Thematic synthesis — what people with GAD report¶

Themes are reported in aggregate, drawn from published, public sources. Each theme rests on at least two independent sources. Quotations are ≤15 words and attributed to the publication, never to an individual. Where a source covers anxiety disorders as a group rather than GAD specifically, that is stated.

Last curated: 2026-09-02.

T1. Anxiety is experienced first as a physical illness, and that shapes the whole care pathway¶

In a GAD-specific qualitative study centred on values and preferences, most participants reported physical symptoms of anxiety as the presenting experience, across five thematic categories spanning onset, daily life, coping, healthcare demand and treatment decisions (Hurtado 2020, PMID 31908140). The live audit also located GAD-specific qualitative work on primary-care implementation, perinatal behaviours and treatment experiences (PMIDs: 33218311, 35716017, 21644188, 41587136).
In primary-care focus groups, patients described being "often puzzled by their symptoms, not recognising their emotional concerns", and needing the GP to act as detective and validator before any emotional framing was possible (Parker 2020, PMID 32059636).
Quantitatively consistent: painful physical symptoms were present in 59.0% of primary-care GAD patients versus 28.3% of controls, rising to 78.0% with comorbid depression (Romera 2010, PMID 20541811).

Why it matters. The disorder is defined by a cognitive criterion (excessive, uncontrollable worry) but is presented somatically. This is a plausible partial explanation for the 34.4% correct-diagnosis rate in primary care (comorbidity and primary care; Wittchen 2002, PMID 12044105).

T2. Patients are rarely offered a real choice, and pharmacotherapy is the default¶

GAD participants reported that their involvement in decisions was "scarce" and that pharmacological treatment was always the first option offered; the study's stated conclusion is that there is little elicitation of patient preferences across the GAD care pathway (Hurtado 2020, PMID 31908140). Its title takes the patient's framing directly: "I have anxiety, but I have values and preferences".
Treatment preference is strong enough to break trials: a commissioned UK RCT of medication versus psychological therapy for GAD was terminated for recruitment failure, with three-quarters of potentially suitable patients declining, "the great majority" because they did not want to risk being randomised to medication (Kalpakidou 2019, PMID 31126337).
Patients in primary care valued not relying solely on medication, along with acknowledgement of stressful life circumstances and the GP booking follow-up appointments (Parker 2020, PMID 32059636).

Why it matters. Guidelines say patient preference should decide between modalities (guidelines; Bandelow 2014, PMID 25138725). The patient-reported evidence says it does not, and the trial-recruitment evidence quantifies how strong that preference is.

T3. Self-reliance and stigma delay help-seeking, and the pattern is gendered¶

Across 25 studies of men's anxiety, men "commonly reported self-reliance over formal help-seeking", managed symptoms through problem-based coping, and experienced significant social and self-stigma because seeking help transgressed masculinity norms; adherence to emotional restrictiveness was positively associated with anxiety while adherence to toughness norms could be protective (Fisher 2021, PMID 34517242).
In a systematic review of help-seeking among university students with depression or anxiety, the most common barriers were self-reliance, stigma, and poor mental health literacy; good mental health literacy and social encouragement were the significant facilitators (Lui 2024, PMID 36084266).
In perinatal anxiety, disclosure, help-seeking and access to support were "interconnected … and pervaded by stigma" (Oh 2020, PMID 33380483).

Why it matters. The treatment gap is usually described as a supply problem (epidemiology and burden). These sources locate a large part of it on the demand side, in mechanisms that screening programmes do not address.

GAD participants perceived little social support from their environment and reported occupational interference as a defining feature of daily life (Hurtado 2020, PMID 31908140).
Canadian youth living with anxiety described a "fractured sense of self", intense self-scrutiny, lack of self-compassion, a profound sense of responsibility for others at the cost of themselves, and difficulty navigating their social sphere (n=58, ages 10–22; Woodgate 2020, PMID 32004336).
Anxiety UK's public self-description positions peer connection as the core offering — lived-experience advisors, groups and courses — and its published testimony section is organised around no longer feeling alone (Anxiety UK — "Home", https://www.anxietyuk.org.uk/, accessed 2026-09-02).

