“Development is the missing link between genotype and phenotype, a place too often occupied by metaphors in the past … But a strong emphasis on the genome means that environmental influence is systematically ignored. If you begin with DNA and view development as “hard-wired,” you overlook the flexible phenotype and the causes of its variation that are the mainsprings of adaptive evolution.” (Mary Jane West-Eberhard, 2003: 89-90)

“Genes, unlike gods, are conditional. They are exquisitely good at simple if-then logic: if in a certain environment, then develop in a certain way… So here is the first moral of the tale: Don’t be frightened of genes. They are not gods; they are cogs. (Matt Ridley, 2003: 250)

 

In his book The Triple Helix, Richard Lewontin told the story of the molecular biologist and Nobel laureate Sydney Brenner, who – while speaking at a conference – predicted that one day we would be able to “compute” an organism (2002). All we would need are two things: the organism’s full genome and powerful enough computers that were up to the task.

The idea is seductive. Genes are sometimes seen as self-sufficient molecules, almost existing in a vacuum, that contain all the information necessary to code for proteins. From there, it’s not a very big logical leap to think that if you had the genome, you could enter the code in some database, hit “run,” and then watch some digitized version of the organism unfold.

In fact, scientists are doing something much like this for the tiny roundworm C. elegans with the project OpenWorm. Yet even for a relatively simple organism such as this, with only about a thousand cells in total, there are reasons to be cautious. As The Economist warned in its write-up of OpenWorm: “Attempting to simulate everything faithfully would bring even a supercomputer to its knees.” However, this isn’t due solely to the limits of computing power (what if we had a super-duper computer!?). Rather, it’s a matter of how the question is framed.

To go back to Brenner, I don’t know what else he said that day, so it isn’t fair to caricature the rest of his comments as quite so reductionist. If asked, I think he would acknowledge that if we play the scenario out, we are immediately confronted by the fact that there is no vacuum. All organisms must reside somewhere, or – more realistically – many somewheres over a lifetime. And, whatever the genome is, environmental circumstances can have profound influences on phenotypes.

 

“The Environment”

“There are more things in heaven and earth, Horatio, than are dreamt of in your database.”

 

Here is a straightforward example. Barry Bogin and Ines Varela Silva found that ethnic Maya children residing in Los Angeles and Florida were on average 11.5 centimeters (4.5 inches) taller than those who lived in Guatemala (Bogin et al 2002). This was one of the largest population shifts in height ever recorded in that short a period of time (for those in the US, most of their parents had migrated only within the last 20 years). In fact, the heights of Maya children in the US were not significantly different from American reference data, despite their being perceived as genetically ‘short.’

 

On the other hand, body build (the proportion of total height that was due to leg length) did not change much. Some studies have in fact found that average body proportions can change in a generation or two, such as in the case of Japan (Tanner et al 1982), but there are reasons that this pattern is inconsistent across studies. Phenotypes are not necessarily set in stone, though some may be more responsive to change than others depending on the age of exposure, and degree of severity, to different environmental variables.

What accounts for the difference in height in the Maya? We can dismiss the idea of a selective migration effect – that the tallest decided to migrate to the US, while shorter ones remained in Guatemala. That possibility seems highly unlikely, as most entered the US following indiscriminate forced displacement during a bloody, decade-long civil war. Another strike against that explanation is that heights of US Maya were even closer to US reference values in 2000 than they were in 1992, indicating an incremental shift in growth patterns. Instead, the authors suggested that the differences were likely due to factors such as nutrition, health care, and the quality of drinking water in the US compared to Guatemala. This is consistent with research demonstrating changes in growth patterns in many populations over time (Bogin 1999; Steckel and Rose 2002).

Certainly, genes are an important source of variation in traits like height, but this is complex, with many genes involved. One of the first to be consistently associated with variations in height across populations (the UK, Sweden and Finland), as well as in adults and children, was reported in 2007 (Weedon et al 2007). However, the effects were modest. Having one copy of a particular allele was roughly equal to an additional 0.4 cm in adult stature, accounting for only about 0.3% of the total variation observed.

