Doctors have long puzzled over an uncomfortable observation: patients with a high body mass index often respond better to immune checkpoint inhibitors, cancer drugs that unleash the immune system against tumors, than leaner patients do, in cancers including lung cancer and melanoma. A study published in Nature offers a reassuring reframing of that "obesity paradox": the advantage appears to come not from excess weight or disturbed metabolism, but from what certain diets do to the microbes living in the gut.
A team led by Lysanne Desharnais at McGill University in Montreal set out to model human eating with unusual precision, designing 12 mouse diets to mirror real-world patterns, among them Mediterranean, Japanese, vegan, American and ketogenic, alongside classic low-fat, high-fat and Western diets. After 15 weeks the animals varied widely in body weight, insulin and leptin. But how well immunotherapy worked lined up poorly with those metabolic measures.
What mattered instead was the "diet-gut axis." Obesogenic diets nurtured a robust, lasting microbial community that made tumors more responsive to treatment. Crucially, the effect was portable: fecal transplants from human donors with a high BMI improved response in mice more than transplants from normal-BMI donors, and even a short switch to an obesogenic diet could restore sensitivity in animals that had received microbes from a non-responder.
Diet and microbes together
The researchers found the two factors work best in combination. Colonizing germ-free mice with a beneficial species, Lactobacillus johnsonii, and pairing it with the right diet drove tumor regression far more effectively than either alone, an effect traced to aromatic amino-acid metabolites the microbes produce, which sharpen the killing power of T cells.
The picture was nuanced, not a licence to eat badly. A Mediterranean diet rich in olive oil kept a lean-like microbiome and stayed unresponsive, while a diet high in the plant fiber inulin was both lean and responsive. That specificity is the hopeful part.
The authors are careful: prolonged obesogenic diets carry well-known health risks and are not a proposed treatment. Rather, they write, the work points to "short-term dietary modulation, and of specific bacteria or microbial-derived metabolites," as a tractable way to build "an optimal host ecosystem" so that more patients, across diverse populations, might benefit from immunotherapy.