That would require such evidence to not be classified. Which it presumably is, as one possibility for the cause is "some other military surreptitiously pointing directed-energy weapons at us while we're out-and-about"; and — regardless of the likelihood of that explanation — just in case it is the explanation, they wouldn't want to do anything that could provide a public-channel feedback loop for said other military to learn exactly how effective their new weapon is.
Also, if you look for statistical anomalies in a random group of people, you just have to pre-screen for 30 variables that you don't mention in your report to get one with p<0.05