Stress and substance use disordersrisk, relapse, and treatment outcomes

For most of the twentieth century, addiction was treated as a failure of willpower — a moral problem given a clinical coat of paint. That framing made treatment almost impossible to improve, because it pointed at the wrong lever entirely. Rajita Sinha's sweeping review changes the frame: two decades of converging evidence show that the real driver of addiction risk, relapse, and treatment failure is a hijacked stress system. Once you see that mechanism clearly, the path to better outcomes comes into focus. Start with what a healthy stress response actually looks like, because the whole story depends on understanding what goes wrong. Sinha organizes this around three phases. First, a baseline state — the brain and body at relative equilibrium. Second, an acute reaction phase, in which the organism mobilizes resources to meet a threat: heart rate and blood pressure spike, the noradrenergic system fires for alertness and vigilance, and the hypothalamic-pituitary-adrenal axis, or HPA axis, kicks in. Corticotropin-releasing factor, or CRF, is released first, triggering a cascade that eventually pushes cortisol into the bloodstream. That cortisol signal is deliberately delayed relative to the fast autonomic response, and the delay matters: it allows prefrontal circuits to engage during the recovery phase, pulling the system back toward equilibrium. Third, that recovery involves prefrontal regulation bringing the HPA and autonomic systems back to baseline. The whole thing is an inverted U: moderate stress sharpens performance, while overwhelming or prolonged stress degrades it. The key word is recovery. Without it, the system does not return to baseline. It recalibrates. That recalibration is where addiction risk begins. Chronic adversity, trauma, and adverse childhood experiences, or ACEs, activate the stress response repeatedly without adequate recovery, and the system shifts toward a persistent distress state. The CDC-Kaiser ACE Study and the longitudinal work it spawned link greater numbers of adverse childhood experiences directly to earlier substance use initiation and higher rates of alcohol use disorder and other substance use disorders. At the peripheral level, chronic early-life stress produces durable changes in HPA and autonomic function: flattened diurnal cortisol rhythms, blunted cortisol and cardiovascular reactivity to laboratory stressors, and altered heart rate variability. At the neural level, structural and functional magnetic resonance imaging studies document reductions in gray matter volume across the orbitofrontal cortex, ventromedial and dorsolateral prefrontal cortex, anterior cingulate cortex, insula, hippocampus, and amygdala in people exposed to psychosocial adversity. Functional imaging adds a complementary picture: reduced activation in prefrontal regions alongside heightened reactivity in limbic-striatal circuits during stress and drug-cue exposure. The net result is reduced top-down control over threat and reward circuitry — exactly the neural configuration that impairs flexible coping and amplifies the drive toward immediate relief. Then drugs enter the picture, and the disruption deepens. Sinha frames the stress-drug cycle as a feed-forward loop. Acute drug use initially blunts the stress response — that is part of the relief it provides. But chronic use and repeated withdrawal progressively remodel peripheral and central stress systems so that each cycle of use makes the next one more likely. During withdrawal and abstinence from nicotine, alcohol, opiates, cocaine, and cannabis, multiple studies show blunted ACTH and cortisol responses to laboratory stressors and to CRF administration, while basal HPA markers and autonomic arousal are elevated. The system becomes tonically high but phasically flat: the resting set point is raised, but the capacity for a sharp, reactive response — and crucially, a recovery — is degraded. Disrupted heart rate variability, elevated inflammatory cytokines, reduced resting heart rate variability, and altered reactive heart rate variability all emerge in this withdrawal profile and are linked to greater craving and higher relapse risk. The brain circuits follow the same trajectory. Neuroimaging in people with substance use disorders shows hyperactivity in limbic-striatal regions during stress and cue exposure, alongside disrupted function in the ventromedial prefrontal cortex, dorsolateral prefrontal cortex, insula, and anterior cingulate cortex. Structural magnetic resonance imaging finds greater atrophy in the orbitofrontal cortex, right medial prefrontal cortex, and anterior cingulate in individuals who relapse compared to those who maintain abstinence. A key mechanistic node tying this together is GABAergic