Why We Run: The Neuroscience of Running Addiction

The concept of the runner's high entered the popular lexicon in the 1970s, coinciding with the first American running boom. Journalists and early exercise researchers attributed the euphoric state — the sudden lift in mood, the sense of effortlessness, the mild time distortion that experienced runners described after long efforts — to endorphins. Beta-endorphin and enkephalin, natural opioid peptides produced by the pituitary gland and hypothalamus during physical stress, seemed the obvious candidates. They were chemically similar to morphine, they were produced in response to exercise, and they were known to reduce pain. The endorphin hypothesis was elegant, intuitive, and almost immediately accepted as fact. Within a decade it had been repeated so often in popular media that questioning it felt like denying the obvious.

The problem was always a biochemical one: endorphins are large peptide molecules. Beta-endorphin, with 31 amino acids and a molecular weight of approximately 3,500 daltons, cannot readily cross the blood-brain barrier under normal physiological conditions. The blood-brain barrier is a selective permeability mechanism that excludes large peptide molecules from the central nervous system. If endorphins produced during exercise remain in the peripheral bloodstream and cannot reach the opioid receptors in the brain's limbic system, they cannot directly cause the subjective feeling of euphoria that runners describe. Researchers raised this objection repeatedly through the 1980s and 1990s, but the endorphin explanation had become so embedded in public understanding that it persisted regardless.

In 2008, Boecker and colleagues published what appeared to be the definitive evidence for the endorphin hypothesis. Using PET scanning, they imaged the brains of 10 trained athletes before and after a 2-hour run and found significantly elevated opioid receptor binding in the prefrontal and limbic regions — areas associated with mood and emotional processing. Self-reported euphoria correlated with the degree of opioid receptor activation. The study was widely interpreted as proof: endorphins cause the runner's high. However, a closer reading reveals an important nuance. The study showed that opioid receptors were activated during running euphoria — but it could not distinguish whether this activation came from endorphins crossing the blood-brain barrier or from other endogenous opioid-like molecules with different pharmacokinetics. The mechanism was still unsettled.

The real story proved to be more nuanced and more interesting than the endorphin hypothesis ever was. The subjective experience runners call the high is not produced by a single neurochemical system but by a complex interaction of multiple systems — opioid, endocannabinoid, serotonergic, and dopaminergic — each contributing different qualities to the experience. Endorphins appear to contribute primarily to peripheral analgesia: they reduce pain signals at the spinal cord level, explaining why runners can push through discomfort they would otherwise find unbearable. The euphoric, anxiolytic, and mood-elevating components of the runner's high turned out to have a different, more direct mechanism — and it required a fundamentally different class of molecule to explain it.

The definitive shift in understanding came in 2015 with a landmark paper by Fuss and colleagues published in the Proceedings of the National Academy of Sciences. The researchers used genetically modified mice with selectively blocked receptor systems to disentangle which neurochemical pathway was responsible for exercise-induced euphoria. When they blocked the opioid receptors (the endorphin pathway), the mice still showed all the behavioral signs of exercise-induced euphoria after running: reduced anxiety in open-field tests, elevated pain thresholds, and increased exploratory behavior. When they blocked the cannabinoid receptors — specifically the CB1 receptors targeted by endocannabinoids — the euphoric behavior disappeared entirely. The mice ran the same amount but showed none of the mood elevation or anxiolysis. The finding was unambiguous: the runner's high requires intact cannabinoid signaling, not opioid signaling.

The key molecule is anandamide, formally named N-arachidonoylethanolamine. Its common name derives from the Sanskrit word 'ananda,' meaning bliss — a remarkably apt description. Anandamide is an endocannabinoid: a lipid-soluble signaling molecule produced naturally by the body that acts on the same cannabinoid receptors (CB1 and CB2) activated by THC from cannabis. Critically, anandamide is lipid-soluble, which means it can cross the blood-brain barrier — unlike large peptide endorphins. When anandamide reaches the CB1 receptors in the limbic system (amygdala, nucleus accumbens, hippocampus, and prefrontal cortex), it produces anxiolysis (reduced anxiety), mild euphoria, pain relief, and a sense of calm well-being. These are precisely the qualities runners describe during a runner's high. A second endocannabinoid, 2-arachidonoylglycerol (2-AG), also activates CB1 receptors and contributes to the exercise response, though anandamide appears to be the primary driver.

