Treatment of nausea in pregnancy: a cross-sectional multinational web-based study of pregnant women and new mothers

Nausea and vomiting during pregnancy (NVP) is one of the most common pregnancy-related complaints, affecting millions of pregnant women worldwide each year. Approximately 7 out of 10 women experience nausea during pregnancy [1] and 50 % experience both nausea and vomiting [2‐4]. For most patients, symptoms appear around the sixth week of pregnancy and gradually decline during the second trimester, peaking at 8–13 weeks [2, 3]. However, 10 % of women will still experience symptoms after 20–22 weeks of pregnancy [3, 4]. The most severe form of NVP, hyperemesis gravidarum, is characterised by severe and persistent nausea and vomiting leading to weight loss, ketonuria, nutritional deficiencies, dehydration and electrolyte imbalance, often so severe as merits hospitalisation, and affects about 1.1% of the pregnant women [1].

NVP has been shown to greatly impact a woman’s life, negatively affecting daily activities, relationship with partner, parenting, occupation and social functioning [5‐7]. Feelings of isolation, fatigue, depression, and helplessness due to nausea have also been described [7‐9]. Nausea has been reported to be responsible for 33 % of all sick leave during pregnancy [10]. In the USA and Canada, NVP is a significant economic burden to women and society [11, 12].

Because NVP is often most intense during the first trimester when organogenesis occurs, teratogenic effects are a concern in treatment. This may lead to caution in prescribing and taking conventional medicines to treat this condition, despite the proven safety of use during pregnancy of many medicines. Pregnant women often overestimate the teratogenic risk associated with the use of medicines in general [13]. Consequently, many women may turn to complementary and alternative medicine (CAM) to alleviate their symptoms or choose to not treat their symptoms due to the fear that taking anything during pregnancy may harm the baby. Despite the high prevalence of NVP, little is known about differences in NVP treatments, i.e. conventional and herbal medicines, across countries. An informal survey in various European countries in 1998 found wide variations in the types of treatment used against mild and moderate nausea and vomiting, whereas hyperemesis gravidarum was treated in a similar fashion in the vast majority of countries [14].

No study has investigated the factors related to NVP treatment in several countries simultaneously. Studies are available from various countries, but data collection methods vary [5, 7, 15‐20], making direct comparisons impossible. In addition different therapies have been included in the definition of CAM [15, 16, 18, 19].

New possibilities for uniform data collection in several countries are emerging that are advantageous for the field of e-epidemiology [21, 22]. Large potential gains in wellbeing for the mother and society as a whole can be achieved with better knowledge on how nausea in pregnancy is being treated. This study is the first to investigate the factors related to the treatment of nausea during pregnancy at a multinational level.

The present study aimed to describe differences in self-reported nausea during pregnancy, as well as the patterns of use of both conventional and herbal medicine across countries in Western, Northern, and Eastern Europe, North America, and Australia. The study also aimed to investigate the factors related to nausea and its treatment, as well as the relationships between different self-reported co-morbidities and nausea.

This cross-sectional study was based on data from a web-based questionnaire covering nausea, medicines against nausea, herbal medicines against nausea, sociodemographic factors, maternal health, and lifestyle during pregnancy [23]. The online questionnaire was distributed simultaneously in 18 countries: Austria, Australia, Croatia, Canada, France, Finland, Iceland, Italy, Norway, Poland, Russia, Serbia, Slovenia, Sweden, Switzerland, the Netherlands, United Kingdom, and USA. The original data file consisted of 9459 women, including 346 South American women who accessed the questionnaire via North American web-sites [23]. For this sub-study, the 346 women from South America were excluded in effort to reduce selection bias as this was considered a special group of women, resulting in a final study population of 9113 women.

Women who were pregnant or had a child who was less than one year old were eligible to participate in the study. An advert containing a link to the online self-completed questionnaire was posted on commonly visited pregnancy and baby related websites in the participating countries. National coordinators selected the most relevant national websites, social networks, and pregnancy forums [23]. The questionnaire was available for 2 months in each participating country between 1st of October 2011 and 29th of February 2012.

