Long-term outcomes and harms: natural history, deprescribing durability, and combination safety

Synthesis of what is actually known about (1) the multi-year natural course of RLS, with and without treatment, and (2) the long-term harms of the drugs patients on a de-escalation path will stay on or taper through - dopamine agonists, alpha-2-delta ligands (gabapentin, pregabalin, gabapentin enacarbil), opioids, and IV iron - including the combinations this knowledge base's own protocol relies on. All claims below cite a study note filename (in knowledge-base/studies/) or a PMID. Medical disclaimer: nothing here is individual medical advice. Any change to RLS medication, especially reducing or stopping a dopamine agonist, must be done only under physician supervision and never abruptly - dopamine agonist withdrawal syndrome (DAWS) can be severe and is refractory to treatment (2013-nirenberg-dopamine-agonist-withdrawal-review.md; 2026-lobinska-dopamine-agonist-withdrawal-syndrome-review.md).

What a skeptical neurologist would object to

1. The natural-history baseline is essentially one Finnish cohort, plus a qualitative echo. Documented spontaneous remission/fluctuation evidence rests almost entirely on a single 10-year population cohort with 55% follow-up participation (2011-juuti-ten-year-natural-course-remission.md), plus two small, retrospective, single-clinic long-term series (2012-lipford-pramipexole-eight-year-cohort.md; 2016-miranda-chile-long-term-follow-up.md) that describe treated rather than natural course. A 2007 narrative review from the project's own home jurisdiction independently states that RLS's course is "generally chronic with phases of spontaneous remission" (2007-droste-diederich-luxembourg-rls-review.md), but offers no quantitative rate. A separate 10-year US clinic cohort adds a partial second quantitative data point from the treatment side: even under continuous treatment, annual dopamine-agonist discontinuation and augmentation rates stayed roughly constant (not declining) through 10 years, while methadone showed neither (2011-silver-ten-year-dopamine-agonist-methadone.md). Without a robust untreated/general-population fluctuation rate, any claim that a non-drug intervention produced improvement is close to uninterpretable - roughly half of Juuti's frequent-RLS baseline cases remitted to rare-or-no symptoms over 10 years with no described intervention at all.
2. "Long-term" deprescribing data mostly means one month, not years. The most detailed recent dopamine-agonist-withdrawal success-rate data in this knowledge base (2026-martin-garcia-structured-dopamine-agonist-withdrawal.md) reports only 1-month outcomes. The best multi-year-adjacent evidence (2022-yeung-laiwah-augmentation-guideline-effectiveness.md) reaches a mean of 28 months (range 5-59) in a single-center, uncontrolled, retrospectively-scored case series of 63 patients - informative, but nowhere near a 5-10 year prospective cohort. No study meeting that bar was found despite multiple independent search sweeps.
3. The protocol's own rescue combination is the one drug pair with the clearest respiratory-harm signal, and regulators on both sides of the Atlantic agree. This knowledge base's tapering and opioid guidance routinely pairs an alpha-2-delta ligand with a low-dose opioid bridge (see tapering-deprescribing.md, opioids-other-drugs.md). That is exactly the combination the FDA warned about in a December 19, 2019 Drug Safety Communication (49 FAERS respiratory-depression cases, 2012-2017, 24% fatal; full warning text confirmed via WebSearch at https://www.fda.gov/safety/medical-product-safety-information/neurontin-gralise-horizant-gabapentin-and-lyrica-lyrica-cr-pregabalin-drug-safety-communication, accessed 2026-09-12; underlying pharmacovigilance analysis in 2020-shrestha-palaian-fda-gabapentinoid-respiratory-warning.md). The EU/UK regulatory system acted even earlier: per WebSearch, the EMA's Pharmacovigilance Risk Assessment Committee (PRAC) recommended in 2017 that gabapentin product information be amended to warn of respiratory depression, and the UK MHRA published a Drug Safety Update in October 2017 flagging the risk even without concomitant opioids - though this review could not locate the primary EMA/PRAC document URL directly and states this honestly rather than inventing a citation. A 2022 meta-analysis found the combination more than doubles mortality odds in non-randomized data (2022-hahn-gabapentinoid-opioid-combination-meta-analysis.md), and roughly a quarter of long-term opioid-treated US Medicare patients were still receiving it as of 2022, years after the warning (2026-raji-medicare-opioid-gabapentinoid-coprescribing-trends.md). Two independent gabapentinoid-misuse systematic reviews add a second layer to this same combination: opioid use disorder is consistently the single greatest risk factor for gabapentinoid misuse, and co-ingestion with opioids/sedatives is what turns an otherwise low-lethality gabapentinoid overdose dangerous (2017-bonnet-scherbaum-gabapentinoid-addiction-review.md; 2021-evoy-gabapentinoid-abuse-review-update.md).
