• Embed Doc
  • Copy Link
  • Readcast
  • Collections
  • CommentGo Back
Download
 
RESEARCH
Thevenousthromboticriskoforalcontraceptives,effectsof oestrogendose and progestogentype: results of the MEGAcase-control study
A van Hylckama Vlieg, research fellow,
1
F M Helmerhorst, professor of clinical epidemiology of fertility,
1,2
 J P Vandenbroucke, professor of clinical epidemiology,
1
C J M Doggen, research fellow,
1
F R Rosendaal,professor of clinical epidemiology, head of department
1,3,4
 ABSTRACT
Objective
To assess the thrombotic risk associated withoral contraceptive use with a focus on dose of oestrogenand type of progestogen of oral contraceptives availablein the Netherlands.
Design
Population based case-control study.
Setting 
Six participating anticoagulation clinics in theNetherlands(Amersfoort,Amsterdam,TheHague,Leiden,Rotterdam, and Utrecht).
Participants
Premenopausal women <50 years old whowerenotpregnant,notwithinfourweekspostpartum,andnot using a hormone excreting intrauterine device or depot contraceptive. Analysis included 1524 patientsand 1760 controls.
Main outcome measures
First objectively diagnosedepisodes of deep venous thrombosis of the leg or pulmonary embolism. Odds ratios calculated by cross-tabulation with a 95% confidence interval according toWoolf 
’
s method; adjusted odds ratios estimated byunconditionallogisticregression,standarderrorsderivedfrom the model.
Results
Currently available oral contraceptives increasedthe risk of venous thrombosis fivefold compared withnon-use (odds ratio 5.0, 95% CI 4.2 to 5.8). The riskclearly differed by type of progestogen and dose of oestrogen. The use of oral contraceptives containing levonorgestrel was associated with an almost fourfoldincreased risk of venous thrombosis (odds ratio 3.6, 2.9to 4.6) relative to non-users, whereas the risk of venousthrombosis compared with non-use was increased 5.6-fold for gestodene (5.6, 3.7 to 8.4), 7.3-fold for desogestrel (7.3, 5.3 to 10.0), 6.8-fold for cyproteroneacetate (6.8, 4.7 to 10.0), and 6.3-fold for drospirenone(6.3, 2.9 to 13.7). The risk of venous thrombosis waspositively associated with oestrogen dose. We confirmedahighriskofvenousthrombosisduringthefirstmonthsof oral contraceptive use irrespective of the type of oralcontraceptives.
Conclusions
Currently available oral contraceptives stillhaveamajorimpactonthrombosisoccurrenceandmanywomendonotusethesafestbrandswithregardtoriskof venous thrombosis.
INTRODUCTION
The first report of an increased risk of venous throm-bosis associated with oral contraceptives appeared in1961.
1
Sincethen,severallargestudieshaveconfirmeda twofold to sixfold increased risk of deep venousthrombosis associated with current oral contraceptiveuse.
2-5
To decrease the risk of thrombosis, the oestro-gendoseincombinedoralcontraceptiveswasstepwisereduced over the years. A lowering of the oestrogendose from 100
 
g to 50
 
g has been associated with a decreased risk of venous thrombosis.
6-8
There is noclear evidence that the lowering of the oestrogen doseto 30
 
