The overlap between anxiety symptoms and medical pathology is one of the most important diagnostic pitfalls in primary care — studies show that approximately 10–15% of patients presenting to primary care with anxiety symptoms have an identifiable organic cause. Hyperthyroidism and cardiac arrhythmias are the most common. TSH is the single most important investigation — it is abnormal in a clinically significant proportion of patients presenting with new anxiety, and hyperthyroidism is entirely treatable, removing the anxiety without psychiatric intervention. Phaeochromocytoma is rare (1–2 per 100,000) but is one of the most dangerous missed diagnoses — it causes episodic hypertensive crises with severe anxiety, headache, palpitations, and sweating that perfectly mimic panic disorder. Any patient with episodic anxiety accompanied by severe hypertension (systolic >200 mmHg during episodes) or headache must have plasma metanephrines or 24-hour urinary catecholamines before the diagnosis of panic disorder is confirmed. The PTSD/GAD distinction has direct treatment implications — trauma-focused CBT (TF-CBT) and Eye Movement Desensitisation and Reprocessing (EMDR) are the NICE-recommended treatments for PTSD and are fundamentally different from the CBT used for GAD. Treating PTSD with relaxation techniques or standard anxiety management can be actively harmful (exposure to triggers without proper trauma processing can worsen flashbacks). NICE NG116 (PTSD, 2018) specifies that trauma-focused psychological therapy should be offered to all patients with PTSD — not pharmacological management as first-line. GPs must screen for PTSD in any patient with anxiety and a history of trauma before referring to generic anxiety services.

The GAD-7 was developed and validated by Spitzer et al. (2006, Archives of Internal Medicine) as a 7-item brief measure for GAD in primary care — it has since been validated for use across all anxiety disorders including panic disorder, social anxiety disorder, and PTSD, making it a universally applicable screening tool. The cut-off of ≥10 for clinically significant anxiety has a sensitivity of 89% and specificity of 82% for GAD in primary care populations. Like the PHQ-9 for depression, serial GAD-7 documentation is the primary mechanism for tracking treatment response — a 5-point reduction represents a clinically meaningful improvement, and a score below 8 at 8–12 weeks represents a treatment response. NICE NG197 (GAD and Panic Disorder, 2011, updated 2019) recommends recording GAD-7 at every contact. The comorbidity of anxiety and depression is a central clinical fact — population studies consistently show that 50–60% of patients with an anxiety disorder also meet criteria for a depressive disorder, and vice versa. This has three implications: (1) always score both GAD-7 and PHQ-9 together; (2) SSRIs treat both, so pharmacological treatment is simple; (3) the prognosis for comorbid anxiety-depression is worse than either alone, and more intensive treatment (higher dose SSRI, combined pharmacological + psychological treatment) is often needed.

The anxiety disorder classification matters clinically because NICE has different treatment guidelines for each specific anxiety disorder, and the psychological therapy required is distinct. For GAD: CBT focusing on worry postponement, cognitive restructuring, and applied relaxation (NICE NG197). For panic disorder: CBT with interoceptive exposure (deliberately inducing feared bodily sensations to demonstrate they are harmless) and panic cycle psychoeducation (NICE NG197). For PTSD: trauma-focused CBT or EMDR — generic CBT for anxiety is not adequate and not recommended (NICE NG116). For OCD: CBT with exposure and response prevention (ERP) — the distinctive feature of OCD treatment is that the patient must be exposed to the feared stimulus and PREVENTED from performing the compulsion, which is fundamentally different from anxiety management techniques. For health anxiety: CBT specifically targeting the reassurance-seeking cycle and illness worry, using techniques from CBT for hypochondriasis — which is counterintuitive because the clinical instinct is to investigate and reassure, whereas the evidence shows that investigation and reassurance maintain and worsen health anxiety by reinforcing the belief that symptoms require investigation. Social anxiety disorder is the most under-recognised anxiety disorder in primary care — it affects approximately 12% of the general population at some point in their lifetime, but is frequently attributed to "personality" or "shyness" rather than recognised as a treatable disorder. The occupational impairment (inability to speak in meetings, give presentations, or function socially at work) is substantial. SSRI treatment (particularly sertraline and paroxetine — both licensed specifically for social anxiety disorder) combined with CBT produces remission in the majority of patients.

