologists.1 This disease can progress to cause maternal liver dysfunction,1 renal impairment2 and ultimately seizures and death.3 The

foetus may suffer intrauterine growth restriction and, when born preterm, is more likely to struggle with the consequences of premature delivery.4 Women with pre-eclampsia are also more likely to suffer stillbirth or neonatal death.5 Of equal importance is the consistent nding that these women have an increased lifetime risk of cardiovascular disease compared with the rest of the population.6,7 Dyslipidaemia (especially hypertriglyceridaemia) has been reported as being part of the pre-eclampsia disease process.8 Hypertriglyceridaemia is well documented as an endothelial disruptor in atherosclerosis and is a potential candidate for the endothelial dysfunction seen in this

2013 RCOG

1321

Gallos et al.

disease. The aim of this study was therefore to perform a systematic review of the literature and meta-analysis to test the hypothesis that elevated triglycerides correlate with increased likelihood of pre-eclampsia. One of the leading hypotheses in the aetiology of pre-eclampsia is that circulating factors, released from the placenta, alter endothelial function in the maternal circulation.9 These factors may subsequently alter vasomotor function,10 angiogenesis,11 endothelial permeability and downstream activation of other cascades such as thrombosis.12 A key variable that may be equally important in the pathogenesis of the disease is the overall sensitivity of the maternal endothelium to these circulating factors. This sensitivity may be modulated by maternal disease, including diabetes,13 chronic hypertension,14 obesity and, importantly, altered lipid prole.15 In pregnancy, as a result of both insulin resistance and increased oestrogen, metabolic changes in both the liver and adipose tissue alter circulating triglycerides, fatty acid, cholesterol and phospholipids.16 As pregnancy continues, this causes hyperlipidaemia consisting principally of increased triglycerides. The mother and foetus can subsequently use these, and so the increase in triglyceride concentrations represents an accessible energy reservoir. Several reports have suggested that women with pre-eclampsia display further changes in lipid metabolism with increases in circulating levels of triglycerides and non-esteried fatty acids.17 These changes have been reported to be present at early gestation in women who subsequently develop pre-eclampsia, with the dyslipidaemia notably preceding clinical diagnosis far earlier than the presence of known circulating factors associated with pre-eclampsia such as soluble Flt-1 (sFlt-1) or soluble endoglin (sEng).18,19 Moreover, serum from women with pre-eclampsia induces greater lipid accumulation in endothelial cells than serum from normal women.20

along with their associated Medical Subject and Emtree Headings were used to search MEDLINE, EMBASE and CINAHL. These two populations of keywords were combined using the AND function of the database. The Web of Science and The Cochrane Library were searched using the keywords pre-eclampsia and triglycerides. There were no limits or philtres placed on the searches, to ensure maximal sensitivity and no language restrictions were applied. All the reference sections of all articles were reviewed to also identify relevant studies.

Selection of articles
Articles were selected if they included a population of pregnant women, tested for triglyceride levels and followed up until the diagnosis of pre-eclampsia. These studies were expected to be of cohort design. We also selected studies that they measured the triglyceride levels on women with known pre-eclampsia and compared those with controls. Of the 1017 identied articles, 965 did not match our selection criteria based on review of their titles and abstracts conducted by two authors (MV and IDG). These two authors then independently reviewed the full text of the remaining 52 articles to determine inclusion or exclusion (Figure 1). We excluded 23 studies after evaluation of the full manuscripts. The most common reason for exclusion was our inability to extract raw data from the published reports (18 studies). Finally, 29 studies were deemed eligible for inclusion of which, 24 casecontrol studies2144 and ve comparative cohort studies.4549 When duplicate data were published, only the most up-to-date, larger series was included. Any disagreements about study eligibility were resolved by consensus, with arbitration by a third reviewer (AC) if necessary.

Data extraction
Data were extracted from the eligible studies by two authors (MV and IDG) using a piloted data extraction form. We collected information on denition and diagnosis of pre-eclampsia, gestational age at testing and diagnosis, timing and method of triglyceride measurements and fasting or non-fasting status of the participants. The majority of the papers reported triglyceride measurements in millimolar and for few papers that reported data in milligram per decilitre measurements were converted to millimolar. From eligible studies we extracted mean and standard deviations (SDs) of triglyceride measurements from women with pre-eclampsia compared with controls. When medians and 95% condence intervals (95% CI) were reported instead we assessed the skewness and if acceptable we presumed a normal distribution of the triglyceride levels across women included in the study and we computed the means and SDs. Two reviewers (MV and IDG) completed the quality assessment using the NewcastleOttawa Quality

Methods
Data sources and search strategy
We conducted a thorough search to identify eligible studies that measured and reported the triglyceride levels in pregnant women and women were followed up until the development of pre-eclampsia or selected on the basis of presence of pre-eclampsia and compared with controls. The hypothesis is to explore the association of hypertriglyceridaemia with pre-eclampsia. The databases searched included MEDLINE, EMBASE, Excerpta Medica Database, ISI Web of Knowledge, Cumulative Index to Nursing and Allied Health Literature (CINAHL) and The Cochrane Library from inception until June 2012. A combination of keywords for pre-eclampsia (Pre-eclampsia, pregnancy-induced hypertension, eclampsia, pregnancy and hypertension), for triglycerides (triglycerides, lipids, hyperlipid*, dyslipidemia, cholesterol)

1322

2013 RCOG

Triglycerides and pre-eclampsia

Figure 1. Flowchart of the study selection process.

