The effect of heartburn and acid reflux on the severity of nausea and vomiting of pregnancy

1The Motherisk Program, The Hospital for Sick Children; 2Department of Pharmacology, University of Toronto, Toronto, Ontario Correspondence: Dr Gideon Koren, The Motherisk Program, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8. Telephone 416-813-5781, fax 416-813-7562, e-mail gkoren@sickkids.ca Received for publication July 14, 2008. Accepted July 17, 2008 original arTicle

Heartburn (HB) and acid reflux (RF) are common medical disorders; it has been estimated that the prevalence of gastroesophageal reflux disorders in pregnancy ranges from 40% to 85% (8)(9)(10)(11)(12)(13).The onset of HB/RF can occur at any time during pregnancy.In one study (14), 52% of symptoms began in the first trimester and almost all appeared by the second trimester, with only 8% of symptoms beginning in the third trimester.
Other studies (15,16) reported increased severity and frequency of symptoms as gestational age increases.Regardless of the time of onset, anecdotal clinical evidence suggests that the presence of pre-existing gastrointestinal (GI) conditions and/or symptoms, as well as HB/RF during pregnancy aggravate nausea and vomiting (10,11).
Although the symptoms of HB/RF do not differ in the pregnant versus the nonpregnant population, their etiology may (10,11).Changes in motility of the GI tract due to increased levels of circulating female hormones result in increased GI tract symptoms in pregnancy (10,11,(17)(18)(19).Specifically, a decrease in lower esophageal sphincter pressure during pregnancy accompanied by HB has been demonstrated (17).Furthermore, a decrease in lower esophageal sphincter pressure occurs in the presence of estrogen and progesterone among female volunteers using oral contraceptives (18,19).Similar changes in gastric motility and dysrhythmias have been observed in NVP (8).Therefore, it is biologically plausible that HB/RF can contribute to the severity of NVP.If confirmed, this hypothesis can lead to improvement in the management of NVP by treating symptoms of RF.
The objective of the present study was to determine whether pregnant women suffering from HB, RF or both experience increased severity of NVP.

MeTHODS
The Motherisk Program, located at the Hospital for Sick Children in Toronto, Ontario, has a specialized helpline for the management of NVP.Women from Canada and the United States experiencing NVP can call a toll-free service to receive pharmacological and nonpharmacological advice regarding the management of NVP.This evidence-based counselling is based on research and continuous systematic review of all emerging clinical and experimental evidence.
For the purpose of the present study, a prospective cohort of women counselled by the NVP helpline from January 1, 2007 to December 31, 2007, was analyzed.The study group consisted of women who experienced HB/RF while suffering from NVP.A comparison cohort consisting of women who experienced NVP, but not HB/RF was also collected from the same time period.A standard interview was conducted in which detailed quantification of symptoms was obtained using the following tools: the Pregnancy-Unique Quantification of Emesis and Nausea • (PUQE) score (20) (Table 1); the Well-being score (21) ranging from 0 (the worst possible) • to 10 (the best possible) was recorded based on how the woman felt overall compared with how she felt before pregnancy; and a self-report of how the woman perceived her NVP symptoms • (mild, moderate or severe).
In addition, the time of onset of the NVP symptoms, gravidity, maternal age at conception, gestational age at the initial interview, other medical conditions, pre-existing GI conditions or symptoms, medication use and the severity of NVP in previous pregnancies were recorded.
The unpaired t test was used to compare the mean scores of different characteristics between the two groups.Characteristics that were found to be statistically different between the two groups were included in a multiple linear regression model to determine factors that contributed to the differences in the PUQE and Well-being scores observed between the two groups.
Statistical significance was set at P<0.05 and conducted with the SigmaStat program version 3.1 (Systat Software Inc, USA).

