The more I read about research into the causes and treatments of dementia, the more concerned I am that the frenzy around creating “a world without Alzheimer’s” and the magnitude of funds directed at this goal are causing a lot of scientists to forget their basic rules of science.
Here is a case in point: recently, WedMD’s Medscape online newsletter sent out a bulletin with the headline, “Are Fewer People Developing Dementia?” The article quoted recent data from the Rotterdam study showing an overall 25% decline in the incidence of new cases diagnosed between 2000-2005, compared with a similar study conducted from 1990-1995. The article quoted the study team as saying, “Although the differences were nonsignificant, our study suggests that dementia incidence has decreased….” The team suggested that much of this decline might be due to improvements in cardiovascular health.
Now, I happen to believe that much dementia, including that which we call Alzheimer’s, is at least partly connected to cardiovascular health; therefore I am quite optimistic that improvements in this area may indeed be leading to a lower incidence of dementia. The problem is that the results were nonsignificant; scientifically, this means that the study did not prove a thing!
The finding of larger brain volumes in the newer cohort lends support to their theory, but the primary outcome did not reach statistical significance. This means that the “25% decline” is unproven, and yet it is trumpeted in a media report filled with speculation by the research community.
This thought process echoes the rather sketchy data that has been accepted as gospel by much of the health care profession over the years. For example, in spite of the volumes of data showing little benefit and much potential for harm from antipsychotics, most physicians believe them to be helpful in many people living with dementia, and the 2011 Agency for Healthcare Research and Quality statement reports “high strength of evidence” of benefit in “behavioral symptoms of dementia.”
Similar claims of overreaching conclusions have been made about the benefits of cholinesterase inhibitor drugs, such as donepazil. And recent accusations of researchers (including a claim that the data on the benefits of long-acting donepazil were exaggerated, and a “whistleblower” suit claiming inaccurate data reporting in a 2000 Massachusetts brain volume study), suggest that there is at the least a lot of “groupthink” going on, and perhaps even worse in some cases.
Another example of groupthink is the rush toward treatment of amyloid, with little or no discussion of studies (like the nun study) that have shown the presence of large amounts of amyloid in cognitively normal individuals as well.
The media feeds the frenzy by highlighting every study with even a whiff of possibility as the next breakthrough. And the researchers certainly pick up on this and use the media hype to get their names out there in the public eye.
It is time to get real–both in setting the goals of research and in the way we conduct and report these studies. Irresponsible reporting does not help our cause and actually harms our efforts to improve the lives of people with dementia.