regulation — GABA interneurons normally constrain excitatory stress signals across hypothalamic, amygdalar, striatal, and prefrontal circuits. Chronic drug exposure and high-intensity stress downregulate these GABA circuits, removing the inhibitory brake on stress arousal. Each round of use and withdrawal tightens the loop: worse stress biology drives more use, and more use worsens stress biology. The clinical consequence of that loop shows up most vividly in the relapse data. Stress is not a marginal trigger — it sits at the center of why patients fail treatment. Laboratory stress provocation reliably produces high levels of drug craving and enhanced negative mood and anxiety in people with substance use disorders. Peripheral biomarkers nail down the connection: Junghanns and colleagues showed that an impaired serum cortisol stress response predicted early relapse in alcohol-dependent patients. Back and colleagues found that reactivity to laboratory stress provocation predicted relapse to cocaine. Fox and colleagues documented that recent frequency of cocaine and alcohol use directly affects drug craving and responses to stress and drug-related cues. Inflammatory markers join the picture too — interleukin six, interleukin ten, and tumor necrosis factor alpha are among the cytokines Sinha lists as predicting time to relapse and frequency of use. The pattern across studies is consistent: high basal tone, blunted phasic reactivity, disrupted heart rate variability, and elevated subjective craving form a multilevel stress profile that maps reliably onto increased relapse risk. An intervention that does not engage stress biology leaves that profile intact — and the patient structurally exposed. That is the opening for the treatment implications, and Sinha surveys a concrete pharmacological toolkit for exploiting it. Three classes of agents stand out. First, HPA-axis modulators: mifepristone has been shown to normalize peripheral HPA responses and decrease alcohol intake in alcohol-dependent individuals. Second, noradrenergic agents that blunt the sympathetic stress signal: the alpha-one antagonists prazosin and doxazosin reduced stress-induced alcohol craving and negative emotions, and doxazosin improved abstinence in pilot work with cocaine use disorder. Prazosin's benefit was most pronounced in individuals with greater alcohol withdrawal severity — an early precision medicine signal. The alpha-two agonist lofexidine reduced stress-induced opioid craving, and guanfacine improved stress- and cue-related craving and cortisol dynamics, with effects particularly strong in women. Third, neuroactive steroids: micronized progesterone at four hundred milligrams per day for five days reduced cocaine craving and cortisol responses and improved inhibitory performance on the Stroop task, effects attributed to increases in the GABAergic neuroactive steroid allopregnanolone. Pregnenolone, its precursor, reduced stress- and cue-related craving and normalized HPA and autonomic stress measures in people with alcohol use disorder and cocaine use disorder, with initial signals of improved alcohol outcomes. What unites these agents is that their effects are measurable against the same biomarker panel — basal and phasic cortisol and ACTH, autonomic indices like heart rate and blood pressure, inflammatory cytokines, and neuroimaging markers of prefrontal engagement. That panel is also the basis for a precision medicine strategy. Sex and gender, withdrawal severity, trauma history, substance use disorder severity, and comorbidity all moderate which agents work for whom. Naltrexone showed greater effects in men and in those with pretreatment abstinence, but not in alcohol-dependent women. A network meta-analysis in smoking cessation by Smith and colleagues found that varenicline outperformed bupropion in women, while the two were comparable in men. Moderators like these can match patients to interventions — but only if stress biomarkers are measured in the first place. Sinha's review makes the case that addiction is a disorder of stress pathophysiology, not a deficit of character. The multilevel disruptions it describes — peripheral HPA and autonomic dysregulation, degraded GABAergic control, limbic-prefrontal circuit remodeling — are measurable, mechanistically linked to craving and relapse, and increasingly targetable. Behavioral, social, and pharmacologic approaches all have purchase on this system. The evidence argues for cross-level research connecting animal models of HPA and autonomic change to clinical trials testing whether normalizing stress biology translates into better outcomes for patients. For the millions of people who cycle through treatment and relapse, that translation is the whole point. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.