- Email: [email protected]
Maternal characteristics largely explain poor pregnancy outcome after hyperemesis gravidarum
European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Maternal characteristics largely explain poor pregnancy outcome after hyperemesis gravidarum Tessa J. Roseboom a,b,*, Anita C.J. Ravelli c, Joris A. van der Post b, Rebecca C. Painter b a
Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Academic Medical Centre, Amsterdam, The Netherlands Department of Obstetrics and Gynecology, Academic Medical Centre, Amsterdam, The Netherlands c Department of Medical Informatics, Academic Medical Centre, Amsterdam, The Netherlands b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 26 October 2010 Received in revised form 14 December 2010 Accepted 7 January 2011
Objective: To describe the characteristics of women who suffer from hyperemesis gravidarum, and explore the independent effect of hyperemesis gravidarum on pregnancy outcome. Study design: In The Netherlands Perinatal Registry, we used all data on singleton pregnancies of at least 24 weeks and 500 g without congenital anomalies in the years 2000–2006. We examined the characteristics of women who suffered from hyperemesis gravidarum and their children. Results: Women who suffered from hyperemesis gravidarum were slightly younger; more often primiparous, of lower socio-economic status, of non-Western descent and substance abusers; had more often conceived through assisted reproduction techniques and more often had pre-existing hypertension, diabetes mellitus and psychiatric diseases than women who did not suffer from hyperemesis gravidarum. Also, their pregnancies were more often complicated by hypertension and diabetes and they more often carried a female fetus. Pregnancies complicated by hyperemesis gravidarum significantly more often had an adverse outcome (prematurity or birth weight below the 10th percentile). The increased risk of adverse pregnancy outcomes after hyperemesis gravidarum was largely explained by the differences in maternal characteristics (crude OR 1.22 (95% CI 1.10–1.36), adjusted OR was 1.07 (95% CI 0.95–1.19)). Conclusion: Hyperemesis gravidarum is associated with adverse pregnancy outcomes. This is largely explained by differences in maternal characteristics. Given the impact of the early environment on later health (which is independent of size at birth), studies that aim to assess the long-term consequences of hyperemesis gravidarum need to be given high priority. ß 2011 Elsevier Ireland Ltd. All rights reserved.
Keywords: Pregnancy Hyperemesis gravidarum Prematurity Birth weight
1. Introduction Nausea and vomiting in pregnancy is the most common pregnancy complication, affecting more than half of all women during the first trimester of pregnancy [1]. A more severe form of nausea and vomiting, known as hyperemesis gravidarum, is much less common, occurring in 0.5–3% of pregnancies [2–4] but is the most common cause of hospitalization in the first half of pregnancy and is the second only to preterm labour for pregnancy overall [5]. Both the aetiology and pathogenesis of hyperemesis gravidarum remain unknown [6,7]. Hyperemesis gravidarum generally occurs more often among young women in their first pregnancy, and many [2,4,8–10], but
not all [7], studies found that babies born after hyperemetic pregnancies had lower birth weights. Young and primiparous women are known to have lighter babies, but none of the previous studies has taken this into account when investigating the outcome of pregnancies complicated by hyperemesis gravidarum. Therefore this study aims to investigate the independent contribution of hyperemesis gravidarum on pregnancy outcome. We describe the characteristics of women who suffer from hyperemesis gravidarum, as well as the neonatal outcomes of their babies, and investigate the independent role of hyperemesis gravidarum on pregnancy outcome. 2. Materials and methods 2.1. Netherlands Perinatal Registry
* Corresponding author at: Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Department of Obstetrics and Gynecology, Academic Medical Centre, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. Tel.: +31 20 5666934; fax: +31 20 6912683. E-mail address: [email protected] (T.J. Roseboom). 0301-2115/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2011.01.010
This was a population-based retrospective cohort study of all singleton deliveries after 24 weeks of gestation or more and a birth weight above 500 g in The Netherlands, between 2000 and 2006. Pregnancies with congenital anomalies (2.3%) were excluded. All
T.J. Roseboom et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
