I consider myself a pretty skeptical person, but I have to confess I accepted at face value claims that girls were reaching puberty. Surely such a claim, based on studies published in some relatively prestigious journals, had a lot of data behind them. Most of the debate in the media over this claim was what was causing early puberty. Environmentalists suggested manmade chemicals as a possible cause, some doctors suggested the high levels of obesity in children, while some feminists and conservatives suggested that the high level of sexual images in contemporary media might somehow be responsible.
It turns out, however, that the alleged evidence for early puberty is severely flawed and the quick acceptance of early puberty may be preventing some girls from obtaining the best health care available.
The New York Times‘ Gina Kolata, who has helped debunk other junk science issues in the past, reported on the debate over early puberty this week. The evidence for the phenomenon is appallingly absent. The claim is based entirely on a single study by Dr. Marcia E. Herman-Giddens that was published in Pediatrics in 1997. Herman-Giddens study looked at 17,077 girls aged 3 to 12. The girls had gone to a pediatricians office from July 1992 through September 1993. Herman-Giddens managed to persuade 225 health care providers at 65 private practices in the United States to evaluate any girls they saw for early onset of puberty.
As she reported in her 1997 article, of that sample population, Black girls began developing breasts at 8.87 years while white girls began developing breasts at 9.96 years. Many pediatrics textbooks, taking these numbers at face value, pushed back the lower end of normal puberty to age 6 for black girls and age 7 for white girls. But did they have a good basis for doing so? No.
The first major problem with the study is that the sample is incredibly biased. By focusing exclusively on girls who were brought to see a pediatrician, it is possible that Herman-Giddens is simply measuring the rate of premature puberty brought on by other health problems. After all, parents who are worried that their daughters are experiencing puberty early are more likely to bring their children to see a pediatrician about that problem. In statistics this is called a selection bias.
The second major problem is that the only marker of puberty for which reliable data exists for a long period of time — the onsent of menstruation — has remained steady for the last 50 years. There simply is no reliable long-term data on the average age that breasts start developing, but there is reliable long-term data for menstruation. That age has remained stable, even in Herman-Giddens study, at 12 years and 8 months for white girls and a bit earlier for black girls. If puberty is starting earlier and earlier it is very odd that the average age of menstruation has remained so consistent.
A group of pediatric endocrinologists wrote a letter to Pediatrics recently lamenting that doctors who believe this data are giving incorrect health advice to young girls. As they note, early puberty can be the result of serious health problems, such as brain tumors and other disorders, but parents who bring in girls who are experiencing early puberty are simply told that this is normal and no further tests or investigation are warranted. As Dr. Laura K. Bachrach told The New York Times, “You don’t just dismiss a child in that age range [under 7]. I think it is potentially medically dangerous.”
Doubters fault theory finding earlier puberty. Gina Kolata, The New York Times, February 20, 2001.