What shifts the oxygen dissociation curve to the left?
Physiology — Oxygen Carriage, the Dissociation Curve and Tissue Delivery
The question
A patient is rewarming on the intensive care unit after prolonged surgery with massive haemorrhage. She has received twelve units of red cells stored for three weeks, her core temperature is 33 degrees Celsius, and mechanical ventilation has produced a respiratory alkalosis with a pH of 7.52. Her haemoglobin concentration and arterial oxygen saturation are both now normal. What is the most likely effect of this combination on oxygen carriage?
- The P50 falls and haemoglobin releases less oxygen at any given tissue oxygen tension
- The P50 rises and oxygen offloading to the tissues is enhanced
- Arterial oxygen content falls substantially because less oxygen binds to haemoglobin
- The Bohr effect is exaggerated, so more oxygen is released in metabolically active tissue
- The pulse oximeter reading will fall, revealing the change in affinity
The answer
The P50 falls and haemoglobin releases less oxygen at any given tissue oxygen tension
The P50 falls and haemoglobin releases less oxygen at any given tissue oxygen tension — Correct. Alkalosis, hypothermia, and 2,3-DPG depletion from stored blood all increase haemoglobin oxygen affinity. P50 and affinity move in opposite directions: a left-shifted curve has a lower P50 and releases oxygen less readily at the tissues.
The P50 rises and oxygen offloading to the tissues is enhanced — This describes a right shift. Acidosis, hypercapnia, fever, and raised 2,3-DPG all shift the curve to the right and facilitate unloading. Every factor in this stem acts in the opposite direction.
Arterial oxygen content falls substantially because less oxygen binds to haemoglobin — A left shift increases binding in the lung, not decreases it. With a normal haemoglobin and saturation, the arterial oxygen content is essentially unchanged. The problem in this scenario is unloading at the tissues, not loading at the alveolus.
The Bohr effect is exaggerated, so more oxygen is released in metabolically active tissue — The Bohr effect is the rightward shift that occurs locally when CO2 rises and pH falls in metabolically active tissue. A systemic alkalosis opposes it. In this patient, the Bohr response is blunted, not exaggerated.
The pulse oximeter reading will fall, revealing the change in affinity — Saturation is read from the flat upper portion of the curve, where a shift barely alters the value. A left shift may even nudge the saturation slightly higher. This abnormality is invisible to the pulse oximeter, which is precisely what makes it clinically dangerous.
Explanation
The oxyhaemoglobin dissociation curve plots saturation against oxygen tension, and its sigmoid shape arises from cooperative binding. As each of the four haem groups picks up oxygen, the molecule transitions from the tense to the relaxed conformation, and the remaining binding sites become progressively more avid. The P50 (the tension at which haemoglobin is half saturated, normally about 3.5 kPa or 26-27 mmHg) indicates where the curve sits. P50 and affinity are reciprocals: high affinity means low P50, which means the curve is shifted to the left.
A right shift facilitates tissue unloading. It is caused by acidosis, hypercapnia, fever, and raised 2,3-DPG (which binds the beta chains of deoxyhaemoglobin and stabilises the tense form). Chronic anaemia, chronic hypoxia, and altitude acclimatisation all raise 2,3-DPG for exactly this reason. Sickle haemoglobin is also right-shifted.
A left shift produces the opposite effect: haemoglobin holds on more tightly and unloading suffers. The causes are the mirror image: alkalosis, hypocapnia, hypothermia, and low 2,3-DPG. Then there are the abnormal haemoglobins, where the clinically important examples reside. Fetal haemoglobin binds 2,3-DPG poorly and is left-shifted, which suits placental oxygen pickup. Carbon monoxide produces two effects simultaneously: it occupies binding sites (reducing capacity) and it shifts the remaining sites to the left (impairing release from haemoglobin that is still functional). That double mechanism is why CO poisoning is worse than the equivalent degree of anaemia. Methaemoglobin similarly left-shifts the unaffected haem groups.
Stored blood is the factor candidates most often overlook. Red cells lose their 2,3-DPG during storage, so freshly transfused stored blood carries oxygen adequately but releases it poorly. The 2,3-DPG regenerates over a day or two. Combining depleted stored blood with hypothermia and an iatrogenic respiratory alkalosis (as in this stem) produces a patient with a reassuring haemoglobin, a reassuring saturation, and a reassuring oxygen content, while tissue oxygen delivery is impaired. Pursuing a low PaCO2 and tolerating hypothermia in a massively transfused patient works against effective tissue oxygenation. The oxygen cascade does not end at the arterial blood gas.
Tips for the candidate
Establish the direction first. The curve plots saturation upwards against oxygen tension along the horizontal axis. A LEFT shift means the same saturation is reached at a LOWER tension: tighter binding, higher affinity, lower P50. Candidates who attempt to derive this under pressure often invert it. Committing "left = lower P50 = holds on tighter" to memory is more reliable than reasoning it through each time.
Both lists are best constructed as mirror images rather than independently. Everything that occurs in hard-working, hypoxic, acidotic, hot tissue shifts the curve to the right and facilitates unloading. That grouping has a physiological logic that can be reconstructed in the examination. The left-shift list is simply the negation, plus three special cases that are not merely the opposite of anything: fetal haemoglobin, carbon monoxide and methaemoglobin, and stored blood via 2,3-DPG depletion.
Stems that stack multiple factors and expect the candidate to determine the net effect are common. Each clause should be tagged as left or right, and the net direction taken. Where factors genuinely conflict, the question is usually asking the candidate to identify the dominant one. Three factors all pointing the same way (as in this stem) is a clear signal, not a conflict.
Any option that confuses position with capacity can be discarded on principle. A shift changes how readily haemoglobin releases oxygen. It does not meaningfully alter how much oxygen the arterial blood contains, because saturation is determined on the flat upper portion of the curve. An option claiming that a left shift causes a substantial fall in arterial content, or that the pulse oximeter will detect it, can be eliminated on that basis.