Skip to content

Anorexia nervosa — master index

Last curated: 2026-09-02 · status: built (Codex), independently audited (Claude), deepened (Claude), then independently re-audited (Codex) · 18/18 wiki pages written · 17 pages curated; guidelines.md remains draft because its carried-forward APA recommendation text could not be retrieved · 282 unique PMIDs cited condition-wide (113 before the depth pass; 278 at re-audit intake) · 46 ClinicalTrials.gov records and the live 337-record interventional portfolio verified through API queries · 4 landmark notes

The condition in five sentences. AN is a restrictive eating disorder in which low energy intake, weight-gain fear or persistent weight-preventing behaviour, and disturbed weight/shape experience interact with starvation-mediated physiology. Mortality is markedly elevated: a 36-study meta-analysis estimated 5.1 deaths per 1,000 person-years and an SMR of 5.86 with suicide accounting for one fifth of deaths, and two 2026 syntheses on far larger samples report an all-cause risk ratio of 5.52 (95% CI 4.47–6.82) and a pooled SMR of 5.06 (3.47–7.38), although their overlapping cohorts and heterogeneity do not establish a secular trend (Arcelus 2011, PMID 21727255; Semchishen 2026, PMID 41536100; Lai 2026, PMID 41277145). Genetic evidence supports a metabo-psychiatric research model: a 16,992-case GWAS identified eight loci and psychiatric plus BMI-independent metabolic correlations (Watson 2019, PMID 31308545). Family-based approaches have the clearest adolescent psychotherapy evidence, whereas adult CBT-E, MANTRA, SSCM and focal psychodynamic trials do not establish a consistent winner — the adult network meta-analysis screened 14,003 reports to find 16 trials and rated its own evidence low to very low (Lock 2010, PMID 20921118; Solmi 2021, PMID 33600749). Atypical AN is in scope because comparable psychopathology and clinically important complications can occur without low BMI, exposing weight-based diagnostic bias (Walsh 2023, PMID 36508318).

Scope

This condition covers anorexia nervosa, including atypical AN. Bulimia nervosa, binge-eating disorder and ARFID appear only as differential diagnoses. The ethics page treats capacity, compulsory care and the contested terminal-AN proposal as unresolved content and sources both sides.

Reading paths

Pages

# Section Page Scope Status
1 Foundations overview.md Map, burden, treatment frame curated
2 Foundations diagnosis-and-classification.md DSM/ICD framing, atypical AN, differential curated
3 Foundations epidemiology-and-incidence.md Frequency, age, sex/gender, ascertainment curated
4 Outcome mortality-and-long-term-outcome.md Mortality, suicide, recovery, relapse curated
5 Mechanism genetics.md Heritability, GWAS, metabolic correlations curated
6 Mechanism neurobiology-and-cognition.md Reward, habit, cognition, state confounding curated
7 Clinical medical-complications.md Cardiac, bone, endocrine, renal, GI curated
8 Clinical refeeding-and-nutritional-rehabilitation.md Refeeding risk and protocols curated
9 Treatment treatment-in-adolescents.md FBT and parent-focused variants curated
10 Treatment treatment-in-adults.md CBT-E, MANTRA, SSCM, FPT curated
11 Treatment pharmacotherapy.md Olanzapine and adjunctive evidence curated
12 Treatment service-models-and-setting.md Outpatient, day, inpatient curated
13 Frontier neuromodulation-and-experimental-therapy.md DBS, rTMS, psychedelics curated
14 Ethics severe-enduring-illness-and-compulsory-treatment.md Capacity, coercion, terminal-AN dispute curated
15 Human patient-experience-and-advocacy.md Lived experience, carers, organizations curated
16 Reference guidelines.md Guideline synthesis draft — APA 2023 text unretrievable
17 Frontier clinical-trials-landscape.md Live registrations and design gaps curated
18 Safety red-flags-and-safety-concerns.md Instability, refeeding and suicide risk curated

Literature layer

Artifact Contents Status
BIBLIOGRAPHY.md 282 papers, topic-grouped, with page map; non-primary publication types (comments, letters, editorials, protocols) labelled; depth-pass additions marked independently re-audited
notes/ Four landmark paper notes, each extended at audit with the replication or follow-up evidence published since audited
guideline registry 15 documents including NICE, APA, RANZCP, Canada (×2), German S3, ANZAED (×3), USPSTF, ASPEN and the AED Nine Truths; disagreements and watch list; NICE and RANZCP text read directly, APA text still not retrievable deepened, one gap
statistics Source-specific quantitative tables across prevalence/incidence, mortality and outcome, heritability and comorbidity, complications at extreme malnutrition, pregnancy, bone/fracture, treatment, economics and the trial portfolio independently re-audited
patient voice Method, organizations, themes and sources audited

Evidence frontier

OPEN-QUESTIONS.md contains 25 stable questions and a fourteen-row “Dots not yet connected” table. The audit re-tested every asserted absence by live search; two were stale and were rewritten as answers (one-year refeeding outcomes, and the empirical status of the “terminal AN” proposal), and two questions were added (OQ-16 weight-indexed refeeding, OQ-17 the course of atypical AN). The depth pass opened eight further questions (OQ-18 to OQ-25) and four junctions (D11–D14), and corrected one stale absence: a small four-year comparison of involuntary and voluntary treatment pathways does exist (Abry 2024, PMID 37690079) and finds no quality-of-life or mortality difference, against a national register showing elevated post-compulsion mortality. Highest-leverage junctions are metabolic-genetic translation, developmental transport of family-treatment mechanisms, registry-linked mortality outcomes, the entirely missing longitudinal evidence on atypical AN, weight-indexed refeeding, empirical capacity/compulsion outcomes, microbiome causality, and the 80% of registered AN trials that never reach publication.

Depth-pass and re-audit standing (2026-09-02)

Measure After build + first audit At depth-pass submission After independent re-audit
Distinct PMIDs condition-wide 113 278 282
Distinct records per page (condition-wide records ÷ 18) 6.3 15.4 15.7
Total wiki lines 1,545 2,443 2,451
Lines per page (mean) 86 136 136
Bibliography PMID records 113 278 282
Guideline registry documents 7 15 15
Open questions / junctions 17 / 10 25 / 14 25 / 14
Pages at curated 17 0 17

The depth pass substantially widened the evidence base rather than merely increasing line count: its submitted condition-wide figure rose from 113 to 278 distinct PMIDs. The re-audit found 167—not the logged 165—bibliography records marked as depth-pass additions, retrieved every one live, removed one irrelevant padded record, and added six records needed to correct stale claims and absences. Details are in LOG.md.

Audit standing (2026-09-02)

Check Result
Depth-pass PMIDs 167 at intake, all resolved live via PubMed E-utilities; 166 retained after removing one irrelevant citation
Unique PMIDs after corrections 282, exactly matched between page citations and the bibliography
ClinicalTrials.gov records 46 distinct NCT IDs, all resolved live with matching titles/status/enrollment; portfolio count re-run live
Newly asserted absences 21 targeted PubMed searches re-run; five stale or materially incomplete claims rewritten
Padding One irrelevant citation removed; one reference orphaned by an audit rewrite removed
Broken relative links 0
Pages promoted to curated 17 of 18
Pages held at draft guidelines.md — APA 2023 recommendation text could not be retrieved for verification