There is a longitudinal study in Sweden that tracked a couple hundred thousand girls/women from childhood to old age (multigenerational, following every 10 years... this is why I LOVE soc.med.country's studies... a "big & long" study in the US is 3 years with 6000 people in it, vs hundreds of thousands over decades).
The study was notable for 3 things*:
1) Girls about to hit puberty gain between (apx) 15-30lbs on average
2) Those who diet or exercise to lose that weight are typically 15-30lbs overweight/ struggle with their weight ***until menopause***
3) Those that do NOT diet or exercise to lose that weight, it melts off with no effort in about a year and do NOT struggle with their weight until AFTER menopause
It's a pretty mindblowing piece of research, especially from a sociological perspective (medically it's just common sense, the body is prepping for childbearing years and if it's 'starved' it tries to create a safety net assuming the person is living in an area with famine/food supply issues, and if it's not is assumes that the person is living in a bountiful area where food for itself and unborn children will be plentiful). Anyhow, sociologically... the boomers were the FIRST group of adolescents to wear FORM FITTING CLOTHES before apx 16. Girls prior to the boomers in the 50's were kept in smock type dresses. Boomers were the first ADOLESCENTS to restrict their eating/try to 'have a figure' to fit into those dresses.
Anyhow... it's a really cool, really interesting thing to think about, for setting up for life long (well, the next 40 years of fertile years, in any event) healthy weight
* The study is in swedish, and one of my professors (a fertility specialist) brought it in a translation for us to study for our midterms, and then again it was used in our childhood nutrition class. It's on the peer review database (in swedish) but I don't have my english version handy, so my numbers aren't 100% accurate, but approximate. I've used the actual study on here before, but that was before my basement flood, and MP doesn't let one search old posts very easily :P The weight may be 10-15, or 20-30... but I went with the more midline option from my memory to be 'safe'. There's actually a series of weights listed in the study with both the mean and average and I just can't pull the exact numbers right now, so I just wanted to make sure this caveat was in there. It's only a few pounds in either direction, but you get the idea.)