Note: Responses below are paraphrased responses from an interview between Victoria Guazzelli Williamson (VGW) and Dr. Anne-Lise Goddings. VGW made her best efforts to take thorough and accurate notes during the interview so that responses could be accessed by developmental scientists interested in pubertal measurement.
Biography: Dr. Anne-Lise Goddings is a physician and a National Institute for Health Research Clinical Lecturer in the Population, Policy & Practice Department of the University College London, Great Ormond Street Institute of Child Health. Following her medical training, she received her Doctorate from the University College London in 2015 focusing on the role of puberty in adolescent brain development. She has over 20 peer-reviewed publications in prestigious journals including The Lancet, Neuroimage, and Journal of Neuroscience.
Q: Why do you study puberty? Why is puberty something scientists are interested in?
A: People who study pediatrics tend to think of it being about small children, so adolescents are often overlooked. I got interested in studying adolescents as they haven’t been given as much focus. I think it is difficult to study the period of adolescence without studying puberty because puberty influences so much of adolescence and it is linked to what adolescents think about and find important, and it also impacts their development. However, as I’ve studied puberty, I’ve realized it is much more complicated to study than we originally thought.
Q: How do researchers measure puberty? Is hormonal or physical measurement better?
A: Measuring puberty is actually quite difficult and often we try to simplify how we measure it. There are hormonal and physical measures.
Physical measures are the original way to measure puberty and involves us finding out how much someone’s body is changing and the advantage is that everyone understands the sort of things that develop in our bodies (i.e., secondary sex characteristics). But on the flip side most people have very little understanding of what actually happens (i.e., all the stages our bodies go through from being child-like to adult-like). We expect young people to tell us about their pubertal development when they are going through it and don’t have much idea of what to expect, and when they haven’t completed it themselves. Culturally, we are becoming more and more body shy which can make it very difficult if you ask a young person about their breast or body hair development. Young people may not know the answers to those questions and may be wondering those same questions themselves.
Physician-informed physical measures can address some of these problems but, culturally, we are moving away from being okay with having doctors examine young people. Thus, it is a much less straightforward test to do. Physician-assessed Tanner staging is thought to be the best, mostly because it is closest to what was originally done and it involves both looking and feeling to assess physical, pubertal development. Breast tissue looks similar to fat tissue for example; so for someone with obesity you cannot tell if it is fat or breast tissue without palpating it. The same goes for boys for checking for the size of testes; it is not directly related to the size of the scrotum.
Hormone measures are an alternative but the problem with hormones is that they only tell you what is happening at that moment. Physical development tells you over months or years whereas hormones only give you a spot number. Additionally, there is huge variation based on time of day, time of month, and there are huge variations between individuals in normative levels of these hormones. Ultimately, we need to look at both hormonal and physical measures to measure puberty.
Q: What are some of the biggest issues with pubertal measurement?
A: The other aspect of hormones that we tend to gloss over is how they interact neurally. Within the brain there are receptors that change hormones, like estrogen and testosterone. So just because we are measuring hormone levels peripherally, we may not be measuring the aspects of puberty that we are actually interested in (if we are interested in brain development). It’s a reasonable assumption that as the hormones are mostly made peripherally (gonads), that changes in hormones in the body are going to impact pubertal hormones in the brain.
One of the biggest challenges for researchers interested in measuring pubertal development has to do with the opinion of young people and their parents. I did a study with girls aged 11-13 where I palpated their breasts to see their breast development. It was very difficult to recruit for this study and most of this challenge came from their parents rather than the girls themselves. The girls seemed to understand the purpose behind it. This may explain why we have moved more to self-report which is very variable in how it relates to physician assessment. Self-report varies based on how you deliver it, and it is less accurate among populations with overweight/obesity, as well as for people from some cultural backgrounds - but it is not clear what differences in reported score are due to, both form young people and physicians.
The Pubertal Development Scale is commonly used and accepted as a tool but a big challenge with it is that for each process you are asking them for their stage in development including whether it has just started, has definitely started or has finished. Asking if development is finished, for someone who has not fully gone through it is very difficult to judge. Having said this, it does seem to correlate moderately well with other measures. All of our puberty measures have a certain amount of error, and it is difficult to compare any of these measures with something that is highly accurate. For instance, puberty always loses to age because age can be measured so accurately.
One of the challenges for measuring hormones via salivary data, especially estradiol, is that it is broken down in saliva at room temperature. So what happens between data collection and analysis is really important. Additionally, low flow rate can impact DHEA-S levels. Even something as innocuous as what you are thinking about when you collect the saliva can change the hormone levels in the sample.
