Familial Risk of Preeclampsia in Newfoundland: A Population-Based Study
Lesa M. Dawson*,
Patrick S. Parfrey,,
Donna Hefferton,,
Elizabeth L. Dicks,,
M. Jim Cooper,,
David Young* and
Philip A. Marsden
*Department of Obstetrics, Division of Nephrology, and Clinical Epidemiology Unit, Memorial University, St. Johns, Newfoundland, Canada; Renal Division and Department of Medicine, St. Michaels Hospital and University of Toronto, Toronto, Ontario, Canada.
Correspondence to: Dr. Patrick S. Parfrey, Clinical Epidemiology Unit, Health Sciences Center, St. Johns, Newfoundland A1B 3V6, Canada. Phone: 709-777-7261; Fax: 709-777-6995; E-mail pparfrey{at}mun.ca
ABSTRACT. This study sought to quantify the familial risk ofpreeclampsia (proteinuric hypertension) in Newfoundland andto identify characteristics in probands that predict increasedfamilial risk. Reviewed were 5173 obstetric charts from 10 hospitals,representing 99% of deliveries on the island of Newfoundlandfor a 1-yr period from April 1996 to March 1997; pregnancy-inducedhypertension was diagnosed according to strict criteria. Familyobstetric histories were obtained from identified probands withpreeclampsia, and sisters and mothers of probands were interviewed.In addition, the obstetric charts from sisters and mothers werereviewed to identify preeclampsia. The incidence of preeclampsiain the population was 5.6% (n = 292), and in primiparous womenit was 7.9%. Factors independently associated with increasedrisk of preeclampsia included primiparous delivery, multiplegestation, pregestational and gestational diabetes, maternalage of more than 35 yr, and region of the province. Of 330 sistersidentified, 217 had 445 pregnancies, with 331 charts locatedfor review. The incidence of preeclampsia (based on chart review)in 163 primiparous sisters was 20.2%. The relative risk of preeclampsiain primiparous sisters of probands with preeclampsia comparedwith primiparous women in the population was 2.6 (95% confidenceinterval, 1.8 to 3.6). Factors in probands independently associatedwith a higher risk of preeclampsia in sisters included at least2+ proteinuria and region of the province. This population-basedstudy, which used unbiased ascertainment and strict diagnosticcriteria, demonstrated a significantly higher risk of preeclampsiain sisters of probands with preeclampsia, particularly whenprobands were defined by severity of preeclampsia and by geographicregion.
Preeclampsia is an important cause of morbidity and mortalityin mothers and newborns. The pathophysiology of the systemicvasospasm, proteinuric hypertension, and multiple end-organischemia remains poorly understood. Modest progress in prediction,prophylaxis or treatment has occurred during the past 30 yr.Preeclampsia occurs at higher rates in sisters, daughters, andmothers of affected women (17), but genetic study ofthis disease remains an enormous challenge, requiring not onlyfemale gender and reproduction to manifest the phenotype, butconsideration of fetal-maternal and fetal-paternal genotypeinteractions (811). To date, the molecular determinantsimportant in the inherited predisposition to preeclampsia remainunknown.
The population of Newfoundland has arisen from Irish and Englishsettlement in the late 18th and early 19th centuries, with littlesubsequent outmigration or immigration (12). A total of 90%of the islands current 550,000 residents are descendedfrom approximately 20,000 original settlers. Until recent decades,isolation was enhanced by dependence on outport fisheries, lackof roads, and segregation by religion. This persistent geneticisolation produced a high coefficient of kinship (13,14), which,together with relatively large families, close family ties,and a modern health care system, facilitates investigation ofgenetic disease.
Our objective was to quantify the risk for the development ofpreeclampsia in the sisters and mother of probands with preeclampsiain the population, to identify predictors of enhanced familialrisk in Newfoundland, and to determine whether a search forsusceptibility genes should be undertaken. Strict diagnosticcriteria (15) were applied by using data derived from the primarydocuments, and probands were identified from the total provincialpopulation at risk during 1 yr.
Identification of preeclamptic probands and sisters at riskof developing preeclampsia is shown in Figure 1. A total of5173 obstetric charts from 10 hospitals were reviewed, constituting99% of all third-trimester deliveries on the island of Newfoundlandfrom April 1996 to March 1997. Each record was reviewed by aresearch nurse or trained research assistant, who summarizedrelevant information into a detailed data summary sheet. Thisincluded demographic and clinical information, data regardingobstetric history, diabetes, renal disease, previous preeclampsia,hypertension, BP during pregnancy, and results of urine tests.ICD-9 hospital codes for hypertension as used by Canadian Institutefor Health Information were collected for comparison with researchresults. One chart contained too few data to include in theanalysis. Consequently, our study population size was 5172.All data sheets were reviewed by a specialist obstetrician (L.D.),who assigned the diagnoses for hypertensive diseases in pregnancy.Strict diagnostic criteria were used in concordance with guidelinesfrom the 1997 Canadian Hypertension Society Consensus Conference(16).
Figure 1. Charts identifying the preeclampsia probands in the Newfoundland population (left) and identifying sisters of these probands at risk of preeclampsia (right).
Study design was approved by the hospital ethics committee.Once probands with preeclampsia had been identified, each wassent an initial explanatory contact letter from the deliveringphysician or their family physician. Each proband was then contactedby telephone, and after consent had been provided, detailedpregnancy history and family history were obtained. Contactinformation for female first-degree relatives was obtained.Mothers and sisters were then interviewed in the same manneras for probands. Written consent for obstetric chart reviewwas obtained, regardless of history of hypertensive pregnancy.Consent forms were forwarded to delivering hospitals. Chartson the island of Newfoundland were reviewed by research staffin person. For deliveries occurring elsewhere, charts were forwardedto the research team for analysis. Multiple attempts were madein every case to locate older records or those in other provincesor countries. We reviewed all available charts from deliveriesin sisters and mothers by using the same procedure and criteriaas for the provincial population. The same research staff andclinician reviewed both population and relative data. Each recordwas allocated to one diagnostic category on the bases of definedclinical criteria, in concordance with guidelines from the 1997Canadian Hypertension Society Consensus Conference (16) (Table1).
Table 1. Definitions for diagnostic categories of pregnancy-induced hypertensiona
Of 292 probands who developed preeclampsia in this population,15 probands were excluded from the family study because theyhad a factor predisposing to preeclampsia9 had twin pregnancies,and 6 had insulin-dependent diabetes mellitus. Of 277 eligibleprobands, 210 (76%) were interviewed. The population under studydid not reflect the important contribution of donated gametesto the incidence of preeclampsia (17). Women were not interviewedfor the following reasons: 6 refused, 55 had inaccurate contactinformation, and 6 had no physician contact available. The 67probands not included in the family study differed from the210 included, in that 76 versus 67% were primiparous and 12versus 4% were aged <19 yr. There were no differences forregions of domicile.
Of the 330 sisters of probands identified from these 210 interviews,95 were nulliparous. Of 235 parous sisters, 217 (92%) were interviewed,and 445 pregnancies with third-trimester births were confirmed.Seventy-four percent (n = 331) of pregnancies were reviewedby using the hospital chart as the source document (Figure 1).
Two hundred eight mothers of probands were located for interviewand consented for chart review of their 821 deliveries. Of those,446 charts (54%) were located for analysis. Of 375 pregnancieswithout charts available, 269 occurred before 1965 or had noavailable delivery date. Forty-one were home deliveries.
Univariate comparison of continuous variables was undertakenby t tests and of ordinal variables by 2 tests. Incidence rates(with 95% confidence intervals [CI]) for all types of hypertensionin pregnancy were calculated in the population. Relative risk(with 95% CI) for each risk factor potentially contributingto development of preeclampsia was calculated. Predictors forthe development of preeclampsia in probands and in primiparoussisters of probands were sought by multiple logistic regression,the forward stepwise method, and SPSS software version 10.0(SPSS Inc., Chicago, IL).
