Heavier young men had daughters and granddaughters whose epigenetic clocks ran slower

Can what happened to your grandfather when he was young show up in your own body? Older studies based on harvest records suggested so: grandparents' food supply in childhood was linked to their grandchildren's health, but only along certain family lines. In mice, a father's diet can change the metabolism of his offspring, and some of these changes last several generations. How this could work in people is still unclear.

A team in Finland looked for such a link using a rare set of records: army draft measurements of young men taken almost a century ago, matched to the blood of their children, grandchildren and great-grandchildren. They found a consistent but small link between the men's body mass index (BMI, weight relative to height) and the pace of aging in their female descendants, and no sign of the molecular mechanism they were looking for. The work is a preprint, not yet peer-reviewed.

What they did

• The men. Finnish military draft records from 1927 to 1957 gave the height, weight, BMI and chest circumference of about 1,000 men aged 18 to 24. Almost all were in what is now called the normal range: 95% had a BMI between 18.5 and 25.
• Their descendants. The men were linked to three generations of their families in the Young Finns Study, a long-running Finnish cohort that began in 1980 with children and teenagers. Blood samples came from 362 of the men's children (aged 59 to 80), 421 grandchildren (34 to 49) and 200 great-grandchildren (7 to 30). Sperm came from 51 sons and 98 grandsons.
• The clocks. From the blood, the team read DNA methylation, small chemical tags on DNA that change with age, and used it to calculate three epigenetic clocks: PhenoAge, GrimAge, and DunedinPACE, which estimates how fast a person is aging right now (1 means one year of biological aging per calendar year).
• Accounting for other differences. Models were adjusted for the descendants' age, sex, BMI and smoking, for the men's social class, region of birth and birth year, and, in the strictest model, for the descendants' years of schooling.
• Looking for a mechanism. The team also scanned hundreds of thousands of methylation sites in the descendants' blood and sperm for links with the men's measurements.

What they found

• Higher BMI, slower clock, in the female line. A higher BMI in the young men was linked to a slower DunedinPACE in their children overall, in their daughters (169 people) and in their daughters' children. In the daughters' daughters (126 people) the link held in the simpler models but not after adjusting for schooling, except when smokers were left out.
• Nothing in the male line. The men's BMI was not linked to the clocks of their sons or of their male-line descendants.
• Driven by the upper end. The link came mainly from men in the top quarter of BMI (above 23). Their daughters aged, by DunedinPACE, a little over two weeks less per year than the daughters of the other men; the granddaughters showed a similar gap.
• Other measures were weaker. PhenoAge showed no link with BMI, and GrimAge only a weak one in the simplest models. Height and weight gave a few scattered results; chest circumference gave none.
• A partial repeat. In the Young Finns Study itself, mothers whose BMI had been higher from age 13 on had children with a slower DunedinPACE (255 mothers). Split by sex, this reached significance only in sons. Fathers' BMI (191 fathers) showed no link.
• No inherited methylation pattern. The scans of blood and sperm found almost nothing: one site in the blood of the daughters' children and one site each in the sperm of sons and grandsons. The team concludes this does not support DNA methylation as the carrier of the link.

What to keep in mind

• Small groups. The authors call the number of people per generation a clear limitation, and say their methylation scans may have lacked the power to find milder signals.
• Unknown reasons for BMI. The records do not say why some young men weighed more, so BMI cannot be read simply as a sign of better food supply. Some of the men had not yet reached their adult height.
• Selection. Families could only be linked if the man was alive in 1964, when Finland's population register began, which may bias the sample.
• Observational. The study cannot tell biological inheritance from social inheritance. The authors suggest the link might reflect wealth or status passed down the female line, although their models adjusted for social class and schooling.
• Not about being overweight. Almost all the men were in the normal range, and in adults a higher BMI usually goes with a faster DunedinPACE. A recent study found the opposite pattern, linking a father's overweight in puberty to faster aging in his children. The authors call a developmental explanation of their results "highly speculative".

Why it matters

This study offers rare data across four generations to test whether a grandparent's youth leaves a mark on grandchildren. It found a modest link between young men's BMI and slower biological aging in their daughters and granddaughters, but no molecular trace of inheritance in blood or sperm. As the authors put it, the results point to traits passed down the generations, not to inheritance through DNA methylation. Larger studies will need to show whether the pattern holds.