Friday, November 30th 2012

Audiences, Trolls, and Getting Some Science Onto the Internet

Earlier this week, the Women in Science group at the University of Illinois, Urbana-Champaign convened a panel on using social media to promote science. Melanie Tannenbaum, Bill Hammack, Joanne Manaster and I were the panelists, and Jo Holley was the organizer.

There were a few things that I found interesting about our varying responses as panelists to the questions we were asked. That variation seemed to map onto the kind of social media we used, and the audience we sought to reach. I am guessing Melanie and I reach the most educated audience, and that the vast majority are adults. Joanne and Bill both have a significant cadre of younger fans – I liked Bill’s analysis that those who watch his great YouTube videos are young gamers… and old gamers. I have a feeling that despite variation in age or education, we are all reaching a highly intelligent, motivated audience.

Anyway, the rest of my comments are interspersed through a Storify I’ve made, where you’ll find more links, pictures, and (almost) full video of the panel. Enjoy! (I cannot seem to get the Storify to embed, so just follow this link for now.)

[View the story “Nov 26th Panel on Using Social Media to Promote Science, University of Illinois” on Storify]

Comments Off on Audiences, Trolls, and Getting Some Science Onto the Internet

Monday, November 26th 2012

Today: Using Social Media to Promote Science

Join me, Joanne Manaster, Melanie Tannenbaum and Prof. Bill Hammack today from 4-6pm at the University of Illinois, Urbana-Champaign Institute for Genomic Biology, room 612 (that’s the room right next to Array Cafe, in the Gatehouse).

The University of Illinois has a surprising number of academics who are successful science writers, bloggers, photographers and social media users (I know I’m going to forget a few, but others who come to mind are Christine Herman, Dan Simons, and Alex Wild). I suspect the proportion of online science outreach folks is quite high here, and we’re only starting to tap into our collective potential (Midwest Science Online, anyone?).

This afternoon’s talk should be a great introduction to the panelists, our perspectives, and the concept of online science outreach. For my part, I’ll likely discuss my own academic trajectory, how that led to blogging, my thoughts about the blogosphere versus academia, and the importance of historicizing and contextualizing science.

Perhaps you’ll get a few ideas about how to get started, or to take your own work in a new direction. Join us!

Comments Off on Today: Using Social Media to Promote Science

Friday, November 16th 2012

Bone Mineral Density Status: It’s Complicated

Femur_fibre_arrangement_for_strength wikimedia commons

The femoral head. Public domain image via Wikimedia Commons.

Lady bones are delicate, selfless, dense connective tissue that hold lady bodies up to help the lady vessel carry babies. When we are pregnant, the thinking goes, lady bones give up their calcium in vast quantities, depleting themselves for the good of their darling fetuses. We ladies mostly replete this calcium between pregnancies. Then of course, when we lose our essential lady-ness in the form of menopause, our lady bones go all to hell, so Sally Field gives us a monthly dose of Boniva, which we gobble gratefully until some of us start getting Boniva fractures.*

Plenty of the wisdom above is true. Pregnancy does often cause a loss of calcium and bone density in skeletal bones, and mothers usually get it back between pregnancies. Bone density is lost after menopause, when estrogen isn’t around anymore to help with mineral deposition or aid intestinal calcium absorption.

But as with many things, women’s bone health is a lot more complicated than that. Often, when you perform a study to a high theoretical and technical standard, you are left with as many questions as answers.

Madimenos et al (2012) recently published an article in the American Journal of Human Biology about bone health among the Shuar of Amazonian Ecuador. The Shuar are forager-horticulturalists – for the most part they are subsistence farmers, which means they are farming enough for their families to eat, but some also do small-scale farming and pastoralism to sell at market. This Shuar field site is well known within anthropology for its rigorous methodology and outreach with local participants. It’s also an interdisciplinary site that seems to be great for tackling both biological and cultural anthropology questions. As with many indigenous groups, the Shuar are moving towards market integration, and so questions about major cultural and nutritional transitions are especially relevant there.

