Thanks Jin for this thoughtful response. Indeed, I think language development is one of the nicer examples of a sensitive period. Your point that "one vulnerability is associated with other vulnerabilities" is really a central problem in doing causal research on the sources of inequalities. We need to do rigorous research to identify what we could change/intervene upon to improve health outcomes and eliminate inequalities, but distinguishing between the effects of intervening at age 2 years versus age 12 years turns out to be very tricky. The fact is that children who are in disadvantaged settings at age 2 are usually also in disadvantaged settings at age 12 (and also age 42, incidentally), so it's very tough to tell observationally when an intervention could be beneficial. This leads to a lot of key debates in the field: how critical is it to get interventions to very young children? Is there still a benefit of intervening to help older children or even adults. Some argue no - it's a waste of money and effort after a certain age because it's very unlikely to have a large benefit.