Leukopenia (Low White Blood Cell Count): Should You Worry?

Your lab report shows a white blood cell count below the low end of the printed range, probably flagged with an L or an arrow. This page does not repeat what white blood cells are — that is in our guide to the WBC blood test. It answers four different questions: is this number actually abnormal, which line on your report is the one that matters, what explains it most often, and when to seek care without waiting. If your result is above the upper limit instead, read high WBC (leukocytosis).

First: is this actually abnormal?

A value just under the cutoff is not a disease

A reference interval is not a border between healthy and sick. It is the range covering 95% of healthy people — which means that by construction, one perfectly healthy person in twenty falls outside it. A count of 3.6 or 3.8 K/µL against a lower limit of 4.0 very often belongs to that margin, especially since the count varies in the same person from day to day and with the time of the draw.¹

Cutoffs differ from lab to lab

This is something we verify guide after guide: printed intervals differ by analyzer and by the reference population the lab used. Some U.S. labs open the normal range at 4,500 cells/µL, others at 4,000 or 3,800. Compare yourself to the limit on YOUR report, not to one on a website — ours included.

What the draw itself can change

Unlike potassium, the WBC count is not very sensitive to how the blood was collected. One thing is worth knowing: exercise, stress, or pain push the number up for a few hours, because neutrophils resting along vessel walls are released into circulation. A calm morning draw can therefore look "lower" than a previous one without anything having happened.¹

The number that matters is the ANC, not the WBC

This is probably the most useful thing on this page. The total white count says almost nothing by itself. It adds up five cell families that do not share a role or a weight: neutrophils, lymphocytes, monocytes, eosinophils, and basophils. A total of 3,500/µL with neutrophils at 2,200 (the rest, a post-viral lymphocyte dip) has nothing in common with a total of 3,500/µL whose neutrophils are 400.

It is the neutrophils that carry defense against bacteria, and therefore the infection risk. Find the neutrophil line as an absolute number — the absolute neutrophil count (ANC), in cells/µL — not the percentage: a high percentage inside a low total can still mean very few cells. Most U.S. labs print the ANC for you on a CBC with differential.

Two honest caveats. The 2023 European Hematology Association / EuNet-INNOCHRON consensus defines adult neutropenia at 1.8 × 10⁹/L (1,800/µL), following the WHO, while most U.S. labs print 1,500 — so a result between 1,500 and 1,800 sits in genuinely contested territory.² And "agranulocytosis" is not a synonym for severe neutropenia: the same consensus reserves it for counts below 200/µL, whereas severe neutropenia starts below 500.²³ What drives management is depth, duration, and context — not the label.

A low count that is normal: Duffy-null associated neutrophil count

This section matters more in the United States than anywhere else, and it is the single most likely explanation for a stable, isolated low ANC here.

A common variant of the ACKR1/DARC gene (the rs2814778 polymorphism) removes the Duffy antigen from red blood cells — the Duffy-null phenotype. Carriers distribute their neutrophils differently between blood and tissue: the marrow produces a normal number, fewer are counted in a tube. The result is a durably lower ANC with normal bone marrow and no increased risk of infection — the American Society of Hematology states plainly that these individuals "have no increased risk of infection but are often incorrectly labeled as having neutropenia."⁴⁵

The figures are concrete, and most of them are American:

• In U.S. primary care, among 120 healthy adults who identified as Black, 66.7% carried the Duffy-null phenotype. Their median ANC was 2,820/µL versus 5,005/µL in the Duffy non-null group; 23.8% of them fell below 2,000/µL, against none of the non-null participants.⁵
• The first Duffy-null-specific reference intervals were published in JAMA in 2023: a central 95% interval of 1,210 – 5,390/µL, against the institutional standard of 1,920 – 7,600/µL.⁶ They are still not printed on routine reports.
• In two large population cohorts, 21% of people with the Duffy-null genotype sat below the neutrophil threshold commonly used to gate access to treatment, against 1% of non-carriers — and the genotype was not associated with any increased risk of infection, viral or bacterial.⁷
• The consequences are measurable. A 2024 cross-sectional study of contemporary phase 3 cancer trials found that 76.5% of eligible trials excluded patients whose ANC fell in the Duffy-null reference range, and 53.5% of standard anticancer regimens called for a dose modification at those counts.⁸

A note on wording, which is not cosmetic. This was long called "benign ethnic neutropenia." Professional bodies now name it by its cause: ACKR1/DARC-associated neutropenia in the European consensus, Duffy-null associated neutrophil count (DANC) in the American literature — explicitly "to emphasize the genetic rather than ethnic basis."²⁵ What is being described is the frequency of a gene variant, not a property of a racial group. The variant is common in people of African, Caribbean, Middle Eastern, and Yemenite Jewish ancestry, and it is a matter of individual genotype, not of appearance.

The old term survives in regulatory text: FDA clozapine labeling still says "benign ethnic neutropenia (BEN)" — while doing the practical thing, setting a separate baseline ANC of ≥1,000/µL to start treatment rather than ≥1,500/µL, with its own monitoring table.⁹ If this may apply to you, tell your clinician. It changes how your number should be read, and Duffy testing exists.

