The European Journal of Public Health Advance Access published online on June 10, 2008
The European Journal of Public Health, doi:10.1093/eurpub/ckn051
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Length of residence and risk of developing hyperemesis gravidarum among first generation immigrants to Norway
Åse Vikanes1,2, Andrej M. Grjibovski1, Siri Vangen1,3 and Per Magnus1
1 Division of Epidemiology, Norwegian Institute of Public Health, Post box 4404 Nydalen, 0403 Oslo, Norway
2 Department of Gynecology and Obstetrics, Akershus University Hospital, 1478 Lørenskog, Norway
3 Department of Gynecology and Obstetrics, Ullevål University Hospital, 0407 Oslo, Norway
Correspondence: Åse Vikanes, Division of Epidemiology, Norwegian Institute of Public Health, Post box 4404 Nydalen, 0403 Oslo, Norway, tel: +47 23408328, fax: +47 23408252, e-mail: ase.vigdis.vikanes{at}fhi.no
Received August 20, 2007, accepted May 8, 2008
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Background: To estimate the risk of hyperemesis gravidarum (HG) among first generation immigrants to Norway by length of residence. Design: A cross-sectional study. Methods: The sample consisted of first generation immigrants with a prevalence of HG exceeding ethnic Norwegians by 50%, registered in the Medical Birth Registry of Norway (MBRN). The women were born in Turkey, Middle East, North Africa, Other Africa, Iran, Pakistan, India and Sri Lanka and Central and South America, the total sample size is 50 904. MBRN data on HG, age, parity, plurality and sex of the baby were linked to information on country of birth, maternal education and duration of stay obtained from Statistics Norway. Independent associations were studied for each immigrant group and adjusted for potential confounders. To account for dependencies in the sample, a generalized estimating equations (GEE) approach was used. Results: For women from Central and South America, adjusted analysis showed a decrease in the risk of HG by longer residency (P for trend = 0.026). A similar but not significant trend was observed for women born in the Middle East (P for trend = 0.097). Women born in Turkey who had been living in Norway for 6–8 years had a higher risk of HG than newcomers, though no trend was observed (P for trend = 0.127). Women born in Iran and North Africa who lived longer in Norway tended to have a higher risk of developing HG than newcomers (P for trend = 0.083 and 0.118, respectively) Conclusion: Associations between HG and duration of residence in Norway did not show a universal pattern across immigrant groups. Women born in Central and South America had a lower risk of HG with increasing length of residence. Some evidence to the contrary was found for women born in Iran, North Africa and Turkey.
Keywords: hyperemesis gravidarum, immigrants, length of residency, Norway, prevalence
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Nausea and vomiting may occur in as many as 90% of pregnancies, whereas hyperemesis gravidarum (HG) represents a more severe condition and is potentially lethal if not treated.1 HG affects 0.5–2% of all pregnancies, and is defined as persisting nausea and vomiting leading to dehydration, weight loss and nutritional deficiencies starting before the 22nd week of gestation.2 The etiology is unknown. Earlier studies have shown that HG can lead to an increased risk of low birth weight, preterm birth and lower 5 min Apgar scores.3,4 Moreover, HG affects a woman's quality of life and daily functioning, and is the most common reason for hospitalization during early pregnancy.5,6
Ethnic differences in the prevalence of HG have been described.7–9 In New Zealand, the incidence of HG among the Pacific Islanders was higher than in women of European descent.10,11 In the USA, being non-white was associated with a higher probability of developing HG.12 In Norway, an increased risk was associated with having a non-Norwegian name.13 Another study showed that Pakistani immigrant women in Norway were 3.7 times more likely to report HG than ethnic Norwegian women.14 However, these studies suffer from relatively small sample sizes, different definitions of HG and low precision in the description of ethnic origin. In our previous study on 900 000 first births in Norway we found that all immigrants, except those born in Europe and North America, had a higher prevalence of HG than the ethnic Norwegian women, who had an HG prevalence of 0.9%.15 However, all immigrant women were included regardless of the duration of their residence in Norway.
The interaction between migration and health is complex and constitutes a growing global challenge as the number of migrants is increasing. In 2000, 175 million people or 3% of the world's population were immigrants.16 These numbers reflect migrant workers, permanent migrants, refugees and asylum applicants, but not illegal migrants who go through a highly heterogeneous process of adjustment.17 Migration is known to influence health, and migrants’ disease patterns often become more similar to the host populations’ over time.18–22 Despite this fact, we only found one study addressing changes in pregnancy outcomes in immigrants related to duration of stay in a new country. Mexican immigrants who had lived 5 years or longer in the USA were 1.9 and 1.5 times more likely to deliver preterm and low birth weight infants, respectively, than those who had lived there for a shorter period.23 To our knowledge, there have been no studies on pregnancy-related conditions or outcomes among immigrants in relation to the length of residence in European countries.