The questionnaire was originally developed in Norwegian and English before being translated into the relevant languages, and is available online as an appendix to the paper by Lupattelli et al. [23]. A pilot study was performed during September 2011 in Norway, Finland, Italy, and Sweden (n = 47) but resulted in no major changes to the questionnaire. All national coordinators assured the quality of their version of the questionnaire.

The representativeness of the study population was assessed by comparing the sociodemographic and lifestyle characteristics (i.e., age, marital status, education, and smoking) of the study population to the general birthing population in the corresponding country. The similarities were satisfactory with the exception that the study participants were generally more educated than the general birthing population, as described in detail elsewhere [23].

During the 2-month study period in each country, a total of 9113 women were included in the study. Respondents who were residents of Europe (Western, n = 3201; Northern, n = 2820; Eastern, n = 2342) constituted the largest proportion of the total study population, followed by North America (n = 533) and Australia (n = 217).

At the time of completing the questionnaire, 4938 (54.2 %) of the women were pregnant and 4175 (45.8 %) had given birth during the previous year. Among the pregnant respondents, 1067 (21.6 %), 1656 (33.5 %), and 2214 (44.8 %) were in the first, second, and third trimester of their pregnancy, respectively, and 182 (3.7 %) were less than 6 weeks pregnant. A total of 1913 (45.8 %) of the mothers had an infant less than 24 weeks of age.

Among the respondents, 6701 (73.5 %) had experienced nausea during pregnancy; 1828 (27.3 %) used some form of treatment against nausea and 4873 (72.7 %) did not. Conventional medicines against nausea were used by 1201 (17.9 %) of the women and herbal medicines by 556 (8.3 %).

Both the prevalence of nausea and its treatment varied by country and region (Table 2 and Fig. 1). The prevalence of nausea ranged from 62.0 % in Russia to 84.5 % in Iceland. The proportion of respondents treated among those who suffered from nausea ranged from 10.4 % in Finland to 53.2 % in France; the next highest proportions were in Switzerland (48.9 %), Canada (48.0 %), and Australia (48.0 %). In 11 countries the treatment rates were below 30 % (Fig. 1). Among the regions, Australia (48.0 %) and North America (41.2 %) had the highest rates of treatment.

The most commonly used conventional medicines against nausea in the total population were antihistamines, which were used by 613 respondents (6.7 %) (Table 2 and Additional file 2). Metoclopramide was the second most commonly used medicine with 268 respondents (2.9 %). Antacids (ATC-group A02), ondansetron, and domperidone were used by 176 (2.6 %), 54 (0.6 %), and 48 (0.5 %) respondents, respectively. Conventional medicines were most commonly used against nausea in Canada, France, Switzerland and Sweden. The type of conventional medicine most commonly used among women with nausea differed by region and country, but in the majority of countries it was either antihistamines or metoclopramide. An exception was ondansetron, which was the most commonly used medicine in the United States, closely followed by antihistamines. In Croatia, Russia, Serbia, and Slovenia the most common medicines were antacids, despite heartburn and reflux problems being less prevalent in these countries (60.5 %, 59.6 %, 52.7 %, and 57.7 %, respectively) compared to the total population (66.0 %). The countries in Eastern Europe had a low frequency of conventional medicines use. One respondent from Slovenia reported the use of conventional medicines. Interestingly, metopimazine was reported to be used against nausea by 29 respondents, all in France. In Canada, 54 respondents had used Diclectin®, which is a combination of the antihistamine doxylamine and pyridoxine.

Among the five regions, Australia had the highest frequency of herbal medicine use (21.7 %). Ginger was the most commonly used herbal medicine in the total population (5.1 %) and in most regions and countries. However, in Russia the most commonly used herbal medicine was artichoke (2.8 %). In most countries, herbal medicines were used to a lesser extent than conventional medicines, with the exception of Slovenia, Poland, Austria, and Italy.