What accounts for the rest? There must be other genes involved, but across the globe, several studies have found correlations between many environmental variables – large and small – with how a fetus, infant, or child physically develops. This is true not only for height, but for things like adiposity and overall health. I’m going to group these together, not to imply they are all equivalent, but only because they all deal with somatic development and it would be too cumbersome to treat them all separately:

• altitude (Greska 1990)
• seasonality (Little et al 1993)
• pollutants (Schell et al 2012)
• religious food restrictions such as during Ramadan (Reiches et al 2014)
• socioeconomic status (Leatherman et al 1995)
• neighborhood wealth (Drewnowski et al 2007)
• homelessness/availability of shelter (Smith & Richards 2008)
• number of siblings (Ochiai et al 2012)
• infectious diseases (Moore et al 2001)
• maternal nutrition during pregnancy (Barker 1998)
• maternal smoking during pregnancy (von Kries et al 2002)
• maternal nutrition as a child (Stein et al 2004)
• whether one’s mother was a twin or triplet (if you’re a marmoset; Rutherford et al 2014)
• grandmother’s exposure to famine (Stein and Lumey 2000)
• psychological stress (Gohlke et al 2004)
• our gut microbiota (Turnbaugh et al 2006)
• amount of sleep (Taveras et al 2014)
• physical activity levels (Cardoso and Garcia 2009)
• political instability and war (Clarkin 2012; Devakumar et al 2014)
• cultural beliefs about food and household distribution of resources (Dettwyler 1993)

This is just a partial list; a complete one is probably an exercise in futility. One of my advisors in graduate school, Gary James, liked to say that “the best model of reality is reality itself.” What he meant was that we can never account for all of the variables that influence our biology. We can pick out a few major ones, but it’s beyond our capacity to model everything, even for a simple organism like C. elegans. This doesn’t mean we can’t know anything; otherwise what’s the point? In fact, we know a great deal about growth and development. It’s just a reminder about the complexity of organisms, and how many moving pieces there are.

 

Seamless and Inseparable

 “What is inherited is DNA. Everything else is developed.” – James Tanner, “Foetus Into Man” (1990 :119)

 

As nearly everyone would acknowledge, BOTH genetics and environment are essential, not because one is more important than the other, but because the two are inextricably intertwined. There is no organism without a genome; but there is also no such thing as an organism without an environment.

This also puts a serious damper on the idea of ‘computing’ organisms based solely on genomes, at least not in any absolute sense. If you had the complete genome of an extinct mammoth, and could somehow find it a suitable place to gestate and then find someone to care for it until maturity, you’d get something we would identify as a mammoth, rather than, say, a jellyfish. But to fully account for its individual biology – what kind of mammoth or jellyfish would we get – we would need much more than the genome. We would also need a crystal ball to predict all future environmental variables the organism would ever encounter.

Speaking of jellyfish, if you grow one in space, you will still get a jellyfish. But without being exposed to gravity in the early stages of its life, even for a few days, there is a decent chance that they will move abnormally once they’ve returned to earth (Spangenberg et al 1994). Of course, most jellyfish will never experience space, but the point is that being able to navigate something even as basic as gravity does not come automatically, but may require exposure to it.

We’re then forced to confront the reality that the effects of the environment on development are not merely noise interfering with the ‘true’ nature of the organism. They are integral. It makes sense that jellyfish genomes should ‘expect’ gravity because, before space travel, it had never been otherwise in the billions of years of life on earth. In other words, genomes are sculpted by natural selection to navigate probable environments, but with some degree of unpredictability. The same is true for us. As the developmental psychologist Michael Tomasello once said: “Fish are born expecting water, and humans are born expecting culture.”