We have previously found the prevalence of HG to vary according to the mother's country of birth.15 As part of an expanded study on HG, we want to explore if length of residence in Norway influences the risk of developing the disease among first generation immigrants. HG has been associated with several factors that can be influenced by migration, such as psychological stress, an increased intake of fat before pregnancy and exposure to Helicobacter pylori.2,7,24–27 Moreover, the stress associated with being in a new country may change with length of residence. Given that the prevalence of HG in Norway is higher among immigrants, we hypothesize that their risk of developing HG may decrease with an increased length of residence in Norway. To examine this hypothesis, we studied whether or not women with a longer duration of stay in Norway had a lower risk of developing HG.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Sample
The data on all pregnancies (n = 55 283) among first generation immigrants were obtained from the Medical Birth Registry of Norway (MBRN); from 16th gestational week (1967–98) and 12th gestational week (June, 1999–2005). Statistics Norway defines a first generation immigrant as an immigrant who was born outside Norway with two parents born in countries other than Norway.28 The mother's unique identification number was used to link the MBRN to the data on maternal country of birth and education as registered by Statistics Norway, which routinely registers these data for all immigrant women at the time of immigration to Norway. To make a logical follow-up of our previous study, we selected only those groups of immigrants in which the prevalence of HG exceeded the prevalence among ethnic Norwegian women by 50%. Altogether, eight groups were selected including women born in Turkey, Middle East, North Africa, Other Africa (Africa except North Africa), Iran, Pakistan, India and Sri Lanka and Central and South America.15 We excluded all records with missing data for maternal age, parity, plurality, marital status and sex of the baby because they were expected to have low validity of HG registration. Given that the unit of observation in this study is pregnancy, only one record was used in cases of multiple births. The final sample consisted of 50 904 pregnancies or 92.1% of the initial sample.
Variables
The HG data (ICD-8 codes 638.0, 638.9 and 784.1 and ICD-10 codes O21.0, O21.1 and O21.9), maternal age, parity, plurality, marital status, sex of the baby and year of delivery were obtained from MBRN. Notification to MBRN is compulsory and provided by midwives and physicians attending the birth who completed a standardized form. The duration of residence in Norway was calculated using the information on the date of immigration to Norway. The date of delivery is categorized into 4 groups: 0–2 years, 3–5 years, 6–8 years and 9 years or more. Women who had been living in Norway for 2 years or less comprised the reference group. Maternal age was classified as <25 years, 20–25 years, 26–29 years, 30–34 and 35 years or older. By parity, women were categorized as primiparas or paras. By plurality, pregnancies were dichotomized as singleton or plural. Education was classified as 
9 years and 10 or more years. Women with missing data on education comprised a separate group. Marital status was classified into two groups: married and unmarried, where the latter included cohabiting, single, divorced, separated and widows due to small numbers in each of the groups. Years of delivery were grouped into four periods: before 1990, 1990–96, 1997–2000 and after 2000.
Data analyses
Univariate associations between the variables were studied by Pearson's chi-squared tests. Independent effects of the length of residence in Norway on the risk of developing HG were explored separately for each immigrant group by calculating the relative risks with adjustment for maternal age, education, marital status, parity, plurality, sex of the baby and year of delivery. Given that the same woman may be included more than once in the analysis, a generalized estimating equations (GEE) approach was applied to account for dependencies in the sample.29 Population averaged GEE models with log link function and exchangeable correlation structure were fit to obtain crude and adjusted relative risks (RR) for developing HG. Standard errors used for calculating 95% confidence intervals (CI) were adjusted for clustering on maternal individual number. In addition, HG trends across the duration of stay categories were studied by introducing years spent in Norway as a continuous variable in the model. GEE models were fit using STATA software version 8 (STATA Corporation, TX, USA). All other analyses were performed using SPSS version 14.0 (SPSS Inc, Chicago, IL, USA).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Women born in Pakistan constituted the largest group (26%), followed by women born on the African continent (except North Africa), in India and Sri Lanka (table 1). A third of the women were aged between 25 years and 29 years, whereas 28.8% were below 25 years and 12.2% were aged 35 years or older. Most women were paras and married. The information on education level was missing for almost half of the women, whereas 40.4% had 10 years of education or more. The vast majority immigrated to Norway after 1990, 50% after1997. A detailed description of the sample is presented in table 1.
