REM-suppressing antidepressants linked to improved survival in ALS patients
ALS patients who were prescribed antidepressants that suppress REM sleep showed modestly better survival rates than those on other antidepressants, according to a large retrospective cohort study published in Neurological Sciences. But the authors caution the association is statistically fragile and may reflect differences in why the drugs were prescribed rather than a direct biological effect.
What the Study Found
Researchers led by Cosmo Fowler at Emory University, in collaboration with Piedmont Atlanta and MetroHealth/Case Western Reserve, used the TriNetX U.S. Collaborative Network to identify 14,441 ALS patients with a riluzole prescription between May 2014 and May 2024. Among them, 4,177 were taking REM-suppressing antidepressants (RSAs) and 880 were on non-REM-suppressing antidepressants (NRSAs).
REM sleep is known to be a period of physiological vulnerability for patients with neuromuscular disease. During REM, the brain imposes generalized loss of skeletal muscle tone that spares only the diaphragm and extraocular muscles. For ALS patients with already compromised respiratory function, each REM episode may carry an elevated risk of hypoventilation, oxygen desaturation, or aspiration. Several antidepressant classes, including certain tricyclics, MAOIs, and SSRIs, are known to reduce REM sleep fraction on polysomnography, sometimes by 30-50%. The study tested whether using these drugs might translate into a survival benefit by reducing cumulative time spent in this physiologically precarious state.
In the unmatched analysis, the NRSA cohort had a 1.28 times higher risk of death (HR 1.28, 95% CI 1.11-1.46). One-year survival was 68.61% in the RSA group versus 60.73% in the NRSA group, a difference that was statistically significant (log-rank p<0.001). The relative risk reduction was roughly 13% at one year.
After propensity score matching to control for demographic and clinical differences between the groups, the gap narrowed: 65.92% versus 60.97%, with a p-value of 0.035, borderline by conventional standards. A risk difference analysis found the result was no longer statistically significant (RR 1.07, 95% CI 0.92-1.25), meaning the finding could plausibly be due to chance.
Why It Matters
The idea is biologically plausible. REM sleep produces near-complete skeletal muscle atonia, and in ALS patients with compromised respiratory function, each REM episode carries a risk of hypoventilation or aspiration. Reducing REM fraction could, in theory, reduce cumulative exposure to that risk over months and years.
But ALS is also a disease in which depression and anxiety are common, and the choice of antidepressant is rarely random. Clinicians may prescribe different classes of antidepressants based on symptom profiles, comorbid conditions, and side effect tolerability. Those same clinical factors may independently affect survival, making it difficult to disentangle drug effects from patient selection.
The study touches on a broader open question in sleep medicine: whether modifying REM sleep duration through pharmacologic means produces net clinical benefit or harm. For most healthy individuals, REM sleep is considered essential for emotional regulation and memory consolidation. But for patients with neuromuscular respiratory compromise, the risk-benefit calculus may be different.
Limitations
The authors are explicit that this is a hypothesis-generating study, not a demonstration of causation. The association was modest, and the statistical significance disappeared in the more conservative risk difference analysis. Residual confounding by indication, the possibility that the underlying reason for choosing one antidepressant over another drives the survival difference, not the drug’s effect on REM sleep, cannot be ruled out.
The study is also retrospective and relies on electronic health record data from the TriNetX network, which may contain coding errors and incomplete medication histories. The use of riluzole prescription as an inclusion criterion is a proxy for ALS diagnosis but may not capture all cases. And while the analysis adjusted for many covariates, unmeasured confounders such as disease severity at baseline, pulmonary function, and non-pharmacologic interventions could still bias the results.
Bottom Line
This large observational study found that ALS patients taking REM-suppressing antidepressants had modestly better one-year survival than those on non-REM-suppressing antidepressants. The association was statistically fragile and attenuated after propensity score matching, leaving causation unproven. The finding is best interpreted as hypothesis-generating, a signal that warrants prospective investigation, not a basis for changing clinical practice.
Source
Fowler C, et al. REM-suppressing antidepressants and survival in amyotrophic lateral sclerosis. Neurol Sci. 2026 Jun 21;47(8):646. DOI: 10.1007/s10072-026-09202-1. PMID: 42472717. Funding: UM1TR004528 (Clinical and Translational Science Collaborative of Cleveland, Case Western Reserve University).