Assessment Scales for observational studies. We awarded studies a maximum of nine stars for casecontrol studies and eight for cohort studies (Figure 2). Any differences were resolved by consensus.

Data synthesis
Triglyceride levels between women with pre-eclampsia compared with healthy women were compared by weighed mean differences (WMDs). The WMDs from individual studies were meta-analysed using a random effects model. Studies were weighted by the inverse of the variance and random effects models were used as standard, as they give conservative estimates of effect.50 We planned a priori subgroup analyses for important confounders that include gestational age, fasting status and body mass index (BMI), at the time of triglyceride measurement, and study design for potential clinical heterogeneity across the studies. Statistical analyses were performed using REVMAN 5.0 (Cochrane Collaboration, Oxford, UK) and STATA 9.0 (Stata Corp, College Station, TX, USA).

ter (Table 1). The cohort studies recruited women prospectively in the second trimester and followed up women during their pregnancy until the diagnosis of pre-eclampsia (Table 2). In 17 studies the measurements of the triglyceride concentrations were carried out on fasting blood samples. The denition of cases and controls was considered adequate in most casecontrol studies (2/24 and 23/ 24, respectively). Often the recruitment of the cases and controls was poorly dened and it was not representative of the population (15/29 for both quality criteria). Controls were commonly matched for gestational and/or maternal age (6/24) and the triglyceride concentrations were measured in similar manner with a similar non-response rate. The ve cohort studies were considered of high quality except for three studies that did not adequately describe the selection of the cases and the controls.

Association between raised triglycerides and pre-eclampsia

Meta-analysis of the results of the 24 casecontrol studies shows that pre-eclampsia is associated with higher levels of serum triglycerides (WMD 0.78 mmol/l, 95% CI 0.600.96, P < 0.00001) (Figure 3). In this meta-analysis, we encountered signicant heterogeneity (I2 = 94%, P < 0.00001). We explored the possible reasons for this heterogeneity and we identied the gestational age of triglyceride measurement as a possible factor. We therefore undertook a subgroup analysis of studies according to the gestational age and found

Results
The studies involved 5857 participants: 1467 women with pre-eclampsia and 4400 healthy women. The main study characteristics of the studies included in this review are summarised in Tables 1 and 2. The included studies were mainly casecontrol studies carried out in the third trimes-

2013 RCOG

1323

Gallos et al.

Figure 2. NewcastleOttawa quality assessment of the studies.

that triglycerides were signicantly higher in the third trimester compared with the second trimester or postpartum (third trimester, WMD 0.86 mmol/l, 95% CI 0.641.09 versus second trimester, WMD 0.23, 95% CI 0.100.36 and postpartum, WMD 0.41, 95% CI 0.300.53, P < 0.00001). Meta-analysis of the ve prospective cohort studies conrms the association of hypertriglyceridaemia, when measured in the second trimester, with pre-eclampsia (WMD 0.24 mmol/l, 95% CI 0.130.34, P < 0.0001). We encountered moderate heterogeneity in this analysis (I2 = 62%, P = 0.03). The triglyceride levels were signicantly different across studies according to the fasting status of the women when the blood samples were taken (v2 = 15.73, P < 0.00001). Our planned adjustment of our inferences for BMI was not performed, as the primary studies did not stratify the results according to BMI.

Discussion
In this systematic review we found that hypertriglyceridaemia is associated with and precedes the onset of pre-eclampsia. We found this association mostly in case control studies performed in the third trimester, but also in cohort studies that included women from the second trimester of pregnancy. From this study, we add epidemiological evidence supporting that hypertriglyceridaemia may be involved in the causal pathway of pre-eclampsia. This inference is justied primarily by the strength of the association found in this study for both second and third trimesters. All the included studies were consistent in suggesting this association and in only three studies (3/29) the 95% condence intervals marginally crossed the line of the null hypothesis being true. Even so, a constellation of metabolic

1324

2013 RCOG

Table 1. Characteristics of the casecontrol studies Cases Women with pre-eclampsia (n = 62) after 20 weeks of gestation Excluded: None Women with pre-eclampsia (n = 51) at a mean gestational age of 34 weeks (SD 4.8) Excluded: Women with signicant obstetric disease or pregnancyunrelated complications Primigravidas at different gestational ages with samples taken at third trimester (n = 67) Age and gestation-matched controls (n = 84) Controls Triglyceride (TG) test

