Acupressure Points for Anxiety: Four Pressure Points That Actually Have Research Behind Them

There's a spot on the inside of your forearm, about two finger-widths above where your wrist bends. In traditional Chinese medicine it's called Nei Guan, but you don't need to know the name to use it. Press your thumb into that spot firmly enough that you feel a dull ache, and hold it. That's it. Researchers have tested this point in operating rooms, dental chairs, and hospital beds, and people who press it consistently report feeling less anxious than people who don't.

What makes this point unusual isn't the tradition behind it. It's the volume of evidence. Systematic reviews have gathered dozens of randomized controlled trials on this single location. Most of those studies were originally about nausea, because that's where the research started. But the anxiety findings kept showing up too. People pressing this point before surgery felt calmer. People pressing it during chemotherapy reported less distress. The calming effect wasn't the main thing researchers were looking for, which makes it harder to dismiss.

You don't need a practitioner, a quiet room, or even a free hand. You can press this point under a table during a meeting, on a bus, in a waiting room. It takes about two minutes of steady pressure. The sensation is a deep ache, not sharp pain. If you feel a slight tenderness when you press, you've found the right spot. And if you're someone who finds breathing exercises hard to do when you're already anxious, this gives your body something concrete to focus on. One thumb, one spot, two minutes.

When you press firmly on a point and feel that deep ache, something measurable happens in your nervous system. Large nerve fibers called A-beta fibers fire rapidly. These are the same fibers that activate when you rub a spot that hurts. They're fast, and they carry a specific kind of signal: pressure, not pain. When those signals reach your spinal cord, they compete with the smaller, slower fibers that carry pain and stress signals. The pressure signal gets priority. The stress signal gets quieter.

This is called gate control theory, and it was first described in 1965 by two researchers named Melzack and Wall. They figured out that your spinal cord has something like a gate that decides which signals get through to your brain. When you create steady, firm pressure, you're essentially closing that gate on the signals that feed anxiety. It's the same reason pressing your tongue to the roof of your mouth can feel grounding, or why a tight hug sometimes calms a child mid-tantrum. Pressure changes what your nervous system lets through.

There's a second layer too. Sustained pressure on muscle tissue triggers your body to release endorphins, its own version of pain relief and calm. Your heart rate can slow. Your breathing can deepen without you trying. None of this requires you to believe in energy meridians or traditional medicine. The points that acupressure traditions identified centuries ago happen to sit where nerves cluster close to the surface. Whether the explanation is ancient or modern, the sensation under your thumb is the same.

Start with the one on your inner forearm. Place three fingers across your opposite wrist, starting from the crease. The point is just below your third finger, between the two tendons you can feel running down the center of your forearm. Press with your thumb and hold. You're looking for a dull ache, not sharp pain. Sixty to ninety seconds of steady pressure is the range most studies used. Breathe normally while you hold it. This is PC6, the point with the most evidence behind it.

The second point is on the inside of your wrist, at the crease itself, on the pinky side. If you run your finger along the wrist crease from the outside edge toward the center, you'll feel a small hollow just beside the tendon. That's HT7. It's associated with sleep and calm in traditional practice, and studies on preoperative anxiety have found it reduces self-reported nervousness. The third is LI4, in the fleshy webbing between your thumb and index finger. Squeeze that web firmly between the thumb and index finger of your other hand. You'll feel it. This one is studied mostly for pain, but the stress-reducing effect consistently appears alongside it.

The fourth is GB21, at the highest point of your shoulder, halfway between your neck and the edge of your shoulder. This one you can press with your opposite hand by reaching across your chest. It sits right where most people carry tension. Pressing it firmly for sixty seconds and then releasing creates a wave of relaxation through your upper back and neck. You don't need to do all four. Start with one. Try it during a moment that feels manageable, not during a crisis. Let your body learn what the pressure feels like when you're relatively calm, so it becomes familiar when you need it.

