Unbelievably bad reporting on top of weak studies. From the article:

"...showed that young men who smoked just before puberty produced sons who were more likely to be overweight, beginning in adolescence."

So I looked at the study. They "found" that fathers who began smoking at very young age produced sons (but not daughters) with a higher BMI at age 9 (but not at 7). Note the fine distinctions, all from a sample size of just 330 such fathers. This reeks of a Munchhausen Grid.

But it gets better: the difference is between a BMI of 17.23 and one of 18.15. What an enormous difference! And both values are considered underweight.

So, guys, better start smoking before you're ten years old, or you risk siring underweight sons!