information for this study was obtained from The Netherlands Perinatal Registry [11–15]. This registry contains population-based information on pregnancies, deliveries and (re)admissions of newborns. It is based on a validated probabilistic linkage of the national midwifery registry (LVR1), the obstetrics registry (LVR2) and the neonatology/paediatrics registry (LNR). The midwifery and obstetrics registries start at the first (booking) prenatal visit. If the child is admitted to a neonatal ward after birth, the information is collected in the neonatology registry. The registry contains information about the mother (demographic information, lifestyle, medical conditions), the pregnancy (the mode of conception, the course of pregnancy) delivery, and the baby, and covers 96% of all deliveries in The Netherlands. 2.2. Co-variates We defined hyperemetic pregnancies as those pregnancies that were labeled during any time of pregnancy as complicated by hyperemesis gravidarum by the caregiver (either the midwife or the gynecologist). Information on parity, ethnicity, hypertension (pre-existing or pregnancy-induced), diabetes (pre-existing or pregnancy-related), thrombo-embolic disease, psychiatric disease (depression requiring medication, bipolar depression and all psychotic disorders), substance abuse (daily use of cannabis/opiates/benzodiazepines, and frequent use of hard drugs such as cocaine and heroin), and mode of conception (natural, IVF or other assisted reproduction techniques) all came from the medical records. The urbanization degree and socioeconomic status were based on the four-digit postcode of the pregnant woman. Urbanization was based on the number of addresses per square kilometre and categorised as very urban (2500 addresses/km2), intermediate urban/rural (between 500 and 2500 addresses/km2) and very rural (<500 addresses/km2). The socio-economic status was categorised according to average price of housing based on postcode using data from The Netherlands Institute for Social Research in three categories (10th, 10–90, and 90th percentile). 2.3. Statistical methods We compared the characteristics of women with hyperemesis gravidarum to those of women without hyperemesis using a Chi square test or a t-test. Secondly, we compared characteristics of the babies born to these women. Analyses were repeated for the severe cases (the women with hyperemesis gravidarum who were hospitalized during gestation). Subsequently, we performed linear and logistic regression analyses to adjust for maternal and fetal characteristics and year of birth. We examined the association between hyperemesis gravidarum and pregnancy outcome for the total population and for the liveborns only. 3. Results In the period 2000–2006, of the 1,199,218 singleton deliveries after more than 24 weeks of gestation in The Netherlands, 2190 (0.2%) were complicated by hyperemesis gravidarum (753 as indicated by the midwife without hospital admission) and 1437 by a gynecologist as requiring hospital admission (including 607 referrals from midwives). Table 1 shows the maternal characteristics of the 1,197,028 women without hyperemesis, and those of the 2190 with hyperemesis. Women who suffered from hyperemesis gravidarum were slightly younger, more often primiparous, more often of nonWestern origin, had a lower socio-economic status, and had more often conceived through assisted reproduction techniques. Also, they reported substance abuse more often and more often had a
57
Table 1 Demographic characteristics of women with and without hyperemesis gravidarum.
N Age (SD) Primiparous Non-Western ethnicity Low socio-economic status (10th percentile) Living in very urban areas Pre-existing hypertension Hypertension (pre-existing or pregnancy-related) Diabetes mellitus (type 1 and type 2) Diabetes mellitus and gestational diabetes Thrombo-embolic diseases Psychiatric illness Substance abuse during pregnancy Assisted reproduction Female gender child
Hyperemesis
No hyperemesis
2190 29.4 (4.8) 49.6% 28.7% 13.2%
1,197,028 30.4 (4.8) 46.3% 16.0% 9.9%
P value
<0.0001 <0.001 <0.0001 <0.0001
20.5% 0.87% 15.6%
19.5% 0.32% 8.5%
0.26 <0.0001 <0.0001
1.0%
0.5%
0.0005
1.8%
0.9%
<0.0001
0.3% 2.1% 0.3%
0.2% 0.3% 0.1%
0.36 <0.0001 <0.0005
2.4% 53.7%
1.7% 48.8%
<0.01 <0.0001
psychiatric illness, pre-existing hypertension and diabetes mellitus. They more often carried a female fetus. Table 2 shows the perinatal outcomes of pregnancies complicated by hyperemesis, and those without hyperemesis. Adverse pregnancy outcomes were more prevalent among women who had suffered from hyperemesis gravidarum. Women who suffered from hyperemesis gravidarum more often delivered prematurely and more often had a baby that was small for gestational age (<10th percentile). Perinatal mortality and NICU admissions did not differ. The maternal characteristics associated with hyperemesis (age, parity, ethnicity, mode of conception, socio-economic status, hypertension, diabetes, psychiatric illness and drugs) are all known to be associated with adverse perinatal outcomes. Therefore, we investigated whether the association between hyperemesis gravidarum and adverse pregnancy outcome was attenuated by
Table 2 Pregnancy outcomes of hyperemetic and non-hyperemetic pregnancies.
N Fetal mortality Perinatal mortality Birth weight in grams (SD) Premature (<37 weeks) Weight below 10th percentile Adverse fetal outcomea incl mortality Liveborn babies only N Neonatal mortality Birth weight Premature (<37 weeks) Weight below 10th percentile NICU admission Apgar <7 Adverse fetal outcomea incl mortality
Hyperemesis
No hyperemesis
P value
2190 0.2% 0.3% 3337 (587) 7.6% 10.8% 17.9%
1,197,028 0.5% 0.6% 3453 (579) 5.7% 9.8% 15.1%
0.06 0.10 <0.0001 <0.0001 0.11 0.0003
2186 0.14% 3337 (587) 7.6% 10.8%
1,191,519 0.13% 3459 (569) 5.5% 9.7%
0.96 <0.0001 <0.0001 0.09
1.2% 2.3% 3.2%
0.9% 2.4% 3.0%
0.20 0.77 0.68
Fetal mortality, death during gestation and during labour divided by 1000 total births; neonatal mortality, death within first week after birth divided by 1000 live births. a Adverse fetal outcome, prematurity (<37 weeks) or birth weight below p10 and/or perinatal mortality.
