Well, probably not. I think I spent more time outdoors than did the kids in this interventional study in China. This is an interesting study, both from the scientific question of whether outdoor time can prevent/delay development of myopia in children, as well as from consideration of how the study design might introduce unintended bias in the outcomes.
Source: He M, Xiang F, Zeng Y, et al. Effect of time spent outdoors at school on the development of myopia among children in China: a randomized clinical trial. JAMA. 2015:314(11):1142-1148; doi:10.1001/jama.2015.10803. See AAP Grand Rounds commentary by Dr. Carrie Phillipi (subscription required).
PICO Question: Among school-aged children, can increased time outdoors prevent the development of myopia as compared to controls?
Question type: Intervention
Study design: Cluster randomized controlled
The authors of this study in Guangzhou, China, randomized 6-year-old children (in a cluster pattern of schools) to receive either no intervention, or an extra 40-minute mandatory outdoor activity class at the end of the school day. They measured visual outcomes, including refractive examinations in a subset of children, and concluded that the extra outdoor activity group had a lower cumulative (3-year period) incidence rate of myopia than did controls: 30.4% versus 39.5%, which for you EBM geeks is a number needed to treat of about 11 (pretty good, but not as robust as expected based on other studies). The authors did their best to ensure that the outdoor activities actually happened, that children had close follow up, and teachers and students completed outdoor activity reports and diaries.
It gets more interesting when you think about how cluster randomization might alter study outcomes. Simply stated (and the link in the previous sentence explains this in better detail), analyzing data based on clusters invokes a different set of statistical analyses than does analyzing individual study subjects. One of the problems is that individuals within a cluster may not behave completely independently, which is an assumption of standard statistical testing. By virtue of simply being in the same cluster, the individuals may have more factors in common with one another, and thus the same interventions may have different impacts on different clusters.
I think there is a clue to this in Table 1 in the current study. It simply lists a comparison of 13 characteristics like age, gender, and visual exams of the intervention versus the control group. Normally, in a randomized study, any differences between the 2 groups seen at baseline is by definition due to chance alone; that's what randomization does. In fact, one usually doesn't report p values for the baseline data differences, because it is irrelevant. However, in this case the authors and/or JAMA's editors wanted to report p values here, 3 of the 13 characteristics had differences yielding p values of 0.05 or less. That raises the possibility that randomization by cluster, in this case, might not have been ideal.
I was also intrigued by what mechanism could possibly account for a benefit of outdoor activity on myopia incidence. It turns out that there is some support for this hypothesis, from a few other trials as well as from biologic conjecture that exposure to outdoor light might favorably alter retinal neurotransmitters, or just that better light decreases near activities. Very importantly, it isn't known how long any true benefits of delayed myopia development will last.
So, this study fills yet another lacuna in my knowledge base, namely a myopia prevention strategy that I didn't know existed. The study itself hasn't convinced me, but it will be most interesting to follow this research thread in the future.
And, since my own visual problems relate to astigmatism, I don't think I missed anything even if I did indeed waste some of my childhood daylight hours indoors!
