Pith. sign in

REVIEW 4 major objections 5 minor 43 references

dsLassoCov: a federated machine learning approach incorporating covariate control

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read dsLassoCov is claimed to be the first federated learning algorithm that controls for covariate effects in high-dimensional linear models while exchanging only aggregated gradients and losses.

desk verdict A useful federated Lasso implementation is undercut by an algebraically wrong proximal update in the pseudocode, plus soft simulation evidence, though the real-data deployment is a genuine practical contribution worth refereeing. read the letter →

arxiv 2412.07991 v1 pith:DJDTYCBL submitted 2024-12-11 q-bio.QM cs.LG

Han Cao , Augusto Anguita , Charline Warembourg , Xavier Escriba-Montagut , Martine Vrijheid , Juan R. Gonzalez , Tim Cadman , Verena Schneider-Lindner
show 3 more authors
Daniel Durstewitz Xavier Basagana Emanuel Schwarz
This is my paper · ORCID
classification q-bio.QMcs.LG MSC 62J07
keywords federatedlearningLASSOcovariateadjustmentconfoundingfeatureselectionexposomeprivacy-preservinganalysishigh-dimensionalregression
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Federated learning trains models across distributed databases by exchanging model updates rather than raw data, and the open problem is controlling for confounders in that setting. dsLassoCov is a federated implementation of LASSO that adjusts for covariates while the model trains, so data owners never need to run per-feature residualizations or reveal individual-level data. The paper's central claim is that minimizing a penalized loss over feature coefficients and unpenalized covariate coefficients yields feature coefficients that are independent of linear associations with the covariates. Because the covariate block is profiled out algebraically, the optimizer can be federated by exchanging only gradients and loss values, keeping communication cost low. Simulations and a six-cohort exposome analysis, which reproduces previously reported hypertension associations, support the method's efficiency and feature-selection accuracy.

What carries the argument

The central mechanism is the profiled covariate objective: because the unpenalized covariate block is linear, the optimal covariate coefficients are a closed-form least-squares projection of the residual $y - Xw$ onto the covariate space. Plugging that solution into formulation (1) yields formulation (2), in which the loss is the squared distance between the outcome and the prediction, each with linear covariate associations removed. This identity carries the covariate-control claim, since it makes the feature solution independent of covariate associations; it also dictates the federated protocol, where each server computes local gradients and losses of that same objective and the client aggregates them with weights proportional to local sample sizes. The same profiled objective extends to classification by replacing squared error with logistic loss.

What would settle it

Generate simulated data in which a confounder affects the outcome nonlinearly (for example, quadratically or through a threshold) while also shifting a subset of features, run dsLassoCov and a correctly adjusted comparator, and compare feature-selection accuracy; if dsLassoCov's accuracy drops materially, the linear-covariate assumption is the point of failure.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a mathematical reduction plus a distributed implementation. For the objective $\min_{w,w^{(c)}} \frac{1}{2n}\sum_i (y_i - x_i w - x_i^{(c)} w^{(c)})^2 + \lambda|w|$, the first-order condition for $w^{(c)}$ gives $w^{(c)} = (X^{(c)T}X^{(c)})^{-1}X^{(c)T}(y - Xw)$. Substituting this back turns the objective into a comparison between the covariate-residualized outcome and the covariate-residualized prediction, so the optimal $w$ is the same as if the covariates had been regressed out of both sides. The paper calls this a double-machine-learning-like debiased property and notes that a similar device appears in double-LASSO. The federated algorithm then needs only shared gradients, losses, and sample counts, not the data themselves. In simulation, the method matches the conventional federated residualization pipeline on regression feature selection and beats it in classification, where residualizing a binary outcome is not possible; in the real-world exposome application it selects exposures whose odds ratios align with earlier studies.

Load-bearing premise

The load-bearing premise is that the included covariates capture all confounding and act linearly on the outcome, because the derivation that the feature solution is independent of covariate associations holds only for that linear model.