2013 RCOG

Study, year

Barden, 1999 (n = 146)

Belo, 2004 (n = 118)

De, 2006 (n = 100) Women with normal uncomplicated pregnancies (n = 50)

Women with pre-eclampsia (n = 50) Excluded: Women with cardiac, renal, hepatic dysfunction or dyslipidaemia

Francoual, 1999 (n = 50)

TGs were determined antenatally without fasting. Determined enzymatically using Abbott reagents on a COVASMIRA analyser (Roche Diagnostics, Basel, Switzerland) Non-fasted serum samples were obtained at the third trimester of gestation. Serum lipid analysis was performed in an auto-analyser (Cobas Mira S, Roche; Basel, Switzerland) using commercially available kits. TG concentrations were determined by enzymatic colorimetric tests (GPOPAP methods, Roche) Blood samples were drawn from all the women following a fast of 12 hours and analysed for TG by enzymatic methods with the help of Glaxo kits on ERBA Chem-5 semi auto analyser Blood samples were obtained after 12 hours of fasting. Serum total TG were analysed on a Hitachi 717 analyser (Boehringer Mannheim, Mannheim, Germany) Blood samples were obtained after overnight fast. Serum TG were measured by enzymatic methods (Trinder, Bayer Diagnostics, Tarrytown, NY, USA) adapted to a Cobas Mira automated analyser (Hoffmann Larroche, Basel, Switzerland) Fasting serum samples were analysed and TG levels were measured by enzymatic methods (Trinder, Bayer Diagnostics) adapted to a Cobas Mira automated analyser (Hoffmann)

Gratacos, 1996 (n = 155)

Normotensive women (n = 15) throughout gestation without proteinuria or hyperuricaemia with no history of chronic hypertension, diabetes, renal or cardiovascular disease Normotensive healthy pregnant women(n = 115) delivering at term without any of the exclusion criteria and with similar blood samples. Matched for age and BMI Women with normal pregnancies matched for age, BMI, smoking and parity (n = 35)

Gratacos, 2003 (n = 70)

Triglycerides and pre-eclampsia

Harsem, 2007 (n = 81)

Women with mild (n = 23) and severe (n = 12) pre-eclampsia at a mean gestational of 29 3 and 30 5 weeks, respectively Excluded: None Women with mild (n = 20) or severe (n = 20) pre-eclampsia and for whom samples were available for all three trimesters Excluded: Women with diabetes, chronic hypertension endocrine or metabolic disorder Women with severe pre-eclampsia and singleton pregnancies (n = 35) Excluded: Women with pre-existent hypertension, diabetes or other chronic disease, previous pregnancy within 6 months and use of oral contraception Women with pre-eclampsia and singleton pregnancies delivered by caesarean section (n = 40) Women with uncomplicated pregnancies undergoing elective caesarean section (n = 41) between 37 and 42 weeks

Blood was sampled before caesarean section from 88 of the 102 women in gel vials and left at room temperature for 3060 minutes until centrifugation at 2000 g for 10 minutes at room temperature, and serum was stored at 76C until analysis. Serum concentration of TGs was determined by routine enzymatic methods on a Cobas Integra instrument (Roche)

1325

1326
Cases Women with normal pregnancies matched for gestational age (n = 20) Blood samples were obtained before labour and were non-fasting. Serum was obtained using dry sterile tubes in which the blood was allowed to coagulate for 60 minutes at room temperature before centrifugation. Serum and plasma samples were immediately stored at 70C (without thawing) until enzymatic analysis of TG concentrations Blood samples were collected early morning on empty stomach. The concentration of serum TG was measured by GPO-PAP method Controls Triglyceride (TG) test Women with normal pregnancies (n = 20) Normotensive pregnant women (n = 25) Blood samples were taken before antihypertensive treatment Blood samples were collected after women had fasted for 810 hours for the analysis of TG concentrations Venous blood samples were drawn after 8 hours fast. Blood was centrifuged and plasma lipid prole (total cholesterol and triglycerides) and uric acid were measured by quantitative enzymatic assays (Sigma; St Louis, MO, USA)) Normotensive women at the time of admission to hospital (n = 200) at a median gestational age of 38 and range 3840 weeks Consecutive women undergoing routine 3rd trimester blood analysis and with none of the exclusion criteria (n = 30) TGs were measured after 810 hours fast between 17 and 19 weeks Women with pre-eclampsia (n = 20) at the time of admittance to hospital at a median gestational age of 34.5 weeks Excluded: Women with cigarette or illicit drug use, chronic hypertension, renal disease, or a previous history of metabolic disorders Women with pre-eclampsia (n = 40) and singleton pregnancies and not on any medications Excluded: Women with cardiac, hepatic, renal or metabolic or history of hypertension Women with pre-eclampsia (n = 20) Excluded: Women with diabetes, renal disease, primary hypertension or other systemic disease Women with singleton pregnancies and pre-eclampsia (n = 33) at a median gestational age of 36 and range 3339 weeks Women with singleton pregnancies and pre-eclampsia (n = 53) Excluded: Women in labour, with ruptured membranes, multiple pregnancies, smokers or any concurrent medical complications before or developing during pregnancy, such as diabetes mellitus or inammatory diseases Nulliparous women with pre-eclampsia (n = 17) attending the ultrasound screening (1719 weeks) Excluded: None Women with pre-eclampsia (n = 23) with mean gestational age 36 and SD 2 weeks Excluded: None Normotensive healthy controls (n = 17) matched for age, BMI, parity and gestation Women with normal pregnancies (n = 22) with mean gestational age 35 and SD 3 weeks Fasting blood was taken in the morning. Serum lipoproteins were fractionated by ultracentrifugation. TGs were measured with chemicals of reagent grade

Table 1. (Continued)

Gallos et al.

Study, year

Hubel, 1998 (n = 40)

Khaliq, 2000 (n = 60)

Kharb, 1998 (n = 45)

Lei, 2011 (n = 233)

Llurba, 2004 (n = 83)

Lorentzen, 1995 (n = 34)

Maseki, 198152 (n = 45)

2013 RCOG

Table 1. (Continued) Cases Controls were recruited from the antenatal clinics (n = 46) Controls Triglyceride (TG) test Fasting blood samples were obtained and TG measurements were made using standardised techniques

2013 RCOG

Study, year

Mikhail, 1995

Murai, 1997 (n = 62) Women with normal pregnancies matched for nulliparity, race, maternal age at delivery (within 3 years), and duration of pregnancy (within 2 weeks) (n = 31)

Pregnant women with pre-eclampsia (n = 29) and singleton pregnancies from 28 to 42 weeks Excluded: Women with ruptured membranes, in labour or with medical complications Women with singleton pregnancies and pre-eclampsia (n = 31) Excluded: None

Ozan, 1997 (n = 59) Women with uncomplicated pregnancies (n = 20)

Non-fasting serum samples were collected during the late third trimester of pregnancy before labour or administration of antihypertensive agents or intravenous uids. Blood samples were drawn at 3638 weeks of gestation from asymptomatic women. Plasma TGs were quantied in matched women using a microtitre-plate modication of the Sigma triglycerides diagnostic kit Fasting serum TG level measurements were obtained at admission to hospital. Serum triglyceride levels were determined by enzymatic-spectrophotometric methods (BioSystems, American Biosystems, Inc., Roanoke, VA, USA)

Powers, 1998 (n = 54)

Women with pre-eclampsia (n = 39) Excluded: Women with a history of antihypertensive medication, cardiac disorder, diabetes mellitus, hyperlipidaemia, hepatic or renal disorder, the smokers and the drinkers Women with pre-eclampsia (n = 21) Excluded: None Women with uncomplicated pregnancy with no medical problems (n = 21)

Raijmakers, 2004 (n = 64)

Women with pre-eclampsia (n = 40) with a median gestational age of 30 weeks and range from 24 to 36 weeks

Plasma samples were obtained 6 hours after each womans most recent meal and all measurements were performed with appropriate diagnostic kits supplied by Sigma Chemical Co Fasting status is unreported or technique of measurement

Sahu, 2009 (n = 60)

Women with uncomplicated pregnancies matched for gestational age (n = 24) with a median gestational age of 31 weeks and range from 30 to 32 weeks Primigravida women matched for age in their third trimester (n = 30)

Fasting samples were analysed by enzymatic colorimetric methods in the autoanalyser (Dimension AR, Dade Behring Limited, Milton Keynes, UK)

Triglycerides and pre-eclampsia

Schjetlein, 1999 (n = 297)

Primigravida women with pre-eclampsia (n = 30) attending the hospital in their third trimester Excluded: Women with diabetes mellitus with or without treatment, obesity, severe anaemia, eclampsia or women suffering from any other systemic or endocrine disorder Women with pre-eclampsia (n = 200) with a mean gestational age of 35.6 weeks and range from 27 to 40 weeks Excluded: Women with essential hypertension, diabetes, nephrotic syndrome or thyrotoxicosis

Women with uncomplicated pregnancies (n = 97) with a mean gestational age of 36.3 weeks and range from 24 to 42 weeks

