1) Yes, to some extent, research has shown that humans have similar “sensitive” or “critical” periods during which exposure to a certain behavior or environment has a larger effect on the individual. This is shown in a baby’s inability to thrive without hugs and bonding with its mother or in the long-lasting effects of adverse childhood events, like abuse, (vs. adverse adult events) on health.
2) Gruenewald shows that higher allostatic load is associated with both cumulative and persistent lifetime SES adversity in people, while Weaver shows that “positive” maternal attention can impact a rat’s glucocorticoid receptor gene if delivered in a sensitive time window. The findings differ in many aspects (ie. study subjects, time sensitive vs. life course, biomarker changes vs. gene expression as an outcome), but they both seem to support the conclusion that environment can affect biological functioning.
3) This is a difficult question to answer because I think the gene-by-environment argument has a lot of potential. The concept is logical and easy to observe anecdotally, that neither genes nor environment but rather the interaction of the two completely explains a health outcome. Additionally, the article does well to point out examples in which the phenomenon has proven true (MAOA allele with childhood maltreatment, serotonin transporter polymorphism with depression). However, race is not a perfect indicator for genetics, just as socioeconomic status and geographic region are not perfect indicators for environment. In our attempt to apply this phenomenon to more generalizable groups, we introduce error. I will admit that observational studies show similarities in health outcomes among people in these socially-constructed groups (by race or SES or geography) that could be explained by gene-by-environment interactions, but the possibility also exists that people in the same “group” are experiencing different gene-by-environment interactions that simply result in the same health outcome.