Hyperemesis Gravidarum (HG) is a severe and debilitating condition in pregnancy affecting around 0·8–3·3 % of pregnancies(Reference Fiaschi, Nelson-Piercy and Tata1–Reference Einarson, Piwko and Koren4). HG is characterised by nausea and vomiting in pregnancy, distinguishing from common nausea and vomiting in pregnancy by intensity and persistence. In 2021, global consensus on the definition of HG was reached, defining it as ‘nausea and vomiting, with at least one of them being severe; starting before 16 weeks gestation, causing an inability to eat and/or drink normally and strongly limiting daily living activities’(Reference Jansen, Koot and Van’t Hooft5). Until recently, the cause of the condition was suggestive. However, findings by Fejzo et al. confirm the linkage between levels of growth differentiation factor 15, the mother’s sensitivity to it and severe nausea and vomiting in pregnancy(Reference Fejzo, Rocha and Cimino6).
Although few publications have described the nutritional intake of women with HG, it is generally assumed that prolonged nausea and vomiting significantly impair the ability to maintain adequate nutrition and hydration status. Severe weight loss is common in HG and has been associated with gallbladder and liver dysfunction, as well as renal failure(Reference Fejzo, Poursharif and Korst7). In addition to physiological outcomes of HG, psychological health is commonly impaired and affected women frequently report a lack of understanding from healthcare professionals(Reference Beirne, Andrews and Murtagh8). Other adverse outcomes related to HG include low birthweight of the child, small for gestational age and premature delivery(Reference Veenendaal, van Abeelen and Painter9). Additionally, long-term effects of HG to the offspring include an increased risk of a neurodevelopmental disorder, such as autism spectrum disorder(Reference Fejzo, Kam and Laguna10,Reference Nijsten, Jansen and Limpens11) , possibly due to malnutrition in pregnancy during critical time points for fetal brain development(Reference Getahun, Fassett and Jacobsen12).
Given the vital role of nutrition in supporting both maternal health and fetal development, understanding the nutritional intakes of women with HG is of paramount importance. However, there is a paucity of research in this field. Indeed, an international consensus document stressed the importance of understanding and including nutritional intakes in HG in future research(Reference Jansen, Koot and Van’t Hooft13). The primary aim of this research was therefore to compare the energy intakes of women with HG in Ireland to estimated requirements. Additionally, we investigated the intake of macronutrients, micronutrients and fluid in comparison with recommendations. We also aimed to investigate if nutritional intakes of women with HG are related to symptom severity. We hypothesised that the nutritional intakes are lower compared with estimated requirements and that intakes of energy, nutrients and fluid correlates negatively with the 24-h Pregnancy-Unique Quantification of Emesis and Nausea (PUQE-24) scores in total and in parts.
Materials and methods
Participants
Data for this cross-sectional observational study were collected between March 2021 and January 2022 at the day assessment unit (DAU) of the Rotunda Hospital, a large urban maternity hospital in Dublin, Ireland. Eligibility for study participation required women to be over 18 years of age, to understand both written and spoken English and to have a diagnosis of HG. The diagnosis of HG was guided by the 2015 National Clinical Practice Guidelines for Hyperemesis and Nausea and Vomiting in Pregnancy(14), which exclude potential differential diagnoses or predisposing conditions. In accordance with the aforementioned guideline, criteria for a HG-related admission to the DAU included weight loss, signs of dehydration and/or ketonuria as well as moderate or severe nausea and vomiting in pregnancy. This was assessed using the 24 h Pregnancy-Unique Quantification of Emesis and Nausea (PUQE-24) score, which is a validated tool that quantifies the intensity and frequency of nausea, retching and vomiting during the preceding 24 h(Reference Ebrahimi, Maltepe and Bournissen15). A total PUQE-24 score of ≥7 indicates moderate- to- severe HG.
Ethics
Ethical approval was granted by the Rotunda Hospital Research Ethics Committee (REC-2021-005). Informed, written consent was obtained from all participants at study inclusion. The research is adherent with the declaration of Helsinki.
Demographics
Information on age, parity, gestational age, gravida, weight, height, and pre-pregnancy weight was collected by the midwife at the time of inclusion and documented in the electronic medical record, from which the data was subsequently retrieved. Pre-pregnancy weight was self-reported and weight at inclusion was measured upon arrival to the DAU before administration of intravenous (IV)-fluids. Weight at inclusion and pre-pregnancy weight were used to calculate the weight change from before pregnancy until inclusion. Height was measured by the midwife as part of the clinical assessment and recorded in the electronic medical record. Pre-pregnancy BMI as well as current BMI were calculated as weight divided by height in meters squared, using the weights and heights recorded in the electronic chart.