Of the 5172 deliveries, 2500 were primiparous. Multiple gestationoccurred in 1.3% (n = 67) pregnancies. Mean age was 27 ±5 yr. There were 0.7% (n = 8) pregnancies in which deliveryoccurred at less than 30 wk, and 7.0% (n = 364) less than 37wk. Gestational diabetes mellitus was noted in 3.7% (n = 192).Twenty-eight patients (0.5%) had pregestational diabetes.
The incidence of preeclampsia and hypertension, assessed bythe predefined criteria, in all deliveries and in primiparousdeliveries in the province is listed in Table 2. A total of292 (5.6%) women were classified as having preeclampsia and283 (5.5%) as having nonproteinuric hypertension. There wereno cases of eclampsia. Of the 2500 primiparous patients, 198(7.9%) developed preeclampsia and 6.7% nonproteinuric hypertension(Table 2). Of 3479 normotensive mothers who had urine testsperformed, 527 (15.1%, 95% CI, 14.0 to 16.3) had 1+ proteinuria.Of 1529 primiparous pregnancies with a urine test available,248 (16.2%, 95% CI, 14.4 to 18.1) had 1+ proteinuria.
Table 2. Incidence of pregnancy-induced and preexisting hypertension in primiparous (n = 2500) and in all deliveries (n = 5172) in Newfoundland during 1996 to 1997a
The incidence of preeclampsia with at least 2+ proteinuria was2.3% (n = 119), and of preeclampsia with diastolic BP at least110 mmHg, it was 1.0% (n = 50). Evidence of hemolysis, elevatedliver enzymes, and low platelets (HELLP syndrome) was observedin 38 (0.7%) patients, of whom 9 patients had increases in serumaspartate transaminase only and 3 more had thrombocytopeniawithout documentation of liver disease.
Geographic variations in the rates of preeclampsia were observed,with rates varying from 11 to 4% (Figure 2), with significantlyhigher risk observed in the Peninsulas and Western regions.Multiple logistic regression was undertaken to identify independentrisk factors for preeclampsia. For this analysis, only womenwho did not develop pregnancy-induced hypertension (n = 3767)and those who developed preeclampsia were included (n = 292).It excluded those who developed nonproteinuric hypertension,intrapartum hypertension, those whose hypertension could notbe classified, and those who had preexisting hypertension. Allpredictors included in the model are listed in Table 3. Independentrisk factors include primiparity, multiple gestation, pregestationaldiabetes, gestational diabetes, maternal age of more than 35yr, and two regions of the island (Peninsulas and Western) (overall2 = 167, P < 0.001) (Table 3).
Table 3. Independent predictors of preeclampsia in Newfoundland identified by multiple logistic regressiona
Preeclampsia in Mothers
Mean age at first delivery was 20 ± 4 yr. At interview,46 (22%) of 208 mothers reported current hypertension. Eighty-sevenfirst pregnancy records were classified, with preeclampsia occurringin 11 (12.6%). The risk of preeclampsia in primiparous motherswas 1.6 more than in primiparous deliveries in the provincialpopulation (95% CI, 0.9 to 3.2).
Preeclampsia in Sisters
Among sisters of probands with preeclampsia, 163 primiparousdeliveries were reviewed. The mean age of delivery was 24 ±5 yr. The rate of pregestational diabetes was unchanged fromthe population rates. Gestational diabetes occurred more frequentlyat 6.3%. There were two sets of twins (0.6%). Preeclampsia occurredin 33 (20.2%) of first pregnancies. The risk of preeclampsiain primiparous sisters was 2.6-fold more than in primiparousdeliveries in the provincial population (95% CI, 1.8 to 3.6).
Predictors of Increased Family Risk
The incidence of preeclampsia in sisters of probands with preeclampsiaincreased when preeclampsia probands were characterized as havingat least 2+ proteinuria, at least one diastolic BP of more than110 mmHg, HELLP syndrome, or delivery in hospital in the Peninsulasregion (Table 4). Preeclampsia occurred in 29.2% of primiparoussisters of probands who had preeclampsia and at least 2+ proteinuria.The relative risk of preeclampsia in the sisters of these probandswas 3.7 (95% CI, 2.5 to 5.5), relative to that in primiparousdeliveries in the provincial population.
