Once you account for what's in the bag, sex becomes irrelevant
For decades, a quiet uncertainty haunted transfusion medicine: does the identity of the donor shape the fate of the patient? A large Swedish study of nearly 370,000 recipients has now offered a clarifying answer — donor sex and pregnancy history carry no bearing on survival, and the confusion that preceded this finding was itself a lesson in how easily an unmeasured variable can masquerade as meaning.
- Conflicting studies had left clinicians without a clear answer on whether blood from female donors — particularly those who had been pregnant — raised mortality risk in recipients.
- The tension was real: men receiving blood from previously pregnant women appeared in some analyses to face higher death rates, a finding with serious implications for how blood banks might allocate donations.
- The Karolinska team identified the hidden culprit — female donors have measurably lower hemoglobin levels, meaning their blood delivers less oxygen per unit and more often requires a follow-up transfusion within 24 hours.
- Once hemoglobin differences were properly accounted for, the survival gap disappeared entirely, exposing prior alarming findings as a statistical artifact rather than a biological truth.
- Sweden's practice of distributing donor blood without regard to sex created near-random assignment, lending the observational study an unusual degree of causal credibility.
For years, a stubborn question shadowed blood transfusion research: does the sex of the donor — or whether a female donor had been pregnant — affect a patient's chances of survival? Some studies pointed toward higher mortality when men received blood from women who had been pregnant; others found no such pattern. The contradiction held, unresolved, until now.
Researchers at Karolinska Institutet analyzed nearly 370,000 patients who received their first red blood cell transfusion in Sweden between 2010 and 2018, tracking survival over two years through national registries. Their conclusion, published in JAMA Internal Medicine, was clear: donor sex and pregnancy history have no effect on patient survival. The earlier conflicts, they found, traced back to a single overlooked variable — hemoglobin.
Women's blood contains less hemoglobin on average (135 g/l versus 149 g/l in men), meaning each unit carries less oxygen-carrying capacity. Patients transfused with female-donor blood were 12 percent more likely to need a second transfusion within 24 hours — not because of anything about the donor's sex per se, but because of what was physically in the bag. Once the analysis adjusted for hemoglobin levels, the apparent survival difference vanished. Lead author Jingcheng Zhao described the earlier findings as a statistical illusion: a real-looking difference that reflected blood composition, not donor identity.
The study's design lent it unusual strength. Because Swedish blood banks do not sort donations by donor sex or pregnancy history, blood was effectively distributed at random — allowing the researchers to draw conclusions with a confidence rarely available from observational data. Zhao noted the results could be interpreted as equivalent to a randomized controlled trial.
One limitation: the study could not examine unfiltered blood products, since leukoreduction has been standard in Sweden since the 1990s. The team plans to build on this work by developing better methods for studying transfusion outcomes and exploring other donor and storage variables. For now, one long-standing uncertainty has been put to rest.
For years, blood banks and transfusion researchers have wrestled with a nagging question: does it matter who gives the blood? Specifically, does the sex of the donor, or whether a female donor has been pregnant, change the odds that a patient will survive after receiving a transfusion? The medical literature offered no clear answer. Some studies suggested that men who received blood from women—especially women who had been pregnant—faced higher mortality rates. Other research found no such pattern. The contradiction persisted, unresolved.
Now a large Swedish study has cut through the confusion. Researchers at Karolinska Institutet examined nearly 370,000 patients who received their first red blood cell transfusion between 2010 and 2018, drawing on national registries that tracked their survival over the following two years. The finding, published in JAMA Internal Medicine, is straightforward: donor sex makes no difference to patient survival. Neither does a female donor's pregnancy history. The apparent conflicts in earlier research, the team concluded, stemmed from a factor that previous investigators had overlooked.