Fasting samples were analysed for TG levels with conventional routine methods

1327

1328
Cases Caucasian primiparous women with uncomplicated pregnancies (n = 29) Controls Triglyceride (TG) test Fasting samples were analysed by standard automated laboratory techniques (Beckman Coulter LX20 PRO, Fullerton, CA, USA) 23 years after pregnancy Caucasian primiparous women with pre-eclampsia (n = 22) Excluded: Women with pre-existent hypertension, diabetes mellitus, renal disease, current cancer therapy or chronic use of corticosteroid medication Women with pre-eclampsia (n = 43) at a mean gestational age of 35.7 and a standard error of 4 weeks Excluded: None Women with pre-eclampsia (n = 125) Excluded: Women with chronic hypertension and postpartum Pregnant women with no obvious medical problems (n = 143) at a mean gestational age of 31.6 and a standard error of 7 weeks Women with uncomplicated pregnancies matched for gestational and maternal age (n = 179) Women in their postpartum with pre-eclampsia (n = 173) Excluded: Women with chronic hypertension Women with uncomplicated pregnancies delivered within 2 hours of the cases (n = 186) TG concentration was determined by the method of Spayd and co-workers51 with a Vitros analyzer clinical chemistry slide and a Vitros 950 analyzer (Ortho Diagnostics, Rochester, NY, USA) TG concentrations were mostly fasting (94%) and measured enzymatically employing assays standardised by the Lipid Standardization Programme of the Centres for Disease Control and Prevention, Atlanta, GA, USA TG concentrations were measured 1272 hours postpartum enzymatically using assays standardised by the Lipid Standardization Programme of the Centres for Disease Control and Prevention, Atlanta, GA, USA

Gallos et al.

Table 1. (Continued)

Study, year

Spaan, 2010 (n = 51)

Vanderjagt, 2004 (n = 173)

Ware-Jauregui, 1999 (n = 304)

Williams, 2003 (n = 359)

2013 RCOG

Table 2. Characteristics of the cohort studies Population Women were observed routinely for pre-eclampsia, pregnancy-induced hypertension according to early or late onset (before or after 37 weeks) Triglyceride (TG) test Outcome Study Design/Follow up Prospective cohort study Follow up: 6.7% of women delivered elsewhere

2013 RCOG

Study, year

Clausen, 2001 (n = 2157)

Pregnant women at their ultrasound scan (1719 weeks) Excluded: Women with twin pregnancies

Enquobahrie, 2004 (n = 567) Women were observed routinely for pre-eclampsia, pregnancy-induced hypertension according to early or late onset (before or after 37 weeks)

Pregnant women attending prenatal care clinic before 16 weeks of gestation Excluded: Women with chronic hypertension, diabetes mellitus and spontaneous or induced abortions

Prospective cohort study Follow up: n = 45 of women delivered elsewhere, moved or their medical records were missing

Setareh, 2009 (n = 343)

Triglycerides were measured at the time of recruitment in a non-fasting state (1719 weeks) and allowed to coagulate before centrifugation at 400 g for 10 minutes. The serum samples were transferred on ice to a 708C freezer within 140 minutes after venepuncture Maternal nonfasting samples, collected at an average of 13 weeks gestation. Maternal plasma TG concentrations were measured enzymatically using assays standardised by the Lipid Standardization Programme of the Centres for Disease Control and Prevention (Atlanta, GA, USA) Fasting blood samples were taken on week 16 for TG measurement. Enzymatic colorimetric test was used to dene TG

Prospective cohort study Follow up: None reported

Takahashi, 2008 (n = 48) In the rst trimester (until the 13th week) and second trimester TGs were measured. The measurement of serum TG was done using the enzymatic colorimetric method on an Advia 1650 (Bayer) device Triglycerides were measured at 2832 weeks of pregnancy

In follow up, blood pressure of both severe pre-eclampsia and control group were recorded again in week 38 of pregnancy and pre delivery. Only severe pre-eclampsia is compared with control Women were observed and groups were separated into those who developed pre-eclampsia and those who did not

Prospective cohort study Follow up: No losses

Ziaei, 2006 (n = 470)

colname="col2">Pregnant women referred to antenatal care before the week 16 Excluded: Women with diabetes, chronic hypertension, previous obstetric complications or other systemic disease Pregnant over 35 years of age women with ongoing gestations (>13 weeks) of single foetuses Excluded: Women with a history of cardiovascular or kidney disease, and/or diabetes mellitus Women in their rst pregnancy between 28 and 32 weeks (n = 470) Exclusion: Women with diabetes mellitus or any other endocrine or metabolic disorder, any history of cardiovascular disease and hypertension, smoking, and nonsingleton pregnancy

Women were observed once every other week until 36 weeks of gestational age. Afterward, they were checked once a week until 40 weeks of gestational age. For each woman who experienced pre-eclampsia, a non-hypertensive healthy pregnant participant delivering at term was matched for age, body mass index, educational level, family income, and occupation. Triglyceride levels from women with pre-eclampsia were compared with the matched women with normal pregnancies

Prospective cohort study Follow up: No losses

Triglycerides and pre-eclampsia

1329

Gallos et al.