Maternal characteristics as predictors of nausea and the use of conventional medicines against nausea during pregnancy are shown in Table 1. Respondents who had more than one previous live birth, worked as health care personnel, or were unemployed were more likely to experience nausea, whereas respondents who used folic acid during pregnancy only or smoked during pregnancy were less likely to experience nausea according to adjusted models. Respondents who had primary school as their highest completed education, were health care personnel, or had used folic acid before the pregnancy were more likely to have used conventional medicines against nausea than respondents with characteristics within the respective reference categories. Multiple pregnancy was also associated with use of medicines against nausea.

Women who experienced nausea during pregnancy were more likely to have any of the acute short-term illnesses listed in Table 3. These respondents were also more likely to have four or more co-morbidities in terms of acute short-term illnesses, any chronic illness, or to have taken sick leave during pregnancy. This pattern was similar when comparing respondents who had nausea and used conventional medicines against nausea to the respondents who experienced nausea without using any treatment. However, the effect estimates were generally lower than for nausea alone. Sub analyses including only respondents pregnant at the time of participation in the study revealed that respondents who experienced nausea were more likely to have symptoms of depression (EPDS score ≥13) than respondents without nausea. This was also true among respondents who had nausea and used conventional medicines compared to respondents who experienced nausea without treatment.

In an additional sub analysis we found that time of gestation acted as a plausible effect modifier of the association between medicated nausea and comorbidity. Specifically, respondents early in their pregnancy (≤1 trimester) who treated nausea with conventional medicines presented a significant 3.1-, 2.8-, and 2.1-fold increased likelihood of taking sick leave (crude OR: 3.1, 95 % CI: 1.8-5.5), having depressive symptoms during pregnancy (crude OR: 2.8, 95 % CI: 1.7-4.6), and having heartburn and reflux problems (crude OR: 2.1, 95 % CI: 1.6-2.6), respectively, compared to respondents with non-medicated nausea. Such measures of association were of a much smaller magnitude (30-70 % increased likelihood) in the stratum comprising only respondents later in their pregnancy (>1 trimester). Similar results were observed when respondents with nausea were compared to those without nausea (data not shown).

Variations were found across countries and regions in the prevalence of nausea, treatment rates, and types of treatment used against nausea during pregnancy. Cultural differences reflected in different treatment traditions, differences between countries with respect to the women’s and general practitioners’ willingness to treat, and variations in access to prenatal care and treatments and their relative costs may explain several of our findings. Among respondents suffering from nausea, less than one in three used any form of treatment, and only 18 % had used any medicine against this complaint. We do not have data on the severity of nausea, and the respondents may generally suffer from mild symptoms that are sufficiently managed by non-pharmacological treatments, such as dietary changes. However, the low prevalence of treatment may also be explained by a reluctance of many general practitioners to treat these women [34], or by an overestimation of the risk of medicines among pregnant women with nausea [35]. The overall prevalence of nausea (73.5 %) in this study is in accordance with a recent meta-analysis of NVP including 59 studies from various countries [1].

Canada, followed by France, Switzerland and Sweden, had the highest prevalence of conventional medicines use against nausea. This finding may be due to the clear and well known guidelines in this country [36, 37] and the antiemetic Diclectin®, which is approved for use against NVP. Therefore, simplifying the identification of safe and effective treatments may possibly increase the use of treatment. Among the European countries, medicine use was highest in France, which is in line with the results of a study of drug utilisation in pregnancy that included six European Registries of Congenital Anomalies in four European countries: France, Great Britain, Italy, and the Netherlands [38]. The authors of the study found that the two centres in France had the highest prevalence of medicine use (80.8 % and 74.2 %), and that antinauseants were the most frequently consumed drugs in this country (20.9 % and 15.0 %) [38]. We found that several French respondents used metopimazine, a dopamine antagonist, which was not reported in any of the other countries. This finding is in accordance with a comparative study by Einarson and colleagues in 1998 in which France was the only country to list metopimazine as a treatment option [14]. In Sweden, antihistamines were the most frequently used medicines against nausea, which is also in line with previous findings [17].