4. Mortality evidence is genuinely contradictory, not just "mixed." One veteran cohort reports RLS almost doubling all-cause mortality (HR 1.88; 2016-molnar-veterans-cohort-mortality-comorbidity.md); two other analyses - including a systematic review restricted to the four highest-quality available cohorts - find no significant association at all (pooled HR 1.09, 95% CI 0.80-1.78; 2017-kendzerska-plms-mortality-systematic-review.md; 2012-szentkiralyi-four-cohort-mortality-null.md). A protocol cannot invoke the alarming number to motivate urgency and quietly drop it when it complicates the picture.
5. Repeated IV iron dosing has essentially no RLS-specific long-term safety data. The best-quality evidence on ferric carboxymaltose's most quantifiable long-term harm - hypophosphatemia - comes from general iron-deficiency-anemia trials, not RLS patients, and shows a single course causes hypophosphatemia in about half of recipients, persisting at 5 weeks in nearly a third (2018-wolf-ferric-carboxymaltose-hypophosphatemia-rct.md), with osteomalacia and fractures reported at the tail of cumulative dosing (2025-magagnoli-fcm-hypophosphatemia-systematic-review.md). A protocol built around repeated multi-year FCM dosing is extrapolating past its evidence.
6. The flagship device solution's "long-term" study is one year, unpublished. TOMAC's cost-effectiveness case rests on a US-payer Markov model (2023-ryschon-tomac-cost-effectiveness.md), and its manufacturer's own post-market study explicitly named for "long-term effectiveness and safety" defines its primary endpoint at 1 year and has posted no results as of September 2026 (2023-noctrix-thrive-tomac-long-term-observational-registration.md).
7. The evidence base leans on a drug that is not accessible in Europe. Gabapentin enacarbil - the alpha-2-delta ligand with the strongest guideline-cited augmentation-avoidance trial data - still lacks EMA marketing authorization (2019-wanner-nondopaminergic-vs-dopaminergic-europe-access.md; confirmed unchanged via WebSearch, accessed 2026-09-11). A Luxembourg-based patient cannot obtain the specific product much of the trial evidence describes.
8. Impulse control disorders and driving impairment are real but thinly quantified. ICD prevalence estimates (7-10%) and sleep-attack/driving-impairment rates (~10%) both come from small, single-center, largely self-reported cohorts (2011-voon-impulse-control-behaviours-cohort.md; 2011-dang-impulse-control-disorders-case-series.md; 2012-lipford-pramipexole-eight-year-cohort.md). No prospective, multi-center, denominator-based incidence study exists, and medico-legal/financial harm from ICDs is documented only qualitatively in a 12-patient case series. A further, more general search for RLS-specific motor-vehicle-crash or driving-impairment cohort data (as opposed to self-reported sleep attacks) found no additional denominator-based studies beyond what is already cited here.
9. Gabapentinoid dependence risk is real but not RLS-specific, and "long-term dose stability" data is survivorship-biased. Two systematic reviews of gabapentinoid misuse (106 and 55 studies respectively) both conclude that general-population addiction risk is low but concentrated in patients with prior substance-use disorder, especially opioid users - exactly this project's own taper-bridge population (2017-bonnet-scherbaum-gabapentinoid-addiction-review.md; 2021-evoy-gabapentinoid-abuse-review-update.md). Neither review is RLS-specific. Separately, a US claims cohort found early-onset RLS patients started on a gabapentinoid had a higher adjusted incidence of substance-use-disorder diagnoses over 2 years than those started on a dopamine agonist (2023-costales-mental-health-gabapentinoid-vs-agonist.md) - one of the only RLS-specific data points on this question, and still only 2 years, not 5-10. Multi-year dose-stability figures this knowledge base cites (e.g., a 2-year rotigotine extension with only a 2.9% dose increase, 2010-hogl-rotigotine-two-year-extension.md) describe only the minority of patients who remained in the trial: a real-world US claims cohort found just 22.3% of newly treated early-onset RLS patients stayed on any single first-line drug for even one full year (2022-costales-early-onset-treatment-persistence.md), meaning trial-extension "stability" figures are not representative of typical patient trajectories.