The ECG before starting an SSRI or SNRI in a patient with anxiety presenting with palpitations serves two purposes: (1) it excludes a cardiac arrhythmia as the organic cause (SVT, WPW syndrome, AF — all cause palpitations that are misattributed to anxiety); and (2) it provides a baseline QTc measurement before prescribing citalopram or escitalopram, which cause dose-dependent QTc prolongation. WPW syndrome (Wolff-Parkinson-White — short PR interval + delta wave on ECG) causes paroxysmal SVT that typically presents in young adults as sudden-onset palpitations + anxiety + near-syncope — it is a potentially life-threatening arrhythmia that is entirely manageable with catheter ablation and should not be managed with anxiolytics. The caffeine-anxiety relationship is clinically important and underappreciated — caffeine is a competitive adenosine receptor antagonist that increases adrenaline release, heart rate, and cortisol, directly worsening anxiety and panic symptoms. Studies show that patients with panic disorder are significantly more sensitive to caffeine's anxiogenic effects than non-anxious individuals, and caffeine reduction of as little as 200 mg/day can significantly reduce panic attack frequency. Advising a patient to reduce caffeine intake from 600 mg/day (3 large coffees + 2 energy drinks) to 100 mg/day is a simple, cost-free intervention that can have a dramatic impact. Temporal lobe epilepsy presenting as episodic anxiety (ictal anxiety) is one of the most important neurological differentials for panic disorder — the features that distinguish it from panic disorder include: very brief episodes (<2 minutes vs panic attacks lasting 5–20 minutes), automatisms (lip smacking, hand movements), déjà vu or jamais vu, olfactory hallucinations (metallic or burning smell), and post-ictal confusion. Any patient with atypical panic attacks (very brief, stereotyped, with neurological features) needs EEG referral before the diagnosis of panic disorder is made.

The stepped care model for anxiety in NICE NG197 is important to understand because it determines the appropriate referral level and avoids over-medicalisation of mild anxiety or under-treatment of severe anxiety. Step 1 (GP recognition and psychoeducation + signposting to self-help resources) is appropriate for subthreshold anxiety (GAD-7 <10) without functional impairment. Step 2 (guided self-help, computerised CBT) is appropriate for mild anxiety (GAD-7 5–9). Step 3 (individual CBT, group CBT, SSRI) is appropriate for moderate anxiety (GAD-7 10–14). Step 4 (CMHT, complex psychological therapy, medication combinations) is for severe or treatment-resistant anxiety. The key principle is not to start at Step 3 for a patient who could be appropriately managed at Step 2 — this uses scarce resources inefficiently and can over-pathologise normal anxiety responses to life stressors. However, patients with moderate-severe anxiety who are clearly not going to benefit from self-help (those with complex trauma, comorbid depression, or significant functional impairment) should be referred directly to Step 3 rather than going through a Step 2 pathway first. The trauma-PTSD IAPT distinction requires emphasis — many GP referrals for "anxiety" reach generic IAPT and are treated with standard GAD/anxiety CBT, which is not effective for PTSD and can be distressing. Specifying "possible PTSD — please assess for trauma-focused pathway" on the IAPT referral is the clinical action that ensures the right treatment is provided.

The transient worsening of anxiety in the first 1–2 weeks after starting an SSRI is a well-documented and clinically important pharmacological effect — it is caused by acute stimulation of 5-HT2C receptors (which mediate anxiogenic effects) before the anxiolytic down-regulation of 5-HT1A receptors (which takes 1–2 weeks) has occurred. Patients who are not warned about this effect frequently stop their SSRI within the first week, concluding it is "making them worse." A brief psychoeducation at prescription ("You may feel slightly more anxious or jittery for the first 1–2 weeks — this is a common and temporary effect, and the medication will start to help your anxiety from week 2–4") dramatically improves adherence. Starting at a lower dose (25 mg sertraline rather than 50 mg) for the first 1–2 weeks can reduce this effect in highly anxious patients. Pregabalin as a Schedule 3 controlled drug for GAD is an important prescribing consideration — the 2019 scheduling of pregabalin and gabapentin as Class C controlled drugs (Schedule 3 in England) followed evidence of significant misuse, particularly in patients with opioid use disorder. GPs must screen for substance misuse and addiction history before prescribing pregabalin — it should not be prescribed to patients with active opioid use disorder or history of pregabalin/gabapentin misuse. The prescription must be on a CD prescription pad in England. The benzodiazepine guidance for anxiety is unambiguous in NICE NG197 — benzodiazepines should not be used for long-term anxiety management due to tolerance, dependence, and the evidence that they prevent the long-term habituation to anxiety that is required for recovery (they block the learning that occurs during exposure-based therapies). Short-term use (maximum 2–4 weeks) at the lowest effective dose may be appropriate while awaiting IAPT or SSRI effect. The prescription should specify a fixed end date, not "PRN" without limit.