Antihistamines and metoclopramide were the most commonly used conventional medicines against nausea in the vast majority of the countries. The exceptions were the USA, Croatia, Russia, Serbia, and Slovenia, where ondansetron (USA) and antacids (Croatia, Russia, Serbia, and Slovenia) were the most commonly used medicines against nausea. This is in line with findings of a study of hyperemesis gravidarum treatments detecting inter-country variations of frequency of different treatments of which serotonin inhibitors were most frequently used in the USA, antihistamines in Canada, whereas Australia had the highest reported use of promotility agents such as metoclopramide [39]. Meta-analyses and epidemiological studies have not found a higher risk of malformations with antihistamines and metoclopramide [40‐43], and antihistamines are regarded as a first line treatment according to guidelines in both North America and Europe [36, 37, 44‐46]. Recently, the safety of metoclopramide and ondansetron has been questioned [47, 48]. In July 2013, the European Medicines Agency (EMA) recommended changing the use of metoclopramide to 10 mg three times a day up to 5 days to reduce the risk of extrapyramidal side effects [48]. As this is seldom long enough to treat NVP, the change will limit this medicine’s usefulness for nausea and vomiting in the pregnant population. In 2011, the U.S. Food and Drug Administration (FDA) raised concerns over cardiovascular safety, suggesting that ondansetron could cause prolonged QT interval, which can lead to Torsade de Pointes [49]. Electrocardiogram (ECG) monitoring in patients with electrolyte abnormalities is advised. A recent Danish registry study of ondansetron use during the first trimester did not detect any increased risk of malformations [50]. Another unpublished study based on the same registries detected a 2-fold increase in the prevalence of major congenital heart defects after exposure to ondansetron [51]. Notably, the data in our study were collected during winter 2011–2012. Therefore, the medication utilisation pattern may have changed due to the warnings issued by the EMA (2013) and FDA (2011) with respect to metoclopramide and ondansetron, respectively.

Australia had the highest rate of herbal medicine use against nausea, followed by Iceland and Slovenia. Australia has previously been reported to have a high prevalence of herbal medicine and CAM use in pregnancy in general [52, 53], and also more specifically a high use of ginger during pregnancy [52]. The results with respect to Slovenia were special; 12.1 % of respondents had used herbal medicine, but only one respondent had used a conventional medicine against nausea. This finding may indicate a long tradition of herbal medicine in Slovenia or a lack of access to conventional medicines.

Ginger was the dominant herbal medicine used against nausea. Ginger has been reported to be more effective than placebo and equally effective as vitamin B6 and dimenhydrinate against nausea in pregnancy [54]. With respect to safety, a cohort study with 1020 ginger-exposed pregnancies (466 in the first trimester) found no increased risk of malformations, stillbirth/perinatal death, low birth weight, preterm birth, or low Apgar score [16]. Russia was the only country to report the use of artichoke against nausea. However, no studies of artichoke use in pregnancy were found, though artichoke has been observed to have an antiemetic effect in outpatients with dyspeptic syndrome [55].

Various maternal characteristics were associated with nausea and its treatment. Having more than one previous live birth was associated with nausea, probably because having additional children results in less time to rest and relieve the nausea. This finding is in accordance with previous research [56, 57], but the data are conflicting [4]. Other factors associated with nausea were working as health care personnel or being unemployed, which is in line with previous research that found an association between being a housewife or out of work and nausea [57, 58]. Respondents who smoked during pregnancy or who used folic acid during pregnancy were less likely to report nausea. Decreased risk of nausea among smokers was observed in several earlier studies [4, 56, 57]. Use of vitamins in early pregnancy was previously found to be protective against nausea [57, 59]. Women who take folic acid before, as well as during, pregnancy are most likely planning to become pregnant, and this may imply that they are more attentive to early symptoms of pregnancy than women who use folic acid only during pregnancy.

