Urea-stimulated copeptin: a novel diagnostic approach in polyuria polydipsia syndrome

Abstract

Background

Distinguishing arginine vasopressin (AVP) deficiency from primary polydipsia remains challenging. While hypertonic saline-stimulated copeptin testing offers high diagnostic accuracy, it is complex and limited to specialized centers. Intravenous urea is known to stimulate AVP secretion, but the effect of oral urea on copeptin levels is unknown.

Methods

Twenty-two healthy adults were included in a randomized, double-blind, placebo-controlled cross-over trial receiving a single dose of urea (0.5 g/kg; minimum 30 g, maximum 45 g) and placebo. Serum copeptin was measured at 30-min intervals for 2.5 h. In a second step, 13 patients with AVP-deficiency and 13 patients with primary polydipsia were included in an open-label pilot study, receiving urea only. The primary endpoint was maximum copeptin within 150 min.

Results

In healthy adults, median [IQR] copeptin significantly increased from 4.6 [3.0-5.7] pmol/L at baseline to a maximum of 10.1 [7.2-11.6] pmol/L at 120 min after ingestion of urea, while it remained stable at 3.8 [2.9-6.6] pmol/L after placebo intake (P < .001). In patients with AVP-deficiency, copeptin remained below detection limit throughout the test, while in patients with primary polydipsia the peak was seen 150 min after ingestion of urea at 7.4 pmol/L [4.3, 10.3]. The best copeptin cut-off for differentiating AVP-deficiency from primary polydipsia was 3.5 pmol/L after 120 min, with 93% sensitivity and specificity.

Conclusion

Oral urea stimulates copeptin in healthy adults and patients with primary polydipsia, but not in patients with AVP-deficiency, establishing the first oral copeptin-based test in differentiating primary polydipsia from AVP-deficiency.

Introduction

Arginine vasopressin (AVP) deficiency, formerly known as central diabetes insipidus,¹ is a neuroendocrine condition characterized by excessive urine output (polyuria) and increased fluid intake (polydipsia).² Similar symptoms are seen in patients with primary polydipsia, where, however, AVP secretion is intact.³ Discriminating those conditions is challenging, yet crucial, as treatment options differ and misdiagnosis may have serious consequences.⁴

Traditionally, the water deprivation test was considered the diagnostic gold standard.⁵ However, several studies showed its limited diagnostic accuracy, indicating the need for more reliable testing methods.^6,7 Direct AVP measurement, while conceptually promising, is impractical in clinical practice due to its instability and technical measurement challenges.⁸ The discovery of copeptin, the c-terminal end of the AVP-prohormone, has provided new diagnostic opportunities.⁹ Copeptin measurement upon osmotic stimulation with hypertonic saline infusion has demonstrated high diagnostic accuracy in differentiating AVP-deficiency from primary polydipsia and is considered the current gold standard.⁶ However, the need for repeated venous blood gas analysis and close monitoring restricts its availability to specialized centers, highlighting the need for a simpler and more widely accessible test.

Urea, a byproduct of protein metabolism, is a key contributor to plasma osmolality. Historically, it has been used as a diuretic for heart failure¹⁰ and as hyperosmotic agent to reduce brain swelling.¹¹ Nowadays urea is implemented for the treatment of the syndrome of inappropriate antidiuresis.¹² Administered orally as a beverage, urea is cost-effective and safe.¹³

Already in 1983, intravenous urea was found to elevate plasma osmolality and stimulate AVP release.¹⁴ However, no study has investigated the effect of orally administered urea on AVP or copeptin levels. In this trial, we aimed to investigate the effect of oral urea on copeptin levels in healthy adults and gather first data on the diagnostic performance of urea-stimulated copeptin in differentiating AVP-deficiency from primary polydipsia. We hypothesized that urea would stimulate copeptin levels in healthy adults and patients with primary polydipsia, whereas there would be no stimulation in patients with AVP-deficiency.

Methods

Trial design and study participants

This trial, conducted at the University Hospital Basel, Switzerland from June 2023 to June 2024, consisted of 2 parts. Part one was a double-blind, randomized, placebo-controlled cross-over study in healthy adults serving as a proof-of-concept study. Part two was an open-label, single-arm pilot study in patients with AVP-deficiency or primary polydipsia. The local ethics committee approved the protocol (EKNZ 2023-00751), and written informed consent was obtained from all participants prior to any study procedures. The study was conducted in accordance with the Declaration of Helsinki and was registered on ClinicalTrials.gov (NCT05890690).