|
Different patterns of HG prevalence in relation to the duration of stay in Norway reflected maternal ethnic background (table 2). For women born in the Middle East and Central and South America, the HG prevalence was lower for those who had lived in Norway for 9 years than for those with fewer years of residence, 2.0 and 1.7% versus 0.9 and 0.9%, respectively. In contrast, Iranian women with the longest stay in Norway had a higher prevalence of HG than the newcomers (1.3% versus 2.6%). However, none of the differences in the prevalence of HG were statistically significant.
|
In crude analysis, women from the Middle East with longer residencies in Norway were less likely to have HG (P for trend = 0.037). A similar association was found for women born in Central and South America and Other Africa, although not reaching the level of statistical significance (P for trend = 0.083 and 0.095, respectively). The opposite trend was observed for women from Iran, but the findings were inconclusive (P = 0.082). Women born in India or Sri Lanka who had lived in Norway between 3 years and 5 years were significantly less likely to have HG than the newcomers, but no trend was observed.
Adjustment for other maternal characteristics weakened the association between the duration of stay in Norway and HG for women from the Middle East to non-significant level (P for trend = 0.097), although the estimates remained virtually unchanged. At the same time, this association of interest for women from Central and South America was strengthened (P for trend = 0.026). The risk of developing HG after 8 years in Norway among these women was almost 3 times lower than in the reference group.
The associations between duration of stay and HG in other groups were less obvious. Women born in Turkey who had been living in Norway for 6–8 years had more than a 2 times higher risk of HG than the reference group (adjusted RR = 2.06, 95% CI: 1.06–4.02). Similarly, the highest risk of HG compared to the reference group was observed among women born in Iran with the longest time of residency in Norway (adjusted RR = 2. 28, 95% CI: 0.90–5.82), but the findings were not significant (P for trend = 0.083). Women from North Africa showed a similar pattern although not significant. The detailed data on crude and adjusted RR for HG by duration of residence in Norway are presented for all studied groups in table 3.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
The associations between HG and duration of residence in Norway observed in this study do not have a universal pattern across immigrant groups. Our main hypothesis was supported by data on women from Central and South America and probably from the Middle East that showed a decreasing risk of developing HG with an increasing length of time in Norway. No clear patterns were found in other immigrant groups. There were some indications, however, of the opposite trend for women born in Iran, North Africa and Turkey, although the results were inconclusive.
The use of data from high quality national registers that include virtually all immigrant women is the main strength of this study.30 Notification to MBRN is compulsory and provided by midwives and physicians attending the birth who complete a standardized form. The diagnostic criteria are expected to be the same during the different periods of births. The validity of the HG registration in MBRN was previously discussed.31 Furthermore, it was tested in a study showing the sensitivity, specificity and total accuracy of the registration in non-Norwegian women to be 67, 98 and 92%, respectively.15 The registration of country of birth by Statistics Norway is performed at the time of immigration and is likely to have a higher validity than self-reported data on ethnic background.
Despite the fact that the overall sample in this study was rather large, the statistical power was insufficient to show significant differences in such rare outcomes as HG by the length of residence for some groups, even when the relative risks were high. For example, women born in North Africa and Iran who had lived in Norway for 9 years had 76% and 128% higher risk of HG than the reference groups. However, the CIs were too wide and included 1 because the groups were small consisting of only 48 and 41 cases of HG, respectively.
Longer residence in Norway was associated with higher education and higher parity (data not shown). As in our previous study that revealed negative associations between HG and maternal education as well as HG and parity, both variables could be considered negative confounders in this study.15 Adjustment for these variables resulted in an elevated RR for those who lived longer in Norway (table 3).15,32 Education could have a different impact in different groups of immigrants, further supporting our strategy to stratify the data by country of birth. However, there is still a risk of residual confounding due to the large proportion of missing data for education.
A history of HG in previous pregnancies is the most important determinant of HG in the index pregnancy.31 However, more than a third of the women were primiparas and most of the paras had their first pregnancy outside Norway making adequate adjustment for prior HG impossible.
