No, but I'm trying to explain the sampling bias, not talking about growing up obese. If you randomly sample obese people at any age, 18 or 81, your sample will

No, but I'm trying to explain the sampling bias, not talking about growing up obese. If you randomly sample obese people at any age, 18 or 81, your sample will still be biased by years-spent-obese. There probably is some effect from growing up obese, but the sampling bias is hugely relevant to interpreting the statistics, and it's not talked about much.