More Sleep = Lower BMI

I read an article today from the New York Times, about a study that "suggests that adolescent obesity could be decreased if teenagers got more sleep, and the heaviest would benefit most."
The study, published in the journal Pediatrics, took 1,429 ninth-graders from suburban Philadelphia high schools, and followed up on them every six months for four years. The students' heights, weights, and hours of sleep were all self-reported at each follow-up. The researchers controlled for physical activity, sex, race, and socioeconomic status, but did not collect data about caloric intake.
They found a negative correlation between hours of sleep and BMI, showing that each additional hour of sleep was associated with a lower BMI, ranging from a 0.04 to a 0.28 difference. The largest difference occurred in kids with BMI's over the 90th percentile.
The authors extrapolated their data and predicted that an increase in teens' average sleep from 8 hours to 10 hours would result in a 4% reduction in US teen obesity rates (or 500,000 fewer overweight teens).
I'm wary of all the self-reporting done in this study, especially of weight, and suspect that the weights are often being reported as lower than they really are. The study being done every six months is also somewhat odd considering that most teenagers sleep extremely differently during the school year and during summer break, which takes up almost three months of the year. In order to trust this study more, I would hope that the teens had their heights and weights recorded by either the researchers or their doctors; that their sleep was electronically monitored in some way; and that steps were taken to separate consistent sleep schedules and varied sleep schedules, which could have some kind of effect as well.

NY Times Article: http://well.blogs.nytimes.com/2013/04/10/less-sleep-more-weight/?ref=health
Original Study: http://pediatrics.aappublications.org/content/early/2013/04/03/peds.2012-2368.full.pdf+html