g or 20
 
g led to a furtherdecrease of the risk of deep venous thrombosis.Oral contraceptives may contain different types of progestogens.Firstgenerationoralcontraceptivescon-tained lynestrenol, but these are now little used. Sec-ond generation oral contraceptives, which are widelyused, contain levonorgestrel or, less often, norgestrel.Thirdgenerationoralcontraceptives,containingdeso-gestrel or gestodene, which became available in the1980s, are also widely used. Two other types of oralcontraceptives are not included in this classification.Preparations containing cyproterone acetate are usedfortreatmentofacnevulgaris,seborrhoea,ormildhir-sutismandhaveanti-ovulatoryactionsimilartothatof a progestogen.
9-11
Preparations containing drospire-none, which is an antimineralocorticoid, also inhibit ovulation and have been on the market since2001.
1213
Since 1995, numerous reports have been availableon the difference in thrombotic risk associated withsecond and third generation oral contraceptives.
4714
Most reported an increased risk of venous thrombosisassociatedwiththenewerthirdgenerationoralcontra-ceptives. Some, however, did not confirm this finding or suggested that the risk difference between third andsecond generation oral contraceptives was overesti-mated because of bias or confounding such as referralor prescription bias.
515
Kemmeren et al performed a meta-analysisoncohortandcase-controlstudiesasses-sing the risk of venous thrombosis among women
1
Department of ClinicalEpidemiology, Leiden UniversityMedical Center, C7-P,PO Box 9600, NL-2300RC Leiden, Netherlands
2
Department of Gynaecology andReproductive Medicine, LeidenUniversity Medical Center
3
Department of Thrombosis andHaemostasis, Leiden UniversityMedical Center
4
Einthoven Laboratory forExperimental Vascular Medicine,Leiden University Medical Center
Correspondence to: F R RosendaalF.R.Rosendaal@lumc.nl
Cite this as:
BMJ 
2009;339:b2921
doi:10.1136/bmj.b2921BMJ |
ONLINE FIRST | bmj.com page 1 of 8
 
using oral contraceptives before 1995, and found a twofold increased risk of venous thrombosis for thirdgeneration oral contraceptives compared with oralcontraceptives containing levonorgestrel.
16
As severalauthors pointed out, none of the arguments that thisfindingwascausedbybiasstooduptosubsequentana-lyses or reasoning.
17-20
Oralcontraceptivescontainingcyproteroneacetate,on the market since 1988, have been associated with a highly increased risk of venous thrombosis in somestudies,
21-23
butnotall.
2425
Limitedinformationisavail-able for the thrombotic risk associated with the newest oralcontraceptivescontainingdrospirenone,butaser-ies of reported cases of venous thrombosis after itsintroduction raised concern about an increased riskassociated with this oral contraceptive.
2627
Two recent studies sponsored by the manufacturer, however,reported a similar thrombotic risk for drospirenonecompared with levonorgestrel.
2829
All these studiesincluded few cases of thrombosis and therefore hadconsiderable uncertainty around the risk estimates.Theaimofthepresentstudywastoassessthethrom-botic risk associated with current oral contraceptiveuse with a focus on dose of oestrogen and type of pro-gestogen. We set out to determine which hormonalcontraceptiveissafestwithregardtotheriskofvenousthrombosis using data from a large case-control studyof patients with a first deep venous thrombosis andhealthy controls.
METHODS
Study design
This analysis was performed using data from theMEGA study (multiple environmental and geneticassessment of risk factors for venous thrombosis-study), a large, population based, case-control studyon risk factors for venous thrombosis.BetweenMarch1999andSeptember2004,consecu-tivepatientsaged<70yearswithafirstepisodeofdeepvenous thrombosis (leg or arm) or pulmonary embo-lism were included from the files of six participating anticoagulation clinics in the Netherlands (Amers-foort, Amsterdam, The Hague, Leiden, Rotterdam,andUtrecht).Informationonthediagnosticprocedurewasobtainedfromhospitalrecordsandgeneralpracti-tioners.AdeepvenousthrombosiswasconfirmedwithDoppler ultrasonography. A pulmonary embolismwas confirmed by a ventilation perfusion lung scan,spiral computed tomography, or angiogram. Exclu-sion criteria were severe psychiatric problems and theinability to speak Dutch. Of the 6567 eligible patients,310 died soon after the venous thrombosis. Of theremaining 6257 patients, 5184 participated (83%). Of the non-participants, 86 were in the end stage of fataldisease and 987 refused to participate or could not belocated. Of the participants, 4752 (91.7%) returned a fullquestionnaireand431completedashortquestion-naire by telephone.Partners of patients <70 years old were invited toparticipate as controls. Of the 5184 participating patients, 3735 had an eligible partner. One partnerdied soon after the request for participation. Of theremaining 3734 partners, 3039 participated (81.4%).Of the non-participants, 20 were in the end stage of disease and 675 refused to participate or could not belocated.Afullquestionnairewasreturnedby2873par-ticipating partners (94.5%), and 164 completed a short questionnaire by phone.From January 2002 until September 2004, addi-tionalcontrolswererecruitedbyrandomdigitdialling.Phonenumbersweredialledatrandomwithinthegeo-graphical inclusion area of the patients. The randomcontrols were frequency matched to the patients withrespecttoageandsex.Onlycontrolsbetweentheagesof 18 and 70 years with no history of deep venousthrombosis were included and the same exclusion cri-teria were applied as for the patients.Of the 4350 eligible random controls, four diedbefore they were able to participate. Of the remaining 4346individuals,3000participated(69%).Ofthenon-participants, 15 were in the end stage of disease and1331 refused to participate or could not be located. Aquestionnairewasreturnedby2788(93%) oftheparti-cipating random controls.For the current analyses, only women aged18-50 years were included. Women who were post-menopausal, pregnant, or within 4 weeks postpartumat the time of the thrombotic event or index date (seebelow) and women using hormonal contraceptionother than oral contraceptives were excluded. Fromthe total patient group, we included 1524 femalepatients. From the partner control group, we included712 female partner controls. From the random digit dialling control group, we selected 1048 female con-trols. Of the total control group described in the cur-rent study (n
=
1760), 40.5% were partner controls and59.5%wererandomdigitdiallingcontrols.Fortheana-lyses described in this manuscript, we pooled the twocontrol groups into a single group and adjusted forinclusion date.
Data collection
All participants filled in a standardised questionnaireon risk factors for venous thrombosis such as familyhistory of thrombosis, pregnancy, and oral contracep-tive use in the year before the index date. The indexdate was the date of the thrombotic event for patientsand their partners and the date of filling in the ques-tionnaire for the random controls. The questionnairewas sent to all patients and their partners within a few weeks after the index date. For the random controls,thequestionnairewassentaftertheiragreementtopar-ticipate.At least three months after discontinuation of theoral anticoagulation therapy, patients and their part-ners were invited to the anticoagulation clinic for a blood sample and an interview. During the interviews(inpersonorbytelephone),detailsoncurrentoralcon-traceptive use were verified.
RESEARCH
page 2 of 8
BMJ |
ONLINE FIRST | bmj.com
 