The causes, from most common to rarest

1. A recent or ongoing viral infection. By far the leading cause in an otherwise well person. Influenza, mononucleosis, CMV, COVID-19, viral hepatitis, parvovirus: the drop is usually modest and transient and corrects on its own within weeks.²³ That is exactly why a recheck is the central step.
2. Medications. Beyond chemotherapy, a long list of ordinary drugs lowers neutrophils: antithyroid drugs (methimazole, propylthiouracil), sulfasalazine, trimethoprim-sulfamethoxazole, certain beta-lactams, vancomycin, NSAIDs, ticlopidine, clozapine, rituximab and other biologics.³¹⁰ Two very different things coexist: a mild, frequent dip with no consequence, and idiosyncratic drug-induced agranulocytosis — rare, abrupt, serious, with mortality still reaching 20% in older adults.¹⁰
3. Nutrient deficiencies. Vitamin B12 and folate deficiency affect all three cell lines, so leukopenia typically comes alongside a high MCV, anemia, or a low platelet count. Copper deficiency and malnutrition do the same, more rarely.³²
4. Autoimmune disease. Lupus, rheumatoid arthritis (Felty syndrome), Sjögren's, thyroid and inflammatory bowel disease: the neutropenia is usually moderate and accompanied by other findings.²
5. Hypersplenism. An enlarged spleen — cirrhosis, portal hypertension, some blood disorders — sequesters circulating cells. Platelets usually fall too, which is why liver function tests belong in the work-up.²
6. Bone marrow disorders, last because they are the rarest: aplastic anemia, myelodysplastic syndromes, leukemia, lymphoma. They almost always show more than one abnormal cell line — low hemoglobin, low platelets — or abnormal cells on the smear, not an isolated, mild leukopenia.²³

When to seek care without waiting

• An ANC below 500/µL with a fever. This is the one true emergency on this page. The IDSA defines fever in this setting as a single oral temperature ≥ 101 °F (38.3 °C), or ≥ 100.4 °F (38.0 °C) sustained an hour, with an ANC below 500/µL or expected to fall below it within 48 hours. It calls for same-day evaluation and empiric broad-spectrum antibiotics started before culture results.¹¹¹² If you are on chemotherapy, follow the instructions your oncology team gave you and go to the ER — do not wait.
• A sore throat, mouth ulcers, or an unexplained fever while taking a high-risk drug, especially an antithyroid medication or clozapine. Do not self-treat with a leftover antibiotic: get a CBC the same day.¹⁰⁹
• Leukopenia together with anemia, low platelets, enlarged lymph nodes, an enlarged spleen, unexplained weight loss, or night sweats.
• Recurrent bacterial infections — abscesses, pneumonia, skin infections, severe gum disease. That pattern matters more than the number itself.

What your doctor will do next

The real pathway is simpler than most people expect. Facing an isolated, mild leukopenia, the first move is almost always to look at the differential and compare with your previous CBCs: a count low but stable for years means something entirely different from a recent drop.² A repeat CBC with a manual smear review settles most cases.

If the drop is confirmed, first-line testing is standardized: CRP, liver and kidney panels, vitamin B12 and folate, viral serologies (HIV, hepatitis B and C, EBV, CMV, parvovirus), thyroid testing with TSH, antinuclear and antineutrophil antibodies.² A full medication review — over-the-counter drugs and supplements included — is part of it, not an afterthought.³

A bone marrow biopsy is not a required step. The European consensus explicitly recommends against it in adults with mild, isolated, long-standing, stable neutropenia.² It is discussed when several cell lines are affected, when the count is worsening, or when the smear is abnormal.

What this result does NOT mean

It is not leukemia. Blood cancers almost always announce themselves with several abnormal cell lines, abnormal cells on the smear, or clinical signs — not with a WBC of 3,600/µL, a balanced differential, and a normal hemoglobin.

It does not mean "a weakened immune system." The link between the number and infection risk only becomes meaningful at deep neutropenia. Between 1,000 and 1,500/µL there is no measurable excess risk — and with the Duffy-null variant, none even lower.⁷⁴

It is not the explanation for your fatigue. Fatigue usually comes from the viral infection that lowered the count; the count itself does not cause tiredness. Our guide to blood tests for fatigue covers what is actually worth looking for.

And it is not caused by pregnancy. Pregnancy tends to raise the white count, so a low value in a pregnant patient is not explained by the pregnancy and deserves the same reasoning as in anyone else.¹

Get your CBC interpreted by AI DiagMe

A white blood cell count is never read alone: what matters is the differential, the ANC, the rest of your CBC, and your personal context.

👉 AI DiagMe interprets your lab results in plain language, taking your whole profile into account. An informational service that does not provide a diagnosis and complements, never replaces, your physician.

Should you repeat the test, and when?

Yes — and it is the single most useful step on this page. Reasonable intervals:

• Mild, isolated leukopenia, no fever, no symptoms: recheck in 2 to 4 weeks. That gives a post-viral dip time to correct — if the illness has only just ended, wait the full three weeks, or you are still measuring the virus.
• ANC between 500 and 1,000/µL: a closer recheck, within days, timed by your clinician.
• ANC below 500/µL, or any fever: not "recheck next month." Same day.

Persisting beyond three months, it changes category: this is chronic neutropenia, warranting a hematology opinion and CBCs every 3 to 4 months.² Many mild chronic neutropenias stay stable for life — and a share of them are, precisely, Duffy-null counts never recognized as such.

Frequently asked questions

Bottom line

A low WBC is not read on its own: the ANC is what counts, with landmarks at 1,500 / 1,000 / 500 cells/µL. The commonest cause is a transient viral infection, then medications, then deficiencies; marrow disease comes last and shows up as several abnormal cell lines. One case deserves to be widely known here: Duffy-null associated neutrophil count, a normal low ANC with no added infection risk. The only emergency is an ANC under 500/µL with fever; otherwise, the right move is a recheck in 2 to 4 weeks.

Sources

Official sources and peer-reviewed publications (PubMed) used for this guide: