Does smoking cause polycythemia? Yes, smoking is a well-established cause of secondary polycythemia, raising hematocrit through chronic carbon monoxide exposure and tissue hypoxia, but the diagnosis is often missed or misattributed. The rise is usually mild to moderate and hard to spot. A hemoglobin taken at face value can hide the real cause.
Polycythemia means too many red blood cells. It is not one disease but a group of conditions united by elevated hemoglobin, hematocrit, or red cell mass. Smoking sits firmly in the acquired, secondary category. That means the excess red cells are a response to something outside the marrow, not a primary marrow disorder. This distinction drives every downstream decision. The cause of polycythemia determines the correct treatment.
The carbon monoxide mechanism
Cigarette smoke contains carbon monoxide at concentrations of roughly 10,000 to 40,000 parts per million in the inhaled smoke stream. What matters is the delivered dose. Chronic smokers commonly carry carboxyhemoglobin levels of 3% to 10%, and heavy smokers can exceed 10%. Carboxyhemoglobin is hemoglobin bound to carbon monoxide instead of oxygen. Its binding affinity for hemoglobin is about 200 to 250 times that of oxygen.
The consequence is a functional anemia. Each carboxyhemoglobin molecule is unavailable to carry oxygen. It also shifts the oxygen-hemoglobin dissociation curve to the left. When the curve shifts left, hemoglobin holds oxygen more tightly and releases less to tissues. So the oxygen is not just carried away from the bloodstream; it is also delivered less readily. Tissue oxygenation falls even when the arterial oxygen tension looks normal.
The body senses this as hypoxia at the tissue level. The kidney responds by increasing erythropoietin, the hormone that tells the bone marrow to make red cells. Erythropoietin is produced by peritubular interstitial cells in the kidney in response to cellular oxygen sensing. The sensor is a family of prolyl hydroxylase enzymes that degrade hypoxia-inducible factor when oxygen is plentiful. When oxygen is scarce, hypoxia-inducible factor survives, erythropoietin transcription rises, and red cell production increases.
The net effect is a compensatory erythrocytosis. The plasma volume often falls at the same time because smoking causes hemoconcentration, so the rise in measured hemoglobin overstates the rise in total red cell mass. Plasma volume contraction is common in smokers and can push hematocrit upward even before red cell production increases. Both mechanisms work together.
Smoking also triggers chronic airway disease and sleep-disordered breathing in some people. That combination adds true hypoxemia to the carbon monoxide effect. Severe chronic obstructive pulmonary disease can produce hypoxemic polycythemia independent of carbon monoxide. The two mechanisms frequently coexist in the same patient. This is why smoking-related polycythemia is usually multifactorial rather than a single pathway.
Carboxyhemoglobin falls with a half-life of about 4 to 6 hours on room air. It falls faster with supplemental oxygen, roughly 60 to 90 minutes. That kinetics matters for testing: a carboxyhemoglobin level drawn many hours after the last cigarette can underestimate exposure. Timing of the blood draw affects interpretation.
How large is the effect
Population-level estimates show the direction and size of the smoking effect. Large cohort and cross-sectional studies report mean hemoglobin and hematocrit values that are modestly higher in smokers than in nonsmokers. Typical differences are on the order of a few tenths of a gram per deciliter for hemoglobin and 1% to 3% for hematocrit. The effect is dose-dependent: more cigarettes per day and longer smoking duration produce larger differences. Heavy smokers can show hematocrit values well above the reference range.
The prevalence of erythrocytosis among smokers depends on the definition. When erythrocytosis is defined as hematocrit above 51% in men or above 48% in women, large referral series consistently identify smoking as one of the most common causes of apparent secondary polycythemia. In patients evaluated for elevated hematocrit, smoking is often the leading identifiable secondary cause after excluding chronic lung disease and altitude. The exact proportion varies with the population and the threshold used.
The hematocrit threshold itself is not uniform. The World Health Organization defines elevated hematocrit as above 49% in men and above 48% in women, but that reference frame originates from a 1960s population and may not match contemporary smokers. Many laboratories and clinical guidelines use different cutoffs for investigation. For men, common thresholds for further workup range from 52% to 54%. For women, common thresholds range from 48% to 51%. Some hematology societies have moved toward hemoglobin-based rather than hematocrit-based definitions to reduce hydration-related variability.
The discrepancy between these thresholds is not trivial. A man with a hematocrit of 51% is above the WHO definition but below many workup thresholds. A woman at 50% is above the WHO definition but below some guidance. Which threshold applies depends on the guideline being followed and the clinical context. This is why interpretation should combine the absolute value, the trend over time, the smoking history, and the presence of symptoms or complications.
