E3 PreliminaryPreliminaryPEM unclearMethods-PaperPeer-reviewedMachine draft
Causality on longitudinal data: Stable specification search in constrained structural equation modeling.
Rahmadi, Ridho, Groot, Perry, van Rijn, Marieke Hc et al. · Statistical methods in medical research · 2018 · DOI
Quick Summary
This study developed a new statistical method for finding cause-and-effect relationships in long-term health data, even when sample sizes are small. The researchers tested their method on data from people with chronic fatigue syndrome, Alzheimer's disease, and chronic kidney disease, and found it could identify stable patterns of how different health factors influence each other over time.
Why It Matters
Understanding causal relationships in ME/CFS is challenging because the disease involves multiple interconnected physiological systems that change over time. This methodological advance provides researchers with a more stable statistical tool to identify which factors actually cause changes in ME/CFS symptoms and disease progression, potentially leading to better targeted interventions.
Observed Findings
- The proposed algorithm achieved comparable or better performance than existing causal modeling approaches on simulated data with known ground truth.
- When applied to chronic fatigue syndrome longitudinal data, the method identified causal relationships between health variables that align with previous hypothesis-driven research.
- The approach revealed potential novel causal relationships in ME/CFS and other chronic disease datasets that had not been previously identified.
Inferred Conclusions
- Stability selection through subsampling substantially improves the robustness of causal model learning in finite longitudinal datasets.
- The multi-objective evolutionary approach successfully balances model fit and parsimony to identify biologically plausible causal structures.
- Existential causal modeling can integrate both exploratory discovery and prior knowledge constraints to generate clinically relevant hypotheses in chronic disease research.
Remaining Questions
- Which specific novel causal relationships identified in the ME/CFS data have biological plausibility and deserve experimental validation?
- How does this method perform with different sample sizes, missing data patterns, and measurement error typical in ME/CFS studies?
- Can findings from stability selection be prospectively validated in independent ME/CFS cohorts?
What This Study Does Not Prove
This is a methodological paper, not a clinical outcomes study; it does not prove specific causal relationships exist in ME/CFS, only that the statistical method can identify stable causal patterns when applied to real disease data. The stability of findings across subsamples does not confirm biological causation—it only indicates which relationships are most robust in the data. Findings still require validation through independent studies and mechanistic investigation.
Tags
Method Flag:Exploratory Only
Metadata
- DOI
- 10.1177/0962280217713347
- PMID
- 28657454
- Review status
- Machine draft
- Evidence level
- Early hypothesis, preprint, editorial, or weak support
- Last updated
- 8 April 2026
About the PEM badge: “PEM required” means post-exertional malaise was an explicit required diagnostic criterion for participant inclusion in this study — not that PEM was studied, observed, or discussed. Studies using criteria that do not require PEM (e.g. Fukuda, Oxford) are tagged “PEM not required”. How the atlas works →
Spotted an error in this entry? Report it →