Anxiety: symptoms, causes, treatment and 2025 evidence

Anxiety is the mind and body’s anticipatory alarm, the cluster of worry, physical arousal and avoidance that fires when the brain predicts a threat that may or may not arrive. In small doses it is the system working as designed: it gets you to revise for the exam, leave on time for the flight, and pay attention to the email you would rather ignore. It becomes a clinical problem when the alarm runs too often, fires too loudly, or refuses to switch off after the threat has passed.

This hub is the anchor page for everything we publish on anxiety. It explains what anxiety actually is, how clinicians distinguish the major presentations, what 2025 evidence says about causes and treatments, and where the field stands on emerging questions such as algorithm-driven feeds and the wider influence of digital life on the anxious nervous system.

What anxiety is, clinically

In a clinical sense, anxiety is not the same thing as feeling stressed before a job interview. The diagnostic threshold turns on three features: the worry is excessive relative to the situation, it is difficult to control, and it causes meaningful interference with work, sleep, relationships or daily functioning. Most people who meet criteria for an anxiety disorder describe the experience as something that has stopped feeling like a useful warning and started feeling like a noise their nervous system cannot turn down.

The body part matters as much as the mental part. Anxiety travels with a recognisable physiology, a faster heart rate, shallow breathing, tense shoulders and jaw, tight chest, a queasy stomach, restless sleep, and a mind that keeps generating the next thing to worry about. Patients often describe the feeling as being braced for impact even when nothing is hitting them. That mismatch, between the body’s preparation and the absence of an actual threat, is the central engine of what makes anxiety disorders distinct from everyday worry.

A clinician’s job is to work out whether the pattern fits one of the recognised presentations below, what is driving it, and what to do about it. The tools to do that have improved substantially in the past five years, and the evidence base for treatment is now strong enough that “you’ll have to live with it” is no longer an honest thing to say to most people.

The major anxiety presentations

Anxiety is not a single condition. It is a family of related disorders that share the alarm physiology but differ in what triggers it, how it shows up, and what helps most. The five most common presentations sit on the same continuum but warrant different treatment plans.

Generalised anxiety disorder (GAD) is the prototype the rest of the family is compared to. The worry is broad, finances, health, relationships, the news, the future, and the person typically describes it as “always on” rather than tied to one situation. GAD is what most people mean when they describe anxiety as the everyday worry that won’t shut off, which is also why it sits at the centre of this cluster.

Social anxiety disorder narrows the alarm onto situations involving other people’s judgement: speaking up in meetings, eating in public, networking events, dating. The fear is specifically of being negatively evaluated; the avoidance pattern that follows is the part that usually does the most harm to a person’s career, friendships and confidence.

Panic disorder is defined by recurring panic attacks, short, intense waves of physical symptoms that often feel like a heart attack or imminent collapse, and the anticipatory fear of when the next one will hit. Many people who have one or two panic attacks never develop the disorder; what crosses the threshold is the build-up of fear about the attacks themselves.

Specific phobias are intense, focused fears of a particular object or situation: heights, flying, needles, dogs, vomiting, enclosed spaces. They are the most prevalent anxiety presentations in the population and also the most treatable, because the exposure-based protocols that work for them are well-mapped and short.

Health anxiety sits slightly off to the side of the family. It is driven less by a fear of external threats and more by an internal preoccupation with bodily sensations and the meaning they might carry. Health-anxious people tend to over-monitor, over-research and over-consult, and the reassurance they receive tends to fade quickly and need to be sought again. We will publish dedicated cluster pages for each of these presentations over time; this hub keeps the picture deliberately wide.

Comorbidity is the rule rather than the exception. Roughly half of people who meet criteria for one anxiety disorder also meet criteria for a second, and overlap with depression sits near the same rate. That does not mean the labels are useless, they still guide treatment selection, but it does mean a tidy single-diagnosis frame is often a clinical fiction. The practical implication is that an honest first conversation with a therapist usually maps the whole picture before settling on a treatment plan.

How anxiety differs from worry, fear and stress

These four words get used interchangeably in everyday language, and clinically they describe slightly different things. Keeping the distinctions in mind helps when reading both your own experience and what the research says.