One problem is that we often remain stuck in dichotomous “nature versus nurture” thinking. Some of us lean more toward one side than another, depending on the biological trait in question, but as Anne Fausto-Sterling warned, that view is too simplistic:

Most people have only one model to think about how human traits develop. Whether they call themselves determinists or constructionists, nativists or empiricists; whether they emphasize genes or environment, nature or nurture, or think of important traits as hardwired or plastic, they envision a seesaw. Sometimes nature (genes, wiring, etc.) is heavier, and the seesaw bounces onto nature’s ground. Sometimes the environment (culture, nurture) is heavier and weighs down that end of the seesaw. Frustratingly, even though a process approach to understanding motor development is pretty much settled science, in most other arenas (sex differences, intelligence, athletic and musical ability, and more), people can’t stop arguing—adding a little weight to one side or there to the other—and the seesaw teeters back and forth with each new research finding.

But there is no seesaw. Human behaviors are not “things” more or less influenced by one or another source. They are processes that are simultaneously embodied and shaped by experience. The behavior of walking, as just one example, differs from one individual to the next depending on health history, body size and shape (for example, limb-to-trunk length ratios), past physical experience, the character of the walking surface, clothing, and many other factors.

An emphasis on development as a process underscores how interwoven nature and nurture are. Perhaps this concept doesn’t come easily to us, in part, because time is invisible. But it helps to reflect on the obvious fact that we aren’t born fully-formed. Rather, all organisms grow into maturity, responding to stimuli and challenges, while exchanging matter and energy with the world around us. And as humans, we can take a rather long time to develop into maturity.

As we grow, we incorporate our environmental circumstances. Researchers have given this idea different names such as ‘embodiment’ (Krieger 2005), experiences getting ‘under the skin,’ racial inequalities ‘becoming biology’ (Gravlee 2009), or culture becoming ‘en-brained’ throughout our lives (Downey and Lende 2012: 32). And though I have to admit that I never quite got what Pierre Bourdieu was trying to say, I do remember liking his phrase ‘the internalization of externality.’ In subtle and not-so-subtle ways, our outsides get in our insides. And, sometimes it can seem like everything in our inside wants to be on our outside. Insides and outsides may be in more intimate contact than we are sometimes aware.

On a personal note, when I look at my young children now, of course I see their current selves. Even as a father who has been present for nearly every day of their lives, it can be hard to remember how much they’ve changed in appearance and behavior in only a few short years. Their present selves might be even more different, had their past selves contracted a nasty infection, attended a different school, etc. Each transient state impacts the next, with the earliest ones having a disparate effect on all the subsequent ones.

Some have emphasized the importance of the first 1,000 days as a critical period of life. There’s nothing magical about 1,000 days, but it’s a good approximation of when our development is most vulnerable, and when our developmental trajectory is most alterable, depending on what part of our biology we’re talking about. Of course, events later in life can also have profound effects. Rachel Yehuda and Linda Bierer (2009) have referred to the development of post-traumatic stress disorder as akin to an ‘existential transformation.’ In a way, falling in love is also transformational, at least for prairie voles. There are cultural examples too. Someone who has undergone a rite of passage – an initiation, marriage, a rise in rank – is often conferred a new status and set of rights and responsibilities (van Gennep 1960). A person (or vole) who has fallen in love, been initiated into a society as an adult, or diagnosed with PTSD is obviously still the same person they were before. Yet, in another sense they can also be seen as qualitatively different.

 

Hardwired for Plasticity

Over at The Mermaid’s Tale, Ken Weiss once wrote that “we are hard-wired not to be hard-wired” for our behavioral repertoire. Bill Leonard and colleagues made a similar claim about our diet: that we were selected to be flexible omnivores, rather than to have a highly specialized dietary niche (Leonard et al 2010). As Matt Ridley succinctly put it, plasticity is evolution’s “masterstroke” (2003:174).

This brings us to a fundamental question: how does a genome accommodate so much unpredictable environmental complexity? In a nutshell, plasticity. Variants of genes (alleles) that thrive in given set of ecological conditions will be passed on at higher rates than others. But this is not the only force affecting allele frequencies; nor is it the only way to adapt. I once heard the biological anthropologist R. Brooke Thomas say that, without a doubt, natural selection was an essential mechanism in shaping our biology. However, he also felt it was “too clunky” to try to explain all variation and adaptation. And he was right.