Figure 3. Forest plot showing the results of meta-analysis of studies along with calculated exact binomial condence intervals that examine the difference in triglyceride concentrations in women with pre-eclampsia compared with normal controls. Results subgrouped according to the study design.

changes may happen that lead to pre-eclampsia, and hypertriglyceridaemia may only explain a part of this pathway. This prevents us from drawing strong conclusions about causality from this study. The temporality, though, where hypertriglyceridaemia clearly precedes the onset of pre-eclampsia, leads us to generate the hypothesis that we may be able to change the natural history of the disease if we intervene early by lowering the triglyceride levels. Before such an intervention it would be important to dene the normal triglyceride levels in pregnancy and correctly identify women that could benet most from this therapy. A weakness, which is difcult to account for, is that the observed association may be overestimated because of the study design in casecontrol studies, but this was also proven in ve prospective cohort studies when analysed separately. The casecontrol studies were signicantly

different between themselves, which is reected in the high heterogeneity we encountered in this meta-analysis. This was partially explained from the different gestational age and fasting status of the targeted populations across the studies. The selection of controls varied across the studies, which introduced further heterogeneity. Potential bias is also possible in the casecontrol studies because the cases were not always representative of women with pre-eclampsia. The selection of controls did not include community controls and convenient hospital controls were often used. This introduces bias and in most of the studies the comparability of cases and controls was found to be poor. The cohort studies were of higher quality and their results are likely to be more reliable. Of note, is the fact that hypertriglyceridaemia may be associated with nutrition, and indeed, it was our aim also to adjust our estimates for BMI, which

1330

2013 RCOG

Triglycerides and pre-eclampsia

is a potential confounder, but the primary studies did not stratify their results according to BMI. However, BMI in women with pre-eclampsia compared with healthy women was reported and statistically tested in 12 studies. In the majority of the studies there was only weak evidence that groups were different for BMI with P-values >0.05 in 11 studies. However, the direction of the association was not different for any of the included studies and only the strength of the association differed.

Conclusion
The association between hypertriglyceridaemia and pre-eclampsia were signicant in both analyses of case control and cohort studies. The cohort design of ve included studies also highlights the temporality of this association where hypertriglyceridaemia present in the second trimester preceded the onset of pre-eclampsia, which was often diagnosed in the third trimester. This is clinically attractive because measurement of triglycerides is well established in all clinical laboratories and may represent a cost-effective way of identifying at-risk pregnancies. The role of hypertriglyceridaemia in the pathogenesis of the disease and particularly potential mechanisms by which it might be modulated are potential avenues for further research.

Disclosure of interests
None to be declared.

Contribution to authorship
IDG and MV conceptualised this study. IDG, KS and MV performed the search, selected abstracts, obtained the full manuscripts and extracted the data. IDG performed the meta-analysis and wrote all versions of the manuscript. MK, AC, ST and MV critically revised the manuscript and all authors approved the nal version.

Details of ethics approval

Not required.

Funding
No funding was sought for this study.

Acknowledgements
None. &

References
1 MacGillivray I, Davey DA. Classication of hypertensive disorders of pregnancy. In: Sharp F, Symonds EM, editors. Proceedings of the 16th Study Group of RCOG. Ithaca, NY: Perinatology Press; 1986. pp. 4018.