Raichlen and colleagues published a striking evolutionary study in 2012 that provided convergent evidence for the endocannabinoid-exercise link. They measured endocannabinoid levels in three species — humans, dogs, and ferrets — before and after exercise. Humans and dogs, both of whom evolved as persistence hunters (capable of long-distance sustained running to pursue prey), showed significant increases in circulating anandamide levels after moderate running exercise. Ferrets, who are not persistence hunters and did not evolve for sustained running, showed no meaningful endocannabinoid response to the same exercise protocol. The study suggested that the endocannabinoid reward system is not simply a generic exercise response but a specific adaptation to the type of locomotion — sustained moderate-intensity running — that was critical for survival in persistence-hunting species. The pleasure of running, from this perspective, evolved specifically to reward the behavior of long-distance running.

The dose-response relationship between running intensity and endocannabinoid release is practically important for runners. Research by Sparling and colleagues demonstrated that endocannabinoid levels increase most substantially during moderate-intensity exercise — approximately 60-80% of maximum heart rate, corresponding to Zone 2-3 in a standard five-zone model. At very low intensities (gentle walking, below 55% max HR), the endocannabinoid response is minimal. At very high intensities (above 85% max HR, anaerobic effort), the endocannabinoid response also diminishes as other physiological stress responses dominate. Duration also matters substantially: the endocannabinoid effect typically manifests after 20-30 minutes of sustained moderate exercise and peaks between 30 and 60 minutes. This explains a common frustration among beginner runners: 15-minute runs at an effort that is uncomfortably hard do not produce the runner's high. The neurochemistry requires sustained moderate effort, not brief intense exertion.

Dopamine is widely mischaracterized as the 'pleasure chemical.' Kent Berridge's decades of research at the University of Michigan, using sophisticated lesioning and pharmacological techniques in rats, established a crucial distinction: dopamine is not about the pleasure of experiencing a reward. It is about the anticipation of reward, the motivational drive to pursue a goal, and the learning signal that encodes 'this action led to something good — do it again.' Berridge separated what he called 'wanting' (dopaminergic, motivational) from 'liking' (opioid and endocannabinoid, hedonic) — two systems that normally work together but are neurochemically distinct. Running powerfully engages both, but understanding dopamine's role helps explain why running becomes compulsive in a way that mere pleasure does not.

When a runner completes a long run, achieves a personal record, or finishes a challenging training session, the dopaminergic reward circuits — particularly the mesolimbic pathway from the ventral tegmental area to the nucleus accumbens — fire robustly. This dopamine release encodes the completed action as rewarding and drives the formation of a habit trace. Over time, through repetition, the brain begins to release dopamine not just at the completion of a run but at the anticipatory cues that precede it: the sight of running shoes by the door, the morning alarm, the specific route. This is the classical Pavlovian conditioning of dopamine systems — cue-triggered wanting — and it is the neurological substrate of what runners describe as 'needing' their morning run. The run itself becomes almost secondary; the wanting loop is powerful enough to drive behavior even when the immediate prospect of running is not appealing.

The same dopamine circuitry implicated in running habit formation — the striatum, nucleus accumbens, and prefrontal cortex — is the circuitry involved in all forms of behavioral and substance addiction. This is not coincidence; it is a shared neural architecture. The brain's reward system evolved to motivate survival behaviors (eating, mating, social bonding, exploration of rewarding environments) and does not distinguish between adaptive and maladaptive compulsions at the circuit level. Running activates this system through a genuinely healthy mechanism, but the structural similarity to addiction pathways explains why runners speak of 'needing' a run in ways that sound more like addiction than recreation. When runners describe feeling 'off' or irritable after missing training, they are accurately describing a dopamine-system state: the expected reward signal did not arrive, and the brain responds with a motivational deficit that manifests as mood disruption.

Lichtenstein and colleagues, in a 2017 review, examined the distinction between healthy running passion and exercise addiction. Exercise addiction — defined by running despite injury, running at the expense of work, family, or social obligations, experiencing withdrawal distress, and being unable to cut back despite wanting to — affects an estimated 3-5% of regular runners. It correlates strongly with anxiety disorders, perfectionism, and high neuroticism scores. Importantly, exercise addiction is distinct from the healthy compulsion most committed runners feel. The critical differentiation is whether running serves life goals or undermines them — whether the runner experiences the compulsion as autonomous and positive or as ego-dystonic and distressing. Dopamine system dysregulation, particularly in high sensation-seeking individuals, may predispose some runners to cross from healthy dedication into addictive patterns. Awareness of this distinction allows coaches and athletes to distinguish drive from disorder.

