Sunday, January 31, 2016

Wait, there's more: milk adiponectin and child growth varies!

It was a big day for the Biomarkers & Milk group. We had an important paper accepted at Annals of Human Biology. This paper is the second paper to come out of the Milk with Altitude Project (2013 edition), and I assure you, there will be more! 

This paper is important to me because it is further evidence for a hypothesis I have been working on. As I have written before, as an anthropologist, I tend to study non-western populations, and my research is primarily concentrated with two groups: Filipinos living in Cebu, Philippines, and ethnic Tibetans living in Nepal. Recently, working with an outstanding team of undergraduates, we published a paper looking at the hormone adiponectin in milk samples from 113 Filipino women.
So why are we excited about this? In the Cebu adiponectin article, we reported something new and different about milk adiponectin and child growth. Prior to our work, a phenomenal piece by Woo et al., (2012) had shown a significant, inverse association between milk adiponectin and infant weight for age z-score. Simply put: more adiponectin in milk, less growth in the child. BUT – and here’s the cool part – they found that the slower growing children had faster growth from 1-2 years of age.
Katie Hinde over at Mammals Suck suggested that this may be an adaptive strategy to allow mothers to mediate their energetic investment in offspring. By down regulating growth while it is directly dependent on her metabolism, mothers may be able to conceive sooner. The catch-up growth from 1-2 years then, makes complete sense – children are then growing when other community members can contribute to their energetic budgets.

And we think we’ve found the corollary – such trade-offs are environmentally sensitive. In a high resource (low pathogen?) environment such as that of the mothers Woo et al., studied, this is an excellent reproductive strategy, as it allows mothers to defer some of the costs of reproduction to others. However, in a low resource (high pathogen?) environment, the resources necessary to support later growth may be unavailable, and it may be a better reproductive strategy to promote early growth – when maternal metabolism can support it.

For the Cebu study, we had mean adiponectin levels of 7 ng/mL of milk, compared to 21 ng/mL for the US women Woo et al., studied. Child weight for age (centiles) also differed – the 50th percentile for our sample was equivalent to the 5th percentile for US infants. However, when we used a similar analytical technique as Woo et al., a really interesting picture emerged (Figure 1). Our data fit almost perfectly with that of Woo et al. In their model, there is a plateau for the association between milk adiponectin and child growth at lower levels of milk adiponectin (10-13 ng/mL) followed by a decline in weight for age as milk adiponectin increases. We found a positive association between milk adiponectin and infant weight for age at even lower levels of milk adiponectin (2 ng/mL to 8 ng/mL) – however there was a plateauing of the association from about 8-10 ng/mL, or the max of our range. The distribution of milk adiponectin in our sample barely overlapped with that of Woo et al.
Figure 1: Comparison of weight-for-age percentiles by milk adiponectin between the Woo et al., study and the Cebu study. C

When you put the two datasets together, instead of forming alternative models, they form a complementary model suggesting that the association between milk adiponectin and child weight for age z-score is not a linear association but rather a bell-shaped association, where at a very low levels (<7ng/mL) milk adiponectin is positively associated with growth, a flattening of the association at levels of 7-12 ng/mL (intermediate levels) and then an inverse association between milk adiponectin and growth at higher (>12 ng/mL) levels. And because milk adiponectin had not generally been studied in non-Western populations, such association was largely missed. This is why studying milk composition in a diversity of human populations and ecological zones is so important.

We’ve replicated this in our Tibetan study, but that’s a separate post because the relationships get even more interesting. Once the Tibetan study is out, I’ll continue the discussion started here with that piece. 

References cited
Anderson J, McKinley K, Onugha J, Duazo P, Chernoff M, Quinn EA. (2015) Lower levels of human milk adiponectin predict offspring weight for age: a study in a lean population of Filipinos. Matern Child Nutr. 2015 Oct 7. doi: 10.1111/mcn.12216. [Epub ahead of print]

Hinde K. (2012) Mother’ fat sends love letter to baby via the milk express., accessed Jan 26, 2016.

Martin LJ, Woo JG, Geraghty SR, Altaye M, Davidson BS, Banach W, Dolan LM, Ruiz-Palacios GM, Morrow AL. (2006) Adiponectin is present in human milk and is associated with maternal factors. Am J Clin Nutr. 2006 May;83(5):1106-11.

Woo JG, Guerrero ML, Guo F, Martin LJ, Davidson BS, Ortega H, Ruiz-Palacios GM, Morrow AL. (2012) Human milk adiponectin affects infant weight trajectory during the second year of life. J Pediatr Gastroenterol Nutr. 2012 Apr;54(4):532-9. doi: 10.1097/MPG.0b013e31823fde04.


  1. Have you published any of your research from Nepal?

  2. Yes, the first paper from Nepal is in this month's issue of the American Journal of Physical Anthropology (; a paper on milk hormones was just accepted at Annuals of Human Biology and should be available soon.