Low Blood Counts That Never Recover: How Myelodysplastic Syndrome Gets Missed

This post walks the finding from the top: why a low count gets blamed on age, what myelodysplastic syndrome is, which reversible causes have to be cleared first, what the marrow test is being asked, and how risk scoring drives what follows. Most persistent low counts are not MDS. They still deserve an answer.

Why a low count gets blamed on age

Anemia is common in older adults, and that commonness is the problem. Prevalence exceeds 20 percent in people 85 and older (per Guralnik et al., Blood, and the NHANES III analysis), so it gets accepted as expected. It should not be. Anemia is not an inevitable consequence of aging, and a cause is identified in about two-thirds of older patients who get a proper workup (per the American Academy of Family Physicians).

Roughly a third of anemic older adults have an iron, folate, or B12 deficiency, a third have kidney insufficiency or chronic inflammation, and a third are unexplained on the first pass. That last third is where cases get dropped. A single low hemoglobin means little without the rest of the panel and the values before it; our guide on how to read a complete blood count covers how each line is read against the others.

MDS sits in that unexplained third. United States incidence climbs to about 58 per 100,000 in people over 80 (NIH StatPearls), exactly the decades where a low count gets waved through as normal aging.

What myelodysplastic syndrome actually is

MDS is a clonal disorder of the blood-forming stem cells. The marrow keeps producing cells, but they are abnormal and fail to mature, so they die before reaching the bloodstream. That is ineffective hematopoiesis (NIH StatPearls), and it gives the disease its counterintuitive picture: a crowded marrow feeding a blood count that stays low.

Anemia is usually the first and loudest of the low counts, followed by low platelets and low neutrophils. The red cells are often large rather than small, so the MCV runs high and does not answer to B12 or folate, a combination that points away from a simple deficiency and toward primary marrow disease (per Merck Manual and NIH StatPearls).

Two numbers set the boundaries. Dysplasia must reach at least 10 percent of cells in at least one lineage (WHO 2022 classification). And blasts, the immature precursors, must stay under 20 percent of nucleated marrow cells; at 20 percent or above the diagnosis becomes acute myeloid leukemia (NIH StatPearls). In between, the blast count divides the disease: 5 to 9 percent is MDS with increased blasts type 1 and 10 to 19 percent is type 2 under the WHO 2022 classification, while the 2022 International Consensus Classification puts that same 10 to 19 percent band into a separate category it calls MDS/AML.

That gradient is why MDS is taken seriously. Roughly 30 percent of de novo MDS progresses to acute myeloid leukemia, and treatment-related MDS progresses far more often; in one retrospective series of 112 patients, 55 percent transformed (NIH StatPearls). The word syndrome undersells it.

The reversible causes that have to go first

Before a low count can be called MDS, the fixable things must be excluded. Several look identical to MDS under a microscope and disappear when corrected. The list includes vitamin B12 and folate deficiency, copper deficiency, excess zinc, HIV, alcohol, medications, autoimmune destruction, hemolysis, and paroxysmal nocturnal hemoglobinuria (NIH StatPearls and published diagnostic guidance). Copper deficiency earns a specific mention: it produces a dysplastic-looking, cytopenic marrow repeatedly mistaken for MDS, and it reverses with repletion. The distinguishing detail is that copper deficiency reverses with repletion and, in reported cases, comes with a normal karyotype, while clonal chromosome abnormalities show up in about half of MDS cases (Blood, 2007). Suspect it after gastric bypass, in malabsorption, and in anyone taking zinc.

B12 deficiency is the other classic impostor, producing megaloblastic changes that mimic dysplasia. There are documented cases where it caused chromosome abnormalities that resolved once the B12 was replaced (per a 2017 case report in BMJ Case Reports). Which is why B12 and folate are corrected before the marrow is interpreted.

A trial of iron that does nothing is not a rule-out. It is one hypothesis tested. The rest of the list is still open.

What the bone marrow is actually being asked

The marrow test is where a low count either becomes a diagnosis or stops being one, and it is not one test. An aspirate and a core biopsy are taken together and sent for several readings: morphology with an iron stain, blast percentage, flow cytometry, cytogenetics by karyotype plus FISH, and a somatic mutation panel (per the National Cancer Institute and NIH StatPearls).

Each answers a different question. Morphology asks whether the cells are dysplastic and in how many lineages. The blast count asks how far along the process is. Cytogenetics and mutations ask whether there is a clone and which one, since that carries most of the prognostic weight. Flow cytometry separates this from other marrow diseases, and our post on reactive versus malignant blood counts covers that separation.

There is a middle ground: an unexplained cytopenia with a myeloid mutation but too little dysplasia for MDS is clonal cytopenia of undetermined significance, where a myeloid neoplasm is roughly 14 times more likely to develop with a clone than without one (hazard ratio 13.9; Malcovati et al., Blood 2017). Not MDS. Still followed.

Not aplastic anemia, and not a simple deficiency

Two conditions sit close enough to be confused with MDS, and both are separated in the marrow rather than the blood. Aplastic anemia is marrow failure from stem cell injury, with precursors diminished or absent, so the biopsy is markedly hypocellular: an empty marrow making too few cells. MDS is usually the opposite, a normocellular or hypercellular marrow making plenty of cells badly. Roughly 10 to 20 percent of MDS cases are hypocellular, the hypoplastic variant that genuinely does overlap (per an American Society of Hematology education review, 2019). Clonality separates them further: clonal chromosome abnormalities show up in about half of MDS cases and are much less common in aplastic anemia, where the karyotype is usually normal, though a minority of aplastic anemia patients do carry one.

Deficiency anemia is separated by the simplest test in the sequence: a deficiency corrects when the deficiency is corrected. MDS does not, because nothing is missing. The machinery making the cells is the problem, and that non-response is where the question is supposed to change.

How risk scoring decides what happens next

MDS is not staged the way a solid tumor is. It is risk-scored, and the score drives everything. The standard tool is the Revised International Prognostic Scoring System, built from five inputs: marrow blast percentage, chromosome findings, hemoglobin, platelet count, and neutrophil count. It sorts patients into five groups, very low, low, intermediate, high, and very high (Greenberg et al., Blood 2012, and the American Cancer Society). The newer IPSS-M folds in gene mutations and sorts into six.

Management follows the group. Lower-risk disease is managed for livable counts: transfusion, erythropoiesis-stimulating agents, lenalidomide for the 5q deletion, immunosuppressive therapy in selected patients (per the National Cancer Institute). Higher-risk disease is managed to slow transformation, with hypomethylating agents such as azacitidine and decitabine (NCI). Allogeneic stem cell transplant is the only potentially curative treatment for MDS, so eligibility is assessed early rather than after other options run out (per the National Cancer Institute).

This is why a hemoglobin by itself never tells you what happens next. Two people with the same hemoglobin land in different risk groups on a chromosome finding neither of them can feel.

How Healz reads a count that will not recover

A count that will not recover is a differential, not a diagnosis, and the differential is long. Healz is equipped with root-cause technology, so it does not stop at the first explanation on the chart. It treats the iron label as a hypothesis, cross-checks your case against more than a million rare cases, and drills toward what the iron never addressed: copper, B12, a drug on your list, or a clonal marrow only a biopsy will show.

The whole picture stays in one place. As a blood test ai analyzer and ai lab report reader, Healz reads the CBC, the smear, and the marrow report together, so the MCV, the neutrophil shape, and the blast percentage are weighed against each other instead of one line at a time. Frontier AI works your case. Healz has memory that keeps every count you upload and connects them across years, so four months of persistence is a fact on the record, not a recollection. Everything is in one place, one chat, not ten apps. When you want a human in the loop, you can bring a board-certified doctor into the same chat for a second opinion.