Issue: June 2015
June 17, 2015
14 min read

Decreasing age of puberty may lead to poorer health for a generation of women

Issue: June 2015
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

It is not just parents who look at girls today and think they grow up much faster than they did a generation ago. For 20 years, researchers and pediatric specialists have suspected that many normal American girls are entering puberty when younger than 8 years, the age before which the medical community has considered breast budding in girls to be precocious.

Last fall, publication of The New Puberty by pediatric endocrinologist Louise Greenspan, MD, and psychologist Julianna Deardorff, PhD, brought this suspicion — now backed by research — to the general public. Aimed at parents, the book offered families practical strategies for supporting girls entering this socially and psychologically vulnerable life stage. The authors emphasized that, despite several articles in the popular press conflating the trend toward earlier puberty with examples of extremely young girls with pronounced breast development or menarche, the new puberty is not that. Although puberty at a younger age raises some concerns, it is not necessarily bad.

“If one went back 100 years, one would find that there’s an awful lot of data where the age of menarche was diminishing, for example, and that was always perceived as a positive thing, due to better health and nutrition and better sanitation,” Mark Palmert, MD, PhD, division head for endocrinology at the University of Toronto’s Hospital for Sick Children, told Endocrine Today. “At some inflection point, we as a community began to say that’s too early. … But how does society define too early?”

Most children with early puberty on the cusp of the normal range do not have causative underlying medical problems. For these otherwise normal children, neither Palmert nor Greenspan, an associate clinical professor at the University of California, San Francisco and a physician at Kaiser Permanente, advocates viewing early development as a medical problem for the child. Rather, known health implications of early puberty appear to occur later in life with increased risks for breast cancer, type 2 diabetes and heart disease.

Louise Greenspan, MD, cautions that in an otherwise healthy girl, medical treatment for early normal puberty is not likely to be necessary but psychological support may be helpful.

Photo by  Robert M. Ortner


Effect of race, ethnicity

Experts agree that normal puberty now appears at an earlier age in girls in the United States and Europe. A 1997 study conducted by Herman-Giddens and colleagues in the Pediatric Research in Office Settings (PROS) network was the first to raise concern about pubertal timing in girls and to reset the line defining early and early-normal development.

“That was the pivotal study that woke everyone up and changed perceptions,” Mitchell Geffner, MD, immediate past president of the Pediatric Endocrine Society, told Endocrine Today.

The study involved 17,077 girls aged 3 to 12 years — 9.6% black and 90.4% white — presenting at pediatrician offices across the United States. By age 8 years, 48.3% of black girls and 14.7% of white girls were showing breast and/or pubic hair development, with mean age of onset of breast development at 8.87 years and menarche at 12.16 years for black girls and 9.96 and 12.88 years, respectively, for white girls. The researchers concluded, “Practitioners may need to revise their criteria for referral of girls with precocious puberty, with attention to racial differences.”

Until that study, pediatricians considered any pubertal signs in girls before age 8 years to be too early and in need of medical explanation, Geffner said.

Herman-Giddens’ statistical analysis changed the cut point that separated early from early-normal puberty to 7 years for white girls and 6 years for black girls, according to Geffner. “When you translate that in the real world, that’s the first grade, which is a rather shocking way to look at it. So if you had small breast buds or a little bit of pubic hair and you were 6 [years old] and you were African American, that was statistically defined as normal,” Geffner said.


Since that initial report, several studies have confirmed its results while attempting to address questions about its methodology and to extend findings to other populations.

A longitudinal cohort study conducted by Biro and colleagues, including Greenspan, attempted to clarify the effect of BMI and race/ethnicity in declining age at puberty. The investigators enrolled more than 1,200 girls aged 6 to 8 years in the San Francisco Bay area, greater Cincinnati and New York City. They confirmed that ages of Tanner breast stage 2 onset varied by race/ethnicity for white, black, Hispanic and Asian girls. They also found that girls with higher BMI reached this stage at a younger age and that higher BMI was the strongest predictor of breast budding.

Mitchell Geffner


“Obesity is associated with earlier thelarche,” Greenspan told Endocrine Today. “In the past, some of the studies were done by visual inspection, so there was some question as to whether it was true breast tissue.” This study controlled for obesity by using palpation to assess for breast tissue and using two testers in girls with BMI greater than the 85th percentile.

“Obesity is definitely associated with early puberty, but it’s not the only thing,” Greenspan said. “I think it was a factor about 80% of the time, but we still have 20% that’s coming from another cause.”

Several factors besides adiposity are being explored as influencers for the declining age of puberty, including endocrine-disrupting chemicals, sugar consumption, family stress, fetal environment, size at birth and maternal characteristics, such as pre-pregnancy BMI and gestational weight gain, among others.

Precocious vs. early normal

This shift toward earlier thelarche has not meant that pediatric endocrinologists are seeing more troubling cases of precocious puberty at younger and younger ages, according to Palmert.

“We’re not seeing large numbers of these precocious kids, we’re just seeing more kids having development at the tail of the normal range. That’s an important distinction because those [very young] kids are rare to start with.”

For clinicians, the cutoff for normal puberty matters because any development of secondary sexual characteristics under that age will need to be accounted for, according to Palmert. “I don’t know that there’s a universal definition anymore,” he said.

Geffner agrees that physicians cannot ignore significant pubertal development at the margins of the new early normal. “The trick is that while it’s probably normal most of the time, in my practice, if I see a girl who’s 6 or 7 who has a lot of puberty, not just a pubic hair or a little breast bud, that could still have a significant underlying reason,” he said. “So one has to use clinical acumen and order lab tests as appropriate and be careful that not just because some magical age number has been met that means it has to be normal.”

Health implications of early puberty

Apart from medical problems that might initiate the early appearance of secondary sexual characteristics, earlier age at puberty has increasingly been associated with health concerns later in life. The earliest to arise — and one cited as a reason for delaying pubertal onset — is short stature, which has a clearly understood mechanism.

According to Geffner, a child’s adult height can be predicted to some extent based on her past growth pattern, current age and bone maturation, as determined by X-ray. A girl undergoing puberty produces hormones that cause the bone epiphyses and growth plates to fuse, and her adult height will depend on how tall she was when she began producing those hormones.

“Girls, for example, will complete their height growth by bone age 15 and for the most part by bone age 13, which normally occurs when they’re 15 and 13 [years old], respectively,” Geffner said. “But if an 8-year-old has a bone age of 13, she’s not going to grow much anymore.”


Such a girl might be tall at the moment because she had a growth spurt brought on by puberty, but that spurt will end and she will most likely end up shorter than her adult contemporaries.

Geffner states that short stature might be a good reason for pausing pubertal development to allow the girl to catch up to her bone age. However, Palmert warns that evidence is scarce that this works in older children. “Under age 6 for sure, 6 to 8 questionable, after 8 probably not. So if you’re looking at a 4-year-old, if you intervene, that child will most likely have a better adult height. If you’re looking at a 7-year-old and you intervene, it’s a question mark,” he said.

Risk for breast cancer

Other conditions have murkier ties to early puberty, including breast and other reproductive cancers, cardiovascular disease and type 2 diabetes.

According to Greenspan, estrogen may be the mediator between early puberty and increased risk for breast cancer, and thus thelarche rather than menarche is likely to be the key moment. However, most studies that rely on recall use menarche as a surrogate for thelarche because women have a better memory for when they experienced their first period than when they first developed breast buds.

“Theoretically, the earlier you are at menarche the earlier you were at thelarche, and your body has been exposed to [increased] estrogen [levels] for longer,” Greenspan said. “The idea is that it’s lifetime exposure to estrogen, and the earlier you are when you reach thelarche, that’s a reflection of more exposure to estrogen.”

Although the age of pubertal onset is clearly decreasing, ongoing studies are finding evidence that the age of menarche appears to be falling less rapidly, Greenspan said, which could mean that a crucial period of breast development is longer, allowing greater exposure to cancer-causing agents.

“The concern would be that then the tempo of puberty has changed, that the tempo of puberty might be longer ... and the concern about having a longer tempo of puberty is that your window of susceptibility to environmental agents is longer or your window of susceptibility to a lot of things endogenous and exogenous is longer,” he said. “So instead of having 2 to 3 years of massive breast growth, for example, you might have 3 to 4 years.”

Risk for cardiometabolic issues

Estrogen exposure plays a role in breast cancer development, but women’s reproductive hormones have been considered protective against CVD. However, early menarche appears to increase risk for coronary heart disease, stroke and hypertension later in life.

Dexter Canoy


Dexter Canoy, MD, PhD, a cardiovascular epidemiologist in the Nuffield department of population health at the University of Oxford, and colleagues analyzed data from the Million Women Study, which enrolled 1.2 million women in England and Scotland. The women who experienced menarche by age 10 years or younger had a higher RR for CHD than those with menarche at age 13 years (adjusted RR = 1.27; 95% CI, 1.22-1.31). Late menarche, at age 17 years or older, had an RR of 1.23 (95% CI, 1.16-1.3). The same relationships were seen for early and late menarche with cerebrovascular disease and hypertension, although these risks were smaller.