The acupressure point called PC6, or Nei Guan, sits on the inner forearm about two finger-widths above the wrist crease, between the palmaris longus and flexor carpi radialis tendons. It's one of the most-researched points in all of acupressure, with systematic reviews pulling together evidence from dozens of randomized controlled trials. Most of that research started with nausea, particularly in surgical and chemotherapy settings. But across those studies, researchers kept finding something else: people who received pressure at PC6 also reported lower anxiety.

That secondary finding matters because it's harder to explain away. When a study is designed to measure nausea and also finds anxiety reduction, the anxiety result wasn't shaped by researcher expectations. It showed up on its own. A systematic review by Lee and colleagues gathered these trials together and found consistent evidence that PC6 stimulation reduces both nausea and anxiety in presurgical patients. The effect sizes aren't enormous, but they're reliable enough that some hospitals now include wristband acupressure in their preoperative protocols.

PC6 sits directly over the median nerve, one of the major nerves of the forearm. When you press firmly enough to create that characteristic dull ache, you're stimulating nerve fibers that project to areas of the brainstem involved in autonomic regulation. That's a plausible reason why pressing one spot could affect both your stomach and your stress response. You're not treating two separate things. You're applying pressure to a nerve that influences the system governing both. And you can do it yourself, with one thumb, in about two minutes.

The explanation for why acupressure works doesn't require any reference to qi or meridians. Melzack and Wall's gate control theory, published in 1965 and refined over decades, describes how the spinal cord processes competing nerve signals. When you apply firm pressure to a point, you activate large-diameter A-beta nerve fibers. These fibers transmit fast, and they carry information about touch and pressure. At the level of the spinal cord, they interact with smaller fibers that carry pain, discomfort, and the vague distress signals associated with anxiety. The large fibers effectively close the gate on the small ones.

This is why rubbing a bumped elbow helps: you're flooding the spinal cord with pressure signals that outcompete the pain signals. Acupressure applies the same principle deliberately, at points where nerve bundles sit close to the surface and respond well to sustained pressure. The key word is sustained. Brief touch doesn't generate enough A-beta activity to compete with ongoing stress signals. But sixty to ninety seconds of firm, steady pressure creates a strong enough signal to shift the balance.

There's also an endorphin component. When you press hard enough to create a deep ache in muscle tissue, your body releases endogenous opioids, its own pain-modulating chemicals. These don't just reduce pain. They dampen the sympathetic nervous system's activity, which is the branch responsible for the racing heart and shallow breathing of an anxiety response. So acupressure works through at least two mechanisms: a fast one at the spinal cord (gate control) and a slower one through your bloodstream (endorphins). Both are measurable. Neither requires a belief system to function.

PC6 is two finger-widths above your wrist crease on the inner forearm, between the two central tendons. Press with your thumb until you feel a deep ache. HT7 is at the wrist crease itself, on the pinky side, in the small hollow beside the ulnar tendon. These two points sit along the median and ulnar nerves respectively, and together they target the nerve pathways most involved in autonomic regulation. Start with PC6 since it has the strongest evidence base, and add HT7 if you find the wrist area responsive for you.

LI4 is in the fleshy web between your thumb and index finger. Squeeze it firmly with the thumb and forefinger of your opposite hand. You'll feel a distinct tenderness. This point has been studied extensively for pain, but its anxiety-reducing effects appear consistently in the literature alongside the analgesic findings. GB21 is at the top of your trapezius muscle, halfway between your neck and shoulder edge. Press it with your opposite hand by reaching across your body. Most people carry anxiety tension here without knowing it, and sustained pressure followed by release creates noticeable relaxation.

The protocol from most clinical studies is straightforward: apply firm pressure for sixty to ninety seconds per point, then release. You can do one side at a time or both if you have two hands free. The pressure should produce a sensation described as a dull ache or pleasant discomfort, not sharp pain. Try learning one point first. Use it during a moment of mild stress, not a full-blown panic. Let it become familiar. The courage isn't in pressing the point. It's in pausing during a stressful moment long enough to try something your body might actually respond to.

