LETTER TO THE EDITOR |
We were a mite surprised by the letter from Dr Reich stating that our article on the respiratory effects of infant swimming1 had a "number of epidemiologic and statistical problems" and had been published mainly for "political rather than scientific purposes." At the reading of Reich's letter, it seems, however, that the "number of epidemiologic and statistical problems" is reduced to 2 points. First, he wonders why the risk of a positive exercise-induced bronchoconstriction (EIB) test was adjusted for parental smoking but not the risk of doctor-diagnosed asthma (DDA) or of DDA and/or EIB test. Actually, if we did so, it was simply because we adjusted the outcomes for predictors entering in the model with a P value of <.1. In the case of parental smoking, this level of statistical significance was reached for the EIB test (P = .08) but not for DDA (P = .82) or DDA and/or EIB test (P = .41). The second point raised by Reich is the lack of swimming information for those children who had not swum when they were infants. Yet, this information was indicated in our Table 1, giving for the infant swimming group and their controls the lifetime cumulative attendance of indoor chlorinated pools (in hours) as well as the proportions of children who had access to a backyard pool. The materials and methods section of our article clearly referred to another article2 for a full description of the protocol of our study. In that article,2 it was explained that the non–infant swimming group had learned swimming at school or with their parents. Thus, the comparison in our study was exactly that suggested by Reich (ie, between children having learned to swim before the age of 2 years and children who learned swimming later, at school or with their parents).
We can understand that Reich has revised his opinion regarding infant swimming after having seen his niece almost drown. However, as scientists, we should refrain from basing our judgment on our personal experience or on some anecdotal observations. We are committed to support our analyses and conclusions with sound scientific evidence. Of course, as Reich says, there is no need of evidence to state that the ability of swimming reduces the risk of drowning and that the earlier a child can swim the better it is for his or her safety. By contrast, it is not self-evident that to give swimming lessons to an infant who is not developmentally ready is an effective measure for reducing the risks of drowning, knowing that there are no better preventive measures than pool-fencing and parental supervision.
Although based on a relatively small population (43 infant swimmers, not 25 as stated by Reich), our study is consistent with the mounting epidemiologic evidence that some chlorination products can be detrimental to the airways of pool attendees (for review see ref 3). Looking at some recent findings, clearly there remains much to explore in this area. For instance, a link between attendance at chlorinated pools during childhood and the risk of hay fever later in life was found in a recent German study.4 In a new study,5 we also found that the risk of asthma is actually not limited to indoor pools but also concerns outdoor pools. Given these emerging risks and the well-known vulnerability of infants to chemical stressors, we can only recommend parents to choose for their children well-managed swimming pools with low levels of chlorination products in the water and air.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||