The Evidence Base

In a German insurance-claims cohort spanning roughly 10.7–14.3 million antibiotic treatment episodes, fluoroquinolone use was associated with modestly elevated overall risk of neurological and neuropsychiatric outcomes compared with reference antibiotics (adjusted relative risks approximately 1.04–1.10, comparator-dependent). Drug-induced polyneuropathy showed a stronger association — adjusted hazard ratio of 1.68 — with excess risk concentrated in the first 92 days after exposure.

Population-scale human evidence, not a case series or a spontaneous-report signal, that fluoroquinolone exposure carries measurable, timing-specific neurological risk.

Using US MarketScan commercial and Medicare supplemental claims data, the authors 1:1 propensity-score matched 3,174,620 patients (1,587,310 per group) who filled an oral fluoroquinolone or macrolide prescription. Crude incidence of aortic aneurysm or dissection within the 60-day follow-up was 1.9 cases per 1000 person-years among fluoroquinolone users versus 1.2 cases per 1000 person-years among macrolide users; a subsequent replication study reports the study's adjusted hazard ratio as 1.34 (95% CI 1.17–1.54).

A large, propensity-matched US nationwide cohort contributing to an observational literature on fluoroquinolones and aortic aneurysm/dissection risk that is not uniformly settled — other studies using different comparators and populations have reported smaller, larger, or null associations.

Positive disproportionality signals for several peripheral-neuropathy manifestations linked to ciprofloxacin, levofloxacin, moxifloxacin, and ofloxacin, with additional signals for gemifloxacin — plus demographic and outcome-pattern differences across agents.

Broadens the peripheral-neuropathy signal beyond any single fluoroquinolone, indicating a class-wide reporting pattern rather than an isolated-drug issue.

A systematic review identified 12 included studies, comprising 4 clinical trials and publications describing 11 individual case reports/cases linking fluoroquinolone use to panic attacks. A companion active-comparator FAERS analysis found reporting rates roughly sixfold higher than azithromycin and twelvefold higher than trimethoprim-sulfamethoxazole. The authors explicitly caution that spontaneous-report data cannot establish causality.

Extends the neuropsychiatric signal into an underrecognized symptom domain — acute panic and anxiety — with the authors modeling exactly the careful causality language this page aims to use.

Analysis of FAERS reports identified disproportionate reporting signals involving psychiatric outcomes such as anxiety, panic, delirium, hallucinations, insomnia, and suicidal ideation. The analysis also identified notable ocular signals, particularly with moxifloxacin, including iris-transillumination and pigment-dispersion abnormalities.

Broadens the documented postmarketing safety pattern beyond tendons and peripheral nerves, identifying psychiatric and ocular event categories that may otherwise be separated across specialties or overlooked as part of a drug-related pattern.

A patient with cystic fibrosis developed recurrent musculoskeletal symptoms associated with inhaled levofloxacin treatment cycles. Symptoms followed the treatment pattern and improved after discontinuation (dechallenge).

Demonstrates that the evidence library includes different routes of fluoroquinolone administration — the temporal association and dechallenge pattern support clinical plausibility, but this does not imply that inhaled and oral routes carry identical systemic risk, and it is not proof of FQAD or mitochondrial injury.

Two weeks of low-dose ciprofloxacin exposure in human cortical organoids increased reactive oxygen species, reduced mitochondrial membrane potential, and disrupted cortical development, GABAergic network formation, and neuronal firing patterns. Ciprofloxacin also significantly reduced expression of the transcription factor FOXG1, and molecular docking simulations suggested a potential CPFX–FOXG1 interaction — the authors propose FOXG1 disruption as a potential contributor to ciprofloxacin neurotoxicity, not a confirmed direct-binding target.

Direct evidence in a human cortical-organoid model — rather than an animal model — linking ciprofloxacin exposure to mitochondrial disturbance, disrupted neurodevelopment, and altered excitatory/inhibitory network formation.

In mouse tendon cells, fluoroquinolone exposure reduced type I collagen production and assembly, hydroxyproline and fibronectin levels, lysyl oxidase activity, collagen-fibril stiffness and diameter, and active β1-integrin expression. The authors propose disrupted cell-matrix adhesion as part of the tendon-injury mechanism.

Establishes tendon injury as a parallel fluoroquinolone-toxicity pathway operating through extracellular-matrix and cell-adhesion disruption. This is not forced into the mitochondrial mechanism — it is presented as a distinct, potentially interacting mechanism.

In C2C12 and primary human skeletal-muscle myotubes, statin exposure downregulated TOMM40 and TOMM22 — components of the mitochondrial protein-import machinery — and impaired mitochondrial function, dynamics, and mitophagy, with increased superoxide production. Restoring TOMM40/TOMM22 expression partially rescued these effects.

Extends the DIMD framework's cross-drug pattern beyond fluoroquinolones: statins reach mitochondrial biology through protein import and quality control, not just the classical CoQ10-depletion mechanism. The rescue experiment (restoring TOMM40/TOMM22) supports a direct, not merely correlative, role for this pathway.

In rat astrocytes, statin-induced coenzyme Q depletion reduced ATP-linked respiration and ATP levels and increased proton leak, while driving mitochondrial fission, altered biogenesis, and broader metabolic reprogramming — an active adaptive remodeling response, not simply energetic failure.

A second, independent statin/CoQ model in a different cell type (astrocytes rather than myotubes) showing that mitochondrial responses to drug-induced CoQ loss can include active remodeling — reinforcing that "mitochondrial injury" is not a single uniform phenotype across tissues.

Using high-resolution respirometry in human HEK293T cells, gentamicin and ciprofloxacin markedly increased mitochondrial leak respiration, while amoxicillin showed no significant effect in the same experimental system.

Direct, same-system evidence that mitochondrial effects are drug-specific rather than a blanket property of an entire antibiotic class — amoxicillin's null result is the control that makes the ciprofloxacin and gentamicin findings meaningful.

Using a mouse model carrying pre-existing mitochondrial DNA polymerase-γ (Polg) dysfunction, together with human myocarditis biopsy findings, the investigators found that underlying mitochondrial vulnerability increased susceptibility to cardiac injury following mRNA vaccination. The experimental work implicated mitochondrial oxidative stress and ROS-associated necroptosis as contributing mechanisms.

This study provides experimental evidence for an important broader principle: mitochondrial dysfunction may remain compensated or clinically silent until an additional biological stressor exposes that vulnerability — a second-hit dynamic directly relevant to the DIMD framework's core logic, even though the exposure studied here is mRNA vaccination, not a fluoroquinolone.

In patient-derived fibroblasts and cybrids carrying the m.3243A>G mtDNA mutation, impaired mitophagy was associated with persistence of mutant mtDNA; pharmacologically restoring mitophagy capacity via PI3K-AKT-MTORC1 axis inhibition progressively reduced the mutant mtDNA burden and improved mitochondrial function.

This is not a drug-induced mitochondrial injury study and is not evidence that any drug caused mitochondrial disease. Its relevance is purely mechanistic: it demonstrates that mitochondrial quality-control activity can itself shift mitochondrial population dynamics (heteroplasmy) over time — a biological principle the DIMD framework's persistence/quality-control reasoning draws on, studied here in an inherited-disease model unrelated to any drug exposure.

Following a request involving patient representation, France's Haute Autorité de Santé formally evaluated the need for clinical guidance concerning serious fluoroquinolone adverse effects. HAS documented that available scientific evidence was insufficient to support formal treatment recommendations, and proceeded toward professional information and guidance focused on recognition, reporting, patient information, and management using the available evidence.

A national health authority has formally documented the clinical-management evidence gap — movement beyond warnings and prescribing restrictions toward the unresolved question of how affected patients should be recognized, evaluated, and managed.