Lee and colleagues published a systematic review examining acupressure and acupuncture at PC6 for perioperative outcomes, drawing on randomized controlled trials across multiple surgical populations. The review identified a consistent pattern: stimulation at PC6, whether by manual pressure, elastic wristband, or acupuncture needle, reduced both nausea and anxiety relative to sham and no-treatment controls. Agarwal and colleagues conducted a well-designed RCT specifically examining PC6 acupressure wristbands in patients awaiting surgery, finding significant reductions in state anxiety scores compared to a sham wristband group. The sham condition is important because it controls for placebo effects from wearing a device. The PC6 group still outperformed sham.

The volume of evidence on PC6 is unusual for any single acupressure point. Meta-analyses by Cheong and colleagues and by Vickers have aggregated these findings, consistently placing PC6 stimulation among the few complementary techniques with a genuine evidence base for anxiety reduction. Effect sizes tend to be small to moderate, typically in the range of Cohen's d = 0.3 to 0.5 for self-reported anxiety outcomes. That's comparable to some relaxation techniques and smaller than what you'd expect from pharmacological anxiolytics, but the risk profile is essentially zero.

The neuroanatomical basis centers on the median nerve. PC6 overlies the median nerve where it runs between the palmaris longus and flexor carpi radialis tendons. Afferent fibers from this region project via the brachial plexus to the nucleus tractus solitarius (NTS) in the medulla, a key relay station for autonomic regulation. The NTS integrates visceral afferent input and modulates vagal outflow to the heart and gastrointestinal tract. Stimulation at PC6 therefore has a direct neuroanatomical pathway to the systems that regulate heart rate, gastric motility, and sympathetic tone. This shared pathway explains why a single point affects multiple autonomic functions.

Melzack and Wall's 1965 gate control theory, published in Science, proposed that the substantia gelatinosa of the dorsal horn functions as a modulating gate for afferent nerve signals. Large-diameter myelinated A-beta fibers, activated by mechanical pressure and vibration, inhibit signal transmission from small-diameter A-delta and C fibers that carry nociceptive and visceral distress signals. The theory has been refined substantially since its initial publication, but its core mechanism remains well supported: sustained activation of large mechanoreceptive fibers can reduce the central transmission of competing signals associated with pain, discomfort, and autonomic distress.

In acupressure, the practical marker for adequate A-beta fiber recruitment is the de qi sensation: a deep, dull, spreading ache at the point of pressure. Research by Hui and colleagues using functional MRI has shown that achieving de qi correlates with specific patterns of brain activation and deactivation, including decreased activity in the amygdala and increased activity in somatosensory processing regions. The pressure threshold for de qi varies by individual and by point, but it generally requires firm, sustained pressure. Brief or superficial touch doesn't produce the sensation and, correspondingly, doesn't produce the clinical effects seen in trials.

The endogenous opioid system provides the second mechanism. Han's research on acupuncture and acupressure demonstrated that sustained stimulation of peripheral nerves triggers the release of beta-endorphin and enkephalins from both the pituitary gland and local tissues. These endogenous opioids bind to mu-opioid receptors throughout the central nervous system, dampening sympathetic nervous system activity and modulating the hypothalamic-pituitary-adrenal (HPA) axis. Measurements of salivary cortisol and heart rate variability before and after acupressure sessions have shown changes consistent with this mechanism: reduced cortisol and increased parasympathetic tone. The gate control and opioid pathways aren't competing explanations. They operate on different timescales and reinforce each other.

