Khat’s Hidden Cost: Rodent Studies Reveal Dose-Dependent Damage to Spatial Memory

Khat, the leafy green shrub chewed daily by an estimated 20 million people across East Africa and the Arabian Peninsula, has long carried a dual reputation: a culturally cherished social stimulant and a controlled amphetamine-like drug. Now, a comprehensive synthesis of preclinical research suggests the plant may exact a quieter, more insidious toll on the brain, one that accumulates silently with every session of chewing. A narrative review published in Discover Toxicology by Khalid Bekri and Alfoalem Araba Abiye of Addis Ababa University has systematically pulled together every controlled rodent study examining whether Catha edulis, the botanical source of khat, compromises the brain circuits that allow animals to learn and remember spatial environments. The verdict is sobering: across fourteen studies encompassing 610 rodents, higher doses and longer exposures consistently eroded spatial learning and memory, with chronic exposure producing the most severe and persistent deficits.

The stakes of this question extend far beyond the laboratory. Spatial cognition, the capacity to acquire, store, and retrieve information about environmental geometry and object location, depends on a distributed brain network anchored by the hippocampal formation, with entorhinal grid cells and hippocampal place cells jointly constructing the cognitive maps that guide navigation. At the cellular level, the formation and consolidation of spatial memories rely on long-term potentiation at Schaffer collateral synapses connecting the CA3 and CA1 fields of the hippocampus, a process governed by NMDA and AMPA receptor-mediated glutamatergic transmission and finely tuned by dopaminergic, noradrenergic, and serotonergic inputs. Cathinone, khat’s principal psychoactive alkaloid, is a beta-keto phenethylamine structurally related to amphetamine that promotes the release of dopamine and norepinephrine and blocks their reuptake. Because dopamine and norepinephrine dynamically regulate hippocampal synaptic plasticity, cathinone’s pharmacological profile provides direct mechanistic grounds for suspecting that chronic khat exposure could disrupt the very synaptic machinery that spatial memory requires.

Human studies have long hinted at trouble. Chronic khat users perform worse than khat-free controls on working memory and cognitive flexibility tasks, on verbal learning and delayed recall, and across broader domains including attention, motor speed, decision-making, and inhibitory control, and a meta-analysis has linked khat use to measurable memory impairment. These deficits carry real functional consequences: compromised occupational performance, impaired driving safety, and reduced academic achievement among the university students and young adults among whom use is common. Yet human data are notoriously difficult to interpret. Polysubstance use, small and non-representative samples, and uncontrolled dosing histories all muddy the causal picture, leaving open the question of whether khat itself, rather than lifestyle factors correlated with its use, drives the cognitive decline. Controlled rodent experiments, which permit precise manipulation of dose, duration, and exposure route while excluding such confounds, are therefore essential for establishing a reliable dose-response relationship, and it is precisely this synthesis that the new review delivers.

To build their evidence base, the authors conducted a structured literature search across PubMed, ResearchGate, Elsevier, EMBASE, and Web of Science, screening records according to PRISMA 2020 methodology with a search end date of May 25, 2025. From an initial pool of 504 records, the investigators winnowed the field to fourteen primary studies meeting strict inclusion criteria: use of Catha edulis extract, juice, or purified cathinone; an in-vivo rodent model; validated behavioral assessment of spatial cognition; and quantifiable outcomes. The behavioral paradigms represented were dominated by the Morris Water Maze, employed in nine studies, which measures both spatial learning through escape latency and path length during acquisition trials and spatial memory retention through time spent in the target quadrant during probe trials. Three studies used the Radial Arm Maze, which tracks spatial working and reference memory through arm-entry patterns, while the T-maze, Y-maze, and Novel Object Recognition task each appeared in smaller numbers. Study quality was assessed with the SYRCLE Risk of Bias tool, validated specifically for animal intervention studies.

The pattern that emerged across the fourteen studies is strikingly duration-dependent. Acute exposure, defined as a single dose ranging from 40 to 500 milligrams per kilogram, produced bidirectional effects that initially appeared to follow an inverted-U curve. At the low end, doses of 40 to 100 milligrams per kilogram modestly improved exploratory behavior and reduced first-trial water maze latency, but the review’s authors interpret these gains as arousal-mediated motivation rather than genuine memory enhancement. At 150 to 500 milligrams per kilogram, every study reported consistent deficits across maze paradigms, including increased path length and latency and reduced working memory scores, with no graded response within this range. The reviewers argue this pattern reflects a threshold effect rather than a true inverted-U relationship: once dopaminergic stimulation exceeds a critical level, it disrupts the NMDA receptor-dependent processes required for spatial encoding. Finer dose stratification within the low-to-moderate range will be needed to settle the question formally.