This project tested hypotheses about four potential predictors of adult bone health: age at menarche (first period), age at first birth, lactational duration (how long you breastfeed), and interbirth intervals (IBI, time between births). Each one of these variables are themselves hugely variable based on environment and culture. For instance, age at menarche is driven late by energetic constraint, but early by certain psychosocial factors (e.g., Belachew et al., 2011; Chisholm et al., 2005; Ellis et al., 2011; Ellison, 1982). So, seeing if a variable like age at menarche has predictive value for bone health is about seeing if environment during development, gene expression, social environment, and nutritional status play a role in bone health. There’s a lot going on.

At the same time, trying to make connections between variables like age at menarche and bone density are important because age at menarche also signals three other important components of a woman’s life history: the conditions under which she lived at puberty, which is an important developmental set point (Ellis, 2004; Ellison, 1982), her lifetime estrogen exposure, and how quickly she moves out of adolescent subfecundity, as the earlier your menarche the more quickly you attain regular ovulatory cycles (Apter and Vihko, 1983; Vihko and Apter, 1984).

I’m going to talk about what Madimenos et al (2012) found, why their study design was great, and what aspiring researchers can learn from this kind of work.

Bone Health: As You’d Expect, and Not

The major hypotheses of this project were that 1) earlier menarche would be correlated with higher bone density; 2) older age at first birth would be correlated with higher bone density; 3) longer lactational duration would be correlated with lower bone density; and 4) longer IBIs would be correlated with higher bone density. They found 1) support; 2) partial support; 3) partial support; and 4) no support.

Now, for those of you who don’t do scientific research, this might seem like a bummer. In fact, the places where Madimenos et al (2012) found partial or no support were some of the more interesting parts of the paper, and showed how complex the relationships are between these life history variables, environment, and bone health. If there was one thing I wish more non-scientists understood (and scientists, and journal editors, so kudos to Editor-in-chief Dr. Peter Ellison and AJHB), it’s that partial and null results are where the party begins.

In the second hypothesis on age at first birth, Madimenos et al (2012) found that a later age at first birth was only correlated with greater bone health in younger, lactating women (14-24 year old mothers). That time period represents both adolescence and young adulthood, and so it makes sense that the youngest mothers of this group would be the ones not only with the lower bone density (as they may not have even achieved full skeletal maturation before getting pregnant), but the closest to that first pregnancy time-wise. A very young mom sampled soon after her first pregnancy wouldn’t have enough time to replete much of the bone lost during that pregnancy, particularly if she was already at a disadvantage in terms of her bone health going into it.

This project’s third hypothesis posited that the longer an individual spends breastfeeding in her lifetime, the lower her bone density. The authors found support for this hypothesis among currently lactating women. They found the opposite relationship among 35-44 year old, non-lactating women. What was interesting here was that the older group not currently lactating had the longest average lactational duration. Might this indicate a shift in cultural practices around breastfeeding, if they were breastfeeding longer per pregnancy than the younger women? And why did they see an improvement in bone health among this group? Might this relationship be simply because women who were able to breastfeed longer had better nutritional status to begin with, or because the way bones bounce back from lactation overcompensates for pregnancy or lactation-related bone loss?

Finally, in testing the fourth hypothesis the authors found no relationship between IBI and bone health. Average IBI in this population is about 31 months, which may simply be plenty of time to replete lost calcium. Perhaps in a more marginal environment, or one where women are more closely spacing their births, we would find this relationship.

It’s About How You Use It

I’ve written quite a bit in the past about the mistakes of translating research on the behavior or physiology of WEIRD, undergraduate, white populations across all humans, or thinking you can understand much of anything about evolution or adaptive scenarios from the way this very specific class of people thinks about love, mating and dating, or success (here, here, here, here and here). Though this paper isn’t about behavior, I think anyone who studies humans can learn a lot from this paper’s study design. Further, Madimenos et al (2012) do not overstate conclusions, and if anything organize their data in a way less likely, rather than more likely, to lead to significant results.