ReSULTS
Of 194 women in the HB/RF group, 60 experienced only HB, 42 experienced only RF, and 92 reported both HB and RF.Selfreported severity of NVP was statistically different between the groups, with 75% of women experiencing both HB and RF classifying their NVP as severe, whereas only 48% of women in the control group (90 of 188) classified their NVP as severe (P<0.0001)(Figure 1).Seventy per cent of women with RF only self-reported their NVP as severe, which was significantly higher than among controls (P=0.009)(Figure 1).Additionally, in the HB/RF group, there were significantly fewer women who classified their NVP as moderate (30%) compared with controls (40%) (P=0.04), and no women experiencing both HB and RF classified their NVP as mild compared with 5% of controls (P=0.002)(Figure 1).
The PUQE scores corroborated the self-reports: the mean PUQE score of women experiencing both HB and RF was higher (9.6±2.6)compared with controls (8.9±2.6); this difference in score was statistically significant (P=0.02)(Figure 2).Women experiencing both HB and RF had the highest mean PUQE score (10.0±2.4) which was found to be significantly higher than in controls (P=0.0004)(Figure 2).
The Well-being scores reflected a similar pattern.The mean score of controls (4.9±2.0) was higher than in the HB/RF group (4.3±2.1;P=0.01), and significantly higher than the mean Well-being score of women experiencing both HB and RF (3.9±2.1;P=0.0004) (Figure 3).
Potential confounding factors were compared; no differences between control and HB/RF groups were found for maternal age at conception, gravidity, gestational age at onset of symptoms, gestational age at interview, multiple gestation and vitamin use.In contrast, there were higher rates of antiemetic use (P=0.0001) and the use of antacids (P<0.0001) in the HB/RF group.Similarly, there were significantly higher prevalences of pre-existing GI conditions/symptoms (P=0.004),comorbidities (P<0.0001) and history of severe NVP or hyperemesis gravidarum in previous pregnancies (P=0.03) in the HB/RF group compared with the controls.
The multiple linear regression analysis revealed that the presence of HB/RF could predict PUQE scores (P=0.03) as did pre-existing GI conditions/symptoms (P=0.045).Similarly, the How many times in the past 24 hours did you vomit?≥7 How many times in the past 24 hours did you experience gagging, retching or dry heaves? ≥7 PUQE score: _____________ *The PUQE scale is a validated scoring system used to quantify the severity of nausea and vomiting of pregnancy based on quantification of three physical symptoms of nausea and vomiting of pregnancy (nausea, vomiting and retching) for the past 12 hours (21).We have used the modified PUQE scale that quantifies nausea, vomiting and retching for the past 24 hours presence of HB/RF solely accounted for the ability to predict Well-being scores (P=0.02).When only the subset of women experiencing both HB and RF was analyzed, again, the presence of HB/RF solely predicted PUQE scores (P=0.002), and Wellbeing scores (P=0.006).

DiSCUSSiON
Our data demonstrate that women experiencing HB and/or RF in pregnancy experience increased severity of NVP as indicated by validated tools for NVP.Furthermore, assuming women experiencing both HB and RF have increased severity of disease compared with women who have heartburn alone, trends were observed in which more women experiencing both HB and RF classified their NVP as severe compared with control, and compared with women only experiencing either HB or RF.These same trends were observed in both the PUQE scores and the Well-being scores, indicating that more severe HB/RF symptoms result in increased severity of nausea and vomiting.Despite the dearth of studies examining the relationship between HB and RF and severity of NVP, several references recommend treatment of HB/RF during pregnancy to reduce pregnancy complications (9)(10)(11)(12)(13).
Certain potential confounding factors were considered in our multiple linear regression analysis results and demonstrated that the only factor that could consistently account for the increased PUQE scores and the decreased Well-being scores observed in the study group was the presence of HB/RF.Furthermore, the increased use of antiemetics and antacids observed in the HB/RF group confirms the increased severity of NVP, and the presence of HB and/or RF, respectively.
Our study documents for the first time, a correlation between HB/RF and increased severity of NVP.Future controlled studies should determine whether treatment of HB/RF in pregnancy can reduce the severity of NVP.
CONTRiBUTiONS: SG wrote the paper, conceived of the research proposal, collected the data, conducted the analyses and interpreted the results.CM conceived of the research proposal, collected the data, conducted the analyses and revised the paper.GK conceived of the research proposal, analyzed and interpreted the results, and revised the paper.

Figure 1 )
Figure 1) Self-report of severity of nausea and vomiting of pregnancy (NVP) as mild, moderate or severe by women experiencing heartburn (HB), acid reflux (RF), both HB and RF, total disease (including HB, RF and both), or no HB/RF.*P<0.05 compared with control

DiSCLOSURe OF iNTeReSTS AND FUNDiNG:
The Motherisk NVP helpline is supported by an unrestricted grant from Duchesnay Inc, Canada.GK is holder of the Research Leadership for Better Pharmacotherapy during Pregnancy and Lactation (Hospital for Sick Children, Toronto, Ontario) and the Ivey Chair in Molecular Toxicology (Department of Medicine, University of Western Ontario, London, Ontario).The study procedures received ethics approval from the Hospital for Sick Children Ethics Committee, the Research Ethics Board.Ethics approval was obtained on May 22, 2008.The Research Ethics Board application number is 1000012298.
eTHiCS AppROvAL: