Several readers have pointed out, correctly, that the army is not demographically simliar to the civilian U.S. population, and therefore my comparison was in error. One example: the suicide rate for American males has been above 20 per 100,000 for years, and most of the army is male, so wouldn’t that be a better comparison?

Upon further consideration, I think we’re all wrong. I was in error to compare the Army suicide rate to that of the U.S. population, but the answer isn’t just to change the sample being compared. As a statistician would say, the important thing about a rate is the trend.

It would not be hard to believe, for example, that members of the military are generally more motivated about life and less likely than similar members of the civilian population (mostly male, white, etc.) to commit suicide. But if army suicides spike when others do not, then that may say something about the plight of soldiers at this point in time. And conversely, if all suicides are up, then it would actually be unusual for military suicides not to follow suit. So let’s look at the trends, based on the stats I was able to gather on the fly:

Between 2001 and 2006, the U.S. suicide rate went from 10.8 to 13.4 (per CNN) per 100,000, an increase of about 25 percent.

The rate for American men aged 17-45 went from 17.6 to 21.1 per 100,000 (per CNN), an increase of 20 percent.

During the same period, the new report says, the Army rate went from 9.1 to 17.3 per 100,000. That’s an increase of 90 percent.

It may be unfair to use 2001 as the comparison, since that was the Army’s all-time low, but the numbers may well be an indication of how much pain our soldiers are currently facing. Or, it could just be that 2006 is the fluke — after all, the rate stayed between about 10 and 13 suicides per 100,000 right up through 2005, before spiking in 2006.