A neuroeconomic theory of bidirectional synaptic plasticity and addiction

Neuronal mechanisms underlying addiction have been attracting attention in neurobiology, economics, neuropsychiatry, and neuroeconomics. This paper proposes a possible link between economic theory of addiction (Becker and Murphy, 1988) and neurobiological theory of bidirectional synaptic plasticity (Bienenstock, Cooper, Munro, 1982) based on recent findings in neuroeconomics and neurobiology of addiction. Furthermore, it is suggested that several neurobiological substrates such as cortisol (a stress hormone), NMDA and AMPA receptors/subunits and intracellular calcium in the postsynaptic neurons are critical factors determining parameters in Becker and Murphy’s economic theory of addiction. Future directions in the application of the theory to studies in neuroeconomics and neuropsychiatry of addiction and its relation to stress at the molecular level are discussed.

💡 Research Summary

**
The paper proposes a novel integrative framework that bridges the rational addiction model of Becker and Murphy (1988) with the Bienenstock‑Cooper‑Munro (BCM) theory of bidirectional synaptic plasticity. In the Becker‑Murphy (BM) model, an addict’s decision problem is formalized as the maximization of discounted utility over a horizon, where the state variable “consumption capital” S captures the accumulated physiological impact of past drug use. S decays at a rate δ (the depreciation rate) and interacts with the discount factor r to shape intertemporal choices. While behavioral studies have linked r to delay discounting in smokers, the neurobiological substrates of S and δ have remained speculative.

BCM theory, originally devised to explain visual cortical development, describes synaptic weight changes (Δw) as a function of instantaneous postsynaptic activity y relative to a sliding threshold Mθ. When y < Mθ, long‑term depression (LTD) occurs; when y > Mθ, long‑term potentiation (LTP) is induced. Crucially, Mθ itself is a monotonic increasing function of the time‑averaged postsynaptic activity, and empirical work has tied Mθ to NMDA‑mediated calcium influx, the NMDA/AMPA receptor ratio, and the proportion of NR2B‑containing NMDA receptors. Stress hormones such as cortisol elevate persistent intracellular Ca²⁺, thereby raising Mθ and biasing synapses toward depression. Conversely, chronic exposure to addictive substances (e.g., nicotine) lowers Mθ, facilitating LTP in dopaminergic circuits.

The author postulates a generalized mapping Mθ = f_g(y, S) where ∂f_g/∂y < 0 (higher activity reduces the threshold) and ∂f_g/∂S > 0 (greater consumption capital lowers the threshold). Under this mapping, accumulation of S through repeated drug intake reduces Mθ, promoting LTP and thus providing a synaptic substrate for habit formation. An increase in the depreciation rate δ accelerates the decay of S, causing a rapid rise in Mθ, which would favor LTD and potentially neurodegeneration—offering a mechanistic explanation for the BM prediction that high δ individuals are more vulnerable to addiction.

From this unified perspective, several testable predictions emerge: (1) resting intracellular Ca²⁺ concentration (