2 Hay JE. Liver disease in pregnancy. Hepatology 2008;47:106776. 3 Maynard SE, Thadhani R. Pregnancy and the kidney. J Am Soc Nephrol 2009;20:1422. 4 Stennett AK, Khalil RA. Neurovascular mechanisms of hypertension in pregnancy. Curr Neurovasc Res 2006;3:13148. 5 Gruslin A, Lemyre B. Pre-eclampsia: fetal assessment and neonatal outcomes. Best Pract Res Clin Obstet Gynaecol 2011;25:491507. 6 Andersgaard AB, Acharya G, Mathiesen EB, Johnsen SH, Straume B, ian P. Recurrence and long-term maternal health risks of hypertensive disorders of pregnancy: a population-based study. Am J Obstet Gynecol 2012;206:143.e1143.e8. 7 Ray JG, Vermeulen MJ, Schull MJ, Redelmeier DA. Cardiovascular health after maternal placental syndromes (CHAMPS): population-based retrospective cohort study. Lancet 2005;366: 1797803. 8 Ghio A, Bertolotto A, Resi V, Volpe L, Di Cianni G. Triglyceride metabolism in pregnancy. Adv Clin Chem 2011;55:13353. 9 Silasi M, Cohen B, Karumanchi SA, Rana S. Abnormal placentation, angiogenic factors, and the pathogenesis of preeclampsia. Obstet Gynecol Clin North Am 2010;37:23953. 10 Sweeney M, Wareing M, Mills TA, Baker PN, Taggart MJ. Characterisation of tone oscillations in placental and myometrial arteries from normal pregnancies and those complicated by pre-eclampsia and growth restriction. Placenta 2008;29:35665. 11 Foidart JM, Schaaps JP, Chantraine F, Munaut C, Lorquet S. Dysregulation of anti-angiogenic agents (sFlt-1, PLGF, and sEngoglin) in preeclampsiaa step forward but not the denitive answer. J Reprod Immunol 2009;82:10611. 12 Wang Y, Lewis DF, Gu Y, Zhang Y, Alexander JS, Granger DN. Placental trophoblast-derived factors diminish endothelial barrier function. J Clin Endocrinol 2004;89:24218. 13 Dunne F. Type 2 diabetes and pregnancy. Semin Fetal Neonatal Med 2005;10:3339. 14 Coomarasamy A, Honest H, Papaioannou S, Gee H, Khan KS. Aspirin for prevention of preeclampsia in women with historical risk factors: a systematic review. Obstet Gynecol 2003;101:131932. 15 Sohlberg S, Stephansson O, Cnattingius S, Wikstrom AK. Maternal body mass index, height, and risks of preeclampsia. Am J Hypertens 2012;25:1205. 16 Di Cianni G, Miccoli R, Volpe L, Lencioni C, Del Prato S. Intermediate metabolism in normal pregnancy and in gestational diabetes. Diabetes Metab Res Rev 2003;19:25970. 17 Montes A, Walden CE, Knopp RH, Cheung M, Chapman MB, Albers JJ. Physiologic and supraphysiologic increases in lipoprotein lipids and apoproteins in late pregnancy and postpartum. Possible markers for the diagnosis of prelipemia. Arteriosclerosis 1984;4:40717. 18 Maynard SE, Min JY, Merchan J, Lim KH, Li J, Mondal S, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003;111:64958. 19 Levine RJ, Lam C, Qian C, Yu KF, Maynard SE, Sachs BP, et al. Soluble endoglin and other circulating antiangiogenic factors in preeclampsia. N Engl J Med 2006;355:9921005. 20 Robinson NJ, Minchell LJ, Myers JE, Hubel CA, Crocker IP. A potential role for free fatty acids in the pathogenesis of preeclampsia. J Hypertens 2009;27:1293302. 21 Barden AE, Beilin LJ, Ritchie J, Walters BN, Michael CJ. Does a predisposition to the metabolic syndrome sensitize women to develop pre-eclampsia? Hypertension 1999;17:130715. 22 Belo L, Gaffney D, Caslake M, Santos-Silva A, Pereira-Leite L, Quintanilha A, et al. Apolipoprotein E and cholesteryl ester transfer protein polymorphisms in normal and preeclamptic pregnancies. Eur J Obstet Gynecol Reprod Biol 2004;112:915.

2013 RCOG

1331

Gallos et al.

23 De J, Mukhopadhyay AK, Saha PK. Study of serum lipid prole in pregnancy induced hypertension. Indian J Clin Biochem 2006;21:1658. 24 Francoual J, Audibert F, Claise C, Chalas J, Trioche P, Frydman R, et al. Implication of apolipoprotein E and the L-arginine-nitric oxide system in preeclampsia. Hypertens Pregnancy 1999;18:22937. s E, Casals E, Sanllehy C, Cararach V, Alonso PL, Fortuny A. 25 Grataco Variation in lipid levels during pregnancy in women with different types of hypertension. Acta Obstet Gynecol Scand 1996;75:896 901. s E, Casals E, Go mez O, Llurba E, Mercader I, Cararach V, 26 Grataco et al. Increased susceptibility to low density lipoprotein oxidation in women with a history of pre-eclampsia. BJOG 2003;110:4004. 27 Harsem NK, Roald B, Braekke K, Staff AC. Acute atherosis in decidual tissue: not associated with systemic oxidative stress in preeclampsia. Placenta 2007;28:95864. 28 Hubel CA, Lyall F, Weissfeld L, Gandley RE, Roberts JM. Small low-density lipoproteins and vascular cell adhesion molecule-1 are increased in association with hyperlipidemia in preeclampsia. Metabolism 1998;47:12818. 29 Khaliq F, Singhal U, Arshad Z, Hossain MM. Study of serum lipid and lipoprotein in pre-eclampsia with special reference to parity. Indian J Physiol Pharmacol 2000;44:1926. 30 Kharb S, Gulati N, Singh V, Singh GP. Lipid peroxidation and vitamin E levels in preeclampsia. Gynecol Obstet Invest 1998;46:23840. 31 Lei Q, Lv LJ, Zhang BY, Wen JY, Liu GC, Lin XH, et al. Ante-partum and post-partum markers of metabolic syndrome in pre-eclampsia. J Hum Hypertens 2011;25:117. s E, Mart 32 Llurba E, Grataco n-Gall an P, Cabero L, Dominguez C. A comprehensive study of oxidative stress and antioxidant status in preeclampsia and normal pregnancy. Free Radic Biol Med 2004;37:55770. 33 Lorentzen B, Drevon CA, Endresen MJ, Henriksen T. Fatty acid pattern of esteried and free fatty acids in sera of women with normal and pre-eclamptic pregnancy. Br J Obstet Gynaecol 1995;102:5307. 34 Mikhail MS, Basu J, Palan PR, Furgiuele J, Romney SL, Anyaegbunam A. Lipid prole in women with preeclampsia: relationship between plasma triglyceride levels and severity of preeclampsia. J Assoc Acad Minor Phys 1995;6:435. 35 Murai JT, Muzykanskiy E, Taylor RN. Maternal and fetal modulators of lipid metabolism correlate with the development of preeclampsia. Metabolism 1997;46:9637. 36 Ozan H, Esmer A, Kolsal N, Copur OU, Ediz B. Plasma ascorbic acid level and erythrocyte fragility in preeclampsia and eclampsia. Eur J Obstet Gynecol Reprod Biol 1997;71:3540. 37 Powers RW, Evans RW, Majors AK, Ojimba JI, Ness RB, Crombleholme WR, et al. Plasma homocysteine concentration is increased in preeclampsia and is associated with evidence of endothelial activation. Am J Obstet Gynecol 1998;179:160511.