Nutritional intake
Upon arrival to the DAU, participants were provided with a paper-based food and fluid diary to complete. These diaries covered the food and fluid intake during the preceding 24-hours. The women were instructed to be as clear as possible on what they ate and to not include food that was not eaten (i.e. leftovers). Examples of how to report on portion sizes were included in the food and fluid diaries (e.g. ‘1 medium potato, 1 large portion of lasagne, 150 g of salmon, dessert spoon of mayonnaise’). Other suggestions of how to report portion size were to indicate how much of the plate was covered, or to use handfuls or cup measurements. Pictures of a 200 mL disposable cup, a 300 mL mug and a 500 mL water bottle were included to use as references to estimate the fluid intake. Participants were also instructed to write down how the food was cooked. The food and fluid diaries were later analysed using Nutritics software version 5·7. All food and fluid diaries were reviewed and analysed by the first author. One participant with a completely blank food and fluid diary was excluded from the nutritional intake analysis, as it was not possible to determine whether the blank diary reflected no nutritional intake or non-completion of the diary. For participants reporting no or very low intake, written comments in the diary confirmed that they had been unable to eat, and these entries were therefore retained in the analysis. Any decisions on portion sizes, specific product or recipes were documented to ensure all diaries were analysed systematically. If portion size was not specified in a food and fluid diary, an average portion size was chosen using Nutritics software or by using the portion size specified on the product website (e.g. ‘bread’ was inputted as a medium slice (36 g) of bread (white average) with 7 g of salted butter per reported bread slice). Data on analysed intakes of total fluid (from food as well as drinks) and energy as well as macro- and micronutrient intake were then collected from the Nutritics report of the analyses. Data on micronutrient supplements were not captured in the food and fluid diaries.
Nutritional requirements
Nutrient intakes were compared with European Food Safety Authority (EFSA) dietary reference values because these provide harmonised European nutrient reference values and are consistent with the Irish clinical practice guideline on nutrition in pregnancy, which draws on EFSA recommendations(16). Weight at inclusion was used to calculate estimated energy and protein requirement. Basal metabolic rate was calculated with the Henry Oxford equation(Reference Henry17). The calculated basal metabolic rate was multiplied by a conservative physical activity level of 1·4. An additional 5 %, 10 % or 25 % was added for first, second or third trimester(Reference Butte and King18). Based on population reference intake (PRI) guidelines from the EFSA, protein requirements were calculated as 0·83 g/kg body weight, plus an additional 1, 9 or 28 g in first, second or third trimester(19). No further adjustments were made for one woman carrying multiple fetuses, since she was pregnant in the first trimester. The minimum requirements of carbohydrates were 175 g/day,(20) and requirements of fluid during pregnancy were 2300 mL/day in accordance with EFSA adequate intake guidelines(21). A fibre intake of 25 g/day was considered adequate in alignment with EFSA recommendations for adults(20). Thiamine PRI for pregnant women, age 30–39, with a physical activity level of 1·4, in the first trimester was 0·8 mg/day (0·77 + 0·03)(22) and the Na value of 2000 mg is the value considered by EFSA to be ‘safe and adequate’(23). Other daily requirements of micronutrients were according to EFSA recommendations for intakes during pregnancy(24–32). PRI was used or adequate intake if PRI was not available.
Severity of Hyperemesis Gravidarum
The severity of HG was assessed using PUQE-24 score reflecting nausea, vomiting and retching during the preceding 24 h. The PUQE-24 score includes three questions relating to nausea, retching and vomiting, with each question scoring between one to five. The three questions are summed, and the total PUQE-24 score, between three to 15, reflects symptom severity (≤6: mild, 7–12: moderate and ≥13: severe). The PUQE-24 score also includes a separate question on general well-being, rated from 0 to 10. The PUQE-24 score was assessed by the midwife upon arrival to the DAU and recorded in the patient’s electronic chart, from where it was later collected.