“Although there has been no single reason why we would suspect early ages in menarche would be linked to CVD later in life, there have been a number of observations suggesting that those with early menarche seem to be overweight or obese and have other less favorable CV risk profiles, such as having high blood pressure or cholesterol even during young ages,” Canoy told Endocrine Today.

Earlier age at menarche appears to raise the risk for developing type 2 diabetes later in life, as well. Although excess body weight has been associated with both early pubertal onset and diabetes, BMI does not entirely explain the connection between the two.


“It is not an obvious link. Many people will be surprised at such a long-term disease association,” Ken K. Ong, FRCPCH, PhD, head of the medical research council (MRC) growth and development program at the University of Cambridge, told Endocrine Today. “They may understand that overweight and obese girls are more likely to have early puberty and have higher risk of later type 2 diabetes, but may not have linked early puberty directly to later health risks.”

In one large prospective study, Ong and colleagues from other institutions analyzed data from more than 15,000 women across eight European countries. Participants with the earliest age at menarche (8 to 11 years) had a 70% higher incidence of type 2 diabetes than those who reached menarche at 13 years; only 41% of this elevated risk could be attributed to adult BMI. No reduction in diabetes risk was observed in participants with later menarche.

Mark Palmert


“Although avoidance of adult overweight and obesity may attenuate the risk of type 2 diabetes in women with early menarche, our findings suggest that strategies to prevent early menarche may be important in their own right,” the researchers wrote.

“Personally, I don’t think that avoidance of diabetes or other diseases when you are an older adult is an effective motivation for children and adolescents to change their lifestyle or take other preventive actions,” Ong told Endocrine Today. “But it is possible that they may see more visible short-term goals as being more attractive, for example, avoiding early onset of monthly bleeding, and possibly taller adult height due to a longer growing period. If this in turn also benefits later health, then it will be a ‘win-win’ approach.”

When to delay puberty

In the meantime, health care providers and researchers advise girls and parents to control the risk factors they can.

Endocrinologists typically consider medically delaying puberty in some children for two reasons, according to Palmert. The first is to prevent short stature, and the second is psychosocial considerations, such as distress or difficult adjustments to early puberty.

However, Palmert said each case is different, but he would intervene medically in early-normal puberty, as opposed to true precocious puberty, only in an extraordinary situation, such as a young girl with neurologic compromise. In cases where a girl is entering puberty at a very young age, Palmert will discuss options with the family, asking, “How much of this can we handle by you handling this at home as a family and a parent through education and anticipatory guidance and being aware of risks, such as problematic peer interactions and increased risk taking behaviors, and working to mitigate these vs. using medication?”

Greenspan agrees that in an otherwise normal girl, medical treatment is not likely to be necessary. But she cautioned, “Not treating medically doesn’t mean you’re going to ignore the problem. … Pediatric endocrinologists owe it to their patients to be a resource in the psychological aspects of supporting these girls.” – by Jill Rollet

  • Biro FM, et al. Pediatrics. 2013;doi:10.1542/peds.2012-3773.
  • Breast Cancer and the Environment Research Program website. Available at: Accessed May 10, 2015.
  • Canoy D, et al. Circulation. 2014;doi:10.1161/CIRCULATIONAHA.114.010070.
  • Elks CE, et al. Diabetes Care. 2013;doi:10.2337/dc13-0446.
  • Greenspan L, Deardorff J. The New Puberty. Emmaus, Pennsylvania: Rodale Books; 2014.
  • Herman-Giddens ME, et al. Pediatrics. 1997;99:505-512.
For more information:
  • Dexter Canoy, MD, PhD, can be reached at the Nuffield Department of Population Health, University of Oxford, Old Road Campus, Oxford OX3 7LF, UK; email:
  • Mitchell Geffner, MD, can be reached at Children’s Hospital Los Angeles, 4650 Sunset Blvd., Los Angeles, CA 90027; email:
  • Louise Greenspan, MD, can be reached at Kaiser Permanente Medical Center, 2200 O’Farrell St., San Francisco, CA 94115; email:
  • Ken K. Ong, FRCPCH, PhD, can be reached at the University of Cambridge School of Clinical Medicine, Box 285, Institute of Metabolic Science, Cambridge Biomedical Campus, Cambridge CB2 0QQ, United Kingdom; email:
  • Mark Palmert, MD, PhD, can be reached at the Hospital for Sick Children, 555 University Ave., Toronto, Ontario, Canada M5G 1X8; email:
  • Disclosures: Geffner reports serving on an advisory board for Endo Pharmaceuticals. Canoy, Greenspan, Ong and Palmert report no relevant financial disclosures.


