Understand cancer risk

Five years cannot rule out a cancer that takes longer to appear, and the authors confine their conclusion to that window. The interval reaches a rate ratio of 1.13, so a modest increase is not excluded, only a large one. Almost everyone enrolled was a middle-aged or older adult in a heart-disease or bowel-polyp trial, so the reassurance reaches women of reproductive age by argument rather than by measurement.

The shape comes from observational data, so it maps an association rather than a demonstrated causal curve. Nonlinearity cleared significance for recurrence only; for breast-cancer death and death from any cause it did not, and most of the reduction there sat between 20 and 40 mg a day.

Diet was self-reported, and soy eaters may differ from others in weight, diet quality and health behaviour that no adjustment removes. Only recurrence reached significance; both death-rate estimates crossed the line of no effect. Diagnoses ran from 1991 to 2006.

The tamoxifen-user estimate, unlike the non-user one, still includes no effect within its range of uncertainty, so the near-identity is between point estimates rather than between two equally firm results. Every study pooled is observational, with self-reported diet.

Neither trial screened people in for low vitamin D, so both test adding more to people who mostly had enough rather than correcting a shortfall. The Australian participants averaged 77 nanomoles per litre before they started, in a country with a great deal of sun. Five years may be short for cancers that take longer than that to surface.

Four cases against one is too few to size the risk; it gives the order of magnitude and nothing more. The rodent finding is why anyone with a personal or family history of medullary thyroid cancer, a rare tumour of the thyroid, is told not to take it. The label lists further warnings the trials were never built to quantify, among them gallbladder disease, acute kidney injury and suicidal thoughts.

Twelve of the 18 studies are case-control, a design in which people recall their past diet after already knowing their diagnosis. The literature stops in 2004. Eight studies gave no menopause-specific result, and six of those did not support an association at all.

Most outcomes are synthesised narratively rather than pooled, so there are few numbers behind the verdict. The reassurance covers soy as food, not concentrated extracts, which is the distinction the 100 mg line marks. It was conducted at a naturopathic college, which has an interest in the answer.

Every included study is observational with self-reported intake, so soy eaters may simply differ from other patients. The authors describe their own overall-survival finding as limited evidence and confine it to postmenopausal patients. The search stops in January 2018.

No pooled effect size was reported for these five trials, so the size of the null cannot be stated. The wider review they sit inside is a narrative synthesis rather than a meta-analysis for most of its outcomes. It was conducted at a naturopathic college, which has an interest in the answer.

Every study on both sides is observational, so men who take these drugs differ from men who do not in ways no adjustment removes, sun exposure among them. They also see doctors more often, which alone yields more diagnoses of a cancer found by looking. The newer pooling's studies barely agree.

Men on erectile-dysfunction drugs are diagnosed with melanoma more often. What that means is unsettled: one analysis reads it as shared sun exposure, and a later one urges at-risk men to avoid the drugs without answering that argument.

It appeared only for standalone high-dose supplements, not for the same vitamins inside a multivitamin, which is hard to explain if the vitamin is doing it. Intake was reported once, at the start. The review clearing B12 includes authors with declared ties to manufacturers.

Whether high-dose B12 supplements raise lung cancer risk is unresolved. One cohort saw about double the risk in men, none in women, a split nobody has explained, and the risk ran higher in smokers, so something besides B12 may explain it.

The cohort recorded what people chose to consume rather than assigning it, and its own authors flag selection effects, unmeasured differences between users and non-users, and the possibility that illness came first. The percentage rises are real, but the absolute difference in risk between users and non-users is small. The rodent work is the original researchers re-examining their own contested findings, not an independent repeat. The review pointing the other way came from a consultancy regularly retained by regulated industries.

None of the three studies on sweeteners and cancer could settle it: a cohort of people's own choices, a team re-analysing its own rodent work, an industry consultancy's review. The study that could has not been run.