Fear is the response to a present, identifiable threat, the car coming towards you, the dog with its hackles up. It is precise, time-limited and usually ends when the threat passes. Anxiety is the response to a possible or imagined future threat. It can have no clear endpoint because the predicted danger has not arrived and may never arrive. Worry is the cognitive content of anxiety: the chain of “what if” thoughts the mind generates to try to control the unknown. Stress is the body’s response to demand, anything from a deadline to a new baby to an illness, and may or may not include anxiety, depending on how the person appraises the situation.

The reason this matters: treatment depends on which one you are dealing with. Fear responses to genuine danger should be respected, not pathologised. Stress often responds to changes in workload, sleep and recovery time. Worry as a cognitive habit responds to specific psychotherapeutic techniques that don’t always work for stress. And anxiety as a clinical disorder usually needs the structured packages described later in this guide.

What causes anxiety: the 2025 biopsychosocial picture

The honest answer to “what causes anxiety” in 2025 is that it is multiply determined. No single gene, brain region, life event or cultural pressure explains the variance on its own. The most useful frame is biopsychosocial: biology sets the susceptibility, psychology shapes the pattern, and the social environment provides the load.

Biological contributions include heritable temperament (anxious people often describe at least one anxious parent), differences in amygdala–prefrontal cortex coupling, autonomic reactivity, and the gut–brain axis. The genetics are polygenic, many small contributions rather than one big one, and they interact heavily with environment. Sleep deprivation, alcohol withdrawal, caffeine load, thyroid dysfunction and some prescription medications can all push an otherwise stable system into clinical anxiety.

Psychological contributions include early attachment patterns, learned beliefs about safety and control, cognitive habits such as catastrophising and intolerance of uncertainty, and the specific avoidances a person has built over time to manage previous bouts. These are the modifiable ingredients that most psychotherapy targets.

Social contributions in 2025 are receiving fresh attention. Economic instability, climate anxiety, the post-pandemic shape of work and education, and, crucially, the texture of life inside algorithmically curated feeds are all candidate contributors. The literature on the social media side has matured enough that it now has its own section below.

It is worth saying clearly: anxiety is not a sign of personal weakness, a failure of resilience, or something a person should be able to “think themselves out of”. It is a real disorder of an overactive prediction system, and the treatments that work treat it as such.

Evidence-based treatments: what actually works

The treatment evidence base for anxiety is, by some distance, the strongest in mental health. Three modalities have the deepest support, and they work best in combination rather than competition.

Cognitive-behavioural therapy (CBT) remains the first-line psychological treatment for almost every anxiety presentation. Meta-analyses across two decades show consistent moderate-to-large effects for GAD, panic disorder, social anxiety, specific phobias and health anxiety. The active ingredients are not surprising: identifying the cognitive habits that maintain the worry, testing them against reality, and using gradual exposure to dismantle the avoidances the anxiety has built. Modern CBT delivery has expanded to brief, group and videoconference formats with effects comparable to traditional weekly sessions, which has helped the access problem this field has historically struggled with.

Medication has a clear role in moderate to severe presentations. SSRIs and SNRIs are the most commonly prescribed first-line agents and produce reductions in symptom severity comparable to CBT in many head-to-head trials. The decision to medicate is individual and best made jointly with a prescriber who can weigh side-effect profiles, comorbid depression, prior treatment history and personal preference. Benzodiazepines have a narrow short-term role and are not appropriate for long-term maintenance.

Mindfulness-based approaches have moved from “complementary” to mainstream over the past decade. Mindfulness-Based Stress Reduction (MBSR) and Mindfulness-Based Cognitive Therapy (MBCT) both have substantial trial evidence for generalised anxiety and recurrent depression, and they integrate naturally with CBT for many patients. The mechanism is not mystical, it is structured practice in observing thoughts without acting on them, which is one of the most useful skills an anxious mind can build.

For people whose anxiety is intertwined with their social media use, our deep dive on the evidence-based treatment protocols for that specific pattern sets out which of the above modalities adapt best, and a separate piece on the five-rule protocol therapists are using in clinic translates the research into the day-to-day behavioural changes patients actually try first.

