Action Potential - June 12, 2017
The newest alcohol study…
I’ll preface this to say that I don’t intend any of this to be healthcare advice, and I’m not that kind of doctor.
Any aficionado of science news will have spent roughly half of their time listening to the latest on the rollercoaster of good and bad news about all our favorite tasty treats and their effects on biology. Common favorites include fat, cheese, chocolate, and coffee, but perhaps the most discussed of all is alcohol.
Common lore has cast moderate (~a small glass of wine or two a day) alcohol consumption as generally good for the brain. The newest finding from Topiwala and friends says, “nuh-uh”. Over the 30 year observational study, they found that even moderate alcohol consumption predicted hippocampal shrinkage, decreased structural-integrity in the corpus callosum, and faster decline in lexical fluency. In non-neurosciency terms, it looked to them like alcohol damaged the brain, and that the people who drank more got worse faster at naming as many words starting with the letter “C” as fast as they could.
To me, the brain imaging in this study seems robust, but the behavioral results are a little less convincing. Here are the reasons for my skepticism:
The researchers collected information about how many drinks their subjects typically consumed per week. If I was doing the research, I would probably do my stats by using all of those reported amounts of alcohol for each subject and adding that variable directly to my statistical model. Instead, they reduce these numbers to categories (light alcohol use, moderate alcohol use, heavy alcohol use, etc.), and then do their stats across those categories. Why is this bad? Well depending on where you draw the lines for your categories of data, you might get a different answer. For instance, if I arbitrarily decided that moderate drinking should be defined as 6-13 drinks per week instead of 7-14, which is very similar, I don’t know for sure that I would have seen the same results. You’re just losing information that doesn’t need to be lost.
The last thing I’m skeptical about is in regards to the actual alcohol usage. It is well-known that heavy alcohol consumption is bad for the brain. It’s also been shown that people underestimate their alcohol usage (e.g. doi:10.1093/eurpub/ckt016). If people are underestimating their alcohol usage, and are called “moderate” drinkers, when they’re actually “heavy” drinkers, this paper is not adding too much to what is known. Now the authors say that their design helps prevent this underestimation, but it sounded like the participants responded about their average weekly alcohol usage in ~5 year intervals, and I could totally see that leading to underestimation.
Okay, let’s look at the researchers' plot for the behavioral data (below). They use “change in lexical fluency” as the dependent variable related to alcohol consumption category. The super striking thing to me about this graph is the rather surprising finding that at the beginning of the study, the heavy drinkers seem to be way better than the non-drinkers, and that the performance on the task seems to be dose-dependent on the amount of alcohol consumed. Now the authors addressed this, and said that there were no statistically significant differences between these groups at the beginning of the study. But to me, it seems very likely that those initial differences between the groups are driving the differences in the changes of the scores over the 30 year study.
Whatever the case may be about the truth of the behavioral result in this study, here are some general things to consider about alcohol effects. I don’t think many people were arguing that alcohol was good for neurons, in the first place. Moderate drinking might be good for socializing which seems to be good for the brain. Also, moderate drinking might be good for cardiovascular health which seems to be good for the brain. These positives were always being balanced with the negative effects. So maybe, at least this can be that last jolt of motivation you need to stop drinking alone.
doi: https://doi.org/10.1136/bmj.j2645
DAM! That’s cool.
So if you have kept up with your review of general neuroscience, you might recall that there are different types of cells in the brain. One of those classes of cells are known as “microglia”, which roughly means “tiny, yucky glutinous stuff”, to help distinguish it from the rest of the yucky glutinous stuff in the brain. And that’s important because microglia do wayyyy different things. But exactly why they do what they do is still not well understood in some circumstances.
Let me explain.
The brain doesn’t trust the rest of the body. It demands more blood when the rest of the body is running out of oxygen. It protects its cells from all that dirty signaling propaganda flowing through the rest of the blood stream. And it has its own immune system to help do all the cleaning, which is under very strict control. Microglia are the brain’s immune cells. Now, if the brain ever has inflammation, that tends to be bad news. Sometimes if the brain has local inflammation (such as in certain diseases like Alzheimer’s), scientists look at the microglia, and they find that they have become angry, ultra-aggressive little beasts. So naturally, researchers are curious whether the things are out of control, rampaging and wreaking havoc, or if they’re working really hard in order to help, or maybe that they don’t really contribute to the disease at all.
Enter Professor Schwartz’s team. These researchers were able to sequence RNA from all the immune cells in the brains of mice models of Alzheimer’s patients. They repeated this for different stages of the disease’s progression, and found a subset of microglia that looked different from the microglia of healthy mice, which change as the disease progresses. These special microglia were dubbed, “DAM,” for “disease-associated microglia”, but I’m going to call them DAMs even though “microglia” is already plural. These microglia express different proteins than the proteins expressed in healthy control mice, and those proteins seem to help with the clean-up. If those helpful proteins are mutated, then the disease occurs worse and earlier.
So what’s so cool about this. Well first of all, being able to sequence the RNA from all the immune cells in a mouse brain is really hard, and it’s really sweet that we can do this now. The findings of the study are also interesting, though, because they point to dysfunctioning microglia as being important to the disease. Now that we know that DAMs exist, we can try to figure out how to make them better and faster and hopefully repair damage involved in Alzheimer’s disease that often snowballs out of control.
Weizmann Institute of Science. (2017, June 9). The brain's rejuvenating cells. ScienceDaily. Retrieved June 12, 2017 from www.sciencedaily.com/releases/2017/06/170609135903.htm