58
T.J. Roseboom et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
Table 3 Crude and adjusted odds ratios of hyperemesis gravidarum on different outcomes. Outcome
Perinatal mortality Premature birth SGA 10th percentile Adverse fetal outcomea
Crude
Adjusted
Odds ratio
95% CI
Odds ratio
95% CI
0.54 1.37 1.11 1.22
0.26–1.13 1.17–1.60 0.97–1.27 1.10–1.36
0.46 1.18 0.98 1.07
0.22–0.96 1.01–1.38 0.85–1.12 0.95–1.19
Adjusted for parity, maternal age, socio-economic status, non-Western ethnicity, pre-existing hypertension, diabetes mellitus, psychiatric disorders, substance abuse, assisted reproduction, urbanization, hypertensive disorders, diabetes gravidarum, female child gender and year. a Adverse fetal outcome, perinatal mortality and/or SGA and/or premature birth.
adjustment for these characteristics. Table 3 shows that the crude association between hyperemesis gravidarum and adverse outcome (premature birth, weight <10th percentile and/or perinatal mortality) had an OR of 1.22 (95% CI 1.10–1.36). After adjustment for maternal characteristics (age, parity, socio-economic status, ethnicity, mode of conception, urbanization, substance abuse, hypertension, diabetes and psychiatric disease and sex of the baby), the OR was 1.07 (95% CI 0.95–1.19). 4. Comment Hyperemesis gravidarum is associated with adverse pregnancy outcomes. Women who suffer from hyperemesis gravidarum more often deliver prematurely, and more often have children who are small for gestational age. This is largely explained by differences in age, parity, ethnicity, socio-economic status, substance abuse, and the prevalence of hypertension, diabetes and psychiatric illnesses. The main strength of this study is the large sample size and its representativeness. With 96% coverage and over a million pregnancies included, it is – to our knowledge – the largest cohort study on hyperemesis gravidarum. Because we used information from a national registry, its weakness is the quality of the data. The presence or absence of hyperemesis may have been underreported or selectively reported since it is not an obligatory field in the registry. The relatively low prevalence of hyperemesis compared to other European studies is consistent with this [2– 4]. This will have weakened the associations found, and may have led to an underestimation of the effects. The estimate, however, of the reduction of birth weight after hyperemesis gravidarum found in our study is similar to that found in other studies [4,5]. Inevitably, there has been under-reporting of some risk factors like smoking and the diseases in the obstetric history, which will have weakened the associations, and did not allow us to fully adjust for these effects. Also, the absence of information on weight gain in the registry prevented us from examining whether the effects of hyperemesis are mediated through reduced weight gain. We defined hyperemetic pregnancies as those pregnancies that were labeled during any time of pregnancy as complicated by hyperemesis gravidarum by either the midwife or the gynecologist. One-third of the cases were labeled as hyperemesis gravidarum by the midwife, but did not require hospital admission, one-third of the cases were referred by the midwife to the gynecologist and admitted to hospital, and one-third were labeled as hyperemesis gravidarum by the gynecologist, and consequently admitted. The pregnancies that were labeled as complicated by hyperemesis gravidarum but that did not require hospital admissions may have been less severe. Analyses of only the pregnancies that did require hospital admission for hyperemesis gravidarum, however, showed similar results (results not shown). The finding that women who suffer from hyperemesis gravidarum are different from women who do not suffer from it are in agreement with other reports in the literature [3,6,7,10], although the associations with substance abuse and assisted
reproduction had not been reported before. The association between hyperemesis gravidarum and prematurity and reduced birth weights has been reported previously [2,4,7–10], but what this study adds is that this association is largely dependent on the various maternal characteristics associated with hyperemesis gravidarum. Although it may seem reassuring that hyperemesis gravidarum does not have an independent negative effect on pregnancy outcome, this may be misleading. There is a large body of evidence suggesting that the environment in utero has lasting consequences for health in later life, and that this may be independent of fetal growth. We have shown that maternal under nutrition especially in early gestation is associated with increased risks of chronic degenerative diseases such as type 2 diabetes, cardiovascular disease and breast-cancer [16–19]. The effects of maternal under-nutrition during gestation on later health were present even without effects on maternal weight gain and size of the baby at birth. Animal studies have shown that under-nutrition during gestation has lasting adverse consequences for health and lifespan [20]. Given the fact that hyperemesis gravidarum is associated with metabolic changes in the mother which are similar to those seen in starvation, it is surprising that the long term consequences of hyperemesis gravidarum are almost entirely unstudied. To our knowledge there is only one report of an increased risk of testicular cancer among boys born to mothers who suffered from hyperemesis gravidarum which was hypothesized to be due to the increased levels of estrogens at the time of differentiation of the testis [21]. The effects of severe morning sickness on the offspring’s markers of cardiovascular risk have, however, never been studied. Follow-up studies in large population-based cohorts may provide the first evidence of an increase of cardiovascular risk markers among the children of mothers who suffered severe hyperemesis gravidarum, and should be given high priority. A randomized controlled trial investigating the effects of nasogastric feeding strategies for hyperemesis gravidarum on short term outcomes, including improvement of nausea and vomiting, duration of hospitalization and birth weight, has never been conducted [22]. The follow-up of the cardiovascular risk profile of the offspring from a trial of this kind may resolve whether nutritional support may lower cardiovascular risk in the offspring. Hyperemesis gravidarum is associated with adverse pregnancy outcomes. Notwithstanding the fact that this is largely explained by differences in maternal characteristics, and given the impact of the early environment for later health, studies that aim to assess the long-term consequences of hyperemesis gravidarum need to be given high priority. Acknowledgements The Netherlands Perinatal Registry has given permission for the analysis of the data described in this paper. This permission is gratefully acknowledged.