The 2025 evidence on social-media-driven anxiety

The most active research area in anxiety this year has been the relationship between social media use and adolescent mental health. The reason the field is moving fast is that the methodology finally caught up with the question. After a decade of cross-sectional surveys, 2025 produced the first generation of properly pooled meta-analyses, longitudinal cohorts and randomised abstinence trials that can sustain causal weight.

A 2025 systematic review and meta-analysis in Behavioral Sciences pooled 24 studies for 68 effect sizes and reported an aggregate correlation of roughly r = 0.22 between exposure to social media risk factors and adolescent mental disorder outcomes, small to moderate, but consistent and robust. A 2024 Journal of Adolescent Health review focused specifically on adolescent anxiety found stronger correlations, around r = 0.35, when the exposure variable was problematic rather than total use. The distinction matters enormously: it is not the time on the app, it is the behavioural pattern.

The 2023 US Surgeon General’s advisory had already framed adolescent social media use as a public-health priority, and the 2025 evidence has largely supported that framing while sharpening it. A 2025 Scientific Reports meta-analysis of social media abstinence trials found that structured breaks produce small but real improvements in affective well-being, meaningful at population scale, modest at the individual level.

If you want to follow the threads here in detail, we have ranked the seven 2025 findings most worth knowing, set out the contrarian case that the panic has been overstated, looked at the demographic split between girls and boys in Gen Z, and traced the neuroscience of the dopamine loop that helps explain why the feeds are sticky in the first place. The 30-day detox question tests what happens to anxiety symptoms when someone steps off, and the wider overdiagnosis debate sits inside our broader editorial position that headline numbers and lived experience often pull in different directions.

The honest summary of the 2025 evidence is that algorithmic platforms do appear to act as a load amplifier on already-anxious nervous systems, the effect is real but smaller than the loudest headlines suggest, and the people most affected, particularly mid-adolescent girls and existing problematic users, are also the people for whom the available interventions work best.

Two further threads from the 2025 literature are worth flagging because they often surprise people. First, the way anxious teens use social media is genuinely different from how non-anxious peers use it, more time, more dependence, more reactivity to what they see, which complicates any clean one-way causal story. Second, adults are not exempt: a 2025 finding that has caught researchers off guard is that mothers of teenagers often report higher social-media-driven anxiety than their teenagers do, driven by a mix of parenting worry and direct platform use. Both findings push the picture further from “phones are ruining the kids” and closer to a system-level account of how attention, emotion and platform design interact across whole households.

Where this all leaves the everyday reader is somewhere genuinely useful. If you suspect your own anxiety is being amplified by your feed, the most actionable 2025 advice is not to argue about screen-time policy but to watch for the problematic-use indicators, loss of control, reaching for the app on autopilot, mood crashes after sessions, sleep disruption, and to test a structured break against the rest of an evidence-based plan rather than as a standalone fix.

When to seek professional help

A reasonable rule of thumb: if anxiety has been getting in the way of work, sleep, relationships or how you live your life for more than a couple of weeks, and self-directed approaches are not shifting it, it is time to talk to a professional. A GP or family doctor is a sensible first stop in most health systems; they can rule out medical contributors, signpost to talking therapies, and start a conversation about medication if appropriate.

Some signals warrant faster action. If you are experiencing panic attacks that are restricting where you go or what you do, persistent avoidance of work or school, intrusive thoughts of self-harm, or anxiety that is severe enough to disrupt eating or sleep for more than a week, do not wait. Anxiety is highly treatable, and treatment works faster the earlier it starts. If you are in immediate distress, the crisis support resources on our topics page list appropriate hotlines and routes to urgent help.

For people in milder territory who simply want to act on what they are noticing, the most useful first move is usually to write down what the anxiety is about, how often it shows up, what makes it better and what makes it worse, and what you have already tried. Bringing that to a first appointment compresses several sessions of work and tends to lead to a more focused plan.

Editorial note

This hub is updated as new evidence lands. The anxiety research base is moving quickly, the 2025 syntheses are already reshaping how clinicians think about screen-based contributors, and the next two years are likely to bring further refinement on which interventions work best for which presentations. When the picture shifts, this page shifts with it, and the satellite articles below are versioned individually so they can be updated without disturbing the overall hub.

If you read one related piece next, we suggest the biggest findings ranked for the current state of the social media question, or the signs and symptoms walkthrough if you came here trying to make sense of what you are feeling.