Can Puberty Affect the Development of Eating Disorders?

Eating disorders typically begin in adolescence. One common explanation for this is that during adolescence females are increasingly exposed to the media, thin models, and dieting. While this is probably true to some extent, it doesn’t explain why the rates of eating disorders are quite low despite the high levels of exposure to thin models in the media. Out of 100 girls, only a handful develop eating disorders, yet all of them are exposed to the same magazines and TV shows.

This means there must be some other factors that differ between this group of girls. One hypothesis is that hormonal changes during puberty may modulate the genetic risk factors for eating disorders. These changes may “turn on” genes that predispose individuals to eating disorders. Previous research has shown that genetic factors modulate disordered eating (eating disorders have a high heritability), but how? What are the mechanisms of this modulation?

Exploring this idea, Dr. Kelly Klump and colleagues sought to focus on the role of estradiol–the predominant estrogen during reproductive years in females. Estradiol (and other hormones, such as progesterone) becomes really important during puberty in females. Moreover, it has also been shown to modulate appetite and binge eating in animals and humans (Asarian & Geary, 2006; Edler, 2007; Klump et al, 2008). Finally, estradiol has also been shown to regulate the turning “on” and “off” of some genes in the brain(Ostlund, 2003).

Together, this suggests that estradiol might modulate the genetic influences on eating disorders during puberty. In order to test this idea, Klump and colleagues recruited 99 pairs of same-sex female twins between the ages of 10 and 15. Forty-nine were monozygotic twins (MZ) who share ~100% of their genetic make-up and fifty were dizygotic (DZ) who share roughly 50% of theirs.

Since there were relatively few twin pairs, Klump et al split up the sample into two groups: low estradiol and high estradiol.

Eating attitudes and behaviours were assessed using the Minnesota Eating Behaviours Survey (MEBS). MEBS assessed weight preoccupation, body dissatisfaction, binge eating, and compensatory behaviours. Physical changes of puberty were assessed using the Pubertal Development Scale, and estradiol was assessed using salivary samples.

There were no significant differences in demographic characteristics between samples with regards to ethnicity and parental incomes, but the group with high estradiol levels were older (as expected, since they were in the later stages of puberty. The two groups didn’t differ in BMI and have a wide range of disordered eating attitudes (which is a good thing, as it enables the researches to measure the effect of estradiol on these differences).

There were also no differences in the MEBS scores between the low estradiol and high estradiol groups. At first, this might seem to suggest that estradiol doesn’t modulate the genetic influences on disordered eating. But actually, this is not the case.

Previous research has shown that pre-puberty, genetic influences have almost no effect on the variability of disordered eating, but after puberty, genetics plays a significant role in explaining the variability of disordered eating among individuals.

To say this in another way, the differences in disordered eating between two 8-year-olds are likely due to shared or non-shared environmental factors (literally, environmental factors they share and don’t share together, such as family life, friends, schooling, any life experiences ) but during adolescence, the differences in disordered eating between two 15-year-olds can be partly explained by the differences in their genetic make-up.

So, what did Klump and colleagues find?


  • In the low (young) estradiol group, the MZ and DZ twin correlations (between estradiol and MEBS scores) were approximately equal. This suggests that genetic effects are unlikely to contribute to the differences in disordered eating behaviours, as one would expect higher correlations in the MZ group versus the DZ group (because the former shares ~100% of the genetic make-up and the latter only ~50%). This suggests that genetic differences do not affect disordered eating in twins with low estradiol.
  • In the high (older) estradiol group, the MZ twin correlations were more than double the DZ twin correlations (between estradiol and MEBS scores). This suggests that genetics differences do affect disordered eating in twins with high estradiol. 

This means the differences in disordered eating that are observed in the younger (low estradiol) group are likely due to environmental factors (like different family environments, as one example), but in the older (high estradiol) group, these differences might be partly (not fully) explained by differences in estradiol.

Findings provide preliminary support for our hypothesis that levels of estradiol moderate genetic effects on several types of disordered eating attitudes and behaviors. Using a pilot sample of twins, disordered eating exhibited little-to-no genetic influence in twins with low estradiol levels during puberty, but moderate-to-substantial genetic effects in twins with higher estradiol levels. These effects were present even when controlling for age, BMI, and the physical changes of puberty.

This study is important because we’ve known for a while that eating disorders are highly heritable disorders,and  this shows that part of this heritability might be due to different levels of estradiol during puberty.

