A tape measure, a stopwatch, and a face photo

Expectant parents now have two ways to shield a newborn from RSV. The mother can get a vaccine during pregnancy, called RSVpreF, or the baby can get nirsevimab, a long-acting antibody given after birth. Which works better?

Researchers at Kaiser Permanente Northern California looked at full-term babies born across the 2023–24 and 2024–25 RSV seasons. Their main comparison matched 11,293 infants whose mothers were vaccinated with 3,831 who received the antibody.

Compared with babies who had neither, maternal vaccination was associated with 69.3% effectiveness against infection and 86.4% against hospitalization. For nirsevimab, the figures were 65.2% and 73.8%.

Head to head, the study found no statistically significant difference. That doesn't prove the two are identical. Just 12 babies were hospitalized in the main comparison, too few to rule out smaller differences.

This was a look back at health records, not a trial. Families who chose each option may differ in ways the data didn't capture, and testing was at clinicians' discretion. The results cover full-term infants in one insured California system. Some authors reported outside grants from vaccine makers, including Pfizer, Moderna, Sanofi, and GSK.

BCG, a tuberculosis vaccine, is also used to treat bladder cancer. Observational studies of those patients turned up something odd: the treatment was linked to less dementia. Nobody could say whether the vaccine deserved the credit, or how it might work.

A study in Science Advances tested the question in PS19 mice, which are bred to develop tangles of tau, a protein that builds up abnormally in the brain. The mice received BCG or saline at 3 months old and were examined months later.

The vaccinated mice had less chemically altered tau in the hippocampus, a brain region central to memory. Their brains shrank less, they did better on a test of recognizing new objects, and they survived longer.

BCG reprogrammed immune cells, which then took up more tau fibrils. That could help explain the findings, but the mechanism remains unconfirmed. PS19 mice capture tau damage, not the whole of Alzheimer's disease. The human hint still comes from looking back at bladder cancer patients rather than from a trial. Whether BCG safely protects people from dementia remains unknown.

Doctors have long sized up how well a patient looks. Researchers at Mass General Brigham built a model to put a number on that impression. FAHR-Face learned from more than 40 million facial photographs. It was then fine-tuned for two jobs: estimating biological age and predicting survival.

The survival version was tuned on 34,363 photos of people with cancer. Among cancer patients, those it placed in the highest-risk quarter had more than three times the mortality of those in the lowest. The two models picked up different information, and using both together improved predictions.

So far, it has been tested mainly in cancer patients, so its performance in the general population is unknown. The authors present it as support for a doctor's judgment and plan broader testing.

A health estimate drawn from a selfie raises another question: who else might run your photo through a model like this, and what they would do with the answer.

A tape measure around the calf takes seconds. Researchers in China asked whether that measurement, taken twice a few years apart, could pick out who is more likely to fall.

They followed 7,068 community-dwelling adults aged 60 or older from 2018 to 2021, a median of 3.07 years. Their average age was 79.7. By the end, 23.6% reported a fall in the past year.

Those whose calves shrank by 5 cm or more had 33% higher odds of falling than those with no shrinkage. The pattern held after researchers set aside people who had already fallen at the start.

A tape can't tell muscle from fat or fluid. Only the right leg was measured, and falls were self-reported. Because shrinkage and falls happened over the same stretch, a fall, or the illness behind it, could have caused the shrinking. The follow-up overlapped the pandemic, and everyone lived in China. This study cannot show that rebuilding calves prevents falls.

On one side, a nurse with a stopwatch timing someone's walk over a 2.44-meter course. On the other, a wrist accelerometer worn for nine days, recording cadence, or steps per minute, through ordinary life. Which tells you more about survival?

Researchers tested both in 4,067 Whitehall II participants, average age 69.3. They followed them from 2012–2013 for a median of 11.8 years; 739 died.

Faster clinic walking speed, typical everyday cadence and peak cadence all went with lower mortality after accounting for other factors. All three ranked people about equally well. But only the clinic walk significantly improved prediction beyond standard risk factors. It nudged the C index, a score of how well a model sorts people by risk, from 0.742 to 0.747.

That is a small gain. The two kinds of walking also lined up weakly, with correlations around 0.27–0.28. A deliberate walk under observation and everyday walking seem to capture different parts of mobility. An accompanying commentary suggested using clinic speed as a first-line measure, with wrist data as a second layer.

The study is observational, so it can't show that walking faster changes anyone's odds. Participants were mostly White (92.5%) and male (74.4%). The accelerometer group was healthier than the wider cohort, and there was no outside dataset for validation. There is no agreed method for measuring cadence from a wrist. Could a better tracker measure close the gap?