Do small family gatherings really spread the virus?
Yes. In this Danish registry study, grandparents had a 9.9 percent higher hazard of catching SARS-CoV-2 in the days around a grandchild’s birthday than at other times. The risk was greater when the birthday belonged to a preschool-aged child.
This is one of those studies where the design is as interesting as the result. Researchers cannot randomly assign families to hold parties or skip them. But a child’s birthday is close to random. It falls on the same date every year, it has nothing to do with how careful a family is about germs, and it reliably pulls relatives into one room. So the researchers treated each grandchild’s birthday as a natural experiment: an occasion for a gathering that arrives on a fixed date regardless of what a virus is doing.
Denmark keeps national registries that link people to their children and grandchildren, and that record every PCR and antigen test carried out in the country. That let the team follow more than a million grandparents and ask what happened to their infection risk in the window after each grandchild’s birthday.
What the data show
The study followed 1,106,493 grandparents linked to 961,294 grandchildren between February 19, 2020 and February 28, 2022. The grandparents had a median age of 67 years, with an interquartile range of 60 to 73, and 54 percent were female. The exposure window was defined as the seven days running from two to eight days after the birthday, which allows for incubation time and for parties that do not land exactly on the date.
Across that window, grandparents had a 9.9 percent higher hazard of testing positive for SARS-CoV-2 compared with grandparents in non-birthday periods at the same moment, with a 95 percent confidence interval of 7.9 to 12.0 percent. In plain terms, the birthday window was a measurably riskier stretch of days.
The age of the birthday child mattered. When the birthday belonged to a preschool-aged grandchild, the adjusted hazard ratio was 1.15 (95 percent CI, 1.12 to 1.18), meaning about a 15 percent higher risk. For school-aged grandchildren it was 1.07 (95 percent CI, 1.04 to 1.09), closer to a 7 percent bump. Younger children spread more virus to visiting grandparents, which fits what is known about how toddlers behave: they are close-contact machines, and they are less likely to be masked, tested, or kept home for a runny nose.
The effect was not constant. It varied over calendar time and by viral variant, and it was absent during the period when Alpha dominated. So this was a pathway that opened and closed depending on which virus was circulating and what people were doing at the time.
Dr. Kumar’s Take
I like this study because it measures something many people suspected but rarely quantified. Public health messaging spent years talking about crowds, bars, and stadiums. Meanwhile the transmission that actually reached my older patients usually came from a living room, a car ride, or a birthday cake.
The effect size is honest about itself. Roughly a 10 percent higher hazard is not a catastrophe, and I would never tell a grandparent to skip their grandchild’s fourth birthday. The point is not avoidance. The point is that the risk is real, it is concentrated in a predictable window, and it is therefore one of the few risks you can actually plan around. You know the date months in advance.
I would also say the logic generalizes beyond COVID, though the data here do not. Nothing about a birthday party is specific to SARS-CoV-2. The same room, the same young child, and the same close contact drive other respiratory viruses too.
How strong is the evidence?
The design is the strength here. Because birthdays are fixed and unrelated to a family’s health habits, this comes much closer to a real experiment than most observational research on gatherings. Calendar time was the underlying time scale, so seasonal conditions, pandemic phase, and the restrictions in force at any given moment are tightly controlled. The authors also reassigned birthdays at random 100 times and reestimated the hazard ratios to check that the association was not an artifact of the data structure, and they varied the boundaries of the risk window to see whether the result depended on how they drew it.
The limits are worth naming. The study measures a window of time, not the party itself. Researchers did not observe who attended, how long anyone stayed, or whether a window was open. Some grandparents surely skipped the gathering, and including them would push the measured effect downward, which suggests the true risk of actually attending is somewhat higher than the pooled figure. And antigen tests done at home were not part of Danish surveillance, so infections caught only that way would not appear in the registry.
Who should pay attention
If you have preschool-aged grandchildren, you are the person this study describes. Birthdays for children turning up to six years carried the clearest risk, higher than birthdays for school-aged grandchildren. The cohort itself was broad: any grandparent living in Denmark with one to five grandchildren aged 0 to 15, with half of them younger than 67. This is not a finding restricted to the oldest grandparents. It does not mean staying away, either. It means treating the week of the party as a week worth a little extra care.
Practical Takeaways
- Time your vaccines around the calendar you already have. If a grandchild’s birthday falls in the fall or winter, make sure your COVID and flu shots are current a couple of weeks beforehand.
- Pay closest attention to parties for toddlers and preschoolers, since those birthdays carried a 15 percent higher risk compared with 7 percent for school-aged children.
- Ask the simple question ahead of time about whether any child at the party has been sick that week, and consider holding the gathering outdoors or opening windows when the weather allows.
- The authors suggest pre-event testing as one way to reduce spillover to older adults at small family events. It is a low-effort step for a date you know in advance.
- If you develop symptoms in the days after a family gathering, test early rather than waiting it out, because early antiviral treatment depends on catching the infection quickly.
Related Studies and Research
- Tart cherry juice improves sleep duration and quality in older adults
- Long daytime naps linked to higher death risk in older adults
- High-dose vitamin D boluses in preschoolers with asthma: safe but not sufficient
- Chronic L-theanine administration for major depressive disorder
FAQs
Should grandparents stop attending birthday parties?
No, and I do not read this study as an argument for that. A 9.9 percent higher hazard during a single window is a modest increase, and the value of seeing your grandchildren is not something a hazard ratio can capture. The useful response is preparation rather than avoidance: keep vaccinations current, know that the toddler parties carry the most risk, consider testing before the event, and be quick to test if you feel unwell afterward. Risk you can predict is risk you can manage, and a birthday is about as predictable as it gets.
Why were preschoolers riskier than school-aged children?
The study reports the difference, an adjusted hazard ratio of 1.15 for birthdays of grandchildren turning up to six years versus 1.07 for those turning seven to sixteen. Why is my inference, not the study’s finding. The most likely explanation is contact behavior. Very young children are held, carried, and hugged constantly, they do not cover coughs reliably, and they often attend daycare where respiratory viruses circulate freely. School-aged children have more independence and more distance from adults. The gap is a reminder that how a child interacts with people matters as much as how many people are in the room.
Does this finding apply to viruses other than COVID?
The study only measured SARS-CoV-2, so anything beyond that is inference rather than evidence. That said, nothing about the mechanism is unique to COVID. A crowded indoor room, a young child shedding a respiratory virus, and an older adult in close contact is the same setup that drives influenza and RSV transmission. The authors frame their result as a lesson for future respiratory pandemics rather than for COVID alone. I would treat the birthday-window idea as a general one about winter respiratory viruses, while acknowledging the data here are specific to COVID.
Bottom Line
Using more than a million grandparents and the fact that birthdays land on the same day every year, Danish researchers showed that the days after a grandchild’s birthday carried a 9.9 percent higher hazard of SARS-CoV-2 infection, rising to a 15 percent higher risk when the child was preschool-aged, with the size of the effect shifting over time and by variant. The message is not to skip the party. It is that ordinary family life, not just crowds and nightclubs, moves respiratory viruses, and that a date you already have circled on the calendar is a date you can prepare for.