Study procedures

For part one, we included healthy adults (≥18 years) with no medication except hormonal contraception. Exclusion criteria included polyuria (urine output >40 mL/kg body weight per 24 h) or polydipsia (fluid intake >3L per 24 h), pregnancy, breastfeeding, or allergies to components of the study drink.

Participants presented for 2 main study visits, separated by a washout phase of at least 3 days. During these visits, participants received a drink containing urea or placebo in random order. The urea drink (OMANDA AG, CH-3072 Ostermundigen) contained a dose of 0.5 g/kg body weight of urea, with a minimum of 30 g and a maximum of 45 g, dissolved in 200 mL of water along with citrus flavored additives to improve palatability. The placebo drink was designed to mimic the taste of the urea drink and was prepared by dissolving bitter-tasting nutritional supplements (Ergytonyl®, Carmol® and BitterLiebe®) and citrus flavored additives in 200 mL of water. The study drink was prepared by unblinded study staff members otherwise not involved in the study, ensuring blinding for participants and investigators. Participants presented in the morning after overnight food fasting and 2 h of fluid restriction. Prior to the visits, participants were asked to refrain from physical exercise and alcohol consumption for 24 h, and from smoking on the test day.

Baseline blood samples were collected at least 20 min after venous catheter insertion. Immediately after, participants received the study drink to consume within a maximum of 10 min, followed by 50 mL of orange juice to cleanse the taste of the study drink. No additional fluid or food consumption was allowed during the test. Blood samples were taken 30, 60, 90, 120, and 150 min after the ingestion of the study drink. Concurrently, blood pressure and heart rate were measured, and pre-defined adverse effects (ie, nausea, headache, dizziness, dysgeusia, thirst, fatigue, and indigestion) were assessed using a numeric rating scale (NRS) from 0 (no perception) to 10 (very strong perception).

Urine samples were collected at baseline and at the 150-min mark. Additionally, participants completed a 24-h urine collection and a drinking protocol, starting at the beginning of the test.

For part two, eligible participants were adult patients (≥18 years) with a confirmed diagnosis of AVP deficiency or primary polydipsia, confirmed by a water deprivation, hypertonic saline infusion, or arginine infusion test. Exclusion criteria included pregnancy, breastfeeding, and allergies to components of the study drink.

Each participant attended a single study visit, during which they received an open-label urea drink prepared identically to the formulation described for healthy adults. The study procedures were conducted similarly to those of healthy adults, except that no 24-h urine collection or drinking protocol was performed. Additionally, patients on desmopressin medication were asked to stop the drug at least 12 h before the test, and patients on cortisol replacement therapy were asked to take their stress doses before the visit.

Outcomes

For part one, the primary outcome was the difference in the maximal increase in copeptin (ΔCopeptin) within 150 min after ingestion of urea versus placebo in healthy adults. To derive the primary outcome, the maximal copeptin value between 30 and 150 min was identified for each participant, and the associated baseline value was subtracted. For part two, the primary outcome was ΔCopeptin within 150 min after the ingestion of urea in patients with AVP-deficiency or primary polydipsia.

Secondary outcomes included changes in copeptin levels at all measured timepoints, changes in parameters associated with fluid balance (ie, urea, sodium, potassium, osmolality, estimated glomerular filtration rate, glucose), changes in vital parameters, and adverse effects reported during the test.

Laboratory measurements

All laboratory measurements were performed by the laboratory of the University Hospital Basel. Copeptin was measured in batch analysis at the end of each study part using a commercial automated immunofluorescence assay (B.R.A.H.M.S. Copeptin-proAVP KRYPTOR, Thermo Fisher Scientific, Hennigsdorf Germany). Information on further laboratory measurement techniques is provided in the Appendix.

Sample size estimation

The sample size estimation was based on previous data following arginine infusion in healthy adults.¹⁵ We assumed a slightly lower increase in copeptin following urea intake, with an expected mean value of 10 pmol/L and a slightly larger standard deviation (SD) of 8 pmol/L at 150 min. We estimated copeptin 150 min after ingestion of placebo to remain comparable to baseline levels, with a mean of 5 pmol/L (SD 4). Based on these assumptions, the effect size (Cohen's d) was 0.79. The sample size was calculated by a paired 2-sided t-test using the pwr.t.test function (package “pwr”) in R, with a significance level α of 5% and a power β of 90%. Accounting for a 10% drop-out rate, the required sample size was n = 22.

Randomization

Participants were randomized according to a pre-defined randomization list. Randomization scheme was 1:1 using randomly permuted block sizes.

