The MARSIPAN criteria (Management of Really Sick Patients with Anorexia Nervosa) define when hospital admission is required in anorexia nervosa — the key physical parameters that mandate same-day medical admission are: BMI <13 kg/m², heart rate <40 bpm, blood pressure <80/50 mmHg, QTc >450 ms on ECG, temperature <35.5°C, electrolyte abnormalities (K+ <3.0 mmol/L, PO4 <0.5 mmol/L, Mg2+ <0.5 mmol/L), syncope, or muscle weakness preventing standing from sitting. GPs who see patients with anorexia nervosa must perform or arrange an urgent physical assessment including ECG, electrolytes, glucose, and lying/standing blood pressure. The cardiac risk in severe anorexia nervosa is substantial — bradyarrhythmias (from vagal overactivity + electrolyte imbalances) and QTc prolongation (from hypokalaemia, hypomagnesaemia) can cause sudden cardiac death, particularly during refeeding. The NICE CG9 guidance on eating disorders and MARSIPAN guidelines provide the framework. Refeeding syndrome is the most important acute complication of treating severe malnutrition and anorexia — the pathophysiology involves: during prolonged starvation, the body shifts to fat/protein catabolism. Intracellular electrolytes (phosphate, potassium, magnesium) remain relatively preserved despite depletion of total body stores. When carbohydrates are reintroduced, insulin secretion drives glucose, potassium, phosphate, and magnesium back into cells simultaneously, causing precipitous plasma falls. Hypophosphataemia (<0.3 mmol/L) is the critical trigger for cardiac failure, respiratory muscle weakness, and rhabdomyolysis. The NICE CG32 refeeding protocol specifies: start at 5 kcal/kg/day maximum (not normal intake), increase slowly over 1 week, give IV Pabrinex (thiamine + B vitamins) before any glucose, monitor electrolytes daily for 1 week.

The paraneoplastic anorexia-cachexia syndrome (CACS) deserves emphasis as a clinically important and pathophysiologically distinct entity from simple anorexia from reduced intake — in CACS, the weight loss is disproportionate to the degree of anorexia because of active catabolism of muscle protein (proteolysis-inducing factor, IL-6, TNF-α drive muscle wasting independent of caloric intake). This means that simply increasing caloric intake does not reverse the muscle wasting in CACS — the patient continues to lose lean body mass even with supplemental nutrition. CACS affects approximately 50–80% of cancer patients at diagnosis and is a major negative prognostic indicator (cancer patients with cachexia die earlier than matched patients without cachexia, independent of tumour stage). Understanding CACS changes the management approach: the goal is not to 'feed them more' but to address symptoms (nausea, early satiety, dysphagia, pain), optimise nutritional intake within the physiological constraints, and focus on quality of life. Megestrol acetate and corticosteroids improve appetite temporarily but do not improve survival or lean mass. This is covered in the palliative care algorithm. The GLP-1 agonist (semaglutide, liraglutide) mechanism of appetite suppression is now clinically important because these drugs are widely prescribed for type 2 diabetes and obesity. GLP-1 agonists suppress appetite via: (1) direct hypothalamic GLP-1 receptor activation reducing hunger signals; (2) delayed gastric emptying causing early satiety; (3) inhibiting the ghrelin (hunger hormone) system. The profound appetite suppression they cause (contributing to 10–15% body weight loss) is the therapeutic mechanism, but in some patients — particularly elderly patients, those with low baseline BMI, or those with concurrent illness — this appetite suppression can become a genuine clinical problem causing malnutrition. GPs must document weight at every diabetes review for patients on GLP-1 agonists and reduce the dose or stop the drug if unintended weight loss occurs.

