The Serum Platelet-Albumin Ratio in Hyperemesis Gravidarum: A Retrospective Comparative Study

Introduction

Hyperemesis gravidarum (HG) is a clinical entity characterized by severe and prolonged nausea and/or vomiting that occurs during pregnancy. It occurs in approximately 0.3–2% of pregnancies and can lead to dehydration, ketonuria, ketoacidosis, and weight loss.^1,2 It is also one of the most common reasons for hospitalization in the first half of pregnancy, typically occurring between the fifth and tenth weeks of gestation, with the majority of cases improving by mid-pregnancy. However, symptoms persist until the third trimester in 20% to 25% of cases, and until delivery in 5%.^3,4

The etiology and pathogenesis of HG have not been fully elucidated; however, several risk factors have been proposed for the development of HG. Increased placental load, as seen in molar pregnancies and multiple gestations, has been associated with a higher risk of HG. Additionally, the consumption of estrogen-containing medications, hyperthyroidism, diabetes, and the presence of a female fetus are also considered risk factors. Studies have reported that women with a family history of HG among close relatives are also at a higher risk.¹

The development of HG is attributed to hormones such as human chorionic gonadotropin (hCG), estrogen, progesterone, leptin, placental growth hormone, prolactin, thyroxine, and adrenocortical hormones. Numerous studies have reported a connection between HG and hormones produced by the placenta and genes expressed in it (GDF15, IGFBP7, and PGR). For example, hCG is a hormone secreted by the placental cytotrophoblast cell layer that is associated with fetal growth and various placental, uterine, and fetal functions, serum levels peaking during the 12th to 14th weeks, which coincides with the period of most severe HG symptoms. hCG is therefore considered an important factor in the pathogenesis of HG.⁵

The platelet-to-albumin ratio (PAR) is a novel inflammatory marker reflecting both systemic inflammation and nutritional status. It has been investigated in various cardiovascular, oncological, and inflammatory conditions. However, evidence regarding its role in obstetric disorders remains limited, and data on its association with hyperemesis gravidarum are scarce. Therefore, this study aimed to investigate the relationship between PAR and HG and to evaluate its potential clinical utility.

Materials and Methods

This retrospective comparative study was conducted at Kepez State Hospital Gynecology and obstetrics clinics, Türkiye, between 1 January, 2024, and 31 December, 2025. One hundred sixty women were included in the study (group 1, control n=80, and group 2, hyperemesis gravidarum n=80). The study approved by the local ethical committee (Antalya Training and Research Hospital Ethical Committee, reference number: 2025/487) and it has been conducted in accordance with the principles of the Declaration of Helsinki. Informed consent forms were obtained at the time of admission to hospital for future use.

Hyperemesis gravidarum was diagnosed based on clinical criteria, including persistent nausea and vomiting leading to dehydration, ketonuria, electrolyte imbalance, and weight loss exceeding 5% of pre-pregnancy weight. Patients requiring hospitalization for intravenous fluid therapy and clinical management were classified as having HG.

Healthy women aged between 18 and 35, in the first 20 weeks of pregnancy, with no chronic diseases that might adversely affect fetal development (such as hypertension, diabetes mellitus, and migraines), and with no missing data in the hospital automation system were included in the study. The exclusion criteria included the presence of additional chronic diseases (such as pregestational diabetes and chronic hypertension), obesity (BMI > 35 kg/m²), the detection of anomalies in fetal screening tests, and missing data in the hospital automation system. Only patients requiring hospitalization were included in the HG group, and mild cases managed on an outpatient basis were excluded.

The participants’ sociodemographic data recorded in the database included age at first hospital visit, BMI, personal history, family history, smoking habits, alcohol consumption, and medication habits. Additionally, details concerning obstetric history, gestational age, and were included in the research, together with laboratory.

The evaluation of blood parameters collected at the first hospital visit, including complete blood count, hematocrit, leukocyte count, neutrophils, lymphocytes, platelets, free T4, albumin, ALT, AST, BUN, creatinine, hs-CRP, sodium, and potassium, was performed using a Coulter LH – 750 device (Beckman Coulter, Brea, CA, USA).

Statistical Analysis

SPSS (Statistical Package for the Social Sciences) version 27.0 software was used for statistical analysis. The Kolmogorov Smirnov test was applied to evaluate the normality of distribution of continuous variables. The t test was applied for the analysis of data exhibiting normal distribution, and the Mann Whitney U-test for non-normally distributed data. The chi-square and the Fisher exact tests were used to compare categorical parameters. Categorical measurements were expressed as frequencies and percentages, and continuous measurements as means and standard deviations (with median and 25th-75th percentiles where necessary). Receiver operating characteristics (ROC) analysis was used to determine the predictive value of PAR. A significance level of <0.05 was imposed for all tests.

