Gut Check is a periodic look at health claims. We ask: Should you believe this?

The claim:

M

en who received a blood transfusion from a woman who had ever been pregnant had a higher risk of dying prematurely than men who got blood from a man or a never-pregnant woman, scientists reported on Tuesday.

Tell me more:

Bizarre as this seems, researchers have seen hints of a “mother effect” on blood transfusions before. Half a dozen studies have found that recipients were more likely to die after receiving blood from a woman than from a man — though the biggest and most recent did not.

Deaths resulting from transfusions are rare, but one cause can be transfusion-related acute lung injury. It has been linked to blood specifically from mothers, especially if the transfusion consisted not of red blood cells but plasma, the liquid portion of blood that contains antibodies. Pregnancy changes a woman’s circulating antibodies, as well as other characteristics of her immune system.

advertisement

Scientists in the Netherlands therefore analyzed data on 31,118 patients who received red blood cell transfusions from 2005 to 2015. Most of the donors (88 percent) were male; 6 percent were women who had been pregnant at least once (regardless of outcome), and 6 percent were women who had not been, according to the national blood supply agency, Sanquin. The sex imbalance is because men are allowed to donate blood more often than women and because donations from women with an unknown pregnancy history were excluded from the analysis.

Recipients, ages 42 to 77, were followed for a median of 245 days. Within that time, 3,969, or 13 percent of recipients, died, Sanquin’s Rutger Middelburg and colleagues reported in the Journal of the American Medical Association. The risk of dying was about the same after getting blood from a man or a never-pregnant women, regardless of the recipient’s sex. But every unit of blood that a man got from a woman who had been pregnant raised his chance of dying 13 percent.

The results are “provocative,” said Dr. Ritchard Cable of the American Red Cross, which supplies just over 40 percent of U.S. blood transfusions, who co-authored a commentary on the study. “If true, they would be revolutionary — but that’s a big ‘if’ with capital letters.”

Really?

The “if” starts with the fact that the researchers tested lots of pairings: male and female recipients in four age groups, male and female donors with different pregnancy statuses, men and women who got a single transfusion or many. The only statistically significant combination was males 50 and younger getting blood from ever-pregnant women.

That’s a statistical red flag: When you test a lot of things, you’re supposed to adjust your math to reduce the chance of an association popping up at random. The researchers didn’t, they acknowledge, prompting what Cable called “lively discussion” by the paper’s reviewers and JAMA editors about whether the statistical analysis passed muster. In the end, the paper was accepted.

Newsletters

Sign up for our Daily Recap newsletter

Please enter a valid email address.

Another problem is that if recipients got multiple transfusions from different kinds of donors (a man and then a woman, for instance, or a never-pregnant woman and then a mother), they weren’t counted in the analysis that found the higher mortality risk to men getting blood from an ever-pregnant woman. Needing multiple transfusions is a sign of serious illness, so the researchers used statistical techniques to control for this. That led to what Cable said was “a very complex study design” that a lot of the reviewers and JAMA editors “had trouble with.”

“It’s not at all clear,” he added, whether alternative factors that might explain the higher mortality in these men were ruled out. A JAMA spokesperson did not respond to a request for comment.

If the danger of blood from mothers is real, however, it’s probably because of something immunological. Being pregnant changes the immune system, including producing antibodies to molecules made by genes on the Y chromosome if her fetus is male. It’s not clear, though, how that could prove lethal to men who got a mother’s blood: All the transfusions in the study were of red blood cells, not whole blood, meaning without plasma. Antibodies are in plasma, not cells.

Red blood cell transfusions also contain some white blood cells, said Middelburg, who is also an epidemiologist at Leiden University Medical Center in the Netherlands. But it’s questionable whether there are enough such lymphocytes to trigger a lethal immune reaction. The researchers don’t even know what the transfusion recipients died of, let alone “why the [mortality] effect should be limited to … men under 50,” Middelburg said in an interview. He and his colleagues therefore call their findings “very tentative” in the paper.

The verdict:

Blood banks would do well to study whether blood from ever-pregnant women really puts men at risk, but this study is too weak and preliminary to bar mothers from donating blood or require that blood be labeled with the pregnancy history of donors.

Leave a Reply to Laura Henze Russell Cancel reply

Please enter your name.
Please enter a comment.

  • Ugo, suggest you do your homework before responding. The point was made by in a lecture at Tufts University by Dr. Joanne Berger-Sweeney, who holds a Ph.D. in neurotoxicology from the Johns Hopkins School of Public Health, and served as dean of the School of Arts and Sciences at Tufts, and is now President of Trinity College.

    Or you could google, and review this 2013 post by Stanford Medical School researchers: “In men, high testosterone can mean weakened immune response, study finds,” at https://med.stanford.edu/news/all-news/2013/12/in-men-high-testosterone-can-mean-weakened-immune-response-study-finds.html. A woman needs a stronger immune system in order to allow a fetus to grow within her (which has one-half of her DNA and one-half of the father’s DNA).

    So which do you have, a stronger immune system, or low T?

  • I believe it’s fairly well established that women have stronger immune systems than men, because they need them to bear children. I heard a scientific presentation about some diseases thst only affect girls because boys with the condition all die in utero. Immune system differences are likely why there are higher rates of autism in boys.

    • I am fairly sure that your “fairly well established” is not supported by scientific evidence.
      If anything, a pregnant woman should have a weakened/permissive immune system to allow for the “parasite” to grow in her

  • I find “curious” about why you and/or the authors felt compelled to find some kind of justification for the HUGE bias in gender of donors.
    Clearly, giving blood is a manly man’s job! :-0

  • All fetuses bathe in fluid. I once heard about a woman,who had directly seen male and female hormones given to those of the opposite sex. She said female hormones had little affect on what she called gristly males. She thought male hormones given to small framed narrow shouldered women in particular seemed to have a poisonous affect. What about blood from strongly masculine men to pregnant females?

  • “The sex imbalance is because men are allowed to donate blood more often than women.”

    This statement does not appear consistent with the available data.

    In the Netherlands, men can donate 5 times/yearly. Women can donate 3 times/yearly.

    Per the 2010 paper below, 50% of blood donors in the Netherlands are women.

    https://www.sanquin.nl/en/become-a-donor/can-i-become-a-blood-donor/registration-procedure/how-often-can-i-give-blood/
    “Men can give blood a maximum of five times per year, women three times per year. How often you give blood will depend on your blood type.”

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2957494/

    “Women [play] a more substantial role [in blood donation]: in Spain 46% of the donors are women, in Portugal 43%, in Belgium 45.%, in the Netherlands 50%, in Denmark 50%, in France 50%, in the United Kingdom 53%, and in Finland 55%.

  • While an intriguing and possibly relevant idea, this body of research including the study cited displays systemic problems in biomedical and social sciences where statistical analysis and p values are relied on.
    Intended or not, the Scandinavian study cited amounts to p-hacking, described but not named in the article.
    Causes of death are not described in the article. Were they described in the research report? Specifically are the excess deaths plausibly caused by transfusions?
    Most of all, this represents the fragmentation of a great deal of research. If there is cause to suspect a given causal relationship, e.g. transfusions and deaths, based on low quality or conflicting studies, what does another low quality study add, other than muddying the water further? This is a pervasive problem. A single high quality study is called for, ideally a prospective RCT, but does not serve the publication requirements for academics and conceivably might be difficult to fund.

Sign up for our Morning Rounds newsletter

Your daily dose of news in health and medicine.