Statistical analysis

For the primary outcome, we performed Wilcoxon's signed-rank test. In a supplementary analysis, participants with presyncope before or during the test were excluded, since presyncope is known to elevate copeptin levels.

Part two was of an exploratory nature, aiming to gather first data on copeptin after ingestion of urea in patients with AVP-deficiency or primary polydipsia. Due to considerations of feasibility, we aimed to recruit 26 patients in total, 13 with AVP-deficiency and 13 with primary polydipsia. The primary outcome was assessed using summary statistics and visual representation of the data.

The diagnostic accuracy in part two was analyzed visually and using summary statistics. First, the timepoint with the maximum difference in copeptin levels between patients with AVP-deficiency and patients with primary polydipsia was evaluated. Second, the best cut-off for overall diagnostic accuracy was identified by inspecting receiver operating characteristic (ROC) curves, with sensitivity, specificity and area under the curve reported with 95% confidence intervals (CI). The best cut-off was determined using Youden's J index.

The time course of all laboratory parameters was evaluated using summary statistics and visual representation of the data. Adverse effects were summarized by frequency and proportion of participants experiencing symptoms (NRS > 0), and the level of discomfort was evaluated using summary statistics. All analyses were prespecified and performed in R Version 4.4.0.

Results

Baseline characteristics

For part one, 22 healthy adults were enrolled from June, 1, 2023 to September, 30, 2023. For part two, 13 patients with AVP-deficiency and 13 patients with primary polydipsia were recruited from November, 1, 2023, to June, 30, 2024. Median [IQR] age of healthy adults was 27 years [26, 32], 35 years [28, 47] in patients with AVP-deficiency and 30 years [25, 59] in patients with primary polydipsia. Fifty-five percent of healthy adults were female, 54% and 85% in patients with AVP-deficiency or primary polydipsia, respectively. In patients with AVP-deficiency, replacement therapy for additional deficiencies of anterior pituitary hormones was hydrocortisone (46%), levothyroxine (46%), sex hormones (38%), and somatotropin (15%). Baseline characteristics are summarized in Table 1.

Effects on plasma urea and osmolality

The time course of urea and osmolality in healthy adults, patients with AVP-deficiency and primary polydipsia is presented in Figure 1A and B.

In healthy adults, median plasma urea levels were 4.3 mmol/L [3.5, 5] before ingestion of urea and 4.2 mmol/L [3.7, 5.6] before ingestion of placebo. A peak level of 16.8 mmol/L [14.5, 18] plasma urea was reached 60 min after ingestion of urea, whereas no relevant change was observed after placebo. The day after the test, blood urea was 6.0 mmol/L [4.8, 7.8] after ingestion of urea and 4.7 mmol/L [4.1, 5.7] after ingestion of placebo. Baseline plasma osmolality was 292 mOsm/kg [288, 294] before ingestion of urea and 293 [290, 296] before ingestion of placebo. Similar to urea levels, a plasma osmolality peak of 304 mOsm/kg [301, 307] was reached 90 min after ingestion of urea, whereas no relevant change was observed after placebo.

In patients with AVP-deficiency or primary polydipsia, baseline urea levels were 3.9 mmol/L [3.7, 4] and 3.8 mmol/L [3.4, 5], respectively. A peak of 16.7 mmol/L [16.0, 16.9] was reached after 90 min in patients with AVP-deficiency, and 16.9 mmol/L [15.9, 20.1] after 60 min in patients with primary polydipsia. Plasma osmolality at baseline was 290 mOsm/kg [288, 295] in patients with AVP-deficiency and 289 [287, 290] in patients with primary polydipsia. A peak of 307 mOsm/kg [305, 311] in patients with AVP-deficiency, and of 303 mOsm/kg [301, 304] in patients with primary polydipsia was reached after 60 min.

Effects on copeptin levels

The time course of copeptin levels in healthy adults, patients with AVP-deficiency or primary polydipsia is shown in Figure 1C.

In healthy adults, median copeptin at baseline was 4.6 pmol/L [3.0, 5.7] and reached a peak of 10.1 pmol/L [7.2, 11.6] 120 min upon ingestion of urea, while in the placebo arm copeptin was 3.8 pmol/L [2.9, 6.6] at baseline with no major changes during the test. The maximum median change in copeptin was +4.7 pmol/L [+3.8, +7.1] after ingestion of urea and ±0 pmol/L [−1.3, +0.2] after ingestion of placebo (P = .005).

