from Hacker News

Adult Neurogenesis – A Pointed Review

by mantesso on 4/5/18, 3:10 PM with 36 comments

• by subroutine on 4/5/18, 5:59 PM

  Last week a paper came out saying no adult neurogenesis. But I see today this paper came out:

  http://www.cell.com/cell-stem-cell/fulltext/S1934-5909(18)30...

  It usually flips every month.

• by savanaly on 4/5/18, 5:56 PM

  Clearly humans care a lot about neurogenesis not just because understanding whether it happens in humans would be a stepping stone to building a robust model of how cognition happens, but also because intuitively it has to do with whether we can enhance our own mental abilities in our lifetime-- one of the most fervent desires of many, I'm sure.

  We've already had to write off things like growth mindset as a way to increase our fluid intelligence, it seems. It is dismaying that, based on this study, we'll have to write off things like exercise as well which we now realize only work in rats. What's left that could plausibly help? Perhaps just: building better work and study habits and simply spending more time on mental tasks. At least we're pretty sure those aren't illusory.

• by cobbzilla on 4/5/18, 6:01 PM

  Fun article. Best line: "if we want to believe something, it will accrete a protective layer of positive studies whether it’s true or not."

• by kragen on 4/5/18, 5:41 PM

  I've wondered for a long time if stem cell therapy could, for example, cure my amblyopia — though it hardly seems worth the cancer risk. This very surprising news is a data point in favor of that kind of thing.

  It's also a kind of data point that we really need to improve science a lot in order to be able to advance as a civilization.

  (Amblyopia is due to anomalous neural development caused by, more or less, early childhood early sensory deprivation from one eye. The result is a lifelong lack of visual competence in that eye — it may focus perfectly, as mine does, but recognition is poor, apparently because perception of Gabors at certain spatial frequencies is very poor. I say "more or less" because the eye is still providing sensory input, but because it's pointing in the wrong direction, that input isn't useful.)

• by cdcox on 4/5/18, 6:30 PM

  While this comes to the right conclusion I think it mistraces a lot of the issues. For example: a fair amount of the depression and neurogenesis studies (and indeed neurogenesis manipulation studies) suffered from the same issue: it's really hard to make a population of cells die without generally poisioning the remaining cells. Most of the initial studies relied on things like radiation, ie radiate a rat and it starts acting depressed. There is now increasing evidence that radiation kills off a bunch of synapses and causes some wicked disregualtion across the system so that turned out to be wrong. Conversely BDNF enhances neurogenesis but it also enhances synaptic learning so even positive manipulations had this issue but in reverse. Newer timed genetic manipulation studies were more precise but the depression-neurogenesis link has been slowly losing steam in light of these studies.

  But the BIG issue boils down to this: there was never strong evidence for the importance of neurogenesis. The dentate gyrus is a tiny region of the brain with neurons that act in unusual ways. In humans it's an even tinier region. While it is probably required, at least as a pass through, to learn new memories, it's surprising how much weight people were placing on neurogenesis in this tiny tiny structure. Given, networks can make small zones have big effects, but the weight of evidence should have always been on people pushing the area to prove the dentate was this massively important structure. This has not been conclusively done (studies have linked it to some specific subtypes of learning but those have limited it as much as they have found an important role for it). This is a structural issue in science broadly, there is an incentive to push forward with the next big finding but no incentive to go back and confirm the gulf of assumptions that a literature is resting on.

  The good news is, human brains could always change, synaptic plasticity is present throughout life and has been shown by many good studies to occur throughout adulthood. We don't need new cells in a tiny part of the brain to learn new things, networks rewire and while that rewiring isn't something as dramatic as new cells, small effects can lead to big outcomes in dynamic systems. People obviously make new habits, learn new skills, and make new memories throughout life. The onus should have always been on science to show why it occurs, not to lend credibility that it occurs.

• by garyrob on 4/5/18, 6:43 PM

  A recent study[0] seems to provide meaningful evidence that while adult neurogenesis occurs in rats, rabbits and some other species, as mentioned in the research referred to by the OP, it does not actually occur in humans. I'm not claiming to know the truth on this, but I thought this research should be part of the discussion.

  [0] https://www.theatlantic.com/science/archive/2018/03/do-adult...