Genetic adaptations certainly exist. But a genome that can respond to environmental feedback and operate in many possible, unpredictable conditions could be even more likely to survive and reproduce than a rigid one, hardwired to just do its thing regardless of circumstances. Sometimes those responses are due to developmental constraints; sometimes they result in pathologies; and other times they may be truly adaptive. Though it’s not always easy to differentiate between them (Ellison and Jasienska 2007).

Take body temperature regulation, for example. There do seem to be genetic adaptations for this across human populations. Near the equator, body builds tend to be thinner and longer-limbed – on average – in order to dissipate heat more easily. But this has always been a very imperfect correlation, and there are exceptions to this trend. The strength of this association has also weakened in recent decades, as globalization and changes in diet and activity patterns have affected patterns of body mass around the world (Katzmarzyk and Leonard 1998). Rather than relying solely on body build to regulate temperature, there are other levels of adaptation (Frisancho 1993):

• Behavioral: Is it hot out? Go find some shade.
• Cultural: Are you cold? Build a fire. Put on a parka.
• Physiological: Sweating, shivering. Vasodilation/vasconstriction.
• Developmental: Older Quechua in high-altitude Peru are more resistant to cold than young Quechua, as they had more time to develop under cold conditions (Little et al 1971).

All of these adaptations take place on different timescales. Whereas natural selection takes generations, by definition, behaviors can take just a few seconds. Developmental adaptations occur while the organism is growing, and may take years, but this is still within a single lifetime. Finally, there may be another, intermediate level of adaptation, which Chris Kuzawa referred to as “phenotypic inertia” (Kuzawa 2005).

It’s an intriguing idea. Kuzawa hypothesized that: “As a mode of adaptation, phenotypic inertia may help the organism cope with ecologic trends too gradual to be tracked by conventional developmental plasticity, but too rapid to be tracked by natural selection.” In other words, the experiences of parental generations can carry over, perhaps by developmental constraint, perhaps epigenetically (yes, I realize that word is overused).

Last example. Stewart et al (1980) conducted an experiment where one group of rats was given a low protein diet for ten to twelve generations, while a control group was given a normal diet. At the end of that time, the poor-diet group weighed about half of what the controls did by adulthood. When the researchers tried to rehabilitate the poor-diet lineage by re-introducing a normal diet, some interesting patterns emerged.

First, timing was important, and the earlier the rehabilitation, the better the results. Rats given good diets after weaning did not fare as well in terms of physical growth as those who were given good diets in infancy (‘cross-fostered’ to control mothers at birth) or in utero. Second, full recovery took two to three generations in both of the groups that were rehabilitated postnatally. One generation was not sufficient. By comparison, the prenatally rehabilitated group actually overshot the growth seen in the control group, after which they came back to ‘normal’ by generation three.

Finally, the low protein group also suffered in terms of learning (a Lashley jumping platform test). Many never achieved complete proficiency on the various tests, which consisted of choosing between and jumping through one of two doors. By contrast, 100% of the control group did so, and in many fewer trials (average of 170 vs. 230). For the rehabbed offspring, those well-fed after weaning showed no improvement (230 trials), while the early postnatal (190) and prenatal groups (210) fared slightly better, though they never caught up completely, at least not within three generations.  

So where does this leave us? Stewart et al. hedged somewhat against making superficial comparisons between their study on rats and humans, but they also suggested there could be parallels with deprived populations, such as in developing countries facing generations of poverty. Recovery might not happen right away in terms of growth or learning, even if nutritional quality (or other factors) change radically.    