T.J. Roseboom et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
References [1] Arsenault MY, Lane CA, MacKinnon CJ, et al. The management of nausea and vomiting of pregnancy. J Obstet Gynaecol Can 2002;24:817–31. [2] Bashiri A, Neumann L, Maymon E, Katz M. Hyperemesis gravidarum: epidemiologic features, complications and outcome. Eur J Obstet Gynecol Reprod Biol 1995;63:135–8. [3] Vikanes A, Grjibovski AM, Vangen S, Magnus P. Variations in prevalence of hyperemesis gravidarum by country of birth: a study of 900074 pregnancies in Norway, 1967–2005. Scand J Public Health 2008;36:135–42. [4] Bailit JL. Hyperemesis gravidarum: epidemiologic findings from a large cohort. Am J Obstet Gynaecol 2005;193:811–4. [5] Gazmararian JA, Petersen R, Jamieson DJ, et al. Hospitalizations during pregnancy among managed care enrollees. Obstet Gynaecol 2002;100:94–100. [6] Fell DB, Dodds L, Joseph KS, Allen WM, Butler B. Risk factors for hyperemesis gravidarum requiring hospital admission during pregnancy. Obstet Gynaecol 2006;107:277–84. [7] Verberg MF, Gillott DJ, Al-Fardan N, Grudzinskas JG. Hyperemesis gravidarum, a literature review. Hum Reprod Update 2005;11(September–October (5)): 527–39. [8] Dodds L, Fell DB, Joseph KS. Outcome of pregnancies complicated by hyperemesis gravidarum. Obstet Gynaecol 2006;107:285–92. [9] Tan PC, Jacob R, Quek KF, Omar SZ. Pregnancy outcome in hyperemesis gravidarum and the effect of laboratory clinical indicators of hyperemesis severity. J Obstet Gynaecol Res 2007;33:457–64. [10] Depue RH, Bernstein L, Loss RK, Judd HL, Henderson BE. Hyperemesis gravidarum in relation to estradiol levels, pregnancy outcome, and other maternal factors: a seroepidemiologic study. Am J Obstet Gynaecol 1987;156:1137–41. [11] Stichting Perinatale Registratie Nederland. Perinatal care in The Netherlands 2006. Utrecht: Stichting Perinatale Registratie Nederland; 2008.
59
[12] Meray N, Reitsma JB, Ravelli AC, Bonsel GJ. Probabilistic record linkage is a valid and transparent tool to combine databases without a patient identification number. J Clin Epidemiol 2007;60(9):883–91. [13] Tromp M, Ravelli AC, Meray N, Reitsma JB, Bonsel GJ. An efficient validation method of probabilistic record linkage including readmissions and twins. Methods Inf Med 2008;47(4):356–63. [14] Ravelli ACJ, Tromp M, Eskes M, van Huis AM, Steegers EAP, Tamminga P, Bonsel GJ. Perinatal mortality in The Netherlands 2000–2006; risk factors and risk selection. Ned Tijdschr Geneeskd 2008;152(50):2728–33. [15] Ravelli ACJ, Tromp M, Eskes M, et al. Decreasing perinatal mortality in The Netherlands, 2000–2006: a record linkage study. JECH 2009. [16] Ravelli AC, Van Der Meulen JH, Michels RP, et al. Glucose tolerance in adults after prenatal exposure to famine. Lancet 1998;351(9097):173–7. [17] de Rooij SR, Painter RC, Roseboom TJ, Phillips DI, Osmond C, Barker DJP, Tanck M, Michels R, Bossuyt P, Bleker OP. Glucose tolerance at age 58 and the decline of glucose tolerance in comparison with age 50 in people prenatally exposed to the Dutch famine. Diabetologia 2006;49(4):637–43. [18] Painter RC, de Rooij SR, Roseboom TJ, Bossuyt P, Simmers TA, Osmond C, Barker DJP, Bleker OP. Early onset of coronary artery disease after prenatal exposure to the Dutch famine. Am J Clin Nutr 2006;84(2):322–7. [19] Painter RC, de Rooij SR, Bossuyt PM, Osmond C, Barker DJP, Bleker OP, Roseboom TJ. A possible link between prenatal exposure to famine and breast cancer – a preliminary study. Am J Hum Biol 2006;18:853–6. [20] Ozanne SE, Hales CN. Lifespan: catch-up and obesity in male mice. Nature 2004;427:411–2. [21] Henderson BE, Benton B, Jing J, Yu MC, Pike MC. Risk factors for cancer of the testis in young men. Int J Cancer 1979;15:598–602. [22] Jewell D, Young G. Interventions for nausea and vomiting in early pregnancy. Cochrane Database Syst Rev 2003;4. CD000145.