Respondents who had a lower education were more likely to use medicines against nausea. This finding is in accordance with a Swedish study [17]. Respondents working as health care professionals were also more likely to use medicines against nausea, which can be explained by this group being aware of safe and effective treatment options for nausea. Multiple pregnancy was associated with the use of medicines. This may indicate that use of medicines may act as a marker of severe forms of nausea, as it is previously found that multiple pregnancy increase the risk of nausea [56]. In addition, the severity of NVP symptoms has been associated with the use of antiemetics [7, 60].

We found a high burden of co-morbidity among respondents experiencing nausea during pregnancy. The association with symptoms of depression and sick leave in particular warrants attention. Women who suffer from any pregnancy-related complaint may tend to seek information on the internet to a greater extent than women who feel well. Therefore, the respondents may be seeking information and responding at the peak of their discomfort. However, symptoms of depression have also been associated with nausea during pregnancy [7, 9, 60]. Similarly, the association with heartburn and reflux problems is in agreement with previous studies [61]. Clinicians should be aware of the high degree of co-morbidity with nausea and routinely ask women with nausea whether they have reflux problems or other pregnancy-related ailments.

Our findings indicate that women who have nausea in early pregnancy, especially those who treat nausea with medicines, have a high likelihood of experiencing depressive symptoms, heartburn and reflux problems, and taking sick leave. This is an important clinical finding and emphasises how debilitating nausea during pregnancy can be for these women. General practitioners in contact with women with NVP should be aware of the high degree of co-morbidity, examine these women for symptoms of depression and heartburn and reflux problems, and treat these conditions if present. Special attention should be paid to women in early pregnancy. Treating heartburn and reflux problems may alleviate symptoms of NVP and increase the women’s wellbeing [62]. Major guidelines suggest antacids as adjunctive therapy against NVP [36, 37, 44].

This study has several strengths and limitations that should be acknowledged. This is the first multinational study to simultaneously collect data on the prevalence of nausea and its treatment, which enables direct comparisons between countries and regions. A large number of women from a variety of countries in different regions of the world were reached due to the utilisation of a web-based questionnaire posted on various pregnancy-related websites. These data provide valuable insights into the prevalence of nausea and the treatment of this complaint across countries and regions. Furthermore, the study population was reasonably comparable to the general birthing population with respect to age, parity, and smoking habits, though the women in the study population had a higher education on average [23]. However, the possibility that the respondents differ from the general birthing population in ways that our analysis cannot control for cannot be excluded. In some of the countries (Australia, Canada, France, Russia, the Netherlands, and the USA), the study sample was a small proportion of the general birthing population. For these specific countries, our findings should be generalised with caution.

There are some other limitations that need to be addressed. First, a conventional response rate could not be calculated because the questionnaire was only accessible through websites. However, epidemiological studies have indicated that web-based recruitment methods have reasonable validity [63, 64]. In addition, women of childbearing age generally have a relatively high internet penetration rate [65‐67]. The fact that we found a prevalence of nausea very similar to the prevalence reported in the literature, and that the comparison with the birthing population in each participating country had high external validity, supports our approach. However, the higher education of the respondents may have had an impact on their choice of treatment. Second, including women at an early stage in their pregnancy may underestimate the prevalence of nausea, as this complaint often does not occur before gestational weeks 6–8. However, this only applies to the 182 women (2.0 %) who were less than 6 weeks pregnant at the time of participation. Thirdly, although we tried to minimise the risk of recall bias by excluding women with a youngest child aged >1 year, this risk cannot be ruled out. In addition, the EPDS was only measured at one time point during the pregnancy and two time points are considered more valid [29]. Finally, we lack information on the severity of nausea and our results should be interpreted with these limitations in mind.

The prevalence of nausea was high across all participating countries, but its treatment varied, possibly due to cultural differences and differences in attitudes towards medicines. Women who reported nausea also had a high burden of co-morbidity, especially heartburn and reflux symptoms. The association with symptoms of depression and sick leave warrants attention. These findings will be helpful to health care personnel involved in the care of pregnant women with nausea.