In 1999, van Praag and colleagues published a study in Nature Neuroscience that overturned a long-held dogma of neuroscience: the belief that adult brains cannot grow new neurons. The researchers placed running wheels in the cages of adult mice and compared their brains after several weeks to sedentary controls. The running mice showed more than twice as much neurogenesis — the growth of new neurons — in the hippocampus as the sedentary mice. This finding, replicated many times since in both animals and humans, established that voluntary aerobic exercise is one of the most powerful known stimulants of adult hippocampal neurogenesis. The hippocampus is central to spatial memory, episodic memory, and emotional regulation — and it is one of the brain regions most severely affected by depression, chronic stress, and aging. The implication was profound: running does not just make you feel better, it physically grows your brain.

In humans, Erickson and colleagues published a landmark 2011 study in PNAS that made international news. In a randomized controlled trial, older adults (ages 55-80) who engaged in aerobic exercise for one year showed a 2% increase in hippocampal volume as measured by MRI, compared to a 1.4% decrease in the control group that did stretching only. A 2% increase may sound modest, but hippocampal volume typically decreases with age at a rate of approximately 1-2% per year — meaning one year of running effectively reversed 1-3 years of hippocampal aging. The exercise group also performed better on spatial memory tasks. Critically, the hippocampal volume changes were directly correlated with changes in fitness (VO2 max improvements) and with changes in serum BDNF levels, establishing the mechanistic chain: running → increased fitness → increased BDNF → hippocampal growth → improved memory.

BDNF — Brain-Derived Neurotrophic Factor — is the molecular key linking exercise to neuroplasticity. John Ratey, a Harvard psychiatrist, popularized it as 'Miracle-Gro for the brain' in his 2008 book 'Spark.' BDNF is a protein that promotes the survival, growth, and differentiation of neurons and synapses. Running causes an acute 2-3x spike in serum BDNF during and immediately after exercise, and chronic training roughly doubles resting BDNF concentrations in the bloodstream. In the brain, BDNF activates the TrkB receptor on neurons, triggering intracellular signaling cascades that promote synaptic strengthening, dendritic branching, and new neuron survival in the hippocampus. BDNF levels are consistently lower in people with depression, anxiety disorders, and Alzheimer's disease — and increasing BDNF through exercise is one proposed mechanism for running's antidepressant and cognitive-protective effects.

Beyond the hippocampus, running produces measurable changes in the prefrontal cortex — the brain region responsible for executive function, working memory, impulse control, and decision-making. Studies using neuroimaging in both cross-sectional and longitudinal designs show that regular aerobic exercisers have greater grey matter volume in prefrontal areas and perform better on executive function tasks than sedentary controls. The prefrontal cortex is also central to top-down emotional regulation — the capacity to modulate emotional responses, tolerate discomfort, and maintain perspective under stress. This may partially explain why regular runners consistently report improvements in stress tolerance, emotional resilience, and self-control that extend beyond the direct neurochemical effects of individual runs. They are not simply feeling better after each run; they are gradually becoming more neurologically capable of feeling better.

In 2004, paleontologists Dennis Bramble and Daniel Lieberman published a landmark paper in Nature titled 'Endurance Running and the Evolution of Homo.' Their thesis was bold and detailed: humans are uniquely and specifically adapted for long-distance endurance running, and these adaptations — absent in our australopithecine ancestors and in our nearest primate relatives — were central to the emergence of the genus Homo approximately 2 million years ago. The paper catalogued at least 26 anatomical features that distinguish human bodies from other great apes in ways that specifically enhance running performance: the enlarged gluteus maximus (the largest muscle in the body, which does almost nothing during walking but is essential during running), the elongated Achilles tendon (which stores and releases elastic energy like a spring), the plantar arch of the foot (a second elastic energy storage device), the nuchal ligament (a neck ligament that stabilizes the head during running), a narrow pelvis and waist that permits rotational momentum transfer, and the extraordinary capacity to sweat and dissipate heat. No other primate shares this constellation of adaptations.