Statistical analysis
Statistical power was calculated for the study’s primary outcome: the difference between energy intake and energy requirements. To achieve 80 % power at a 95 % CI, a minimum of eight participants was required to detect a difference between a mean energy intake of 1500 kcal/day and a mean energy requirement of 2000 kcal/day (sd 500 kcal/day). An intake of 1500 kcal/day was chosen as a plausible conservative estimate for women with HG. Although lower than typical pregnancy energy requirements, it remains higher than intakes reported in some previous HG cohorts(Reference Birkeland, Stokke and Tangvik33,Reference Maslin, Dean and Shawe34) . Actual intakes may therefore be lower, which would reduce the required sample size. For descriptive purposes, frequencies of categorical variables are presented. Additionally, normally distributed continuous variables are presented with mean and sd, and non-normally distributed variables are presented with median and quartiles (25th–75th). The differences in median intakes and estimated requirements for energy and protein were assessed using a Wilcoxon signed-rank test for paired samples, as these are individual-level requirements. For other nutrients with fixed reference values, one-sample Wilcoxon signed-rank tests were used. Estimated intakes were compared with PRI if available or adequate intake if PRI was not available. The correlations between nutrients and dimensions of the PUQE-24 score as well as total PUQE-24 score were assessed using Spearman’s correlation. Women were also divided into two groups; one group with mild to moderate PUQE-24 score (≤12) and one group with severe PUQE-24 score (≥13). Participants with PUQE-24 scores reflecting mild to moderate were grouped together due to the low number with a mild PUQE-24 score. The median intakes of nutrients were compared between groups using a Mann–Whitney U test. Analyses involving PUQE-24 score were restricted to participants with complete PUQE-24 score data, as some women had missing values for this measure. Statistical significance was considered as p < 0·05. Multiple test correction was considered. However, due to the small sample size and the risk of increasing type II error, unadjusted p values are presented. Statistical analyses were carried out using SPSS version 28.0.1.0 and R version 4.5.3.
Results
Demographics
A total of thirty-seven women were recruited. One participant was excluded due to not completing a food and fluid diary, leaving thirty-six women included in the final analysis. The mean (sd) age of the participants were 30·7 (5·5) years. Most participants (n 35, 97·2 %) had a singleton pregnancy, and the majority were pregnant in their first trimester (n 24, 66·7 %). The mean (sd) gestational age was 12·9 (7·0) weeks. Of women with available data for weight change (n 22), seventeen women (77 %) reported weight loss from before pregnancy until inclusion. Most participants had a previous pregnancy and/or childbirth. Demographics of the thirty-six included women are presented in Table 1.
Demographics and descriptives of thirty-six study participants
Table 1. Long description
A table with 25 rows and 4 columns summarizing demographic and descriptive data of thirty-six study participants. The columns are labeled Number of participants with available data, Mean, Median or percent, and sd or IQR as appropriate. The rows include Age (years), Height (cm), Pre-pregnancy weight (kg), Pre-pregnancy BMI kg/m2, Weight at inclusion (kg), Weight change from pre-pregnancy until inclusion (kg), Gestation (weeks), Vomiting score, Nausea score, Retching score, Total PUQE-24 score, Wellbeing score, Multiple pregnancy, Trimester, Gravida, and Parity. Each row provides specific data points for the respective demographic or descriptive category.
IQR, inter quartile range; PUQE-24, 24-hour Pregnancy-Unique Quantification of Emesis and Nausea.
a) Pre-pregnancy weight (kg) was self-reported.
*Mean.
†Median.
‡%.
Estimated dietary intakes compared with estimated requirements
Most of the women consumed less than the estimated requirements for all studied macronutrients and fluid. Specifically, thirty-one participants (93·9 %) did not meet their individually estimated energy requirements, twenty-six participants (78·8 %) did not meet their individually estimated protein requirements, twenty-eight participants (77·8 %) did not meet the minimum carbohydrate requirement and thirty participants (83·3 %) did not meet the fluid recommendation. None of the women consumed the recommended amount of fibre. One woman reported only water intake in the preceding 24 h; hence, this woman consumed no energy or nutrients. The median (interquartile range) energy intake of the participants was significantly lower compared with estimated requirements (1050 (481–1461) kcal v. 2057 (1978–2332) kcal (p < 0·001)). The median intakes of macronutrients and fluid were significantly less than estimated requirements (p < 0·01) (Table 2).
Dietary intakes of thirty-six women with Hyperemesis Gravidarum compared with requirements
Table 2. Long description
A table comparing dietary intakes of women with Hyperemesis Gravidarum to estimated requirements. The table has 17 rows and 7 columns. Column headers are Estimated intakes/day (Median), Estimated intakes/day (IQR), Estimated requirements/day (PRI unless otherwise stated), Estimated requirements/day (PRI unless otherwise stated), P, and % not meeting requirements/day. Row labels include Energy (kcal), Protein (g), Carbohydrate (g), Fibre (g), Fat (g), Na (mg), Potassium (mg), Ca (mg), Fe (mg), Iodine (µg), Vitamin A (retinol equivalents), Vitamin D (µg), Thiamine (B1) (mg), Folates (B9) Total (µg), Vitamin B12 (µg), Vitamin C (mg), and Fluid (mL). Each row lists the estimated intakes, interquartile ranges, and requirements for each nutrient and fluid, along with the percentage of women not meeting the requirements.
IQR, interquartile range; PRI, population reference intake.
P values were calculated using Wilcoxon signed-rank tests. For energy and protein, intakes were compared with individually calculated requirements; for other nutrients, intakes were compared with fixed reference values.
a) All estimated requirements per day were according to the European Food Safety Authority (EFSA) guidelines for pregnant women.
b) Energy and protein requirements were individually calculated; percentages are based on participants with complete data for both intake and requirement (n 33).
c) This recommendation includes supplemental folic acid.
*Median (IQR) of calculated requirements among thirty-three participants with available data.
†Adequate intake (AI) instead of PRI.
None of the women consumed the recommended amount of potassium, Fe, iodine, vitamin D or folate (n 36, 100 % below requirements for each). Most of the women did not meet the requirements for Ca (n 35, 97·2 %), vitamin B12 (n 31, 86·1 %), vitamin A (n 30, 83·3 %), vitamin C (n 29, 80·6 %) and thiamine (n 22, 61·1 %). The median intakes of micronutrients were significantly lower than requirements for all nutrients studied (p < 0·05) (Table 2).
Estimated dietary intakes in relation to symptom severity
Eighty-one percent of participants (n 29) completed the PUQE-24 score. One woman (3 %) reported a PUQE-24 score corresponding to mild symptoms, likely due to antiemetics being effective on the day of data collection. The majority (n 19, 66 %) of women reported a PUQE-24 score corresponding to moderate symptoms and nine women (31 %) reported a severe PUQE-24 score. Spearman’s correlations between nutrients and dimensions of the PUQE-24 score as well as total PUQE-24 score and well-being score can be seen in Figure 1. For nausea, significant negative correlations were observed for energy (r s = −0·45, p = 0·015), protein (r s = −0·44, p = 0·018), carbohydrate (r s = −0·48, p = 0·009) and fat intake (r s = −0·44, p = 0·016). For vomiting, significant negative correlations were observed for energy (r s = −0·54, p = 0·003), protein (r s = −0·46, p = 0·012), carbohydrate (r s = −0·48, p = 0·008) and fat intake (r s = −0·60, p < 0·001). No significant correlations were observed between these macronutrients and retching or wellbeing score. Total PUQE-24 score was significantly negatively correlated with carbohydrate intake (r s = −0·38, p = 0·042).
Spearman’s correlations between nutrients and 24-h Pregnancy-Unique Quantification of Emesis and Nausea (PUQE-24) score among women with Hyperemesis Gravidarum (n 29). *Represents p values <0·05. **Represents p values <0·01.
Figure 1. Long description
A matrix showing correlation values between different nutrients and symptoms among women with Hyperemesis Gravidarum. The matrix has 17 rows and 6 columns, with column labels including Nausea, Vomiting, Retching, Total PUQE-24 score, and Well-being, and row labels including Energy (kcal), Protein (g), Carbohydrate (g), Fibre (g), Fat (g), Sodium (mg), Potassium (mg), Calcium (mg), Iron (mg), Iodine (µg), Vitamin A (retinol equivalents), Vitamin D (µg), Thiamine (B1) (mg), Folates (B9) Total (µg), Vitamin B12 (µg), Vitamin C (mg), and Fluid (mL). The matrix is color-coded with a gradient from blue to red, indicating correlation values ranging from -1 to 1. Notable trends include strong negative correlations between Vomiting and Fat (g) (-0.60), and between Vomiting and Calcium (mg) (-0.64). The matrix also shows significant negative correlations between Nausea and various nutrients such as Carbohydrate (g) (-0.48) and Iron (mg) (-0.54). Positive correlations are observed between Retching and Vitamin D (µg) (0.35), and between Well-being and Fluid (mL) (0.34).