We can say, then, that of course genes matter tremendously, but so does the environment. And “the environment” does not begin in kindergarten, at birth, or even prenatally. The experiences of previous generations may also leave a mark. In truth, it all counts, and for fundamental biological reasons. I’ll end with Matt Ridley:

“It is genes that allow the human mind to learn, to remember, to imitate, to imprint, to absorb culture, and to express instincts. Genes are not puppet masters or blueprints. Nor are they just the carriers of heredity. They are active during life; they switch on and off; they respond to the environment…They are both cause and consequence of our actions. Somehow the adherents of the “nurture” side of the argument have scared themselves silly at the power and inevitability of genes and missed the greatest lesson of all: the genes are on their side.” (2003: 6)

 

 

References

Barker DJP. 1998. Mothers, Babies and Health in Later Life. Edinburgh: Churchill Livingstone. Link

Bogin B. 1999. Patterns of Human Growth. Cambridge. Link

Bogin B, Smith P, Orden AB, Varela Silva MI, Loucky J.  2002. Rapid change in height and body proportions of Maya American children. Am J Hum Biol. 14(6):753-61. Link

Cardoso HF, Garcia S. 2009. The Not-so-Dark Ages: ecology for human growth in medieval and early twentieth century Portugal as inferred from skeletal growth profiles. Am J Phys Anthropol 138(2):136-47. Link

Clarkin PF. 2012. War, forced displacement and growth in Laotian adults. Ann Hum Biol. 39(1):36-45. Link

Dettwyler KM. 1993. Dancing Skeletons: Life and Death in West Africa. Waveland. Link

Devakumar D, Birch M, Osrin D, Sondorp E, Wells JCK. 2014. The intergenerational effects of war on the health of children. BMC Medicine 12:57 Link

Downey G, Lende DH. 2012. Neuroanthropology and the encultured brain. In Lende DH, Downey G (eds.) The Encultured Brain p. 23-65. Cambridge: MIT. Link

Drewnowski A, Rehm CD, Solet D. 2007. Disparities in obesity rates: Analysis by ZIP code area. Social Science & Medicine 65 (12): 2458-63. Link

Ellison PT, Jasienska G. 2007. Constraint, pathology, and adaptation: how can we tell them apart? Am J Hum Biol. 19(5):622-30. Link

Frisancho AR. 1993. Human Adaptation and Accommodation. U Michigan. Link

Gohlke BC, Frazer FL, Stanhope R. 2004. Growth hormone secretion and long-term growth data in children with psychosocial short stature treated by different changes in environment. J Pediatr Endocrinol Metab 17:637–643. Link

Gravlee CC. 2009. How race becomes biology: embodiment of social inequality. Am J Phys Anthropol. 139(1):47-57. Link

Greska LP. 1990. Developmental responses to high-altitude hypoxia in Bolivian children of European ancestry: A test of the developmental adaptation hypothesis. American Journal of Human Biology 2(6): 603–12 Link

Katzmarzyk PT1, Leonard WR. 1998. Climatic influences on human body size and proportions: ecological adaptations and secular trends. Am J Phys Anthropol. 106(4):483-503. Link

Krieger N. 2005. Embodiment: a conceptual glossary for epidemiology. J Epidemiol Community Health 59(5): 350–355. Link

Kuzawa CW. 2005. Fetal origins of developmental plasticity: are fetal cues reliable predictors of future nutritional environments? American Journal of Human Biology 17: 5-21. Link

Leatherman TL, Carey JW, Thomas RB. 1995. Socioeconomic change and patterns of growth in the Andes. Am J Phys Anthropol. 97(3):307-21. Link

Leonard WR, Stock JT, Valeggia CR. 2010. Evolutionary perspectives on human diet and nutrition. Evolutionary Anthropology 19:85–86. Link

Lewontin R. 2002. The Triple Helix: Gene, Organism, and Environment. Harvard. Link

Little MA, Thomas RB, Mazess RB, Baker PT. 1971. Population differences and developmental changes in extremity temperature responses to cold among Andean Indians. Human Biology 43(1): 70-91. Link

Little MA, Gray SJ, Leslie PW. 1993. Growth of nomadic and settled Turkana infants of northwest Kenya.  Am J Phys Anthropol. 92(3):273-89. Link