Combination-safety table

What is established (grade A/B)

• Meaningful spontaneous remission of RLS occurs over years without necessarily invoking any specific intervention: in a 10-year Finnish population cohort, RLS prevalence fell from 21% to 15% with age, and roughly half of participants with frequent RLS at baseline reported rare-or-no symptoms a decade later (2011-juuti-ten-year-natural-course-remission.md).
• The opioid-plus-gabapentinoid combination this project's own protocol uses as an augmentation-rescue bridge (tapering-deprescribing.md, opioids-other-drugs.md) carries a meta-analytically confirmed increase in dizziness, cognitive dysfunction, respiratory depression, and (in non-randomized data) mortality (2022-hahn-gabapentinoid-opioid-combination-meta-analysis.md).
• Opioid-plus-gabapentinoid co-prescribing is independently associated with higher fall risk in a large Medicare cohort of older adults (adjusted HR 1.18, 95% CI 1.02-1.37 vs a benzodiazepine-plus-opioid reference; 2021-shah-raji-opioid-gabapentinoid-falls-medicare.md).
• A single course of intravenous ferric carboxymaltose - the iron formulation this knowledge base's own protocol relies on most heavily - causes hypophosphatemia in roughly half of recipients in a large head-to-head RCT, persisting at 5 weeks in nearly a third, via an FGF23-mediated mechanism (2018-wolf-ferric-carboxymaltose-hypophosphatemia-rct.md).
• Low-dose opioids used as an augmentation-rescue bridge are dose-stable, not escalating, over 2 years in most patients (median dose change 0 morphine milligram equivalents; 2023-winkelman-national-rls-opioid-registry.md), though this does not offset the combination-safety risks documented above when a gabapentinoid is co-prescribed.
• Gabapentinoid misuse/dependence is a real, systematically documented risk that is concentrated almost entirely in patients with current or past substance use disorder (especially opioid use), not the general RLS population - a consistent finding across two independent systematic reviews spanning 2017 and its 2021 update (2017-bonnet-scherbaum-gabapentinoid-addiction-review.md; 2021-evoy-gabapentinoid-abuse-review-update.md).