Why it matters. Trial primary endpoints are symptom scales (HAM-A, GAD-7, PSWQ). Social connection and work functioning are secondary outcomes at best (screening and measurement).

T5. Coming off medication is a distinct, under-supported experience¶

In 20 in-depth interviews with people who had attempted SSRI withdrawal in the previous year, five themes emerged: managing the release from emotional blunting and cognitive suppression; negative impact on close relationships and social interaction; concurrent negative physical symptoms alongside positive health changes (exercise as coping); the importance of GP and family support and of mental-health literacy in others; and the necessity of gradual, flexible tapering with attention to timing (Mahmood 2024, PMID 39102706).
In 483 in-scope free-text survey responses to a James Lind Alliance priority-setting exercise on reducing and stopping psychiatric medication, six themes covered experiences, challenges, strategies, outcomes, emotional context and areas for improvement; the paper's title carries a respondent's paraphrased point that people should be told before prescription how hard stopping is (Boland 2025, PMID 40790920).

Why it matters. The quantitative literature now characterises discontinuation symptoms precisely — ≥1 symptom in 31% versus 17% on placebo, mostly mild and short (Henssler 2024, PMID 38851198; Kalfas 2025, PMID 40632531). The lived-experience literature describes a different object: emotional, cognitive and relational disruption not captured by the DESS. Both are true, and they are not measuring the same thing.

T6. Perinatal anxiety is described as invisible next to perinatal depression¶

Women with perinatal anxiety described barriers to disclosure, difficulty establishing support networks, and an experience "pervaded by stigma"; the authors call for parity with depression to avoid missed opportunities (Oh 2020, PMID 33380483).
A narrative review of postpartum anxiety disorders identified loss, frustration and guilt accompanied by physical tension across studies, with most women having more than one anxiety disorder in addition to postpartum depression — and noted that evidence on clinical and subclinical symptoms and their outcomes is lacking (Ali 2018, PMID 29881312).
In Pakistan, mothers with postnatal anxiety (GAD-7 ≥10) described domestic responsibilities and sociocultural norms of motherhood, and financial constraints, as the perceived drivers, with distinct support needs (Saeed 2024, PMID 38772885).

Why it matters. GAD-specific perinatal prevalence is 4.1% antenatal and 5.7% postnatal (Dennis 2017, PMID 28302701). WPSI found no treatment trials through its 2020 review (Nelson 2020, PMID 32510989); a 2024 systems-level cluster-RCT secondary analysis subsequently measured perinatal GAD symptoms, though not GAD-specific pharmacotherapy (Zimmermann 2024, PMID 38992743) — see special populations.

T7. What patients value in a clinician is continuity and validation, not throughput¶

Patients described the relationship itself as therapeutic — "doctor as drug" — with continuity particularly valued, and wanted a collaborative partnership while often having to relinquish or over-take involvement depending on illness severity and perceived clinician capacity (Parker 2020, PMID 32059636).
The GAD-specific study reaches a compatible conclusion from the opposite direction: absent elicitation of values and preferences produces a system biased toward prescribing and away from the patient's own account of coping (Hurtado 2020, PMID 31908140).

Why it matters. The best-evidenced service intervention in GAD — collaborative care, NNT ~5 for response — is precisely a continuity-and-follow-up intervention (Roy-Byrne 2010, PMID 20483968); the patient-reported literature independently identifies the same ingredients.

Where patient priorities and the research agenda diverge¶

Patients emphasise	The literature measures
Physical symptoms as the presenting problem (T1)	Worry as the diagnostic criterion
Choice between modalities (T2)	Comparative efficacy in randomised samples that exclude preference
Stigma and self-reliance as barriers (T3)	Instrument accuracy and screening coverage
Social connection and work (T4)	HAM-A, GAD-7 and PSWQ scores
The experience of stopping medication (T5)	Discontinuation-symptom counts
Perinatal anxiety as a distinct entity (T6)	Perinatal depression
Continuity and validation (T7)	Session counts and protocol fidelity

Every row is a research question (OPEN-QUESTIONS.md).

Sources cited in this file¶

Full annotations in sources.md.