Some micronutrients showed significant negative correlations with nausea and vomiting scores. Specifically, nausea was significantly negatively correlated with Na (r s = −0·47, p = 0·010), potassium (r s = −0·44, p = 0·017), Ca (r s = −0·51, p = 0·005), Fe (r s = −0·54, p = 0·002) and thiamine intake (r s = −0·40, p = 0·031). Vomiting was significantly negatively correlated with Na (r s = −0·54, p = 0·003), potassium (r s = −0·42, p = 0·023), Ca (r s = −0·64, p < 0·001), Fe (r s = −0·43, p = 0·020), iodine (r s = −0·46, p = 0·012), vitamin D (r s = −0·57, p = 0·001) and vitamin B12 intake (r s = −0·60, p < 0·001). Total PUQE-24 score was significantly negatively correlated with Ca intake (r s = −0·41, p = 0·026). No nutrients were significantly correlated with retching or well-being scores.
Differences in nutrient intakes of women with a mild or moderate total PUQE-24 score compared with those with a severe PUQE-24 score are presented in Table 3. Carbohydrate intake was significantly lower in the group with severe PUQE-24 score. However, for other nutrients, while generally lower intakes in the severe group could be seen, these differences did not reach statistical significance.
Differences in nutrient intakes among 24-h pregnancy-unique quantification of emesis and nausea (PUQE-24) score groups (n 29)
Table 3. Long description
A table comparing nutrient intakes among different PUQE-24 score groups. The table has 17 rows and 6 columns. The columns are labeled as follows: Nutrient, Mild/Moderate (Median) (n 20), Mild/Moderate (IQR) (n 20), Severe (Median) (n 9), Severe (IQR) (n 9), and P. The rows list different nutrients and their intake values. Row 1: Energy (kcal), 1216, 638–1567, 460, 358–1176, 0.06. Row 2: Protein (g), 40, 15–52, 17, 9–60, 0.34. Row 3: Carbohydrate (g), 129, 87–172, 68, 43–129, 0.03. Row 4: Fibre (g), 8, 5–12, 5, 2–8, 0.07. Row 5: Fat (g), 53, 26–77, 30, 14–44, 0.06. Row 6: Na (mg), 1313, 689–2111, 816, 413–1194, 0.10. Row 7: Potassium (mg), 1541, 526–2078, 802, 164–1668, 0.32. Row 8: Ca (mg), 493, 293–604, 252, 178–564, 0.20. Row 9: Fe (mg), 4.9, 2.3–8.9, 2.7, 1.1–5.7, 0.22. Row 10: Iodine (µg), 43, 22–74, 25, 11–103, 0.66. Row 11: Vitamin A (retinol equivalents), 309, 178–615, 272, 121–1328, 1.00. Row 12: Vitamin D (µg), 0.3, 0.1–1.7, 0.1, 0.1–0.6, 0.39. Row 13: Thiamine (B1) (mg), 0.6, 0.4–1.2, 0.7, 0.2–1.2, 0.47. Row 14: Folates (B9) Total (µg), 112, 49–186, 123, 35–230, 0.76. Row 15: Vitamin B12 (µg), 1.5, 0.5–3.5, 0.7, 0.0–2.9, 0.39. Row 16: Vitamin C (mg), 29, 15–82, 21, 3–138, 0.66. Row 17: Fluid (mL), 1218, 907–1867, 787, 560–1565, 0.14.
P values were calculated using the Mann–Whitney U test.
* A PUQE-24 score of ≤12 was considered mild to moderate and ≥13 severe.
Discussion
Summary of results
Among women with HG, we found that the median intake of energy as well as macronutrients and micronutrients were significantly lower than estimated requirements. We also found that intakes of several nutrients, including energy, macronutrients, Na, potassium, Ca, Fe, iodine, vitamin D, thiamine and vitamin B12, were significantly negatively correlated with nausea and/or vomiting dimensions of the PUQE-24 score. Although not significant, we saw lower intakes of nutrients among women with severe PUQE-24 scores as compared with women with low to moderate PUQE-24 scores.