Emerging C

• Long-term (up to 12-year) dopamine agonist treatment carries accumulating, not one-time, risk: augmentation reached 42% of a Mayo Clinic cohort followed a mean of 8 years, impulse control disorders affected 10%, and 10% self-reported sleep attacks while driving (2012-lipford-pramipexole-eight-year-cohort.md). An independent Chilean series with a mean 6.3-year follow-up found an even higher rate (66%) of augmentation-pattern worsening (2016-miranda-chile-long-term-follow-up.md), suggesting trial-based augmentation percentages likely understate real-world, multi-year risk.
• Guideline-based cross-titration off a dopamine agonist works for most patients followed well beyond one month: 78% of 63 augmented patients were responders at a mean 28-month (range 5-59 month) follow-up, and responders were more likely to have fully discontinued the dopamine agonist (2022-yeung-laiwah-augmentation-guideline-effectiveness.md) - the best "years, not weeks" outcome data currently in this knowledge base, though still a single-center, uncontrolled case series.
• Real-world opioid-gabapentinoid co-prescribing has not meaningfully declined since the FDA's 2019 respiratory-depression warning: roughly a quarter of long-term-opioid Medicare beneficiaries still received a concurrent gabapentinoid in 2022, with about a third starting at 1200 mg/day or higher (2026-raji-medicare-opioid-gabapentinoid-coprescribing-trends.md).
• Gabapentinoid harms in older adults extend beyond falls and sedation to include respiratory depression and pneumonia, cognitive and functional impairment, and a possible (unquantified) atrial fibrillation signal, per an independent German geriatric-psychiatry review (2025-gahr-gabapentinoids-geriatric-psychiatry-risks.md).
• Ferric-carboxymaltose-associated hypophosphatemia has real downstream clinical consequences - general weakness, fatigue, bone pain, muscle pain, osteomalacia, and fractures - reported across the wider literature at rates of 50-92% versus 2-8% with alternative IV iron formulations, prompting a 2023 FDA label update that a recent systematic review argues still under-specifies monitoring needs (2025-magagnoli-fcm-hypophosphatemia-systematic-review.md).
• Real-world adherence to de-escalation guidance remains poor even where evidence supports it: 85% of a 2025 Spanish cohort of augmented patients remained on a dopamine agonist, and only 20% of those had a dose reduction (2025-martin-garcia-spanish-augmentation-cohort.md).
• A 10-year US clinic cohort found dopamine-agonist discontinuation and augmentation rates stay roughly constant year-over-year rather than declining, so risk keeps accumulating for as long as a patient remains on the drug (annual augmentation ~7% for pramipexole after year 1, persisting through year 10; only 58% remained on pramipexole past 5 years); methadone, used as a comparator, showed neither augmentation nor efficacy loss over the same decade (2011-silver-ten-year-dopamine-agonist-methadone.md).
• Multi-year dose stability has been directly demonstrated for at least one dopamine agonist: a 2-year open-label rotigotine extension found only a 2.9% mean dose increase from year 1 to year 2, with 64.7% of patients completing the full 2 years (2010-hogl-rotigotine-two-year-extension.md) - though the same drug's longer 5-year extension shows augmentation emerging later (13% by year 5; 2014-bogan-rotigotine-bench-to-bedside-review.md), so 2-year dose stability should not be read as proof augmentation risk has passed.
• Real-world treatment persistence is far lower than trial-extension data suggests: in a US claims cohort of newly treated early-onset RLS patients, only 22.3% remained on their initially prescribed monotherapy for a full year, regardless of drug class (2022-costales-early-onset-treatment-persistence.md) - a reminder that most published "long-term" outcome data describes a self-selected subset of patients who stayed in treatment, not a representative trajectory.
• A gabapentinoid-versus-dopamine-agonist comparative cohort in early-onset RLS found a higher adjusted incidence of substance-use-disorder diagnoses after starting a gabapentinoid, even though the broader composite mental-health-episode difference was not statistically significant after adjustment (2023-costales-mental-health-gabapentinoid-vs-agonist.md).

Contested or negative findings

• RLS's association with all-cause mortality is inconsistent across the best available evidence. One veteran cohort reports an 88% relative increase (HR 1.88; 2016-molnar-veterans-cohort-mortality-comorbidity.md); a meta-analysis of 8 studies found a smaller but still significant confounder-adjusted increase (HR 1.52; 2017-katsanos-meta-analysis-cardiovascular-mortality.md); but four separate general-population cohorts spanning Germany and the USA found no association at all (HRs 0.21-1.07; 2012-szentkiralyi-four-cohort-mortality-null.md), and a systematic review's meta-analysis of the four highest-quality cohorts likewise found no significant pooled effect (HR 1.09, 95% CI 0.80-1.78; 2017-kendzerska-plms-mortality-systematic-review.md) - though periodic limb movements in sleep specifically did show a consistent mortality/cardiovascular signal across all studies examining them.
• The cerebrovascular/cardiovascular event association with RLS does not survive confounder adjustment in the best available meta-analysis, even though the same meta-analysis found a significant adjusted all-cause mortality association - an internally nuanced result that resists a simple "RLS causes heart disease" or "RLS is harmless" headline either way (2017-katsanos-meta-analysis-cardiovascular-mortality.md).
• No RLS-specific incidence figure for dopamine agonist withdrawal syndrome exists; the widely cited 24% figure comes from Parkinson disease taper cohorts, where typical dopamine agonist doses are much higher than in RLS, and may not generalize (2026-lobinska-dopamine-agonist-withdrawal-syndrome-review.md).
• Cost-effectiveness modeling for device therapy (TOMAC) concludes it is cost-effective or cost-saving, but only within a US-payer Markov model built on a single trial's effect sizes and partly authored by the device manufacturer's own health-economics consultancy (2023-ryschon-tomac-cost-effectiveness.md); no equivalent European analysis was found.