Moore SR, Lima AA, Conaway MR, Schorling JB, Soares AM, Guerrant RL. 2001. Early childhood diarrhoea and helminthiases associate with long-term linear growth faltering. Int J Epidemiol. 2001 30(6):1457-64. Link

Ochiai H, Shirasawa T, Ohtsu T, Nishimura R, Morimoto A, Obuchi R, Hoshino H, Tajima N, Kokaze A. 2012. Number of siblings, birth order, and childhood overweight: a population-based cross-sectional study in Japan. BMC Public Health 12:766. Link

Reiches MW, Moore SE, Prentice AM, Ellison PT. 2014. Endocrine responses, weight change, and energy sparing mechanisms during Ramadan among Gambian adolescent women. Am J Hum Biol 26(3):395-400. Link

Ridley M. 2003. Nature Via Nurture: Genes, Experience, and What Makes Us Human. New York: Harper Collins. Link

Rutherford JN, deMartelly VA, Layne Colon DG, Ross CN, Tardif SD. 2014. Developmental origins of pregnancy loss in the adult female common marmoset monkey (Callithrix jacchus). PLoS One.  28;9(5):e96845. Link

Schell LM, Burnitz KK, Gallo MV. 2012. Growth as a mirror: is endocrine disruption challenging Tanner’s concept? Ann Hum Biol. 39(5):361-71. Link

Smith C, Richards R. 2008. Dietary intake, overweight status, and perceptions of food insecurity among homeless Minnesotan youth. Am J Hum Biol. 20(5):550-63. Link

Spangenberg DB, Jernigan T, McCombs R, Lowe BT, Sampson M, Slusser J. 1994.  Development studies of Aurelia (Jellyfish) ephyrae which developed during the SLS-1 mission. Advances in Space Research 14 (8):239–47. Link

Steckel RH and Rose JC. 2002. The Backbone of History: Health and Nutrition in the Western Hemisphere. Cambridge University Press. Link

Stein AD, Barnhart HX, Wang M, Hoshen MB, Ologoudou K, Ramakrishnan U, Grajeda R, Ramirez-Zea M, Martorell R. 2004. Comparison of linear growth patterns in the first three years of life across two generations in Guatemala. Pediatrics 113(3 Pt 1):e270-5. Link

Stein AD, Lumey LH. 2000. The relationship between maternal and offspring birth weights after maternal prenatal famine exposure: the Dutch Famine Birth Cohort Study. Hum Biol 72:641–654. Link

Stewart RJC, Sheppard H, Preece R, Waterlow JC. 1980. The effect of rehabilitation at different stages of development of rats marginally malnourished for ten to twelve generations. British Journal of Nutrition 43: 403-11. Link

Tanner JM. 1990. Foetus Into Man: Physical Growth from Conception to Maturity. Harvard. Link 

Tanner JM, Hayashi T, Preece MA, Cameron N. 1982. Increase in length of leg relative to trunk in Japanese children and adults from 1957 to 1977: comparison with British and with Japanese Americans. Annals of Human Biology 9: 411–23. Link

Taveras EM, Gillman MW, Peña MM, Redline S, Rifas-Shiman SL. 2014. Chronic Sleep Curtailment and Adiposity. Pediatrics doi:10.1542/peds.2013-3065 Link

Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI. 2006. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature. 444(7122):1027-31. Link

van Gennep A. 1960. The Rites of Passage. Psychology Press. Link

von Kries R, Toschke AM, Koletzko B, Slikker W Jr. 2002. Maternal smoking during pregnancy and childhood obesity. Am J Epidemiol 156(10):954-61. Link

Weedon et al. 2007. A common variant of HMGA2 is associated with adult and childhood height in the general population. Nature Genetics 39 (10): 1245-50. Link

West-Eberhard MJ. 2003. Developmental Plasticity and Evolution. Oxford Link

Yehuda R, Bierer LM. 2009. The relevance of epigenetics to PTSD: implications for the DSM-V. J Trauma Stress. 22(5):427-34. Link