Statistical analysis
Relative risks were assessed by calculating odds ratiosand 95% confidence intervals. Risk estimates wereadjustedbyunconditionallogisticregression,andcon-fidence intervals were derived from the model. Theoddsratios in theoverall analysisoftherisk associatedwith current oral contraceptive use were adjusted forage. When analysing the thrombotic risk associatedwithdoseofoestrogenortypeofprogestogen,anaddi-tional adjustment was made for date of inclusion(divided in a total of 12 periods of 6 calendar monthsspanning 1999-2004). For levonorgestrel, gestodene,desogestrel, and lynestrenol, preparations can containdifferentdosesofoestrogenandanalysiswasrestrictedtoonedose
 — 
themostfrequentlyusedof30
 
gethinyl-oestradiol for levonorgestrel, gestodene, and desoges-treland37.5
 
gethinyloestradiolforlynestrenol.Inallanalyses, we pooled the two control groups into onegroup and adjusted for inclusion date. For current oral contraceptive use, the risk of venous thrombosiswas calculated for all users of oral contraceptives, andseparatelyforthedifferenttypesoforalcontraceptives,compared with non-users (never users and past userscombined).Becauseapositivefamilyhistoryofvenousthrombosis has been hypothesised to lead to preferen-tial prescription of specific types of oral contraceptive,we took a positive family history into account. A posi-tive family history was defined as a participant having at least one parent or sibling with a history of venousthrombosisasreportedbythe participants.Body massindex (weight (kg)/(height (m)
2
)), also a potential con-founder in the association between different types of oral contraceptives and the risk of venous thrombosis,wascalculatedusingweightandheightasstatedbytheparticipantsinthequestionnaire.Smokingwasdefinedascurrentsmoking(comparedwithneversmokersandpast smokers combined).All oral contraceptives were classified according tothe dose of oestrogen and the type of progestogen.Comparisons were made relative to non-users, andbetween users of oral contraceptives containing differ-ent oestrogen doses and progestogen types. Becausesomecombinationsofoestrogendoseandprogestogentypedonotexistorwerenotusedinthisstudypopula-tion, we also performed stratified analyses, such as a comparison between several types of progestogenswith the same dose of oestrogen (30
 
