iiinova — where your digital self keeps evolving

Thanks to @prateekgv.bsky.social for his perseverance and tremendous attention to detail throughout this work, and to the rest of team. Zhenghao Chen, Kevin Wright, Andrea Di Francesco, Vladimir Jojic, and Gary Churchill.

000

What are the key methodological insights from this work? Dense longitudinal sampling combined with state-space modeling revealed temporal dynamics invisible to static snapshots. Derived traits provided novel lifespan predictors independent of traditional static measurements.

120

We finemapped these loci to specific genes including Pzp (hepatokine regulating liver-adipose tissue energy homeostasis), Man2a1 (glycosyl hydrolase linked to IBD), and St8sia2 (polysialic acid synthesis affecting hypothalamic energy balance regulation).

100

Genome-wide mapping identified 10 heritable traits capturing body weight dynamics associated with lifespan, mapping to 8 genomic loci. Notably, none of these loci overlapped with previously identified body weight QTLs (elifesciences.org/articles/64329), suggesting distinct genetic architectures.

100

We leveraged our high temporal resolution measurements and quantified recovery rates following stressors (handling, tests, environmental changes). Consistent with our earlier study, CR and IF mice showed faster return to steady state across multiple stressors compared to mice fed ad-libitum.

100

What about loss of body weight? Time spent steadily losing weight was negatively associated with lifespan (log hazard ratio ~+0.065 per percentage point increase during 6-12 months). The effect was strongest in middle age, indicating sustained weight loss during this period increases mortality.

100

Which dynamic state was associated with longer lifespan? Time spent in steady body weight homeostasis showed positive age-dependent associations with lifespan after controlling for diet and body weight. The association strengthened from early to late life, particularly after 18 months of age.

100

We developed an autoregressive hidden Markov model to classify weight dynamics into three physiological states: growth (weight gain), steady (maintenance), and decline (weight loss).

100

An organism’s health, however, is highly dynamic. Even lab mice face daily fluctuations in energy intake, environmental stress, and metabolic demands. Their ability to maintain equilibrium, or homeostasis, may be just as important as their annual health "snapshots" in predicting lifespan.

100

We found that the strongest trait associations with lifespan included retention of body weight through periods of handling—an indicator of stress resilience, high lymphocyte proportion, low red blood cell distribution width and high adiposity in late life.

daphnarothschild.bsky.social @daphnarothschild.bsky.social · 19/11/2025

Over a year ago, along with Gary Churchill at Jackson Labs, we explored the impact of dietary restriction on lifespan in one of the largest studies using 960 female Diversity Outbred mice. Caloric restriction and intermittent fasting both extended lifespan proportionally to the level of restriction.

152

Reposted by Anil Raj

Ribosomal RNA (rRNA) genes are found in hundreds of copies in the human genome. Do sequence variations in these paralogs change the ribosome function? Yes! I am excited to share our new preprint @mbarnalab @jkpritch in collaboration with Calico: www.medrxiv.org/content/10.1... 1/8

276

A reminder from Calico Life Sciences that the path from model organisms to human biology is shorter than it looks! Thanks to the team! @dghendrickson.bsky.social, Jordan Brown, Nathaniel Thayer, Manuel Hotz, @daphnarothschild.bsky.social , @jkpritch.bsky.social, @mbarnalab.bsky.social + others

140

Our companion paper shows heritable rRNA sequence variants within these paralogs associated with distinct trait categories: es15l with adiposity, es39l with body dimensions, es27l with blood phenotypes. [See excellent thread by @daphnarothschild.bsky.social bsky.app/profile/daph...

What happens to cellular function? In primary human pancreatic islets, we found opposing effects: Higher 45S → increased glucose-stimulated insulin secretion (especially in females). Higher 5S → reduced insulin response. Same molecular machine, opposite outcomes.

100

We traced these effects across multiple scales: clinical and blood markers from 490K individuals → MRI-derived body composition traits from 100K → 54K proteomes → gene expression in 57 tissues from 948 GTEx donors → functional assays in primary human islets.

Analysis of the plasma proteome in UK Biobank participants showed that 45S copy number impacted high-output secretory cells operating at the limits of protein synthesis: pancreatic islets, intestinal goblet cells, bone marrow.

What does 5S copy number actually affect then? Body scaling. Higher 5S correlates with increased lean mass, larger organ volumes (heart, liver, kidneys, spleen), greater height, and stronger grip. It's proportional organismal growth without disease risk.

100

Does rDNA copy number impact human health? Higher 45S associates with type 2 diabetes, cardiovascular disease, chronic kidney disease, and inflammatory markers. But 5S showed no disease associations, despite equivalent statistical power.

100

We quantified copy numbers for both 5S and 45S in ~490K UK Biobank participants using whole genome sequencing data and found they are essentially uncorrelated. The two components of the same molecular machine, despite each being highly heritable, segregate as independent genetic factors.