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Maternal characteristics largely explain poor pregnancy outcome after hyperemesis gravidarum Tessa J. Roseboom a,b,*, Anita C.J. Ravelli c, Joris A. van der Post b, Rebecca C. Painter b a
Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Academic Medical Centre, Amsterdam, The Netherlands Department of Obstetrics and Gynecology, Academic Medical Centre, Amsterdam, The Netherlands c Department of Medical Informatics, Academic Medical Centre, Amsterdam, The Netherlands b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 26 October 2010 Received in revised form 14 December 2010 Accepted 7 January 2011
Objective: To describe the characteristics of women who suffer from hyperemesis gravidarum, and explore the independent effect of hyperemesis gravidarum on pregnancy outcome. Study design: In The Netherlands Perinatal Registry, we used all data on singleton pregnancies of at least 24 weeks and 500 g without congenital anomalies in the years 2000–2006. We examined the characteristics of women who suffered from hyperemesis gravidarum and their children. Results: Women who suffered from hyperemesis gravidarum were slightly younger; more often primiparous, of lower socio-economic status, of non-Western descent and substance abusers; had more often conceived through assisted reproduction techniques and more often had pre-existing hypertension, diabetes mellitus and psychiatric diseases than women who did not suffer from hyperemesis gravidarum. Also, their pregnancies were more often complicated by hypertension and diabetes and they more often carried a female fetus. Pregnancies complicated by hyperemesis gravidarum significantly more often had an adverse outcome (prematurity or birth weight below the 10th percentile). The increased risk of adverse pregnancy outcomes after hyperemesis gravidarum was largely explained by the differences in maternal characteristics (crude OR 1.22 (95% CI 1.10–1.36), adjusted OR was 1.07 (95% CI 0.95–1.19)). Conclusion: Hyperemesis gravidarum is associated with adverse pregnancy outcomes. This is largely explained by differences in maternal characteristics. Given the impact of the early environment on later health (which is independent of size at birth), studies that aim to assess the long-term consequences of hyperemesis gravidarum need to be given high priority. ß 2011 Elsevier Ireland Ltd. All rights reserved.
Keywords: Pregnancy Hyperemesis gravidarum Prematurity Birth weight
1. Introduction Nausea and vomiting in pregnancy is the most common pregnancy complication, affecting more than half of all women during the first trimester of pregnancy [1]. A more severe form of nausea and vomiting, known as hyperemesis gravidarum, is much less common, occurring in 0.5–3% of pregnancies [2–4] but is the most common cause of hospitalization in the first half of pregnancy and is the second only to preterm labour for pregnancy overall [5]. Both the aetiology and pathogenesis of hyperemesis gravidarum remain unknown [6,7]. Hyperemesis gravidarum generally occurs more often among young women in their first pregnancy, and many [2,4,8–10], but
not all [7], studies found that babies born after hyperemetic pregnancies had lower birth weights. Young and primiparous women are known to have lighter babies, but none of the previous studies has taken this into account when investigating the outcome of pregnancies complicated by hyperemesis gravidarum. Therefore this study aims to investigate the independent contribution of hyperemesis gravidarum on pregnancy outcome. We describe the characteristics of women who suffer from hyperemesis gravidarum, as well as the neonatal outcomes of their babies, and investigate the independent role of hyperemesis gravidarum on pregnancy outcome. 2. Materials and methods 2.1. Netherlands Perinatal Registry
* Corresponding author at: Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Department of Obstetrics and Gynecology, Academic Medical Centre, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. Tel.: +31 20 5666934; fax: +31 20 6912683. E-mail address: [email protected] (T.J. Roseboom). 0301-2115/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2011.01.010
This was a population-based retrospective cohort study of all singleton deliveries after 24 weeks of gestation or more and a birth weight above 500 g in The Netherlands, between 2000 and 2006. Pregnancies with congenital anomalies (2.3%) were excluded. All
T.J. Roseboom et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
information for this study was obtained from The Netherlands Perinatal Registry [11–15]. This registry contains population-based information on pregnancies, deliveries and (re)admissions of newborns. It is based on a validated probabilistic linkage of the national midwifery registry (LVR1), the obstetrics registry (LVR2) and the neonatology/paediatrics registry (LNR). The midwifery and obstetrics registries start at the first (booking) prenatal visit. If the child is admitted to a neonatal ward after birth, the information is collected in the neonatology registry. The registry contains information about the mother (demographic information, lifestyle, medical conditions), the pregnancy (the mode of conception, the course of pregnancy) delivery, and the baby, and covers 96% of all deliveries in The Netherlands. 