Welcome to Summer! It’s post-Memorial day. Which means time to dig out the white pants, put on our flip flops, and switch our (non-COVID) parenting concerns to sunscreen and swimming.
If you’re wondering about sunscreen, I wrote about it in a less well timed newsletter in September. Subsequent to that I have gotten a key follow-up question: How can you get sunscreen onto the face of a small wiggly child? My initial answer was “Ask another person to hold them down” but a number of better parents than I have suggested a cheap makeup brush or makeup sponge. Effective and apparently makes it fun. Good tip. Too late for me.
Today we will tackle the second question: swimming. And, more specifically, the question of the safety value of swim lessons.
Background: Drowning among children
Drowning is a common cause of accidental death; the CDC reports an estimated 10 unintentional drowning deaths per day in the US. Children are at high risk. I used the CDC cause of death calculator to produce the table below, looking at the counts of unintentional drowning deaths in the US in 2019 by age group. Among kids, the highest risk groups are children under 4 and those 15 to 19. (This is just the US; drowning is an even larger risk in much of the developing world.)
In thinking about prevention, it is helpful (if anxiety provoking) to break down what is happening in these different age groups. In very young children (under 1) the most common drowning location is the bathtub. Infants or babies left unattended can drown, even if they are in seemingly “safe” infant tubs. In slightly older kids the most common drowning location is pools. A toddler falls into a pool unsupervised and drowns.
This can happen incredibly fast. You’re cleaning up at the end of the day, you take the three year old’s water wings off and tell them not to go in the pool. But they walk down the steps anyway, as you’re turned away. And they go one step too far, and you do not turn around to find them in time. Drowning is quiet.
I am not trying to scare you. Well, maybe I am a little. I’ll talk below about swim lessons, etc, but a key thing to say which I think the above makes clear is that supervision of young kids around pools is just really, really important.
Drowning risks are lower for older children, partly because some of the behavior mention above is just less likely with a 9 year old. If they cannot swim, they know it. With teenagers, drowning episodes more often involve open water and, in many cases, alcohol.
Pools, fences, and swim lessons
Given the importance of swimming pools in drowning in younger children, it’s not surprising that much of the discussion about drowning prevention focuses on pool safety. Notably pool fences and swim lessons.
Both of these prevention approaches are difficult to study convincingly with data because drowning deaths are rare. Although they are a highly ranked cause of death among kids, the total number of deaths in this age group is very low. This makes it difficult to study with something like a randomized trial, for some of the reasons I discussed in this article on statistical power. You’d need an enormous randomized trial to have any hope of seeing an impact of swim lessons on deaths. Even an observational study — say, comparing kids who have taken swim lessons to those who did not — is likely implausible just given the sample size necessary.
Instead, most of the findings on these questions come from case-control studies. In this type of study, researchers identify cases — here, pool drowning deaths — and collect information about the circumstances — for example, whether the child had swim lessons. They then identify some comparable children (similar, perhaps, on demographics or other characteristics) who were not drowning victims, and collect the same data on them. By comparing the groups, they hope to find features which differ across the groups and then conclude something about their relative risk.
This approach has been used to study pool fencing (see review article here) and conclude that pool fencing lowers drowning deaths considerably (perhaps by as much as 70%), and that four sided fencing is better than partial fencing. As the review article notes, the data isn’t especially high quality and samples are small. It seems clear why fencing would matter, and why full fencing would matter more, but it’s hard to show in data.
Turning to swim lessons: the most widely cited study is a 2009 article which identified 88 cases of unintentional drowning, and compared the cases to 273 controls. They looked separately at the 1 to 4 age group and the 5 to 19 age group. The differences in formal swim lessons between the cases and controls were extremely large. For the 1 to 4 age group, they calculated an 88% reduction in the risk of drowning from formal swim lessons.
However: the sample size means the estimates are extremely noisy. For the 1 to 4 age group, the data is consistent with anything from only a 3% reduction in drowning risk to a 99% reduction. For the 5 to 19 age group, the effect is about a 60% reduction in risk, but they cannot rule out even a 50% risk increase. Put simply, the results are suggestive, especially for the younger group, but very imprecise.
There are other problems with case-control studies. Notably, the persistent worry that there are other difference between the cases and controls which are driving the effect. Formal swim lessons are likely associated with other family characteristics which could drive differences in risk. The authors try to adjust for this but it challenging.
An alternative is to run a more convincing causal study on an intermediate outcome. An example of this is this randomized trial of swim lessons, which randomized children between 2 and 3.5 years into 8 or 12 weeks of swim lessons (unfortunately there is no no-lesson control group). They find that swimming ability and water safety reactions improve in both groups, and more so in the 12 weeks group.
From this we conclude that, basically, you can teach kids this age to swim. Will this translate to lower risk of drowning? It certainty seems like it could! But it is worth remembering that the circumstances in which a child is at risk of pool drowning — they are unsupervised, they fall in unexpectedly — may be quite different than even a “fall in pool simulation.” So, again, suggestive and encouraging but not conclusive.
You might think that there were millions of studies of this but, in fact, these two are the main ones. You can read a longer review here but there aren’t significant additional studies to speak of. The AAP recommends swim lessons for children 1 to 4, based largely on these two studies, and also strongly recommends them for older children for less well stated reasons.
I’ll summarize at the end, but two notes first.
Children Under 1
The AAP does not recommend lessons for children under 1. The reason is that children this age cannot learn to swim. It’s also true that this age group is much less likely to wander off and fall into the pool. And, if they did, they would be unlikely to have the presence of mind to swim to the edge.
To be clear — it’s fine to bring your baby into the pool (if you’re holding them) and they might like the water. But starting formal safety-oriented swim lessons before this age isn’t likely to be very helpful.
Dry Drowning
Every year there is a little bit of panic around the idea of “dry drowning”, which refers to a situation in which a child has (I’m summarizing) a delayed reaction to inhaling water. They seem okay, and then are not later. This is also called “secondary drowning”.
And, every year, doctors have to reassure parents not to freak out about this. Yes, kids can have a delayed reaction to water inhalation. But! First: it is very rare — only 1 to 2 percent of drowning cases. And, second, it’s accompanied by symptoms you’d notice and worry about — persistent coughing, shortness of breath, fatigue, vomiting. If you notice this after a risky water exposure, even a few hours after, you should react. But, again, very rare.
The Bottom Line
One of the review articles on the question of swim lessons rates the evidence as a “B”, meaning it’s inconsistent and not very statistically compelling. I do not disagree, by and large, although I think there is an underlying logic to the idea that knowing how to swim could, in certain situations, matter. Also, the evidence on swim lessons shows that young kids can learn to swim. Since swimming is fun, this is one reason to do it.
Do you need to put your six month old in swim safety class? No. If the three year old camp has swim lessons is that a good thing? Perhaps yes.
However: I think the most important bottom line here is that swim lessons are by no means a panacea. Even if your 3 year old’s swim instructor says they are the next Katie Ledecky they should not be left unsupervised with access to a pool. Even if your nine year old is a good swimmer, they should still not be swimming alone, or at least not without someone knowing where they are.
Pools are awesome and fun and I love them. But they are also one of the comparatively few things I do worry about.