Impact of maternal nutrition on fetal and child health
Optimal maternal nutrition during pregnancy is essential for promoting healthy fetal growth and development, and inadequate nutrition can adversely affect both mother and infant. Inadequate weight gain, as a result of inadequate energy intake and excessive losses, is common in HG(Reference Meinich and Trovik35,Reference Doherty, Killeen and Sheehy36) . Among women with HG and inadequate weight gain, small for gestational age occurs in 16·7 % of infants, compared with 4·2 % of those with adequate weight gain(Reference Meinich and Trovik35). These statistics demonstrate that it is likely the malnutrition, as a consequence of HG, that drives poor intrauterine growth. In non-HG pregnancies, malnutrition can lead to long-term health consequences for both the mother and the infant, including poor fetal growth and an increased risk of chronic diseases later in life. Barker’s ‘fetal origins hypothesis’ suggests that inadequate nutrition in the womb predisposes individuals to conditions such as cardiovascular disease (CVD), diabetes and hypertension in adulthood(Reference Barker37). The Dutch Hunger Winter of 1944–1945 provided key evidence, showing that children exposed to famine in utero had higher rates of metabolic disorders and lower birth weights, reinforcing the critical role of maternal nutrition during pregnancy(Reference Roseboom, van der Meulen and Ravelli38). Despite the biological plausibility of malnutrition as a result of HG leading to cardiometabolic disturbances among offspring in later life, a recent systematic review did not find consistent evidence for an association(Reference Nijsten, Jansen and Limpens11). Further research, specifically with long-term follow-up of infants exposed to HG, is required to fully elucidate the impact of the condition on body composition, CVD and diabetes risk.
Dietary intakes in our cohort compared with other Hyperemesis Gravidarum cohorts
Our findings support previous findings that women with HG have inadequate nutritional intakes. The low-energy intakes observed in this and other cohorts may be partly explained by the underlying aetiology of HG. Evidence implicates fetally derived growth differentiation factor fifteen and maternal sensitivity to it, with activation of brainstem pathways involved in nausea, vomiting and aversive responses(Reference Fejzo, Rocha and Cimino6). This may help explain the food aversion, reduced appetite and inability to meet nutritional requirements observed in women with HG, reinforcing that inadequate intake reflects the underlying pathophysiology of the condition rather than poor dietary choice or lack of effort. Van Stuijvenberg et al. reported even lower energy intakes (443 kcal) compared with our study (1050 kcal), along with lower intakes of all comparable nutrients(Reference van Stuijvenberg, Schabort and Labadarios39). Notably, their study population included women living in areas of low socio-economic status, with 45 % having a BMI of <20. In another study by Birkeland et al., the dietary intakes were comparable with those of our group. However, the methodologies of the two studies differ(Reference Birkeland, Stokke and Tangvik33). Birkeland et al. compared the dietary intakes of women with HG against those of pregnant controls without HG, finding that all nutrients studied were significantly lower in women with HG compared with the controls. In contrast, our study compared the intakes of women with HG to dietary recommendations/estimated requirements. A study by Maslin et al. compared dietary intakes of women of different HG categories and overall, to dietary recommendations and found, in line with our findings, that women with HG have significantly lower intakes of most nutrients compared with requirements(Reference Maslin, Dean and Shawe34). Another study by Fatma et al. found that antioxidant intake among women with HG was significantly lower compared with controls(Reference Celik, Guzel and Kuyumcuoglu40). A study conducted in the UK by Crozier et al.(Reference Crozier, Inskip and Godfrey41) found poorer diet scores among women with severe nausea and vomiting in pregnancy compared with their pre-pregnancy diet. While these results are not directly comparable to ours, as they reported differences in diet scores and consumption of food groups, they suggest that the diet quality of women with HG is poor, which aligns with our findings. Although previous studies on the topic are heterogeneous in methodology and context, the conclusion that women with HG consume inadequate amounts of nutrients holds.
To our knowledge, no previous study has compared fluid intakes of women with HG to recommendations. Ensuring adequate hydration is crucial for maintaining maternal health, preventing complications and improving the overall well-being of pregnant women. Dehydration can cause venous thromboembolism, electrolyte imbalances, fatigue, dizziness and headaches, reducing the quality of life for pregnant women. The hypercoagulable state of pregnancy, in combination with dehydration, significantly increases the risk of venous thromboembolism, making it a critical condition to prevent and manage effectively(Reference Bukhari, Fatima and Barakat42). Indeed, the leading cause of pregnancy related death in UK and Ireland continues to be venous thromboembolism(Reference Felker, Patel and Kotnis43). Our findings of inadequate fluid intakes among women with HG, regardless of severity, highlight the need for access to IV- fluids and tailored dietary advice.