Mechanistic picture in 5 bullet points

1. Chronic dopamine agonist exposure (D3-receptor desensitization, compensatory D1-receptor upregulation) is the leading mechanistic explanation for augmentation and appears to leave a lasting signature: even dopamine-agonist-pretreated patients not yet meeting formal augmentation criteria respond worse to subsequent non-dopaminergic drugs than treatment-naive patients (see dopamine-agonists-augmentation.md for the full mechanistic evidence base).
2. Impulse control disorders and dopamine agonist withdrawal syndrome likely share an underlying mesolimbic reward-circuitry mechanism: ICDs are reported as a major risk factor for DAWS, and DAWS is refractory to non-dopaminergic rescue drugs, consistent with a relative dopaminergic-reward deficiency once the agonist is reduced or stopped (2013-nirenberg-dopamine-agonist-withdrawal-review.md).
3. Ferric-carboxymaltose-induced hypophosphatemia is mechanistically explained by a sharp, transient rise in biologically active fibroblast growth factor 23 (FGF23), which drives renal phosphate wasting, lowers calcitriol and calcium, and raises parathyroid hormone - a pathway shared with hereditary hypophosphatemic disorders (2018-wolf-ferric-carboxymaltose-hypophosphatemia-rct.md).
4. Gabapentinoid-opioid respiratory depression is understood as additive-to-synergistic CNS and brainstem-respiratory-center depression, with risk concentrated in patients who already carry a respiratory risk factor (age, COPD, renal impairment) or are combining the drugs with a third CNS depressant - 92% of FDA-reported fatal/serious cases had such a factor (2020-shrestha-palaian-fda-gabapentinoid-respiratory-warning.md).
5. Gabapentin's markedly prolonged elimination half-life in renal impairment (5-9h to as much as 132h in dialysis) is a straightforward pharmacokinetic accumulation mechanism, distinct from the opioid-combination and FGF23 mechanisms above, and explains why standard non-renal dosing can produce severe, reversible neurotoxicity in this specific, RLS-enriched population (2022-lehmann-gabapentin-overdose-dialysis-case-report.md).

Implications for a non-drug long-term plan for people currently on dopamine agonists or alpha-2-delta ligands

• Set expectations using the best available natural-history data, not silence: roughly half of people with frequent RLS symptoms in a general population can be expected to see substantial spontaneous improvement over a decade even without a specific new intervention (2011-juuti-ten-year-natural-course-remission.md). Any non-drug protocol's outcome claims should be framed against this base rate, not presented as if all improvement is attributable to the protocol.
• Treat the opioid-plus-gabapentinoid rescue combination as requiring active, ongoing respiratory and fall-risk monitoring, not a one-time prescribing decision: document risk factors, counsel explicitly on the FDA's warning, and do not assume the combination is being avoided elsewhere in a patient's care (2022-hahn-gabapentinoid-opioid-combination-meta-analysis.md; 2020-shrestha-palaian-fda-gabapentinoid-respiratory-warning.md; 2021-shah-raji-opioid-gabapentinoid-falls-medicare.md).
• Screen for impulse control disorders and driving impairment at baseline and at every dopamine agonist dose change, not only when augmentation is suspected - both risks accumulate with years of exposure, not just with dose escalation at augmentation (2012-lipford-pramipexole-eight-year-cohort.md; 2011-voon-impulse-control-behaviours-cohort.md).
• If repeated IV ferric carboxymaltose dosing is part of the plan, add serum phosphate to the monitoring panel alongside ferritin/transferrin saturation, and consider an alternative formulation (ferric derisomaltose) for patients expected to need multiple courses over years (2018-wolf-ferric-carboxymaltose-hypophosphatemia-rct.md; 2025-magagnoli-fcm-hypophosphatemia-systematic-review.md).
• Adjust gabapentinoid dosing aggressively for renal function and consider closer monitoring in dialysis or CKD patients, a population with both elevated RLS prevalence and elevated gabapentinoid-accumulation risk (2022-lehmann-gabapentin-overdose-dialysis-case-report.md).
• Any tapering or dose-reduction step for a dopamine agonist must be physician-supervised, slow, and individualized, with explicit screening for pre-existing impulse control disorder as a DAWS risk marker; never present abrupt discontinuation as an option (2013-nirenberg-dopamine-agonist-withdrawal-review.md; 2026-lobinska-dopamine-agonist-withdrawal-syndrome-review.md).
• For a Luxembourg/EU-based plan specifically, do not assume access to gabapentin enacarbil (not EMA-authorized) or to US-style TOMAC/Nidra reimbursement pathways; default to off-label gabapentin or pregabalin dosing and treat device cost-effectiveness as unproven in a European payer context until a European analysis exists (2019-wanner-nondopaminergic-vs-dopaminergic-europe-access.md; 2023-ryschon-tomac-cost-effectiveness.md). Targeted French- and German-language web searches for Luxembourg CNS (Caisse Nationale de Sante) reimbursement of IV ferric carboxymaltose specifically for RLS, and for German statutory reimbursement of RLS devices, did not surface a citable, RLS-specific reimbursement determination as of September 2026; this should be stated honestly as unconfirmed rather than assumed either way.

