Behavioral Genetics I. 1h 38m Robert Sapolsky video.

This video explores the ways in which we can determine the genetic bases of behavior. Sapolsky defines behavioral genetics: "look for patterns of shared traits among individuals who have differing degrees of shared genes and infer relatedness and infer genetic influences from that." The lecture then documents and exposes deep flaws in several of the approaches which have been attempted (some of which, despite their flaws, have significantly influenced our culture). The lecture not only warns us that many behavioral genetics studies do not stand up to reasoned biological scrutiny, it helps us to deeply understand the challenges that behavioral genetics studies need to address to succeed. The fundamental concern in all these studies is how effectively does the study control for the influence of environment. And the short answer is that many have failed to be convincing upon close examination.

Topics discussed include twin studies, the 1980 study on "Sex differences in mathematical ability", adoption studies (with animals this is known as "cross fostering"), and that weird and wonderful world of identical twins separated at birth (yes, Sapolsky finds deep, deep flaws in these studies too!!!). One of the most fascinating discussions is on the amazing discoveries about how important the pre-natal environment is in fixing phenotypes that we might otherwise naively assume to be genetic! Another is on "non-mendelian inheritance of traits". What? Has Lamarck been vindicated? Partially: modern biology now has evidence for some limited Lamarckian inheritance!!! Blow me away! Finally, he examines the effects of the cellular material in an egg cell and indirect genetic effects.

Detailed Notes:

The first naive approach to behavioral genetics discussed is "if you see a trait that is universal in a species, obviously its genetic." But environment runs in families and in species too: shared genes and shared environment tend to covary. How can we control for environment?

Twin studies. There are monozygotic (identical) and dizygotic (fraternal) twins. We know that dizygotics are treated differently by parents, differential traits may be caused by parental behavior and not genetics!  Moreover, if monozygotic twins split within 5 days, they each get a separate placenta (dichorionic). That happens 1/3 of the time. The other 2/3 of the time, the split happens after 5 days and they share a placenta (monochorionic). Does your twin study account for this environmental difference? Evidently many of them do not! We know that IQ is more highly correlated in monochorionic twins. So environment can certainly be a factor even in twins.

At 1 hour of birth, the average rate of movement of limbs is greater among boys than girls. Sex differences in behavior manifests as early as one hour of birth. Sex differences in the behavior of parents has been seen within moments of birth. Hmm, it is going to be very hard to control for environment even with twin studies.

Benbow & Stanley wrote an enormously influential study in 1980 in Science "Sex differences in mathematical ability: Fact or artifact?". Reader's Digest called it the "math gene" though the study makes much more modest conclusions.  The study thought they had controlled for environment by testing "gifted" children before specialization into separate math tracks. Sapolsky observes that as early as first grade, boys are called on and praised more than girls and guidance councellors tend to recommend more math to boys. There are huge environmental differences well before grade 7!

Another approach is with adoption studies (in animal studies this is called cross fostering). Seymour Kety of Harvard studied adoptees in Denmark and determined that in the general population (no family history, and no history in one's adopted family) schizophrenia occurs at a 1% rate. If one's adoptive family has schizophrenia but your biological family has none, then the incidence rate is 3%. If the adoptive family has no schizophrenia but your biological family does, the rate is 9%. If both the adoptive family and your biological family have schizophrenia, the rate is 17%. Sapolsky tears it apart: since adoption doesn't typically happen within nanoseconds of birth, both the prenatal effects and early effects after birth are not controlled. In addition, adoption is not random: adoptees tend to be placed in families that are like the child's biological family.

Tom Bouchard at U Minnesota has studied identical twins separated at birth. 40 pairs were included in the initial studies although 200 have now been identified. Due to the penchant of humans to find pattern even where it doesn't exist, the first studies were bizarre (that's Sapolsky's "technical" term). For example, it was reported that some twins both flush the toilet before and after use. The literature has also provided some more noble results: IQ, introversion/extroversion, and degree of aggression are all about 50-50 genetic-environmental. Since adoption is not random and the sample size is so small, it is not clear how much capital to invest in such studies.

Some behavioral traits occur in the absence of any learning or experience. For example, all babies and fetuses smile. It is a basic motoric pattern.  Congenitally blind babies start smiling at the same time of life. Congenitally deaf kids start babling at the same age as all kids. That suggests it could be genetic.

Then he discusses the fascinating and important world of pre-natal environmental effects. Blood in the prenatal environment is shared. Fred vom Saal showed that there is a different hormonal environment depending on the sex of neighboring fetuses (rats can have 12 pups at once). The more male siblings around a female yields a later onset of puberty. These endocrine (blood-based hormone system) effects can have long-lasting impact.