What is a larger factor in early puberty: environment or genetics?


About half of the population variability in menarche timing is thought to be attributable to genetic factors.

Complex traits such as pubertal timing have both genetic and environmental components. For these traits, the estimated heritability is around 40% to 70%. However, specifically identified genetic variants for age at menarche currently only explain around 3% of the observed population variation in this trait.

There is evidence that the genetic contribution to menarche is co-inherited with genetic factors implicated in obesity. For example, the heritable component implicated in menarche timing estimated from studies of related individuals is thought to overlap with the heritable component of BMI. This is backed up by the fact that a number of genetic variants identified for age at menarche are also associated with BMI. What is not clear is whether these genetic variants have independent effects on these traits — termed pleiotropy — or if the association with BMI is driven through an effect on developmental timing or vice versa.

Cathy E. Elks


There is currently a lot of focus on a technique called Mendelian randomization, a genetic approach that mimics a randomized controlled trial, to try to establish whether a risk factor has a causal effect on an outcome. The idea is that genetic variants are randomly allocated at conception and can therefore be used as an unbiased instrument for an exposure, such as menarche, and tested for association with later disease outcomes.

There are several challenges with this. First, is that it still might be the case that these genetic variants have independent effects on the exposure and outcome we are interested in rather than through the gene-exposure-outcome pathway. Another challenge in the case of menarche is that our genetic instrument is still currently rather weak — as we can’t explain much of the variability in menarche with known variants — so we need to do more work to improve this, both through bigger genetic studies (with the imminent availability of genetic data from UK BioBank on half a million people this will be possible), and also the identification of rarer genetic variants that might have bigger effects.

Cathy E. Elks, PhD, is a postdoctoral research associate in genetic and epigenetic epidemiology of women’s health at the University of Bristol, United Kingdom. Disclosure: Elks reports no relevant financial disclosures.


Numerous factors in our environment are potentially contributing to pubertal timing, but to pinpoint a specific trigger in any given child is almost impossible.

There are reports of multiple factors associated with precocious puberty, but these are hard to prove because there are no randomized controlled trials isolating one particular factor.

Molly Regelmann


Obesity and postnatal weight gain are certainly associated with the onset of puberty in girls. Data from the National Health and Nutrition Examination Survey III showed that children who gained more weight in the first year of life had a higher risk for obesity and earlier onset of puberty. It is difficult to determine whether the excessive weight gain in infancy or the subsequent obesity that is also associated with rapid weight gain in early life leads to the earlier onset of puberty. The implication is that interventions that address early nutrition and prevention of overfeeding could contribute to lowering the risk for precocious puberty.

Other environmental influences, such as endocrine disrupting chemicals, have also been reported to be associated with precocious puberty. For example, elevation of urinary BPA has been widely reported to be associated with idiopathic precocious puberty. BPA has been associated with obesity, and BPA tends to be found in canned foods and previously in plastic baby bottles. Canned foods tend to be more processed and formula-fed infants tend to gain weight more readily than breast fed infants. It is difficult to determine whether BPA directly or indirectly leads to early pubertal development. Maybe children exposed to BPA in canned foods generally have lower quality foods in their diet than those eating fresh produce and the higher calorie diet leads to obesity and early puberty? The BPA may be an indicator of a poorer diet, higher risk for obesity and therefore higher risk for early puberty but may not be the direct cause of the early puberty. That’s not to say that eliminating BPA from food containers seems like an unreasonable approach. It is equally important to make sure that the chemicals replacing BPA in plastic food containers are safe.

Tea tree oil and lavender oil have also been reported in the literature to be associated with precocious central puberty. Tea tree oils are commonly found in “natural” lice repellent hair products. Anecdotally, I have had experience with two young girls with evidence of breast development after using leave-in tea tree oil products for more than a year. One had resolution of breast development and the other had progression of precocious puberty following cessation of treatment. Can I prove that the tea tree oil caused the premature thelarche and precocious puberty? No, and certainly there are other children getting exposed to the same products daily with no evidence of early pubertal development. It is likely that a confluence of environmental triggers, which may or may not include the tea tree oil products, and a genetic predisposition, lead to any given child experiencing early puberty.

Molly Regelmann, MD, is an assistant professor of pediatrics, endocrinology and diabetes at Mount Sinai Hospital in New York. Disclosure: Regelmann reports no relevant financial disclosures.