New UNSW research following thousands of Australian families over 23 years shows parents' drinking matters most when teenagers are 15 to 17 – and again when those teenagers grow up and become parents themselves.
If you grew up in a house where winding down with a wine or two was just part of normal adult life, you might recognise those patterns in how you drink now.
A new study from UNSW Sydney shows that's no coincidence – but it also says the story is not as simple as "alcoholism runs in families".
What the study did
In a study published today in the journal, Health Economics, health economist Dr Sergey Alexeev used data collected in the long-running HILDA Survey to analyse drinking data from 6650 young Australians and their parents over 23 years. The children's drinking habits were tracked from age 15 and into their late 30s, while their drinking throughout that time was measured in relation to their parents' typical drinking patterns. This created 10s of thousands of parent-child data points from the original sample – a much larger dataset than previous studies.
"Most studies on this topic are small or short‑term, or they only ask about drinking at a single point in time," Dr Alexeev says. "Here we finally have a national panel that's been running long enough – and asking detailed questions about alcohol every year – to see both generations properly."
What the study found
Parents' influence on adolescents' drinking patterns doesn't rise steadily as children grow, rather, it's strongest at two key stages of life.
The first is middle adolescence, around ages 15-17. At that point, most teens still live at home, are starting to go to parties and experiment with alcohol, and are watching closely how adults behave.
"In the data, 15-17 really jumps out," Dr Alexeev says. "At those ages, heavier‑drinking parents are most likely to have heavier‑drinking teens, when compared to their peers. And similarly, light‑drinking parents are most likely to have light‑drinking teens."
After that, through the late teens and early 20s, the match between parents and kids weakens. Friends, partners, flatmates, workmates and classmates all become more important. Young adults' drinking starts to look less like their parents' and more like the people around them.
But in the late 20s and 30s, something surprising happens: as many of those young adults become parents themselves, the parental imprint reappears.
"When people are working out what a 'normal' adult and parent looks like, they seem to revert a bit towards the patterns they grew up with," Dr Alexeev says. "It's like the template you learned at home lies dormant for a decade, then switches back on when you set up your own family life."
Like mother, like daughter – like father, like son
The second big result is about who influences who.
The study found drinking is passed from one generation to the next along same‑sex lines. Daughters tend to resemble their mothers while sons' drinking resembles their fathers.
While the links are modest, Dr Alexeev says they're consistent in the two life stages that stand out.
"Families where mothers typically drink more tend to have daughters who also drink a little more, on average," he says.
"The same is true for fathers and sons. There is little evidence that fathers shape daughters' drinking in the same way, while, interestingly, mothers have a smaller but still important influence on sons, especially in the two key age windows."
Dr Alexeev says it's difficult to say exactly why mums also influence sons.
"But it fits a story about how drinking norms are set early: kids learn what 'normal' looks like at home, and mums often shape the day-to-day routines and rules around alcohol."
When compared with other traits measured in the HILDA survey, alcohol sits in the middle, Dr Alexeev says.
"It's less consistently passed on than something like religiosity, which parents can directly control, but more persistent than earnings or mental health, which are heavily shaped by the wider world."
Not just in the genes
To work out whether these patterns are mainly genetic or about social learning, the study looked at families where parents and children are not biologically related – such as step‑parents or adoptive parents.
Here, the results largely echo the same effect. Daughters resemble their mothers' drinking even when there is no biological link. But for sons, the father-son resemblance weakens sharply in non‑birth families, while the mother-son link is more stable.
One likely reason is exposure: step‑parents often enter a family later, so they may miss the mid‑teen years when drinking habits are first taking shape. The data can't test that directly, but it would help explain why daughters still mirror the mother‑figure in the household, while the father-son link looks more contingent in non‑birth families.
"That's hard to explain by genes alone," Dr Alexeev says. "It fits more naturally with social learning and gender norms."
Habits that stick
The study also shows just how persistent drinking patterns are once they're set.
By tracking people from early adulthood into mid‑life, Dr Alexeev found that most Australians stay in the same broad drinking range for decades. In fact, people are roughly twice as likely to change social class as to change their drinking level.
"Once your drinking pattern has set in early adulthood, it's remarkably sticky," he says. "That's why those short windows of parental influence can cast such a long shadow.
"Still, this isn't destiny. These are average effects across thousands of families. A 10% increase in a parent's drinking is linked to roughly a 1% increase in an adult child's drinking, and many people end up drinking very differently to their parents."
What this means for parents and policy
For parents, the message isn't "never let your teenager see you drink". Rather, be aware when drinking patterns are more likely to be influential.
As the data shows, there are two acute times: the mid‑teens, when young people are first working out what 'normal' drinking looks like at home, and later, when those same teenagers grow up and become parents themselves. This is the time, the data suggests, they may find themselves defaulting back to the drinking patterns they learned earlier, as they shape a home environment for their own children.
For policymakers, the findings point to the value of timing and context. School‑based programs that speak to both parents and teens, and services that support new parents in setting up healthy routines at home, may have effects that ripple across generations.
"Children aren't just passive victims of 'alcoholic genes'," Dr Alexeev says. "They're active learners, watching what adults do and copying the strategies that seem to work."
Looking ahead
Dr Alexeev says future work will expand the HILDA-based intergenerational framework beyond alcohol to other traits such as mental health, resilience and risk-taking, to better understand how family influences shape behaviour across generations.
Another promising direction, he says, is Indigenous-led research examining multi-generational patterns in Aboriginal and Torres Strait Islander communities, where concepts such as intergenerational trauma resonate with the cyclical patterns seen in drinking behaviour.
"Working in partnership with communities could help us understand how these patterns operate across different family structures and cultural contexts, and translate that evidence into more effective policy."