The MUST (Malnutrition Universal Screening Tool) score is the recommended nutritional screening tool for primary care and hospital outpatients — it takes 3 minutes to calculate and provides a standardised risk score that determines the urgency of dietitian referral and intervention. MUST score ≥2 = high malnutrition risk = dietitian referral within 1 week. The three components are: (1) BMI score: BMI >20 = 0 points; 18.5–20 = 1 point; <18.5 = 2 points. (2) Unintentional weight loss score: <5% = 0 points; 5–10% = 1 point; >10% = 2 points. (3) Acute illness score: if acutely ill and no nutritional intake for >5 days = 2 points. Total MUST score: 0 = low risk (routine care); 1 = medium risk (dietary advice, monitor); ≥2 = high risk (intensive dietetic input). The SCOFF questionnaire is the validated screening tool for eating disorders in primary care — it has sensitivity of 78% and specificity of 94% for anorexia nervosa: (1) Do you make yourself Sick because you feel uncomfortably full? (2) Do you worry you have lost Control over how much you eat? (3) Have you recently lost more than One stone in a 3-month period? (4) Do you believe yourself to be Fat when others say you are too thin? (5) Would you say that Food dominates your life? A score of ≥2 is a positive screen. The SCOFF should be administered sensitively, with a non-judgmental framing — patients with eating disorders frequently deny or minimise symptoms, particularly weight-related questions.

The succession splash test for gastroparesis is a rarely performed but diagnostically useful bedside test — the patient fasts for 4 hours, the examiner places a stethoscope over the epigastrium and rapidly shakes the patient's torso. A splashing sound heard more than 4 hours after the last meal indicates retained gastric fluid (gastroparesis, pyloric stenosis, gastric outlet obstruction). It is quick, non-invasive, and can guide whether urgent upper GI imaging is needed. The CT chest/abdomen/pelvis investigation threshold for unexplained significant weight loss is lower than many GPs realise — NICE NG12 and the 2017 Suspected Cancer guidance support arranging urgent CT when a patient over 40 has unexplained weight loss of >5% over 3 months with any of: anaemia, elevated ESR/CRP, elevated LFTs, hypercalcaemia, or abnormal tumour markers. The CT is the single most yield-per-test investigation in this clinical scenario because it can identify gastric, pancreatic, hepatic, renal, lymph node, and lung malignancies in a single study. The argument for early CT in unexplained weight loss (rather than sequential organ-specific investigations over months) is that it often shortens the diagnostic pathway from several months to days and allows earlier treatment initiation, which directly affects survival in most solid organ cancers.

The NICE CG9 guidance on eating disorders contains a specific instruction that GPs must not delay referral to eating disorders services while waiting for the patient to become 'ready' — this reflects the evidence that early intervention in anorexia nervosa produces significantly better outcomes (higher rates of full recovery, lower mortality) than delayed treatment. The mortality of anorexia nervosa is approximately 5% at 10 years (all-cause mortality) — higher than any other psychiatric condition — and the risk of death is highest in the first years after diagnosis. Every month of delay in accessing specialist eating disorder treatment represents increased risk. GPs should make the CEDS referral immediately on identification of a likely eating disorder, while continuing to provide primary care support, regular physical monitoring, and engagement. In the under-18s, the NHS eating disorders waiting time standard (2019) specifies that young people with suspected eating disorders should be seen within 1 week for urgent cases and 4 weeks for routine referrals — GPs should specify 'urgent' on the referral if there is any physical compromise or rapid weight loss. ARFID (Avoidant/Restrictive Food Intake Disorder) is a DSM-5 diagnosis that describes food refusal based on sensory characteristics, fear of aversive consequences (choking, vomiting), or apparent lack of interest in food — without the body image disturbance of anorexia nervosa. It is increasingly recognised in adults (not just children) and is seen more frequently in patients with autism spectrum disorder, anxiety disorders, and OCD. GPs who encounter patients with longstanding selective eating patterns causing nutritional deficiency or weight loss should consider ARFID and refer to the appropriate local service (eating disorders team or psychology).

Mirtazapine is a particularly useful antidepressant for the combination of depression + anorexia + insomnia that frequently presents in elderly patients or those with chronic illness — its mechanism (noradrenergic and specific serotonergic antidepressant, NASSA) includes potent H1 histamine receptor antagonism at 15 mg, which increases appetite and promotes weight gain. At higher doses (30–45 mg), the noradrenergic effects are more prominent and the appetite-stimulating H1 effect is relatively less pronounced — so the counterintuitive principle applies: start at 15 mg (not the usual 'start low' approach for most antidepressants) if appetite stimulation is a primary treatment goal, then increase if antidepressant effect is insufficient. Clinical trials in cancer cachexia and geriatric anorexia show weight gain of 1–3 kg over 4–8 weeks with mirtazapine. The olanzapine low-dose strategy for refractory anorexia nervosa is increasingly supported by evidence — olanzapine at 2.5–5 mg OD produces meaningful weight gain through appetite stimulation and anxiolytic effects, and also reduces the obsessional food-related thoughts that are a central feature of AN. NICE CG9 (2017) does not yet strongly recommend olanzapine as routine for AN, but the 2022 updated evidence review acknowledges its utility in patients with severe, refractory AN where other treatments have failed. It is initiated by the eating disorders team, not in primary care, but GPs should be aware of this option when discussing treatment options with patients.