2.2. Co-variates We defined hyperemetic pregnancies as those pregnancies that were labeled during any time of pregnancy as complicated by hyperemesis gravidarum by the caregiver (either the midwife or the gynecologist). Information on parity, ethnicity, hypertension (pre-existing or pregnancy-induced), diabetes (pre-existing or pregnancy-related), thrombo-embolic disease, psychiatric disease (depression requiring medication, bipolar depression and all psychotic disorders), substance abuse (daily use of cannabis/opiates/benzodiazepines, and frequent use of hard drugs such as cocaine and heroin), and mode of conception (natural, IVF or other assisted reproduction techniques) all came from the medical records. The urbanization degree and socioeconomic status were based on the four-digit postcode of the pregnant woman. Urbanization was based on the number of addresses per square kilometre and categorised as very urban (2500 addresses/km2), intermediate urban/rural (between 500 and 2500 addresses/km2) and very rural (<500 addresses/km2). The socio-economic status was categorised according to average price of housing based on postcode using data from The Netherlands Institute for Social Research in three categories (10th, 10–90, and 90th percentile). 2.3. Statistical methods We compared the characteristics of women with hyperemesis gravidarum to those of women without hyperemesis using a Chi square test or a t-test. Secondly, we compared characteristics of the babies born to these women. Analyses were repeated for the severe cases (the women with hyperemesis gravidarum who were hospitalized during gestation). Subsequently, we performed linear and logistic regression analyses to adjust for maternal and fetal characteristics and year of birth. We examined the association between hyperemesis gravidarum and pregnancy outcome for the total population and for the liveborns only. 3. Results In the period 2000–2006, of the 1,199,218 singleton deliveries after more than 24 weeks of gestation in The Netherlands, 2190 (0.2%) were complicated by hyperemesis gravidarum (753 as indicated by the midwife without hospital admission) and 1437 by a gynecologist as requiring hospital admission (including 607 referrals from midwives). Table 1 shows the maternal characteristics of the 1,197,028 women without hyperemesis, and those of the 2190 with hyperemesis. Women who suffered from hyperemesis gravidarum were slightly younger, more often primiparous, more often of nonWestern origin, had a lower socio-economic status, and had more often conceived through assisted reproduction techniques. Also, they reported substance abuse more often and more often had a
57
Table 1 Demographic characteristics of women with and without hyperemesis gravidarum.
N Age (SD) Primiparous Non-Western ethnicity Low socio-economic status (10th percentile) Living in very urban areas Pre-existing hypertension Hypertension (pre-existing or pregnancy-related) Diabetes mellitus (type 1 and type 2) Diabetes mellitus and gestational diabetes Thrombo-embolic diseases Psychiatric illness Substance abuse during pregnancy Assisted reproduction Female gender child
Hyperemesis
No hyperemesis
2190 29.4 (4.8) 49.6% 28.7% 13.2%
1,197,028 30.4 (4.8) 46.3% 16.0% 9.9%
P value
<0.0001 <0.001 <0.0001 <0.0001
20.5% 0.87% 15.6%
19.5% 0.32% 8.5%
0.26 <0.0001 <0.0001
1.0%
0.5%
0.0005
1.8%
0.9%
<0.0001
0.3% 2.1% 0.3%
0.2% 0.3% 0.1%
0.36 <0.0001 <0.0005
2.4% 53.7%
1.7% 48.8%
<0.01 <0.0001
psychiatric illness, pre-existing hypertension and diabetes mellitus. They more often carried a female fetus. Table 2 shows the perinatal outcomes of pregnancies complicated by hyperemesis, and those without hyperemesis. Adverse pregnancy outcomes were more prevalent among women who had suffered from hyperemesis gravidarum. Women who suffered from hyperemesis gravidarum more often delivered prematurely and more often had a baby that was small for gestational age (<10th percentile). Perinatal mortality and NICU admissions did not differ. The maternal characteristics associated with hyperemesis (age, parity, ethnicity, mode of conception, socio-economic status, hypertension, diabetes, psychiatric illness and drugs) are all known to be associated with adverse perinatal outcomes. Therefore, we investigated whether the association between hyperemesis gravidarum and adverse pregnancy outcome was attenuated by
Table 2 Pregnancy outcomes of hyperemetic and non-hyperemetic pregnancies.
N Fetal mortality Perinatal mortality Birth weight in grams (SD) Premature (<37 weeks) Weight below 10th percentile Adverse fetal outcomea incl mortality Liveborn babies only N Neonatal mortality Birth weight Premature (<37 weeks) Weight below 10th percentile NICU admission Apgar <7 Adverse fetal outcomea incl mortality
Hyperemesis
No hyperemesis
P value
2190 0.2% 0.3% 3337 (587) 7.6% 10.8% 17.9%
1,197,028 0.5% 0.6% 3453 (579) 5.7% 9.8% 15.1%
0.06 0.10 <0.0001 <0.0001 0.11 0.0003
2186 0.14% 3337 (587) 7.6% 10.8%
1,191,519 0.13% 3459 (569) 5.5% 9.7%
0.96 <0.0001 <0.0001 0.09
1.2% 2.3% 3.2%
0.9% 2.4% 3.0%
0.20 0.77 0.68
Fetal mortality, death during gestation and during labour divided by 1000 total births; neonatal mortality, death within first week after birth divided by 1000 live births. a Adverse fetal outcome, prematurity (<37 weeks) or birth weight below p10 and/or perinatal mortality.