38 Raijmakers MT, van Tits BJ, Hak-Lemmers HL, Roes EM, Steegers EA, Peters WH. Low plasma levels of oxidized low density lipoprotein in preeclampsia. Acta Obstet Gynecol Scand 2004;83:11737. 39 Sahu S, Abraham R, Vedavalli R, Daniel M. Study of lipid prole, lipid peroxidation and vitamin E in pregnancy induced hypertension. Indian J Physiol Pharmacol 2009;53:3659. 40 Schjetlein R, Abdelnoor M, Haugen G, Husby H, Sandset PM, Wislff F. Hemostatic variables as independent predictors for fetal growth retardation in preeclampsia. Acta Obstet Gynecol Scand 1999;78:1917. 41 Spaan JJ, Houben AJ, Musella A, Ekhart T, Spaanderman ME, Peeters LL. Insulin resistance relates to microvascular reactivity 23 years after preeclampsia. Microvasc Res 2010;80:41721. 42 Vanderjagt DJ, Patel RJ, El-Nafaty AU, Melah GS, Crossey MJ, Glew RH. High-density lipoprotein and homocysteine levels correlate inversely in preeclamptic women in northern Nigeria. Acta Obstet Gynecol Scand 2004;83:53642. 43 Ware-Jauregui S, Sanchez SE, Zhang C, Laraburre G, King IB, Williams MA. Plasma lipid concentrations in pre-eclamptic and normotensive Peruvian women. Int J Gynaecol Obstet 1999;67:14755. 44 Williams MA, Woelk GB, King IB, Jenkins L, Mahomed K. Plasma carotenoids, retinol, tocopherols, and lipoproteins in preeclamptic and normotensive pregnant Zimbabwean women. Am J Hypertens 2003;16:66572. 45 Clausen T, Djurovic S, Henriksen T. Dyslipidemia in early second trimester is mainly a feature of women with early onset pre-eclampsia. BJOG 2001;108:10817. 46 Enquobahrie DA, Williams MA, Butler CL, Frederick IO, Miller RS, Luthy DA. Maternal plasma lipid concentrations in early pregnancy and risk of preeclampsia. Am J Hypertens 2004;17:57481. 47 Setareh A, Mitra MG, Sedigheh B, Shoaleh S, Vahid Y, Siroos S. Maternal plasma lipid concentrations in rst trimester of pregnancy and risk of severe preeclampsia. Pakistan J Med Sci 2009;25:5637. 48 Takahashi WH, Martinelli S, Khoury MY, Lopes RGC, Garcia SAL, Lippi UG. Assessment of serum lipids in pregnant women aged over 35 years and their relation with pre-eclampsia. Einstein 2008;6:637. 49 Ziaei S, Bonab KM, Kazemnejad A. Serum lipid levels at 28 32 weeks gestation and hypertensive disorders. Hypertens in Pregnancy 2006;25:310. 50 Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. The Cochrane Collaboration, 2011. [www.cochrane-handbook.org]. Accessed 15 February 2013. 51 Spayd RW, Bruschi B, Burdick BA, Dappen GM, Eikenberry JN, Esders TW, et al. Multilayer lm elements for clinical analysis. Clin Chem 1978;24:134850. 52 Maseki M, Nishigaki I, Hagihara M, Tomoda Y, Yagi K. Lipid peroxide levels and lipid content of serum lipoprotein fractions of pregnant subjects with or without preeclampsia. Clin Chim Acta 1981;115:15561.

1332

2013 RCOG