The functional impact is not simply proportional to the number. A hematocrit of 54% raises blood viscosity substantially, and viscosity rises steeply as hematocrit climbs above 50%. The relationship between viscosity and hematocrit is nonlinear. Small further increases at high hematocrit produce disproportionately large viscosity changes. That is one reason clinicians care about thresholds.
The counterintuitive finding
The detail that surprises many clinicians is that smoking-related erythrocytosis can reduce rather than increase measured oxygen saturation on pulse oximetry. Pulse oximeters cannot reliably distinguish carboxyhemoglobin from oxyhemoglobin. In some devices, carbon monoxide bound to hemoglobin is read as oxygenated hemoglobin, so the oximeter reads falsely high. In other circumstances, especially with concurrent anemia, the reading is low. Caregivers often cite falsely reassuring or falsely alarming numbers without knowing which direction the error runs.
The safer conclusion is that pulse oximetry cannot be trusted to assess oxygenation in smokers. A normal saturation reading does not exclude significant carboxyhemoglobin. A carboxyhemoglobin level, or co-oximetry, is the appropriate test when carbon monoxide exposure is suspected. Co-oximetry measures the percentage of hemoglobin bound to carbon monoxide directly rather than relying on the two-wavelength approximation used by standard pulse oximeters. This is a professional caveat worth repeating: a reassuring pulse oximeter in a heavy smoker is not evidence against carbon monoxide burden.
When polycythemia is not from smoking
A hematocrit just above the reference range is common and often benign, but some elevated values signal serious disease. The diagnostic task is to separate smoking-related erythrocytosis from conditions that need specific treatment. The key test is serum erythropoietin. A low erythropoietin level in a patient with elevated red cell mass points toward a primary marrow disorder rather than a secondary response. A normal or high erythropoietin level supports a secondary cause.
JAK2 mutation testing is central to the workup of suspected polycythemia vera. The JAK2 V617F mutation is present in roughly 95% of polycythemia vera cases, and exon 12 mutations account for most of the remainder. The presence of a JAK2 mutation with an otherwise unexplained erythrocytosis establishes the diagnosis in the appropriate clinical setting. The absence of a mutation does not fully exclude a myeloproliferative neoplasm, but it substantially lowers the probability.
The table below contrasts the main categories.
| Category | Typical erythropoietin | Key distinguishing feature | Main risk |
|---|---|---|---|
| Smoking-related secondary | Normal or high | Active smoking, elevated carboxyhemoglobin | Thrombosis with high hematocrit |
| Hypoxic lung disease | High | Chronic hypoxemia, COPD, sleep apnea | Cor pulmonale, thrombosis |
| High-altitude | High | Sustained altitude exposure | Thrombosis, hyperviscosity |
| Polycythemia vera | Low | JAK2 mutation, splenomegaly, pruritus | Thrombosis, progression |
| Renal or tumor EPO excess | High | Renal mass, tumor, or cyst | Local tumor effects |
| Relative (hemoconcentration) | Normal | Dehydration, plasma volume loss | Usually benign |
Each row represents a different mechanism and a different decision. The table should be read as a prompt for testing, not as a diagnosis.
Another edge case is the smoker with a JAK2 mutation. Smoking and polycythemia vera can coexist. A patient can have both a secondary drive and a primary marrow disorder. In that situation, treating the smoking effect does not remove the myeloproliferative risk, and JAK2 testing is essential before attributing the hematocrit to cigarettes. This overlap is one of the most important clinical traps.
A third edge case is sleep apnea in a smoker. Obstructive sleep apnea produces intermittent nocturnal hypoxemia that can drive erythropoietin and raise hematocrit. When sleep apnea is untreated, the smoking-related rise compounds the hypoxemic rise. Treating the apnea can lower the hematocrit even if smoking continues, so the smoking history alone may overstate the causal contribution.
Red flags that change urgency
Most smoking-related erythrocytosis is mild and managed in outpatient settings. Some findings change urgency and require prompt evaluation.
Hematocrit above 60% in men or above 55% in women, which warrants urgent evaluation for hyperviscosity and thrombosis. New neurological symptoms, chest pain, or visual disturbance in a patient with elevated hematocrit. Marked splenomegaly, unexplained weight loss, or night sweats, which suggest a myeloproliferative neoplasm. Low serum erythropoietin with elevated red cell mass, which points away from a secondary cause. A JAK2 mutation detected in the setting of elevated hematocrit. Carboxyhemoglobin above 15%, which suggests substantial exposure and possible carbon monoxide poisoning.