Q: What can we do to address or mitigate issues with generalizability related to puberty?
A: We should use as many of the different tools in combination as we can so that we can work out what is similar and what is different across populations. When using hormone measures, we should use them as well as we can (e.g., repeating sampling across consecutive days and using assay as accurately as possible). Hair and urine can also be used to measure puberty in earlier stages of development. In self-assessment questionnaires, we should think about what it is that each question is asking and be open to the idea that there are lots of different ways to analyze them. For instance, Tanner stage 4 in girls and Tanner stage 5 in boys are so different. They are driven by different hormones and it is unlikely that they have the same effect. At baseline, we should assume that different hormones do different things first and then if we find evidence that they do the same thing, we can adjust our understanding based on that.
I think with different populations, we really need to think about how we present information and provide as much information as we can to young people (if showing them pictures, provide as much information in the pictures as you can). If giving the Pubertal Development Scale, researchers should describe to participants that puberty is a long process that takes 2-3 years. A mini education class about what puberty is and how long it takes and what changes they may or may not have noticed would be helpful in these cases. Additionally, we should work to reassure young people that all of the developmental changes associated with puberty are normal. Honestly, puberty assessments need to come with a little health warning. We need to have the young person think about what they’ve noticed is changing before we have them fill out a form. Unless we have a mini-education session to inform them about these changes being okay and totally normal, then we can increase anxiety for young people and we will be less likely to capture accurate results.
There’s an opportunity to educate young people. If they are going to give us their own version of their development, it is important to tell them why we are asking for this and what is important about their development. If puberty is one of the focuses of what we are doing our research on, then giving it a bit of time and context is really important. Even just saying that “I know that this sketch is obviously not the same body shape as all young people, but ignore that and just focus on this (e.g., for the boy’s picture about hair development, tell them to ignore the size of penis and scrotum and just focus on the hair).” Being quite explicit about what we are trying to measure can get us the best answers we can with the tools we have until we have better tools. The processes and stage of puberty are likely to be the same but how it appears on one’s body based on demographics may make it harder to work out what the right answer is for a researcher.
Q: Tell me about the time you went to the library to check out the Tanner stage book.
A: Dr. Tanner was at University College London where I did my research. University College London has copies of the original book they produced alongside the papers, which essentially has the original photographs of the young people they used for the study. I had to get special clearance to see the book in a special room in the library. I’ve also had my internet searches flagged because of how much I had looked at puberty which is obviously a sensitive topic, especially for images.
We see Tanner staging so often as this concrete thing but when you look at the pictures you see there is a lot of variability and, unsurprisingly, they grouped people into the developmental stages very liberally. All the girls from the original study were caucasian and tended to be underweight. They were typical for a specific population but not for a global population.
Q: What can we do to address these issues of generalizability?
A: We can't throw out the whole thing. What the Tanner stage did pretty accurately was document that there is a progression of pubertal hair and breast development. What is less clear is extrapolating this to other ethnic groups and weight and social groups and the six month development period might not be accurate. There are other scales but they tend to use Tanner stage as a baseline. If you have darker skin, noticing a few darker hairs is more difficult. Weight is an issue for breast development. Boys from different ethnic groups, may have different proportions (e.g., penis size). A kind of summary drawing mightapply to the population it was intended for, but certainly doesn’t account for individual differences in where you are starting and ending. In general, we researchers tend to study girls but we are terrible at studying boys' pubertal development. Documenting the details of physical pubertal development in boys involves an intimate conversation and physical examination and this hasn’t been done in boys as much as girls. On the plus side, testosterone is an easier hormone to measure.
The Tanner staging gave us the baseline knowledge that puberty is a process that goes through various stages, but does not allow for individual differences. Translating physician ratings is also challenging. Technology can allow us to assess physical puberty with more than a four-step scale. One way around it could be to give young people a picture that looks like them (which will make it more accessible and approachable to them), and then have them assess the pubertal development on top of that. Another aspect of it is that we know the final product of this group of children is not necessarily accurate. A Tanner stage 5 does not necessarily mark the final stage of puberty for everyone (e.g., there can be more breast and hair development than usually depicted in a stage 5).
There is an interesting difference between measuring puberty clinically and in research. Clinically, you mostly need to know if they have started or stopped puberty. Research-wise you may want a more nuanced perspective (like timing or tempo). The difficulty is that, as an isolated topic, puberty is considered kind of done. What Tanner did so well is that he came up with something that is considered a Gold Standard so it is difficult to fight against and say that it can be improved.