In patients, median copeptin levels at baseline were under the detection limit with <2.7 [<2.7, <2.7] pmol/L in AVP-deficiency and <2.7 [<2.7, 3.9] pmol/L in primary polydipsia. Median copeptin peaked after 150 min in patients with primary polydipsia at 7.4 pmol/L [4.3, 10.3], while showing no change in patients with AVP-deficiency. The maximum median change in copeptin was +4.1 pmol/L [+1.3, +6.1] in patients with primary polydipsia and ±0 pmol/L [±0, +0.5] in patients with AVP-deficiency.

Diagnostic accuracy of urea-stimulated copeptin

The best copeptin cut-off for distinguishing patients with AVP-deficiency from patients with primary polydipsia was observed 120 min after ingestion of urea at 3.5 pmol/L, with a sensitivity of 92% (CI: 77%-100%) and specificity of 92% (CI: 77%-100%). Maximum specificity (100%) was observed at a copeptin cut-off of 2.7 pmol/L, with a sensitivity of 77% (CI: 54%-100%). Conversely, maximum sensitivity (100%) was observed at a copeptin cut-off of 4.7 pmol/L with a specificity of 69% (CI: 46%-92%). The distribution of copeptin at 120 min, ROC curve, and best copeptin cut-off are presented in Figure 2.

Effects on sodium, potassium, and glucose

The time course of sodium, potassium, and glucose in healthy adults, patients with AVP-deficiency or primary polydipsia is presented in Figure 3.

In healthy adults, median sodium levels were 139 mmol/L [139, 141] and 139 mmol/L [139, 140] before ingestion of urea or placebo, respectively, remaining stable throughout the test. Potassium was 3.8 mmol/L [3.7, 4] and 3.8 mmol/L [3.7, 4.1] before ingestion of urea or placebo, respectively. A slight increase was observed in both groups, with a median of +0.4 mmol/L [+0.2, +0.5] after urea and +0.4 mmol/L [+0.1, +0.5] after placebo. Median glucose was 4.7 mmol/L [4.4, 4.8] before ingestion of urea and 4.6 mmol/L [4.5, 4.9] before ingestion of placebo. In both urea and placebo, median glucose levels peaked at 30 min at 5.9 mmol/L [5.4, 6.4] and 6.7 mmol/L [5.9, 7.6], respectively.

In patients, sodium levels before ingestion of urea were 140 mmol/L [138, 141] in AVP-deficiency and 137 mmol/L [137, 139] in primary polydipsia. During the test, sodium slightly increased in both groups, with a median maximum change of +2 mmol/L [+2, +3] in AVP-deficiency and +1 mmol/L [0, 2] in primary polydipsia. Potassium before ingestion of urea was 4 mmol/L [3.7, 4] in AVP-deficiency and 3.9 mmol/L [3.7, 4] in primary polydipsia. Similar to healthy adults, potassium slightly increased during the test with a median maximum change of +0.3 mmol/L [+0.2, +0.4] in AVP-deficiency and +0.2 mmol/L [+0.1, +0.3] in primary polydipsia.

Median glucose before ingestion of urea was 5.2 mmol/L [4.8, 5.4] in AVP-deficiency and 5.1 mmol/L [4.8, 5.5] in primary polydipsia. In both groups, glucose levels peaked at 30 minutes with 6.0 mmol/L [5.5, 7.0] in AVP-deficiency and 6.4 mmol/L [5.4, 7.2] in primary polydipsia.

Effects on kidney function, urine output and fluid intake in healthy adults

Median creatinine clearance calculated from 24-h urine collection was 105 mL/min/1.73 m² [88, 113] after ingestion of urea and 108 [89, 120] mL/min/1.73 m² after ingestion of placebo. Median 24-h urine volume was higher after ingestion of urea with 2025 mL [1820, 2575] compared with placebo with 1550 mL [1112, 2325]. Fluid intake following the ingestion of urea or placebo as documented by 24-h drinking protocol was similar, with 2650 mL [2274, 3148] after urea and 2508 mL [1952, 2760] after placebo.

Safety and tolerability

Frequency and severity of all adverse effects are presented in Table 2.

The ingestion of urea was generally well tolerated in all 3 groups. In healthy adults, nausea (50% vs. 9%), headache (64% vs. 41%), and indigestion (64% vs. 41%) were more common after ingesting urea compared with placebo; the severity was generally mild with median NRS after ingestion of urea being 2 [1, 2] for nausea, 2 [1, 2] for headache, and 1 [1, 3] for indigestion.