Electrolyte imbalances can result from dehydration, insufficient intake or the loss of electrolytes through vomiting. The most common electrolyte imbalance in HG is hypokalemia(Reference Maslin and Dean44), commonly caused by inadequate intakes or losses of potassium through vomiting. Hypokalaemia can lead to severe outcomes such as cardiac arrest, stressing the importance of an adequate intake(Reference Walch, Duke and Auty45). Our findings highlight that women with HG does not have an adequate intake of potassium. Therefore, potassium levels should be monitored and, if necessary, supplemented to prevent severe complications following hypokalaemia.
Dietary intakes compared with severity of Hyperemesis Gravidarum
We used the PUQE-24 score to assess nausea and vomiting severity in pregnancy. Although the score is a validated and widely used tool(Reference Ebrahimi, Maltepe and Bournissen15), it should be noted that it does not necessarily capture the full clinical burden of HG. PUQE-24 reflects nausea, vomiting and retching during the preceding 24 h, but does not fully account for other clinically important indicators of HG severity, such as weight loss, dehydration, electrolyte disturbances, nutritional status, duration of symptoms, functional impairment or quality of life. However, PUQE-24 score was deemed appropriate for the purpose of this study as it captures symptom severity during the same 24-h period covered by the food and fluid diaries. Our research adds to the few studies that have assessed dietary intakes by severity of HG symptoms. We found that energy intake was significantly correlated with the vomiting and nausea dimensions of the PUQE-24 score. Although we did not reach statistical significance for the correlation between energy intake and total PUQE-24 score a negative relationship was seen which is consistent with the findings of Birkeland et al. who found that energy intake was significantly negatively correlated with PUQE-24 score(Reference Birkeland, Stokke and Tangvik33). Similarly, Maslin et al. found that increasing symptom severity, in a UK-based cohort, led to lower dietary intakes(Reference Maslin, Dean and Shawe34). However, they compared intakes between mild, moderate and severe PUQE-24 score groups, which we could not do due to a limited number of participants with low PUQE-24 scores. Consequently, our group is more homogeneous in PUQE-24 scores compared with Maslin et al., potentially accounting for some differences in findings. The small number of women in the severe PUQE-24 group in our cohort may also have also limited statistical power for group comparisons by symptom severity. Accordingly, non-significant subgroup differences should not be interpreted as evidence of no clinically meaningful difference. The findings should be confirmed in larger cohorts.
Ebrahim et al. have previously explored the relationship between PUQE-24 scores and fluid intake. They found a relationship between well-being scores and fluid intake, but not between fluid intake and other parts of the PUQE-24 score or total PUQE-24 score(Reference Ebrahimi, Maltepe and Bournissen15). The fluid intakes of the women in that study were based on the reported amounts of cups of liquid the woman had been drinking during the preceding 24 h, and the assumption that one cup was equal to 200 mL. This differs from the method in this present study where women were asked to estimate the amount of fluid she had been drinking on each occasion. Our study also measured fluid intake from foods as well as fluids, meaning the measured fluid intake is possibly more detailed in our study. However, we did not find a significant correlation between fluid intakes and well-being score or PUQE-24 score. Regardless of PUQE-24 score, the majority of women with HG are at risk of dehydration and will likely require IV fluids for rehydration.
Challenges and strategies for creating nutritional interventions for women with Hyperemesis Gravidarum
Creating nutritional interventions for women with HG presents several challenges, including the severity and unpredictability of nausea and vomiting, which often make oral intake difficult. Individual variability in symptoms, limited research on effective treatments and the need for tailored, flexible solutions further complicate intervention development. Additionally, a lack of awareness among healthcare providers and the psychological burden on patients can hinder timely and effective nutritional management(Reference Beirne, Andrews and Murtagh8,Reference Doherty, McHale and Killeen46) .
Given the associations between HG, nutritional and hydration deficiencies and poor pregnancy outcomes, a proactive approach to nutritional management is necessary. Nutritional interventions for women with HG should focus on maintaining hydration and addressing nutrient deficiencies through tailored dietary modifications, supplementation and, in severe cases, enteral or parenteral nutrition. The low intake of key micronutrients observed in this cohort highlights the need to assess tolerance of standard oral supplementation and, where clinically appropriate, consider alternative formulations or routes of administration. Given the need for tailored dietary advice, the inclusion of a dietitian in the multidisciplinary team managing HG is paramount. Qualitative research has shown that dietitians play a significant role in supporting women with HG, offering more than just nutrition advice(Reference Beirne, Andrews and Murtagh8,Reference Groleau, Benady-Chorney and Panaitoiu47) . They provide validation, information and reassurance. It is suggested that a strong patient–clinician relationship forms, as dietitians are well-placed to address many of the patient’s main concerns, including dietary issues and coping strategies for physical discomfort.
