Beyond endpoint lists in rat elevated plus maze pharmacology: Reporting instability and domain-structured...

Rat elevated plus maze (EPM) pharmacology is often interpreted through endpoint lists, yet the construct assignment of individual variables is rarely tested against pharmacological calibration and reporting practice. This creates a contemporary reproducibility problem: secondary endpoints may be treated as anxiety indices without evidence that they belong to the same domain as proportional open-arm measures. We address this problem with a targeted critical framework synthesis combining a purposive calibration corpus (Corpus A, 14 foundational rat EPM studies, 1985-1998) and a structured sampled reporting audit (Corpus B, 57 post-2000 rat EPM pharmacology papers). The synthesis is explicitly not a systematic review and not a pooled meta-analysis; no figure digitization was used. The audit indicates a median of four EPM parameters reported per study and that 88% of studies present EPM data exclusively in figures without tabulated mean ± SEM values, which is consistent with the no-digitization rule rather than a justification for it. Primary proportional open-arm measures show complete terminological stability in the sampled corpus (Terminological Instability Index, TII = 0.00), whereas closed-arm entries, rearing, head dipping, and stretched-attend posture/risk assessment show high instability (TII ≥ 0.43). Pharmacological calibration across benzodiazepine, yohimbine, and FG-7142/PTZ anchor classes is consistent with a proportional open-arm core, a motor-control and mixed-variable layer, and a context-dependent ethological periphery. We propose that the rat acute EPM should be interpreted not as a retrospective catalogue of anxiety-like readouts, but as a domain-structured behavioral pharmacology assay in which primary claims should be anchored in the proportional open-arm core, checked against motor/mixed variables, and extended to ethological measures only under explicit calibration and design conditions.