Table 4. Incidence and relative risk of preeclampsia in primiparous sisters of probands with preeclampsia, the latter defined by severity of preeclampsia or by region of domicile
A multiple logistic analysis was performed to identify independentrisk factors for preeclampsia in sisters of probands. Includedin the model were the presence or absence of the following characteristicsin the proband of each sister: 2+ proteinuria, diastolic BPat least 110 mmHg, HELLP syndrome, and living in Peninsulasregion. Excluded from analysis were sisters who had gestationaldiabetes or who had multiple gestation (n = 11). Table 5 revealsthat independent risk factors for preeclampsia in sisters werethe presence of at least 2+ proteinuria in probands and theregion in which the proband was living.
Table 5. Independent predictors of preeclampsia in 152 sisters of probands with preeclampsia, defined by characteristic of probanda
In mothers of probands with preeclampsia and at least 2+ proteinuria,9 (22.5%) of 40 first pregnancies in mothers were associatedwith preeclampsia (relative risk, 2.8; 95% CI, 1.6 to 5.1),and in mothers of probands with HELLP, 6 (42.9%) of 14 firstpregnancies were associated with preeclampsia (relative risk,4.5, 95% CI, 2.2 to 9.2)
Families with Preeclampsia
Fourteen families with 3 or 4 members with preeclampsia and36 families with 2 members with preeclampsia (including proband)have been identified. Two sets of sisters were also probandsdelivering in the study period, leading to a total of 51 familieswith 2 or more affected members.
Accurate phenotypic identification of preeclampsia remains achallenge in the genetic studies of the pathogenesis of preeclampsia(18). We have rigorously applied strict definitions (16) usingprimary records as sources of data. The design of the studywe report here stressed a population-based recruitment versusa center-based cohort recruitment, thus eliminating ascertainmentbias in the identification of probands. Ascertainment of sistersof probands was good because of the excellent participationrate of probands identified from the population and the highresponse rate from sisters. Ascertainment in mothers was poorbecause of practical difficulties in locating records, becauseof home deliveries, or because of remote time frames. In ourstudy, location of pregnancy information was successful in 77%of hospital deliveries occurring after 1965. HELLP syndromewas not well defined as a clinical entity before 1980, and thereforeinformation on liver function or platelet count was not locatedin charts before this period.
The population rate of preeclampsia in primiparous Newfoundlandwomen (7.9%) is similar to that in the American Calcium to PreventPreeclampsia trial (7.1%) (15). Previous epidemiologic studyof preeclampsia has described factors linked with higher incidenceof disease, particularly primiparity, multiple gestation, oldermaternal age, and pregestational diabetes (1921). Observationsin the Newfoundland population further confirm these risk factors.
The regional differences observed for incidence rates of preeclampsiain the population and for incidence rates in sisters of probands(defined by the probands region) suggest multiple geneticisolates may be present in Newfoundland. After the major migrationsto Newfoundland before 1840, natural increase became the dominantmechanism of population growth (12). Community expansion occurredpredominantly by partition of ancestral lands among heirs orby movement to other nearby uninhabited lands near the settlementcore, close to fishing grounds. This kept related families together.Even in 1982, 50% of the population lived in communities offewer than 2500 people, and 41% lived in communities of fewerthan 1000 people (13,14). Thus, the presence of multiple geneticallysimplified isolates, with strong founder effects and a highcoefficient of kinship, may explain the higher sibling riskfor preeclampsia observed in certain regions of the province.
However, environmental factors specific to the Western and Peninsulasregions could also be at work here, particularly because thereare substantial socioeconomic differences between these regionsand the Eastern region, the latter being more prosperous, withless dependence on fishing and with higher employment rates.Our data further confirm that sisters of probands with preeclampsiahave an elevated risk of preeclampsia, and this is so even withthe use of more precise definitions of preeclampsia than inother studies (4,6). Characterization of probands by the severityof preeclampsia (presence of at least 2+ proteinuria, HELLPsyndrome) is associated with higher risk in sisters of probands,an observation not previously reported. This increased risksuggests the inheritance of molecular determinants that contributeto development of the disease.
In most populations, a sibling risk of 2 to 4, as seen in thisstudy, would be considered too low for a linkage analysis approach.However, genetically simplified isolates are the optimal choiceof study population when mapping genes in complex diseases (22).It has been argued that genetic studies in very young founderpopulations, such as for Newfoundland, Quebec, and Costa Rica,may be able to exploit less dense marker maps for linkage dysequilibriummapping (23) and that inbreeding may increase the power to detectthe role of a disease susceptibility factor by comparison ofgenotypic distributions among case patients and controls (24).In Newfoundland, the presence of potential founder effects,as identified in other inherited disorders (2527), maybe associated with higher allele frequency of particular genotypesassociated with preeclampsia, thereby facilitating successfulgene mapping. This study suggests that in Newfoundland, familiesat high risk for preeclampsia can be identified, and that asearch for susceptibility genes for preeclampsia could be undertaken.
We conclude that the incidence of preeclampsia in primiparousNewfoundland women is 7.9%, that the incidence varies by region,and that primiparous sisters of probands with preeclampsia areat increased risk of developing preeclampsia, particularly whenprobands are defined by the severity of preeclampsia and byregion of domicile.
Acknowledgments
Funding was provided by the Janeway Foundation and MemorialUniversity. This study was supported by CIHR operating grantMOP 37778 (P.A.M.) and a CIHR Distinguished Scientist (RPP)award (P.S.P.).
Chesley LC, Cosgrove RA, Annitto JE: Pregnancy in the sisters and daughters of eclamptic women. Pathol Microbiol 24: 662666, 1961
Sutherland A, Cooper DW, Howie PW, Liston WA, MacGillivray I: The incidence of severe preeclampsia amongst mothers and mothers-in-law of pre-eclamptics and controls. Br J Obstet Gynecol 88: 785790, 1981[Medline]
Cooper DW, Liston WA: Genetic control of severe preeclampsia. J Med Genet 16: 409416, 1979[Abstract/Free Full Text]
Chesley LC, Cooper DW: Genetics of hypertension in pregnancy: Possible single gene control of preeclampsia and eclampsia in the descendants of eclamptic women. Br J Obstet Gynecol 93: 898908, 1986[Medline]
Arngrimsson R, Björnsson S, Geirsson RT, Björnsson H, Walker JJ, Snaedal G: Genetic and familial predisposition to eclampsia and preeclampsia in a defined population. Br J Obstet Gynecol 97: 762769, 1990[Medline]
Cooper DW, Brennecke SP, Wilton AN: Genetics of preeclampsia. Hypertens Pregnancy 12: 123, 1993
Cincotta RB, Brennecke SP: Family history of preeclampsia as a predictor for preeclampsia in primigravidas. Int J Gynecol Obstet 60: 2327, 1998[CrossRef][Medline]
Lie RT, Rasmussen S, Brunborg H, Gjessing HK, Lie-Nielsen E, Irgens LM: Fetal and maternal contributions to risk of preeclampsia: Population based study. Br Med J 316: 13431347, 1998[Abstract/Free Full Text]
Mogren I, Högberg U, Winkvist A, Stenlund H: Familial occurrence of preeclampsia. Epidemiology 10: 518522, 1999[Medline]
Morgan T, Ward K: New insights into the genetics of preeclampsia. Semin Perinatol 23: 1423, 1999[CrossRef][Medline]
Mannion JJ, ed. The Peopling of Newfoundland: Essays in Geography. Toronto: University of Toronto Press, 1986
Bear JC, Nemec TF, Kennedy JC, Marshall JC, Power AA, Lolonel VM, Burke GB: Persistent genetic isolation in outport Newfoundland. Am J Med Genet 27: 807830, 1987[CrossRef][Medline]
Bear JC, Nemec TF, Kennedy JC, Marshall WH, Power AA, Lolonel VM, Burke GB: Inbreeding in outport Newfoundland. Am J Med Genet 29: 649660, 1988[CrossRef][Medline]
Levine RJ, Hauth JC, Curet LB, Sibai BM, Catalano PM, Morris CD, DerSimonian R, Esterlitz JR, Raymond EG, Bild DE, Clemens JD, Cutler JA: Trial of calcium to prevent preeclampsia. N Engl J Med 337: 6976, 1997[Abstract/Free Full Text]
Helewa ME, Burrows RF, Smith J, Williams K, Brain P, Rabkui SW: Report of the Canadian Hypertension Society Consensus Conference: 1. Definitions, evaluation, and classification of hypertensive disorders in pregnancy. CMAJ 157: 715725, 1997[Abstract]
Salha O, Sharma V, Dada T, Nugent D, Rutherford AJ, Tomlinson AJ, Philips S, Allgar V, Walker JJ: The influence of donated gametes in the incidence of hypertensive disorders of pregnancy. Hum Reprod 14: 22682273, 1999[Abstract/Free Full Text]
Broughton PF: What is the place of genetics in the pathogenesis of preeclampsia? Biol Neonate 76: 325330, 1999[CrossRef][Medline]
Ros HS, Cnattingius S, Lipworth L: Comparison of risk factors for preeclampsia and gestational hypertension in a population-based cohort study. Am J Epidemiol 147: 10621070, 1998[Abstract/Free Full Text]
Wright AF, Carothers AD, Pirastu M: Population choice in mapping for complex diseases. Nat Genet 23: 397404, 1999[CrossRef][Medline]
Service SK, Lang DW, Freimer NB, Sandkuijl L: Linkage disequilibrium mapping of disease genes by reconstruction of ancestral haplotypes in founder populations. Am J Hum Genet 64: 17281738, 1999[CrossRef][Medline]
Genin E, Clerget-Darpoux F: Association studies in consanguineous populations. Am J Hum Genet 58: 861866, 1996[Medline]
Olufemi SE, Green JS, Manickam P, Guru SC, Agarwal SK, Kester MB, Dong Q, Burns AL, Spiegel AM, Marx SJ, Collins FS, Chandrasekharappa SC: Common Ancestral mutation in the MENI gene is likely responsible for the prolactinoma variant of MENI (MENI Burin) in four kindreds from Newfoundland. Hum Mutat 11: 264269, 1998[CrossRef][Medline]
Froggett NJ, Green J, Bassett C, Gareth RE, Bishop DT, Kolodner R, Maher ER: A common MSH2 mutation in English and North American HNPCC families: Origin, phenotypic expression, and sex specific differences in colorectal cancer. J Med Genet 36: 97102, 1999[Abstract/Free Full Text]
Young T-L, Woods MO, Parfrey PS, Green JS, Hefferton D, Davidson WS: A founder effect in the Newfoundland population reduced the Bardet-Biedl Syndrome I (BBS1) interval to 1 cM. Am J Hum Genet 65: 16801687, 1999[CrossRef][Medline]
Received for publication October 29, 2001.
Accepted for publication March 4, 2002.
This article has been cited by other articles:
C. B. Tempfer, S. Jirecek, E. K. Riener, H. Zeisler, D. Denschlag, L. Hefler, and P. W. Husslein Polymorphisms of Thrombophilic and Vasoactive Genes and Severe Preeclampsia: A Pilot Study
Reproductive Sciences,
May 1, 2004;
11(4):
227 - 231.
[Abstract][PDF]
R. Thadhani, W. P. Mutter, M. Wolf, R. J. Levine, R. N. Taylor, V. P. Sukhatme, J. Ecker, and S. A. Karumanchi First Trimester Placental Growth Factor and Soluble Fms-Like Tyrosine Kinase 1 and Risk for Preeclampsia
J. Clin. Endocrinol. Metab.,
February 1, 2004;
89(2):
770 - 775.
[Abstract][Full Text][PDF]