Of note, I think it is generally unusual for people to genuinely go backwards in pubertal development, and if it occurs it is often associated with pathology. On the flip side, it is quite common to have self-assessment and even physician assessment go backwards, because you realize you over-categorized someone looking backwards.
Q: Why is puberty so difficult to study?
A: One of the reasons puberty is so difficult to study is because we ignore the social side of it. There is such a big difference between what you think your breasts look like and what they actually look like. Generally, humans are bad at reporting what they look like. For instance, trying to get a teenager to look in the mirror and say what they look like is quite difficult to achieve accurately. Even hormonally, we know that hormones are impacted by stress hormones and having someone come into a lab can be stressful thus impacting the hormonal samples we get. We often miss the most interesting part of puberty which is how someone's body is impacted by how someone feels about their body. These changes are not happening in an isolated manner; they are happening to someone who is hyper-aware of them and impacted by them as well. Physiologically, it is difficult for puberty to objectively regress but it is easy for this to happen in measurement.
There are a lot of publications showing differences in how physician-assessed versus self-assessed puberty. But the message that comes out of a lot of it is that they’ve seen a regression to the mean. If you ask a group of 11 year olds what their pubertal stage is, they will tend to over or under report the mean (perhaps even because they want to be perceived as normal). For example, if a teen has no development, they want to have the same amount as their friends, and report what most people have. There is something really anxiety-provoking about whether one’s body is developing the same as everyone else’s; there is a huge weight of cognition and complexity and emotional angst that comes into play in filling out this apparently simple questionnaire. We see the same thing in parental questionnaires as well (e.g., parents will report that their 10 year old hasn't reported puberty even if they have because they are their parents’ little kid).
Q: Is puberty happening earlier in girls?
A: Some of it. Menarche has been happening earlier in girls (although that seems to be evening out now). Pubertal hair development has not necessarily happened much earlier. Determining the onset of breast development is difficult because we are measuring different populations (e.g., everybody else versus just causasian people in the original studies). For instance, breast development appears to be happening earlier but this could just be because we are capturing more diverse populations now. Interestingly, menarche development seems to track with nutrition, health, and weight; we need sufficient nutrients to be able to go through menarche earlier.
Q: What, if anything, do you think is critically missing regarding education about puberty in middle/high schools? (Question from Akhila Nekkanti)
A: I think that we need to be educating children about puberty earlier than we think because there are a significant minority of girls that are having notable changes at 8 or 9 years so they need a heads up. It should be a conversation that continues to happen as they progress through puberty. It should be a conversation that happens at home and school but if that doesn't happen, this conversation can end up taking place on the internet which can lead to inaccurate ideas. This often happens because people think that young people aren't old enough to learn about pubertal development. There are some books written for quite young children about what your body is and how it changes which can be helpful. Another aspect of puberty is that it is not just physical changes, even if places teach you that your body changes they may not touch on emotional changes (like how young people may get anxious/worried about things they weren’t before as well as interested in things they weren't before). We need to teach people that these changes are okay and normal. We need to also give them skills on how to manage all of that.
Q: Do children ever react negatively to the physical examination? How is this process conducted in a trauma-informed way? (Question from Theresa Cheng)
A: When I’ve done them both clinically and research-wise, I only once had a girl decide she wanted to stop during the examination. That’s the only time it has happened to me. I have had lots of people decide they don't want to do the study in advance. This is an area where you have to be very clear with what you are going to do. Part of giving young people ownership over their body, is giving them time and opportunity to get out of it. If a young person changes their mind in front of me, they may be less likely to say no (perhaps because they may not feel comfortable to stop). We send them the Tanner stage questionnaire in advance to show them what we are going to be looking for. It’s also important to always have a chaperone (normally a parent). There were some girls who didn’t want their parents (particularly fathers) to be with them so we always made sure there was a female researcher to chaperone in case. Overwhelmingly, they say it was not as bad as they thought it would be.
We did learn in some of our studies about keeping parents in the loop (especially if we were recruiting over email). We had one or two young people concerned about what we were going to do. This emphasized to us the need to be very clear about the boundaries and rationale for what we were going to do.
- end of interview -
Acknowledgements: I thank Dr. Anne-Lise Goddings for her time and for sharing her expertise. I would also like to thank Theresa Cheng and Akhila Nekkanti for their important questions and Dr. Kate Mills for helping to facilitate this interview.
