Hyperemesis gravidarum (HG) is a severe form of nausea and vomiting in pregnancy associated with maternal and fetal morbidity. Recent evidence links HG to the emetic hormone, growth and differentiation factor 15 (GDF15), and affected individuals have significantly higher concentrations than unaffected individuals. ^, Thus, a medication-induced reduction in GDF15 concentrations may be a biologically plausible antiemetic response. In nonpregnant individuals, glucocorticoids lower circulating GDF15 concentrations in a dose-dependent manner. While patient-reported survey data indicate that glucocorticoids are “very effective” for HG, their use is reserved for treatment-resistant cases due to possible maternal and fetal safety concerns. ^,,,, Because the antiemetic mechanism is unknown in HG, we hypothesized that glucocorticoid effectiveness may be accompanied by a reduction in GDF15.

This retrospective cohort study included 22 patients with HG treated at The Morning Sickness Clinic. Institutional approval and ethics committee oversight was obtained, and all patients provided informed consent.

Since circulating GDF15 concentrations rise throughout early pregnancy, cases and controls were individually matched by gestational age. The cohort comprised 11 cases with refractory symptoms who received glucocorticoids after failing other antiemetics and 11 controls managed with other antiemetics. Due to the retrospective nature, data on parity, prepregnancy body mass index, and HG history were unavailable, and patients were not matched for baseline severity.

Glucocorticoid treatment consisted of intravenous solumedrol (40 mg; Solu-Medrol) followed by a 19-day oral prednisone taper: 20 mg 3 times daily (3 days), 20 mg twice daily (3 days), 10 mg 3 times daily (3 days), 10 mg twice daily (3 days), 10 mg daily (3 days), and 5 mg daily (4 days). No patients required repeat courses. Controls received nonglucocorticoid antiemetics. Concomitant antiemetic therapy was comparable between groups, with ondansetron used in >50% of both cohorts.

Clinical data and blood were collected at initial visit (mean gestational age, 13 [standard deviation {SD}, 3] weeks for cases; 12 [SD, 2] weeks for controls) and follow-up visit (mean gestational age, 14 [SD, 2] weeks for cases; 14 [SD, 3] weeks for controls). While these represented follow-up intervals of 1 week for cases and 2 weeks for controls, both cohorts were assessed at an average of 14 weeks. Clinical data included circulating GDF15 concentrations, symptom scores, gestational age, and weight. For consistency, all blood draws were performed prior to intravenous fluid hydration.

GDF15 concentrations were quantified in duplicate using enzyme-linked immunosorbent assays (AnshLabs, AL-1014-r; sensitivity=2.2 pg/mL; dynamic range=15–2906 pg/mL; within-run and between-run variation <4.3% and <5.8%, respectively). Serum samples were diluted 1:15 to meet the assay’s dynamic range, and results converted to ng/mL (1 ng=1000 pg). Symptom severity was measured using the 12-question HyperEmesis Level Prediction (HELP) tool (range, 0–60), categorized as mild (0–20), moderate (20–32), or severe (33–60). Improvements were evaluated as a reduction in score from baseline since a threshold for clinically meaningful change has not been established.

Between-group comparisons of GDF15 concentrations and symptom severity used independent 2-tailed t tests (normal) and Mann-Whitney U tests (non-normal). Within-group symptom changes used paired t tests due to their dependent nature and normality was confirmed by Shapiro-Wilk and visual inspection.

Primary analysis employed Ordinary Least Squares regression to assess the association between glucocorticoid treatment and circulating GDF15 concentrations, adjusting for baseline weight to control for the influence of body mass on GDF15 concentrations. Because raw GDF15 values violated assumptions of normality and homoscedasticity, concentrations were log-transformed prior to modeling. Coefficients and 95% confidence intervals were subsequently back transformed to express the treatment effect as a percentage reduction. Model integrity was verified through visual and analytic diagnostic tests ( Supplemental Note 1 ). Sensitivity analyses were also performed to evaluate the inclusion of additional baseline covariates and the impact of influential outliers. A P value <.05 was considered statistically significant.

At first clinic visit, the groups did not have significant differences in baseline characteristics. Mean GDF15 concentrations were 14.4 (SD, 6.1) ng/mL in cases compared to 18.4 (SD, 8.7) ng/mL in controls ( P =.23), and mean HELP scores were 36.6 (SD, 8.4) and 32.9 (SD, 6.1), respectively ( P =.25).

Follow-up assessments at an average of 14 weeks’ gestation showed circulating GDF15 concentrations were significantly lower in the glucocorticoid group (median, 10.6 ng/mL; 95% confidence interval [CI], 8.9–16.7 vs 14.6 ng/mL; 95% CI, 12.1–31; P =.04) ( Figure 1 A). Visual inspection of individual trajectories showed a consistent downward trend following glucocorticoids, whereas controls exhibited marked variability ( Figure 1 B).

Lower GDF15 concentrations and symptom resolution in glucocorticoid treated cases

( A ) Significantly lower GDF15 concentrations in cases at follow-up. ( B ) Individual trajectories show a consistent downward trend in cases, not controls. ( C ) Symptom improvement in controls and treatment-resistant cases.

GDF15 , growth and differentiation factor 15.

Stay updated, free articles. Join our Telegram channel

Join