The higher SSRI dose requirement in OCD is one of the most clinically important pharmacological principles in anxiety disorder management — OCD requires doses of serotonin reuptake inhibition that are 2–3 times higher than those required for depression. Sertraline 200 mg/day for OCD vs 50–100 mg/day for depression; fluoxetine 60 mg/day for OCD vs 20 mg/day for depression. The reason is thought to be that OCD involves a different neurobiological substrate (orbito-frontal cortex — caudate nucleus circuit dysregulation) that requires more intensive serotonergic modulation. GPs who prescribe sertraline 50 mg for OCD and see no response have under-dosed the medication — the therapeutic dose has not been reached. The panic disorder management principle of avoiding A&E reinforcement is clinically significant — patients who attend A&E during panic attacks receive extensive investigation (ECG, troponin, blood gases, nebulisers) which provides powerful positive reinforcement for the belief that panic attacks are medically dangerous. This worsens the disorder long-term. A GP who explains the panic cycle clearly (sensations → misinterpretation → fear → more sensations → escalating fear) and teaches controlled breathing gives the patient something that 100 A&E visits cannot. The anxiety-menopause connection is frequently missed because GPs do not routinely ask about menstrual cycle changes in women aged 45–55 presenting with new-onset anxiety. Perimenopausal anxiety (new onset, without prior anxiety history, in a woman in the 40s–50s with menstrual irregularity, hot flushes, or sleep disturbance) should be managed with HRT as first-line, not SSRIs, as this addresses the hormonal aetiology rather than the symptom.

The worry postponement technique is one of the most evidence-based and most under-used brief interventions in primary care — it was developed by Borkovec et al. as a CBT component for GAD and has been shown in RCTs to significantly reduce worry frequency and duration within 1–2 weeks of consistent practice. Its mechanism is based on the principle that most GAD worries are not solved by prolonged rumination — they are sustained by the erroneous belief that worrying prevents bad outcomes ("if I stop worrying, something bad will happen"). Scheduled worry time disrupts this belief by demonstrating that the world does not fall apart when worry is deliberately postponed. A GP who spends 5 minutes teaching this technique in a consultation provides something that may be more therapeutically effective than a prescription pad. The nature exposure (green exercise) evidence is worth knowing clinically — a series of studies including the "Blue Mind" research on blue spaces (coastal, riverside) and "green exercise" studies consistently show that exercise in natural settings has significantly greater anxiolytic effect than the same exercise in urban settings, mediated through reduced amygdala activation. This gives GPs a specific recommendation: "Walk in the park for 30 minutes 5 times per week" rather than "do some exercise." The distinction matters for patients who find urban gym environments themselves anxiety-provoking.

The 2-week SSRI review for anxiety carries the same patient safety imperative as for depression — the MHRA warning about increased suicidal ideation in the first weeks of SSRI/SNRI treatment applies equally to anxiety disorders. Patients with anxiety have an independently elevated suicide risk, and the activation effects of early SSRI treatment (increased agitation, impulsivity, energy before anxiolytic effects take hold) can be particularly dangerous in anxiety patients who have never had suicidal ideation before. This is particularly relevant in younger patients (under 30) where MHRA guidance specifically highlights the risk. A 2-week review (phone call acceptable if face-to-face not possible) is the minimum standard and should be documented. The benzodiazepine dependence management pathway deserves specific attention — benzodiazepine dependence is a major iatrogenic problem in UK primary care, with approximately 1.5 million people in the UK estimated to be dependent on prescribed benzodiazepines or Z-drugs. The Heather Ashton Manual (available free online — originally from Newcastle University) is the gold-standard guide for benzodiazepine withdrawal, recommended by Public Health England. The key principles are: switch to diazepam equivalent (longer half-life, smoother withdrawal), then reduce by 10% of current dose every 1–4 weeks (not 10% of original dose). The process takes months for patients who have been on benzodiazepines for years. Community drug and alcohol teams manage this — GPs should not attempt rapid benzodiazepine withdrawal in the community without specialist support, due to seizure risk.