Some researchers have speculated that genetic factors may explain the differences in the prevalence of HG between ethnic groups.33,34 The variations in HG by length of residence in Norway in some groups, especially among women from Central and South America, observed in this study, may suggest that environmental factors are important in the aetiology of HG and can modify the genetic risk of HG. Interestingly, the prevalence of HG among Norwegian-born women is about 0.9%, i.e. similar to the prevalence of HG among women born in Central and South America and the Middle East who have lived in Norway for 9 years or longer, which supports our main hypothesis. However, no negative associations between length of stay in Norway and HG were found in other groups. On the contrary, Turkish women, and most likely women born in Iran and North Africa, who had lived in Norway for some time had a higher probability of developing HG. These results suggest that the same environmental factors may have a different influence on different ethnic groups, or that different environmental factors are important in different groups. A variety of environmental factors associated with HG such as Helicobacter pylori, fat intake before pregnancy, psychological aspects as well as hormonal and immunological disturbances have been explored over the last decades, but different studies have shown conflicting evidence.24–27,35
Previously, immigration to Norway was mainly driven by the labour market and most immigrants came from Europe and North America. Since 1981, immigrants to Norway have come from a wide range of geographical areas for different reasons including refuge, asylum or permanent residence for family unification.28 In addition to personal vulnerabilities, immigrants often have a pattern of disease similar to that observed in their country of origin. It is not known whether differences in risk of HG by length of residence are associated with their immigrant background or reflect their psychological stress level. Migration, however, is known to increase the level of psychological stress for some groups.21
Acculturation describes the process by which a racial or ethnic group, usually a minority, adopts the cultural patterns of a dominant or host group.36 Health outcomes among immigrants might differ by generational status and duration of residence, whereas behavioural risk factors seem to converge towards the host population.18,20 Whether or not the considerable variations in risk of HG associated with the length of residence in Norway reflect varying degrees of acculturation remains to be explored. One can, for instance, speculate that a change in diet such as increased intake of fat may play a role.
In conclusion, the data do not fully support our original hypothesis on the negative association between the length of residence and the risk of HG. The patterns were different for different immigrant groups. While a clear negative association was found for women born in Central and South America, and a similar pattern was observed for women born in the Middle East, some evidence to the contrary was found for women born in Iran, North Africa and probably Turkey, although our data do not permit firm conclusions for these groups.
Further research should address the underlying factors that may explain the observed patterns of HG in different ethnic groups. These studies might explore the influence of different environmental factors on HG or aim to reveal the genetic aspects of the disease.
|
Acknowledgement |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Financial support for this study was obtained from the Research Council of Norway.
Conflict of interest: None declared.
Key points
|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
1 Ogunyemi DA, Michelini GA. Hyperemesis gravidarum. Emedicine. webMD. Ref Type: Internet Communication. Available online at http://www.emedicine.com/med/topic1075.htm (8 August 2006, date accessed).
2 Eliakim R, Abulafia O, Sherer DM. Hyperemesis gravidarum: a current review. Am J Perinatol (2000) 17:207–18.[CrossRef][Web of Science][Medline]
3 Godsey RK, Newman RB. Hyperemesis gravidarum. A comparison of single and multiple admissions. J Reprod Med (1991) 36:287–90.[Web of Science][Medline]
4 Dodds L, Fell DB, Joseph KS, et al. Outcomes of pregnancies complicated by hyperemesis gravidarum. Obstet Gynecol (2006) 107:285–92.[Web of Science][Medline]
5 Attard CL, Kohli MA, Coleman S, et al. The burden of illness of severe nausea and vomiting of pregnancy in the United States. Am J Obstet Gynecol (2002) 186(5 Suppl Understanding):S220–7.[CrossRef][Web of Science][Medline]
6 Bacak SJ, Callaghan WM, Dietz PM, Crouse C. Pregnancy-associated hospitalizations in the United States, 1999-2000. Am J Obstet Gynecol (2005) 192:592–97.[CrossRef][Web of Science][Medline]
7 Fairweather DV. Nausea and vomiting in pregnancy. Am J Obstet Gynecol (1968) 102:135–75.[Web of Science][Medline]
8 Louw JX. Aspects of vomiting in the African. Postoperative vomiting and vomiting of pregnancy. S Afr Med J (1966) 40:957–9.[Medline]
9 Jimenez V, Marleau JD. Is hyperemesis gravidarum related to country of origin? Short report. Can Fam Physician (2000) 46:1607–8.
10 Jordan V, MacDonald J, Crichton S, et al. The incidence of hyperemesis gravidarum is increased among Pacific Islanders living in Wellington. N Z Med J (1995) 108:342–4.[Web of Science][Medline]
11 Browning J, North R, Hayward P, et al. Hyperemesis gravidarum: a particular problem for Pacific Islanders. N Z Med J (1991) 104:480.[Web of Science][Medline]
12 Paauw JD, Bierling S, Cook CR, Davis AT. Hyperemesis gravidarum and fetal outcome. JPEN J Parenter Enteral Nutr (2005) 29:93–6.
