After creatinine, the line that most often turns red on a CKD cat's blood work is HCT (hematocrit). Anemia affects 30–65% of cats with CKD and becomes more common as the stage rises. It is tied directly to appetite and energy, so most households eventually hear the words "let's try the red-cell injection (darbepoetin)".

This article covers what to know before and after that injection — especially two things textbooks rarely spell out: stopping the injection shows up as a drop about a month later, and the injection and iron belong together. We show both with a year and a half of records from our own cat.

1. Why CKD causes anemia — four strands

• Too little red-cell hormone — the kidney makes the hormone (EPO) that tells the bone marrow to make red cells. Less kidney tissue means a weaker signal. This is the core of CKD anemia.
• Shorter red-cell life — feline red cells normally live about 80 days (shorter than the human 120). In uremic blood they live even less.
• Iron that cannot be used — CKD and chronic inflammation close the gate through which iron enters from the gut (via a hormone called hepcidin). Iron in the body cannot reach the marrow.
• Blood that is lost — small gastrointestinal bleeds (CKD, steroids, uremic gastritis) and frequent blood draws. A cat has roughly 55–65 ml of blood per kg, so a 5 kg cat has only about 300 ml in total.

The first two are "making too few, losing them too fast"; the last two are "no material, or a leak". Treatment splits the same way — a drug that runs the factory (darbepoetin) and the material (iron).

2. Reading the CBC — count, size, contents, today's output

Two of these are easiest to remember as pictures. Reticulocytes are "today's factory output". A cat's freshly released (aggregate) reticulocytes mature within a day or two, so this number shows only the last few days of production. RDW is "how mixed the warehouse stock is". Red cells live about 80 days, so RDW reflects what came in over the past several weeks to two months. On the same report, the two numbers look at different time windows.

And MCH can be calculated even when the report leaves it out. For example, HGB 10.6 and RBC 9.3 gives 10.6 ÷ 9.3 × 10 = 11.4 pg. If this number drifts down over months, the factory is making more cells but filling each one less — the picture of a factory short of iron.

3. When to treat — HCT 25%

The 2026 IRIS guidelines say to start treatment at HCT below 25%, or when it stays at 25–28%. Quality of life is reported to fall below about 27%. The guideline is not "hold out until 20%".

There are three options.

• Darbepoetin (brand name Aranesp) — an injection. The most used and best evidenced today, and the focus of this article.
• Molidustat — an oral drug that makes the body produce more of its own EPO, conditionally approved in the US. See the drug and supplement compendium and new treatments.
• Transfusion — buys time when HCT falls fast or signs are severe. Darbepoetin takes 2–3 weeks to act, so a transfusion sometimes bridges that gap.

4. Using darbepoetin — do not stop at the peak

Dose and what to expect

Median survival was about 8 months in cats that responded and about 3 months in those that did not. But non-responders were probably sicker to begin with, so this should not be read as "the injection bought five months". Whether a cat responds is also a marker of how that cat is doing.

For reference, the human CKD label starts at 0.45 µg/kg once weekly. Some cats respond below the feline standard (0.75–1 µg/kg) — ours did (section 6).

The sawtooth — the peak comes 1–2 weeks after the last injection

Darbepoetin tells the marrow to produce, and the cells it signals take time to become red cells. So HCT peaks 1–2 weeks after the last injection. The problem is what comes next. Stopping halts production, while the cells already made last a few more weeks and then fade. The number falls 3–5 weeks after stopping.

Repeating "stop at target, restart when it falls" turns the graph into a sawtooth. After restarting, the effect again takes 2–3 weeks — and that gap is the lowest stretch. That is why the guidance is to stretch the interval (weekly → every 2 weeks) instead of stopping.

5. Iron — why it pairs with darbepoetin

The "iron (Fe)" on a blood test sees only about 0.1% of the body's iron. Most body iron sits inside hemoglobin in red cells, and the rest is stored in the liver and marrow. Iron floating in the blood is the tiny amount moving between the two. So serum iron can read normal while stores are empty.

Think of stored iron as an emergency savings account. Day to day the body recycles iron from old red cells and barely touches it. It draws on the account at three times — after blood loss, during a fast self-recovery, and when darbepoetin pushes production. Darbepoetin runs the factory several times faster than usual, so material runs out quickly. A factory short of material raises the count but fills each cell less — RBC going up while MCH and MCV go down is the sign.

Oral iron, injected iron

• Absorption — in people, roughly 10% of swallowed iron is absorbed and the rest leaves in the stool (which may look dark). Once absorbed, iron has no exit, so the body controls the amount at the entry gate (the gut).
• Alternate days absorb more — a dose of iron raises the gate-closing hormone (hepcidin) for 24 hours. In a human study, alternate-day dosing absorbed 34% more than daily dosing (Stoffel 2017). There is no cat study, but the mechanism is the same, so alternate days is a reasonable choice.
• With inflammation the gate is shut — oral iron then gets in poorly. That is why many vets give iron dextran 50 mg per cat intramuscularly every 1–2 months. It fits well on a darbepoetin day.
• How to give it — iron tastes metallic. Mixed into the main meal it can turn the cat against that food (food aversion), so give it separately in a small favourite treat.