g ethinylestra-diol) and the effect of oestrogen dose stratified for thedifferent types of progestogens.
RESULTS
Table 1 shows the general characteristics of the studypopulation. Of all 1524 patients, 859 (56.4%) had a deep venous thrombosis of the leg, 495 (32.5%) had a pulmonary embolism, 111 (7.3%) had both, and 59(3.9%) had a deep venous thrombosis of the arm. Themean age of the 1524 patients was 37.1 years, ranging from 18 to 49. The mean age of the 1760 controls wassimilar to that of patients
 — 
namely, 37.4 years (range18-49). Among the controls, women who were using hormonal contraceptives (mean age 33.6) were about sixyearsyoungerthanwomenwhodidnotusehormo-nal contraceptives (mean age 39.6, mean difference6.0, 95% CI 5.3 to 6.8).Of the 1524 patients, 1103 (72.4%) were using oralcontraceptives at the time of thrombosis, comparedwith658/1760(37.4%)ofthecontrols.Thepercentageof women using oral contraceptives was higher inyounger women than in older women (table 1).Overall, current oral contraceptive use was asso-ciated with a fivefold increased risk of venous throm-bosis (odds ratio 5.0, 95% CI 4.2 to 5.8). Additionaladjustment for smoking and body mass index resultedin a relative risk of 5.4 (95% CI 4.5 to 6.4). Restrictionto individuals without a positive family history of venous thrombosis resulted in a relative risk of 5.8(4.7 to 7.2) for all oral contraceptive users combinedversus non-users.Toshowtheabsoluteeffectoforalcontraceptiveuseamong women in different age categories, we esti-mated the incidence of venous thrombosis in womennot using oral contraceptives aged <30 years,30-40years,and40-50yearsold.Theoverallincidenceofvenousthrombosisperagecategory(I)canbecalcu-lated as I
=
(p
0
×
I
0
)+(p
1
×
I
1
), relative risk
=
I
1
/I
0
, and
Table 1
|
General characteristics of study population. Data are number (%) unless otherwiseindicated
Thrombosis patients (n
=
1524) Controls (n
=
1760)
Deep venousthrombosis oflegonly 859 (56.4)
— 
Pulmonary embolismonly 495 (32.5)
— 
Deep venousthrombosis ofleg+ pulmonaryembolism111 (7.3)
— 
Deep venousthrombosis ofarm 59 (3.9)
— 
Mean (range)age (years) 37.1 (18-49) 37.4(18-49)Oralcontraceptive use 1103 (72.4) 658 (37.4)Oralcontraceptive use by age:<30 years 327 (87.2) 266 (68.0)30-40 years 346 (73.0) 210 (35.5)40-50 years 430 (63.7) 182 (23.4)Mean (range)body mass index* 26.8 (16.0-57.8) 24.4(15.7-50.7)Positive family history
†
309 (26.4) 186 (14.3)Smoking 
‡
380 (26.8) 500 (30.9)
*Data available for 1391 (91.3%) patients and 1591 (90.4%) controls.
†
Defined as having one or more parents or siblings who experienced deep venous thrombosis or pulmonaryembolism; information about family history of venous thrombosis available for 1171 (76.8%) patients and 1298(73.8%) controls.
‡
Defined as current smoking at index date; information on smoking available for 1418 patients (93.0%) and1618 controls (91.9%).
Table 2
|
 Absolute risk of venous thrombosis associated with oral contraceptive use by agecategory
 Age categoryIncidence of venousthrombosis in non-users of oral contraceptives (I
0
 ) per 10 000 person-years*Relative risk (95% CI) of oral contraceptive use
†
Incidence of venousthrombosis in oralcontraceptive users (I
1
 ) per 10 000 person-years
‡
<30 years 1.2 3.1 (2.2to 4.6) 3.730-40years 2.0 5.0 (3.8to 6.5) 10.040-50years 2.3 5.8 (4.6to 7.3) 13.3
*I
0
is based on incidences published by Naess et al.
30
†
Non-users of oral contraceptives are used as the reference category.
‡
I
1
=
I
0
×
relative risk.
RESEARCH
BMJ |
ONLINE FIRST | bmj.com page 3 of 8
of 00

Leave a Comment

You must be to leave a comment.
Submit
Characters: ...
You must be to leave a comment.
Submit
Characters: ...