calming supplements & the placebo effect

I had a really tough ride on Suzy this week, which led me into a discussion with my barn manager about possibly giving her another dose of depo.  Barn manager suggested that since I spend a lot of time with Suzy, I’d be able to help Suzy’s owner decide if her behavior warranted another dose.  Now, there’s not a lot (almost none, but perhaps there is something I haven’t dug up yet) of evidence that depo actually does anything in horses, despite many, many, many anecdotes to the contrary.  Absence of evidence isn’t evidence of absence, but I’m not really one to go adding hormones to someone’s lifestyle without a pretty good reason.

seemed appropriate here

This led me to think about the hundreds of calming supplements and treatments available for horses that have little to no evidence of a significant (or even potentially biologically relevant) effect upon behavior.  Hell, I use one of them!  Yet so many of us cling to these supplements and swear up and down that they are doing something.

But the placebo effect is a strong and very legitimate thing. Whether or not the placebo effect is causing actual physiological changes or is being exploited by companies to sell sham products, it is a very easy trap to fall into. Most of us are going to add calming supplements to a horse’s diet after they’ve been particularly wild and crazy for some amount of time — maybe a week’s worth of rides.  But if those particularly wild and crazy days are outliers, or even just somewhere on the edge of “normal” for that horse, then his behavior is going to trend back toward normal whether you add the calming supplement or not.  Those days were abnormal, and unless there’s something else significant going on to change the horse’s behavior (totally plausible — but not necessarily the case), the added calming supplement most likely did nothing.  The horse’s behavior was going to go back toward “normal” anyway.

big spikes of bad behavior like this == abnormal

Let’s say that we rate Murray’s badness on a very scientific scale of 1-10 each day: 1 is totally normal, no bad behaviors performed, 10 is rearing and striking.  (For the record, I’ve never experienced either a 1 or a 10 day.)  If Murray tends to hover around a 3.5 because he likes to get down with his bad self, but suddenly spikes up to an 8 one day, I’m likely to dump a bunch of extra magnesium in his bucket for a few days.  And then, because being a level-8 bad boy is a really unlikely occurrence, we just slide back toward the average. Which is both statistically and realistically much more likely to occur on any given day.

Especially because the way we think of behavior (or anything with a gradient, basically) tends to fall on a bell curve.  If we assume that each behavioral category is roughly on standard deviation away from the next, it means that behaviors in the “average” category occur 68% of the time. Behaviors one more standard deviation away from average occur with 95% frequency.  If being slightly nutty or slightly better than average is occurring with 95% frequency, any time you have a bad day, you have something approaching a 95% chance that your next ride will be better (assuming the two days are independent, which they aren’t).

So when we’re adding calming supplements, or more invasive approaches to calming behavior, how are we to know if we are really doing anything?  There are ways to test it — you could blindly rank your horse’s behavior every single day that you spend with him while another person either gives him that supplement or a sham supplement. Then you’d compare the distributions of behaviors demonstrated.  I just turn to the science.  Of course, there’s really not a whole lot of peer reviewed research in this area, which is frustrating.

All of this is not to say that I don’t believe in calming supplements or that diet can change behavior.  Obviously I do, and there are some behaviors that I have seen decrease rapidly with the right change in diet.  I am completely certain that if we mapped Murray’s behavior when he’s on alfalfa and compared it to his behavior off alfalfa, we’d see a significant difference.

even walking is hard when Murray eats alfalfa

But it does mean that I approach adding supplements like this with a fair bit of caution, and I pretty much don’t believe what I read from testimonials on the internet.  People are notoriously bad at understanding probability and statistics without training in those areas, and we love just-so stories.  We also really, really want to help our horses, and they can’t talk to us and say “I’m feeling a little extra girthy today, how about you take it extra slow while we tack up?” (I mean, sure, he does say that to me. It’s just usually while I’m tacking up and it’s a leeetle bit too late.)  It also means that I don’t pay too much attention to any one ride, or even any one week, when making decisions about this stuff.  One day is an outlier.  A whole week… could easily be a rough week.  Especially when there’s other stuff going on in the world that might influence behavior more than supplements do — changes in turnout, routine, weather… I’m seriously convinced that when the barometer drops, Murray’s brain swells and hurts his little head, because he’s usually much worse before a storm than during!

I may also use this logic to aggressively convince myself out of using supplements that might help. When barn manager originally told me to take Murray off alfalfa I was like “NAW THAT’S NOT A THING”.  Same with adding magnesium.  And both of those things had a huge positive effect on Murray’s general outlook on life.

So, how do you approach adding and subtracting supplements like this?  Just go with the prevailing wisdom? Trust your gut? Appeal to nature?