The behavioral context for these anatomical adaptations is persistence hunting — a hunting strategy in which human hunters pursue large prey at moderate speed over several hours, often in the heat of the day, until the animal overheats and collapses from hyperthermia. While most large mammals can outrun humans in short sprints, humans' ability to thermoregulate through sweating while running gives us a unique advantage in sustained pursuit: we can chase prey indefinitely at a speed they cannot maintain without overheating. This strategy was apparently used by Homo erectus and persists among some modern hunter-gatherer groups, including the San Bushmen of the Kalahari and the Tarahumara of Mexico. Anthropologists have documented persistence hunts in which men ran prey (kudu, wildebeest) for 2-5 hours at 6-8 km/h across open savanna in 35-40°C heat. The prey walked, trotted, and then collapsed; the hunters ran — moderately, sustainably, without stopping.

The endocannabinoid reward system, which produces euphoria specifically during sustained moderate-intensity running (the persistence-hunting intensity range), appears to be an evolutionary motivational mechanism. Lieberman's group at Harvard has argued that human ancestors who found running pleasurable survived better — they were more likely to engage in the sustained pursuit that resulted in successful hunts and caloric intake. The pleasure of running, from this perspective, is not incidental or hedonistic but adaptive: it is a neurochemical reward system that motivated a critical survival behavior. This evolutionary framing explains several observations that are otherwise puzzling: why the runner's high appears at exactly the intensity and duration of persistence hunting (moderate pace, 30-60 minutes), why it does not appear during sprinting or very slow walking, and why humans and dogs (another persistence-hunting species) show the endocannabinoid response while ferrets (a sprint predator) do not.

Running was also profoundly social in the evolutionary context. Persistence hunting was a cooperative activity: multiple hunters worked together, with different individuals taking point when others tired, tracking the prey across terrain, and coordinating without modern communication tools. The social dimension of running appears to activate additional neurochemical pathways — mirror neuron systems, oxytocin release from synchronized movement, and the endorphin-mediated social bonding that has been documented in group exercise research. This evolutionary context gives modern phenomena — running clubs, parkrun communities, mass marathon events — a deeper significance. They are not merely recreational preferences; they recreate the social structure of persistence hunting, activating ancient motivational systems that our brains are specifically wired to respond to. The surge of emotion many runners feel crossing a marathon finish line in a crowd has evolutionary antecedents in the social celebration of a successful group hunt.

One of the most counterintuitive aspects of running — particularly for non-runners — is that suffering and satisfaction coexist during and after hard efforts. Long-distance runners speak of the agony of miles 20-25 in a marathon and the profound fulfillment of crossing the finish line with genuine affection for the experience. Ultra runners describe hour upon hour of physical misery followed by deep peace and joy. How does pain become pleasurable? The answer involves both the peripheral biology of beta-endorphins and a fundamental principle of neurological adaptation called opponent-process theory.

Beta-endorphins, while unable to cross the blood-brain barrier easily, do perform an important function during intense running: they bind to opioid receptors in the peripheral nervous system and spinal cord, reducing afferent pain signaling from muscles, joints, and connective tissue. This peripheral analgesia is real and substantial. Trained runners can sustain forces and discomforts during hard training that would be intolerable without this endogenous pain-suppression system. The analgesic effect increases with both intensity and duration, which is why pain tolerance tends to be higher during the final miles of a marathon (when endorphin production is at its peak) than in the opening miles. This is not the euphoria of the runner's high — it is the numbing of pain signals that allows performance to continue despite significant physical stress.

Opponent-process theory, developed by Solomon and Corbit in 1974, proposes that the brain actively counteracts any strong emotional state with an opposing process of opposite valence. When you experience an unpleasant stimulus (the physical discomfort of running hard), the nervous system mounts an opponent response (relief, calm, sometimes euphoria) that activates during and persists after the stimulus ends. Critically, the theory predicts that with repeated exposure, the opponent response grows stronger while the primary response diminishes — which is exactly what experienced runners describe: the suffering of running becomes easier to tolerate over time, while the post-run satisfaction intensifies. The first run of a training block hurts and produces modest pleasure afterward. After months of training, the same run hurts less and produces greater satisfaction. The neurological economics of suffering improve with experience.

Physical pain tolerance itself changes with training in measurable ways. Assa and colleagues demonstrated in 2019 that trained endurance athletes have significantly higher pain thresholds and pain tolerance (the ability to maintain exposure to a pain stimulus) than matched sedentary controls, even when tested with stimuli unrelated to exercise (cold pressor tests, pressure pain). This change in pain processing appears to be partly mediated by BDNF-driven changes in central pain processing pathways, and partly by repeated exposure and psychological adaptation. Runners genuinely feel pain differently than non-runners — not because they suppress the signal, but because their brains process and contextualize pain signals through a more experienced, less alarmed nervous system. This altered pain relationship is one reason many elite runners describe sensations during racing that would seem agonizing to an outside observer as simply 'part of the experience' — they have trained their pain response as deliberately as they have trained their cardiovascular system.

Running alongside other people activates neurological systems that solo running does not. When humans move in synchrony — matching stride for stride, breathing in rhythm, turning together — mirror neuron systems in the premotor cortex fire as if the observer is performing the observed movements. This neural mirroring creates a functional resonance between moving bodies that is associated with empathy, social cohesion, and prosocial behavior. The synchronized movement of group running is not merely convenient or motivating in a superficial sense; it engages ancient neural machinery for social bonding that our brains evolved to use in the context of cooperative physical activity — exactly the activity our ancestors performed when hunting together.

Tarr and colleagues published a striking experimental demonstration of this in 2015 in Biology Letters. Participants were assigned to exercise in groups either in synchrony (matching movements) or in the same space but asynchronously. After exercise, researchers measured pain thresholds using a pressure algometer — a standard measure of endorphin and endocannabinoid system activity. The synchronous group had significantly higher pain thresholds than the asynchronous group despite performing the same physical work at the same intensity. The study directly implicated social synchrony — not just the presence of others — as an independent driver of endorphin and endocannabinoid release during exercise. Running with others, at the same pace, activates neurochemical pathways that solo running at the same effort does not.

The formation of running identity around groups mirrors the anthropology of tribal cohesion. Running clubs, parkrun communities, online Strava segments and leaderboards, marathon training groups, and relay team cultures share a structural feature: they create in-group identity around a shared physical practice. Research on group identity formation consistently shows that shared suffering creates stronger social bonds than shared pleasure — the phenomenon that military researchers have called 'stress inoculation bonding,' where units who train together under physical hardship develop cohesion that outlasts the training. The 'suffer together' dimension of a hard group workout, a cold morning club run, or a challenging race is not incidental to the social bonding of running communities — it is the mechanism. Shared physical challenge activates oxytocin release, reduces competitive anxiety among group members, and creates a memory of collective achievement that functions as a social identity anchor.

Races amplify the social neurochemistry of running to an extraordinary degree. The combination of synchronized movement with thousands of other runners, the crowd response at key moments (start, finish, wall sections), the shared suffering of late-race fatigue, and the cultural ritual of the mass start creates a social neurochemical environment that most runners describe as qualitatively different from any training run — regardless of performance. Runners who race in silence often report a less powerful emotional experience than those who make eye contact, exchange encouragement, or high-five spectators. The finish line emotion — that overwhelming surge of feeling that causes even experienced, analytical runners to cry — is not produced by physical exertion alone. It is the convergence of endocannabinoid euphoria, dopaminergic reward at goal achievement, endorphin analgesia releasing, and oxytocin-mediated social completion: a neurochemical peak that requires the social context of racing to reach its full intensity.

The most common misconception about running is that people who love it simply have greater willpower or discipline than those who don't. The neuroscience suggests a more complex picture. Whether running 'takes hold' for any given individual depends on the interplay of genetic factors (dopamine and endocannabinoid receptor density and sensitivity), early experience (was the first running environment supportive or punishing?), starting intensity (too hard reduces endocannabinoid response), and duration (too short to reach the neurochemical threshold). People who give up running after a few weeks and conclude it's 'not for them' may simply have encountered the worst possible conditions for their reward systems: high-intensity short runs that never activated the endocannabinoid system, in isolating conditions that deprived them of the social neurochemistry of group movement.

Initial experience appears to be disproportionately influential. The neurological response to a first running experience creates a prediction that the brain uses to evaluate all subsequent running experiences. If the first runs are characterized by misery without reward — common when beginners run too hard, too far, too soon, in isolation — the brain encodes running as a net negative experience and the motivational system assigns it low priority. If early runs are at appropriate intensity (conversational pace, Zone 1-2), in a social context, with achievable goals and clear progress signals, the brain's prediction of reward is confirmed and strengthened. This is why the structure of beginning running programs (Couch to 5K, beginner club training groups) matters neurologically: they are, inadvertently, optimizing for the neurochemical conditions most likely to hook the reward system before the beginner quits.

The first 2-4 weeks of running present a particular neurological challenge. Before significant mitochondrial adaptation, cardiac efficiency improvements, and musculoskeletal strengthening occur, running is physiologically hard at almost any pace. The aerobic system is undertrained, the muscles are unaccustomed, and the joints are adapting to impact loads they have rarely experienced. During this period, the body has not yet built the fitness required to produce a meaningful endocannabinoid response at comfortable intensities — which requires sustained moderate effort, not beginner-level struggling. Most people who quit running do so in this window. They quit before their physiology catches up to their neurochemistry requirements. Coaches and running programs that explicitly address this gap — by setting realistic intensity targets, emphasizing social context, celebrating non-performance milestones, and normalizing the difficulty of the first weeks — are working with the neuroscience of habit formation rather than against it.

Key Neurochemicals in Running

The World Health Organization's 2022 global guidelines on physical activity recommend 150-300 minutes per week of moderate-intensity aerobic exercise for adults, specifically citing prevention of depression, anxiety, and cognitive decline alongside cardiovascular and metabolic benefits. This recommendation represents a watershed moment: a global health body formally recognizing exercise as a preventive intervention for mental health conditions, not merely physical ones. The evidence base supporting this recommendation has accumulated over three decades and is now extensive enough that most psychiatric professional organizations include exercise recommendations in their clinical guidelines for mild to moderate depression and anxiety.

The most influential early clinical trial was conducted by Blumenthal and colleagues, published in JAMA in 1999. In a randomized controlled trial of 156 adults with major depressive disorder, participants were assigned to 16 weeks of aerobic exercise, sertraline (a standard SSRI antidepressant), or a combination of both. At 16 weeks, all three groups showed comparable remission rates, with no statistically significant difference between the exercise-only group and the sertraline group. The finding was remarkable: a 16-week program of aerobic exercise (30 minutes, three times per week at 70-85% max HR) was as effective as a frontline antidepressant medication for major depression in older adults. A 10-month follow-up analysis was even more striking: the exercise-only group had significantly lower relapse rates than the sertraline group, suggesting that exercise produced more durable remission. Participants who had continued exercising independently after the trial ended had substantially lower depression scores than those who had stopped.

A 2016 meta-analysis by Schuch and colleagues in the Journal of Psychiatric Research synthesized 25 randomized controlled trials of exercise versus placebo control for depression. The pooled effect size was approximately 0.7 — a large effect in clinical psychology terms, comparable to the 0.8 effect size typically reported for antidepressant medication versus placebo. The analysis included quality adjustments for methodological rigor and found that even when applying conservative corrections, the effect size remained clinically meaningful. The mechanisms proposed include: serotonin upregulation (specifically 5-HT1A receptor sensitivity), normalization of HPA axis hyperactivity (the stress-cortisol dysregulation common in depression), BDNF-driven hippocampal neurogenesis (hippocampal volume is measurably reduced in clinical depression and recovers with both antidepressants and exercise), and norepinephrine system normalization. No single mechanism fully accounts for the antidepressant effect; it appears to be a multi-system intervention.

Running vs Medication for Mild-Moderate Depression

An important clinical caveat must accompany this evidence: running is well-supported as a treatment for mild to moderate depression, but the evidence for severe depression is considerably weaker. People with severe major depressive episodes often cannot initiate or sustain exercise due to anergia, psychomotor retardation, and profound motivational deficits — the very symptoms that exercise would eventually help. For severe depression, pharmacological or psychotherapeutic intervention is typically required before exercise becomes feasible. Additionally, running should be understood as a complement to, rather than a replacement for, professional mental health care for anyone who is suffering significantly. The growing clinical evidence does, however, argue strongly that exercise should be routinely prescribed alongside conventional treatments, and that the common medical practice of failing to discuss exercise as a therapeutic intervention for depression represents a missed opportunity of substantial magnitude.