Open questions

• No prospective, multi-center cohort following patients through dopamine-agonist de-escalation or gabapentinoid maintenance for 5-10 years was found despite repeated, independently run search sweeps. The closest candidate, a manufacturer-sponsored device post-market study, defines "long-term" as one year and has posted no results (2023-noctrix-thrive-tomac-long-term-observational-registration.md).
• What is the true, RLS-specific (not Parkinson-disease-extrapolated) incidence of dopamine agonist withdrawal syndrome during a slow, guideline-concordant taper, and does iron repletion or a specific taper rate reduce it? Not established.
• Does long-term (years, repeated-dose) IV ferric carboxymaltose use in RLS patients specifically show the same ~50% hypophosphatemia rate seen in single-course, non-RLS iron-deficiency-anemia trials, and does cumulative dosing raise osteomalacia/fracture risk in this population? No RLS-specific long-term iron-safety cohort was found.
• Is the elevated fall risk seen with opioid-gabapentinoid co-prescribing in a general Medicare chronic-pain population (2021-shah-raji-opioid-gabapentinoid-falls-medicare.md) similar in magnitude in RLS patients specifically, who may differ in baseline frailty and treatment duration?
• What is a denominator-based, prospective, multi-center incidence rate for impulse control disorders and for sleep-attack-related driving impairment on dopamine agonists, and how often does either resolve after dose reduction versus complete withdrawal? Only small, single-center, largely self-reported data exist.
• Is there any citable, RLS-specific cost-effectiveness or reimbursement analysis for IV iron, gabapentin enacarbil, or neuromodulation devices in an EU/Luxembourg payer context? None was found in PubMed or in French/German-language web searches conducted for this review, and a repeat PubMed search ("restless legs syndrome" AND cost-effectiveness/economic evaluation/budget impact AND Europe) run for this update again returned no RLS-specific European economic evaluation.
• Is there a prospective, denominator-based cohort quantifying motor-vehicle crash risk (not just self-reported sleep attacks or driving impairment) in RLS patients on dopamine agonists? A dedicated PubMed search run for this update found only case reports, letters, and Parkinson-disease-focused sleep reviews, not an RLS-specific crash-incidence study.
• Does repeated, long-term (years) intravenous iron dosing in RLS patients specifically carry an iron overload (as opposed to hypophosphatemia) risk requiring ferritin-ceiling monitoring? A dedicated PubMed search for this update found no RLS-specific long-term iron-overload safety study.
• Does all-cause mortality risk in RLS vary meaningfully by treatment status (untreated vs dopamine-agonist-treated vs alpha-2-delta-treated), which none of the mortality cohorts in this knowledge base were designed to test?

