"You've written a number of times about sleep regression at four months and eight months. Since we're a few weeks away from four months, I was wondering what exactly is sleep regression? Are there theories as to what causes it, like some developmental milestone that all kids hit at the same time? How long does it typically last? Do kids typically return to their previous sleeping patterns afterwards, regardless of how you handle a sleep regression? Anything else that would be useful to know?"
I think "regression" is just the term people use so that parents know that the blow-up of sleep is pretty universal at certain times and isn't just something that's affecting one child.
I actually think regression is a complete misnomer for several reasons. First, it implies that babies' sleep progresses in a linear fashion, with better sleep each week until the fateful day at which they sleep all the way through every single night. To put it mildly, that isn't true. Secondly, it indicates that it's a negative trend, when in reality it's just a reaction to things cooking on other levels. Kind of the way you shouldn't run right after you eat a big meal, since your body is busy digesting the food and can't spare the energy for running.
Babies have a whole lot going on with them in general. On one level we have the physical milestones. Rachael at SparkPlugDance.org has done this amazing breakdown of the physical milestones from birth through walking and how they affect neural pathways. (She also has a great article up about how to make tummy time fun for even the biggest resisters.) On another level we have the whole feeding thing, which is quite a bit to get coordinated. First they have to get down the sucking part, then they all seem to go through that nasty "does my baby have colic or is it a GI problem?" phase at 6 weeks, then they have the growth spurts (at 3, 6, and 12 weeks and again at 6 months). They're learning to trust you and love you. Then there's teething, which kids can start as early as 6 weeks (mine did) and can last on and off for months before one pops through.
All of these things can affect sleep. But all of these things are kind of obvious, so if your baby's up all night grabbing her ears and drooling, you know it's teething. If he can't sleep because he keeps waking up on his hands and knees rocking back and forth, you know it's because his body is learning to crawl. Since you can see it and label it you can understand it.
Your baby is also going through enormous developmental spurts that you can't see, because they're dealing with cognitive processes. They work through these spurts the same way they work through the physical spurts, but when your baby is practicing recognizing patterns, you can't see that. Leading up to the actual new skill the baby is going to go through several weeks of intense brain work and prep that you can't necessarily see (unless you know specifically what to look for). One of the side effects of this brain work is that they don't sleep as well as they do during times in which they're not about to master a new skill. They may seem restless in the night (like they do sometimes when mastering a physical skill) but it's just nothing you can see and label. So we call it a sleep regression.
Once a baby has learned the new skill, s/he will often sleep through the entire night for 1-3 nights after mastering the skill. Which is freaky from the parents' perspective, because you go from waking 10 times one night to sleeping 12 hours the next. But it makes complete sense if you know that the baby was working on this really tough challenge and couldn't help but wake up so often (in fact it's probably a miracle that the baby could even fall asleep in the first place with so much going on in the noggin), and then once they've got it they relax and sleep it off for a couple of days, like sleeping off a crazy party. Then they'll most likely go back to sleeping the way they were before they started working on the new developmental skill.
Everything I know about when and what the actual developmental spurts are I learned from the book The Wonder Weeks by the Dutch researchers Hetty Vanderijt and Frans Plooij. They tell you when the spurts happen (based on a 40-week baby, so if your baby was born early s/he'll hit the spurts a little later and if s/he was born later than 40 weeks s/he'll hit them earlier), what kinds of symptoms the baby will show while working on the new skill, how you'll feel when the spurt is happening, how to recognize what's going on, and what kinds of games you can play with your baby to help develop or practice the new skill. One of the things I love about this book is that they don't tell you what to do (aside from suggesting games to play) they just tell you what's going on. I have a definite bias against books that tell you how to raise your kids, and I love that this book is just a roadmap of what's going on when.
According to Vanderijt and Plooij, the spurts happen at weeks 5, 8, 12, 19, 26, 37, 46, and 55 (that's as far as they studied).
I think the reason the "4-month sleep regression" is such a big deal is that babies wake up a lot in the few weeks it takes while they're working on the 19-week spurt, and then once they've gone through that spurt they only have a week or two (if that) of respite before they start prepping for the 26-week spurt. So it's kind of one long stretch of bad sleep and cranky moods during that stage of your baby's life.
The "8-month sleep regression" (which for some babies is closer to a 9-month sleep regression) is related to the 37-week spurt. For some reason that one just seems to cause more waking, too, than some of the other spurts do. It might also be particularly hard because many babies are smack in the middle of working on crawling or walking, and also teething. (At Casa Moxie we've had probably 8 weeks of crappy sleep between teething, the 37-week spurt, crawling, teething, and now pulling up. Every now and then he'll have an easy night, but boy is it rough being a 9-month-old.)
Bear in mind that individual kids have different reactions to all kind of spurts (physical, developmental, etc.). Some teeth painfully for months, while others just pop a tooth with no symptoms. Some will wake in the night practicing crawling for weeks, while others never do and just take off one day with no warning to you. The developmental spurts are the same, so you might have a kid who has 3-4 nights of wacky sleep and then learns a new skill, or you could have one that spends 3-4 weeks waking up before every spurt.
If you want to be really prepared for the "regressions," I'd say borrow or buy the Wonder Weeks book, read it, and pay attention to the signals your baby is giving you. Once you know what's going on it becomes fasciating and annoying instead of worrisome and torturous.