13 Vilming B, Nesheim BI. Hyperemesis gravidarum in a contemporary population in Oslo. Acta Obstet Gynecol Scand (2000) 79:640–3.[CrossRef][Web of Science][Medline]
14 Vangen S, Stoltenberg C, Stray-Pedersen B. Complaints and complications in pregnancy: a study of ethnic Norwegian and ethnic Pakistani women in Oslo. Ethn Health (1999) 4:19–28.[CrossRef][Medline]
15 Vikanes ÅV, Grjibovski A, Vangen S, Magnus P. Variations in prevalence of hyperemesis gravidarum by country of birth: a study of 900,074 birth in Norway, 1967–2005. Scand J Public Health (2008) 36:135–42.
16 Norredam ML, Nielsen AS, Krasnik A. Access to health care for migrants. Ugeskr Laeger (2006) 168:3008–11.[Medline]
17 Bhugra D. Migration and mental health. Acta Psychiatr Scand (2004) 109:243–58.[CrossRef][Web of Science][Medline]
18 Singh GK, Siahpush M. Ethnic-immigrant differentials in health behaviors, morbidity, and cause-specific mortality in the United States: an analysis of two national data bases. Hum Biol (2002) 74:83–109.[Web of Science][Medline]
19 Thomas SL, Thomas SD. Displacement and health. Br Med Bull (2004) 69:115–27.
20 Hosper K, Nierkens V, Nicolaou M, Stronks K. Behavioural risk factors in two generations of non-Western migrants: do trends converge towards the host population? Eur J Epidemiol (2007) 22:163–72.[CrossRef][Web of Science][Medline]
21 Singh GK, Siahpush M. All-cause and cause-specific mortality of immigrants and native born in the United States. Am J Public Health (2001) 91:392–9.
22 Singh GK, Yu SM. Adverse pregnancy outcomes: differences between US- and foreign-born women in major US racial and ethnic groups. Am J Public Health (1996) 86:837–43.
23 Guendelman S, English PB. Effect of United States residence on birth outcomes among Mexican immigrants: an exploratory study. Am J Epidemiol (1995) 142(Suppl 9):S30–8.[Web of Science][Medline]
24 Simpson SW, Goodwin TM, Robins SB, et al. Psychological factors and hyperemesis gravidarum. J Womens Health Gend Based Med (2001) 10:471–7.[CrossRef][Web of Science][Medline]
25 Signorello LB, Harlow BL, Wang S, Erick MA. Saturated fat intake and the risk of severe hyperemesis gravidarum. Epidemiology (1998) 9:636–40.[Web of Science][Medline]
26 Penney DS. Helicobacter pylori and severe nausea and vomiting during pregnancy. J Midwifery Womens Health (2005) 50:418–22.[CrossRef][Web of Science][Medline]
27 Verberg MF, Gillott DJ, Al Fardan N, Grudzinskas JG. Hyperemesis gravidarum, a literature review. Hum Reprod Update (2005) 11:527–39.
28 Statistics Norway. Ref Type: Internet Communication. Available online at www.ssb.no (13 November 2006, date accessed).
29 Twisk J. Applied longitudinal analysis for epidemiology. In: A practical guide (2003) Cambridge, UK: Cambrigde University Press.
30 Irgens LM. The medical birth registry of Norway. Epidemiological research and surveillance throughout 30 years. Acta Obstet Gynecol Scand (2000) 79:435–9.[CrossRef][Web of Science][Medline]
31 Trogstad LI, Stoltenberg C, Magnus P, et al. Recurrence risk in hyperemesis gravidarum. BJOG (2005) 112:1641–5.[CrossRef][Web of Science][Medline]
32 Mehio-Sibai A, Feinleib M, Sibai TA, Armenian HK. A positive or a negative confounding variable? A simple teaching aid for clinicians and students. Ann Epidemiol (2005) 15:421–3.[CrossRef][Web of Science][Medline]
33 Price A, Davies R, Heller SR, et al. Asian women are at increased risk of gestational thyrotoxicosis. J Clin Endocrinol Metab (1996) 81:1160–3.[Abstract]
34 Jordan V, Grebe SK, Cooke RR, et al. Acidic isoforms of chorionic gonadotrophin in European and Samoan women are associated with hyperemesis gravidarum and may be thyrotrophic. Clin Endocrinol (1999) 50:619–27.[CrossRef][Medline]
35 Frigo P, Lang C, Reisenberger K, et al. Hyperemesis gravidarum associated with Helicobacter pylori seropositivity. Obstet Gynecol (1998) 91:615–7.[CrossRef][Web of Science][Medline]
36 Satia-Abouta J, Patterson RE, Neuhouser ML, Elder J. Dietary acculturation: applications to nutrition research and dietetics. J Am Diet Assoc (2002) 102:1105–18.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||