Strengths and limitations
To the best of our knowledge, our study is the first to study nutrient and fluid intakes among women with HG in an Irish setting. Additionally, it is the first study to explore fluid intakes in comparison to recommendations, among women with HG. The 24-h recall method was chosen for its ability to capture the dietary intake of women with HG at that exact moment, reflecting the changeable nature of the condition. The approach is patient centred and enhances the completeness of data collection. Unlike a Food Frequency Questionnaire, which measures habitual intake over a longer period (e.g., 1 year) and may not capture specific foods chosen by women suffering from HG, the 24-h recall provides a more accurate snapshot of their current dietary habits.
Our study has some limitations. When calculating energy requirements, a physical activity level value was assumed rather than measured, with a conservative low physical activity level chosen due to the debilitating nature of HG. However, given that HG is often associated with severe fatigue and reduced activity, actual physical activity levels may have been lower in some participants, potentially leading to overestimation of energy requirements. Future studies should measure or estimate physical activity levels directly to improve the precision of individual energy requirement calculations. Physical activity levels in women with HG have not been explored in this or previous studies.
Anti-emetic medication use could not be reliably accounted for in the analyses. Although prescribed anti-emetic medications were recorded, it was not known whether these medications were taken, or when, and data on dose and adherence were unavailable.
Micronutrient supplement use, including commonly recommended supplements in pregnancy and HG, was not captured accurately in the food and fluid diaries and were therefore not included in the data analysis. This may have led to underestimation of total micronutrient intake and may error conclusions regarding the proportion of women not meeting requirements for some nutrients, particularly those commonly consumed as supplements. Future studies should systematically record supplement type, dose, frequency, formulation and route of administration to better estimate total micronutrient exposure in women with HG. Additionally, vomiting of specific food items was not accounted for in the food diary analysis, which means the nutrition uptake may have been even lower than calculated intakes. Our estimates may therefore be conservative and further reinforce the concern that women with HG are at high risk of inadequate nutritional intake and malnutrition.
The generalisability of our findings may be constrained by the fact that data collection was conducted exclusively within an urban Irish setting. Further research in diverse geographical and socio-demographic contexts is warranted to enhance the external validity and applicability of these results to broader populations. Furthermore, the relatively small sample size may have limited the robustness of our results for some micronutrients. Additionally, as multiple nutrient and fluid comparisons were conducted without adjustment for multiple testing, some statistically significant findings, particularly those with p values close to 0·05, may represent chance findings. These findings should therefore be interpreted cautiously and confirmed in larger studies. Lastly, the small number of women in the severe PUQE-24 group also limited statistical power for subgroup comparisons by symptom severity. Therefore, non-significant differences between PUQE-24 groups should be interpreted cautiously. Future research with larger sample size is therefore warranted.
Conclusion
Our study confirms that women with HG consume inadequate amounts of energy, fluids and key macro- and micronutrients during pregnancy. Although mean intakes of nutrients between PUQE-24 score groups generally were not significantly different, many nutrients correlated negatively with nausea and vomiting score as well as total PUQE-24 score. These findings highlight the importance of routine nutrition assessment and individualised nutritional interventions to improve nutritional status among women with HG.
Data availability statement
Data is available from corresponding author upon reasonable request.
Acknowledgements
We wish to thank all participants who generously gave their time to contribute to this study. Additionally, we are grateful to the staff at the Rotunda Hospital for their dedicated assistance with data collection.
Financial support
NA.
Authors contributions
E.O.B. and L.K. were involved in the conceptualisation and design. Data collection was carried out by A.C. E.O.B. and H.A. contributed with supervision. Analysis was performed by M.S. and E.O.B. M.S. drafted the first version of the paper. All authors edited and revised critically. All authors approved the final version.
Competing interests
EOB is a board member of Hyperemesis Ireland, charity, unpaid. All other authors report no conflict of interest.
Ethical standards
Ethical approval has been granted by the Rotunda Hospital Research Ethics Committee (REC-2021-005).
Consent
Informed written consent was obtained from all participants.
Open access