Notes in this topic

New notes written for this topic (long-term-outcomes-harms-and-combination-safety):

• 2012-lipford-pramipexole-eight-year-cohort.md - Mayo Clinic 8-year (up to 12-year) pramipexole cohort: 42% augmentation, 10% ICD, 10% self-reported sleep attacks while driving (PMID 23036265).
• 2016-miranda-chile-long-term-follow-up.md - Chilean 6.3-year (mean) retrospective series: 66% augmentation-pattern worsening, 6% ICD (PMID 28393990).
• 2021-shah-raji-opioid-gabapentinoid-falls-medicare.md - 47,964-patient Medicare cohort: opioid+gabapentinoid co-prescribing raises fall risk (aHR 1.18) (PMID 35476819).
• 2026-raji-medicare-opioid-gabapentinoid-coprescribing-trends.md - Medicare trends 2017-2022: opioid-gabapentinoid co-prescribing unchanged (~25-28%) despite FDA warning (PMID 42506314).
• 2025-gahr-gabapentinoids-geriatric-psychiatry-risks.md - German review: falls, respiratory depression/pneumonia, cognitive impairment, possible atrial fibrillation with gabapentinoids in elderly (PMID 40504234).
• 2020-shrestha-palaian-fda-gabapentinoid-respiratory-warning.md - FDA FAERS review: 49 respiratory-depression cases (2012-2017), 24% fatal, underlying the December 2019 warning (PMID 33728039).
• 2018-wolf-ferric-carboxymaltose-hypophosphatemia-rct.md - Landmark 1997-patient RCT: FCM causes hypophosphatemia in 50.8% vs 0.9% with ferumoxytol, via FGF23 (PMID 30518682).
• 2025-magagnoli-fcm-hypophosphatemia-systematic-review.md - Systematic review: FCM-hypophosphatemia causes osteomalacia/fractures; FDA label updated 2023 (PMID 39935027).
• 2012-szentkiralyi-four-cohort-mortality-null.md - Four prospective cohorts (Germany/USA): RLS not associated with all-cause mortality (PMID 23129573).
• 2017-kendzerska-plms-mortality-systematic-review.md - Systematic review/meta-analysis: null pooled all-cause mortality HR (1.09) for RLS; PLMS independently associated (PMID 28329227).
• 2019-wanner-nondopaminergic-vs-dopaminergic-europe-access.md - Review confirming gabapentin enacarbil is not available in Europe (PMID 31229171).
• 2023-noctrix-thrive-tomac-long-term-observational-registration.md - TOMAC "long-term" post-market study (NCT06076499): 1-year primary endpoint, no results posted as of September 2026.
• 2011-silver-ten-year-dopamine-agonist-methadone.md - 10-year clinic cohort: pramipexole/pergolide augmentation and discontinuation rates stay constant, not declining, through year 10; methadone shows neither (PMID 21239226).
• 2023-costales-mental-health-gabapentinoid-vs-agonist.md - Early-onset RLS cohort: gabapentinoid initiation linked to higher adjusted substance-use-disorder incidence than dopamine-agonist initiation over 2 years (PMID 37708806).
• 2017-bonnet-scherbaum-gabapentinoid-addiction-review.md - Systematic review (106 studies): gabapentinoid addiction risk low in general population, concentrated in opioid/SUD patients (PMID 28988943).
• 2021-evoy-gabapentinoid-abuse-review-update.md - Systematic review update (55 studies, 2016-2020): rising gabapentinoid misuse, opioid use disorder the dominant risk factor, increased hospitalization/overdose-mortality signal (PMID 33215352).
• 2010-hogl-rotigotine-two-year-extension.md - 2-year rotigotine open-label extension: dose stable (+2.9% year-over-year), 64.7% completion, augmentation not assessed (PMID 20920156).
• 2007-droste-diederich-luxembourg-rls-review.md - Luxembourg-authored narrative review: RLS course "generally chronic with phases of spontaneous remission" (PMID 17953080).
• 2022-costales-early-onset-treatment-persistence.md - US claims cohort: only 22.3% of newly treated early-onset RLS patients persisted on any first-line monotherapy for a full year (PMID 35605349).