This is further interesting because it also provides a clue as to why many more women suffer from eating disorders than men. (And men do suffer from eating disorders.) As I mentioned in the beginning, the common explanation for this is that men are not pressured to be thin, whereas women are. While I don’t want to completely discount that idea, as it probably holds some (though in my opinion, limited) truth, there might be an additional factor that contributes to the gender differences in the prevalence of eating disorders: estradiol.

How can estradiol modulate genetic effects that may predispose some individuals to develop an eating disorder? Although estradiol is mostly known for its role in reproduction, it also has a critical function during development on gene regulation in the central nervous system. This regulation (or turning genes “on” and “off”) may lead to changes in neurotransmitters and neuronal circuitry.

Animal data show that genes and gonadal hormones [such as estradiol] influence the pattern and timing of changes in brain structure and function during puberty (Ahmed et al., 2008, Nunez et al., 2002, Primus and Kellogg, 1991, Spear, 2000, Zehr et al., 2006). Emerging data suggest that these same effects may be present in humans, as levels of estradiol are inversely correlated with changes in neural structures in girls during puberty, and these neural changes appear to be influenced by genetic factors (Lenroot et al., 2009, Peper et al., 2009a, Peper et al., 2009b). Taken together, data across disparate sources suggest that increases in estradiol during puberty may increase genetic influences on disordered eating through differential organization of neural circuitry underlying risk.

Of course, more research is needed. Future studies should examine larger groups, if possible, and also obtain multiple measures of estradiol (in this study, estradiol was only measured once). Moreover, future research can also measure other hormones and whether they have modulating effects on the genetic influences on disordered eating. Finally, eating attitudes and behaviours be assessed using a different scale. These studies will be important to determine whether findings in this study are replicable.

Most importantly, if these findings are replicable, it will be interesting and important to find out what genes estradiol regulates (switches “on” and “off”), and precisely how this regulation might modulate disordered eating.


Klump, K.L., Keel, P.K., Sisk, C., & Burt, S.A. (2010). Preliminary evidence that estradiol moderates genetic influences on disordered eating attitudes and behaviors during puberty. Psychological Medicine, 40 (10), 1745-53 PMID: 20059800


Tetyana is the creator and manager of the blog.


  1. Ok, I have to comment (and maybe I’ve missed something) – but the higher estrogen group was EXCLUSIVELY older? This could be a confound simply based on age (though they mentioned they “controlled for this statistically” – but I’m assuming as you stated the groups stratified by estrogen levels still had a significant age difference?

    I mean the high estrogen group was farther along in the developmental process, but one cannot state that the difference between groups was SOLELY estrogen status or pubertal level. . .if I sampled a similar group of 8 year olds experiencing precocious puberty – would there be similar risk? This would more strongly support the ED-estrogen link (although that would be a difficult sample to recruit).

    In addition, if I wanted to rule out simply age and puberty as factors influencing ED development (ie. implicate estrogen), it would have also been fantastic to follow a similar group of male boys – if the trigger is non-gender specific and puberty specific, we should see the same effect. . .

    However, the relation between estrogen and b/p is interesting because small children/teens early in their ED are less likely to display b/p behaviour. . .

    I dunno, just some quickly written down thoughts!

    • Hi A:)

      So, the mean age for the low estradiol group was 11.7 and for the high estradiol it was 12.23 with standard deviation between 1.3-1.4 for both, so there was definitely an age overlap.

      I think what you are missing (or what I didn’t make clear enough) is that they are not comparing low estradiol versus high estradiol groups. They are comparing the correlation between estradiol and MEBS scores in MZ and DZ twins in low estradiol group and the same correlation in the high estradiol group. It is a twin study looking at how estradiol modulates the heritability of disordered eating, not how estradiol itself modulates it.

      Essentially, in the low estradiol group, there were no differences between the levels of estradiol and MEBS scores in MZ or DZ twins. This suggests low genetic effects.

      In the high estradiol group, there was a much higher correlation with estradiol and MEBS scores in the MZ twins (who share roughly ~100% of their DNA) verus DZ twins (who share ~50%). This suggests that estradiol modulates the genetic effects. I think it is a confusing study at first, maybe looking at the original paper ( will help.

      Since previous studies have shown that heritability of disordered eating pre-puberty is almost zero, and during early adolescents/early puberty, it rises up to 50%, what they are suggesting with these data is that estradiol modulates this heritability.

      Hopefully that makes sense.

      The next thing would be to do a study on male twins, and I’m not sure if that’s in progress or not (this paper was published in 2010).


Comments are closed.