58
T.J. Roseboom et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
Table 3 Crude and adjusted odds ratios of hyperemesis gravidarum on different outcomes. Outcome
Perinatal mortality Premature birth SGA 10th percentile Adverse fetal outcomea
Crude
Adjusted
Odds ratio
95% CI
Odds ratio
95% CI
0.54 1.37 1.11 1.22
0.26–1.13 1.17–1.60 0.97–1.27 1.10–1.36
0.46 1.18 0.98 1.07
0.22–0.96 1.01–1.38 0.85–1.12 0.95–1.19
Adjusted for parity, maternal age, socio-economic status, non-Western ethnicity, pre-existing hypertension, diabetes mellitus, psychiatric disorders, substance abuse, assisted reproduction, urbanization, hypertensive disorders, diabetes gravidarum, female child gender and year. a Adverse fetal outcome, perinatal mortality and/or SGA and/or premature birth.
adjustment for these characteristics. Table 3 shows that the crude association between hyperemesis gravidarum and adverse outcome (premature birth, weight <10th percentile and/or perinatal mortality) had an OR of 1.22 (95% CI 1.10–1.36). After adjustment for maternal characteristics (age, parity, socio-economic status, ethnicity, mode of conception, urbanization, substance abuse, hypertension, diabetes and psychiatric disease and sex of the baby), the OR was 1.07 (95% CI 0.95–1.19). 4. Comment Hyperemesis gravidarum is associated with adverse pregnancy outcomes. Women who suffer from hyperemesis gravidarum more often deliver prematurely, and more often have children who are small for gestational age. This is largely explained by differences in age, parity, ethnicity, socio-economic status, substance abuse, and the prevalence of hypertension, diabetes and psychiatric illnesses. The main strength of this study is the large sample size and its representativeness. With 96% coverage and over a million pregnancies included, it is – to our knowledge – the largest cohort study on hyperemesis gravidarum. Because we used information from a national registry, its weakness is the quality of the data. The presence or absence of hyperemesis may have been underreported or selectively reported since it is not an obligatory field in the registry. The relatively low prevalence of hyperemesis compared to other European studies is consistent with this [2– 4]. This will have weakened the associations found, and may have led to an underestimation of the effects. The estimate, however, of the reduction of birth weight after hyperemesis gravidarum found in our study is similar to that found in other studies [4,5]. Inevitably, there has been under-reporting of some risk factors like smoking and the diseases in the obstetric history, which will have weakened the associations, and did not allow us to fully adjust for these effects. Also, the absence of information on weight gain in the registry prevented us from examining whether the effects of hyperemesis are mediated through reduced weight gain. We defined hyperemetic pregnancies as those pregnancies that were labeled during any time of pregnancy as complicated by hyperemesis gravidarum by either the midwife or the gynecologist. One-third of the cases were labeled as hyperemesis gravidarum by the midwife, but did not require hospital admission, one-third of the cases were referred by the midwife to the gynecologist and admitted to hospital, and one-third were labeled as hyperemesis gravidarum by the gynecologist, and consequently admitted. The pregnancies that were labeled as complicated by hyperemesis gravidarum but that did not require hospital admissions may have been less severe. Analyses of only the pregnancies that did require hospital admission for hyperemesis gravidarum, however, showed similar results (results not shown). The finding that women who suffer from hyperemesis gravidarum are different from women who do not suffer from it are in agreement with other reports in the literature [3,6,7,10], although the associations with substance abuse and assisted
reproduction had not been reported before. The association between hyperemesis gravidarum and prematurity and reduced birth weights has been reported previously [2,4,7–10], but what this study adds is that this association is largely dependent on the various maternal characteristics associated with hyperemesis gravidarum. Although it may seem reassuring that hyperemesis gravidarum does not have an independent negative effect on pregnancy outcome, this may be misleading. There is a large body of evidence suggesting that the environment in utero has lasting consequences for health in later life, and that this may be independent of fetal growth. We have shown that maternal under nutrition especially in early gestation is associated with increased risks of chronic degenerative diseases such as type 2 diabetes, cardiovascular disease and breast-cancer [16–19]. The effects of maternal under-nutrition during gestation on later health were present even without effects on maternal weight gain and size of the baby at birth. Animal studies have shown that under-nutrition during gestation has lasting adverse consequences for health and lifespan [20]. Given the fact that hyperemesis gravidarum is associated with metabolic changes in the mother which are similar to those seen in starvation, it is surprising that the long term consequences of hyperemesis gravidarum are almost entirely unstudied. To our knowledge there is only one report of an increased risk of testicular cancer among boys born to mothers who suffered from hyperemesis gravidarum which was hypothesized to be due to the increased levels of estrogens at the time of differentiation of the testis [21]. The effects of severe morning sickness on the offspring’s markers of cardiovascular risk have, however, never been studied. Follow-up studies in large population-based cohorts may provide the first evidence of an increase of cardiovascular risk markers among the children of mothers who suffered severe hyperemesis gravidarum, and should be given high priority. A randomized controlled trial investigating the effects of nasogastric feeding strategies for hyperemesis gravidarum on short term outcomes, including improvement of nausea and vomiting, duration of hospitalization and birth weight, has never been conducted [22]. The follow-up of the cardiovascular risk profile of the offspring from a trial of this kind may resolve whether nutritional support may lower cardiovascular risk in the offspring. Hyperemesis gravidarum is associated with adverse pregnancy outcomes. Notwithstanding the fact that this is largely explained by differences in maternal characteristics, and given the impact of the early environment for later health, studies that aim to assess the long-term consequences of hyperemesis gravidarum need to be given high priority. Acknowledgements The Netherlands Perinatal Registry has given permission for the analysis of the data described in this paper. This permission is gratefully acknowledged.
