Q1. Which fasting plasma glucose is diagnostic of diabetes? A. ≥ 5.6 mmol/L B. ≥ 6.1 mmol/L C. ≥ 7.0 mmol/L D. ≥ 11.1 mmol/L ANSWER: C — ≥ 7.0 mmol/L. Why: 6.1–6.9 is impaired fasting glucose (prediabetes). 11.1 is the threshold for a RANDOM or 2-hour value, not a fasting one. LEARNING POINT: keep the fasting and random thresholds separate — mixing them is a common slip.

Q2. Why is type 2 diabetes ketosis-resistant? A. Type 2 patients cannot produce ketones B. Residual insulin suppresses lipolysis C. Glucose is not raised in type 2 D. Type 2 patients have more glucagon ANSWER: B — residual insulin suppresses lipolysis. Why: it takes very little insulin to inhibit fat breakdown. Type 2 patients retain enough for that even when glucose control is poor, so they develop HHS rather than DKA. LEARNING POINT: this is the mechanistic reason behind the type 1 / type 2 emergency split.

Q3. An asymptomatic 55-year-old has a screening fasting glucose of 7.4 mmol/L. The correct next step is: A. Diagnose diabetes and start metformin B. Repeat the test on a separate day to confirm C. Reassure — this is normal D. Start insulin ANSWER: B — repeat on a separate day. Why: in an ASYMPTOMATIC patient a single abnormal result is insufficient; it may reflect intercurrent illness or an assay issue. With classic symptoms, one value would suffice. A: premature without confirmation. C: 7.4 is above the diagnostic threshold, not normal. LEARNING POINT: the presence or absence of symptoms changes how many abnormal results you need.

Q4. A 63-year-old with type 2 diabetes has HbA1c 52 mmol/mol on metformin, BP 158/96, cholesterol 6.2, no statin. Which intervention most reduces mortality? A. Add a second glucose-lowering agent B. Intensify to insulin C. Treat blood pressure and start a statin D. Increase the metformin dose ANSWER: C — blood pressure and statin. Why: most people with type 2 diabetes die of cardiovascular disease. With a reasonable HbA1c already, treating untreated hypertension and dyslipidaemia has far more impact than further glucose lowering. LEARNING POINT: diabetes management is cardiovascular risk management. Do not be led by the glucose number alone.

Q5. A 60-year-old on metformin and dapagliflozin presents with vomiting and abdominal pain. Glucose 10.2 mmol/L, and she is breathing deeply. The most important next investigation is: A. Nothing — the glucose excludes DKA B. Blood ketones and a blood gas C. Abdominal ultrasound only D. Repeat glucose in six hours ANSWER: B — ketones and a gas. Why: SGLT2 inhibitors excrete glucose renally, so ketoacidosis can occur at a normal or near-normal glucose — euglycaemic DKA. Deep breathing suggests respiratory compensation for a metabolic acidosis. A: the trap this question is built on. LEARNING POINT: on an SGLT2 inhibitor, check ketones whatever the glucose reads.

Q6. A patient with type 1 diabetes has had several severe hypoglycaemic episodes with no warning symptoms. The most appropriate response is: A. Tighten glycaemic targets further B. Relax glycaemic targets to allow awareness to recover C. Stop insulin entirely D. Add a sulfonylurea ANSWER: B — relax the targets. Why: repeated hypoglycaemia blunts the autonomic warning response. Deliberately avoiding hypoglycaemia for a period allows awareness to return, which is safer than the tighter control that caused it. A: worsens the problem. C: dangerous — type 1 requires insulin. D: sulfonylureas are for type 2 and cause more hypos. LEARNING POINT: hypoglycaemia unawareness is a reason to loosen control, not tighten it.

Q7. A patient with sickle cell disease has an HbA1c of 41 mmol/mol but fasting glucose repeatedly 8.9 mmol/L. The best interpretation is: A. The patient does not have diabetes B. HbA1c is unreliable here; diagnose on glucose criteria C. The glucose results are erroneous D. Repeat the HbA1c in three months ANSWER: B — HbA1c is unreliable in haemoglobinopathy. Why: HbA1c depends on red cell lifespan and on normal haemoglobin. In sickle cell disease and other haemoglobinopathies, and in haemolysis, it can read falsely low. Glucose-based criteria should be used instead. LEARNING POINT: know when NOT to trust HbA1c — anaemia, haemoglobinopathy, pregnancy, recent transfusion, CKD.