Sweets eaten in the two days before a stool sample were linked to a further 24.1% drop in gut bacterial diversity among hospitalized blood cancer patients who had received broad-spectrum antibiotics, according to a paper published in Nature on Sept. 30, 2026. The same sugar showed no such association in patients who had not been given antibiotics. Because the patients were confined to hospital wards, every item on their trays could be logged, which is what made the comparison possible.
The analysis drew on 173 patients and 9,419 recorded meals. Jonas Schluter of NYU Langone Health, Marcel van den Brink of City of Hope and Jonathan Peled of Memorial Sloan Kettering Cancer Center are the co-senior authors; Anqi Dai is first author.
Meal logs and daily stool samples
All 173 patients were in hospital for allogeneic hematopoietic cell transplantation, the donor stem cell procedure used mainly for leukemia, lymphoma and other blood cancers. The National Cancer Institute notes that the pre-transplant regimen leaves patients at increased risk of infection and that the stay can run weeks, which is why antibiotics are routine on these wards and why diet can be recorded meal by meal. Dietary intake was logged at every meal and microbiome samples were collected daily, a cadence that is almost impossible to reproduce with outpatients who recall their diets from memory. That design let the team ask a narrow question: when antibiotics and sugar arrive together, does the gut lose more than antibiotics alone would take?
The Bayesian analysis in the Nature paper used 1,009 stool samples from 158 of the patients, each matched to the sweets eaten in the preceding two days.
After exposure to broad-spectrum antibiotics, every additional 100 grams of sweets was associated with a further 24.1% decrease in mean alpha diversity, a measure of how many different bacterial species are present, with a 95% confidence interval of 7.7% to 38.1%. NYU Langone’s own summary puts the same loss at 21% per 100 grams; the paper’s abstract figure is the one used here. Sweets without antibiotic exposure showed no such link.
Enterococcus faecium and the biapenem mice
The organism that moved most was Enterococcus faecium. Following antibiotic exposure, each extra 100 grams of sweets was associated with a 0.89-unit rise in its centered log-ratio abundance (95% confidence interval 0.22 to 1.58). The Centers for Disease Control and Prevention describes the closely related vancomycin-resistant enterococci as germs that cause most of their infections in hospitals, particularly among patients with weakened immune systems, and estimated about 54,500 such infections and 5,400 deaths among hospitalized U.S. patients in 2017.
A separate mouse experiment tested the idea directly. Animals given the antibiotic biapenem along with dietary sucrose showed a 16.3-fold expansion of Enterococcus on day 3 and a 33.4-fold expansion on day 6, while sucrose without an antibiotic had no effect on the enterococcal burden. The pairing mirrors what the human data showed: sugar alone did little, and the damage appeared only when an antibiotic had already thinned out the competing bacteria.
Limits of an observational ward cohort
William Jogia, a PhD student in Schluter’s lab and the paper’s second author, has said it is not yet clear whether sugar directly nourishes Enterococcus, weakens its competitors or acts some other way in patients. The human data are observational, drawn from one kind of patient, and the mouse arm was a separate experiment, so the results describe an association in a specific hospital population rather than a rule for anyone taking a course of amoxicillin. Nothing in the paper shows that a patient who skips dessert during a transplant admission avoids an infection, and no trial has yet randomized patients to different sugar intakes while they are on antibiotics.
Schluter nonetheless told NYU Langone that cutting back on sweets during and after antibiotic treatment could help protect the gut. The NYU Langone release also discloses that Schluter holds intellectual property filings related to the microbiome, sits on an advisory board and holds equity in Jona Health, and co-founded Postbiotics Plus Research, relationships the university says it manages under its policies. A syndicated account lists multiple National Institutes of Health grants among the funders, alongside Gabrielle’s Angel Foundation, the Starr Cancer Consortium and the American Society of Clinical Oncology. Authors from MD Anderson, the University of Siena, the University of Minnesota and Fred Hutchinson Cancer Center also contributed.
Peled, a Memorial Sloan Kettering co-senior author, notes that microbiome injury in blood cancer patients has been tied to higher mortality and more infections, and he has called for trials of short-term dietary changes. The 24.1% figure currently rests on 1,009 stool samples from 158 patients.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
