The Barnes et al. TBI/dementia study has been making its way around Twitter again lately. I'm sure the authors are great epidemiologists, but there are some problems that I wanted to share with the world, or at least my 48 followers. bit.ly/2xRsRy7 /1
First, dementia “diagnosis” in someone’s VA chart can range from well-formed to completely arbitrary. Some people are diagnosed based on an appropriate diagnostic process. /2
Many, though, are classified with a dementia diagnosis because they tell their primary care physician, or other provider who is not qualified to diagnose neurodegenerative disorders, that they have memory problems, so the provider codes it as “dementia”. /3
This sample could, then, include any number of people whose diagnoses are based on self-report of memory concerns. And especially for a sample that includes a very substantial number of recent (young) OEF/OIF veterans, and a mean age of 49, this is a very substantial problem. /4
It looks like, for example, at age 40, more than 5% of the mTBI sample had “dementia”. /5
Second, “dementia” was not broken down by diagnosis. I would very, very, very much like to have seen the numbers for AD, for example, rather than collapsed numbers that include any number of patients with NOS dementia labels. /6
The conclusions, interestingly, focus entirely on AD and CTE, even though there is no indication of dementia type in the data. /7
Third, note that the mean time from index date (meaning date of most severe TBI) to dementia diagnosis was 3.6 years (SD=3.0). Seriously? This means that these were, mostly, very recent head injuries. /8
This amplifies my 1st concern, that people are being given “dementia” labels due to self-reported complaints post-mTBI. I would much prefer a study that is able to parse out remote vs. recent TBIs, rather than collapsing them all and simply reporting on the effect of recency. /9
Fourth, and the authors do note this possibility—that complaints (and “dementia” diagnosis) are actually reflecting persisting cognitive concerns post-injury. /10
Their interpretation is that even very mTBI “is associated with greater risk of long-term cognitive and functional impairment in these veterans.” This ignores the literature that highlights the importance of distinguishing between objective & subjective symptoms. /11
Not trying to unfairly criticize a study that was well-done from a technical standpoint. But the other day I removed a diagnosis of CTE (based just on history of playing hockey) from the chart of a cognitively intact veteran. He would've had "dementia" in this study. /12