B12 and folate — not material but machinery

If iron is the material for red cells, B12 (cobalamin) and folate are the factory machinery. When the machinery fails, cells are released large and under-divided, so — the opposite of iron shortage — MCV goes up. Cats make the factor needed to absorb B12 only in the pancreas, so CKD cats with pancreatitis or chronic enteropathy run low easily. B12 shortage, though, usually shows first as poor appetite and weight loss rather than anemia. See cobalamin in the compendium.

6. A real record — a year and a half, three courses of darbepoetin

This is our CKD cat (16 years old, about 5.7 kg, stage 4 since May 2026). Values are from a reference lab; days measured on the in-clinic analyser are marked. MCH is calculated as HGB ÷ RBC × 10.

Four things can be read from this table.

① For the first year it was a "fewer cells" anemia. From April 2025 to early February 2026, HCT fell from 31.7 to 27.6, but MCH stayed around 14 and MCV around 38. Each cell was normal; there were just fewer of them — the textbook picture of hormone-deficiency anemia. An iron supplement was given on and off during this period (about 7 months, 44 doses) without changing the numbers. Iron was not the bottleneck then, so that was expected.

② The February drop was a different event. Red cells fell 20% in 17 days. CKD anemia does not move that fast. At the same time creatinine went down while urea went up, and steroids were at their highest dose for the pancreas. A single gastrointestinal bleed fits best (no black stool was recorded, so it is an inference). Then in March the cat recovered 36% on its own, without darbepoetin — a good sign that the marrow was still alive, but a recovery like that would have drawn down the savings account. The iron supplement had stopped at the end of February.

③ MCH began to fall. From March: 14.4 → 13.4 → 13.0 → 12.8 → 11.4. By September 8, RBC was 63% higher than in February, but HGB only 29% higher. MCV shrank from 38.4 to 32.3, and RDW, 15–19 until February, rose to 29.7 in May and 26.2 in September — small, pale new cells mixing in: a factory short of iron. On that same September 8, serum iron was 25.9, normal. This is the "normal serum iron ≠ enough iron" of section 5. Courses 1 and 2 were both run without iron.

④ The sawtooth repeated twice. The peak of course 2 (Sep 8, HCT 30.1, the highest reference-lab value in a year) came 16 days after the last dose. Then on day 37 off the drug (Sep 29), HGB had fallen from 10.6 to 8.1 — back to the February low. Reticulocytes were already 14,000/µL on Sep 8, below the regenerative threshold of 50,000, and near zero on the in-clinic analyser on Sep 29 — the factory had stopped. Course 1 had the same shape.

About the dose — a quarter rather than a half

This cat used 10 µg prefilled syringes. The first instruction was half (5 µg, about 0.9 µg/kg — the feline standard). But the day after a half dose, the cat screamed and limped when its leg was touched. Arthritis was diagnosed around the same time, so the cause cannot be pinned down; afterwards the dose was changed to a quarter (2.5 µg, about 0.44 µg/kg) weekly, and HCT rose without pain. That is almost exactly the human starting dose (0.45 µg/kg). Half every two weeks and a quarter weekly give the same total, and weekly keeps blood levels steadier.

What changed from course 3

• Darbepoetin with iron for the first time — oral iron (iron bisglycinate, about 7 mg, given separately in a treat). The test is whether MCH stops falling or rises after 3–4 weeks. If not, injected iron is next to discuss.
• No stopping at the peak — once on target weekly, the plan is to stretch to every 2 weeks.
• Watching blood pressure — pressure rose in each month on darbepoetin (June, August), so it is measured more often in the first 2 weeks.

Course 3 began on October 1, 2026. When the late-October recheck is in, we will add the ending to this section.

7. What to watch at home

• Gum and nose colour — on a good day, photograph the gums in bright light. Without a reference you cannot see "paler".
• Sleeping breathing rate — as anemia deepens, breathing speeds up to carry more oxygen. In cats on fluids, fluid overload raises it too; a sudden weight gain (100–150 g in a day or two) helps tell them apart.
• Energy and appetite — anemia is one of the things that lowers appetite. In our records, though, the link was weaker than expected within HCT 22–30%. If appetite falls, suspect nausea and pain as well (appetite stimulants).
• Pica — licking litter, walls or concrete can be a sign of anemia, especially iron shortage.

8. Questions to ask your vet

• "What HCT are we aiming for? Once we reach it, can we stretch to every 2 weeks rather than stop?"
• "Should we add iron with darbepoetin? Oral, or injected (iron dextran)?"
• "Can the blood test include an absolute reticulocyte count and a blood smear?"
• "Can we follow anemia numbers at the same lab from now on?"
• "After starting darbepoetin, when and how often should blood pressure be checked?"
• "(If MCV is high) Should we check B12?"

9. Further reading

• Kidney values on the blood test — reading the whole report
• Blood pressure — the number that pairs with darbepoetin
• Metabolic acidosis — another complication in chapter 07
• Appetite stimulants — the appetite problem that overlaps with anemia
• Getting the most from your vet — recheck schedule and question list
• Compendium — red-cell injections and iron