The overlooked factor was hemoglobin itself. Blood from women contains less hemoglobin on average—135 grams per liter compared to 149 in men's blood. This is a biological reality with practical consequences. Patients who received transfusions from female donors had a 12 percent higher risk of needing another transfusion within 24 hours, simply because the first unit contained less oxygen-carrying capacity. But when the researchers adjusted their analysis to account for these hemoglobin differences, the sex gap vanished entirely. Survival rates were identical whether the blood came from a man or a woman, whether the woman had been pregnant once, twice, or never.
Jingcheng Zhao, the study's lead author, explained the significance of this correction. Previous studies had failed to factor in hemoglobin variation, creating a statistical illusion—a difference that looked like it mattered but actually reflected nothing more than the physical properties of the blood itself. Once you account for what's actually in the bag, sex becomes irrelevant to outcomes. "When we take into account the lower hemoglobin levels in blood from women, we see no difference in survival," Zhao said, "regardless of how many times the female donors had been pregnant and of the patients' sex and age."
The study's design gave the researchers unusual confidence in their conclusions. Blood banks in Sweden don't consider donor sex or pregnancy history when deciding which units to send to hospitals. This meant the assignment of donor blood to patients was essentially random—as close to a controlled experiment as observational data can get. Zhao noted that this randomness allowed them to draw causal conclusions from what would normally be considered merely correlational evidence. "It means that our results can be interpreted as if our study was a randomized controlled study," he said.
The research was financed by the Swedish Research Council, Karolinska Institutet, and Region Stockholm, with no reported conflicts of interest. One limitation: the study could not separately examine transfusions using red blood cells that had not undergone leukoreduction—the filtering out of white blood cells—because this procedure has been standard in Sweden since the 1990s. The researchers cautioned against generalizing their findings to unfiltered blood products, though such products are now relatively rare in modern transfusion practice.
Zhao and his colleagues say the work opens new ground. They plan to develop better methods for studying causal relationships in transfusion epidemiology using observational data, examining other donor characteristics and how blood is handled during storage and transport. "There's still much we don't know about blood transfusion and its effects," Zhao said. For now, though, one persistent question has been answered: the person whose blood you receive need not be your demographic match to give you the best chance of survival.
Notable Quotes
When we take into account the lower hemoglobin levels in blood from women, we see no difference in survival among patients who received a blood transfusion from women compared with from men, regardless of how many times the female donors had been pregnant.— Jingcheng Zhao, lead author, Karolinska Institutet
Our results can be interpreted as if our study was a randomized controlled study, and that we can draw conclusions on causality.— Jingcheng Zhao
The Hearth Conversation Another angle on the story
Why did earlier studies reach different conclusions about female donors?
They measured the outcome—mortality—without accounting for the fact that female blood simply contains less hemoglobin. So they saw more transfusions needed and higher mortality risk, but they attributed it to the donor's sex rather than to the hemoglobin difference.
But if female blood has less hemoglobin, doesn't that mean it's actually less effective?
It's less concentrated, yes. A unit of female blood carries less oxygen-carrying capacity. But that's not a quality problem—it's just a quantity issue. You can give more units. Once you adjust for that in the analysis, the survival outcomes are identical.
How confident are you in this finding?
Very. They studied 370,000 patients over eight years using national registries. And because Swedish blood banks assign donors randomly without regard to sex or pregnancy history, the study design is nearly as strong as a randomized trial.
What about the patients who needed a second transfusion within 24 hours?
That's real—12 percent higher risk. But it's expected and explained. You need more units of lower-hemoglobin blood to reach the same total. It's not a survival problem; it's a logistics problem.
So what changes in practice?
Probably nothing immediate. Blood banks already don't consider donor sex or pregnancy when distributing blood. But the research clarifies why earlier studies were wrong, and it opens the door to studying other donor factors more carefully—things we actually don't understand yet.
What don't we understand?
How blood is stored and handled, how other donor characteristics affect outcomes, whether there are subgroups of patients who do respond differently to certain donor types. The researchers say there's still much to learn.