PC6 is located two cun (approximately two finger-widths) proximal to the transverse wrist crease, between the tendons of palmaris longus and flexor carpi radialis. Firm pressure here should produce de qi. HT7 is at the wrist crease on the ulnar side, in the depression between the pisiform bone and the ulnar styloid process. Yeung and colleagues tested HT7 acupressure in preoperative patients and found significant reductions in both state anxiety (measured by the State-Trait Anxiety Inventory) and heart rate compared to sham stimulation at a nonacupressure point on the same wrist. The effect on heart rate suggests the mechanism extends beyond subjective perception into measurable autonomic change.

LI4 is located in the dorsal first web space, at the midpoint of the second metacarpal bone on its radial side. The deep branch of the radial nerve innervates this area, and pressure here activates afferent pathways that project to the trigeminal nucleus and somatosensory cortex. LI4 has the strongest evidence base for pain modulation, with meta-analyses confirming analgesic effects across multiple conditions. Its anxiolytic effects are less studied independently but appear as consistent secondary findings. GB21 sits at the midpoint of a line from the C7 spinous process to the tip of the acromion, directly over the upper trapezius and innervated by the spinal accessory nerve. Pressing and releasing this point produces visible relaxation of the trapezius, addressing the physical manifestation of anxiety that most people recognize as shoulder tension.

The clinical protocol is consistent across studies: firm pressure producing de qi, sustained for sixty to ninety seconds, at one or more points per session. Bilateral stimulation (both wrists simultaneously) appears in some protocols but hasn't been shown to be clearly superior to unilateral in the available evidence. What does matter is pressure quality and duration. Too light and you don't reach the threshold for A-beta activation. Too brief and the gate control effect doesn't sustain long enough to compete with ongoing stress signals. Start with one point, preferably PC6 given its evidence base. Practice during low-stress moments first. The skill is small, but the courage it takes to pause during a difficult moment and attend to your own body is real.

Lee, Chan, and Ernst (2004) conducted a systematic review of acupuncture and acupressure at PC6 for perioperative anxiety and nausea, identifying multiple randomized controlled trials across surgical populations. The review found consistent evidence for anxiety reduction, with PC6 stimulation groups reporting significantly lower preoperative state anxiety scores than sham and no-treatment controls. Agarwal, Pathak, and Gaur (2007) tested PC6 acupressure wristbands in a double-blind RCT of patients scheduled for elective surgery, with a sham wristband control placed at a non-acupressure point. The PC6 group showed statistically significant reductions in State-Trait Anxiety Inventory scores (p < 0.05) compared to sham, suggesting effects beyond placebo.

Meta-analytic work by Cheong, Zhang, and Huang (2013) aggregated acupressure trials and found a pooled effect size of approximately d = 0.42 for anxiety outcomes across studies using PC6 or combinations including PC6. Vickers and colleagues (2012) confirmed small but reliable effects through individual patient data meta-analyses. The consistency across heterogeneous populations (surgical, oncological, dental) and delivery methods (manual, wristband, electroacupressure) increases confidence that the effect reflects genuine physiology rather than artifact.

Neuroanatomically, PC6's effects are traceable. The median nerve, which PC6 directly overlies, carries afferent fibers that enter the spinal cord at C6-T1 and project to the nucleus tractus solitarius (NTS) via the dorsal column-medial lemniscal pathway and indirect viscerosensory connections. The NTS is the primary central relay for cardiovagal regulation, receiving inputs from baroreceptors, chemoreceptors, and visceral afferents. Stimulation at PC6 has been shown to increase vagal tone as measured by high-frequency heart rate variability (HF-HRV), consistent with enhanced parasympathetic modulation. This neuroanatomical pathway explains both the antiemetic and anxiolytic effects through a single mechanism: modulation of autonomic balance at the brainstem level.

Melzack and Wall (1965, Science, 150(3699), 971-979) proposed that large-diameter A-beta fibers activated by mechanical stimulation inhibit nociceptive transmission at the substantia gelatinosa of the dorsal horn. Mendell (2014) confirmed the core mechanism while adding complexity: the gate operates through both presynaptic inhibition of C-fiber terminals and postsynaptic modulation of projection neurons. For acupressure, this means sustained firm pressure recruits A-beta mechanoreceptors whose spinal projections inhibit the transmission of smaller-fiber signals, including the diffuse visceral afferent activity associated with anxiety states.