T.J. Roseboom et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 56–59
References [1] Arsenault MY, Lane CA, MacKinnon CJ, et al. The management of nausea and vomiting of pregnancy. J Obstet Gynaecol Can 2002;24:817–31. [2] Bashiri A, Neumann L, Maymon E, Katz M. Hyperemesis gravidarum: epidemiologic features, complications and outcome. Eur J Obstet Gynecol Reprod Biol 1995;63:135–8. [3] Vikanes A, Grjibovski AM, Vangen S, Magnus P. Variations in prevalence of hyperemesis gravidarum by country of birth: a study of 900074 pregnancies in Norway, 1967–2005. Scand J Public Health 2008;36:135–42. [4] Bailit JL. Hyperemesis gravidarum: epidemiologic findings from a large cohort. Am J Obstet Gynaecol 2005;193:811–4. [5] Gazmararian JA, Petersen R, Jamieson DJ, et al. Hospitalizations during pregnancy among managed care enrollees. Obstet Gynaecol 2002;100:94–100. [6] Fell DB, Dodds L, Joseph KS, Allen WM, Butler B. Risk factors for hyperemesis gravidarum requiring hospital admission during pregnancy. Obstet Gynaecol 2006;107:277–84. [7] Verberg MF, Gillott DJ, Al-Fardan N, Grudzinskas JG. Hyperemesis gravidarum, a literature review. Hum Reprod Update 2005;11(September–October (5)): 527–39. [8] Dodds L, Fell DB, Joseph KS. Outcome of pregnancies complicated by hyperemesis gravidarum. Obstet Gynaecol 2006;107:285–92. [9] Tan PC, Jacob R, Quek KF, Omar SZ. Pregnancy outcome in hyperemesis gravidarum and the effect of laboratory clinical indicators of hyperemesis severity. J Obstet Gynaecol Res 2007;33:457–64. [10] Depue RH, Bernstein L, Loss RK, Judd HL, Henderson BE. Hyperemesis gravidarum in relation to estradiol levels, pregnancy outcome, and other maternal factors: a seroepidemiologic study. Am J Obstet Gynaecol 1987;156:1137–41. [11] Stichting Perinatale Registratie Nederland. Perinatal care in The Netherlands 2006. Utrecht: Stichting Perinatale Registratie Nederland; 2008.
59
[12] Meray N, Reitsma JB, Ravelli AC, Bonsel GJ. Probabilistic record linkage is a valid and transparent tool to combine databases without a patient identification number. J Clin Epidemiol 2007;60(9):883–91. [13] Tromp M, Ravelli AC, Meray N, Reitsma JB, Bonsel GJ. An efficient validation method of probabilistic record linkage including readmissions and twins. Methods Inf Med 2008;47(4):356–63. [14] Ravelli ACJ, Tromp M, Eskes M, van Huis AM, Steegers EAP, Tamminga P, Bonsel GJ. Perinatal mortality in The Netherlands 2000–2006; risk factors and risk selection. Ned Tijdschr Geneeskd 2008;152(50):2728–33. [15] Ravelli ACJ, Tromp M, Eskes M, et al. Decreasing perinatal mortality in The Netherlands, 2000–2006: a record linkage study. JECH 2009. [16] Ravelli AC, Van Der Meulen JH, Michels RP, et al. Glucose tolerance in adults after prenatal exposure to famine. Lancet 1998;351(9097):173–7. [17] de Rooij SR, Painter RC, Roseboom TJ, Phillips DI, Osmond C, Barker DJP, Tanck M, Michels R, Bossuyt P, Bleker OP. Glucose tolerance at age 58 and the decline of glucose tolerance in comparison with age 50 in people prenatally exposed to the Dutch famine. Diabetologia 2006;49(4):637–43. [18] Painter RC, de Rooij SR, Roseboom TJ, Bossuyt P, Simmers TA, Osmond C, Barker DJP, Bleker OP. Early onset of coronary artery disease after prenatal exposure to the Dutch famine. Am J Clin Nutr 2006;84(2):322–7. [19] Painter RC, de Rooij SR, Bossuyt PM, Osmond C, Barker DJP, Bleker OP, Roseboom TJ. A possible link between prenatal exposure to famine and breast cancer – a preliminary study. Am J Hum Biol 2006;18:853–6. [20] Ozanne SE, Hales CN. Lifespan: catch-up and obesity in male mice. Nature 2004;427:411–2. [21] Henderson BE, Benton B, Jing J, Yu MC, Pike MC. Risk factors for cancer of the testis in young men. Int J Cancer 1979;15:598–602. [22] Jewell D, Young G. Interventions for nausea and vomiting in early pregnancy. Cochrane Database Syst Rev 2003;4. CD000145.