In patients with AVP-deficiency, ingestion of urea led to a higher prevalence of nausea (62% vs. 46%), fatigue (100% vs. 85%), dysgeusia (85% vs. 69%), headache (69% vs. 54%) and indigestion (54% vs. 46%) than in primary polydipsia; the severity was mild to moderate with median NRS of 2 [2, 2] for nausea, 5 [2, 7] for fatigue, 5 [1, 6] for dysgeusia, 4 [3, 6] for headache, and 3 [2, 6] for indigestion and was comparable to the severity of symptoms in primary polydipsia.

Discussion

Our study has 3 main findings. First, we confirm a significant increase in plasma osmolality following the ingestion of urea. Second, urea stimulates copeptin secretion in healthy adults and patients with primary polydipsia but not in patients with AVP-deficiency. Third, and most importantly, copeptin levels measured 2 h post-urea ingestion distinguished AVP deficiency from primary polydipsia with over 90% accuracy.

Despite decades of clinical use,¹⁶ pharmacokinetic data on oral urea administration remain limited. Our findings provide detailed insights into plasma urea dynamics during the first 2.5 h post-ingestion, along with data on excretion and 24-h urine output. The rise in plasma osmolality following oral urea ingestion aligns with previous reports on intravenous urea administration by Zerbe et al.¹⁴ after the infusion of 15% urea¹⁴ and confirms the rapid absorption and high bioavailability of oral urea. The dose used in our study is coherent with the clinical use in hyponatremia with 15-60 g/day.¹⁷ Upon administration of a single oral dose of urea, peak levels in plasma are observed within 60 to 90 min, mostly remaining below 20 mmol/L, and levels return to near baseline within 24 h.

The stimulatory effect of urea ingestion on copeptin is in accordance with our hypothesis. Interestingly, while plasma osmolality peaks within 60 min, copeptin levels peak later at 120 minutes. Urea is a polar molecule that crosses cell membranes through transporters present in many tissues, including the central nervous system. This permeability contributes to its reputation as an “ineffective osmole” compared with less membrane-permeable solutes such as sodium or mannitol.^13,18 Consequently, the greatest copeptin increase would be expected to coincide with or shortly follow the peak in plasma osmolality when the difference in intra- and extracellular osmotic pressure is at its peak. As the delayed response of copeptin contrasts with other osmotic stimuli,^14,19 additional stimulatory effects should be considered. One potential mechanism involves the effect of urea-induced osmotic diuresis on blood volume. By increasing the osmotic load in the renal tubules, urea promotes water excretion without significant electrolyte loss.¹³ This osmotic diuresis may further result in mild hypovolemia. A resulting drop in arterial blood pressure could trigger baroreceptor-mediated copeptin release.²⁰ However, in our study, stable blood pressure and renal function throughout the test period make this explanation less plausible (see Appendix).

Alternatively, cerebral dehydration caused by urea may play a role. Urea does not cross the blood-brain barrier as easily as other cell membranes.¹³ This discrepancy in permeability creates an osmotic gradient, driving water from the brain into circulation. Historically, this mechanism was exploited in hyperosmolar therapies using urea to reduce cerebral edema.¹¹ The observed stimulation of copeptin may be a consequence of this cerebral dehydration, potentially explaining the delayed response as dehydration intensifies over time. However, no such effect has been reported previously, and this hypothesis requires further investigation.

Generally, the copeptin response to oral urea was comparable to the effect of arginine infusion^15,21 but less pronounced than the increase observed after hypertonic saline infusion.⁶ This difference has already been observed by Zerbe et al.¹⁴ and is likely attributable to variations in membrane permeability, as previously discussed. The copeptin response was more pronounced in healthy adults than in patients with primary polydipsia, consistent with findings from other stimulatory tests of the posterior pituitary.^6,15,19 Still, our proposed copeptin cut-off of 3.5 pmol/L at 120 min post-ingestion exhibits promising results with a specificity and sensitivity of 92%, similar to the arginine infusion test.¹⁵ Since our findings on diagnostic performance were exploratory, a validation study with prespecified time measurement timepoint and copeptin cut-off is needed to establish the clinical utility of urea testing. Our recent head-to-head comparison of arginine and hypertonic saline infusion revealed the superiority of the hypertonic saline test,²¹ leading to the recommendation of employing both upper and lower cut-offs for arginine infusion and conducting follow-up testing with hypertonic saline for indeterminate results. A similar approach could be applied for testing with urea.