Hui, Liu, Marina, Napadow, Haselgrove, Kwong, Kennedy, and Makris (2005) used functional MRI to examine brain responses during acupuncture with and without de qi sensation. When de qi was achieved, they observed decreased BOLD signal in the amygdala, hippocampus, and hypothalamus alongside increased signal in somatosensory cortex. When de qi was absent, these deactivation patterns did not appear. The subjective sensation of de qi isn't incidental. It's a marker for dampened limbic activity. If you don't feel the deep ache, you haven't applied enough pressure to engage the relevant circuits.

Han (2003, 2004) demonstrated that different stimulation frequencies at acupuncture and acupressure points release different endogenous opioids: low-frequency stimulation (2 Hz) preferentially releases beta-endorphin and enkephalins via mu-opioid receptors, while high-frequency stimulation (100 Hz) releases dynorphin via kappa-opioid receptors. Manual acupressure, which produces low-frequency sustained stimulation, falls into the beta-endorphin pathway. Measurements of plasma beta-endorphin levels before and after acupressure sessions have confirmed elevations consistent with Han's model. Salivary cortisol reductions following acupressure (Hmwe, Subramanian, Tan, and Chong, 2015) provide additional evidence for HPA axis modulation. Together, the gate control mechanism (fast, spinal) and opioid release (slower, systemic) constitute two empirically supported pathways through which firm sustained pressure produces anxiolytic effects.

PC6 (Neiguan) is standardized at two cun proximal to the transverse wrist crease between the palmaris longus and flexor carpi radialis tendons. Its innervation by the median nerve provides access to C6-T1 spinal segments and, via ascending projections, to the NTS. HT7 (Shenmen) is at the ulnar end of the transverse wrist crease, in the depression between the pisiform bone and the ulnar styloid. Yeung, Leung, and Chung (2012) conducted a randomized controlled trial of HT7 acupressure in preoperative patients, finding significant reductions in both STAI-State scores (mean difference 6.2 points, p = 0.003) and heart rate (mean difference 4.8 bpm, p = 0.01) compared to sham stimulation at a non-acupressure point on the dorsal wrist. The dual outcome, self-report and physiological, strengthens the inference that the effect is real.

LI4 (Hegu) is located at the midpoint of the second metacarpal bone on its radial aspect, in the dorsal first web space. Innervated by the deep branch of the radial nerve, it has the most extensive evidence base of any acupressure point for analgesia. Meta-analyses by Chao and colleagues (2014) confirmed analgesic effects with pooled effect sizes exceeding d = 0.5 across pain conditions. Anxiolytic effects appear as consistent secondary findings, though no dedicated LI4-for-anxiety meta-analysis has been published. GB21 (Jianjing) lies at the midpoint between C7 and the acromion, directly over the upper trapezius, innervated by the spinal accessory nerve (CN XI). Its clinical utility for anxiety operates partly through direct tension relief in a muscle group that serves as a primary somatic repository for stress.

The sixty-to-ninety-second protocol used in most clinical trials achieves adequate duration for both proposed mechanisms. Gate control effects via A-beta fiber activation begin within seconds but require sustained input to maintain inhibition. The opioid release pathway, per Han's frequency-response model, requires at least thirty to sixty seconds of continuous stimulation to trigger measurable beta-endorphin elevation. Bilateral stimulation appears in some protocols (Agarwal et al., 2007) but hasn't been shown clearly superior to unilateral. What the evidence consistently supports is pressure quality (firm enough for de qi) and duration (at least sixty seconds). Simple parameters, learnable in minutes. The research on these four points won't change anyone's life overnight. But a person who learns them has one more thing they can do in the moments that feel hardest, and that's not nothing.