The prescribing of oral nutritional supplements (ONS) on NHS prescription is an important and underused intervention in primary care — the ACBS (Advisory Committee on Borderline Substances) specifies the conditions in which ONS can be prescribed on FP10: disease-related malnutrition, pre-operative preparation of malnourished patients, dysphagia, proven inflammatory bowel disease, following total gastrectomy, short bowel syndrome, and intractable malabsorption (specifically, not for general 'feeling unwell' or 'poor appetite' without nutritional compromise). For GP prescribers, the principle is: MUST ≥2 + any medical condition causing the malnutrition = prescribable. The ONS must be cost-effectively selected — Fresubin 2 kcal (400 kcal/200 ml, prescribable) is more energy-dense than Fortisip (300 kcal/200 ml) and may be appropriate for patients needing maximum energy in minimum volume. Dietitian guidance on ONS selection and monitoring is best practice. Ondansetron-induced constipation is a very common and practically important side effect — ondansetron (a 5-HT3 receptor antagonist) dramatically reduces bowel motility. In already anorexic patients, constipation further reduces appetite (abdominal fullness, bloating). Co-prescribing a stimulant laxative (senna 15 mg nocte or bisacodyl 10 mg OD) alongside ondansetron is recommended in all patients except those with diarrhoea.

The social eating evidence is strong and clinically relevant — studies consistently show that elderly people living alone eat significantly less (20–30% fewer calories) than those living with others, with social mealtimes producing automatic appetite improvement independent of the food available. The mechanism involves social facilitation of eating behaviour (mirroring, reduced self-monitoring, improved mood) and the structure and ritual of a shared meal. Community interventions (lunch clubs, Meals on Wheels, befriending services) produce measurable improvements in nutritional intake and quality of life in elderly isolated patients. GPs who identify social isolation as a contributing factor in elderly anorexia should actively refer to social services and community befriending services — these are evidence-based nutritional interventions even though they are social rather than medical. The ginger antiemetic evidence base is worth knowing — a Cochrane review of ginger preparations for nausea and vomiting (including chemotherapy-induced nausea) found Level 1b evidence for efficacy superior to placebo. The mechanism involves 5-HT3 receptor antagonism (similar to ondansetron) and cholinergic effects on the gastric myenteric plexus. Ginger 250 mg capsules QDS is the studied dose. As a safe, inexpensive, widely available supplement with evidence for efficacy, it is a reasonable first-line recommendation for mild-moderate nausea contributing to anorexia in primary care, before escalating to prescription antiemetics.

The 6-week maximum watchful waiting principle for unexplained weight loss is a useful clinical governance standard — unexplained significant weight loss (>5% over 3 months) in an adult over 40 should not be managed with 'watch and wait' for more than 6 weeks without a clear diagnosis. This is because the most common serious cause — upper GI malignancy — is a rapidly progressive disease where weeks matter for treatment outcome. If first-line investigations are reassuring but the weight loss continues, a CT chest/abdomen/pelvis is the next step — its yield in unexplained weight loss + normal first-line investigations is approximately 30–40% (revealing an occult malignancy in 1 in 3 patients). The family counselling around reduced oral intake in dying patients is one of the most important and emotionally charged conversations in palliative primary care — many families interpret reduced eating in a dying patient as 'giving up' or 'not being fed properly' and request interventional feeding (NG tube, PEG, TPN). Providing accurate, compassionate information that reduced oral intake is a normal part of the dying process (reduced metabolic demand, reduced gut absorption, central appetite suppression from disease) and that forced feeding in the dying causes discomfort without benefit or survival extension is a core GP communication skill. The Palliative Care Pain algorithm addresses this context further. NICE NG31 (Care of Dying Adults) specifically states that artificial nutrition and hydration should not be started routinely in the last days of life unless there is a specific clinical indication and it is consistent with the patient's wishes.