Reclaimed Inefficiency, Not Created Capacity: A Bioenergetic Account of Limited Pharmacological Cognitive Enhancement in Healthy Adults

Non-medical use of prescription stimulants for cognitive enhancement continues to rise, sustained by an expectation of performance gain that the efficacy literature does not support. Meta-analyses in healthy, non-sleep-deprived adults report standardised mean differences of approximately 0.12 for modafinil and 0.21 for methylphenidate, with no detectable effect for d-amphetamine. This review examines whether the bioenergetics of the human brain explains that ceiling. We review three converging bodies of evidence: meta-analytic efficacy stratified by cognitive domain; neuroimaging of cerebral glucose metabolism and network dynamics under stimulant administration; and cardiovascular safety data from network meta-analysis. The pivotal observation is that methylphenidate does not augment cerebral energy supply. In accuracy-controlled conditions it reduces the metabolic cost of task performance, principally by attenuating activation in default-mode regions, and improves performance selectively in individuals who engaged those regions under placebo, with the potential for detriment in those already optimally focused. We argue that the resource these agents mobilise is recovered inefficiency rather than created capacity. On this account the ceiling on enhancement in healthy adults is not pharmacological but bioenergetic, and is therefore not addressable through dose optimisation or novel formulation; meanwhile the sympathomimetic cost is incurred irrespective of whether recoverable inefficiency exists. The resulting benefit–risk asymmetry falls most heavily on the population least likely to benefit. The account is integrated with the Connectivity Maintenance Cost (CMC) framework: under a fixed cerebral energy budget, stimulant effects reflect reallocation of connectivity funding rather than increased supply — making the enhancement ceiling a structural prediction of network bioenergetics. We identify the evidentiary limits of this account, including the age and heterogeneity of the metabolic imaging literature, and specify what would refute it.