Pre-existing notes drawn on for this topic:

• 2011-juuti-ten-year-natural-course-remission.md - 10-year Finnish population cohort: RLS prevalence 21%->15%, ~50% remission in frequent-RLS cases (PMID 21725862).
• 2016-molnar-veterans-cohort-mortality-comorbidity.md - US veterans cohort: RLS associated with 88% higher all-cause mortality (PMID 26377112).
• 2017-katsanos-meta-analysis-cardiovascular-mortality.md - Meta-analysis: adjusted all-cause mortality HR 1.52 significant; cerebrovascular/cardiovascular association not significant after adjustment (PMID 28948600).
• 2022-hahn-gabapentinoid-opioid-combination-meta-analysis.md - Meta-analysis: opioid+gabapentinoid combination raises dizziness, cognitive dysfunction, respiratory depression, and non-RCT mortality risk (PMID 36304170).
• 2022-lehmann-gabapentin-overdose-dialysis-case-report.md - Case series: gabapentin accumulation/overdose in peritoneal dialysis patients due to 15-25x prolonged half-life (PMID 36518357).
• 2025-richards-gabapentin-enacarbil-dementia-agitation-rct.md - RCT in dementia patients: gabapentin enacarbil improves agitation/sleep but with a non-significant falls trend (PMID 39909068).
• 2011-dang-impulse-control-disorders-case-series.md - 12-patient case series: devastating impulse control disorders on dopamine agonists, some persisting after withdrawal (PMID 21412200).
• 2011-voon-impulse-control-behaviours-cohort.md - German cohort: 7.1% impulse control behavior prevalence on dopaminergic therapy (PMID 21955669).
• 2013-nirenberg-dopamine-agonist-withdrawal-review.md - Review: DAWS is severe, refractory, and ICD is a major risk factor (PMID 23686524).
• 2026-lobinska-dopamine-agonist-withdrawal-syndrome-review.md - Review: DAWS up to 24% in Parkinson disease taper cohorts, RLS-specific rate unknown (PMID 41870480).
• 2013-garcia-borreguero-irlssg-long-term-guidelines.md - IRLSSG long-term guideline: evidence graded by duration, up to 5 years for rotigotine (PMID 23859128).
• 2021-tambon-gabapentinoid-addictovigilance-france.md - French pharmacovigilance: rising pregabalin-specific misuse and mortality signal (PMID 33613345).
• 2023-ryschon-tomac-cost-effectiveness.md - US Markov model: TOMAC cost-effective/cost-saving vs sham over 3-year and lifetime horizons (PMID 37824047).
• 2021-sieminski-therapy-satisfaction-willingness-to-pay.md - Polish cohort: RLS therapy costs ~3% of household income; willingness to pay up to 8.3% (PMID 34333733).
• 2023-iqwig-nichtmedikamentoese-verfahren-rls.md - German HTA: non-drug RLS interventions underpowered; German reimbursement context for iron and devices.
• 2022-yeung-laiwah-augmentation-guideline-effectiveness.md - 63-patient cohort followed 5-59 months: 78% responders to guideline-based cross-titration off dopamine agonists (PMID 35532181).
• 2026-martin-garcia-structured-dopamine-agonist-withdrawal.md - Structured 1-month DA withdrawal protocol: 77% complete withdrawal success (PMID 42673880).
• 2025-martin-garcia-spanish-augmentation-cohort.md - Real-world gap: 85% of augmented patients remain on a dopamine agonist (PMID 41130145).
• 2011-ellenbogen-gabapentin-enacarbil-52-week-extension.md - 52-week extension: one death (fall, 25 days post-discontinuation, judged unrelated) among 573 subjects (PMID 21242741).
• 2010-bogan-gabapentin-enacarbil-long-term-maintenance.md - RCT: gabapentin enacarbil maintains efficacy for up to 9 months (PMID 20511481).
• 2026-phan-iron-therapy-narrative-review.md - Review: IV FCM effective and "generally mild" adverse events including transient hypophosphatemia (PMID 42371291).
• 2023-winkelman-national-rls-opioid-registry.md - 448-patient, 2-year registry: opioid dose stable (median change 0 MME) (PMID 36697248).
• 2014-bogan-rotigotine-bench-to-bedside-review.md - Review including a 5-year rotigotine open-label extension: 57% premature discontinuation, 13% clinically significant augmentation by year 5 (PMID 24636821).
• 2017-evoy-gabapentinoid-abuse-misuse-review.md - Original 2017 systematic review (59 studies): 1.6% general-population gabapentinoid abuse prevalence, rising to 3-68% among opioid abusers (PMID 28144823).