Notice that the passage uses both absolute numbers and percentages. It tells us that the absolute number of obese children has been steadily increasing over the past 15 years, but according to the obesity definition they are always in the top 15th percentile, so 85% of children have less body fat than them. The only way it is possible for the absolute number to go up while the percentage stays the same is for the overall number of children to increase.
To illustrate this point let's say in year one we have a population of 200 children, 15% are obese or 30 children, and 85% or 170 children are not. Then in year two we are told there are now 45 obese children, well if the population stayed at 200 children, that would be 20+% rather than 15%, hence if we are told that 45 children are now obese and they still represent 15%, the whole population of children must have gone up - both obese and non-obese groups to 300 - 45 of them obese (15%) and 255 (85%) are not obese. Hence, we can conclude that (C) must be true.
The issue with B is that it attempts to determine the cause of obesity but the obesity numbers alone are insufficient to conclude that obesity rates have been going up due to lack of exercises, considering there might have been other factors in play - poor diet, socioeconomic conditions, genetics etc. Since we cannot infer anything about the drivers of increasing obesity based on numbers alone, B is an improper inference.
Let me know if this makes sense and if you have any further questions.