Repeated exposure told a darker story. Subacute dosing, at 40 to 360 milligrams per kilogram over five to seventeen days, was consistently associated with spatial learning deficits, though with task-specific nuances: at 40 milligrams per kilogram over ten days, CBA mice showed disrupted reversal learning and impaired reference memory in the water maze, while the same dose over a shorter five-day protocol spared learning but impaired memory retention. Subchronic exposure of 29 to 90 days produced the most robust impairment profiles across multiple paradigms, with 100 milligrams per kilogram driving increased thigmotaxis, reduced spontaneous alternation, and elevated error rates in later training phases, though one study at the same dose reported no significant deficits, a discrepancy the authors attribute to methodological heterogeneity. Chronic exposure beyond 90 days, examined in a single study of 36 rats given 100, 200, or 300 milligrams per kilogram for twelve weeks, caused persistent spatial learning and memory impairment: at the highest dose, rats took significantly longer and swam significantly farther to reach the hidden platform, and spent significantly less time in the target quadrant during the probe trial. Critically, no study in the entire evidence base incorporated a washout period, so whether these deficits reverse after cessation remains entirely unknown.

Beneath the behavioral findings lie several converging mechanistic pathways that the review outlines with technical precision. As an indirect monoaminergic agonist, cathinone elevates extracellular dopamine, norepinephrine, and serotonin, and supranormal dopaminergic signaling can disrupt NMDA receptor-dependent long-term potentiation at hippocampal CA3-CA1 synapses, the cellular substrate of spatial learning. Chronic exposure has additionally been linked to reduced brain-derived neurotrophic factor signaling through its tropomyosin receptor kinase B, impaired neurogenesis, and altered dendritic spine morphology in the hippocampus, changes that would structurally erode the synaptic architecture on which memory consolidation depends. Oxidative stress and microglial activation may compound the damage by releasing pro-inflammatory cytokines such as interleukin-1 beta and tumor necrosis factor alpha, further impairing glutamatergic transmission. The authors caution, however, that direct mechanistic evidence from khat-exposed animals remains scarce, and they call for future studies combining behavioral testing with electrophysiology and molecular assays of BDNF/TrkB signaling, phosphorylated CREB, neuroinflammation, and oxidative stress markers.

The review is equally candid about the weaknesses of the evidence it synthesizes. The risk-of-bias assessment revealed that 78.6 percent of studies were at high risk of performance and detection bias because outcome assessment was not blinded, a serious concern in behavioral research where experimenter expectations can subtly inflate treatment effects. Allocation concealment was unreported in all fourteen studies, and random housing was described in only two. The evidence base is also narrow: 78.6 percent of studies used Swiss albino mice, an outbred stock with high inter-individual genetic variability, and nearly two-thirds originated from Ethiopia, with locally sourced khat of variable cathinone content that may range from 0.1 to 1.2 percent of dry weight and degrades rapidly after harvest into the weaker alkaloids cathine and norephedrine. Without standardized extract preparation and chemical fingerprinting, the nominal dose-response relationships reported across studies may not be directly comparable. All studies also relied on forced oral dosing rather than voluntary consumption models that better mirror human chewing behavior.

Translation to human exposure adds further nuance. Using body surface area scaling, the reviewers estimated that the rodent doses of 40 to 500 milligrams per kilogram correspond to human equivalent doses of roughly 6.5 to 81 milligrams per kilogram, a range whose upper end substantially exceeds realistic exposure from habitual khat chewing, estimated at approximately 7 to 28 milligrams of cathinone equivalent per session in a 70-kilogram adult. This suggests the most severe impairment profiles may reflect supraphysiological exposure scenarios, while the lower-dose studies more faithfully model everyday use. The authors argue that the cumulative message nonetheless justifies practical action: incorporating khat-use history into clinical cognitive screening, targeted public health education for youth, and epidemiological surveillance in endemic regions. Their research agenda calls for standardized cathinone-equivalent dosing, washout and recovery cohorts to test reversibility, multi-strain and both-sex comparisons, voluntary self-administration models, and longitudinal human studies integrating neuropsychological assessment with neuroimaging and biomarkers. For a plant woven into the social and economic fabric of an entire region, the emerging picture is one of meaningful cognitive risk that science is only beginning to map.

Subject of Research: Preclinical effects of the stimulant plant Catha edulis on spatial learning and memory in rodent models

Article Title: A narrative review of preclinical evidence on the effects of Catha edulis on spatial cognition and memory in rodents

Article References: Bekri, K., & Abiye, A. A. (2026). A narrative review of preclinical evidence on the effects of Catha edulis on spatial cognition and memory in rodents. Discover Toxicology, 3(1), Article 21. https://doi.org/10.1007/s44339-026-00066-w

Image Credits: AI Generated

DOI: 10.1007/s44339-026-00066-w

Keywords: Catha edulis, khat, cathinone, spatial memory, spatial cognition, Morris water maze, hippocampus, long-term potentiation, neurotoxicology, BDNF, rodent models, narrative review