Editorial extensions

If this is right

  • In high-dimensional federated biomarker screens, dsLassoCov removes the need for one generalized-linear-model fit per feature to residualize covariates, which is the main cost of the conventional pipeline.
  • The method supplies a principled covariate adjustment for federated classification, a setting where residualizing a binary outcome is not well-defined.
  • Simulation results support that, as the number of features grows, dsLassoCov keeps runtime lower than the conventional federated approach while retaining at least equal feature-selection accuracy in regression and higher accuracy in classification.
  • The real-data replication suggests that federated selection of early-life exposures for hypertension is stable enough to reproduce previously reported odds ratios across six distinct databases.
  • Adoption could lower the communication ceiling for consortium-wide exposome and omics analyses, since per-iteration exchanged messages do not scale with the number of residualization regressions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension the paper does not develop: the same profiling trick could be applied to other penalties such as ridge or elastic net, but the linear-in-covariates form of the inner adjustment would need to change before those extensions inherit the current guarantees.
  • If confounders act nonlinearly, the profiled linear term will only approximate the true confounding surface; sensitivity analyses varying the covariate set or adding nonlinear terms would be needed to know how much residual confounding remains in real data.
  • The communication advantage is largest when the number of covariates is small relative to the number of features; when covariates are numerous, the closed-form projection may need to be computed or approximated carefully in the federated protocol.
  • A testable consequence: on a cohort with known technical batch structure, replacing per-feature residualization with dsLassoCov should produce nearly identical marginal effect estimates while cutting wall-clock time, a prediction that can be checked directly on existing consortium data.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper introduces dsLassoCov, a federated learning method that fits a LASSO-type linear model with an additional set of unpenalized covariates for confounder control, and applies it to high-dimensional biomedical data in the DataSHIELD environment. The proposed objective is the standard penalized regression with an unpenalized covariate block, and the optimization is a federated proximal-gradient method using exact aggregated gradients and a regularization path with warm starts. The authors provide a profiling argument that the solution is equivalent to minimizing the distance between outcome residuals and prediction residuals after removing linear covariate associations. The empirical part compares dsLassoCov with a per-feature residualization baseline ('ds.glm + ds.Lasso') and local meta-analysis methods in simulations, and reports a replication of a HELIX exposome analysis of childhood hypertension.

Significance. If the implementation were verified to solve the stated objective, the paper would be a useful practical contribution: it packages a known statistical construction (LASSO with unpenalized covariates, equivalent to residualization in the linear case) into a communication-efficient federated algorithm, reports run-time gains over per-feature residualization, and provides an open-source DataSHIELD implementation. The mathematical core is not deep, but the engineering and application value to exposome consortia is real. The paper also ships concrete code repositories and a tutorial, which is commendable. However, as written, the supporting evidence cannot be fully trusted because the central pseudocode appears to implement a different proximal update than the one derived in the supplement, and the simulation data-generation equations are dimensionally inconsistent.

major comments (4)
  1. [Supplementary Methods, Algorithm 1, line 8] Line 8 of Algorithm 1 does not implement the soft-threshold update derived in Supplementary Eq. (8). For z = S_i - G_w/L, the correct proximal update is sign(z) max(|z| - λ/L, 0), equivalently max(|z| - λ/L, 0) * sign(z). The pseudocode instead computes max{S_i - (G_w - λ)/L, 0} * sign(S_i - G_w/L). These differ: for the scalar example S_i=1, G_w=0, λ=0.1, L=1, the correct update gives 0.9 while the pseudocode gives 1.1, and negative arguments are also handled incorrectly. Consequently, if the shipped code follows the pseudocode, the algorithm is not minimizing the stated objective and the covariate-independence argument in Eq. (2) does not apply to the computed solution; if the code is correct, the pseudocode is inconsistent with Eq. (8). Either way, the reported simulation results cannot be verified from the manuscript as written. Please correct the pseudocode and provide a minimal numerical check against a standard solver.
  2. [Supplementary Methods, 'Confounded data simulation', Eq. (12)] The data-generation equation is dimensionally inconsistent. After defining Xsig as n×p_t and Xc as n×p_c, the expression Y = [Xsig Xc a] [w; 0_c] does not conform: the scalar intensity a cannot be appended as a column to the design matrix, and [w; 0_c] has length p_t + p_c, not p_t + p_c + 1. The intended generative model presumably includes the confounding vector c ∈ R^{n×1} multiplied by the scalar a, but this is not what is written. Since every simulation in the paper depends on this generative scheme, please rewrite Eqs. (11)-(12) with explicit dimensions and state the exact R code used, so the simulations are reproducible.
  3. [Methods, 'Covariate control mechanism' and Discussion, first paragraph] The paper claims that dsLassoCov is 'mathematically equivalent to an iterative procedure that alternates between a covariate removal step and a parameter update step,' and that this 'guarantees the estimation of covariate-free parameters at each iteration.' The derivation in Eq. (2) is a static profiling identity at the optimum of the joint objective, not a property of the iterates of Algorithm 1. In Algorithm 1, the covariate coefficients w(c) are updated by a gradient step (line 9), not by residualization or by solving the closed-form conditional minimizer. If the per-iteration equivalence claim is intended, a proof should be supplied; otherwise the claim should be revised to describe the profiled objective and its solution.
  4. [Results, 'Real data analysis'; Table S3; Figure 6 caption] The text and Figure 6 caption state that selected exposures are those with odds ratio below 0.95 or above 1.05, but Table S3 lists many exposures with odds ratios within [0.95, 1.05], for example Cobalt 0.969, PBDE 153 0.970, Lead 0.978, DMTP 0.985, and MEHP 0.994. The selection criterion is therefore not reproduced by the displayed results. Please define the exact selection rule (including how confounders are treated), reconcile Table S3 with Figure 6, and report the confidence intervals or stability of the selected set if a threshold is used.
minor comments (5)
  1. [Abstract and throughout] There are recurring language and typographical issues: 'this allow' should be 'this allows', 'odd ratio' should be 'odds ratio', 'refered' should be 'referred', 'counfounding' should be 'confounding', and 'fedeated' should be 'federated'.
  2. [Supplementary Methods, 'Read data analysis' and 'Data cohorts and preprocessing'] The supplementary text contains duplicated paragraphs: the 'Read data analysis' section repeats the HELIX description, and the 'Data cohorts and preprocessing' section contains two nearly identical paragraphs describing the exposure categories. Please remove the duplicates.
  3. [Supplementary Methods, Algorithm 1, line 5] The line-search candidate sequence is given as {L_{i-1}, 2L_{i-1}, 4L_{i-1}, 16L_{i-1}, ...}; the jump from 4 to 16 is presumably a typo for 8, and the Armijo-Goldstein condition itself is not stated explicitly. Please clarify the line-search rule and the stopping criterion used in Algorithm 1 and Algorithm 2.
  4. [Supplementary Methods, 'The non-federated training protocol', lambda_max derivation] The displayed equation 'λ_p×1 − 1/n | x^T y − x^T x_c (x_c^T x_c)^{-1} x_c^T y / σ | = 0' is not a well-formed mathematical statement. The final formula for λ_max is correct in substance, but the intermediate line should be rewritten componentwise so that the role of σ ∈ [-1,1] is clear.
  5. [Methods, 'Modeling, optimization and algorithm'] The classification extension is asserted by saying that 'one can derive all equations ... by switching to the logistic loss', but no logistic objective or proximal update is given. Since classification simulations are reported, please add the explicit classification objective and confirm that the same exact-gradient aggregation and line search apply unchanged.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the covariate-control derivation is a self-contained variable projection of the stated objective, and the self-citations are infrastructure rather than load-bearing.

full rationale

The paper's central claim—that dsLassoCov controls covariates by leaving the covariate block unpenalized and profiled out—is derived self-containedly. Formulation (1) defines a convex LASSO objective with an unpenalized covariate block, and the covariate-control mechanism solves the first-order optimality condition for w(c), substitutes the solution back, and obtains the residualized objective (2). This is a standard variable-projection / Frisch-Waugh-Lovell-style equivalence rather than a definition of the target in terms of the method's own outputs. The statement that 'the solution w is independent of associations with x(c)' is a mathematical consequence of the profiled objective, and the paper explicitly credits the classical and double-machine-learning ancestry (refs 17, 25, 26), so no known result is being renamed as a new discovery. The only self-citation is to dsMTL (ref 19), which supports the federated optimization framework and the lambda-sequence warm-start; it is not the evidence for the covariate-control claim. Real-data findings are benchmarked against external HELIX publications (refs 27 and 28), and the simulations compare against independent federated and meta-analysis baselines. The apparent proximal-update algebra discrepancy in Algorithm 1 is a correctness issue, not a circular reduction, so it does not raise the circularity score.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central derivation uses standard convex optimization and the classical projection/residualization equivalence. The only data-dependent quantity is the Lasso penalty lambda. The paper's main unexamined assumption is the linearity and sufficiency of covariate adjustment, plus an unproved logistic analogue.

free parameters (1)
  • lambda (regularization strength) = selected by 5-fold CV over 100 values; e.g., 0.009 reported for local-dsLassoCov in the real-data analysis
    Controls sparsity of feature coefficients. This is a standard Lasso hyperparameter, data-dependent but not an ad hoc constant hidden in the method.
assumptions (5)
  • domain assumption Covariate effects are linear
    Stated before Formulation (1): 'Here, we assume that the effect of the covariates is linear.' The closed-form profiling of w_c and the 'covariate-free' interpretation rely on squared-error loss and linear adjustment.
  • standard math Frisch-Waugh-Lovell / projection equivalence
    Invoked in the Introduction (ref 17) and implicit in Eq. (2). The method's covariate-control property is this classical equivalence, not a new result.
  • standard math Proximal gradient with Nesterov acceleration converges to the optimum of the convex nonsmooth objective
    Assumed for Algorithm 1; standard for L1-regularized convex problems with smooth loss.
  • domain assumption The 13 confounders in the HELIX analysis are correctly identified and sufficient via the DAG in Figure S1
    The real-data conclusion that findings replicate previous studies depends on no unmeasured confounding beyond the DAG; the paper states confounders were identified a priori from previous literature.
  • domain assumption The logistic-loss classification extension preserves the same covariate-control interpretation
    The paper only derives the residualization equivalence for squared loss and asserts 'one can derive all equations for the classification algorithm in the same approach by switching to the logistic loss'. No proof is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of dsLassoCov: a federated machine learning approach incorporating covariate control." pith.science (2026). https://pith.science/paper/DJDTYCBL

@misc{pith2026241207991,
  author       = {Pith},
  title        = {Pith review of: dsLassoCov: a federated machine learning approach incorporating covariate control},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DJDTYCBL}},
  note         = {Machine review of arXiv:2412.07991}
}
read the original abstract

Machine learning has been widely adopted in biomedical research, fueled by the increasing availability of data. However, integrating datasets across institutions is challenging due to legal restrictions and data governance complexities. Federated learning allows the direct, privacy preserving training of machine learning models using geographically distributed datasets, but faces the challenge of how to appropriately control for covariate effects. The naive implementation of conventional covariate control methods in federated learning scenarios is often impractical due to the substantial communication costs, particularly with high-dimensional data. To address this issue, we introduce dsLassoCov, a machine learning approach designed to control for covariate effects and allow an efficient training in federated learning. In biomedical analysis, this allow the biomarker selection against the confounding effects. Using simulated data, we demonstrate that dsLassoCov can efficiently and effectively manage confounding effects during model training. In our real-world data analysis, we replicated a large-scale Exposome analysis using data from six geographically distinct databases, achieving results consistent with previous studies. By resolving the challenge of covariate control, our proposed approach can accelerate the application of federated learning in large-scale biomedical studies.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 42 canonical work pages

  1. [31]

    dsLassoCov: a federated learning approach incorporating covariate control

    Elze, M.C. et al. Comparison of propensity score methods and covariate adjustment: evaluation in 4 cardiovascular studies. Journal of the American College of Cardiology 69, 345-357 (2017). Figures Figure 1 . The regularization path of dsLassoCov . The regularization path refered to outputs of dsLassoCov model training, creating a series models indexed by ...

    work page 2017

  2. [32]

    & Zou, Q

    Sun, S., Wang, C., Ding, H. & Zou, Q. Machine learning and its applications in plant molecular studies. Briefings in functional genomics 19, 40-48 (2020)

    work page 2020

  3. [33]

    Watson, D.S. et al. Clinical applications of machine learning algorithms: beyond the black box. Bmj 364(2019)

    work page 2019

  4. [34]

    Asiimwe, R. et al. From biobank and data silos into a data commons: convergence to support translational medicine. Journal of Translational Medicine 19, 1-13 (2021)

    work page 2021

  5. [35]

    & Von dem Bussche, A

    Voigt, P. & Von dem Bussche, A. The eu general data protection regulation (gdpr). A Practical Guide, 1st Ed., Cham: Springer International Publishing 10, 10-5555 (2017)

    work page 2017

  6. [36]

    & Sangiorgi, L

    Decherchi, S., Pedrini, E., Mordenti, M., Cavalli, A. & Sangiorgi, L. Opportunities and challenges for machine learning in rare diseases. Frontiers in medicine 8, 747612 (2021)

    work page 2021

  7. [37]

    Federated learning: Opportunities and challenges

    Mammen, P.M. Federated learning: Opportunities and challenges. arXiv preprint arXiv:2101.05428 (2021)

    arXiv 2021

  8. [38]

    Doiron, D. et al. Data harmonization and federated analysis of population-based studies: the BioSHaRE project. Emerg Themes Epidemiol 10, 12 (2013)

    work page 2013

Show all 43 references
  1. [39]

    Wilson, R.C. et al. DataSHIELD – New Directions and Dimensions. Data Science Journal 16(2017)

    2017

  2. [40]

    Nasirigerdeh, R. et al. sPLINK: A Federated, Privacy-Preserving Tool as a Robust Alternative to Meta-Analysis in Genome-Wide Association Studies. BioRxiv (2020)

    2020

  3. [41]

    Marcon, Y. et al. Orchestrating privacy-protected big data analyses of data from different resources with R and DataSHIELD. PLoS Comput Biol 17, e1008880 (2021)

    2021

  4. [42]

    Wang, J. & Ma, F. Federated learning for rare disease detection: a survey. Rare Disease and Orphan Drugs Journal 16(2023)

    2023

  5. [43]

    Li, T. et al. Federated optimization in heterogeneous networks. Proceedings of Machine learning and systems 2, 429-450 (2020)

    2020

  6. [44]

    Leek, J.T. et al. Tackling the widespread and critical impact of batch effects in high-throughput data. Nature Reviews Genetics 11, 733-739 (2010)

    2010

  7. [45]

    & Dekker, F.W

    Jager, K.J., Zoccali, C., Macleod, A. & Dekker, F.W. Confounding: what it is and how to deal with it. Kidney Int 73, 256-60 (2008)

    2008

  8. [46]

    & Scholte, H.S

    Snoek, L., Miletić, S. & Scholte, H.S. How to control for confounds in decoding analyses of neuroimaging data. Neuroimage 184, 741-760 (2019)

    2019

  9. [47]

    & Blettner, M

    Brenner, H. & Blettner, M. Controlling for continuous confounders in epidemiologic research. Epidemiology 8, 429-434 (1997)

    1997

  10. [48]

    & Waugh, F.V

    Frisch, R. & Waugh, F.V. Partial time regressions as compared with individual trends. Econometrica: Journal of the Econometric Society, 387-401 (1933)

    1933

  11. [49]

    Maitre, L. et al. Human Early Life Exposome (HELIX) study: a European population-based exposome cohort. BMJ Open 8, e021311 (2018)

    2018

  12. [50]

    Cao, H. et al. dsMTL - a computational framework for privacy-preserving, distributed multi-task machine learning. Bioinformatics (2022)

    2022

  13. [51]

    Gaye, A. et al. DataSHIELD: taking the analysis to the data, not the data to the analysis. International journal of epidemiology 43, 1929-1944 (2014)

    2014

  14. [52]

    Jaddoe, V.W.V. et al. The LifeCycle Project-EU Child Cohort Network: a federated analysis infrastructure and harmonized data of more than 250,000 children and parents. Eur J Epidemiol 35, 709-724 (2020)

    2020

  15. [53]

    Regression shrinkage and selection via the lasso

    Tibshirani, R. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society. Series B (Methodological), 267-288 (1996)

    1996

  16. [54]

    & Hastie, T

    Zou, H. & Hastie, T. Regularization and variable selection via the elastic net. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 67, 301-320 (2005)

    2005

  17. [55]

    Regression shrinkage and selection via the lasso: a retrospective

    Tibshirani, R. Regression shrinkage and selection via the lasso: a retrospective. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 73, 273-282 (2011)

    2011

  18. [56]

    Chernozhukov, V. et al. Double/debiased machine learning for treatment and structural parameters. (Oxford University Press Oxford, UK, 2018)

    2018

  19. [57]

    & Chernozhukov, V

    Urminsky, O., Hansen, C. & Chernozhukov, V. Using Double-Lasso Regression for Principled Variable Selection. SSRN Electronic Journal (2016)

    2016

  20. [58]

    Warembourg, C. et al. Urban environment during early-life and blood pressure in young children. Environ Int 146, 106174 (2021)

    2021

  21. [59]

    Warembourg, C. et al. Early-Life Environmental Exposures and Blood Pressure in Children. J Am Coll Cardiol 74, 1317-1328 (2019)

    2019

  22. [60]

    Vrijheid, M. et al. Advancing tools for human early lifecourse exposome research and translation (ATHLETE): Project overview. Environ Epidemiol 5, e166 (2021)

    2021

  23. [61]

    & Schwarz, E

    Cao, H., Chen, J., Meyer-Lindenberg, A. & Schwarz, E. A polygenic score for schizophrenia predicts glycemic control. Transl Psychiatry 7, 1295 (2017)

    2017

  24. [62]

    Elze, M.C. et al. Comparison of propensity score methods and covariate adjustment: evaluation in 4 cardiovascular studies. Journal of the American College of Cardiology 69, 345-357 (2017)

    2017

  25. [63]

    Gradient methods for minimizing composite functions

    Nesterov, Y. Gradient methods for minimizing composite functions. Mathematical Programming 140, 125-161 (2012)

    2012

  26. [64]

    & Teboulle, M

    Beck, A. & Teboulle, M. A fast iterative shrinkage-thresholding algorithm for linear inverse problems. SIAM journal on imaging sciences 2, 183-202 (2009)

    2009

  27. [65]

    & Jieping, Y

    Liu, J. & Jieping, Y. Efficient L1/Lq Norm Regularization

  28. [66]

    & Yin, W

    Peng, Z., Xu, Y., Yan, M. & Yin, W. ARock: An Algorithmic Framework for Asynchronous Parallel Coordinate Updates. SIAM Journal on Scientific Computing 38, A2851-A2879 (2016)

    2016

  29. [67]

    Random forests

    Breiman, L. Random forests. Machine learning 45, 5-32 (2001)

    2001

  30. [68]

    & Lin, C.-J

    Chang, C.-C. & Lin, C.-J. LIBSVM: a library for support vector machines. ACM transactions on intelligent systems and technology (TIST) 2, 1-27 (2011)

    2011

  31. [69]

    Tamayo-Uria, I. et al. The early-life exposome: description and patterns in six European countries. Environment international 123, 189-200 (2019)

    2019

  32. [70]

    Mullie, L. et al. CODA: an open-source platform for federated analysis and machine learning on distributed healthcare data. Journal of the American Medical Informatics Association 31, 651- 665 (2024)

    2024

  33. [71]

    Wolfson, M. et al. DataSHIELD: resolving a conflict in contemporary bioscience—performing a pooled analysis of individual-level data without sharing the data. International journal of epidemiology 39, 1372-1382 (2010)

    2010

  34. [72]

    Packages developed in DataSHIELD

    community, D. Packages developed in DataSHIELD

  35. [73]

    Santos, S. et al. Applying the exposome concept in birth cohort research: a review of statistical approaches. European journal of epidemiology 35, 193-204 (2020). Supplementary Tables Table S1 – Description of the study population Variable name N (%) Min Q1 Median Q3 Max Mater...

    2020

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.