Common questions
Short answers to what people ask most. Where the evidence is strong, it says so. Where a finding comes from mice, it says so.
No single microbe has been shown to cause cancer the way tobacco causes lung cancer. Certain bacteria are found more often in some tumours, and Fusobacterium nucleatum is enriched in colorectal cancer tissue, but enrichment is an association. Whether these bacteria drive the disease, arrive after it starts, or simply thrive in tumour conditions is still being worked out.
For a few specific indications, yes. Particular strains reduce antibiotic-associated diarrhoea and help in some cases of infant colic. Most probiotic products on shelves have not been tested for the benefits printed on the label, and effects that hold for one strain do not transfer to another. Strain, dose, and the condition being treated all matter.
Evidence suggests it can. Patients receiving checkpoint inhibitors who took broad-spectrum antibiotics around the start of treatment tended to do worse in several cohorts, and gut microbial composition tracked with response to anti-PD-1 therapy in melanoma. These are observational findings in human patients, supported by mouse experiments that show mechanism but do not by themselves establish clinical benefit.
The pathobiome describes a microbial community that has shifted into a disease-promoting state, rather than a single pathogen acting alone. It reframes the question from which germ caused this to what changed in the whole community. The concept is useful, and it is younger and less settled than most coverage implies.
For most people, no. Consumer kits report which organisms are present, but the science linking those profiles to personalised diet or health advice is not mature enough to act on. Results also vary with sampling method and the laboratory processing them. They are interesting; they are not diagnostic.
A large share of the most striking microbiome findings. Mouse studies establish mechanism and generate hypotheses, and they routinely fail to reproduce in humans. Half Human marks the distinction every time it matters, because a result in mice and a result in patients are different claims.
Readers who want to know what the microbiome evidence supports, told by a physician who spent twenty-five years in oncology drug development watching most promising results fail. It assumes no scientific background and does not simplify to the point of distortion. It is also written for clinicians and researchers who want the claims separated from the marketing.
No. The author is a practising physician who spent his career developing cancer drugs. The book argues for holding microbiome claims to the same evidentiary standard as any drug, which is a defence of the standard rather than an attack on medicine.
Each of these gets a chapter in the book, with the studies named and their limits spelled out.