Making hormesis visible in new approach methodologies (NAMs): a framework for botanical mixtures, biological context, and risk interpretation

Risk characterization of mixtures such as botanicals is complicated by their complex and variable composition, where there are multiple possibilities for interactions among components, even at low doses. Most traditional toxicological testing and risk-assessment workflows remain optimized for monotonic toxicity. New approach methodologies (NAMs) are much better able to detect low-concentration and time-dependent stimulation or inhibition in complex mixtures, which can vary across donors, sexes, biological states, and production lots. We propose H-NAM, a hormesis-aware overlay for existing NAM batteries and integrated approaches to testing and assessment. H-NAM uses the commonly reported 30-60% maximum stimulation and the often narrow, sub-decade stimulatory concentration window as empirical design heuristics rather than universal biological rules. It integrates two-stage concentration design, time-resolved measurements, relevant biological-context strata, botanical identity and multi-lot controls, assay-interference testing, comparison of monotonic, biphasic, and flexible response models, and quantitative in vitro-in vivo extrapolation. H-NAM separates four questions: (1) whether a biphasic response pattern is credible; (2) how it varies across concentration, time, and biological context; (3) whether the stimulation is adaptive, an early-warning signal, ambiguous, or adverse; and (4) whether the relevant response region is attainable under plausible internal exposure. Exposure relevance is evaluated independently using time-matched exposure estimates and an exploratory stimulatory bioactivity-exposure ratio (sBER). Joint evaluation of biological consequence and exposure relevance is then used to prioritize follow-up testing, hazard characterization, candidate point-of-departure development, or risk management, while a lot-restricted response prompts compositional and manufacturing investigation. H-NAM is intended to reduce both the dismissal of reproducible low-concentration signals and the unsupported interpretation of stimulation as benefit or evidence of safety. An accompanying reporting checklist is suggested for specifying the minimum information needed for independent evaluation, replication, and regulatory interpretation.