If mom is stressed during pregnancy, she will have more glucocorticoids in her blood which is shared with her fetus. Her kids will have smaller brain size, a thinner cortex, less learning ability, will be more prone to anxiety, fewer benzodiazepine receptors, more cognitive decline in old age, and girl children will have elevated glucocorticoids in their pregnancies and so the effect is passed on to the grandkids. What?! Inheriting a non-genetic effect!! That's not supposed to happen, Lamarckian inheritance was thoroughly rejected in my high school biology course. The biologists are sensitive to this history and so they call the phenomenon "non-Mendelian inheritance of traits".

Nutrients are also shared with the fetus. In the Winter of 1944 the Nazis diverted food from Holland to Germany and the result is now called the Dutch Hunger Winter. Third trimester fetuses during the Dutch Hunger Winter have a thrifty phenotype: their pancreas more aggressively secretes insulin to store all available calories, their kidneys retain more salt. Adults from these third trimester fetuses of the Dutch Hunger Winter have a 19-fold increase in obesity, hypertension, diabetes, and metabolic syndrome (a complex of several, usually at least three, of the following factors: abnormal blood fat, high blood pressure, fatty liver, insulin resistance, and new fat deposition). This phenomenon has launched a new field "fetal origins of adult disease".  Biologists have traced the effect of the thrifty phenotype to the epigenetic coding for insulin. Interestingly, the effect only occurs if the starvation happens during the third trimester, in the Lennigrad starvation which was more gradual there is no thrifty phenotype effect.

Another example of pre-natal environmental effects is when fetuses are exposed to more phytoestrogens (plant-based "dietary estrogens") there is a higher risk of estrogen dependent breast cancer (Wikipedia says there are conflicting studies). If flavored sucrose is added to the amniotic fluid (evidently they drink it!), after birth the baby will prefer that flavor (tested by measuring the "sucky motions" of newborns which is believed to represent an indication of preference).

Mothers reading to their fetuses results in their newborns responding with "sucky motions" when the same story is read after birth. Father's voices don't carry through the thoracic diaphram so well and their stories have no effect.

The age of the mother at the birth of her child correlates to her estrogen levels (lower when she is young and tapering off to be lower again in old age). The age of a child reaching puberty is closely correlated to this effect (higher age of puberty when the estrogen levels are low). I wonder how this correlates to the historical averages for age of onset of puberty and what this tells us about how Mom's age has changed through history?

Rats have been bred for different types of anxiety traits: apparently these are genetic traits. Darlene Francis of Berkeley developed a surgical technique to transfer fetuses to foster mothers. It turns out that the anxiety traits were not genetic, something about the fetal-mother relationship spreads the anxiety trait. Another instance of "non-Mendelian inheritance of traits" or Lamarckian inheritance!

You get more genetic material from your mother than your father! Your mitochondria come exclusively from Mom. Your father's genes may not come from your father (according to Bob Wyman about 10-15% of children don't have the father they're supposed to have!). The egg that became you was a fully functioning cell which in addition to mitochondria has a full set of proteins including transcription factors (proteins that promote or block genetic expression). If an environmental toxin knocks out a transcription factor or other vital protein in the egg that becomes you (or at any vital step during development), genetic expression may be unalterably affected for generations to come (Lamarckian inheritance).

The idea of indirect genetic effects comes from psychologist Judith Rich Harris' book "The Nurture Assumption" which has a section on behavioral genetics. She argues that some genetic effects can in turn cause other behavioral effects. For example, one's score on the introversion-extroversion scale may be mediated by the genetic effect of height: people who are treated better become more extroverted (studies have shown that height is correlated in this way). In birds, heritability of pecking order has been shown to be mediated by the melanism or color and iridescence of the feathers. The heritability of chicks to eat grubs is mediated by their interest in their feet which are genetically programmed to be "interesting". There is a 70% heritability in Americans of political party affiliation which is apparently mediated by one's tolerance for ambiguity (conservatives are less tolerant of ambiguity than liberals --- what does this have to do with Kohlberg's scale of moral development?). Rat strains that show heritability of aggression appear to be mediated by tolerance to pain (more aggressive rats have a lower threshold for pain sensitivity). Many behavioral and personality traits are now known to be caused by indirect genetic effects.

Michael Meaney at McGill University found that the effect of petting rats for 3 minutes (vs. a deleterious effect from long petting/removal sessions) causing greater resistance to several neurological insults, lower glucocorticoids, bigger brains, and better learning ability is mediated by the mother's response to the absense of the pup (she gives prolonged licking and attention to the ones that have been temporarily removed and neglects the ones that were absent for long periods).  The effect is multi-generational:  Mom's attention affects transcription factors for receptors to stress hormones and estrogen.  The effect is reversible:  if half of one's childhood is spent with a negligent mom and the other half with an attentive mom, the epigentic pattern changes back.  Behavior & personality traits is often caused by indirect genetic effects.

6. Behavioral Genetics I