These are not diagnoses in themselves. They are prompts for a defined workup.
The time course of reversal
The most clinically useful fact is that smoking-related erythrocytosis is largely reversible. Smoking cessation removes the carbon monoxide source, and carboxyhemoglobin declines within a day. Red cell mass declines more slowly because red cells live about 120 days. Hematocrit typically falls over weeks to months rather than days.
The timeline table below summarizes typical expectations. The specific values are approximate and vary by individual.
| Time after cessation | Expected change | Mechanism |
|---|---|---|
| 4 to 12 hours | Carboxyhemoglobin falls toward normal | Carbon monoxide eliminated |
| 1 to 3 days | Blood pressure and heart rate trend downward | Nicotine and sympathetic effect wane |
| 2 to 8 weeks | Hematocrit begins to decline measurably | Red cell production slows below replacement |
| 1 to 3 months | Hematocrit approaches baseline in many patients | Red cell mass contracts |
| 3 to 6 months | Continued normalization of red cell parameters | Full turnover of red cell pool |
| Ongoing | Need for therapeutic phlebotomy often decreases | Removes the primary drive |
The exact pace depends on red cell lifespan, baseline hematocrit, and whether other drivers such as lung disease or sleep apnea persist. Patients with concurrent hypoxemia may not normalize fully. That is an important expectation to set.
Practical actions and thresholds
The practical approach starts with confirming the exposure before treating the number.
Measure carboxyhemoglobin by co-oximetry rather than relying on pulse oximetry. A level above 3% supports recent smoking exposure. Repeat the complete blood count after adequate hydration to distinguish relative hemoconcentration from true erythrocytosis. Check serum erythropoietin before any therapeutic phlebotomy in a stable patient. A low level changes the diagnostic path. Order JAK2 testing when erythropoietin is low or when splenomegaly or thrombocytosis is present. Screen for chronic lung disease and sleep apnea when hypoxemia is suspected, since treatment changes the trajectory. Counsel cessation as the primary intervention. No drug treats smoking-related erythrocytosis more effectively than stopping smoking.
For symptomatic hyperviscosity or very high hematocrit, therapeutic phlebotomy is used to lower the red cell mass. The procedure removes a unit of blood and reduces hematocrit. The decision threshold is not standardized across guidelines. Common practice targets a hematocrit below 45% in polycythemia vera and uses higher targets for secondary causes. In smoking-related erythrocytosis, phlebotomy is usually reserved for symptomatic patients or those with complications.
Phlebotomy treats the number but not the cause. If smoking continues, the hematocrit will rise again because the erythropoietin drive persists. The same logic applies to phlebotomy in hypoxic lung disease. The underlying driver must be addressed.
Cost is a secondary consideration but real in practice. A carboxyhemoglobin measurement is inexpensive. JAK2 mutation testing is more costly and is ordered selectively. Therapeutic phlebotomy is generally low-cost. The primary expense is diagnostic testing for the less common causes. That cost justifies the workup because missing a treatable myeloproliferative neoplasm has significant downstream consequences.
What the evidence cannot settle
The causal link between smoking and elevated red cell mass is strong at the mechanistic and population level. What the evidence cannot do is assign an exact hematocrit rise to an individual smoker. The magnitude varies with cigarettes per day, years of smoking, carboxyhemoglobin level, plasma volume, lung function, and genetics. Individual prediction from population averages is unreliable.
Nor can the literature settle a single universal hematocrit threshold for investigation. Guideline bodies differ, and the choice of threshold trades sensitivity against specificity. A lower threshold catches more cases of both benign and serious disease. A higher threshold misses some secondary causes but reduces unnecessary testing. The appropriate threshold depends on local resources and clinical context.
Longitudinal data on reversal after cessation are also limited in granularity. Most studies report hematocrit trends over months rather than precise day-by-day trajectories. The qualitative picture is clear; the quantitative details are not. That distinction should shape patient counseling.
Finally, smoking-related erythrocytosis is not equivalent to polycythemia vera in risk. The thrombosis risk of secondary erythrocytosis is driven largely by viscosity and coexisting vascular disease, not by the intrinsic myeloproliferative risk of polycythemia vera. Conflating the two categories is a common error that leads to over-treatment or under-treatment. The right move is to confirm the category before deciding on therapy.
The evidence supports a clear answer and a clear clinical path. Smoking causes secondary polycythemia through carbon monoxide, functional anemia, and erythropoietin-driven red cell production. The effect is reversible with cessation. The diagnostic task is to confirm the category and exclude the conditions that resemble it. The clinical task is to treat the cause, not just the number.