This project takes place in Ecuador, among a group of indigenous forager-horticulturalists. That doesn’t necessarily make these data more translatable across populations than WEIRD datasets, but it expands the diversity of our information. When we observe the relationship between genotype, stress, life history and phenotype in several environments, it makes it possible for us to develop predictions about these relationships, their causality, and their meaning.

The next great thing about this research design is that, even though the life history variables they’re measuring have lots of embedded meaning, they are, indeed, measurable. Few people are going to dispute whether using a participant’s age at menarche is in fact a way to operationalize… age at menarche (some will complain about self-report, but it turns out self-report for age at menarche is pretty good, see Must et al (2002)). Same with age at first birth, interbirth interval, and lactational duration. In this project these measures were all self-report, but there are certainly many studies that follow women longitudinally to gather these kinds of data. And of course, the outcome variable – different measures of bone health – are all measurable as well. No one is asking these women if they feel like their bones are healthy. That may certainly be an interesting qualitative measure to add, but reliable, quantifiable measures are what make projects feasible and results last a long time in the literature.

I would love to see future work testing these questions a few different ways. How would a longitudinal project inform our perspective? Do we see meaningful genetic variation, or variation in gene expression, that explains variability in bone health? Are there other biomarkers that would help us understand environmental variation that might lead to some women allocating less to bone health? In the end, this paper demonstrates that bodies are embedded in cultural, developmental and environmental contexts, and that those contexts make for complex storytelling.


*There appears to be a causal link between Boniva and awful-sounding femur fractures and osteonecrosis of the jaw, having to do with the way bisphosphonates slow osteoclast-related bone resorption. The problem is that some bone breakdown and resorption is important to normal remodeling processes, and bisphosphonates don’t exactly pick and choose between the good and bad resorption.


Apter D, Vihko R. 1983. Early menarche, a risk factor for breast cancer, indicates early onset of ovulatory cycles. Journal of Clinical Endocrinology & Metabolism 57(1):82-86.

Belachew T, Hadley C, Lindstrom D, Getachew Y, Duchateau L, Kolsteren P. 2011. Food insecurity and age at menarche among adolescent girls in Jimma Zone Southwest Ethiopia: a longitudinal study. Reproductive biology and endocrinology: RB&E 9:125.

Chisholm JS, Quinlivan JA, Petersen RW, Coall DA. 2005. Early Stress Predicts Age at Menarche and First Birth, Adult Attachment, and Expected Lifespan. Human Nature 16(3):233-265.

Ellis BJ. 2004. Timing of pubertal maturation in girls: an integrated life history approach. Psychological Bulletin 130(6):920.

Ellis BJ, Shirtcliff EA, Boyce WT, Deardorff J, Essex MJ. 2011. Quality of early family relationships and the timing and tempo of puberty: Effects depend on biological sensitivity to context. Development and psychopathology 23(1):85.

Ellison P. 1982. Skeletal growth, fatness, and menarcheal age: a comparison of two hypotheses. Human Biology 54:269-281.

Madimenos FC, Snodgrass JJ, Liebert MA, Cepon TJ, Sugiyama LS. 2012. Reproductive effects on skeletal health in Shuar women of Amazonian Ecuador: A life history perspective. American Journal of Human Biology 24(6):841-852.

Must A, Phillips SM, Naumova EN, Blum M, Harris S, Dawson-Hughes B, Rand WM. 2002. Recall of Early Menstrual History and Menarcheal Body Size: After 30 Years, How Well Do Women Remember? American Journal of Epidemiology 155(7):672-679.

Vihko R, Apter D. 1984. Endocrine characteristics of adolescent menstrual cycles: impact of early menarche. J Steroid Biochem 20(1):231-236.

Comments Off on Bone Mineral Density Status: It’s Complicated