Dataset Information

Application of regularized regression to identify novel predictors of mortality in a cohort of hemodialysis patients.

ABSTRACT: Cohort studies often provide a large array of data on study participants. The techniques of statistical learning can allow an efficient way to analyze large datasets in order to uncover previously unknown, clinically relevant predictors of morbidity or mortality. We applied a combination of elastic net penalized Cox regression and stability selection with the aim of identifying novel predictors of mortality in a cohort of prevalent hemodialysis patients. In our analysis we included 475 patients from the "rISk strAtification in end-stage Renal disease" (ISAR) study, who we split into derivation and confirmation cohorts. A wide array of examinations was available for study participants, resulting in over a hundred potential predictors. In the selection approach many of the well established predictors were retrieved in the derivation cohort. Additionally, the serum levels of IL-12p70 and AST were selected as mortality predictors and confirmed in the withheld subgroup. High IL-12p70 levels were specifically prognostic of infection-related mortality. In summary, we demonstrate an approach how statistical learning can be applied to a cohort study to derive novel hypotheses in a data-driven way. Our results suggest a novel role of IL-12p70 in infection-related mortality, while AST is a promising additional biomarker in patients undergoing hemodialysis.

SUBMITTER: Werfel S

PROVIDER: S-EPMC8085040 | biostudies-literature | 2021 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Application of regularized regression to identify novel predictors of mortality in a cohort of hemodialysis patients.

Werfel Stanislas S Lorenz Georg G Haller Bernhard B Günthner Roman R Matschkal Julia J Braunisch Matthias C MC Schaller Carolin C Gundel Peter P Kemmner Stephan S Hayek Salim S SS Nusshag Christian C Reiser Jochen J Moog Philipp P Heemann Uwe U Schmaderer Christoph C

Scientific reports 20210429 1

Cohort studies often provide a large array of data on study participants. The techniques of statistical learning can allow an efficient way to analyze large datasets in order to uncover previously unknown, clinically relevant predictors of morbidity or mortality. We applied a combination of elastic net penalized Cox regression and stability selection with the aim of identifying novel predictors of mortality in a cohort of prevalent hemodialysis patients. In our analysis we included 475 patients ...[more]

PMID: 33927289

Similar Datasets

Project description:Rationale & objectiveBoth hypo- and hyperkalemia can cause fatal cardiac arrhythmias. Although predialysis serum potassium level is a known modifiable risk factor for death in patients receiving hemodialysis, especially for hypokalemia, this risk may be underestimated. Therefore, we investigated the relationship between predialysis serum potassium level and death in incident hemodialysis patients and whether there is an optimum level.Study designProspective multicenter cohort study.Setting & participants1,117 incident hemodialysis patients (aged >18 years) from the Netherlands Cooperative Study on the Adequacy of Dialysis-2 study were included and followed from their first hemodialysis treatment until death, transplantation, switch to peritoneal dialysis, or a maximum of 10 years.ExposurePredialysis serum potassium levels were obtained every 6 months and divided into 6 categories: ≤4.0 mmol/L, >4.0 mmol/L to ≤4.5 mmol/L, >4.5 mmol/L to ≤5.0 mmol/L, >5.0 mmol/L to ≤5.5 mmol/L (reference), >5.5 mmol/L to ≤6.0 mmol/L, and >6.0 mmol/L.Outcomes6-month all-cause mortality.Analytical approachCox proportional hazards and restricted cubic spline analyses with time-dependent predialysis serum potassium levels were used to calculate the adjusted HRs for death.ResultsAt baseline, the mean age of the patients was 63 years (standard deviation, 14 years), 58% were men, 26% smoked, 24% had diabetes, 32% had cardiovascular disease, the mean serum potassium level was 5.0 mmol/L (standard deviation, 0.8 mmol/L), 7% had a low subjective global assessment score, and the median residual kidney function was 3.5 mL/min/1.73 m2 (IQR, 1.4-4.8 mL/min/1.73 m2). During the 10-year follow-up, 555 (50%) deaths were observed. Multivariable adjusted HRs for death according to the 6 potassium categories were as follows: 1.42 (95% CI, 1.01-1.99), 1.09 (95% CI, 0.82-1.45), 1.21 (95% CI, 0.94-1.56), 1 (reference), 0.95 (95% CI, 0.71-1.28), and 1.32 (95% CI, 0.97-1.81).LimitationsShorter intervals between potassium measurements would have allowed for more precise mortality risk estimations.ConclusionsWe found a U-shaped relationship between serum potassium level and death in incident hemodialysis patients. A low predialysis serum potassium level was associated with a 1.4-fold stronger risk of death than the optimal level of approximately 5.1 mmol/L. These results may imply the cautious use of potassium-lowering therapy and a potassium-restricted diet in patients receiving hemodialysis.

Dataset Information

Application of regularized regression to identify novel predictors of mortality in a cohort of hemodialysis patients.

Publications

Application of regularized regression to identify novel predictors of mortality in a cohort of hemodialysis patients.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets