This study provides evidence that sleeping with night-lights doesn’t cause nearsightedness in infants. But the author never assumed that it did. The conclusion is just that if there’s a causal relationship, that effect disappears with age.

This helps eliminate the possibility that children who are already nearsighted might sleep with night-lights at a higher rate than those who aren’t nearsighted. But this doesn’t relate to disappearance with age.

This helps show that sleeping with night-lights as an older child does not cause nearsightedness. But these children didn’t sleep with night-lights as infants. So it doesn’t help examine whether effects disappear as a child gets older.

This weakens by reducing the reliability of the two studies. If they didn’t examine enough children to provide significant support for a conclusion about cause, then the author can’t rely on them to conclude that the effect of night-lights disappears with age.

That “several” of the children were nearsighted does not establish a correlation between nearsightedness and night-lights in the study. It’s possible the almost all who slept with night-lights didn’t have nearsightedness, even if several did.