Urea offers several advantages over arginine infusion. First, it is administered orally, making the preparation straightforward and the procedure easy to perform, even in centers with less specialized settings or primary care. Second, urea is highly cost-effective and widely available. Third, we observed only mild side effects providing a high safety and tolerability profile.²² While the bitter taste is generally a limitation of urea use, in the case of a single dose it may be tolerated. Additionally, more palatable formulations are now available.²³

Our study has several strengths, including the robust design and rigorous execution, ensuring high-quality data analyzed using current statistical best practices.²⁴ However, certain limitations exist. Establishing diagnostic testing with urea in children might be challenging due to the bitter taste of the beverage. Further, stress doses in patients on cortisol replacement therapy might have influenced copeptin levels, since glucocorticoids are known to inhibit AVP secretion.²⁵ However, no differences in copeptin levels were observed between patients with and without stress doses, mitigating this potential effect. Considering the good tolerability of urea in our study, stress doses might not be needed in clinical practice. External factors, such as vasovagal reactions to catheter placement or nausea during the test, are known to elevate copeptin levels.^26,27 However, a supplementary analysis (see Appendix) excluding participants who experienced presyncope before or during the test showed no differences in the results, mitigating this potential confounding. Although nausea occurred more frequently following urea ingestion compared with placebo, the intensity was generally mild and peaked at 30 min, before the copeptin peak at 120 min (see Appendix). Therefore, nausea is unlikely to have influenced the outcome. In patients, no placebo testing was performed. However, placebo intake in healthy adults showed no changes in copeptin, and prior studies involving placebo in patients demonstrated no impact on copeptin,²⁸ supporting the validity of our findings.

In conclusion, urea provides a simple, cost-effective alternative diagnostic test in the differential diagnosis of polyuria-polydipsia syndrome and introduces the first oral copeptin-based test in differentiating primary polydipsia from AVP-deficiency.

Acknowledgments

The authors thank Cemile Bathelt, Nina Hutter, Flavia Hasenböhler, Hualin Lüthi, and Jonas Hüllstrung for their help with conducting the study.

Supplementary material

Supplementary material is available at European Journal of Endocrinology online.

Funding

M.C.-C received a grant from the Swiss National Science Foundation (32473B162608). C.A. received the Young Talents in Clinical Research grants from the Swiss Academy of Medical Sciences and the Gottfried und Julia Bangerter-Rhyner Foundation, the University Hospital Basel, and the Hemmi Foundation. Urea and citric-flavored additives were generously provided free of charge by OMANDA AG (3072 Ostermundigen). OMANDA AG had no involvement in any other aspects of this study, including its conception, design, planning, execution, data analysis, or manuscript preparation.

Authors’ contributions

S.L. contributed to data collection, covered all statistical aspects, data analysis, and interpretation, did the literature search, and wrote the first draft of the manuscript. S.L., C.A., and M.C.-C. wrote the protocol. C.A. and M.C.-C. contributed to data analysis and interpretation, edited the manuscript, and supervised the conduct of the study. All other authors contributed to data collection, and data interpretation, and revised the manuscript. S.L., C.A., and M.C.-C. verified the data and had access to all raw data. All authors had final responsibility for the decision to submit for publication.

Sven Lustenberger (Conceptualization [lead], Data curation [lead], Formal analysis [lead], Investigation [lead], Methodology [lead], Project administration [lead], Visualization [lead], Writing—original draft [lead]), Cihan Atila (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Funding acquisition [equal], Methodology [equal], Supervision [equal], Writing—review & editing [equal]), Juliana Baumgartner (Data curation [equal], Investigation [equal], Project administration [equal], Writing—review & editing [equal]), Sophie Monnerat (Methodology [equal], Writing—review & editing [equal]), Julia Beck (Investigation [supporting], Writing—review & editing [equal]), Joyce Santos de Jesus (Conceptualization [equal], Investigation [equal], Methodology [equal], Project administration [lead], Writing—review & editing [equal]), and Mirjam Christ-Crain (Conceptualization [equal], Funding acquisition [lead], Methodology [equal], Supervision [lead], Validation [equal], Writing—review & editing [lead])

Data availability

We may share de-identified, individual participant-level data that underlie the results reported in this article and related documents, including the study protocol and the statistical analysis plan. Data will be available with the publication of our main manuscript upon request. All requests should be sent to the corresponding author. The steering committee of this study will discuss all requests and decide, based on the scientific rigor of the proposal, whether data sharing is appropriate. All applicants are asked to sign a data access agreement.

References

Author notes