The conceptualization of substance addiction is linked to diverse theories, with competing models emphasizing distinct psychological, behavioral, and neurobiological mechanisms [1,2]. As illustrated in Table 1, the Biphasic model of addiction proposed by Aggarwal, Naik, and Lindquist [3] offers an integrative framework that systematically organizes addiction processes across five domains: reinforcement, motivation, associative learning, tolerance, and inhibitory control.

Each domain captures a specific neurobehavioral transition, which collectively map the gradual shift from casual, voluntary psychoactive drug use to compulsive, pathological drug intake [3]. The model offers a unified foundation that aligns and builds on established theories of addiction, including those focused on the underlying neurocircuitry [4], incentive sensitization [5], and habit learning [6].

Importantly, the model hypothesizes the phase transitions to be dynamic and overlapping rather than sequential. This is a notable strength, as empirical evidence suggests that multiple processes—such as craving, habit formation, and impaired inhibitory control—co-occur across the addiction cycle, from drug use to abuse to addiction [7]. By avoiding rigid staging, the model better reflects the heterogeneity and fluidity observed in clinical populations [8].

The Biphasic model effectively links shifts in cognition, emotion, and behavior to experience-dependent alterations in the relevant brain regions and circuits. Early-stage substance use is characterized by the activation of mesocortical and mesolimbic pathways, particularly dopaminergic projections from the ventral tegmental area to the nucleus accumbens [9]. Over time, addictive behavior takes hold as progressive neuroadaptations diminish mesolimbic reward sensitivity, weaken mesocortical executive regulation, and shift behavioral control toward fronto-striatal habit circuitry [10].

With respect to the first domain transition, occasional recreational drug use is driven by positive reinforcement associated with elevated dopamine release in the mesolimbic reward circuit [11]. Continued use diminishes reward sensitivity, reduces responsiveness to natural reinforcers, and contributes to a shift toward negative reinforcement [12]. Substance use becomes increasingly driven by the desire to avoid or alleviate the withdrawal symptoms, dysphoria, or stress that emerge in the absence of the drug [13].

After its first use, an individual likely continues using a specific drug because its immediate effects are experienced as rewarding or pleasurable. This hedonic "liking" is mediated primarily by opioid and endocannabinoid activity, while concurrent dopaminergic signaling reinforces repeated drug use [14]. Chronic drug exposure sensitizes neural systems underlying motivation and incentive salience, increasing the urge to seek out and utilize the drug [15]. Consequently, the subjective pleasure derived from the drug compound decreases over time, with a corresponding increase in wanting or craving [16].

The development of drug-seeking behavior involves two forms of associative learning [17]. In the early stages, classical conditioning—via the amygdala and hippocampus—links environmental cues with the drug’s psychological and physiological outcomes [18]. Subsequent exposure, in the absence of the drug, can induce a sense of expectation, incentivizing the individual to pursue further drug use. Commensurately, the neural loci of behavioral control shifts from the ventral striatum, containing the nucleus accumbens, to the dorsal striatum—i.e., from reward-driven behavior to habit-like responding [19]. Instrumental conditioning gradually automates drug-seeking behaviors, which become increasingly habitual and less dependent on conscious decision-making.

As drug use becomes more frequent, progressively larger quantities are required to achieve the same high or level of intoxication. Physiological tolerance reflects homeostatic neuroadaptations, including diminished dopamine release and postsynaptic responsiveness within reward circuitry [20], blunting the drug’s pleasurable effects. In addition, conditioned tolerance occurs when the environmental cues associated with drug use evoke anticipatory compensatory responses that counteract the drug’s physiological effects [21]. These learned adaptations contribute to escalating drug intake and may help explain the increased risk of overdose when drugs are taken in a novel location—i.e., in the absence of the conditioned compensatory responses that offset the forthcoming drug [22].

Lastly, the Biphasic model describes a transition in inhibitory control. As drug use transitions toward abuse, the development of hypofrontality impairs executive control and increases vulnerability to impulsive behavior [23]. Immediate gratification increasingly drives behavior despite clear and mounting adverse consequences. Chronic exposure further disrupts executive-control networks dependent on mesocortical and fronto-striatal pathways [24]. Diminished top-down neural control, combined with heightened responsiveness to drug-related cues, fosters compulsive behavior tied to drug seeking and taking [25]—a defining feature of advanced addiction.

Clinically, the Biphasic model has clear implications for phase-specific interventions. It closely aligns, for instance, with precision psychiatry approaches that seek to match treatment strategies to the neurobiological and behavioral mechanisms responsible for addiction [26].

To that end, treatment for individuals in the earlier stages of problematic drug use may benefit from targeting positive reinforcement processes through contingency management or behavioral activation approaches [27]. Late-stage addictive behavior, alternatively, may require strategies that address negative reinforcement, withdrawal, craving, or compulsive responding [28], including pharmacotherapy and relapse prevention approaches.

Associative learning processes are also a critical target for intervention. Cue-exposure therapies, extinction-based approaches, and mindfulness interventions may help weaken conditioned or instrumental (habitual) responding and reduce relapse vulnerability [29]. Similarly, conditioned tolerance underscores the importance of contextual factors in relapse prevention and overdose education.

Finally, deficits in inhibitory control suggest the importance of interventions aimed at strengthening prefrontal-mediated executive functioning. Cognitive remediation, emotion regulation training, problem-solving interventions, and emerging neuromodulation techniques may enhance cortical function and connectivity, thereby improving top-down regulatory control over behavior [30].

Despite the Biphasic model’s conceptual strengths, several limitations warrant consideration. First, it is primarily integrative rather than mechanistically novel. Most of the proposed phases are well-established in the literature, with the model primarily providing conceptual synthesis. Second, empirical validation remains limited. Experimental and longitudinal studies are needed to determine whether the proposed phase transitions can be reliably identified and whether they predict treatment outcomes or relapse trajectories.

Third, the model does not account for individual differences, such as genetic vulnerability, early-life stress, trauma exposure, or comorbid psychiatric conditions [31]. Incorporation of these moderators will enhance the model’s explanatory power and clinical relevance. Fourth, owing to similarities in the behavioral and neural mechanisms underlying substance and non-substance addictions [32], Aggarwal et al. [3] proposed that the model may also apply to the latter. This extension remains speculative and requires empirical validation.

The Biphasic model represents a significant step toward conceptual integration in the addiction sciences. Its emphasis on dynamic phase transitions, underlying neural circuitry, and behavioral expression offers both conceptual clarity and translational utility. By organizing complex neurobehavioral adaptations into a coherent, phase-specific framework, it provides a useful perspective through which to view the transition from early, voluntary drug use to uncontrollable, compulsive addictive behavior. Following empirical validation, the model has the potential to inform personalized approaches to addiction assessment, treatment, and prevention.