Results

Twenty-eight of the 188 participants initially evaluated between 1 January, 2023 and 31 December, 2025, were excluded due to ineligibility. The remaining 160 were divided into two groups, Group 1, healthy pregnancies (n=80) and Group 2, women with HG (n=80). All 160 were evaluated in the final analysis, as shown in Figure 1.

Figure 1 Enrollment and follow-up of the study participants.

The participants’ sociodemographic characteristics and clinical features are shown in Table 1. No difference was observed between the groups in age (28.45±5.31 vs 27.48±4.78, p=0.226), BMI (24.31±4.17 vs 23.41±4.64, p=0.194), smoking status [9 smokers (11.3%) vs 15 (18.8%), p=0.268], alcohol consumption [3 (3.8%) vs 7 (8.8%), p=0.328], or gestational age [11.5±3.2 vs 10.6±2.6, p=0.115]. However, gravidity [2.0 (1.0–3.0) vs 1.0 (0–1.0), p<0.001] and parity [1.0 (1.0–2.0) vs 1.0 (1.0–1.0), p<0.001] differed significantly at the p<0.001 level.

Table 1 The Participants’ Sociodemographic Characteristics

The participants’ laboratory results are given in Table 2. While serum hemoglobin, Htc, WBC counts, NLR, PLR, PIV, free thyroxine, alanine aminotransferase, aspartate aminotransferase, hs-CRP, and sodium levels were comparable between HG and healthy pregnancies, significant intergroup differences were observed in terms of TSH (1.64±0.53 vs 1.20±0.39, p =0.006), albumin (42.69±5.72 vs 39.59±8.56, p<0.001), and potassium levels (3.84±0.45 vs 3.58±0.60, p=0.010), and PAR (6.15±1.49 vs 9.26±2.89, p<0.001).

Table 2 The Participants’ Laboratory Values

The recommended cut-off value for PAR was found to be ≥6.51 (AUC = 0.825, 95% CI 0.701–0.794, sensitivity 82.1%, specificity 66.2%) in the ROC analysis (Figure 2). The positive predictive value (PPV) of PAR was 80.9%, while the negative predictive value (NPV) was 79.1%.

Figure 2 ROC curve for PAR.

Discussion

The main finding of this study is that PAR levels were significantly higher in pregnant women with HG compared to healthy controls, whereas other commonly used inflammatory markers such as NLR, PLR, and PIV did not differ significantly. These findings suggest that PAR may be a more sensitive indicator of the inflammatory and nutritional alterations associated with HG. Approximately 50–80% of pregnant women may experience nausea and vomiting, which can lead to fluid-electrolyte imbalance, weight loss, and nutritional deficiencies in those with HG. Symptoms of HG typically begin before 12^th week of pregnancies and may resolve by the 20th week of pregnancy. In some cases, women with HG may experience persistent episodes of nausea and vomiting. This condition can lead to severe morbidities, such as dehydration and vitamin K deficiency, and may even result in life-threatening conditions such as Wernicke’s encephalopathy and central pontine myelinolysis due to thiamine deficiency.^1–3

There is currently no specific biomarker for diagnosis and monitoring of treatment in HG. This first retrospective comparative study in the literature aimed to evaluate whether changes occur in PAR and other inflammatory markers in pregnant women with HG. The results indicated that there was no difference between HG and healthy pregnant women in terms of sociodemographic, clinical, or obstetric characteristics, as well as serum blood parameters indicating inflammation. However, PAR was significantly higher in women with HG.

Hyperemesis gravidarum is an urgent obstetric problem with a frequently occurrence, particularly during the first half of pregnancy, the pathophysiology and etiology of which are not fully understood. Factors implicated in HG include nutritional deficiencies, immunological causes, abnormally elevated levels of hCG, estrogen, progesterone, free thyroxine hormone, and PGE2, as well as Helicobacter pylori infection. Serum hCG levels peak at 11–13 weeks of gestation, and symptoms of HG are also most commonly observed during that week. Due to the similarity between the alpha subunit of hCG and TSH, hCG increases thyroid hormone synthesis, thereby exacerbating the symptoms of HG.^6,7 Elevated levels of estrogen during pregnancy directly affect the central nervous system and slow gastric emptying by delaying passage from the stomach to the duodenum, while progesterone relaxes smooth muscles, reducing motility. These combined factors contribute to the development of HG.⁶ Studies have shown that H. pylori infection, which is closely associated with peptic ulcers, also plays a role in HG.^2,7 In this study, while serum TSH levels were lower in women with HG, no significant difference was found in T4 hormone levels.

Previous studies have shown that HG is associated with adverse maternal complications such as preterm labor, hypertensive disorders of pregnancy, placental abruption, as well as negative perinatal outcomes including miscarriage, prematurity due to preterm labor, respiratory distress syndrome, and fetal growth restriction.^8,9 Timely and effective medical treatment for HG is crucial in mitigating these conditions that exacerbate maternal and fetal morbidity and mortality. If appropriate medical interventions are implemented, both maternal and fetal morbidity and mortality can be reduced, thereby lowering personal, financial, and emotional burdens, as well as the socioeconomic impact on national economies.

The first line of treatment in HG is lifestyle modification, while medical treatment aims to reduce nausea and vomiting, improve quality of life, maintain fluid-electrolyte and acid-base balance, and replenish deficient vitamins and minerals to prevent adverse maternal and fetal outcomes.^10,11 Increased levels of certain inflammatory markers (IL-6 and CRP) associated with HG in pregnant women suggest that inflammation may play a role in the etiology and pathogenesis.¹² The maternal immune system undergoes changes essential for the healthy continuation of the pregnancy. Throughout pregnancy, granulocytes, natural killer cells, and extrathymic cells become activated. Neutrophils, which constitute 50–60% of white blood cells, have an average lifespan of approximately 5.4 days, which may vary due to inflammation in HG, potentially altering the NLR.¹³ Inflammatory markers such as NLR and PLR are valuable not only due to their low cost in diagnosing acute inflammatory diseases, but also because they may provide insights into the clinical predictability and prognosis of hypertensive disorders of pregnancy, cholestasis of pregnancy, inflammatory bowel diseases, systemic lupus erythematosus, and ACS.^10,14–16 Additionally, a sensitivity of 80% and specificity of 85% have been reported for NLR in the preoperative period for acute appendicitis during pregnancy, while PLR exhibits 100% sensitivity and 45% specificity.¹⁷ Furthermore, elevated NLR and PLR have been determined in cases of miscarriage compared to healthy pregnancies.¹⁸ Some studies have also reported increased NLR and PLR in HG,^19,20 while others have found similar results to those in healthy pregnancies.^4,21 No significant differences in NLR, PLR, or PIV between women with HG and healthy pregnant women in the present study.

The underlying mechanism linking elevated PAR to HG may be explained by both inflammatory activation and nutritional impairment. Increased platelet counts reflect inflammatory response and thrombopoietic activity, while decreased albumin levels may result from reduced oral intake, dehydration, and metabolic stress. The combination of these changes may explain the elevated PAR observed in HG patients. PAR is a practical indicator formed by the ratio of platelets, an inflammation and thrombopoietic activity marker, to serum albumin, an acute phase reactant. Although studies have shown that low PAR levels are associated with adverse outcomes in malignancy, cardiovascular diseases, and systemic infections, data from the obstetric field are limited.²² A retrospective case-control study comparing 60 women with threatened abortion (TA) and 60 healthy pregnant women reported lower PAR levels in the TA group. The authors also observed that early pregnancy loss occurred in 30% of the TA cases and concluded that PAR was capable of predicting early pregnancy loss with sensitivity of 66.7% and specificity of 67.7% at a cut-off value of 4.96. Additionally, those authors noted that infants of women with TA had lower birth weights, higher rates of neonatal intensive care admissions, and increased cesarean section rates.²³ Early identification of patients at risk of severe HG requiring hospitalization is clinically important, particularly in settings with limited access to prenatal care. In this context, PAR may serve as a practical and cost-effective biomarker to aid in early risk stratification.

Although widely used inflammatory markers such as NLR and PLR were evaluated in this study, other hemogram-derived indices, including the systemic immune-inflammation index (SII) and systemic inflammation response index (SIRI), were not assessed. Future studies incorporating these markers may provide a more comprehensive evaluation of systemic inflammation in HG.

In cases of HG, hypoalbuminemia caused by nausea and vomiting may be expected to lead to an increase in PLR values. Indeed, serum albumin levels in the present research were lower in pregnant women with HG compared to healthy pregnant women. Furthermore, PLR levels were significantly higher in women with HG compared to the healthy controls (6.15±1.49 vs 9.26±2.89, respectively). According to ROC analysis, the diagnostic value of PLR in predicting HG was 6.51, with an AUC of 0.824 (CI: 0.701–0.874), sensitivity of 82.1%, specificity of 66.2%, a positive predictive value (PPV) of 70.9%, and a negative predictive value (NPV) of 79.1%.

This study has several limitations. First, its retrospective design may introduce selection bias. Second, the sample size is relatively small and derived from a single center, which may limit generalizability. Third, some clinical variables and follow-up data were not available due to the retrospective nature of the study. Finally, additional inflammatory markers such as SII and SIRI were not evaluated.

Conclusion

PAR appears to be a promising, easily accessible biomarker associated with hyperemesis gravidarum. It may have potential clinical utility in identifying patients at risk of more severe disease. However, prospective, multicenter studies with larger sample sizes are required to confirm these findings and to clarify its role in clinical practice.